summaryrefslogtreecommitdiff
path: root/drivers/gpu/drm/amd/display
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/gpu/drm/amd/display')
-rw-r--r--drivers/gpu/drm/amd/display/Kconfig2
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/Makefile1
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c1602
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h84
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c564
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.h55
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c30
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h2
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c140
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c123
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c54
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c66
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h6
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq_params.h4
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c119
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c97
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.h7
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c52
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h3
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_quirks.c178
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_trace.h26
-rw-r--r--drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/Makefile43
-rw-r--r--drivers/gpu/drm/amd/display/dc/basics/dc_common.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/basics/dce_calcs.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c5
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/bios_parser.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c24
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/command_table2.c57
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c104
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/command_table_helper.h8
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/dce110/command_table_helper_dce110.c104
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper2_dce112.c106
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper_dce112.c104
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/dce60/command_table_helper_dce60.c104
-rw-r--r--drivers/gpu/drm/amd/display/dc/bios/dce80/command_table_helper_dce80.c104
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile4
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c8
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c5
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c14
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c14
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dalsmc.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr.c36
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr.h31
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr_smu_msg.c118
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr_smu_msg.h34
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c22
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c15
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn351_clk_mgr.c141
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c297
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c6
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h41
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dalsmc.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.c431
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.c37
-rw-r--r--drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/core/dc.c595
-rw-r--r--drivers/gpu/drm/amd/display/dc/core/dc_debug.c124
-rw-r--r--drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c146
-rw-r--r--drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c27
-rw-r--r--drivers/gpu/drm/amd/display/dc/core/dc_link_exports.c28
-rw-r--r--drivers/gpu/drm/amd/display/dc/core/dc_resource.c445
-rw-r--r--drivers/gpu/drm/amd/display/dc/core/dc_state.c132
-rw-r--r--drivers/gpu/drm/amd/display/dc/core/dc_stream.c161
-rw-r--r--drivers/gpu/drm/amd/display/dc/core/dc_surface.c52
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc.h523
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c804
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h81
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_dp_types.h149
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_dsc.h5
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_fused_io.c148
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_fused_io.h31
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_helper.c10
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_hw_types.h188
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_plane.h16
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_spl_translate.c39
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_state_priv.h20
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_stream.h49
-rw-r--r--drivers/gpu/drm/amd/display/dc/dc_types.h77
-rw-r--r--drivers/gpu/drm/amd/display/dc/dccg/dcn20/dcn20_dccg.h94
-rw-r--r--drivers/gpu/drm/amd/display/dc/dccg/dcn35/dcn35_dccg.c62
-rw-r--r--drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.c26
-rw-r--r--drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.h40
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c12
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c65
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c44
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dmub_abm_lcd.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c15
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c7
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c62
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce/dmub_replay.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c13
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c8
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c201
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c124
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce60/Makefile3
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce60/dce60_timing_generator.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c7
-rw-r--r--drivers/gpu/drm/amd/display/dc/dcn31/dcn31_panel_cntl.c34
-rw-r--r--drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.h6
-rw-r--r--drivers/gpu/drm/amd/display/dc/dio/dcn31/dcn31_dio_link_encoder.c5
-rw-r--r--drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.c130
-rw-r--r--drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.h41
-rw-r--r--drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c17
-rw-r--r--drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.h23
-rw-r--r--drivers/gpu/drm/amd/display/dc/dm_helpers.h13
-rw-r--r--drivers/gpu/drm/amd/display/dc/dm_services.h7
-rw-r--r--drivers/gpu/drm/amd/display/dc/dm_services_types.h26
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/Makefile18
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c138
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c45
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.h12
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c82
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c13
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn30/display_rq_dlg_calc_30.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c7
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c32
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c12
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c5
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn401/dcn401_fpu.c239
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dcn401/dcn401_fpu.h14
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c11
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h8
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/Makefile41
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c58
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/display_mode_core_structs.h111
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/display_mode_util.c6
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c697
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.h8
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.c218
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.h8
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.c151
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.h77
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/bounding_boxes/dcn3_soc_bb.h401
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/bounding_boxes/dcn4_soc_bb.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_dchub_registers.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_display_cfg_types.h57
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_soc_parameter_types.h7
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_types.h99
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.c85
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c5310
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_shared.c12413
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_shared_types.h322
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.c364
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.h6
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.c66
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_dcn4.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_factory.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn3.c12
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c456
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.h7
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_factory.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_interfaces.c49
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_legacy.c10
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_legacy.h9
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_optimization.c307
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_optimization.h33
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_soc15.c1170
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_soc15.h (renamed from drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top_mcache.h)19
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top.c354
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top_mcache.c549
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_debug.c31
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_debug.h182
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_internal_shared_types.h78
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.c50
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml2_mall_phantom.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml2_policy.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c100
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c320
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h9
-rw-r--r--drivers/gpu/drm/amd/display/dc/dml2/dml_display_rq_dlg_calc.c12
-rw-r--r--drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c10
-rw-r--r--drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c20
-rw-r--r--drivers/gpu/drm/amd/display/dc/dpp/dcn32/dcn32_dpp.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.h151
-rw-r--r--drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c35
-rw-r--r--drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c242
-rw-r--r--drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.c7
-rw-r--r--drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.h6
-rw-r--r--drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.c53
-rw-r--r--drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.h8
-rw-r--r--drivers/gpu/drm/amd/display/dc/dsc/dsc.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.c13
-rw-r--r--drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c14
-rw-r--r--drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/gpio/hw_hpd.c8
-rw-r--r--drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.c10
-rw-r--r--drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.h6
-rw-r--r--drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_stream_encoder.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.c18
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubbub/dcn21/dcn21_hubbub.c6
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubbub/dcn30/dcn30_hubbub.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubbub/dcn32/dcn32_hubbub.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c17
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.h16
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.h27
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.c28
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.h9
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn201/dcn201_hubp.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn21/dcn21_hubp.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.c26
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.h6
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn31/dcn31_hubp.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn32/dcn32_hubp.c6
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.c6
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.c146
-rw-r--r--drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.h64
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/Makefile26
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dce/dce_hwseq.h9
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.c30
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c75
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dce120/dce120_hwseq.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dce60/dce60_hwseq.c (renamed from drivers/gpu/drm/amd/display/dc/dce60/dce60_hw_sequencer.c)3
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dce60/dce60_hwseq.h (renamed from drivers/gpu/drm/amd/display/dc/dce60/dce60_hw_sequencer.h)0
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dce80/dce80_hwseq.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c293
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.h11
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_init.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c349
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.h17
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_init.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn201/dcn201_init.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_init.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c24
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c36
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c26
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c31
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c112
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c9
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c6
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c1035
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.h19
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c14
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h45
-rw-r--r--drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/core_status.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/core_types.h50
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h6
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h19
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h61
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h6
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h20
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h8
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h6
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h93
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/optc.h33
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/pg_cntl.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h35
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/hw/transform.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/link.h17
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/link_enc_cfg.h5
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/reg_helper.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/inc/resource.h18
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/Makefile9
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dce60/irq_service_dce60.c31
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dce80/irq_service_dce80.c31
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn20/irq_service_dcn20.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn201/irq_service_dcn201.c33
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn21/irq_service_dcn21.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn30/irq_service_dcn30.c30
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn302/irq_service_dcn302.c19
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn303/irq_service_dcn303.c19
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn31/irq_service_dcn31.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn314/irq_service_dcn314.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn315/irq_service_dcn315.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn32/irq_service_dcn32.c90
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn35/irq_service_dcn35.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn351/irq_service_dcn351.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn36/irq_service_dcn36.c381
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn36/irq_service_dcn36.h12
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/dcn401/irq_service_dcn401.c89
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/irq_service.c64
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq/irq_service.h8
-rw-r--r--drivers/gpu/drm/amd/display/dc/irq_types.h17
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_cts.c67
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio.c69
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio_fixed_vs_pe_retimer.c7
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.c81
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_hpo_fixed_vs_pe_retimer_dp.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/link_detection.c32
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/link_dpms.c128
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/link_factory.c21
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/link_validation.c297
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/link_validation.h11
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_ddc.c2
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c232
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.c64
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.h6
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.c538
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.h55
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c52
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.c26
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c35
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.c114
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_auxless.c1
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_dpia.c14
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c76
-rw-r--r--drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/mmhubbub/Makefile2
-rw-r--r--drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.c16
-rw-r--r--drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.c18
-rw-r--r--drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.h7
-rw-r--r--drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.c385
-rw-r--r--drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.h8
-rw-r--r--drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.c25
-rw-r--r--drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.h27
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.c127
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h232
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn201/dcn201_optc.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn301/dcn301_optc.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.c73
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.h7
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c3
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.c105
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h3
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.c29
-rw-r--r--drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.h22
-rw-r--r--drivers/gpu/drm/amd/display/dc/os_types.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/Makefile34
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dce100/dce100_resource.c8
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dce110/dce110_resource.c6
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c6
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c17
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dce60/dce60_resource.c (renamed from drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.c)42
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dce60/dce60_resource.h (renamed from drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.h)0
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c6
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c21
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.h1
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c64
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c9
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c23
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.h2
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c35
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.h8
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn301/dcn301_resource.c13
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c4
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c74
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.h15
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c31
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c52
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c10
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c108
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h13
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c26
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c47
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.h4
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c38
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn36/dcn36_resource.c2181
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn36/dcn36_resource.h73
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c175
-rw-r--r--drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h7
-rw-r--r--drivers/gpu/drm/amd/display/dc/spl/dc_spl.h18
-rw-r--r--drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_filters.h22
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/Makefile (renamed from drivers/gpu/drm/amd/display/dc/spl/Makefile)2
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl.c (renamed from drivers/gpu/drm/amd/display/dc/spl/dc_spl.c)321
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl.h27
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl_filters.c (renamed from drivers/gpu/drm/amd/display/dc/spl/dc_spl_filters.c)0
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl_filters.h (renamed from drivers/gpu/drm/amd/display/dc/spl/dc_spl_filters.h)0
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl_isharp_filters.c (renamed from drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.c)452
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl_isharp_filters.h (renamed from drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h)18
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_easf_filters.c (renamed from drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_easf_filters.c)1058
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_easf_filters.h (renamed from drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_easf_filters.h)9
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_filters.c (renamed from drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_filters.c)232
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_filters.h13
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/dc_spl_types.h (renamed from drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h)34
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/spl_custom_float.c (renamed from drivers/gpu/drm/amd/display/dc/spl/spl_custom_float.c)0
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/spl_custom_float.h (renamed from drivers/gpu/drm/amd/display/dc/spl/spl_custom_float.h)0
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/spl_debug.h (renamed from drivers/gpu/drm/amd/display/dc/spl/spl_debug.h)0
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/spl_fixpt31_32.c (renamed from drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.c)6
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/spl_fixpt31_32.h (renamed from drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.h)4
-rw-r--r--drivers/gpu/drm/amd/display/dc/sspl/spl_os_types.h (renamed from drivers/gpu/drm/amd/display/dc/spl/spl_os_types.h)0
-rw-r--r--drivers/gpu/drm/amd/display/dmub/dmub_srv.h199
-rw-r--r--drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h820
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/Makefile1
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c85
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.h2
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c116
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.h2
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c98
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.h2
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c118
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.h2
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn36.c34
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn36.h13
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.c276
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.h9
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c381
-rw-r--r--drivers/gpu/drm/amd/display/dmub/src/dmub_srv_stat.c21
-rw-r--r--drivers/gpu/drm/amd/display/include/dal_asic_id.h1
-rw-r--r--drivers/gpu/drm/amd/display/include/dal_types.h1
-rw-r--r--drivers/gpu/drm/amd/display/include/gpio_service_interface.h3
-rw-r--r--drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h2
-rw-r--r--drivers/gpu/drm/amd/display/include/link_service_types.h2
-rw-r--r--drivers/gpu/drm/amd/display/include/logger_interface.h5
-rw-r--r--drivers/gpu/drm/amd/display/include/logger_types.h1
-rw-r--r--drivers/gpu/drm/amd/display/modules/freesync/freesync.c160
-rw-r--r--drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h1
-rw-r--r--drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c53
-rw-r--r--drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_transition.c48
-rw-r--r--drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c73
-rw-r--r--drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c3
-rw-r--r--drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h26
-rw-r--r--drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h38
-rw-r--r--drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c4
-rw-r--r--drivers/gpu/drm/amd/display/modules/power/power_helpers.c9
-rw-r--r--drivers/gpu/drm/amd/display/modules/power/power_helpers.h5
476 files changed, 24904 insertions, 26102 deletions
diff --git a/drivers/gpu/drm/amd/display/Kconfig b/drivers/gpu/drm/amd/display/Kconfig
index 11e3f2f3b174..abd3b6564373 100644
--- a/drivers/gpu/drm/amd/display/Kconfig
+++ b/drivers/gpu/drm/amd/display/Kconfig
@@ -8,6 +8,8 @@ config DRM_AMD_DC
bool "AMD DC - Enable new display engine"
default y
depends on BROKEN || !CC_IS_CLANG || ARM64 || LOONGARCH || RISCV || SPARC64 || X86_64
+ select CEC_CORE
+ select CEC_NOTIFIER
select SND_HDA_COMPONENT if SND_HDA_CORE
# !CC_IS_CLANG: https://github.com/ClangBuiltLinux/linux/issues/1752
select DRM_AMD_DC_FP if ARCH_HAS_KERNEL_FPU_SUPPORT && !(CC_IS_CLANG && (ARM64 || LOONGARCH || RISCV))
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
index ab2a97e354da..7329b8cc2576 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/Makefile
@@ -38,6 +38,7 @@ AMDGPUDM = \
amdgpu_dm_pp_smu.o \
amdgpu_dm_psr.o \
amdgpu_dm_replay.o \
+ amdgpu_dm_quirks.o \
amdgpu_dm_wb.o
ifdef CONFIG_DRM_AMD_DC_FP
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
index f0a6816709ca..cd0e2976e268 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -80,7 +80,6 @@
#include <linux/power_supply.h>
#include <linux/firmware.h>
#include <linux/component.h>
-#include <linux/dmi.h>
#include <linux/sort.h>
#include <drm/display/drm_dp_mst_helper.h>
@@ -93,10 +92,12 @@
#include <drm/drm_fourcc.h>
#include <drm/drm_edid.h>
#include <drm/drm_eld.h>
+#include <drm/drm_utils.h>
#include <drm/drm_vblank.h>
#include <drm/drm_audio_component.h>
#include <drm/drm_gem_atomic_helper.h>
+#include <media/cec-notifier.h>
#include <acpi/video.h>
#include "ivsrcid/dcn/irqsrcs_dcn_1_0.h"
@@ -113,6 +114,8 @@
#include "modules/inc/mod_freesync.h"
#include "modules/power/power_helpers.h"
+static_assert(AMDGPU_DMUB_NOTIFICATION_MAX == DMUB_NOTIFICATION_MAX, "AMDGPU_DMUB_NOTIFICATION_MAX mismatch");
+
#define FIRMWARE_RENOIR_DMUB "amdgpu/renoir_dmcub.bin"
MODULE_FIRMWARE(FIRMWARE_RENOIR_DMUB);
#define FIRMWARE_SIENNA_CICHLID_DMUB "amdgpu/sienna_cichlid_dmcub.bin"
@@ -153,6 +156,9 @@ MODULE_FIRMWARE(FIRMWARE_DCN_35_DMUB);
#define FIRMWARE_DCN_351_DMUB "amdgpu/dcn_3_5_1_dmcub.bin"
MODULE_FIRMWARE(FIRMWARE_DCN_351_DMUB);
+#define FIRMWARE_DCN_36_DMUB "amdgpu/dcn_3_6_dmcub.bin"
+MODULE_FIRMWARE(FIRMWARE_DCN_36_DMUB);
+
#define FIRMWARE_DCN_401_DMUB "amdgpu/dcn_4_0_1_dmcub.bin"
MODULE_FIRMWARE(FIRMWARE_DCN_401_DMUB);
@@ -177,6 +183,8 @@ static int amdgpu_dm_init(struct amdgpu_device *adev);
static void amdgpu_dm_fini(struct amdgpu_device *adev);
static bool is_freesync_video_mode(const struct drm_display_mode *mode, struct amdgpu_dm_connector *aconnector);
static void reset_freesync_config_for_crtc(struct dm_crtc_state *new_crtc_state);
+static struct amdgpu_i2c_adapter *
+create_i2c(struct ddc_service *ddc_service, bool oem);
static enum drm_mode_subconnector get_subconnector_type(struct dc_link *link)
{
@@ -243,6 +251,10 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev,
static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector);
static void handle_hpd_rx_irq(void *param);
+static void amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm,
+ int bl_idx,
+ u32 user_brightness);
+
static bool
is_timing_unchanged_for_freesync(struct drm_crtc_state *old_crtc_state,
struct drm_crtc_state *new_crtc_state);
@@ -269,7 +281,7 @@ static u32 dm_vblank_get_counter(struct amdgpu_device *adev, int crtc)
acrtc = adev->mode_info.crtcs[crtc];
if (!acrtc->dm_irq_params.stream) {
- DRM_ERROR("dc_stream_state is NULL for crtc '%d'!\n",
+ drm_err(adev_to_drm(adev), "dc_stream_state is NULL for crtc '%d'!\n",
crtc);
return 0;
}
@@ -290,7 +302,7 @@ static int dm_crtc_get_scanoutpos(struct amdgpu_device *adev, int crtc,
acrtc = adev->mode_info.crtcs[crtc];
if (!acrtc->dm_irq_params.stream) {
- DRM_ERROR("dc_stream_state is NULL for crtc '%d'!\n",
+ drm_err(adev_to_drm(adev), "dc_stream_state is NULL for crtc '%d'!\n",
crtc);
return 0;
}
@@ -314,7 +326,7 @@ static int dm_crtc_get_scanoutpos(struct amdgpu_device *adev, int crtc,
return 0;
}
-static bool dm_is_idle(void *handle)
+static bool dm_is_idle(struct amdgpu_ip_block *ip_block)
{
/* XXX todo */
return true;
@@ -361,6 +373,8 @@ get_crtc_by_otg_inst(struct amdgpu_device *adev,
static inline bool is_dc_timing_adjust_needed(struct dm_crtc_state *old_state,
struct dm_crtc_state *new_state)
{
+ if (new_state->stream->adjust.timing_adjust_pending)
+ return true;
if (new_state->freesync_config.state == VRR_STATE_ACTIVE_FIXED)
return true;
else if (amdgpu_dm_crtc_vrr_active(old_state) != amdgpu_dm_crtc_vrr_active(new_state))
@@ -738,6 +752,29 @@ static void dmub_aux_setconfig_callback(struct amdgpu_device *adev,
complete(&adev->dm.dmub_aux_transfer_done);
}
+static void dmub_aux_fused_io_callback(struct amdgpu_device *adev,
+ struct dmub_notification *notify)
+{
+ if (!adev || !notify) {
+ ASSERT(false);
+ return;
+ }
+
+ const struct dmub_cmd_fused_request *req = &notify->fused_request;
+ const uint8_t ddc_line = req->u.aux.ddc_line;
+
+ if (ddc_line >= ARRAY_SIZE(adev->dm.fused_io)) {
+ ASSERT(false);
+ return;
+ }
+
+ struct fused_io_sync *sync = &adev->dm.fused_io[ddc_line];
+
+ static_assert(sizeof(*req) <= sizeof(sync->reply_data), "Size mismatch");
+ memcpy(sync->reply_data, req, sizeof(*req));
+ complete(&sync->replied);
+}
+
/**
* dmub_hpd_callback - DMUB HPD interrupt processing callback.
* @adev: amdgpu_device pointer
@@ -761,18 +798,18 @@ static void dmub_hpd_callback(struct amdgpu_device *adev,
return;
if (notify == NULL) {
- DRM_ERROR("DMUB HPD callback notification was NULL");
+ drm_err(adev_to_drm(adev), "DMUB HPD callback notification was NULL");
return;
}
if (notify->link_index > adev->dm.dc->link_count) {
- DRM_ERROR("DMUB HPD index (%u)is abnormal", notify->link_index);
+ drm_err(adev_to_drm(adev), "DMUB HPD index (%u)is abnormal", notify->link_index);
return;
}
/* Skip DMUB HPD IRQ in suspend/resume. We will probe them later. */
if (notify->type == DMUB_NOTIFICATION_HPD && adev->in_suspend) {
- DRM_INFO("Skip DMUB HPD IRQ callback in suspend/resume\n");
+ drm_info(adev_to_drm(adev), "Skip DMUB HPD IRQ callback in suspend/resume\n");
return;
}
@@ -789,11 +826,11 @@ static void dmub_hpd_callback(struct amdgpu_device *adev,
aconnector = to_amdgpu_dm_connector(connector);
if (link && aconnector->dc_link == link) {
if (notify->type == DMUB_NOTIFICATION_HPD)
- DRM_INFO("DMUB HPD IRQ callback: link_index=%u\n", link_index);
+ drm_info(adev_to_drm(adev), "DMUB HPD IRQ callback: link_index=%u\n", link_index);
else if (notify->type == DMUB_NOTIFICATION_HPD_IRQ)
- DRM_INFO("DMUB HPD RX IRQ callback: link_index=%u\n", link_index);
+ drm_info(adev_to_drm(adev), "DMUB HPD RX IRQ callback: link_index=%u\n", link_index);
else
- DRM_WARN("DMUB Unknown HPD callback type %d, link_index=%u\n",
+ drm_warn(adev_to_drm(adev), "DMUB Unknown HPD callback type %d, link_index=%u\n",
notify->type, link_index);
hpd_aconnector = aconnector;
@@ -805,7 +842,7 @@ static void dmub_hpd_callback(struct amdgpu_device *adev,
if (hpd_aconnector) {
if (notify->type == DMUB_NOTIFICATION_HPD) {
if (hpd_aconnector->dc_link->hpd_status == (notify->hpd_status == DP_HPD_PLUG))
- DRM_WARN("DMUB reported hpd status unchanged. link_index=%u\n", link_index);
+ drm_warn(adev_to_drm(adev), "DMUB reported hpd status unchanged. link_index=%u\n", link_index);
handle_hpd_irq_helper(hpd_aconnector);
} else if (notify->type == DMUB_NOTIFICATION_HPD_IRQ) {
handle_hpd_rx_irq(hpd_aconnector);
@@ -824,7 +861,7 @@ static void dmub_hpd_callback(struct amdgpu_device *adev,
static void dmub_hpd_sense_callback(struct amdgpu_device *adev,
struct dmub_notification *notify)
{
- DRM_DEBUG_DRIVER("DMUB HPD SENSE callback.\n");
+ drm_dbg_driver(adev_to_drm(adev), "DMUB HPD SENSE callback.\n");
}
/**
@@ -860,7 +897,7 @@ static void dm_handle_hpd_work(struct work_struct *work)
dmub_hpd_wrk = container_of(work, struct dmub_hpd_work, handle_hpd_work);
if (!dmub_hpd_wrk->dmub_notify) {
- DRM_ERROR("dmub_hpd_wrk dmub_notify is NULL");
+ drm_err(adev_to_drm(dmub_hpd_wrk->adev), "dmub_hpd_wrk dmub_notify is NULL");
return;
}
@@ -874,6 +911,30 @@ static void dm_handle_hpd_work(struct work_struct *work)
}
+static const char *dmub_notification_type_str(enum dmub_notification_type e)
+{
+ switch (e) {
+ case DMUB_NOTIFICATION_NO_DATA:
+ return "NO_DATA";
+ case DMUB_NOTIFICATION_AUX_REPLY:
+ return "AUX_REPLY";
+ case DMUB_NOTIFICATION_HPD:
+ return "HPD";
+ case DMUB_NOTIFICATION_HPD_IRQ:
+ return "HPD_IRQ";
+ case DMUB_NOTIFICATION_SET_CONFIG_REPLY:
+ return "SET_CONFIG_REPLY";
+ case DMUB_NOTIFICATION_DPIA_NOTIFICATION:
+ return "DPIA_NOTIFICATION";
+ case DMUB_NOTIFICATION_HPD_SENSE_NOTIFY:
+ return "HPD_SENSE_NOTIFY";
+ case DMUB_NOTIFICATION_FUSED_IO:
+ return "FUSED_IO";
+ default:
+ return "<unknown>";
+ }
+}
+
#define DMUB_TRACE_MAX_READ 64
/**
* dm_dmub_outbox1_low_irq() - Handles Outbox interrupt
@@ -891,22 +952,13 @@ static void dm_dmub_outbox1_low_irq(void *interrupt_params)
struct dmcub_trace_buf_entry entry = { 0 };
u32 count = 0;
struct dmub_hpd_work *dmub_hpd_wrk;
- static const char *const event_type[] = {
- "NO_DATA",
- "AUX_REPLY",
- "HPD",
- "HPD_IRQ",
- "SET_CONFIGC_REPLY",
- "DPIA_NOTIFICATION",
- "HPD_SENSE_NOTIFY",
- };
do {
if (dc_dmub_srv_get_dmub_outbox0_msg(dm->dc, &entry)) {
trace_amdgpu_dmub_trace_high_irq(entry.trace_code, entry.tick_count,
entry.param0, entry.param1);
- DRM_DEBUG_DRIVER("trace_code:%u, tick_count:%u, param0:%u, param1:%u\n",
+ drm_dbg_driver(adev_to_drm(adev), "trace_code:%u, tick_count:%u, param0:%u, param1:%u\n",
entry.trace_code, entry.tick_count, entry.param0, entry.param1);
} else
break;
@@ -916,7 +968,7 @@ static void dm_dmub_outbox1_low_irq(void *interrupt_params)
} while (count <= DMUB_TRACE_MAX_READ);
if (count > DMUB_TRACE_MAX_READ)
- DRM_DEBUG_DRIVER("Warning : count > DMUB_TRACE_MAX_READ");
+ drm_dbg_driver(adev_to_drm(adev), "Warning : count > DMUB_TRACE_MAX_READ");
if (dc_enable_dmub_notifications(adev->dm.dc) &&
irq_params->irq_src == DC_IRQ_SOURCE_DMCUB_OUTBOX) {
@@ -924,25 +976,25 @@ static void dm_dmub_outbox1_low_irq(void *interrupt_params)
do {
dc_stat_get_dmub_notification(adev->dm.dc, &notify);
if (notify.type >= ARRAY_SIZE(dm->dmub_thread_offload)) {
- DRM_ERROR("DM: notify type %d invalid!", notify.type);
+ drm_err(adev_to_drm(adev), "DM: notify type %d invalid!", notify.type);
continue;
}
if (!dm->dmub_callback[notify.type]) {
- DRM_WARN("DMUB notification skipped due to no handler: type=%s\n",
- event_type[notify.type]);
+ drm_warn(adev_to_drm(adev), "DMUB notification skipped due to no handler: type=%s\n",
+ dmub_notification_type_str(notify.type));
continue;
}
if (dm->dmub_thread_offload[notify.type] == true) {
dmub_hpd_wrk = kzalloc(sizeof(*dmub_hpd_wrk), GFP_ATOMIC);
if (!dmub_hpd_wrk) {
- DRM_ERROR("Failed to allocate dmub_hpd_wrk");
+ drm_err(adev_to_drm(adev), "Failed to allocate dmub_hpd_wrk");
return;
}
dmub_hpd_wrk->dmub_notify = kmemdup(&notify, sizeof(struct dmub_notification),
GFP_ATOMIC);
if (!dmub_hpd_wrk->dmub_notify) {
kfree(dmub_hpd_wrk);
- DRM_ERROR("Failed to allocate dmub_hpd_wrk->dmub_notify");
+ drm_err(adev_to_drm(adev), "Failed to allocate dmub_hpd_wrk->dmub_notify");
return;
}
INIT_WORK(&dmub_hpd_wrk->handle_hpd_work, dm_handle_hpd_work);
@@ -955,13 +1007,13 @@ static void dm_dmub_outbox1_low_irq(void *interrupt_params)
}
}
-static int dm_set_clockgating_state(void *handle,
+static int dm_set_clockgating_state(struct amdgpu_ip_block *ip_block,
enum amd_clockgating_state state)
{
return 0;
}
-static int dm_set_powergating_state(void *handle,
+static int dm_set_powergating_state(struct amdgpu_ip_block *ip_block,
enum amd_powergating_state state)
{
return 0;
@@ -1000,10 +1052,10 @@ static void amdgpu_dm_fbc_init(struct drm_connector *connector)
&compressor->gpu_addr, &compressor->cpu_addr);
if (r)
- DRM_ERROR("DM: Failed to initialize FBC\n");
+ drm_err(adev_to_drm(adev), "DM: Failed to initialize FBC\n");
else {
adev->dm.dc->ctx->fbc_gpu_addr = compressor->gpu_addr;
- DRM_INFO("DM: FBC alloc %lu\n", max_size*4);
+ drm_info(adev_to_drm(adev), "DM: FBC alloc %lu\n", max_size*4);
}
}
@@ -1036,8 +1088,10 @@ static int amdgpu_dm_audio_component_get_eld(struct device *kdev, int port,
continue;
*enabled = true;
+ mutex_lock(&connector->eld_mutex);
ret = drm_eld_size(connector->eld);
memcpy(buf, connector->eld, min(max_bytes, ret));
+ mutex_unlock(&connector->eld_mutex);
break;
}
@@ -1166,13 +1220,13 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
return 0;
if (!fb_info) {
- DRM_ERROR("No framebuffer info for DMUB service.\n");
+ drm_err(adev_to_drm(adev), "No framebuffer info for DMUB service.\n");
return -EINVAL;
}
if (!dmub_fw) {
/* Firmware required for DMUB support. */
- DRM_ERROR("No firmware provided for DMUB.\n");
+ drm_err(adev_to_drm(adev), "No firmware provided for DMUB.\n");
return -EINVAL;
}
@@ -1182,19 +1236,19 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
status = dmub_srv_has_hw_support(dmub_srv, &has_hw_support);
if (status != DMUB_STATUS_OK) {
- DRM_ERROR("Error checking HW support for DMUB: %d\n", status);
+ drm_err(adev_to_drm(adev), "Error checking HW support for DMUB: %d\n", status);
return -EINVAL;
}
if (!has_hw_support) {
- DRM_INFO("DMUB unsupported on ASIC\n");
+ drm_info(adev_to_drm(adev), "DMUB unsupported on ASIC\n");
return 0;
}
/* Reset DMCUB if it was previously running - before we overwrite its memory. */
status = dmub_srv_hw_reset(dmub_srv);
if (status != DMUB_STATUS_OK)
- DRM_WARN("Error resetting DMUB HW: %d\n", status);
+ drm_warn(adev_to_drm(adev), "Error resetting DMUB HW: %d\n", status);
hdr = (const struct dmcub_firmware_header_v1_0 *)dmub_fw->data;
@@ -1263,6 +1317,7 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
case IP_VERSION(3, 1, 4):
case IP_VERSION(3, 5, 0):
case IP_VERSION(3, 5, 1):
+ case IP_VERSION(3, 6, 0):
case IP_VERSION(4, 0, 1):
hw_params.dpia_supported = true;
hw_params.disable_dpia = adev->dm.dc->debug.dpia_debug.bits.disable_dpia;
@@ -1274,7 +1329,9 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
switch (amdgpu_ip_version(adev, DCE_HWIP, 0)) {
case IP_VERSION(3, 5, 0):
case IP_VERSION(3, 5, 1):
+ case IP_VERSION(3, 6, 0):
hw_params.ips_sequential_ono = adev->external_rev_id > 0x10;
+ hw_params.lower_hbr3_phy_ssc = true;
break;
default:
break;
@@ -1282,14 +1339,14 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
status = dmub_srv_hw_init(dmub_srv, &hw_params);
if (status != DMUB_STATUS_OK) {
- DRM_ERROR("Error initializing DMUB HW: %d\n", status);
+ drm_err(adev_to_drm(adev), "Error initializing DMUB HW: %d\n", status);
return -EINVAL;
}
/* Wait for firmware load to finish. */
status = dmub_srv_wait_for_auto_load(dmub_srv, 100000);
if (status != DMUB_STATUS_OK)
- DRM_WARN("Wait for DMUB auto-load failed: %d\n", status);
+ drm_warn(adev_to_drm(adev), "Wait for DMUB auto-load failed: %d\n", status);
/* Init DMCU and ABM if available. */
if (dmcu && abm) {
@@ -1300,11 +1357,11 @@ static int dm_dmub_hw_init(struct amdgpu_device *adev)
if (!adev->dm.dc->ctx->dmub_srv)
adev->dm.dc->ctx->dmub_srv = dc_dmub_srv_create(adev->dm.dc, dmub_srv);
if (!adev->dm.dc->ctx->dmub_srv) {
- DRM_ERROR("Couldn't allocate DC DMUB server!\n");
+ drm_err(adev_to_drm(adev), "Couldn't allocate DC DMUB server!\n");
return -ENOMEM;
}
- DRM_INFO("DMUB hardware initialized: version=0x%08X\n",
+ drm_info(adev_to_drm(adev), "DMUB hardware initialized: version=0x%08X\n",
adev->dm.dmcub_fw_version);
/* Keeping sanity checks off if
@@ -1347,18 +1404,18 @@ static void dm_dmub_hw_resume(struct amdgpu_device *adev)
status = dmub_srv_is_hw_init(dmub_srv, &init);
if (status != DMUB_STATUS_OK)
- DRM_WARN("DMUB hardware init check failed: %d\n", status);
+ drm_warn(adev_to_drm(adev), "DMUB hardware init check failed: %d\n", status);
if (status == DMUB_STATUS_OK && init) {
/* Wait for firmware load to finish. */
status = dmub_srv_wait_for_auto_load(dmub_srv, 100000);
if (status != DMUB_STATUS_OK)
- DRM_WARN("Wait for DMUB auto-load failed: %d\n", status);
+ drm_warn(adev_to_drm(adev), "Wait for DMUB auto-load failed: %d\n", status);
} else {
/* Perform the full hardware initialization. */
r = dm_dmub_hw_init(adev);
if (r)
- DRM_ERROR("DMUB interface failed to initialize: status=%d\n", r);
+ drm_err(adev_to_drm(adev), "DMUB interface failed to initialize: status=%d\n", r);
}
}
@@ -1468,18 +1525,18 @@ static void dm_handle_hpd_rx_offload_work(struct work_struct *work)
offload_work = container_of(work, struct hpd_rx_irq_offload_work, work);
aconnector = offload_work->offload_wq->aconnector;
+ adev = offload_work->adev;
if (!aconnector) {
- DRM_ERROR("Can't retrieve aconnector in hpd_rx_irq_offload_work");
+ drm_err(adev_to_drm(adev), "Can't retrieve aconnector in hpd_rx_irq_offload_work");
goto skip;
}
- adev = drm_to_adev(aconnector->base.dev);
dc_link = aconnector->dc_link;
mutex_lock(&aconnector->hpd_lock);
if (!dc_link_detect_connection_type(dc_link, &new_connection_type))
- DRM_ERROR("KMS: Failed to detect connector\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to detect connector\n");
mutex_unlock(&aconnector->hpd_lock);
if (new_connection_type == dc_connection_none)
@@ -1548,8 +1605,9 @@ skip:
}
-static struct hpd_rx_irq_offload_work_queue *hpd_rx_irq_create_workqueue(struct dc *dc)
+static struct hpd_rx_irq_offload_work_queue *hpd_rx_irq_create_workqueue(struct amdgpu_device *adev)
{
+ struct dc *dc = adev->dm.dc;
int max_caps = dc->caps.max_links;
int i = 0;
struct hpd_rx_irq_offload_work_queue *hpd_rx_offload_wq = NULL;
@@ -1565,7 +1623,7 @@ static struct hpd_rx_irq_offload_work_queue *hpd_rx_irq_create_workqueue(struct
create_singlethread_workqueue("amdgpu_dm_hpd_rx_offload_wq");
if (hpd_rx_offload_wq[i].wq == NULL) {
- DRM_ERROR("create amdgpu_dm_hpd_rx_offload_wq fail!");
+ drm_err(adev_to_drm(adev), "create amdgpu_dm_hpd_rx_offload_wq fail!");
goto out_err;
}
@@ -1614,77 +1672,6 @@ static bool dm_should_disable_stutter(struct pci_dev *pdev)
return false;
}
-static const struct dmi_system_id hpd_disconnect_quirk_table[] = {
- {
- .matches = {
- DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
- DMI_MATCH(DMI_PRODUCT_NAME, "Precision 3660"),
- },
- },
- {
- .matches = {
- DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
- DMI_MATCH(DMI_PRODUCT_NAME, "Precision 3260"),
- },
- },
- {
- .matches = {
- DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
- DMI_MATCH(DMI_PRODUCT_NAME, "Precision 3460"),
- },
- },
- {
- .matches = {
- DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
- DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex Tower Plus 7010"),
- },
- },
- {
- .matches = {
- DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
- DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex Tower 7010"),
- },
- },
- {
- .matches = {
- DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
- DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex SFF Plus 7010"),
- },
- },
- {
- .matches = {
- DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
- DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex SFF 7010"),
- },
- },
- {
- .matches = {
- DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
- DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex Micro Plus 7010"),
- },
- },
- {
- .matches = {
- DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
- DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex Micro 7010"),
- },
- },
- {}
- /* TODO: refactor this from a fixed table to a dynamic option */
-};
-
-static void retrieve_dmi_info(struct amdgpu_display_manager *dm)
-{
- const struct dmi_system_id *dmi_id;
-
- dm->aux_hpd_discon_quirk = false;
-
- dmi_id = dmi_first_match(hpd_disconnect_quirk_table);
- if (dmi_id) {
- dm->aux_hpd_discon_quirk = true;
- DRM_INFO("aux_hpd_discon_quirk attached\n");
- }
-}
void*
dm_allocate_gpu_mem(
@@ -1771,10 +1758,11 @@ dm_dmub_send_vbios_gpint_command(struct amdgpu_device *adev,
return DMUB_STATUS_TIMEOUT;
}
-static struct dml2_soc_bb *dm_dmub_get_vbios_bounding_box(struct amdgpu_device *adev)
+static void *dm_dmub_get_vbios_bounding_box(struct amdgpu_device *adev)
{
- struct dml2_soc_bb *bb;
+ void *bb;
long long addr;
+ unsigned int bb_size;
int i = 0;
uint16_t chunk;
enum dmub_gpint_command send_addrs[] = {
@@ -1787,6 +1775,7 @@ static struct dml2_soc_bb *dm_dmub_get_vbios_bounding_box(struct amdgpu_device *
switch (amdgpu_ip_version(adev, DCE_HWIP, 0)) {
case IP_VERSION(4, 0, 1):
+ bb_size = sizeof(struct dml2_soc_bb);
break;
default:
return NULL;
@@ -1794,7 +1783,7 @@ static struct dml2_soc_bb *dm_dmub_get_vbios_bounding_box(struct amdgpu_device *
bb = dm_allocate_gpu_mem(adev,
DC_MEM_ALLOC_TYPE_GART,
- sizeof(struct dml2_soc_bb),
+ bb_size,
&addr);
if (!bb)
return NULL;
@@ -1828,26 +1817,7 @@ static enum dmub_ips_disable_type dm_get_default_ips_mode(
switch (amdgpu_ip_version(adev, DCE_HWIP, 0)) {
case IP_VERSION(3, 5, 0):
- /*
- * On DCN35 systems with Z8 enabled, it's possible for IPS2 + Z8 to
- * cause a hard hang. A fix exists for newer PMFW.
- *
- * As a workaround, for non-fixed PMFW, force IPS1+RCG as the deepest
- * IPS state in all cases, except for s0ix and all displays off (DPMS),
- * where IPS2 is allowed.
- *
- * When checking pmfw version, use the major and minor only.
- */
- if ((adev->pm.fw_version & 0x00FFFF00) < 0x005D6300)
- ret = DMUB_IPS_RCG_IN_ACTIVE_IPS2_IN_OFF;
- else if (amdgpu_ip_version(adev, GC_HWIP, 0) > IP_VERSION(11, 5, 0))
- /*
- * Other ASICs with DCN35 that have residency issues with
- * IPS2 in idle.
- * We want them to use IPS2 only in display off cases.
- */
- ret = DMUB_IPS_RCG_IN_ACTIVE_IPS2_IN_OFF;
- break;
+ case IP_VERSION(3, 6, 0):
case IP_VERSION(3, 5, 1):
ret = DMUB_IPS_RCG_IN_ACTIVE_IPS2_IN_OFF;
break;
@@ -1879,7 +1849,7 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
mutex_init(&adev->dm.audio_lock);
if (amdgpu_dm_irq_init(adev)) {
- DRM_ERROR("amdgpu: failed to initialize DM IRQ support.\n");
+ drm_err(adev_to_drm(adev), "failed to initialize DM IRQ support.\n");
goto error;
}
@@ -1964,7 +1934,7 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
if (amdgpu_device_seamless_boot_supported(adev)) {
init_data.flags.seamless_boot_edp_requested = true;
init_data.flags.allow_seamless_boot_optimization = true;
- DRM_INFO("Seamless boot condition check passed\n");
+ drm_dbg(adev->dm.ddev, "Seamless boot requested\n");
}
init_data.flags.enable_mipi_converter_optimization = true;
@@ -1991,6 +1961,8 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
init_data.num_virtual_links = 1;
retrieve_dmi_info(&adev->dm);
+ if (adev->dm.edp0_on_dp1_quirk)
+ init_data.flags.support_edp0_on_dp1 = true;
if (adev->dm.bb_from_dmub)
init_data.bb_from_dmub = adev->dm.bb_from_dmub;
@@ -2001,10 +1973,10 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
adev->dm.dc = dc_create(&init_data);
if (adev->dm.dc) {
- DRM_INFO("Display Core v%s initialized on %s\n", DC_VER,
+ drm_info(adev_to_drm(adev), "Display Core v%s initialized on %s\n", DC_VER,
dce_version_to_string(adev->dm.dc->ctx->dce_version));
} else {
- DRM_INFO("Display Core failed to initialize with v%s!\n", DC_VER);
+ drm_info(adev_to_drm(adev), "Display Core failed to initialize with v%s!\n", DC_VER);
goto error;
}
@@ -2030,30 +2002,44 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
if (amdgpu_dc_debug_mask & DC_FORCE_SUBVP_MCLK_SWITCH)
adev->dm.dc->debug.force_subvp_mclk_switch = true;
+ if (amdgpu_dc_debug_mask & DC_DISABLE_SUBVP_FAMS) {
+ adev->dm.dc->debug.force_disable_subvp = true;
+ adev->dm.dc->debug.fams2_config.bits.enable = false;
+ }
+
if (amdgpu_dc_debug_mask & DC_ENABLE_DML2) {
adev->dm.dc->debug.using_dml2 = true;
adev->dm.dc->debug.using_dml21 = true;
}
+ if (amdgpu_dc_debug_mask & DC_HDCP_LC_FORCE_FW_ENABLE)
+ adev->dm.dc->debug.hdcp_lc_force_fw_enable = true;
+
+ if (amdgpu_dc_debug_mask & DC_HDCP_LC_ENABLE_SW_FALLBACK)
+ adev->dm.dc->debug.hdcp_lc_enable_sw_fallback = true;
+
+ if (amdgpu_dc_debug_mask & DC_SKIP_DETECTION_LT)
+ adev->dm.dc->debug.skip_detection_link_training = true;
+
adev->dm.dc->debug.visual_confirm = amdgpu_dc_visual_confirm;
/* TODO: Remove after DP2 receiver gets proper support of Cable ID feature */
adev->dm.dc->debug.ignore_cable_id = true;
if (adev->dm.dc->caps.dp_hdmi21_pcon_support)
- DRM_INFO("DP-HDMI FRL PCON supported\n");
+ drm_info(adev_to_drm(adev), "DP-HDMI FRL PCON supported\n");
r = dm_dmub_hw_init(adev);
if (r) {
- DRM_ERROR("DMUB interface failed to initialize: status=%d\n", r);
+ drm_err(adev_to_drm(adev), "DMUB interface failed to initialize: status=%d\n", r);
goto error;
}
dc_hardware_init(adev->dm.dc);
- adev->dm.hpd_rx_offload_wq = hpd_rx_irq_create_workqueue(adev->dm.dc);
+ adev->dm.hpd_rx_offload_wq = hpd_rx_irq_create_workqueue(adev);
if (!adev->dm.hpd_rx_offload_wq) {
- DRM_ERROR("amdgpu: failed to create hpd rx offload workqueue.\n");
+ drm_err(adev_to_drm(adev), "failed to create hpd rx offload workqueue.\n");
goto error;
}
@@ -2068,10 +2054,10 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
adev->dm.freesync_module = mod_freesync_create(adev->dm.dc);
if (!adev->dm.freesync_module) {
- DRM_ERROR(
- "amdgpu: failed to initialize freesync_module.\n");
+ drm_err(adev_to_drm(adev),
+ "failed to initialize freesync_module.\n");
} else
- DRM_DEBUG_DRIVER("amdgpu: freesync_module init done %p.\n",
+ drm_dbg_driver(adev_to_drm(adev), "amdgpu: freesync_module init done %p.\n",
adev->dm.freesync_module);
amdgpu_dm_init_color_mod();
@@ -2080,7 +2066,7 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
adev->dm.vblank_control_workqueue =
create_singlethread_workqueue("dm_vblank_control_workqueue");
if (!adev->dm.vblank_control_workqueue)
- DRM_ERROR("amdgpu: failed to initialize vblank_workqueue.\n");
+ drm_err(adev_to_drm(adev), "failed to initialize vblank_workqueue.\n");
}
if (adev->dm.dc->caps.ips_support &&
@@ -2091,9 +2077,9 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
adev->dm.hdcp_workqueue = hdcp_create_workqueue(adev, &init_params.cp_psp, adev->dm.dc);
if (!adev->dm.hdcp_workqueue)
- DRM_ERROR("amdgpu: failed to initialize hdcp_workqueue.\n");
+ drm_err(adev_to_drm(adev), "failed to initialize hdcp_workqueue.\n");
else
- DRM_DEBUG_DRIVER("amdgpu: hdcp_workqueue init done %p.\n", adev->dm.hdcp_workqueue);
+ drm_dbg_driver(adev_to_drm(adev), "amdgpu: hdcp_workqueue init done %p.\n", adev->dm.hdcp_workqueue);
dc_init_callbacks(adev->dm.dc, &init_params);
}
@@ -2101,20 +2087,29 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
init_completion(&adev->dm.dmub_aux_transfer_done);
adev->dm.dmub_notify = kzalloc(sizeof(struct dmub_notification), GFP_KERNEL);
if (!adev->dm.dmub_notify) {
- DRM_INFO("amdgpu: fail to allocate adev->dm.dmub_notify");
+ drm_info(adev_to_drm(adev), "fail to allocate adev->dm.dmub_notify");
goto error;
}
adev->dm.delayed_hpd_wq = create_singlethread_workqueue("amdgpu_dm_hpd_wq");
if (!adev->dm.delayed_hpd_wq) {
- DRM_ERROR("amdgpu: failed to create hpd offload workqueue.\n");
+ drm_err(adev_to_drm(adev), "failed to create hpd offload workqueue.\n");
goto error;
}
amdgpu_dm_outbox_init(adev);
if (!register_dmub_notify_callback(adev, DMUB_NOTIFICATION_AUX_REPLY,
dmub_aux_setconfig_callback, false)) {
- DRM_ERROR("amdgpu: fail to register dmub aux callback");
+ drm_err(adev_to_drm(adev), "fail to register dmub aux callback");
+ goto error;
+ }
+
+ for (size_t i = 0; i < ARRAY_SIZE(adev->dm.fused_io); i++)
+ init_completion(&adev->dm.fused_io[i].replied);
+
+ if (!register_dmub_notify_callback(adev, DMUB_NOTIFICATION_FUSED_IO,
+ dmub_aux_fused_io_callback, false)) {
+ drm_err(adev_to_drm(adev), "fail to register dmub fused io callback");
goto error;
}
/* Enable outbox notification only after IRQ handlers are registered and DMUB is alive.
@@ -2131,8 +2126,8 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
}
if (amdgpu_dm_initialize_drm_device(adev)) {
- DRM_ERROR(
- "amdgpu: failed to initialize sw for display support.\n");
+ drm_err(adev_to_drm(adev),
+ "failed to initialize sw for display support.\n");
goto error;
}
@@ -2146,18 +2141,22 @@ static int amdgpu_dm_init(struct amdgpu_device *adev)
adev_to_drm(adev)->mode_config.cursor_height = adev->dm.dc->caps.max_cursor_size;
if (drm_vblank_init(adev_to_drm(adev), adev->dm.display_indexes_num)) {
- DRM_ERROR(
- "amdgpu: failed to initialize sw for display support.\n");
+ drm_err(adev_to_drm(adev),
+ "failed to initialize sw for display support.\n");
goto error;
}
#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
- adev->dm.secure_display_ctxs = amdgpu_dm_crtc_secure_display_create_contexts(adev);
- if (!adev->dm.secure_display_ctxs)
- DRM_ERROR("amdgpu: failed to initialize secure display contexts.\n");
+ amdgpu_dm_crtc_secure_display_create_contexts(adev);
+ if (!adev->dm.secure_display_ctx.crtc_ctx)
+ drm_err(adev_to_drm(adev), "failed to initialize secure display contexts.\n");
+
+ if (amdgpu_ip_version(adev, DCE_HWIP, 0) >= IP_VERSION(4, 0, 1))
+ adev->dm.secure_display_ctx.support_mul_roi = true;
+
#endif
- DRM_DEBUG_DRIVER("KMS initialized.\n");
+ drm_dbg_driver(adev_to_drm(adev), "KMS initialized.\n");
return 0;
error:
@@ -2197,15 +2196,15 @@ static void amdgpu_dm_fini(struct amdgpu_device *adev)
amdgpu_dm_destroy_drm_device(&adev->dm);
#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
- if (adev->dm.secure_display_ctxs) {
+ if (adev->dm.secure_display_ctx.crtc_ctx) {
for (i = 0; i < adev->mode_info.num_crtc; i++) {
- if (adev->dm.secure_display_ctxs[i].crtc) {
- flush_work(&adev->dm.secure_display_ctxs[i].notify_ta_work);
- flush_work(&adev->dm.secure_display_ctxs[i].forward_roi_work);
+ if (adev->dm.secure_display_ctx.crtc_ctx[i].crtc) {
+ flush_work(&adev->dm.secure_display_ctx.crtc_ctx[i].notify_ta_work);
+ flush_work(&adev->dm.secure_display_ctx.crtc_ctx[i].forward_roi_work);
}
}
- kfree(adev->dm.secure_display_ctxs);
- adev->dm.secure_display_ctxs = NULL;
+ kfree(adev->dm.secure_display_ctx.crtc_ctx);
+ adev->dm.secure_display_ctx.crtc_ctx = NULL;
}
#endif
if (adev->dm.hdcp_workqueue) {
@@ -2324,12 +2323,13 @@ static int load_dmcu_fw(struct amdgpu_device *adev)
case IP_VERSION(3, 2, 1):
case IP_VERSION(3, 5, 0):
case IP_VERSION(3, 5, 1):
+ case IP_VERSION(3, 6, 0):
case IP_VERSION(4, 0, 1):
return 0;
default:
break;
}
- DRM_ERROR("Unsupported ASIC type: 0x%X\n", adev->asic_type);
+ drm_err(adev_to_drm(adev), "Unsupported ASIC type: 0x%X\n", adev->asic_type);
return -EINVAL;
}
@@ -2338,7 +2338,8 @@ static int load_dmcu_fw(struct amdgpu_device *adev)
return 0;
}
- r = amdgpu_ucode_request(adev, &adev->dm.fw_dmcu, "%s", fw_name_dmcu);
+ r = amdgpu_ucode_request(adev, &adev->dm.fw_dmcu, AMDGPU_UCODE_REQUIRED,
+ "%s", fw_name_dmcu);
if (r == -ENODEV) {
/* DMCU firmware is not necessary, so don't raise a fuss if it's missing */
DRM_DEBUG_KMS("dm: DMCU firmware not found\n");
@@ -2346,7 +2347,7 @@ static int load_dmcu_fw(struct amdgpu_device *adev)
return 0;
}
if (r) {
- dev_err(adev->dev, "amdgpu_dm: Can't validate firmware \"%s\"\n",
+ drm_err(adev_to_drm(adev), "amdgpu_dm: Can't validate firmware \"%s\"\n",
fw_name_dmcu);
amdgpu_ucode_release(&adev->dm.fw_dmcu);
return r;
@@ -2405,6 +2406,7 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev)
DMUB_WINDOW_MEMORY_TYPE_FB, //DMUB_WINDOW_5_TRACEBUFF
DMUB_WINDOW_MEMORY_TYPE_FB, //DMUB_WINDOW_6_FW_STATE
DMUB_WINDOW_MEMORY_TYPE_FB, //DMUB_WINDOW_7_SCRATCH_MEM
+ DMUB_WINDOW_MEMORY_TYPE_FB, //DMUB_WINDOW_IB_MEM
DMUB_WINDOW_MEMORY_TYPE_FB, //DMUB_WINDOW_SHARED_STATE
};
int r;
@@ -2448,6 +2450,9 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev)
case IP_VERSION(3, 5, 1):
dmub_asic = DMUB_ASIC_DCN35;
break;
+ case IP_VERSION(3, 6, 0):
+ dmub_asic = DMUB_ASIC_DCN36;
+ break;
case IP_VERSION(4, 0, 1):
dmub_asic = DMUB_ASIC_DCN401;
break;
@@ -2468,7 +2473,7 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev)
adev->firmware.fw_size +=
ALIGN(le32_to_cpu(hdr->inst_const_bytes), PAGE_SIZE);
- DRM_INFO("Loading DMUB firmware via PSP: version=0x%08X\n",
+ drm_info(adev_to_drm(adev), "Loading DMUB firmware via PSP: version=0x%08X\n",
adev->dm.dmcub_fw_version);
}
@@ -2477,7 +2482,7 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev)
dmub_srv = adev->dm.dmub_srv;
if (!dmub_srv) {
- DRM_ERROR("Failed to allocate DMUB service!\n");
+ drm_err(adev_to_drm(adev), "Failed to allocate DMUB service!\n");
return -ENOMEM;
}
@@ -2490,7 +2495,7 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev)
/* Create the DMUB service. */
status = dmub_srv_create(dmub_srv, &create_params);
if (status != DMUB_STATUS_OK) {
- DRM_ERROR("Error creating DMUB service: %d\n", status);
+ drm_err(adev_to_drm(adev), "Error creating DMUB service: %d\n", status);
return -EINVAL;
}
@@ -2515,7 +2520,7 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev)
&region_info);
if (status != DMUB_STATUS_OK) {
- DRM_ERROR("Error calculating DMUB region info: %d\n", status);
+ drm_err(adev_to_drm(adev), "Error calculating DMUB region info: %d\n", status);
return -EINVAL;
}
@@ -2544,14 +2549,14 @@ static int dm_dmub_sw_init(struct amdgpu_device *adev)
fb_info = adev->dm.dmub_fb_info;
if (!fb_info) {
- DRM_ERROR(
+ drm_err(adev_to_drm(adev),
"Failed to allocate framebuffer info for DMUB service!\n");
return -ENOMEM;
}
status = dmub_srv_calc_mem_info(dmub_srv, &memory_params, fb_info);
if (status != DMUB_STATUS_OK) {
- DRM_ERROR("Error calculating DMUB FB info: %d\n", status);
+ drm_err(adev_to_drm(adev), "Error calculating DMUB FB info: %d\n", status);
return -EINVAL;
}
@@ -2568,7 +2573,7 @@ static int dm_sw_init(struct amdgpu_ip_block *ip_block)
adev->dm.cgs_device = amdgpu_cgs_create_device(adev);
if (!adev->dm.cgs_device) {
- DRM_ERROR("amdgpu: failed to create cgs device.\n");
+ drm_err(adev_to_drm(adev), "failed to create cgs device.\n");
return -EINVAL;
}
@@ -2746,6 +2751,48 @@ out_fail:
mutex_unlock(&mgr->lock);
}
+void hdmi_cec_unset_edid(struct amdgpu_dm_connector *aconnector)
+{
+ struct cec_notifier *n = aconnector->notifier;
+
+ if (!n)
+ return;
+
+ cec_notifier_phys_addr_invalidate(n);
+}
+
+void hdmi_cec_set_edid(struct amdgpu_dm_connector *aconnector)
+{
+ struct drm_connector *connector = &aconnector->base;
+ struct cec_notifier *n = aconnector->notifier;
+
+ if (!n)
+ return;
+
+ cec_notifier_set_phys_addr(n,
+ connector->display_info.source_physical_address);
+}
+
+static void s3_handle_hdmi_cec(struct drm_device *ddev, bool suspend)
+{
+ struct amdgpu_dm_connector *aconnector;
+ struct drm_connector *connector;
+ struct drm_connector_list_iter conn_iter;
+
+ drm_connector_list_iter_begin(ddev, &conn_iter);
+ drm_for_each_connector_iter(connector, &conn_iter) {
+ if (connector->connector_type == DRM_MODE_CONNECTOR_WRITEBACK)
+ continue;
+
+ aconnector = to_amdgpu_dm_connector(connector);
+ if (suspend)
+ hdmi_cec_unset_edid(aconnector);
+ else
+ hdmi_cec_set_edid(aconnector);
+ }
+ drm_connector_list_iter_end(&conn_iter);
+}
+
static void s3_handle_mst(struct drm_device *dev, bool suspend)
{
struct amdgpu_dm_connector *aconnector;
@@ -2832,13 +2879,40 @@ static int amdgpu_dm_smu_write_watermarks_table(struct amdgpu_device *adev)
ret = amdgpu_dpm_write_watermarks_table(adev);
if (ret) {
- DRM_ERROR("Failed to update WMTABLE!\n");
+ drm_err(adev_to_drm(adev), "Failed to update WMTABLE!\n");
return ret;
}
return 0;
}
+static int dm_oem_i2c_hw_init(struct amdgpu_device *adev)
+{
+ struct amdgpu_display_manager *dm = &adev->dm;
+ struct amdgpu_i2c_adapter *oem_i2c;
+ struct ddc_service *oem_ddc_service;
+ int r;
+
+ oem_ddc_service = dc_get_oem_i2c_device(adev->dm.dc);
+ if (oem_ddc_service) {
+ oem_i2c = create_i2c(oem_ddc_service, true);
+ if (!oem_i2c) {
+ drm_info(adev_to_drm(adev), "Failed to create oem i2c adapter data\n");
+ return -ENOMEM;
+ }
+
+ r = i2c_add_adapter(&oem_i2c->base);
+ if (r) {
+ drm_info(adev_to_drm(adev), "Failed to register oem i2c\n");
+ kfree(oem_i2c);
+ return r;
+ }
+ dm->oem_i2c = oem_i2c;
+ }
+
+ return 0;
+}
+
/**
* dm_hw_init() - Initialize DC device
* @ip_block: Pointer to the amdgpu_ip_block for this hw instance.
@@ -2870,6 +2944,10 @@ static int dm_hw_init(struct amdgpu_ip_block *ip_block)
return r;
amdgpu_dm_hpd_init(adev);
+ r = dm_oem_i2c_hw_init(adev);
+ if (r)
+ drm_info(adev_to_drm(adev), "Failed to add OEM i2c bus\n");
+
return 0;
}
@@ -2885,6 +2963,8 @@ static int dm_hw_fini(struct amdgpu_ip_block *ip_block)
{
struct amdgpu_device *adev = ip_block->adev;
+ kfree(adev->dm.oem_i2c);
+
amdgpu_dm_hpd_fini(adev);
amdgpu_dm_irq_fini(adev);
@@ -2909,7 +2989,7 @@ static void dm_gpureset_toggle_interrupts(struct amdgpu_device *adev,
irq_source = IRQ_TYPE_PFLIP + acrtc->otg_inst;
rc = dc_interrupt_set(adev->dm.dc, irq_source, enable) ? 0 : -EBUSY;
if (rc)
- DRM_WARN("Failed to %s pflip interrupts\n",
+ drm_warn(adev_to_drm(adev), "Failed to %s pflip interrupts\n",
enable ? "enable" : "disable");
if (enable) {
@@ -2919,23 +2999,24 @@ static void dm_gpureset_toggle_interrupts(struct amdgpu_device *adev,
rc = amdgpu_dm_crtc_set_vupdate_irq(&acrtc->base, false);
if (rc)
- DRM_WARN("Failed to %sable vupdate interrupt\n", enable ? "en" : "dis");
+ drm_warn(adev_to_drm(adev), "Failed to %sable vupdate interrupt\n", enable ? "en" : "dis");
irq_source = IRQ_TYPE_VBLANK + acrtc->otg_inst;
/* During gpu-reset we disable and then enable vblank irq, so
* don't use amdgpu_irq_get/put() to avoid refcount change.
*/
if (!dc_interrupt_set(adev->dm.dc, irq_source, enable))
- DRM_WARN("Failed to %sable vblank interrupt\n", enable ? "en" : "dis");
+ drm_warn(adev_to_drm(adev), "Failed to %sable vblank interrupt\n", enable ? "en" : "dis");
}
}
}
+DEFINE_FREE(state_release, struct dc_state *, if (_T) dc_state_release(_T))
+
static enum dc_status amdgpu_dm_commit_zero_streams(struct dc *dc)
{
- struct dc_state *context = NULL;
- enum dc_status res = DC_ERROR_UNEXPECTED;
+ struct dc_state *context __free(state_release) = NULL;
int i;
struct dc_stream_state *del_streams[MAX_PIPES];
int del_streams_count = 0;
@@ -2945,7 +3026,7 @@ static enum dc_status amdgpu_dm_commit_zero_streams(struct dc *dc)
context = dc_state_create_current_copy(dc);
if (context == NULL)
- goto context_alloc_fail;
+ return DC_ERROR_UNEXPECTED;
/* First remove from context all streams */
for (i = 0; i < context->stream_count; i++) {
@@ -2956,25 +3037,20 @@ static enum dc_status amdgpu_dm_commit_zero_streams(struct dc *dc)
/* Remove all planes for removed streams and then remove the streams */
for (i = 0; i < del_streams_count; i++) {
- if (!dc_state_rem_all_planes_for_stream(dc, del_streams[i], context)) {
- res = DC_FAIL_DETACH_SURFACES;
- goto fail;
- }
+ enum dc_status res;
+
+ if (!dc_state_rem_all_planes_for_stream(dc, del_streams[i], context))
+ return DC_FAIL_DETACH_SURFACES;
res = dc_state_remove_stream(dc, context, del_streams[i]);
if (res != DC_OK)
- goto fail;
+ return res;
}
params.streams = context->streams;
params.stream_count = context->stream_count;
- res = dc_commit_streams(dc, &params);
-fail:
- dc_state_release(context);
-
-context_alloc_fail:
- return res;
+ return dc_commit_streams(dc, &params);
}
static void hpd_rx_irq_work_suspend(struct amdgpu_display_manager *dm)
@@ -2987,13 +3063,97 @@ static void hpd_rx_irq_work_suspend(struct amdgpu_display_manager *dm)
}
}
+static int dm_cache_state(struct amdgpu_device *adev)
+{
+ int r;
+
+ adev->dm.cached_state = drm_atomic_helper_suspend(adev_to_drm(adev));
+ if (IS_ERR(adev->dm.cached_state)) {
+ r = PTR_ERR(adev->dm.cached_state);
+ adev->dm.cached_state = NULL;
+ }
+
+ return adev->dm.cached_state ? 0 : r;
+}
+
+static void dm_destroy_cached_state(struct amdgpu_device *adev)
+{
+ struct amdgpu_display_manager *dm = &adev->dm;
+ struct drm_device *ddev = adev_to_drm(adev);
+ struct dm_plane_state *dm_new_plane_state;
+ struct drm_plane_state *new_plane_state;
+ struct dm_crtc_state *dm_new_crtc_state;
+ struct drm_crtc_state *new_crtc_state;
+ struct drm_plane *plane;
+ struct drm_crtc *crtc;
+ int i;
+
+ if (!dm->cached_state)
+ return;
+
+ /* Force mode set in atomic commit */
+ for_each_new_crtc_in_state(dm->cached_state, crtc, new_crtc_state, i) {
+ new_crtc_state->active_changed = true;
+ dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
+ reset_freesync_config_for_crtc(dm_new_crtc_state);
+ }
+
+ /*
+ * atomic_check is expected to create the dc states. We need to release
+ * them here, since they were duplicated as part of the suspend
+ * procedure.
+ */
+ for_each_new_crtc_in_state(dm->cached_state, crtc, new_crtc_state, i) {
+ dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
+ if (dm_new_crtc_state->stream) {
+ WARN_ON(kref_read(&dm_new_crtc_state->stream->refcount) > 1);
+ dc_stream_release(dm_new_crtc_state->stream);
+ dm_new_crtc_state->stream = NULL;
+ }
+ dm_new_crtc_state->base.color_mgmt_changed = true;
+ }
+
+ for_each_new_plane_in_state(dm->cached_state, plane, new_plane_state, i) {
+ dm_new_plane_state = to_dm_plane_state(new_plane_state);
+ if (dm_new_plane_state->dc_state) {
+ WARN_ON(kref_read(&dm_new_plane_state->dc_state->refcount) > 1);
+ dc_plane_state_release(dm_new_plane_state->dc_state);
+ dm_new_plane_state->dc_state = NULL;
+ }
+ }
+
+ drm_atomic_helper_resume(ddev, dm->cached_state);
+
+ dm->cached_state = NULL;
+}
+
+static void dm_complete(struct amdgpu_ip_block *ip_block)
+{
+ struct amdgpu_device *adev = ip_block->adev;
+
+ dm_destroy_cached_state(adev);
+}
+
+static int dm_prepare_suspend(struct amdgpu_ip_block *ip_block)
+{
+ struct amdgpu_device *adev = ip_block->adev;
+
+ if (amdgpu_in_reset(adev))
+ return 0;
+
+ WARN_ON(adev->dm.cached_state);
+
+ return dm_cache_state(adev);
+}
+
static int dm_suspend(struct amdgpu_ip_block *ip_block)
{
struct amdgpu_device *adev = ip_block->adev;
struct amdgpu_display_manager *dm = &adev->dm;
- int ret = 0;
if (amdgpu_in_reset(adev)) {
+ enum dc_status res;
+
mutex_lock(&dm->dc_lock);
dc_allow_idle_optimizations(adev->dm.dc, false);
@@ -3003,19 +3163,27 @@ static int dm_suspend(struct amdgpu_ip_block *ip_block)
if (dm->cached_dc_state)
dm_gpureset_toggle_interrupts(adev, dm->cached_dc_state, false);
- amdgpu_dm_commit_zero_streams(dm->dc);
+ res = amdgpu_dm_commit_zero_streams(dm->dc);
+ if (res != DC_OK) {
+ drm_err(adev_to_drm(adev), "Failed to commit zero streams: %d\n", res);
+ return -EINVAL;
+ }
amdgpu_dm_irq_suspend(adev);
hpd_rx_irq_work_suspend(dm);
- return ret;
+ return 0;
}
- WARN_ON(adev->dm.cached_state);
- adev->dm.cached_state = drm_atomic_helper_suspend(adev_to_drm(adev));
- if (IS_ERR(adev->dm.cached_state))
- return PTR_ERR(adev->dm.cached_state);
+ if (!adev->dm.cached_state) {
+ int r = dm_cache_state(adev);
+
+ if (r)
+ return r;
+ }
+
+ s3_handle_hdmi_cec(adev_to_drm(adev), true);
s3_handle_mst(adev_to_drm(adev), true);
@@ -3144,36 +3312,51 @@ static void dm_gpureset_commit_state(struct dc_state *dc_state,
struct dc_scaling_info scaling_infos[MAX_SURFACES];
struct dc_flip_addrs flip_addrs[MAX_SURFACES];
struct dc_stream_update stream_update;
- } *bundle;
+ } *bundle __free(kfree);
int k, m;
bundle = kzalloc(sizeof(*bundle), GFP_KERNEL);
if (!bundle) {
drm_err(dm->ddev, "Failed to allocate update bundle\n");
- goto cleanup;
+ return;
}
for (k = 0; k < dc_state->stream_count; k++) {
bundle->stream_update.stream = dc_state->streams[k];
- for (m = 0; m < dc_state->stream_status->plane_count; m++) {
+ for (m = 0; m < dc_state->stream_status[k].plane_count; m++) {
bundle->surface_updates[m].surface =
- dc_state->stream_status->plane_states[m];
+ dc_state->stream_status[k].plane_states[m];
bundle->surface_updates[m].surface->force_full_update =
true;
}
update_planes_and_stream_adapter(dm->dc,
UPDATE_TYPE_FULL,
- dc_state->stream_status->plane_count,
+ dc_state->stream_status[k].plane_count,
dc_state->streams[k],
&bundle->stream_update,
bundle->surface_updates);
}
+}
-cleanup:
- kfree(bundle);
+static void apply_delay_after_dpcd_poweroff(struct amdgpu_device *adev,
+ struct dc_sink *sink)
+{
+ struct dc_panel_patch *ppatch = NULL;
+
+ if (!sink)
+ return;
+
+ ppatch = &sink->edid_caps.panel_patch;
+ if (ppatch->wait_after_dpcd_poweroff_ms) {
+ msleep(ppatch->wait_after_dpcd_poweroff_ms);
+ drm_dbg_driver(adev_to_drm(adev),
+ "%s: adding a %ds delay as w/a for panel\n",
+ __func__,
+ ppatch->wait_after_dpcd_poweroff_ms / 1000);
+ }
}
static int dm_resume(struct amdgpu_ip_block *ip_block)
@@ -3184,12 +3367,6 @@ static int dm_resume(struct amdgpu_ip_block *ip_block)
struct amdgpu_dm_connector *aconnector;
struct drm_connector *connector;
struct drm_connector_list_iter iter;
- struct drm_crtc *crtc;
- struct drm_crtc_state *new_crtc_state;
- struct dm_crtc_state *dm_new_crtc_state;
- struct drm_plane *plane;
- struct drm_plane_state *new_plane_state;
- struct dm_plane_state *dm_new_plane_state;
struct dm_atomic_state *dm_state = to_dm_atomic_state(dm->atomic_obj.state);
enum dc_connection_type new_connection_type = dc_connection_none;
struct dc_state *dc_state;
@@ -3221,8 +3398,10 @@ static int dm_resume(struct amdgpu_ip_block *ip_block)
link_enc_cfg_copy(adev->dm.dc->current_state, dc_state);
r = dm_dmub_hw_init(adev);
- if (r)
- DRM_ERROR("DMUB interface failed to initialize: status=%d\n", r);
+ if (r) {
+ drm_err(adev_to_drm(adev), "DMUB interface failed to initialize: status=%d\n", r);
+ return r;
+ }
dc_dmub_srv_set_power_state(dm->dc->ctx->dmub_srv, DC_ACPI_CM_POWER_STATE_D0);
dc_set_power_state(dm->dc, DC_ACPI_CM_POWER_STATE_D0);
@@ -3260,6 +3439,12 @@ static int dm_resume(struct amdgpu_ip_block *ip_block)
mutex_unlock(&dm->dc_lock);
+ /* set the backlight after a reset */
+ for (i = 0; i < dm->num_of_edps; i++) {
+ if (dm->backlight_dev[i])
+ amdgpu_dm_backlight_set_level(dm, i, dm->brightness[i]);
+ }
+
return 0;
}
/* Recreate dc_state - DC invalidates it when setting power state to S3. */
@@ -3289,12 +3474,15 @@ static int dm_resume(struct amdgpu_ip_block *ip_block)
*/
amdgpu_dm_irq_resume_early(adev);
+ s3_handle_hdmi_cec(ddev, false);
+
/* On resume we need to rewrite the MSTM control bits to enable MST*/
s3_handle_mst(ddev, false);
/* Do detection*/
drm_connector_list_iter_begin(ddev, &iter);
drm_for_each_connector_iter(connector, &iter) {
+ bool ret;
if (connector->connector_type == DRM_MODE_CONNECTOR_WRITEBACK)
continue;
@@ -3311,17 +3499,20 @@ static int dm_resume(struct amdgpu_ip_block *ip_block)
if (aconnector->mst_root)
continue;
- mutex_lock(&aconnector->hpd_lock);
+ guard(mutex)(&aconnector->hpd_lock);
if (!dc_link_detect_connection_type(aconnector->dc_link, &new_connection_type))
- DRM_ERROR("KMS: Failed to detect connector\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to detect connector\n");
if (aconnector->base.force && new_connection_type == dc_connection_none) {
emulated_link_detect(aconnector->dc_link);
} else {
- mutex_lock(&dm->dc_lock);
+ guard(mutex)(&dm->dc_lock);
dc_exit_ips_for_hw_access(dm->dc);
- dc_link_detect(aconnector->dc_link, DETECT_REASON_RESUMEFROMS3S4);
- mutex_unlock(&dm->dc_lock);
+ ret = dc_link_detect(aconnector->dc_link, DETECT_REASON_RESUMEFROMS3S4);
+ if (ret) {
+ /* w/a delay for certain panels */
+ apply_delay_after_dpcd_poweroff(adev, aconnector->dc_sink);
+ }
}
if (aconnector->fake_enable && aconnector->dc_link->local_sink)
@@ -3331,44 +3522,10 @@ static int dm_resume(struct amdgpu_ip_block *ip_block)
dc_sink_release(aconnector->dc_sink);
aconnector->dc_sink = NULL;
amdgpu_dm_update_connector_after_detect(aconnector);
- mutex_unlock(&aconnector->hpd_lock);
}
drm_connector_list_iter_end(&iter);
- /* Force mode set in atomic commit */
- for_each_new_crtc_in_state(dm->cached_state, crtc, new_crtc_state, i) {
- new_crtc_state->active_changed = true;
- dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
- reset_freesync_config_for_crtc(dm_new_crtc_state);
- }
-
- /*
- * atomic_check is expected to create the dc states. We need to release
- * them here, since they were duplicated as part of the suspend
- * procedure.
- */
- for_each_new_crtc_in_state(dm->cached_state, crtc, new_crtc_state, i) {
- dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
- if (dm_new_crtc_state->stream) {
- WARN_ON(kref_read(&dm_new_crtc_state->stream->refcount) > 1);
- dc_stream_release(dm_new_crtc_state->stream);
- dm_new_crtc_state->stream = NULL;
- }
- dm_new_crtc_state->base.color_mgmt_changed = true;
- }
-
- for_each_new_plane_in_state(dm->cached_state, plane, new_plane_state, i) {
- dm_new_plane_state = to_dm_plane_state(new_plane_state);
- if (dm_new_plane_state->dc_state) {
- WARN_ON(kref_read(&dm_new_plane_state->dc_state->refcount) > 1);
- dc_plane_state_release(dm_new_plane_state->dc_state);
- dm_new_plane_state->dc_state = NULL;
- }
- }
-
- drm_atomic_helper_resume(ddev, dm->cached_state);
-
- dm->cached_state = NULL;
+ dm_destroy_cached_state(adev);
/* Do mst topology probing after resuming cached state*/
drm_connector_list_iter_begin(ddev, &iter);
@@ -3414,8 +3571,10 @@ static const struct amd_ip_funcs amdgpu_dm_funcs = {
.early_fini = amdgpu_dm_early_fini,
.hw_init = dm_hw_init,
.hw_fini = dm_hw_fini,
+ .prepare_suspend = dm_prepare_suspend,
.suspend = dm_suspend,
.resume = dm_resume,
+ .complete = dm_complete,
.is_idle = dm_is_idle,
.wait_for_idle = dm_wait_for_idle,
.check_soft_reset = dm_check_soft_reset,
@@ -3457,6 +3616,7 @@ static void update_connector_ext_caps(struct amdgpu_dm_connector *aconnector)
struct drm_connector *conn_base;
struct amdgpu_device *adev;
struct drm_luminance_range_info *luminance_range;
+ int min_input_signal_override;
if (aconnector->bl_idx == -1 ||
aconnector->dc_link->connector_signal != SIGNAL_TYPE_EDP)
@@ -3481,24 +3641,34 @@ static void update_connector_ext_caps(struct amdgpu_dm_connector *aconnector)
caps->aux_support = false;
else if (amdgpu_backlight == 1)
caps->aux_support = true;
+ if (caps->aux_support)
+ aconnector->dc_link->backlight_control_type = BACKLIGHT_CONTROL_AMD_AUX;
luminance_range = &conn_base->display_info.luminance_range;
- if (luminance_range->max_luminance) {
- caps->aux_min_input_signal = luminance_range->min_luminance;
+ if (luminance_range->max_luminance)
caps->aux_max_input_signal = luminance_range->max_luminance;
- } else {
- caps->aux_min_input_signal = 0;
+ else
caps->aux_max_input_signal = 512;
- }
+
+ if (luminance_range->min_luminance)
+ caps->aux_min_input_signal = luminance_range->min_luminance;
+ else
+ caps->aux_min_input_signal = 1;
+
+ min_input_signal_override = drm_get_panel_min_brightness_quirk(aconnector->drm_edid);
+ if (min_input_signal_override >= 0)
+ caps->min_input_signal = min_input_signal_override;
}
+DEFINE_FREE(sink_release, struct dc_sink *, if (_T) dc_sink_release(_T))
+
void amdgpu_dm_update_connector_after_detect(
struct amdgpu_dm_connector *aconnector)
{
struct drm_connector *connector = &aconnector->base;
+ struct dc_sink *sink __free(sink_release) = NULL;
struct drm_device *dev = connector->dev;
- struct dc_sink *sink;
/* MST handled by drm_mst framework */
if (aconnector->mst_mgr.mst_state == true)
@@ -3520,7 +3690,7 @@ void amdgpu_dm_update_connector_after_detect(
* For S3 resume with headless use eml_sink to fake stream
* because on resume connector->sink is set to NULL
*/
- mutex_lock(&dev->mode_config.mutex);
+ guard(mutex)(&dev->mode_config.mutex);
if (sink) {
if (aconnector->dc_sink) {
@@ -3545,10 +3715,6 @@ void amdgpu_dm_update_connector_after_detect(
}
}
- mutex_unlock(&dev->mode_config.mutex);
-
- if (sink)
- dc_sink_release(sink);
return;
}
@@ -3556,10 +3722,8 @@ void amdgpu_dm_update_connector_after_detect(
* TODO: temporary guard to look for proper fix
* if this sink is MST sink, we should not do anything
*/
- if (sink && sink->sink_signal == SIGNAL_TYPE_DISPLAY_PORT_MST) {
- dc_sink_release(sink);
+ if (sink && sink->sink_signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
return;
- }
if (aconnector->dc_sink == sink) {
/*
@@ -3568,15 +3732,13 @@ void amdgpu_dm_update_connector_after_detect(
*/
drm_dbg_kms(dev, "DCHPD: connector_id=%d: dc_sink didn't change.\n",
aconnector->connector_id);
- if (sink)
- dc_sink_release(sink);
return;
}
drm_dbg_kms(dev, "DCHPD: connector_id=%d: Old sink=%p New sink=%p\n",
aconnector->connector_id, aconnector->dc_sink, sink);
- mutex_lock(&dev->mode_config.mutex);
+ guard(mutex)(&dev->mode_config.mutex);
/*
* 1. Update status of the drm connector
@@ -3596,6 +3758,7 @@ void amdgpu_dm_update_connector_after_detect(
dc_sink_retain(aconnector->dc_sink);
if (sink->dc_edid.length == 0) {
aconnector->drm_edid = NULL;
+ hdmi_cec_unset_edid(aconnector);
if (aconnector->dc_link->aux_mode) {
drm_dp_cec_unset_edid(&aconnector->dm_dp_aux.aux);
}
@@ -3605,6 +3768,7 @@ void amdgpu_dm_update_connector_after_detect(
aconnector->drm_edid = drm_edid_alloc(edid, sink->dc_edid.length);
drm_edid_connector_update(connector, aconnector->drm_edid);
+ hdmi_cec_set_edid(aconnector);
if (aconnector->dc_link->aux_mode)
drm_dp_cec_attach(&aconnector->dm_dp_aux.aux,
connector->display_info.source_physical_address);
@@ -3621,6 +3785,7 @@ void amdgpu_dm_update_connector_after_detect(
amdgpu_dm_update_freesync_caps(connector, aconnector->drm_edid);
update_connector_ext_caps(aconnector);
} else {
+ hdmi_cec_unset_edid(aconnector);
drm_dp_cec_unset_edid(&aconnector->dm_dp_aux.aux);
amdgpu_dm_update_freesync_caps(connector, NULL);
aconnector->num_modes = 0;
@@ -3635,12 +3800,7 @@ void amdgpu_dm_update_connector_after_detect(
connector->state->content_protection = DRM_MODE_CONTENT_PROTECTION_DESIRED;
}
- mutex_unlock(&dev->mode_config.mutex);
-
update_subconnector_property(aconnector);
-
- if (sink)
- dc_sink_release(sink);
}
static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector)
@@ -3660,7 +3820,7 @@ static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector)
* In case of failure or MST no need to update connector status or notify the OS
* since (for MST case) MST does this in its own context.
*/
- mutex_lock(&aconnector->hpd_lock);
+ guard(mutex)(&aconnector->hpd_lock);
if (adev->dm.hdcp_workqueue) {
hdcp_reset_display(adev->dm.hdcp_workqueue, aconnector->dc_link->link_index);
@@ -3672,7 +3832,7 @@ static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector)
aconnector->timing_changed = false;
if (!dc_link_detect_connection_type(aconnector->dc_link, &new_connection_type))
- DRM_ERROR("KMS: Failed to detect connector\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to detect connector\n");
if (aconnector->base.force && new_connection_type == dc_connection_none) {
emulated_link_detect(aconnector->dc_link);
@@ -3684,11 +3844,13 @@ static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector)
if (aconnector->base.force == DRM_FORCE_UNSPECIFIED)
drm_kms_helper_connector_hotplug_event(connector);
} else {
- mutex_lock(&adev->dm.dc_lock);
- dc_exit_ips_for_hw_access(dc);
- ret = dc_link_detect(aconnector->dc_link, DETECT_REASON_HPD);
- mutex_unlock(&adev->dm.dc_lock);
+ scoped_guard(mutex, &adev->dm.dc_lock) {
+ dc_exit_ips_for_hw_access(dc);
+ ret = dc_link_detect(aconnector->dc_link, DETECT_REASON_HPD);
+ }
if (ret) {
+ /* w/a delay for certain panels */
+ apply_delay_after_dpcd_poweroff(adev, aconnector->dc_sink);
amdgpu_dm_update_connector_after_detect(aconnector);
drm_modeset_lock_all(dev);
@@ -3699,8 +3861,6 @@ static void handle_hpd_irq_helper(struct amdgpu_dm_connector *aconnector)
drm_kms_helper_connector_hotplug_event(connector);
}
}
- mutex_unlock(&aconnector->hpd_lock);
-
}
static void handle_hpd_irq(void *param)
@@ -3711,20 +3871,21 @@ static void handle_hpd_irq(void *param)
}
-static void schedule_hpd_rx_offload_work(struct hpd_rx_irq_offload_work_queue *offload_wq,
+static void schedule_hpd_rx_offload_work(struct amdgpu_device *adev, struct hpd_rx_irq_offload_work_queue *offload_wq,
union hpd_irq_data hpd_irq_data)
{
struct hpd_rx_irq_offload_work *offload_work =
kzalloc(sizeof(*offload_work), GFP_KERNEL);
if (!offload_work) {
- DRM_ERROR("Failed to allocate hpd_rx_irq_offload_work.\n");
+ drm_err(adev_to_drm(adev), "Failed to allocate hpd_rx_irq_offload_work.\n");
return;
}
INIT_WORK(&offload_work->work, dm_handle_hpd_rx_offload_work);
offload_work->data = hpd_irq_data;
offload_work->offload_wq = offload_wq;
+ offload_work->adev = adev;
queue_work(offload_wq->wq, &offload_work->work);
DRM_DEBUG_KMS("queue work to handle hpd_rx offload work");
@@ -3766,7 +3927,7 @@ static void handle_hpd_rx_irq(void *param)
goto out;
if (hpd_irq_data.bytes.device_service_irq.bits.AUTOMATED_TEST) {
- schedule_hpd_rx_offload_work(offload_wq, hpd_irq_data);
+ schedule_hpd_rx_offload_work(adev, offload_wq, hpd_irq_data);
goto out;
}
@@ -3788,7 +3949,7 @@ static void handle_hpd_rx_irq(void *param)
spin_unlock(&offload_wq->offload_lock);
if (!skip)
- schedule_hpd_rx_offload_work(offload_wq, hpd_irq_data);
+ schedule_hpd_rx_offload_work(adev, offload_wq, hpd_irq_data);
goto out;
}
@@ -3805,7 +3966,7 @@ static void handle_hpd_rx_irq(void *param)
spin_unlock(&offload_wq->offload_lock);
if (!skip)
- schedule_hpd_rx_offload_work(offload_wq, hpd_irq_data);
+ schedule_hpd_rx_offload_work(adev, offload_wq, hpd_irq_data);
goto out;
}
@@ -3815,7 +3976,7 @@ out:
if (result && !is_mst_root_connector) {
/* Downstream Port status changed. */
if (!dc_link_detect_connection_type(dc_link, &new_connection_type))
- DRM_ERROR("KMS: Failed to detect connector\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to detect connector\n");
if (aconnector->base.force && new_connection_type == dc_connection_none) {
emulated_link_detect(dc_link);
@@ -3878,19 +4039,19 @@ static int register_hpd_handlers(struct amdgpu_device *adev)
if (dc_is_dmub_outbox_supported(adev->dm.dc)) {
if (!register_dmub_notify_callback(adev, DMUB_NOTIFICATION_HPD,
dmub_hpd_callback, true)) {
- DRM_ERROR("amdgpu: fail to register dmub hpd callback");
+ drm_err(adev_to_drm(adev), "fail to register dmub hpd callback");
return -EINVAL;
}
if (!register_dmub_notify_callback(adev, DMUB_NOTIFICATION_HPD_IRQ,
dmub_hpd_callback, true)) {
- DRM_ERROR("amdgpu: fail to register dmub hpd callback");
+ drm_err(adev_to_drm(adev), "fail to register dmub hpd callback");
return -EINVAL;
}
if (!register_dmub_notify_callback(adev, DMUB_NOTIFICATION_HPD_SENSE_NOTIFY,
dmub_hpd_sense_callback, true)) {
- DRM_ERROR("amdgpu: fail to register dmub hpd sense callback");
+ drm_err(adev_to_drm(adev), "fail to register dmub hpd sense callback");
return -EINVAL;
}
}
@@ -3911,7 +4072,7 @@ static int register_hpd_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_HPD1 ||
int_params.irq_source > DC_IRQ_SOURCE_HPD6) {
- DRM_ERROR("Failed to register hpd irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register hpd irq!\n");
return -EINVAL;
}
@@ -3929,7 +4090,7 @@ static int register_hpd_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_HPD1RX ||
int_params.irq_source > DC_IRQ_SOURCE_HPD6RX) {
- DRM_ERROR("Failed to register hpd rx irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register hpd rx irq!\n");
return -EINVAL;
}
@@ -3971,7 +4132,7 @@ static int dce60_register_irq_handlers(struct amdgpu_device *adev)
for (i = 0; i < adev->mode_info.num_crtc; i++) {
r = amdgpu_irq_add_id(adev, client_id, i + 1, &adev->crtc_irq);
if (r) {
- DRM_ERROR("Failed to add crtc irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add crtc irq id!\n");
return r;
}
@@ -3982,7 +4143,7 @@ static int dce60_register_irq_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_VBLANK1 ||
int_params.irq_source > DC_IRQ_SOURCE_VBLANK6) {
- DRM_ERROR("Failed to register vblank irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register vblank irq!\n");
return -EINVAL;
}
@@ -4001,7 +4162,7 @@ static int dce60_register_irq_handlers(struct amdgpu_device *adev)
i <= VISLANDS30_IV_SRCID_D6_GRPH_PFLIP; i += 2) {
r = amdgpu_irq_add_id(adev, client_id, i, &adev->pageflip_irq);
if (r) {
- DRM_ERROR("Failed to add page flip irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add page flip irq id!\n");
return r;
}
@@ -4012,7 +4173,7 @@ static int dce60_register_irq_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_PFLIP_FIRST ||
int_params.irq_source > DC_IRQ_SOURCE_PFLIP_LAST) {
- DRM_ERROR("Failed to register pflip irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register pflip irq!\n");
return -EINVAL;
}
@@ -4030,7 +4191,7 @@ static int dce60_register_irq_handlers(struct amdgpu_device *adev)
r = amdgpu_irq_add_id(adev, client_id,
VISLANDS30_IV_SRCID_HOTPLUG_DETECT_A, &adev->hpd_irq);
if (r) {
- DRM_ERROR("Failed to add hpd irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add hpd irq id!\n");
return r;
}
@@ -4072,7 +4233,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
for (i = VISLANDS30_IV_SRCID_D1_VERTICAL_INTERRUPT0; i <= VISLANDS30_IV_SRCID_D6_VERTICAL_INTERRUPT0; i++) {
r = amdgpu_irq_add_id(adev, client_id, i, &adev->crtc_irq);
if (r) {
- DRM_ERROR("Failed to add crtc irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add crtc irq id!\n");
return r;
}
@@ -4083,7 +4244,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_VBLANK1 ||
int_params.irq_source > DC_IRQ_SOURCE_VBLANK6) {
- DRM_ERROR("Failed to register vblank irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register vblank irq!\n");
return -EINVAL;
}
@@ -4101,7 +4262,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
for (i = VISLANDS30_IV_SRCID_D1_V_UPDATE_INT; i <= VISLANDS30_IV_SRCID_D6_V_UPDATE_INT; i += 2) {
r = amdgpu_irq_add_id(adev, client_id, i, &adev->vupdate_irq);
if (r) {
- DRM_ERROR("Failed to add vupdate irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add vupdate irq id!\n");
return r;
}
@@ -4112,7 +4273,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_VUPDATE1 ||
int_params.irq_source > DC_IRQ_SOURCE_VUPDATE6) {
- DRM_ERROR("Failed to register vupdate irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register vupdate irq!\n");
return -EINVAL;
}
@@ -4131,7 +4292,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
i <= VISLANDS30_IV_SRCID_D6_GRPH_PFLIP; i += 2) {
r = amdgpu_irq_add_id(adev, client_id, i, &adev->pageflip_irq);
if (r) {
- DRM_ERROR("Failed to add page flip irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add page flip irq id!\n");
return r;
}
@@ -4142,7 +4303,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_PFLIP_FIRST ||
int_params.irq_source > DC_IRQ_SOURCE_PFLIP_LAST) {
- DRM_ERROR("Failed to register pflip irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register pflip irq!\n");
return -EINVAL;
}
@@ -4160,7 +4321,7 @@ static int dce110_register_irq_handlers(struct amdgpu_device *adev)
r = amdgpu_irq_add_id(adev, client_id,
VISLANDS30_IV_SRCID_HOTPLUG_DETECT_A, &adev->hpd_irq);
if (r) {
- DRM_ERROR("Failed to add hpd irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add hpd irq id!\n");
return r;
}
@@ -4210,7 +4371,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, i, &adev->crtc_irq);
if (r) {
- DRM_ERROR("Failed to add crtc irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add crtc irq id!\n");
return r;
}
@@ -4221,7 +4382,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_VBLANK1 ||
int_params.irq_source > DC_IRQ_SOURCE_VBLANK6) {
- DRM_ERROR("Failed to register vblank irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register vblank irq!\n");
return -EINVAL;
}
@@ -4242,7 +4403,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
vrtl_int_srcid[i], &adev->vline0_irq);
if (r) {
- DRM_ERROR("Failed to add vline0 irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add vline0 irq id!\n");
return r;
}
@@ -4253,7 +4414,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_DC1_VLINE0 ||
int_params.irq_source > DC_IRQ_SOURCE_DC6_VLINE0) {
- DRM_ERROR("Failed to register vline0 irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register vline0 irq!\n");
return -EINVAL;
}
@@ -4281,7 +4442,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, i, &adev->vupdate_irq);
if (r) {
- DRM_ERROR("Failed to add vupdate irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add vupdate irq id!\n");
return r;
}
@@ -4292,7 +4453,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_VUPDATE1 ||
int_params.irq_source > DC_IRQ_SOURCE_VUPDATE6) {
- DRM_ERROR("Failed to register vupdate irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register vupdate irq!\n");
return -EINVAL;
}
@@ -4312,7 +4473,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
i++) {
r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, i, &adev->pageflip_irq);
if (r) {
- DRM_ERROR("Failed to add page flip irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add page flip irq id!\n");
return r;
}
@@ -4323,7 +4484,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
if (int_params.irq_source == DC_IRQ_SOURCE_INVALID ||
int_params.irq_source < DC_IRQ_SOURCE_PFLIP_FIRST ||
int_params.irq_source > DC_IRQ_SOURCE_PFLIP_LAST) {
- DRM_ERROR("Failed to register pflip irq!\n");
+ drm_err(adev_to_drm(adev), "Failed to register pflip irq!\n");
return -EINVAL;
}
@@ -4341,7 +4502,7 @@ static int dcn10_register_irq_handlers(struct amdgpu_device *adev)
r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, DCN_1_0__SRCID__DC_HPD1_INT,
&adev->hpd_irq);
if (r) {
- DRM_ERROR("Failed to add hpd irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add hpd irq id!\n");
return r;
}
@@ -4363,7 +4524,7 @@ static int register_outbox_irq_handlers(struct amdgpu_device *adev)
r = amdgpu_irq_add_id(adev, SOC15_IH_CLIENTID_DCE, DCN_1_0__SRCID__DMCUB_OUTBOX_LOW_PRIORITY_READY_INT,
&adev->dmub_outbox_irq);
if (r) {
- DRM_ERROR("Failed to add outbox irq id!\n");
+ drm_err(adev_to_drm(adev), "Failed to add outbox irq id!\n");
return r;
}
@@ -4540,48 +4701,40 @@ static int amdgpu_dm_mode_config_init(struct amdgpu_device *adev)
static void amdgpu_dm_update_backlight_caps(struct amdgpu_display_manager *dm,
int bl_idx)
{
-#if defined(CONFIG_ACPI)
- struct amdgpu_dm_backlight_caps caps;
-
- memset(&caps, 0, sizeof(caps));
+ struct amdgpu_dm_backlight_caps *caps = &dm->backlight_caps[bl_idx];
- if (dm->backlight_caps[bl_idx].caps_valid)
+ if (caps->caps_valid)
return;
- amdgpu_acpi_get_backlight_caps(&caps);
+#if defined(CONFIG_ACPI)
+ amdgpu_acpi_get_backlight_caps(caps);
/* validate the firmware value is sane */
- if (caps.caps_valid) {
- int spread = caps.max_input_signal - caps.min_input_signal;
+ if (caps->caps_valid) {
+ int spread = caps->max_input_signal - caps->min_input_signal;
- if (caps.max_input_signal > AMDGPU_DM_DEFAULT_MAX_BACKLIGHT ||
- caps.min_input_signal < 0 ||
+ if (caps->max_input_signal > AMDGPU_DM_DEFAULT_MAX_BACKLIGHT ||
+ caps->min_input_signal < 0 ||
spread > AMDGPU_DM_DEFAULT_MAX_BACKLIGHT ||
spread < AMDGPU_DM_MIN_SPREAD) {
DRM_DEBUG_KMS("DM: Invalid backlight caps: min=%d, max=%d\n",
- caps.min_input_signal, caps.max_input_signal);
- caps.caps_valid = false;
+ caps->min_input_signal, caps->max_input_signal);
+ caps->caps_valid = false;
}
}
- if (caps.caps_valid) {
- dm->backlight_caps[bl_idx].caps_valid = true;
- if (caps.aux_support)
- return;
- dm->backlight_caps[bl_idx].min_input_signal = caps.min_input_signal;
- dm->backlight_caps[bl_idx].max_input_signal = caps.max_input_signal;
- } else {
- dm->backlight_caps[bl_idx].min_input_signal =
- AMDGPU_DM_DEFAULT_MIN_BACKLIGHT;
- dm->backlight_caps[bl_idx].max_input_signal =
- AMDGPU_DM_DEFAULT_MAX_BACKLIGHT;
+ if (!caps->caps_valid) {
+ caps->min_input_signal = AMDGPU_DM_DEFAULT_MIN_BACKLIGHT;
+ caps->max_input_signal = AMDGPU_DM_DEFAULT_MAX_BACKLIGHT;
+ caps->caps_valid = true;
}
#else
- if (dm->backlight_caps[bl_idx].aux_support)
+ if (caps->aux_support)
return;
- dm->backlight_caps[bl_idx].min_input_signal = AMDGPU_DM_DEFAULT_MIN_BACKLIGHT;
- dm->backlight_caps[bl_idx].max_input_signal = AMDGPU_DM_DEFAULT_MAX_BACKLIGHT;
+ caps->min_input_signal = AMDGPU_DM_DEFAULT_MIN_BACKLIGHT;
+ caps->max_input_signal = AMDGPU_DM_DEFAULT_MAX_BACKLIGHT;
+ caps->caps_valid = true;
#endif
}
@@ -4603,6 +4756,60 @@ static int get_brightness_range(const struct amdgpu_dm_backlight_caps *caps,
return 1;
}
+/* Rescale from [min..max] to [0..AMDGPU_MAX_BL_LEVEL] */
+static inline u32 scale_input_to_fw(int min, int max, u64 input)
+{
+ return DIV_ROUND_CLOSEST_ULL(input * AMDGPU_MAX_BL_LEVEL, max - min);
+}
+
+/* Rescale from [0..AMDGPU_MAX_BL_LEVEL] to [min..max] */
+static inline u32 scale_fw_to_input(int min, int max, u64 input)
+{
+ return min + DIV_ROUND_CLOSEST_ULL(input * (max - min), AMDGPU_MAX_BL_LEVEL);
+}
+
+static void convert_custom_brightness(const struct amdgpu_dm_backlight_caps *caps,
+ unsigned int min, unsigned int max,
+ uint32_t *user_brightness)
+{
+ u32 brightness = scale_input_to_fw(min, max, *user_brightness);
+ u8 prev_signal = 0, prev_lum = 0;
+ int i = 0;
+
+ if (amdgpu_dc_debug_mask & DC_DISABLE_CUSTOM_BRIGHTNESS_CURVE)
+ return;
+
+ if (!caps->data_points)
+ return;
+
+ /* choose start to run less interpolation steps */
+ if (caps->luminance_data[caps->data_points/2].input_signal > brightness)
+ i = caps->data_points/2;
+ do {
+ u8 signal = caps->luminance_data[i].input_signal;
+ u8 lum = caps->luminance_data[i].luminance;
+
+ /*
+ * brightness == signal: luminance is percent numerator
+ * brightness < signal: interpolate between previous and current luminance numerator
+ * brightness > signal: find next data point
+ */
+ if (brightness > signal) {
+ prev_signal = signal;
+ prev_lum = lum;
+ i++;
+ continue;
+ }
+ if (brightness < signal)
+ lum = prev_lum + DIV_ROUND_CLOSEST((lum - prev_lum) *
+ (brightness - prev_signal),
+ signal - prev_signal);
+ *user_brightness = scale_fw_to_input(min, max,
+ DIV_ROUND_CLOSEST(lum * brightness, 101));
+ return;
+ } while (i < caps->data_points);
+}
+
static u32 convert_brightness_from_user(const struct amdgpu_dm_backlight_caps *caps,
uint32_t brightness)
{
@@ -4611,9 +4818,10 @@ static u32 convert_brightness_from_user(const struct amdgpu_dm_backlight_caps *c
if (!get_brightness_range(caps, &min, &max))
return brightness;
- // Rescale 0..255 to min..max
- return min + DIV_ROUND_CLOSEST((max - min) * brightness,
- AMDGPU_MAX_BL_LEVEL);
+ convert_custom_brightness(caps, min, max, &brightness);
+
+ // Rescale 0..max to min..max
+ return min + DIV_ROUND_CLOSEST_ULL((u64)(max - min) * brightness, max);
}
static u32 convert_brightness_to_user(const struct amdgpu_dm_backlight_caps *caps,
@@ -4626,8 +4834,8 @@ static u32 convert_brightness_to_user(const struct amdgpu_dm_backlight_caps *cap
if (brightness < min)
return 0;
- // Rescale min..max to 0..255
- return DIV_ROUND_CLOSEST(AMDGPU_MAX_BL_LEVEL * (brightness - min),
+ // Rescale min..max to 0..max
+ return DIV_ROUND_CLOSEST_ULL((u64)max * (brightness - min),
max - min);
}
@@ -4635,19 +4843,19 @@ static void amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm,
int bl_idx,
u32 user_brightness)
{
- struct amdgpu_dm_backlight_caps caps;
+ struct amdgpu_dm_backlight_caps *caps;
struct dc_link *link;
u32 brightness;
bool rc, reallow_idle = false;
amdgpu_dm_update_backlight_caps(dm, bl_idx);
- caps = dm->backlight_caps[bl_idx];
+ caps = &dm->backlight_caps[bl_idx];
dm->brightness[bl_idx] = user_brightness;
/* update scratch register */
if (bl_idx == 0)
amdgpu_atombios_scratch_regs_set_backlight_level(dm->adev, dm->brightness[bl_idx]);
- brightness = convert_brightness_from_user(&caps, dm->brightness[bl_idx]);
+ brightness = convert_brightness_from_user(caps, dm->brightness[bl_idx]);
link = (struct dc_link *)dm->backlight_link[bl_idx];
/* Change brightness based on AUX property */
@@ -4657,7 +4865,15 @@ static void amdgpu_dm_backlight_set_level(struct amdgpu_display_manager *dm,
reallow_idle = true;
}
- if (caps.aux_support) {
+ if (trace_amdgpu_dm_brightness_enabled()) {
+ trace_amdgpu_dm_brightness(__builtin_return_address(0),
+ user_brightness,
+ brightness,
+ caps->aux_support,
+ power_supply_is_system_supplied() > 0);
+ }
+
+ if (caps->aux_support) {
rc = dc_link_set_backlight_level_nits(link, true, brightness,
AUX_BL_DEFAULT_TRANSITION_TIME_MS);
if (!rc)
@@ -4752,8 +4968,9 @@ amdgpu_dm_register_backlight_device(struct amdgpu_dm_connector *aconnector)
struct drm_device *drm = aconnector->base.dev;
struct amdgpu_display_manager *dm = &drm_to_adev(drm)->dm;
struct backlight_properties props = { 0 };
- struct amdgpu_dm_backlight_caps caps = { 0 };
+ struct amdgpu_dm_backlight_caps *caps;
char bl_name[16];
+ int min, max;
if (aconnector->bl_idx == -1)
return;
@@ -4765,16 +4982,21 @@ amdgpu_dm_register_backlight_device(struct amdgpu_dm_connector *aconnector)
return;
}
- amdgpu_acpi_get_backlight_caps(&caps);
- if (caps.caps_valid) {
+ caps = &dm->backlight_caps[aconnector->bl_idx];
+ if (get_brightness_range(caps, &min, &max)) {
if (power_supply_is_system_supplied() > 0)
- props.brightness = caps.ac_level;
+ props.brightness = DIV_ROUND_CLOSEST((max - min) * caps->ac_level, 100);
else
- props.brightness = caps.dc_level;
+ props.brightness = DIV_ROUND_CLOSEST((max - min) * caps->dc_level, 100);
+ /* min is zero, so max needs to be adjusted */
+ props.max_brightness = max - min;
+ drm_dbg(drm, "Backlight caps: min: %d, max: %d, ac %d, dc %d\n", min, max,
+ caps->ac_level, caps->dc_level);
} else
- props.brightness = AMDGPU_MAX_BL_LEVEL;
+ props.brightness = props.max_brightness = MAX_BACKLIGHT_LEVEL;
- props.max_brightness = AMDGPU_MAX_BL_LEVEL;
+ if (caps->data_points && !(amdgpu_dc_debug_mask & DC_DISABLE_CUSTOM_BRIGHTNESS_CURVE))
+ drm_info(drm, "Using custom brightness curve\n");
props.type = BACKLIGHT_RAW;
snprintf(bl_name, sizeof(bl_name), "amdgpu_bl%d",
@@ -4783,12 +5005,13 @@ amdgpu_dm_register_backlight_device(struct amdgpu_dm_connector *aconnector)
dm->backlight_dev[aconnector->bl_idx] =
backlight_device_register(bl_name, aconnector->base.kdev, dm,
&amdgpu_dm_backlight_ops, &props);
+ dm->brightness[aconnector->bl_idx] = props.brightness;
if (IS_ERR(dm->backlight_dev[aconnector->bl_idx])) {
- DRM_ERROR("DM: Backlight registration failed!\n");
+ drm_err(drm, "DM: Backlight registration failed!\n");
dm->backlight_dev[aconnector->bl_idx] = NULL;
} else
- DRM_DEBUG_DRIVER("DM: Registered Backlight device: %s\n", bl_name);
+ drm_dbg_driver(drm, "DM: Registered Backlight device: %s\n", bl_name);
}
static int initialize_plane(struct amdgpu_display_manager *dm,
@@ -4802,7 +5025,7 @@ static int initialize_plane(struct amdgpu_display_manager *dm,
plane = kzalloc(sizeof(struct drm_plane), GFP_KERNEL);
if (!plane) {
- DRM_ERROR("KMS: Failed to allocate plane\n");
+ drm_err(adev_to_drm(dm->adev), "KMS: Failed to allocate plane\n");
return -ENOMEM;
}
plane->type = plane_type;
@@ -4820,7 +5043,7 @@ static int initialize_plane(struct amdgpu_display_manager *dm,
ret = amdgpu_dm_plane_init(dm, plane, possible_crtcs, plane_cap);
if (ret) {
- DRM_ERROR("KMS: Failed to initialize plane\n");
+ drm_err(adev_to_drm(dm->adev), "KMS: Failed to initialize plane\n");
kfree(plane);
return ret;
}
@@ -4850,7 +5073,6 @@ static void setup_backlight_device(struct amdgpu_display_manager *dm,
aconnector->bl_idx = bl_idx;
amdgpu_dm_update_backlight_caps(dm, bl_idx);
- dm->brightness[bl_idx] = AMDGPU_MAX_BL_LEVEL;
dm->backlight_link[bl_idx] = link;
dm->num_of_edps++;
@@ -4890,14 +5112,14 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
link_cnt = dm->dc->caps.max_links;
if (amdgpu_dm_mode_config_init(dm->adev)) {
- DRM_ERROR("DM: Failed to initialize mode config\n");
+ drm_err(adev_to_drm(adev), "DM: Failed to initialize mode config\n");
return -EINVAL;
}
/* There is one primary plane per CRTC */
primary_planes = dm->dc->caps.max_streams;
if (primary_planes > AMDGPU_MAX_PLANES) {
- DRM_ERROR("DM: Plane nums out of 6 planes\n");
+ drm_err(adev_to_drm(adev), "DM: Plane nums out of 6 planes\n");
return -EINVAL;
}
@@ -4910,7 +5132,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
if (initialize_plane(dm, mode_info, i,
DRM_PLANE_TYPE_PRIMARY, plane)) {
- DRM_ERROR("KMS: Failed to initialize primary plane\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to initialize primary plane\n");
goto fail;
}
}
@@ -4942,14 +5164,14 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
if (initialize_plane(dm, NULL, primary_planes + i,
DRM_PLANE_TYPE_OVERLAY, plane)) {
- DRM_ERROR("KMS: Failed to initialize overlay plane\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to initialize overlay plane\n");
goto fail;
}
}
for (i = 0; i < dm->dc->caps.max_streams; i++)
if (amdgpu_dm_crtc_init(dm, mode_info->planes[i], i)) {
- DRM_ERROR("KMS: Failed to initialize crtc\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to initialize crtc\n");
goto fail;
}
@@ -4966,9 +5188,10 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
case IP_VERSION(2, 1, 0):
case IP_VERSION(3, 5, 0):
case IP_VERSION(3, 5, 1):
+ case IP_VERSION(3, 6, 0):
case IP_VERSION(4, 0, 1):
if (register_outbox_irq_handlers(dm->adev)) {
- DRM_ERROR("DM: Failed to initialize IRQ\n");
+ drm_err(adev_to_drm(adev), "DM: Failed to initialize IRQ\n");
goto fail;
}
break;
@@ -4989,6 +5212,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
case IP_VERSION(3, 2, 1):
case IP_VERSION(3, 5, 0):
case IP_VERSION(3, 5, 1):
+ case IP_VERSION(3, 6, 0):
case IP_VERSION(4, 0, 1):
psr_feature_enabled = true;
break;
@@ -5006,6 +5230,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
case IP_VERSION(3, 2, 1):
case IP_VERSION(3, 5, 0):
case IP_VERSION(3, 5, 1):
+ case IP_VERSION(3, 6, 0):
replay_feature_enabled = true;
break;
@@ -5016,7 +5241,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
}
if (link_cnt > MAX_LINKS) {
- DRM_ERROR(
+ drm_err(adev_to_drm(adev),
"KMS: Cannot support more than %d display indexes\n",
MAX_LINKS);
goto fail;
@@ -5032,12 +5257,12 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
struct amdgpu_dm_wb_connector *wbcon = kzalloc(sizeof(*wbcon), GFP_KERNEL);
if (!wbcon) {
- DRM_ERROR("KMS: Failed to allocate writeback connector\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to allocate writeback connector\n");
continue;
}
if (amdgpu_dm_wb_connector_init(dm, wbcon, i)) {
- DRM_ERROR("KMS: Failed to initialize writeback connector\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to initialize writeback connector\n");
kfree(wbcon);
continue;
}
@@ -5057,12 +5282,12 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
goto fail;
if (amdgpu_dm_encoder_init(dm->ddev, aencoder, i)) {
- DRM_ERROR("KMS: Failed to initialize encoder\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to initialize encoder\n");
goto fail;
}
if (amdgpu_dm_connector_init(dm, aconnector, i, aencoder)) {
- DRM_ERROR("KMS: Failed to initialize connector\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to initialize connector\n");
goto fail;
}
@@ -5071,7 +5296,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
aconnector;
if (!dc_link_detect_connection_type(link, &new_connection_type))
- DRM_ERROR("KMS: Failed to detect connector\n");
+ drm_err(adev_to_drm(adev), "KMS: Failed to detect connector\n");
if (aconnector->base.force && new_connection_type == dc_connection_none) {
emulated_link_detect(link);
@@ -5093,8 +5318,15 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
if (amdgpu_dm_set_replay_caps(link, aconnector))
psr_feature_enabled = false;
- if (psr_feature_enabled)
+ if (psr_feature_enabled) {
amdgpu_dm_set_psr_caps(link);
+ drm_info(adev_to_drm(adev), "PSR support %d, DC PSR ver %d, sink PSR ver %d DPCD caps 0x%x su_y_granularity %d\n",
+ link->psr_settings.psr_feature_enabled,
+ link->psr_settings.psr_version,
+ link->dpcd_caps.psr_info.psr_version,
+ link->dpcd_caps.psr_info.psr_dpcd_caps.raw,
+ link->dpcd_caps.psr_info.psr2_su_y_granularity_cap);
+ }
}
}
amdgpu_set_panel_orientation(&aconnector->base);
@@ -5108,7 +5340,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
case CHIP_VERDE:
case CHIP_OLAND:
if (dce60_register_irq_handlers(dm->adev)) {
- DRM_ERROR("DM: Failed to initialize IRQ\n");
+ drm_err(adev_to_drm(adev), "DM: Failed to initialize IRQ\n");
goto fail;
}
break;
@@ -5130,7 +5362,7 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
case CHIP_VEGA12:
case CHIP_VEGA20:
if (dce110_register_irq_handlers(dm->adev)) {
- DRM_ERROR("DM: Failed to initialize IRQ\n");
+ drm_err(adev_to_drm(adev), "DM: Failed to initialize IRQ\n");
goto fail;
}
break;
@@ -5155,14 +5387,15 @@ static int amdgpu_dm_initialize_drm_device(struct amdgpu_device *adev)
case IP_VERSION(3, 2, 1):
case IP_VERSION(3, 5, 0):
case IP_VERSION(3, 5, 1):
+ case IP_VERSION(3, 6, 0):
case IP_VERSION(4, 0, 1):
if (dcn10_register_irq_handlers(dm->adev)) {
- DRM_ERROR("DM: Failed to initialize IRQ\n");
+ drm_err(adev_to_drm(adev), "DM: Failed to initialize IRQ\n");
goto fail;
}
break;
default:
- DRM_ERROR("Unsupported DCE IP versions: 0x%X\n",
+ drm_err(adev_to_drm(adev), "Unsupported DCE IP versions: 0x%X\n",
amdgpu_ip_version(adev, DCE_HWIP, 0));
goto fail;
}
@@ -5179,7 +5412,8 @@ fail:
static void amdgpu_dm_destroy_drm_device(struct amdgpu_display_manager *dm)
{
- drm_atomic_private_obj_fini(&dm->atomic_obj);
+ if (dm->atomic_obj.state)
+ drm_atomic_private_obj_fini(&dm->atomic_obj);
}
/******************************************************************************
@@ -5297,6 +5531,9 @@ static int dm_init_microcode(struct amdgpu_device *adev)
case IP_VERSION(3, 5, 1):
fw_name_dmub = FIRMWARE_DCN_351_DMUB;
break;
+ case IP_VERSION(3, 6, 0):
+ fw_name_dmub = FIRMWARE_DCN_36_DMUB;
+ break;
case IP_VERSION(4, 0, 1):
fw_name_dmub = FIRMWARE_DCN_401_DMUB;
break;
@@ -5304,7 +5541,8 @@ static int dm_init_microcode(struct amdgpu_device *adev)
/* ASIC doesn't support DMUB. */
return 0;
}
- r = amdgpu_ucode_request(adev, &adev->dm.dmub_fw, "%s", fw_name_dmub);
+ r = amdgpu_ucode_request(adev, &adev->dm.dmub_fw, AMDGPU_UCODE_REQUIRED,
+ "%s", fw_name_dmub);
return r;
}
@@ -5319,7 +5557,7 @@ static int dm_early_init(struct amdgpu_ip_block *ip_block)
/* if there is no object header, skip DM */
if (!amdgpu_atom_parse_data_header(ctx, index, NULL, NULL, NULL, &data_offset)) {
adev->harvest_ip_mask |= AMD_HARVEST_IP_DMU_MASK;
- dev_info(adev->dev, "No object header, skipping DM\n");
+ drm_info(adev_to_drm(adev), "No object header, skipping DM\n");
return -ENOENT;
}
@@ -5424,13 +5662,14 @@ static int dm_early_init(struct amdgpu_ip_block *ip_block)
case IP_VERSION(3, 2, 1):
case IP_VERSION(3, 5, 0):
case IP_VERSION(3, 5, 1):
+ case IP_VERSION(3, 6, 0):
case IP_VERSION(4, 0, 1):
adev->mode_info.num_crtc = 4;
adev->mode_info.num_hpd = 4;
adev->mode_info.num_dig = 4;
break;
default:
- DRM_ERROR("Unsupported DCE IP versions: 0x%x\n",
+ drm_err(adev_to_drm(adev), "Unsupported DCE IP versions: 0x%x\n",
amdgpu_ip_version(adev, DCE_HWIP, 0));
return -EINVAL;
}
@@ -5502,9 +5741,9 @@ fill_plane_color_attributes(const struct drm_plane_state *plane_state,
case DRM_COLOR_YCBCR_BT2020:
if (full_range)
- *color_space = COLOR_SPACE_2020_YCBCR;
+ *color_space = COLOR_SPACE_2020_YCBCR_FULL;
else
- return -EINVAL;
+ *color_space = COLOR_SPACE_2020_YCBCR_LIMITED;
break;
default:
@@ -5520,8 +5759,7 @@ fill_dc_plane_info_and_addr(struct amdgpu_device *adev,
const u64 tiling_flags,
struct dc_plane_info *plane_info,
struct dc_plane_address *address,
- bool tmz_surface,
- bool force_disable_dcc)
+ bool tmz_surface)
{
const struct drm_framebuffer *fb = plane_state->fb;
const struct amdgpu_framebuffer *afb =
@@ -5580,7 +5818,7 @@ fill_dc_plane_info_and_addr(struct amdgpu_device *adev,
plane_info->format = SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616;
break;
default:
- DRM_ERROR(
+ drm_err(adev_to_drm(adev),
"Unsupported screen format %p4cc\n",
&fb->format->format);
return -EINVAL;
@@ -5620,7 +5858,7 @@ fill_dc_plane_info_and_addr(struct amdgpu_device *adev,
&plane_info->tiling_info,
&plane_info->plane_size,
&plane_info->dcc, address,
- tmz_surface, force_disable_dcc);
+ tmz_surface);
if (ret)
return ret;
@@ -5641,7 +5879,6 @@ static int fill_dc_plane_attributes(struct amdgpu_device *adev,
struct dc_scaling_info scaling_info;
struct dc_plane_info plane_info;
int ret;
- bool force_disable_dcc = false;
ret = amdgpu_dm_plane_fill_dc_scaling_info(adev, plane_state, &scaling_info);
if (ret)
@@ -5652,13 +5889,11 @@ static int fill_dc_plane_attributes(struct amdgpu_device *adev,
dc_plane_state->clip_rect = scaling_info.clip_rect;
dc_plane_state->scaling_quality = scaling_info.scaling_quality;
- force_disable_dcc = adev->asic_type == CHIP_RAVEN && adev->in_suspend;
ret = fill_dc_plane_info_and_addr(adev, plane_state,
afb->tiling_flags,
&plane_info,
&dc_plane_state->address,
- afb->tmz_surface,
- force_disable_dcc);
+ afb->tmz_surface);
if (ret)
return ret;
@@ -6004,12 +6239,14 @@ get_output_color_space(const struct dc_crtc_timing *dc_crtc_timing,
if (dc_crtc_timing->pixel_encoding == PIXEL_ENCODING_RGB)
color_space = COLOR_SPACE_2020_RGB_FULLRANGE;
else
- color_space = COLOR_SPACE_2020_YCBCR;
+ color_space = COLOR_SPACE_2020_YCBCR_LIMITED;
break;
case DRM_MODE_COLORIMETRY_DEFAULT: // ITU601
default:
if (dc_crtc_timing->pixel_encoding == PIXEL_ENCODING_RGB) {
color_space = COLOR_SPACE_SRGB;
+ if (connector_state->hdmi.broadcast_rgb == DRM_HDMI_BROADCAST_RGB_LIMITED)
+ color_space = COLOR_SPACE_SRGB_LIMITED;
/*
* 27030khz is the separation point between HDTV and SDTV
* according to HDMI spec, we use YCbCr709 and YCbCr601
@@ -6102,6 +6339,7 @@ static void fill_stream_properties_from_drm_display_mode(
struct amdgpu_dm_connector *aconnector = NULL;
struct hdmi_vendor_infoframe hv_frame;
struct hdmi_avi_infoframe avi_frame;
+ ssize_t err;
if (connector->connector_type != DRM_MODE_CONNECTOR_WRITEBACK)
aconnector = to_amdgpu_dm_connector(connector);
@@ -6148,9 +6386,17 @@ static void fill_stream_properties_from_drm_display_mode(
}
if (stream->signal == SIGNAL_TYPE_HDMI_TYPE_A) {
- drm_hdmi_avi_infoframe_from_display_mode(&avi_frame, (struct drm_connector *)connector, mode_in);
+ err = drm_hdmi_avi_infoframe_from_display_mode(&avi_frame,
+ (struct drm_connector *)connector,
+ mode_in);
+ if (err < 0)
+ drm_warn_once(connector->dev, "Failed to setup avi infoframe on connector %s: %zd \n", connector->name, err);
timing_out->vic = avi_frame.video_code;
- drm_hdmi_vendor_infoframe_from_display_mode(&hv_frame, (struct drm_connector *)connector, mode_in);
+ err = drm_hdmi_vendor_infoframe_from_display_mode(&hv_frame,
+ (struct drm_connector *)connector,
+ mode_in);
+ if (err < 0)
+ drm_warn_once(connector->dev, "Failed to setup vendor infoframe on connector %s: %zd \n", connector->name, err);
timing_out->hdmi_vic = hv_frame.vic;
}
@@ -6263,19 +6509,19 @@ decide_crtc_timing_for_drm_display_mode(struct drm_display_mode *drm_mode,
const struct drm_display_mode *native_mode,
bool scale_enabled)
{
- if (scale_enabled) {
- copy_crtc_timing_for_drm_display_mode(native_mode, drm_mode);
- } else if (native_mode->clock == drm_mode->clock &&
- native_mode->htotal == drm_mode->htotal &&
- native_mode->vtotal == drm_mode->vtotal) {
- copy_crtc_timing_for_drm_display_mode(native_mode, drm_mode);
+ if (scale_enabled || (
+ native_mode->clock == drm_mode->clock &&
+ native_mode->htotal == drm_mode->htotal &&
+ native_mode->vtotal == drm_mode->vtotal)) {
+ if (native_mode->crtc_clock)
+ copy_crtc_timing_for_drm_display_mode(native_mode, drm_mode);
} else {
/* no scaling nor amdgpu inserted, no need to patch */
}
}
static struct dc_sink *
-create_fake_sink(struct dc_link *link)
+create_fake_sink(struct drm_device *dev, struct dc_link *link)
{
struct dc_sink_init_data sink_init_data = { 0 };
struct dc_sink *sink = NULL;
@@ -6285,7 +6531,7 @@ create_fake_sink(struct dc_link *link)
sink = dc_sink_create(&sink_init_data);
if (!sink) {
- DRM_ERROR("Failed to create sink!\n");
+ drm_err(dev, "Failed to create sink!\n");
return NULL;
}
sink->sink_signal = SIGNAL_TYPE_VIRTUAL;
@@ -6418,7 +6664,7 @@ get_highest_refresh_rate_mode(struct amdgpu_dm_connector *aconnector,
m_pref = list_first_entry_or_null(
&aconnector->base.modes, struct drm_display_mode, head);
if (!m_pref) {
- DRM_DEBUG_DRIVER("No preferred mode found in EDID\n");
+ drm_dbg_driver(aconnector->base.dev, "No preferred mode found in EDID\n");
return NULL;
}
}
@@ -6593,7 +6839,7 @@ static void apply_dsc_policy_for_stream(struct amdgpu_dm_connector *aconnector,
dc_link_get_highest_encoding_format(aconnector->dc_link),
&stream->timing.dsc_cfg)) {
stream->timing.flags.DSC = 1;
- DRM_DEBUG_DRIVER("%s: SST_DSC [%s] DSC is selected from SST RX\n",
+ drm_dbg_driver(drm_connector->dev, "%s: SST_DSC [%s] DSC is selected from SST RX\n",
__func__, drm_connector->name);
}
} else if (sink->link->dpcd_caps.dongle_type == DISPLAY_DONGLE_DP_HDMI_CONVERTER) {
@@ -6613,7 +6859,7 @@ static void apply_dsc_policy_for_stream(struct amdgpu_dm_connector *aconnector,
dc_link_get_highest_encoding_format(aconnector->dc_link),
&stream->timing.dsc_cfg)) {
stream->timing.flags.DSC = 1;
- DRM_DEBUG_DRIVER("%s: SST_DSC [%s] DSC is selected from DP-HDMI PCON\n",
+ drm_dbg_driver(drm_connector->dev, "%s: SST_DSC [%s] DSC is selected from DP-HDMI PCON\n",
__func__, drm_connector->name);
}
}
@@ -6641,6 +6887,7 @@ create_stream_for_sink(struct drm_connector *connector,
const struct dc_stream_state *old_stream,
int requested_bpc)
{
+ struct drm_device *dev = connector->dev;
struct amdgpu_dm_connector *aconnector = NULL;
struct drm_display_mode *preferred_mode = NULL;
const struct drm_connector_state *con_state = &dm_state->base;
@@ -6663,11 +6910,6 @@ create_stream_for_sink(struct drm_connector *connector,
drm_mode_init(&mode, drm_mode);
memset(&saved_mode, 0, sizeof(saved_mode));
- if (connector == NULL) {
- DRM_ERROR("connector is NULL!\n");
- return stream;
- }
-
if (connector->connector_type != DRM_MODE_CONNECTOR_WRITEBACK) {
aconnector = NULL;
aconnector = to_amdgpu_dm_connector(connector);
@@ -6682,7 +6924,7 @@ create_stream_for_sink(struct drm_connector *connector,
}
if (!aconnector || !aconnector->dc_sink) {
- sink = create_fake_sink(link);
+ sink = create_fake_sink(dev, link);
if (!sink)
return stream;
@@ -6694,7 +6936,7 @@ create_stream_for_sink(struct drm_connector *connector,
stream = dc_create_stream_for_sink(sink);
if (stream == NULL) {
- DRM_ERROR("Failed to create stream for sink!\n");
+ drm_err(dev, "Failed to create stream for sink!\n");
goto finish;
}
@@ -6726,7 +6968,7 @@ create_stream_for_sink(struct drm_connector *connector,
* case, we call set mode ourselves to restore the previous mode
* and the modelist may not be filled in time.
*/
- DRM_DEBUG_DRIVER("No preferred mode found\n");
+ drm_dbg_driver(dev, "No preferred mode found\n");
} else if (aconnector) {
recalculate_timing = amdgpu_freesync_vid_mode &&
is_freesync_video_mode(&mode, aconnector);
@@ -7040,6 +7282,7 @@ static void amdgpu_dm_connector_unregister(struct drm_connector *connector)
if (amdgpu_dm_should_create_sysfs(amdgpu_dm_connector))
sysfs_remove_group(&connector->kdev->kobj, &amdgpu_group);
+ cec_notifier_conn_unregister(amdgpu_dm_connector->notifier);
drm_dp_aux_unregister(&amdgpu_dm_connector->dm_dp_aux.aux);
}
@@ -7174,11 +7417,18 @@ static void amdgpu_dm_connector_funcs_force(struct drm_connector *connector)
struct dc_link *dc_link = aconnector->dc_link;
struct dc_sink *dc_em_sink = aconnector->dc_em_sink;
const struct drm_edid *drm_edid;
+ struct i2c_adapter *ddc;
+ struct drm_device *dev = connector->dev;
- drm_edid = drm_edid_read(connector);
+ if (dc_link && dc_link->aux_mode)
+ ddc = &aconnector->dm_dp_aux.aux.ddc;
+ else
+ ddc = &aconnector->i2c->base;
+
+ drm_edid = drm_edid_read_ddc(connector, ddc);
drm_edid_connector_update(connector, drm_edid);
if (!drm_edid) {
- DRM_ERROR("No EDID found on connector: %s.\n", connector->name);
+ drm_err(dev, "No EDID found on connector: %s.\n", connector->name);
return;
}
@@ -7220,17 +7470,24 @@ static int get_modes(struct drm_connector *connector)
static void create_eml_sink(struct amdgpu_dm_connector *aconnector)
{
struct drm_connector *connector = &aconnector->base;
+ struct dc_link *dc_link = aconnector->dc_link;
struct dc_sink_init_data init_params = {
.link = aconnector->dc_link,
.sink_signal = SIGNAL_TYPE_VIRTUAL
};
const struct drm_edid *drm_edid;
const struct edid *edid;
+ struct i2c_adapter *ddc;
+
+ if (dc_link && dc_link->aux_mode)
+ ddc = &aconnector->dm_dp_aux.aux.ddc;
+ else
+ ddc = &aconnector->i2c->base;
- drm_edid = drm_edid_read(connector);
+ drm_edid = drm_edid_read_ddc(connector, ddc);
drm_edid_connector_update(connector, drm_edid);
if (!drm_edid) {
- DRM_ERROR("No EDID found on connector: %s.\n", connector->name);
+ drm_err(connector->dev, "No EDID found on connector: %s.\n", connector->name);
return;
}
@@ -7322,7 +7579,7 @@ static enum dc_status dm_validate_stream_and_context(struct dc *dc,
dc_result = DC_FAIL_ATTACH_SURFACES;
if (dc_result == DC_OK)
- dc_result = dc_validate_global_state(dc, dc_state, true);
+ dc_result = dc_validate_global_state(dc, dc_state, DC_VALIDATE_MODE_ONLY);
cleanup:
if (dc_state)
@@ -7335,12 +7592,12 @@ cleanup:
}
struct dc_stream_state *
-create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector,
+create_validate_stream_for_sink(struct drm_connector *connector,
const struct drm_display_mode *drm_mode,
const struct dm_connector_state *dm_state,
const struct dc_stream_state *old_stream)
{
- struct drm_connector *connector = &aconnector->base;
+ struct amdgpu_dm_connector *aconnector = NULL;
struct amdgpu_device *adev = drm_to_adev(connector->dev);
struct dc_stream_state *stream;
const struct drm_connector_state *drm_state = dm_state ? &dm_state->base : NULL;
@@ -7351,8 +7608,12 @@ create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector,
if (!dm_state)
return NULL;
- if (aconnector->dc_link->connector_signal == SIGNAL_TYPE_HDMI_TYPE_A ||
- aconnector->dc_link->dpcd_caps.dongle_type == DISPLAY_DONGLE_DP_HDMI_CONVERTER)
+ if (connector->connector_type != DRM_MODE_CONNECTOR_WRITEBACK)
+ aconnector = to_amdgpu_dm_connector(connector);
+
+ if (aconnector &&
+ (aconnector->dc_link->connector_signal == SIGNAL_TYPE_HDMI_TYPE_A ||
+ aconnector->dc_link->dpcd_caps.dongle_type == DISPLAY_DONGLE_DP_HDMI_CONVERTER))
bpc_limit = 8;
do {
@@ -7360,14 +7621,15 @@ create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector,
dm_state, old_stream,
requested_bpc);
if (stream == NULL) {
- DRM_ERROR("Failed to create stream for sink!\n");
+ drm_err(adev_to_drm(adev), "Failed to create stream for sink!\n");
break;
}
- if (aconnector->base.connector_type == DRM_MODE_CONNECTOR_WRITEBACK)
+ dc_result = dc_validate_stream(adev->dm.dc, stream);
+
+ if (!aconnector) /* writeback connector */
return stream;
- dc_result = dc_validate_stream(adev->dm.dc, stream);
if (dc_result == DC_OK && stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
dc_result = dm_dp_mst_is_port_support_mode(aconnector, stream);
@@ -7375,7 +7637,7 @@ create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector,
dc_result = dm_validate_stream_and_context(adev->dm.dc, stream);
if (dc_result != DC_OK) {
- DRM_DEBUG_KMS("Mode %dx%d (clk %d) pixel_encoding:%s color_depth:%s failed validation -- %s\n",
+ DRM_DEBUG_KMS("Pruned mode %d x %d (clk %d) %s %s -- %s\n",
drm_mode->hdisplay,
drm_mode->vdisplay,
drm_mode->clock,
@@ -7397,7 +7659,7 @@ create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector,
__func__, __LINE__);
aconnector->force_yuv420_output = true;
- stream = create_validate_stream_for_sink(aconnector, drm_mode,
+ stream = create_validate_stream_for_sink(connector, drm_mode,
dm_state, old_stream);
aconnector->force_yuv420_output = false;
}
@@ -7406,12 +7668,16 @@ create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector,
}
enum drm_mode_status amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
- struct drm_display_mode *mode)
+ const struct drm_display_mode *mode)
{
int result = MODE_ERROR;
struct dc_sink *dc_sink;
+ struct drm_display_mode *test_mode;
/* TODO: Unhardcode stream count */
struct dc_stream_state *stream;
+ /* we always have an amdgpu_dm_connector here since we got
+ * here via the amdgpu_dm_connector_helper_funcs
+ */
struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
if ((mode->flags & DRM_MODE_FLAG_INTERLACE) ||
@@ -7430,15 +7696,20 @@ enum drm_mode_status amdgpu_dm_connector_mode_valid(struct drm_connector *connec
if (dc_sink == NULL && aconnector->base.force != DRM_FORCE_ON_DIGITAL &&
aconnector->base.force != DRM_FORCE_ON) {
- DRM_ERROR("dc_sink is NULL!\n");
+ drm_err(connector->dev, "dc_sink is NULL!\n");
goto fail;
}
- drm_mode_set_crtcinfo(mode, 0);
+ test_mode = drm_mode_duplicate(connector->dev, mode);
+ if (!test_mode)
+ goto fail;
- stream = create_validate_stream_for_sink(aconnector, mode,
+ drm_mode_set_crtcinfo(test_mode, 0);
+
+ stream = create_validate_stream_for_sink(connector, test_mode,
to_dm_connector_state(connector->state),
NULL);
+ drm_mode_destroy(connector->dev, test_mode);
if (stream) {
dc_stream_release(stream);
result = MODE_OK;
@@ -7633,6 +7904,23 @@ static int dm_encoder_helper_atomic_check(struct drm_encoder *encoder,
int clock, bpp = 0;
bool is_y420 = false;
+ if ((connector->connector_type == DRM_MODE_CONNECTOR_eDP) ||
+ (connector->connector_type == DRM_MODE_CONNECTOR_LVDS)) {
+ struct amdgpu_encoder *amdgpu_encoder = to_amdgpu_encoder(encoder);
+ struct drm_display_mode *native_mode = &amdgpu_encoder->native_mode;
+ enum drm_mode_status result;
+
+ result = drm_crtc_helper_mode_valid_fixed(encoder->crtc, adjusted_mode, native_mode);
+ if (result != MODE_OK && dm_new_connector_state->scaling == RMX_OFF) {
+ drm_dbg_driver(encoder->dev,
+ "mode %dx%d@%dHz is not native, enabling scaling\n",
+ adjusted_mode->hdisplay, adjusted_mode->vdisplay,
+ drm_mode_vrefresh(adjusted_mode));
+ dm_new_connector_state->scaling = RMX_FULL;
+ }
+ return 0;
+ }
+
if (!aconnector->mst_output_port)
return 0;
@@ -8090,7 +8378,8 @@ static int amdgpu_dm_connector_get_modes(struct drm_connector *connector)
drm_add_modes_noedid(connector, 1920, 1080);
} else {
amdgpu_dm_connector_ddc_get_modes(connector, drm_edid);
- if (encoder)
+ if (encoder && (connector->connector_type != DRM_MODE_CONNECTOR_eDP) &&
+ (connector->connector_type != DRM_MODE_CONNECTOR_LVDS))
amdgpu_dm_connector_add_common_modes(encoder, connector);
amdgpu_dm_connector_add_freesync_modes(connector, drm_edid);
}
@@ -8164,6 +8453,10 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
dm->ddev->mode_config.scaling_mode_property,
DRM_MODE_SCALE_NONE);
+ if (connector_type == DRM_MODE_CONNECTOR_HDMIA
+ || (connector_type == DRM_MODE_CONNECTOR_DisplayPort && !aconnector->mst_root))
+ drm_connector_attach_broadcast_rgb_property(&aconnector->base);
+
drm_object_attach_property(&aconnector->base.base,
adev->mode_info.underscan_property,
UNDERSCAN_OFF);
@@ -8216,7 +8509,7 @@ static int amdgpu_dm_i2c_xfer(struct i2c_adapter *i2c_adap,
int i;
int result = -EIO;
- if (!ddc_service->ddc_pin || !ddc_service->ddc_pin->hw_info.hw_supported)
+ if (!ddc_service->ddc_pin)
return result;
cmd.payloads = kcalloc(num, sizeof(struct i2c_payload), GFP_KERNEL);
@@ -8235,11 +8528,18 @@ static int amdgpu_dm_i2c_xfer(struct i2c_adapter *i2c_adap,
cmd.payloads[i].data = msgs[i].buf;
}
- if (dc_submit_i2c(
- ddc_service->ctx->dc,
- ddc_service->link->link_index,
- &cmd))
- result = num;
+ if (i2c->oem) {
+ if (dc_submit_i2c_oem(
+ ddc_service->ctx->dc,
+ &cmd))
+ result = num;
+ } else {
+ if (dc_submit_i2c(
+ ddc_service->ctx->dc,
+ ddc_service->link->link_index,
+ &cmd))
+ result = num;
+ }
kfree(cmd.payloads);
return result;
@@ -8256,9 +8556,7 @@ static const struct i2c_algorithm amdgpu_dm_i2c_algo = {
};
static struct amdgpu_i2c_adapter *
-create_i2c(struct ddc_service *ddc_service,
- int link_index,
- int *res)
+create_i2c(struct ddc_service *ddc_service, bool oem)
{
struct amdgpu_device *adev = ddc_service->ctx->driver_context;
struct amdgpu_i2c_adapter *i2c;
@@ -8269,13 +8567,39 @@ create_i2c(struct ddc_service *ddc_service,
i2c->base.owner = THIS_MODULE;
i2c->base.dev.parent = &adev->pdev->dev;
i2c->base.algo = &amdgpu_dm_i2c_algo;
- snprintf(i2c->base.name, sizeof(i2c->base.name), "AMDGPU DM i2c hw bus %d", link_index);
+ if (oem)
+ snprintf(i2c->base.name, sizeof(i2c->base.name), "AMDGPU DM i2c OEM bus");
+ else
+ snprintf(i2c->base.name, sizeof(i2c->base.name), "AMDGPU DM i2c hw bus %d",
+ ddc_service->link->link_index);
i2c_set_adapdata(&i2c->base, i2c);
i2c->ddc_service = ddc_service;
+ i2c->oem = oem;
return i2c;
}
+int amdgpu_dm_initialize_hdmi_connector(struct amdgpu_dm_connector *aconnector)
+{
+ struct cec_connector_info conn_info;
+ struct drm_device *ddev = aconnector->base.dev;
+ struct device *hdmi_dev = ddev->dev;
+
+ if (amdgpu_dc_debug_mask & DC_DISABLE_HDMI_CEC) {
+ drm_info(ddev, "HDMI-CEC feature masked\n");
+ return -EINVAL;
+ }
+
+ cec_fill_conn_info_from_drm(&conn_info, &aconnector->base);
+ aconnector->notifier =
+ cec_notifier_conn_register(hdmi_dev, NULL, &conn_info);
+ if (!aconnector->notifier) {
+ drm_err(ddev, "Failed to create cec notifier\n");
+ return -ENOMEM;
+ }
+
+ return 0;
+}
/*
* Note: this function assumes that dc_link_detect() was called for the
@@ -8296,9 +8620,9 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm,
link->priv = aconnector;
- i2c = create_i2c(link->ddc, link->link_index, &res);
+ i2c = create_i2c(link->ddc, false);
if (!i2c) {
- DRM_ERROR("Failed to create i2c adapter data\n");
+ drm_err(adev_to_drm(dm->adev), "Failed to create i2c adapter data\n");
return -ENOMEM;
}
@@ -8306,7 +8630,7 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm,
res = i2c_add_adapter(&i2c->base);
if (res) {
- DRM_ERROR("Failed to register hw i2c %d\n", link->link_index);
+ drm_err(adev_to_drm(dm->adev), "Failed to register hw i2c %d\n", link->link_index);
goto out_free;
}
@@ -8320,7 +8644,7 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm,
&i2c->base);
if (res) {
- DRM_ERROR("connector_init failed\n");
+ drm_err(adev_to_drm(dm->adev), "connector_init failed\n");
aconnector->connector_id = -1;
goto out_free;
}
@@ -8339,6 +8663,10 @@ static int amdgpu_dm_connector_init(struct amdgpu_display_manager *dm,
drm_connector_attach_encoder(
&aconnector->base, &aencoder->base);
+ if (connector_type == DRM_MODE_CONNECTOR_HDMIA ||
+ connector_type == DRM_MODE_CONNECTOR_HDMIB)
+ amdgpu_dm_initialize_hdmi_connector(aconnector);
+
if (connector_type == DRM_MODE_CONNECTOR_DisplayPort
|| connector_type == DRM_MODE_CONNECTOR_eDP)
amdgpu_dm_initialize_dp_connector(dm, aconnector, link->link_index);
@@ -8398,29 +8726,44 @@ static void manage_dm_interrupts(struct amdgpu_device *adev,
struct amdgpu_crtc *acrtc,
struct dm_crtc_state *acrtc_state)
{
- /*
- * We have no guarantee that the frontend index maps to the same
- * backend index - some even map to more than one.
- *
- * TODO: Use a different interrupt or check DC itself for the mapping.
- */
- int irq_type =
- amdgpu_display_crtc_idx_to_irq_type(
- adev,
- acrtc->crtc_id);
struct drm_vblank_crtc_config config = {0};
struct dc_crtc_timing *timing;
int offdelay;
if (acrtc_state) {
- if (amdgpu_ip_version(adev, DCE_HWIP, 0) <
- IP_VERSION(3, 5, 0) ||
- acrtc_state->stream->link->psr_settings.psr_version <
- DC_PSR_VERSION_UNSUPPORTED ||
- !(adev->flags & AMD_IS_APU)) {
- timing = &acrtc_state->stream->timing;
-
- /* at least 2 frames */
+ timing = &acrtc_state->stream->timing;
+
+ /*
+ * Depending on when the HW latching event of double-buffered
+ * registers happen relative to the PSR SDP deadline, and how
+ * bad the Panel clock has drifted since the last ALPM off
+ * event, there can be up to 3 frames of delay between sending
+ * the PSR exit cmd to DMUB fw, and when the panel starts
+ * displaying live frames.
+ *
+ * We can set:
+ *
+ * 20/100 * offdelay_ms = 3_frames_ms
+ * => offdelay_ms = 5 * 3_frames_ms
+ *
+ * This ensures that `3_frames_ms` will only be experienced as a
+ * 20% delay on top how long the display has been static, and
+ * thus make the delay less perceivable.
+ */
+ if (acrtc_state->stream->link->psr_settings.psr_version <
+ DC_PSR_VERSION_UNSUPPORTED) {
+ offdelay = DIV64_U64_ROUND_UP((u64)5 * 3 * 10 *
+ timing->v_total *
+ timing->h_total,
+ timing->pix_clk_100hz);
+ config.offdelay_ms = offdelay ?: 30;
+ } else if (amdgpu_ip_version(adev, DCE_HWIP, 0) <
+ IP_VERSION(3, 5, 0) ||
+ !(adev->flags & AMD_IS_APU)) {
+ /*
+ * Older HW and DGPU have issues with instant off;
+ * use a 2 frame offdelay.
+ */
offdelay = DIV64_U64_ROUND_UP((u64)20 *
timing->v_total *
timing->h_total,
@@ -8428,33 +8771,14 @@ static void manage_dm_interrupts(struct amdgpu_device *adev,
config.offdelay_ms = offdelay ?: 30;
} else {
+ /* offdelay_ms = 0 will never disable vblank */
+ config.offdelay_ms = 1;
config.disable_immediate = true;
}
drm_crtc_vblank_on_config(&acrtc->base,
&config);
-
- amdgpu_irq_get(
- adev,
- &adev->pageflip_irq,
- irq_type);
-#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
- amdgpu_irq_get(
- adev,
- &adev->vline0_irq,
- irq_type);
-#endif
} else {
-#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
- amdgpu_irq_put(
- adev,
- &adev->vline0_irq,
- irq_type);
-#endif
- amdgpu_irq_put(
- adev,
- &adev->pageflip_irq,
- irq_type);
drm_crtc_vblank_off(&acrtc->base);
}
}
@@ -8810,7 +9134,7 @@ static void amdgpu_dm_handle_vrr_transition(struct dm_crtc_state *old_state,
*/
WARN_ON(amdgpu_dm_crtc_set_vupdate_irq(new_state->base.crtc, true) != 0);
WARN_ON(drm_crtc_vblank_get(new_state->base.crtc) != 0);
- DRM_DEBUG_DRIVER("%s: crtc=%u VRR off->on: Get vblank ref\n",
+ drm_dbg_driver(new_state->base.crtc->dev, "%s: crtc=%u VRR off->on: Get vblank ref\n",
__func__, new_state->base.crtc->base.id);
} else if (old_vrr_active && !new_vrr_active) {
/* Transition VRR active -> inactive:
@@ -8818,7 +9142,7 @@ static void amdgpu_dm_handle_vrr_transition(struct dm_crtc_state *old_state,
*/
WARN_ON(amdgpu_dm_crtc_set_vupdate_irq(new_state->base.crtc, false) != 0);
drm_crtc_vblank_put(new_state->base.crtc);
- DRM_DEBUG_DRIVER("%s: crtc=%u VRR on->off: Drop vblank ref\n",
+ drm_dbg_driver(new_state->base.crtc->dev, "%s: crtc=%u VRR on->off: Drop vblank ref\n",
__func__, new_state->base.crtc->base.id);
}
}
@@ -8905,13 +9229,13 @@ static void amdgpu_dm_update_cursor(struct drm_plane *plane,
if (crtc_state->stream) {
if (!dc_stream_set_cursor_attributes(crtc_state->stream,
&attributes))
- DRM_ERROR("DC failed to set cursor attributes\n");
+ drm_err(adev_to_drm(adev), "DC failed to set cursor attributes\n");
update->cursor_attributes = &crtc_state->stream->cursor_attributes;
if (!dc_stream_set_cursor_position(crtc_state->stream,
&position))
- DRM_ERROR("DC failed to set cursor position\n");
+ drm_err(adev_to_drm(adev), "DC failed to set cursor position\n");
update->cursor_position = &crtc_state->stream->cursor_position;
}
@@ -8925,6 +9249,7 @@ static void amdgpu_dm_enable_self_refresh(struct amdgpu_crtc *acrtc_attach,
struct replay_settings *pr = &acrtc_state->stream->link->replay_settings;
struct amdgpu_dm_connector *aconn =
(struct amdgpu_dm_connector *)acrtc_state->stream->dm_stream_context;
+ bool vrr_active = amdgpu_dm_crtc_vrr_active(acrtc_state);
if (acrtc_state->update_type > UPDATE_TYPE_FAST) {
if (pr->config.replay_supported && !pr->replay_feature_enabled)
@@ -8951,14 +9276,15 @@ static void amdgpu_dm_enable_self_refresh(struct amdgpu_crtc *acrtc_attach,
* adequate number of fast atomic commits to notify KMD
* of update events. See `vblank_control_worker()`.
*/
- if (acrtc_attach->dm_irq_params.allow_sr_entry &&
+ if (!vrr_active &&
+ acrtc_attach->dm_irq_params.allow_sr_entry &&
#ifdef CONFIG_DRM_AMD_SECURE_DISPLAY
!amdgpu_dm_crc_window_is_activated(acrtc_state->base.crtc) &&
#endif
(current_ts - psr->psr_dirty_rects_change_timestamp_ns) > 500000000) {
if (pr->replay_feature_enabled && !pr->replay_allow_active)
amdgpu_dm_replay_enable(acrtc_state->stream, true);
- if (psr->psr_version >= DC_PSR_VERSION_SU_1 &&
+ if (psr->psr_version == DC_PSR_VERSION_SU_1 &&
!psr->psr_allow_active && !aconn->disallow_edp_enter_psr)
amdgpu_dm_psr_enable(acrtc_state->stream);
}
@@ -9095,7 +9421,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
afb->tiling_flags,
&bundle->plane_infos[planes_count],
&bundle->flip_addrs[planes_count].address,
- afb->tmz_surface, false);
+ afb->tmz_surface);
drm_dbg_state(state->dev, "plane: id=%d dcc_en=%d\n",
new_plane_state->plane->index,
@@ -9129,7 +9455,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
acrtc_state->stream->link->psr_settings.psr_dirty_rects_change_timestamp_ns =
timestamp_ns;
if (acrtc_state->stream->link->psr_settings.psr_allow_active)
- amdgpu_dm_psr_disable(acrtc_state->stream);
+ amdgpu_dm_psr_disable(acrtc_state->stream, true);
mutex_unlock(&dm->dc_lock);
}
}
@@ -9160,7 +9486,7 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
bundle->surface_updates[planes_count].surface = dc_plane;
if (!bundle->surface_updates[planes_count].surface) {
- DRM_ERROR("No surface for CRTC: id=%d\n",
+ drm_err(dev, "No surface for CRTC: id=%d\n",
acrtc_attach->crtc_id);
continue;
}
@@ -9295,11 +9621,11 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
bundle->stream_update.abm_level = &acrtc_state->abm_level;
mutex_lock(&dm->dc_lock);
- if (acrtc_state->update_type > UPDATE_TYPE_FAST) {
+ if ((acrtc_state->update_type > UPDATE_TYPE_FAST) || vrr_active) {
if (acrtc_state->stream->link->replay_settings.replay_allow_active)
amdgpu_dm_replay_disable(acrtc_state->stream);
if (acrtc_state->stream->link->psr_settings.psr_allow_active)
- amdgpu_dm_psr_disable(acrtc_state->stream);
+ amdgpu_dm_psr_disable(acrtc_state->stream, true);
}
mutex_unlock(&dm->dc_lock);
@@ -9676,20 +10002,20 @@ static void dm_set_writeback(struct amdgpu_display_manager *dm,
wb_info = kzalloc(sizeof(*wb_info), GFP_KERNEL);
if (!wb_info) {
- DRM_ERROR("Failed to allocate wb_info\n");
+ drm_err(adev_to_drm(adev), "Failed to allocate wb_info\n");
return;
}
acrtc = to_amdgpu_crtc(wb_conn->encoder.crtc);
if (!acrtc) {
- DRM_ERROR("no amdgpu_crtc found\n");
+ drm_err(adev_to_drm(adev), "no amdgpu_crtc found\n");
kfree(wb_info);
return;
}
afb = to_amdgpu_framebuffer(new_con_state->writeback_job->fb);
if (!afb) {
- DRM_ERROR("No amdgpu_framebuffer found\n");
+ drm_err(adev_to_drm(adev), "No amdgpu_framebuffer found\n");
kfree(wb_info);
return;
}
@@ -9910,7 +10236,7 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
new_con_state->content_protection >= DRM_MODE_CONTENT_PROTECTION_DESIRED)
enable_encryption = true;
- DRM_INFO("[HDCP_DM] hdcp_update_display enable_encryption = %x\n", enable_encryption);
+ drm_info(adev_to_drm(adev), "[HDCP_DM] hdcp_update_display enable_encryption = %x\n", enable_encryption);
if (aconnector->dc_link)
hdcp_update_display(
@@ -9928,7 +10254,7 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
struct dc_stream_update stream_update;
struct dc_info_packet hdr_packet;
struct dc_stream_status *status = NULL;
- bool abm_changed, hdr_changed, scaling_changed;
+ bool abm_changed, hdr_changed, scaling_changed, output_color_space_changed = false;
memset(&stream_update, 0, sizeof(stream_update));
@@ -9947,13 +10273,18 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
scaling_changed = is_scaling_state_different(dm_new_con_state,
dm_old_con_state);
+ if ((new_con_state->hdmi.broadcast_rgb != old_con_state->hdmi.broadcast_rgb) &&
+ (dm_old_crtc_state->stream->output_color_space !=
+ get_output_color_space(&dm_new_crtc_state->stream->timing, new_con_state)))
+ output_color_space_changed = true;
+
abm_changed = dm_new_crtc_state->abm_level !=
dm_old_crtc_state->abm_level;
hdr_changed =
!drm_connector_atomic_hdr_metadata_equal(old_con_state, new_con_state);
- if (!scaling_changed && !abm_changed && !hdr_changed)
+ if (!scaling_changed && !abm_changed && !hdr_changed && !output_color_space_changed)
continue;
stream_update.stream = dm_new_crtc_state->stream;
@@ -9965,6 +10296,13 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
stream_update.dst = dm_new_crtc_state->stream->dst;
}
+ if (output_color_space_changed) {
+ dm_new_crtc_state->stream->output_color_space
+ = get_output_color_space(&dm_new_crtc_state->stream->timing, new_con_state);
+
+ stream_update.output_color_space = &dm_new_crtc_state->stream->output_color_space;
+ }
+
if (abm_changed) {
dm_new_crtc_state->stream->abm_level = dm_new_crtc_state->abm_level;
@@ -9990,7 +10328,7 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
*/
dummy_updates = kzalloc(sizeof(struct dc_surface_update) * MAX_SURFACES, GFP_ATOMIC);
if (!dummy_updates) {
- DRM_ERROR("Failed to allocate memory for dummy_updates.\n");
+ drm_err(adev_to_drm(adev), "Failed to allocate memory for dummy_updates.\n");
continue;
}
for (j = 0; j < status->plane_count; j++)
@@ -10058,14 +10396,19 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
if (amdgpu_dm_is_valid_crc_source(cur_crc_src)) {
#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
if (amdgpu_dm_crc_window_is_activated(crtc)) {
+ uint8_t cnt;
spin_lock_irqsave(&adev_to_drm(adev)->event_lock, flags);
- acrtc->dm_irq_params.window_param.update_win = true;
-
- /**
- * It takes 2 frames for HW to stably generate CRC when
- * resuming from suspend, so we set skip_frame_cnt 2.
- */
- acrtc->dm_irq_params.window_param.skip_frame_cnt = 2;
+ for (cnt = 0; cnt < MAX_CRC_WINDOW_NUM; cnt++) {
+ if (acrtc->dm_irq_params.window_param[cnt].enable) {
+ acrtc->dm_irq_params.window_param[cnt].update_win = true;
+
+ /**
+ * It takes 2 frames for HW to stably generate CRC when
+ * resuming from suspend, so we set skip_frame_cnt 2.
+ */
+ acrtc->dm_irq_params.window_param[cnt].skip_frame_cnt = 2;
+ }
+ }
spin_unlock_irqrestore(&adev_to_drm(adev)->event_lock, flags);
}
#endif
@@ -10193,16 +10536,20 @@ static int dm_force_atomic_commit(struct drm_connector *connector)
*/
conn_state = drm_atomic_get_connector_state(state, connector);
- ret = PTR_ERR_OR_ZERO(conn_state);
- if (ret)
+ /* Check for error in getting connector state */
+ if (IS_ERR(conn_state)) {
+ ret = PTR_ERR(conn_state);
goto out;
+ }
/* Attach crtc to drm_atomic_state*/
crtc_state = drm_atomic_get_crtc_state(state, &disconnected_acrtc->base);
- ret = PTR_ERR_OR_ZERO(crtc_state);
- if (ret)
+ /* Check for error in getting crtc state */
+ if (IS_ERR(crtc_state)) {
+ ret = PTR_ERR(crtc_state);
goto out;
+ }
/* force a restore */
crtc_state->mode_changed = true;
@@ -10210,9 +10557,11 @@ static int dm_force_atomic_commit(struct drm_connector *connector)
/* Attach plane to drm_atomic_state */
plane_state = drm_atomic_get_plane_state(state, plane);
- ret = PTR_ERR_OR_ZERO(plane_state);
- if (ret)
+ /* Check for error in getting plane state */
+ if (IS_ERR(plane_state)) {
+ ret = PTR_ERR(plane_state);
goto out;
+ }
/* Call commit internally with the state we just constructed */
ret = drm_atomic_commit(state);
@@ -10220,7 +10569,7 @@ static int dm_force_atomic_commit(struct drm_connector *connector)
out:
drm_atomic_state_put(state);
if (ret)
- DRM_ERROR("Restoring old state failed with %i\n", ret);
+ drm_err(ddev, "Restoring old state failed with %i\n", ret);
return ret;
}
@@ -10304,7 +10653,7 @@ static int do_aquire_global_lock(struct drm_device *dev,
&commit->flip_done, 10*HZ);
if (ret == 0)
- DRM_ERROR("[CRTC:%d:%s] hw_done or flip_done timed out\n",
+ drm_err(dev, "[CRTC:%d:%s] hw_done or flip_done timed out\n",
crtc->base.id, crtc->name);
drm_crtc_commit_put(commit);
@@ -10420,6 +10769,7 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm,
struct dm_atomic_state *dm_state = NULL;
struct dm_crtc_state *dm_old_crtc_state, *dm_new_crtc_state;
struct dc_stream_state *new_stream;
+ struct amdgpu_device *adev = dm->adev;
int ret = 0;
/*
@@ -10449,8 +10799,8 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm,
drm_old_conn_state = drm_atomic_get_old_connector_state(state,
connector);
- if (IS_ERR(drm_new_conn_state)) {
- ret = PTR_ERR_OR_ZERO(drm_new_conn_state);
+ if (WARN_ON(!drm_new_conn_state)) {
+ ret = -EINVAL;
goto fail;
}
@@ -10460,7 +10810,7 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm,
if (!drm_atomic_crtc_needs_modeset(new_crtc_state))
goto skip_modeset;
- new_stream = create_validate_stream_for_sink(aconnector,
+ new_stream = create_validate_stream_for_sink(connector,
&new_crtc_state->mode,
dm_new_conn_state,
dm_old_crtc_state->stream);
@@ -10473,7 +10823,7 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm,
*/
if (!new_stream) {
- DRM_DEBUG_DRIVER("%s: Failed to create new stream for crtc %d\n",
+ drm_dbg_driver(adev_to_drm(adev), "%s: Failed to create new stream for crtc %d\n",
__func__, acrtc->base.base.id);
ret = -ENOMEM;
goto fail;
@@ -10511,7 +10861,7 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm,
dc_is_stream_unchanged(new_stream, dm_old_crtc_state->stream) &&
dc_is_stream_scaling_unchanged(new_stream, dm_old_crtc_state->stream)) {
new_crtc_state->mode_changed = false;
- DRM_DEBUG_DRIVER("Mode change not required, setting mode_changed to %d",
+ drm_dbg_driver(adev_to_drm(adev), "Mode change not required, setting mode_changed to %d",
new_crtc_state->mode_changed);
}
}
@@ -10549,7 +10899,7 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm,
is_timing_unchanged_for_freesync(new_crtc_state,
old_crtc_state)) {
new_crtc_state->mode_changed = false;
- DRM_DEBUG_DRIVER(
+ drm_dbg_driver(adev_to_drm(adev),
"Mode change not required for front porch change, setting mode_changed to %d",
new_crtc_state->mode_changed);
@@ -10570,7 +10920,7 @@ static int dm_update_crtc_state(struct amdgpu_display_manager *dm,
if (ret)
goto fail;
- DRM_DEBUG_DRIVER("Disabling DRM crtc: %d\n",
+ drm_dbg_driver(adev_to_drm(adev), "Disabling DRM crtc: %d\n",
crtc->base.id);
/* i.e. reset mode */
@@ -10703,6 +11053,9 @@ static bool should_reset_plane(struct drm_atomic_state *state,
state->allow_modeset)
return true;
+ if (amdgpu_in_reset(adev) && state->allow_modeset)
+ return true;
+
/* Exit early if we know that we're adding or removing the plane. */
if (old_plane_state->crtc != new_plane_state->crtc)
return true;
@@ -11153,8 +11506,8 @@ dm_get_plane_scale(struct drm_plane_state *plane_state,
int plane_src_w, plane_src_h;
dm_get_oriented_plane_size(plane_state, &plane_src_w, &plane_src_h);
- *out_plane_scale_w = plane_state->crtc_w * 1000 / plane_src_w;
- *out_plane_scale_h = plane_state->crtc_h * 1000 / plane_src_h;
+ *out_plane_scale_w = plane_src_w ? plane_state->crtc_w * 1000 / plane_src_w : 0;
+ *out_plane_scale_h = plane_src_h ? plane_state->crtc_h * 1000 / plane_src_h : 0;
}
/*
@@ -11408,6 +11761,30 @@ static int dm_crtc_get_cursor_mode(struct amdgpu_device *adev,
return 0;
}
+static bool amdgpu_dm_crtc_mem_type_changed(struct drm_device *dev,
+ struct drm_atomic_state *state,
+ struct drm_crtc_state *crtc_state)
+{
+ struct drm_plane *plane;
+ struct drm_plane_state *new_plane_state, *old_plane_state;
+
+ drm_for_each_plane_mask(plane, dev, crtc_state->plane_mask) {
+ new_plane_state = drm_atomic_get_plane_state(state, plane);
+ old_plane_state = drm_atomic_get_plane_state(state, plane);
+
+ if (IS_ERR(new_plane_state) || IS_ERR(old_plane_state)) {
+ drm_err(dev, "Failed to get plane state for plane %s\n", plane->name);
+ return false;
+ }
+
+ if (old_plane_state->fb && new_plane_state->fb &&
+ get_mem_type(old_plane_state->fb) != get_mem_type(new_plane_state->fb))
+ return true;
+ }
+
+ return false;
+}
+
/**
* amdgpu_dm_atomic_check() - Atomic check implementation for AMDgpu DM.
*
@@ -11605,10 +11982,6 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev,
/* Remove exiting planes if they are modified */
for_each_oldnew_plane_in_descending_zpos(state, plane, old_plane_state, new_plane_state) {
- if (old_plane_state->fb && new_plane_state->fb &&
- get_mem_type(old_plane_state->fb) !=
- get_mem_type(new_plane_state->fb))
- lock_and_validation_needed = true;
ret = dm_update_plane_state(dc, state, plane,
old_plane_state,
@@ -11846,7 +12219,7 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev,
drm_dbg_atomic(dev, "MST drm_dp_mst_atomic_check() failed\n");
goto fail;
}
- status = dc_validate_global_state(dc, dm_state->context, true);
+ status = dc_validate_global_state(dc, dm_state->context, DC_VALIDATE_MODE_ONLY);
if (status != DC_OK) {
drm_dbg_atomic(dev, "DC global validation failure: %s (%d)",
dc_status_to_str(status), status);
@@ -11903,9 +12276,11 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev,
/*
* Only allow async flips for fast updates that don't change
- * the FB pitch, the DCC state, rotation, etc.
+ * the FB pitch, the DCC state, rotation, mem_type, etc.
*/
- if (new_crtc_state->async_flip && lock_and_validation_needed) {
+ if (new_crtc_state->async_flip &&
+ (lock_and_validation_needed ||
+ amdgpu_dm_crtc_mem_type_changed(dev, state, new_crtc_state))) {
drm_dbg_atomic(crtc->dev,
"[CRTC:%d:%s] async flips are only supported for fast updates\n",
crtc->base.id, crtc->name);
@@ -11967,7 +12342,7 @@ static bool dm_edid_parser_send_cea(struct amdgpu_display_manager *dm,
res = dc_wake_and_execute_dmub_cmd(dm->dc->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY);
if (!res) {
- DRM_ERROR("EDID CEA parser failed\n");
+ drm_err(adev_to_drm(dm->adev), "EDID CEA parser failed\n");
return false;
}
@@ -11975,7 +12350,7 @@ static bool dm_edid_parser_send_cea(struct amdgpu_display_manager *dm,
if (output->type == DMUB_CMD__EDID_CEA_ACK) {
if (!output->ack.success) {
- DRM_ERROR("EDID CEA ack failed at offset %d\n",
+ drm_err(adev_to_drm(dm->adev), "EDID CEA ack failed at offset %d\n",
output->ack.offset);
}
} else if (output->type == DMUB_CMD__EDID_CEA_AMD_VSDB) {
@@ -11987,7 +12362,7 @@ static bool dm_edid_parser_send_cea(struct amdgpu_display_manager *dm,
vsdb->min_refresh_rate_hz = output->amd_vsdb.min_frame_rate;
vsdb->max_refresh_rate_hz = output->amd_vsdb.max_frame_rate;
} else {
- DRM_WARN("Unknown EDID CEA parser results\n");
+ drm_warn(adev_to_drm(dm->adev), "Unknown EDID CEA parser results\n");
return false;
}
@@ -12203,7 +12578,7 @@ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector,
enum adaptive_sync_type as_type = ADAPTIVE_SYNC_TYPE_NONE;
if (!connector->state) {
- DRM_ERROR("%s - Connector has no state", __func__);
+ drm_err(adev_to_drm(adev), "%s - Connector has no state", __func__);
goto update;
}
@@ -12237,10 +12612,14 @@ void amdgpu_dm_update_freesync_caps(struct drm_connector *connector,
if (edid && (sink->sink_signal == SIGNAL_TYPE_DISPLAY_PORT ||
sink->sink_signal == SIGNAL_TYPE_EDP)) {
- amdgpu_dm_connector->min_vfreq = connector->display_info.monitor_range.min_vfreq;
- amdgpu_dm_connector->max_vfreq = connector->display_info.monitor_range.max_vfreq;
- if (amdgpu_dm_connector->max_vfreq - amdgpu_dm_connector->min_vfreq > 10)
- freesync_capable = true;
+ if (amdgpu_dm_connector->dc_link &&
+ amdgpu_dm_connector->dc_link->dpcd_caps.allow_invalid_MSA_timing_param) {
+ amdgpu_dm_connector->min_vfreq = connector->display_info.monitor_range.min_vfreq;
+ amdgpu_dm_connector->max_vfreq = connector->display_info.monitor_range.max_vfreq;
+ if (amdgpu_dm_connector->max_vfreq - amdgpu_dm_connector->min_vfreq > 10)
+ freesync_capable = true;
+ }
+
parse_amd_vsdb(amdgpu_dm_connector, edid, &vsdb_info);
if (vsdb_info.replay_mode) {
@@ -12384,7 +12763,7 @@ int amdgpu_dm_process_dmub_aux_transfer_sync(
}
if (!wait_for_completion_timeout(&adev->dm.dmub_aux_transfer_done, 10 * HZ)) {
- DRM_ERROR("wait_for_completion_timeout timeout!");
+ drm_err(adev_to_drm(adev), "wait_for_completion_timeout timeout!");
*operation_result = AUX_RET_ERROR_TIMEOUT;
goto out;
}
@@ -12394,31 +12773,24 @@ int amdgpu_dm_process_dmub_aux_transfer_sync(
* Transient states before tunneling is enabled could
* lead to this error. We can ignore this for now.
*/
- if (p_notify->result != AUX_RET_ERROR_PROTOCOL_ERROR) {
- DRM_WARN("DPIA AUX failed on 0x%x(%d), error %d\n",
+ if (p_notify->result == AUX_RET_ERROR_PROTOCOL_ERROR) {
+ drm_warn(adev_to_drm(adev), "DPIA AUX failed on 0x%x(%d), error %d\n",
payload->address, payload->length,
p_notify->result);
}
- *operation_result = AUX_RET_ERROR_INVALID_REPLY;
+ *operation_result = p_notify->result;
goto out;
}
+ payload->reply[0] = adev->dm.dmub_notify->aux_reply.command & 0xF;
+ if (adev->dm.dmub_notify->aux_reply.command & 0xF0)
+ /* The reply is stored in the top nibble of the command. */
+ payload->reply[0] = (adev->dm.dmub_notify->aux_reply.command >> 4) & 0xF;
- payload->reply[0] = adev->dm.dmub_notify->aux_reply.command;
- if (!payload->write && p_notify->aux_reply.length &&
- (payload->reply[0] == AUX_TRANSACTION_REPLY_AUX_ACK)) {
-
- if (payload->length != p_notify->aux_reply.length) {
- DRM_WARN("invalid read length %d from DPIA AUX 0x%x(%d)!\n",
- p_notify->aux_reply.length,
- payload->address, payload->length);
- *operation_result = AUX_RET_ERROR_INVALID_REPLY;
- goto out;
- }
-
+ /*write req may receive a byte indicating partially written number as well*/
+ if (p_notify->aux_reply.length)
memcpy(payload->data, p_notify->aux_reply.data,
p_notify->aux_reply.length);
- }
/* success */
ret = p_notify->aux_reply.length;
@@ -12429,6 +12801,79 @@ out:
return ret;
}
+static void abort_fused_io(
+ struct dc_context *ctx,
+ const struct dmub_cmd_fused_request *request
+)
+{
+ union dmub_rb_cmd command = { 0 };
+ struct dmub_rb_cmd_fused_io *io = &command.fused_io;
+
+ io->header.type = DMUB_CMD__FUSED_IO;
+ io->header.sub_type = DMUB_CMD__FUSED_IO_ABORT;
+ io->header.payload_bytes = sizeof(*io) - sizeof(io->header);
+ io->request = *request;
+ dm_execute_dmub_cmd(ctx, &command, DM_DMUB_WAIT_TYPE_NO_WAIT);
+}
+
+static bool execute_fused_io(
+ struct amdgpu_device *dev,
+ struct dc_context *ctx,
+ union dmub_rb_cmd *commands,
+ uint8_t count,
+ uint32_t timeout_us
+)
+{
+ const uint8_t ddc_line = commands[0].fused_io.request.u.aux.ddc_line;
+
+ if (ddc_line >= ARRAY_SIZE(dev->dm.fused_io))
+ return false;
+
+ struct fused_io_sync *sync = &dev->dm.fused_io[ddc_line];
+ struct dmub_rb_cmd_fused_io *first = &commands[0].fused_io;
+ const bool result = dm_execute_dmub_cmd_list(ctx, count, commands, DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)
+ && first->header.ret_status
+ && first->request.status == FUSED_REQUEST_STATUS_SUCCESS;
+
+ if (!result)
+ return false;
+
+ while (wait_for_completion_timeout(&sync->replied, usecs_to_jiffies(timeout_us))) {
+ reinit_completion(&sync->replied);
+
+ struct dmub_cmd_fused_request *reply = (struct dmub_cmd_fused_request *) sync->reply_data;
+
+ static_assert(sizeof(*reply) <= sizeof(sync->reply_data), "Size mismatch");
+
+ if (reply->identifier == first->request.identifier) {
+ first->request = *reply;
+ return true;
+ }
+ }
+
+ reinit_completion(&sync->replied);
+ first->request.status = FUSED_REQUEST_STATUS_TIMEOUT;
+ abort_fused_io(ctx, &first->request);
+ return false;
+}
+
+bool amdgpu_dm_execute_fused_io(
+ struct amdgpu_device *dev,
+ struct dc_link *link,
+ union dmub_rb_cmd *commands,
+ uint8_t count,
+ uint32_t timeout_us)
+{
+ struct amdgpu_display_manager *dm = &dev->dm;
+
+ mutex_lock(&dm->dpia_aux_lock);
+
+ const bool result = execute_fused_io(dev, link->ctx, commands, count, timeout_us);
+
+ mutex_unlock(&dm->dpia_aux_lock);
+ return result;
+}
+
int amdgpu_dm_process_dmub_set_config_sync(
struct dc_context *ctx,
unsigned int link_index,
@@ -12447,7 +12892,7 @@ int amdgpu_dm_process_dmub_set_config_sync(
ret = 0;
*operation_result = adev->dm.dmub_notify->sc_status;
} else {
- DRM_ERROR("wait_for_completion_timeout timeout!");
+ drm_err(adev_to_drm(adev), "wait_for_completion_timeout timeout!");
ret = -1;
*operation_result = SET_CONFIG_UNKNOWN_ERROR;
}
@@ -12467,3 +12912,10 @@ bool dm_execute_dmub_cmd_list(const struct dc_context *ctx, unsigned int count,
{
return dc_dmub_srv_cmd_run_list(ctx->dmub_srv, count, cmd, wait_type);
}
+
+void dm_acpi_process_phy_transition_interlock(
+ const struct dc_context *ctx,
+ struct dm_process_phy_transition_init_params process_phy_transition_init_params)
+{
+ // Not yet implemented
+}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
index 6464a8378387..b937da0a4e4a 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -50,7 +50,7 @@
#define AMDGPU_DM_MAX_NUM_EDP 2
-#define AMDGPU_DMUB_NOTIFICATION_MAX 7
+#define AMDGPU_DMUB_NOTIFICATION_MAX 8
#define HDMI_AMD_VENDOR_SPECIFIC_DATA_BLOCK_IEEE_REGISTRATION_ID 0x00001A
#define AMD_VSDB_VERSION_3_FEATURECAP_REPLAYMODE 0x40
@@ -81,6 +81,7 @@ struct amdgpu_bo;
struct dmub_srv;
struct dc_plane_state;
struct dmub_notification;
+struct dmub_cmd_fused_request;
struct amd_vsdb_block {
unsigned char ieee_id[3];
@@ -151,6 +152,18 @@ struct idle_workqueue {
bool running;
};
+#define MAX_LUMINANCE_DATA_POINTS 99
+
+/**
+ * struct amdgpu_dm_luminance_data - Custom luminance data
+ * @luminance: Luminance in percent
+ * @input_signal: Input signal in range 0-255
+ */
+struct amdgpu_dm_luminance_data {
+ u8 luminance;
+ u8 input_signal;
+} __packed;
+
/**
* struct amdgpu_dm_backlight_caps - Information about backlight
*
@@ -195,6 +208,14 @@ struct amdgpu_dm_backlight_caps {
* @dc_level: the default brightness if booted on DC
*/
u8 dc_level;
+ /**
+ * @data_points: the number of custom luminance data points
+ */
+ u8 data_points;
+ /**
+ * @luminance_data: custom luminance data
+ */
+ struct amdgpu_dm_luminance_data luminance_data[MAX_LUMINANCE_DATA_POINTS];
};
/**
@@ -256,6 +277,10 @@ struct hpd_rx_irq_offload_work {
* @offload_wq: offload work queue that this work is queued to
*/
struct hpd_rx_irq_offload_work_queue *offload_wq;
+ /**
+ * @adev: amdgpu_device pointer
+ */
+ struct amdgpu_device *adev;
};
/**
@@ -541,12 +566,12 @@ struct amdgpu_display_manager {
#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
/**
- * @secure_display_ctxs:
+ * @secure_display_ctx:
*
- * Store the ROI information and the work_struct to command dmub and psp for
- * all crtcs.
+ * Store secure display relevant info. e.g. the ROI information
+ * , the work_struct to command dmub, etc.
*/
- struct secure_display_context *secure_display_ctxs;
+ struct secure_display_context secure_display_ctx;
#endif
/**
* @hpd_rx_offload_wq:
@@ -594,6 +619,13 @@ struct amdgpu_display_manager {
bool aux_hpd_discon_quirk;
/**
+ * @edp0_on_dp1_quirk:
+ *
+ * quirk for platforms that put edp0 on DP1.
+ */
+ bool edp0_on_dp1_quirk;
+
+ /**
* @dpia_aux_lock:
*
* Guards access to DPIA AUX
@@ -604,8 +636,26 @@ struct amdgpu_display_manager {
* @bb_from_dmub:
*
* Bounding box data read from dmub during early initialization for DCN4+
+ * Data is stored as a byte array that should be casted to the appropriate bb struct
*/
- struct dml2_soc_bb *bb_from_dmub;
+ void *bb_from_dmub;
+
+ /**
+ * @oem_i2c:
+ *
+ * OEM i2c bus
+ */
+ struct amdgpu_i2c_adapter *oem_i2c;
+
+ /**
+ * @fused_io:
+ *
+ * dmub fused io interface
+ */
+ struct fused_io_sync {
+ struct completion replied;
+ char reply_data[0x40]; // Cannot include dmub_cmd here
+ } fused_io[8];
};
enum dsc_clock_force_state {
@@ -671,6 +721,8 @@ struct amdgpu_dm_connector {
uint32_t connector_id;
int bl_idx;
+ struct cec_notifier *notifier;
+
/* we need to mind the EDID between detect
and get modes due to analog/digital/tvencoder */
const struct drm_edid *drm_edid;
@@ -697,6 +749,8 @@ struct amdgpu_dm_connector {
struct drm_dp_mst_port *mst_output_port;
struct amdgpu_dm_connector *mst_root;
struct drm_dp_aux *dsc_aux;
+ uint32_t mst_local_bw;
+ uint16_t vc_full_pbn;
struct mutex handle_mst_msg_ready;
/* TODO see if we can merge with ddc_bus or make a dm_connector */
@@ -945,7 +999,7 @@ void amdgpu_dm_connector_init_helper(struct amdgpu_display_manager *dm,
int link_index);
enum drm_mode_status amdgpu_dm_connector_mode_valid(struct drm_connector *connector,
- struct drm_display_mode *mode);
+ const struct drm_display_mode *mode);
void dm_restore_drm_connector_state(struct drm_device *dev,
struct drm_connector *connector);
@@ -981,11 +1035,19 @@ extern const struct drm_encoder_helper_funcs amdgpu_dm_encoder_helper_funcs;
int amdgpu_dm_process_dmub_aux_transfer_sync(struct dc_context *ctx, unsigned int link_index,
struct aux_payload *payload, enum aux_return_code_type *operation_result);
+bool amdgpu_dm_execute_fused_io(
+ struct amdgpu_device *dev,
+ struct dc_link *link,
+ union dmub_rb_cmd *commands,
+ uint8_t count,
+ uint32_t timeout_us
+);
+
int amdgpu_dm_process_dmub_set_config_sync(struct dc_context *ctx, unsigned int link_index,
struct set_config_cmd_payload *payload, enum set_config_status *operation_result);
struct dc_stream_state *
- create_validate_stream_for_sink(struct amdgpu_dm_connector *aconnector,
+ create_validate_stream_for_sink(struct drm_connector *connector,
const struct drm_display_mode *drm_mode,
const struct dm_connector_state *dm_state,
const struct dc_stream_state *old_stream);
@@ -1010,4 +1072,10 @@ void dm_free_gpu_mem(struct amdgpu_device *adev,
bool amdgpu_dm_is_headless(struct amdgpu_device *adev);
+void hdmi_cec_set_edid(struct amdgpu_dm_connector *aconnector);
+void hdmi_cec_unset_edid(struct amdgpu_dm_connector *aconnector);
+int amdgpu_dm_initialize_hdmi_connector(struct amdgpu_dm_connector *aconnector);
+
+void retrieve_dmi_info(struct amdgpu_display_manager *dm);
+
#endif /* __AMDGPU_DM_H__ */
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
index f936a35fa9eb..033bd817d871 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
@@ -30,6 +30,7 @@
#include "amdgpu_dm.h"
#include "dc.h"
#include "amdgpu_securedisplay.h"
+#include "amdgpu_dm_psr.h"
static const char *const pipe_crc_sources[] = {
"none",
@@ -83,45 +84,274 @@ const char *const *amdgpu_dm_crtc_get_crc_sources(struct drm_crtc *crtc,
}
#ifdef CONFIG_DRM_AMD_SECURE_DISPLAY
+static void update_phy_id_mapping(struct amdgpu_device *adev)
+{
+ struct drm_device *ddev = adev_to_drm(adev);
+ struct amdgpu_display_manager *dm = &adev->dm;
+ struct drm_connector *connector;
+ struct amdgpu_dm_connector *aconnector;
+ struct amdgpu_dm_connector *sort_connector[AMDGPU_DM_MAX_CRTC] = {NULL};
+ struct drm_connector_list_iter iter;
+ uint8_t idx = 0, idx_2 = 0, connector_cnt = 0;
+
+ dm->secure_display_ctx.phy_mapping_updated = false;
+
+ mutex_lock(&ddev->mode_config.mutex);
+ drm_connector_list_iter_begin(ddev, &iter);
+ drm_for_each_connector_iter(connector, &iter) {
+
+ if (connector->status != connector_status_connected)
+ continue;
+
+ if (idx >= AMDGPU_DM_MAX_CRTC) {
+ DRM_WARN("%s connected connectors exceed max crtc\n", __func__);
+ mutex_unlock(&ddev->mode_config.mutex);
+ return;
+ }
+
+ aconnector = to_amdgpu_dm_connector(connector);
+
+ sort_connector[idx] = aconnector;
+ idx++;
+ connector_cnt++;
+ }
+ drm_connector_list_iter_end(&iter);
+
+ /* sort connectors by link_enc_hw_instance first */
+ for (idx = connector_cnt; idx > 1 ; idx--) {
+ for (idx_2 = 0; idx_2 < (idx - 1); idx_2++) {
+ if (sort_connector[idx_2]->dc_link->link_enc_hw_inst >
+ sort_connector[idx_2 + 1]->dc_link->link_enc_hw_inst)
+ swap(sort_connector[idx_2], sort_connector[idx_2 + 1]);
+ }
+ }
+
+ /*
+ * Sort mst connectors by RAD. mst connectors with the same enc_hw_instance are already
+ * sorted together above.
+ */
+ for (idx = 0; idx < connector_cnt; /*Do nothing*/) {
+ if (sort_connector[idx]->mst_root) {
+ uint8_t i, j, k;
+ uint8_t mst_con_cnt = 1;
+
+ for (idx_2 = (idx + 1); idx_2 < connector_cnt; idx_2++) {
+ if (sort_connector[idx_2]->mst_root == sort_connector[idx]->mst_root)
+ mst_con_cnt++;
+ else
+ break;
+ }
+
+ for (i = mst_con_cnt; i > 1; i--) {
+ for (j = idx; j < (idx + i - 2); j++) {
+ int mstb_lct = sort_connector[j]->mst_output_port->parent->lct;
+ int next_mstb_lct = sort_connector[j + 1]->mst_output_port->parent->lct;
+ u8 *rad;
+ u8 *next_rad;
+ bool swap = false;
+
+ /* Sort by mst tree depth first. Then compare RAD if depth is the same*/
+ if (mstb_lct > next_mstb_lct) {
+ swap = true;
+ } else if (mstb_lct == next_mstb_lct) {
+ if (mstb_lct == 1) {
+ if (sort_connector[j]->mst_output_port->port_num > sort_connector[j + 1]->mst_output_port->port_num)
+ swap = true;
+ } else if (mstb_lct > 1) {
+ rad = sort_connector[j]->mst_output_port->parent->rad;
+ next_rad = sort_connector[j + 1]->mst_output_port->parent->rad;
+
+ for (k = 0; k < mstb_lct - 1; k++) {
+ int shift = (k % 2) ? 0 : 4;
+ int port_num = (rad[k / 2] >> shift) & 0xf;
+ int next_port_num = (next_rad[k / 2] >> shift) & 0xf;
+
+ if (port_num > next_port_num) {
+ swap = true;
+ break;
+ }
+ }
+ } else {
+ DRM_ERROR("MST LCT shouldn't be set as < 1");
+ mutex_unlock(&ddev->mode_config.mutex);
+ return;
+ }
+ }
+
+ if (swap)
+ swap(sort_connector[j], sort_connector[j + 1]);
+ }
+ }
+
+ idx += mst_con_cnt;
+ } else {
+ idx++;
+ }
+ }
+
+ /* Complete sorting. Assign relavant result to dm->secure_display_ctx.phy_id_mapping[]*/
+ memset(dm->secure_display_ctx.phy_id_mapping, 0, sizeof(dm->secure_display_ctx.phy_id_mapping));
+ for (idx = 0; idx < connector_cnt; idx++) {
+ aconnector = sort_connector[idx];
+
+ dm->secure_display_ctx.phy_id_mapping[idx].assigned = true;
+ dm->secure_display_ctx.phy_id_mapping[idx].is_mst = false;
+ dm->secure_display_ctx.phy_id_mapping[idx].enc_hw_inst = aconnector->dc_link->link_enc_hw_inst;
+
+ if (sort_connector[idx]->mst_root) {
+ dm->secure_display_ctx.phy_id_mapping[idx].is_mst = true;
+ dm->secure_display_ctx.phy_id_mapping[idx].lct = aconnector->mst_output_port->parent->lct;
+ dm->secure_display_ctx.phy_id_mapping[idx].port_num = aconnector->mst_output_port->port_num;
+ memcpy(dm->secure_display_ctx.phy_id_mapping[idx].rad,
+ aconnector->mst_output_port->parent->rad, sizeof(aconnector->mst_output_port->parent->rad));
+ }
+ }
+ mutex_unlock(&ddev->mode_config.mutex);
+
+ dm->secure_display_ctx.phy_id_mapping_cnt = connector_cnt;
+ dm->secure_display_ctx.phy_mapping_updated = true;
+}
+
+static bool get_phy_id(struct amdgpu_display_manager *dm,
+ struct amdgpu_dm_connector *aconnector, uint8_t *phy_id)
+{
+ int idx, idx_2;
+ bool found = false;
+
+ /*
+ * Assume secure display start after all connectors are probed. The connection
+ * config is static as well
+ */
+ if (!dm->secure_display_ctx.phy_mapping_updated) {
+ DRM_WARN("%s Should update the phy id table before get it's value", __func__);
+ return false;
+ }
+
+ for (idx = 0; idx < dm->secure_display_ctx.phy_id_mapping_cnt; idx++) {
+ if (!dm->secure_display_ctx.phy_id_mapping[idx].assigned) {
+ DRM_ERROR("phy_id_mapping[%d] should be assigned", idx);
+ return false;
+ }
+
+ if (aconnector->dc_link->link_enc_hw_inst ==
+ dm->secure_display_ctx.phy_id_mapping[idx].enc_hw_inst) {
+ if (!dm->secure_display_ctx.phy_id_mapping[idx].is_mst) {
+ found = true;
+ goto out;
+ } else {
+ /* Could caused by wrongly pass mst root connector */
+ if (!aconnector->mst_output_port) {
+ DRM_ERROR("%s Check mst case but connector without a port assigned", __func__);
+ return false;
+ }
+
+ if (aconnector->mst_root &&
+ aconnector->mst_root->mst_mgr.mst_primary == NULL) {
+ DRM_WARN("%s pass in a stale mst connector", __func__);
+ }
+
+ if (aconnector->mst_output_port->parent->lct == dm->secure_display_ctx.phy_id_mapping[idx].lct &&
+ aconnector->mst_output_port->port_num == dm->secure_display_ctx.phy_id_mapping[idx].port_num) {
+ if (aconnector->mst_output_port->parent->lct == 1) {
+ found = true;
+ goto out;
+ } else if (aconnector->mst_output_port->parent->lct > 1) {
+ /* Check RAD */
+ for (idx_2 = 0; idx_2 < aconnector->mst_output_port->parent->lct - 1; idx_2++) {
+ int shift = (idx_2 % 2) ? 0 : 4;
+ int port_num = (aconnector->mst_output_port->parent->rad[idx_2 / 2] >> shift) & 0xf;
+ int port_num2 = (dm->secure_display_ctx.phy_id_mapping[idx].rad[idx_2 / 2] >> shift) & 0xf;
+
+ if (port_num != port_num2)
+ break;
+ }
+
+ if (idx_2 == aconnector->mst_output_port->parent->lct - 1) {
+ found = true;
+ goto out;
+ }
+ } else {
+ DRM_ERROR("lCT should be >= 1");
+ return false;
+ }
+ }
+ }
+ }
+ }
+
+out:
+ if (found) {
+ DRM_DEBUG_DRIVER("Associated secure display PHY ID as %d", idx);
+ *phy_id = idx;
+ } else {
+ DRM_WARN("Can't find associated phy ID");
+ return false;
+ }
+
+ return true;
+}
+
static void amdgpu_dm_set_crc_window_default(struct drm_crtc *crtc, struct dc_stream_state *stream)
{
struct drm_device *drm_dev = crtc->dev;
struct amdgpu_display_manager *dm = &drm_to_adev(drm_dev)->dm;
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
+ struct amdgpu_dm_connector *aconnector;
bool was_activated;
+ uint8_t phy_id;
+ unsigned long flags;
+ int i;
- spin_lock_irq(&drm_dev->event_lock);
- was_activated = acrtc->dm_irq_params.window_param.activated;
- acrtc->dm_irq_params.window_param.x_start = 0;
- acrtc->dm_irq_params.window_param.y_start = 0;
- acrtc->dm_irq_params.window_param.x_end = 0;
- acrtc->dm_irq_params.window_param.y_end = 0;
- acrtc->dm_irq_params.window_param.activated = false;
- acrtc->dm_irq_params.window_param.update_win = false;
- acrtc->dm_irq_params.window_param.skip_frame_cnt = 0;
- spin_unlock_irq(&drm_dev->event_lock);
+ spin_lock_irqsave(&drm_dev->event_lock, flags);
+ was_activated = acrtc->dm_irq_params.crc_window_activated;
+ for (i = 0; i < MAX_CRC_WINDOW_NUM; i++) {
+ acrtc->dm_irq_params.window_param[i].x_start = 0;
+ acrtc->dm_irq_params.window_param[i].y_start = 0;
+ acrtc->dm_irq_params.window_param[i].x_end = 0;
+ acrtc->dm_irq_params.window_param[i].y_end = 0;
+ acrtc->dm_irq_params.window_param[i].enable = false;
+ acrtc->dm_irq_params.window_param[i].update_win = false;
+ acrtc->dm_irq_params.window_param[i].skip_frame_cnt = 0;
+ }
+ acrtc->dm_irq_params.crc_window_activated = false;
+ spin_unlock_irqrestore(&drm_dev->event_lock, flags);
/* Disable secure_display if it was enabled */
- if (was_activated) {
+ if (was_activated && dm->secure_display_ctx.op_mode == LEGACY_MODE) {
/* stop ROI update on this crtc */
- flush_work(&dm->secure_display_ctxs[crtc->index].notify_ta_work);
- flush_work(&dm->secure_display_ctxs[crtc->index].forward_roi_work);
- dc_stream_forward_crc_window(stream, NULL, true);
+ flush_work(&dm->secure_display_ctx.crtc_ctx[crtc->index].notify_ta_work);
+ flush_work(&dm->secure_display_ctx.crtc_ctx[crtc->index].forward_roi_work);
+ aconnector = (struct amdgpu_dm_connector *)stream->dm_stream_context;
+
+ if (aconnector && get_phy_id(dm, aconnector, &phy_id)) {
+ if (dm->secure_display_ctx.support_mul_roi)
+ dc_stream_forward_multiple_crc_window(stream, NULL, phy_id, true);
+ else
+ dc_stream_forward_crc_window(stream, NULL, phy_id, true);
+ } else {
+ DRM_DEBUG_DRIVER("%s Can't find matching phy id", __func__);
+ }
}
}
static void amdgpu_dm_crtc_notify_ta_to_read(struct work_struct *work)
{
- struct secure_display_context *secure_display_ctx;
+ struct secure_display_crtc_context *crtc_ctx;
struct psp_context *psp;
struct ta_securedisplay_cmd *securedisplay_cmd;
struct drm_crtc *crtc;
struct dc_stream_state *stream;
+ struct amdgpu_dm_connector *aconnector;
uint8_t phy_inst;
+ struct amdgpu_display_manager *dm;
+ struct crc_data crc_cpy[MAX_CRC_WINDOW_NUM];
+ unsigned long flags;
+ uint8_t roi_idx = 0;
int ret;
+ int i;
- secure_display_ctx = container_of(work, struct secure_display_context, notify_ta_work);
- crtc = secure_display_ctx->crtc;
+ crtc_ctx = container_of(work, struct secure_display_crtc_context, notify_ta_work);
+ crtc = crtc_ctx->crtc;
if (!crtc)
return;
@@ -133,21 +363,50 @@ static void amdgpu_dm_crtc_notify_ta_to_read(struct work_struct *work)
return;
}
+ dm = &drm_to_adev(crtc->dev)->dm;
stream = to_amdgpu_crtc(crtc)->dm_irq_params.stream;
- phy_inst = stream->link->link_enc_hw_inst;
-
- /* need lock for multiple crtcs to use the command buffer */
- mutex_lock(&psp->securedisplay_context.mutex);
+ aconnector = (struct amdgpu_dm_connector *)stream->dm_stream_context;
+ if (!aconnector)
+ return;
- psp_prep_securedisplay_cmd_buf(psp, &securedisplay_cmd,
- TA_SECUREDISPLAY_COMMAND__SEND_ROI_CRC);
+ mutex_lock(&crtc->dev->mode_config.mutex);
+ if (!get_phy_id(dm, aconnector, &phy_inst)) {
+ DRM_WARN("%s Can't find mapping phy id!", __func__);
+ mutex_unlock(&crtc->dev->mode_config.mutex);
+ return;
+ }
+ mutex_unlock(&crtc->dev->mode_config.mutex);
- securedisplay_cmd->securedisplay_in_message.send_roi_crc.phy_id = phy_inst;
+ spin_lock_irqsave(&crtc->dev->event_lock, flags);
+ memcpy(crc_cpy, crtc_ctx->crc_info.crc, sizeof(struct crc_data) * MAX_CRC_WINDOW_NUM);
+ spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
+ /* need lock for multiple crtcs to use the command buffer */
+ mutex_lock(&psp->securedisplay_context.mutex);
/* PSP TA is expected to finish data transmission over I2C within current frame,
* even there are up to 4 crtcs request to send in this frame.
*/
- ret = psp_securedisplay_invoke(psp, TA_SECUREDISPLAY_COMMAND__SEND_ROI_CRC);
+ if (dm->secure_display_ctx.support_mul_roi) {
+ psp_prep_securedisplay_cmd_buf(psp, &securedisplay_cmd,
+ TA_SECUREDISPLAY_COMMAND__SEND_ROI_CRC_V2);
+
+ securedisplay_cmd->securedisplay_in_message.send_roi_crc_v2.phy_id = phy_inst;
+
+ for (i = 0; i < MAX_CRC_WINDOW_NUM; i++) {
+ if (crc_cpy[i].crc_ready)
+ roi_idx |= 1 << i;
+ }
+ securedisplay_cmd->securedisplay_in_message.send_roi_crc_v2.roi_idx = roi_idx;
+
+ ret = psp_securedisplay_invoke(psp, TA_SECUREDISPLAY_COMMAND__SEND_ROI_CRC_V2);
+ } else {
+ psp_prep_securedisplay_cmd_buf(psp, &securedisplay_cmd,
+ TA_SECUREDISPLAY_COMMAND__SEND_ROI_CRC);
+
+ securedisplay_cmd->securedisplay_in_message.send_roi_crc.phy_id = phy_inst;
+
+ ret = psp_securedisplay_invoke(psp, TA_SECUREDISPLAY_COMMAND__SEND_ROI_CRC);
+ }
if (!ret) {
if (securedisplay_cmd->status != TA_SECUREDISPLAY_STATUS__SUCCESS)
@@ -160,22 +419,47 @@ static void amdgpu_dm_crtc_notify_ta_to_read(struct work_struct *work)
static void
amdgpu_dm_forward_crc_window(struct work_struct *work)
{
- struct secure_display_context *secure_display_ctx;
+ struct secure_display_crtc_context *crtc_ctx;
struct amdgpu_display_manager *dm;
struct drm_crtc *crtc;
struct dc_stream_state *stream;
+ struct amdgpu_dm_connector *aconnector;
+ struct crc_window roi_cpy[MAX_CRC_WINDOW_NUM];
+ unsigned long flags;
+ uint8_t phy_id;
- secure_display_ctx = container_of(work, struct secure_display_context, forward_roi_work);
- crtc = secure_display_ctx->crtc;
+ crtc_ctx = container_of(work, struct secure_display_crtc_context, forward_roi_work);
+ crtc = crtc_ctx->crtc;
if (!crtc)
return;
dm = &drm_to_adev(crtc->dev)->dm;
stream = to_amdgpu_crtc(crtc)->dm_irq_params.stream;
+ aconnector = (struct amdgpu_dm_connector *)stream->dm_stream_context;
+
+ if (!aconnector)
+ return;
+
+ mutex_lock(&crtc->dev->mode_config.mutex);
+ if (!get_phy_id(dm, aconnector, &phy_id)) {
+ DRM_WARN("%s Can't find mapping phy id!", __func__);
+ mutex_unlock(&crtc->dev->mode_config.mutex);
+ return;
+ }
+ mutex_unlock(&crtc->dev->mode_config.mutex);
+
+ spin_lock_irqsave(&crtc->dev->event_lock, flags);
+ memcpy(roi_cpy, crtc_ctx->roi, sizeof(struct crc_window) * MAX_CRC_WINDOW_NUM);
+ spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
mutex_lock(&dm->dc_lock);
- dc_stream_forward_crc_window(stream, &secure_display_ctx->rect, false);
+ if (dm->secure_display_ctx.support_mul_roi)
+ dc_stream_forward_multiple_crc_window(stream, roi_cpy,
+ phy_id, false);
+ else
+ dc_stream_forward_crc_window(stream, &roi_cpy[0].rect,
+ phy_id, false);
mutex_unlock(&dm->dc_lock);
}
@@ -186,7 +470,7 @@ bool amdgpu_dm_crc_window_is_activated(struct drm_crtc *crtc)
bool ret = false;
spin_lock_irq(&drm_dev->event_lock);
- ret = acrtc->dm_irq_params.window_param.activated;
+ ret = acrtc->dm_irq_params.crc_window_activated;
spin_unlock_irq(&drm_dev->event_lock);
return ret;
@@ -224,10 +508,14 @@ int amdgpu_dm_crtc_configure_crc_source(struct drm_crtc *crtc,
mutex_lock(&adev->dm.dc_lock);
+ /* For PSR1, check that the panel has exited PSR */
+ if (stream_state->link->psr_settings.psr_version < DC_PSR_VERSION_SU_1)
+ amdgpu_dm_psr_wait_disable(stream_state);
+
/* Enable or disable CRTC CRC generation */
if (dm_is_crc_source_crtc(source) || source == AMDGPU_DM_PIPE_CRC_SOURCE_NONE) {
if (!dc_stream_configure_crc(stream_state->ctx->dc,
- stream_state, NULL, enable, enable)) {
+ stream_state, NULL, enable, enable, 0, true)) {
ret = -EINVAL;
goto unlock;
}
@@ -258,6 +546,10 @@ int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name)
struct drm_crtc_commit *commit;
struct dm_crtc_state *crtc_state;
struct drm_device *drm_dev = crtc->dev;
+#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
+ struct amdgpu_device *adev = drm_to_adev(drm_dev);
+ struct amdgpu_display_manager *dm = &adev->dm;
+#endif
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
struct drm_dp_aux *aux = NULL;
bool enable = false;
@@ -357,6 +649,17 @@ int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name)
}
+ /*
+ * Reading the CRC requires the vblank interrupt handler to be
+ * enabled. Keep a reference until CRC capture stops.
+ */
+ enabled = amdgpu_dm_is_valid_crc_source(cur_crc_src);
+ if (!enabled && enable) {
+ ret = drm_crtc_vblank_get(crtc);
+ if (ret)
+ goto cleanup;
+ }
+
#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
/* Reset secure_display when we change crc source from debugfs */
amdgpu_dm_set_crc_window_default(crtc, crtc_state->stream);
@@ -367,16 +670,7 @@ int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name)
goto cleanup;
}
- /*
- * Reading the CRC requires the vblank interrupt handler to be
- * enabled. Keep a reference until CRC capture stops.
- */
- enabled = amdgpu_dm_is_valid_crc_source(cur_crc_src);
if (!enabled && enable) {
- ret = drm_crtc_vblank_get(crtc);
- if (ret)
- goto cleanup;
-
if (dm_is_crc_source_dprx(source)) {
if (drm_dp_start_crc(aux, crtc)) {
DRM_DEBUG_DRIVER("dp start crc failed\n");
@@ -402,6 +696,13 @@ int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name)
/* Reset crc_skipped on dm state */
crtc_state->crc_skip_count = 0;
+#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
+ /* Initialize phy id mapping table for secure display*/
+ if (dm->secure_display_ctx.op_mode == LEGACY_MODE &&
+ !dm->secure_display_ctx.phy_mapping_updated)
+ update_phy_id_mapping(adev);
+#endif
+
cleanup:
if (commit)
drm_crtc_commit_put(commit);
@@ -456,7 +757,7 @@ void amdgpu_dm_crtc_handle_crc_irq(struct drm_crtc *crtc)
}
if (dm_is_crc_source_crtc(cur_crc_src)) {
- if (!dc_stream_get_crc(stream_state->ctx->dc, stream_state,
+ if (!dc_stream_get_crc(stream_state->ctx->dc, stream_state, 0,
&crcs[0], &crcs[1], &crcs[2]))
return;
@@ -472,8 +773,17 @@ void amdgpu_dm_crtc_handle_crc_window_irq(struct drm_crtc *crtc)
enum amdgpu_dm_pipe_crc_source cur_crc_src;
struct amdgpu_crtc *acrtc = NULL;
struct amdgpu_device *adev = NULL;
- struct secure_display_context *secure_display_ctx = NULL;
+ struct secure_display_crtc_context *crtc_ctx = NULL;
+ bool reset_crc_frame_count[MAX_CRC_WINDOW_NUM] = {false};
+ uint32_t crc_r[MAX_CRC_WINDOW_NUM] = {0};
+ uint32_t crc_g[MAX_CRC_WINDOW_NUM] = {0};
+ uint32_t crc_b[MAX_CRC_WINDOW_NUM] = {0};
unsigned long flags1;
+ bool forward_roi_change = false;
+ bool notify_ta = false;
+ bool all_crc_ready = true;
+ struct dc_stream_state *stream_state;
+ int i;
if (crtc == NULL)
return;
@@ -481,78 +791,160 @@ void amdgpu_dm_crtc_handle_crc_window_irq(struct drm_crtc *crtc)
acrtc = to_amdgpu_crtc(crtc);
adev = drm_to_adev(crtc->dev);
drm_dev = crtc->dev;
+ stream_state = to_dm_crtc_state(crtc->state)->stream;
spin_lock_irqsave(&drm_dev->event_lock, flags1);
cur_crc_src = acrtc->dm_irq_params.crc_src;
/* Early return if CRC capture is not enabled. */
if (!amdgpu_dm_is_valid_crc_source(cur_crc_src) ||
- !dm_is_crc_source_crtc(cur_crc_src))
- goto cleanup;
-
- if (!acrtc->dm_irq_params.window_param.activated)
- goto cleanup;
+ !dm_is_crc_source_crtc(cur_crc_src)) {
+ spin_unlock_irqrestore(&drm_dev->event_lock, flags1);
+ return;
+ }
- if (acrtc->dm_irq_params.window_param.skip_frame_cnt) {
- acrtc->dm_irq_params.window_param.skip_frame_cnt -= 1;
- goto cleanup;
+ if (!acrtc->dm_irq_params.crc_window_activated) {
+ spin_unlock_irqrestore(&drm_dev->event_lock, flags1);
+ return;
}
- secure_display_ctx = &adev->dm.secure_display_ctxs[acrtc->crtc_id];
- if (WARN_ON(secure_display_ctx->crtc != crtc)) {
- /* We have set the crtc when creating secure_display_context,
+ crtc_ctx = &adev->dm.secure_display_ctx.crtc_ctx[acrtc->crtc_id];
+ if (WARN_ON(crtc_ctx->crtc != crtc)) {
+ /* We have set the crtc when creating secure_display_crtc_context,
* don't expect it to be changed here.
*/
- secure_display_ctx->crtc = crtc;
+ crtc_ctx->crtc = crtc;
}
- if (acrtc->dm_irq_params.window_param.update_win) {
- /* prepare work for dmub to update ROI */
- secure_display_ctx->rect.x = acrtc->dm_irq_params.window_param.x_start;
- secure_display_ctx->rect.y = acrtc->dm_irq_params.window_param.y_start;
- secure_display_ctx->rect.width = acrtc->dm_irq_params.window_param.x_end -
- acrtc->dm_irq_params.window_param.x_start;
- secure_display_ctx->rect.height = acrtc->dm_irq_params.window_param.y_end -
- acrtc->dm_irq_params.window_param.y_start;
- schedule_work(&secure_display_ctx->forward_roi_work);
-
- acrtc->dm_irq_params.window_param.update_win = false;
+ for (i = 0; i < MAX_CRC_WINDOW_NUM; i++) {
+ struct crc_params crc_window = {
+ .windowa_x_start = acrtc->dm_irq_params.window_param[i].x_start,
+ .windowa_y_start = acrtc->dm_irq_params.window_param[i].y_start,
+ .windowa_x_end = acrtc->dm_irq_params.window_param[i].x_end,
+ .windowa_y_end = acrtc->dm_irq_params.window_param[i].y_end,
+ .windowb_x_start = acrtc->dm_irq_params.window_param[i].x_start,
+ .windowb_y_start = acrtc->dm_irq_params.window_param[i].y_start,
+ .windowb_x_end = acrtc->dm_irq_params.window_param[i].x_end,
+ .windowb_y_end = acrtc->dm_irq_params.window_param[i].y_end,
+ };
+
+ crtc_ctx->roi[i].enable = acrtc->dm_irq_params.window_param[i].enable;
+
+ if (!acrtc->dm_irq_params.window_param[i].enable) {
+ crtc_ctx->crc_info.crc[i].crc_ready = false;
+ continue;
+ }
- /* Statically skip 1 frame, because we may need to wait below things
- * before sending ROI to dmub:
- * 1. We defer the work by using system workqueue.
- * 2. We may need to wait for dc_lock before accessing dmub.
- */
- acrtc->dm_irq_params.window_param.skip_frame_cnt = 1;
+ if (acrtc->dm_irq_params.window_param[i].skip_frame_cnt) {
+ acrtc->dm_irq_params.window_param[i].skip_frame_cnt -= 1;
+ crtc_ctx->crc_info.crc[i].crc_ready = false;
+ continue;
+ }
- } else {
- /* prepare work for psp to read ROI/CRC and send to I2C */
- schedule_work(&secure_display_ctx->notify_ta_work);
+ if (acrtc->dm_irq_params.window_param[i].update_win) {
+ crtc_ctx->roi[i].rect.x = crc_window.windowa_x_start;
+ crtc_ctx->roi[i].rect.y = crc_window.windowa_y_start;
+ crtc_ctx->roi[i].rect.width = crc_window.windowa_x_end -
+ crc_window.windowa_x_start;
+ crtc_ctx->roi[i].rect.height = crc_window.windowa_y_end -
+ crc_window.windowa_y_start;
+
+ if (adev->dm.secure_display_ctx.op_mode == LEGACY_MODE)
+ /* forward task to dmub to update ROI */
+ forward_roi_change = true;
+ else if (adev->dm.secure_display_ctx.op_mode == DISPLAY_CRC_MODE)
+ /* update ROI via dm*/
+ dc_stream_configure_crc(stream_state->ctx->dc, stream_state,
+ &crc_window, true, true, i, false);
+
+ reset_crc_frame_count[i] = true;
+
+ acrtc->dm_irq_params.window_param[i].update_win = false;
+
+ /* Statically skip 1 frame, because we may need to wait below things
+ * before sending ROI to dmub:
+ * 1. We defer the work by using system workqueue.
+ * 2. We may need to wait for dc_lock before accessing dmub.
+ */
+ acrtc->dm_irq_params.window_param[i].skip_frame_cnt = 1;
+ crtc_ctx->crc_info.crc[i].crc_ready = false;
+ } else {
+ if (!dc_stream_get_crc(stream_state->ctx->dc, stream_state, i,
+ &crc_r[i], &crc_g[i], &crc_b[i]))
+ DRM_ERROR("Secure Display: fail to get crc from engine %d\n", i);
+
+ if (adev->dm.secure_display_ctx.op_mode == LEGACY_MODE)
+ /* forward task to psp to read ROI/CRC and output via I2C */
+ notify_ta = true;
+ else if (adev->dm.secure_display_ctx.op_mode == DISPLAY_CRC_MODE)
+ /* Avoid ROI window get changed, keep overwriting. */
+ dc_stream_configure_crc(stream_state->ctx->dc, stream_state,
+ &crc_window, true, true, i, false);
+
+ /* crc ready for psp to read out */
+ crtc_ctx->crc_info.crc[i].crc_ready = true;
+ }
}
-cleanup:
spin_unlock_irqrestore(&drm_dev->event_lock, flags1);
+
+ if (forward_roi_change)
+ schedule_work(&crtc_ctx->forward_roi_work);
+
+ if (notify_ta)
+ schedule_work(&crtc_ctx->notify_ta_work);
+
+ spin_lock_irqsave(&crtc_ctx->crc_info.lock, flags1);
+ for (i = 0; i < MAX_CRC_WINDOW_NUM; i++) {
+ crtc_ctx->crc_info.crc[i].crc_R = crc_r[i];
+ crtc_ctx->crc_info.crc[i].crc_G = crc_g[i];
+ crtc_ctx->crc_info.crc[i].crc_B = crc_b[i];
+
+ if (!crtc_ctx->roi[i].enable) {
+ crtc_ctx->crc_info.crc[i].frame_count = 0;
+ continue;
+ }
+
+ if (!crtc_ctx->crc_info.crc[i].crc_ready)
+ all_crc_ready = false;
+
+ if (reset_crc_frame_count[i] || crtc_ctx->crc_info.crc[i].frame_count == UINT_MAX)
+ /* Reset the reference frame count after user update the ROI
+ * or it reaches the maximum value.
+ */
+ crtc_ctx->crc_info.crc[i].frame_count = 0;
+ else
+ crtc_ctx->crc_info.crc[i].frame_count += 1;
+ }
+ spin_unlock_irqrestore(&crtc_ctx->crc_info.lock, flags1);
+
+ if (all_crc_ready)
+ complete_all(&crtc_ctx->crc_info.completion);
}
-struct secure_display_context *
-amdgpu_dm_crtc_secure_display_create_contexts(struct amdgpu_device *adev)
+void amdgpu_dm_crtc_secure_display_create_contexts(struct amdgpu_device *adev)
{
- struct secure_display_context *secure_display_ctxs = NULL;
+ struct secure_display_crtc_context *crtc_ctx = NULL;
int i;
- secure_display_ctxs = kcalloc(adev->mode_info.num_crtc,
- sizeof(struct secure_display_context),
+ crtc_ctx = kcalloc(adev->mode_info.num_crtc,
+ sizeof(struct secure_display_crtc_context),
GFP_KERNEL);
- if (!secure_display_ctxs)
- return NULL;
+ if (!crtc_ctx) {
+ adev->dm.secure_display_ctx.crtc_ctx = NULL;
+ return;
+ }
for (i = 0; i < adev->mode_info.num_crtc; i++) {
- INIT_WORK(&secure_display_ctxs[i].forward_roi_work, amdgpu_dm_forward_crc_window);
- INIT_WORK(&secure_display_ctxs[i].notify_ta_work, amdgpu_dm_crtc_notify_ta_to_read);
- secure_display_ctxs[i].crtc = &adev->mode_info.crtcs[i]->base;
+ INIT_WORK(&crtc_ctx[i].forward_roi_work, amdgpu_dm_forward_crc_window);
+ INIT_WORK(&crtc_ctx[i].notify_ta_work, amdgpu_dm_crtc_notify_ta_to_read);
+ crtc_ctx[i].crtc = &adev->mode_info.crtcs[i]->base;
+ spin_lock_init(&crtc_ctx[i].crc_info.lock);
}
- return secure_display_ctxs;
+ adev->dm.secure_display_ctx.crtc_ctx = crtc_ctx;
+
+ adev->dm.secure_display_ctx.op_mode = DISPLAY_CRC_MODE;
}
#endif
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.h
index 748e80ef40d0..3da056c8d20b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.h
@@ -40,20 +40,53 @@ enum amdgpu_dm_pipe_crc_source {
};
#ifdef CONFIG_DRM_AMD_SECURE_DISPLAY
+#define MAX_CRTC 6
+
+enum secure_display_mode {
+ /* via dmub + psp */
+ LEGACY_MODE = 0,
+ /* driver directly */
+ DISPLAY_CRC_MODE,
+ SECURE_DISPLAY_MODE_MAX,
+};
+
+struct phy_id_mapping {
+ bool assigned;
+ bool is_mst;
+ uint8_t enc_hw_inst;
+ u8 lct;
+ u8 port_num;
+ u8 rad[8];
+};
+
+struct crc_data {
+ uint32_t crc_R;
+ uint32_t crc_G;
+ uint32_t crc_B;
+ uint32_t frame_count;
+ bool crc_ready;
+};
+
+struct crc_info {
+ struct crc_data crc[MAX_CRC_WINDOW_NUM];
+ struct completion completion;
+ spinlock_t lock;
+};
+
struct crc_window_param {
uint16_t x_start;
uint16_t y_start;
uint16_t x_end;
uint16_t y_end;
/* CRC window is activated or not*/
- bool activated;
+ bool enable;
/* Update crc window during vertical blank or not */
bool update_win;
/* skip reading/writing for few frames */
int skip_frame_cnt;
};
-struct secure_display_context {
+struct secure_display_crtc_context {
/* work to notify PSP TA*/
struct work_struct notify_ta_work;
@@ -63,7 +96,20 @@ struct secure_display_context {
struct drm_crtc *crtc;
/* Region of Interest (ROI) */
- struct rect rect;
+ struct crc_window roi[MAX_CRC_WINDOW_NUM];
+
+ struct crc_info crc_info;
+};
+
+struct secure_display_context {
+
+ struct secure_display_crtc_context *crtc_ctx;
+ /* Whether dmub support multiple ROI setting */
+ bool support_mul_roi;
+ enum secure_display_mode op_mode;
+ bool phy_mapping_updated;
+ int phy_id_mapping_cnt;
+ struct phy_id_mapping phy_id_mapping[MAX_CRTC];
};
#endif
@@ -95,8 +141,7 @@ void amdgpu_dm_crtc_handle_crc_irq(struct drm_crtc *crtc);
#ifdef CONFIG_DRM_AMD_SECURE_DISPLAY
bool amdgpu_dm_crc_window_is_activated(struct drm_crtc *crtc);
void amdgpu_dm_crtc_handle_crc_window_irq(struct drm_crtc *crtc);
-struct secure_display_context *amdgpu_dm_crtc_secure_display_create_contexts(
- struct amdgpu_device *adev);
+void amdgpu_dm_crtc_secure_display_create_contexts(struct amdgpu_device *adev);
#else
#define amdgpu_dm_crc_window_is_activated(x)
#define amdgpu_dm_crtc_handle_crc_window_irq(x)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c
index 64a041c2af05..010172f930ae 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.c
@@ -93,7 +93,7 @@ int amdgpu_dm_crtc_set_vupdate_irq(struct drm_crtc *crtc, bool enable)
return rc;
}
-bool amdgpu_dm_crtc_vrr_active(struct dm_crtc_state *dm_state)
+bool amdgpu_dm_crtc_vrr_active(const struct dm_crtc_state *dm_state)
{
return dm_state->freesync_config.state == VRR_STATE_ACTIVE_VARIABLE ||
dm_state->freesync_config.state == VRR_STATE_ACTIVE_FIXED;
@@ -113,6 +113,7 @@ bool amdgpu_dm_crtc_vrr_active(struct dm_crtc_state *dm_state)
*
* Panel Replay and PSR SU
* - Enable when:
+ * - VRR is disabled
* - vblank counter is disabled
* - entry is allowed: usermode demonstrates an adequate number of fast
* commits)
@@ -131,19 +132,20 @@ static void amdgpu_dm_crtc_set_panel_sr_feature(
bool is_sr_active = (link->replay_settings.replay_allow_active ||
link->psr_settings.psr_allow_active);
bool is_crc_window_active = false;
+ bool vrr_active = amdgpu_dm_crtc_vrr_active_irq(vblank_work->acrtc);
#ifdef CONFIG_DRM_AMD_SECURE_DISPLAY
is_crc_window_active =
amdgpu_dm_crc_window_is_activated(&vblank_work->acrtc->base);
#endif
- if (link->replay_settings.replay_feature_enabled &&
+ if (link->replay_settings.replay_feature_enabled && !vrr_active &&
allow_sr_entry && !is_sr_active && !is_crc_window_active) {
amdgpu_dm_replay_enable(vblank_work->stream, true);
} else if (vblank_enabled) {
if (link->psr_settings.psr_version < DC_PSR_VERSION_SU_1 && is_sr_active)
- amdgpu_dm_psr_disable(vblank_work->stream);
- } else if (link->psr_settings.psr_feature_enabled &&
+ amdgpu_dm_psr_disable(vblank_work->stream, false);
+ } else if (link->psr_settings.psr_feature_enabled && !vrr_active &&
allow_sr_entry && !is_sr_active && !is_crc_window_active) {
struct amdgpu_dm_connector *aconn =
@@ -659,6 +661,15 @@ static int amdgpu_dm_crtc_helper_atomic_check(struct drm_crtc *crtc,
return -EINVAL;
}
+ if (!state->legacy_cursor_update && amdgpu_dm_crtc_vrr_active(dm_crtc_state)) {
+ struct drm_plane_state *primary_state;
+
+ /* Pull in primary plane for correct VRR handling */
+ primary_state = drm_atomic_get_plane_state(state, crtc->primary);
+ if (IS_ERR(primary_state))
+ return PTR_ERR(primary_state);
+ }
+
/* In some use cases, like reset, no stream is attached */
if (!dm_crtc_state->stream)
return 0;
@@ -726,7 +737,16 @@ int amdgpu_dm_crtc_init(struct amdgpu_display_manager *dm,
* support programmable degamma anywhere.
*/
is_dcn = dm->adev->dm.dc->caps.color.dpp.dcn_arch;
- drm_crtc_enable_color_mgmt(&acrtc->base, is_dcn ? MAX_COLOR_LUT_ENTRIES : 0,
+ /* Dont't enable DRM CRTC degamma property for DCN401 since the
+ * pre-blending degamma LUT doesn't apply to cursor, and therefore
+ * can't work similar to a post-blending degamma LUT as in other hw
+ * versions.
+ * TODO: revisit it once KMS plane color API is merged.
+ */
+ drm_crtc_enable_color_mgmt(&acrtc->base,
+ (is_dcn &&
+ dm->adev->dm.dc->ctx->dce_version != DCN_VERSION_4_01) ?
+ MAX_COLOR_LUT_ENTRIES : 0,
true, MAX_COLOR_LUT_ENTRIES);
drm_mode_crtc_set_gamma_size(&acrtc->base, MAX_COLOR_LEGACY_LUT_ENTRIES);
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h
index 17e948753f59..c1212947a77b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crtc.h
@@ -37,7 +37,7 @@ int amdgpu_dm_crtc_set_vupdate_irq(struct drm_crtc *crtc, bool enable);
bool amdgpu_dm_crtc_vrr_active_irq(struct amdgpu_crtc *acrtc);
-bool amdgpu_dm_crtc_vrr_active(struct dm_crtc_state *dm_state);
+bool amdgpu_dm_crtc_vrr_active(const struct dm_crtc_state *dm_state);
int amdgpu_dm_crtc_enable_vblank(struct drm_crtc *crtc);
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
index 6a97bb2d9160..b726bcd18e29 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
@@ -25,6 +25,7 @@
#include <linux/string_helpers.h>
#include <linux/uaccess.h>
+#include <media/cec-notifier.h>
#include "dc.h"
#include "amdgpu.h"
@@ -258,7 +259,7 @@ static ssize_t dp_link_settings_write(struct file *f, const char __user *buf,
struct dc_link *link = connector->dc_link;
struct amdgpu_device *adev = drm_to_adev(connector->base.dev);
struct dc *dc = (struct dc *)link->dc;
- struct dc_link_settings prefer_link_settings;
+ struct dc_link_settings prefer_link_settings = {0};
char *wr_buf = NULL;
const uint32_t wr_buf_size = 40;
/* 0: lane_count; 1: link_rate */
@@ -389,7 +390,7 @@ static ssize_t dp_mst_link_setting(struct file *f, const char __user *buf,
struct dc_link *link = aconnector->dc_link;
struct amdgpu_device *adev = drm_to_adev(aconnector->base.dev);
struct dc *dc = (struct dc *)link->dc;
- struct dc_link_settings prefer_link_settings;
+ struct dc_link_settings prefer_link_settings = {0};
char *wr_buf = NULL;
const uint32_t wr_buf_size = 40;
/* 0: lane_count; 1: link_rate */
@@ -613,7 +614,7 @@ static ssize_t dp_phy_settings_write(struct file *f, const char __user *buf,
uint32_t wr_buf_size = 40;
long param[3];
bool use_prefer_link_setting;
- struct link_training_settings link_lane_settings;
+ struct link_training_settings link_lane_settings = {0};
int max_param_num = 3;
uint8_t param_nums = 0;
int r = 0;
@@ -768,7 +769,7 @@ static ssize_t dp_phy_test_pattern_debugfs_write(struct file *f, const char __us
LINK_RATE_UNKNOWN, LINK_SPREAD_DISABLED};
struct dc_link_settings cur_link_settings = {LANE_COUNT_UNKNOWN,
LINK_RATE_UNKNOWN, LINK_SPREAD_DISABLED};
- struct link_training_settings link_training_settings;
+ struct link_training_settings link_training_settings = {0};
int i;
if (size == 0)
@@ -902,9 +903,10 @@ static int dmub_tracebuffer_show(struct seq_file *m, void *data)
{
struct amdgpu_device *adev = m->private;
struct dmub_srv_fb_info *fb_info = adev->dm.dmub_fb_info;
+ struct dmub_fw_meta_info *fw_meta_info = NULL;
struct dmub_debugfs_trace_entry *entries;
uint8_t *tbuf_base;
- uint32_t tbuf_size, max_entries, num_entries, i;
+ uint32_t tbuf_size, max_entries, num_entries, first_entry, i;
if (!fb_info)
return 0;
@@ -913,20 +915,42 @@ static int dmub_tracebuffer_show(struct seq_file *m, void *data)
if (!tbuf_base)
return 0;
- tbuf_size = fb_info->fb[DMUB_WINDOW_5_TRACEBUFF].size;
+ if (adev->dm.dmub_srv)
+ fw_meta_info = &adev->dm.dmub_srv->meta_info;
+
+ tbuf_size = fw_meta_info ? fw_meta_info->trace_buffer_size :
+ DMUB_TRACE_BUFFER_SIZE;
max_entries = (tbuf_size - sizeof(struct dmub_debugfs_trace_header)) /
sizeof(struct dmub_debugfs_trace_entry);
num_entries =
((struct dmub_debugfs_trace_header *)tbuf_base)->entry_count;
+ /* DMCUB tracebuffer is a ring. If it rolled over, print a hint that
+ * entries are being overwritten.
+ */
+ if (num_entries > max_entries)
+ seq_printf(m, "...\n");
+
+ first_entry = num_entries % max_entries;
num_entries = min(num_entries, max_entries);
entries = (struct dmub_debugfs_trace_entry
*)(tbuf_base +
sizeof(struct dmub_debugfs_trace_header));
- for (i = 0; i < num_entries; ++i) {
+ /* To print entries chronologically, start from the first entry till the
+ * top of buffer, then from base of buffer to first entry.
+ */
+ for (i = first_entry; i < num_entries; ++i) {
+ struct dmub_debugfs_trace_entry *entry = &entries[i];
+
+ seq_printf(m,
+ "trace_code=%u tick_count=%u param0=%u param1=%u\n",
+ entry->trace_code, entry->tick_count, entry->param0,
+ entry->param1);
+ }
+ for (i = 0; i < first_entry; ++i) {
struct dmub_debugfs_trace_entry *entry = &entries[i];
seq_printf(m,
@@ -1145,7 +1169,7 @@ static int amdgpu_current_colorspace_show(struct seq_file *m, void *data)
case COLOR_SPACE_2020_RGB_FULLRANGE:
seq_puts(m, "BT2020_RGB");
break;
- case COLOR_SPACE_2020_YCBCR:
+ case COLOR_SPACE_2020_YCBCR_LIMITED:
seq_puts(m, "BT2020_YCC");
break;
default:
@@ -2825,6 +2849,67 @@ static int is_dpia_link_show(struct seq_file *m, void *data)
return 0;
}
+/**
+ * hdmi_cec_state_show - Read out the HDMI-CEC feature status
+ * @m: sequence file.
+ * @data: unused.
+ *
+ * Return 0 on success
+ */
+static int hdmi_cec_state_show(struct seq_file *m, void *data)
+{
+ struct drm_connector *connector = m->private;
+ struct amdgpu_dm_connector *aconnector = to_amdgpu_dm_connector(connector);
+
+ seq_printf(m, "%s:%d\n", connector->name, connector->base.id);
+ seq_printf(m, "HDMI-CEC status: %d\n", aconnector->notifier ? 1 : 0);
+
+ return 0;
+}
+
+/**
+ * hdmi_cec_state_write - Enable/Disable HDMI-CEC feature from driver side
+ * @f: file structure.
+ * @buf: userspace buffer. set to '1' to enable; '0' to disable cec feature.
+ * @size: size of buffer from userpsace.
+ * @pos: unused.
+ *
+ * Return size on success, error code on failure
+ */
+static ssize_t hdmi_cec_state_write(struct file *f, const char __user *buf,
+ size_t size, loff_t *pos)
+{
+ int ret;
+ bool enable;
+ struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private;
+ struct drm_device *ddev = aconnector->base.dev;
+
+ if (size == 0)
+ return -EINVAL;
+
+ ret = kstrtobool_from_user(buf, size, &enable);
+ if (ret) {
+ drm_dbg_driver(ddev, "invalid user data !\n");
+ return ret;
+ }
+
+ if (enable) {
+ if (aconnector->notifier)
+ return -EINVAL;
+ ret = amdgpu_dm_initialize_hdmi_connector(aconnector);
+ if (ret)
+ return ret;
+ hdmi_cec_set_edid(aconnector);
+ } else {
+ if (!aconnector->notifier)
+ return -EINVAL;
+ cec_notifier_conn_unregister(aconnector->notifier);
+ aconnector->notifier = NULL;
+ }
+
+ return size;
+}
+
DEFINE_SHOW_ATTRIBUTE(dp_dsc_fec_support);
DEFINE_SHOW_ATTRIBUTE(dmub_fw_state);
DEFINE_SHOW_ATTRIBUTE(dmub_tracebuffer);
@@ -2837,6 +2922,7 @@ DEFINE_SHOW_ATTRIBUTE(psr_capability);
DEFINE_SHOW_ATTRIBUTE(dp_is_mst_connector);
DEFINE_SHOW_ATTRIBUTE(dp_mst_progress_status);
DEFINE_SHOW_ATTRIBUTE(is_dpia_link);
+DEFINE_SHOW_STORE_ATTRIBUTE(hdmi_cec_state);
static const struct file_operations dp_dsc_clock_en_debugfs_fops = {
.owner = THIS_MODULE,
@@ -2972,7 +3058,8 @@ static const struct {
char *name;
const struct file_operations *fops;
} hdmi_debugfs_entries[] = {
- {"hdcp_sink_capability", &hdcp_sink_capability_fops}
+ {"hdcp_sink_capability", &hdcp_sink_capability_fops},
+ {"hdmi_cec_state", &hdmi_cec_state_fops}
};
/*
@@ -3457,8 +3544,8 @@ static int crc_win_x_start_set(void *data, u64 val)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
spin_lock_irq(&drm_dev->event_lock);
- acrtc->dm_irq_params.window_param.x_start = (uint16_t) val;
- acrtc->dm_irq_params.window_param.update_win = false;
+ acrtc->dm_irq_params.window_param[0].x_start = (uint16_t) val;
+ acrtc->dm_irq_params.window_param[0].update_win = false;
spin_unlock_irq(&drm_dev->event_lock);
return 0;
@@ -3474,7 +3561,7 @@ static int crc_win_x_start_get(void *data, u64 *val)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
spin_lock_irq(&drm_dev->event_lock);
- *val = acrtc->dm_irq_params.window_param.x_start;
+ *val = acrtc->dm_irq_params.window_param[0].x_start;
spin_unlock_irq(&drm_dev->event_lock);
return 0;
@@ -3494,8 +3581,8 @@ static int crc_win_y_start_set(void *data, u64 val)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
spin_lock_irq(&drm_dev->event_lock);
- acrtc->dm_irq_params.window_param.y_start = (uint16_t) val;
- acrtc->dm_irq_params.window_param.update_win = false;
+ acrtc->dm_irq_params.window_param[0].y_start = (uint16_t) val;
+ acrtc->dm_irq_params.window_param[0].update_win = false;
spin_unlock_irq(&drm_dev->event_lock);
return 0;
@@ -3511,7 +3598,7 @@ static int crc_win_y_start_get(void *data, u64 *val)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
spin_lock_irq(&drm_dev->event_lock);
- *val = acrtc->dm_irq_params.window_param.y_start;
+ *val = acrtc->dm_irq_params.window_param[0].y_start;
spin_unlock_irq(&drm_dev->event_lock);
return 0;
@@ -3530,8 +3617,8 @@ static int crc_win_x_end_set(void *data, u64 val)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
spin_lock_irq(&drm_dev->event_lock);
- acrtc->dm_irq_params.window_param.x_end = (uint16_t) val;
- acrtc->dm_irq_params.window_param.update_win = false;
+ acrtc->dm_irq_params.window_param[0].x_end = (uint16_t) val;
+ acrtc->dm_irq_params.window_param[0].update_win = false;
spin_unlock_irq(&drm_dev->event_lock);
return 0;
@@ -3547,7 +3634,7 @@ static int crc_win_x_end_get(void *data, u64 *val)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
spin_lock_irq(&drm_dev->event_lock);
- *val = acrtc->dm_irq_params.window_param.x_end;
+ *val = acrtc->dm_irq_params.window_param[0].x_end;
spin_unlock_irq(&drm_dev->event_lock);
return 0;
@@ -3566,8 +3653,8 @@ static int crc_win_y_end_set(void *data, u64 val)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
spin_lock_irq(&drm_dev->event_lock);
- acrtc->dm_irq_params.window_param.y_end = (uint16_t) val;
- acrtc->dm_irq_params.window_param.update_win = false;
+ acrtc->dm_irq_params.window_param[0].y_end = (uint16_t) val;
+ acrtc->dm_irq_params.window_param[0].update_win = false;
spin_unlock_irq(&drm_dev->event_lock);
return 0;
@@ -3583,7 +3670,7 @@ static int crc_win_y_end_get(void *data, u64 *val)
struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
spin_lock_irq(&drm_dev->event_lock);
- *val = acrtc->dm_irq_params.window_param.y_end;
+ *val = acrtc->dm_irq_params.window_param[0].y_end;
spin_unlock_irq(&drm_dev->event_lock);
return 0;
@@ -3606,13 +3693,14 @@ static int crc_win_update_set(void *data, u64 val)
/* PSR may write to OTG CRC window control register,
* so close it before starting secure_display.
*/
- amdgpu_dm_psr_disable(acrtc->dm_irq_params.stream);
+ amdgpu_dm_psr_disable(acrtc->dm_irq_params.stream, true);
spin_lock_irq(&adev_to_drm(adev)->event_lock);
- acrtc->dm_irq_params.window_param.activated = true;
- acrtc->dm_irq_params.window_param.update_win = true;
- acrtc->dm_irq_params.window_param.skip_frame_cnt = 0;
+ acrtc->dm_irq_params.window_param[0].enable = true;
+ acrtc->dm_irq_params.window_param[0].update_win = true;
+ acrtc->dm_irq_params.window_param[0].skip_frame_cnt = 0;
+ acrtc->dm_irq_params.crc_window_activated = true;
spin_unlock_irq(&adev_to_drm(adev)->event_lock);
mutex_unlock(&adev->dm.dc_lock);
@@ -3900,7 +3988,7 @@ static int capabilities_show(struct seq_file *m, void *unused)
struct hubbub *hubbub = dc->res_pool->hubbub;
- if (hubbub->funcs->get_mall_en)
+ if (hubbub && hubbub->funcs->get_mall_en)
hubbub->funcs->get_mall_en(hubbub, &mall_in_use);
if (dc->cap_funcs.get_subvp_en)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
index e339c7a8d541..b1d1897f5eaf 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_hdcp.c
@@ -26,6 +26,7 @@
#include "amdgpu_dm_hdcp.h"
#include "amdgpu.h"
#include "amdgpu_dm.h"
+#include "dc_fused_io.h"
#include "dm_helpers.h"
#include <drm/display/drm_hdcp_helper.h>
#include "hdcp_psp.h"
@@ -76,6 +77,34 @@ lp_read_dpcd(void *handle, uint32_t address, uint8_t *data, uint32_t size)
return dm_helpers_dp_read_dpcd(link->ctx, link, address, data, size);
}
+static bool lp_atomic_write_poll_read_i2c(
+ void *handle,
+ const struct mod_hdcp_atomic_op_i2c *write,
+ const struct mod_hdcp_atomic_op_i2c *poll,
+ struct mod_hdcp_atomic_op_i2c *read,
+ uint32_t poll_timeout_us,
+ uint8_t poll_mask_msb
+)
+{
+ struct dc_link *link = handle;
+
+ return dm_atomic_write_poll_read_i2c(link, write, poll, read, poll_timeout_us, poll_mask_msb);
+}
+
+static bool lp_atomic_write_poll_read_aux(
+ void *handle,
+ const struct mod_hdcp_atomic_op_aux *write,
+ const struct mod_hdcp_atomic_op_aux *poll,
+ struct mod_hdcp_atomic_op_aux *read,
+ uint32_t poll_timeout_us,
+ uint8_t poll_mask_msb
+)
+{
+ struct dc_link *link = handle;
+
+ return dm_atomic_write_poll_read_aux(link, write, poll, read, poll_timeout_us, poll_mask_msb);
+}
+
static uint8_t *psp_get_srm(struct psp_context *psp, uint32_t *srm_version, uint32_t *srm_size)
{
struct ta_hdcp_shared_memory *hdcp_cmd;
@@ -172,7 +201,10 @@ void hdcp_update_display(struct hdcp_workqueue *hdcp_work,
struct mod_hdcp_display_adjustment display_adjust;
unsigned int conn_index = aconnector->base.index;
- mutex_lock(&hdcp_w->mutex);
+ guard(mutex)(&hdcp_w->mutex);
+ drm_connector_get(&aconnector->base);
+ if (hdcp_w->aconnector[conn_index])
+ drm_connector_put(&hdcp_w->aconnector[conn_index]->base);
hdcp_w->aconnector[conn_index] = aconnector;
memset(&link_adjust, 0, sizeof(link_adjust));
@@ -209,7 +241,6 @@ void hdcp_update_display(struct hdcp_workqueue *hdcp_work,
mod_hdcp_update_display(&hdcp_w->hdcp, conn_index, &link_adjust, &display_adjust, &hdcp_w->output);
process_output(hdcp_w);
- mutex_unlock(&hdcp_w->mutex);
}
static void hdcp_remove_display(struct hdcp_workqueue *hdcp_work,
@@ -220,8 +251,7 @@ static void hdcp_remove_display(struct hdcp_workqueue *hdcp_work,
struct drm_connector_state *conn_state = aconnector->base.state;
unsigned int conn_index = aconnector->base.index;
- mutex_lock(&hdcp_w->mutex);
- hdcp_w->aconnector[conn_index] = aconnector;
+ guard(mutex)(&hdcp_w->mutex);
/* the removal of display will invoke auth reset -> hdcp destroy and
* we'd expect the Content Protection (CP) property changed back to
@@ -237,9 +267,11 @@ static void hdcp_remove_display(struct hdcp_workqueue *hdcp_work,
}
mod_hdcp_remove_display(&hdcp_w->hdcp, aconnector->base.index, &hdcp_w->output);
-
+ if (hdcp_w->aconnector[conn_index]) {
+ drm_connector_put(&hdcp_w->aconnector[conn_index]->base);
+ hdcp_w->aconnector[conn_index] = NULL;
+ }
process_output(hdcp_w);
- mutex_unlock(&hdcp_w->mutex);
}
void hdcp_reset_display(struct hdcp_workqueue *hdcp_work, unsigned int link_index)
@@ -247,7 +279,7 @@ void hdcp_reset_display(struct hdcp_workqueue *hdcp_work, unsigned int link_inde
struct hdcp_workqueue *hdcp_w = &hdcp_work[link_index];
unsigned int conn_index;
- mutex_lock(&hdcp_w->mutex);
+ guard(mutex)(&hdcp_w->mutex);
mod_hdcp_reset_connection(&hdcp_w->hdcp, &hdcp_w->output);
@@ -256,11 +288,13 @@ void hdcp_reset_display(struct hdcp_workqueue *hdcp_work, unsigned int link_inde
for (conn_index = 0; conn_index < AMDGPU_DM_MAX_DISPLAY_INDEX; conn_index++) {
hdcp_w->encryption_status[conn_index] =
MOD_HDCP_ENCRYPTION_STATUS_HDCP_OFF;
+ if (hdcp_w->aconnector[conn_index]) {
+ drm_connector_put(&hdcp_w->aconnector[conn_index]->base);
+ hdcp_w->aconnector[conn_index] = NULL;
+ }
}
process_output(hdcp_w);
-
- mutex_unlock(&hdcp_w->mutex);
}
void hdcp_handle_cpirq(struct hdcp_workqueue *hdcp_work, unsigned int link_index)
@@ -277,7 +311,7 @@ static void event_callback(struct work_struct *work)
hdcp_work = container_of(to_delayed_work(work), struct hdcp_workqueue,
callback_dwork);
- mutex_lock(&hdcp_work->mutex);
+ guard(mutex)(&hdcp_work->mutex);
cancel_delayed_work(&hdcp_work->callback_dwork);
@@ -285,8 +319,6 @@ static void event_callback(struct work_struct *work)
&hdcp_work->output);
process_output(hdcp_work);
-
- mutex_unlock(&hdcp_work->mutex);
}
static void event_property_update(struct work_struct *work)
@@ -323,7 +355,7 @@ static void event_property_update(struct work_struct *work)
continue;
drm_modeset_lock(&dev->mode_config.connection_mutex, NULL);
- mutex_lock(&hdcp_work->mutex);
+ guard(mutex)(&hdcp_work->mutex);
if (conn_state->commit) {
ret = wait_for_completion_interruptible_timeout(&conn_state->commit->hw_done,
@@ -355,7 +387,6 @@ static void event_property_update(struct work_struct *work)
drm_hdcp_update_content_protection(connector,
DRM_MODE_CONTENT_PROTECTION_DESIRED);
}
- mutex_unlock(&hdcp_work->mutex);
drm_modeset_unlock(&dev->mode_config.connection_mutex);
}
}
@@ -368,7 +399,7 @@ static void event_property_validate(struct work_struct *work)
struct amdgpu_dm_connector *aconnector;
unsigned int conn_index;
- mutex_lock(&hdcp_work->mutex);
+ guard(mutex)(&hdcp_work->mutex);
for (conn_index = 0; conn_index < AMDGPU_DM_MAX_DISPLAY_INDEX;
conn_index++) {
@@ -408,8 +439,6 @@ static void event_property_validate(struct work_struct *work)
schedule_work(&hdcp_work->property_update_work);
}
}
-
- mutex_unlock(&hdcp_work->mutex);
}
static void event_watchdog_timer(struct work_struct *work)
@@ -420,7 +449,7 @@ static void event_watchdog_timer(struct work_struct *work)
struct hdcp_workqueue,
watchdog_timer_dwork);
- mutex_lock(&hdcp_work->mutex);
+ guard(mutex)(&hdcp_work->mutex);
cancel_delayed_work(&hdcp_work->watchdog_timer_dwork);
@@ -429,8 +458,6 @@ static void event_watchdog_timer(struct work_struct *work)
&hdcp_work->output);
process_output(hdcp_work);
-
- mutex_unlock(&hdcp_work->mutex);
}
static void event_cpirq(struct work_struct *work)
@@ -439,13 +466,11 @@ static void event_cpirq(struct work_struct *work)
hdcp_work = container_of(work, struct hdcp_workqueue, cpirq_work);
- mutex_lock(&hdcp_work->mutex);
+ guard(mutex)(&hdcp_work->mutex);
mod_hdcp_process_event(&hdcp_work->hdcp, MOD_HDCP_EVENT_CPIRQ, &hdcp_work->output);
process_output(hdcp_work);
-
- mutex_unlock(&hdcp_work->mutex);
}
void hdcp_destroy(struct kobject *kobj, struct hdcp_workqueue *hdcp_work)
@@ -455,6 +480,7 @@ void hdcp_destroy(struct kobject *kobj, struct hdcp_workqueue *hdcp_work)
for (i = 0; i < hdcp_work->max_link; i++) {
cancel_delayed_work_sync(&hdcp_work[i].callback_dwork);
cancel_delayed_work_sync(&hdcp_work[i].watchdog_timer_dwork);
+ cancel_delayed_work_sync(&hdcp_work[i].property_validate_dwork);
}
sysfs_remove_bin_file(kobj, &hdcp_work[0].attr);
@@ -469,7 +495,6 @@ static bool enable_assr(void *handle, struct dc_link *link)
struct mod_hdcp hdcp = hdcp_work->hdcp;
struct psp_context *psp = hdcp.config.psp.handle;
struct ta_dtm_shared_memory *dtm_cmd;
- bool res = true;
if (!psp->dtm_context.context.initialized) {
DRM_INFO("Failed to enable ASSR, DTM TA is not initialized.");
@@ -478,7 +503,7 @@ static bool enable_assr(void *handle, struct dc_link *link)
dtm_cmd = (struct ta_dtm_shared_memory *)psp->dtm_context.context.mem_context.shared_buf;
- mutex_lock(&psp->dtm_context.mutex);
+ guard(mutex)(&psp->dtm_context.mutex);
memset(dtm_cmd, 0, sizeof(struct ta_dtm_shared_memory));
dtm_cmd->cmd_id = TA_DTM_COMMAND__TOPOLOGY_ASSR_ENABLE;
@@ -490,12 +515,10 @@ static bool enable_assr(void *handle, struct dc_link *link)
if (dtm_cmd->dtm_status != TA_DTM_STATUS__SUCCESS) {
DRM_INFO("Failed to enable ASSR");
- res = false;
+ return false;
}
- mutex_unlock(&psp->dtm_context.mutex);
-
- return res;
+ return true;
}
static void update_config(void *handle, struct cp_psp_stream_config *config)
@@ -503,6 +526,7 @@ static void update_config(void *handle, struct cp_psp_stream_config *config)
struct hdcp_workqueue *hdcp_work = handle;
struct amdgpu_dm_connector *aconnector = config->dm_stream_ctx;
int link_index = aconnector->dc_link->link_index;
+ unsigned int conn_index = aconnector->base.index;
struct mod_hdcp_display *display = &hdcp_work[link_index].display;
struct mod_hdcp_link *link = &hdcp_work[link_index].link;
struct hdcp_workqueue *hdcp_w = &hdcp_work[link_index];
@@ -556,13 +580,14 @@ static void update_config(void *handle, struct cp_psp_stream_config *config)
(!!aconnector->base.state) ?
aconnector->base.state->hdcp_content_type : -1);
- mutex_lock(&hdcp_w->mutex);
+ guard(mutex)(&hdcp_w->mutex);
mod_hdcp_add_display(&hdcp_w->hdcp, link, display, &hdcp_w->output);
-
+ drm_connector_get(&aconnector->base);
+ if (hdcp_w->aconnector[conn_index])
+ drm_connector_put(&hdcp_w->aconnector[conn_index]->base);
+ hdcp_w->aconnector[conn_index] = aconnector;
process_output(hdcp_w);
- mutex_unlock(&hdcp_w->mutex);
-
}
/**
@@ -614,7 +639,7 @@ static void update_config(void *handle, struct cp_psp_stream_config *config)
* incorrect/corrupted and we should correct our SRM by getting it from PSP
*/
static ssize_t srm_data_write(struct file *filp, struct kobject *kobj,
- struct bin_attribute *bin_attr, char *buffer,
+ const struct bin_attribute *bin_attr, char *buffer,
loff_t pos, size_t count)
{
struct hdcp_workqueue *work;
@@ -638,7 +663,7 @@ static ssize_t srm_data_write(struct file *filp, struct kobject *kobj,
}
static ssize_t srm_data_read(struct file *filp, struct kobject *kobj,
- struct bin_attribute *bin_attr, char *buffer,
+ const struct bin_attribute *bin_attr, char *buffer,
loff_t pos, size_t count)
{
struct hdcp_workqueue *work;
@@ -736,19 +761,33 @@ struct hdcp_workqueue *hdcp_create_workqueue(struct amdgpu_device *adev,
INIT_DELAYED_WORK(&hdcp_work[i].watchdog_timer_dwork, event_watchdog_timer);
INIT_DELAYED_WORK(&hdcp_work[i].property_validate_dwork, event_property_validate);
- hdcp_work[i].hdcp.config.psp.handle = &adev->psp;
+ struct mod_hdcp_config *config = &hdcp_work[i].hdcp.config;
+ struct mod_hdcp_ddc_funcs *ddc_funcs = &config->ddc.funcs;
+
+ config->psp.handle = &adev->psp;
if (dc->ctx->dce_version == DCN_VERSION_3_1 ||
dc->ctx->dce_version == DCN_VERSION_3_14 ||
dc->ctx->dce_version == DCN_VERSION_3_15 ||
dc->ctx->dce_version == DCN_VERSION_3_5 ||
dc->ctx->dce_version == DCN_VERSION_3_51 ||
+ dc->ctx->dce_version == DCN_VERSION_3_6 ||
dc->ctx->dce_version == DCN_VERSION_3_16)
- hdcp_work[i].hdcp.config.psp.caps.dtm_v3_supported = 1;
- hdcp_work[i].hdcp.config.ddc.handle = dc_get_link_at_index(dc, i);
- hdcp_work[i].hdcp.config.ddc.funcs.write_i2c = lp_write_i2c;
- hdcp_work[i].hdcp.config.ddc.funcs.read_i2c = lp_read_i2c;
- hdcp_work[i].hdcp.config.ddc.funcs.write_dpcd = lp_write_dpcd;
- hdcp_work[i].hdcp.config.ddc.funcs.read_dpcd = lp_read_dpcd;
+ config->psp.caps.dtm_v3_supported = 1;
+ config->ddc.handle = dc_get_link_at_index(dc, i);
+
+ ddc_funcs->write_i2c = lp_write_i2c;
+ ddc_funcs->read_i2c = lp_read_i2c;
+ ddc_funcs->write_dpcd = lp_write_dpcd;
+ ddc_funcs->read_dpcd = lp_read_dpcd;
+
+ config->debug.lc_enable_sw_fallback = dc->debug.hdcp_lc_enable_sw_fallback;
+ if (dc->caps.fused_io_supported || dc->debug.hdcp_lc_force_fw_enable) {
+ ddc_funcs->atomic_write_poll_read_i2c = lp_atomic_write_poll_read_i2c;
+ ddc_funcs->atomic_write_poll_read_aux = lp_atomic_write_poll_read_aux;
+ } else {
+ ddc_funcs->atomic_write_poll_read_i2c = NULL;
+ ddc_funcs->atomic_write_poll_read_aux = NULL;
+ }
memset(hdcp_work[i].aconnector, 0,
sizeof(struct amdgpu_dm_connector *) *
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
index b0fea0856866..9e3e51a2dc49 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
@@ -55,16 +55,21 @@ static u32 edid_extract_panel_id(struct edid *edid)
(u32)EDID_PRODUCT_ID(edid);
}
-static void apply_edid_quirks(struct edid *edid, struct dc_edid_caps *edid_caps)
+static void apply_edid_quirks(struct drm_device *dev, struct edid *edid, struct dc_edid_caps *edid_caps)
{
uint32_t panel_id = edid_extract_panel_id(edid);
switch (panel_id) {
+ /* Workaround for monitors that need a delay after detecting the link */
+ case drm_edid_encode_panel_id('G', 'B', 'T', 0x3215):
+ drm_dbg_driver(dev, "Add 10s delay for link detection for panel id %X\n", panel_id);
+ edid_caps->panel_patch.wait_after_dpcd_poweroff_ms = 10000;
+ break;
/* Workaround for some monitors which does not work well with FAMS */
case drm_edid_encode_panel_id('S', 'A', 'M', 0x0E5E):
case drm_edid_encode_panel_id('S', 'A', 'M', 0x7053):
case drm_edid_encode_panel_id('S', 'A', 'M', 0x71AC):
- DRM_DEBUG_DRIVER("Disabling FAMS on monitor with panel id %X\n", panel_id);
+ drm_dbg_driver(dev, "Disabling FAMS on monitor with panel id %X\n", panel_id);
edid_caps->panel_patch.disable_fams = true;
break;
/* Workaround for some monitors that do not clear DPCD 0x317 if FreeSync is unsupported */
@@ -73,11 +78,11 @@ static void apply_edid_quirks(struct edid *edid, struct dc_edid_caps *edid_caps)
case drm_edid_encode_panel_id('B', 'O', 'E', 0x092A):
case drm_edid_encode_panel_id('L', 'G', 'D', 0x06D1):
case drm_edid_encode_panel_id('M', 'S', 'F', 0x1003):
- DRM_DEBUG_DRIVER("Clearing DPCD 0x317 on monitor with panel id %X\n", panel_id);
+ drm_dbg_driver(dev, "Clearing DPCD 0x317 on monitor with panel id %X\n", panel_id);
edid_caps->panel_patch.remove_sink_ext_caps = true;
break;
case drm_edid_encode_panel_id('S', 'D', 'C', 0x4154):
- DRM_DEBUG_DRIVER("Disabling VSC on monitor with panel id %X\n", panel_id);
+ drm_dbg_driver(dev, "Disabling VSC on monitor with panel id %X\n", panel_id);
edid_caps->panel_patch.disable_colorimetry = true;
break;
default:
@@ -101,6 +106,7 @@ enum dc_edid_status dm_helpers_parse_edid_caps(
{
struct amdgpu_dm_connector *aconnector = link->priv;
struct drm_connector *connector = &aconnector->base;
+ struct drm_device *dev = connector->dev;
struct edid *edid_buf = edid ? (struct edid *) edid->raw_edid : NULL;
struct cea_sad *sads;
int sad_count = -1;
@@ -130,7 +136,7 @@ enum dc_edid_status dm_helpers_parse_edid_caps(
edid_caps->edid_hdmi = connector->display_info.is_hdmi;
- apply_edid_quirks(edid_buf, edid_caps);
+ apply_edid_quirks(dev, edid_buf, edid_caps);
sad_count = drm_edid_to_sad((struct edid *) edid->raw_edid, &sads);
if (sad_count <= 0)
@@ -624,6 +630,19 @@ bool dm_helpers_submit_i2c(
return result;
}
+bool dm_helpers_execute_fused_io(
+ struct dc_context *ctx,
+ struct dc_link *link,
+ union dmub_rb_cmd *commands,
+ uint8_t count,
+ uint32_t timeout_us
+)
+{
+ struct amdgpu_device *dev = ctx->driver_context;
+
+ return amdgpu_dm_execute_fused_io(dev, link, commands, count, timeout_us);
+}
+
static bool execute_synaptics_rc_command(struct drm_dp_aux *aux,
bool is_write_cmd,
unsigned char cmd,
@@ -885,6 +904,12 @@ bool dm_helpers_dp_write_dsc_enable(
return ret;
}
+bool dm_helpers_dp_write_hblank_reduction(struct dc_context *ctx, const struct dc_stream_state *stream)
+{
+ // TODO
+ return false;
+}
+
bool dm_helpers_is_dp_sink_present(struct dc_link *link)
{
bool dp_sink_present;
@@ -906,15 +931,15 @@ dm_helpers_probe_acpi_edid(void *data, u8 *buf, unsigned int block, size_t len)
{
struct drm_connector *connector = data;
struct acpi_device *acpidev = ACPI_COMPANION(connector->dev->dev);
- unsigned char start = block * EDID_LENGTH;
- void *edid;
+ unsigned short start = block * EDID_LENGTH;
+ struct edid *edid;
int r;
if (!acpidev)
return -ENODEV;
/* fetch the entire edid from BIOS */
- r = acpi_video_get_edid(acpidev, ACPI_VIDEO_DISPLAY_LCD, -1, &edid);
+ r = acpi_video_get_edid(acpidev, ACPI_VIDEO_DISPLAY_LCD, -1, (void *)&edid);
if (r < 0) {
drm_dbg(connector->dev, "Failed to get EDID from ACPI: %d\n", r);
return r;
@@ -924,7 +949,14 @@ dm_helpers_probe_acpi_edid(void *data, u8 *buf, unsigned int block, size_t len)
goto cleanup;
}
- memcpy(buf, edid + start, len);
+ /* sanity check */
+ if (edid->revision < 4 || !(edid->input & DRM_EDID_INPUT_DIGITAL) ||
+ (edid->input & DRM_EDID_DIGITAL_TYPE_MASK) == DRM_EDID_DIGITAL_TYPE_UNDEF) {
+ r = -EINVAL;
+ goto cleanup;
+ }
+
+ memcpy(buf, (void *)edid + start, len);
r = 0;
cleanup:
@@ -997,6 +1029,10 @@ enum dc_edid_status dm_helpers_read_local_edid(
return EDID_NO_RESPONSE;
edid = drm_edid_raw(drm_edid); // FIXME: Get rid of drm_edid_raw()
+ if (!edid ||
+ edid->extensions >= sizeof(sink->dc_edid.raw_edid) / EDID_LENGTH)
+ return EDID_BAD_INPUT;
+
sink->dc_edid.length = EDID_LENGTH * (edid->extensions + 1);
memmove(sink->dc_edid.raw_edid, (uint8_t *)edid, sink->dc_edid.length);
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
index 3390f0d8420a..b61e210f6246 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.c
@@ -473,7 +473,7 @@ void amdgpu_dm_irq_fini(struct amdgpu_device *adev)
unregister_all_irq_handlers(adev);
}
-int amdgpu_dm_irq_suspend(struct amdgpu_device *adev)
+void amdgpu_dm_irq_suspend(struct amdgpu_device *adev)
{
int src;
struct list_head *hnd_list_h;
@@ -511,10 +511,9 @@ int amdgpu_dm_irq_suspend(struct amdgpu_device *adev)
}
DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags);
- return 0;
}
-int amdgpu_dm_irq_resume_early(struct amdgpu_device *adev)
+void amdgpu_dm_irq_resume_early(struct amdgpu_device *adev)
{
int src;
struct list_head *hnd_list_h, *hnd_list_l;
@@ -522,7 +521,7 @@ int amdgpu_dm_irq_resume_early(struct amdgpu_device *adev)
DM_IRQ_TABLE_LOCK(adev, irq_table_flags);
- DRM_DEBUG_KMS("DM_IRQ: early resume\n");
+ drm_dbg(adev_to_drm(adev), "DM_IRQ: early resume\n");
/* re-enable short pulse interrupts HW interrupt */
for (src = DC_IRQ_SOURCE_HPD1RX; src <= DC_IRQ_SOURCE_HPD6RX; src++) {
@@ -533,11 +532,9 @@ int amdgpu_dm_irq_resume_early(struct amdgpu_device *adev)
}
DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags);
-
- return 0;
}
-int amdgpu_dm_irq_resume_late(struct amdgpu_device *adev)
+void amdgpu_dm_irq_resume_late(struct amdgpu_device *adev)
{
int src;
struct list_head *hnd_list_h, *hnd_list_l;
@@ -545,7 +542,7 @@ int amdgpu_dm_irq_resume_late(struct amdgpu_device *adev)
DM_IRQ_TABLE_LOCK(adev, irq_table_flags);
- DRM_DEBUG_KMS("DM_IRQ: resume\n");
+ drm_dbg(adev_to_drm(adev), "DM_IRQ: resume\n");
/**
* Renable HW interrupt for HPD and only since FLIP and VBLANK
@@ -559,7 +556,6 @@ int amdgpu_dm_irq_resume_late(struct amdgpu_device *adev)
}
DM_IRQ_TABLE_UNLOCK(adev, irq_table_flags);
- return 0;
}
/*
@@ -894,6 +890,15 @@ void amdgpu_dm_hpd_init(struct amdgpu_device *adev)
struct drm_device *dev = adev_to_drm(adev);
struct drm_connector *connector;
struct drm_connector_list_iter iter;
+ int irq_type;
+ int i;
+
+ /* First, clear all hpd and hpdrx interrupts */
+ for (i = DC_IRQ_SOURCE_HPD1; i <= DC_IRQ_SOURCE_HPD6RX; i++) {
+ if (!dc_interrupt_set(adev->dm.dc, i, false))
+ drm_err(dev, "Failed to clear hpd(rx) source=%d on init\n",
+ i);
+ }
drm_connector_list_iter_begin(dev, &iter);
drm_for_each_connector_iter(connector, &iter) {
@@ -907,10 +912,31 @@ void amdgpu_dm_hpd_init(struct amdgpu_device *adev)
dc_link = amdgpu_dm_connector->dc_link;
+ /*
+ * Get a base driver irq reference for hpd ints for the lifetime
+ * of dm. Note that only hpd interrupt types are registered with
+ * base driver; hpd_rx types aren't. IOW, amdgpu_irq_get/put on
+ * hpd_rx isn't available. DM currently controls hpd_rx
+ * explicitly with dc_interrupt_set()
+ */
if (dc_link->irq_source_hpd != DC_IRQ_SOURCE_INVALID) {
- dc_interrupt_set(adev->dm.dc,
- dc_link->irq_source_hpd,
- true);
+ irq_type = dc_link->irq_source_hpd - DC_IRQ_SOURCE_HPD1;
+ /*
+ * TODO: There's a mismatch between mode_info.num_hpd
+ * and what bios reports as the # of connectors with hpd
+ * sources. Since the # of hpd source types registered
+ * with base driver == mode_info.num_hpd, we have to
+ * fallback to dc_interrupt_set for the remaining types.
+ */
+ if (irq_type < adev->mode_info.num_hpd) {
+ if (amdgpu_irq_get(adev, &adev->hpd_irq, irq_type))
+ drm_err(dev, "DM_IRQ: Failed get HPD for source=%d)!\n",
+ dc_link->irq_source_hpd);
+ } else {
+ dc_interrupt_set(adev->dm.dc,
+ dc_link->irq_source_hpd,
+ true);
+ }
}
if (dc_link->irq_source_hpd_rx != DC_IRQ_SOURCE_INVALID) {
@@ -935,6 +961,7 @@ void amdgpu_dm_hpd_fini(struct amdgpu_device *adev)
struct drm_device *dev = adev_to_drm(adev);
struct drm_connector *connector;
struct drm_connector_list_iter iter;
+ int irq_type;
drm_connector_list_iter_begin(dev, &iter);
drm_for_each_connector_iter(connector, &iter) {
@@ -948,9 +975,18 @@ void amdgpu_dm_hpd_fini(struct amdgpu_device *adev)
dc_link = amdgpu_dm_connector->dc_link;
if (dc_link->irq_source_hpd != DC_IRQ_SOURCE_INVALID) {
- dc_interrupt_set(adev->dm.dc,
- dc_link->irq_source_hpd,
- false);
+ irq_type = dc_link->irq_source_hpd - DC_IRQ_SOURCE_HPD1;
+
+ /* TODO: See same TODO in amdgpu_dm_hpd_init() */
+ if (irq_type < adev->mode_info.num_hpd) {
+ if (amdgpu_irq_put(adev, &adev->hpd_irq, irq_type))
+ drm_err(dev, "DM_IRQ: Failed put HPD for source=%d!\n",
+ dc_link->irq_source_hpd);
+ } else {
+ dc_interrupt_set(adev->dm.dc,
+ dc_link->irq_source_hpd,
+ false);
+ }
}
if (dc_link->irq_source_hpd_rx != DC_IRQ_SOURCE_INVALID) {
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h
index 2349238a626b..ba17c23b2706 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq.h
@@ -90,14 +90,14 @@ void amdgpu_dm_hpd_fini(struct amdgpu_device *adev);
* amdgpu_dm_irq_suspend - disable ASIC interrupt during suspend.
*
*/
-int amdgpu_dm_irq_suspend(struct amdgpu_device *adev);
+void amdgpu_dm_irq_suspend(struct amdgpu_device *adev);
/**
* amdgpu_dm_irq_resume_early - enable HPDRX ASIC interrupts during resume.
* amdgpu_dm_irq_resume - enable ASIC interrupt during resume.
*
*/
-int amdgpu_dm_irq_resume_early(struct amdgpu_device *adev);
-int amdgpu_dm_irq_resume_late(struct amdgpu_device *adev);
+void amdgpu_dm_irq_resume_early(struct amdgpu_device *adev);
+void amdgpu_dm_irq_resume_late(struct amdgpu_device *adev);
#endif /* __AMDGPU_DM_IRQ_H__ */
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq_params.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq_params.h
index 6a7ecc1e4602..6c9de834455b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq_params.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_irq_params.h
@@ -39,7 +39,9 @@ struct dm_irq_params {
#ifdef CONFIG_DEBUG_FS
enum amdgpu_dm_pipe_crc_source crc_src;
#ifdef CONFIG_DRM_AMD_SECURE_DISPLAY
- struct crc_window_param window_param;
+ struct crc_window_param window_param[MAX_CRC_WINDOW_NUM];
+ /* At least one CRC window is activated or not*/
+ bool crc_window_activated;
#endif
#endif
};
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 6e4359490613..7187d5aedf0a 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -51,6 +51,9 @@
#define PEAK_FACTOR_X1000 1006
+/*
+ * This function handles both native AUX and I2C-Over-AUX transactions.
+ */
static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
struct drm_dp_aux_msg *msg)
{
@@ -59,6 +62,7 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
enum aux_return_code_type operation_result;
struct amdgpu_device *adev;
struct ddc_service *ddc;
+ uint8_t copy[16];
if (WARN_ON(msg->size > 16))
return -E2BIG;
@@ -74,6 +78,11 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
(msg->request & DP_AUX_I2C_WRITE_STATUS_UPDATE) != 0;
payload.defer_delay = 0;
+ if (payload.write) {
+ memcpy(copy, msg->buffer, msg->size);
+ payload.data = copy;
+ }
+
result = dc_link_aux_transfer_raw(TO_DM_AUX(aux)->ddc_service, &payload,
&operation_result);
@@ -87,15 +96,25 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
if (adev->dm.aux_hpd_discon_quirk) {
if (msg->address == DP_SIDEBAND_MSG_DOWN_REQ_BASE &&
operation_result == AUX_RET_ERROR_HPD_DISCON) {
- result = 0;
+ result = msg->size;
operation_result = AUX_RET_SUCCESS;
}
}
- if (payload.write && result >= 0)
- result = msg->size;
+ /*
+ * result equals to 0 includes the cases of AUX_DEFER/I2C_DEFER
+ */
+ if (payload.write && result >= 0) {
+ if (result) {
+ /*one byte indicating partially written bytes*/
+ drm_dbg_dp(adev_to_drm(adev), "AUX partially written\n");
+ result = payload.data[0];
+ } else if (!payload.reply[0])
+ /*I2C_ACK|AUX_ACK*/
+ result = msg->size;
+ }
- if (result < 0)
+ if (result < 0) {
switch (operation_result) {
case AUX_RET_SUCCESS:
break;
@@ -114,6 +133,13 @@ static ssize_t dm_dp_aux_transfer(struct drm_dp_aux *aux,
break;
}
+ drm_dbg_dp(adev_to_drm(adev), "DP AUX transfer fail:%d\n", operation_result);
+ }
+
+ if (payload.reply[0])
+ drm_dbg_dp(adev_to_drm(adev), "AUX reply command not ACK: 0x%02x.",
+ payload.reply[0]);
+
return result;
}
@@ -155,6 +181,17 @@ amdgpu_dm_mst_connector_late_register(struct drm_connector *connector)
return 0;
}
+
+static inline void
+amdgpu_dm_mst_reset_mst_connector_setting(struct amdgpu_dm_connector *aconnector)
+{
+ aconnector->drm_edid = NULL;
+ aconnector->dsc_aux = NULL;
+ aconnector->mst_output_port->passthrough_aux = NULL;
+ aconnector->mst_local_bw = 0;
+ aconnector->vc_full_pbn = 0;
+}
+
static void
amdgpu_dm_mst_connector_early_unregister(struct drm_connector *connector)
{
@@ -182,9 +219,7 @@ amdgpu_dm_mst_connector_early_unregister(struct drm_connector *connector)
dc_sink_release(dc_sink);
aconnector->dc_sink = NULL;
- aconnector->drm_edid = NULL;
- aconnector->dsc_aux = NULL;
- port->passthrough_aux = NULL;
+ amdgpu_dm_mst_reset_mst_connector_setting(aconnector);
}
aconnector->mst_status = MST_STATUS_DEFAULT;
@@ -504,9 +539,7 @@ dm_dp_mst_detect(struct drm_connector *connector,
dc_sink_release(aconnector->dc_sink);
aconnector->dc_sink = NULL;
- aconnector->drm_edid = NULL;
- aconnector->dsc_aux = NULL;
- port->passthrough_aux = NULL;
+ amdgpu_dm_mst_reset_mst_connector_setting(aconnector);
amdgpu_dm_set_mst_status(&aconnector->mst_status,
MST_REMOTE_EDID | MST_ALLOCATE_NEW_PAYLOAD | MST_CLEAR_ALLOCATED_PAYLOAD,
@@ -590,11 +623,12 @@ dm_dp_add_mst_connector(struct drm_dp_mst_topology_mgr *mgr,
amdgpu_dm_set_mst_status(&aconnector->mst_status,
MST_PROBE, true);
- if (drm_connector_init(
+ if (drm_connector_dynamic_init(
dev,
connector,
&dm_dp_mst_connector_funcs,
- DRM_MODE_CONNECTOR_DisplayPort)) {
+ DRM_MODE_CONNECTOR_DisplayPort,
+ NULL)) {
kfree(aconnector);
return NULL;
}
@@ -1617,7 +1651,6 @@ int pre_validate_dsc(struct drm_atomic_state *state,
if (ind >= 0) {
struct drm_connector *connector;
- struct amdgpu_dm_connector *aconnector;
struct drm_connector_state *drm_new_conn_state;
struct dm_connector_state *dm_new_conn_state;
struct dm_crtc_state *dm_old_crtc_state;
@@ -1625,15 +1658,17 @@ int pre_validate_dsc(struct drm_atomic_state *state,
connector =
amdgpu_dm_find_first_crtc_matching_connector(state,
state->crtcs[ind].ptr);
- aconnector = to_amdgpu_dm_connector(connector);
+ if (!connector)
+ continue;
+
drm_new_conn_state =
drm_atomic_get_new_connector_state(state,
- &aconnector->base);
+ connector);
dm_new_conn_state = to_dm_connector_state(drm_new_conn_state);
dm_old_crtc_state = to_dm_crtc_state(state->crtcs[ind].old_state);
local_dc_state->streams[i] =
- create_validate_stream_for_sink(aconnector,
+ create_validate_stream_for_sink(connector,
&state->crtcs[ind].new_state->mode,
dm_new_conn_state,
dm_old_crtc_state->stream);
@@ -1688,32 +1723,33 @@ clean_exit:
return ret;
}
-static unsigned int kbps_from_pbn(unsigned int pbn)
+static uint32_t kbps_from_pbn(unsigned int pbn)
{
- unsigned int kbps = pbn;
+ uint64_t kbps = (uint64_t)pbn;
kbps *= (1000000 / PEAK_FACTOR_X1000);
kbps *= 8;
kbps *= 54;
kbps /= 64;
- return kbps;
+ return (uint32_t)kbps;
}
static bool is_dsc_common_config_possible(struct dc_stream_state *stream,
struct dc_dsc_bw_range *bw_range)
{
struct dc_dsc_policy dsc_policy = {0};
+ bool is_dsc_possible;
dc_dsc_get_policy_for_timing(&stream->timing, 0, &dsc_policy, dc_link_get_highest_encoding_format(stream->link));
- dc_dsc_compute_bandwidth_range(stream->sink->ctx->dc->res_pool->dscs[0],
- stream->sink->ctx->dc->debug.dsc_min_slice_height_override,
- dsc_policy.min_target_bpp * 16,
- dsc_policy.max_target_bpp * 16,
- &stream->sink->dsc_caps.dsc_dec_caps,
- &stream->timing, dc_link_get_highest_encoding_format(stream->link), bw_range);
-
- return bw_range->max_target_bpp_x16 && bw_range->min_target_bpp_x16;
+ is_dsc_possible = dc_dsc_compute_bandwidth_range(stream->sink->ctx->dc->res_pool->dscs[0],
+ stream->sink->ctx->dc->debug.dsc_min_slice_height_override,
+ dsc_policy.min_target_bpp * 16,
+ dsc_policy.max_target_bpp * 16,
+ &stream->sink->dsc_caps.dsc_dec_caps,
+ &stream->timing, dc_link_get_highest_encoding_format(stream->link), bw_range);
+
+ return is_dsc_possible;
}
#endif
@@ -1819,9 +1855,18 @@ enum dc_status dm_dp_mst_is_port_support_mode(
struct drm_dp_mst_port *immediate_upstream_port = NULL;
uint32_t end_link_bw = 0;
- /*Get last DP link BW capability*/
- if (dp_get_link_current_set_bw(&aconnector->mst_output_port->aux, &end_link_bw)) {
- if (stream_kbps > end_link_bw) {
+ /*Get last DP link BW capability. Mode shall be supported by Legacy peer*/
+ if (aconnector->mst_output_port->pdt != DP_PEER_DEVICE_DP_LEGACY_CONV &&
+ aconnector->mst_output_port->pdt != DP_PEER_DEVICE_NONE) {
+ if (aconnector->vc_full_pbn != aconnector->mst_output_port->full_pbn) {
+ dp_get_link_current_set_bw(&aconnector->mst_output_port->aux, &end_link_bw);
+ aconnector->vc_full_pbn = aconnector->mst_output_port->full_pbn;
+ aconnector->mst_local_bw = end_link_bw;
+ } else {
+ end_link_bw = aconnector->mst_local_bw;
+ }
+
+ if (end_link_bw > 0 && stream_kbps > end_link_bw) {
DRM_DEBUG_DRIVER("MST_DSC dsc decode at last link."
"Mode required bw can't fit into last link\n");
return DC_FAIL_BANDWIDTH_VALIDATE;
@@ -1835,11 +1880,15 @@ enum dc_status dm_dp_mst_is_port_support_mode(
if (immediate_upstream_port) {
virtual_channel_bw_in_kbps = kbps_from_pbn(immediate_upstream_port->full_pbn);
virtual_channel_bw_in_kbps = min(root_link_bw_in_kbps, virtual_channel_bw_in_kbps);
- if (bw_range.min_kbps > virtual_channel_bw_in_kbps) {
- DRM_DEBUG_DRIVER("MST_DSC dsc decode at last link."
- "Max dsc compression can't fit into MST available bw\n");
- return DC_FAIL_BANDWIDTH_VALIDATE;
- }
+ } else {
+ /* For topology LCT 1 case - only one mstb*/
+ virtual_channel_bw_in_kbps = root_link_bw_in_kbps;
+ }
+
+ if (bw_range.min_kbps > virtual_channel_bw_in_kbps) {
+ DRM_DEBUG_DRIVER("MST_DSC dsc decode at last link."
+ "Max dsc compression can't fit into MST available bw\n");
+ return DC_FAIL_BANDWIDTH_VALIDATE;
}
}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
index 495e3cd70426..eef51652ca35 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.c
@@ -26,6 +26,7 @@
#include <drm/drm_atomic_helper.h>
#include <drm/drm_blend.h>
+#include "drm/drm_framebuffer.h"
#include <drm/drm_gem_atomic_helper.h>
#include <drm/drm_plane_helper.h>
#include <drm/drm_gem_framebuffer_helper.h>
@@ -91,9 +92,9 @@ enum dm_micro_swizzle {
MICRO_SWIZZLE_R = 3
};
-const struct drm_format_info *amdgpu_dm_plane_get_format_info(const struct drm_mode_fb_cmd2 *cmd)
+const struct drm_format_info *amdgpu_dm_plane_get_format_info(u32 pixel_format, u64 modifier)
{
- return amdgpu_lookup_format_info(cmd->pixel_format, cmd->modifier[0]);
+ return amdgpu_lookup_format_info(pixel_format, modifier);
}
void amdgpu_dm_plane_fill_blending_from_plane_state(const struct drm_plane_state *plane_state,
@@ -176,7 +177,7 @@ static unsigned int amdgpu_dm_plane_modifier_gfx9_swizzle_mode(uint64_t modifier
return AMD_FMT_MOD_GET(TILE, modifier);
}
-static void amdgpu_dm_plane_fill_gfx8_tiling_info_from_flags(union dc_tiling_info *tiling_info,
+static void amdgpu_dm_plane_fill_gfx8_tiling_info_from_flags(struct dc_tiling_info *tiling_info,
uint64_t tiling_flags)
{
/* Fill GFX8 params */
@@ -189,6 +190,7 @@ static void amdgpu_dm_plane_fill_gfx8_tiling_info_from_flags(union dc_tiling_inf
tile_split = AMDGPU_TILING_GET(tiling_flags, TILE_SPLIT);
num_banks = AMDGPU_TILING_GET(tiling_flags, NUM_BANKS);
+ tiling_info->gfxversion = DcGfxVersion8;
/* XXX fix me for VI */
tiling_info->gfx8.num_banks = num_banks;
tiling_info->gfx8.array_mode =
@@ -209,7 +211,7 @@ static void amdgpu_dm_plane_fill_gfx8_tiling_info_from_flags(union dc_tiling_inf
}
static void amdgpu_dm_plane_fill_gfx9_tiling_info_from_device(const struct amdgpu_device *adev,
- union dc_tiling_info *tiling_info)
+ struct dc_tiling_info *tiling_info)
{
/* Fill GFX9 params */
tiling_info->gfx9.num_pipes =
@@ -230,7 +232,7 @@ static void amdgpu_dm_plane_fill_gfx9_tiling_info_from_device(const struct amdgp
}
static void amdgpu_dm_plane_fill_gfx9_tiling_info_from_modifier(const struct amdgpu_device *adev,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
uint64_t modifier)
{
unsigned int mod_bank_xor_bits = AMD_FMT_MOD_GET(BANK_XOR_BITS, modifier);
@@ -260,7 +262,7 @@ static void amdgpu_dm_plane_fill_gfx9_tiling_info_from_modifier(const struct amd
static int amdgpu_dm_plane_validate_dcc(struct amdgpu_device *adev,
const enum surface_pixel_format format,
const enum dc_rotation_angle rotation,
- const union dc_tiling_info *tiling_info,
+ const struct dc_tiling_info *tiling_info,
const struct dc_plane_dcc_param *dcc,
const struct dc_plane_address *address,
const struct plane_size *plane_size)
@@ -275,8 +277,11 @@ static int amdgpu_dm_plane_validate_dcc(struct amdgpu_device *adev,
if (!dcc->enable)
return 0;
- if (format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN ||
- !dc->cap_funcs.get_dcc_compression_cap)
+ if (adev->family < AMDGPU_FAMILY_GC_12_0_0 &&
+ format >= SURFACE_PIXEL_FORMAT_VIDEO_BEGIN)
+ return -EINVAL;
+
+ if (!dc->cap_funcs.get_dcc_compression_cap)
return -EINVAL;
input.format = format;
@@ -307,18 +312,18 @@ static int amdgpu_dm_plane_fill_gfx9_plane_attributes_from_modifiers(struct amdg
const enum surface_pixel_format format,
const enum dc_rotation_angle rotation,
const struct plane_size *plane_size,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct dc_plane_dcc_param *dcc,
- struct dc_plane_address *address,
- const bool force_disable_dcc)
+ struct dc_plane_address *address)
{
const uint64_t modifier = afb->base.modifier;
int ret = 0;
amdgpu_dm_plane_fill_gfx9_tiling_info_from_modifier(adev, tiling_info, modifier);
tiling_info->gfx9.swizzle = amdgpu_dm_plane_modifier_gfx9_swizzle_mode(modifier);
+ tiling_info->gfxversion = DcGfxVersion9;
- if (amdgpu_dm_plane_modifier_has_dcc(modifier) && !force_disable_dcc) {
+ if (amdgpu_dm_plane_modifier_has_dcc(modifier)) {
uint64_t dcc_address = afb->address + afb->base.offsets[1];
bool independent_64b_blks = AMD_FMT_MOD_GET(DCC_INDEPENDENT_64B, modifier);
bool independent_128b_blks = AMD_FMT_MOD_GET(DCC_INDEPENDENT_128B, modifier);
@@ -358,10 +363,9 @@ static int amdgpu_dm_plane_fill_gfx12_plane_attributes_from_modifiers(struct amd
const enum surface_pixel_format format,
const enum dc_rotation_angle rotation,
const struct plane_size *plane_size,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct dc_plane_dcc_param *dcc,
- struct dc_plane_address *address,
- const bool force_disable_dcc)
+ struct dc_plane_address *address)
{
const uint64_t modifier = afb->base.modifier;
int ret = 0;
@@ -370,8 +374,9 @@ static int amdgpu_dm_plane_fill_gfx12_plane_attributes_from_modifiers(struct amd
amdgpu_dm_plane_fill_gfx9_tiling_info_from_device(adev, tiling_info);
tiling_info->gfx9.swizzle = amdgpu_dm_plane_modifier_gfx9_swizzle_mode(modifier);
+ tiling_info->gfxversion = DcGfxAddr3;
- if (amdgpu_dm_plane_modifier_has_dcc(modifier) && !force_disable_dcc) {
+ if (amdgpu_dm_plane_modifier_has_dcc(modifier)) {
int max_compressed_block = AMD_FMT_MOD_GET(DCC_MAX_COMPRESSED_BLOCK, modifier);
dcc->enable = 1;
@@ -695,7 +700,7 @@ static void amdgpu_dm_plane_add_gfx12_modifiers(struct amdgpu_device *adev,
uint64_t mod_4k = ver | AMD_FMT_MOD_SET(TILE, AMD_FMT_MOD_TILE_GFX12_4K_2D);
uint64_t mod_256b = ver | AMD_FMT_MOD_SET(TILE, AMD_FMT_MOD_TILE_GFX12_256B_2D);
uint64_t dcc = ver | AMD_FMT_MOD_SET(DCC, 1);
- uint8_t max_comp_block[] = {1, 0};
+ uint8_t max_comp_block[] = {2, 1, 0};
uint64_t max_comp_block_mod[ARRAY_SIZE(max_comp_block)] = {0};
uint8_t i = 0, j = 0;
uint64_t gfx12_modifiers[] = {mod_256k, mod_64k, mod_4k, mod_256b, DRM_FORMAT_MOD_LINEAR};
@@ -835,12 +840,11 @@ int amdgpu_dm_plane_fill_plane_buffer_attributes(struct amdgpu_device *adev,
const enum surface_pixel_format format,
const enum dc_rotation_angle rotation,
const uint64_t tiling_flags,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
struct dc_plane_dcc_param *dcc,
struct dc_plane_address *address,
- bool tmz_surface,
- bool force_disable_dcc)
+ bool tmz_surface)
{
const struct drm_framebuffer *fb = &afb->base;
int ret;
@@ -900,16 +904,14 @@ int amdgpu_dm_plane_fill_plane_buffer_attributes(struct amdgpu_device *adev,
ret = amdgpu_dm_plane_fill_gfx12_plane_attributes_from_modifiers(adev, afb, format,
rotation, plane_size,
tiling_info, dcc,
- address,
- force_disable_dcc);
+ address);
if (ret)
return ret;
} else if (adev->family >= AMDGPU_FAMILY_AI) {
ret = amdgpu_dm_plane_fill_gfx9_plane_attributes_from_modifiers(adev, afb, format,
rotation, plane_size,
tiling_info, dcc,
- address,
- force_disable_dcc);
+ address);
if (ret)
return ret;
} else {
@@ -946,13 +948,13 @@ static int amdgpu_dm_plane_helper_prepare_fb(struct drm_plane *plane,
adev = amdgpu_ttm_adev(rbo->tbo.bdev);
r = amdgpu_bo_reserve(rbo, true);
if (r) {
- dev_err(adev->dev, "fail to reserve bo (%d)\n", r);
+ drm_err(adev_to_drm(adev), "fail to reserve bo (%d)\n", r);
return r;
}
r = dma_resv_reserve_fences(rbo->tbo.base.resv, 1);
if (r) {
- dev_err(adev->dev, "reserving fence slot failed (%d)\n", r);
+ drm_err(adev_to_drm(adev), "reserving fence slot failed (%d)\n", r);
goto error_unlock;
}
@@ -1000,14 +1002,13 @@ static int amdgpu_dm_plane_helper_prepare_fb(struct drm_plane *plane,
dm_plane_state_old->dc_state != dm_plane_state_new->dc_state) {
struct dc_plane_state *plane_state =
dm_plane_state_new->dc_state;
- bool force_disable_dcc = !plane_state->dcc.enable;
amdgpu_dm_plane_fill_plane_buffer_attributes(
adev, afb, plane_state->format, plane_state->rotation,
afb->tiling_flags,
&plane_state->tiling_info, &plane_state->plane_size,
&plane_state->dcc, &plane_state->address,
- afb->tmz_surface, force_disable_dcc);
+ afb->tmz_surface);
}
return 0;
@@ -1260,21 +1261,24 @@ static int amdgpu_dm_plane_atomic_check(struct drm_plane *plane,
}
static int amdgpu_dm_plane_atomic_async_check(struct drm_plane *plane,
- struct drm_atomic_state *state)
+ struct drm_atomic_state *state, bool flip)
{
struct drm_crtc_state *new_crtc_state;
struct drm_plane_state *new_plane_state;
struct dm_crtc_state *dm_new_crtc_state;
- /* Only support async updates on cursor planes. */
- if (plane->type != DRM_PLANE_TYPE_CURSOR)
+ if (flip) {
+ if (plane->type != DRM_PLANE_TYPE_OVERLAY)
+ return -EINVAL;
+ } else if (plane->type != DRM_PLANE_TYPE_CURSOR) {
return -EINVAL;
+ }
new_plane_state = drm_atomic_get_new_plane_state(state, plane);
new_crtc_state = drm_atomic_get_new_crtc_state(state, new_plane_state->crtc);
dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
/* Reject overlay cursors for now*/
- if (dm_new_crtc_state->cursor_mode == DM_CURSOR_OVERLAY_MODE)
+ if (!flip && dm_new_crtc_state->cursor_mode == DM_CURSOR_OVERLAY_MODE)
return -EINVAL;
return 0;
@@ -1421,6 +1425,20 @@ static void amdgpu_dm_plane_atomic_async_update(struct drm_plane *plane,
amdgpu_dm_plane_handle_cursor_update(plane, old_state);
}
+static void amdgpu_dm_plane_panic_flush(struct drm_plane *plane)
+{
+ struct dm_plane_state *dm_plane_state = to_dm_plane_state(plane->state);
+ struct drm_framebuffer *fb = plane->state->fb;
+ struct dc_plane_state *dc_plane_state;
+
+ if (!dm_plane_state || !dm_plane_state->dc_state)
+ return;
+
+ dc_plane_state = dm_plane_state->dc_state;
+
+ dc_plane_force_dcc_and_tiling_disable(dc_plane_state, fb->modifier ? true : false);
+}
+
static const struct drm_plane_helper_funcs dm_plane_helper_funcs = {
.prepare_fb = amdgpu_dm_plane_helper_prepare_fb,
.cleanup_fb = amdgpu_dm_plane_helper_cleanup_fb,
@@ -1429,6 +1447,16 @@ static const struct drm_plane_helper_funcs dm_plane_helper_funcs = {
.atomic_async_update = amdgpu_dm_plane_atomic_async_update
};
+static const struct drm_plane_helper_funcs dm_primary_plane_helper_funcs = {
+ .prepare_fb = amdgpu_dm_plane_helper_prepare_fb,
+ .cleanup_fb = amdgpu_dm_plane_helper_cleanup_fb,
+ .atomic_check = amdgpu_dm_plane_atomic_check,
+ .atomic_async_check = amdgpu_dm_plane_atomic_async_check,
+ .atomic_async_update = amdgpu_dm_plane_atomic_async_update,
+ .get_scanout_buffer = amdgpu_display_get_scanout_buffer,
+ .panic_flush = amdgpu_dm_plane_panic_flush,
+};
+
static void amdgpu_dm_plane_drm_plane_reset(struct drm_plane *plane)
{
struct dm_plane_state *amdgpu_state = NULL;
@@ -1855,7 +1883,10 @@ int amdgpu_dm_plane_init(struct amdgpu_display_manager *dm,
plane->type != DRM_PLANE_TYPE_CURSOR)
drm_plane_enable_fb_damage_clips(plane);
- drm_plane_helper_add(plane, &dm_plane_helper_funcs);
+ if (plane->type == DRM_PLANE_TYPE_PRIMARY)
+ drm_plane_helper_add(plane, &dm_primary_plane_helper_funcs);
+ else
+ drm_plane_helper_add(plane, &dm_plane_helper_funcs);
#ifdef AMD_PRIVATE_COLOR
dm_atomic_plane_attach_color_mgmt_properties(dm, plane);
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.h
index 6498359bff6f..ea2619b507db 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_plane.h
@@ -47,19 +47,18 @@ int amdgpu_dm_plane_fill_plane_buffer_attributes(struct amdgpu_device *adev,
const enum surface_pixel_format format,
const enum dc_rotation_angle rotation,
const uint64_t tiling_flags,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
struct dc_plane_dcc_param *dcc,
struct dc_plane_address *address,
- bool tmz_surface,
- bool force_disable_dcc);
+ bool tmz_surface);
int amdgpu_dm_plane_init(struct amdgpu_display_manager *dm,
struct drm_plane *plane,
unsigned long possible_crtcs,
const struct dc_plane_cap *plane_cap);
-const struct drm_format_info *amdgpu_dm_plane_get_format_info(const struct drm_mode_fb_cmd2 *cmd);
+const struct drm_format_info *amdgpu_dm_plane_get_format_info(u32 pixel_format, u64 modifier);
void amdgpu_dm_plane_fill_blending_from_plane_state(const struct drm_plane_state *plane_state,
bool *per_pixel_alpha, bool *pre_multiplied_alpha,
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c
index f40240aafe98..ff7b867ae98b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.c
@@ -54,7 +54,8 @@ static bool link_supports_psrsu(struct dc_link *link)
if (amdgpu_dc_debug_mask & DC_DISABLE_PSR_SU)
return false;
- return dc_dmub_check_min_version(dc->ctx->dmub_srv->dmub);
+ /* Temporarily disable PSR-SU to avoid glitches */
+ return false;
}
/*
@@ -86,14 +87,6 @@ void amdgpu_dm_set_psr_caps(struct dc_link *link)
link->psr_settings.psr_feature_enabled = true;
}
-
- DRM_INFO("PSR support %d, DC PSR ver %d, sink PSR ver %d DPCD caps 0x%x su_y_granularity %d\n",
- link->psr_settings.psr_feature_enabled,
- link->psr_settings.psr_version,
- link->dpcd_caps.psr_info.psr_version,
- link->dpcd_caps.psr_info.psr_dpcd_caps.raw,
- link->dpcd_caps.psr_info.psr2_su_y_granularity_cap);
-
}
/*
@@ -126,8 +119,10 @@ bool amdgpu_dm_link_setup_psr(struct dc_stream_state *stream)
psr_config.allow_multi_disp_optimizations =
(amdgpu_dc_feature_mask & DC_PSR_ALLOW_MULTI_DISP_OPT);
- if (!psr_su_set_dsc_slice_height(dc, link, stream, &psr_config))
- return false;
+ if (link->psr_settings.psr_version == DC_PSR_VERSION_SU_1) {
+ if (!psr_su_set_dsc_slice_height(dc, link, stream, &psr_config))
+ return false;
+ }
ret = dc_link_setup_psr(link, stream, &psr_config, &psr_context);
@@ -201,14 +196,13 @@ void amdgpu_dm_psr_enable(struct dc_stream_state *stream)
*
* Return: true if success
*/
-bool amdgpu_dm_psr_disable(struct dc_stream_state *stream)
+bool amdgpu_dm_psr_disable(struct dc_stream_state *stream, bool wait)
{
- unsigned int power_opt = 0;
bool psr_enable = false;
DRM_DEBUG_DRIVER("Disabling psr...\n");
- return dc_link_set_psr_allow_active(stream->link, &psr_enable, true, false, &power_opt);
+ return dc_link_set_psr_allow_active(stream->link, &psr_enable, wait, false, NULL);
}
/*
@@ -251,3 +245,33 @@ bool amdgpu_dm_psr_is_active_allowed(struct amdgpu_display_manager *dm)
return allow_active;
}
+
+/**
+ * amdgpu_dm_psr_wait_disable() - Wait for eDP panel to exit PSR
+ * @stream: stream state attached to the eDP link
+ *
+ * Waits for a max of 500ms for the eDP panel to exit PSR.
+ *
+ * Return: true if panel exited PSR, false otherwise.
+ */
+bool amdgpu_dm_psr_wait_disable(struct dc_stream_state *stream)
+{
+ enum dc_psr_state psr_state = PSR_STATE0;
+ struct dc_link *link = stream->link;
+ int retry_count;
+
+ if (link == NULL)
+ return false;
+
+ for (retry_count = 0; retry_count <= 1000; retry_count++) {
+ dc_link_get_psr_state(link, &psr_state);
+ if (psr_state == PSR_STATE0)
+ break;
+ udelay(500);
+ }
+
+ if (retry_count == 1000)
+ return false;
+
+ return true;
+}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h
index cd2d45c2b5ef..e2366321a3c1 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_psr.h
@@ -34,8 +34,9 @@
void amdgpu_dm_set_psr_caps(struct dc_link *link);
void amdgpu_dm_psr_enable(struct dc_stream_state *stream);
bool amdgpu_dm_link_setup_psr(struct dc_stream_state *stream);
-bool amdgpu_dm_psr_disable(struct dc_stream_state *stream);
+bool amdgpu_dm_psr_disable(struct dc_stream_state *stream, bool wait);
bool amdgpu_dm_psr_disable_all(struct amdgpu_display_manager *dm);
bool amdgpu_dm_psr_is_active_allowed(struct amdgpu_display_manager *dm);
+bool amdgpu_dm_psr_wait_disable(struct dc_stream_state *stream);
#endif /* AMDGPU_DM_AMDGPU_DM_PSR_H_ */
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_quirks.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_quirks.c
new file mode 100644
index 000000000000..1da07ebf9217
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_quirks.c
@@ -0,0 +1,178 @@
+// SPDX-License-Identifier: MIT
+/*
+ * Copyright 2025 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include <linux/dmi.h>
+
+#include "amdgpu.h"
+#include "amdgpu_dm.h"
+
+struct amdgpu_dm_quirks {
+ bool aux_hpd_discon;
+ bool support_edp0_on_dp1;
+};
+
+static struct amdgpu_dm_quirks quirk_entries = {
+ .aux_hpd_discon = false,
+ .support_edp0_on_dp1 = false
+};
+
+static int edp0_on_dp1_callback(const struct dmi_system_id *id)
+{
+ quirk_entries.support_edp0_on_dp1 = true;
+ return 0;
+}
+
+static int aux_hpd_discon_callback(const struct dmi_system_id *id)
+{
+ quirk_entries.aux_hpd_discon = true;
+ return 0;
+}
+
+static const struct dmi_system_id dmi_quirk_table[] = {
+ {
+ .callback = aux_hpd_discon_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "Precision 3660"),
+ },
+ },
+ {
+ .callback = aux_hpd_discon_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "Precision 3260"),
+ },
+ },
+ {
+ .callback = aux_hpd_discon_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "Precision 3460"),
+ },
+ },
+ {
+ .callback = aux_hpd_discon_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex Tower Plus 7010"),
+ },
+ },
+ {
+ .callback = aux_hpd_discon_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex Tower 7010"),
+ },
+ },
+ {
+ .callback = aux_hpd_discon_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex SFF Plus 7010"),
+ },
+ },
+ {
+ .callback = aux_hpd_discon_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex SFF 7010"),
+ },
+ },
+ {
+ .callback = aux_hpd_discon_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex Micro Plus 7010"),
+ },
+ },
+ {
+ .callback = aux_hpd_discon_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "Dell Inc."),
+ DMI_MATCH(DMI_PRODUCT_NAME, "OptiPlex Micro 7010"),
+ },
+ },
+ {
+ .callback = edp0_on_dp1_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "HP"),
+ DMI_MATCH(DMI_PRODUCT_NAME, "HP Elite mt645 G8 Mobile Thin Client"),
+ },
+ },
+ {
+ .callback = edp0_on_dp1_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "HP"),
+ DMI_MATCH(DMI_PRODUCT_NAME, "HP EliteBook 645 14 inch G11 Notebook PC"),
+ },
+ },
+ {
+ .callback = edp0_on_dp1_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "HP"),
+ DMI_MATCH(DMI_PRODUCT_NAME, "HP EliteBook 665 16 inch G11 Notebook PC"),
+ },
+ },
+ {
+ .callback = edp0_on_dp1_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "HP"),
+ DMI_MATCH(DMI_PRODUCT_NAME, "HP ProBook 445 14 inch G11 Notebook PC"),
+ },
+ },
+ {
+ .callback = edp0_on_dp1_callback,
+ .matches = {
+ DMI_MATCH(DMI_SYS_VENDOR, "HP"),
+ DMI_MATCH(DMI_PRODUCT_NAME, "HP ProBook 465 16 inch G11 Notebook PC"),
+ },
+ },
+ {}
+ /* TODO: refactor this from a fixed table to a dynamic option */
+};
+
+void retrieve_dmi_info(struct amdgpu_display_manager *dm)
+{
+ struct drm_device *dev = dm->ddev;
+ int dmi_id;
+
+ dm->aux_hpd_discon_quirk = false;
+ dm->edp0_on_dp1_quirk = false;
+
+ dmi_id = dmi_check_system(dmi_quirk_table);
+
+ if (!dmi_id)
+ return;
+
+ if (quirk_entries.aux_hpd_discon) {
+ dm->aux_hpd_discon_quirk = true;
+ drm_info(dev, "aux_hpd_discon_quirk attached\n");
+ }
+ if (quirk_entries.support_edp0_on_dp1) {
+ dm->edp0_on_dp1_quirk = true;
+ drm_info(dev, "support_edp0_on_dp1 attached\n");
+ }
+}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_trace.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_trace.h
index 4686d4b0cbad..95f890fda8aa 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_trace.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_trace.h
@@ -726,6 +726,32 @@ TRACE_EVENT(dcn_optc_lock_unlock_state,
)
);
+TRACE_EVENT(amdgpu_dm_brightness,
+ TP_PROTO(void *function, u32 user_brightness, u32 converted_brightness, bool aux, bool ac),
+ TP_ARGS(function, user_brightness, converted_brightness, aux, ac),
+ TP_STRUCT__entry(
+ __field(void *, function)
+ __field(u32, user_brightness)
+ __field(u32, converted_brightness)
+ __field(bool, aux)
+ __field(bool, ac)
+ ),
+ TP_fast_assign(
+ __entry->function = function;
+ __entry->user_brightness = user_brightness;
+ __entry->converted_brightness = converted_brightness;
+ __entry->aux = aux;
+ __entry->ac = ac;
+ ),
+ TP_printk("%ps: brightness requested=%u converted=%u aux=%s power=%s",
+ (void *)__entry->function,
+ (u32)__entry->user_brightness,
+ (u32)__entry->converted_brightness,
+ (__entry->aux) ? "true" : "false",
+ (__entry->ac) ? "AC" : "DC"
+ )
+);
+
#endif /* _AMDGPU_DM_TRACE_H_ */
#undef TRACE_INCLUDE_PATH
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c
index 0d5fefb0f591..d9527c05fc87 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_wb.c
@@ -102,13 +102,13 @@ static int amdgpu_dm_wb_prepare_job(struct drm_writeback_connector *wb_connector
r = amdgpu_bo_reserve(rbo, true);
if (r) {
- dev_err(adev->dev, "fail to reserve bo (%d)\n", r);
+ drm_err(adev_to_drm(adev), "fail to reserve bo (%d)\n", r);
return r;
}
r = dma_resv_reserve_fences(rbo->tbo.base.resv, 1);
if (r) {
- dev_err(adev->dev, "reserving fence slot failed (%d)\n", r);
+ drm_err(adev_to_drm(adev), "reserving fence slot failed (%d)\n", r);
goto error_unlock;
}
diff --git a/drivers/gpu/drm/amd/display/dc/Makefile b/drivers/gpu/drm/amd/display/dc/Makefile
index 8992e697759f..3c9ecea7eebc 100644
--- a/drivers/gpu/drm/amd/display/dc/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/Makefile
@@ -52,32 +52,31 @@ endif
DC_LIBS += hdcp
ifdef CONFIG_DRM_AMD_DC_FP
-DC_LIBS += spl
-DC_SPL_TRANS += dc_spl_translate.o
+DC_LIBS += sspl
+AMD_DISPLAY_FILES += $(addprefix $(AMDDALPATH)/dc/, dc_spl_translate.o)
endif
AMD_DC = $(addsuffix /Makefile, $(addprefix $(FULL_AMD_DISPLAY_PATH)/dc/,$(DC_LIBS)))
include $(AMD_DC)
-DISPLAY_CORE = dc.o dc_stat.o dc_resource.o dc_hw_sequencer.o dc_sink.o \
-dc_surface.o dc_debug.o dc_stream.o dc_link_enc_cfg.o dc_link_exports.o dc_state.o
+FILES =
+FILES += dc_dmub_srv.o
+FILES += dc_edid_parser.o
+FILES += dc_fused_io.o
+FILES += dc_helper.o
+FILES += core/dc.o
+FILES += core/dc_debug.o
+FILES += core/dc_hw_sequencer.o
+FILES += core/dc_link_enc_cfg.o
+FILES += core/dc_link_exports.o
+FILES += core/dc_resource.o
+FILES += core/dc_sink.o
+FILES += core/dc_stat.o
+FILES += core/dc_state.o
+FILES += core/dc_stream.o
+FILES += core/dc_surface.o
+FILES += core/dc_vm_helper.o
+
+AMD_DISPLAY_FILES += $(addprefix $(AMDDALPATH)/dc/, $(FILES))
-DISPLAY_CORE += dc_vm_helper.o
-
-AMD_DISPLAY_CORE = $(addprefix $(AMDDALPATH)/dc/core/,$(DISPLAY_CORE))
-
-AMD_DM_REG_UPDATE = $(addprefix $(AMDDALPATH)/dc/,dc_helper.o)
-
-AMD_DC_SPL_TRANS = $(addprefix $(AMDDALPATH)/dc/,$(DC_SPL_TRANS))
-
-AMD_DISPLAY_FILES += $(AMD_DISPLAY_CORE)
-AMD_DISPLAY_FILES += $(AMD_DM_REG_UPDATE)
-
-DC_DMUB += dc_dmub_srv.o
-DC_EDID += dc_edid_parser.o
-AMD_DISPLAY_DMUB = $(addprefix $(AMDDALPATH)/dc/,$(DC_DMUB))
-AMD_DISPLAY_EDID = $(addprefix $(AMDDALPATH)/dc/,$(DC_EDID))
-AMD_DISPLAY_FILES += $(AMD_DISPLAY_DMUB) $(AMD_DISPLAY_EDID)
-
-AMD_DISPLAY_FILES += $(AMD_DC_SPL_TRANS)
diff --git a/drivers/gpu/drm/amd/display/dc/basics/dc_common.c b/drivers/gpu/drm/amd/display/dc/basics/dc_common.c
index b2fc4f8e6482..a51c2701da24 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/dc_common.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/dc_common.c
@@ -40,7 +40,8 @@ bool is_rgb_cspace(enum dc_color_space output_color_space)
case COLOR_SPACE_YCBCR709:
case COLOR_SPACE_YCBCR601_LIMITED:
case COLOR_SPACE_YCBCR709_LIMITED:
- case COLOR_SPACE_2020_YCBCR:
+ case COLOR_SPACE_2020_YCBCR_LIMITED:
+ case COLOR_SPACE_2020_YCBCR_FULL:
return false;
default:
/* Add a case to switch */
diff --git a/drivers/gpu/drm/amd/display/dc/basics/dce_calcs.c b/drivers/gpu/drm/amd/display/dc/basics/dce_calcs.c
index 681799468487..d897f8a30ede 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/dce_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/dce_calcs.c
@@ -1393,7 +1393,7 @@ static void calculate_bandwidth(
if ((bw_mtn(data->dram_speed_change_margin, bw_int_to_fixed(0)) && bw_ltn(data->dram_speed_change_margin, bw_int_to_fixed(9999)))) {
/*determine the minimum dram clock change margin for each set of clock frequencies*/
data->min_dram_speed_change_margin[i][j] = bw_min2(data->min_dram_speed_change_margin[i][j], data->dram_speed_change_margin);
- /*compute the maximum clock frequuency required for the dram clock change at each set of clock frequencies*/
+ /*compute the maximum clock frequency required for the dram clock change at each set of clock frequencies*/
data->dispclk_required_for_dram_speed_change_pipe[i][j] = bw_max2(bw_div(bw_div(bw_mul(data->src_pixels_for_first_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]))), bw_div(bw_div(bw_mul(data->src_pixels_for_last_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_add(bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]), data->active_time[k]))));
if ((bw_ltn(data->dispclk_required_for_dram_speed_change_pipe[i][j], vbios->high_voltage_max_dispclk))) {
data->display_pstate_change_enable[k] = 1;
@@ -1407,7 +1407,7 @@ static void calculate_bandwidth(
if ((bw_mtn(data->dram_speed_change_margin, bw_int_to_fixed(0)) && bw_ltn(data->dram_speed_change_margin, bw_int_to_fixed(9999)))) {
/*determine the minimum dram clock change margin for each display pipe*/
data->min_dram_speed_change_margin[i][j] = bw_min2(data->min_dram_speed_change_margin[i][j], data->dram_speed_change_margin);
- /*compute the maximum clock frequuency required for the dram clock change at each set of clock frequencies*/
+ /*compute the maximum clock frequency required for the dram clock change at each set of clock frequencies*/
data->dispclk_required_for_dram_speed_change_pipe[i][j] = bw_max2(bw_div(bw_div(bw_mul(data->src_pixels_for_first_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_sub(bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]), data->mcifwr_burst_time[i][j]))), bw_div(bw_div(bw_mul(data->src_pixels_for_last_output_pixel[k], dceip->display_pipe_throughput_factor), dceip->lb_write_pixels_per_dispclk), (bw_add(bw_sub(bw_sub(bw_sub(bw_sub(data->maximum_latency_hiding_with_cursor[k], vbios->nbp_state_change_latency), data->dmif_burst_time[i][j]), data->dram_speed_change_line_source_transfer_time[k][i][j]), data->mcifwr_burst_time[i][j]), data->active_time[k]))));
if ((bw_ltn(data->dispclk_required_for_dram_speed_change_pipe[i][j], vbios->high_voltage_max_dispclk))) {
data->display_pstate_change_enable[k] = 1;
diff --git a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
index 88d3f9d7dd55..452206b5095e 100644
--- a/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
+++ b/drivers/gpu/drm/amd/display/dc/basics/fixpt31_32.c
@@ -51,8 +51,6 @@ static inline unsigned long long complete_integer_division_u64(
{
unsigned long long result;
- ASSERT(divisor);
-
result = div64_u64_rem(dividend, divisor, remainder);
return result;
@@ -213,9 +211,6 @@ struct fixed31_32 dc_fixpt_recip(struct fixed31_32 arg)
* @note
* Good idea to use Newton's method
*/
-
- ASSERT(arg.value);
-
return dc_fixpt_from_fraction(
dc_fixpt_one.value,
arg.value);
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
index 3bacf470f7c5..67f08495b7e6 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser.c
@@ -2384,10 +2384,10 @@ static enum bp_result get_integrated_info_v8(
}
/*
- * get_integrated_info_v8
+ * get_integrated_info_v9
*
* @brief
- * Get V8 integrated BIOS information
+ * Get V9 integrated BIOS information
*
* @param
* bios_parser *bp - [in]BIOS parser handler to get master data table
diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
index c9a6de110b74..04eb647acc4e 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -1778,6 +1778,7 @@ static enum bp_result get_firmware_info_v3_1(
struct dc_firmware_info *info)
{
struct atom_firmware_info_v3_1 *firmware_info;
+ struct atom_firmware_info_v3_2 *firmware_info32;
struct atom_display_controller_info_v4_1 *dce_info = NULL;
if (!info)
@@ -1785,11 +1786,13 @@ static enum bp_result get_firmware_info_v3_1(
firmware_info = GET_IMAGE(struct atom_firmware_info_v3_1,
DATA_TABLES(firmwareinfo));
+ firmware_info32 = GET_IMAGE(struct atom_firmware_info_v3_2,
+ DATA_TABLES(firmwareinfo));
dce_info = GET_IMAGE(struct atom_display_controller_info_v4_1,
DATA_TABLES(dce_info));
- if (!firmware_info || !dce_info)
+ if (!firmware_info || !firmware_info32 || !dce_info)
return BP_RESULT_BADBIOSTABLE;
memset(info, 0, sizeof(*info));
@@ -1817,7 +1820,15 @@ static enum bp_result get_firmware_info_v3_1(
bp->cmd_tbl.get_smu_clock_info(bp, SMU9_SYSPLL0_ID) * 10;
}
- info->oem_i2c_present = false;
+ /* These fields are marked as reserved in v3_1, but they appear to be populated
+ * properly.
+ */
+ if (firmware_info32 && firmware_info32->board_i2c_feature_id == 0x2) {
+ info->oem_i2c_present = true;
+ info->oem_i2c_obj_id = firmware_info32->board_i2c_feature_gpio_id;
+ } else {
+ info->oem_i2c_present = false;
+ }
return BP_RESULT_OK;
}
@@ -3088,11 +3099,12 @@ static enum bp_result construct_integrated_info(
info->ext_disp_conn_info.path[i].ext_encoder_obj_id.id,
info->ext_disp_conn_info.path[i].caps
);
- if (info->ext_disp_conn_info.path[i].caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN)
- DC_LOG_BIOS("BIOS EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN on path %d\n", i);
+ if ((info->ext_disp_conn_info.path[i].caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN)
+ DC_LOG_BIOS("BIOS AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN on path %d\n", i);
else if (bp->base.ctx->dc->config.force_bios_fixed_vs) {
- info->ext_disp_conn_info.path[i].caps |= EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN;
- DC_LOG_BIOS("driver forced EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN on path %d\n", i);
+ info->ext_disp_conn_info.path[i].caps &= ~AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK;
+ info->ext_disp_conn_info.path[i].caps |= AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN;
+ DC_LOG_BIOS("driver forced AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN on path %d\n", i);
}
}
// Log the Checksum and Voltage Swing
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
index 7d18f372ce7a..f2b1720a6a66 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table2.c
@@ -101,7 +101,6 @@ static void init_dig_encoder_control(struct bios_parser *bp)
bp->cmd_tbl.dig_encoder_control = encoder_control_digx_v1_5;
break;
default:
- dm_output_to_console("Don't have dig_encoder_control for v%d\n", version);
bp->cmd_tbl.dig_encoder_control = encoder_control_fallback;
break;
}
@@ -210,6 +209,7 @@ static enum bp_result encoder_control_fallback(
******************************************************************************
*****************************************************************************/
+
static enum bp_result transmitter_control_v1_6(
struct bios_parser *bp,
struct bp_transmitter_control *cntl);
@@ -238,7 +238,6 @@ static void init_transmitter_control(struct bios_parser *bp)
bp->cmd_tbl.transmitter_control = transmitter_control_v1_7;
break;
default:
- dm_output_to_console("Don't have transmitter_control for v%d\n", crev);
bp->cmd_tbl.transmitter_control = transmitter_control_fallback;
break;
}
@@ -325,6 +324,21 @@ static void transmitter_control_dmcub_v1_7(
dc_wake_and_execute_dmub_cmd(dmcub->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
}
+static struct dc_link *get_link_by_phy_id(struct dc *p_dc, uint32_t phy_id)
+{
+ struct dc_link *link = NULL;
+
+ // Get Transition Bitmask from dc_link structure associated with PHY
+ for (uint8_t link_id = 0; link_id < MAX_LINKS; link_id++) {
+ if (phy_id == p_dc->links[link_id]->link_enc->transmitter) {
+ link = p_dc->links[link_id];
+ break;
+ }
+ }
+
+ return link;
+}
+
static enum bp_result transmitter_control_v1_7(
struct bios_parser *bp,
struct bp_transmitter_control *cntl)
@@ -363,7 +377,38 @@ static enum bp_result transmitter_control_v1_7(
if (bp->base.ctx->dc->ctx->dmub_srv &&
bp->base.ctx->dc->debug.dmub_command_table) {
+ struct dm_process_phy_transition_init_params process_phy_transition_init_params = {0};
+ struct dc_link *link = get_link_by_phy_id(bp->base.ctx->dc, dig_v1_7.phyid);
+ bool is_phy_transition_interlock_allowed = false;
+ uint8_t action = dig_v1_7.action;
+
+ if (link) {
+ if (link->phy_transition_bitmask &&
+ (action == TRANSMITTER_CONTROL_ENABLE || action == TRANSMITTER_CONTROL_DISABLE)) {
+ is_phy_transition_interlock_allowed = true;
+
+ // Prepare input parameters for processing ACPI retimers
+ process_phy_transition_init_params.action = action;
+ process_phy_transition_init_params.display_port_lanes_count = cntl->lanes_number;
+ process_phy_transition_init_params.phy_id = dig_v1_7.phyid;
+ process_phy_transition_init_params.signal = cntl->signal;
+ process_phy_transition_init_params.sym_clock_10khz = dig_v1_7.symclk_units.symclk_10khz;
+ process_phy_transition_init_params.display_port_link_rate = link->cur_link_settings.link_rate;
+ process_phy_transition_init_params.transition_bitmask = link->phy_transition_bitmask;
+ }
+ dig_v1_7.skip_phy_ssc_reduction = link->wa_flags.skip_phy_ssc_reduction;
+ }
+
+ // Handle PRE_OFF_TO_ON: Process ACPI PHY Transition Interlock
+ if (is_phy_transition_interlock_allowed && action == TRANSMITTER_CONTROL_ENABLE)
+ dm_acpi_process_phy_transition_interlock(bp->base.ctx, process_phy_transition_init_params);
+
transmitter_control_dmcub_v1_7(bp->base.ctx->dmub_srv, &dig_v1_7);
+
+ // Handle POST_ON_TO_OFF: Process ACPI PHY Transition Interlock
+ if (is_phy_transition_interlock_allowed && action == TRANSMITTER_CONTROL_DISABLE)
+ dm_acpi_process_phy_transition_interlock(bp->base.ctx, process_phy_transition_init_params);
+
return BP_RESULT_OK;
}
@@ -408,8 +453,6 @@ static void init_set_pixel_clock(struct bios_parser *bp)
bp->cmd_tbl.set_pixel_clock = set_pixel_clock_v7;
break;
default:
- dm_output_to_console("Don't have set_pixel_clock for v%d\n",
- BIOS_CMD_TABLE_PARA_REVISION(setpixelclock));
bp->cmd_tbl.set_pixel_clock = set_pixel_clock_fallback;
break;
}
@@ -554,7 +597,6 @@ static void init_set_crtc_timing(struct bios_parser *bp)
set_crtc_using_dtd_timing_v3;
break;
default:
- dm_output_to_console("Don't have set_crtc_timing for v%d\n", dtd_version);
bp->cmd_tbl.set_crtc_timing = NULL;
break;
}
@@ -671,8 +713,6 @@ static void init_enable_crtc(struct bios_parser *bp)
bp->cmd_tbl.enable_crtc = enable_crtc_v1;
break;
default:
- dm_output_to_console("Don't have enable_crtc for v%d\n",
- BIOS_CMD_TABLE_PARA_REVISION(enablecrtc));
bp->cmd_tbl.enable_crtc = NULL;
break;
}
@@ -864,8 +904,6 @@ static void init_set_dce_clock(struct bios_parser *bp)
bp->cmd_tbl.set_dce_clock = set_dce_clock_v2_1;
break;
default:
- dm_output_to_console("Don't have set_dce_clock for v%d\n",
- BIOS_CMD_TABLE_PARA_REVISION(setdceclock));
bp->cmd_tbl.set_dce_clock = NULL;
break;
}
@@ -1046,3 +1084,4 @@ void dal_firmware_parser_init_cmd_tbl(struct bios_parser *bp)
init_enable_lvtma_control(bp);
}
+
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
index e317a3615147..91bc8a06e2cf 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.c
@@ -293,3 +293,107 @@ uint8_t dal_cmd_table_helper_encoder_id_to_atom(
return ENCODER_OBJECT_ID_NONE;
}
}
+
+uint8_t phy_id_to_atom(enum transmitter t)
+{
+ uint8_t atom_phy_id;
+
+ switch (t) {
+ case TRANSMITTER_UNIPHY_A:
+ atom_phy_id = ATOM_PHY_ID_UNIPHYA;
+ break;
+ case TRANSMITTER_UNIPHY_B:
+ atom_phy_id = ATOM_PHY_ID_UNIPHYB;
+ break;
+ case TRANSMITTER_UNIPHY_C:
+ atom_phy_id = ATOM_PHY_ID_UNIPHYC;
+ break;
+ case TRANSMITTER_UNIPHY_D:
+ atom_phy_id = ATOM_PHY_ID_UNIPHYD;
+ break;
+ case TRANSMITTER_UNIPHY_E:
+ atom_phy_id = ATOM_PHY_ID_UNIPHYE;
+ break;
+ case TRANSMITTER_UNIPHY_F:
+ atom_phy_id = ATOM_PHY_ID_UNIPHYF;
+ break;
+ case TRANSMITTER_UNIPHY_G:
+ atom_phy_id = ATOM_PHY_ID_UNIPHYG;
+ break;
+ default:
+ atom_phy_id = ATOM_PHY_ID_UNIPHYA;
+ break;
+ }
+ return atom_phy_id;
+}
+
+uint8_t clock_source_id_to_atom_phy_clk_src_id(
+ enum clock_source_id id)
+{
+ uint8_t atom_phy_clk_src_id = 0;
+
+ switch (id) {
+ case CLOCK_SOURCE_ID_PLL0:
+ atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P0PLL;
+ break;
+ case CLOCK_SOURCE_ID_PLL1:
+ atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
+ break;
+ case CLOCK_SOURCE_ID_PLL2:
+ atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P2PLL;
+ break;
+ case CLOCK_SOURCE_ID_EXTERNAL:
+ atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_REFCLK_SRC_EXT;
+ break;
+ default:
+ atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
+ break;
+ }
+
+ return atom_phy_clk_src_id >> 2;
+}
+
+bool engine_bp_to_atom(enum engine_id id, uint32_t *atom_engine_id)
+{
+ bool result = false;
+
+ if (atom_engine_id != NULL)
+ switch (id) {
+ case ENGINE_ID_DIGA:
+ *atom_engine_id = ASIC_INT_DIG1_ENCODER_ID;
+ result = true;
+ break;
+ case ENGINE_ID_DIGB:
+ *atom_engine_id = ASIC_INT_DIG2_ENCODER_ID;
+ result = true;
+ break;
+ case ENGINE_ID_DIGC:
+ *atom_engine_id = ASIC_INT_DIG3_ENCODER_ID;
+ result = true;
+ break;
+ case ENGINE_ID_DIGD:
+ *atom_engine_id = ASIC_INT_DIG4_ENCODER_ID;
+ result = true;
+ break;
+ case ENGINE_ID_DIGE:
+ *atom_engine_id = ASIC_INT_DIG5_ENCODER_ID;
+ result = true;
+ break;
+ case ENGINE_ID_DIGF:
+ *atom_engine_id = ASIC_INT_DIG6_ENCODER_ID;
+ result = true;
+ break;
+ case ENGINE_ID_DIGG:
+ *atom_engine_id = ASIC_INT_DIG7_ENCODER_ID;
+ result = true;
+ break;
+ case ENGINE_ID_DACA:
+ *atom_engine_id = ASIC_INT_DAC1_ENCODER_ID;
+ result = true;
+ break;
+ default:
+ break;
+ }
+
+ return result;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.h b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.h
index dfd30aaf4032..547700e119a6 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.h
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper.h
@@ -59,4 +59,12 @@ uint8_t dal_cmd_table_helper_transmitter_bp_to_atom(
uint8_t dal_cmd_table_helper_encoder_id_to_atom(
enum encoder_id id);
+
+uint8_t phy_id_to_atom(enum transmitter t);
+
+uint8_t clock_source_id_to_atom_phy_clk_src_id(
+ enum clock_source_id id);
+
+bool engine_bp_to_atom(enum engine_id id, uint32_t *atom_engine_id);
+
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
index 73458e295103..268e2414b34f 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/command_table_helper2.c
@@ -82,13 +82,13 @@ bool dal_bios_parser_init_cmd_tbl_helper2(
case DCN_VERSION_3_21:
case DCN_VERSION_3_5:
case DCN_VERSION_3_51:
+ case DCN_VERSION_3_6:
case DCN_VERSION_4_01:
*h = dal_cmd_tbl_helper_dce112_get_table2();
return true;
default:
- /* Unsupported DCE */
- BREAK_TO_DEBUGGER();
+ *h = dal_cmd_tbl_helper_dce112_get_table2();
return false;
}
}
diff --git a/drivers/gpu/drm/amd/display/dc/bios/dce110/command_table_helper_dce110.c b/drivers/gpu/drm/amd/display/dc/bios/dce110/command_table_helper_dce110.c
index 11bf247bb180..3099128223df 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/dce110/command_table_helper_dce110.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/dce110/command_table_helper_dce110.c
@@ -31,39 +31,6 @@
#include "../command_table_helper.h"
-static uint8_t phy_id_to_atom(enum transmitter t)
-{
- uint8_t atom_phy_id;
-
- switch (t) {
- case TRANSMITTER_UNIPHY_A:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- case TRANSMITTER_UNIPHY_B:
- atom_phy_id = ATOM_PHY_ID_UNIPHYB;
- break;
- case TRANSMITTER_UNIPHY_C:
- atom_phy_id = ATOM_PHY_ID_UNIPHYC;
- break;
- case TRANSMITTER_UNIPHY_D:
- atom_phy_id = ATOM_PHY_ID_UNIPHYD;
- break;
- case TRANSMITTER_UNIPHY_E:
- atom_phy_id = ATOM_PHY_ID_UNIPHYE;
- break;
- case TRANSMITTER_UNIPHY_F:
- atom_phy_id = ATOM_PHY_ID_UNIPHYF;
- break;
- case TRANSMITTER_UNIPHY_G:
- atom_phy_id = ATOM_PHY_ID_UNIPHYG;
- break;
- default:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- }
- return atom_phy_id;
-}
-
static uint8_t signal_type_to_atom_dig_mode(enum signal_type s)
{
uint8_t atom_dig_mode = ATOM_TRANSMITTER_DIGMODE_V5_DP;
@@ -94,32 +61,6 @@ static uint8_t signal_type_to_atom_dig_mode(enum signal_type s)
return atom_dig_mode;
}
-static uint8_t clock_source_id_to_atom_phy_clk_src_id(
- enum clock_source_id id)
-{
- uint8_t atom_phy_clk_src_id = 0;
-
- switch (id) {
- case CLOCK_SOURCE_ID_PLL0:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P0PLL;
- break;
- case CLOCK_SOURCE_ID_PLL1:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- case CLOCK_SOURCE_ID_PLL2:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P2PLL;
- break;
- case CLOCK_SOURCE_ID_EXTERNAL:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_REFCLK_SRC_EXT;
- break;
- default:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- }
-
- return atom_phy_clk_src_id >> 2;
-}
-
static uint8_t hpd_sel_to_atom(enum hpd_source_id id)
{
uint8_t atom_hpd_sel = 0;
@@ -207,51 +148,6 @@ static bool clock_source_id_to_atom(
return result;
}
-static bool engine_bp_to_atom(enum engine_id id, uint32_t *atom_engine_id)
-{
- bool result = false;
-
- if (atom_engine_id != NULL)
- switch (id) {
- case ENGINE_ID_DIGA:
- *atom_engine_id = ASIC_INT_DIG1_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGB:
- *atom_engine_id = ASIC_INT_DIG2_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGC:
- *atom_engine_id = ASIC_INT_DIG3_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGD:
- *atom_engine_id = ASIC_INT_DIG4_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGE:
- *atom_engine_id = ASIC_INT_DIG5_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGF:
- *atom_engine_id = ASIC_INT_DIG6_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGG:
- *atom_engine_id = ASIC_INT_DIG7_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DACA:
- *atom_engine_id = ASIC_INT_DAC1_ENCODER_ID;
- result = true;
- break;
- default:
- break;
- }
-
- return result;
-}
-
static uint8_t encoder_action_to_atom(enum bp_encoder_control_action action)
{
uint8_t atom_action = 0;
diff --git a/drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper2_dce112.c b/drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper2_dce112.c
index 755b6e33140a..349f0e5d5856 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper2_dce112.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper2_dce112.c
@@ -29,40 +29,9 @@
#include "include/bios_parser_types.h"
-#include "../command_table_helper2.h"
-
-static uint8_t phy_id_to_atom(enum transmitter t)
-{
- uint8_t atom_phy_id;
+#include "../command_table_helper.h"
- switch (t) {
- case TRANSMITTER_UNIPHY_A:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- case TRANSMITTER_UNIPHY_B:
- atom_phy_id = ATOM_PHY_ID_UNIPHYB;
- break;
- case TRANSMITTER_UNIPHY_C:
- atom_phy_id = ATOM_PHY_ID_UNIPHYC;
- break;
- case TRANSMITTER_UNIPHY_D:
- atom_phy_id = ATOM_PHY_ID_UNIPHYD;
- break;
- case TRANSMITTER_UNIPHY_E:
- atom_phy_id = ATOM_PHY_ID_UNIPHYE;
- break;
- case TRANSMITTER_UNIPHY_F:
- atom_phy_id = ATOM_PHY_ID_UNIPHYF;
- break;
- case TRANSMITTER_UNIPHY_G:
- atom_phy_id = ATOM_PHY_ID_UNIPHYG;
- break;
- default:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- }
- return atom_phy_id;
-}
+#include "../command_table_helper2.h"
static uint8_t signal_type_to_atom_dig_mode(enum signal_type s)
{
@@ -91,32 +60,6 @@ static uint8_t signal_type_to_atom_dig_mode(enum signal_type s)
return atom_dig_mode;
}
-static uint8_t clock_source_id_to_atom_phy_clk_src_id(
- enum clock_source_id id)
-{
- uint8_t atom_phy_clk_src_id = 0;
-
- switch (id) {
- case CLOCK_SOURCE_ID_PLL0:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P0PLL;
- break;
- case CLOCK_SOURCE_ID_PLL1:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- case CLOCK_SOURCE_ID_PLL2:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P2PLL;
- break;
- case CLOCK_SOURCE_ID_EXTERNAL:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_REFCLK_SRC_EXT;
- break;
- default:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- }
-
- return atom_phy_clk_src_id >> 2;
-}
-
static uint8_t hpd_sel_to_atom(enum hpd_source_id id)
{
uint8_t atom_hpd_sel = 0;
@@ -209,51 +152,6 @@ static bool clock_source_id_to_atom(
return result;
}
-static bool engine_bp_to_atom(enum engine_id id, uint32_t *atom_engine_id)
-{
- bool result = false;
-
- if (atom_engine_id != NULL)
- switch (id) {
- case ENGINE_ID_DIGA:
- *atom_engine_id = ASIC_INT_DIG1_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGB:
- *atom_engine_id = ASIC_INT_DIG2_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGC:
- *atom_engine_id = ASIC_INT_DIG3_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGD:
- *atom_engine_id = ASIC_INT_DIG4_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGE:
- *atom_engine_id = ASIC_INT_DIG5_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGF:
- *atom_engine_id = ASIC_INT_DIG6_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGG:
- *atom_engine_id = ASIC_INT_DIG7_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DACA:
- *atom_engine_id = ASIC_INT_DAC1_ENCODER_ID;
- result = true;
- break;
- default:
- break;
- }
-
- return result;
-}
-
static uint8_t encoder_action_to_atom(enum bp_encoder_control_action action)
{
uint8_t atom_action = 0;
diff --git a/drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper_dce112.c b/drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper_dce112.c
index 06b4f7fa4a50..1a5fefcde8af 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper_dce112.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/dce112/command_table_helper_dce112.c
@@ -31,39 +31,6 @@
#include "../command_table_helper.h"
-static uint8_t phy_id_to_atom(enum transmitter t)
-{
- uint8_t atom_phy_id;
-
- switch (t) {
- case TRANSMITTER_UNIPHY_A:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- case TRANSMITTER_UNIPHY_B:
- atom_phy_id = ATOM_PHY_ID_UNIPHYB;
- break;
- case TRANSMITTER_UNIPHY_C:
- atom_phy_id = ATOM_PHY_ID_UNIPHYC;
- break;
- case TRANSMITTER_UNIPHY_D:
- atom_phy_id = ATOM_PHY_ID_UNIPHYD;
- break;
- case TRANSMITTER_UNIPHY_E:
- atom_phy_id = ATOM_PHY_ID_UNIPHYE;
- break;
- case TRANSMITTER_UNIPHY_F:
- atom_phy_id = ATOM_PHY_ID_UNIPHYF;
- break;
- case TRANSMITTER_UNIPHY_G:
- atom_phy_id = ATOM_PHY_ID_UNIPHYG;
- break;
- default:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- }
- return atom_phy_id;
-}
-
static uint8_t signal_type_to_atom_dig_mode(enum signal_type s)
{
uint8_t atom_dig_mode = ATOM_TRANSMITTER_DIGMODE_V6_DP;
@@ -91,32 +58,6 @@ static uint8_t signal_type_to_atom_dig_mode(enum signal_type s)
return atom_dig_mode;
}
-static uint8_t clock_source_id_to_atom_phy_clk_src_id(
- enum clock_source_id id)
-{
- uint8_t atom_phy_clk_src_id = 0;
-
- switch (id) {
- case CLOCK_SOURCE_ID_PLL0:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P0PLL;
- break;
- case CLOCK_SOURCE_ID_PLL1:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- case CLOCK_SOURCE_ID_PLL2:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P2PLL;
- break;
- case CLOCK_SOURCE_ID_EXTERNAL:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_REFCLK_SRC_EXT;
- break;
- default:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- }
-
- return atom_phy_clk_src_id >> 2;
-}
-
static uint8_t hpd_sel_to_atom(enum hpd_source_id id)
{
uint8_t atom_hpd_sel = 0;
@@ -209,51 +150,6 @@ static bool clock_source_id_to_atom(
return result;
}
-static bool engine_bp_to_atom(enum engine_id id, uint32_t *atom_engine_id)
-{
- bool result = false;
-
- if (atom_engine_id != NULL)
- switch (id) {
- case ENGINE_ID_DIGA:
- *atom_engine_id = ASIC_INT_DIG1_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGB:
- *atom_engine_id = ASIC_INT_DIG2_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGC:
- *atom_engine_id = ASIC_INT_DIG3_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGD:
- *atom_engine_id = ASIC_INT_DIG4_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGE:
- *atom_engine_id = ASIC_INT_DIG5_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGF:
- *atom_engine_id = ASIC_INT_DIG6_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGG:
- *atom_engine_id = ASIC_INT_DIG7_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DACA:
- *atom_engine_id = ASIC_INT_DAC1_ENCODER_ID;
- result = true;
- break;
- default:
- break;
- }
-
- return result;
-}
-
static uint8_t encoder_action_to_atom(enum bp_encoder_control_action action)
{
uint8_t atom_action = 0;
diff --git a/drivers/gpu/drm/amd/display/dc/bios/dce60/command_table_helper_dce60.c b/drivers/gpu/drm/amd/display/dc/bios/dce60/command_table_helper_dce60.c
index 710221b4f5c5..01ccc803040c 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/dce60/command_table_helper_dce60.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/dce60/command_table_helper_dce60.c
@@ -58,51 +58,6 @@ static uint8_t encoder_action_to_atom(enum bp_encoder_control_action action)
return atom_action;
}
-static bool engine_bp_to_atom(enum engine_id id, uint32_t *atom_engine_id)
-{
- bool result = false;
-
- if (atom_engine_id != NULL)
- switch (id) {
- case ENGINE_ID_DIGA:
- *atom_engine_id = ASIC_INT_DIG1_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGB:
- *atom_engine_id = ASIC_INT_DIG2_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGC:
- *atom_engine_id = ASIC_INT_DIG3_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGD:
- *atom_engine_id = ASIC_INT_DIG4_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGE:
- *atom_engine_id = ASIC_INT_DIG5_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGF:
- *atom_engine_id = ASIC_INT_DIG6_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGG:
- *atom_engine_id = ASIC_INT_DIG7_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DACA:
- *atom_engine_id = ASIC_INT_DAC1_ENCODER_ID;
- result = true;
- break;
- default:
- break;
- }
-
- return result;
-}
-
static bool clock_source_id_to_atom(
enum clock_source_id id,
uint32_t *atom_pll_id)
@@ -149,32 +104,6 @@ static bool clock_source_id_to_atom(
return result;
}
-static uint8_t clock_source_id_to_atom_phy_clk_src_id(
- enum clock_source_id id)
-{
- uint8_t atom_phy_clk_src_id = 0;
-
- switch (id) {
- case CLOCK_SOURCE_ID_PLL0:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P0PLL;
- break;
- case CLOCK_SOURCE_ID_PLL1:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- case CLOCK_SOURCE_ID_PLL2:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P2PLL;
- break;
- case CLOCK_SOURCE_ID_EXTERNAL:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_REFCLK_SRC_EXT;
- break;
- default:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- }
-
- return atom_phy_clk_src_id >> 2;
-}
-
static uint8_t signal_type_to_atom_dig_mode(enum signal_type s)
{
uint8_t atom_dig_mode = ATOM_TRANSMITTER_DIGMODE_V5_DP;
@@ -270,39 +199,6 @@ static uint8_t dig_encoder_sel_to_atom(enum engine_id id)
return atom_dig_encoder_sel;
}
-static uint8_t phy_id_to_atom(enum transmitter t)
-{
- uint8_t atom_phy_id;
-
- switch (t) {
- case TRANSMITTER_UNIPHY_A:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- case TRANSMITTER_UNIPHY_B:
- atom_phy_id = ATOM_PHY_ID_UNIPHYB;
- break;
- case TRANSMITTER_UNIPHY_C:
- atom_phy_id = ATOM_PHY_ID_UNIPHYC;
- break;
- case TRANSMITTER_UNIPHY_D:
- atom_phy_id = ATOM_PHY_ID_UNIPHYD;
- break;
- case TRANSMITTER_UNIPHY_E:
- atom_phy_id = ATOM_PHY_ID_UNIPHYE;
- break;
- case TRANSMITTER_UNIPHY_F:
- atom_phy_id = ATOM_PHY_ID_UNIPHYF;
- break;
- case TRANSMITTER_UNIPHY_G:
- atom_phy_id = ATOM_PHY_ID_UNIPHYG;
- break;
- default:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- }
- return atom_phy_id;
-}
-
static uint8_t disp_power_gating_action_to_atom(
enum bp_pipe_control_action action)
{
diff --git a/drivers/gpu/drm/amd/display/dc/bios/dce80/command_table_helper_dce80.c b/drivers/gpu/drm/amd/display/dc/bios/dce80/command_table_helper_dce80.c
index 8b30b558cf1f..2ec5264536c7 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/dce80/command_table_helper_dce80.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/dce80/command_table_helper_dce80.c
@@ -58,51 +58,6 @@ static uint8_t encoder_action_to_atom(enum bp_encoder_control_action action)
return atom_action;
}
-static bool engine_bp_to_atom(enum engine_id id, uint32_t *atom_engine_id)
-{
- bool result = false;
-
- if (atom_engine_id != NULL)
- switch (id) {
- case ENGINE_ID_DIGA:
- *atom_engine_id = ASIC_INT_DIG1_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGB:
- *atom_engine_id = ASIC_INT_DIG2_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGC:
- *atom_engine_id = ASIC_INT_DIG3_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGD:
- *atom_engine_id = ASIC_INT_DIG4_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGE:
- *atom_engine_id = ASIC_INT_DIG5_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGF:
- *atom_engine_id = ASIC_INT_DIG6_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DIGG:
- *atom_engine_id = ASIC_INT_DIG7_ENCODER_ID;
- result = true;
- break;
- case ENGINE_ID_DACA:
- *atom_engine_id = ASIC_INT_DAC1_ENCODER_ID;
- result = true;
- break;
- default:
- break;
- }
-
- return result;
-}
-
static bool clock_source_id_to_atom(
enum clock_source_id id,
uint32_t *atom_pll_id)
@@ -149,32 +104,6 @@ static bool clock_source_id_to_atom(
return result;
}
-static uint8_t clock_source_id_to_atom_phy_clk_src_id(
- enum clock_source_id id)
-{
- uint8_t atom_phy_clk_src_id = 0;
-
- switch (id) {
- case CLOCK_SOURCE_ID_PLL0:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P0PLL;
- break;
- case CLOCK_SOURCE_ID_PLL1:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- case CLOCK_SOURCE_ID_PLL2:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P2PLL;
- break;
- case CLOCK_SOURCE_ID_EXTERNAL:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_REFCLK_SRC_EXT;
- break;
- default:
- atom_phy_clk_src_id = ATOM_TRANSMITTER_CONFIG_V5_P1PLL;
- break;
- }
-
- return atom_phy_clk_src_id >> 2;
-}
-
static uint8_t signal_type_to_atom_dig_mode(enum signal_type s)
{
uint8_t atom_dig_mode = ATOM_TRANSMITTER_DIGMODE_V5_DP;
@@ -270,39 +199,6 @@ static uint8_t dig_encoder_sel_to_atom(enum engine_id id)
return atom_dig_encoder_sel;
}
-static uint8_t phy_id_to_atom(enum transmitter t)
-{
- uint8_t atom_phy_id;
-
- switch (t) {
- case TRANSMITTER_UNIPHY_A:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- case TRANSMITTER_UNIPHY_B:
- atom_phy_id = ATOM_PHY_ID_UNIPHYB;
- break;
- case TRANSMITTER_UNIPHY_C:
- atom_phy_id = ATOM_PHY_ID_UNIPHYC;
- break;
- case TRANSMITTER_UNIPHY_D:
- atom_phy_id = ATOM_PHY_ID_UNIPHYD;
- break;
- case TRANSMITTER_UNIPHY_E:
- atom_phy_id = ATOM_PHY_ID_UNIPHYE;
- break;
- case TRANSMITTER_UNIPHY_F:
- atom_phy_id = ATOM_PHY_ID_UNIPHYF;
- break;
- case TRANSMITTER_UNIPHY_G:
- atom_phy_id = ATOM_PHY_ID_UNIPHYG;
- break;
- default:
- atom_phy_id = ATOM_PHY_ID_UNIPHYA;
- break;
- }
- return atom_phy_id;
-}
-
static uint8_t disp_power_gating_action_to_atom(
enum bp_pipe_control_action action)
{
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile b/drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile
index ab1132bc896a..60021671b386 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/Makefile
@@ -112,7 +112,7 @@ AMD_DISPLAY_FILES += $(AMD_DAL_CLK_MGR_DCN21)
###############################################################################
# DCN30
###############################################################################
-CLK_MGR_DCN30 = dcn30_clk_mgr.o dcn30_clk_mgr_smu_msg.o
+CLK_MGR_DCN30 = dcn30_clk_mgr.o dcn30_clk_mgr_smu_msg.o dcn30m_clk_mgr.o dcn30m_clk_mgr_smu_msg.o
AMD_DAL_CLK_MGR_DCN30 = $(addprefix $(AMDDALPATH)/dc/clk_mgr/dcn30/,$(CLK_MGR_DCN30))
@@ -174,7 +174,7 @@ AMD_DISPLAY_FILES += $(AMD_DAL_CLK_MGR_DCN32)
###############################################################################
# DCN35
###############################################################################
-CLK_MGR_DCN35 = dcn35_smu.o dcn35_clk_mgr.o
+CLK_MGR_DCN35 = dcn35_smu.o dcn351_clk_mgr.o dcn35_clk_mgr.o
AMD_DAL_CLK_MGR_DCN35 = $(addprefix $(AMDDALPATH)/dc/clk_mgr/dcn35/,$(CLK_MGR_DCN35))
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c
index 0e243f4344d0..4071851f9e86 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/clk_mgr.c
@@ -67,7 +67,7 @@ int clk_mgr_helper_get_active_display_cnt(
if (dc_state_get_stream_subvp_type(context, stream) == SUBVP_PHANTOM)
continue;
- if (!stream->dpms_off || (stream_status && stream_status->plane_count))
+ if (!stream->dpms_off || dc->is_switch_in_progress_dest || (stream_status && stream_status->plane_count))
display_count++;
}
@@ -158,7 +158,6 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p
return NULL;
}
dce60_clk_mgr_construct(ctx, clk_mgr);
- dce_clk_mgr_construct(ctx, clk_mgr);
return &clk_mgr->base;
}
#endif
@@ -355,8 +354,11 @@ struct clk_mgr *dc_clk_mgr_create(struct dc_context *ctx, struct pp_smu_funcs *p
BREAK_TO_DEBUGGER();
return NULL;
}
+ if (ctx->dce_version == DCN_VERSION_3_51)
+ dcn351_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg);
+ else
+ dcn35_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg);
- dcn35_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg);
return &clk_mgr->base.base;
}
break;
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c
index 26feefbb8990..f5ad0a177038 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dce100/dce_clk_mgr.c
@@ -245,6 +245,11 @@ int dce_set_clock(
pxl_clk_params.target_pixel_clock_100hz = requested_clk_khz * 10;
pxl_clk_params.pll_id = CLOCK_SOURCE_ID_DFS;
+ /* DCE 6.0, DCE 6.4: engine clock is the same as PLL0 */
+ if (clk_mgr_base->ctx->dce_version == DCE_VERSION_6_0 ||
+ clk_mgr_base->ctx->dce_version == DCE_VERSION_6_4)
+ pxl_clk_params.pll_id = CLOCK_SOURCE_ID_PLL0;
+
if (clk_mgr_dce->dfs_bypass_active)
pxl_clk_params.flags.SET_DISPCLK_DFS_BYPASS = true;
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c
index 19897fa52e7e..d82a52319088 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.c
@@ -142,17 +142,3 @@ int rv1_vbios_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_di
return actual_dispclk_set_mhz * 1000;
}
-
-int rv1_vbios_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr)
-{
- int actual_dprefclk_set_mhz = -1;
-
- actual_dprefclk_set_mhz = rv1_vbios_smu_send_msg_with_param(
- clk_mgr,
- VBIOSSMC_MSG_SetDprefclkFreq,
- khz_to_mhz_ceil(clk_mgr->base.dprefclk_khz));
-
- /* TODO: add code for programing DP DTO, currently this is down by command table */
-
- return actual_dprefclk_set_mhz * 1000;
-}
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.h
index 083cb3158859..81d7c912549c 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr_vbios_smu.h
@@ -27,6 +27,5 @@
#define DAL_DC_DCN10_RV1_CLK_MGR_VBIOS_SMU_H_
int rv1_vbios_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dispclk_khz);
-int rv1_vbios_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr);
#endif /* DAL_DC_DCN10_RV1_CLK_MGR_VBIOS_SMU_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c
index 7920f6f1aa62..76c612ecfe3c 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn201/dcn201_clk_mgr.c
@@ -34,8 +34,8 @@
#include "dm_services.h"
#include "cyan_skillfish_ip_offset.h"
-#include "dcn/dcn_2_0_3_offset.h"
-#include "dcn/dcn_2_0_3_sh_mask.h"
+#include "dcn/dcn_2_0_1_offset.h"
+#include "dcn/dcn_2_0_1_sh_mask.h"
#include "clk/clk_11_0_1_offset.h"
#include "clk/clk_11_0_1_sh_mask.h"
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c
index 23b390245b5d..5a633333dbb5 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.c
@@ -164,20 +164,6 @@ int rn_vbios_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dis
return actual_dispclk_set_mhz * 1000;
}
-int rn_vbios_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr)
-{
- int actual_dprefclk_set_mhz = -1;
-
- actual_dprefclk_set_mhz = rn_vbios_smu_send_msg_with_param(
- clk_mgr,
- VBIOSSMC_MSG_SetDprefclkFreq,
- khz_to_mhz_ceil(clk_mgr->base.dprefclk_khz));
-
- /* TODO: add code for programing DP DTO, currently this is down by command table */
-
- return actual_dprefclk_set_mhz * 1000;
-}
-
int rn_vbios_smu_set_hard_min_dcfclk(struct clk_mgr_internal *clk_mgr, int requested_dcfclk_khz)
{
int actual_dcfclk_set_mhz = -1;
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.h
index 1ce19d875358..f76fad87f0e1 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn21/rn_clk_mgr_vbios_smu.h
@@ -30,7 +30,6 @@ enum dcn_pwr_state;
int rn_vbios_smu_get_smu_version(struct clk_mgr_internal *clk_mgr);
int rn_vbios_smu_set_dispclk(struct clk_mgr_internal *clk_mgr, int requested_dispclk_khz);
-int rn_vbios_smu_set_dprefclk(struct clk_mgr_internal *clk_mgr);
int rn_vbios_smu_set_hard_min_dcfclk(struct clk_mgr_internal *clk_mgr, int requested_dcfclk_khz);
int rn_vbios_smu_set_min_deep_sleep_dcfclk(struct clk_mgr_internal *clk_mgr, int requested_min_ds_dcfclk_khz);
void rn_vbios_smu_set_phyclk(struct clk_mgr_internal *clk_mgr, int requested_phyclk_khz);
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dalsmc.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dalsmc.h
index fa09c594fd36..06da34676965 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dalsmc.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dalsmc.h
@@ -56,6 +56,7 @@
#define DALSMC_MSG_SetDisplayRefreshFromMall 0xF
#define DALSMC_MSG_SetExternalClientDfCstateAllow 0x10
#define DALSMC_MSG_BacoAudioD3PME 0x11
-#define DALSMC_Message_Count 0x12
+#define DALSMC_MSG_SmartAccess 0x12
+#define DALSMC_Message_Count 0x13
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c
index 8083a553c60e..ef77fcd164ed 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c
@@ -30,6 +30,7 @@
#include "dce100/dce_clk_mgr.h"
#include "dcn30/dcn30_clk_mgr.h"
#include "dml/dcn30/dcn30_fpu.h"
+#include "dcn30/dcn30m_clk_mgr.h"
#include "reg_helper.h"
#include "core_types.h"
#include "dm_helpers.h"
@@ -498,7 +499,8 @@ static struct clk_mgr_funcs dcn3_funcs = {
.are_clock_states_equal = dcn3_are_clock_states_equal,
.enable_pme_wa = dcn3_enable_pme_wa,
.notify_link_rate_change = dcn30_notify_link_rate_change,
- .is_smu_present = dcn3_is_smu_present
+ .is_smu_present = dcn3_is_smu_present,
+ .set_smartmux_switch = dcn30m_set_smartmux_switch
};
static void dcn3_init_clocks_fpga(struct clk_mgr *clk_mgr)
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr.c
new file mode 100644
index 000000000000..8e8a11c7437e
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr.c
@@ -0,0 +1,36 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "clk_mgr_internal.h"
+#include "dcn30/dcn30m_clk_mgr.h"
+#include "dcn30m_clk_mgr_smu_msg.h"
+
+
+uint32_t dcn30m_set_smartmux_switch(struct clk_mgr *clk_mgr_base, uint32_t pins_to_set)
+{
+ struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
+
+ return dcn30m_smu_set_smart_mux_switch(clk_mgr, pins_to_set);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr.h
new file mode 100644
index 000000000000..757985b2eadc
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr.h
@@ -0,0 +1,31 @@
+/*
+ * Copyright 2019 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef __DCN30M_CLK_MGR_H__
+#define __DCN30M_CLK_MGR_H__
+
+uint32_t dcn30m_set_smartmux_switch(struct clk_mgr *clk_mgr_base, uint32_t pins_to_set);
+
+#endif //__DCN30M_CLK_MGR_H__
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr_smu_msg.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr_smu_msg.c
new file mode 100644
index 000000000000..0dd0583ff21e
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr_smu_msg.c
@@ -0,0 +1,118 @@
+/*
+ * Copyright 2020 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "dcn30m_clk_mgr_smu_msg.h"
+
+#include "clk_mgr_internal.h"
+#include "reg_helper.h"
+#include "dm_helpers.h"
+
+#include "dalsmc.h"
+
+#define mmDAL_MSG_REG 0x1628A
+#define mmDAL_ARG_REG 0x16273
+#define mmDAL_RESP_REG 0x16274
+
+#define REG(reg_name) \
+ mm ## reg_name
+
+#include "logger_types.h"
+#undef DC_LOGGER
+#define DC_LOGGER \
+ CTX->logger
+#define smu_print(str, ...) {DC_LOG_SMU(str, ##__VA_ARGS__); }
+
+
+/*
+ * Function to be used instead of REG_WAIT macro because the wait ends when
+ * the register is NOT EQUAL to zero, and because the translation in msg_if.h
+ * won't work with REG_WAIT.
+ */
+static uint32_t dcn30m_smu_wait_for_response(struct clk_mgr_internal *clk_mgr,
+ unsigned int delay_us, unsigned int max_retries)
+{
+ uint32_t reg = 0;
+
+ do {
+ reg = REG_READ(DAL_RESP_REG);
+ if (reg)
+ break;
+
+ if (delay_us >= 1000)
+ msleep(delay_us/1000);
+ else if (delay_us > 0)
+ udelay(delay_us);
+ } while (max_retries--);
+
+ /* handle DALSMC_Result_CmdRejectedBusy? */
+
+ /* Log? */
+
+ return reg;
+}
+
+static bool dcn30m_smu_send_msg_with_param(struct clk_mgr_internal *clk_mgr,
+ uint32_t msg_id, uint32_t param_in, uint32_t *param_out)
+{
+ uint32_t result;
+ /* Wait for response register to be ready */
+ dcn30m_smu_wait_for_response(clk_mgr, 10, 200000);
+
+ /* Clear response register */
+ REG_WRITE(DAL_RESP_REG, 0);
+
+ /* Set the parameter register for the SMU message */
+ REG_WRITE(DAL_ARG_REG, param_in);
+
+ /* Trigger the message transaction by writing the message ID */
+ REG_WRITE(DAL_MSG_REG, msg_id);
+
+ result = dcn30m_smu_wait_for_response(clk_mgr, 10, 200000);
+
+ if (IS_SMU_TIMEOUT(result))
+ dm_helpers_smu_timeout(CTX, msg_id, param_in, 10 * 200000);
+
+ /* Wait for response */
+ if (result == DALSMC_Result_OK) {
+ if (param_out)
+ *param_out = REG_READ(DAL_ARG_REG);
+
+ return true;
+ }
+
+ return false;
+}
+
+uint32_t dcn30m_smu_set_smart_mux_switch(struct clk_mgr_internal *clk_mgr, uint32_t pins_to_set)
+{
+ uint32_t response = 0;
+
+ smu_print("SMU Set SmartMux Switch: switch_dgpu = %d\n", pins_to_set);
+
+ dcn30m_smu_send_msg_with_param(clk_mgr,
+ DALSMC_MSG_SmartAccess, pins_to_set, &response);
+
+ return response;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr_smu_msg.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr_smu_msg.h
new file mode 100644
index 000000000000..8a59a473fc5e
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30m_clk_mgr_smu_msg.h
@@ -0,0 +1,34 @@
+/*
+ * Copyright 2020 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#ifndef DAL_DC_DCN30M_CLK_MGR_SMU_MSG_H_
+#define DAL_DC_DCN30M_CLK_MGR_SMU_MSG_H_
+
+#include "core_types.h"
+
+struct clk_mgr_internal;
+
+uint32_t dcn30m_smu_set_smart_mux_switch(struct clk_mgr_internal *clk_mgr, uint32_t pins_to_set);
+#endif /* DAL_DC_DCN30M_CLK_MGR_SMU_MSG_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c
index a0fb4481d2f1..e4d22f74f986 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_clk_mgr.c
@@ -130,7 +130,7 @@ static void dcn315_update_clocks(struct clk_mgr *clk_mgr_base,
struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
struct dc_clocks *new_clocks = &context->bw_ctx.bw.dcn.clk;
struct dc *dc = clk_mgr_base->ctx->dc;
- int display_count;
+ int display_count = 0;
bool update_dppclk = false;
bool update_dispclk = false;
bool dpp_clock_lowered = false;
@@ -194,8 +194,6 @@ static void dcn315_update_clocks(struct clk_mgr *clk_mgr_base,
// workaround: Limit dppclk to 100Mhz to avoid lower eDP panel switch to plus 4K monitor underflow.
if (new_clocks->dppclk_khz < MIN_DPP_DISP_CLK)
new_clocks->dppclk_khz = MIN_DPP_DISP_CLK;
- if (new_clocks->dispclk_khz < MIN_DPP_DISP_CLK)
- new_clocks->dispclk_khz = MIN_DPP_DISP_CLK;
if (should_set_clock(safe_to_lower, new_clocks->dppclk_khz, clk_mgr->base.clks.dppclk_khz)) {
if (clk_mgr->base.clks.dppclk_khz > new_clocks->dppclk_khz)
@@ -204,15 +202,19 @@ static void dcn315_update_clocks(struct clk_mgr *clk_mgr_base,
update_dppclk = true;
}
- if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)) {
- /* No need to apply the w/a if we haven't taken over from bios yet */
- if (clk_mgr_base->clks.dispclk_khz)
- dcn315_disable_otg_wa(clk_mgr_base, context, true);
+ if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz) &&
+ (new_clocks->dispclk_khz > 0 || (safe_to_lower && display_count == 0))) {
+ int requested_dispclk_khz = new_clocks->dispclk_khz;
+ dcn315_disable_otg_wa(clk_mgr_base, context, true);
+
+ /* Clamp the requested clock to PMFW based on their limit. */
+ if (dc->debug.min_disp_clk_khz > 0 && requested_dispclk_khz < dc->debug.min_disp_clk_khz)
+ requested_dispclk_khz = dc->debug.min_disp_clk_khz;
+
+ dcn315_smu_set_dispclk(clk_mgr, requested_dispclk_khz);
clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz;
- dcn315_smu_set_dispclk(clk_mgr, clk_mgr_base->clks.dispclk_khz);
- if (clk_mgr_base->clks.dispclk_khz)
- dcn315_disable_otg_wa(clk_mgr_base, context, false);
+ dcn315_disable_otg_wa(clk_mgr_base, context, false);
update_dispclk = true;
}
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c
index 2d14346b680e..478b4d6a3544 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn315/dcn315_smu.c
@@ -49,12 +49,9 @@ static const struct IP_BASE MP0_BASE = { { { { 0x00016000, 0x00DC0000, 0x00E0000
{ { 0, 0, 0, 0, 0, 0 } },
{ { 0, 0, 0, 0, 0, 0 } },
{ { 0, 0, 0, 0, 0, 0 } } } };
-static const struct IP_BASE NBIO_BASE = { { { { 0x00000000, 0x00000014, 0x00000D20, 0x00010400, 0x0241B000, 0x04040000 } },
- { { 0, 0, 0, 0, 0, 0 } },
- { { 0, 0, 0, 0, 0, 0 } },
- { { 0, 0, 0, 0, 0, 0 } },
- { { 0, 0, 0, 0, 0, 0 } },
- { { 0, 0, 0, 0, 0, 0 } } } };
+
+#define CTX clk_mgr->base.ctx
+#define IND_REG(offset) offset
#define regBIF_BX_PF2_RSMU_INDEX 0x0000
#define regBIF_BX_PF2_RSMU_INDEX_BASE_IDX 1
@@ -67,9 +64,6 @@ static const struct IP_BASE NBIO_BASE = { { { { 0x00000000, 0x00000014, 0x00000D
#define FN(reg_name, field) \
FD(reg_name##__##field)
-#define REG_NBIO(reg_name) \
- (NBIO_BASE.instance[0].segment[regBIF_BX_PF2_ ## reg_name ## _BASE_IDX] + regBIF_BX_PF2_ ## reg_name)
-
#undef DC_LOGGER
#define DC_LOGGER \
CTX->logger
@@ -77,6 +71,13 @@ static const struct IP_BASE NBIO_BASE = { { { { 0x00000000, 0x00000014, 0x00000D
#define mmMP1_C2PMSG_3 0x3B1050C
+#define reg__MP1_C2PMSG_3_MASK (0xFFFFFFFF)
+#define reg__MP1_C2PMSG_3__SHIFT (0)
+
+
+#define data_reg_name__MP1_C2PMSG_3_MASK (0xFFFFFFFF)
+#define data_reg_name__MP1_C2PMSG_3__SHIFT (0)
+
#define VBIOSSMC_MSG_TestMessage 0x01 ///< To check if PMFW is alive and responding. Requirement specified by PMFW team
#define VBIOSSMC_MSG_GetPmfwVersion 0x02 ///< Get PMFW version
#define VBIOSSMC_MSG_Spare0 0x03 ///< Spare0
@@ -153,12 +154,10 @@ static int dcn315_smu_send_msg_with_param(
for (i = 0; i < SMU_REGISTER_WRITE_RETRY_COUNT; i++) {
/* Trigger the message transaction by writing the message ID */
- generic_write_indirect_reg(CTX,
- REG_NBIO(RSMU_INDEX), REG_NBIO(RSMU_DATA),
- mmMP1_C2PMSG_3, msg_id);
- read_back_data = generic_read_indirect_reg(CTX,
- REG_NBIO(RSMU_INDEX), REG_NBIO(RSMU_DATA),
- mmMP1_C2PMSG_3);
+ IX_REG_SET_SYNC(mmMP1_C2PMSG_3, 0,
+ MP1_C2PMSG_3, msg_id);
+ IX_REG_GET_SYNC(mmMP1_C2PMSG_3,
+ MP1_C2PMSG_3, &read_back_data);
if (read_back_data == msg_id)
break;
udelay(2);
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c
index c3e50c3aaa60..49efea0c8fcf 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn316/dcn316_clk_mgr.c
@@ -140,7 +140,7 @@ static void dcn316_update_clocks(struct clk_mgr *clk_mgr_base,
struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
struct dc_clocks *new_clocks = &context->bw_ctx.bw.dcn.clk;
struct dc *dc = clk_mgr_base->ctx->dc;
- int display_count;
+ int display_count = 0;
bool update_dppclk = false;
bool update_dispclk = false;
bool dpp_clock_lowered = false;
@@ -201,8 +201,6 @@ static void dcn316_update_clocks(struct clk_mgr *clk_mgr_base,
// workaround: Limit dppclk to 100Mhz to avoid lower eDP panel switch to plus 4K monitor underflow.
if (new_clocks->dppclk_khz < 100000)
new_clocks->dppclk_khz = 100000;
- if (new_clocks->dispclk_khz < 100000)
- new_clocks->dispclk_khz = 100000;
if (should_set_clock(safe_to_lower, new_clocks->dppclk_khz, clk_mgr->base.clks.dppclk_khz)) {
if (clk_mgr->base.clks.dppclk_khz > new_clocks->dppclk_khz)
@@ -211,11 +209,18 @@ static void dcn316_update_clocks(struct clk_mgr *clk_mgr_base,
update_dppclk = true;
}
- if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)) {
+ if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz) &&
+ (new_clocks->dispclk_khz > 0 || (safe_to_lower && display_count == 0))) {
+ int requested_dispclk_khz = new_clocks->dispclk_khz;
+
dcn316_disable_otg_wa(clk_mgr_base, context, safe_to_lower, true);
+ /* Clamp the requested clock to PMFW based on their limit. */
+ if (dc->debug.min_disp_clk_khz > 0 && requested_dispclk_khz < dc->debug.min_disp_clk_khz)
+ requested_dispclk_khz = dc->debug.min_disp_clk_khz;
+
+ dcn316_smu_set_dispclk(clk_mgr, requested_dispclk_khz);
clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz;
- dcn316_smu_set_dispclk(clk_mgr, clk_mgr_base->clks.dispclk_khz);
dcn316_disable_otg_wa(clk_mgr_base, context, safe_to_lower, false);
update_dispclk = true;
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn351_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn351_clk_mgr.c
new file mode 100644
index 000000000000..4607eff07253
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn351_clk_mgr.c
@@ -0,0 +1,141 @@
+/*
+ * Copyright 2024 Advanced Micro Devices, Inc.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the "Software"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
+ * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
+ * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
+ * OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * Authors: AMD
+ *
+ */
+
+#include "core_types.h"
+#include "dcn35_clk_mgr.h"
+
+#define DCN_BASE__INST0_SEG1 0x000000C0
+#define mmCLK1_CLK_PLL_REQ 0x16E37
+
+#define mmCLK1_CLK0_DFS_CNTL 0x16E69
+#define mmCLK1_CLK1_DFS_CNTL 0x16E6C
+#define mmCLK1_CLK2_DFS_CNTL 0x16E6F
+#define mmCLK1_CLK3_DFS_CNTL 0x16E72
+#define mmCLK1_CLK4_DFS_CNTL 0x16E75
+#define mmCLK1_CLK5_DFS_CNTL 0x16E78
+
+#define mmCLK1_CLK0_CURRENT_CNT 0x16EFC
+#define mmCLK1_CLK1_CURRENT_CNT 0x16EFD
+#define mmCLK1_CLK2_CURRENT_CNT 0x16EFE
+#define mmCLK1_CLK3_CURRENT_CNT 0x16EFF
+#define mmCLK1_CLK4_CURRENT_CNT 0x16F00
+#define mmCLK1_CLK5_CURRENT_CNT 0x16F01
+
+#define mmCLK1_CLK0_BYPASS_CNTL 0x16E8A
+#define mmCLK1_CLK1_BYPASS_CNTL 0x16E93
+#define mmCLK1_CLK2_BYPASS_CNTL 0x16E9C
+#define mmCLK1_CLK3_BYPASS_CNTL 0x16EA5
+#define mmCLK1_CLK4_BYPASS_CNTL 0x16EAE
+#define mmCLK1_CLK5_BYPASS_CNTL 0x16EB7
+
+#define mmCLK1_CLK0_DS_CNTL 0x16E83
+#define mmCLK1_CLK1_DS_CNTL 0x16E8C
+#define mmCLK1_CLK2_DS_CNTL 0x16E95
+#define mmCLK1_CLK3_DS_CNTL 0x16E9E
+#define mmCLK1_CLK4_DS_CNTL 0x16EA7
+#define mmCLK1_CLK5_DS_CNTL 0x16EB0
+
+#define mmCLK1_CLK0_ALLOW_DS 0x16E84
+#define mmCLK1_CLK1_ALLOW_DS 0x16E8D
+#define mmCLK1_CLK2_ALLOW_DS 0x16E96
+#define mmCLK1_CLK3_ALLOW_DS 0x16E9F
+#define mmCLK1_CLK4_ALLOW_DS 0x16EA8
+#define mmCLK1_CLK5_ALLOW_DS 0x16EB1
+
+#define mmCLK5_spll_field_8 0x1B04B
+#define mmCLK6_spll_field_8 0x1B24B
+#define mmDENTIST_DISPCLK_CNTL 0x0124
+#define regDENTIST_DISPCLK_CNTL 0x0064
+#define regDENTIST_DISPCLK_CNTL_BASE_IDX 1
+
+#define CLK1_CLK_PLL_REQ__FbMult_int__SHIFT 0x0
+#define CLK1_CLK_PLL_REQ__PllSpineDiv__SHIFT 0xc
+#define CLK1_CLK_PLL_REQ__FbMult_frac__SHIFT 0x10
+#define CLK1_CLK_PLL_REQ__FbMult_int_MASK 0x000001FFL
+#define CLK1_CLK_PLL_REQ__PllSpineDiv_MASK 0x0000F000L
+#define CLK1_CLK_PLL_REQ__FbMult_frac_MASK 0xFFFF0000L
+
+#define CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_SEL_MASK 0x00000007L
+
+// DENTIST_DISPCLK_CNTL
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_WDIVIDER__SHIFT 0x0
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_RDIVIDER__SHIFT 0x8
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_CHG_DONE__SHIFT 0x13
+#define DENTIST_DISPCLK_CNTL__DENTIST_DPPCLK_CHG_DONE__SHIFT 0x14
+#define DENTIST_DISPCLK_CNTL__DENTIST_DPPCLK_WDIVIDER__SHIFT 0x18
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_WDIVIDER_MASK 0x0000007FL
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_RDIVIDER_MASK 0x00007F00L
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_CHG_DONE_MASK 0x00080000L
+#define DENTIST_DISPCLK_CNTL__DENTIST_DPPCLK_CHG_DONE_MASK 0x00100000L
+#define DENTIST_DISPCLK_CNTL__DENTIST_DPPCLK_WDIVIDER_MASK 0x7F000000L
+
+#define CLK5_spll_field_8__spll_ssc_en_MASK 0x00002000L
+
+#define REG(reg) \
+ (clk_mgr->regs->reg)
+
+#define BASE_INNER(seg) DCN_BASE__INST0_SEG ## seg
+
+#define BASE(seg) BASE_INNER(seg)
+
+#define SR(reg_name)\
+ .reg_name = BASE(reg ## reg_name ## _BASE_IDX) + \
+ reg ## reg_name
+
+#define CLK_SR_DCN35(reg_name)\
+ .reg_name = mm ## reg_name
+
+static const struct clk_mgr_registers clk_mgr_regs_dcn351 = {
+ CLK_REG_LIST_DCN35()
+};
+
+static const struct clk_mgr_shift clk_mgr_shift_dcn351 = {
+ CLK_COMMON_MASK_SH_LIST_DCN32(__SHIFT)
+};
+
+static const struct clk_mgr_mask clk_mgr_mask_dcn351 = {
+ CLK_COMMON_MASK_SH_LIST_DCN32(_MASK)
+};
+
+#define TO_CLK_MGR_DCN35(clk_mgr)\
+ container_of(clk_mgr, struct clk_mgr_dcn35, base)
+
+
+void dcn351_clk_mgr_construct(
+ struct dc_context *ctx,
+ struct clk_mgr_dcn35 *clk_mgr,
+ struct pp_smu_funcs *pp_smu,
+ struct dccg *dccg)
+{
+ /*register offset changed*/
+ clk_mgr->base.regs = &clk_mgr_regs_dcn351;
+ clk_mgr->base.clk_mgr_shift = &clk_mgr_shift_dcn351;
+ clk_mgr->base.clk_mgr_mask = &clk_mgr_mask_dcn351;
+
+ dcn35_clk_mgr_construct(ctx, clk_mgr, pp_smu, dccg);
+
+}
+
+
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c
index b77333817f18..bb1ac12a2b09 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c
@@ -36,15 +36,11 @@
#include "dcn20/dcn20_clk_mgr.h"
-
-
#include "reg_helper.h"
#include "core_types.h"
#include "dcn35_smu.h"
#include "dm_helpers.h"
-/* TODO: remove this include once we ported over remaining clk mgr functions*/
-#include "dcn30/dcn30_clk_mgr.h"
#include "dcn31/dcn31_clk_mgr.h"
#include "dc_dmub_srv.h"
@@ -55,35 +51,104 @@
#define DC_LOGGER \
clk_mgr->base.base.ctx->logger
+#define DCN_BASE__INST0_SEG1 0x000000C0
+#define mmCLK1_CLK_PLL_REQ 0x16E37
+
+#define mmCLK1_CLK0_DFS_CNTL 0x16E69
+#define mmCLK1_CLK1_DFS_CNTL 0x16E6C
+#define mmCLK1_CLK2_DFS_CNTL 0x16E6F
+#define mmCLK1_CLK3_DFS_CNTL 0x16E72
+#define mmCLK1_CLK4_DFS_CNTL 0x16E75
+#define mmCLK1_CLK5_DFS_CNTL 0x16E78
+
+#define mmCLK1_CLK0_CURRENT_CNT 0x16EFB
+#define mmCLK1_CLK1_CURRENT_CNT 0x16EFC
+#define mmCLK1_CLK2_CURRENT_CNT 0x16EFD
+#define mmCLK1_CLK3_CURRENT_CNT 0x16EFE
+#define mmCLK1_CLK4_CURRENT_CNT 0x16EFF
+#define mmCLK1_CLK5_CURRENT_CNT 0x16F00
+
+#define mmCLK1_CLK0_BYPASS_CNTL 0x16E8A
+#define mmCLK1_CLK1_BYPASS_CNTL 0x16E93
+#define mmCLK1_CLK2_BYPASS_CNTL 0x16E9C
+#define mmCLK1_CLK3_BYPASS_CNTL 0x16EA5
+#define mmCLK1_CLK4_BYPASS_CNTL 0x16EAE
+#define mmCLK1_CLK5_BYPASS_CNTL 0x16EB7
+
+#define mmCLK1_CLK0_DS_CNTL 0x16E83
+#define mmCLK1_CLK1_DS_CNTL 0x16E8C
+#define mmCLK1_CLK2_DS_CNTL 0x16E95
+#define mmCLK1_CLK3_DS_CNTL 0x16E9E
+#define mmCLK1_CLK4_DS_CNTL 0x16EA7
+#define mmCLK1_CLK5_DS_CNTL 0x16EB0
+
+#define mmCLK1_CLK0_ALLOW_DS 0x16E84
+#define mmCLK1_CLK1_ALLOW_DS 0x16E8D
+#define mmCLK1_CLK2_ALLOW_DS 0x16E96
+#define mmCLK1_CLK3_ALLOW_DS 0x16E9F
+#define mmCLK1_CLK4_ALLOW_DS 0x16EA8
+#define mmCLK1_CLK5_ALLOW_DS 0x16EB1
+
+#define mmCLK5_spll_field_8 0x1B24B
+#define mmCLK6_spll_field_8 0x1B24B
+#define mmDENTIST_DISPCLK_CNTL 0x0124
+#define regDENTIST_DISPCLK_CNTL 0x0064
+#define regDENTIST_DISPCLK_CNTL_BASE_IDX 1
+
+#define CLK1_CLK_PLL_REQ__FbMult_int__SHIFT 0x0
+#define CLK1_CLK_PLL_REQ__PllSpineDiv__SHIFT 0xc
+#define CLK1_CLK_PLL_REQ__FbMult_frac__SHIFT 0x10
+#define CLK1_CLK_PLL_REQ__FbMult_int_MASK 0x000001FFL
+#define CLK1_CLK_PLL_REQ__PllSpineDiv_MASK 0x0000F000L
+#define CLK1_CLK_PLL_REQ__FbMult_frac_MASK 0xFFFF0000L
+
+#define CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_SEL_MASK 0x00000007L
+#define CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_DIV_MASK 0x000F0000L
+// DENTIST_DISPCLK_CNTL
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_WDIVIDER__SHIFT 0x0
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_RDIVIDER__SHIFT 0x8
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_CHG_DONE__SHIFT 0x13
+#define DENTIST_DISPCLK_CNTL__DENTIST_DPPCLK_CHG_DONE__SHIFT 0x14
+#define DENTIST_DISPCLK_CNTL__DENTIST_DPPCLK_WDIVIDER__SHIFT 0x18
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_WDIVIDER_MASK 0x0000007FL
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_RDIVIDER_MASK 0x00007F00L
+#define DENTIST_DISPCLK_CNTL__DENTIST_DISPCLK_CHG_DONE_MASK 0x00080000L
+#define DENTIST_DISPCLK_CNTL__DENTIST_DPPCLK_CHG_DONE_MASK 0x00100000L
+#define DENTIST_DISPCLK_CNTL__DENTIST_DPPCLK_WDIVIDER_MASK 0x7F000000L
+
+#define CLK5_spll_field_8__spll_ssc_en_MASK 0x00002000L
+#define CLK6_spll_field_8__spll_ssc_en_MASK 0x00002000L
-#define regCLK1_CLK_PLL_REQ 0x0237
-#define regCLK1_CLK_PLL_REQ_BASE_IDX 0
+#define SMU_VER_THRESHOLD 0x5D4A00 //93.74.0
+#undef FN
+#define FN(reg_name, field_name) \
+ clk_mgr->clk_mgr_shift->field_name, clk_mgr->clk_mgr_mask->field_name
-#define CLK1_CLK_PLL_REQ__FbMult_int__SHIFT 0x0
-#define CLK1_CLK_PLL_REQ__PllSpineDiv__SHIFT 0xc
-#define CLK1_CLK_PLL_REQ__FbMult_frac__SHIFT 0x10
-#define CLK1_CLK_PLL_REQ__FbMult_int_MASK 0x000001FFL
-#define CLK1_CLK_PLL_REQ__PllSpineDiv_MASK 0x0000F000L
-#define CLK1_CLK_PLL_REQ__FbMult_frac_MASK 0xFFFF0000L
+#define REG(reg) \
+ (clk_mgr->regs->reg)
-#define regCLK1_CLK2_BYPASS_CNTL 0x029c
-#define regCLK1_CLK2_BYPASS_CNTL_BASE_IDX 0
+#define BASE_INNER(seg) DCN_BASE__INST0_SEG ## seg
-#define CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_SEL__SHIFT 0x0
-#define CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_DIV__SHIFT 0x10
-#define CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_SEL_MASK 0x00000007L
-#define CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_DIV_MASK 0x000F0000L
+#define BASE(seg) BASE_INNER(seg)
-#define regCLK5_0_CLK5_spll_field_8 0x464b
-#define regCLK5_0_CLK5_spll_field_8_BASE_IDX 0
+#define SR(reg_name)\
+ .reg_name = BASE(reg ## reg_name ## _BASE_IDX) + \
+ reg ## reg_name
-#define CLK5_0_CLK5_spll_field_8__spll_ssc_en__SHIFT 0xd
-#define CLK5_0_CLK5_spll_field_8__spll_ssc_en_MASK 0x00002000L
+#define CLK_SR_DCN35(reg_name)\
+ .reg_name = mm ## reg_name
-#define SMU_VER_THRESHOLD 0x5D4A00 //93.74.0
+static const struct clk_mgr_registers clk_mgr_regs_dcn35 = {
+ CLK_REG_LIST_DCN35()
+};
-#define REG(reg_name) \
- (ctx->clk_reg_offsets[reg ## reg_name ## _BASE_IDX] + reg ## reg_name)
+static const struct clk_mgr_shift clk_mgr_shift_dcn35 = {
+ CLK_COMMON_MASK_SH_LIST_DCN32(__SHIFT)
+};
+
+static const struct clk_mgr_mask clk_mgr_mask_dcn35 = {
+ CLK_COMMON_MASK_SH_LIST_DCN32(_MASK)
+};
#define TO_CLK_MGR_DCN35(clk_mgr)\
container_of(clk_mgr, struct clk_mgr_dcn35, base)
@@ -138,34 +203,41 @@ static void dcn35_disable_otg_wa(struct clk_mgr *clk_mgr_base, struct dc_state *
struct pipe_ctx *pipe = safe_to_lower
? &context->res_ctx.pipe_ctx[i]
: &dc->current_state->res_ctx.pipe_ctx[i];
+ struct link_encoder *new_pipe_link_enc = new_pipe->link_res.dio_link_enc;
+ struct link_encoder *pipe_link_enc = pipe->link_res.dio_link_enc;
bool stream_changed_otg_dig_on = false;
+ bool has_active_hpo = false;
+
if (pipe->top_pipe || pipe->prev_odm_pipe)
continue;
+
+ if (!dc->config.unify_link_enc_assignment) {
+ if (new_pipe->stream)
+ new_pipe_link_enc = new_pipe->stream->link_enc;
+ if (pipe->stream)
+ pipe_link_enc = pipe->stream->link_enc;
+ }
+
stream_changed_otg_dig_on = old_pipe->stream && new_pipe->stream &&
old_pipe->stream != new_pipe->stream &&
old_pipe->stream_res.tg == new_pipe->stream_res.tg &&
- new_pipe->stream->link_enc && !new_pipe->stream->dpms_off &&
- new_pipe->stream->link_enc->funcs->is_dig_enabled &&
- new_pipe->stream->link_enc->funcs->is_dig_enabled(
- new_pipe->stream->link_enc) &&
+ new_pipe_link_enc && !new_pipe->stream->dpms_off &&
+ new_pipe_link_enc->funcs->is_dig_enabled &&
+ new_pipe_link_enc->funcs->is_dig_enabled(
+ new_pipe_link_enc) &&
new_pipe->stream_res.stream_enc &&
new_pipe->stream_res.stream_enc->funcs->is_fifo_enabled &&
new_pipe->stream_res.stream_enc->funcs->is_fifo_enabled(new_pipe->stream_res.stream_enc);
- bool has_active_hpo = false;
-
if (old_pipe->stream && new_pipe->stream && old_pipe->stream == new_pipe->stream) {
has_active_hpo = dccg->ctx->dc->link_srv->dp_is_128b_132b_signal(old_pipe) &&
dccg->ctx->dc->link_srv->dp_is_128b_132b_signal(new_pipe);
- }
-
-
- if (!has_active_hpo && !dccg->ctx->dc->link_srv->dp_is_128b_132b_signal(pipe) &&
- (pipe->stream && (pipe->stream->dpms_off || dc_is_virtual_signal(pipe->stream->signal) ||
- !pipe->stream->link_enc) && !stream_changed_otg_dig_on)) {
-
+ }
+ if (!has_active_hpo && !stream_changed_otg_dig_on && pipe->stream &&
+ (pipe->stream->dpms_off || dc_is_virtual_signal(pipe->stream->signal) || !pipe_link_enc) &&
+ !dccg->ctx->dc->link_srv->dp_is_128b_132b_signal(pipe)) {
/* This w/a should not trigger when we have a dig active */
if (disable) {
if (pipe->stream_res.tg && pipe->stream_res.tg->funcs->immediate_disable_crtc)
@@ -338,6 +410,7 @@ void dcn35_update_clocks(struct clk_mgr *clk_mgr_base,
if (clk_mgr_base->clks.dtbclk_en && !new_clocks->dtbclk_en) {
if (clk_mgr->base.ctx->dc->config.allow_0_dtb_clk)
dcn35_smu_set_dtbclk(clk_mgr, false);
+
clk_mgr_base->clks.dtbclk_en = new_clocks->dtbclk_en;
}
/* check that we're not already in lower */
@@ -355,11 +428,17 @@ void dcn35_update_clocks(struct clk_mgr *clk_mgr_base,
}
if (!clk_mgr_base->clks.dtbclk_en && new_clocks->dtbclk_en) {
- dcn35_smu_set_dtbclk(clk_mgr, true);
- clk_mgr_base->clks.dtbclk_en = new_clocks->dtbclk_en;
+ int actual_dtbclk = 0;
dcn35_update_clocks_update_dtb_dto(clk_mgr, context, new_clocks->ref_dtbclk_khz);
- clk_mgr_base->clks.ref_dtbclk_khz = new_clocks->ref_dtbclk_khz;
+ dcn35_smu_set_dtbclk(clk_mgr, true);
+
+ actual_dtbclk = REG_READ(CLK1_CLK4_CURRENT_CNT);
+
+ if (actual_dtbclk) {
+ clk_mgr_base->clks.ref_dtbclk_khz = new_clocks->ref_dtbclk_khz;
+ clk_mgr_base->clks.dtbclk_en = new_clocks->dtbclk_en;
+ }
}
/* check that we're not already in D0 */
@@ -397,14 +476,19 @@ void dcn35_update_clocks(struct clk_mgr *clk_mgr_base,
update_dppclk = true;
}
- if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)) {
+ if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz) &&
+ (new_clocks->dispclk_khz > 0 || (safe_to_lower && display_count == 0))) {
+ int requested_dispclk_khz = new_clocks->dispclk_khz;
+
dcn35_disable_otg_wa(clk_mgr_base, context, safe_to_lower, true);
- if (dc->debug.min_disp_clk_khz > 0 && new_clocks->dispclk_khz < dc->debug.min_disp_clk_khz)
- new_clocks->dispclk_khz = dc->debug.min_disp_clk_khz;
+ /* Clamp the requested clock to PMFW based on their limit. */
+ if (dc->debug.min_disp_clk_khz > 0 && requested_dispclk_khz < dc->debug.min_disp_clk_khz)
+ requested_dispclk_khz = dc->debug.min_disp_clk_khz;
+ dcn35_smu_set_dispclk(clk_mgr, requested_dispclk_khz);
clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz;
- dcn35_smu_set_dispclk(clk_mgr, clk_mgr_base->clks.dispclk_khz);
+
dcn35_disable_otg_wa(clk_mgr_base, context, safe_to_lower, false);
update_dispclk = true;
@@ -452,7 +536,6 @@ static int get_vco_frequency_from_reg(struct clk_mgr_internal *clk_mgr)
struct fixed31_32 pll_req;
unsigned int fbmult_frac_val = 0;
unsigned int fbmult_int_val = 0;
- struct dc_context *ctx = clk_mgr->base.ctx;
/*
* Register value of fbmult is in 8.16 format, we are converting to 314.32
@@ -512,22 +595,24 @@ static void dcn35_dump_clk_registers(struct clk_state_registers_and_bypass *regs
static bool dcn35_is_spll_ssc_enabled(struct clk_mgr *clk_mgr_base)
{
struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
- struct dc_context *ctx = clk_mgr->base.ctx;
+
uint32_t ssc_enable;
- REG_GET(CLK5_0_CLK5_spll_field_8, spll_ssc_en, &ssc_enable);
+ if (clk_mgr_base->ctx->dce_version == DCN_VERSION_3_51) {
+ ssc_enable = REG_READ(CLK6_spll_field_8) & CLK6_spll_field_8__spll_ssc_en_MASK;
+ } else {
+ ssc_enable = REG_READ(CLK5_spll_field_8) & CLK5_spll_field_8__spll_ssc_en_MASK;
+ }
- return ssc_enable == 1;
+ return ssc_enable != 0;
}
static void init_clk_states(struct clk_mgr *clk_mgr)
{
- struct clk_mgr_internal *clk_mgr_int = TO_CLK_MGR_INTERNAL(clk_mgr);
uint32_t ref_dtbclk = clk_mgr->clks.ref_dtbclk_khz;
+
memset(&(clk_mgr->clks), 0, sizeof(struct dc_clocks));
- if (clk_mgr_int->smu_ver >= SMU_VER_THRESHOLD)
- clk_mgr->clks.dtbclk_en = true; // request DTBCLK disable on first commit
clk_mgr->clks.ref_dtbclk_khz = ref_dtbclk; // restore ref_dtbclk
clk_mgr->clks.p_state_change_support = true;
clk_mgr->clks.prev_p_state_change_support = true;
@@ -538,6 +623,7 @@ static void init_clk_states(struct clk_mgr *clk_mgr)
void dcn35_init_clocks(struct clk_mgr *clk_mgr)
{
struct clk_mgr_internal *clk_mgr_int = TO_CLK_MGR_INTERNAL(clk_mgr);
+
init_clk_states(clk_mgr);
// to adjust dp_dto reference clock if ssc is enable otherwise to apply dprefclk
@@ -632,6 +718,7 @@ static struct wm_table lpddr5_wm_table = {
};
static DpmClocks_t_dcn35 dummy_clocks;
+static DpmClocks_t_dcn351 dummy_clocks_dcn351;
static struct dcn35_watermarks dummy_wms = { 0 };
@@ -642,10 +729,10 @@ static struct dcn35_ss_info_table ss_info_table = {
static void dcn35_read_ss_info_from_lut(struct clk_mgr_internal *clk_mgr)
{
- struct dc_context *ctx = clk_mgr->base.ctx;
- uint32_t clock_source;
+ uint32_t clock_source = 0;
+
+ clock_source = REG_READ(CLK1_CLK2_BYPASS_CNTL) & CLK1_CLK2_BYPASS_CNTL__CLK2_BYPASS_SEL_MASK;
- REG_GET(CLK1_CLK2_BYPASS_CNTL, CLK2_BYPASS_SEL, &clock_source);
// If it's DFS mode, clock_source is 0.
if (dcn35_is_spll_ssc_enabled(&clk_mgr->base) && (clock_source < ARRAY_SIZE(ss_info_table.ss_percentage))) {
clk_mgr->dprefclk_ss_percentage = ss_info_table.ss_percentage[clock_source];
@@ -755,6 +842,22 @@ static void dcn35_get_dpm_table_from_smu(struct clk_mgr_internal *clk_mgr,
dcn35_smu_transfer_dpm_table_smu_2_dram(clk_mgr);
}
+static void dcn351_get_dpm_table_from_smu(struct clk_mgr_internal *clk_mgr,
+ struct dcn351_smu_dpm_clks *smu_dpm_clks)
+{
+ DpmClocks_t_dcn351 *table = smu_dpm_clks->dpm_clks;
+
+ if (!clk_mgr->smu_ver)
+ return;
+ if (!table || smu_dpm_clks->mc_address.quad_part == 0)
+ return;
+ memset(table, 0, sizeof(*table));
+ dcn35_smu_set_dram_addr_high(clk_mgr,
+ smu_dpm_clks->mc_address.high_part);
+ dcn35_smu_set_dram_addr_low(clk_mgr,
+ smu_dpm_clks->mc_address.low_part);
+ dcn35_smu_transfer_dpm_table_smu_2_dram(clk_mgr);
+}
static uint32_t find_max_clk_value(const uint32_t clocks[], uint32_t num_clocks)
{
uint32_t max = 0;
@@ -1093,6 +1196,57 @@ struct clk_mgr_funcs dcn35_fpga_funcs = {
.get_dtb_ref_clk_frequency = dcn31_get_dtb_ref_freq_khz,
};
+static void translate_to_DpmClocks_t_dcn35(struct dcn351_smu_dpm_clks *smu_dpm_clks_a,
+ struct dcn35_smu_dpm_clks *smu_dpm_clks_b)
+{
+ /*translate two structures and only take need clock tables*/
+ uint8_t i;
+
+ if (smu_dpm_clks_a == NULL || smu_dpm_clks_b == NULL ||
+ smu_dpm_clks_a->dpm_clks == NULL || smu_dpm_clks_b->dpm_clks == NULL)
+ return;
+
+ for (i = 0; i < NUM_DCFCLK_DPM_LEVELS; i++)
+ smu_dpm_clks_b->dpm_clks->DcfClocks[i] = smu_dpm_clks_a->dpm_clks->DcfClocks[i];
+
+ for (i = 0; i < NUM_DISPCLK_DPM_LEVELS; i++)
+ smu_dpm_clks_b->dpm_clks->DispClocks[i] = smu_dpm_clks_a->dpm_clks->DispClocks[i];
+
+ for (i = 0; i < NUM_DPPCLK_DPM_LEVELS; i++)
+ smu_dpm_clks_b->dpm_clks->DppClocks[i] = smu_dpm_clks_a->dpm_clks->DppClocks[i];
+
+ for (i = 0; i < NUM_FCLK_DPM_LEVELS; i++) {
+ smu_dpm_clks_b->dpm_clks->FclkClocks_Freq[i] = smu_dpm_clks_a->dpm_clks->FclkClocks_Freq[i];
+ smu_dpm_clks_b->dpm_clks->FclkClocks_Voltage[i] = smu_dpm_clks_a->dpm_clks->FclkClocks_Voltage[i];
+ }
+ for (i = 0; i < NUM_MEM_PSTATE_LEVELS; i++) {
+ smu_dpm_clks_b->dpm_clks->MemPstateTable[i].MemClk =
+ smu_dpm_clks_a->dpm_clks->MemPstateTable[i].MemClk;
+ smu_dpm_clks_b->dpm_clks->MemPstateTable[i].UClk =
+ smu_dpm_clks_a->dpm_clks->MemPstateTable[i].UClk;
+ smu_dpm_clks_b->dpm_clks->MemPstateTable[i].Voltage =
+ smu_dpm_clks_a->dpm_clks->MemPstateTable[i].Voltage;
+ smu_dpm_clks_b->dpm_clks->MemPstateTable[i].WckRatio =
+ smu_dpm_clks_a->dpm_clks->MemPstateTable[i].WckRatio;
+ }
+ smu_dpm_clks_b->dpm_clks->MaxGfxClk = smu_dpm_clks_a->dpm_clks->MaxGfxClk;
+ smu_dpm_clks_b->dpm_clks->MinGfxClk = smu_dpm_clks_a->dpm_clks->MinGfxClk;
+ smu_dpm_clks_b->dpm_clks->NumDcfClkLevelsEnabled =
+ smu_dpm_clks_a->dpm_clks->NumDcfClkLevelsEnabled;
+ smu_dpm_clks_b->dpm_clks->NumDispClkLevelsEnabled =
+ smu_dpm_clks_a->dpm_clks->NumDispClkLevelsEnabled;
+ smu_dpm_clks_b->dpm_clks->NumFclkLevelsEnabled =
+ smu_dpm_clks_a->dpm_clks->NumFclkLevelsEnabled;
+ smu_dpm_clks_b->dpm_clks->NumMemPstatesEnabled =
+ smu_dpm_clks_a->dpm_clks->NumMemPstatesEnabled;
+ smu_dpm_clks_b->dpm_clks->NumSocClkLevelsEnabled =
+ smu_dpm_clks_a->dpm_clks->NumSocClkLevelsEnabled;
+
+ for (i = 0; i < NUM_SOC_VOLTAGE_LEVELS; i++) {
+ smu_dpm_clks_b->dpm_clks->SocClocks[i] = smu_dpm_clks_a->dpm_clks->SocClocks[i];
+ smu_dpm_clks_b->dpm_clks->SocVoltage[i] = smu_dpm_clks_a->dpm_clks->SocVoltage[i];
+ }
+}
void dcn35_clk_mgr_construct(
struct dc_context *ctx,
struct clk_mgr_dcn35 *clk_mgr,
@@ -1100,6 +1254,7 @@ void dcn35_clk_mgr_construct(
struct dccg *dccg)
{
struct dcn35_smu_dpm_clks smu_dpm_clks = { 0 };
+ struct dcn351_smu_dpm_clks smu_dpm_clks_dcn351 = { 0 };
clk_mgr->base.base.ctx = ctx;
clk_mgr->base.base.funcs = &dcn35_funcs;
@@ -1112,6 +1267,12 @@ void dcn35_clk_mgr_construct(
clk_mgr->base.dprefclk_ss_divider = 1000;
clk_mgr->base.ss_on_dprefclk = false;
clk_mgr->base.dfs_ref_freq_khz = 48000;
+ if (ctx->dce_version != DCN_VERSION_3_51) {
+ clk_mgr->base.regs = &clk_mgr_regs_dcn35;
+ clk_mgr->base.clk_mgr_shift = &clk_mgr_shift_dcn35;
+ clk_mgr->base.clk_mgr_mask = &clk_mgr_mask_dcn35;
+ }
+
clk_mgr->smu_wm_set.wm_set = (struct dcn35_watermarks *)dm_helpers_allocate_gpu_mem(
clk_mgr->base.base.ctx,
@@ -1130,14 +1291,24 @@ void dcn35_clk_mgr_construct(
DC_MEM_ALLOC_TYPE_GART,
sizeof(DpmClocks_t_dcn35),
&smu_dpm_clks.mc_address.quad_part);
-
if (smu_dpm_clks.dpm_clks == NULL) {
smu_dpm_clks.dpm_clks = &dummy_clocks;
smu_dpm_clks.mc_address.quad_part = 0;
}
-
ASSERT(smu_dpm_clks.dpm_clks);
+ if (ctx->dce_version == DCN_VERSION_3_51) {
+ smu_dpm_clks_dcn351.dpm_clks = (DpmClocks_t_dcn351 *)dm_helpers_allocate_gpu_mem(
+ clk_mgr->base.base.ctx,
+ DC_MEM_ALLOC_TYPE_GART,
+ sizeof(DpmClocks_t_dcn351),
+ &smu_dpm_clks_dcn351.mc_address.quad_part);
+ if (smu_dpm_clks_dcn351.dpm_clks == NULL) {
+ smu_dpm_clks_dcn351.dpm_clks = &dummy_clocks_dcn351;
+ smu_dpm_clks_dcn351.mc_address.quad_part = 0;
+ }
+ }
+
clk_mgr->base.smu_ver = dcn35_smu_get_smu_version(&clk_mgr->base);
if (clk_mgr->base.smu_ver)
@@ -1166,7 +1337,11 @@ void dcn35_clk_mgr_construct(
if (clk_mgr->base.base.ctx->dc->debug.pstate_enabled) {
int i;
- dcn35_get_dpm_table_from_smu(&clk_mgr->base, &smu_dpm_clks);
+ if (ctx->dce_version == DCN_VERSION_3_51) {
+ dcn351_get_dpm_table_from_smu(&clk_mgr->base, &smu_dpm_clks_dcn351);
+ translate_to_DpmClocks_t_dcn35(&smu_dpm_clks_dcn351, &smu_dpm_clks);
+ } else
+ dcn35_get_dpm_table_from_smu(&clk_mgr->base, &smu_dpm_clks);
DC_LOG_SMU("NumDcfClkLevelsEnabled: %d\n"
"NumDispClkLevelsEnabled: %d\n"
"NumSocClkLevelsEnabled: %d\n"
@@ -1227,6 +1402,10 @@ void dcn35_clk_mgr_construct(
dm_helpers_free_gpu_mem(clk_mgr->base.base.ctx, DC_MEM_ALLOC_TYPE_GART,
smu_dpm_clks.dpm_clks);
+ if (smu_dpm_clks_dcn351.dpm_clks && smu_dpm_clks_dcn351.mc_address.quad_part != 0)
+ dm_helpers_free_gpu_mem(clk_mgr->base.base.ctx, DC_MEM_ALLOC_TYPE_GART,
+ smu_dpm_clks_dcn351.dpm_clks);
+
if (ctx->dc->config.disable_ips != DMUB_IPS_DISABLE_ALL) {
bool ips_support = false;
@@ -1240,7 +1419,7 @@ void dcn35_clk_mgr_construct(
/* Disable dynamic IPS2 in older PMFW (93.12) for Z8 interop. */
if (ctx->dc->config.disable_ips == DMUB_IPS_ENABLE &&
- ctx->dce_version == DCN_VERSION_3_5 &&
+ ctx->dce_version != DCN_VERSION_3_51 &&
((clk_mgr->base.smu_ver & 0x00FFFFFF) <= 0x005d0c00))
ctx->dc->config.disable_ips = DMUB_IPS_RCG_IN_ACTIVE_IPS2_IN_OFF;
} else {
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.h
index 1203dc605b12..a12a9bf90806 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.h
@@ -60,4 +60,8 @@ void dcn35_clk_mgr_construct(struct dc_context *ctx,
void dcn35_clk_mgr_destroy(struct clk_mgr_internal *clk_mgr_int);
+void dcn351_clk_mgr_construct(struct dc_context *ctx,
+ struct clk_mgr_dcn35 *clk_mgr,
+ struct pp_smu_funcs *pp_smu,
+ struct dccg *dccg);
#endif //__DCN35_CLK_MGR_H__
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c
index f6f0e6a33001..604d256cb47a 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c
@@ -84,8 +84,8 @@
#define VBIOSSMC_MSG_AllowZstatesEntry 0x15
#define VBIOSSMC_MSG_DisallowZstatesEntry 0x16
#define VBIOSSMC_MSG_SetDtbClk 0x17
-#define VBIOSSMC_MSG_DispPsrEntry 0x18 ///< Display PSR entry, DMU
-#define VBIOSSMC_MSG_DispPsrExit 0x19 ///< Display PSR exit, DMU
+#define VBIOSSMC_MSG_DispIPS2Entry 0x18 ///< Display IPS2 entry, DMU
+#define VBIOSSMC_MSG_DispIPS2Exit 0x19 ///< Display IPS2 exit, DMU
#define VBIOSSMC_MSG_DisableLSdma 0x1A ///< Disable LSDMA; only sent by VBIOS
#define VBIOSSMC_MSG_DpControllerPhyStatus 0x1B ///< Inform PMFW about the pre conditions for turning SLDO2 on/off . bit[0]==1 precondition is met, bit[1-2] are for DPPHY number
#define VBIOSSMC_MSG_QueryIPS2Support 0x1C ///< Return 1: support; else not supported
@@ -475,7 +475,7 @@ int dcn35_smu_exit_low_power_state(struct clk_mgr_internal *clk_mgr)
retv = dcn35_smu_send_msg_with_param(
clk_mgr,
- VBIOSSMC_MSG_DispPsrExit,
+ VBIOSSMC_MSG_DispIPS2Exit,
0);
smu_print("%s: smu_exit_low_power_state return = %d\n", __func__, retv);
return retv;
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h
index 3fae13c73934..ab9d21ba0c43 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h
@@ -126,18 +126,31 @@ typedef struct {
uint32_t MaxGfxClk;
} DpmClocks_t_dcn35;
-
-// Throttler Status Bitmask
-
-
-
-
-
-
-
-
-
-
+typedef struct {
+ uint32_t DcfClocks[NUM_DCFCLK_DPM_LEVELS];
+ uint32_t DispClocks[NUM_DISPCLK_DPM_LEVELS];
+ uint32_t DppClocks[NUM_DPPCLK_DPM_LEVELS];
+ uint32_t SocClocks[NUM_SOCCLK_DPM_LEVELS];
+ uint32_t VClocks0[NUM_VCN_DPM_LEVELS];
+ uint32_t VClocks1[NUM_VCN_DPM_LEVELS];
+ uint32_t DClocks0[NUM_VCN_DPM_LEVELS];
+ uint32_t DClocks1[NUM_VCN_DPM_LEVELS];
+ uint32_t VPEClocks[NUM_VPE_DPM_LEVELS];
+ uint32_t FclkClocks_Freq[NUM_FCLK_DPM_LEVELS];
+ uint32_t FclkClocks_Voltage[NUM_FCLK_DPM_LEVELS];
+ uint32_t SocVoltage[NUM_SOC_VOLTAGE_LEVELS];
+ MemPstateTable_t MemPstateTable[NUM_MEM_PSTATE_LEVELS];
+ uint8_t NumDcfClkLevelsEnabled;
+ uint8_t NumDispClkLevelsEnabled; // Applies to both Dispclk and Dppclk
+ uint8_t NumSocClkLevelsEnabled;
+ uint8_t Vcn0ClkLevelsEnabled; // Applies to both Vclk0 and Dclk0
+ uint8_t Vcn1ClkLevelsEnabled; // Applies to both Vclk1 and Dclk1
+ uint8_t VpeClkLevelsEnabled;
+ uint8_t NumMemPstatesEnabled;
+ uint8_t NumFclkLevelsEnabled;
+ uint32_t MinGfxClk;
+ uint32_t MaxGfxClk;
+} DpmClocks_t_dcn351;
#define TABLE_BIOS_IF 0 // Called by BIOS
#define TABLE_WATERMARKS 1 // Called by DAL through VBIOS
@@ -163,6 +176,10 @@ struct dcn35_smu_dpm_clks {
union large_integer mc_address;
};
+struct dcn351_smu_dpm_clks {
+ DpmClocks_t_dcn351 *dpm_clks;
+ union large_integer mc_address;
+};
/* TODO: taken from vgh, may not be correct */
struct display_idle_optimization {
unsigned int df_request_disabled : 1;
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dalsmc.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dalsmc.h
index dbfdd3487da5..2e0d34fd7512 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dalsmc.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dalsmc.h
@@ -43,7 +43,9 @@
#define DALSMC_MSG_ActiveUclkFclk 0x18
#define DALSMC_MSG_IdleUclkFclk 0x19
#define DALSMC_MSG_SetUclkPstateAllow 0x1A
-#define DALSMC_Message_Count 0x1B
+#define DALSMC_MSG_SubvpUclkFclk 0x1B
+#define DALSMC_MSG_GetNumUmcChannels 0x1C
+#define DALSMC_Message_Count 0x1D
typedef enum {
FCLK_SWITCH_DISALLOW,
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.c
index 8cfc5f435937..b59703467128 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.c
@@ -22,7 +22,7 @@
#include "dcn/dcn_4_1_0_offset.h"
#include "dcn/dcn_4_1_0_sh_mask.h"
-#include "dml/dcn401/dcn401_fpu.h"
+#define DCN_BASE__INST0_SEG1 0x000000C0
#define mmCLK01_CLK0_CLK_PLL_REQ 0x16E37
#define mmCLK01_CLK0_CLK0_DFS_CNTL 0x16E69
@@ -141,6 +141,20 @@ static bool dcn401_is_ppclk_idle_dpm_enabled(struct clk_mgr_internal *clk_mgr, P
return ppclk_idle_dpm_enabled;
}
+static bool dcn401_is_df_throttle_opt_enabled(struct clk_mgr_internal *clk_mgr)
+{
+ bool is_df_throttle_opt_enabled = false;
+
+ if (ASICREV_IS_GC_12_0_1_A0(clk_mgr->base.ctx->asic_id.hw_internal_rev) &&
+ clk_mgr->smu_ver >= 0x663500) {
+ is_df_throttle_opt_enabled = !clk_mgr->base.ctx->dc->debug.force_subvp_df_throttle;
+ }
+
+ is_df_throttle_opt_enabled &= clk_mgr->smu_present;
+
+ return is_df_throttle_opt_enabled;
+}
+
/* Query SMU for all clock states for a particular clock */
static void dcn401_init_single_clock(struct clk_mgr_internal *clk_mgr, PPCLK_e clk, unsigned int *entry_0,
unsigned int *num_levels)
@@ -167,43 +181,36 @@ static void dcn401_init_single_clock(struct clk_mgr_internal *clk_mgr, PPCLK_e c
static void dcn401_build_wm_range_table(struct clk_mgr *clk_mgr)
{
- /* legacy */
- DC_FP_START();
- dcn401_build_wm_range_table_fpu(clk_mgr);
- DC_FP_END();
-
- if (clk_mgr->ctx->dc->debug.using_dml21) {
- /* For min clocks use as reported by PM FW and report those as min */
- uint16_t min_uclk_mhz = clk_mgr->bw_params->clk_table.entries[0].memclk_mhz;
- uint16_t min_dcfclk_mhz = clk_mgr->bw_params->clk_table.entries[0].dcfclk_mhz;
-
- /* Set A - Normal - default values */
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].valid = true;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.wm_type = WATERMARKS_CLOCK_RANGE;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.min_dcfclk = min_dcfclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.max_dcfclk = 0xFFFF;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.min_uclk = min_uclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.max_uclk = 0xFFFF;
-
- /* Set B - Unused on dcn4 */
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].valid = false;
-
- /* Set 1A - Dummy P-State - P-State latency set to "dummy p-state" value */
- /* 'DalDummyClockChangeLatencyNs' registry key option set to 0x7FFFFFFF can be used to disable Set C for dummy p-state */
- if (clk_mgr->ctx->dc->bb_overrides.dummy_clock_change_latency_ns != 0x7FFFFFFF) {
- clk_mgr->bw_params->wm_table.nv_entries[WM_1A].valid = true;
- clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.wm_type = WATERMARKS_DUMMY_PSTATE;
- clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.min_dcfclk = min_dcfclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.max_dcfclk = 0xFFFF;
- clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.min_uclk = min_uclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.max_uclk = 0xFFFF;
- } else {
- clk_mgr->bw_params->wm_table.nv_entries[WM_1A].valid = false;
- }
-
- /* Set 1B - Unused on dcn4 */
- clk_mgr->bw_params->wm_table.nv_entries[WM_1B].valid = false;
+ /* For min clocks use as reported by PM FW and report those as min */
+ uint16_t min_uclk_mhz = clk_mgr->bw_params->clk_table.entries[0].memclk_mhz;
+ uint16_t min_dcfclk_mhz = clk_mgr->bw_params->clk_table.entries[0].dcfclk_mhz;
+
+ /* Set A - Normal - default values */
+ clk_mgr->bw_params->wm_table.nv_entries[WM_A].valid = true;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.wm_type = WATERMARKS_CLOCK_RANGE;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.min_dcfclk = min_dcfclk_mhz;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.max_dcfclk = 0xFFFF;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.min_uclk = min_uclk_mhz;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.max_uclk = 0xFFFF;
+
+ /* Set B - Unused on dcn4 */
+ clk_mgr->bw_params->wm_table.nv_entries[WM_B].valid = false;
+
+ /* Set 1A - Dummy P-State - P-State latency set to "dummy p-state" value */
+ /* 'DalDummyClockChangeLatencyNs' registry key option set to 0x7FFFFFFF can be used to disable Set C for dummy p-state */
+ if (clk_mgr->ctx->dc->bb_overrides.dummy_clock_change_latency_ns != 0x7FFFFFFF) {
+ clk_mgr->bw_params->wm_table.nv_entries[WM_1A].valid = true;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.wm_type = WATERMARKS_DUMMY_PSTATE;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.min_dcfclk = min_dcfclk_mhz;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.max_dcfclk = 0xFFFF;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.min_uclk = min_uclk_mhz;
+ clk_mgr->bw_params->wm_table.nv_entries[WM_1A].pmfw_breakdown.max_uclk = 0xFFFF;
+ } else {
+ clk_mgr->bw_params->wm_table.nv_entries[WM_1A].valid = false;
}
+
+ /* Set 1B - Unused on dcn4 */
+ clk_mgr->bw_params->wm_table.nv_entries[WM_1B].valid = false;
}
void dcn401_init_clocks(struct clk_mgr *clk_mgr_base)
@@ -304,6 +311,25 @@ void dcn401_init_clocks(struct clk_mgr *clk_mgr_base)
dcn401_build_wm_range_table(clk_mgr_base);
}
+bool dcn401_is_dc_mode_present(struct clk_mgr *clk_mgr_base)
+{
+ struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
+
+ return clk_mgr->smu_present && clk_mgr->dpm_present &&
+ ((clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_dcfclk_levels &&
+ clk_mgr_base->bw_params->dc_mode_limit.dcfclk_mhz) ||
+ (clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_dispclk_levels &&
+ clk_mgr_base->bw_params->dc_mode_limit.dispclk_mhz) ||
+ (clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_dtbclk_levels &&
+ clk_mgr_base->bw_params->dc_mode_limit.dtbclk_mhz) ||
+ (clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_fclk_levels &&
+ clk_mgr_base->bw_params->dc_mode_limit.fclk_mhz) ||
+ (clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_memclk_levels &&
+ clk_mgr_base->bw_params->dc_mode_limit.memclk_mhz) ||
+ (clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_socclk_levels &&
+ clk_mgr_base->bw_params->dc_mode_limit.socclk_mhz));
+}
+
static void dcn401_dump_clk_registers(struct clk_state_registers_and_bypass *regs_and_bypass,
struct clk_mgr *clk_mgr_base, struct clk_log_info *log_info)
{
@@ -614,207 +640,6 @@ static void dcn401_update_clocks_update_dentist(
}
-static void dcn401_update_clocks_legacy(struct clk_mgr *clk_mgr_base,
- struct dc_state *context,
- bool safe_to_lower)
-{
- struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
- struct dc_clocks *new_clocks = &context->bw_ctx.bw.dcn.clk;
- struct dc *dc = clk_mgr_base->ctx->dc;
- int display_count;
- bool update_dppclk = false;
- bool update_dispclk = false;
- bool enter_display_off = false;
- bool dpp_clock_lowered = false;
- struct dmcu *dmcu = clk_mgr_base->ctx->dc->res_pool->dmcu;
- bool force_reset = false;
- bool update_uclk = false, update_fclk = false;
- bool p_state_change_support;
- bool fclk_p_state_change_support;
- int total_plane_count;
-
- if (dc->work_arounds.skip_clock_update)
- return;
-
- if (clk_mgr_base->clks.dispclk_khz == 0 ||
- (dc->debug.force_clock_mode & 0x1)) {
- /* This is from resume or boot up, if forced_clock cfg option used,
- * we bypass program dispclk and DPPCLK, but need set them for S3.
- */
- force_reset = true;
-
- dcn2_read_clocks_from_hw_dentist(clk_mgr_base);
-
- /* Force_clock_mode 0x1: force reset the clock even it is the same clock
- * as long as it is in Passive level.
- */
- }
- display_count = clk_mgr_helper_get_active_display_cnt(dc, context);
-
- if (display_count == 0)
- enter_display_off = true;
-
- if (clk_mgr->smu_present) {
- if (enter_display_off == safe_to_lower)
- dcn401_smu_set_num_of_displays(clk_mgr, display_count);
-
- clk_mgr_base->clks.fclk_prev_p_state_change_support = clk_mgr_base->clks.fclk_p_state_change_support;
-
- total_plane_count = clk_mgr_helper_get_active_plane_cnt(dc, context);
- fclk_p_state_change_support = new_clocks->fclk_p_state_change_support || (total_plane_count == 0);
-
- if (should_update_pstate_support(safe_to_lower, fclk_p_state_change_support, clk_mgr_base->clks.fclk_p_state_change_support)) {
- clk_mgr_base->clks.fclk_p_state_change_support = fclk_p_state_change_support;
-
- /* To enable FCLK P-state switching, send PSTATE_SUPPORTED message to PMFW */
- if (clk_mgr_base->clks.fclk_p_state_change_support) {
- /* Handle the code for sending a message to PMFW that FCLK P-state change is supported */
- dcn401_smu_send_fclk_pstate_message(clk_mgr, true);
- }
- }
-
- if (dc->debug.force_min_dcfclk_mhz > 0)
- new_clocks->dcfclk_khz = (new_clocks->dcfclk_khz > (dc->debug.force_min_dcfclk_mhz * 1000)) ?
- new_clocks->dcfclk_khz : (dc->debug.force_min_dcfclk_mhz * 1000);
-
- if (should_set_clock(safe_to_lower, new_clocks->dcfclk_khz, clk_mgr_base->clks.dcfclk_khz)) {
- clk_mgr_base->clks.dcfclk_khz = new_clocks->dcfclk_khz;
- if (dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DCFCLK))
- dcn401_smu_set_hard_min_by_freq(clk_mgr, PPCLK_DCFCLK, khz_to_mhz_ceil(clk_mgr_base->clks.dcfclk_khz));
- }
-
- if (should_set_clock(safe_to_lower, new_clocks->dcfclk_deep_sleep_khz, clk_mgr_base->clks.dcfclk_deep_sleep_khz)) {
- clk_mgr_base->clks.dcfclk_deep_sleep_khz = new_clocks->dcfclk_deep_sleep_khz;
- if (dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DCFCLK))
- dcn401_smu_set_min_deep_sleep_dcef_clk(clk_mgr, khz_to_mhz_ceil(clk_mgr_base->clks.dcfclk_deep_sleep_khz));
- }
-
- if (should_set_clock(safe_to_lower, new_clocks->socclk_khz, clk_mgr_base->clks.socclk_khz))
- /* We don't actually care about socclk, don't notify SMU of hard min */
- clk_mgr_base->clks.socclk_khz = new_clocks->socclk_khz;
-
- clk_mgr_base->clks.prev_p_state_change_support = clk_mgr_base->clks.p_state_change_support;
- clk_mgr_base->clks.prev_num_ways = clk_mgr_base->clks.num_ways;
-
- if (clk_mgr_base->clks.num_ways != new_clocks->num_ways &&
- clk_mgr_base->clks.num_ways < new_clocks->num_ways) {
- clk_mgr_base->clks.num_ways = new_clocks->num_ways;
- if (dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_UCLK))
- dcn401_smu_send_cab_for_uclk_message(clk_mgr, clk_mgr_base->clks.num_ways);
- }
-
-
- p_state_change_support = new_clocks->p_state_change_support || (total_plane_count == 0);
- if (should_update_pstate_support(safe_to_lower, p_state_change_support, clk_mgr_base->clks.prev_p_state_change_support)) {
- clk_mgr_base->clks.p_state_change_support = p_state_change_support;
- clk_mgr_base->clks.fw_based_mclk_switching = p_state_change_support && new_clocks->fw_based_mclk_switching;
-
- /* to disable P-State switching, set UCLK min = max */
- if (!clk_mgr_base->clks.p_state_change_support && dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_UCLK))
- dcn401_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK,
- clk_mgr_base->bw_params->clk_table.entries[clk_mgr_base->bw_params->clk_table.num_entries_per_clk.num_memclk_levels - 1].memclk_mhz);
- }
-
- /* Always update saved value, even if new value not set due to P-State switching unsupported. Also check safe_to_lower for FCLK */
- if (safe_to_lower && (clk_mgr_base->clks.fclk_p_state_change_support != clk_mgr_base->clks.fclk_prev_p_state_change_support)) {
- update_fclk = true;
- }
-
- if (!clk_mgr_base->clks.fclk_p_state_change_support &&
- update_fclk &&
- dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_FCLK)) {
- /* Handle code for sending a message to PMFW that FCLK P-state change is not supported */
- dcn401_smu_send_fclk_pstate_message(clk_mgr, false);
- }
-
- /* Always update saved value, even if new value not set due to P-State switching unsupported */
- if (should_set_clock(safe_to_lower, new_clocks->dramclk_khz, clk_mgr_base->clks.dramclk_khz)) {
- clk_mgr_base->clks.dramclk_khz = new_clocks->dramclk_khz;
- update_uclk = true;
- }
-
- /* set UCLK to requested value if P-State switching is supported, or to re-enable P-State switching */
- if (clk_mgr_base->clks.p_state_change_support &&
- (update_uclk || !clk_mgr_base->clks.prev_p_state_change_support) &&
- dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_UCLK))
- dcn401_smu_set_hard_min_by_freq(clk_mgr, PPCLK_UCLK, khz_to_mhz_ceil(clk_mgr_base->clks.dramclk_khz));
-
- if (clk_mgr_base->clks.num_ways != new_clocks->num_ways &&
- clk_mgr_base->clks.num_ways > new_clocks->num_ways) {
- clk_mgr_base->clks.num_ways = new_clocks->num_ways;
- if (dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_UCLK))
- dcn401_smu_send_cab_for_uclk_message(clk_mgr, clk_mgr_base->clks.num_ways);
- }
- }
-
- if (should_set_clock(safe_to_lower, new_clocks->dppclk_khz, clk_mgr_base->clks.dppclk_khz)) {
- if (clk_mgr_base->clks.dppclk_khz > new_clocks->dppclk_khz)
- dpp_clock_lowered = true;
-
- clk_mgr_base->clks.dppclk_khz = new_clocks->dppclk_khz;
- clk_mgr_base->clks.actual_dppclk_khz = new_clocks->dppclk_khz;
-
- if (clk_mgr->smu_present && !dpp_clock_lowered && dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DPPCLK))
- clk_mgr_base->clks.actual_dppclk_khz = dcn401_set_hard_min_by_freq_optimized(clk_mgr, PPCLK_DPPCLK, clk_mgr_base->clks.dppclk_khz);
- update_dppclk = true;
- }
-
- if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)) {
- clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz;
-
- if (clk_mgr->smu_present && dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DISPCLK))
- clk_mgr_base->clks.actual_dispclk_khz = dcn401_set_hard_min_by_freq_optimized(clk_mgr, PPCLK_DISPCLK, clk_mgr_base->clks.dispclk_khz);
-
- update_dispclk = true;
- }
-
- if (!new_clocks->dtbclk_en && dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DTBCLK)) {
- new_clocks->ref_dtbclk_khz = clk_mgr_base->bw_params->clk_table.entries[0].dtbclk_mhz * 1000;
- }
-
- /* clock limits are received with MHz precision, divide by 1000 to prevent setting clocks at every call */
- if (!dc->debug.disable_dtb_ref_clk_switch &&
- should_set_clock(safe_to_lower, new_clocks->ref_dtbclk_khz / 1000, clk_mgr_base->clks.ref_dtbclk_khz / 1000) &&
- dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DTBCLK)) {
- /* DCCG requires KHz precision for DTBCLK */
- clk_mgr_base->clks.ref_dtbclk_khz =
- dcn401_smu_set_hard_min_by_freq(clk_mgr, PPCLK_DTBCLK, khz_to_mhz_ceil(new_clocks->ref_dtbclk_khz));
-
- dcn401_update_clocks_update_dtb_dto(clk_mgr, context, clk_mgr_base->clks.ref_dtbclk_khz);
- }
-
- if (dc->config.forced_clocks == false || (force_reset && safe_to_lower)) {
- if (dpp_clock_lowered) {
- /* if clock is being lowered, increase DTO before lowering refclk */
- dcn401_update_clocks_update_dpp_dto(clk_mgr, context,
- safe_to_lower, clk_mgr_base->clks.dppclk_khz);
- dcn401_update_clocks_update_dentist(clk_mgr, context);
- if (clk_mgr->smu_present && dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DPPCLK)) {
- clk_mgr_base->clks.actual_dppclk_khz = dcn401_set_hard_min_by_freq_optimized(clk_mgr, PPCLK_DPPCLK,
- clk_mgr_base->clks.dppclk_khz);
- dcn401_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower,
- clk_mgr_base->clks.actual_dppclk_khz);
- }
-
- } else {
- /* if clock is being raised, increase refclk before lowering DTO */
- if (update_dppclk || update_dispclk)
- dcn401_update_clocks_update_dentist(clk_mgr, context);
- /* There is a check inside dcn20_update_clocks_update_dpp_dto which ensures
- * that we do not lower dto when it is not safe to lower. We do not need to
- * compare the current and new dppclk before calling this function.
- */
- dcn401_update_clocks_update_dpp_dto(clk_mgr, context,
- safe_to_lower, clk_mgr_base->clks.actual_dppclk_khz);
- }
- }
-
- if (update_dispclk && dmcu && dmcu->funcs->is_dmcu_initialized(dmcu))
- /*update dmcu for wait_loop count*/
- dmcu->funcs->set_psr_wait_loop(dmcu,
- clk_mgr_base->clks.dispclk_khz / 1000 / 7);
-}
-
static void dcn401_execute_block_sequence(struct clk_mgr *clk_mgr_base, unsigned int num_steps)
{
struct clk_mgr_internal *clk_mgr_internal = TO_CLK_MGR_INTERNAL(clk_mgr_base);
@@ -869,6 +694,12 @@ static void dcn401_execute_block_sequence(struct clk_mgr *clk_mgr_base, unsigned
params->update_idle_hardmin_params.uclk_mhz,
params->update_idle_hardmin_params.fclk_mhz);
break;
+ case CLK_MGR401_UPDATE_SUBVP_HARDMINS:
+ dcn401_smu_set_subvp_uclk_fclk_hardmin(
+ clk_mgr_internal,
+ params->update_idle_hardmin_params.uclk_mhz,
+ params->update_idle_hardmin_params.fclk_mhz);
+ break;
case CLK_MGR401_UPDATE_DEEP_SLEEP_DCFCLK:
dcn401_smu_set_min_deep_sleep_dcef_clk(
clk_mgr_internal,
@@ -945,15 +776,21 @@ static unsigned int dcn401_build_update_bandwidth_clocks_sequence(
bool update_active_uclk = false;
bool update_idle_fclk = false;
bool update_idle_uclk = false;
+ bool update_subvp_prefetch_dramclk = false;
+ bool update_subvp_prefetch_fclk = false;
bool is_idle_dpm_enabled = dcn401_is_ppclk_dpm_enabled(clk_mgr_internal, PPCLK_UCLK) &&
dcn401_is_ppclk_dpm_enabled(clk_mgr_internal, PPCLK_FCLK) &&
dcn401_is_ppclk_idle_dpm_enabled(clk_mgr_internal, PPCLK_UCLK) &&
dcn401_is_ppclk_idle_dpm_enabled(clk_mgr_internal, PPCLK_FCLK);
+ bool is_df_throttle_opt_enabled = is_idle_dpm_enabled &&
+ dcn401_is_df_throttle_opt_enabled(clk_mgr_internal);
int total_plane_count = clk_mgr_helper_get_active_plane_cnt(dc, context);
int active_uclk_mhz = khz_to_mhz_ceil(clk_mgr_base->clks.dramclk_khz);
int active_fclk_mhz = khz_to_mhz_ceil(clk_mgr_base->clks.fclk_khz);
int idle_uclk_mhz = khz_to_mhz_ceil(clk_mgr_base->clks.idle_dramclk_khz);
int idle_fclk_mhz = khz_to_mhz_ceil(clk_mgr_base->clks.idle_fclk_khz);
+ int subvp_prefetch_dramclk_mhz = khz_to_mhz_ceil(clk_mgr_base->clks.subvp_prefetch_dramclk_khz);
+ int subvp_prefetch_fclk_mhz = khz_to_mhz_ceil(clk_mgr_base->clks.subvp_prefetch_fclk_khz);
unsigned int num_steps = 0;
@@ -982,15 +819,15 @@ static unsigned int dcn401_build_update_bandwidth_clocks_sequence(
update_active_fclk = true;
update_idle_fclk = true;
- /* To enable FCLK P-state switching, send PSTATE_SUPPORTED message to PMFW */
- if (clk_mgr_base->clks.fclk_p_state_change_support) {
- /* Handle the code for sending a message to PMFW that FCLK P-state change is supported */
- if (dcn401_is_ppclk_dpm_enabled(clk_mgr_internal, PPCLK_FCLK)) {
- block_sequence[num_steps].params.update_pstate_support_params.support = true;
- block_sequence[num_steps].func = CLK_MGR401_UPDATE_FCLK_PSTATE_SUPPORT;
- num_steps++;
- }
- }
+ /* To enable FCLK P-state switching, send PSTATE_SUPPORTED message to PMFW (message not supported on DCN401)*/
+ // if (clk_mgr_base->clks.fclk_p_state_change_support) {
+ // /* Handle the code for sending a message to PMFW that FCLK P-state change is supported */
+ // if (dcn401_is_ppclk_dpm_enabled(clk_mgr_internal, PPCLK_FCLK)) {
+ // block_sequence[num_steps].params.update_pstate_support_params.support = true;
+ // block_sequence[num_steps].func = CLK_MGR401_UPDATE_FCLK_PSTATE_SUPPORT;
+ // num_steps++;
+ // }
+ // }
}
if (!clk_mgr_base->clks.fclk_p_state_change_support && dcn401_is_ppclk_dpm_enabled(clk_mgr_internal, PPCLK_FCLK)) {
@@ -1109,6 +946,12 @@ static unsigned int dcn401_build_update_bandwidth_clocks_sequence(
}
}
+ if (should_set_clock(safe_to_lower, new_clocks->subvp_prefetch_dramclk_khz, clk_mgr_base->clks.subvp_prefetch_dramclk_khz)) {
+ clk_mgr_base->clks.subvp_prefetch_dramclk_khz = new_clocks->subvp_prefetch_dramclk_khz;
+ update_subvp_prefetch_dramclk = true;
+ subvp_prefetch_dramclk_mhz = khz_to_mhz_ceil(clk_mgr_base->clks.subvp_prefetch_dramclk_khz);
+ }
+
/* FCLK */
/* Always update saved value, even if new value not set due to P-State switching unsupported */
if (should_set_clock(safe_to_lower, new_clocks->fclk_khz, clk_mgr_base->clks.fclk_khz)) {
@@ -1129,6 +972,12 @@ static unsigned int dcn401_build_update_bandwidth_clocks_sequence(
}
}
+ if (should_set_clock(safe_to_lower, new_clocks->subvp_prefetch_fclk_khz, clk_mgr_base->clks.subvp_prefetch_fclk_khz)) {
+ clk_mgr_base->clks.subvp_prefetch_fclk_khz = new_clocks->subvp_prefetch_fclk_khz;
+ update_subvp_prefetch_fclk = true;
+ subvp_prefetch_fclk_mhz = khz_to_mhz_ceil(clk_mgr_base->clks.subvp_prefetch_fclk_khz);
+ }
+
/* When idle DPM is enabled, need to send active and idle hardmins separately */
/* CLK_MGR401_UPDATE_ACTIVE_HARDMINS */
if ((update_active_uclk || update_active_fclk) && is_idle_dpm_enabled) {
@@ -1146,6 +995,14 @@ static unsigned int dcn401_build_update_bandwidth_clocks_sequence(
num_steps++;
}
+ /* CLK_MGR401_UPDATE_SUBVP_HARDMINS */
+ if ((update_subvp_prefetch_dramclk || update_subvp_prefetch_fclk) && is_df_throttle_opt_enabled) {
+ block_sequence[num_steps].params.update_idle_hardmin_params.uclk_mhz = subvp_prefetch_dramclk_mhz;
+ block_sequence[num_steps].params.update_idle_hardmin_params.fclk_mhz = subvp_prefetch_fclk_mhz;
+ block_sequence[num_steps].func = CLK_MGR401_UPDATE_SUBVP_HARDMINS;
+ num_steps++;
+ }
+
/* set UCLK to requested value if P-State switching is supported, or to re-enable P-State switching */
if (update_active_uclk || update_idle_uclk) {
if (!is_idle_dpm_enabled) {
@@ -1178,14 +1035,14 @@ static unsigned int dcn401_build_update_bandwidth_clocks_sequence(
// (*num_steps)++;
// }
- /* disable FCLK P-State support if needed */
- if (!fclk_p_state_change_support &&
- should_update_pstate_support(safe_to_lower, fclk_p_state_change_support, clk_mgr_base->clks.fclk_prev_p_state_change_support) &&
- dcn401_is_ppclk_dpm_enabled(clk_mgr_internal, PPCLK_FCLK)) {
- block_sequence[num_steps].params.update_pstate_support_params.support = false;
- block_sequence[num_steps].func = CLK_MGR401_UPDATE_FCLK_PSTATE_SUPPORT;
- num_steps++;
- }
+ /* disable FCLK P-State support if needed (message not supported on DCN401)*/
+ // if (!fclk_p_state_change_support &&
+ // should_update_pstate_support(safe_to_lower, fclk_p_state_change_support, clk_mgr_base->clks.fclk_prev_p_state_change_support) &&
+ // dcn401_is_ppclk_dpm_enabled(clk_mgr_internal, PPCLK_FCLK)) {
+ // block_sequence[num_steps].params.update_pstate_support_params.support = false;
+ // block_sequence[num_steps].func = CLK_MGR401_UPDATE_FCLK_PSTATE_SUPPORT;
+ // num_steps++;
+ // }
}
if (new_clocks->fw_based_mclk_switching != clk_mgr_base->clks.fw_based_mclk_switching &&
@@ -1366,11 +1223,6 @@ static void dcn401_update_clocks(struct clk_mgr *clk_mgr_base,
unsigned int num_steps = 0;
- if (dc->debug.enable_legacy_clock_update) {
- dcn401_update_clocks_legacy(clk_mgr_base, context, safe_to_lower);
- return;
- }
-
/* build bandwidth related clocks update sequence */
num_steps = dcn401_build_update_bandwidth_clocks_sequence(clk_mgr_base,
context,
@@ -1505,6 +1357,20 @@ static void dcn401_set_hard_min_memclk(struct clk_mgr *clk_mgr_base, bool curren
dcn401_execute_block_sequence(clk_mgr_base, num_steps);
}
+static int dcn401_get_hard_min_memclk(struct clk_mgr *clk_mgr_base)
+{
+ struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
+
+ return clk_mgr->base.ctx->dc->current_state->bw_ctx.bw.dcn.clk.dramclk_khz;
+}
+
+static int dcn401_get_hard_min_fclk(struct clk_mgr *clk_mgr_base)
+{
+ struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
+
+ return clk_mgr->base.ctx->dc->current_state->bw_ctx.bw.dcn.clk.fclk_khz;
+}
+
/* Get current memclk states, update bounding box */
static void dcn401_get_memclk_states_from_smu(struct clk_mgr *clk_mgr_base)
{
@@ -1549,6 +1415,15 @@ static void dcn401_get_memclk_states_from_smu(struct clk_mgr *clk_mgr_base)
if (clk_mgr->dpm_present && !num_levels)
clk_mgr->dpm_present = false;
+ clk_mgr_base->bw_params->num_channels = dcn401_smu_get_num_of_umc_channels(clk_mgr);
+ if (clk_mgr_base->ctx->dc_bios) {
+ /* use BIOS values if none provided by PMFW */
+ if (clk_mgr_base->bw_params->num_channels == 0) {
+ clk_mgr_base->bw_params->num_channels = clk_mgr_base->ctx->dc_bios->vram_info.num_chans;
+ }
+ clk_mgr_base->bw_params->dram_channel_width_bytes = clk_mgr_base->ctx->dc_bios->vram_info.dram_channel_width_bytes;
+ }
+
/* Refresh bounding box */
clk_mgr_base->ctx->dc->res_pool->funcs->update_bw_bounding_box(
clk_mgr->base.ctx->dc, clk_mgr_base->bw_params);
@@ -1625,6 +1500,35 @@ static int dcn401_get_dispclk_from_dentist(struct clk_mgr *clk_mgr_base)
return 0;
}
+unsigned int dcn401_get_max_clock_khz(struct clk_mgr *clk_mgr_base, enum clk_type clk_type)
+{
+ struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
+
+ unsigned int num_clk_levels;
+
+ switch (clk_type) {
+ case CLK_TYPE_DISPCLK:
+ num_clk_levels = clk_mgr->base.bw_params->clk_table.num_entries_per_clk.num_dispclk_levels;
+ return dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DISPCLK) ?
+ clk_mgr->base.bw_params->clk_table.entries[num_clk_levels - 1].dispclk_mhz * 1000 :
+ clk_mgr->base.boot_snapshot.dispclk;
+ case CLK_TYPE_DPPCLK:
+ num_clk_levels = clk_mgr->base.bw_params->clk_table.num_entries_per_clk.num_dppclk_levels;
+ return dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DPPCLK) ?
+ clk_mgr->base.bw_params->clk_table.entries[num_clk_levels - 1].dppclk_mhz * 1000 :
+ clk_mgr->base.boot_snapshot.dppclk;
+ case CLK_TYPE_DSCCLK:
+ num_clk_levels = clk_mgr->base.bw_params->clk_table.num_entries_per_clk.num_dispclk_levels;
+ return dcn401_is_ppclk_dpm_enabled(clk_mgr, PPCLK_DISPCLK) ?
+ clk_mgr->base.bw_params->clk_table.entries[num_clk_levels - 1].dispclk_mhz * 1000 / 3 :
+ clk_mgr->base.boot_snapshot.dispclk / 3;
+ default:
+ break;
+ }
+
+ return 0;
+}
+
static struct clk_mgr_funcs dcn401_funcs = {
.get_dp_ref_clk_frequency = dce12_get_dp_ref_freq_khz,
.get_dtb_ref_clk_frequency = dcn401_get_dtb_ref_freq_khz,
@@ -1638,6 +1542,10 @@ static struct clk_mgr_funcs dcn401_funcs = {
.enable_pme_wa = dcn401_enable_pme_wa,
.is_smu_present = dcn401_is_smu_present,
.get_dispclk_from_dentist = dcn401_get_dispclk_from_dentist,
+ .get_hard_min_memclk = dcn401_get_hard_min_memclk,
+ .get_hard_min_fclk = dcn401_get_hard_min_fclk,
+ .is_dc_mode_present = dcn401_is_dc_mode_present,
+ .get_max_clock_khz = dcn401_get_max_clock_khz,
};
struct clk_mgr_internal *dcn401_clk_mgr_construct(
@@ -1698,7 +1606,7 @@ struct clk_mgr_internal *dcn401_clk_mgr_construct(
clk_mgr->base.bw_params = kzalloc(sizeof(*clk_mgr->base.bw_params), GFP_KERNEL);
if (!clk_mgr->base.bw_params) {
BREAK_TO_DEBUGGER();
- kfree(clk_mgr);
+ kfree(clk_mgr401);
return NULL;
}
@@ -1709,6 +1617,7 @@ struct clk_mgr_internal *dcn401_clk_mgr_construct(
if (!clk_mgr->wm_range_table) {
BREAK_TO_DEBUGGER();
kfree(clk_mgr->base.bw_params);
+ kfree(clk_mgr401);
return NULL;
}
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.h
index 8b0461992b22..97a1ce1e8a9e 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr.h
@@ -90,6 +90,7 @@ enum dcn401_clk_mgr_block_sequence_func {
CLK_MGR401_UPDATE_DTBCLK_DTO,
CLK_MGR401_UPDATE_DENTIST,
CLK_MGR401_UPDATE_PSR_WAIT_LOOP,
+ CLK_MGR401_UPDATE_SUBVP_HARDMINS,
};
struct dcn401_clk_mgr_block_sequence {
@@ -104,10 +105,13 @@ struct dcn401_clk_mgr {
};
void dcn401_init_clocks(struct clk_mgr *clk_mgr_base);
+bool dcn401_is_dc_mode_present(struct clk_mgr *clk_mgr_base);
struct clk_mgr_internal *dcn401_clk_mgr_construct(struct dc_context *ctx,
struct dccg *dccg);
void dcn401_clk_mgr_destroy(struct clk_mgr_internal *clk_mgr);
+unsigned int dcn401_get_max_clock_khz(struct clk_mgr *clk_mgr_base, enum clk_type clk_type);
+
#endif /* __DCN401_CLK_MGR_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.c
index 7700477d019b..21c35528f61f 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.c
@@ -21,6 +21,14 @@
#define smu_print(str, ...) {DC_LOG_SMU(str, ##__VA_ARGS__); }
+/* temporary define */
+#ifndef DALSMC_MSG_SubvpUclkFclk
+#define DALSMC_MSG_SubvpUclkFclk 0x1B
+#endif
+#ifndef DALSMC_MSG_GetNumUmcChannels
+#define DALSMC_MSG_GetNumUmcChannels 0x1C
+#endif
+
/*
* Function to be used instead of REG_WAIT macro because the wait ends when
* the register is NOT EQUAL to zero, and because the translation in msg_if.h
@@ -296,6 +304,24 @@ bool dcn401_smu_set_active_uclk_fclk_hardmin(struct clk_mgr_internal *clk_mgr,
return success;
}
+bool dcn401_smu_set_subvp_uclk_fclk_hardmin(struct clk_mgr_internal *clk_mgr,
+ uint16_t uclk_freq_mhz,
+ uint16_t fclk_freq_mhz)
+{
+ uint32_t response = 0;
+ bool success;
+
+ /* 15:0 for uclk, 32:16 for fclk */
+ uint32_t param = (fclk_freq_mhz << 16) | uclk_freq_mhz;
+
+ smu_print("SMU Set active hardmin by freq: uclk_freq_mhz = %d MHz, fclk_freq_mhz = %d MHz\n", uclk_freq_mhz, fclk_freq_mhz);
+
+ success = dcn401_smu_send_msg_with_param(clk_mgr,
+ DALSMC_MSG_SubvpUclkFclk, param, &response);
+
+ return success;
+}
+
void dcn401_smu_set_min_deep_sleep_dcef_clk(struct clk_mgr_internal *clk_mgr, uint32_t freq_mhz)
{
smu_print("SMU Set min deep sleep dcef clk: freq_mhz = %d MHz\n", freq_mhz);
@@ -311,3 +337,14 @@ void dcn401_smu_set_num_of_displays(struct clk_mgr_internal *clk_mgr, uint32_t n
dcn401_smu_send_msg_with_param(clk_mgr,
DALSMC_MSG_NumOfDisplays, num_displays, NULL);
}
+
+unsigned int dcn401_smu_get_num_of_umc_channels(struct clk_mgr_internal *clk_mgr)
+{
+ unsigned int response = 0;
+
+ dcn401_smu_send_msg_with_param(clk_mgr, DALSMC_MSG_GetNumUmcChannels, 0, &response);
+
+ smu_print("SMU Get Num UMC Channels: num_umc_channels = %d\n", response);
+
+ return response;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.h
index 651fb8d62864..e02eb1294b37 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn401/dcn401_clk_mgr_smu_msg.h
@@ -23,7 +23,11 @@ bool dcn401_smu_set_idle_uclk_fclk_hardmin(struct clk_mgr_internal *clk_mgr,
bool dcn401_smu_set_active_uclk_fclk_hardmin(struct clk_mgr_internal *clk_mgr,
uint16_t uclk_freq_mhz,
uint16_t fclk_freq_mhz);
+bool dcn401_smu_set_subvp_uclk_fclk_hardmin(struct clk_mgr_internal *clk_mgr,
+ uint16_t uclk_freq_mhz,
+ uint16_t fclk_freq_mhz);
void dcn401_smu_set_min_deep_sleep_dcef_clk(struct clk_mgr_internal *clk_mgr, uint32_t freq_mhz);
void dcn401_smu_set_num_of_displays(struct clk_mgr_internal *clk_mgr, uint32_t num_displays);
+unsigned int dcn401_smu_get_num_of_umc_channels(struct clk_mgr_internal *clk_mgr);
#endif /* __DCN401_CLK_MGR_SMU_MSG_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 1dd26d5df6b9..9ab0ee20ca6f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -36,7 +36,9 @@
#include "resource.h"
#include "dc_state.h"
#include "dc_state_priv.h"
+#include "dc_plane.h"
#include "dc_plane_priv.h"
+#include "dc_stream_priv.h"
#include "gpio_service_interface.h"
#include "clk_mgr.h"
@@ -239,6 +241,7 @@ static bool create_links(
DC_LOG_DC("BIOS object table - end");
/* Create a link for each usb4 dpia port */
+ dc->lowest_dpia_link_index = MAX_LINKS;
for (i = 0; i < dc->res_pool->usb4_dpia_count; i++) {
struct link_init_data link_init_params = {0};
struct dc_link *link;
@@ -251,6 +254,9 @@ static bool create_links(
link = dc->link_srv->create_link(&link_init_params);
if (link) {
+ if (dc->lowest_dpia_link_index > dc->link_count)
+ dc->lowest_dpia_link_index = dc->link_count;
+
dc->links[dc->link_count] = link;
link->dc = dc;
++dc->link_count;
@@ -276,6 +282,7 @@ static bool create_links(
link->link_id.type = OBJECT_TYPE_CONNECTOR;
link->link_id.id = CONNECTOR_ID_VIRTUAL;
link->link_id.enum_id = ENUM_ID_1;
+ link->psr_settings.psr_version = DC_PSR_VERSION_UNSUPPORTED;
link->link_enc = kzalloc(sizeof(*link->link_enc), GFP_KERNEL);
if (!link->link_enc) {
@@ -438,9 +445,12 @@ bool dc_stream_adjust_vmin_vmax(struct dc *dc,
* Don't adjust DRR while there's bandwidth optimizations pending to
* avoid conflicting with firmware updates.
*/
- if (dc->ctx->dce_version > DCE_VERSION_MAX)
- if (dc->optimized_required || dc->wm_optimized_required)
+ if (dc->ctx->dce_version > DCE_VERSION_MAX) {
+ if (dc->optimized_required || dc->wm_optimized_required) {
+ stream->adjust.timing_adjust_pending = true;
return false;
+ }
+ }
dc_exit_ips_for_hw_access(dc);
@@ -452,6 +462,7 @@ bool dc_stream_adjust_vmin_vmax(struct dc *dc,
if (dc->caps.max_v_total != 0 &&
(adjust->v_total_max > dc->caps.max_v_total || adjust->v_total_min > dc->caps.max_v_total)) {
+ stream->adjust.timing_adjust_pending = false;
if (adjust->allow_otg_v_count_halt)
return set_long_vtotal(dc, stream, adjust);
else
@@ -465,7 +476,7 @@ bool dc_stream_adjust_vmin_vmax(struct dc *dc,
dc->hwss.set_drr(&pipe,
1,
*adjust);
-
+ stream->adjust.timing_adjust_pending = false;
return true;
}
}
@@ -515,33 +526,6 @@ bool dc_stream_get_last_used_drr_vtotal(struct dc *dc,
return status;
}
-bool dc_stream_get_crtc_position(struct dc *dc,
- struct dc_stream_state **streams, int num_streams,
- unsigned int *v_pos, unsigned int *nom_v_pos)
-{
- /* TODO: Support multiple streams */
- const struct dc_stream_state *stream = streams[0];
- int i;
- bool ret = false;
- struct crtc_position position;
-
- dc_exit_ips_for_hw_access(dc);
-
- for (i = 0; i < MAX_PIPES; i++) {
- struct pipe_ctx *pipe =
- &dc->current_state->res_ctx.pipe_ctx[i];
-
- if (pipe->stream == stream && pipe->stream_res.stream_enc) {
- dc->hwss.get_position(&pipe, 1, &position);
-
- *v_pos = position.vertical_count;
- *nom_v_pos = position.nominal_vcount;
- ret = true;
- }
- }
- return ret;
-}
-
#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
static inline void
dc_stream_forward_dmub_crc_window(struct dc_dmub_srv *dmub_srv,
@@ -579,7 +563,7 @@ dc_stream_forward_dmcu_crc_window(struct dmcu *dmcu,
bool
dc_stream_forward_crc_window(struct dc_stream_state *stream,
- struct rect *rect, bool is_stop)
+ struct rect *rect, uint8_t phy_id, bool is_stop)
{
struct dmcu *dmcu;
struct dc_dmub_srv *dmub_srv;
@@ -598,7 +582,7 @@ dc_stream_forward_crc_window(struct dc_stream_state *stream,
if (i == MAX_PIPES)
return false;
- mux_mapping.phy_output_num = stream->link->link_enc_hw_inst;
+ mux_mapping.phy_output_num = phy_id;
mux_mapping.otg_output_num = pipe->stream_res.tg->inst;
dmcu = dc->res_pool->dmcu;
@@ -615,6 +599,68 @@ dc_stream_forward_crc_window(struct dc_stream_state *stream,
return true;
}
+
+static void
+dc_stream_forward_dmub_multiple_crc_window(struct dc_dmub_srv *dmub_srv,
+ struct crc_window *window, struct otg_phy_mux *mux_mapping, bool stop)
+{
+ int i;
+ union dmub_rb_cmd cmd = {0};
+
+ cmd.secure_display.mul_roi_ctl.phy_id = mux_mapping->phy_output_num;
+ cmd.secure_display.mul_roi_ctl.otg_id = mux_mapping->otg_output_num;
+
+ cmd.secure_display.header.type = DMUB_CMD__SECURE_DISPLAY;
+
+ if (stop) {
+ cmd.secure_display.header.sub_type = DMUB_CMD__SECURE_DISPLAY_MULTIPLE_CRC_STOP_UPDATE;
+ } else {
+ cmd.secure_display.header.sub_type = DMUB_CMD__SECURE_DISPLAY_MULTIPLE_CRC_WIN_NOTIFY;
+ for (i = 0; i < MAX_CRC_WINDOW_NUM; i++) {
+ cmd.secure_display.mul_roi_ctl.roi_ctl[i].x_start = window[i].rect.x;
+ cmd.secure_display.mul_roi_ctl.roi_ctl[i].y_start = window[i].rect.y;
+ cmd.secure_display.mul_roi_ctl.roi_ctl[i].x_end = window[i].rect.x + window[i].rect.width;
+ cmd.secure_display.mul_roi_ctl.roi_ctl[i].y_end = window[i].rect.y + window[i].rect.height;
+ cmd.secure_display.mul_roi_ctl.roi_ctl[i].enable = window[i].enable;
+ }
+ }
+
+ dc_wake_and_execute_dmub_cmd(dmub_srv->ctx, &cmd, DM_DMUB_WAIT_TYPE_NO_WAIT);
+}
+
+bool
+dc_stream_forward_multiple_crc_window(struct dc_stream_state *stream,
+ struct crc_window *window, uint8_t phy_id, bool stop)
+{
+ struct dc_dmub_srv *dmub_srv;
+ struct otg_phy_mux mux_mapping;
+ struct pipe_ctx *pipe;
+ int i;
+ struct dc *dc = stream->ctx->dc;
+
+ for (i = 0; i < MAX_PIPES; i++) {
+ pipe = &dc->current_state->res_ctx.pipe_ctx[i];
+ if (pipe->stream == stream && !pipe->top_pipe && !pipe->prev_odm_pipe)
+ break;
+ }
+
+ /* Stream not found */
+ if (i == MAX_PIPES)
+ return false;
+
+ mux_mapping.phy_output_num = phy_id;
+ mux_mapping.otg_output_num = pipe->stream_res.tg->inst;
+
+ dmub_srv = dc->ctx->dmub_srv;
+
+ /* forward to dmub only. no dmcu support*/
+ if (dmub_srv)
+ dc_stream_forward_dmub_multiple_crc_window(dmub_srv, window, &mux_mapping, stop);
+ else
+ return false;
+
+ return true;
+}
#endif /* CONFIG_DRM_AMD_SECURE_DISPLAY */
/**
@@ -625,15 +671,17 @@ dc_stream_forward_crc_window(struct dc_stream_state *stream,
* @enable: Enable CRC if true, disable otherwise.
* @continuous: Capture CRC on every frame if true. Otherwise, only capture
* once.
+ * @idx: Capture CRC on which CRC engine instance
+ * @reset: Reset CRC engine before the configuration
*
- * By default, only CRC0 is configured, and the entire frame is used to
- * calculate the CRC.
+ * By default, the entire frame is used to calculate the CRC.
*
* Return: %false if the stream is not found or CRC capture is not supported;
* %true if the stream has been configured.
*/
bool dc_stream_configure_crc(struct dc *dc, struct dc_stream_state *stream,
- struct crc_params *crc_window, bool enable, bool continuous)
+ struct crc_params *crc_window, bool enable, bool continuous,
+ uint8_t idx, bool reset)
{
struct pipe_ctx *pipe;
struct crc_params param;
@@ -677,6 +725,9 @@ bool dc_stream_configure_crc(struct dc *dc, struct dc_stream_state *stream,
param.continuous_mode = continuous;
param.enable = enable;
+ param.crc_eng_inst = idx;
+ param.reset = reset;
+
tg = pipe->stream_res.tg;
/* Only call if supported */
@@ -691,6 +742,7 @@ bool dc_stream_configure_crc(struct dc *dc, struct dc_stream_state *stream,
*
* @dc: DC object.
* @stream: The DC stream state of the stream to get CRCs from.
+ * @idx: index of crc engine to get CRC from
* @r_cr: CRC value for the red component.
* @g_y: CRC value for the green component.
* @b_cb: CRC value for the blue component.
@@ -700,7 +752,7 @@ bool dc_stream_configure_crc(struct dc *dc, struct dc_stream_state *stream,
* Return:
* %false if stream is not found, or if CRCs are not enabled.
*/
-bool dc_stream_get_crc(struct dc *dc, struct dc_stream_state *stream,
+bool dc_stream_get_crc(struct dc *dc, struct dc_stream_state *stream, uint8_t idx,
uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb)
{
int i;
@@ -721,7 +773,7 @@ bool dc_stream_get_crc(struct dc *dc, struct dc_stream_state *stream,
tg = pipe->stream_res.tg;
if (tg->funcs->get_crc)
- return tg->funcs->get_crc(tg, r_cr, g_y, b_cb);
+ return tg->funcs->get_crc(tg, idx, r_cr, g_y, b_cb);
DC_LOG_WARNING("CRC capture not supported.");
return false;
}
@@ -863,7 +915,8 @@ void dc_stream_set_static_screen_params(struct dc *dc,
static void dc_destruct(struct dc *dc)
{
// reset link encoder assignment table on destruct
- if (dc->res_pool && dc->res_pool->funcs->link_encs_assign)
+ if (dc->res_pool && dc->res_pool->funcs->link_encs_assign &&
+ !dc->config.unify_link_enc_assignment)
link_enc_cfg_init(dc, dc->current_state);
if (dc->current_state) {
@@ -885,17 +938,18 @@ static void dc_destruct(struct dc *dc)
if (dc->link_srv)
link_destroy_link_service(&dc->link_srv);
- if (dc->ctx->gpio_service)
- dal_gpio_service_destroy(&dc->ctx->gpio_service);
+ if (dc->ctx) {
+ if (dc->ctx->gpio_service)
+ dal_gpio_service_destroy(&dc->ctx->gpio_service);
- if (dc->ctx->created_bios)
- dal_bios_parser_destroy(&dc->ctx->dc_bios);
+ if (dc->ctx->created_bios)
+ dal_bios_parser_destroy(&dc->ctx->dc_bios);
+ kfree(dc->ctx->logger);
+ dc_perf_trace_destroy(&dc->ctx->perf_trace);
- kfree(dc->ctx->logger);
- dc_perf_trace_destroy(&dc->ctx->perf_trace);
-
- kfree(dc->ctx);
- dc->ctx = NULL;
+ kfree(dc->ctx);
+ dc->ctx = NULL;
+ }
kfree(dc->bw_vbios);
dc->bw_vbios = NULL;
@@ -923,6 +977,8 @@ static bool dc_construct_ctx(struct dc *dc,
if (!dc_ctx)
return false;
+ dc_stream_init_rmcm_3dlut(dc);
+
dc_ctx->cgs_device = init_params->cgs_device;
dc_ctx->driver_context = init_params->driver;
dc_ctx->dc = dc;
@@ -1148,6 +1204,12 @@ static void apply_ctx_interdependent_lock(struct dc *dc,
static void dc_update_visual_confirm_color(struct dc *dc, struct dc_state *context, struct pipe_ctx *pipe_ctx)
{
+ if (dc->debug.visual_confirm & VISUAL_CONFIRM_EXPLICIT) {
+ memcpy(&pipe_ctx->visual_confirm_color, &pipe_ctx->plane_state->visual_confirm_color,
+ sizeof(pipe_ctx->visual_confirm_color));
+ return;
+ }
+
if (dc->ctx->dce_version >= DCN_VERSION_1_0) {
memset(&pipe_ctx->visual_confirm_color, 0, sizeof(struct tg_color));
@@ -1159,6 +1221,8 @@ static void dc_update_visual_confirm_color(struct dc *dc, struct dc_state *conte
get_surface_tile_visual_confirm_color(pipe_ctx, &(pipe_ctx->visual_confirm_color));
else if (dc->debug.visual_confirm == VISUAL_CONFIRM_HW_CURSOR)
get_cursor_visual_confirm_color(pipe_ctx, &(pipe_ctx->visual_confirm_color));
+ else if (dc->debug.visual_confirm == VISUAL_CONFIRM_DCC)
+ get_dcc_visual_confirm_color(dc, pipe_ctx, &(pipe_ctx->visual_confirm_color));
else {
if (dc->ctx->dce_version < DCN_VERSION_2_0)
color_space_to_black_color(
@@ -1173,6 +1237,53 @@ static void dc_update_visual_confirm_color(struct dc *dc, struct dc_state *conte
get_mclk_switch_visual_confirm_color(pipe_ctx, &(pipe_ctx->visual_confirm_color));
else if (dc->debug.visual_confirm == VISUAL_CONFIRM_FAMS2)
get_fams2_visual_confirm_color(dc, context, pipe_ctx, &(pipe_ctx->visual_confirm_color));
+ else if (dc->debug.visual_confirm == VISUAL_CONFIRM_VABC)
+ get_vabc_visual_confirm_color(pipe_ctx, &(pipe_ctx->visual_confirm_color));
+ }
+ }
+}
+
+void dc_get_visual_confirm_for_stream(
+ struct dc *dc,
+ struct dc_stream_state *stream_state,
+ struct tg_color *color)
+{
+ struct dc_stream_status *stream_status = dc_stream_get_status(stream_state);
+ struct pipe_ctx *pipe_ctx;
+ int i;
+ struct dc_plane_state *plane_state = NULL;
+
+ if (!stream_status)
+ return;
+
+ switch (dc->debug.visual_confirm) {
+ case VISUAL_CONFIRM_DISABLE:
+ return;
+ case VISUAL_CONFIRM_PSR:
+ case VISUAL_CONFIRM_FAMS:
+ pipe_ctx = dc_stream_get_pipe_ctx(stream_state);
+ if (!pipe_ctx)
+ return;
+ dc_dmub_srv_get_visual_confirm_color_cmd(dc, pipe_ctx);
+ memcpy(color, &dc->ctx->dmub_srv->dmub->visual_confirm_color, sizeof(struct tg_color));
+ return;
+
+ default:
+ /* find plane with highest layer_index */
+ for (i = 0; i < stream_status->plane_count; i++) {
+ if (stream_status->plane_states[i]->visible)
+ plane_state = stream_status->plane_states[i];
+ }
+ if (!plane_state)
+ return;
+ /* find pipe that contains plane with highest layer index */
+ for (i = 0; i < MAX_PIPES; i++) {
+ struct pipe_ctx *pipe = &dc->current_state->res_ctx.pipe_ctx[i];
+
+ if (pipe->plane_state == plane_state) {
+ memcpy(color, &pipe->visual_confirm_color, sizeof(struct tg_color));
+ return;
+ }
}
}
}
@@ -1650,17 +1761,23 @@ bool dc_validate_boot_timing(const struct dc *dc,
return false;
}
- if (dc->debug.force_odm_combine)
+ if (dc->debug.force_odm_combine) {
+ DC_LOG_DEBUG("boot timing validation failed due to force_odm_combine\n");
return false;
+ }
/* Check for enabled DIG to identify enabled display */
- if (!link->link_enc->funcs->is_dig_enabled(link->link_enc))
+ if (!link->link_enc->funcs->is_dig_enabled(link->link_enc)) {
+ DC_LOG_DEBUG("boot timing validation failed due to disabled DIG\n");
return false;
+ }
enc_inst = link->link_enc->funcs->get_dig_frontend(link->link_enc);
- if (enc_inst == ENGINE_ID_UNKNOWN)
+ if (enc_inst == ENGINE_ID_UNKNOWN) {
+ DC_LOG_DEBUG("boot timing validation failed due to unknown DIG engine ID\n");
return false;
+ }
for (i = 0; i < dc->res_pool->stream_enc_count; i++) {
if (dc->res_pool->stream_enc[i]->id == enc_inst) {
@@ -1674,62 +1791,98 @@ bool dc_validate_boot_timing(const struct dc *dc,
}
// tg_inst not found
- if (i == dc->res_pool->stream_enc_count)
+ if (i == dc->res_pool->stream_enc_count) {
+ DC_LOG_DEBUG("boot timing validation failed due to timing generator instance not found\n");
return false;
+ }
- if (tg_inst >= dc->res_pool->timing_generator_count)
+ if (tg_inst >= dc->res_pool->timing_generator_count) {
+ DC_LOG_DEBUG("boot timing validation failed due to invalid timing generator count\n");
return false;
+ }
- if (tg_inst != link->link_enc->preferred_engine)
+ if (tg_inst != link->link_enc->preferred_engine) {
+ DC_LOG_DEBUG("boot timing validation failed due to non-preferred timing generator\n");
return false;
+ }
tg = dc->res_pool->timing_generators[tg_inst];
- if (!tg->funcs->get_hw_timing)
+ if (!tg->funcs->get_hw_timing) {
+ DC_LOG_DEBUG("boot timing validation failed due to missing get_hw_timing callback\n");
return false;
+ }
- if (!tg->funcs->get_hw_timing(tg, &hw_crtc_timing))
+ if (!tg->funcs->get_hw_timing(tg, &hw_crtc_timing)) {
+ DC_LOG_DEBUG("boot timing validation failed due to failed get_hw_timing return\n");
return false;
+ }
- if (crtc_timing->h_total != hw_crtc_timing.h_total)
+ if (crtc_timing->h_total != hw_crtc_timing.h_total) {
+ DC_LOG_DEBUG("boot timing validation failed due to h_total mismatch\n");
return false;
+ }
- if (crtc_timing->h_border_left != hw_crtc_timing.h_border_left)
+ if (crtc_timing->h_border_left != hw_crtc_timing.h_border_left) {
+ DC_LOG_DEBUG("boot timing validation failed due to h_border_left mismatch\n");
return false;
+ }
- if (crtc_timing->h_addressable != hw_crtc_timing.h_addressable)
+ if (crtc_timing->h_addressable != hw_crtc_timing.h_addressable) {
+ DC_LOG_DEBUG("boot timing validation failed due to h_addressable mismatch\n");
return false;
+ }
- if (crtc_timing->h_border_right != hw_crtc_timing.h_border_right)
+ if (crtc_timing->h_border_right != hw_crtc_timing.h_border_right) {
+ DC_LOG_DEBUG("boot timing validation failed due to h_border_right mismatch\n");
return false;
+ }
- if (crtc_timing->h_front_porch != hw_crtc_timing.h_front_porch)
+ if (crtc_timing->h_front_porch != hw_crtc_timing.h_front_porch) {
+ DC_LOG_DEBUG("boot timing validation failed due to h_front_porch mismatch\n");
return false;
+ }
- if (crtc_timing->h_sync_width != hw_crtc_timing.h_sync_width)
+ if (crtc_timing->h_sync_width != hw_crtc_timing.h_sync_width) {
+ DC_LOG_DEBUG("boot timing validation failed due to h_sync_width mismatch\n");
return false;
+ }
- if (crtc_timing->v_total != hw_crtc_timing.v_total)
+ if (crtc_timing->v_total != hw_crtc_timing.v_total) {
+ DC_LOG_DEBUG("boot timing validation failed due to v_total mismatch\n");
return false;
+ }
- if (crtc_timing->v_border_top != hw_crtc_timing.v_border_top)
+ if (crtc_timing->v_border_top != hw_crtc_timing.v_border_top) {
+ DC_LOG_DEBUG("boot timing validation failed due to v_border_top mismatch\n");
return false;
+ }
- if (crtc_timing->v_addressable != hw_crtc_timing.v_addressable)
+ if (crtc_timing->v_addressable != hw_crtc_timing.v_addressable) {
+ DC_LOG_DEBUG("boot timing validation failed due to v_addressable mismatch\n");
return false;
+ }
- if (crtc_timing->v_border_bottom != hw_crtc_timing.v_border_bottom)
+ if (crtc_timing->v_border_bottom != hw_crtc_timing.v_border_bottom) {
+ DC_LOG_DEBUG("boot timing validation failed due to v_border_bottom mismatch\n");
return false;
+ }
- if (crtc_timing->v_front_porch != hw_crtc_timing.v_front_porch)
+ if (crtc_timing->v_front_porch != hw_crtc_timing.v_front_porch) {
+ DC_LOG_DEBUG("boot timing validation failed due to v_front_porch mismatch\n");
return false;
+ }
- if (crtc_timing->v_sync_width != hw_crtc_timing.v_sync_width)
+ if (crtc_timing->v_sync_width != hw_crtc_timing.v_sync_width) {
+ DC_LOG_DEBUG("boot timing validation failed due to v_sync_width mismatch\n");
return false;
+ }
/* block DSC for now, as VBIOS does not currently support DSC timings */
- if (crtc_timing->flags.DSC)
+ if (crtc_timing->flags.DSC) {
+ DC_LOG_DEBUG("boot timing validation failed due to DSC\n");
return false;
+ }
if (dc_is_dp_signal(link->connector_signal)) {
unsigned int pix_clk_100hz = 0;
@@ -1751,39 +1904,55 @@ bool dc_validate_boot_timing(const struct dc *dc,
} else if (se && se->funcs->get_pixels_per_cycle) {
uint32_t pixels_per_cycle = se->funcs->get_pixels_per_cycle(se);
- if (pixels_per_cycle != 1 && !dc->debug.enable_dp_dig_pixel_rate_div_policy)
+ if (pixels_per_cycle != 1 && !dc->debug.enable_dp_dig_pixel_rate_div_policy) {
+ DC_LOG_DEBUG("boot timing validation failed due to pixels_per_cycle\n");
return false;
+ }
pix_clk_100hz *= pixels_per_cycle;
}
// Note: In rare cases, HW pixclk may differ from crtc's pixclk
// slightly due to rounding issues in 10 kHz units.
- if (crtc_timing->pix_clk_100hz != pix_clk_100hz)
+ if (crtc_timing->pix_clk_100hz != pix_clk_100hz) {
+ DC_LOG_DEBUG("boot timing validation failed due to pix_clk_100hz mismatch\n");
return false;
+ }
- if (!se || !se->funcs->dp_get_pixel_format)
+ if (!se || !se->funcs->dp_get_pixel_format) {
+ DC_LOG_DEBUG("boot timing validation failed due to missing dp_get_pixel_format\n");
return false;
+ }
if (!se->funcs->dp_get_pixel_format(
se,
&hw_crtc_timing.pixel_encoding,
- &hw_crtc_timing.display_color_depth))
+ &hw_crtc_timing.display_color_depth)) {
+ DC_LOG_DEBUG("boot timing validation failed due to dp_get_pixel_format failure\n");
return false;
+ }
- if (hw_crtc_timing.display_color_depth != crtc_timing->display_color_depth)
+ if (hw_crtc_timing.display_color_depth != crtc_timing->display_color_depth) {
+ DC_LOG_DEBUG("boot timing validation failed due to display_color_depth mismatch\n");
return false;
+ }
- if (hw_crtc_timing.pixel_encoding != crtc_timing->pixel_encoding)
+ if (hw_crtc_timing.pixel_encoding != crtc_timing->pixel_encoding) {
+ DC_LOG_DEBUG("boot timing validation failed due to pixel_encoding mismatch\n");
return false;
+ }
}
+
if (link->dpcd_caps.dprx_feature.bits.VSC_SDP_COLORIMETRY_SUPPORTED) {
+ DC_LOG_DEBUG("boot timing validation failed due to VSC SDP colorimetry\n");
return false;
}
- if (link->dpcd_caps.channel_coding_cap.bits.DP_128b_132b_SUPPORTED)
+ if (link->dpcd_caps.channel_coding_cap.bits.DP_128b_132b_SUPPORTED) {
+ DC_LOG_DEBUG("boot timing validation failed due to DP 128b/132b\n");
return false;
+ }
if (dc->link_srv->edp_is_ilr_optimization_required(link, crtc_timing)) {
DC_LOG_EVENT_LINK_TRAINING("Seamless boot disabled to optimize eDP link rate\n");
@@ -1947,6 +2116,18 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
dc->hwss.enable_accelerated_mode(dc, context);
}
+ if (dc->hwseq->funcs.wait_for_pipe_update_if_needed) {
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ pipe = &context->res_ctx.pipe_ctx[i];
+ //Only delay otg master for a given config
+ if (resource_is_pipe_type(pipe, OTG_MASTER)) {
+ //dc_commit_state_no_check is always a full update
+ dc->hwseq->funcs.wait_for_pipe_update_if_needed(dc, pipe, false);
+ break;
+ }
+ }
+ }
+
if (context->stream_count > get_seamless_boot_stream_count(context) ||
context->stream_count == 0)
dc->hwss.prepare_bandwidth(dc, context);
@@ -2011,6 +2192,14 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
if (dc->hwss.program_front_end_for_ctx) {
dc->hwss.interdependent_update_lock(dc, context, true);
dc->hwss.program_front_end_for_ctx(dc, context);
+
+ if (dc->hwseq->funcs.set_wait_for_update_needed_for_pipe) {
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ pipe = &context->res_ctx.pipe_ctx[i];
+ dc->hwseq->funcs.set_wait_for_update_needed_for_pipe(dc, pipe);
+ }
+ }
+
dc->hwss.interdependent_update_lock(dc, context, false);
dc->hwss.post_unlock_program_front_end(dc, context);
}
@@ -2063,7 +2252,7 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
dc_enable_stereo(dc, context, dc_streams, context->stream_count);
- if (context->stream_count > get_seamless_boot_stream_count(context) ||
+ if (get_seamless_boot_stream_count(context) == 0 ||
context->stream_count == 0) {
/* Must wait for no flips to be pending before doing optimize bw */
hwss_wait_for_no_pipes_pending(dc, context);
@@ -2152,6 +2341,15 @@ enum dc_status dc_commit_streams(struct dc *dc, struct dc_commit_streams_params
for (i = 0; i < params->stream_count; i++) {
struct dc_stream_state *stream = params->streams[i];
struct dc_stream_status *status = dc_stream_get_status(stream);
+ struct dc_sink *sink = stream->sink;
+
+ /* revalidate streams */
+ if (!dc_is_virtual_signal(sink->sink_signal)) {
+ res = dc_validate_stream(dc, stream);
+ if (res != DC_OK)
+ return res;
+ }
+
dc_stream_log(dc, stream);
@@ -2186,12 +2384,12 @@ enum dc_status dc_commit_streams(struct dc *dc, struct dc_commit_streams_params
context->power_source = params->power_source;
- res = dc_validate_with_context(dc, set, params->stream_count, context, false);
+ res = dc_validate_with_context(dc, set, params->stream_count, context, DC_VALIDATE_MODE_AND_PROGRAMMING);
/*
* Only update link encoder to stream assignment after bandwidth validation passed.
*/
- if (res == DC_OK && dc->res_pool->funcs->link_encs_assign)
+ if (res == DC_OK && dc->res_pool->funcs->link_encs_assign && !dc->config.unify_link_enc_assignment)
dc->res_pool->funcs->link_encs_assign(
dc, context, context->streams, context->stream_count);
@@ -2487,7 +2685,7 @@ static enum surface_update_type get_plane_info_update_type(const struct dc *dc,
if (memcmp(&u->plane_info->tiling_info, &u->surface->tiling_info,
- sizeof(union dc_tiling_info)) != 0) {
+ sizeof(struct dc_tiling_info)) != 0) {
update_flags->bits.swizzle_change = 1;
elevate_update_type(&update_type, UPDATE_TYPE_MED);
@@ -2704,7 +2902,7 @@ static enum surface_update_type check_update_surfaces_for_stream(
int i;
enum surface_update_type overall_type = UPDATE_TYPE_FAST;
- if (dc->idle_optimizations_allowed)
+ if (dc->idle_optimizations_allowed || dc_can_clear_cursor_limit(dc))
overall_type = UPDATE_TYPE_FULL;
if (stream_status == NULL || stream_status->plane_count != surface_count)
@@ -2759,10 +2957,13 @@ static enum surface_update_type check_update_surfaces_for_stream(
if (stream_update->sharpening_required)
su_flags->bits.sharpening_required = 1;
+ if (stream_update->output_color_space)
+ su_flags->bits.out_csc = 1;
+
if (su_flags->raw != 0)
overall_type = UPDATE_TYPE_FULL;
- if (stream_update->output_csc_transform || stream_update->output_color_space)
+ if (stream_update->output_csc_transform)
su_flags->bits.out_csc = 1;
/* Output transfer function changes do not require bandwidth recalculation,
@@ -2982,6 +3183,10 @@ static void copy_surface_update_to_plane(
if (srf_update->cursor_csc_color_matrix)
surface->cursor_csc_color_matrix =
*srf_update->cursor_csc_color_matrix;
+
+ if (srf_update->bias_and_scale.bias_and_scale_valid)
+ surface->bias_and_scale =
+ srf_update->bias_and_scale;
}
static void copy_stream_update_to_stream(struct dc *dc,
@@ -3048,8 +3253,14 @@ static void copy_stream_update_to_stream(struct dc *dc,
if (update->vrr_active_fixed)
stream->vrr_active_fixed = *update->vrr_active_fixed;
- if (update->crtc_timing_adjust)
+ if (update->crtc_timing_adjust) {
+ if (stream->adjust.v_total_min != update->crtc_timing_adjust->v_total_min ||
+ stream->adjust.v_total_max != update->crtc_timing_adjust->v_total_max ||
+ stream->adjust.timing_adjust_pending)
+ update->crtc_timing_adjust->timing_adjust_pending = true;
stream->adjust = *update->crtc_timing_adjust;
+ update->crtc_timing_adjust->timing_adjust_pending = false;
+ }
if (update->dpms_off)
stream->dpms_off = *update->dpms_off;
@@ -3096,7 +3307,8 @@ static void copy_stream_update_to_stream(struct dc *dc,
if (dsc_validate_context) {
stream->timing.dsc_cfg = *update->dsc_config;
stream->timing.flags.DSC = enable_dsc;
- if (!dc->res_pool->funcs->validate_bandwidth(dc, dsc_validate_context, true)) {
+ if (dc->res_pool->funcs->validate_bandwidth(dc, dsc_validate_context,
+ DC_VALIDATE_MODE_ONLY) != DC_OK) {
stream->timing.dsc_cfg = old_dsc_cfg;
stream->timing.flags.DSC = old_dsc_enabled;
update->dsc_config = NULL;
@@ -3125,7 +3337,7 @@ static void backup_planes_and_stream_state(
return;
for (i = 0; i < status->plane_count; i++) {
- scratch->plane_states[i] = *status->plane_states[i];
+ dc_plane_copy_config(&scratch->plane_states[i], status->plane_states[i]);
}
scratch->stream_state = *stream;
}
@@ -3141,10 +3353,7 @@ static void restore_planes_and_stream_state(
return;
for (i = 0; i < status->plane_count; i++) {
- /* refcount will always be valid, restore everything else */
- struct kref refcount = status->plane_states[i]->refcount;
- *status->plane_states[i] = scratch->plane_states[i];
- status->plane_states[i]->refcount = refcount;
+ dc_plane_copy_config(status->plane_states[i], &scratch->plane_states[i]);
}
*stream = scratch->stream_state;
}
@@ -3321,7 +3530,7 @@ static bool update_planes_and_stream_state(struct dc *dc,
}
if (update_type == UPDATE_TYPE_FULL) {
- if (!dc->res_pool->funcs->validate_bandwidth(dc, context, false)) {
+ if (dc->res_pool->funcs->validate_bandwidth(dc, context, DC_VALIDATE_MODE_AND_PROGRAMMING) != DC_OK) {
BREAK_TO_DEBUGGER();
goto fail;
}
@@ -3841,6 +4050,9 @@ static void commit_planes_for_stream(struct dc *dc,
if (update_type == UPDATE_TYPE_FULL && dc->optimized_required)
hwss_process_outstanding_hw_updates(dc, dc->current_state);
+ if (update_type != UPDATE_TYPE_FAST && dc->res_pool->funcs->prepare_mcache_programming)
+ dc->res_pool->funcs->prepare_mcache_programming(dc, context);
+
for (i = 0; i < dc->res_pool->pipe_count; i++) {
struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
@@ -3872,6 +4084,7 @@ static void commit_planes_for_stream(struct dc *dc,
&context->res_ctx,
stream);
ASSERT(top_pipe_to_program != NULL);
+
for (i = 0; i < dc->res_pool->pipe_count; i++) {
struct pipe_ctx *old_pipe = &dc->current_state->res_ctx.pipe_ctx[i];
@@ -3899,9 +4112,6 @@ static void commit_planes_for_stream(struct dc *dc,
odm_pipe->ttu_regs.min_ttu_vblank = MAX_TTU;
}
- if (update_type != UPDATE_TYPE_FAST && dc->res_pool->funcs->prepare_mcache_programming)
- dc->res_pool->funcs->prepare_mcache_programming(dc, context);
-
if ((update_type != UPDATE_TYPE_FAST) && stream->update_flags.bits.dsc_changed)
if (top_pipe_to_program &&
top_pipe_to_program->stream_res.tg->funcs->lock_doublebuffer_enable) {
@@ -3925,6 +4135,9 @@ static void commit_planes_for_stream(struct dc *dc,
dc->hwss.wait_for_dcc_meta_propagation(dc, top_pipe_to_program);
}
+ if (dc->hwseq->funcs.wait_for_pipe_update_if_needed)
+ dc->hwseq->funcs.wait_for_pipe_update_if_needed(dc, top_pipe_to_program, update_type == UPDATE_TYPE_FAST);
+
if (should_lock_all_pipes && dc->hwss.interdependent_update_lock) {
if (dc->hwss.subvp_pipe_control_lock)
dc->hwss.subvp_pipe_control_lock(dc, context, true, should_lock_all_pipes, NULL, subvp_prev_use);
@@ -4045,12 +4258,6 @@ static void commit_planes_for_stream(struct dc *dc,
if (update_type == UPDATE_TYPE_FAST)
continue;
- ASSERT(!pipe_ctx->plane_state->triplebuffer_flips);
- if (dc->hwss.program_triplebuffer != NULL && dc->debug.enable_tri_buf) {
- /*turn off triple buffer for full update*/
- dc->hwss.program_triplebuffer(
- dc, pipe_ctx, pipe_ctx->plane_state->triplebuffer_flips);
- }
stream_status =
stream_get_status(context, pipe_ctx->stream);
@@ -4059,8 +4266,37 @@ static void commit_planes_for_stream(struct dc *dc,
dc, pipe_ctx->stream, stream_status->plane_count, context);
}
}
+
+ for (j = 0; j < dc->res_pool->pipe_count; j++) {
+ struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j];
+
+ if (!pipe_ctx->plane_state)
+ continue;
+
+ /* Full fe update*/
+ if (update_type == UPDATE_TYPE_FAST)
+ continue;
+
+ ASSERT(!pipe_ctx->plane_state->triplebuffer_flips);
+ if (dc->hwss.program_triplebuffer != NULL && dc->debug.enable_tri_buf) {
+ /*turn off triple buffer for full update*/
+ dc->hwss.program_triplebuffer(
+ dc, pipe_ctx, pipe_ctx->plane_state->triplebuffer_flips);
+ }
+ }
+
if (dc->hwss.program_front_end_for_ctx && update_type != UPDATE_TYPE_FAST) {
dc->hwss.program_front_end_for_ctx(dc, context);
+
+ //Pipe busy until some frame and line #
+ if (dc->hwseq->funcs.set_wait_for_update_needed_for_pipe && update_type == UPDATE_TYPE_FULL) {
+ for (j = 0; j < dc->res_pool->pipe_count; j++) {
+ struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j];
+
+ dc->hwseq->funcs.set_wait_for_update_needed_for_pipe(dc, pipe_ctx);
+ }
+ }
+
if (dc->debug.validate_dml_output) {
for (i = 0; i < dc->res_pool->pipe_count; i++) {
struct pipe_ctx *cur_pipe = &context->res_ctx.pipe_ctx[i];
@@ -4400,7 +4636,8 @@ static struct dc_state *create_minimal_transition_state(struct dc *dc,
backup_and_set_minimal_pipe_split_policy(dc, base_context, policy);
/* commit minimal state */
- if (dc->res_pool->funcs->validate_bandwidth(dc, minimal_transition_context, false)) {
+ if (dc->res_pool->funcs->validate_bandwidth(dc, minimal_transition_context,
+ DC_VALIDATE_MODE_AND_PROGRAMMING) == DC_OK) {
/* prevent underflow and corruption when reconfiguring pipes */
force_vsync_flip_in_minimal_transition_context(minimal_transition_context);
} else {
@@ -4510,7 +4747,7 @@ static bool commit_minimal_transition_based_on_current_context(struct dc *dc,
struct pipe_split_policy_backup policy;
struct dc_state *intermediate_context;
struct dc_state *old_current_state = dc->current_state;
- struct dc_surface_update srf_updates[MAX_SURFACE_NUM] = {0};
+ struct dc_surface_update srf_updates[MAX_SURFACES] = {0};
int surface_count;
/*
@@ -4823,7 +5060,8 @@ static bool full_update_required(struct dc *dc,
stream_update->lut3d_func ||
stream_update->pending_test_pattern ||
stream_update->crtc_timing_adjust ||
- stream_update->scaler_sharpener_update))
+ stream_update->scaler_sharpener_update ||
+ stream_update->hw_cursor_req))
return true;
if (stream) {
@@ -4834,6 +5072,9 @@ static bool full_update_required(struct dc *dc,
if (dc->idle_optimizations_allowed)
return true;
+ if (dc_can_clear_cursor_limit(dc))
+ return true;
+
return false;
}
@@ -4919,7 +5160,7 @@ static bool update_planes_and_stream_v1(struct dc *dc,
copy_stream_update_to_stream(dc, context, stream, stream_update);
if (update_type >= UPDATE_TYPE_FULL) {
- if (!dc->res_pool->funcs->validate_bandwidth(dc, context, false)) {
+ if (dc->res_pool->funcs->validate_bandwidth(dc, context, DC_VALIDATE_MODE_AND_PROGRAMMING) != DC_OK) {
DC_ERROR("Mode validation failed for stream update!\n");
dc_state_release(context);
return false;
@@ -5203,8 +5444,8 @@ bool dc_update_planes_and_stream(struct dc *dc,
else
ret = update_planes_and_stream_v2(dc, srf_updates,
surface_count, stream, stream_update);
-
- if (ret)
+ if (ret && (dc->ctx->dce_version >= DCN_VERSION_3_2 ||
+ dc->ctx->dce_version == DCN_VERSION_3_01))
clear_update_flags(srf_updates, surface_count, stream);
return ret;
@@ -5235,7 +5476,7 @@ void dc_commit_updates_for_stream(struct dc *dc,
ret = update_planes_and_stream_v1(dc, srf_updates, surface_count, stream,
stream_update, state);
- if (ret)
+ if (ret && dc->ctx->dce_version >= DCN_VERSION_3_2)
clear_update_flags(srf_updates, surface_count, stream);
}
@@ -5307,11 +5548,18 @@ void dc_set_power_state(struct dc *dc, enum dc_acpi_cm_power_state power_state)
dc->vm_pa_config.valid) {
dc->hwss.init_sys_ctx(dc->hwseq, dc, &dc->vm_pa_config);
}
+ break;
+ case DC_ACPI_CM_POWER_STATE_D3:
+ if (dc->caps.ips_support)
+ dc_dmub_srv_notify_fw_dc_power_state(dc->ctx->dmub_srv, DC_ACPI_CM_POWER_STATE_D3);
+ if (dc->caps.ips_v2_support) {
+ if (dc->clk_mgr->funcs->set_low_power_state)
+ dc->clk_mgr->funcs->set_low_power_state(dc->clk_mgr);
+ }
break;
default:
ASSERT(dc->current_state->stream_count == 0);
-
dc_dmub_srv_notify_fw_dc_power_state(dc->ctx->dmub_srv, power_state);
dc_state_destruct(dc->current_state);
@@ -5337,18 +5585,6 @@ bool dc_is_dmcu_initialized(struct dc *dc)
return false;
}
-void get_clock_requirements_for_state(struct dc_state *state, struct AsicStateEx *info)
-{
- info->displayClock = (unsigned int)state->bw_ctx.bw.dcn.clk.dispclk_khz;
- info->engineClock = (unsigned int)state->bw_ctx.bw.dcn.clk.dcfclk_khz;
- info->memoryClock = (unsigned int)state->bw_ctx.bw.dcn.clk.dramclk_khz;
- info->maxSupportedDppClock = (unsigned int)state->bw_ctx.bw.dcn.clk.max_supported_dppclk_khz;
- info->dppClock = (unsigned int)state->bw_ctx.bw.dcn.clk.dppclk_khz;
- info->socClock = (unsigned int)state->bw_ctx.bw.dcn.clk.socclk_khz;
- info->dcfClockDeepSleep = (unsigned int)state->bw_ctx.bw.dcn.clk.dcfclk_deep_sleep_khz;
- info->fClock = (unsigned int)state->bw_ctx.bw.dcn.clk.fclk_khz;
- info->phyClock = (unsigned int)state->bw_ctx.bw.dcn.clk.phyclk_khz;
-}
enum dc_status dc_set_clock(struct dc *dc, enum dc_clock_type clock_type, uint32_t clk_khz, uint32_t stepping)
{
if (dc->hwss.set_clock)
@@ -5435,6 +5671,11 @@ bool dc_set_ips_disable(struct dc *dc, unsigned int disable_ips)
void dc_allow_idle_optimizations_internal(struct dc *dc, bool allow, char const *caller_name)
{
+ int idle_fclk_khz = 0, idle_dramclk_khz = 0, i = 0;
+ enum mall_stream_type subvp_pipe_type[MAX_PIPES] = {0};
+ struct pipe_ctx *pipe = NULL;
+ struct dc_state *context = dc->current_state;
+
if (dc->debug.disable_idle_power_optimizations) {
DC_LOG_DEBUG("%s: disabled\n", __func__);
return;
@@ -5459,6 +5700,25 @@ void dc_allow_idle_optimizations_internal(struct dc *dc, bool allow, char const
dc->idle_optimizations_allowed = allow;
DC_LOG_DEBUG("%s: %s\n", __func__, allow ? "enabled" : "disabled");
}
+
+ // log idle clocks and sub vp pipe types at idle optimization time
+ if (dc->clk_mgr != NULL && dc->clk_mgr->funcs->get_hard_min_fclk)
+ idle_fclk_khz = dc->clk_mgr->funcs->get_hard_min_fclk(dc->clk_mgr);
+
+ if (dc->clk_mgr != NULL && dc->clk_mgr->funcs->get_hard_min_memclk)
+ idle_dramclk_khz = dc->clk_mgr->funcs->get_hard_min_memclk(dc->clk_mgr);
+
+ if (dc->res_pool && context) {
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ pipe = &context->res_ctx.pipe_ctx[i];
+ subvp_pipe_type[i] = dc_state_get_pipe_subvp_type(context, pipe);
+ }
+ }
+
+ DC_LOG_DC("%s: allow_idle=%d\n HardMinUClk_Khz=%d HardMinDramclk_Khz=%d\n Pipe_0=%d Pipe_1=%d Pipe_2=%d Pipe_3=%d Pipe_4=%d Pipe_5=%d (caller=%s)\n",
+ __func__, allow, idle_fclk_khz, idle_dramclk_khz, subvp_pipe_type[0], subvp_pipe_type[1], subvp_pipe_type[2],
+ subvp_pipe_type[3], subvp_pipe_type[4], subvp_pipe_type[5], caller_name);
+
}
void dc_exit_ips_for_hw_access_internal(struct dc *dc, const char *caller_name)
@@ -6053,15 +6313,22 @@ bool dc_abm_save_restore(
void dc_query_current_properties(struct dc *dc, struct dc_current_properties *properties)
{
unsigned int i;
- bool subvp_sw_cursor_req = false;
+ unsigned int max_cursor_size = dc->caps.max_cursor_size;
+ unsigned int stream_cursor_size;
- for (i = 0; i < dc->current_state->stream_count; i++) {
- if (check_subvp_sw_cursor_fallback_req(dc, dc->current_state->streams[i])) {
- subvp_sw_cursor_req = true;
- break;
+ if (dc->debug.allow_sw_cursor_fallback && dc->res_pool->funcs->get_max_hw_cursor_size) {
+ for (i = 0; i < dc->current_state->stream_count; i++) {
+ stream_cursor_size = dc->res_pool->funcs->get_max_hw_cursor_size(dc,
+ dc->current_state,
+ dc->current_state->streams[i]);
+
+ if (stream_cursor_size < max_cursor_size) {
+ max_cursor_size = stream_cursor_size;
+ }
}
}
- properties->cursor_size_limit = subvp_sw_cursor_req ? 64 : dc->caps.max_cursor_size;
+
+ properties->cursor_size_limit = max_cursor_size;
}
/**
@@ -6088,13 +6355,14 @@ void dc_set_edp_power(const struct dc *dc, struct dc_link *edp_link,
edp_link->dc->link_srv->edp_set_panel_power(edp_link, powerOn);
}
-/*
- *****************************************************************************
+/**
* dc_get_power_profile_for_dc_state() - extracts power profile from dc state
*
* Called when DM wants to make power policy decisions based on dc_state
*
- *****************************************************************************
+ * @context: Pointer to the dc_state from which the power profile is extracted.
+ *
+ * Return: The power profile structure containing the power level information.
*/
struct dc_power_profile dc_get_power_profile_for_dc_state(const struct dc_state *context)
{
@@ -6109,3 +6377,76 @@ struct dc_power_profile dc_get_power_profile_for_dc_state(const struct dc_state
profile.power_level = dc->res_pool->funcs->get_power_profile(context);
return profile;
}
+
+/**
+ * dc_get_det_buffer_size_from_state() - extracts detile buffer size from dc state
+ *
+ * This function is called to log the detile buffer size from the dc_state.
+ *
+ * @context: a pointer to the dc_state from which the detile buffer size is extracted.
+ *
+ * Return: the size of the detile buffer, or 0 if not available.
+ */
+unsigned int dc_get_det_buffer_size_from_state(const struct dc_state *context)
+{
+ struct dc *dc = context->clk_mgr->ctx->dc;
+
+ if (dc->res_pool->funcs->get_det_buffer_size)
+ return dc->res_pool->funcs->get_det_buffer_size(context);
+ else
+ return 0;
+}
+
+/**
+ * dc_get_host_router_index: Get index of host router from a dpia link
+ *
+ * This function return a host router index of the target link. If the target link is dpia link.
+ *
+ * @link: Pointer to the target link (input)
+ * @host_router_index: Pointer to store the host router index of the target link (output).
+ *
+ * Return: true if the host router index is found and valid.
+ *
+ */
+bool dc_get_host_router_index(const struct dc_link *link, unsigned int *host_router_index)
+{
+ struct dc *dc;
+
+ if (!link || !host_router_index || link->ep_type != DISPLAY_ENDPOINT_USB4_DPIA)
+ return false;
+
+ dc = link->ctx->dc;
+
+ if (link->link_index < dc->lowest_dpia_link_index)
+ return false;
+
+ *host_router_index = (link->link_index - dc->lowest_dpia_link_index) / dc->caps.num_of_dpias_per_host_router;
+ if (*host_router_index < dc->caps.num_of_host_routers)
+ return true;
+ else
+ return false;
+}
+
+bool dc_is_cursor_limit_pending(struct dc *dc)
+{
+ uint32_t i;
+
+ for (i = 0; i < dc->current_state->stream_count; i++) {
+ if (dc_stream_is_cursor_limit_pending(dc, dc->current_state->streams[i]))
+ return true;
+ }
+
+ return false;
+}
+
+bool dc_can_clear_cursor_limit(struct dc *dc)
+{
+ uint32_t i;
+
+ for (i = 0; i < dc->current_state->stream_count; i++) {
+ if (dc_state_can_clear_stream_cursor_subvp_limit(dc->current_state->streams[i], dc->current_state))
+ return true;
+ }
+
+ return false;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
index af1ea5792560..bbce751b485f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_debug.c
@@ -51,126 +51,6 @@
DC_LOG_BANDWIDTH_CALCS(__VA_ARGS__); \
} while (0)
-void pre_surface_trace(
- struct dc *dc,
- const struct dc_plane_state *const *plane_states,
- int surface_count)
-{
- int i;
- DC_LOGGER_INIT(dc->ctx->logger);
-
- for (i = 0; i < surface_count; i++) {
- const struct dc_plane_state *plane_state = plane_states[i];
-
- SURFACE_TRACE("Planes %d:\n", i);
-
- SURFACE_TRACE(
- "plane_state->visible = %d;\n"
- "plane_state->flip_immediate = %d;\n"
- "plane_state->address.type = %d;\n"
- "plane_state->address.grph.addr.quad_part = 0x%llX;\n"
- "plane_state->address.grph.meta_addr.quad_part = 0x%llX;\n"
- "plane_state->scaling_quality.h_taps = %d;\n"
- "plane_state->scaling_quality.v_taps = %d;\n"
- "plane_state->scaling_quality.h_taps_c = %d;\n"
- "plane_state->scaling_quality.v_taps_c = %d;\n",
- plane_state->visible,
- plane_state->flip_immediate,
- plane_state->address.type,
- plane_state->address.grph.addr.quad_part,
- plane_state->address.grph.meta_addr.quad_part,
- plane_state->scaling_quality.h_taps,
- plane_state->scaling_quality.v_taps,
- plane_state->scaling_quality.h_taps_c,
- plane_state->scaling_quality.v_taps_c);
-
- SURFACE_TRACE(
- "plane_state->src_rect.x = %d;\n"
- "plane_state->src_rect.y = %d;\n"
- "plane_state->src_rect.width = %d;\n"
- "plane_state->src_rect.height = %d;\n"
- "plane_state->dst_rect.x = %d;\n"
- "plane_state->dst_rect.y = %d;\n"
- "plane_state->dst_rect.width = %d;\n"
- "plane_state->dst_rect.height = %d;\n"
- "plane_state->clip_rect.x = %d;\n"
- "plane_state->clip_rect.y = %d;\n"
- "plane_state->clip_rect.width = %d;\n"
- "plane_state->clip_rect.height = %d;\n",
- plane_state->src_rect.x,
- plane_state->src_rect.y,
- plane_state->src_rect.width,
- plane_state->src_rect.height,
- plane_state->dst_rect.x,
- plane_state->dst_rect.y,
- plane_state->dst_rect.width,
- plane_state->dst_rect.height,
- plane_state->clip_rect.x,
- plane_state->clip_rect.y,
- plane_state->clip_rect.width,
- plane_state->clip_rect.height);
-
- SURFACE_TRACE(
- "plane_state->plane_size.surface_size.x = %d;\n"
- "plane_state->plane_size.surface_size.y = %d;\n"
- "plane_state->plane_size.surface_size.width = %d;\n"
- "plane_state->plane_size.surface_size.height = %d;\n"
- "plane_state->plane_size.surface_pitch = %d;\n",
- plane_state->plane_size.surface_size.x,
- plane_state->plane_size.surface_size.y,
- plane_state->plane_size.surface_size.width,
- plane_state->plane_size.surface_size.height,
- plane_state->plane_size.surface_pitch);
-
-
- SURFACE_TRACE(
- "plane_state->tiling_info.gfx8.num_banks = %d;\n"
- "plane_state->tiling_info.gfx8.bank_width = %d;\n"
- "plane_state->tiling_info.gfx8.bank_width_c = %d;\n"
- "plane_state->tiling_info.gfx8.bank_height = %d;\n"
- "plane_state->tiling_info.gfx8.bank_height_c = %d;\n"
- "plane_state->tiling_info.gfx8.tile_aspect = %d;\n"
- "plane_state->tiling_info.gfx8.tile_aspect_c = %d;\n"
- "plane_state->tiling_info.gfx8.tile_split = %d;\n"
- "plane_state->tiling_info.gfx8.tile_split_c = %d;\n"
- "plane_state->tiling_info.gfx8.tile_mode = %d;\n"
- "plane_state->tiling_info.gfx8.tile_mode_c = %d;\n",
- plane_state->tiling_info.gfx8.num_banks,
- plane_state->tiling_info.gfx8.bank_width,
- plane_state->tiling_info.gfx8.bank_width_c,
- plane_state->tiling_info.gfx8.bank_height,
- plane_state->tiling_info.gfx8.bank_height_c,
- plane_state->tiling_info.gfx8.tile_aspect,
- plane_state->tiling_info.gfx8.tile_aspect_c,
- plane_state->tiling_info.gfx8.tile_split,
- plane_state->tiling_info.gfx8.tile_split_c,
- plane_state->tiling_info.gfx8.tile_mode,
- plane_state->tiling_info.gfx8.tile_mode_c);
-
- SURFACE_TRACE(
- "plane_state->tiling_info.gfx8.pipe_config = %d;\n"
- "plane_state->tiling_info.gfx8.array_mode = %d;\n"
- "plane_state->color_space = %d;\n"
- "plane_state->dcc.enable = %d;\n"
- "plane_state->format = %d;\n"
- "plane_state->rotation = %d;\n"
- "plane_state->stereo_format = %d;\n",
- plane_state->tiling_info.gfx8.pipe_config,
- plane_state->tiling_info.gfx8.array_mode,
- plane_state->color_space,
- plane_state->dcc.enable,
- plane_state->format,
- plane_state->rotation,
- plane_state->stereo_format);
-
- SURFACE_TRACE("plane_state->tiling_info.gfx9.swizzle = %d;\n",
- plane_state->tiling_info.gfx9.swizzle);
-
- SURFACE_TRACE("\n");
- }
- SURFACE_TRACE("\n");
-}
-
void update_surface_trace(
struct dc *dc,
const struct dc_surface_update *updates,
@@ -386,6 +266,10 @@ char *dc_status_to_str(enum dc_status status)
return "Fail dp payload allocation";
case DC_FAIL_DP_LINK_BANDWIDTH:
return "Insufficient DP link bandwidth";
+ case DC_FAIL_HW_CURSOR_SUPPORT:
+ return "HW Cursor not supported";
+ case DC_FAIL_DP_TUNNEL_BW_VALIDATE:
+ return "Fail DP Tunnel BW validation";
case DC_ERROR_UNEXPECTED:
return "Unexpected error";
}
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
index 252af83e34a5..ec4e80e5b6eb 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_hw_sequencer.c
@@ -34,6 +34,7 @@
#include "dc_state_priv.h"
#define NUM_ELEMENTS(a) (sizeof(a) / sizeof((a)[0]))
+#define MAX_NUM_MCACHE 8
/* used as index in array of black_color_format */
enum black_color_format {
@@ -176,7 +177,7 @@ static bool is_ycbcr2020_type(
{
bool ret = false;
- if (color_space == COLOR_SPACE_2020_YCBCR)
+ if (color_space == COLOR_SPACE_2020_YCBCR_LIMITED || color_space == COLOR_SPACE_2020_YCBCR_FULL)
ret = true;
return ret;
}
@@ -247,7 +248,8 @@ void color_space_to_black_color(
case COLOR_SPACE_YCBCR709_BLACK:
case COLOR_SPACE_YCBCR601_LIMITED:
case COLOR_SPACE_YCBCR709_LIMITED:
- case COLOR_SPACE_2020_YCBCR:
+ case COLOR_SPACE_2020_YCBCR_LIMITED:
+ case COLOR_SPACE_2020_YCBCR_FULL:
*black_color = black_color_format[BLACK_COLOR_FORMAT_YUV_CV];
break;
@@ -425,6 +427,70 @@ void get_hdr_visual_confirm_color(
}
}
+/* Visual Confirm color definition for Smart Mux */
+void get_smartmux_visual_confirm_color(
+ struct dc *dc,
+ struct tg_color *color)
+{
+ uint32_t color_value = MAX_TG_COLOR_VALUE;
+
+ const struct tg_color sm_ver_colors[5] = {
+ {0, 0, 0}, /* SMUX_MUXCONTROL_UNSUPPORTED - Black */
+ {0, MAX_TG_COLOR_VALUE, 0}, /* SMUX_MUXCONTROL_v10 - Green */
+ {0, MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE}, /* SMUX_MUXCONTROL_v15 - Cyan */
+ {MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE, 0}, /* SMUX_MUXCONTROL_MDM - Yellow */
+ {MAX_TG_COLOR_VALUE, 0, MAX_TG_COLOR_VALUE}, /* SMUX_MUXCONTROL_vUNKNOWN - Magenta*/
+ };
+
+ if (dc->caps.is_apu) {
+ /* APU driving the eDP */
+ *color = sm_ver_colors[dc->config.smart_mux_version];
+ } else {
+ /* dGPU driving the eDP - red */
+ color->color_r_cr = color_value;
+ color->color_g_y = 0;
+ color->color_b_cb = 0;
+ }
+}
+
+/* Visual Confirm color definition for VABC */
+void get_vabc_visual_confirm_color(
+ struct pipe_ctx *pipe_ctx,
+ struct tg_color *color)
+{
+ uint32_t color_value = MAX_TG_COLOR_VALUE;
+ struct dc_link *edp_link = NULL;
+
+ if (pipe_ctx && pipe_ctx->stream && pipe_ctx->stream->link) {
+ if (pipe_ctx->stream->link->connector_signal == SIGNAL_TYPE_EDP)
+ edp_link = pipe_ctx->stream->link;
+ }
+
+ if (edp_link) {
+ switch (edp_link->backlight_control_type) {
+ case BACKLIGHT_CONTROL_PWM:
+ color->color_r_cr = color_value;
+ color->color_g_y = 0;
+ color->color_b_cb = 0;
+ break;
+ case BACKLIGHT_CONTROL_AMD_AUX:
+ color->color_r_cr = 0;
+ color->color_g_y = color_value;
+ color->color_b_cb = 0;
+ break;
+ case BACKLIGHT_CONTROL_VESA_AUX:
+ color->color_r_cr = 0;
+ color->color_g_y = 0;
+ color->color_b_cb = color_value;
+ break;
+ }
+ } else {
+ color->color_r_cr = 0;
+ color->color_g_y = 0;
+ color->color_b_cb = 0;
+ }
+}
+
void get_subvp_visual_confirm_color(
struct pipe_ctx *pipe_ctx,
struct tg_color *color)
@@ -514,6 +580,53 @@ void get_cursor_visual_confirm_color(
}
}
+void get_dcc_visual_confirm_color(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct tg_color *color)
+{
+ const uint32_t MCACHE_ID_UNASSIGNED = 0xF;
+
+ if (!pipe_ctx->plane_state->dcc.enable) {
+ color->color_r_cr = 0; /* black - DCC disabled */
+ color->color_g_y = 0;
+ color->color_b_cb = 0;
+ return;
+ }
+
+ if (dc->ctx->dce_version < DCN_VERSION_4_01) {
+ color->color_r_cr = MAX_TG_COLOR_VALUE; /* red - DCC enabled */
+ color->color_g_y = 0;
+ color->color_b_cb = 0;
+ return;
+ }
+
+ uint32_t first_id = pipe_ctx->mcache_regs.main.p0.mcache_id_first;
+ uint32_t second_id = pipe_ctx->mcache_regs.main.p0.mcache_id_second;
+
+ if (first_id != MCACHE_ID_UNASSIGNED && second_id != MCACHE_ID_UNASSIGNED && first_id != second_id) {
+ color->color_r_cr = MAX_TG_COLOR_VALUE/2; /* grey - 2 mcache */
+ color->color_g_y = MAX_TG_COLOR_VALUE/2;
+ color->color_b_cb = MAX_TG_COLOR_VALUE/2;
+ }
+
+ else if (first_id != MCACHE_ID_UNASSIGNED || second_id != MCACHE_ID_UNASSIGNED) {
+ const struct tg_color id_colors[MAX_NUM_MCACHE] = {
+ {0, MAX_TG_COLOR_VALUE, 0}, /* green */
+ {0, 0, MAX_TG_COLOR_VALUE}, /* blue */
+ {MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE, 0}, /* yellow */
+ {MAX_TG_COLOR_VALUE, 0, MAX_TG_COLOR_VALUE}, /* magenta */
+ {0, MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE}, /* cyan */
+ {MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE, MAX_TG_COLOR_VALUE}, /* white */
+ {MAX_TG_COLOR_VALUE/2, 0, 0}, /* dark red */
+ {0, MAX_TG_COLOR_VALUE/2, 0}, /* dark green */
+ };
+
+ uint32_t assigned_id = (first_id != MCACHE_ID_UNASSIGNED) ? first_id : second_id;
+ *color = id_colors[assigned_id];
+ }
+}
+
void set_p_state_switch_method(
struct dc *dc,
struct dc_state *context,
@@ -525,6 +638,7 @@ void set_p_state_switch_method(
if (!dc->ctx || !dc->ctx->dmub_srv || !pipe_ctx || !vba)
return;
+ pipe_ctx->p_state_type = P_STATE_UNKNOWN;
if (vba->DRAMClockChangeSupport[vba->VoltageLevel][vba->maxMpcComb] !=
dm_dram_clock_change_unsupported) {
/* MCLK switching is supported */
@@ -571,6 +685,21 @@ void set_p_state_switch_method(
}
}
+void set_drr_and_clear_adjust_pending(
+ struct pipe_ctx *pipe_ctx,
+ struct dc_stream_state *stream,
+ struct drr_params *params)
+{
+ /* params can be null.*/
+ if (pipe_ctx && pipe_ctx->stream_res.tg &&
+ pipe_ctx->stream_res.tg->funcs->set_drr)
+ pipe_ctx->stream_res.tg->funcs->set_drr(
+ pipe_ctx->stream_res.tg, params);
+
+ if (stream)
+ stream->adjust.timing_adjust_pending = false;
+}
+
void get_fams2_visual_confirm_color(
struct dc *dc,
struct dc_state *context,
@@ -594,7 +723,7 @@ void get_fams2_visual_confirm_color(
void hwss_build_fast_sequence(struct dc *dc,
struct dc_dmub_cmd *dc_dmub_cmd,
unsigned int dmub_cmd_count,
- struct block_sequence block_sequence[],
+ struct block_sequence block_sequence[MAX_HWSS_BLOCK_SEQUENCE_SIZE],
unsigned int *num_steps,
struct pipe_ctx *pipe_ctx,
struct dc_stream_status *stream_status,
@@ -714,7 +843,14 @@ void hwss_build_fast_sequence(struct dc *dc,
block_sequence[*num_steps].func = DPP_SET_OUTPUT_TRANSFER_FUNC;
(*num_steps)++;
}
-
+ if (dc->debug.visual_confirm != VISUAL_CONFIRM_DISABLE &&
+ dc->hwss.update_visual_confirm_color) {
+ block_sequence[*num_steps].params.update_visual_confirm_params.dc = dc;
+ block_sequence[*num_steps].params.update_visual_confirm_params.pipe_ctx = current_mpc_pipe;
+ block_sequence[*num_steps].params.update_visual_confirm_params.mpcc_id = current_mpc_pipe->plane_res.hubp->inst;
+ block_sequence[*num_steps].func = MPC_UPDATE_VISUAL_CONFIRM;
+ (*num_steps)++;
+ }
if (current_mpc_pipe->stream->update_flags.bits.out_csc) {
block_sequence[*num_steps].params.power_on_mpc_mem_pwr_params.mpc = dc->res_pool->mpc;
block_sequence[*num_steps].params.power_on_mpc_mem_pwr_params.mpcc_id = current_mpc_pipe->plane_res.hubp->inst;
@@ -786,7 +922,7 @@ void hwss_build_fast_sequence(struct dc *dc,
}
void hwss_execute_sequence(struct dc *dc,
- struct block_sequence block_sequence[],
+ struct block_sequence block_sequence[MAX_HWSS_BLOCK_SEQUENCE_SIZE],
int num_steps)
{
unsigned int i;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c
index 039b176e086d..814f68d76257 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_enc_cfg.c
@@ -44,20 +44,8 @@ static bool is_dig_link_enc_stream(struct dc_stream_state *stream)
* yet match.
*/
if (link_enc && ((uint32_t)stream->link->connector_signal & link_enc->output_signals)) {
- if (dc_is_dp_signal(stream->signal)) {
- /* DIGs do not support DP2.0 streams with 128b/132b encoding. */
- struct dc_link_settings link_settings = {0};
-
- stream->ctx->dc->link_srv->dp_decide_link_settings(stream, &link_settings);
- if ((link_settings.link_rate >= LINK_RATE_LOW) &&
- link_settings.link_rate <= LINK_RATE_HIGH3) {
- is_dig_stream = true;
- break;
- }
- } else {
- is_dig_stream = true;
- break;
- }
+ is_dig_stream = true;
+ break;
}
}
}
@@ -559,17 +547,6 @@ struct link_encoder *link_enc_cfg_get_next_avail_link_enc(struct dc *dc)
return link_enc;
}
-struct link_encoder *link_enc_cfg_get_link_enc_used_by_stream(
- struct dc *dc,
- const struct dc_stream_state *stream)
-{
- struct link_encoder *link_enc;
-
- link_enc = link_enc_cfg_get_link_enc_used_by_link(dc, stream->link);
-
- return link_enc;
-}
-
struct link_encoder *link_enc_cfg_get_link_enc(
const struct dc_link *link)
{
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_exports.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_exports.c
index 457d60eeb486..130455f2802a 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_exports.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_exports.c
@@ -125,6 +125,14 @@ uint32_t dc_link_bandwidth_kbps(
return link->dc->link_srv->dp_link_bandwidth_kbps(link, link_settings);
}
+uint32_t dc_link_required_hblank_size_bytes(
+ const struct dc_link *link,
+ struct dp_audio_bandwidth_params *audio_params)
+{
+ return link->dc->link_srv->dp_required_hblank_size_bytes(link,
+ audio_params);
+}
+
void dc_get_cur_link_res_map(const struct dc *dc, uint32_t *map)
{
dc->link_srv->get_cur_res_map(dc, map);
@@ -142,6 +150,12 @@ bool dc_link_update_dsc_config(struct pipe_ctx *pipe_ctx)
return link->dc->link_srv->update_dsc_config(pipe_ctx);
}
+struct ddc_service *
+dc_get_oem_i2c_device(struct dc *dc)
+{
+ return dc->res_pool->oem_device;
+}
+
bool dc_is_oem_i2c_device_present(
struct dc *dc,
size_t slave_address)
@@ -356,15 +370,10 @@ bool dc_link_should_enable_fec(const struct dc_link *link)
return link->dc->link_srv->dp_should_enable_fec(link);
}
-int dc_link_dp_dpia_handle_usb4_bandwidth_allocation_for_link(
+void dc_link_dp_dpia_handle_usb4_bandwidth_allocation_for_link(
struct dc_link *link, int peak_bw)
{
- return link->dc->link_srv->dpia_handle_usb4_bandwidth_allocation_for_link(link, peak_bw);
-}
-
-void dc_link_handle_usb4_bw_alloc_response(struct dc_link *link, uint8_t bw, uint8_t result)
-{
- link->dc->link_srv->dpia_handle_bw_alloc_response(link, bw, result);
+ link->dc->link_srv->dpia_handle_usb4_bandwidth_allocation_for_link(link, peak_bw);
}
bool dc_link_check_link_loss_status(
@@ -506,7 +515,8 @@ void dc_link_enable_hpd_filter(struct dc_link *link, bool enable)
link->dc->link_srv->enable_hpd_filter(link, enable);
}
-bool dc_link_dp_dpia_validate(struct dc *dc, const struct dc_stream_state *streams, const unsigned int count)
+enum dc_status dc_link_validate_dp_tunneling_bandwidth(const struct dc *dc, const struct dc_state *new_ctx)
{
- return dc->link_srv->validate_dpia_bandwidth(streams, count);
+ return dc->link_srv->validate_dp_tunnel_bandwidth(dc, new_ctx);
}
+
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 619fad17de55..4d6181e7c612 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -76,6 +76,7 @@
#include "dcn321/dcn321_resource.h"
#include "dcn35/dcn35_resource.h"
#include "dcn351/dcn351_resource.h"
+#include "dcn36/dcn36_resource.h"
#include "dcn401/dcn401_resource.h"
#if defined(CONFIG_DRM_AMD_DC_FP)
#include "dc_spl_translate.h"
@@ -204,6 +205,8 @@ enum dce_version resource_parse_asic_id(struct hw_asic_id asic_id)
dc_version = DCN_VERSION_3_5;
if (ASICREV_IS_GC_11_0_4(asic_id.hw_internal_rev))
dc_version = DCN_VERSION_3_51;
+ if (ASICREV_IS_DCN36(asic_id.hw_internal_rev))
+ dc_version = DCN_VERSION_3_6;
break;
case AMDGPU_FAMILY_GC_12_0_0:
if (ASICREV_IS_GC_12_0_1_A0(asic_id.hw_internal_rev) ||
@@ -320,6 +323,9 @@ struct resource_pool *dc_create_resource_pool(struct dc *dc,
case DCN_VERSION_3_51:
res_pool = dcn351_create_resource_pool(init_data, dc);
break;
+ case DCN_VERSION_3_6:
+ res_pool = dcn36_create_resource_pool(init_data, dc);
+ break;
case DCN_VERSION_4_01:
res_pool = dcn401_create_resource_pool(init_data, dc);
break;
@@ -941,6 +947,17 @@ static void calculate_adjust_recout_for_visual_confirm(struct pipe_ctx *pipe_ctx
*base_offset = VISUAL_CONFIRM_BASE_DEFAULT;
}
+static void reverse_adjust_recout_for_visual_confirm(struct rect *recout,
+ struct pipe_ctx *pipe_ctx)
+{
+ int dpp_offset, base_offset;
+
+ calculate_adjust_recout_for_visual_confirm(pipe_ctx, &base_offset,
+ &dpp_offset);
+ recout->height += base_offset;
+ recout->height += dpp_offset;
+}
+
static void adjust_recout_for_visual_confirm(struct rect *recout,
struct pipe_ctx *pipe_ctx)
{
@@ -1325,32 +1342,6 @@ static void calculate_inits_and_viewports(struct pipe_ctx *pipe_ctx)
data->viewport_c.y += src.y / vpc_div;
}
-static bool is_subvp_high_refresh_candidate(struct dc_stream_state *stream)
-{
- uint32_t refresh_rate;
- struct dc *dc = stream->ctx->dc;
-
- refresh_rate = (stream->timing.pix_clk_100hz * (uint64_t)100 +
- stream->timing.v_total * stream->timing.h_total - (uint64_t)1);
- refresh_rate = div_u64(refresh_rate, stream->timing.v_total);
- refresh_rate = div_u64(refresh_rate, stream->timing.h_total);
-
- /* If there's any stream that fits the SubVP high refresh criteria,
- * we must return true. This is because cursor updates are asynchronous
- * with full updates, so we could transition into a SubVP config and
- * remain in HW cursor mode if there's no cursor update which will
- * then cause corruption.
- */
- if ((refresh_rate >= 120 && refresh_rate <= 175 &&
- stream->timing.v_addressable >= 1080 &&
- stream->timing.v_addressable <= 2160) &&
- (dc->current_state->stream_count > 1 ||
- (dc->current_state->stream_count == 1 && !stream->allow_freesync)))
- return true;
-
- return false;
-}
-
static enum controller_dp_test_pattern convert_dp_to_controller_test_pattern(
enum dp_test_pattern test_pattern)
{
@@ -1455,7 +1446,8 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
DC_LOGGER_INIT(pipe_ctx->stream->ctx->logger);
/* Invalid input */
- if (!plane_state->dst_rect.width ||
+ if (!plane_state ||
+ !plane_state->dst_rect.width ||
!plane_state->dst_rect.height ||
!plane_state->src_rect.width ||
!plane_state->src_rect.height) {
@@ -1641,6 +1633,62 @@ bool resource_build_scaling_params(struct pipe_ctx *pipe_ctx)
return res;
}
+bool resource_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx)
+{
+ struct pipe_ctx *test_pipe, *split_pipe;
+ struct rect r1 = pipe_ctx->plane_res.scl_data.recout;
+ int r1_right, r1_bottom;
+ int cur_layer = pipe_ctx->plane_state->layer_index;
+
+ reverse_adjust_recout_for_visual_confirm(&r1, pipe_ctx);
+ r1_right = r1.x + r1.width;
+ r1_bottom = r1.y + r1.height;
+
+ /**
+ * Disable the cursor if there's another pipe above this with a
+ * plane that contains this pipe's viewport to prevent double cursor
+ * and incorrect scaling artifacts.
+ */
+ for (test_pipe = pipe_ctx->top_pipe; test_pipe;
+ test_pipe = test_pipe->top_pipe) {
+ struct rect r2;
+ int r2_right, r2_bottom;
+ // Skip invisible layer and pipe-split plane on same layer
+ if (!test_pipe->plane_state ||
+ !test_pipe->plane_state->visible ||
+ test_pipe->plane_state->layer_index == cur_layer)
+ continue;
+
+ r2 = test_pipe->plane_res.scl_data.recout;
+ reverse_adjust_recout_for_visual_confirm(&r2, test_pipe);
+ r2_right = r2.x + r2.width;
+ r2_bottom = r2.y + r2.height;
+
+ /**
+ * There is another half plane on same layer because of
+ * pipe-split, merge together per same height.
+ */
+ for (split_pipe = pipe_ctx->top_pipe; split_pipe;
+ split_pipe = split_pipe->top_pipe)
+ if (split_pipe->plane_state->layer_index == test_pipe->plane_state->layer_index) {
+ struct rect r2_half;
+
+ r2_half = split_pipe->plane_res.scl_data.recout;
+ reverse_adjust_recout_for_visual_confirm(&r2_half, split_pipe);
+ r2.x = min(r2_half.x, r2.x);
+ r2.width = r2.width + r2_half.width;
+ r2_right = r2.x + r2.width;
+ r2_bottom = min(r2_bottom, r2_half.y + r2_half.height);
+ break;
+ }
+
+ if (r1.x >= r2.x && r1.y >= r2.y && r1_right <= r2_right && r1_bottom <= r2_bottom)
+ return true;
+ }
+
+ return false;
+}
+
enum dc_status resource_build_scaling_params_for_context(
const struct dc *dc,
@@ -2094,7 +2142,8 @@ int resource_get_odm_slice_dst_width(struct pipe_ctx *otg_master,
count = resource_get_odm_slice_count(otg_master);
h_active = timing->h_addressable +
timing->h_border_left +
- timing->h_border_right;
+ timing->h_border_right +
+ otg_master->hblank_borrow;
width = h_active / count;
if (otg_master->stream_res.tg)
@@ -2615,6 +2664,162 @@ static void remove_hpo_dp_link_enc_from_ctx(struct resource_context *res_ctx,
}
}
+static inline int find_acquired_dio_link_enc_for_link(
+ const struct resource_context *res_ctx,
+ const struct dc_link *link)
+{
+ int i;
+
+ for (i = 0; i < ARRAY_SIZE(res_ctx->dio_link_enc_ref_cnts); i++)
+ if (res_ctx->dio_link_enc_ref_cnts[i] > 0 &&
+ res_ctx->dio_link_enc_to_link_idx[i] == link->link_index)
+ return i;
+
+ return -1;
+}
+
+static inline int find_fixed_dio_link_enc(const struct dc_link *link)
+{
+ /* the 8b10b dp phy can only use fixed link encoder */
+ return link->eng_id;
+}
+
+static inline int find_free_dio_link_enc(const struct resource_context *res_ctx,
+ const struct dc_link *link, const struct resource_pool *pool)
+{
+ int i;
+ int enc_count = pool->dig_link_enc_count;
+
+ /* for dpia, check preferred encoder first and then the next one */
+ for (i = 0; i < enc_count; i++)
+ if (res_ctx->dio_link_enc_ref_cnts[(link->dpia_preferred_eng_id + i) % enc_count] == 0)
+ break;
+
+ return (i >= 0 && i < enc_count) ? (link->dpia_preferred_eng_id + i) % enc_count : -1;
+}
+
+static inline void acquire_dio_link_enc(
+ struct resource_context *res_ctx,
+ unsigned int link_index,
+ int enc_index)
+{
+ res_ctx->dio_link_enc_to_link_idx[enc_index] = link_index;
+ res_ctx->dio_link_enc_ref_cnts[enc_index] = 1;
+}
+
+static inline void retain_dio_link_enc(
+ struct resource_context *res_ctx,
+ int enc_index)
+{
+ res_ctx->dio_link_enc_ref_cnts[enc_index]++;
+}
+
+static inline void release_dio_link_enc(
+ struct resource_context *res_ctx,
+ int enc_index)
+{
+ ASSERT(res_ctx->dio_link_enc_ref_cnts[enc_index] > 0);
+ res_ctx->dio_link_enc_ref_cnts[enc_index]--;
+}
+
+static bool is_dio_enc_acquired_by_other_link(const struct dc_link *link,
+ int enc_index,
+ int *link_index)
+{
+ const struct dc *dc = link->dc;
+ const struct resource_context *res_ctx = &dc->current_state->res_ctx;
+
+ /* pass the link_index that acquired the enc_index */
+ if (res_ctx->dio_link_enc_ref_cnts[enc_index] > 0 &&
+ res_ctx->dio_link_enc_to_link_idx[enc_index] != link->link_index) {
+ *link_index = res_ctx->dio_link_enc_to_link_idx[enc_index];
+ return true;
+ }
+
+ return false;
+}
+
+static void swap_dio_link_enc_to_muxable_ctx(struct dc_state *context,
+ const struct resource_pool *pool,
+ int new_encoder,
+ int old_encoder)
+{
+ struct resource_context *res_ctx = &context->res_ctx;
+ int stream_count = context->stream_count;
+ int i = 0;
+
+ res_ctx->dio_link_enc_ref_cnts[new_encoder] = res_ctx->dio_link_enc_ref_cnts[old_encoder];
+ res_ctx->dio_link_enc_to_link_idx[new_encoder] = res_ctx->dio_link_enc_to_link_idx[old_encoder];
+ res_ctx->dio_link_enc_ref_cnts[old_encoder] = 0;
+
+ for (i = 0; i < stream_count; i++) {
+ struct dc_stream_state *stream = context->streams[i];
+ struct pipe_ctx *pipe_ctx = resource_get_otg_master_for_stream(&context->res_ctx, stream);
+
+ if (pipe_ctx && pipe_ctx->link_res.dio_link_enc == pool->link_encoders[old_encoder])
+ pipe_ctx->link_res.dio_link_enc = pool->link_encoders[new_encoder];
+ }
+}
+
+static bool add_dio_link_enc_to_ctx(const struct dc *dc,
+ struct dc_state *context,
+ const struct resource_pool *pool,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_stream_state *stream)
+{
+ struct resource_context *res_ctx = &context->res_ctx;
+ int enc_index;
+
+ enc_index = find_acquired_dio_link_enc_for_link(res_ctx, stream->link);
+
+ if (enc_index >= 0) {
+ retain_dio_link_enc(res_ctx, enc_index);
+ } else {
+ if (stream->link->is_dig_mapping_flexible)
+ enc_index = find_free_dio_link_enc(res_ctx, stream->link, pool);
+ else {
+ int link_index = 0;
+
+ enc_index = find_fixed_dio_link_enc(stream->link);
+ /* Fixed mapping link can only use its fixed link encoder.
+ * If the encoder is acquired by other link then get a new free encoder and swap the new
+ * one into the acquiring link.
+ */
+ if (enc_index >= 0 && is_dio_enc_acquired_by_other_link(stream->link, enc_index, &link_index)) {
+ int new_enc_index = find_free_dio_link_enc(res_ctx, dc->links[link_index], pool);
+
+ if (new_enc_index >= 0)
+ swap_dio_link_enc_to_muxable_ctx(context, pool, new_enc_index, enc_index);
+ else
+ return false;
+ }
+ }
+
+ if (enc_index >= 0)
+ acquire_dio_link_enc(res_ctx, stream->link->link_index, enc_index);
+ }
+
+ if (enc_index >= 0)
+ pipe_ctx->link_res.dio_link_enc = pool->link_encoders[enc_index];
+
+ return pipe_ctx->link_res.dio_link_enc != NULL;
+}
+
+static void remove_dio_link_enc_from_ctx(struct resource_context *res_ctx,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_stream_state *stream)
+{
+ int enc_index = -1;
+
+ if (stream->link)
+ enc_index = find_acquired_dio_link_enc_for_link(res_ctx, stream->link);
+
+ if (enc_index >= 0) {
+ release_dio_link_enc(res_ctx, enc_index);
+ pipe_ctx->link_res.dio_link_enc = NULL;
+ }
+}
+
static int get_num_of_free_pipes(const struct resource_pool *pool, const struct dc_state *context)
{
int i;
@@ -2662,6 +2867,10 @@ void resource_remove_otg_master_for_stream_output(struct dc_state *context,
remove_hpo_dp_link_enc_from_ctx(
&context->res_ctx, otg_master, stream);
}
+
+ if (stream->ctx->dc->config.unify_link_enc_assignment)
+ remove_dio_link_enc_from_ctx(&context->res_ctx, otg_master, stream);
+
if (otg_master->stream_res.audio)
update_audio_usage(
&context->res_ctx,
@@ -2676,6 +2885,7 @@ void resource_remove_otg_master_for_stream_output(struct dc_state *context,
if (pool->funcs->remove_stream_from_ctx)
pool->funcs->remove_stream_from_ctx(
stream->ctx->dc, context, stream);
+
memset(otg_master, 0, sizeof(*otg_master));
}
@@ -3387,10 +3597,13 @@ static int get_norm_pix_clk(const struct dc_crtc_timing *timing)
break;
case COLOR_DEPTH_121212:
normalized_pix_clk = (pix_clk * 36) / 24;
- break;
+ break;
+ case COLOR_DEPTH_141414:
+ normalized_pix_clk = (pix_clk * 42) / 24;
+ break;
case COLOR_DEPTH_161616:
normalized_pix_clk = (pix_clk * 48) / 24;
- break;
+ break;
default:
ASSERT(0);
break;
@@ -3524,16 +3737,22 @@ static int acquire_resource_from_hw_enabled_state(
return -1;
}
-static void mark_seamless_boot_stream(
- const struct dc *dc,
- struct dc_stream_state *stream)
+static void mark_seamless_boot_stream(const struct dc *dc,
+ struct dc_stream_state *stream)
{
struct dc_bios *dcb = dc->ctx->dc_bios;
- if (dc->config.allow_seamless_boot_optimization &&
- !dcb->funcs->is_accelerated_mode(dcb)) {
- if (dc_validate_boot_timing(dc, stream->sink, &stream->timing))
- stream->apply_seamless_boot_optimization = true;
+ DC_LOGGER_INIT(dc->ctx->logger);
+
+ if (stream->apply_seamless_boot_optimization)
+ return;
+ if (!dc->config.allow_seamless_boot_optimization)
+ return;
+ if (dcb->funcs->is_accelerated_mode(dcb))
+ return;
+ if (dc_validate_boot_timing(dc, stream->sink, &stream->timing)) {
+ stream->apply_seamless_boot_optimization = true;
+ DC_LOG_DC("Marked stream for seamless boot optimization\n");
}
}
@@ -3644,6 +3863,7 @@ enum dc_status resource_map_pool_resources(
struct pipe_ctx *pipe_ctx = NULL;
int pipe_idx = -1;
bool acquired = false;
+ bool is_dio_encoder = true;
calculate_phy_pix_clks(stream);
@@ -3691,6 +3911,10 @@ enum dc_status resource_map_pool_resources(
if (!dc->link_srv->dp_decide_link_settings(stream,
&pipe_ctx->link_config.dp_link_settings))
return DC_FAIL_DP_LINK_BANDWIDTH;
+
+ dc->link_srv->dp_decide_tunnel_settings(stream,
+ &pipe_ctx->link_config.dp_tunnel_settings);
+
if (dc->link_srv->dp_get_encoding_format(
&pipe_ctx->link_config.dp_link_settings) == DP_128b_132b_ENCODING) {
pipe_ctx->stream_res.hpo_dp_stream_enc =
@@ -3709,10 +3933,16 @@ enum dc_status resource_map_pool_resources(
}
}
+ if (dc->config.unify_link_enc_assignment && is_dio_encoder)
+ if (!add_dio_link_enc_to_ctx(dc, context, pool, pipe_ctx, stream))
+ return DC_NO_LINK_ENC_RESOURCE;
+
/* TODO: Add check if ASIC support and EDID audio */
if (!stream->converter_disable_audio &&
dc_is_audio_capable_signal(pipe_ctx->stream->signal) &&
- stream->audio_info.mode_count && stream->audio_info.flags.all) {
+ stream->audio_info.mode_count &&
+ (stream->audio_info.flags.all ||
+ (stream->sink && stream->sink->edid_caps.panel_patch.skip_audio_sab_check))) {
pipe_ctx->stream_res.audio = find_first_free_audio(
&context->res_ctx, pool, pipe_ctx->stream_res.stream_enc->id, dc_ctx->dce_version);
@@ -3825,7 +4055,7 @@ static bool add_all_planes_for_stream(
* @set: An array of dc_validation_set with all the current streams reference
* @set_count: Total of streams
* @context: New context
- * @fast_validate: Enable or disable fast validation
+ * @validate_mode: identify the validation mode
*
* This function updates the potential new stream in the context object. It
* creates multiple lists for the add, remove, and unchanged streams. In
@@ -3840,7 +4070,7 @@ enum dc_status dc_validate_with_context(struct dc *dc,
const struct dc_validation_set set[],
int set_count,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
struct dc_stream_state *unchanged_streams[MAX_PIPES] = { 0 };
struct dc_stream_state *del_streams[MAX_PIPES] = { 0 };
@@ -4009,7 +4239,12 @@ enum dc_status dc_validate_with_context(struct dc *dc,
}
}
- res = dc_validate_global_state(dc, context, fast_validate);
+ /* clear subvp cursor limitations */
+ for (i = 0; i < context->stream_count; i++) {
+ dc_state_set_stream_subvp_cursor_limit(context->streams[i], context, false);
+ }
+
+ res = dc_validate_global_state(dc, context, validate_mode);
/* calculate pixel rate divider after deciding pxiel clock & odm combine */
if ((dc->hwss.calculate_pix_rate_divider) && (res == DC_OK)) {
@@ -4027,11 +4262,46 @@ fail:
}
/**
+ * decide_hblank_borrow - Decides the horizontal blanking borrow value for a given pipe context.
+ * @pipe_ctx: Pointer to the pipe context structure.
+ *
+ * This function calculates the horizontal blanking borrow value for a given pipe context based on the
+ * display stream compression (DSC) configuration. If the horizontal active pixels (hactive) are less
+ * than the total width of the DSC slices, it sets the hblank_borrow value to the difference. If the
+ * total horizontal timing minus the hblank_borrow value is less than 32, it resets the hblank_borrow
+ * value to 0.
+ */
+static void decide_hblank_borrow(struct pipe_ctx *pipe_ctx)
+{
+ uint32_t hactive;
+ uint32_t ceil_slice_width;
+ struct dc_stream_state *stream = NULL;
+
+ if (!pipe_ctx)
+ return;
+
+ stream = pipe_ctx->stream;
+
+ if (stream->timing.flags.DSC) {
+ hactive = stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right;
+
+ /* Assume if determined slices does not divide Hactive evenly, Hborrow is needed for padding*/
+ if (hactive % stream->timing.dsc_cfg.num_slices_h != 0) {
+ ceil_slice_width = (hactive / stream->timing.dsc_cfg.num_slices_h) + 1;
+ pipe_ctx->hblank_borrow = ceil_slice_width * stream->timing.dsc_cfg.num_slices_h - hactive;
+
+ if (stream->timing.h_total - hactive - pipe_ctx->hblank_borrow < 32)
+ pipe_ctx->hblank_borrow = 0;
+ }
+ }
+}
+
+/**
* dc_validate_global_state() - Determine if hardware can support a given state
*
* @dc: dc struct for this driver
* @new_ctx: state to be validated
- * @fast_validate: set to true if only yes/no to support matters
+ * @validate_mode: identify the validation mode
*
* Checks hardware resource availability and bandwidth requirement.
*
@@ -4041,7 +4311,7 @@ fail:
enum dc_status dc_validate_global_state(
struct dc *dc,
struct dc_state *new_ctx,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
enum dc_status result = DC_ERROR_UNEXPECTED;
int i, j;
@@ -4064,6 +4334,10 @@ enum dc_status dc_validate_global_state(
if (pipe_ctx->stream != stream)
continue;
+ /* Decide whether hblank borrow is needed and save it in pipe_ctx */
+ if (dc->debug.enable_hblank_borrow)
+ decide_hblank_borrow(pipe_ctx);
+
if (dc->res_pool->funcs->patch_unknown_plane_state &&
pipe_ctx->plane_state &&
pipe_ctx->plane_state->tiling_info.gfx9.swizzle == DC_SW_UNKNOWN) {
@@ -4096,8 +4370,7 @@ enum dc_status dc_validate_global_state(
result = resource_build_scaling_params_for_context(dc, new_ctx);
if (result == DC_OK)
- if (!dc->res_pool->funcs->validate_bandwidth(dc, new_ctx, fast_validate))
- result = DC_FAIL_BANDWIDTH_VALIDATE;
+ result = dc->res_pool->funcs->validate_bandwidth(dc, new_ctx, validate_mode);
return result;
}
@@ -4203,7 +4476,7 @@ static void set_avi_info_frame(
break;
case COLOR_SPACE_2020_RGB_FULLRANGE:
case COLOR_SPACE_2020_RGB_LIMITEDRANGE:
- case COLOR_SPACE_2020_YCBCR:
+ case COLOR_SPACE_2020_YCBCR_LIMITED:
hdmi_info.bits.EC0_EC2 = COLORIMETRYEX_BT2020RGBYCBCR;
hdmi_info.bits.C0_C1 = COLORIMETRY_EXTENDED;
break;
@@ -4217,7 +4490,7 @@ static void set_avi_info_frame(
break;
}
- if (pixel_encoding && color_space == COLOR_SPACE_2020_YCBCR &&
+ if (pixel_encoding && color_space == COLOR_SPACE_2020_YCBCR_LIMITED &&
stream->out_transfer_func.tf == TRANSFER_FUNCTION_GAMMA22) {
hdmi_info.bits.EC0_EC2 = 0;
hdmi_info.bits.C0_C1 = COLORIMETRY_ITU709;
@@ -4438,7 +4711,7 @@ static void set_hfvs_info_packet(
static void adaptive_sync_override_dp_info_packets_sdp_line_num(
const struct dc_crtc_timing *timing,
struct enc_sdp_line_num *sdp_line_num,
- struct _vcs_dpi_display_pipe_dest_params_st *pipe_dlg_param)
+ unsigned int vstartup_start)
{
uint32_t asic_blank_start = 0;
uint32_t asic_blank_end = 0;
@@ -4453,8 +4726,8 @@ static void adaptive_sync_override_dp_info_packets_sdp_line_num(
asic_blank_end = (asic_blank_start - tg->v_border_bottom -
tg->v_addressable - tg->v_border_top);
- if (pipe_dlg_param->vstartup_start > asic_blank_end) {
- v_update = (tg->v_total - (pipe_dlg_param->vstartup_start - asic_blank_end));
+ if (vstartup_start > asic_blank_end) {
+ v_update = (tg->v_total - (vstartup_start - asic_blank_end));
sdp_line_num->adaptive_sync_line_num_valid = true;
sdp_line_num->adaptive_sync_line_num = (tg->v_total - v_update - 1);
} else {
@@ -4467,7 +4740,7 @@ static void set_adaptive_sync_info_packet(
struct dc_info_packet *info_packet,
const struct dc_stream_state *stream,
struct encoder_info_frame *info_frame,
- struct _vcs_dpi_display_pipe_dest_params_st *pipe_dlg_param)
+ unsigned int vstartup_start)
{
if (!stream->adaptive_sync_infopacket.valid)
return;
@@ -4475,7 +4748,7 @@ static void set_adaptive_sync_info_packet(
adaptive_sync_override_dp_info_packets_sdp_line_num(
&stream->timing,
&info_frame->sdp_line_num,
- pipe_dlg_param);
+ vstartup_start);
*info_packet = stream->adaptive_sync_infopacket;
}
@@ -4508,6 +4781,7 @@ void resource_build_info_frame(struct pipe_ctx *pipe_ctx)
{
enum signal_type signal = SIGNAL_TYPE_NONE;
struct encoder_info_frame *info = &pipe_ctx->stream_res.encoder_info_frame;
+ unsigned int vstartup_start = 0;
/* default all packets to invalid */
info->avi.valid = false;
@@ -4521,6 +4795,9 @@ void resource_build_info_frame(struct pipe_ctx *pipe_ctx)
info->adaptive_sync.valid = false;
signal = pipe_ctx->stream->signal;
+ if (pipe_ctx->stream->ctx->dc->res_pool->funcs->get_vstartup_for_pipe)
+ vstartup_start = pipe_ctx->stream->ctx->dc->res_pool->funcs->get_vstartup_for_pipe(pipe_ctx);
+
/* HDMi and DP have different info packets*/
if (dc_is_hdmi_signal(signal)) {
set_avi_info_frame(&info->avi, pipe_ctx);
@@ -4542,7 +4819,7 @@ void resource_build_info_frame(struct pipe_ctx *pipe_ctx)
set_adaptive_sync_info_packet(&info->adaptive_sync,
pipe_ctx->stream,
info,
- &pipe_ctx->pipe_dlg_param);
+ vstartup_start);
}
patch_gamut_packet_checksum(&info->gamut);
@@ -4636,7 +4913,10 @@ bool pipe_need_reprogram(
return true;
/* DIG link encoder resource assignment for stream changed. */
- if (pipe_ctx_old->stream->ctx->dc->res_pool->funcs->link_encs_assign) {
+ if (pipe_ctx_old->stream->ctx->dc->config.unify_link_enc_assignment) {
+ if (pipe_ctx_old->link_res.dio_link_enc != pipe_ctx->link_res.dio_link_enc)
+ return true;
+ } else if (pipe_ctx_old->stream->ctx->dc->res_pool->funcs->link_encs_assign) {
bool need_reprogram = false;
struct dc *dc = pipe_ctx_old->stream->ctx->dc;
struct link_encoder *link_enc_prev =
@@ -4902,6 +5182,28 @@ void get_audio_check(struct audio_info *aud_modes,
}
}
+struct link_encoder *get_temp_dio_link_enc(
+ const struct resource_context *res_ctx,
+ const struct resource_pool *const pool,
+ const struct dc_link *link)
+{
+ struct link_encoder *link_enc = NULL;
+ int enc_index;
+
+ if (link->is_dig_mapping_flexible)
+ enc_index = find_acquired_dio_link_enc_for_link(res_ctx, link);
+ else
+ enc_index = link->eng_id;
+
+ if (enc_index < 0)
+ enc_index = find_free_dio_link_enc(res_ctx, link, pool);
+
+ if (enc_index >= 0)
+ link_enc = pool->link_encoders[enc_index];
+
+ return link_enc;
+}
+
static struct hpo_dp_link_encoder *get_temp_hpo_dp_link_enc(
const struct resource_context *res_ctx,
const struct resource_pool *const pool,
@@ -4931,11 +5233,17 @@ bool get_temp_dp_link_res(struct dc_link *link,
memset(link_res, 0, sizeof(*link_res));
if (dc->link_srv->dp_get_encoding_format(link_settings) == DP_128b_132b_ENCODING) {
- link_res->hpo_dp_link_enc = get_temp_hpo_dp_link_enc(res_ctx,
- dc->res_pool, link);
+ link_res->hpo_dp_link_enc = get_temp_hpo_dp_link_enc(res_ctx, dc->res_pool, link);
if (!link_res->hpo_dp_link_enc)
return false;
+ } else if (dc->link_srv->dp_get_encoding_format(link_settings) == DP_8b_10b_ENCODING &&
+ dc->config.unify_link_enc_assignment) {
+ link_res->dio_link_enc = get_temp_dio_link_enc(res_ctx,
+ dc->res_pool, link);
+ if (!link_res->dio_link_enc)
+ return false;
}
+
return true;
}
@@ -5207,26 +5515,24 @@ enum dc_status update_dp_encoder_resources_for_test_harness(const struct dc *dc,
remove_hpo_dp_link_enc_from_ctx(&context->res_ctx, pipe_ctx, pipe_ctx->stream);
}
+ if (pipe_ctx->link_res.dio_link_enc == NULL && dc->config.unify_link_enc_assignment)
+ if (!add_dio_link_enc_to_ctx(dc, context, dc->res_pool, pipe_ctx, pipe_ctx->stream))
+ return DC_NO_LINK_ENC_RESOURCE;
+
return DC_OK;
}
-bool check_subvp_sw_cursor_fallback_req(const struct dc *dc, struct dc_stream_state *stream)
+struct dscl_prog_data *resource_get_dscl_prog_data(struct pipe_ctx *pipe_ctx)
{
- if (!dc->debug.disable_subvp_high_refresh && is_subvp_high_refresh_candidate(stream))
- return true;
- if (dc->current_state->stream_count == 1 && stream->timing.v_addressable >= 2880 &&
- ((stream->timing.pix_clk_100hz * 100) / stream->timing.v_total / stream->timing.h_total) < 120)
- return true;
- else if (dc->current_state->stream_count > 1 && stream->timing.v_addressable >= 1080 &&
- ((stream->timing.pix_clk_100hz * 100) / stream->timing.v_total / stream->timing.h_total) < 120)
- return true;
-
- return false;
+ return &pipe_ctx->plane_res.scl_data.dscl_prog_data;
}
-struct dscl_prog_data *resource_get_dscl_prog_data(struct pipe_ctx *pipe_ctx)
+static bool resource_allocate_mcache(struct dc_state *context, const struct dc_mcache_params *mcache_params)
{
- return &pipe_ctx->plane_res.scl_data.dscl_prog_data;
+ if (context->clk_mgr->ctx->dc->res_pool->funcs->program_mcache_pipe_config)
+ context->clk_mgr->ctx->dc->res_pool->funcs->program_mcache_pipe_config(context, mcache_params);
+
+ return true;
}
void resource_init_common_dml2_callbacks(struct dc *dc, struct dml2_configuration_options *dml2_options)
@@ -5248,6 +5554,7 @@ void resource_init_common_dml2_callbacks(struct dc *dc, struct dml2_configuratio
dml2_options->callbacks.get_stream_status = &dc_state_get_stream_status;
dml2_options->callbacks.get_stream_from_id = &dc_state_get_stream_from_id;
dml2_options->callbacks.get_max_flickerless_instant_vtotal_increase = &dc_stream_get_max_flickerless_instant_vtotal_increase;
+ dml2_options->callbacks.allocate_mcache = &resource_allocate_mcache;
dml2_options->svp_pstate.callbacks.dc = dc;
dml2_options->svp_pstate.callbacks.add_phantom_plane = &dc_state_add_phantom_plane;
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_state.c b/drivers/gpu/drm/amd/display/dc/core/dc_state.c
index e006f816ff2f..883054bb18e7 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_state.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_state.c
@@ -22,6 +22,7 @@
* Authors: AMD
*
*/
+#include "dc_types.h"
#include "core_types.h"
#include "core_status.h"
#include "dc_state.h"
@@ -193,11 +194,6 @@ static void init_state(struct dc *dc, struct dc_state *state)
struct dc_state *dc_state_create(struct dc *dc, struct dc_state_create_params *params)
{
struct dc_state *state;
-#ifdef CONFIG_DRM_AMD_DC_FP
- struct dml2_configuration_options *dml2_opt = &dc->dml2_tmp;
-
- memcpy(dml2_opt, &dc->dml2_options, sizeof(dc->dml2_options));
-#endif
state = kvzalloc(sizeof(struct dc_state), GFP_KERNEL);
@@ -210,14 +206,12 @@ struct dc_state *dc_state_create(struct dc *dc, struct dc_state_create_params *p
#ifdef CONFIG_DRM_AMD_DC_FP
if (dc->debug.using_dml2) {
- dml2_opt->use_clock_dc_limits = false;
- if (!dml2_create(dc, dml2_opt, &state->bw_ctx.dml2)) {
+ if (!dml2_create(dc, &dc->dml2_options, &state->bw_ctx.dml2)) {
dc_state_release(state);
return NULL;
}
- dml2_opt->use_clock_dc_limits = true;
- if (!dml2_create(dc, dml2_opt, &state->bw_ctx.dml2_dc_power_source)) {
+ if (!dml2_create(dc, &dc->dml2_dc_power_options, &state->bw_ctx.dml2_dc_power_source)) {
dc_state_release(state);
return NULL;
}
@@ -433,6 +427,8 @@ enum dc_status dc_state_remove_stream(
return DC_ERROR_UNEXPECTED;
}
+ dc_stream_release_3dlut_for_stream(dc, stream);
+
dc_stream_release(state->streams[i]);
state->stream_count--;
@@ -483,9 +479,9 @@ bool dc_state_add_plane(
if (stream_status == NULL) {
dm_error("Existing stream not found; failed to attach surface!\n");
goto out;
- } else if (stream_status->plane_count == MAX_SURFACE_NUM) {
+ } else if (stream_status->plane_count == MAX_SURFACES) {
dm_error("Surface: can not attach plane_state %p! Maximum is: %d\n",
- plane_state, MAX_SURFACE_NUM);
+ plane_state, MAX_SURFACES);
goto out;
} else if (!otg_master_pipe) {
goto out;
@@ -600,7 +596,7 @@ bool dc_state_rem_all_planes_for_stream(
{
int i, old_plane_count;
struct dc_stream_status *stream_status = NULL;
- struct dc_plane_state *del_planes[MAX_SURFACE_NUM] = { 0 };
+ struct dc_plane_state *del_planes[MAX_SURFACES] = { 0 };
for (i = 0; i < state->stream_count; i++)
if (state->streams[i] == stream) {
@@ -812,8 +808,12 @@ enum dc_status dc_state_add_phantom_stream(const struct dc *dc,
if (phantom_stream_status) {
phantom_stream_status->mall_stream_config.type = SUBVP_PHANTOM;
phantom_stream_status->mall_stream_config.paired_stream = main_stream;
+ phantom_stream_status->mall_stream_config.subvp_limit_cursor_size = false;
+ phantom_stream_status->mall_stream_config.cursor_size_limit_subvp = false;
}
+ dc_state_set_stream_subvp_cursor_limit(main_stream, state, true);
+
return res;
}
@@ -875,7 +875,7 @@ bool dc_state_rem_all_phantom_planes_for_stream(
{
int i, old_plane_count;
struct dc_stream_status *stream_status = NULL;
- struct dc_plane_state *del_planes[MAX_SURFACE_NUM] = { 0 };
+ struct dc_plane_state *del_planes[MAX_SURFACES] = { 0 };
for (i = 0; i < state->stream_count; i++)
if (state->streams[i] == phantom_stream) {
@@ -939,13 +939,20 @@ void dc_state_release_phantom_streams_and_planes(
const struct dc *dc,
struct dc_state *state)
{
+ unsigned int phantom_count;
+ struct dc_stream_state *phantom_streams[MAX_PHANTOM_PIPES];
+ struct dc_plane_state *phantom_planes[MAX_PHANTOM_PIPES];
int i;
- for (i = 0; i < state->phantom_stream_count; i++)
- dc_state_release_phantom_stream(dc, state, state->phantom_streams[i]);
+ phantom_count = state->phantom_stream_count;
+ memcpy(phantom_streams, state->phantom_streams, sizeof(struct dc_stream_state *) * MAX_PHANTOM_PIPES);
+ for (i = 0; i < phantom_count; i++)
+ dc_state_release_phantom_stream(dc, state, phantom_streams[i]);
- for (i = 0; i < state->phantom_plane_count; i++)
- dc_state_release_phantom_plane(dc, state, state->phantom_planes[i]);
+ phantom_count = state->phantom_plane_count;
+ memcpy(phantom_planes, state->phantom_planes, sizeof(struct dc_plane_state *) * MAX_PHANTOM_PIPES);
+ for (i = 0; i < phantom_count; i++)
+ dc_state_release_phantom_plane(dc, state, phantom_planes[i]);
}
struct dc_stream_state *dc_state_get_stream_from_id(const struct dc_state *state, unsigned int id)
@@ -977,3 +984,94 @@ bool dc_state_is_fams2_in_use(
return is_fams2_in_use;
}
+
+void dc_state_set_stream_subvp_cursor_limit(const struct dc_stream_state *stream,
+ struct dc_state *state,
+ bool limit)
+{
+ struct dc_stream_status *stream_status;
+
+ stream_status = dc_state_get_stream_status(state, stream);
+
+ if (stream_status) {
+ stream_status->mall_stream_config.subvp_limit_cursor_size = limit;
+ }
+}
+
+bool dc_state_get_stream_subvp_cursor_limit(const struct dc_stream_state *stream,
+ struct dc_state *state)
+{
+ bool limit = false;
+
+ struct dc_stream_status *stream_status;
+
+ stream_status = dc_state_get_stream_status(state, stream);
+
+ if (stream_status) {
+ limit = stream_status->mall_stream_config.subvp_limit_cursor_size;
+ }
+
+ return limit;
+}
+
+void dc_state_set_stream_cursor_subvp_limit(const struct dc_stream_state *stream,
+ struct dc_state *state,
+ bool limit)
+{
+ struct dc_stream_status *stream_status;
+
+ stream_status = dc_state_get_stream_status(state, stream);
+
+ if (stream_status) {
+ stream_status->mall_stream_config.cursor_size_limit_subvp = limit;
+ }
+}
+
+bool dc_state_get_stream_cursor_subvp_limit(const struct dc_stream_state *stream,
+ struct dc_state *state)
+{
+ bool limit = false;
+
+ struct dc_stream_status *stream_status;
+
+ stream_status = dc_state_get_stream_status(state, stream);
+
+ if (stream_status) {
+ limit = stream_status->mall_stream_config.cursor_size_limit_subvp;
+ }
+
+ return limit;
+}
+
+bool dc_state_can_clear_stream_cursor_subvp_limit(const struct dc_stream_state *stream,
+ struct dc_state *state)
+{
+ bool can_clear_limit = false;
+
+ struct dc_stream_status *stream_status;
+
+ stream_status = dc_state_get_stream_status(state, stream);
+
+ if (stream_status) {
+ can_clear_limit = dc_state_get_stream_cursor_subvp_limit(stream, state) &&
+ (stream_status->mall_stream_config.type == SUBVP_PHANTOM ||
+ stream->hw_cursor_req ||
+ !stream_status->mall_stream_config.subvp_limit_cursor_size ||
+ !stream->cursor_position.enable ||
+ dc_stream_check_cursor_attributes(stream, state, &stream->cursor_attributes));
+ }
+
+ return can_clear_limit;
+}
+
+bool dc_state_is_subvp_in_use(struct dc_state *state)
+{
+ uint32_t i;
+
+ for (i = 0; i < state->stream_count; i++) {
+ if (dc_state_get_stream_subvp_type(state, state->streams[i]) != SUBVP_NONE)
+ return true;
+ }
+
+ return false;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
index 55dc482d9b36..4d6bc9fd4faa 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
@@ -37,6 +37,8 @@
#define DC_LOGGER dc->ctx->logger
#ifndef MIN
#define MIN(X, Y) ((X) < (Y) ? (X) : (Y))
+#endif
+#ifndef MAX
#define MAX(x, y) ((x > y) ? x : y)
#endif
@@ -199,7 +201,8 @@ struct dc_stream_state *dc_copy_stream(const struct dc_stream_state *stream)
dc_stream_assign_stream_id(new_stream);
/* If using dynamic encoder assignment, wait till stream committed to assign encoder. */
- if (new_stream->ctx->dc->res_pool->funcs->link_encs_assign)
+ if (new_stream->ctx->dc->res_pool->funcs->link_encs_assign &&
+ !new_stream->ctx->dc->config.unify_link_enc_assignment)
new_stream->link_enc = NULL;
kref_init(&new_stream->refcount);
@@ -262,13 +265,16 @@ void program_cursor_attributes(
}
/*
- * dc_stream_set_cursor_attributes() - Update cursor attributes and set cursor surface address
+ * dc_stream_check_cursor_attributes() - Check validitity of cursor attributes and surface address
*/
-bool dc_stream_set_cursor_attributes(
- struct dc_stream_state *stream,
+bool dc_stream_check_cursor_attributes(
+ const struct dc_stream_state *stream,
+ struct dc_state *state,
const struct dc_cursor_attributes *attributes)
{
- struct dc *dc;
+ const struct dc *dc;
+
+ unsigned int max_cursor_size;
if (NULL == stream) {
dm_error("DC: dc_stream is NULL!\n");
@@ -286,24 +292,38 @@ bool dc_stream_set_cursor_attributes(
dc = stream->ctx->dc;
- /* SubVP is not compatible with HW cursor larger than 64 x 64 x 4.
- * Therefore, if cursor is greater than 64 x 64 x 4, fallback to SW cursor in the following case:
- * 1. If the config is a candidate for SubVP high refresh (both single an dual display configs)
- * 2. If not subvp high refresh, for single display cases, if resolution is >= 5K and refresh rate < 120hz
- * 3. If not subvp high refresh, for multi display cases, if resolution is >= 4K and refresh rate < 120hz
+ /* SubVP is not compatible with HW cursor larger than what can fit in cursor SRAM.
+ * Therefore, if cursor is greater than this, fallback to SW cursor.
*/
- if (dc->debug.allow_sw_cursor_fallback &&
- attributes->height * attributes->width * 4 > 16384 &&
- !stream->hw_cursor_req) {
- if (check_subvp_sw_cursor_fallback_req(dc, stream))
+ if (dc->debug.allow_sw_cursor_fallback && dc->res_pool->funcs->get_max_hw_cursor_size) {
+ max_cursor_size = dc->res_pool->funcs->get_max_hw_cursor_size(dc, state, stream);
+ max_cursor_size = max_cursor_size * max_cursor_size * 4;
+
+ if (attributes->height * attributes->width * 4 > max_cursor_size) {
return false;
+ }
}
- stream->cursor_attributes = *attributes;
-
return true;
}
+/*
+ * dc_stream_set_cursor_attributes() - Update cursor attributes and set cursor surface address
+ */
+bool dc_stream_set_cursor_attributes(
+ struct dc_stream_state *stream,
+ const struct dc_cursor_attributes *attributes)
+{
+ bool result = false;
+
+ if (dc_stream_check_cursor_attributes(stream, stream->ctx->dc->current_state, attributes)) {
+ stream->cursor_attributes = *attributes;
+ result = true;
+ }
+
+ return result;
+}
+
bool dc_stream_program_cursor_attributes(
struct dc_stream_state *stream,
const struct dc_cursor_attributes *attributes)
@@ -549,6 +569,14 @@ bool dc_stream_fc_disable_writeback(struct dc *dc,
return true;
}
+/**
+ * dc_stream_remove_writeback() - Disables writeback and removes writeback info.
+ * @dc: Display core control structure.
+ * @stream: Display core stream state.
+ * @dwb_pipe_inst: Display writeback pipe.
+ *
+ * Return: returns true on success, false otherwise.
+ */
bool dc_stream_remove_writeback(struct dc *dc,
struct dc_stream_state *stream,
uint32_t dwb_pipe_inst)
@@ -605,17 +633,6 @@ bool dc_stream_remove_writeback(struct dc *dc,
return true;
}
-bool dc_stream_warmup_writeback(struct dc *dc,
- int num_dwb,
- struct dc_writeback_info *wb_info)
-{
- dc_exit_ips_for_hw_access(dc);
-
- if (dc->hwss.mmhubbub_warmup)
- return dc->hwss.mmhubbub_warmup(dc, num_dwb, wb_info);
- else
- return false;
-}
uint32_t dc_stream_get_vblank_counter(const struct dc_stream_state *stream)
{
uint8_t i;
@@ -840,6 +857,73 @@ void dc_stream_log(const struct dc *dc, const struct dc_stream_state *stream)
}
/*
+* dc_stream_get_3dlut()
+* Requirements:
+* 1. Is stream already owns an RMCM instance, return it.
+* 2. If it doesn't and we don't need to allocate, return NULL.
+* 3. If there's a free RMCM instance, assign to stream and return it.
+* 4. If no free RMCM instances, return NULL.
+*/
+
+struct dc_rmcm_3dlut *dc_stream_get_3dlut_for_stream(
+ const struct dc *dc,
+ const struct dc_stream_state *stream,
+ bool allocate_one)
+{
+ unsigned int num_rmcm = dc->caps.color.mpc.num_rmcm_3dluts;
+
+ // see if one is allocated for this stream
+ for (int i = 0; i < num_rmcm; i++) {
+ if (dc->res_pool->rmcm_3dlut[i].isInUse &&
+ dc->res_pool->rmcm_3dlut[i].stream == stream)
+ return &dc->res_pool->rmcm_3dlut[i];
+ }
+
+ //case: not found one, and dont need to allocate
+ if (!allocate_one)
+ return NULL;
+
+ //see if there is an unused 3dlut, allocate
+ for (int i = 0; i < num_rmcm; i++) {
+ if (!dc->res_pool->rmcm_3dlut[i].isInUse) {
+ dc->res_pool->rmcm_3dlut[i].isInUse = true;
+ dc->res_pool->rmcm_3dlut[i].stream = stream;
+ return &dc->res_pool->rmcm_3dlut[i];
+ }
+ }
+
+ //dont have a 3dlut
+ return NULL;
+}
+
+
+void dc_stream_release_3dlut_for_stream(
+ const struct dc *dc,
+ const struct dc_stream_state *stream)
+{
+ struct dc_rmcm_3dlut *rmcm_3dlut =
+ dc_stream_get_3dlut_for_stream(dc, stream, false);
+
+ if (rmcm_3dlut) {
+ rmcm_3dlut->isInUse = false;
+ rmcm_3dlut->stream = NULL;
+ rmcm_3dlut->protection_bits = 0;
+ }
+}
+
+
+void dc_stream_init_rmcm_3dlut(struct dc *dc)
+{
+ unsigned int num_rmcm = dc->caps.color.mpc.num_rmcm_3dluts;
+
+ for (int i = 0; i < num_rmcm; i++) {
+ dc->res_pool->rmcm_3dlut[i].isInUse = false;
+ dc->res_pool->rmcm_3dlut[i].stream = NULL;
+ dc->res_pool->rmcm_3dlut[i].protection_bits = 0;
+ }
+}
+
+/*
* Finds the greatest index in refresh_rate_hz that contains a value <= refresh
*/
static int dc_stream_get_nearest_smallest_index(struct dc_stream_state *stream, int refresh)
@@ -1117,3 +1201,26 @@ unsigned int dc_stream_get_max_flickerless_instant_vtotal_increase(struct dc_str
return dc_stream_get_max_flickerless_instant_vtotal_delta(stream, is_gaming, false);
}
+
+bool dc_stream_is_cursor_limit_pending(struct dc *dc, struct dc_stream_state *stream)
+{
+ bool is_limit_pending = false;
+
+ if (dc->current_state)
+ is_limit_pending = dc_state_get_stream_cursor_subvp_limit(stream, dc->current_state);
+
+ return is_limit_pending;
+}
+
+bool dc_stream_can_clear_cursor_limit(struct dc *dc, struct dc_stream_state *stream)
+{
+ bool can_clear_limit = false;
+
+ if (dc->current_state)
+ can_clear_limit = dc_state_get_stream_cursor_subvp_limit(stream, dc->current_state) &&
+ (stream->hw_cursor_req ||
+ !stream->cursor_position.enable ||
+ dc_stream_check_cursor_attributes(stream, dc->current_state, &stream->cursor_attributes));
+
+ return can_clear_limit;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
index ccbb15f1638c..922f23557f5d 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_surface.c
@@ -83,13 +83,6 @@ uint8_t dc_plane_get_pipe_mask(struct dc_state *dc_state, const struct dc_plane
/*******************************************************************************
* Public functions
******************************************************************************/
-void enable_surface_flip_reporting(struct dc_plane_state *plane_state,
- uint32_t controller_id)
-{
- plane_state->irq_source = controller_id + DC_IRQ_SOURCE_PFLIP1 - 1;
- /*register_flip_interrupt(surface);*/
-}
-
struct dc_plane_state *dc_create_plane_state(const struct dc *dc)
{
struct dc_plane_state *plane_state = kvzalloc(sizeof(*plane_state),
@@ -116,7 +109,8 @@ struct dc_plane_state *dc_create_plane_state(const struct dc *dc)
*****************************************************************************
*/
const struct dc_plane_status *dc_plane_get_status(
- const struct dc_plane_state *plane_state)
+ const struct dc_plane_state *plane_state,
+ union dc_plane_status_update_flags flags)
{
const struct dc_plane_status *plane_status;
struct dc *dc;
@@ -143,7 +137,7 @@ const struct dc_plane_status *dc_plane_get_status(
if (pipe_ctx->plane_state != plane_state)
continue;
- if (pipe_ctx->plane_state)
+ if (pipe_ctx->plane_state && flags.bits.address)
pipe_ctx->plane_state->status.is_flip_pending = false;
break;
@@ -158,7 +152,8 @@ const struct dc_plane_status *dc_plane_get_status(
if (pipe_ctx->plane_state != plane_state)
continue;
- dc->hwss.update_pending_status(pipe_ctx);
+ if (flags.bits.address)
+ dc->hwss.update_pending_status(pipe_ctx);
}
return plane_status;
@@ -277,4 +272,41 @@ void dc_3dlut_func_retain(struct dc_3dlut *lut)
kref_get(&lut->refcount);
}
+void dc_plane_force_dcc_and_tiling_disable(struct dc_plane_state *plane_state,
+ bool clear_tiling)
+{
+ struct dc *dc;
+ int i;
+
+ if (!plane_state)
+ return;
+
+ dc = plane_state->ctx->dc;
+
+ if (!dc || !dc->current_state)
+ return;
+
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ struct pipe_ctx *pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
+ if (!pipe_ctx)
+ continue;
+
+ if (dc->hwss.clear_surface_dcc_and_tiling)
+ dc->hwss.clear_surface_dcc_and_tiling(pipe_ctx, plane_state, clear_tiling);
+ }
+}
+
+void dc_plane_copy_config(struct dc_plane_state *dst, const struct dc_plane_state *src)
+{
+ struct kref temp_refcount;
+
+ /* backup persistent info */
+ memcpy(&temp_refcount, &dst->refcount, sizeof(struct kref));
+
+ /* copy all configuration information */
+ memcpy(dst, src, sizeof(struct dc_plane_state));
+
+ /* restore persistent info */
+ memcpy(&dst->refcount, &temp_refcount, sizeof(struct kref));
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index 104051935884..59c07756130d 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -46,7 +46,7 @@
#include "dmub/inc/dmub_cmd.h"
-#include "spl/dc_spl_types.h"
+#include "sspl/dc_spl_types.h"
struct abm_save_restore;
@@ -55,14 +55,24 @@ struct aux_payload;
struct set_config_cmd_payload;
struct dmub_notification;
-#define DC_VER "3.2.310"
+#define DC_VER "3.2.340"
-#define MAX_SURFACES 3
+/**
+ * MAX_SURFACES - representative of the upper bound of surfaces that can be piped to a single CRTC
+ */
+#define MAX_SURFACES 4
+/**
+ * MAX_PLANES - representative of the upper bound of planes that are supported by the HW
+ */
#define MAX_PLANES 6
#define MAX_STREAMS 6
#define MIN_VIEWPORT_SIZE 12
#define MAX_NUM_EDP 2
-#define MAX_HOST_ROUTERS_NUM 2
+#define MAX_SUPPORTED_FORMATS 7
+
+#define MAX_HOST_ROUTERS_NUM 3
+#define MAX_DPIA_PER_HOST_ROUTER 3
+#define MAX_DPIA_NUM (MAX_HOST_ROUTERS_NUM * MAX_DPIA_PER_HOST_ROUTER)
/* Display Core Interfaces */
struct dc_versions {
@@ -188,6 +198,34 @@ struct dpp_color_caps {
struct rom_curve_caps ogam_rom_caps;
};
+/* Below structure is to describe the HW support for mem layout, extend support
+ range to match what OS could handle in the roadmap */
+struct lut3d_caps {
+ uint32_t dma_3d_lut : 1; /*< DMA mode support for 3D LUT */
+ struct {
+ uint32_t swizzle_3d_rgb : 1;
+ uint32_t swizzle_3d_bgr : 1;
+ uint32_t linear_1d : 1;
+ } mem_layout_support;
+ struct {
+ uint32_t unorm_12msb : 1;
+ uint32_t unorm_12lsb : 1;
+ uint32_t float_fp1_5_10 : 1;
+ } mem_format_support;
+ struct {
+ uint32_t order_rgba : 1;
+ uint32_t order_bgra : 1;
+ } mem_pixel_order_support;
+ /*< size options are 9, 17, 33, 45, 65 */
+ struct {
+ uint32_t dim_9 : 1; /* 3D LUT support for 9x9x9 */
+ uint32_t dim_17 : 1; /* 3D LUT support for 17x17x17 */
+ uint32_t dim_33 : 1; /* 3D LUT support for 33x33x33 */
+ uint32_t dim_45 : 1; /* 3D LUT support for 45x45x45 */
+ uint32_t dim_65 : 1; /* 3D LUT support for 65x65x65 */
+ } lut_dim_caps;
+};
+
/**
* struct mpc_color_caps - color pipeline capabilities for multiple pipe and
* plane combined blocks
@@ -199,14 +237,21 @@ struct dpp_color_caps {
* @shared_3d_lut: shared 3D LUT flag. Can be either DPP or MPC, but single
* instance
* @ogam_rom_caps: pre-definied curve caps for regamma 1D LUT
+ * @mcm_3d_lut_caps: HW support cap for MCM LUT memory
+ * @rmcm_3d_lut_caps: HW support cap for RMCM LUT memory
+ * @preblend: whether color manager supports preblend with MPC
*/
struct mpc_color_caps {
uint16_t gamut_remap : 1;
uint16_t ogam_ram : 1;
uint16_t ocsc : 1;
uint16_t num_3dluts : 3;
+ uint16_t num_rmcm_3dluts : 3;
uint16_t shared_3d_lut:1;
struct rom_curve_caps ogam_rom_caps;
+ struct lut3d_caps mcm_3d_lut_caps;
+ struct lut3d_caps rmcm_3d_lut_caps;
+ bool preblend;
};
/**
@@ -245,6 +290,7 @@ struct dc_caps {
uint32_t i2c_speed_in_khz_hdcp;
uint32_t dmdata_alloc_size;
unsigned int max_cursor_size;
+ unsigned int max_buffered_cursor_size;
unsigned int max_video_width;
/*
* max video plane width that can be safely assumed to be always
@@ -265,6 +311,7 @@ struct dc_caps {
bool dmcub_support;
bool zstate_support;
bool ips_support;
+ bool ips_v2_support;
uint32_t num_of_internal_disp;
enum dp_protocol_version max_dp_protocol_version;
unsigned int mall_size_per_mem_channel;
@@ -278,6 +325,7 @@ struct dc_caps {
bool edp_dsc_support;
bool vbios_lttpr_aware;
bool vbios_lttpr_enable;
+ bool fused_io_supported;
uint32_t max_otg_num;
uint32_t max_cab_allocation_bytes;
uint32_t cache_line_size;
@@ -290,6 +338,7 @@ struct dc_caps {
uint16_t subvp_vertical_int_margin_us;
bool seamless_odm;
uint32_t max_v_total;
+ bool vtotal_limited_by_fp2;
uint32_t max_disp_clock_khz_at_vmin;
uint8_t subvp_drr_vblank_start_margin_us;
bool cursor_not_scaled;
@@ -298,6 +347,10 @@ struct dc_caps {
/* Conservative limit for DCC cases which require ODM4:1 to support*/
uint32_t dcc_plane_width_limit;
struct dc_scl_caps scl_caps;
+ uint8_t num_of_host_routers;
+ uint8_t num_of_dpias_per_host_router;
+ /* limit of the ODM only, could be limited by other factors (like pipe count)*/
+ uint8_t max_odm_combine_factor;
};
struct dc_bug_wa {
@@ -442,6 +495,7 @@ struct dc_config {
bool enable_windowed_mpo_odm;
bool forceHBR2CP2520; // Used for switching between test patterns TPS4 and CP2520
uint32_t allow_edp_hotplug_detection;
+ bool skip_riommu_prefetch_wa;
bool clamp_min_dcfclk;
uint64_t vblank_alignment_dto_params;
uint8_t vblank_alignment_max_frame_time_diff;
@@ -451,6 +505,7 @@ struct dc_config {
bool use_spl;
bool prefer_easf;
bool use_pipe_ctx_sync_logic;
+ int smart_mux_version;
bool ignore_dpref_ss;
bool enable_mipi_converter_optimization;
bool use_default_clock_table;
@@ -461,7 +516,9 @@ struct dc_config {
bool EnableMinDispClkODM;
bool enable_auto_dpm_test_logs;
unsigned int disable_ips;
+ unsigned int disable_ips_rcg;
unsigned int disable_ips_in_vpb;
+ bool disable_ips_in_dpms_off;
bool usb4_bw_alloc_support;
bool allow_0_dtb_clk;
bool use_assr_psp_message;
@@ -470,6 +527,10 @@ struct dc_config {
bool disable_hbr_audio_dp2;
bool consolidated_dpia_dp_lt;
bool set_pipe_unlock_order;
+ bool enable_dpia_pre_training;
+ bool unify_link_enc_assignment;
+ struct spl_sharpness_range dcn_sharpness_range;
+ struct spl_sharpness_range dcn_override_sharpness_range;
};
enum visual_confirm {
@@ -481,11 +542,15 @@ enum visual_confirm {
VISUAL_CONFIRM_SWAPCHAIN = 6,
VISUAL_CONFIRM_FAMS = 7,
VISUAL_CONFIRM_SWIZZLE = 9,
+ VISUAL_CONFIRM_SMARTMUX_DGPU = 10,
VISUAL_CONFIRM_REPLAY = 12,
VISUAL_CONFIRM_SUBVP = 14,
VISUAL_CONFIRM_MCLK_SWITCH = 16,
VISUAL_CONFIRM_FAMS2 = 19,
VISUAL_CONFIRM_HW_CURSOR = 20,
+ VISUAL_CONFIRM_VABC = 21,
+ VISUAL_CONFIRM_DCC = 22,
+ VISUAL_CONFIRM_EXPLICIT = 0x80000000,
};
enum dc_psr_power_opts {
@@ -627,6 +692,8 @@ struct dc_clocks {
int bw_dispclk_khz;
int idle_dramclk_khz;
int idle_fclk_khz;
+ int subvp_prefetch_dramclk_khz;
+ int subvp_prefetch_fclk_khz;
};
struct dc_bw_validation_profile {
@@ -754,6 +821,7 @@ enum pg_hw_resources {
PG_DCHVM,
PG_DWB,
PG_HPO,
+ PG_DCOH,
PG_HW_RESOURCES_NUM_ELEMENT
};
@@ -770,7 +838,7 @@ union dpia_debug_options {
uint32_t disable_mst_dsc_work_around:1; /* bit 3 */
uint32_t enable_force_tbt3_work_around:1; /* bit 4 */
uint32_t disable_usb4_pm_support:1; /* bit 5 */
- uint32_t enable_consolidated_dpia_dp_lt:1; /* bit 6 */
+ uint32_t enable_usb4_bw_zero_alloc_patch:1; /* bit 6 */
uint32_t reserved:25;
} bits;
uint32_t raw;
@@ -888,12 +956,18 @@ struct dc_debug_options {
bool voltage_align_fclk;
bool disable_min_fclk;
+ bool hdcp_lc_force_fw_enable;
+ bool hdcp_lc_enable_sw_fallback;
+
bool disable_dfs_bypass;
bool disable_dpp_power_gate;
bool disable_hubp_power_gate;
bool disable_dsc_power_gate;
bool disable_optc_power_gate;
bool disable_hpo_power_gate;
+ bool disable_io_clk_power_gate;
+ bool disable_mem_power_gate;
+ bool disable_dio_power_gate;
int dsc_min_slice_height_override;
int dsc_bpp_increment_div;
bool disable_pplib_wm_range;
@@ -1054,8 +1128,8 @@ struct dc_debug_options {
bool dml21_force_pstate_method;
uint32_t dml21_force_pstate_method_values[MAX_PIPES];
uint32_t dml21_disable_pstate_method_mask;
+ union fw_assisted_mclk_switch_version fams_version;
union dmub_fams2_global_feature_config fams2_config;
- bool enable_legacy_clock_update;
unsigned int force_cositing;
unsigned int disable_spl;
unsigned int force_easf;
@@ -1068,6 +1142,9 @@ struct dc_debug_options {
unsigned int scale_to_sharpness_policy;
bool skip_full_updated_if_possible;
unsigned int enable_oled_edp_power_up_opt;
+ bool enable_hblank_borrow;
+ bool force_subvp_df_throttle;
+ uint32_t acpi_transition_bitmasks[MAX_PIPES];
};
@@ -1127,7 +1204,7 @@ struct dc_init_data {
uint32_t *dcn_reg_offsets;
uint32_t *nbio_reg_offsets;
uint32_t *clk_reg_offsets;
- struct dml2_soc_bb *bb_from_dmub;
+ void *bb_from_dmub;
};
struct dc_callback_init {
@@ -1228,6 +1305,12 @@ union dc_3dlut_state {
};
+struct dc_rmcm_3dlut {
+ bool isInUse;
+ const struct dc_stream_state *stream;
+ uint8_t protection_bits;
+};
+
struct dc_3dlut {
struct kref refcount;
struct tetrahedral_params lut_3d;
@@ -1298,7 +1381,7 @@ struct dc_plane_state {
struct rect clip_rect;
struct plane_size plane_size;
- union dc_tiling_info tiling_info;
+ struct dc_tiling_info tiling_info;
struct dc_plane_dcc_param dcc;
@@ -1365,11 +1448,13 @@ struct dc_plane_state {
int sharpness_level;
enum linear_light_scaling linear_light_scaling;
unsigned int sdr_white_level_nits;
+ struct spl_sharpness_range sharpness_range;
+ enum sharpness_range_source sharpness_source;
};
struct dc_plane_info {
struct plane_size plane_size;
- union dc_tiling_info tiling_info;
+ struct dc_tiling_info tiling_info;
struct dc_plane_dcc_param dcc;
enum surface_pixel_format format;
enum dc_rotation_angle rotation;
@@ -1396,11 +1481,179 @@ struct dc_scratch_space {
* store current value in plane states so we can still recover
* a valid current state during dc update.
*/
- struct dc_plane_state plane_states[MAX_SURFACE_NUM];
+ struct dc_plane_state plane_states[MAX_SURFACES];
struct dc_stream_state stream_state;
};
+/*
+ * A link contains one or more sinks and their connected status.
+ * The currently active signal type (HDMI, DP-SST, DP-MST) is also reported.
+ */
+ struct dc_link {
+ struct dc_sink *remote_sinks[MAX_SINKS_PER_LINK];
+ unsigned int sink_count;
+ struct dc_sink *local_sink;
+ unsigned int link_index;
+ enum dc_connection_type type;
+ enum signal_type connector_signal;
+ enum dc_irq_source irq_source_hpd;
+ enum dc_irq_source irq_source_hpd_rx;/* aka DP Short Pulse */
+ enum dc_irq_source irq_source_read_request;/* Read Request */
+
+ bool is_hpd_filter_disabled;
+ bool dp_ss_off;
+
+ /**
+ * @link_state_valid:
+ *
+ * If there is no link and local sink, this variable should be set to
+ * false. Otherwise, it should be set to true; usually, the function
+ * core_link_enable_stream sets this field to true.
+ */
+ bool link_state_valid;
+ bool aux_access_disabled;
+ bool sync_lt_in_progress;
+ bool skip_stream_reenable;
+ bool is_internal_display;
+ /** @todo Rename. Flag an endpoint as having a programmable mapping to a DIG encoder. */
+ bool is_dig_mapping_flexible;
+ bool hpd_status; /* HPD status of link without physical HPD pin. */
+ bool is_hpd_pending; /* Indicates a new received hpd */
+
+ /* USB4 DPIA links skip verifying link cap, instead performing the fallback method
+ * for every link training. This is incompatible with DP LL compliance automation,
+ * which expects the same link settings to be used every retry on a link loss.
+ * This flag is used to skip the fallback when link loss occurs during automation.
+ */
+ bool skip_fallback_on_link_loss;
+
+ bool edp_sink_present;
+
+ struct dp_trace dp_trace;
+
+ /* caps is the same as reported_link_cap. link_traing use
+ * reported_link_cap. Will clean up. TODO
+ */
+ struct dc_link_settings reported_link_cap;
+ struct dc_link_settings verified_link_cap;
+ struct dc_link_settings cur_link_settings;
+ struct dc_lane_settings cur_lane_setting[LANE_COUNT_DP_MAX];
+ struct dc_link_settings preferred_link_setting;
+ /* preferred_training_settings are override values that
+ * come from DM. DM is responsible for the memory
+ * management of the override pointers.
+ */
+ struct dc_link_training_overrides preferred_training_settings;
+ struct dp_audio_test_data audio_test_data;
+
+ uint8_t ddc_hw_inst;
+
+ uint8_t hpd_src;
+
+ uint8_t link_enc_hw_inst;
+ /* DIG link encoder ID. Used as index in link encoder resource pool.
+ * For links with fixed mapping to DIG, this is not changed after dc_link
+ * object creation.
+ */
+ enum engine_id eng_id;
+ enum engine_id dpia_preferred_eng_id;
+
+ bool test_pattern_enabled;
+ /* Pending/Current test pattern are only used to perform and track
+ * FIXED_VS retimer test pattern/lane adjustment override state.
+ * Pending allows link HWSS to differentiate PHY vs non-PHY pattern,
+ * to perform specific lane adjust overrides before setting certain
+ * PHY test patterns. In cases when lane adjust and set test pattern
+ * calls are not performed atomically (i.e. performing link training),
+ * pending_test_pattern will be invalid or contain a non-PHY test pattern
+ * and current_test_pattern will contain required context for any future
+ * set pattern/set lane adjust to transition between override state(s).
+ * */
+ enum dp_test_pattern current_test_pattern;
+ enum dp_test_pattern pending_test_pattern;
+
+ union compliance_test_state compliance_test_state;
+
+ void *priv;
+
+ struct ddc_service *ddc;
+
+ enum dp_panel_mode panel_mode;
+ bool aux_mode;
+
+ /* Private to DC core */
+
+ const struct dc *dc;
+
+ struct dc_context *ctx;
+
+ struct panel_cntl *panel_cntl;
+ struct link_encoder *link_enc;
+ struct graphics_object_id link_id;
+ /* Endpoint type distinguishes display endpoints which do not have entries
+ * in the BIOS connector table from those that do. Helps when tracking link
+ * encoder to display endpoint assignments.
+ */
+ enum display_endpoint_type ep_type;
+ union ddi_channel_mapping ddi_channel_mapping;
+ struct connector_device_tag_info device_tag;
+ struct dpcd_caps dpcd_caps;
+ uint32_t dongle_max_pix_clk;
+ unsigned short chip_caps;
+ unsigned int dpcd_sink_count;
+ struct hdcp_caps hdcp_caps;
+ enum edp_revision edp_revision;
+ union dpcd_sink_ext_caps dpcd_sink_ext_caps;
+
+ struct psr_settings psr_settings;
+ struct replay_settings replay_settings;
+
+ /* Drive settings read from integrated info table */
+ struct dc_lane_settings bios_forced_drive_settings;
+
+ /* Vendor specific LTTPR workaround variables */
+ uint8_t vendor_specific_lttpr_link_rate_wa;
+ bool apply_vendor_specific_lttpr_link_rate_wa;
+
+ /* MST record stream using this link */
+ struct link_flags {
+ bool dp_keep_receiver_powered;
+ bool dp_skip_DID2;
+ bool dp_skip_reset_segment;
+ bool dp_skip_fs_144hz;
+ bool dp_mot_reset_segment;
+ /* Some USB4 docks do not handle turning off MST DSC once it has been enabled. */
+ bool dpia_mst_dsc_always_on;
+ /* Forced DPIA into TBT3 compatibility mode. */
+ bool dpia_forced_tbt3_mode;
+ bool dongle_mode_timing_override;
+ bool blank_stream_on_ocs_change;
+ bool read_dpcd204h_on_irq_hpd;
+ bool force_dp_ffe_preset;
+ bool skip_phy_ssc_reduction;
+ } wa_flags;
+ union dc_dp_ffe_preset forced_dp_ffe_preset;
+ struct link_mst_stream_allocation_table mst_stream_alloc_table;
+
+ struct dc_link_status link_status;
+ struct dprx_states dprx_states;
+
+ struct gpio *hpd_gpio;
+ enum dc_link_fec_state fec_state;
+ bool is_dds;
+ bool is_display_mux_present;
+ bool link_powered_externally; // Used to bypass hardware sequencing delays when panel is powered down forcibly
+
+ struct dc_panel_config panel_config;
+ struct phy_state phy_state;
+ uint32_t phy_transition_bitmask;
+ // BW ALLOCATON USB4 ONLY
+ struct dc_dpia_bw_alloc dpia_bw_alloc_config;
+ bool skip_implict_edp_power_control;
+ enum backlight_control_type backlight_control_type;
+};
+
struct dc {
struct dc_debug_options debug;
struct dc_versions versions;
@@ -1414,6 +1667,7 @@ struct dc {
uint8_t link_count;
struct dc_link *links[MAX_LINKS];
+ uint8_t lowest_dpia_link_index;
struct link_service *link_srv;
struct dc_state *current_state;
@@ -1443,6 +1697,10 @@ struct dc {
/* Require to maintain clocks and bandwidth for UEFI enabled HW */
+ /* For eDP to know the switching state of SmartMux */
+ bool is_switch_in_progress_orig;
+ bool is_switch_in_progress_dest;
+
/* FBC compressor */
struct compressor *fbc_compressor;
@@ -1468,11 +1726,12 @@ struct dc {
struct dc_scratch_space current_state;
struct dc_scratch_space new_state;
struct dc_stream_state temp_stream; // Used so we don't need to allocate stream on the stack
+ struct dc_link temp_link;
bool pipes_to_unlock_first[MAX_PIPES]; /* Any of the pipes indicated here should be unlocked first */
} scratch;
struct dml2_configuration_options dml2_options;
- struct dml2_configuration_options dml2_tmp;
+ struct dml2_configuration_options dml2_dc_power_options;
enum dc_acpi_cm_power_state power_state;
};
@@ -1524,6 +1783,7 @@ struct dc_surface_update {
const struct dc_cm2_parameters *cm2_params;
const struct dc_csc_transform *cursor_csc_color_matrix;
unsigned int sdr_white_level_nits;
+ struct dc_bias_and_scale bias_and_scale;
};
/*
@@ -1572,25 +1832,19 @@ bool dc_validate_boot_timing(const struct dc *dc,
enum dc_status dc_validate_plane(struct dc *dc, const struct dc_plane_state *plane_state);
-void get_clock_requirements_for_state(struct dc_state *state, struct AsicStateEx *info);
-
enum dc_status dc_validate_with_context(struct dc *dc,
const struct dc_validation_set set[],
int set_count,
struct dc_state *context,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
bool dc_set_generic_gpio_for_stereo(bool enable,
struct gpio_service *gpio_service);
-/*
- * fast_validate: we return after determining if we can support the new state,
- * but before we populate the programming info
- */
enum dc_status dc_validate_global_state(
struct dc *dc,
struct dc_state *new_ctx,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
bool dc_acquire_release_mpc_3dlut(
struct dc *dc, bool acquire,
@@ -1635,167 +1889,6 @@ uint32_t dc_bandwidth_in_kbps_from_timing(
const enum dc_link_encoding_format link_encoding);
/* Link Interfaces */
-/*
- * A link contains one or more sinks and their connected status.
- * The currently active signal type (HDMI, DP-SST, DP-MST) is also reported.
- */
-struct dc_link {
- struct dc_sink *remote_sinks[MAX_SINKS_PER_LINK];
- unsigned int sink_count;
- struct dc_sink *local_sink;
- unsigned int link_index;
- enum dc_connection_type type;
- enum signal_type connector_signal;
- enum dc_irq_source irq_source_hpd;
- enum dc_irq_source irq_source_hpd_rx;/* aka DP Short Pulse */
-
- bool is_hpd_filter_disabled;
- bool dp_ss_off;
-
- /**
- * @link_state_valid:
- *
- * If there is no link and local sink, this variable should be set to
- * false. Otherwise, it should be set to true; usually, the function
- * core_link_enable_stream sets this field to true.
- */
- bool link_state_valid;
- bool aux_access_disabled;
- bool sync_lt_in_progress;
- bool skip_stream_reenable;
- bool is_internal_display;
- /** @todo Rename. Flag an endpoint as having a programmable mapping to a DIG encoder. */
- bool is_dig_mapping_flexible;
- bool hpd_status; /* HPD status of link without physical HPD pin. */
- bool is_hpd_pending; /* Indicates a new received hpd */
-
- /* USB4 DPIA links skip verifying link cap, instead performing the fallback method
- * for every link training. This is incompatible with DP LL compliance automation,
- * which expects the same link settings to be used every retry on a link loss.
- * This flag is used to skip the fallback when link loss occurs during automation.
- */
- bool skip_fallback_on_link_loss;
-
- bool edp_sink_present;
-
- struct dp_trace dp_trace;
-
- /* caps is the same as reported_link_cap. link_traing use
- * reported_link_cap. Will clean up. TODO
- */
- struct dc_link_settings reported_link_cap;
- struct dc_link_settings verified_link_cap;
- struct dc_link_settings cur_link_settings;
- struct dc_lane_settings cur_lane_setting[LANE_COUNT_DP_MAX];
- struct dc_link_settings preferred_link_setting;
- /* preferred_training_settings are override values that
- * come from DM. DM is responsible for the memory
- * management of the override pointers.
- */
- struct dc_link_training_overrides preferred_training_settings;
- struct dp_audio_test_data audio_test_data;
-
- uint8_t ddc_hw_inst;
-
- uint8_t hpd_src;
-
- uint8_t link_enc_hw_inst;
- /* DIG link encoder ID. Used as index in link encoder resource pool.
- * For links with fixed mapping to DIG, this is not changed after dc_link
- * object creation.
- */
- enum engine_id eng_id;
- enum engine_id dpia_preferred_eng_id;
-
- bool test_pattern_enabled;
- /* Pending/Current test pattern are only used to perform and track
- * FIXED_VS retimer test pattern/lane adjustment override state.
- * Pending allows link HWSS to differentiate PHY vs non-PHY pattern,
- * to perform specific lane adjust overrides before setting certain
- * PHY test patterns. In cases when lane adjust and set test pattern
- * calls are not performed atomically (i.e. performing link training),
- * pending_test_pattern will be invalid or contain a non-PHY test pattern
- * and current_test_pattern will contain required context for any future
- * set pattern/set lane adjust to transition between override state(s).
- * */
- enum dp_test_pattern current_test_pattern;
- enum dp_test_pattern pending_test_pattern;
-
- union compliance_test_state compliance_test_state;
-
- void *priv;
-
- struct ddc_service *ddc;
-
- enum dp_panel_mode panel_mode;
- bool aux_mode;
-
- /* Private to DC core */
-
- const struct dc *dc;
-
- struct dc_context *ctx;
-
- struct panel_cntl *panel_cntl;
- struct link_encoder *link_enc;
- struct graphics_object_id link_id;
- /* Endpoint type distinguishes display endpoints which do not have entries
- * in the BIOS connector table from those that do. Helps when tracking link
- * encoder to display endpoint assignments.
- */
- enum display_endpoint_type ep_type;
- union ddi_channel_mapping ddi_channel_mapping;
- struct connector_device_tag_info device_tag;
- struct dpcd_caps dpcd_caps;
- uint32_t dongle_max_pix_clk;
- unsigned short chip_caps;
- unsigned int dpcd_sink_count;
- struct hdcp_caps hdcp_caps;
- enum edp_revision edp_revision;
- union dpcd_sink_ext_caps dpcd_sink_ext_caps;
-
- struct psr_settings psr_settings;
- struct replay_settings replay_settings;
-
- /* Drive settings read from integrated info table */
- struct dc_lane_settings bios_forced_drive_settings;
-
- /* Vendor specific LTTPR workaround variables */
- uint8_t vendor_specific_lttpr_link_rate_wa;
- bool apply_vendor_specific_lttpr_link_rate_wa;
-
- /* MST record stream using this link */
- struct link_flags {
- bool dp_keep_receiver_powered;
- bool dp_skip_DID2;
- bool dp_skip_reset_segment;
- bool dp_skip_fs_144hz;
- bool dp_mot_reset_segment;
- /* Some USB4 docks do not handle turning off MST DSC once it has been enabled. */
- bool dpia_mst_dsc_always_on;
- /* Forced DPIA into TBT3 compatibility mode. */
- bool dpia_forced_tbt3_mode;
- bool dongle_mode_timing_override;
- bool blank_stream_on_ocs_change;
- bool read_dpcd204h_on_irq_hpd;
- } wa_flags;
- struct link_mst_stream_allocation_table mst_stream_alloc_table;
-
- struct dc_link_status link_status;
- struct dprx_states dprx_states;
-
- struct gpio *hpd_gpio;
- enum dc_link_fec_state fec_state;
- bool link_powered_externally; // Used to bypass hardware sequencing delays when panel is powered down forcibly
-
- struct dc_panel_config panel_config;
- struct phy_state phy_state;
- // BW ALLOCATON USB4 ONLY
- struct dc_dpia_bw_alloc dpia_bw_alloc_config;
- bool skip_implict_edp_power_control;
- enum backlight_control_type backlight_control_type;
-};
-
/* Return an enumerated dc_link.
* dc_link order is constant and determined at
* boot time. They cannot be created or destroyed.
@@ -1937,6 +2030,9 @@ int dc_link_aux_transfer_raw(struct ddc_service *ddc,
struct aux_payload *payload,
enum aux_return_code_type *operation_result);
+struct ddc_service *
+dc_get_oem_i2c_device(struct dc *dc);
+
bool dc_is_oem_i2c_device_present(
struct dc *dc,
size_t slave_address
@@ -2017,6 +2113,24 @@ uint32_t dc_link_bandwidth_kbps(
const struct dc_link *link,
const struct dc_link_settings *link_setting);
+struct dp_audio_bandwidth_params {
+ const struct dc_crtc_timing *crtc_timing;
+ enum dp_link_encoding link_encoding;
+ uint32_t channel_count;
+ uint32_t sample_rate_hz;
+};
+
+/* The function calculates the minimum size of hblank (in bytes) needed to
+ * support the specified channel count and sample rate combination, given the
+ * link encoding and timing to be used. This calculation is not supported
+ * for 8b/10b SST.
+ *
+ * return - min hblank size in bytes, 0 if 8b/10b SST.
+ */
+uint32_t dc_link_required_hblank_size_bytes(
+ const struct dc_link *link,
+ struct dp_audio_bandwidth_params *audio_params);
+
/* The function takes a snapshot of current link resource allocation state
* @dc: pointer to dc of the dm calling this
* @map: a dc link resource snapshot defined internally to dc.
@@ -2314,19 +2428,6 @@ unsigned int dc_dp_trace_get_link_loss_count(struct dc_link *link);
void dc_link_set_usb4_req_bw_req(struct dc_link *link, int req_bw);
/*
- * Handle function for when the status of the Request above is complete.
- * We will find out the result of allocating on CM and update structs.
- *
- * @link: pointer to the dc_link struct instance
- * @bw: Allocated or Estimated BW depending on the result
- * @result: Response type
- *
- * return: none
- */
-void dc_link_handle_usb4_bw_alloc_response(struct dc_link *link,
- uint8_t bw, uint8_t result);
-
-/*
* Handle the USB4 BW Allocation related functionality here:
* Plug => Try to allocate max bw from timing parameters supported by the sink
* Unplug => de-allocate bw
@@ -2334,23 +2435,17 @@ void dc_link_handle_usb4_bw_alloc_response(struct dc_link *link,
* @link: pointer to the dc_link struct instance
* @peak_bw: Peak bw used by the link/sink
*
- * return: allocated bw else return 0
*/
-int dc_link_dp_dpia_handle_usb4_bandwidth_allocation_for_link(
+void dc_link_dp_dpia_handle_usb4_bandwidth_allocation_for_link(
struct dc_link *link, int peak_bw);
/*
- * Validate the BW of all the valid DPIA links to make sure it doesn't exceed
- * available BW for each host router
+ * Calculates the DP tunneling bandwidth required for the stream timing
+ * and aggregates the stream bandwidth for the respective DP tunneling link
*
- * @dc: pointer to dc struct
- * @stream: pointer to all possible streams
- * @count: number of valid DPIA streams
- *
- * return: TRUE if bw used by DPIAs doesn't exceed available BW else return FALSE
+ * return: dc_status
*/
-bool dc_link_dp_dpia_validate(struct dc *dc, const struct dc_stream_state *streams,
- const unsigned int count);
+enum dc_status dc_link_validate_dp_tunneling_bandwidth(const struct dc *dc, const struct dc_state *new_ctx);
/* Sink Interfaces - A sink corresponds to a display output device */
@@ -2376,6 +2471,13 @@ struct dc_sink_dsc_caps {
struct dsc_dec_dpcd_caps dsc_dec_caps;
};
+struct dc_sink_hblank_expansion_caps {
+ // 'true' if these are virtual DPCD's HBlank expansion caps (immediately upstream of sink in MST topology),
+ // 'false' if they are sink's HBlank expansion caps
+ bool is_virtual_dpcd_hblank_expansion;
+ struct hblank_expansion_dpcd_caps dpcd_caps;
+};
+
struct dc_sink_fec_caps {
bool is_rx_fec_supported;
bool is_topology_fec_supported;
@@ -2402,6 +2504,7 @@ struct dc_sink {
struct scdc_caps scdc_caps;
struct dc_sink_dsc_caps dsc_caps;
struct dc_sink_fec_caps fec_caps;
+ struct dc_sink_hblank_expansion_caps hblank_expansion_caps;
bool is_vsc_sdp_colorimetry_supported;
@@ -2550,13 +2653,25 @@ struct dc_power_profile {
struct dc_power_profile dc_get_power_profile_for_dc_state(const struct dc_state *context);
+unsigned int dc_get_det_buffer_size_from_state(const struct dc_state *context);
+
+bool dc_get_host_router_index(const struct dc_link *link, unsigned int *host_router_index);
+
/* DSC Interfaces */
#include "dc_dsc.h"
+void dc_get_visual_confirm_for_stream(
+ struct dc *dc,
+ struct dc_stream_state *stream_state,
+ struct tg_color *color);
+
/* Disable acc mode Interfaces */
void dc_disable_accelerated_mode(struct dc *dc);
bool dc_is_timing_changed(struct dc_stream_state *cur_stream,
struct dc_stream_state *new_stream);
+bool dc_is_cursor_limit_pending(struct dc *dc);
+bool dc_can_clear_cursor_limit(struct dc *dc);
+
#endif /* DC_INTERFACE_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c
index f90fc154549a..f5ef1a07078e 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c
+++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c
@@ -39,6 +39,7 @@
#define CTX dc_dmub_srv->ctx
#define DC_LOGGER CTX->logger
+#define GPINT_RETRY_NUM 20
static void dc_dmub_srv_construct(struct dc_dmub_srv *dc_srv, struct dc *dc,
struct dmub_srv *dmub)
@@ -70,20 +71,28 @@ void dc_dmub_srv_destroy(struct dc_dmub_srv **dmub_srv)
}
}
-void dc_dmub_srv_wait_idle(struct dc_dmub_srv *dc_dmub_srv)
+bool dc_dmub_srv_wait_for_pending(struct dc_dmub_srv *dc_dmub_srv)
{
- struct dmub_srv *dmub = dc_dmub_srv->dmub;
- struct dc_context *dc_ctx = dc_dmub_srv->ctx;
+ struct dmub_srv *dmub;
+ struct dc_context *dc_ctx;
enum dmub_status status;
+ if (!dc_dmub_srv || !dc_dmub_srv->dmub)
+ return false;
+
+ dc_ctx = dc_dmub_srv->ctx;
+ dmub = dc_dmub_srv->dmub;
+
do {
- status = dmub_srv_wait_for_idle(dmub, 100000);
+ status = dmub_srv_wait_for_pending(dmub, 100000);
} while (dc_dmub_srv->ctx->dc->debug.disable_timeout && status != DMUB_STATUS_OK);
if (status != DMUB_STATUS_OK) {
DC_ERROR("Error waiting for DMUB idle: status=%d\n", status);
dc_dmub_srv_log_diagnostic_data(dc_dmub_srv);
}
+
+ return status == DMUB_STATUS_OK;
}
void dc_dmub_srv_clear_inbox0_ack(struct dc_dmub_srv *dc_dmub_srv)
@@ -126,7 +135,49 @@ void dc_dmub_srv_send_inbox0_cmd(struct dc_dmub_srv *dc_dmub_srv,
}
}
-bool dc_dmub_srv_cmd_list_queue_execute(struct dc_dmub_srv *dc_dmub_srv,
+static bool dc_dmub_srv_reg_cmd_list_queue_execute(struct dc_dmub_srv *dc_dmub_srv,
+ unsigned int count,
+ union dmub_rb_cmd *cmd_list)
+{
+ struct dc_context *dc_ctx;
+ struct dmub_srv *dmub;
+ enum dmub_status status = DMUB_STATUS_OK;
+ int i;
+
+ if (!dc_dmub_srv || !dc_dmub_srv->dmub)
+ return false;
+
+ dc_ctx = dc_dmub_srv->ctx;
+ dmub = dc_dmub_srv->dmub;
+
+ for (i = 0 ; i < count; i++) {
+ /* confirm no messages pending */
+ do {
+ status = dmub_srv_wait_for_idle(dmub, 100000);
+ } while (dc_dmub_srv->ctx->dc->debug.disable_timeout && status != DMUB_STATUS_OK);
+
+ /* queue command */
+ if (status == DMUB_STATUS_OK)
+ status = dmub_srv_reg_cmd_execute(dmub, &cmd_list[i]);
+
+ /* check for errors */
+ if (status != DMUB_STATUS_OK) {
+ break;
+ }
+ }
+
+ if (status != DMUB_STATUS_OK) {
+ if (status != DMUB_STATUS_POWER_STATE_D3) {
+ DC_ERROR("Error starting DMUB execution: status=%d\n", status);
+ dc_dmub_srv_log_diagnostic_data(dc_dmub_srv);
+ }
+ return false;
+ }
+
+ return true;
+}
+
+static bool dc_dmub_srv_fb_cmd_list_queue_execute(struct dc_dmub_srv *dc_dmub_srv,
unsigned int count,
union dmub_rb_cmd *cmd_list)
{
@@ -143,20 +194,25 @@ bool dc_dmub_srv_cmd_list_queue_execute(struct dc_dmub_srv *dc_dmub_srv,
for (i = 0 ; i < count; i++) {
// Queue command
- status = dmub_srv_cmd_queue(dmub, &cmd_list[i]);
+ if (!cmd_list[i].cmd_common.header.multi_cmd_pending ||
+ dmub_rb_num_free(&dmub->inbox1.rb) >= count - i) {
+ status = dmub_srv_fb_cmd_queue(dmub, &cmd_list[i]);
+ } else {
+ status = DMUB_STATUS_QUEUE_FULL;
+ }
if (status == DMUB_STATUS_QUEUE_FULL) {
/* Execute and wait for queue to become empty again. */
- status = dmub_srv_cmd_execute(dmub);
+ status = dmub_srv_fb_cmd_execute(dmub);
if (status == DMUB_STATUS_POWER_STATE_D3)
return false;
do {
- status = dmub_srv_wait_for_idle(dmub, 100000);
+ status = dmub_srv_wait_for_inbox_free(dmub, 100000, count - i);
} while (dc_dmub_srv->ctx->dc->debug.disable_timeout && status != DMUB_STATUS_OK);
/* Requeue the command. */
- status = dmub_srv_cmd_queue(dmub, &cmd_list[i]);
+ status = dmub_srv_fb_cmd_queue(dmub, &cmd_list[i]);
}
if (status != DMUB_STATUS_OK) {
@@ -168,7 +224,7 @@ bool dc_dmub_srv_cmd_list_queue_execute(struct dc_dmub_srv *dc_dmub_srv,
}
}
- status = dmub_srv_cmd_execute(dmub);
+ status = dmub_srv_fb_cmd_execute(dmub);
if (status != DMUB_STATUS_OK) {
if (status != DMUB_STATUS_POWER_STATE_D3) {
DC_ERROR("Error starting DMUB execution: status=%d\n", status);
@@ -180,6 +236,26 @@ bool dc_dmub_srv_cmd_list_queue_execute(struct dc_dmub_srv *dc_dmub_srv,
return true;
}
+bool dc_dmub_srv_cmd_list_queue_execute(struct dc_dmub_srv *dc_dmub_srv,
+ unsigned int count,
+ union dmub_rb_cmd *cmd_list)
+{
+ bool res = false;
+
+ if (dc_dmub_srv && dc_dmub_srv->dmub) {
+ if (dc_dmub_srv->dmub->inbox_type == DMUB_CMD_INTERFACE_REG) {
+ res = dc_dmub_srv_reg_cmd_list_queue_execute(dc_dmub_srv, count, cmd_list);
+ } else {
+ res = dc_dmub_srv_fb_cmd_list_queue_execute(dc_dmub_srv, count, cmd_list);
+ }
+
+ if (res)
+ res = dmub_srv_update_inbox_status(dc_dmub_srv->dmub) == DMUB_STATUS_OK;
+ }
+
+ return res;
+}
+
bool dc_dmub_srv_wait_for_idle(struct dc_dmub_srv *dc_dmub_srv,
enum dm_dmub_wait_type wait_type,
union dmub_rb_cmd *cmd_list)
@@ -200,18 +276,20 @@ bool dc_dmub_srv_wait_for_idle(struct dc_dmub_srv *dc_dmub_srv,
if (status != DMUB_STATUS_OK) {
DC_LOG_DEBUG("No reply for DMUB command: status=%d\n", status);
- if (!dmub->debug.timeout_occured) {
- dmub->debug.timeout_occured = true;
- dmub->debug.timeout_cmd = *cmd_list;
- dmub->debug.timestamp = dm_get_timestamp(dc_dmub_srv->ctx);
+ if (!dmub->debug.timeout_info.timeout_occured) {
+ dmub->debug.timeout_info.timeout_occured = true;
+ if (cmd_list)
+ dmub->debug.timeout_info.timeout_cmd = *cmd_list;
+ dmub->debug.timeout_info.timestamp = dm_get_timestamp(dc_dmub_srv->ctx);
}
dc_dmub_srv_log_diagnostic_data(dc_dmub_srv);
return false;
}
// Copy data back from ring buffer into command
- if (wait_type == DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)
- dmub_rb_get_return_data(&dmub->inbox1_rb, cmd_list);
+ if (wait_type == DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY && cmd_list) {
+ dmub_srv_cmd_get_response(dc_dmub_srv->dmub, cmd_list);
+ }
}
return true;
@@ -224,74 +302,10 @@ bool dc_dmub_srv_cmd_run(struct dc_dmub_srv *dc_dmub_srv, union dmub_rb_cmd *cmd
bool dc_dmub_srv_cmd_run_list(struct dc_dmub_srv *dc_dmub_srv, unsigned int count, union dmub_rb_cmd *cmd_list, enum dm_dmub_wait_type wait_type)
{
- struct dc_context *dc_ctx;
- struct dmub_srv *dmub;
- enum dmub_status status;
- int i;
-
- if (!dc_dmub_srv || !dc_dmub_srv->dmub)
- return false;
-
- dc_ctx = dc_dmub_srv->ctx;
- dmub = dc_dmub_srv->dmub;
-
- for (i = 0 ; i < count; i++) {
- // Queue command
- status = dmub_srv_cmd_queue(dmub, &cmd_list[i]);
-
- if (status == DMUB_STATUS_QUEUE_FULL) {
- /* Execute and wait for queue to become empty again. */
- status = dmub_srv_cmd_execute(dmub);
- if (status == DMUB_STATUS_POWER_STATE_D3)
- return false;
-
- status = dmub_srv_wait_for_idle(dmub, 100000);
- if (status != DMUB_STATUS_OK)
- return false;
-
- /* Requeue the command. */
- status = dmub_srv_cmd_queue(dmub, &cmd_list[i]);
- }
-
- if (status != DMUB_STATUS_OK) {
- if (status != DMUB_STATUS_POWER_STATE_D3) {
- DC_ERROR("Error queueing DMUB command: status=%d\n", status);
- dc_dmub_srv_log_diagnostic_data(dc_dmub_srv);
- }
- return false;
- }
- }
-
- status = dmub_srv_cmd_execute(dmub);
- if (status != DMUB_STATUS_OK) {
- if (status != DMUB_STATUS_POWER_STATE_D3) {
- DC_ERROR("Error starting DMUB execution: status=%d\n", status);
- dc_dmub_srv_log_diagnostic_data(dc_dmub_srv);
- }
+ if (!dc_dmub_srv_cmd_list_queue_execute(dc_dmub_srv, count, cmd_list))
return false;
- }
-
- // Wait for DMUB to process command
- if (wait_type != DM_DMUB_WAIT_TYPE_NO_WAIT) {
- if (dc_dmub_srv->ctx->dc->debug.disable_timeout) {
- do {
- status = dmub_srv_wait_for_idle(dmub, 100000);
- } while (status != DMUB_STATUS_OK);
- } else
- status = dmub_srv_wait_for_idle(dmub, 100000);
-
- if (status != DMUB_STATUS_OK) {
- DC_LOG_DEBUG("No reply for DMUB command: status=%d\n", status);
- dc_dmub_srv_log_diagnostic_data(dc_dmub_srv);
- return false;
- }
- // Copy data back from ring buffer into command
- if (wait_type == DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY)
- dmub_rb_get_return_data(&dmub->inbox1_rb, cmd_list);
- }
-
- return true;
+ return dc_dmub_srv_wait_for_idle(dc_dmub_srv, wait_type, cmd_list);
}
bool dc_dmub_srv_optimized_init_done(struct dc_dmub_srv *dc_dmub_srv)
@@ -927,16 +941,15 @@ void dc_dmub_setup_subvp_dmub_command(struct dc *dc,
dc_wake_and_execute_dmub_cmd(dc->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
}
-bool dc_dmub_srv_get_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv, struct dmub_diagnostic_data *diag_data)
+bool dc_dmub_srv_get_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv)
{
- if (!dc_dmub_srv || !dc_dmub_srv->dmub || !diag_data)
+ if (!dc_dmub_srv || !dc_dmub_srv->dmub)
return false;
- return dmub_srv_get_diagnostic_data(dc_dmub_srv->dmub, diag_data);
+ return dmub_srv_get_diagnostic_data(dc_dmub_srv->dmub);
}
void dc_dmub_srv_log_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv)
{
- struct dmub_diagnostic_data diag_data = {0};
uint32_t i;
if (!dc_dmub_srv || !dc_dmub_srv->dmub) {
@@ -946,102 +959,56 @@ void dc_dmub_srv_log_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv)
DC_LOG_ERROR("%s: DMCUB error - collecting diagnostic data\n", __func__);
- if (!dc_dmub_srv_get_diagnostic_data(dc_dmub_srv, &diag_data)) {
+ if (!dc_dmub_srv_get_diagnostic_data(dc_dmub_srv)) {
DC_LOG_ERROR("%s: dc_dmub_srv_get_diagnostic_data failed.", __func__);
return;
}
DC_LOG_DEBUG("DMCUB STATE:");
- DC_LOG_DEBUG(" dmcub_version : %08x", diag_data.dmcub_version);
- DC_LOG_DEBUG(" scratch [0] : %08x", diag_data.scratch[0]);
- DC_LOG_DEBUG(" scratch [1] : %08x", diag_data.scratch[1]);
- DC_LOG_DEBUG(" scratch [2] : %08x", diag_data.scratch[2]);
- DC_LOG_DEBUG(" scratch [3] : %08x", diag_data.scratch[3]);
- DC_LOG_DEBUG(" scratch [4] : %08x", diag_data.scratch[4]);
- DC_LOG_DEBUG(" scratch [5] : %08x", diag_data.scratch[5]);
- DC_LOG_DEBUG(" scratch [6] : %08x", diag_data.scratch[6]);
- DC_LOG_DEBUG(" scratch [7] : %08x", diag_data.scratch[7]);
- DC_LOG_DEBUG(" scratch [8] : %08x", diag_data.scratch[8]);
- DC_LOG_DEBUG(" scratch [9] : %08x", diag_data.scratch[9]);
- DC_LOG_DEBUG(" scratch [10] : %08x", diag_data.scratch[10]);
- DC_LOG_DEBUG(" scratch [11] : %08x", diag_data.scratch[11]);
- DC_LOG_DEBUG(" scratch [12] : %08x", diag_data.scratch[12]);
- DC_LOG_DEBUG(" scratch [13] : %08x", diag_data.scratch[13]);
- DC_LOG_DEBUG(" scratch [14] : %08x", diag_data.scratch[14]);
- DC_LOG_DEBUG(" scratch [15] : %08x", diag_data.scratch[15]);
+ DC_LOG_DEBUG(" dmcub_version : %08x", dc_dmub_srv->dmub->debug.dmcub_version);
+ DC_LOG_DEBUG(" scratch [0] : %08x", dc_dmub_srv->dmub->debug.scratch[0]);
+ DC_LOG_DEBUG(" scratch [1] : %08x", dc_dmub_srv->dmub->debug.scratch[1]);
+ DC_LOG_DEBUG(" scratch [2] : %08x", dc_dmub_srv->dmub->debug.scratch[2]);
+ DC_LOG_DEBUG(" scratch [3] : %08x", dc_dmub_srv->dmub->debug.scratch[3]);
+ DC_LOG_DEBUG(" scratch [4] : %08x", dc_dmub_srv->dmub->debug.scratch[4]);
+ DC_LOG_DEBUG(" scratch [5] : %08x", dc_dmub_srv->dmub->debug.scratch[5]);
+ DC_LOG_DEBUG(" scratch [6] : %08x", dc_dmub_srv->dmub->debug.scratch[6]);
+ DC_LOG_DEBUG(" scratch [7] : %08x", dc_dmub_srv->dmub->debug.scratch[7]);
+ DC_LOG_DEBUG(" scratch [8] : %08x", dc_dmub_srv->dmub->debug.scratch[8]);
+ DC_LOG_DEBUG(" scratch [9] : %08x", dc_dmub_srv->dmub->debug.scratch[9]);
+ DC_LOG_DEBUG(" scratch [10] : %08x", dc_dmub_srv->dmub->debug.scratch[10]);
+ DC_LOG_DEBUG(" scratch [11] : %08x", dc_dmub_srv->dmub->debug.scratch[11]);
+ DC_LOG_DEBUG(" scratch [12] : %08x", dc_dmub_srv->dmub->debug.scratch[12]);
+ DC_LOG_DEBUG(" scratch [13] : %08x", dc_dmub_srv->dmub->debug.scratch[13]);
+ DC_LOG_DEBUG(" scratch [14] : %08x", dc_dmub_srv->dmub->debug.scratch[14]);
+ DC_LOG_DEBUG(" scratch [15] : %08x", dc_dmub_srv->dmub->debug.scratch[15]);
for (i = 0; i < DMUB_PC_SNAPSHOT_COUNT; i++)
- DC_LOG_DEBUG(" pc[%d] : %08x", i, diag_data.pc[i]);
- DC_LOG_DEBUG(" unk_fault_addr : %08x", diag_data.undefined_address_fault_addr);
- DC_LOG_DEBUG(" inst_fault_addr : %08x", diag_data.inst_fetch_fault_addr);
- DC_LOG_DEBUG(" data_fault_addr : %08x", diag_data.data_write_fault_addr);
- DC_LOG_DEBUG(" inbox1_rptr : %08x", diag_data.inbox1_rptr);
- DC_LOG_DEBUG(" inbox1_wptr : %08x", diag_data.inbox1_wptr);
- DC_LOG_DEBUG(" inbox1_size : %08x", diag_data.inbox1_size);
- DC_LOG_DEBUG(" inbox0_rptr : %08x", diag_data.inbox0_rptr);
- DC_LOG_DEBUG(" inbox0_wptr : %08x", diag_data.inbox0_wptr);
- DC_LOG_DEBUG(" inbox0_size : %08x", diag_data.inbox0_size);
- DC_LOG_DEBUG(" outbox1_rptr : %08x", diag_data.outbox1_rptr);
- DC_LOG_DEBUG(" outbox1_wptr : %08x", diag_data.outbox1_wptr);
- DC_LOG_DEBUG(" outbox1_size : %08x", diag_data.outbox1_size);
- DC_LOG_DEBUG(" is_enabled : %d", diag_data.is_dmcub_enabled);
- DC_LOG_DEBUG(" is_soft_reset : %d", diag_data.is_dmcub_soft_reset);
- DC_LOG_DEBUG(" is_secure_reset : %d", diag_data.is_dmcub_secure_reset);
- DC_LOG_DEBUG(" is_traceport_en : %d", diag_data.is_traceport_en);
- DC_LOG_DEBUG(" is_cw0_en : %d", diag_data.is_cw0_enabled);
- DC_LOG_DEBUG(" is_cw6_en : %d", diag_data.is_cw6_enabled);
-}
-
-static bool dc_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx)
-{
- struct pipe_ctx *test_pipe, *split_pipe;
- const struct scaler_data *scl_data = &pipe_ctx->plane_res.scl_data;
- struct rect r1 = scl_data->recout, r2, r2_half;
- int r1_r = r1.x + r1.width, r1_b = r1.y + r1.height, r2_r, r2_b;
- int cur_layer = pipe_ctx->plane_state->layer_index;
-
- /**
- * Disable the cursor if there's another pipe above this with a
- * plane that contains this pipe's viewport to prevent double cursor
- * and incorrect scaling artifacts.
- */
- for (test_pipe = pipe_ctx->top_pipe; test_pipe;
- test_pipe = test_pipe->top_pipe) {
- // Skip invisible layer and pipe-split plane on same layer
- if (!test_pipe->plane_state->visible || test_pipe->plane_state->layer_index == cur_layer)
- continue;
-
- r2 = test_pipe->plane_res.scl_data.recout;
- r2_r = r2.x + r2.width;
- r2_b = r2.y + r2.height;
-
- /**
- * There is another half plane on same layer because of
- * pipe-split, merge together per same height.
- */
- for (split_pipe = pipe_ctx->top_pipe; split_pipe;
- split_pipe = split_pipe->top_pipe)
- if (split_pipe->plane_state->layer_index == test_pipe->plane_state->layer_index) {
- r2_half = split_pipe->plane_res.scl_data.recout;
- r2.x = (r2_half.x < r2.x) ? r2_half.x : r2.x;
- r2.width = r2.width + r2_half.width;
- r2_r = r2.x + r2.width;
- break;
- }
-
- if (r1.x >= r2.x && r1.y >= r2.y && r1_r <= r2_r && r1_b <= r2_b)
- return true;
- }
-
- return false;
+ DC_LOG_DEBUG(" pc[%d] : %08x", i, dc_dmub_srv->dmub->debug.pc[i]);
+ DC_LOG_DEBUG(" unk_fault_addr : %08x", dc_dmub_srv->dmub->debug.undefined_address_fault_addr);
+ DC_LOG_DEBUG(" inst_fault_addr : %08x", dc_dmub_srv->dmub->debug.inst_fetch_fault_addr);
+ DC_LOG_DEBUG(" data_fault_addr : %08x", dc_dmub_srv->dmub->debug.data_write_fault_addr);
+ DC_LOG_DEBUG(" inbox1_rptr : %08x", dc_dmub_srv->dmub->debug.inbox1_rptr);
+ DC_LOG_DEBUG(" inbox1_wptr : %08x", dc_dmub_srv->dmub->debug.inbox1_wptr);
+ DC_LOG_DEBUG(" inbox1_size : %08x", dc_dmub_srv->dmub->debug.inbox1_size);
+ DC_LOG_DEBUG(" inbox0_rptr : %08x", dc_dmub_srv->dmub->debug.inbox0_rptr);
+ DC_LOG_DEBUG(" inbox0_wptr : %08x", dc_dmub_srv->dmub->debug.inbox0_wptr);
+ DC_LOG_DEBUG(" inbox0_size : %08x", dc_dmub_srv->dmub->debug.inbox0_size);
+ DC_LOG_DEBUG(" outbox1_rptr : %08x", dc_dmub_srv->dmub->debug.outbox1_rptr);
+ DC_LOG_DEBUG(" outbox1_wptr : %08x", dc_dmub_srv->dmub->debug.outbox1_wptr);
+ DC_LOG_DEBUG(" outbox1_size : %08x", dc_dmub_srv->dmub->debug.outbox1_size);
+ DC_LOG_DEBUG(" is_enabled : %d", dc_dmub_srv->dmub->debug.is_dmcub_enabled);
+ DC_LOG_DEBUG(" is_soft_reset : %d", dc_dmub_srv->dmub->debug.is_dmcub_soft_reset);
+ DC_LOG_DEBUG(" is_secure_reset : %d", dc_dmub_srv->dmub->debug.is_dmcub_secure_reset);
+ DC_LOG_DEBUG(" is_traceport_en : %d", dc_dmub_srv->dmub->debug.is_traceport_en);
+ DC_LOG_DEBUG(" is_cw0_en : %d", dc_dmub_srv->dmub->debug.is_cw0_enabled);
+ DC_LOG_DEBUG(" is_cw6_en : %d", dc_dmub_srv->dmub->debug.is_cw6_enabled);
}
static bool dc_dmub_should_update_cursor_data(struct pipe_ctx *pipe_ctx)
{
if (pipe_ctx->plane_state != NULL) {
- if (pipe_ctx->plane_state->address.type == PLN_ADDR_TYPE_VIDEO_PROGRESSIVE)
- return false;
-
- if (dc_can_pipe_disable_cursor(pipe_ctx))
+ if (pipe_ctx->plane_state->address.type == PLN_ADDR_TYPE_VIDEO_PROGRESSIVE ||
+ resource_can_pipe_disable_cursor(pipe_ctx))
return false;
}
@@ -1245,7 +1212,7 @@ static int count_active_streams(const struct dc *dc)
for (i = 0; i < dc->current_state->stream_count; ++i) {
struct dc_stream_state *stream = dc->current_state->streams[i];
- if (stream && !stream->dpms_off)
+ if (stream && (!stream->dpms_off || dc->config.disable_ips_in_dpms_off))
count += 1;
}
@@ -1290,7 +1257,7 @@ static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle)
ips_fw->signals.bits.ips1_commit,
ips_fw->signals.bits.ips2_commit);
- dc_dmub_srv_wait_idle(dc->ctx->dmub_srv);
+ dc_dmub_srv_wait_for_idle(dc->ctx->dmub_srv, DM_DMUB_WAIT_TYPE_WAIT, NULL);
memset(&new_signals, 0, sizeof(new_signals));
@@ -1302,12 +1269,16 @@ static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle)
new_signals.bits.allow_ips1 = 1;
new_signals.bits.allow_ips2 = 1;
new_signals.bits.allow_z10 = 1;
+ // New in IPSv2.0
+ new_signals.bits.allow_ips1z8 = 1;
} else if (dc->config.disable_ips == DMUB_IPS_DISABLE_IPS1) {
new_signals.bits.allow_ips1 = 1;
} else if (dc->config.disable_ips == DMUB_IPS_DISABLE_IPS2) {
+ // IPSv1.0 only
new_signals.bits.allow_pg = 1;
new_signals.bits.allow_ips1 = 1;
} else if (dc->config.disable_ips == DMUB_IPS_DISABLE_IPS2_Z10) {
+ // IPSv1.0 only
new_signals.bits.allow_pg = 1;
new_signals.bits.allow_ips1 = 1;
new_signals.bits.allow_ips2 = 1;
@@ -1319,6 +1290,8 @@ static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle)
new_signals.bits.allow_ips1 = 1;
new_signals.bits.allow_ips2 = 1;
new_signals.bits.allow_z10 = 1;
+ // New in IPSv2.0
+ new_signals.bits.allow_ips1z8 = 1;
} else {
/* RCG only */
new_signals.bits.allow_pg = 0;
@@ -1326,8 +1299,28 @@ static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle)
new_signals.bits.allow_ips2 = 0;
new_signals.bits.allow_z10 = 0;
}
+ } else if (dc->config.disable_ips == DMUB_IPS_DISABLE_Z8_RETENTION) {
+ new_signals.bits.allow_pg = 1;
+ new_signals.bits.allow_ips1 = 1;
+ new_signals.bits.allow_ips2 = 1;
+ new_signals.bits.allow_z10 = 1;
+ }
+ // Setting RCG allow bits (IPSv2.0)
+ if (dc->config.disable_ips_rcg == DMUB_IPS_RCG_ENABLE) {
+ new_signals.bits.allow_ips0_rcg = 1;
+ new_signals.bits.allow_ips1_rcg = 1;
+ } else if (dc->config.disable_ips_rcg == DMUB_IPS0_RCG_DISABLE) {
+ new_signals.bits.allow_ips1_rcg = 1;
+ } else if (dc->config.disable_ips_rcg == DMUB_IPS1_RCG_DISABLE) {
+ new_signals.bits.allow_ips0_rcg = 1;
+ }
+ // IPS dynamic allow bits (IPSv2 change, vpb use case)
+ if (dc->config.disable_ips_in_vpb == DMUB_IPS_VPB_ENABLE_IPS1_AND_RCG) {
+ new_signals.bits.allow_dynamic_ips1 = 1;
+ } else if (dc->config.disable_ips_in_vpb == DMUB_IPS_VPB_ENABLE_ALL) {
+ new_signals.bits.allow_dynamic_ips1 = 1;
+ new_signals.bits.allow_dynamic_ips1_z8 = 1;
}
-
ips_driver->signals = new_signals;
dc_dmub_srv->driver_signals = ips_driver->signals;
}
@@ -1351,7 +1344,7 @@ static void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle)
static void dc_dmub_srv_exit_low_power_state(const struct dc *dc)
{
struct dc_dmub_srv *dc_dmub_srv;
- uint32_t rcg_exit_count = 0, ips1_exit_count = 0, ips2_exit_count = 0;
+ uint32_t rcg_exit_count = 0, ips1_exit_count = 0, ips2_exit_count = 0, ips1z8_exit_count = 0;
if (dc->debug.dmcub_emulation)
return;
@@ -1371,45 +1364,49 @@ static void dc_dmub_srv_exit_low_power_state(const struct dc *dc)
rcg_exit_count = ips_fw->rcg_exit_count;
ips1_exit_count = ips_fw->ips1_exit_count;
ips2_exit_count = ips_fw->ips2_exit_count;
+ ips1z8_exit_count = ips_fw->ips1_z8ret_exit_count;
ips_driver->signals.all = 0;
dc_dmub_srv->driver_signals = ips_driver->signals;
DC_LOG_IPS(
- "%s (allow ips1=%u ips2=%u) (commit ips1=%u ips2=%u) (count rcg=%u ips1=%u ips2=%u)",
+ "%s (allow ips1=%u ips2=%u) (commit ips1=%u ips2=%u ips1z8=%u) (count rcg=%u ips1=%u ips2=%u ips1_z8=%u)",
__func__,
ips_driver->signals.bits.allow_ips1,
ips_driver->signals.bits.allow_ips2,
ips_fw->signals.bits.ips1_commit,
ips_fw->signals.bits.ips2_commit,
+ ips_fw->signals.bits.ips1z8_commit,
ips_fw->rcg_entry_count,
ips_fw->ips1_entry_count,
- ips_fw->ips2_entry_count);
+ ips_fw->ips2_entry_count,
+ ips_fw->ips1_z8ret_entry_count);
/* Note: register access has technically not resumed for DCN here, but we
* need to be message PMFW through our standard register interface.
*/
dc_dmub_srv->needs_idle_wake = false;
- if ((prev_driver_signals.bits.allow_ips2 || prev_driver_signals.all == 0) &&
+ if (!dc->caps.ips_v2_support && ((prev_driver_signals.bits.allow_ips2 || prev_driver_signals.all == 0) &&
(!dc->debug.optimize_ips_handshake ||
- ips_fw->signals.bits.ips2_commit || !ips_fw->signals.bits.in_idle)) {
+ ips_fw->signals.bits.ips2_commit || !ips_fw->signals.bits.in_idle))) {
DC_LOG_IPS(
- "wait IPS2 eval (ips1_commit=%u ips2_commit=%u)",
+ "wait IPS2 eval (ips1_commit=%u ips2_commit=%u )",
ips_fw->signals.bits.ips1_commit,
ips_fw->signals.bits.ips2_commit);
if (!dc->debug.optimize_ips_handshake || !ips_fw->signals.bits.ips2_commit)
udelay(dc->debug.ips2_eval_delay_us);
- if (ips_fw->signals.bits.ips2_commit) {
- DC_LOG_IPS(
- "exit IPS2 #1 (ips1_commit=%u ips2_commit=%u)",
- ips_fw->signals.bits.ips1_commit,
- ips_fw->signals.bits.ips2_commit);
+ DC_LOG_IPS(
+ "exit IPS2 #1 (ips1_commit=%u ips2_commit=%u)",
+ ips_fw->signals.bits.ips1_commit,
+ ips_fw->signals.bits.ips2_commit);
- // Tell PMFW to exit low power state
- dc->clk_mgr->funcs->exit_low_power_state(dc->clk_mgr);
+ // Tell PMFW to exit low power state
+ dc->clk_mgr->funcs->exit_low_power_state(dc->clk_mgr);
+
+ if (ips_fw->signals.bits.ips2_commit) {
DC_LOG_IPS(
"wait IPS2 entry delay (ips1_commit=%u ips2_commit=%u)",
@@ -1447,35 +1444,38 @@ static void dc_dmub_srv_exit_low_power_state(const struct dc *dc)
ips_fw->signals.bits.ips1_commit,
ips_fw->signals.bits.ips2_commit);
- dmub_srv_sync_inbox1(dc->ctx->dmub_srv->dmub);
+ dmub_srv_sync_inboxes(dc->ctx->dmub_srv->dmub);
}
}
dc_dmub_srv_notify_idle(dc, false);
if (prev_driver_signals.bits.allow_ips1 || prev_driver_signals.all == 0) {
DC_LOG_IPS(
- "wait for IPS1 commit clear (ips1_commit=%u ips2_commit=%u)",
+ "wait for IPS1 commit clear (ips1_commit=%u ips2_commit=%u ips1z8=%u)",
ips_fw->signals.bits.ips1_commit,
- ips_fw->signals.bits.ips2_commit);
+ ips_fw->signals.bits.ips2_commit,
+ ips_fw->signals.bits.ips1z8_commit);
while (ips_fw->signals.bits.ips1_commit)
udelay(1);
DC_LOG_IPS(
- "wait for IPS1 commit clear done (ips1_commit=%u ips2_commit=%u)",
+ "wait for IPS1 commit clear done (ips1_commit=%u ips2_commit=%u ips1z8=%u)",
ips_fw->signals.bits.ips1_commit,
- ips_fw->signals.bits.ips2_commit);
+ ips_fw->signals.bits.ips2_commit,
+ ips_fw->signals.bits.ips1z8_commit);
}
}
if (!dc_dmub_srv_is_hw_pwr_up(dc->ctx->dmub_srv, true))
ASSERT(0);
- DC_LOG_IPS("%s exit (count rcg=%u ips1=%u ips2=%u)",
+ DC_LOG_IPS("%s exit (count rcg=%u ips1=%u ips2=%u ips1z8=%u)",
__func__,
rcg_exit_count,
ips1_exit_count,
- ips2_exit_count);
+ ips2_exit_count,
+ ips1z8_exit_count);
}
void dc_dmub_srv_set_power_state(struct dc_dmub_srv *dc_dmub_srv, enum dc_acpi_cm_power_state power_state)
@@ -1688,20 +1688,21 @@ bool dc_wake_and_execute_gpint(const struct dc_context *ctx, enum dmub_gpint_com
return result;
}
-void dc_dmub_srv_fams2_update_config(struct dc *dc,
+static void dc_dmub_srv_rb_based_fams2_update_config(struct dc *dc,
struct dc_state *context,
bool enable)
{
uint8_t num_cmds = 1;
uint32_t i;
- union dmub_rb_cmd cmd[MAX_STREAMS + 1];
+ union dmub_rb_cmd cmd[2 * MAX_STREAMS + 1];
struct dmub_rb_cmd_fams2 *global_cmd = &cmd[0].fams2_config;
- memset(cmd, 0, sizeof(union dmub_rb_cmd) * (MAX_STREAMS + 1));
+ memset(cmd, 0, sizeof(union dmub_rb_cmd) * (2 * MAX_STREAMS + 1));
/* fill in generic command header */
global_cmd->header.type = DMUB_CMD__FW_ASSISTED_MCLK_SWITCH;
global_cmd->header.sub_type = DMUB_CMD__FAMS2_CONFIG;
- global_cmd->header.payload_bytes = sizeof(struct dmub_rb_cmd_fams2) - sizeof(struct dmub_cmd_header);
+ global_cmd->header.payload_bytes =
+ sizeof(struct dmub_rb_cmd_fams2) - sizeof(struct dmub_cmd_header);
if (enable) {
/* send global configuration parameters */
@@ -1714,17 +1715,28 @@ void dc_dmub_srv_fams2_update_config(struct dc *dc,
/* construct per-stream configs */
for (i = 0; i < context->bw_ctx.bw.dcn.fams2_global_config.num_streams; i++) {
- struct dmub_rb_cmd_fams2 *stream_cmd = &cmd[i+1].fams2_config;
+ struct dmub_rb_cmd_fams2 *stream_base_cmd = &cmd[i+1].fams2_config;
+ struct dmub_rb_cmd_fams2 *stream_sub_state_cmd = &cmd[i+1+context->bw_ctx.bw.dcn.fams2_global_config.num_streams].fams2_config;
/* configure command header */
- stream_cmd->header.type = DMUB_CMD__FW_ASSISTED_MCLK_SWITCH;
- stream_cmd->header.sub_type = DMUB_CMD__FAMS2_CONFIG;
- stream_cmd->header.payload_bytes = sizeof(struct dmub_rb_cmd_fams2) - sizeof(struct dmub_cmd_header);
- stream_cmd->header.multi_cmd_pending = 1;
- /* copy stream static state */
- memcpy(&stream_cmd->config.stream,
- &context->bw_ctx.bw.dcn.fams2_stream_params[i],
- sizeof(struct dmub_fams2_stream_static_state));
+ stream_base_cmd->header.type = DMUB_CMD__FW_ASSISTED_MCLK_SWITCH;
+ stream_base_cmd->header.sub_type = DMUB_CMD__FAMS2_CONFIG;
+ stream_base_cmd->header.payload_bytes =
+ sizeof(struct dmub_rb_cmd_fams2) - sizeof(struct dmub_cmd_header);
+ stream_base_cmd->header.multi_cmd_pending = 1;
+ stream_sub_state_cmd->header.type = DMUB_CMD__FW_ASSISTED_MCLK_SWITCH;
+ stream_sub_state_cmd->header.sub_type = DMUB_CMD__FAMS2_CONFIG;
+ stream_sub_state_cmd->header.payload_bytes =
+ sizeof(struct dmub_rb_cmd_fams2) - sizeof(struct dmub_cmd_header);
+ stream_sub_state_cmd->header.multi_cmd_pending = 1;
+ /* copy stream static base state */
+ memcpy(&stream_base_cmd->config,
+ &context->bw_ctx.bw.dcn.fams2_stream_base_params[i],
+ sizeof(union dmub_cmd_fams2_config));
+ /* copy stream static sub state */
+ memcpy(&stream_sub_state_cmd->config,
+ &context->bw_ctx.bw.dcn.fams2_stream_sub_params[i],
+ sizeof(union dmub_cmd_fams2_config));
}
}
@@ -1735,13 +1747,70 @@ void dc_dmub_srv_fams2_update_config(struct dc *dc,
if (enable && context->bw_ctx.bw.dcn.fams2_global_config.features.bits.enable) {
/* set multi pending for global, and unset for last stream cmd */
global_cmd->header.multi_cmd_pending = 1;
- cmd[context->bw_ctx.bw.dcn.fams2_global_config.num_streams].fams2_config.header.multi_cmd_pending = 0;
- num_cmds += context->bw_ctx.bw.dcn.fams2_global_config.num_streams;
+ cmd[2 * context->bw_ctx.bw.dcn.fams2_global_config.num_streams].fams2_config.header.multi_cmd_pending = 0;
+ num_cmds += 2 * context->bw_ctx.bw.dcn.fams2_global_config.num_streams;
}
dm_execute_dmub_cmd_list(dc->ctx, num_cmds, cmd, DM_DMUB_WAIT_TYPE_WAIT);
}
+static void dc_dmub_srv_ib_based_fams2_update_config(struct dc *dc,
+ struct dc_state *context,
+ bool enable)
+{
+ struct dmub_fams2_config_v2 *config = (struct dmub_fams2_config_v2 *)dc->ctx->dmub_srv->dmub->ib_mem_gart.cpu_addr;
+ union dmub_rb_cmd cmd;
+ uint32_t i;
+
+ memset(config, 0, sizeof(*config));
+ memset(&cmd, 0, sizeof(cmd));
+
+ cmd.ib_fams2_config.header.type = DMUB_CMD__FW_ASSISTED_MCLK_SWITCH;
+ cmd.ib_fams2_config.header.sub_type = DMUB_CMD__FAMS2_IB_CONFIG;
+
+ cmd.ib_fams2_config.ib_data.src.quad_part = dc->ctx->dmub_srv->dmub->ib_mem_gart.gpu_addr;
+ cmd.ib_fams2_config.ib_data.size = sizeof(*config);
+
+ if (enable && context->bw_ctx.bw.dcn.fams2_global_config.features.bits.enable) {
+ /* copy static feature configuration overrides */
+ config->global.features.bits.enable_stall_recovery = dc->debug.fams2_config.bits.enable_stall_recovery;
+ config->global.features.bits.enable_offload_flip = dc->debug.fams2_config.bits.enable_offload_flip;
+ config->global.features.bits.enable_debug = dc->debug.fams2_config.bits.enable_debug;
+
+ /* send global configuration parameters */
+ memcpy(&config->global, &context->bw_ctx.bw.dcn.fams2_global_config,
+ sizeof(struct dmub_cmd_fams2_global_config));
+
+ /* construct per-stream configs */
+ for (i = 0; i < context->bw_ctx.bw.dcn.fams2_global_config.num_streams; i++) {
+ /* copy stream static base state */
+ memcpy(&config->stream_v1[i].base,
+ &context->bw_ctx.bw.dcn.fams2_stream_base_params[i],
+ sizeof(config->stream_v1[i].base));
+
+ /* copy stream static sub-state */
+ memcpy(&config->stream_v1[i].sub_state,
+ &context->bw_ctx.bw.dcn.fams2_stream_sub_params_v2[i],
+ sizeof(config->stream_v1[i].sub_state));
+ }
+ }
+
+ config->global.features.bits.enable_visual_confirm = dc->debug.visual_confirm == VISUAL_CONFIRM_FAMS2;
+ config->global.features.bits.enable = enable;
+
+ dm_execute_dmub_cmd_list(dc->ctx, 1, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
+}
+
+void dc_dmub_srv_fams2_update_config(struct dc *dc,
+ struct dc_state *context,
+ bool enable)
+{
+ if (dc->debug.fams_version.major == 2)
+ dc_dmub_srv_rb_based_fams2_update_config(dc, context, enable);
+ if (dc->debug.fams_version.major == 3)
+ dc_dmub_srv_ib_based_fams2_update_config(dc, context, enable);
+}
+
void dc_dmub_srv_fams2_drr_update(struct dc *dc,
uint32_t tg_inst,
uint32_t vtotal_min,
@@ -1761,7 +1830,8 @@ void dc_dmub_srv_fams2_drr_update(struct dc *dc,
cmd.fams2_drr_update.dmub_optc_state_req.v_total_mid_frame_num = vtotal_mid_frame_num;
cmd.fams2_drr_update.dmub_optc_state_req.program_manual_trigger = program_manual_trigger;
- cmd.fams2_drr_update.header.payload_bytes = sizeof(cmd.fams2_drr_update) - sizeof(cmd.fams2_drr_update.header);
+ cmd.fams2_drr_update.header.payload_bytes =
+ sizeof(cmd.fams2_drr_update) - sizeof(cmd.fams2_drr_update.header);
dm_execute_dmub_cmd(dc->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
}
@@ -1797,7 +1867,8 @@ void dc_dmub_srv_fams2_passthrough_flip(
/* build command header */
cmds[num_cmds].fams2_flip.header.type = DMUB_CMD__FW_ASSISTED_MCLK_SWITCH;
cmds[num_cmds].fams2_flip.header.sub_type = DMUB_CMD__FAMS2_FLIP;
- cmds[num_cmds].fams2_flip.header.payload_bytes = sizeof(struct dmub_rb_cmd_fams2_flip);
+ cmds[num_cmds].fams2_flip.header.payload_bytes =
+ sizeof(struct dmub_rb_cmd_fams2_flip) - sizeof(struct dmub_cmd_header);
/* for chaining multiple commands, all but last command should set to 1 */
cmds[num_cmds].fams2_flip.header.multi_cmd_pending = 1;
@@ -1865,80 +1936,267 @@ void dc_dmub_srv_fams2_passthrough_flip(
}
}
-bool dc_dmub_srv_ips_residency_cntl(struct dc_dmub_srv *dc_dmub_srv, bool start_measurement)
+
+bool dc_dmub_srv_ips_residency_cntl(const struct dc_context *ctx, uint8_t panel_inst, bool start_measurement)
+{
+ union dmub_rb_cmd cmd;
+
+ memset(&cmd, 0, sizeof(cmd));
+
+ cmd.ips_residency_cntl.header.type = DMUB_CMD__IPS;
+ cmd.ips_residency_cntl.header.sub_type = DMUB_CMD__IPS_RESIDENCY_CNTL;
+ cmd.ips_residency_cntl.header.payload_bytes = sizeof(struct dmub_cmd_ips_residency_cntl_data);
+
+ // only panel_inst=0 is supported at the moment
+ cmd.ips_residency_cntl.cntl_data.panel_inst = panel_inst;
+ cmd.ips_residency_cntl.cntl_data.start_measurement = start_measurement;
+
+ if (!dc_wake_and_execute_dmub_cmd(ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY))
+ return false;
+
+ return true;
+}
+
+bool dc_dmub_srv_ips_query_residency_info(const struct dc_context *ctx, uint8_t panel_inst, struct dmub_ips_residency_info *driver_info,
+ enum ips_residency_mode ips_mode)
{
- bool result;
+ union dmub_rb_cmd cmd;
+ uint32_t bytes = sizeof(struct dmub_ips_residency_info);
- if (!dc_dmub_srv || !dc_dmub_srv->dmub)
+ dmub_flush_buffer_mem(&ctx->dmub_srv->dmub->scratch_mem_fb);
+ memset(&cmd, 0, sizeof(cmd));
+
+ cmd.ips_query_residency_info.header.type = DMUB_CMD__IPS;
+ cmd.ips_query_residency_info.header.sub_type = DMUB_CMD__IPS_QUERY_RESIDENCY_INFO;
+ cmd.ips_query_residency_info.header.payload_bytes = sizeof(struct dmub_cmd_ips_query_residency_info_data);
+
+ cmd.ips_query_residency_info.info_data.dest.quad_part = ctx->dmub_srv->dmub->scratch_mem_fb.gpu_addr;
+ cmd.ips_query_residency_info.info_data.size = bytes;
+ cmd.ips_query_residency_info.info_data.panel_inst = panel_inst;
+ cmd.ips_query_residency_info.info_data.ips_mode = (uint32_t)ips_mode;
+
+ if (!dc_wake_and_execute_dmub_cmd(ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY) ||
+ cmd.ips_query_residency_info.header.ret_status == 0)
return false;
- result = dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__IPS_RESIDENCY,
- start_measurement, NULL, DM_DMUB_WAIT_TYPE_WAIT);
+ // copy the result to the output since ret_status != 0 means the command returned data
+ memcpy(driver_info, ctx->dmub_srv->dmub->scratch_mem_fb.cpu_addr, bytes);
+
+ return true;
+}
+
+bool dmub_lsdma_init(struct dc_dmub_srv *dc_dmub_srv)
+{
+ struct dc_context *dc_ctx = dc_dmub_srv->ctx;
+ union dmub_rb_cmd cmd;
+ enum dm_dmub_wait_type wait_type;
+ struct dmub_cmd_lsdma_data *lsdma_data = &cmd.lsdma.lsdma_data;
+ bool result;
+
+ memset(&cmd, 0, sizeof(cmd));
+
+ cmd.cmd_common.header.type = DMUB_CMD__LSDMA;
+ cmd.cmd_common.header.sub_type = DMUB_CMD__LSDMA_INIT_CONFIG;
+ wait_type = DM_DMUB_WAIT_TYPE_NO_WAIT;
+
+ lsdma_data->u.init_data.gpu_addr_base.quad_part = dc_ctx->dmub_srv->dmub->lsdma_rb_fb.gpu_addr;
+ lsdma_data->u.init_data.ring_size = dc_ctx->dmub_srv->dmub->lsdma_rb_fb.size;
+
+ result = dc_wake_and_execute_dmub_cmd(dc_ctx, &cmd, wait_type);
+
+ if (!result)
+ DC_ERROR("LSDMA Init failed in DMUB");
return result;
}
-void dc_dmub_srv_ips_query_residency_info(struct dc_dmub_srv *dc_dmub_srv, struct ips_residency_info *output)
+bool dmub_lsdma_send_linear_copy_packet(
+ struct dc_dmub_srv *dc_dmub_srv,
+ uint64_t src_addr,
+ uint64_t dst_addr,
+ uint32_t count)
{
- uint32_t i;
- enum dmub_gpint_command command_code;
+ struct dc_context *dc_ctx = dc_dmub_srv->ctx;
+ union dmub_rb_cmd cmd;
+ enum dm_dmub_wait_type wait_type;
+ struct dmub_cmd_lsdma_data *lsdma_data = &cmd.lsdma.lsdma_data;
+ bool result;
- if (!dc_dmub_srv || !dc_dmub_srv->dmub)
- return;
+ memset(&cmd, 0, sizeof(cmd));
- switch (output->ips_mode) {
- case DMUB_IPS_MODE_IPS1_MAX:
- command_code = DMUB_GPINT__GET_IPS1_HISTOGRAM_COUNTER;
- break;
- case DMUB_IPS_MODE_IPS2:
- command_code = DMUB_GPINT__GET_IPS2_HISTOGRAM_COUNTER;
- break;
- case DMUB_IPS_MODE_IPS1_RCG:
- command_code = DMUB_GPINT__GET_IPS1_RCG_HISTOGRAM_COUNTER;
- break;
- case DMUB_IPS_MODE_IPS1_ONO2_ON:
- command_code = DMUB_GPINT__GET_IPS1_ONO2_ON_HISTOGRAM_COUNTER;
- break;
- default:
- command_code = DMUB_GPINT__INVALID_COMMAND;
- break;
- }
+ cmd.cmd_common.header.type = DMUB_CMD__LSDMA;
+ cmd.cmd_common.header.sub_type = DMUB_CMD__LSDMA_LINEAR_COPY;
+ wait_type = DM_DMUB_WAIT_TYPE_NO_WAIT;
+
+ lsdma_data->u.linear_copy_data.count = count - 1; // LSDMA controller expects bytes to copy -1
+ lsdma_data->u.linear_copy_data.src_lo = src_addr & 0xFFFFFFFF;
+ lsdma_data->u.linear_copy_data.src_hi = (src_addr >> 32) & 0xFFFFFFFF;
+ lsdma_data->u.linear_copy_data.dst_lo = dst_addr & 0xFFFFFFFF;
+ lsdma_data->u.linear_copy_data.dst_hi = (dst_addr >> 32) & 0xFFFFFFFF;
+
+ result = dc_wake_and_execute_dmub_cmd(dc_ctx, &cmd, wait_type);
+
+ if (!result)
+ DC_ERROR("LSDMA Linear Copy failed in DMUB");
+
+ return result;
+}
+
+bool dmub_lsdma_send_tiled_to_tiled_copy_command(
+ struct dc_dmub_srv *dc_dmub_srv,
+ struct lsdma_send_tiled_to_tiled_copy_command_params params)
+{
+ struct dc_context *dc_ctx = dc_dmub_srv->ctx;
+ union dmub_rb_cmd cmd;
+ enum dm_dmub_wait_type wait_type;
+ struct dmub_cmd_lsdma_data *lsdma_data = &cmd.lsdma.lsdma_data;
+ bool result;
+
+ memset(&cmd, 0, sizeof(cmd));
+
+ cmd.cmd_common.header.type = DMUB_CMD__LSDMA;
+ cmd.cmd_common.header.sub_type = DMUB_CMD__LSDMA_TILED_TO_TILED_COPY;
+ wait_type = DM_DMUB_WAIT_TYPE_NO_WAIT;
+
+ lsdma_data->u.tiled_copy_data.src_addr_lo = params.src_addr & 0xFFFFFFFF;
+ lsdma_data->u.tiled_copy_data.src_addr_hi = (params.src_addr >> 32) & 0xFFFFFFFF;
+ lsdma_data->u.tiled_copy_data.dst_addr_lo = params.dst_addr & 0xFFFFFFFF;
+ lsdma_data->u.tiled_copy_data.dst_addr_hi = (params.dst_addr >> 32) & 0xFFFFFFFF;
+ lsdma_data->u.tiled_copy_data.src_x = params.src_x;
+ lsdma_data->u.tiled_copy_data.src_y = params.src_y;
+ lsdma_data->u.tiled_copy_data.dst_x = params.dst_x;
+ lsdma_data->u.tiled_copy_data.dst_y = params.dst_y;
+ lsdma_data->u.tiled_copy_data.src_width = params.src_width - 1; // LSDMA controller expects width -1
+ lsdma_data->u.tiled_copy_data.dst_width = params.dst_width - 1; // LSDMA controller expects width -1
+ lsdma_data->u.tiled_copy_data.src_swizzle_mode = params.swizzle_mode;
+ lsdma_data->u.tiled_copy_data.dst_swizzle_mode = params.swizzle_mode;
+ lsdma_data->u.tiled_copy_data.src_element_size = params.element_size;
+ lsdma_data->u.tiled_copy_data.dst_element_size = params.element_size;
+ lsdma_data->u.tiled_copy_data.rect_x = params.rect_x;
+ lsdma_data->u.tiled_copy_data.rect_y = params.rect_y;
+ lsdma_data->u.tiled_copy_data.dcc = params.dcc;
+ lsdma_data->u.tiled_copy_data.tmz = params.tmz;
+ lsdma_data->u.tiled_copy_data.read_compress = params.read_compress;
+ lsdma_data->u.tiled_copy_data.write_compress = params.write_compress;
+ lsdma_data->u.tiled_copy_data.src_height = params.src_height - 1; // LSDMA controller expects height -1
+ lsdma_data->u.tiled_copy_data.dst_height = params.dst_height - 1; // LSDMA controller expects height -1
+ lsdma_data->u.tiled_copy_data.data_format = params.data_format;
+ lsdma_data->u.tiled_copy_data.max_com = params.max_com;
+ lsdma_data->u.tiled_copy_data.max_uncom = params.max_uncom;
+
+ result = dc_wake_and_execute_dmub_cmd(dc_ctx, &cmd, wait_type);
+
+ if (!result)
+ DC_ERROR("LSDMA Tiled to Tiled Copy failed in DMUB");
+
+ return result;
+}
+
+bool dmub_lsdma_send_pio_copy_command(
+ struct dc_dmub_srv *dc_dmub_srv,
+ uint64_t src_addr,
+ uint64_t dst_addr,
+ uint32_t byte_count,
+ uint32_t overlap_disable)
+{
+ struct dc_context *dc_ctx = dc_dmub_srv->ctx;
+ union dmub_rb_cmd cmd;
+ enum dm_dmub_wait_type wait_type;
+ struct dmub_cmd_lsdma_data *lsdma_data = &cmd.lsdma.lsdma_data;
+ bool result;
+
+ memset(&cmd, 0, sizeof(cmd));
+
+ cmd.cmd_common.header.type = DMUB_CMD__LSDMA;
+ cmd.cmd_common.header.sub_type = DMUB_CMD__LSDMA_PIO_COPY;
+ wait_type = DM_DMUB_WAIT_TYPE_NO_WAIT;
+
+ lsdma_data->u.pio_copy_data.packet.fields.byte_count = byte_count;
+ lsdma_data->u.pio_copy_data.packet.fields.overlap_disable = overlap_disable;
+ lsdma_data->u.pio_copy_data.src_lo = src_addr & 0xFFFFFFFF;
+ lsdma_data->u.pio_copy_data.src_hi = (src_addr >> 32) & 0xFFFFFFFF;
+ lsdma_data->u.pio_copy_data.dst_lo = dst_addr & 0xFFFFFFFF;
+ lsdma_data->u.pio_copy_data.dst_hi = (dst_addr >> 32) & 0xFFFFFFFF;
+
+ result = dc_wake_and_execute_dmub_cmd(dc_ctx, &cmd, wait_type);
+
+ if (!result)
+ DC_ERROR("LSDMA PIO Copy failed in DMUB");
+
+ return result;
+}
+
+bool dmub_lsdma_send_pio_constfill_command(
+ struct dc_dmub_srv *dc_dmub_srv,
+ uint64_t dst_addr,
+ uint32_t byte_count,
+ uint32_t data)
+{
+ struct dc_context *dc_ctx = dc_dmub_srv->ctx;
+ union dmub_rb_cmd cmd;
+ enum dm_dmub_wait_type wait_type;
+ struct dmub_cmd_lsdma_data *lsdma_data = &cmd.lsdma.lsdma_data;
+ bool result;
+
+ memset(&cmd, 0, sizeof(cmd));
+
+ cmd.cmd_common.header.type = DMUB_CMD__LSDMA;
+ cmd.cmd_common.header.sub_type = DMUB_CMD__LSDMA_PIO_CONSTFILL;
+ wait_type = DM_DMUB_WAIT_TYPE_NO_WAIT;
+
+ lsdma_data->u.pio_constfill_data.packet.fields.constant_fill = 1;
+ lsdma_data->u.pio_constfill_data.packet.fields.byte_count = byte_count;
+ lsdma_data->u.pio_constfill_data.dst_lo = dst_addr & 0xFFFFFFFF;
+ lsdma_data->u.pio_constfill_data.dst_hi = (dst_addr >> 32) & 0xFFFFFFFF;
+ lsdma_data->u.pio_constfill_data.data = data;
+
+ result = dc_wake_and_execute_dmub_cmd(dc_ctx, &cmd, wait_type);
+
+ if (!result)
+ DC_ERROR("LSDMA PIO Constfill failed in DMUB");
+
+ return result;
+}
+
+bool dmub_lsdma_send_poll_reg_write_command(struct dc_dmub_srv *dc_dmub_srv, uint32_t reg_addr, uint32_t reg_data)
+{
+ struct dc_context *dc_ctx = dc_dmub_srv->ctx;
+ union dmub_rb_cmd cmd;
+ enum dm_dmub_wait_type wait_type;
+ struct dmub_cmd_lsdma_data *lsdma_data = &cmd.lsdma.lsdma_data;
+ bool result;
+
+ memset(&cmd, 0, sizeof(cmd));
- if (command_code == DMUB_GPINT__INVALID_COMMAND)
+ cmd.cmd_common.header.type = DMUB_CMD__LSDMA;
+ cmd.cmd_common.header.sub_type = DMUB_CMD__LSDMA_POLL_REG_WRITE;
+ wait_type = DM_DMUB_WAIT_TYPE_NO_WAIT;
+
+ lsdma_data->u.reg_write_data.reg_addr = reg_addr;
+ lsdma_data->u.reg_write_data.reg_data = reg_data;
+
+ result = dc_wake_and_execute_dmub_cmd(dc_ctx, &cmd, wait_type);
+
+ if (!result)
+ DC_ERROR("LSDMA Poll Reg failed in DMUB");
+
+ return result;
+}
+
+void dc_dmub_srv_release_hw(const struct dc *dc)
+{
+ struct dc_dmub_srv *dc_dmub_srv = dc->ctx->dmub_srv;
+ union dmub_rb_cmd cmd = {0};
+
+ if (!dc_dmub_srv || !dc_dmub_srv->dmub)
return;
- // send gpint commands and wait for ack
- if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_RESIDENCY_PERCENT,
- (uint16_t)(output->ips_mode),
- &output->residency_percent, DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY))
- output->residency_percent = 0;
-
- if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_RESIDENCY_ENTRY_COUNTER,
- (uint16_t)(output->ips_mode),
- &output->entry_counter, DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY))
- output->entry_counter = 0;
-
- if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_RESIDENCY_DURATION_US_LO,
- (uint16_t)(output->ips_mode),
- &output->total_active_time_us[0], DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY))
- output->total_active_time_us[0] = 0;
- if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_RESIDENCY_DURATION_US_HI,
- (uint16_t)(output->ips_mode),
- &output->total_active_time_us[1], DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY))
- output->total_active_time_us[1] = 0;
-
- if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_INACTIVE_RESIDENCY_DURATION_US_LO,
- (uint16_t)(output->ips_mode),
- &output->total_inactive_time_us[0], DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY))
- output->total_inactive_time_us[0] = 0;
- if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, DMUB_GPINT__GET_IPS_INACTIVE_RESIDENCY_DURATION_US_HI,
- (uint16_t)(output->ips_mode),
- &output->total_inactive_time_us[1], DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY))
- output->total_inactive_time_us[1] = 0;
-
- // NUM_IPS_HISTOGRAM_BUCKETS = 16
- for (i = 0; i < 16; i++)
- if (!dc_wake_and_execute_gpint(dc_dmub_srv->ctx, command_code, i, &output->histogram[i],
- DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY))
- output->histogram[i] = 0;
+ memset(&cmd, 0, sizeof(cmd));
+ cmd.idle_opt_notify_idle.header.type = DMUB_CMD__IDLE_OPT;
+ cmd.idle_opt_notify_idle.header.sub_type = DMUB_CMD__IDLE_OPT_RELEASE_HW;
+ cmd.idle_opt_notify_idle.header.payload_bytes =
+ sizeof(cmd.idle_opt_notify_idle) -
+ sizeof(cmd.idle_opt_notify_idle.header);
+
+ dm_execute_dmub_cmd(dc->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
}
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h
index 10b48198b7a6..8ea320f21269 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.h
@@ -58,7 +58,7 @@ struct dc_dmub_srv {
bool needs_idle_wake;
};
-void dc_dmub_srv_wait_idle(struct dc_dmub_srv *dc_dmub_srv);
+bool dc_dmub_srv_wait_for_pending(struct dc_dmub_srv *dc_dmub_srv);
bool dc_dmub_srv_optimized_init_done(struct dc_dmub_srv *dc_dmub_srv);
@@ -94,7 +94,7 @@ void dc_dmub_srv_clear_inbox0_ack(struct dc_dmub_srv *dmub_srv);
void dc_dmub_srv_wait_for_inbox0_ack(struct dc_dmub_srv *dmub_srv);
void dc_dmub_srv_send_inbox0_cmd(struct dc_dmub_srv *dmub_srv, union dmub_inbox0_data_register data);
-bool dc_dmub_srv_get_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv, struct dmub_diagnostic_data *dmub_oca);
+bool dc_dmub_srv_get_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv);
void dc_dmub_setup_subvp_dmub_command(struct dc *dc, struct dc_state *context, bool enable);
void dc_dmub_srv_log_diagnostic_data(struct dc_dmub_srv *dc_dmub_srv);
@@ -210,6 +210,60 @@ void dc_dmub_srv_fams2_passthrough_flip(
struct dc_surface_update *srf_updates,
int surface_count);
+bool dmub_lsdma_init(struct dc_dmub_srv *dc_dmub_srv);
+bool dmub_lsdma_send_linear_copy_packet(
+ struct dc_dmub_srv *dc_dmub_srv,
+ uint64_t src_addr,
+ uint64_t dst_addr,
+ uint32_t count);
+bool dmub_lsdma_send_pio_copy_command(
+ struct dc_dmub_srv *dc_dmub_srv,
+ uint64_t src_addr,
+ uint64_t dst_addr,
+ uint32_t byte_count,
+ uint32_t overlap_disable);
+bool dmub_lsdma_send_pio_constfill_command(
+ struct dc_dmub_srv *dc_dmub_srv,
+ uint64_t dst_addr,
+ uint32_t byte_count,
+ uint32_t data);
+
+struct lsdma_send_tiled_to_tiled_copy_command_params {
+ uint64_t src_addr;
+ uint64_t dst_addr;
+
+ uint32_t src_x : 16;
+ uint32_t src_y : 16;
+
+ uint32_t dst_x : 16;
+ uint32_t dst_y : 16;
+
+ uint32_t src_width : 16;
+ uint32_t dst_width : 16;
+
+ uint32_t rect_x : 16;
+ uint32_t rect_y : 16;
+
+ uint32_t src_height : 16;
+ uint32_t dst_height : 16;
+
+ uint32_t data_format : 6;
+ uint32_t swizzle_mode : 5;
+ uint32_t element_size : 3;
+ uint32_t dcc : 1;
+ uint32_t tmz : 1;
+ uint32_t read_compress : 2;
+ uint32_t write_compress : 2;
+ uint32_t max_com : 2;
+ uint32_t max_uncom : 1;
+ uint32_t padding : 9;
+};
+
+bool dmub_lsdma_send_tiled_to_tiled_copy_command(
+ struct dc_dmub_srv *dc_dmub_srv,
+ struct lsdma_send_tiled_to_tiled_copy_command_params params);
+bool dmub_lsdma_send_poll_reg_write_command(struct dc_dmub_srv *dc_dmub_srv, uint32_t reg_addr, uint32_t reg_data);
+
/**
* struct ips_residency_info - struct containing info from dmub_ips_residency_stats
*
@@ -223,7 +277,7 @@ void dc_dmub_srv_fams2_passthrough_flip(
* @histogram: Histogram of given IPS state durations - bucket definitions in dmub_ips.c
*/
struct ips_residency_info {
- enum dmub_ips_mode ips_mode;
+ enum ips_residency_mode ips_mode;
unsigned int residency_percent;
unsigned int entry_counter;
unsigned int total_active_time_us[2];
@@ -231,21 +285,16 @@ struct ips_residency_info {
unsigned int histogram[16];
};
-/**
- * bool dc_dmub_srv_ips_residency_cntl() - Controls IPS residency measurement status
- *
- * @dc_dmub_srv: The DC DMUB service pointer
- * @start_measurement: Describes whether to start or stop measurement
- *
- * Return: true if GPINT was sent successfully, false otherwise
- */
-bool dc_dmub_srv_ips_residency_cntl(struct dc_dmub_srv *dc_dmub_srv, bool start_measurement);
+bool dc_dmub_srv_ips_residency_cntl(const struct dc_context *ctx, uint8_t panel_inst, bool start_measurement);
+
+bool dc_dmub_srv_ips_query_residency_info(const struct dc_context *ctx, uint8_t panel_inst,
+ struct dmub_ips_residency_info *driver_info,
+ enum ips_residency_mode ips_mode);
/**
- * bool dc_dmub_srv_ips_query_residency_info() - Queries DMCUB for residency info
+ * dc_dmub_srv_release_hw() - Notifies DMUB service that HW access is no longer required.
*
- * @dc_dmub_srv: The DC DMUB service pointer
- * @output: Output struct to copy the the residency info to
+ * @dc - pointer to DC object
*/
-void dc_dmub_srv_ips_query_residency_info(struct dc_dmub_srv *dc_dmub_srv, struct ips_residency_info *output);
+void dc_dmub_srv_release_hw(const struct dc *dc);
#endif /* _DMUB_DC_SRV_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
index 8dd6eb044829..5ce1be362534 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_dp_types.h
@@ -159,6 +159,16 @@ struct dc_link_settings {
uint8_t link_rate_set;
};
+struct dc_tunnel_settings {
+ bool should_enable_dp_tunneling;
+ bool should_use_dp_bw_allocation;
+ uint8_t cm_id;
+ uint8_t group_id;
+ uint32_t bw_granularity;
+ uint32_t estimated_bw;
+ uint32_t allocated_bw;
+};
+
union dc_dp_ffe_preset {
struct {
uint8_t level : 4;
@@ -300,6 +310,19 @@ union lane_align_status_updated {
uint8_t raw;
};
+union link_service_irq_vector_esi0 {
+ struct {
+ uint8_t DP_LINK_RX_CAP_CHANGED:1;
+ uint8_t DP_LINK_STATUS_CHANGED:1;
+ uint8_t DP_LINK_STREAM_STATUS_CHANGED:1;
+ uint8_t DP_LINK_HDMI_LINK_STATUS_CHANGED:1;
+ uint8_t DP_LINK_CONNECTED_OFF_ENTRY_REQUESTED:1;
+ uint8_t DP_LINK_TUNNELING_IRQ:1;
+ uint8_t reserved:2;
+ } bits;
+ uint8_t raw;
+};
+
union lane_adjust {
struct {
uint8_t VOLTAGE_SWING_LANE:2;
@@ -410,14 +433,6 @@ union dwnstream_port_caps_byte3_hdmi {
uint8_t raw;
};
-union hdmi_sink_encoded_link_bw_support {
- struct {
- uint8_t HDMI_SINK_ENCODED_LINK_BW_SUPPORT:3;
- uint8_t RESERVED:5;
- } bits;
- uint8_t raw;
-};
-
union hdmi_encoded_link_bw {
struct {
uint8_t FRL_MODE:1; // Bit 0
@@ -427,7 +442,28 @@ union hdmi_encoded_link_bw {
uint8_t BW_32Gbps:1;
uint8_t BW_40Gbps:1;
uint8_t BW_48Gbps:1;
- uint8_t RESERVED:1; // Bit 7
+ uint8_t FRL_LINK_TRAINING_FINISHED:1; // Bit 7
+ } bits;
+ uint8_t raw;
+};
+
+union hdmi_tx_link_status {
+ struct {
+ uint8_t HDMI_TX_LINK_ACTIVE_STATUS:1;
+ uint8_t HDMI_TX_READY_STATUS:1;
+ uint8_t RESERVED:6;
+ } bits;
+ uint8_t raw;
+};
+
+union autonomous_mode_and_frl_link_status {
+ struct {
+ uint8_t FRL_LT_IN_PROGRESS_STATUS:1;
+ uint8_t FRL_LT_LINK_CONFIG_IN_PROGRESS:3;
+ uint8_t RESERVED:1;
+ uint8_t FALLBACK_POLICY:1;
+ uint8_t FALLBACK_POLICY_VALID:1;
+ uint8_t REGULATED_AUTONOMOUS_MODE_SUPPORTED:1;
} bits;
uint8_t raw;
};
@@ -470,8 +506,10 @@ union sink_status {
uint8_t raw;
};
-/*6-byte structure corresponding to 6 registers (200h-205h)
-read during handling of HPD-IRQ*/
+/* 7-byte structure corresponding to 6 registers (200h-205h)
+ * and LINK_SERVICE_IRQ_ESI0 (2005h) for tunneling IRQ
+ * read during handling of HPD-IRQ
+ */
union hpd_irq_data {
struct {
union sink_count sink_cnt;/* 200h */
@@ -479,9 +517,10 @@ union hpd_irq_data {
union lane_status lane01_status;/* 202h */
union lane_status lane23_status;/* 203h */
union lane_align_status_updated lane_status_updated;/* 204h */
- union sink_status sink_status;
+ union sink_status sink_status;/* 205h */
+ union link_service_irq_vector_esi0 link_service_irq_esi0;/* 2005h */
} bytes;
- uint8_t raw[6];
+ uint8_t raw[7];
};
union down_stream_port_count {
@@ -914,10 +953,30 @@ union dpia_info {
uint8_t raw;
};
+/* DPCD[0xE0020] USB4_DRIVER_BW_CAPABILITY register. */
+union usb4_driver_bw_cap {
+ struct {
+ uint8_t rsvd :7;
+ uint8_t driver_bw_alloc_support :1;
+ } bits;
+ uint8_t raw;
+};
+
+/* DPCD[0xE0021] DP_IN_ADAPTER_TUNNEL_INFORMATION register. */
+union dpia_tunnel_info {
+ struct {
+ uint8_t group_id :3;
+ uint8_t rsvd :5;
+ } bits;
+ uint8_t raw;
+};
+
/* DP Tunneling over USB4 */
struct dpcd_usb4_dp_tunneling_info {
union dp_tun_cap_support dp_tun_cap;
union dpia_info dpia_info;
+ union usb4_driver_bw_cap driver_bw_cap;
+ union dpia_tunnel_info dpia_tunnel_info;
uint8_t usb4_driver_id;
uint8_t usb4_topology_id[DPCD_USB4_TOPOLOGY_ID_LEN];
};
@@ -959,6 +1018,14 @@ union dp_128b_132b_supported_lttpr_link_rates {
uint8_t raw;
};
+union dp_alpm_lttpr_cap {
+ struct {
+ uint8_t AUX_LESS_ALPM_SUPPORTED :1;
+ uint8_t RESERVED :7;
+ } bits;
+ uint8_t raw;
+};
+
union dp_sink_video_fallback_formats {
struct {
uint8_t dp_1024x768_60Hz_24bpp_support :1;
@@ -969,6 +1036,21 @@ union dp_sink_video_fallback_formats {
uint8_t raw;
};
+union dp_receive_port0_cap {
+ struct {
+ uint8_t RESERVED :1;
+ uint8_t LOCAL_EDID_PRESENT :1;
+ uint8_t ASSOCIATED_TO_PRECEDING_PORT:1;
+ uint8_t HBLANK_EXPANSION_CAPABLE :1;
+ uint8_t BUFFER_SIZE_UNIT :1;
+ uint8_t BUFFER_SIZE_PER_PORT :1;
+ uint8_t HBLANK_REDUCTION_CAPABLE :1;
+ uint8_t RESERVED2:1;
+ uint8_t BUFFER_SIZE:8;
+ } bits;
+ uint8_t raw[2];
+};
+
union dpcd_max_uncompressed_pixel_rate_cap {
struct {
uint16_t max_uncompressed_pixel_rate_cap :15;
@@ -1103,7 +1185,10 @@ struct dc_lttpr_caps {
uint8_t max_ext_timeout;
union dp_main_link_channel_coding_lttpr_cap main_link_channel_coding;
union dp_128b_132b_supported_lttpr_link_rates supported_128b_132b_rates;
+ union dp_alpm_lttpr_cap alpm;
uint8_t aux_rd_interval[MAX_REPEATER_CNT - 1];
+ uint8_t lttpr_ieee_oui[3]; // Always read from closest LTTPR to host
+ uint8_t lttpr_device_id[6]; // Always read from closest LTTPR to host
};
struct dc_dongle_dfp_cap_ext {
@@ -1132,6 +1217,7 @@ struct dc_dongle_caps {
uint32_t dp_hdmi_max_bpc;
uint32_t dp_hdmi_max_pixel_clk_in_khz;
uint32_t dp_hdmi_frl_max_link_bw_in_kbps;
+ uint32_t dp_hdmi_regulated_autonomous_mode_support;
struct dc_dongle_dfp_cap_ext dfp_cap_ext;
};
@@ -1193,6 +1279,9 @@ struct dpcd_caps {
struct replay_info pr_info;
uint16_t edp_oled_emission_rate;
+ union dp_receive_port0_cap receive_port0_cap;
+ /* Indicates the number of SST links supported by MSO (Multi-Stream Output) */
+ uint8_t mso_cap_sst_links_supported;
};
union dpcd_sink_ext_caps {
@@ -1354,6 +1443,12 @@ struct dp_trace {
#ifndef DPCD_MAX_UNCOMPRESSED_PIXEL_RATE_CAP
#define DPCD_MAX_UNCOMPRESSED_PIXEL_RATE_CAP 0x221c
#endif
+#ifndef DP_LTTPR_ALPM_CAPABILITIES
+#define DP_LTTPR_ALPM_CAPABILITIES 0xF0009
+#endif
+#ifndef DP_REGULATED_AUTONOMOUS_MODE_SUPPORTED_AND_HDMI_LINK_TRAINING_STATUS
+#define DP_REGULATED_AUTONOMOUS_MODE_SUPPORTED_AND_HDMI_LINK_TRAINING_STATUS 0x303C
+#endif
#ifndef DP_REPEATER_CONFIGURATION_AND_STATUS_SIZE
#define DP_REPEATER_CONFIGURATION_AND_STATUS_SIZE 0x50
#endif
@@ -1363,6 +1458,12 @@ struct dp_trace {
#ifndef DP_BRANCH_VENDOR_SPECIFIC_START
#define DP_BRANCH_VENDOR_SPECIFIC_START 0x50C
#endif
+#ifndef DP_LTTPR_IEEE_OUI
+#define DP_LTTPR_IEEE_OUI 0xF003D
+#endif
+#ifndef DP_LTTPR_DEVICE_ID
+#define DP_LTTPR_DEVICE_ID 0xF0040
+#endif
/** USB4 DPCD BW Allocation Registers Chapter 10.7 **/
#ifndef DP_TUNNELING_CAPABILITIES
#define DP_TUNNELING_CAPABILITIES 0xE000D /* 1.4a */
@@ -1400,4 +1501,26 @@ struct dp_trace {
#ifndef REQUESTED_BW
#define REQUESTED_BW 0xE0031 /* 1.4a */
#endif
+# ifndef DP_TUNNELING_BW_ALLOC_BITS_MASK
+# define DP_TUNNELING_BW_ALLOC_BITS_MASK (0x0F << 0)
+# endif
+# ifndef DP_TUNNELING_BW_REQUEST_FAILED
+# define DP_TUNNELING_BW_REQUEST_FAILED (1 << 0)
+# endif
+# ifndef DP_TUNNELING_BW_REQUEST_SUCCEEDED
+# define DP_TUNNELING_BW_REQUEST_SUCCEEDED (1 << 1)
+# endif
+# ifndef DP_TUNNELING_ESTIMATED_BW_CHANGED
+# define DP_TUNNELING_ESTIMATED_BW_CHANGED (1 << 2)
+# endif
+# ifndef DP_TUNNELING_BW_ALLOC_CAP_CHANGED
+# define DP_TUNNELING_BW_ALLOC_CAP_CHANGED (1 << 3)
+# endif
+# ifndef DPTX_BW_ALLOC_UNMASK_IRQ
+# define DPTX_BW_ALLOC_UNMASK_IRQ (1 << 6)
+# endif
+# ifndef DPTX_BW_ALLOC_MODE_ENABLE
+# define DPTX_BW_ALLOC_MODE_ENABLE (1 << 7)
+# endif
+
#endif /* DC_DP_TYPES_H */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dsc.h b/drivers/gpu/drm/amd/display/dc/dc_dsc.h
index 9014c2409817..9d18f1c08079 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dsc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_dsc.h
@@ -94,6 +94,11 @@ uint32_t dc_dsc_stream_bandwidth_overhead_in_kbps(
const int num_slices_h,
const bool is_dp);
+void dc_dsc_dump_decoder_caps(const struct display_stream_compressor *dsc,
+ const struct dsc_dec_dpcd_caps *dsc_sink_caps);
+void dc_dsc_dump_encoder_caps(const struct display_stream_compressor *dsc,
+ const struct dc_crtc_timing *timing);
+
/* TODO - Hardware/specs limitation should be owned by dc dsc and returned to DM,
* and DM can choose to OVERRIDE the limitation on CASE BY CASE basis.
* Hardware/specs limitation should not be writable by DM.
diff --git a/drivers/gpu/drm/amd/display/dc/dc_fused_io.c b/drivers/gpu/drm/amd/display/dc/dc_fused_io.c
new file mode 100644
index 000000000000..fee69642fb93
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dc_fused_io.c
@@ -0,0 +1,148 @@
+// SPDX-License-Identifier: MIT
+//
+// Copyright 2025 Advanced Micro Devices, Inc.
+
+#include "dc_fused_io.h"
+
+#include "dm_helpers.h"
+#include "gpio.h"
+
+static bool op_i2c_convert(
+ union dmub_rb_cmd *cmd,
+ const struct mod_hdcp_atomic_op_i2c *op,
+ enum dmub_cmd_fused_request_type type,
+ uint32_t ddc_line,
+ bool over_aux
+)
+{
+ struct dmub_cmd_fused_request *req = &cmd->fused_io.request;
+ struct dmub_cmd_fused_request_location_i2c *loc = &req->u.i2c;
+
+ if (!op || op->size > sizeof(req->buffer))
+ return false;
+
+ req->type = type;
+ loc->is_aux = false;
+ loc->ddc_line = ddc_line;
+ loc->over_aux = over_aux;
+ loc->address = op->address;
+ loc->offset = op->offset;
+ loc->length = op->size;
+ memcpy(req->buffer, op->data, op->size);
+
+ return true;
+}
+
+static bool op_aux_convert(
+ union dmub_rb_cmd *cmd,
+ const struct mod_hdcp_atomic_op_aux *op,
+ enum dmub_cmd_fused_request_type type,
+ uint32_t ddc_line
+)
+{
+ struct dmub_cmd_fused_request *req = &cmd->fused_io.request;
+ struct dmub_cmd_fused_request_location_aux *loc = &req->u.aux;
+
+ if (!op || op->size > sizeof(req->buffer))
+ return false;
+
+ req->type = type;
+ loc->is_aux = true;
+ loc->ddc_line = ddc_line;
+ loc->address = op->address;
+ loc->length = op->size;
+ memcpy(req->buffer, op->data, op->size);
+
+ return true;
+}
+
+static bool atomic_write_poll_read(
+ struct dc_link *link,
+ union dmub_rb_cmd commands[3],
+ uint32_t poll_timeout_us,
+ uint8_t poll_mask_msb
+)
+{
+ const uint8_t count = 3;
+ const uint32_t timeout_per_request_us = 10000;
+ const uint32_t timeout_per_aux_transaction_us = 10000;
+ uint64_t timeout_us = 0;
+
+ commands[1].fused_io.request.poll_mask_msb = poll_mask_msb;
+ commands[1].fused_io.request.timeout_us = poll_timeout_us;
+
+ for (uint8_t i = 0; i < count; i++) {
+ struct dmub_rb_cmd_fused_io *io = &commands[i].fused_io;
+
+ io->header.type = DMUB_CMD__FUSED_IO;
+ io->header.sub_type = DMUB_CMD__FUSED_IO_EXECUTE;
+ io->header.multi_cmd_pending = i != count - 1;
+ io->header.payload_bytes = sizeof(commands[i].fused_io) - sizeof(io->header);
+
+ timeout_us += timeout_per_request_us + io->request.timeout_us;
+ if (!io->request.timeout_us && io->request.u.aux.is_aux)
+ timeout_us += timeout_per_aux_transaction_us * (io->request.u.aux.length / 16);
+ }
+
+ if (!dm_helpers_execute_fused_io(link->ctx, link, commands, count, timeout_us))
+ return false;
+
+ return commands[0].fused_io.request.status == FUSED_REQUEST_STATUS_SUCCESS;
+}
+
+bool dm_atomic_write_poll_read_i2c(
+ struct dc_link *link,
+ const struct mod_hdcp_atomic_op_i2c *write,
+ const struct mod_hdcp_atomic_op_i2c *poll,
+ struct mod_hdcp_atomic_op_i2c *read,
+ uint32_t poll_timeout_us,
+ uint8_t poll_mask_msb
+)
+{
+ if (!link)
+ return false;
+
+ const bool over_aux = false;
+ const uint32_t ddc_line = link->ddc->ddc_pin->pin_data->en;
+
+ union dmub_rb_cmd commands[3] = { 0 };
+ const bool converted = op_i2c_convert(&commands[0], write, FUSED_REQUEST_WRITE, ddc_line, over_aux)
+ && op_i2c_convert(&commands[1], poll, FUSED_REQUEST_POLL, ddc_line, over_aux)
+ && op_i2c_convert(&commands[2], read, FUSED_REQUEST_READ, ddc_line, over_aux);
+
+ if (!converted)
+ return false;
+
+ const bool result = atomic_write_poll_read(link, commands, poll_timeout_us, poll_mask_msb);
+
+ memcpy(read->data, commands[0].fused_io.request.buffer, read->size);
+ return result;
+}
+
+bool dm_atomic_write_poll_read_aux(
+ struct dc_link *link,
+ const struct mod_hdcp_atomic_op_aux *write,
+ const struct mod_hdcp_atomic_op_aux *poll,
+ struct mod_hdcp_atomic_op_aux *read,
+ uint32_t poll_timeout_us,
+ uint8_t poll_mask_msb
+)
+{
+ if (!link)
+ return false;
+
+ const uint32_t ddc_line = link->ddc->ddc_pin->pin_data->en;
+ union dmub_rb_cmd commands[3] = { 0 };
+ const bool converted = op_aux_convert(&commands[0], write, FUSED_REQUEST_WRITE, ddc_line)
+ && op_aux_convert(&commands[1], poll, FUSED_REQUEST_POLL, ddc_line)
+ && op_aux_convert(&commands[2], read, FUSED_REQUEST_READ, ddc_line);
+
+ if (!converted)
+ return false;
+
+ const bool result = atomic_write_poll_read(link, commands, poll_timeout_us, poll_mask_msb);
+
+ memcpy(read->data, commands[0].fused_io.request.buffer, read->size);
+ return result;
+}
+
diff --git a/drivers/gpu/drm/amd/display/dc/dc_fused_io.h b/drivers/gpu/drm/amd/display/dc/dc_fused_io.h
new file mode 100644
index 000000000000..c74917240985
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dc_fused_io.h
@@ -0,0 +1,31 @@
+/* SPDX-License-Identifier: MIT */
+/*
+ * Copyright 2025 Advanced Micro Devices, Inc.
+ */
+
+#ifndef __DC_FUSED_IO_H__
+#define __DC_FUSED_IO_H__
+
+#include "dc.h"
+#include "mod_hdcp.h"
+
+bool dm_atomic_write_poll_read_i2c(
+ struct dc_link *link,
+ const struct mod_hdcp_atomic_op_i2c *write,
+ const struct mod_hdcp_atomic_op_i2c *poll,
+ struct mod_hdcp_atomic_op_i2c *read,
+ uint32_t poll_timeout_us,
+ uint8_t poll_mask_msb
+);
+
+bool dm_atomic_write_poll_read_aux(
+ struct dc_link *link,
+ const struct mod_hdcp_atomic_op_aux *write,
+ const struct mod_hdcp_atomic_op_aux *poll,
+ struct mod_hdcp_atomic_op_aux *read,
+ uint32_t poll_timeout_us,
+ uint8_t poll_mask_msb
+);
+
+#endif // __DC_FUSED_IO_H__
+
diff --git a/drivers/gpu/drm/amd/display/dc/dc_helper.c b/drivers/gpu/drm/amd/display/dc/dc_helper.c
index b402be59b2c8..7217de258851 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/dc_helper.c
@@ -682,13 +682,19 @@ void reg_sequence_wait_done(const struct dc_context *ctx)
if (offload &&
ctx->dc->debug.dmub_offload_enabled &&
!ctx->dc->debug.dmcub_emulation) {
- dc_dmub_srv_wait_idle(ctx->dmub_srv);
+ dc_dmub_srv_wait_for_idle(ctx->dmub_srv, DM_DMUB_WAIT_TYPE_WAIT, NULL);
}
}
char *dce_version_to_string(const int version)
{
switch (version) {
+ case DCE_VERSION_6_0:
+ return "DCE 6.0";
+ case DCE_VERSION_6_1:
+ return "DCE 6.1";
+ case DCE_VERSION_6_4:
+ return "DCE 6.4";
case DCE_VERSION_8_0:
return "DCE 8.0";
case DCE_VERSION_8_1:
@@ -741,6 +747,8 @@ char *dce_version_to_string(const int version)
return "DCN 3.5";
case DCN_VERSION_3_51:
return "DCN 3.5.1";
+ case DCN_VERSION_3_6:
+ return "DCN 3.6";
case DCN_VERSION_4_01:
return "DCN 4.0.1";
default:
diff --git a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
index c10567ec1c81..667852517246 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_hw_types.h
@@ -68,7 +68,7 @@ enum dc_plane_addr_type {
struct dc_plane_address {
enum dc_plane_addr_type type;
- bool tmz_surface;
+ uint8_t tmz_surface;
union {
struct{
PHYSICAL_ADDRESS_LOC addr;
@@ -341,89 +341,101 @@ enum swizzle_mode_addr3_values {
DC_ADDR3_SW_UNKNOWN = DC_ADDR3_SW_MAX
};
-union dc_tiling_info {
+enum dc_gfxversion {
+ DcGfxVersion7 = 0,
+ DcGfxVersion8,
+ DcGfxVersion9,
+ DcGfxVersion10,
+ DcGfxVersion11,
+ DcGfxAddr3,
+ DcGfxVersionUnknown
+};
- struct {
- /* Specifies the number of memory banks for tiling
- * purposes.
- * Only applies to 2D and 3D tiling modes.
- * POSSIBLE VALUES: 2,4,8,16
- */
- unsigned int num_banks;
- /* Specifies the number of tiles in the x direction
- * to be incorporated into the same bank.
- * Only applies to 2D and 3D tiling modes.
- * POSSIBLE VALUES: 1,2,4,8
- */
- unsigned int bank_width;
- unsigned int bank_width_c;
- /* Specifies the number of tiles in the y direction to
- * be incorporated into the same bank.
- * Only applies to 2D and 3D tiling modes.
- * POSSIBLE VALUES: 1,2,4,8
- */
- unsigned int bank_height;
- unsigned int bank_height_c;
- /* Specifies the macro tile aspect ratio. Only applies
- * to 2D and 3D tiling modes.
- */
- unsigned int tile_aspect;
- unsigned int tile_aspect_c;
- /* Specifies the number of bytes that will be stored
- * contiguously for each tile.
- * If the tile data requires more storage than this
- * amount, it is split into multiple slices.
- * This field must not be larger than
- * GB_ADDR_CONFIG.DRAM_ROW_SIZE.
- * Only applies to 2D and 3D tiling modes.
- * For color render targets, TILE_SPLIT >= 256B.
- */
- enum tile_split_values tile_split;
- enum tile_split_values tile_split_c;
- /* Specifies the addressing within a tile.
- * 0x0 - DISPLAY_MICRO_TILING
- * 0x1 - THIN_MICRO_TILING
- * 0x2 - DEPTH_MICRO_TILING
- * 0x3 - ROTATED_MICRO_TILING
- */
- enum tile_mode_values tile_mode;
- enum tile_mode_values tile_mode_c;
- /* Specifies the number of pipes and how they are
- * interleaved in the surface.
- * Refer to memory addressing document for complete
- * details and constraints.
- */
- unsigned int pipe_config;
- /* Specifies the tiling mode of the surface.
- * THIN tiles use an 8x8x1 tile size.
- * THICK tiles use an 8x8x4 tile size.
- * 2D tiling modes rotate banks for successive Z slices
- * 3D tiling modes rotate pipes and banks for Z slices
- * Refer to memory addressing document for complete
- * details and constraints.
- */
- enum array_mode_values array_mode;
- } gfx8;
+ struct dc_tiling_info {
+ unsigned int gfxversion; // Specifies which part of the union to use. Must use DalGfxVersion enum
+ union {
+ struct {
+ /* Specifies the number of memory banks for tiling
+ * purposes.
+ * Only applies to 2D and 3D tiling modes.
+ * POSSIBLE VALUES: 2,4,8,16
+ */
+ unsigned int num_banks;
+ /* Specifies the number of tiles in the x direction
+ * to be incorporated into the same bank.
+ * Only applies to 2D and 3D tiling modes.
+ * POSSIBLE VALUES: 1,2,4,8
+ */
+ unsigned int bank_width;
+ unsigned int bank_width_c;
+ /* Specifies the number of tiles in the y direction to
+ * be incorporated into the same bank.
+ * Only applies to 2D and 3D tiling modes.
+ * POSSIBLE VALUES: 1,2,4,8
+ */
+ unsigned int bank_height;
+ unsigned int bank_height_c;
+ /* Specifies the macro tile aspect ratio. Only applies
+ * to 2D and 3D tiling modes.
+ */
+ unsigned int tile_aspect;
+ unsigned int tile_aspect_c;
+ /* Specifies the number of bytes that will be stored
+ * contiguously for each tile.
+ * If the tile data requires more storage than this
+ * amount, it is split into multiple slices.
+ * This field must not be larger than
+ * GB_ADDR_CONFIG.DRAM_ROW_SIZE.
+ * Only applies to 2D and 3D tiling modes.
+ * For color render targets, TILE_SPLIT >= 256B.
+ */
+ enum tile_split_values tile_split;
+ enum tile_split_values tile_split_c;
+ /* Specifies the addressing within a tile.
+ * 0x0 - DISPLAY_MICRO_TILING
+ * 0x1 - THIN_MICRO_TILING
+ * 0x2 - DEPTH_MICRO_TILING
+ * 0x3 - ROTATED_MICRO_TILING
+ */
+ enum tile_mode_values tile_mode;
+ enum tile_mode_values tile_mode_c;
+ /* Specifies the number of pipes and how they are
+ * interleaved in the surface.
+ * Refer to memory addressing document for complete
+ * details and constraints.
+ */
+ unsigned int pipe_config;
+ /* Specifies the tiling mode of the surface.
+ * THIN tiles use an 8x8x1 tile size.
+ * THICK tiles use an 8x8x4 tile size.
+ * 2D tiling modes rotate banks for successive Z slices
+ * 3D tiling modes rotate pipes and banks for Z slices
+ * Refer to memory addressing document for complete
+ * details and constraints.
+ */
+ enum array_mode_values array_mode;
+ } gfx8;
- struct {
- enum swizzle_mode_values swizzle;
- unsigned int num_pipes;
- unsigned int max_compressed_frags;
- unsigned int pipe_interleave;
-
- unsigned int num_banks;
- unsigned int num_shader_engines;
- unsigned int num_rb_per_se;
- bool shaderEnable;
-
- bool meta_linear;
- bool rb_aligned;
- bool pipe_aligned;
- unsigned int num_pkrs;
- } gfx9;/*gfx9, gfx10 and above*/
- struct {
- enum swizzle_mode_addr3_values swizzle;
- } gfx_addr3;/*gfx with addr3 and above*/
+ struct {
+ enum swizzle_mode_values swizzle;
+ unsigned int num_pipes;
+ unsigned int max_compressed_frags;
+ unsigned int pipe_interleave;
+
+ unsigned int num_banks;
+ unsigned int num_shader_engines;
+ unsigned int num_rb_per_se;
+ bool shaderEnable;
+
+ bool meta_linear;
+ bool rb_aligned;
+ bool pipe_aligned;
+ unsigned int num_pkrs;
+ } gfx9;/*gfx9, gfx10 and above*/
+ struct {
+ enum swizzle_mode_addr3_values swizzle;
+ } gfx_addr3;/*gfx with addr3 and above*/
+ };
};
/* Rotation angle */
@@ -641,7 +653,8 @@ enum dc_color_space {
COLOR_SPACE_YCBCR709_LIMITED,
COLOR_SPACE_2020_RGB_FULLRANGE,
COLOR_SPACE_2020_RGB_LIMITEDRANGE,
- COLOR_SPACE_2020_YCBCR,
+ COLOR_SPACE_2020_YCBCR_LIMITED,
+ COLOR_SPACE_2020_YCBCR_FULL,
COLOR_SPACE_ADOBERGB,
COLOR_SPACE_DCIP3,
COLOR_SPACE_DISPLAYNATIVE,
@@ -649,6 +662,7 @@ enum dc_color_space {
COLOR_SPACE_APPCTRL,
COLOR_SPACE_CUSTOMPOINTS,
COLOR_SPACE_YCBCR709_BLACK,
+ COLOR_SPACE_2020_YCBCR = COLOR_SPACE_2020_YCBCR_LIMITED,
};
enum dc_dither_option {
@@ -960,6 +974,7 @@ struct dc_crtc_timing {
uint32_t pix_clk_100hz;
uint32_t min_refresh_in_uhz;
+ uint32_t max_refresh_in_uhz;
uint32_t vic;
uint32_t hdmi_vic;
@@ -975,6 +990,9 @@ struct dc_crtc_timing {
struct dc_crtc_timing_flags flags;
uint32_t dsc_fixed_bits_per_pixel_x16; /* DSC target bitrate in 1/16 of bpp (e.g. 128 -> 8bpp) */
struct dc_dsc_config dsc_cfg;
+
+ /* The number of pixels that HBlank has been expanded by from the original EDID timing. */
+ uint32_t expanded_hblank;
};
enum trigger_delay {
@@ -1000,6 +1018,7 @@ struct dc_crtc_timing_adjust {
uint32_t v_total_mid;
uint32_t v_total_mid_frame_num;
uint32_t allow_otg_v_count_halt;
+ uint8_t timing_adjust_pending;
};
@@ -1085,7 +1104,8 @@ enum mpcc_gamut_remap_mode_select {
enum mpcc_gamut_remap_id {
MPCC_OGAM_GAMUT_REMAP,
MPCC_MCM_FIRST_GAMUT_REMAP,
- MPCC_MCM_SECOND_GAMUT_REMAP
+ MPCC_MCM_SECOND_GAMUT_REMAP,
+ MPCC_RMCM_GAMUT_REMAP,
};
enum cursor_matrix_mode {
diff --git a/drivers/gpu/drm/amd/display/dc/dc_plane.h b/drivers/gpu/drm/amd/display/dc/dc_plane.h
index bd37ec82b42d..14feb843e694 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_plane.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_plane.h
@@ -28,10 +28,24 @@
#include "dc_hw_types.h"
+union dc_plane_status_update_flags {
+ struct {
+ uint32_t address : 1;
+ } bits;
+ uint32_t raw;
+};
+
struct dc_plane_state *dc_create_plane_state(const struct dc *dc);
const struct dc_plane_status *dc_plane_get_status(
- const struct dc_plane_state *plane_state);
+ const struct dc_plane_state *plane_state,
+ union dc_plane_status_update_flags flags);
void dc_plane_state_retain(struct dc_plane_state *plane_state);
void dc_plane_state_release(struct dc_plane_state *plane_state);
+void dc_plane_force_dcc_and_tiling_disable(struct dc_plane_state *plane_state,
+ bool clear_tiling);
+
+
+void dc_plane_copy_config(struct dc_plane_state *dst, const struct dc_plane_state *src);
+
#endif /* _DC_PLANE_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c b/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c
index c8d8e335fa37..7f57661433eb 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c
+++ b/drivers/gpu/drm/amd/display/dc/dc_spl_translate.c
@@ -3,7 +3,6 @@
// Copyright 2024 Advanced Micro Devices, Inc.
#include "dc_spl_translate.h"
-#include "spl/dc_spl_types.h"
#include "dcn20/dcn20_dpp.h"
#include "dcn32/dcn32_dpp.h"
#include "dcn401/dcn401_dpp.h"
@@ -64,6 +63,13 @@ static void populate_inits_from_splinits(struct scl_inits *inits,
inits->h_c = dc_fixpt_from_int_dy(spl_inits->h_filter_init_int_c, spl_inits->h_filter_init_frac_c >> 5, 0, 19);
inits->v_c = dc_fixpt_from_int_dy(spl_inits->v_filter_init_int_c, spl_inits->v_filter_init_frac_c >> 5, 0, 19);
}
+static void populate_splformat_from_format(enum spl_pixel_format *spl_pixel_format, const enum pixel_format pixel_format)
+{
+ if (pixel_format < PIXEL_FORMAT_INVALID)
+ *spl_pixel_format = (enum spl_pixel_format)pixel_format;
+ else
+ *spl_pixel_format = SPL_PIXEL_FORMAT_INVALID;
+}
/// @brief Translate SPL input parameters from pipe context
/// @param pipe_ctx
/// @param spl_in
@@ -89,7 +95,7 @@ void translate_SPL_in_params_from_pipe_ctx(struct pipe_ctx *pipe_ctx, struct spl
spl_in->callbacks = dcn2_spl_callbacks;
}
// Make format field from spl_in point to plane_res scl_data format
- spl_in->basic_in.format = (enum spl_pixel_format)pipe_ctx->plane_res.scl_data.format;
+ populate_splformat_from_format(&spl_in->basic_in.format, pipe_ctx->plane_res.scl_data.format);
// Make view_format from basic_out point to view_format from stream
spl_in->basic_out.view_format = (enum spl_view_3d)stream->view_format;
// Populate spl input basic input clip rect from plane state clip rect
@@ -108,19 +114,21 @@ void translate_SPL_in_params_from_pipe_ctx(struct pipe_ctx *pipe_ctx, struct spl
spl_in->basic_in.horizontal_mirror = plane_state->horizontal_mirror;
// Calculate horizontal splits and split index
- spl_in->basic_in.mpc_combine_h = resource_get_mpc_slice_count(pipe_ctx);
+ spl_in->basic_in.num_h_slices_recout_width_align.use_recout_width_aligned = false;
+ spl_in->basic_in.num_h_slices_recout_width_align.num_slices_recout_width.mpc_num_h_slices =
+ resource_get_mpc_slice_count(pipe_ctx);
if (stream->view_format == VIEW_3D_FORMAT_SIDE_BY_SIDE)
- spl_in->basic_in.mpc_combine_v = 0;
+ spl_in->basic_in.mpc_h_slice_index = 0;
else
- spl_in->basic_in.mpc_combine_v = resource_get_mpc_slice_index(pipe_ctx);
+ spl_in->basic_in.mpc_h_slice_index = resource_get_mpc_slice_index(pipe_ctx);
populate_splrect_from_rect(&spl_in->basic_out.odm_slice_rect, &odm_slice_src);
spl_in->basic_out.odm_combine_factor = 0;
spl_in->odm_slice_index = resource_get_odm_slice_index(pipe_ctx);
// Make spl input basic out info output_size width point to stream h active
spl_in->basic_out.output_size.width =
- stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right;
+ stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right + pipe_ctx->hblank_borrow;
// Make spl input basic out info output_size height point to v active
spl_in->basic_out.output_size.height =
stream->timing.v_addressable + stream->timing.v_border_bottom + stream->timing.v_border_top;
@@ -148,15 +156,16 @@ void translate_SPL_in_params_from_pipe_ctx(struct pipe_ctx *pipe_ctx, struct spl
spl_in->adaptive_sharpness.enable = true;
spl_in->adaptive_sharpness.sharpness_level = 0;
} else if (sharpness_setting == SHARPNESS_CUSTOM) {
- spl_in->adaptive_sharpness.sharpness_range.sdr_rgb_min = 0;
- spl_in->adaptive_sharpness.sharpness_range.sdr_rgb_max = 1750;
- spl_in->adaptive_sharpness.sharpness_range.sdr_rgb_mid = 750;
- spl_in->adaptive_sharpness.sharpness_range.sdr_yuv_min = 0;
- spl_in->adaptive_sharpness.sharpness_range.sdr_yuv_max = 3500;
- spl_in->adaptive_sharpness.sharpness_range.sdr_yuv_mid = 1500;
- spl_in->adaptive_sharpness.sharpness_range.hdr_rgb_min = 0;
- spl_in->adaptive_sharpness.sharpness_range.hdr_rgb_max = 2750;
- spl_in->adaptive_sharpness.sharpness_range.hdr_rgb_mid = 1500;
+ /* SAT: read harpness_range from dc_plane_state */
+ spl_in->adaptive_sharpness.sharpness_range.sdr_rgb_min = plane_state->sharpness_range.sdr_rgb_min;
+ spl_in->adaptive_sharpness.sharpness_range.sdr_rgb_max = plane_state->sharpness_range.sdr_rgb_max;
+ spl_in->adaptive_sharpness.sharpness_range.sdr_rgb_mid = plane_state->sharpness_range.sdr_rgb_mid;
+ spl_in->adaptive_sharpness.sharpness_range.sdr_yuv_min = plane_state->sharpness_range.sdr_yuv_min;
+ spl_in->adaptive_sharpness.sharpness_range.sdr_yuv_max = plane_state->sharpness_range.sdr_yuv_max;
+ spl_in->adaptive_sharpness.sharpness_range.sdr_yuv_mid = plane_state->sharpness_range.sdr_yuv_mid;
+ spl_in->adaptive_sharpness.sharpness_range.hdr_rgb_min = plane_state->sharpness_range.hdr_rgb_min;
+ spl_in->adaptive_sharpness.sharpness_range.hdr_rgb_max = plane_state->sharpness_range.hdr_rgb_max;
+ spl_in->adaptive_sharpness.sharpness_range.hdr_rgb_mid = plane_state->sharpness_range.hdr_rgb_mid;
if (force_sharpness_level > 0) {
if (force_sharpness_level > 10)
diff --git a/drivers/gpu/drm/amd/display/dc/dc_state_priv.h b/drivers/gpu/drm/amd/display/dc/dc_state_priv.h
index 1a12ef579ff4..1d9bae56ff6a 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_state_priv.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_state_priv.h
@@ -105,4 +105,24 @@ bool dc_state_is_fams2_in_use(
const struct dc *dc,
const struct dc_state *state);
+
+void dc_state_set_stream_subvp_cursor_limit(const struct dc_stream_state *stream,
+ struct dc_state *state,
+ bool limit);
+
+bool dc_state_get_stream_subvp_cursor_limit(const struct dc_stream_state *stream,
+ struct dc_state *state);
+
+void dc_state_set_stream_cursor_subvp_limit(const struct dc_stream_state *stream,
+ struct dc_state *state,
+ bool limit);
+
+bool dc_state_get_stream_cursor_subvp_limit(const struct dc_stream_state *stream,
+ struct dc_state *state);
+
+bool dc_state_can_clear_stream_cursor_subvp_limit(const struct dc_stream_state *stream,
+ struct dc_state *state);
+
+bool dc_state_is_subvp_in_use(struct dc_state *state);
+
#endif /* _DC_STATE_PRIV_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index 413970588a26..5fc6fea211de 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -44,6 +44,8 @@ struct mall_stream_config {
*/
enum mall_stream_type type;
struct dc_stream_state *paired_stream; // master / slave stream
+ bool subvp_limit_cursor_size; /* stream has/is using subvp limiting hw cursor support */
+ bool cursor_size_limit_subvp; /* stream is using hw cursor config preventing subvp */
};
struct dc_stream_status {
@@ -56,7 +58,7 @@ struct dc_stream_status {
int plane_count;
int audio_inst;
struct timing_sync_info timing_sync_info;
- struct dc_plane_state *plane_states[MAX_SURFACE_NUM];
+ struct dc_plane_state *plane_states[MAX_SURFACES];
bool is_abm_supported;
struct mall_stream_config mall_stream_config;
bool fpo_in_use;
@@ -447,10 +449,6 @@ enum dc_status dc_stream_add_dsc_to_resource(struct dc *dc,
struct dc_state *state,
struct dc_stream_state *stream);
-bool dc_stream_warmup_writeback(struct dc *dc,
- int num_dwb,
- struct dc_writeback_info *wb_info);
-
bool dc_stream_dmdata_status_done(struct dc *dc, struct dc_stream_state *stream);
bool dc_stream_set_dynamic_metadata(struct dc *dc,
@@ -507,6 +505,11 @@ void program_cursor_position(
struct dc *dc,
struct dc_stream_state *stream);
+bool dc_stream_check_cursor_attributes(
+ const struct dc_stream_state *stream,
+ struct dc_state *state,
+ const struct dc_cursor_attributes *attributes);
+
bool dc_stream_set_cursor_attributes(
struct dc_stream_state *stream,
const struct dc_cursor_attributes *attributes);
@@ -532,26 +535,29 @@ bool dc_stream_get_last_used_drr_vtotal(struct dc *dc,
struct dc_stream_state *stream,
uint32_t *refresh_rate);
-bool dc_stream_get_crtc_position(struct dc *dc,
- struct dc_stream_state **stream,
- int num_streams,
- unsigned int *v_pos,
- unsigned int *nom_v_pos);
-
#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
bool dc_stream_forward_crc_window(struct dc_stream_state *stream,
struct rect *rect,
+ uint8_t phy_id,
bool is_stop);
+
+bool dc_stream_forward_multiple_crc_window(struct dc_stream_state *stream,
+ struct crc_window *window,
+ uint8_t phy_id,
+ bool stop);
#endif
bool dc_stream_configure_crc(struct dc *dc,
struct dc_stream_state *stream,
struct crc_params *crc_window,
bool enable,
- bool continuous);
+ bool continuous,
+ uint8_t idx,
+ bool reset);
bool dc_stream_get_crc(struct dc *dc,
struct dc_stream_state *stream,
+ uint8_t idx,
uint32_t *r_cr,
uint32_t *g_y,
uint32_t *b_cb);
@@ -573,11 +579,16 @@ bool dc_stream_set_gamut_remap(struct dc *dc,
bool dc_stream_program_csc_matrix(struct dc *dc,
struct dc_stream_state *stream);
-bool dc_stream_get_crtc_position(struct dc *dc,
- struct dc_stream_state **stream,
- int num_streams,
- unsigned int *v_pos,
- unsigned int *nom_v_pos);
+struct dc_rmcm_3dlut *dc_stream_get_3dlut_for_stream(
+ const struct dc *dc,
+ const struct dc_stream_state *stream,
+ bool allocate_one);
+
+void dc_stream_release_3dlut_for_stream(
+ const struct dc *dc,
+ const struct dc_stream_state *stream);
+
+void dc_stream_init_rmcm_3dlut(struct dc *dc);
struct pipe_ctx *dc_stream_get_pipe_ctx(struct dc_stream_state *stream);
@@ -586,4 +597,8 @@ void dc_dmub_update_dirty_rect(struct dc *dc,
struct dc_stream_state *stream,
struct dc_surface_update *srf_updates,
struct dc_state *context);
+
+bool dc_stream_is_cursor_limit_pending(struct dc *dc, struct dc_stream_state *stream);
+bool dc_stream_can_clear_cursor_limit(struct dc *dc, struct dc_stream_state *stream);
+
#endif /* DC_STREAM_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dc_types.h b/drivers/gpu/drm/amd/display/dc/dc_types.h
index edf4df1d03b5..375ca2f13b7a 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_types.h
@@ -76,7 +76,6 @@ struct dc_perf_trace {
unsigned long last_entry_write;
};
-#define MAX_SURFACE_NUM 6
#define NUM_PIXEL_FORMATS 10
enum tiling_mode {
@@ -176,12 +175,14 @@ struct dc_panel_patch {
unsigned int embedded_tiled_slave;
unsigned int disable_fams;
unsigned int skip_avmute;
+ unsigned int skip_audio_sab_check;
unsigned int mst_start_top_delay;
unsigned int remove_sink_ext_caps;
unsigned int disable_colorimetry;
uint8_t blankstream_before_otg_off;
bool oled_optimize_display_on;
unsigned int force_mst_blocked_discovery;
+ unsigned int wait_after_dpcd_poweroff_ms;
};
struct dc_edid_caps {
@@ -210,6 +211,7 @@ struct dc_edid_caps {
bool edid_hdmi;
bool hdr_supported;
+ bool rr_capable;
struct dc_panel_patch panel_patch;
};
@@ -262,6 +264,7 @@ enum dc_timing_source {
TIMING_SOURCE_EDID_4BYTE,
TIMING_SOURCE_EDID_CEA_DISPLAYID_VTDB,
TIMING_SOURCE_EDID_CEA_RID,
+ TIMING_SOURCE_EDID_DISPLAYID_TYPE5,
TIMING_SOURCE_VBIOS,
TIMING_SOURCE_CV,
TIMING_SOURCE_TV,
@@ -875,6 +878,14 @@ struct dsc_dec_dpcd_caps {
bool is_dp; /* Decoded format */
};
+struct hblank_expansion_dpcd_caps {
+ bool expansion_supported;
+ bool reduction_supported;
+ bool buffer_unit_bytes; /* True: buffer size in bytes. False: buffer size in pixels*/
+ bool buffer_per_port; /* True: buffer size per port. False: buffer size per lane*/
+ uint32_t buffer_size; /* Add 1 to value and multiply by 32 */
+};
+
struct dc_golden_table {
uint16_t dc_golden_table_ver;
uint32_t aux_dphy_rx_control0_val;
@@ -932,10 +943,17 @@ enum backlight_control_type {
};
#if defined(CONFIG_DRM_AMD_SECURE_DISPLAY)
+#define MAX_CRC_WINDOW_NUM 2
+
struct otg_phy_mux {
uint8_t phy_output_num;
uint8_t otg_output_num;
};
+
+struct crc_window {
+ struct rect rect;
+ bool enable;
+};
#endif
enum dc_detect_reason {
@@ -1019,6 +1037,13 @@ struct psr_settings {
unsigned int psr_sdp_transmit_line_num_deadline;
uint8_t force_ffu_mode;
unsigned int psr_power_opt;
+
+ /**
+ * Some panels cannot handle idle pattern during PSR entry.
+ * To power down phy before disable stream to avoid sending
+ * idle pattern.
+ */
+ uint8_t power_down_phy_before_disable_stream;
};
enum replay_coasting_vtotal_type {
@@ -1052,10 +1077,13 @@ enum replay_FW_Message_type {
union replay_error_status {
struct {
- unsigned char STATE_TRANSITION_ERROR :1;
- unsigned char LINK_CRC_ERROR :1;
- unsigned char DESYNC_ERROR :1;
- unsigned char RESERVED :5;
+ unsigned int STATE_TRANSITION_ERROR :1;
+ unsigned int LINK_CRC_ERROR :1;
+ unsigned int DESYNC_ERROR :1;
+ unsigned int RESERVED_3 :1;
+ unsigned int LOW_RR_INCORRECT_VTOTAL :1;
+ unsigned int NO_DOUBLED_RR :1;
+ unsigned int RESERVED_6_7 :2;
} bits;
unsigned char raw;
};
@@ -1064,7 +1092,8 @@ union replay_low_refresh_rate_enable_options {
struct {
//BIT[0-3]: Replay Low Hz Support control
unsigned int ENABLE_LOW_RR_SUPPORT :1;
- unsigned int RESERVED_1_3 :3;
+ unsigned int SKIP_ASIC_CHECK :1;
+ unsigned int RESERVED_2_3 :2;
//BIT[4-15]: Replay Low Hz Enable Scenarios
unsigned int ENABLE_STATIC_SCREEN :1;
unsigned int ENABLE_FULL_SCREEN_VIDEO :1;
@@ -1102,6 +1131,12 @@ struct replay_config {
union replay_error_status replay_error_status;
/* Replay Low Hz enable Options */
union replay_low_refresh_rate_enable_options low_rr_enable_options;
+ /* Replay coasting vtotal is within low refresh rate range. */
+ bool low_rr_activated;
+ /* Replay low refresh rate supported*/
+ bool low_rr_supported;
+ /* Replay Video Conferencing Optimization Enabled */
+ bool replay_video_conferencing_optimization_enabled;
};
/* Replay feature flags*/
@@ -1126,10 +1161,12 @@ struct replay_settings {
uint32_t defer_update_coasting_vtotal_table[PR_COASTING_TYPE_NUM];
/* Maximum link off frame count */
uint32_t link_off_frame_count;
- /* Replay pseudo vtotal for abm + ips on full screen video which can improve ips residency */
- uint16_t abm_with_ips_on_full_screen_video_pseudo_vtotal;
+ /* Replay pseudo vtotal for low refresh rate*/
+ uint16_t low_rr_full_screen_video_pseudo_vtotal;
/* Replay last pseudo vtotal set to DMUB */
uint16_t last_pseudo_vtotal;
+ /* Replay desync error */
+ uint32_t replay_desync_error_fail_count;
};
/* To split out "global" and "per-panel" config settings.
@@ -1195,7 +1232,6 @@ struct dc_dpia_bw_alloc {
int bw_granularity; // BW Granularity
int dp_overhead; // DP overhead in dp tunneling
bool bw_alloc_enabled; // The BW Alloc Mode Support is turned ON for all 3: DP-Tx & Dpia & CM
- bool response_ready; // Response ready from the CM side
uint8_t nrd_max_lane_count; // Non-reduced max lane count
uint8_t nrd_max_link_rate; // Non-reduced max link rate
};
@@ -1242,7 +1278,7 @@ struct dc_cm2_gpu_mem_format_parameters {
enum dc_cm2_gpu_mem_size {
DC_CM2_GPU_MEM_SIZE_171717,
- DC_CM2_GPU_MEM_SIZE_TRANSFORMED
+ DC_CM2_GPU_MEM_SIZE_TRANSFORMED,
};
struct dc_cm2_gpu_mem_parameters {
@@ -1251,6 +1287,7 @@ struct dc_cm2_gpu_mem_parameters {
struct dc_cm2_gpu_mem_format_parameters format_params;
enum dc_cm2_gpu_mem_pixel_component_order component_order;
enum dc_cm2_gpu_mem_size size;
+ uint16_t bit_depth;
};
enum dc_cm2_transfer_func_source {
@@ -1274,6 +1311,11 @@ struct dc_cm2_func_luts {
const struct dc_3dlut *lut3d_func;
struct dc_cm2_gpu_mem_parameters gpu_mem_params;
};
+ bool rmcm_3dlut_shaper_select;
+ bool mpc_3dlut_enable;
+ bool rmcm_3dlut_enable;
+ bool mpc_mcm_post_blend;
+ uint8_t rmcm_tmz;
} lut3d_data;
const struct dc_transfer_func *lut1d_func;
};
@@ -1331,4 +1373,19 @@ struct set_backlight_level_params {
uint8_t aux_inst;
};
+enum dc_validate_mode {
+ /* validate the mode and program HW */
+ DC_VALIDATE_MODE_AND_PROGRAMMING = 0,
+ /* only validate the mode */
+ DC_VALIDATE_MODE_ONLY = 1,
+ /* validate the mode and get the max state (voltage level) */
+ DC_VALIDATE_MODE_AND_STATE_INDEX = 2,
+};
+
+struct dc_validation_dpia_set {
+ const struct dc_link *link;
+ const struct dc_tunnel_settings *tunnel_settings;
+ uint32_t required_bw;
+};
+
#endif /* DC_TYPES_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dccg/dcn20/dcn20_dccg.h b/drivers/gpu/drm/amd/display/dc/dccg/dcn20/dcn20_dccg.h
index 160c299419b7..a9b88f5e0c04 100644
--- a/drivers/gpu/drm/amd/display/dc/dccg/dcn20/dcn20_dccg.h
+++ b/drivers/gpu/drm/amd/display/dc/dccg/dcn20/dcn20_dccg.h
@@ -379,53 +379,55 @@ struct dccg_mask {
DCCG401_REG_FIELD_LIST(uint32_t)
};
+#define DCCG_REG_VARIABLE_LIST \
+ uint32_t DPPCLK_DTO_CTRL; \
+ uint32_t DPPCLK_DTO_PARAM[6]; \
+ uint32_t REFCLK_CNTL; \
+ uint32_t DISPCLK_FREQ_CHANGE_CNTL; \
+ uint32_t OTG_PIXEL_RATE_CNTL[MAX_PIPES]; \
+ uint32_t HDMICHARCLK_CLOCK_CNTL[6]; \
+ uint32_t PHYASYMCLK_CLOCK_CNTL; \
+ uint32_t PHYBSYMCLK_CLOCK_CNTL; \
+ uint32_t PHYCSYMCLK_CLOCK_CNTL; \
+ uint32_t PHYDSYMCLK_CLOCK_CNTL; \
+ uint32_t PHYESYMCLK_CLOCK_CNTL; \
+ uint32_t DTBCLK_DTO_MODULO[MAX_PIPES]; \
+ uint32_t DTBCLK_DTO_PHASE[MAX_PIPES]; \
+ uint32_t DCCG_AUDIO_DTBCLK_DTO_MODULO; \
+ uint32_t DCCG_AUDIO_DTBCLK_DTO_PHASE; \
+ uint32_t DCCG_AUDIO_DTO_SOURCE; \
+ uint32_t DPSTREAMCLK_CNTL; \
+ uint32_t HDMISTREAMCLK_CNTL; \
+ uint32_t SYMCLK32_SE_CNTL; \
+ uint32_t SYMCLK32_LE_CNTL; \
+ uint32_t DENTIST_DISPCLK_CNTL; \
+ uint32_t DSCCLK_DTO_CTRL; \
+ uint32_t DSCCLK0_DTO_PARAM; \
+ uint32_t DSCCLK1_DTO_PARAM; \
+ uint32_t DSCCLK2_DTO_PARAM; \
+ uint32_t DSCCLK3_DTO_PARAM; \
+ uint32_t DPSTREAMCLK_ROOT_GATE_DISABLE; \
+ uint32_t DPSTREAMCLK_GATE_DISABLE; \
+ uint32_t DCCG_GATE_DISABLE_CNTL; \
+ uint32_t DCCG_GATE_DISABLE_CNTL2; \
+ uint32_t DCCG_GATE_DISABLE_CNTL3; \
+ uint32_t HDMISTREAMCLK0_DTO_PARAM; \
+ uint32_t DCCG_GATE_DISABLE_CNTL4; \
+ uint32_t OTG_PIXEL_RATE_DIV; \
+ uint32_t DTBCLK_P_CNTL; \
+ uint32_t DPPCLK_CTRL; \
+ uint32_t DCCG_GATE_DISABLE_CNTL5; \
+ uint32_t DCCG_GATE_DISABLE_CNTL6; \
+ uint32_t DCCG_GLOBAL_FGCG_REP_CNTL; \
+ uint32_t SYMCLKA_CLOCK_ENABLE; \
+ uint32_t SYMCLKB_CLOCK_ENABLE; \
+ uint32_t SYMCLKC_CLOCK_ENABLE; \
+ uint32_t SYMCLKD_CLOCK_ENABLE; \
+ uint32_t SYMCLKE_CLOCK_ENABLE; \
+ uint32_t DP_DTO_MODULO[MAX_PIPES]; \
+ uint32_t DP_DTO_PHASE[MAX_PIPES]
struct dccg_registers {
- uint32_t DPPCLK_DTO_CTRL;
- uint32_t DPPCLK_DTO_PARAM[6];
- uint32_t REFCLK_CNTL;
- uint32_t DISPCLK_FREQ_CHANGE_CNTL;
- uint32_t OTG_PIXEL_RATE_CNTL[MAX_PIPES];
- uint32_t HDMICHARCLK_CLOCK_CNTL[6];
- uint32_t PHYASYMCLK_CLOCK_CNTL;
- uint32_t PHYBSYMCLK_CLOCK_CNTL;
- uint32_t PHYCSYMCLK_CLOCK_CNTL;
- uint32_t PHYDSYMCLK_CLOCK_CNTL;
- uint32_t PHYESYMCLK_CLOCK_CNTL;
- uint32_t DTBCLK_DTO_MODULO[MAX_PIPES];
- uint32_t DTBCLK_DTO_PHASE[MAX_PIPES];
- uint32_t DCCG_AUDIO_DTBCLK_DTO_MODULO;
- uint32_t DCCG_AUDIO_DTBCLK_DTO_PHASE;
- uint32_t DCCG_AUDIO_DTO_SOURCE;
- uint32_t DPSTREAMCLK_CNTL;
- uint32_t HDMISTREAMCLK_CNTL;
- uint32_t SYMCLK32_SE_CNTL;
- uint32_t SYMCLK32_LE_CNTL;
- uint32_t DENTIST_DISPCLK_CNTL;
- uint32_t DSCCLK_DTO_CTRL;
- uint32_t DSCCLK0_DTO_PARAM;
- uint32_t DSCCLK1_DTO_PARAM;
- uint32_t DSCCLK2_DTO_PARAM;
- uint32_t DSCCLK3_DTO_PARAM;
- uint32_t DPSTREAMCLK_ROOT_GATE_DISABLE;
- uint32_t DPSTREAMCLK_GATE_DISABLE;
- uint32_t DCCG_GATE_DISABLE_CNTL;
- uint32_t DCCG_GATE_DISABLE_CNTL2;
- uint32_t DCCG_GATE_DISABLE_CNTL3;
- uint32_t HDMISTREAMCLK0_DTO_PARAM;
- uint32_t DCCG_GATE_DISABLE_CNTL4;
- uint32_t OTG_PIXEL_RATE_DIV;
- uint32_t DTBCLK_P_CNTL;
- uint32_t DPPCLK_CTRL;
- uint32_t DCCG_GATE_DISABLE_CNTL5;
- uint32_t DCCG_GATE_DISABLE_CNTL6;
- uint32_t DCCG_GLOBAL_FGCG_REP_CNTL;
- uint32_t SYMCLKA_CLOCK_ENABLE;
- uint32_t SYMCLKB_CLOCK_ENABLE;
- uint32_t SYMCLKC_CLOCK_ENABLE;
- uint32_t SYMCLKD_CLOCK_ENABLE;
- uint32_t SYMCLKE_CLOCK_ENABLE;
- uint32_t DP_DTO_MODULO[MAX_PIPES];
- uint32_t DP_DTO_PHASE[MAX_PIPES];
+ DCCG_REG_VARIABLE_LIST;
};
struct dcn_dccg {
diff --git a/drivers/gpu/drm/amd/display/dc/dccg/dcn35/dcn35_dccg.c b/drivers/gpu/drm/amd/display/dc/dccg/dcn35/dcn35_dccg.c
index b363f5360818..58c84f555c0f 100644
--- a/drivers/gpu/drm/amd/display/dc/dccg/dcn35/dcn35_dccg.c
+++ b/drivers/gpu/drm/amd/display/dc/dccg/dcn35/dcn35_dccg.c
@@ -391,6 +391,7 @@ static void dccg35_set_dppclk_rcg(struct dccg *dccg,
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
+
if (!dccg->ctx->dc->debug.root_clock_optimization.bits.dpp && enable)
return;
@@ -411,6 +412,8 @@ static void dccg35_set_dppclk_rcg(struct dccg *dccg,
BREAK_TO_DEBUGGER();
break;
}
+ //DC_LOG_DEBUG("%s: inst(%d) DPPCLK rcg_disable: %d\n", __func__, inst, enable ? 0 : 1);
+
}
static void dccg35_set_dpstreamclk_rcg(
@@ -1035,6 +1038,7 @@ static void dccg35_enable_dpp_clk_new(
DPPCLK0_DTO_MODULO, 0xFF);
}
+
static void dccg35_disable_dpp_clk_new(
struct dccg *dccg,
int inst)
@@ -1112,30 +1116,24 @@ static void dcn35_set_dppclk_enable(struct dccg *dccg,
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
+
switch (dpp_inst) {
case 0:
REG_UPDATE(DPPCLK_CTRL, DPPCLK0_EN, enable);
- if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpp)
- REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK0_ROOT_GATE_DISABLE, enable);
break;
case 1:
REG_UPDATE(DPPCLK_CTRL, DPPCLK1_EN, enable);
- if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpp)
- REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK1_ROOT_GATE_DISABLE, enable);
break;
case 2:
REG_UPDATE(DPPCLK_CTRL, DPPCLK2_EN, enable);
- if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpp)
- REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK2_ROOT_GATE_DISABLE, enable);
break;
case 3:
REG_UPDATE(DPPCLK_CTRL, DPPCLK3_EN, enable);
- if (dccg->ctx->dc->debug.root_clock_optimization.bits.dpp)
- REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK3_ROOT_GATE_DISABLE, enable);
break;
default:
break;
}
+ //DC_LOG_DEBUG("%s: dpp_inst(%d) DPPCLK_EN = %d\n", __func__, dpp_inst, enable);
}
@@ -1163,14 +1161,18 @@ static void dccg35_update_dpp_dto(struct dccg *dccg, int dpp_inst,
ASSERT(false);
phase = 0xff;
}
+ dccg35_set_dppclk_rcg(dccg, dpp_inst, false);
REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0,
DPPCLK0_DTO_PHASE, phase,
DPPCLK0_DTO_MODULO, modulo);
dcn35_set_dppclk_enable(dccg, dpp_inst, true);
- } else
+ } else {
dcn35_set_dppclk_enable(dccg, dpp_inst, false);
+ /*we have this in hwss: disable_plane*/
+ //dccg35_set_dppclk_rcg(dccg, dpp_inst, true);
+ }
dccg->pipe_dppclk_khz[dpp_inst] = req_dppclk;
}
@@ -1182,6 +1184,7 @@ static void dccg35_set_dppclk_root_clock_gating(struct dccg *dccg,
if (!dccg->ctx->dc->debug.root_clock_optimization.bits.dpp)
return;
+
switch (dpp_inst) {
case 0:
REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DPPCLK0_ROOT_GATE_DISABLE, enable);
@@ -1198,6 +1201,8 @@ static void dccg35_set_dppclk_root_clock_gating(struct dccg *dccg,
default:
break;
}
+ //DC_LOG_DEBUG("%s: dpp_inst(%d) rcg: %d\n", __func__, dpp_inst, enable);
+
}
static void dccg35_get_pixel_rate_div(
@@ -1521,28 +1526,30 @@ static void dccg35_set_physymclk_root_clock_gating(
switch (phy_inst) {
case 0:
REG_UPDATE(DCCG_GATE_DISABLE_CNTL2,
- PHYASYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0);
+ PHYASYMCLK_ROOT_GATE_DISABLE, enable ? 0 : 1);
break;
case 1:
REG_UPDATE(DCCG_GATE_DISABLE_CNTL2,
- PHYBSYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0);
+ PHYBSYMCLK_ROOT_GATE_DISABLE, enable ? 0 : 1);
break;
case 2:
REG_UPDATE(DCCG_GATE_DISABLE_CNTL2,
- PHYCSYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0);
+ PHYCSYMCLK_ROOT_GATE_DISABLE, enable ? 0 : 1);
break;
case 3:
REG_UPDATE(DCCG_GATE_DISABLE_CNTL2,
- PHYDSYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0);
+ PHYDSYMCLK_ROOT_GATE_DISABLE, enable ? 0 : 1);
break;
case 4:
REG_UPDATE(DCCG_GATE_DISABLE_CNTL2,
- PHYESYMCLK_ROOT_GATE_DISABLE, enable ? 1 : 0);
+ PHYESYMCLK_ROOT_GATE_DISABLE, enable ? 0 : 1);
break;
default:
BREAK_TO_DEBUGGER();
return;
}
+ //DC_LOG_DEBUG("%s: dpp_inst(%d) PHYESYMCLK_ROOT_GATE_DISABLE:\n", __func__, phy_inst, enable ? 0 : 1);
+
}
static void dccg35_set_physymclk(
@@ -1643,6 +1650,8 @@ static void dccg35_dpp_root_clock_control(
return;
if (clock_on) {
+ dccg35_set_dppclk_rcg(dccg, dpp_inst, false);
+
/* turn off the DTO and leave phase/modulo at max */
dcn35_set_dppclk_enable(dccg, dpp_inst, 1);
REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0,
@@ -1654,6 +1663,8 @@ static void dccg35_dpp_root_clock_control(
REG_SET_2(DPPCLK_DTO_PARAM[dpp_inst], 0,
DPPCLK0_DTO_PHASE, 0,
DPPCLK0_DTO_MODULO, 1);
+ /*we have this in hwss: disable_plane*/
+ //dccg35_set_dppclk_rcg(dccg, dpp_inst, true);
}
dccg->dpp_clock_gated[dpp_inst] = !clock_on;
@@ -1771,36 +1782,40 @@ static void dccg35_enable_dscclk(struct dccg *dccg, int inst)
//Disable DTO
switch (inst) {
case 0:
+ if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc)
+ REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK0_ROOT_GATE_DISABLE, 1);
+
REG_UPDATE_2(DSCCLK0_DTO_PARAM,
DSCCLK0_DTO_PHASE, 0,
DSCCLK0_DTO_MODULO, 0);
REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK0_EN, 1);
- if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc)
- REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK0_ROOT_GATE_DISABLE, 1);
break;
case 1:
+ if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc)
+ REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK1_ROOT_GATE_DISABLE, 1);
+
REG_UPDATE_2(DSCCLK1_DTO_PARAM,
DSCCLK1_DTO_PHASE, 0,
DSCCLK1_DTO_MODULO, 0);
REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK1_EN, 1);
- if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc)
- REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK1_ROOT_GATE_DISABLE, 1);
break;
case 2:
+ if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc)
+ REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK2_ROOT_GATE_DISABLE, 1);
+
REG_UPDATE_2(DSCCLK2_DTO_PARAM,
DSCCLK2_DTO_PHASE, 0,
DSCCLK2_DTO_MODULO, 0);
REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK2_EN, 1);
- if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc)
- REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK2_ROOT_GATE_DISABLE, 1);
break;
case 3:
+ if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc)
+ REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK3_ROOT_GATE_DISABLE, 1);
+
REG_UPDATE_2(DSCCLK3_DTO_PARAM,
DSCCLK3_DTO_PHASE, 0,
DSCCLK3_DTO_MODULO, 0);
REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK3_EN, 1);
- if (dccg->ctx->dc->debug.root_clock_optimization.bits.dsc)
- REG_UPDATE(DCCG_GATE_DISABLE_CNTL6, DSCCLK3_ROOT_GATE_DISABLE, 1);
break;
default:
BREAK_TO_DEBUGGER();
@@ -1813,9 +1828,6 @@ static void dccg35_disable_dscclk(struct dccg *dccg,
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
- if (!dccg->ctx->dc->debug.root_clock_optimization.bits.dsc)
- return;
-
switch (inst) {
case 0:
REG_UPDATE(DSCCLK_DTO_CTRL, DSCCLK0_EN, 0);
diff --git a/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.c b/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.c
index d3e46c3cfa57..668ee2d405fd 100644
--- a/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.c
+++ b/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.c
@@ -116,7 +116,7 @@ static void dccg401_wait_for_dentist_change_done(
REG_WAIT(DENTIST_DISPCLK_CNTL, DENTIST_DISPCLK_CHG_DONE, 1, 50, 2000);
}
-static void dccg401_get_pixel_rate_div(
+void dccg401_get_pixel_rate_div(
struct dccg *dccg,
uint32_t otg_inst,
uint32_t *tmds_div,
@@ -154,7 +154,7 @@ static void dccg401_get_pixel_rate_div(
*tmds_div = val_tmds_div == 0 ? PIXEL_RATE_DIV_BY_2 : PIXEL_RATE_DIV_BY_4;
}
-static void dccg401_set_pixel_rate_div(
+void dccg401_set_pixel_rate_div(
struct dccg *dccg,
uint32_t otg_inst,
enum pixel_rate_div tmds_div,
@@ -209,7 +209,7 @@ static void dccg401_set_pixel_rate_div(
}
-static void dccg401_set_dtbclk_p_src(
+void dccg401_set_dtbclk_p_src(
struct dccg *dccg,
enum streamclk_source src,
uint32_t otg_inst)
@@ -348,7 +348,7 @@ void dccg401_set_physymclk(
}
}
-static void dccg401_get_dccg_ref_freq(struct dccg *dccg,
+void dccg401_get_dccg_ref_freq(struct dccg *dccg,
unsigned int xtalin_freq_inKhz,
unsigned int *dccg_ref_freq_inKhz)
{
@@ -378,7 +378,7 @@ static void dccg401_otg_drop_pixel(struct dccg *dccg,
OTG_DROP_PIXEL[otg_inst], 1);
}
-static void dccg401_enable_symclk32_le(
+void dccg401_enable_symclk32_le(
struct dccg *dccg,
int hpo_le_inst,
enum phyd32clk_clock_source phyd32clk)
@@ -429,7 +429,7 @@ static void dccg401_enable_symclk32_le(
}
}
-static void dccg401_disable_symclk32_le(
+void dccg401_disable_symclk32_le(
struct dccg *dccg,
int hpo_le_inst)
{
@@ -531,7 +531,7 @@ static void dccg401_enable_dpstreamclk(struct dccg *dccg, int otg_inst, int dp_h
DPSTREAMCLK_ROOT_GATE_DISABLE, 1);
}
-static void dccg401_disable_dpstreamclk(struct dccg *dccg, int dp_hpo_inst)
+void dccg401_disable_dpstreamclk(struct dccg *dccg, int dp_hpo_inst)
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
@@ -574,7 +574,7 @@ static void dccg401_disable_dpstreamclk(struct dccg *dccg, int dp_hpo_inst)
}
}
-static void dccg401_set_dpstreamclk(
+void dccg401_set_dpstreamclk(
struct dccg *dccg,
enum streamclk_source src,
int otg_inst,
@@ -587,7 +587,7 @@ static void dccg401_set_dpstreamclk(
dccg401_enable_dpstreamclk(dccg, otg_inst, dp_hpo_inst);
}
-static void dccg401_set_dp_dto(
+void dccg401_set_dp_dto(
struct dccg *dccg,
const struct dp_dto_params *params)
{
@@ -727,7 +727,7 @@ void dccg401_init(struct dccg *dccg)
}
}
-static void dccg401_set_dto_dscclk(struct dccg *dccg, uint32_t inst)
+void dccg401_set_dto_dscclk(struct dccg *dccg, uint32_t inst, uint32_t num_slices_h)
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
@@ -763,7 +763,7 @@ static void dccg401_set_dto_dscclk(struct dccg *dccg, uint32_t inst)
}
}
-static void dccg401_set_ref_dscclk(struct dccg *dccg,
+void dccg401_set_ref_dscclk(struct dccg *dccg,
uint32_t dsc_inst)
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
@@ -798,7 +798,7 @@ static void dccg401_set_ref_dscclk(struct dccg *dccg,
}
}
-static void dccg401_enable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst, uint32_t link_enc_inst)
+void dccg401_enable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst, uint32_t link_enc_inst)
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
@@ -834,7 +834,7 @@ static void dccg401_enable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst
}
}
-static void dccg401_disable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst, uint32_t link_enc_inst)
+void dccg401_disable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst, uint32_t link_enc_inst)
{
struct dcn_dccg *dccg_dcn = TO_DCN_DCCG(dccg);
diff --git a/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.h b/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.h
index a196ce9e8127..5947a35363aa 100644
--- a/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.h
+++ b/drivers/gpu/drm/amd/display/dc/dccg/dcn401/dcn401_dccg.h
@@ -193,11 +193,47 @@
void dccg401_init(struct dccg *dccg);
void dccg401_update_dpp_dto(struct dccg *dccg, int dpp_inst, int req_dppclk);
-
+void dccg401_get_dccg_ref_freq(struct dccg *dccg,
+ unsigned int xtalin_freq_inKhz,
+ unsigned int *dccg_ref_freq_inKhz);
+void dccg401_set_dpstreamclk(
+ struct dccg *dccg,
+ enum streamclk_source src,
+ int otg_inst,
+ int dp_hpo_inst);
+void dccg401_enable_symclk32_le(
+ struct dccg *dccg,
+ int hpo_le_inst,
+ enum phyd32clk_clock_source phyd32clk);
+void dccg401_disable_symclk32_le(
+ struct dccg *dccg,
+ int hpo_le_inst);
+void dccg401_disable_dpstreamclk(struct dccg *dccg, int dp_hpo_inst);
+void dccg401_set_dto_dscclk(struct dccg *dccg, uint32_t inst, uint32_t num_slices_h);
+void dccg401_set_ref_dscclk(struct dccg *dccg,
+ uint32_t dsc_inst);
void dccg401_set_src_sel(
struct dccg *dccg,
const struct dtbclk_dto_params *params);
-
+void dccg401_set_pixel_rate_div(
+ struct dccg *dccg,
+ uint32_t otg_inst,
+ enum pixel_rate_div tmds_div,
+ enum pixel_rate_div unused);
+void dccg401_get_pixel_rate_div(
+ struct dccg *dccg,
+ uint32_t otg_inst,
+ uint32_t *tmds_div,
+ uint32_t *dp_dto_int);
+void dccg401_set_dp_dto(
+ struct dccg *dccg,
+ const struct dp_dto_params *params);
+void dccg401_enable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst, uint32_t link_enc_inst);
+void dccg401_disable_symclk_se(struct dccg *dccg, uint32_t stream_enc_inst, uint32_t link_enc_inst);
+void dccg401_set_dtbclk_p_src(
+ struct dccg *dccg,
+ enum streamclk_source src,
+ uint32_t otg_inst);
struct dccg *dccg401_create(
struct dc_context *ctx,
const struct dccg_registers *regs,
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
index b700608e4240..b4f5b4a6331a 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.c
@@ -976,11 +976,12 @@ static bool dcn31_program_pix_clk(
struct bp_pixel_clock_parameters bp_pc_params = {0};
enum transmitter_color_depth bp_pc_colour_depth = TRANSMITTER_COLOR_DEPTH_24;
- // Apply ssed(spread spectrum) dpref clock for edp only.
- if (clock_source->ctx->dc->clk_mgr->dp_dto_source_clock_in_khz != 0
- && pix_clk_params->signal_type == SIGNAL_TYPE_EDP
- && encoding == DP_8b_10b_ENCODING)
+ // Apply ssed(spread spectrum) dpref clock for edp and dp
+ if (clock_source->ctx->dc->clk_mgr->dp_dto_source_clock_in_khz != 0 &&
+ dc_is_dp_signal(pix_clk_params->signal_type) &&
+ encoding == DP_8b_10b_ENCODING)
dp_dto_ref_khz = clock_source->ctx->dc->clk_mgr->dp_dto_source_clock_in_khz;
+
// For these signal types Driver to program DP_DTO without calling VBIOS Command table
if (dc_is_dp_signal(pix_clk_params->signal_type) || dc_is_virtual_signal(pix_clk_params->signal_type)) {
if (e) {
@@ -1105,6 +1106,9 @@ static bool dcn401_program_pix_clk(
&dto_params);
} else {
+ if (pll_settings->actual_pix_clk_100hz > 6000000UL)
+ return false;
+
/* disables DP DTO when provided with TMDS signal type */
clock_source->ctx->dc->res_pool->dccg->funcs->set_dp_dto(
clock_source->ctx->dc->res_pool->dccg,
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h
index 0721ae895ae9..94128f7a18b1 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_clock_source.h
@@ -257,7 +257,7 @@ bool dce110_clk_src_construct(
struct dce110_clk_src *clk_src,
struct dc_context *ctx,
struct dc_bios *bios,
- enum clock_source_id,
+ enum clock_source_id id,
const struct dce110_clk_src_regs *regs,
const struct dce110_clk_src_shift *cs_shift,
const struct dce110_clk_src_mask *cs_mask);
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c b/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c
index d28826c3ae5f..0421b267a0b5 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_i2c_hw.c
@@ -292,9 +292,35 @@ static void set_speed(
FN(DC_I2C_DDC1_SPEED, DC_I2C_DDC1_THRESHOLD), 2);
}
+static bool acquire_engine(struct dce_i2c_hw *dce_i2c_hw)
+{
+ uint32_t arbitrate = 0;
+
+ REG_GET(DC_I2C_ARBITRATION, DC_I2C_REG_RW_CNTL_STATUS, &arbitrate);
+ switch (arbitrate) {
+ case DC_I2C_STATUS__DC_I2C_STATUS_USED_BY_SW:
+ return true;
+ case DC_I2C_STATUS__DC_I2C_STATUS_USED_BY_HW:
+ return false;
+ case DC_I2C_STATUS__DC_I2C_STATUS_IDLE:
+ default:
+ break;
+ }
+
+ REG_UPDATE(DC_I2C_ARBITRATION, DC_I2C_SW_USE_I2C_REG_REQ, true);
+ REG_GET(DC_I2C_ARBITRATION, DC_I2C_REG_RW_CNTL_STATUS, &arbitrate);
+ if (arbitrate != DC_I2C_STATUS__DC_I2C_STATUS_USED_BY_SW)
+ return false;
+
+ return true;
+}
+
static bool setup_engine(
struct dce_i2c_hw *dce_i2c_hw)
{
+ // Deassert soft reset to unblock I2C engine registers
+ REG_UPDATE(DC_I2C_CONTROL, DC_I2C_SOFT_RESET, false);
+
uint32_t i2c_setup_limit = I2C_SETUP_TIME_LIMIT_DCE;
uint32_t reset_length = 0;
@@ -309,8 +335,8 @@ static bool setup_engine(
REG_UPDATE_N(SETUP, 1,
FN(DC_I2C_DDC1_SETUP, DC_I2C_DDC1_CLK_EN), 1);
- /* we have checked I2c not used by DMCU, set SW use I2C REQ to 1 to indicate SW using it*/
- REG_UPDATE(DC_I2C_ARBITRATION, DC_I2C_SW_USE_I2C_REG_REQ, 1);
+ if (!acquire_engine(dce_i2c_hw))
+ return false;
/*set SW requested I2c speed to default, if API calls in it will be override later*/
set_speed(dce_i2c_hw, dce_i2c_hw->ctx->dc->caps.i2c_speed_in_khz);
@@ -319,9 +345,8 @@ static bool setup_engine(
i2c_setup_limit = dce_i2c_hw->setup_limit;
/* Program pin select */
- REG_UPDATE_6(DC_I2C_CONTROL,
+ REG_UPDATE_5(DC_I2C_CONTROL,
DC_I2C_GO, 0,
- DC_I2C_SOFT_RESET, 0,
DC_I2C_SEND_RESET, 0,
DC_I2C_SW_STATUS_RESET, 1,
DC_I2C_TRANSACTION_COUNT, 0,
@@ -351,6 +376,32 @@ static bool setup_engine(
return true;
}
+/**
+ * cntl_stuck_hw_workaround - Workaround for I2C engine stuck state
+ * @dce_i2c_hw: Pointer to dce_i2c_hw structure
+ *
+ * If we boot without an HDMI display, the I2C engine does not get initialized
+ * correctly. One of its symptoms is that SW_USE_I2C does not get cleared after
+ * acquire. After setting SW_DONE_USING_I2C on release, the engine gets
+ * immediately reacquired by SW, preventing DMUB from using it.
+ *
+ * This function checks the I2C arbitration status and applies a release
+ * workaround if necessary.
+ */
+static void cntl_stuck_hw_workaround(struct dce_i2c_hw *dce_i2c_hw)
+{
+ uint32_t arbitrate = 0;
+
+ REG_GET(DC_I2C_ARBITRATION, DC_I2C_REG_RW_CNTL_STATUS, &arbitrate);
+ if (arbitrate != DC_I2C_STATUS__DC_I2C_STATUS_USED_BY_SW)
+ return;
+
+ // Still acquired after release, release again as a workaround
+ REG_UPDATE(DC_I2C_ARBITRATION, DC_I2C_SW_DONE_USING_I2C_REG, true);
+ REG_GET(DC_I2C_ARBITRATION, DC_I2C_REG_RW_CNTL_STATUS, &arbitrate);
+ ASSERT(arbitrate != DC_I2C_STATUS__DC_I2C_STATUS_USED_BY_SW);
+}
+
static void release_engine(
struct dce_i2c_hw *dce_i2c_hw)
{
@@ -378,9 +429,9 @@ static void release_engine(
/*for HW HDCP Ri polling failure w/a test*/
set_speed(dce_i2c_hw, dce_i2c_hw->ctx->dc->caps.i2c_speed_in_khz_hdcp);
- /* Release I2C after reset, so HW or DMCU could use it */
- REG_UPDATE_2(DC_I2C_ARBITRATION, DC_I2C_SW_DONE_USING_I2C_REG, 1,
- DC_I2C_SW_USE_I2C_REG_REQ, 0);
+ // Release I2C engine so it can be used by HW or DMCU, automatically clears SW_USE_I2C
+ REG_UPDATE(DC_I2C_ARBITRATION, DC_I2C_SW_DONE_USING_I2C_REG, true);
+ cntl_stuck_hw_workaround(dce_i2c_hw);
if (dce_i2c_hw->ctx->dc->debug.enable_mem_low_power.bits.i2c) {
if (dce_i2c_hw->regs->DIO_MEM_PWR_CTRL)
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
index f5e1d9caee4c..1c2009e38aa1 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_mem_input.c
@@ -98,7 +98,7 @@ static enum mi_bits_per_pixel get_mi_bpp(
}
static enum mi_tiling_format get_mi_tiling(
- union dc_tiling_info *tiling_info)
+ struct dc_tiling_info *tiling_info)
{
switch (tiling_info->gfx8.array_mode) {
case DC_ARRAY_1D_TILED_THIN1:
@@ -133,7 +133,7 @@ static bool is_vert_scan(enum dc_rotation_angle rotation)
static void dce_mi_program_pte_vm(
struct mem_input *mi,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
enum dc_rotation_angle rotation)
{
struct dce_mem_input *dce_mi = TO_DCE_MEM_INPUT(mi);
@@ -430,7 +430,7 @@ static void dce120_mi_program_display_marks(struct mem_input *mi,
}
static void program_tiling(
- struct dce_mem_input *dce_mi, const union dc_tiling_info *info)
+ struct dce_mem_input *dce_mi, const struct dc_tiling_info *info)
{
if (dce_mi->masks->GRPH_SW_MODE) { /* GFX9 */
REG_UPDATE_6(GRPH_CONTROL,
@@ -481,7 +481,6 @@ static void program_tiling(
}
}
-
static void program_size_and_rotation(
struct dce_mem_input *dce_mi,
enum dc_rotation_angle rotation,
@@ -627,10 +626,31 @@ static void program_grph_pixel_format(
GRPH_PRESCALE_B_SIGN, sign);
}
+static void dce_mi_clear_tiling(
+ struct mem_input *mi)
+{
+ struct dce_mem_input *dce_mi = TO_DCE_MEM_INPUT(mi);
+
+ if (dce_mi->masks->GRPH_SW_MODE) { /* GFX9 */
+ REG_UPDATE(GRPH_CONTROL,
+ GRPH_SW_MODE, DC_SW_LINEAR);
+ }
+
+ if (dce_mi->masks->GRPH_MICRO_TILE_MODE) { /* GFX8 */
+ REG_UPDATE(GRPH_CONTROL,
+ GRPH_ARRAY_MODE, DC_SW_LINEAR);
+ }
+
+ if (dce_mi->masks->GRPH_ARRAY_MODE) { /* GFX6 but reuses gfx8 struct */
+ REG_UPDATE(GRPH_CONTROL,
+ GRPH_ARRAY_MODE, DC_SW_LINEAR);
+ }
+}
+
static void dce_mi_program_surface_config(
struct mem_input *mi,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -650,7 +670,7 @@ static void dce_mi_program_surface_config(
static void dce60_mi_program_surface_config(
struct mem_input *mi,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation, /* not used in DCE6 */
struct dc_plane_dcc_param *dcc,
@@ -884,7 +904,8 @@ static const struct mem_input_funcs dce_mi_funcs = {
.mem_input_program_pte_vm = dce_mi_program_pte_vm,
.mem_input_program_surface_config =
dce_mi_program_surface_config,
- .mem_input_is_flip_pending = dce_mi_is_flip_pending
+ .mem_input_is_flip_pending = dce_mi_is_flip_pending,
+ .mem_input_clear_tiling = dce_mi_clear_tiling,
};
#if defined(CONFIG_DRM_AMD_DC_SI)
@@ -897,7 +918,8 @@ static const struct mem_input_funcs dce60_mi_funcs = {
.mem_input_program_pte_vm = dce_mi_program_pte_vm,
.mem_input_program_surface_config =
dce60_mi_program_surface_config,
- .mem_input_is_flip_pending = dce_mi_is_flip_pending
+ .mem_input_is_flip_pending = dce_mi_is_flip_pending,
+ .mem_input_clear_tiling = dce_mi_clear_tiling,
};
#endif
@@ -910,7 +932,8 @@ static const struct mem_input_funcs dce112_mi_funcs = {
.mem_input_program_pte_vm = dce_mi_program_pte_vm,
.mem_input_program_surface_config =
dce_mi_program_surface_config,
- .mem_input_is_flip_pending = dce_mi_is_flip_pending
+ .mem_input_is_flip_pending = dce_mi_is_flip_pending,
+ .mem_input_clear_tiling = dce_mi_clear_tiling,
};
static const struct mem_input_funcs dce120_mi_funcs = {
@@ -922,7 +945,8 @@ static const struct mem_input_funcs dce120_mi_funcs = {
.mem_input_program_pte_vm = dce_mi_program_pte_vm,
.mem_input_program_surface_config =
dce_mi_program_surface_config,
- .mem_input_is_flip_pending = dce_mi_is_flip_pending
+ .mem_input_is_flip_pending = dce_mi_is_flip_pending,
+ .mem_input_clear_tiling = dce_mi_clear_tiling,
};
void dce_mem_input_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
index d199e4ed2e59..1130d7619b26 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_stream_encoder.c
@@ -418,7 +418,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
dynamic_range_rgb = 1; /*limited range*/
break;
case COLOR_SPACE_2020_RGB_FULLRANGE:
- case COLOR_SPACE_2020_YCBCR:
+ case COLOR_SPACE_2020_YCBCR_LIMITED:
case COLOR_SPACE_XR_RGB:
case COLOR_SPACE_MSREF_SCRGB:
case COLOR_SPACE_ADOBERGB:
@@ -430,6 +430,7 @@ static void dce110_stream_encoder_dp_set_stream_attribute(
case COLOR_SPACE_APPCTRL:
case COLOR_SPACE_CUSTOMPOINTS:
case COLOR_SPACE_UNKNOWN:
+ default:
/* do nothing */
break;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c
index ccc154b0281c..3b9011ef9b68 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm.c
@@ -28,6 +28,8 @@
#include "dc.h"
#include "core_types.h"
#include "dmub_cmd.h"
+#include "dc_dmub_srv.h"
+#include "dmub/dmub_srv.h"
#define TO_DMUB_ABM(abm)\
container_of(abm, struct dce_abm, base)
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm_lcd.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm_lcd.c
index 0d7e7f3b81a1..a641ae04450c 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_abm_lcd.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_abm_lcd.c
@@ -240,7 +240,8 @@ bool dmub_abm_save_restore(
cmd.abm_save_restore.abm_init_config_data.version = DMUB_CMD_ABM_CONTROL_VERSION_1;
cmd.abm_save_restore.abm_init_config_data.panel_mask = panel_mask;
- cmd.abm_save_restore.header.payload_bytes = sizeof(struct dmub_rb_cmd_abm_save_restore);
+ cmd.abm_save_restore.header.payload_bytes =
+ sizeof(struct dmub_rb_cmd_abm_save_restore) - sizeof(struct dmub_cmd_header);
dc_wake_and_execute_dmub_cmd(dc, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c
index bf636b28e3e1..d37ecfdde4f1 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_hw_lock_mgr.c
@@ -63,11 +63,26 @@ void dmub_hw_lock_mgr_inbox0_cmd(struct dc_dmub_srv *dmub_srv,
bool should_use_dmub_lock(struct dc_link *link)
{
+ /* ASIC doesn't support DMUB */
+ if (!link->ctx->dmub_srv)
+ return false;
+
if (link->psr_settings.psr_version == DC_PSR_VERSION_SU_1)
return true;
if (link->replay_settings.replay_feature_enabled)
return true;
+ /* only use HW lock for PSR1 on single eDP */
+ if (link->psr_settings.psr_version == DC_PSR_VERSION_1) {
+ struct dc_link *edp_links[MAX_NUM_EDP];
+ int edp_num;
+
+ dc_get_edp_links(link->dc, edp_links, &edp_num);
+
+ if (edp_num == 1)
+ return true;
+ }
+
return false;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
index cae18f8c1c9a..ff3b8244ba3d 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
@@ -390,8 +390,7 @@ static bool dmub_psr_copy_settings(struct dmub_psr *dmub,
!memcmp(link->dpcd_caps.sink_dev_id_str, DP_SINK_DEVICE_STR_ID_1,
sizeof(DP_SINK_DEVICE_STR_ID_1)))
link->psr_settings.force_ffu_mode = 1;
- else
- link->psr_settings.force_ffu_mode = 0;
+
copy_settings_data->force_ffu_mode = link->psr_settings.force_ffu_mode;
if (((link->dpcd_caps.fec_cap.bits.FEC_CAPABLE &&
@@ -419,6 +418,10 @@ static bool dmub_psr_copy_settings(struct dmub_psr *dmub,
copy_settings_data->relock_delay_frame_cnt = 0;
if (link->dpcd_caps.sink_dev_id == DP_BRANCH_DEVICE_ID_001CF8)
copy_settings_data->relock_delay_frame_cnt = 2;
+
+ copy_settings_data->power_down_phy_before_disable_stream =
+ link->psr_settings.power_down_phy_before_disable_stream;
+
copy_settings_data->dsc_slice_height = psr_context->dsc_slice_height;
dc_wake_and_execute_dmub_cmd(dc, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c
index c31e4f26a305..e7a318e26d38 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.c
@@ -4,6 +4,7 @@
#include "dc.h"
#include "dc_dmub_srv.h"
+#include "dc_dp_types.h"
#include "dmub/dmub_srv.h"
#include "core_types.h"
#include "dmub_replay.h"
@@ -43,21 +44,45 @@ static void dmub_replay_get_state(struct dmub_replay *dmub, enum replay_state *s
/*
* Enable/Disable Replay.
*/
-static void dmub_replay_enable(struct dmub_replay *dmub, bool enable, bool wait, uint8_t panel_inst)
+static void dmub_replay_enable(struct dmub_replay *dmub, bool enable, bool wait, uint8_t panel_inst,
+ struct dc_link *link)
{
union dmub_rb_cmd cmd;
struct dc_context *dc = dmub->ctx;
uint32_t retry_count;
enum replay_state state = REPLAY_STATE_0;
+ struct pipe_ctx *pipe_ctx = NULL;
+ struct resource_context *res_ctx = &link->ctx->dc->current_state->res_ctx;
+ uint8_t i;
memset(&cmd, 0, sizeof(cmd));
cmd.replay_enable.header.type = DMUB_CMD__REPLAY;
cmd.replay_enable.data.panel_inst = panel_inst;
cmd.replay_enable.header.sub_type = DMUB_CMD__REPLAY_ENABLE;
- if (enable)
+ if (enable) {
cmd.replay_enable.data.enable = REPLAY_ENABLE;
- else
+ // hpo stream/link encoder assignments are not static, need to update everytime we try to enable replay
+ if (link->cur_link_settings.link_rate >= LINK_RATE_UHBR10) {
+ for (i = 0; i < MAX_PIPES; i++) {
+ if (res_ctx &&
+ res_ctx->pipe_ctx[i].stream &&
+ res_ctx->pipe_ctx[i].stream->link &&
+ res_ctx->pipe_ctx[i].stream->link == link &&
+ res_ctx->pipe_ctx[i].stream->link->connector_signal == SIGNAL_TYPE_EDP) {
+ pipe_ctx = &res_ctx->pipe_ctx[i];
+ //TODO: refactor for multi edp support
+ break;
+ }
+ }
+
+ if (!pipe_ctx)
+ return;
+
+ cmd.replay_enable.data.hpo_stream_enc_inst = pipe_ctx->stream_res.hpo_dp_stream_enc->inst;
+ cmd.replay_enable.data.hpo_link_enc_inst = pipe_ctx->link_res.hpo_dp_link_enc->inst;
+ }
+ } else
cmd.replay_enable.data.enable = REPLAY_DISABLE;
cmd.replay_enable.header.payload_bytes = sizeof(struct dmub_rb_cmd_replay_enable_data);
@@ -149,6 +174,17 @@ static bool dmub_replay_copy_settings(struct dmub_replay *dmub,
copy_settings_data->digbe_inst = replay_context->digbe_inst;
copy_settings_data->digfe_inst = replay_context->digfe_inst;
+ if (link->cur_link_settings.link_rate >= LINK_RATE_UHBR10) {
+ if (pipe_ctx->stream_res.hpo_dp_stream_enc)
+ copy_settings_data->hpo_stream_enc_inst = pipe_ctx->stream_res.hpo_dp_stream_enc->inst;
+ else
+ copy_settings_data->hpo_stream_enc_inst = 0;
+ if (pipe_ctx->link_res.hpo_dp_link_enc)
+ copy_settings_data->hpo_link_enc_inst = pipe_ctx->link_res.hpo_dp_link_enc->inst;
+ else
+ copy_settings_data->hpo_link_enc_inst = 0;
+ }
+
if (pipe_ctx->plane_res.dpp)
copy_settings_data->dpp_inst = pipe_ctx->plane_res.dpp->inst;
else
@@ -211,6 +247,7 @@ static void dmub_replay_set_coasting_vtotal(struct dmub_replay *dmub,
pCmd->header.type = DMUB_CMD__REPLAY;
pCmd->header.sub_type = DMUB_CMD__REPLAY_SET_COASTING_VTOTAL;
pCmd->header.payload_bytes = sizeof(struct dmub_cmd_replay_set_coasting_vtotal_data);
+ pCmd->replay_set_coasting_vtotal_data.panel_inst = panel_inst;
pCmd->replay_set_coasting_vtotal_data.coasting_vtotal = (coasting_vtotal & 0xFFFF);
pCmd->replay_set_coasting_vtotal_data.coasting_vtotal_high = (coasting_vtotal & 0xFFFF0000) >> 16;
@@ -280,7 +317,9 @@ static void dmub_replay_set_power_opt_and_coasting_vtotal(struct dmub_replay *dm
memset(&cmd, 0, sizeof(cmd));
pCmd->header.type = DMUB_CMD__REPLAY;
pCmd->header.sub_type = DMUB_CMD__REPLAY_SET_POWER_OPT_AND_COASTING_VTOTAL;
- pCmd->header.payload_bytes = sizeof(struct dmub_rb_cmd_replay_set_power_opt_and_coasting_vtotal);
+ pCmd->header.payload_bytes =
+ sizeof(struct dmub_rb_cmd_replay_set_power_opt_and_coasting_vtotal) -
+ sizeof(struct dmub_cmd_header);
pCmd->replay_set_power_opt_data.power_opt = power_opt;
pCmd->replay_set_power_opt_data.panel_inst = panel_inst;
pCmd->replay_set_coasting_vtotal_data.coasting_vtotal = (coasting_vtotal & 0xFFFF);
@@ -319,7 +358,8 @@ static void dmub_replay_send_cmd(struct dmub_replay *dmub,
cmd.replay_set_timing_sync.header.sub_type =
DMUB_CMD__REPLAY_SET_TIMING_SYNC_SUPPORTED;
cmd.replay_set_timing_sync.header.payload_bytes =
- sizeof(struct dmub_rb_cmd_replay_set_timing_sync);
+ sizeof(struct dmub_rb_cmd_replay_set_timing_sync) -
+ sizeof(struct dmub_cmd_header);
//Cmd Body
cmd.replay_set_timing_sync.replay_set_timing_sync_data.panel_inst =
cmd_element->sync_data.panel_inst;
@@ -331,7 +371,8 @@ static void dmub_replay_send_cmd(struct dmub_replay *dmub,
cmd.replay_set_frameupdate_timer.header.sub_type =
DMUB_CMD__REPLAY_SET_RESIDENCY_FRAMEUPDATE_TIMER;
cmd.replay_set_frameupdate_timer.header.payload_bytes =
- sizeof(struct dmub_rb_cmd_replay_set_frameupdate_timer);
+ sizeof(struct dmub_rb_cmd_replay_set_frameupdate_timer) -
+ sizeof(struct dmub_cmd_header);
//Cmd Body
cmd.replay_set_frameupdate_timer.data.panel_inst =
cmd_element->panel_inst;
@@ -345,7 +386,8 @@ static void dmub_replay_send_cmd(struct dmub_replay *dmub,
cmd.replay_set_pseudo_vtotal.header.sub_type =
DMUB_CMD__REPLAY_SET_PSEUDO_VTOTAL;
cmd.replay_set_pseudo_vtotal.header.payload_bytes =
- sizeof(struct dmub_rb_cmd_replay_set_pseudo_vtotal);
+ sizeof(struct dmub_rb_cmd_replay_set_pseudo_vtotal) -
+ sizeof(struct dmub_cmd_header);
//Cmd Body
cmd.replay_set_pseudo_vtotal.data.panel_inst =
cmd_element->pseudo_vtotal_data.panel_inst;
@@ -357,7 +399,8 @@ static void dmub_replay_send_cmd(struct dmub_replay *dmub,
cmd.replay_disabled_adaptive_sync_sdp.header.sub_type =
DMUB_CMD__REPLAY_DISABLED_ADAPTIVE_SYNC_SDP;
cmd.replay_disabled_adaptive_sync_sdp.header.payload_bytes =
- sizeof(struct dmub_rb_cmd_replay_disabled_adaptive_sync_sdp);
+ sizeof(struct dmub_rb_cmd_replay_disabled_adaptive_sync_sdp) -
+ sizeof(struct dmub_cmd_header);
//Cmd Body
cmd.replay_disabled_adaptive_sync_sdp.data.panel_inst =
cmd_element->disabled_adaptive_sync_sdp_data.panel_inst;
@@ -369,7 +412,8 @@ static void dmub_replay_send_cmd(struct dmub_replay *dmub,
cmd.replay_set_general_cmd.header.sub_type =
DMUB_CMD__REPLAY_SET_GENERAL_CMD;
cmd.replay_set_general_cmd.header.payload_bytes =
- sizeof(struct dmub_rb_cmd_replay_set_general_cmd);
+ sizeof(struct dmub_rb_cmd_replay_set_general_cmd) -
+ sizeof(struct dmub_cmd_header);
//Cmd Body
cmd.replay_set_general_cmd.data.panel_inst =
cmd_element->set_general_cmd_data.panel_inst;
diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.h b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.h
index e6346c0ffc0e..ccbe385e132c 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.h
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_replay.h
@@ -19,7 +19,7 @@ struct dmub_replay_funcs {
void (*replay_get_state)(struct dmub_replay *dmub, enum replay_state *state,
uint8_t panel_inst);
void (*replay_enable)(struct dmub_replay *dmub, bool enable, bool wait,
- uint8_t panel_inst);
+ uint8_t panel_inst, struct dc_link *link);
bool (*replay_copy_settings)(struct dmub_replay *dmub, struct dc_link *link,
struct replay_context *replay_context, uint8_t panel_inst);
void (*replay_set_power_opt)(struct dmub_replay *dmub, unsigned int power_opt,
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
index d241ee13b293..59a0961b49da 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.c
@@ -409,19 +409,6 @@ void dce110_compressor_destroy(struct compressor **compressor)
*compressor = NULL;
}
-void get_max_support_fbc_buffersize(unsigned int *max_x, unsigned int *max_y)
-{
- *max_x = FBC_MAX_X;
- *max_y = FBC_MAX_Y;
-
- /* if (m_smallLocalFrameBufferMemory == 1)
- * {
- * *max_x = FBC_MAX_X_SG;
- * *max_y = FBC_MAX_Y_SG;
- * }
- */
-}
-
static const struct compressor_funcs dce110_compressor_funcs = {
.power_up_fbc = dce110_compressor_power_up_fbc,
.enable_fbc = dce110_compressor_enable_fbc,
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.h b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.h
index 26c7335a1cbf..223c57941e92 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.h
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_compressor.h
@@ -75,7 +75,5 @@ void dce110_compressor_program_lpt_control(struct compressor *cp,
bool dce110_compressor_is_lpt_enabled_in_hw(struct compressor *cp);
-void get_max_support_fbc_buffersize(unsigned int *max_x, unsigned int *max_y);
-
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c
index 8a3fbf95c48f..2c43c2422638 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_mem_input_v.c
@@ -162,7 +162,7 @@ static void enable(struct dce_mem_input *mem_input110)
static void program_tiling(
struct dce_mem_input *mem_input110,
- const union dc_tiling_info *info,
+ const struct dc_tiling_info *info,
const enum surface_pixel_format pixel_format)
{
uint32_t value = 0;
@@ -523,7 +523,7 @@ static const unsigned int dvmm_Hw_Setting_Linear[4][9] = {
/* Helper to get table entry from surface info */
static const unsigned int *get_dvmm_hw_setting(
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
enum surface_pixel_format format,
bool chroma)
{
@@ -563,7 +563,7 @@ static const unsigned int *get_dvmm_hw_setting(
static void dce_mem_input_v_program_pte_vm(
struct mem_input *mem_input,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
enum dc_rotation_angle rotation)
{
struct dce_mem_input *mem_input110 = TO_DCE_MEM_INPUT(mem_input);
@@ -636,7 +636,7 @@ static void dce_mem_input_v_program_pte_vm(
static void dce_mem_input_v_program_surface_config(
struct mem_input *mem_input,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
index fa422a8cbced..61b0807693fb 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.c
@@ -2127,70 +2127,131 @@ bool dce110_configure_crc(struct timing_generator *tg,
cntl_addr = CRTC_REG(mmCRTC_CRC_CNTL);
- /* First, disable CRC before we configure it. */
- dm_write_reg(tg->ctx, cntl_addr, 0);
+ if (!params->enable || params->reset)
+ /* First, disable CRC before we configure it. */
+ dm_write_reg(tg->ctx, cntl_addr, 0);
if (!params->enable)
return true;
/* Program frame boundaries */
- /* Window A x axis start and end. */
- value = 0;
- addr = CRTC_REG(mmCRTC_CRC0_WINDOWA_X_CONTROL);
- set_reg_field_value(value, params->windowa_x_start,
- CRTC_CRC0_WINDOWA_X_CONTROL,
- CRTC_CRC0_WINDOWA_X_START);
- set_reg_field_value(value, params->windowa_x_end,
- CRTC_CRC0_WINDOWA_X_CONTROL,
- CRTC_CRC0_WINDOWA_X_END);
- dm_write_reg(tg->ctx, addr, value);
-
- /* Window A y axis start and end. */
- value = 0;
- addr = CRTC_REG(mmCRTC_CRC0_WINDOWA_Y_CONTROL);
- set_reg_field_value(value, params->windowa_y_start,
- CRTC_CRC0_WINDOWA_Y_CONTROL,
- CRTC_CRC0_WINDOWA_Y_START);
- set_reg_field_value(value, params->windowa_y_end,
- CRTC_CRC0_WINDOWA_Y_CONTROL,
- CRTC_CRC0_WINDOWA_Y_END);
- dm_write_reg(tg->ctx, addr, value);
-
- /* Window B x axis start and end. */
- value = 0;
- addr = CRTC_REG(mmCRTC_CRC0_WINDOWB_X_CONTROL);
- set_reg_field_value(value, params->windowb_x_start,
- CRTC_CRC0_WINDOWB_X_CONTROL,
- CRTC_CRC0_WINDOWB_X_START);
- set_reg_field_value(value, params->windowb_x_end,
- CRTC_CRC0_WINDOWB_X_CONTROL,
- CRTC_CRC0_WINDOWB_X_END);
- dm_write_reg(tg->ctx, addr, value);
-
- /* Window B y axis start and end. */
- value = 0;
- addr = CRTC_REG(mmCRTC_CRC0_WINDOWB_Y_CONTROL);
- set_reg_field_value(value, params->windowb_y_start,
- CRTC_CRC0_WINDOWB_Y_CONTROL,
- CRTC_CRC0_WINDOWB_Y_START);
- set_reg_field_value(value, params->windowb_y_end,
- CRTC_CRC0_WINDOWB_Y_CONTROL,
- CRTC_CRC0_WINDOWB_Y_END);
- dm_write_reg(tg->ctx, addr, value);
-
- /* Set crc mode and selection, and enable. Only using CRC0*/
- value = 0;
- set_reg_field_value(value, params->continuous_mode ? 1 : 0,
- CRTC_CRC_CNTL, CRTC_CRC_CONT_EN);
- set_reg_field_value(value, params->selection,
- CRTC_CRC_CNTL, CRTC_CRC0_SELECT);
- set_reg_field_value(value, 1, CRTC_CRC_CNTL, CRTC_CRC_EN);
- dm_write_reg(tg->ctx, cntl_addr, value);
+ switch (params->crc_eng_inst) {
+ case 0:
+ /* Window A x axis start and end. */
+ value = 0;
+ addr = CRTC_REG(mmCRTC_CRC0_WINDOWA_X_CONTROL);
+ set_reg_field_value(value, params->windowa_x_start,
+ CRTC_CRC0_WINDOWA_X_CONTROL,
+ CRTC_CRC0_WINDOWA_X_START);
+ set_reg_field_value(value, params->windowa_x_end,
+ CRTC_CRC0_WINDOWA_X_CONTROL,
+ CRTC_CRC0_WINDOWA_X_END);
+ dm_write_reg(tg->ctx, addr, value);
+
+ /* Window A y axis start and end. */
+ value = 0;
+ addr = CRTC_REG(mmCRTC_CRC0_WINDOWA_Y_CONTROL);
+ set_reg_field_value(value, params->windowa_y_start,
+ CRTC_CRC0_WINDOWA_Y_CONTROL,
+ CRTC_CRC0_WINDOWA_Y_START);
+ set_reg_field_value(value, params->windowa_y_end,
+ CRTC_CRC0_WINDOWA_Y_CONTROL,
+ CRTC_CRC0_WINDOWA_Y_END);
+ dm_write_reg(tg->ctx, addr, value);
+
+ /* Window B x axis start and end. */
+ value = 0;
+ addr = CRTC_REG(mmCRTC_CRC0_WINDOWB_X_CONTROL);
+ set_reg_field_value(value, params->windowb_x_start,
+ CRTC_CRC0_WINDOWB_X_CONTROL,
+ CRTC_CRC0_WINDOWB_X_START);
+ set_reg_field_value(value, params->windowb_x_end,
+ CRTC_CRC0_WINDOWB_X_CONTROL,
+ CRTC_CRC0_WINDOWB_X_END);
+ dm_write_reg(tg->ctx, addr, value);
+
+ /* Window B y axis start and end. */
+ value = 0;
+ addr = CRTC_REG(mmCRTC_CRC0_WINDOWB_Y_CONTROL);
+ set_reg_field_value(value, params->windowb_y_start,
+ CRTC_CRC0_WINDOWB_Y_CONTROL,
+ CRTC_CRC0_WINDOWB_Y_START);
+ set_reg_field_value(value, params->windowb_y_end,
+ CRTC_CRC0_WINDOWB_Y_CONTROL,
+ CRTC_CRC0_WINDOWB_Y_END);
+ dm_write_reg(tg->ctx, addr, value);
+
+ /* Set crc mode and selection, and enable.*/
+ value = 0;
+ set_reg_field_value(value, params->continuous_mode ? 1 : 0,
+ CRTC_CRC_CNTL, CRTC_CRC_CONT_EN);
+ set_reg_field_value(value, params->selection,
+ CRTC_CRC_CNTL, CRTC_CRC0_SELECT);
+ set_reg_field_value(value, 1, CRTC_CRC_CNTL, CRTC_CRC_EN);
+ dm_write_reg(tg->ctx, cntl_addr, value);
+ break;
+ case 1:
+ /* Window A x axis start and end. */
+ value = 0;
+ addr = CRTC_REG(mmCRTC_CRC1_WINDOWA_X_CONTROL);
+ set_reg_field_value(value, params->windowa_x_start,
+ CRTC_CRC1_WINDOWA_X_CONTROL,
+ CRTC_CRC1_WINDOWA_X_START);
+ set_reg_field_value(value, params->windowa_x_end,
+ CRTC_CRC1_WINDOWA_X_CONTROL,
+ CRTC_CRC1_WINDOWA_X_END);
+ dm_write_reg(tg->ctx, addr, value);
+
+ /* Window A y axis start and end. */
+ value = 0;
+ addr = CRTC_REG(mmCRTC_CRC1_WINDOWA_Y_CONTROL);
+ set_reg_field_value(value, params->windowa_y_start,
+ CRTC_CRC1_WINDOWA_Y_CONTROL,
+ CRTC_CRC1_WINDOWA_Y_START);
+ set_reg_field_value(value, params->windowa_y_end,
+ CRTC_CRC1_WINDOWA_Y_CONTROL,
+ CRTC_CRC1_WINDOWA_Y_END);
+ dm_write_reg(tg->ctx, addr, value);
+
+ /* Window B x axis start and end. */
+ value = 0;
+ addr = CRTC_REG(mmCRTC_CRC1_WINDOWB_X_CONTROL);
+ set_reg_field_value(value, params->windowb_x_start,
+ CRTC_CRC1_WINDOWB_X_CONTROL,
+ CRTC_CRC1_WINDOWB_X_START);
+ set_reg_field_value(value, params->windowb_x_end,
+ CRTC_CRC1_WINDOWB_X_CONTROL,
+ CRTC_CRC1_WINDOWB_X_END);
+ dm_write_reg(tg->ctx, addr, value);
+
+ /* Window B y axis start and end. */
+ value = 0;
+ addr = CRTC_REG(mmCRTC_CRC1_WINDOWB_Y_CONTROL);
+ set_reg_field_value(value, params->windowb_y_start,
+ CRTC_CRC1_WINDOWB_Y_CONTROL,
+ CRTC_CRC1_WINDOWB_Y_START);
+ set_reg_field_value(value, params->windowb_y_end,
+ CRTC_CRC1_WINDOWB_Y_CONTROL,
+ CRTC_CRC1_WINDOWB_Y_END);
+ dm_write_reg(tg->ctx, addr, value);
+
+ /* Set crc mode and selection, and enable.*/
+ value = 0;
+ set_reg_field_value(value, params->continuous_mode ? 1 : 0,
+ CRTC_CRC_CNTL, CRTC_CRC_CONT_EN);
+ set_reg_field_value(value, params->selection,
+ CRTC_CRC_CNTL, CRTC_CRC1_SELECT);
+ set_reg_field_value(value, 1, CRTC_CRC_CNTL, CRTC_CRC_EN);
+ dm_write_reg(tg->ctx, cntl_addr, value);
+ break;
+ default:
+ return false;
+ }
return true;
}
-bool dce110_get_crc(struct timing_generator *tg,
+bool dce110_get_crc(struct timing_generator *tg, uint8_t idx,
uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb)
{
uint32_t addr = 0;
@@ -2206,14 +2267,30 @@ bool dce110_get_crc(struct timing_generator *tg,
if (!field)
return false;
- addr = CRTC_REG(mmCRTC_CRC0_DATA_RG);
- value = dm_read_reg(tg->ctx, addr);
- *r_cr = get_reg_field_value(value, CRTC_CRC0_DATA_RG, CRC0_R_CR);
- *g_y = get_reg_field_value(value, CRTC_CRC0_DATA_RG, CRC0_G_Y);
+ switch (idx) {
+ case 0:
+ addr = CRTC_REG(mmCRTC_CRC0_DATA_RG);
+ value = dm_read_reg(tg->ctx, addr);
+ *r_cr = get_reg_field_value(value, CRTC_CRC0_DATA_RG, CRC0_R_CR);
+ *g_y = get_reg_field_value(value, CRTC_CRC0_DATA_RG, CRC0_G_Y);
- addr = CRTC_REG(mmCRTC_CRC0_DATA_B);
- value = dm_read_reg(tg->ctx, addr);
- *b_cb = get_reg_field_value(value, CRTC_CRC0_DATA_B, CRC0_B_CB);
+ addr = CRTC_REG(mmCRTC_CRC0_DATA_B);
+ value = dm_read_reg(tg->ctx, addr);
+ *b_cb = get_reg_field_value(value, CRTC_CRC0_DATA_B, CRC0_B_CB);
+ break;
+ case 1:
+ addr = CRTC_REG(mmCRTC_CRC1_DATA_RG);
+ value = dm_read_reg(tg->ctx, addr);
+ *r_cr = get_reg_field_value(value, CRTC_CRC1_DATA_RG, CRC1_R_CR);
+ *g_y = get_reg_field_value(value, CRTC_CRC1_DATA_RG, CRC1_G_Y);
+
+ addr = CRTC_REG(mmCRTC_CRC1_DATA_B);
+ value = dm_read_reg(tg->ctx, addr);
+ *b_cb = get_reg_field_value(value, CRTC_CRC1_DATA_B, CRC1_B_CB);
+ break;
+ default:
+ return false;
+ }
return true;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.h b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.h
index ee4de740aceb..e4f5cad64f32 100644
--- a/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.h
+++ b/drivers/gpu/drm/amd/display/dc/dce110/dce110_timing_generator.h
@@ -286,7 +286,7 @@ bool dce110_arm_vert_intr(
bool dce110_configure_crc(struct timing_generator *tg,
const struct crc_params *params);
-bool dce110_get_crc(struct timing_generator *tg,
+bool dce110_get_crc(struct timing_generator *tg, uint8_t idx,
uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb);
bool dce110_is_two_pixels_per_container(const struct dc_crtc_timing *timing);
diff --git a/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
index fcf59348eb62..31c4f44ceaac 100644
--- a/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
+++ b/drivers/gpu/drm/amd/display/dc/dce120/dce120_timing_generator.c
@@ -1100,45 +1100,79 @@ static bool dce120_configure_crc(struct timing_generator *tg,
if (!dce120_is_tg_enabled(tg))
return false;
- /* First, disable CRC before we configure it. */
- dm_write_reg_soc15(tg->ctx, mmCRTC0_CRTC_CRC_CNTL,
- tg110->offsets.crtc, 0);
+ if (!params->enable || params->reset)
+ /* First, disable CRC before we configure it. */
+ dm_write_reg_soc15(tg->ctx, mmCRTC0_CRTC_CRC_CNTL,
+ tg110->offsets.crtc, 0);
if (!params->enable)
return true;
/* Program frame boundaries */
- /* Window A x axis start and end. */
- CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC0_WINDOWA_X_CONTROL,
- CRTC_CRC0_WINDOWA_X_START, params->windowa_x_start,
- CRTC_CRC0_WINDOWA_X_END, params->windowa_x_end);
-
- /* Window A y axis start and end. */
- CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC0_WINDOWA_Y_CONTROL,
- CRTC_CRC0_WINDOWA_Y_START, params->windowa_y_start,
- CRTC_CRC0_WINDOWA_Y_END, params->windowa_y_end);
-
- /* Window B x axis start and end. */
- CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC0_WINDOWB_X_CONTROL,
- CRTC_CRC0_WINDOWB_X_START, params->windowb_x_start,
- CRTC_CRC0_WINDOWB_X_END, params->windowb_x_end);
-
- /* Window B y axis start and end. */
- CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC0_WINDOWB_Y_CONTROL,
- CRTC_CRC0_WINDOWB_Y_START, params->windowb_y_start,
- CRTC_CRC0_WINDOWB_Y_END, params->windowb_y_end);
-
- /* Set crc mode and selection, and enable. Only using CRC0*/
- CRTC_REG_UPDATE_3(CRTC0_CRTC_CRC_CNTL,
- CRTC_CRC_EN, params->continuous_mode ? 1 : 0,
- CRTC_CRC0_SELECT, params->selection,
- CRTC_CRC_EN, 1);
+ switch (params->crc_eng_inst) {
+ case 0:
+ /* Window A x axis start and end. */
+ CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC0_WINDOWA_X_CONTROL,
+ CRTC_CRC0_WINDOWA_X_START, params->windowa_x_start,
+ CRTC_CRC0_WINDOWA_X_END, params->windowa_x_end);
+
+ /* Window A y axis start and end. */
+ CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC0_WINDOWA_Y_CONTROL,
+ CRTC_CRC0_WINDOWA_Y_START, params->windowa_y_start,
+ CRTC_CRC0_WINDOWA_Y_END, params->windowa_y_end);
+
+ /* Window B x axis start and end. */
+ CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC0_WINDOWB_X_CONTROL,
+ CRTC_CRC0_WINDOWB_X_START, params->windowb_x_start,
+ CRTC_CRC0_WINDOWB_X_END, params->windowb_x_end);
+
+ /* Window B y axis start and end. */
+ CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC0_WINDOWB_Y_CONTROL,
+ CRTC_CRC0_WINDOWB_Y_START, params->windowb_y_start,
+ CRTC_CRC0_WINDOWB_Y_END, params->windowb_y_end);
+
+ /* Set crc mode and selection, and enable.*/
+ CRTC_REG_UPDATE_3(CRTC0_CRTC_CRC_CNTL,
+ CRTC_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
+ CRTC_CRC0_SELECT, params->selection,
+ CRTC_CRC_EN, 1);
+ break;
+ case 1:
+ /* Window A x axis start and end. */
+ CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC1_WINDOWA_X_CONTROL,
+ CRTC_CRC1_WINDOWA_X_START, params->windowa_x_start,
+ CRTC_CRC1_WINDOWA_X_END, params->windowa_x_end);
+
+ /* Window A y axis start and end. */
+ CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC1_WINDOWA_Y_CONTROL,
+ CRTC_CRC1_WINDOWA_Y_START, params->windowa_y_start,
+ CRTC_CRC1_WINDOWA_Y_END, params->windowa_y_end);
+
+ /* Window B x axis start and end. */
+ CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC1_WINDOWB_X_CONTROL,
+ CRTC_CRC1_WINDOWB_X_START, params->windowb_x_start,
+ CRTC_CRC1_WINDOWB_X_END, params->windowb_x_end);
+
+ /* Window B y axis start and end. */
+ CRTC_REG_UPDATE_2(CRTC0_CRTC_CRC1_WINDOWB_Y_CONTROL,
+ CRTC_CRC1_WINDOWB_Y_START, params->windowb_y_start,
+ CRTC_CRC1_WINDOWB_Y_END, params->windowb_y_end);
+
+ /* Set crc mode and selection, and enable */
+ CRTC_REG_UPDATE_3(CRTC0_CRTC_CRC_CNTL,
+ CRTC_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
+ CRTC_CRC1_SELECT, params->selection,
+ CRTC_CRC_EN, 1);
+ break;
+ default:
+ return false;
+ }
return true;
}
-static bool dce120_get_crc(struct timing_generator *tg, uint32_t *r_cr,
- uint32_t *g_y, uint32_t *b_cb)
+static bool dce120_get_crc(struct timing_generator *tg, uint8_t idx,
+ uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb)
{
struct dce110_timing_generator *tg110 = DCE110TG_FROM_TG(tg);
uint32_t value, field;
@@ -1151,14 +1185,30 @@ static bool dce120_get_crc(struct timing_generator *tg, uint32_t *r_cr,
if (!field)
return false;
- value = dm_read_reg_soc15(tg->ctx, mmCRTC0_CRTC_CRC0_DATA_RG,
- tg110->offsets.crtc);
- *r_cr = get_reg_field_value(value, CRTC0_CRTC_CRC0_DATA_RG, CRC0_R_CR);
- *g_y = get_reg_field_value(value, CRTC0_CRTC_CRC0_DATA_RG, CRC0_G_Y);
+ switch (idx) {
+ case 0:
+ value = dm_read_reg_soc15(tg->ctx, mmCRTC0_CRTC_CRC0_DATA_RG,
+ tg110->offsets.crtc);
+ *r_cr = get_reg_field_value(value, CRTC0_CRTC_CRC0_DATA_RG, CRC0_R_CR);
+ *g_y = get_reg_field_value(value, CRTC0_CRTC_CRC0_DATA_RG, CRC0_G_Y);
- value = dm_read_reg_soc15(tg->ctx, mmCRTC0_CRTC_CRC0_DATA_B,
- tg110->offsets.crtc);
- *b_cb = get_reg_field_value(value, CRTC0_CRTC_CRC0_DATA_B, CRC0_B_CB);
+ value = dm_read_reg_soc15(tg->ctx, mmCRTC0_CRTC_CRC0_DATA_B,
+ tg110->offsets.crtc);
+ *b_cb = get_reg_field_value(value, CRTC0_CRTC_CRC0_DATA_B, CRC0_B_CB);
+ break;
+ case 1:
+ value = dm_read_reg_soc15(tg->ctx, mmCRTC0_CRTC_CRC1_DATA_RG,
+ tg110->offsets.crtc);
+ *r_cr = get_reg_field_value(value, CRTC0_CRTC_CRC1_DATA_RG, CRC1_R_CR);
+ *g_y = get_reg_field_value(value, CRTC0_CRTC_CRC1_DATA_RG, CRC1_G_Y);
+
+ value = dm_read_reg_soc15(tg->ctx, mmCRTC0_CRTC_CRC1_DATA_B,
+ tg110->offsets.crtc);
+ *b_cb = get_reg_field_value(value, CRTC0_CRTC_CRC1_DATA_B, CRC1_B_CB);
+ break;
+ default:
+ return false;
+ }
return true;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dce60/Makefile b/drivers/gpu/drm/amd/display/dc/dce60/Makefile
index eede83ad91fa..824f73eb3326 100644
--- a/drivers/gpu/drm/amd/display/dc/dce60/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dce60/Makefile
@@ -25,8 +25,7 @@
CFLAGS_$(AMDDALPATH)/dc/dce60/dce60_resource.o = -Wno-override-init
-DCE60 = dce60_timing_generator.o dce60_hw_sequencer.o \
- dce60_resource.o
+DCE60 = dce60_timing_generator.o
AMD_DAL_DCE60 = $(addprefix $(AMDDALPATH)/dc/dce60/,$(DCE60))
diff --git a/drivers/gpu/drm/amd/display/dc/dce60/dce60_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce60/dce60_timing_generator.c
index e5fb0e8333e4..e691a1cf3356 100644
--- a/drivers/gpu/drm/amd/display/dc/dce60/dce60_timing_generator.c
+++ b/drivers/gpu/drm/amd/display/dc/dce60/dce60_timing_generator.c
@@ -239,6 +239,7 @@ static const struct timing_generator_funcs dce60_tg_funcs = {
dce60_timing_generator_enable_advanced_request,
.configure_crc = dce60_configure_crc,
.get_crc = dce110_get_crc,
+ .is_two_pixels_per_container = dce110_is_two_pixels_per_container,
};
void dce60_timing_generator_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.c b/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.c
index 003a9330c286..88e7a1fc9a30 100644
--- a/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.c
+++ b/drivers/gpu/drm/amd/display/dc/dce80/dce80_timing_generator.c
@@ -105,7 +105,7 @@ static void program_pix_dur(struct timing_generator *tg, uint32_t pix_clk_100hz)
dm_write_reg(tg->ctx, addr, value);
}
-static void program_timing(struct timing_generator *tg,
+static void dce80_timing_generator_program_timing(struct timing_generator *tg,
const struct dc_crtc_timing *timing,
int vready_offset,
int vstartup_start,
@@ -185,7 +185,7 @@ static void dce80_timing_generator_enable_advanced_request(
static const struct timing_generator_funcs dce80_tg_funcs = {
.validate_timing = dce110_tg_validate_timing,
- .program_timing = program_timing,
+ .program_timing = dce80_timing_generator_program_timing,
.enable_crtc = dce110_timing_generator_enable_crtc,
.disable_crtc = dce110_timing_generator_disable_crtc,
.is_counter_moving = dce110_timing_generator_is_counter_moving,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c
index 88cf47a5ea75..baf663b661c8 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer_debug.c
@@ -429,7 +429,9 @@ static unsigned int dcn10_get_otg_states(struct dc *dc, char *pBuf, unsigned int
struct dcn_otg_state s = {0};
int pix_clk = 0;
- optc1_read_otg_state(DCN10TG_FROM_TG(tg), &s);
+ if (tg->funcs->read_otg_state)
+ tg->funcs->read_otg_state(tg, &s);
+
pix_clk = dc->current_state->res_ctx.pipe_ctx[i].stream_res.pix_clk_params.requested_pix_clk_100hz / 10;
//only print if OTG master is enabled
@@ -495,7 +497,8 @@ static void dcn10_clear_otpc_underflow(struct dc *dc)
struct timing_generator *tg = pool->timing_generators[i];
struct dcn_otg_state s = {0};
- optc1_read_otg_state(DCN10TG_FROM_TG(tg), &s);
+ if (tg->funcs->read_otg_state)
+ tg->funcs->read_otg_state(tg, &s);
if (s.otg_enabled & 1)
tg->funcs->clear_optc_underflow(tg);
diff --git a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_panel_cntl.c b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_panel_cntl.c
index 573898984726..f9961a6446f3 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_panel_cntl.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn31/dcn31_panel_cntl.c
@@ -168,31 +168,33 @@ void dcn31_panel_cntl_construct(
struct dcn31_panel_cntl *dcn31_panel_cntl,
const struct panel_cntl_init_data *init_data)
{
- uint8_t pwrseq_inst = 0xF;
dcn31_panel_cntl->base.funcs = &dcn31_link_panel_cntl_funcs;
dcn31_panel_cntl->base.ctx = init_data->ctx;
dcn31_panel_cntl->base.inst = init_data->inst;
- switch (init_data->eng_id) {
- case ENGINE_ID_DIGA:
- pwrseq_inst = 0;
- break;
- case ENGINE_ID_DIGB:
- pwrseq_inst = 1;
- break;
- default:
- DC_LOG_WARNING("Unsupported pwrseq engine id: %d!\n", init_data->eng_id);
- ASSERT(false);
- break;
- }
-
- if (dcn31_panel_cntl->base.ctx->dc->config.support_edp0_on_dp1)
+ if (dcn31_panel_cntl->base.ctx->dc->config.support_edp0_on_dp1) {
//If supported, power sequencer mapping shall follow the DIG instance
+ uint8_t pwrseq_inst = 0xF;
+
+ switch (init_data->eng_id) {
+ case ENGINE_ID_DIGA:
+ pwrseq_inst = 0;
+ break;
+ case ENGINE_ID_DIGB:
+ pwrseq_inst = 1;
+ break;
+ default:
+ DC_LOG_WARNING("Unsupported pwrseq engine id: %d!\n", init_data->eng_id);
+ ASSERT(false);
+ break;
+ }
+
dcn31_panel_cntl->base.pwrseq_inst = pwrseq_inst;
- else
+ } else {
/* If not supported, pwrseq will be assigned in order,
* so first pwrseq will be assigned to first panel instance (legacy behavior)
*/
dcn31_panel_cntl->base.pwrseq_inst = dcn31_panel_cntl->base.inst;
+ }
}
diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c
index d01a8b8f9595..22e66b375a7f 100644
--- a/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dio/dcn10/dcn10_stream_encoder.c
@@ -391,7 +391,7 @@ void enc1_stream_encoder_dp_set_stream_attribute(
break;
case COLOR_SPACE_2020_RGB_LIMITEDRANGE:
case COLOR_SPACE_2020_RGB_FULLRANGE:
- case COLOR_SPACE_2020_YCBCR:
+ case COLOR_SPACE_2020_YCBCR_LIMITED:
case COLOR_SPACE_XR_RGB:
case COLOR_SPACE_MSREF_SCRGB:
case COLOR_SPACE_ADOBERGB:
@@ -404,6 +404,7 @@ void enc1_stream_encoder_dp_set_stream_attribute(
case COLOR_SPACE_CUSTOMPOINTS:
case COLOR_SPACE_UNKNOWN:
case COLOR_SPACE_YCBCR709_BLACK:
+ default:
/* do nothing */
break;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.c
index 425b830b88d2..e93be7b6d9b0 100644
--- a/drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.c
@@ -47,7 +47,7 @@
enc1->base.ctx
-static void enc3_update_hdmi_info_packet(
+void enc3_update_hdmi_info_packet(
struct dcn10_stream_encoder *enc1,
uint32_t packet_index,
const struct dc_info_packet *info_packet)
diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.h
index 06310973ded2..830ce7e47035 100644
--- a/drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/dio/dcn30/dcn30_dio_stream_encoder.h
@@ -322,6 +322,10 @@ void enc3_dp_set_dsc_pps_info_packet(
struct stream_encoder *enc,
bool enable,
uint8_t *dsc_packed_pps,
- bool immediate_update);
+ bool immediate_update);
+void enc3_update_hdmi_info_packet(
+ struct dcn10_stream_encoder *enc1,
+ uint32_t packet_index,
+ const struct dc_info_packet *info_packet);
#endif /* __DC_DIO_STREAM_ENCODER_DCN30_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn31/dcn31_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn31/dcn31_dio_link_encoder.c
index b2cea59ba5d4..9a92f73d5b7f 100644
--- a/drivers/gpu/drm/amd/display/dc/dio/dcn31/dcn31_dio_link_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dio/dcn31/dcn31_dio_link_encoder.c
@@ -653,8 +653,9 @@ void dcn31_link_encoder_get_max_link_cap(struct link_encoder *enc, struct dc_lin
if (!query_dp_alt_from_dmub(enc, &cmd))
return;
- if (cmd.query_dp_alt.data.is_usb &&
- cmd.query_dp_alt.data.is_dp4 == 0)
+ if (cmd.query_dp_alt.data.is_dp_alt_disable == 0 &&
+ cmd.query_dp_alt.data.is_usb &&
+ cmd.query_dp_alt.data.is_dp4 == 0)
link_settings->lane_count = MIN(LANE_COUNT_TWO, link_settings->lane_count);
return;
diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.c
index d4a3e811aa39..9972911330b6 100644
--- a/drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.c
@@ -28,6 +28,7 @@
#include "link_encoder.h"
#include "dcn31/dcn31_dio_link_encoder.h"
#include "dcn35_dio_link_encoder.h"
+#include "dc_dmub_srv.h"
#define CTX \
enc10->base.ctx
#define DC_LOGGER \
@@ -136,9 +137,9 @@ static const struct link_encoder_funcs dcn35_link_enc_funcs = {
.hw_init = dcn35_link_encoder_init,
.setup = dcn35_link_encoder_setup,
.enable_tmds_output = dcn10_link_encoder_enable_tmds_output,
- .enable_dp_output = dcn31_link_encoder_enable_dp_output,
- .enable_dp_mst_output = dcn31_link_encoder_enable_dp_mst_output,
- .disable_output = dcn31_link_encoder_disable_output,
+ .enable_dp_output = dcn35_link_encoder_enable_dp_output,
+ .enable_dp_mst_output = dcn35_link_encoder_enable_dp_mst_output,
+ .disable_output = dcn35_link_encoder_disable_output,
.dp_set_lane_settings = dcn10_link_encoder_dp_set_lane_settings,
.dp_set_phy_pattern = dcn10_link_encoder_dp_set_phy_pattern,
.update_mst_stream_allocation_table =
@@ -159,6 +160,8 @@ static const struct link_encoder_funcs dcn35_link_enc_funcs = {
.is_in_alt_mode = dcn31_link_encoder_is_in_alt_mode,
.get_max_link_cap = dcn31_link_encoder_get_max_link_cap,
.set_dio_phy_mux = dcn31_link_encoder_set_dio_phy_mux,
+ .enable_dpia_output = dcn35_link_encoder_enable_dpia_output,
+ .disable_dpia_output = dcn35_link_encoder_disable_dpia_output,
};
void dcn35_link_encoder_construct(
@@ -265,3 +268,124 @@ void dcn35_link_encoder_construct(
enc10->base.features.flags.bits.HDMI_6GB_EN = 0;
}
+
+/* DPIA equivalent of link_transmitter_control. */
+static bool link_dpia_control(struct dc_context *dc_ctx,
+ struct dmub_cmd_dig_dpia_control_data *dpia_control)
+{
+ union dmub_rb_cmd cmd;
+
+ memset(&cmd, 0, sizeof(cmd));
+
+ cmd.dig1_dpia_control.header.type = DMUB_CMD__DPIA;
+ cmd.dig1_dpia_control.header.sub_type =
+ DMUB_CMD__DPIA_DIG1_DPIA_CONTROL;
+ cmd.dig1_dpia_control.header.payload_bytes =
+ sizeof(cmd.dig1_dpia_control) -
+ sizeof(cmd.dig1_dpia_control.header);
+
+ cmd.dig1_dpia_control.dpia_control = *dpia_control;
+
+ dc_wake_and_execute_dmub_cmd(dc_ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
+
+ return true;
+}
+
+static void link_encoder_disable(struct dcn10_link_encoder *enc10)
+{
+ /* reset training complete */
+ REG_UPDATE(DP_LINK_CNTL, DP_LINK_TRAINING_COMPLETE, 0);
+}
+
+void dcn35_link_encoder_enable_dp_output(
+ struct link_encoder *enc,
+ const struct dc_link_settings *link_settings,
+ enum clock_source_id clock_source)
+{
+ struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+ if (!enc->ctx->dc->config.unify_link_enc_assignment)
+ dcn31_link_encoder_enable_dp_output(enc, link_settings, clock_source);
+ else {
+ DC_LOG_DEBUG("%s: enc_id(%d)\n", __func__, enc->preferred_engine);
+ dcn20_link_encoder_enable_dp_output(enc, link_settings, clock_source);
+ }
+}
+
+void dcn35_link_encoder_enable_dp_mst_output(
+ struct link_encoder *enc,
+ const struct dc_link_settings *link_settings,
+ enum clock_source_id clock_source)
+{
+ struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+ if (!enc->ctx->dc->config.unify_link_enc_assignment)
+ dcn31_link_encoder_enable_dp_mst_output(enc, link_settings, clock_source);
+ else {
+ DC_LOG_DEBUG("%s: enc_id(%d)\n", __func__, enc->preferred_engine);
+ dcn10_link_encoder_enable_dp_mst_output(enc, link_settings, clock_source);
+ }
+}
+
+void dcn35_link_encoder_disable_output(
+ struct link_encoder *enc,
+ enum signal_type signal)
+{
+ struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+
+ if (!enc->ctx->dc->config.unify_link_enc_assignment)
+ dcn31_link_encoder_disable_output(enc, signal);
+ else {
+ DC_LOG_DEBUG("%s: enc_id(%d)\n", __func__, enc->preferred_engine);
+ dcn10_link_encoder_disable_output(enc, signal);
+ }
+}
+
+void dcn35_link_encoder_enable_dpia_output(
+ struct link_encoder *enc,
+ const struct dc_link_settings *link_settings,
+ uint8_t dpia_id,
+ uint8_t digmode,
+ uint8_t fec_rdy)
+{
+ struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+ struct dmub_cmd_dig_dpia_control_data dpia_control = { 0 };
+
+ enc1_configure_encoder(enc10, link_settings);
+
+ dpia_control.action = (uint8_t)TRANSMITTER_CONTROL_ENABLE;
+ dpia_control.enc_id = enc->preferred_engine;
+ dpia_control.mode_laneset.digmode = digmode;
+ dpia_control.lanenum = (uint8_t)link_settings->lane_count;
+ dpia_control.symclk_10khz = link_settings->link_rate *
+ LINK_RATE_REF_FREQ_IN_KHZ / 10;
+ /* DIG_BE_CNTL.DIG_HPD_SELECT set to 5 (hpdsel - 1) to indicate HPD pin unused by DPIA. */
+ dpia_control.hpdsel = 6;
+ dpia_control.dpia_id = dpia_id;
+ dpia_control.fec_rdy = fec_rdy;
+
+ DC_LOG_DEBUG("%s: DPIA(%d) - enc_id(%d)\n", __func__, dpia_control.dpia_id, dpia_control.enc_id);
+ link_dpia_control(enc->ctx, &dpia_control);
+}
+
+void dcn35_link_encoder_disable_dpia_output(
+ struct link_encoder *enc,
+ uint8_t dpia_id,
+ uint8_t digmode)
+{
+ struct dcn10_link_encoder *enc10 = TO_DCN10_LINK_ENC(enc);
+ struct dmub_cmd_dig_dpia_control_data dpia_control = { 0 };
+
+ if (enc->funcs->is_dig_enabled && !enc->funcs->is_dig_enabled(enc))
+ return;
+
+ dpia_control.action = (uint8_t)TRANSMITTER_CONTROL_DISABLE;
+ dpia_control.enc_id = enc->preferred_engine;
+ dpia_control.mode_laneset.digmode = digmode;
+ dpia_control.dpia_id = dpia_id;
+
+ DC_LOG_DEBUG("%s: DPIA(%d) - enc_id(%d)\n", __func__, dpia_control.dpia_id, dpia_control.enc_id);
+ link_dpia_control(enc->ctx, &dpia_control);
+
+ link_encoder_disable(enc10);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.h b/drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.h
index d546a3676304..5712e6553fab 100644
--- a/drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/dio/dcn35/dcn35_dio_link_encoder.h
@@ -144,4 +144,45 @@ bool dcn35_is_dig_enabled(struct link_encoder *enc);
enum signal_type dcn35_get_dig_mode(struct link_encoder *enc);
void dcn35_link_encoder_setup(struct link_encoder *enc, enum signal_type signal);
+/*
+ * Enable DP transmitter and its encoder.
+ */
+void dcn35_link_encoder_enable_dp_output(
+ struct link_encoder *enc,
+ const struct dc_link_settings *link_settings,
+ enum clock_source_id clock_source);
+
+/*
+ * Enable DP transmitter and its encoder in MST mode.
+ */
+void dcn35_link_encoder_enable_dp_mst_output(
+ struct link_encoder *enc,
+ const struct dc_link_settings *link_settings,
+ enum clock_source_id clock_source);
+
+/*
+ * Disable transmitter and its encoder.
+ */
+void dcn35_link_encoder_disable_output(
+ struct link_encoder *enc,
+ enum signal_type signal);
+
+/*
+ * Enable DP transmitter and its encoder for dpia port.
+ */
+void dcn35_link_encoder_enable_dpia_output(
+ struct link_encoder *enc,
+ const struct dc_link_settings *link_settings,
+ uint8_t dpia_id,
+ uint8_t digmode,
+ uint8_t fec_rdy);
+
+/*
+ * Disable transmitter and its encoder for dpia port.
+ */
+void dcn35_link_encoder_disable_dpia_output(
+ struct link_encoder *enc,
+ uint8_t dpia_id,
+ uint8_t digmode);
+
#endif /* __DC_LINK_ENCODER__DCN35_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c
index 098c2a01a850..d5fa551dd3c9 100644
--- a/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.c
@@ -60,7 +60,7 @@ static void enc401_dp_set_odm_combine(
}
/* setup stream encoder in dvi mode */
-static void enc401_stream_encoder_dvi_set_stream_attribute(
+void enc401_stream_encoder_dvi_set_stream_attribute(
struct stream_encoder *enc,
struct dc_crtc_timing *crtc_timing,
bool is_dual_link)
@@ -100,7 +100,7 @@ static void enc401_stream_encoder_dvi_set_stream_attribute(
}
/* setup stream encoder in hdmi mode */
-static void enc401_stream_encoder_hdmi_set_stream_attribute(
+void enc401_stream_encoder_hdmi_set_stream_attribute(
struct stream_encoder *enc,
struct dc_crtc_timing *crtc_timing,
int actual_pix_clk_khz,
@@ -229,7 +229,7 @@ static void enc401_stream_encoder_hdmi_set_stream_attribute(
REG_UPDATE(HDMI_GC, HDMI_GC_AVMUTE, 0);
}
-static void enc401_set_dig_input_mode(struct stream_encoder *enc, unsigned int pix_per_container)
+void enc401_set_dig_input_mode(struct stream_encoder *enc, unsigned int pix_per_container)
{
struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
@@ -260,7 +260,7 @@ static bool is_two_pixels_per_containter(const struct dc_crtc_timing *timing)
return two_pix;
}
-static void enc401_stream_encoder_dp_unblank(
+void enc401_stream_encoder_dp_unblank(
struct dc_link *link,
struct stream_encoder *enc,
const struct encoder_unblank_param *param)
@@ -376,7 +376,7 @@ static void enc401_stream_encoder_dp_unblank(
/* this function read dsc related register fields to be logged later in dcn10_log_hw_state
* into a dcn_dsc_state struct.
*/
-static void enc401_read_state(struct stream_encoder *enc, struct enc_state *s)
+void enc401_read_state(struct stream_encoder *enc, struct enc_state *s)
{
struct dcn10_stream_encoder *enc1 = DCN10STRENC_FROM_STRENC(enc);
@@ -394,7 +394,7 @@ static void enc401_read_state(struct stream_encoder *enc, struct enc_state *s)
}
}
-static void enc401_stream_encoder_enable(
+void enc401_stream_encoder_enable(
struct stream_encoder *enc,
enum signal_type signal,
bool enable)
@@ -632,7 +632,7 @@ void enc401_stream_encoder_dp_set_stream_attribute(
break;
case COLOR_SPACE_2020_RGB_LIMITEDRANGE:
case COLOR_SPACE_2020_RGB_FULLRANGE:
- case COLOR_SPACE_2020_YCBCR:
+ case COLOR_SPACE_2020_YCBCR_LIMITED:
case COLOR_SPACE_XR_RGB:
case COLOR_SPACE_MSREF_SCRGB:
case COLOR_SPACE_ADOBERGB:
@@ -645,6 +645,7 @@ void enc401_stream_encoder_dp_set_stream_attribute(
case COLOR_SPACE_CUSTOMPOINTS:
case COLOR_SPACE_UNKNOWN:
case COLOR_SPACE_YCBCR709_BLACK:
+ default:
/* do nothing */
break;
}
@@ -704,7 +705,7 @@ void enc401_stream_encoder_dp_set_stream_attribute(
DP_SST_SDP_SPLITTING, enable_sdp_splitting);
}
-static void enc401_stream_encoder_map_to_link(
+void enc401_stream_encoder_map_to_link(
struct stream_encoder *enc,
uint32_t stream_enc_inst,
uint32_t link_enc_inst)
diff --git a/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.h b/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.h
index d751839598f8..d6b00cd246b1 100644
--- a/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/dio/dcn401/dcn401_dio_stream_encoder.h
@@ -214,4 +214,27 @@ void enc401_stream_encoder_dp_set_stream_attribute(
enum dc_color_space output_color_space,
bool use_vsc_sdp_for_colorimetry,
uint32_t enable_sdp_splitting);
+void enc401_stream_encoder_dvi_set_stream_attribute(
+ struct stream_encoder *enc,
+ struct dc_crtc_timing *crtc_timing,
+ bool is_dual_link);
+void enc401_stream_encoder_dp_unblank(
+ struct dc_link *link,
+ struct stream_encoder *enc,
+ const struct encoder_unblank_param *param);
+void enc401_stream_encoder_enable(
+ struct stream_encoder *enc,
+ enum signal_type signal,
+ bool enable);
+void enc401_set_dig_input_mode(struct stream_encoder *enc, unsigned int pix_per_container);
+void enc401_stream_encoder_map_to_link(
+ struct stream_encoder *enc,
+ uint32_t stream_enc_inst,
+ uint32_t link_enc_inst);
+void enc401_read_state(struct stream_encoder *enc, struct enc_state *s);
+void enc401_stream_encoder_hdmi_set_stream_attribute(
+ struct stream_encoder *enc,
+ struct dc_crtc_timing *crtc_timing,
+ int actual_pix_clk_khz,
+ bool enable_audio);
#endif /* __DC_DIO_STREAM_ENCODER_DCN401_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dm_helpers.h b/drivers/gpu/drm/amd/display/dc/dm_helpers.h
index 2e4a46f1b499..9d160b39e8c5 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_helpers.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_helpers.h
@@ -153,11 +153,24 @@ bool dm_helpers_submit_i2c(
const struct dc_link *link,
struct i2c_command *cmd);
+bool dm_helpers_execute_fused_io(
+ struct dc_context *ctx,
+ struct dc_link *link,
+ union dmub_rb_cmd *commands,
+ uint8_t count,
+ uint32_t timeout_us
+);
+
bool dm_helpers_dp_write_dsc_enable(
struct dc_context *ctx,
const struct dc_stream_state *stream,
bool enable
);
+
+bool dm_helpers_dp_write_hblank_reduction(
+ struct dc_context *ctx,
+ const struct dc_stream_state *stream);
+
bool dm_helpers_is_dp_sink_present(
struct dc_link *link);
diff --git a/drivers/gpu/drm/amd/display/dc/dm_services.h b/drivers/gpu/drm/amd/display/dc/dm_services.h
index f81e5a4e1d6d..7b9c22c45453 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_services.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_services.h
@@ -291,6 +291,13 @@ bool dm_execute_dmub_cmd(const struct dc_context *ctx, union dmub_rb_cmd *cmd, e
bool dm_execute_dmub_cmd_list(const struct dc_context *ctx, unsigned int count, union dmub_rb_cmd *cmd, enum dm_dmub_wait_type wait_type);
/*
+ * ACPI Interfaces
+ */
+void dm_acpi_process_phy_transition_interlock(
+ const struct dc_context *ctx,
+ struct dm_process_phy_transition_init_params process_phy_transition_init_params);
+
+/*
* Debug and verification hooks
*/
diff --git a/drivers/gpu/drm/amd/display/dc/dm_services_types.h b/drivers/gpu/drm/amd/display/dc/dm_services_types.h
index facf269c4326..bf63da266a18 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_services_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_services_types.h
@@ -275,4 +275,30 @@ enum dm_dmub_wait_type {
DM_DMUB_WAIT_TYPE_WAIT_WITH_REPLY,
};
+enum dm_acpi_transition_link_type {
+ hdmi_tmds,
+ hdmi_frl,
+ dp_8b_10b,
+ dp_128b_132b,
+ none
+};
+
+struct dm_process_phy_transition_init_params {
+ uint32_t phy_id;
+ uint8_t action;
+ uint32_t sym_clock_10khz;
+ enum signal_type signal;
+ enum dc_lane_count display_port_lanes_count;
+ enum dc_link_rate display_port_link_rate;
+ uint32_t transition_bitmask;
+ uint8_t hdmi_frl_num_lanes;
+};
+
+struct dm_process_phy_transition_input_params {
+ uint32_t phy_id;
+ uint32_t transition_id;
+ uint32_t phy_configuration;
+ uint32_t data_rate;
+};
+
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile
index 46f9c05de16e..b357683b4255 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile
@@ -29,11 +29,15 @@ dml_ccflags := $(CC_FLAGS_FPU)
dml_rcflags := $(CC_FLAGS_NO_FPU)
ifneq ($(CONFIG_FRAME_WARN),0)
-ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)),y)
-frame_warn_flag := -Wframe-larger-than=3072
-else
-frame_warn_flag := -Wframe-larger-than=2048
-endif
+ ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)),y)
+ frame_warn_limit := 3072
+ else
+ frame_warn_limit := 2048
+ endif
+
+ ifeq ($(call test-lt, $(CONFIG_FRAME_WARN), $(frame_warn_limit)),y)
+ frame_warn_flag := -Wframe-larger-than=$(frame_warn_limit)
+ endif
endif
CFLAGS_$(AMDDALPATH)/dc/dml/display_mode_lib.o := $(dml_ccflags)
@@ -110,9 +114,6 @@ CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/calcs/dcn_calcs.o := $(dml_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_auto.o := $(dml_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/calcs/dcn_calc_math.o := $(dml_rcflags)
-CFLAGS_$(AMDDALPATH)/dc/dml/dcn401/dcn401_fpu.o := $(dml_ccflags)
-CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml/dcn401/dcn401_fpu.o := $(dml_rcflags)
-
ifdef CONFIG_DRM_AMD_DC_FP
DML += display_mode_lib.o display_rq_dlg_helpers.o dml1_display_rq_dlg_calc.o
DML += dcn10/dcn10_fpu.o
@@ -133,7 +134,6 @@ DML += dcn303/dcn303_fpu.o
DML += dcn314/dcn314_fpu.o
DML += dcn35/dcn35_fpu.o
DML += dcn351/dcn351_fpu.o
-DML += dcn401/dcn401_fpu.o
DML += dsc/rc_calc_fpu.o
DML += calcs/dcn_calcs.o calcs/dcn_calc_math.o calcs/dcn_calc_auto.o
endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c b/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c
index 39525721c976..74962791302f 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/calcs/dcn_calcs.c
@@ -748,7 +748,7 @@ static unsigned int get_highest_allowed_voltage_level(bool is_vmin_only_asic)
bool dcn_validate_bandwidth(
struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
/*
* we want a breakdown of the various stages of validation, which the
@@ -1119,7 +1119,7 @@ bool dcn_validate_bandwidth(
BW_VAL_TRACE_END_VOLTAGE_LEVEL();
- if (v->voltage_level != number_of_states_plus_one && !fast_validate) {
+ if (v->voltage_level != number_of_states_plus_one && validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING) {
float bw_consumed = v->total_bandwidth_consumed_gbyte_per_second;
if (bw_consumed < v->fabric_and_dram_bandwidth_vmin0p65)
@@ -1286,7 +1286,7 @@ bool dcn_validate_bandwidth(
}
} else if (v->voltage_level == number_of_states_plus_one) {
BW_VAL_TRACE_SKIP(fail);
- } else if (fast_validate) {
+ } else if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING) {
BW_VAL_TRACE_SKIP(fast);
}
@@ -1312,138 +1312,6 @@ bool dcn_validate_bandwidth(
return false;
}
-static unsigned int dcn_find_normalized_clock_vdd_Level(
- const struct dc *dc,
- enum dm_pp_clock_type clocks_type,
- int clocks_in_khz)
-{
- int vdd_level = dcn_bw_v_min0p65;
-
- if (clocks_in_khz == 0)/*todo some clock not in the considerations*/
- return vdd_level;
-
- switch (clocks_type) {
- case DM_PP_CLOCK_TYPE_DISPLAY_CLK:
- if (clocks_in_khz > dc->dcn_soc->max_dispclk_vmax0p9*1000) {
- vdd_level = dcn_bw_v_max0p91;
- BREAK_TO_DEBUGGER();
- } else if (clocks_in_khz > dc->dcn_soc->max_dispclk_vnom0p8*1000) {
- vdd_level = dcn_bw_v_max0p9;
- } else if (clocks_in_khz > dc->dcn_soc->max_dispclk_vmid0p72*1000) {
- vdd_level = dcn_bw_v_nom0p8;
- } else if (clocks_in_khz > dc->dcn_soc->max_dispclk_vmin0p65*1000) {
- vdd_level = dcn_bw_v_mid0p72;
- } else
- vdd_level = dcn_bw_v_min0p65;
- break;
- case DM_PP_CLOCK_TYPE_DISPLAYPHYCLK:
- if (clocks_in_khz > dc->dcn_soc->phyclkv_max0p9*1000) {
- vdd_level = dcn_bw_v_max0p91;
- BREAK_TO_DEBUGGER();
- } else if (clocks_in_khz > dc->dcn_soc->phyclkv_nom0p8*1000) {
- vdd_level = dcn_bw_v_max0p9;
- } else if (clocks_in_khz > dc->dcn_soc->phyclkv_mid0p72*1000) {
- vdd_level = dcn_bw_v_nom0p8;
- } else if (clocks_in_khz > dc->dcn_soc->phyclkv_min0p65*1000) {
- vdd_level = dcn_bw_v_mid0p72;
- } else
- vdd_level = dcn_bw_v_min0p65;
- break;
-
- case DM_PP_CLOCK_TYPE_DPPCLK:
- if (clocks_in_khz > dc->dcn_soc->max_dppclk_vmax0p9*1000) {
- vdd_level = dcn_bw_v_max0p91;
- BREAK_TO_DEBUGGER();
- } else if (clocks_in_khz > dc->dcn_soc->max_dppclk_vnom0p8*1000) {
- vdd_level = dcn_bw_v_max0p9;
- } else if (clocks_in_khz > dc->dcn_soc->max_dppclk_vmid0p72*1000) {
- vdd_level = dcn_bw_v_nom0p8;
- } else if (clocks_in_khz > dc->dcn_soc->max_dppclk_vmin0p65*1000) {
- vdd_level = dcn_bw_v_mid0p72;
- } else
- vdd_level = dcn_bw_v_min0p65;
- break;
-
- case DM_PP_CLOCK_TYPE_MEMORY_CLK:
- {
- unsigned factor = (ddr4_dram_factor_single_Channel * dc->dcn_soc->number_of_channels);
-
- if (clocks_in_khz > dc->dcn_soc->fabric_and_dram_bandwidth_vmax0p9*1000000/factor) {
- vdd_level = dcn_bw_v_max0p91;
- BREAK_TO_DEBUGGER();
- } else if (clocks_in_khz > dc->dcn_soc->fabric_and_dram_bandwidth_vnom0p8*1000000/factor) {
- vdd_level = dcn_bw_v_max0p9;
- } else if (clocks_in_khz > dc->dcn_soc->fabric_and_dram_bandwidth_vmid0p72*1000000/factor) {
- vdd_level = dcn_bw_v_nom0p8;
- } else if (clocks_in_khz > dc->dcn_soc->fabric_and_dram_bandwidth_vmin0p65*1000000/factor) {
- vdd_level = dcn_bw_v_mid0p72;
- } else
- vdd_level = dcn_bw_v_min0p65;
- }
- break;
-
- case DM_PP_CLOCK_TYPE_DCFCLK:
- if (clocks_in_khz > dc->dcn_soc->dcfclkv_max0p9*1000) {
- vdd_level = dcn_bw_v_max0p91;
- BREAK_TO_DEBUGGER();
- } else if (clocks_in_khz > dc->dcn_soc->dcfclkv_nom0p8*1000) {
- vdd_level = dcn_bw_v_max0p9;
- } else if (clocks_in_khz > dc->dcn_soc->dcfclkv_mid0p72*1000) {
- vdd_level = dcn_bw_v_nom0p8;
- } else if (clocks_in_khz > dc->dcn_soc->dcfclkv_min0p65*1000) {
- vdd_level = dcn_bw_v_mid0p72;
- } else
- vdd_level = dcn_bw_v_min0p65;
- break;
-
- default:
- break;
- }
- return vdd_level;
-}
-
-unsigned int dcn_find_dcfclk_suits_all(
- const struct dc *dc,
- struct dc_clocks *clocks)
-{
- unsigned vdd_level, vdd_level_temp;
- unsigned dcf_clk;
-
- /*find a common supported voltage level*/
- vdd_level = dcn_find_normalized_clock_vdd_Level(
- dc, DM_PP_CLOCK_TYPE_DISPLAY_CLK, clocks->dispclk_khz);
- vdd_level_temp = dcn_find_normalized_clock_vdd_Level(
- dc, DM_PP_CLOCK_TYPE_DISPLAYPHYCLK, clocks->phyclk_khz);
-
- vdd_level = dcn_bw_max(vdd_level, vdd_level_temp);
- vdd_level_temp = dcn_find_normalized_clock_vdd_Level(
- dc, DM_PP_CLOCK_TYPE_DPPCLK, clocks->dppclk_khz);
- vdd_level = dcn_bw_max(vdd_level, vdd_level_temp);
-
- vdd_level_temp = dcn_find_normalized_clock_vdd_Level(
- dc, DM_PP_CLOCK_TYPE_MEMORY_CLK, clocks->fclk_khz);
- vdd_level = dcn_bw_max(vdd_level, vdd_level_temp);
- vdd_level_temp = dcn_find_normalized_clock_vdd_Level(
- dc, DM_PP_CLOCK_TYPE_DCFCLK, clocks->dcfclk_khz);
-
- /*find that level conresponding dcfclk*/
- vdd_level = dcn_bw_max(vdd_level, vdd_level_temp);
- if (vdd_level == dcn_bw_v_max0p91) {
- BREAK_TO_DEBUGGER();
- dcf_clk = dc->dcn_soc->dcfclkv_max0p9*1000;
- } else if (vdd_level == dcn_bw_v_max0p9)
- dcf_clk = dc->dcn_soc->dcfclkv_max0p9*1000;
- else if (vdd_level == dcn_bw_v_nom0p8)
- dcf_clk = dc->dcn_soc->dcfclkv_nom0p8*1000;
- else if (vdd_level == dcn_bw_v_mid0p72)
- dcf_clk = dc->dcn_soc->dcfclkv_mid0p72*1000;
- else
- dcf_clk = dc->dcn_soc->dcfclkv_min0p65*1000;
-
- DC_LOG_BANDWIDTH_CALCS("\tdcf_clk for voltage = %d\n", dcf_clk);
- return dcf_clk;
-}
-
void dcn_bw_update_from_pplib_fclks(
struct dc *dc,
struct dm_pp_clock_levels_with_voltage *fclks)
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
index e9fea9c2162e..2a2eaf6adf26 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.c
@@ -1315,7 +1315,7 @@ static void swizzle_to_dml_params(
int dcn20_populate_dml_pipes_from_context(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int pipe_cnt, i;
bool synchronized_vblank = true;
@@ -1733,7 +1733,7 @@ void dcn20_calculate_wm(struct dc *dc, struct dc_state *context,
int *out_pipe_cnt,
int *pipe_split_from,
int vlevel,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int pipe_cnt, i, pipe_idx;
@@ -1780,10 +1780,10 @@ void dcn20_calculate_wm(struct dc *dc, struct dc_state *context,
if (pipe_cnt != pipe_idx) {
if (dc->res_pool->funcs->populate_dml_pipes)
pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc,
- context, pipes, fast_validate);
+ context, pipes, validate_mode);
else
pipe_cnt = dcn20_populate_dml_pipes_from_context(dc,
- context, pipes, fast_validate);
+ context, pipes, validate_mode);
}
*out_pipe_cnt = pipe_cnt;
@@ -2027,7 +2027,7 @@ void dcn20_patch_bounding_box(struct dc *dc, struct _vcs_dpi_soc_bounding_box_st
}
static bool dcn20_validate_bandwidth_internal(struct dc *dc, struct dc_state *context,
- bool fast_validate, display_e2e_pipe_params_st *pipes)
+ enum dc_validate_mode validate_mode, display_e2e_pipe_params_st *pipes)
{
bool out = false;
@@ -2040,7 +2040,7 @@ static bool dcn20_validate_bandwidth_internal(struct dc *dc, struct dc_state *co
BW_VAL_TRACE_COUNT();
- out = dcn20_fast_validate_bw(dc, context, pipes, &pipe_cnt, pipe_split_from, &vlevel, fast_validate);
+ out = dcn20_fast_validate_bw(dc, context, pipes, &pipe_cnt, pipe_split_from, &vlevel, validate_mode);
if (pipe_cnt == 0)
goto validate_out;
@@ -2050,12 +2050,12 @@ static bool dcn20_validate_bandwidth_internal(struct dc *dc, struct dc_state *co
BW_VAL_TRACE_END_VOLTAGE_LEVEL();
- if (fast_validate) {
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING) {
BW_VAL_TRACE_SKIP(fast);
goto validate_out;
}
- dcn20_calculate_wm(dc, context, pipes, &pipe_cnt, pipe_split_from, vlevel, fast_validate);
+ dcn20_calculate_wm(dc, context, pipes, &pipe_cnt, pipe_split_from, vlevel, validate_mode);
dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel);
BW_VAL_TRACE_END_WATERMARKS();
@@ -2077,7 +2077,7 @@ validate_out:
}
bool dcn20_validate_bandwidth_fp(struct dc *dc, struct dc_state *context,
- bool fast_validate, display_e2e_pipe_params_st *pipes)
+ enum dc_validate_mode validate_mode, display_e2e_pipe_params_st *pipes)
{
bool voltage_supported = false;
bool full_pstate_supported = false;
@@ -2095,12 +2095,11 @@ bool dcn20_validate_bandwidth_fp(struct dc *dc, struct dc_state *context,
/*Unsafe due to current pipe merge and split logic*/
ASSERT(context != dc->current_state);
- if (fast_validate) {
- return dcn20_validate_bandwidth_internal(dc, context, true, pipes);
- }
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING)
+ return dcn20_validate_bandwidth_internal(dc, context, validate_mode, pipes);
// Best case, we support full UCLK switch latency
- voltage_supported = dcn20_validate_bandwidth_internal(dc, context, false, pipes);
+ voltage_supported = dcn20_validate_bandwidth_internal(dc, context, DC_VALIDATE_MODE_AND_PROGRAMMING, pipes);
full_pstate_supported = context->bw_ctx.bw.dcn.clk.p_state_change_support;
if (context->bw_ctx.dml.soc.dummy_pstate_latency_us == 0 ||
@@ -2113,7 +2112,7 @@ bool dcn20_validate_bandwidth_fp(struct dc *dc, struct dc_state *context,
context->bw_ctx.dml.soc.dram_clock_change_latency_us = context->bw_ctx.dml.soc.dummy_pstate_latency_us;
memset(pipes, 0, dc->res_pool->pipe_count * sizeof(display_e2e_pipe_params_st));
- voltage_supported = dcn20_validate_bandwidth_internal(dc, context, false, pipes);
+ voltage_supported = dcn20_validate_bandwidth_internal(dc, context, DC_VALIDATE_MODE_AND_PROGRAMMING, pipes);
dummy_pstate_supported = context->bw_ctx.bw.dcn.clk.p_state_change_support;
if (voltage_supported && (dummy_pstate_supported || !(context->stream_count))) {
@@ -2156,14 +2155,14 @@ void dcn20_fpu_adjust_dppclk(struct vba_vars_st *v,
int dcn21_populate_dml_pipes_from_context(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
uint32_t pipe_cnt;
int i;
dc_assert_fp_enabled();
- pipe_cnt = dcn20_populate_dml_pipes_from_context(dc, context, pipes, fast_validate);
+ pipe_cnt = dcn20_populate_dml_pipes_from_context(dc, context, pipes, validate_mode);
for (i = 0; i < pipe_cnt; i++) {
@@ -2239,7 +2238,7 @@ static void dcn21_calculate_wm(struct dc *dc, struct dc_state *context,
int *out_pipe_cnt,
int *pipe_split_from,
int vlevel_req,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int pipe_cnt, i, pipe_idx;
int vlevel, vlevel_max;
@@ -2281,10 +2280,10 @@ static void dcn21_calculate_wm(struct dc *dc, struct dc_state *context,
if (pipe_cnt != pipe_idx) {
if (dc->res_pool->funcs->populate_dml_pipes)
pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc,
- context, pipes, fast_validate);
+ context, pipes, validate_mode);
else
pipe_cnt = dcn21_populate_dml_pipes_from_context(dc,
- context, pipes, fast_validate);
+ context, pipes, validate_mode);
}
*out_pipe_cnt = pipe_cnt;
@@ -2319,7 +2318,7 @@ static void dcn21_calculate_wm(struct dc *dc, struct dc_state *context,
}
bool dcn21_validate_bandwidth_fp(struct dc *dc, struct dc_state *context,
- bool fast_validate, display_e2e_pipe_params_st *pipes)
+ enum dc_validate_mode validate_mode, display_e2e_pipe_params_st *pipes)
{
bool out = false;
@@ -2337,7 +2336,7 @@ bool dcn21_validate_bandwidth_fp(struct dc *dc, struct dc_state *context,
/*Unsafe due to current pipe merge and split logic*/
ASSERT(context != dc->current_state);
- out = dcn21_fast_validate_bw(dc, context, pipes, &pipe_cnt, pipe_split_from, &vlevel, fast_validate);
+ out = dcn21_fast_validate_bw(dc, context, pipes, &pipe_cnt, pipe_split_from, &vlevel, validate_mode);
if (pipe_cnt == 0)
goto validate_out;
@@ -2347,12 +2346,12 @@ bool dcn21_validate_bandwidth_fp(struct dc *dc, struct dc_state *context,
BW_VAL_TRACE_END_VOLTAGE_LEVEL();
- if (fast_validate) {
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING) {
BW_VAL_TRACE_SKIP(fast);
goto validate_out;
}
- dcn21_calculate_wm(dc, context, pipes, &pipe_cnt, pipe_split_from, vlevel, fast_validate);
+ dcn21_calculate_wm(dc, context, pipes, &pipe_cnt, pipe_split_from, vlevel, validate_mode);
dcn20_calculate_dlg_params(dc, context, pipes, pipe_cnt, vlevel);
BW_VAL_TRACE_END_WATERMARKS();
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.h
index b6c34198ddc8..aed00039ca62 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn20/dcn20_fpu.h
@@ -44,14 +44,14 @@ void dcn20_calculate_dlg_params(struct dc *dc,
int dcn20_populate_dml_pipes_from_context(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void dcn20_calculate_wm(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
int *out_pipe_cnt,
int *pipe_split_from,
int vlevel,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void dcn20_cap_soc_clocks(struct _vcs_dpi_soc_bounding_box_st *bb,
struct pp_smu_nv_clock_table max_clocks);
void dcn20_update_bounding_box(struct dc *dc,
@@ -62,7 +62,7 @@ void dcn20_update_bounding_box(struct dc *dc,
void dcn20_patch_bounding_box(struct dc *dc,
struct _vcs_dpi_soc_bounding_box_st *bb);
bool dcn20_validate_bandwidth_fp(struct dc *dc, struct dc_state *context,
- bool fast_validate, display_e2e_pipe_params_st *pipes);
+ enum dc_validate_mode validate_mode, display_e2e_pipe_params_st *pipes);
void dcn20_fpu_set_wm_ranges(int i,
struct pp_smu_wm_range_sets *ranges,
struct _vcs_dpi_soc_bounding_box_st *loaded_bb);
@@ -75,9 +75,9 @@ void dcn20_fpu_adjust_dppclk(struct vba_vars_st *v,
int dcn21_populate_dml_pipes_from_context(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
-bool dcn21_validate_bandwidth_fp(struct dc *dc, struct dc_state *context, bool
- fast_validate, display_e2e_pipe_params_st *pipes);
+ enum dc_validate_mode validate_mode);
+bool dcn21_validate_bandwidth_fp(struct dc *dc, struct dc_state *context, enum
+ dc_validate_mode, display_e2e_pipe_params_st *pipes);
void dcn21_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params);
void dcn21_clk_mgr_set_bw_params_wm_table(struct clk_bw_params *bw_params);
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c
index aac0a0ae2966..e5f5c0663750 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.c
@@ -178,82 +178,6 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_0_soc = {
};
-void optc3_fpu_set_vrr_m_const(struct timing_generator *optc,
- double vtotal_avg)
-{
- struct optc *optc1 = DCN10TG_FROM_TG(optc);
- double vtotal_min, vtotal_max;
- double ratio, modulo, phase;
- uint32_t vblank_start;
- uint32_t v_total_mask_value = 0;
-
- dc_assert_fp_enabled();
-
- /* Compute VTOTAL_MIN and VTOTAL_MAX, so that
- * VOTAL_MAX - VTOTAL_MIN = 1
- */
- v_total_mask_value = 16;
- vtotal_min = dcn_bw_floor(vtotal_avg);
- vtotal_max = dcn_bw_ceil(vtotal_avg);
-
- /* Check that bottom VBLANK is at least 2 lines tall when running with
- * VTOTAL_MIN. Note that VTOTAL registers are defined as 'total number
- * of lines in a frame - 1'.
- */
- REG_GET(OTG_V_BLANK_START_END, OTG_V_BLANK_START,
- &vblank_start);
- ASSERT(vtotal_min >= vblank_start + 1);
-
- /* Special case where the average frame rate can be achieved
- * without using the DTO
- */
- if (vtotal_min == vtotal_max) {
- REG_SET(OTG_V_TOTAL, 0, OTG_V_TOTAL, (uint32_t)vtotal_min);
-
- optc->funcs->set_vtotal_min_max(optc, 0, 0);
- REG_SET(OTG_M_CONST_DTO0, 0, OTG_M_CONST_DTO_PHASE, 0);
- REG_SET(OTG_M_CONST_DTO1, 0, OTG_M_CONST_DTO_MODULO, 0);
- REG_UPDATE_3(OTG_V_TOTAL_CONTROL,
- OTG_V_TOTAL_MIN_SEL, 0,
- OTG_V_TOTAL_MAX_SEL, 0,
- OTG_SET_V_TOTAL_MIN_MASK_EN, 0);
- return;
- }
-
- ratio = vtotal_max - vtotal_avg;
- modulo = 65536.0 * 65536.0 - 1.0; /* 2^32 - 1 */
- phase = ratio * modulo;
-
- /* Special cases where the DTO phase gets rounded to 0 or
- * to DTO modulo
- */
- if (phase <= 0 || phase >= modulo) {
- REG_SET(OTG_V_TOTAL, 0, OTG_V_TOTAL,
- phase <= 0 ?
- (uint32_t)vtotal_max : (uint32_t)vtotal_min);
- REG_SET(OTG_V_TOTAL_MIN, 0, OTG_V_TOTAL_MIN, 0);
- REG_SET(OTG_V_TOTAL_MAX, 0, OTG_V_TOTAL_MAX, 0);
- REG_SET(OTG_M_CONST_DTO0, 0, OTG_M_CONST_DTO_PHASE, 0);
- REG_SET(OTG_M_CONST_DTO1, 0, OTG_M_CONST_DTO_MODULO, 0);
- REG_UPDATE_3(OTG_V_TOTAL_CONTROL,
- OTG_V_TOTAL_MIN_SEL, 0,
- OTG_V_TOTAL_MAX_SEL, 0,
- OTG_SET_V_TOTAL_MIN_MASK_EN, 0);
- return;
- }
- REG_UPDATE_6(OTG_V_TOTAL_CONTROL,
- OTG_V_TOTAL_MIN_SEL, 1,
- OTG_V_TOTAL_MAX_SEL, 1,
- OTG_SET_V_TOTAL_MIN_MASK_EN, 1,
- OTG_SET_V_TOTAL_MIN_MASK, v_total_mask_value,
- OTG_VTOTAL_MID_REPLACING_MIN_EN, 0,
- OTG_VTOTAL_MID_REPLACING_MAX_EN, 0);
- REG_SET(OTG_V_TOTAL, 0, OTG_V_TOTAL, (uint32_t)vtotal_min);
- optc->funcs->set_vtotal_min_max(optc, vtotal_min, vtotal_max);
- REG_SET(OTG_M_CONST_DTO0, 0, OTG_M_CONST_DTO_PHASE, (uint32_t)phase);
- REG_SET(OTG_M_CONST_DTO1, 0, OTG_M_CONST_DTO_MODULO, (uint32_t)modulo);
-}
-
void dcn30_fpu_populate_dml_writeback_from_context(
struct dc *dc, struct resource_context *res_ctx, display_e2e_pipe_params_st *pipes)
{
@@ -415,7 +339,8 @@ void dcn30_fpu_calculate_wm_and_dlg(
* newly found dummy_latency_index
*/
context->bw_ctx.dml.soc.dram_clock_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.pstate_latency_us;
- dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, false, true);
+ dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel,
+ DC_VALIDATE_MODE_AND_PROGRAMMING, true);
maxMpcComb = context->bw_ctx.dml.vba.maxMpcComb;
dcfclk = context->bw_ctx.dml.vba.DCFCLKState[vlevel][context->bw_ctx.dml.vba.maxMpcComb];
pstate_en = context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][maxMpcComb] != dm_dram_clock_change_unsupported;
@@ -706,7 +631,8 @@ int dcn30_find_dummy_latency_index_for_fw_based_mclk_switch(struct dc *dc,
while (dummy_latency_index < max_latency_table_entries) {
context->bw_ctx.dml.soc.dram_clock_change_latency_us =
dc->clk_mgr->bw_params->dummy_pstate_table[dummy_latency_index].dummy_pstate_latency_us;
- dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, false, true);
+ dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel,
+ DC_VALIDATE_MODE_AND_PROGRAMMING, true);
if (context->bw_ctx.dml.soc.allow_dram_self_refresh_or_dram_clock_change_in_vblank ==
dm_allow_self_refresh_and_mclk_switch)
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.h
index cab864095ce7..e3b6ad6a8784 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/dcn30_fpu.h
@@ -29,9 +29,6 @@
#include "core_types.h"
#include "dcn20/dcn20_optc.h"
-void optc3_fpu_set_vrr_m_const(struct timing_generator *optc,
- double vtotal_avg);
-
void dcn30_fpu_populate_dml_writeback_from_context(
struct dc *dc, struct resource_context *res_ctx, display_e2e_pipe_params_st *pipes);
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c
index cee1b351e105..8d24763938ea 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_mode_vba_30.c
@@ -281,10 +281,10 @@ static void CalculateDynamicMetadataParameters(
double DISPCLK,
double DCFClkDeepSleep,
double PixelClock,
- long HTotal,
- long VBlank,
- long DynamicMetadataTransmittedBytes,
- long DynamicMetadataLinesBeforeActiveRequired,
+ unsigned int HTotal,
+ unsigned int VBlank,
+ unsigned int DynamicMetadataTransmittedBytes,
+ int DynamicMetadataLinesBeforeActiveRequired,
int InterlaceEnable,
bool ProgressiveToInterlaceUnitInOPP,
double *Tsetup,
@@ -1002,6 +1002,7 @@ static bool CalculatePrefetchSchedule(
dst_y_prefetch_equ = VStartup - (Tsetup + dml_max(TWait + TCalc, *Tdmdl)) / LineTime
- (*DSTYAfterScaler + *DSTXAfterScaler / myPipe->HTotal);
+ dst_y_prefetch_equ = dml_min(dst_y_prefetch_equ, 63.75); // limit to the reg limit of U6.2 for DST_Y_PREFETCH
Lsw_oto = dml_max(PrefetchSourceLinesY, PrefetchSourceLinesC);
Tsw_oto = Lsw_oto * LineTime;
@@ -3265,8 +3266,8 @@ static double CalculateWriteBackDelay(
static void CalculateDynamicMetadataParameters(int MaxInterDCNTileRepeaters, double DPPCLK, double DISPCLK,
- double DCFClkDeepSleep, double PixelClock, long HTotal, long VBlank, long DynamicMetadataTransmittedBytes,
- long DynamicMetadataLinesBeforeActiveRequired, int InterlaceEnable, bool ProgressiveToInterlaceUnitInOPP,
+ double DCFClkDeepSleep, double PixelClock, unsigned int HTotal, unsigned int VBlank, unsigned int DynamicMetadataTransmittedBytes,
+ int DynamicMetadataLinesBeforeActiveRequired, int InterlaceEnable, bool ProgressiveToInterlaceUnitInOPP,
double *Tsetup, double *Tdmbf, double *Tdmec, double *Tdmsks)
{
double TotalRepeaterDelayTime = 0;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_rq_dlg_calc_30.c b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_rq_dlg_calc_30.c
index 76d3bb3c9155..8d4873f80df0 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_rq_dlg_calc_30.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn30/display_rq_dlg_calc_30.c
@@ -1562,6 +1562,7 @@ static void dml_rq_dlg_get_dlg_params(struct display_mode_lib *mode_lib,
dml_print("DML_DLG: %s: disp_dlg_regs->dst_y_per_row_vblank = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_row_vblank);
dml_print("DML_DLG: %s: disp_dlg_regs->dst_y_per_vm_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_vm_flip);
dml_print("DML_DLG: %s: disp_dlg_regs->dst_y_per_row_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_row_flip);
+
disp_dlg_regs->refcyc_per_pte_group_vblank_l =
(unsigned int)(dst_y_per_row_vblank * (double)htotal
* ref_freq_to_pix_freq / (double)dpte_groups_per_row_ub_l);
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h
index d2ae43a82ba5..dfcc5d50071e 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/dcn31_fpu.h
@@ -55,5 +55,5 @@ int dcn_get_approx_det_segs_required_for_pstate(
int dcn31x_populate_dml_pipes_from_context(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
#endif /* __DCN31_FPU_H__*/
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c
index f567a9023682..ed59c77bc6f6 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn31/display_mode_vba_31.c
@@ -1105,6 +1105,7 @@ static bool CalculatePrefetchSchedule(
Tr0_oto_lines = dml_ceil(4.0 * Tr0_oto / LineTime, 1) / 4.0;
dst_y_prefetch_oto = Tvm_oto_lines + 2 * Tr0_oto_lines + Lsw_oto;
dst_y_prefetch_equ = VStartup - (*TSetup + dml_max(TWait + TCalc, *Tdmdl)) / LineTime - (*DSTYAfterScaler + *DSTXAfterScaler / myPipe->HTotal);
+ dst_y_prefetch_equ = dml_min(dst_y_prefetch_equ, 63.75); // limit to the reg limit of U6.2 for DST_Y_PREFETCH
dst_y_prefetch_equ = dml_floor(4.0 * (dst_y_prefetch_equ + 0.125), 1) / 4.0;
Tpre_rounded = dst_y_prefetch_equ * LineTime;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c
index 21f637ae4add..df9d50b9b57c 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.c
@@ -306,7 +306,7 @@ static unsigned int get_vertical_back_porch(struct dc_crtc_timing *timing)
int dcn314_populate_dml_pipes_from_context_fpu(struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i, pipe_cnt;
struct resource_context *res_ctx = &context->res_ctx;
@@ -316,7 +316,7 @@ int dcn314_populate_dml_pipes_from_context_fpu(struct dc *dc, struct dc_state *c
dc_assert_fp_enabled();
- dcn31x_populate_dml_pipes_from_context(dc, context, pipes, fast_validate);
+ dcn31x_populate_dml_pipes_from_context(dc, context, pipes, validate_mode);
for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) {
struct dc_crtc_timing *timing;
@@ -409,6 +409,9 @@ int dcn314_populate_dml_pipes_from_context_fpu(struct dc *dc, struct dc_state *c
context->bw_ctx.dml.ip.det_buffer_size_kbytes = 192;
}
+ if (dc->debug.force_odm_combine_4to1)
+ context->bw_ctx.dml.ip.odm_combine_4to1_supported = true;
+
for (i = 0; i < dc->res_pool->pipe_count; i++) {
struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.h
index d32c5bb99f4c..362ac79184ea 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/dcn314_fpu.h
@@ -35,6 +35,6 @@
void dcn314_update_bw_bounding_box_fpu(struct dc *dc, struct clk_bw_params *bw_params);
int dcn314_populate_dml_pipes_from_context_fpu(struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c
index 5865e8fa2d8e..9f3938a50240 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn314/display_mode_vba_314.c
@@ -1123,6 +1123,7 @@ static bool CalculatePrefetchSchedule(
Tr0_oto_lines = dml_ceil(4.0 * Tr0_oto / LineTime, 1) / 4.0;
dst_y_prefetch_oto = Tvm_oto_lines + 2 * Tr0_oto_lines + Lsw_oto;
dst_y_prefetch_equ = VStartup - (*TSetup + dml_max(TWait + TCalc, *Tdmdl)) / LineTime - (*DSTYAfterScaler + *DSTXAfterScaler / myPipe->HTotal);
+ dst_y_prefetch_equ = dml_min(dst_y_prefetch_equ, 63.75); // limit to the reg limit of U6.2 for DST_Y_PREFETCH
dst_y_prefetch_equ = dml_floor(4.0 * (dst_y_prefetch_equ + 0.125), 1) / 4.0;
Tpre_rounded = dst_y_prefetch_equ * LineTime;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
index 6f490d8d7038..6160952245b4 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
@@ -290,7 +290,7 @@ int dcn32_find_dummy_latency_index_for_fw_based_mclk_switch(struct dc *dc,
vba->DRAMClockChangeSupport[vlevel][context->bw_ctx.dml.vba.maxMpcComb] = temp_clock_change_support;
context->bw_ctx.dml.soc.dram_clock_change_latency_us =
dc->clk_mgr->bw_params->dummy_pstate_table[dummy_latency_index].dummy_pstate_latency_us;
- dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, false);
+ dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, DC_VALIDATE_MODE_AND_PROGRAMMING);
/* for subvp + DRR case, if subvp pipes are still present we support pstate */
if (vba->DRAMClockChangeSupport[vlevel][vba->maxMpcComb] == dm_dram_clock_change_unsupported &&
@@ -626,6 +626,8 @@ static bool dcn32_assign_subvp_pipe(struct dc *dc,
* - Not TMZ surface
*/
if (pipe->plane_state && !pipe->top_pipe && !pipe->prev_odm_pipe && !dcn32_is_center_timing(pipe) &&
+ !pipe->stream->hw_cursor_req &&
+ !dc_state_get_stream_cursor_subvp_limit(pipe->stream, context) &&
!(pipe->stream->timing.pix_clk_100hz / 10000 > DCN3_2_MAX_SUBVP_PIXEL_RATE_MHZ) &&
(!dcn32_is_psr_capable(pipe) || (context->stream_count == 1 && dc->caps.dmub_caps.subvp_psr)) &&
dc_state_get_pipe_subvp_type(context, pipe) == SUBVP_NONE &&
@@ -1477,7 +1479,7 @@ static bool dcn32_full_validate_bw_helper(struct dc *dc,
/* Conditions for setting up phantom pipes for SubVP:
* 1. Not force disable SubVP
- * 2. Full update (i.e. !fast_validate)
+ * 2. Full update (i.e. DC_VALIDATE_MODE_AND_PROGRAMMING)
* 3. Enough pipes are available to support SubVP (TODO: Which pipes will use VACTIVE / VBLANK / SUBVP?)
* 4. Display configuration passes validation
* 5. (Config doesn't support MCLK in VACTIVE/VBLANK || dc->debug.force_subvp_mclk_switch)
@@ -1515,7 +1517,8 @@ static bool dcn32_full_validate_bw_helper(struct dc *dc,
dc->res_pool->funcs->add_phantom_pipes(dc, context, pipes, *pipe_cnt, dc_pipe_idx);
- *pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, false);
+ *pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes,
+ DC_VALIDATE_MODE_AND_PROGRAMMING);
// Populate dppclk to trigger a recalculate in dml_get_voltage_level
// so the phantom pipe DLG params can be assigned correctly.
pipes[0].clks_cfg.dppclk_mhz = get_dppclk_calculated(&context->bw_ctx.dml, pipes, *pipe_cnt, 0);
@@ -1558,7 +1561,8 @@ static bool dcn32_full_validate_bw_helper(struct dc *dc,
dc_state_remove_phantom_streams_and_planes(dc, context);
dc_state_release_phantom_streams_and_planes(dc, context);
vba->DRAMClockChangeSupport[*vlevel][vba->maxMpcComb] = dm_dram_clock_change_unsupported;
- *pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, false);
+ *pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes,
+ DC_VALIDATE_MODE_AND_PROGRAMMING);
*vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, *pipe_cnt);
/* This may adjust vlevel and maxMpcComb */
@@ -2136,7 +2140,7 @@ bool dcn32_internal_validate_bw(struct dc *dc,
display_e2e_pipe_params_st *pipes,
int *pipe_cnt_out,
int *vlevel_out,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool out = false;
bool repopulate_pipes = false;
@@ -2160,7 +2164,7 @@ bool dcn32_internal_validate_bw(struct dc *dc,
for (i = 0; i < context->stream_count; i++)
resource_update_pipes_for_stream_with_slice_count(context, dc->current_state, dc->res_pool, context->streams[i], 1);
- pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate);
+ pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, validate_mode);
if (!pipe_cnt) {
out = true;
@@ -2170,13 +2174,13 @@ bool dcn32_internal_validate_bw(struct dc *dc,
dml_log_pipe_params(&context->bw_ctx.dml, pipes, pipe_cnt);
context->bw_ctx.dml.soc.max_vratio_pre = dcn32_determine_max_vratio_prefetch(dc, context);
- if (!fast_validate) {
+ if (validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING) {
if (!dcn32_full_validate_bw_helper(dc, context, pipes, &vlevel, split, merge,
&pipe_cnt, &repopulate_pipes))
goto validate_fail;
}
- if (fast_validate ||
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING ||
(dc->debug.dml_disallow_alternate_prefetch_modes &&
(vlevel == context->bw_ctx.dml.soc.num_states ||
vba->DRAMClockChangeSupport[vlevel][vba->maxMpcComb] == dm_dram_clock_change_unsupported))) {
@@ -2193,7 +2197,7 @@ bool dcn32_internal_validate_bw(struct dc *dc,
context->bw_ctx.dml.soc.allow_for_pstate_or_stutter_in_vblank_final =
dm_prefetch_support_none;
- context->bw_ctx.dml.validate_max_state = fast_validate;
+ context->bw_ctx.dml.validate_max_state = (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING);
vlevel = dml_get_voltage_level(&context->bw_ctx.dml, pipes, pipe_cnt);
context->bw_ctx.dml.validate_max_state = false;
@@ -2245,7 +2249,7 @@ bool dcn32_internal_validate_bw(struct dc *dc,
int flag_vlevel = vlevel;
int i;
- pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate);
+ pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, validate_mode);
if (!dc->config.enable_windowed_mpo_odm)
dcn32_update_dml_pipes_odm_policy_based_on_context(dc, context, pipes);
@@ -2341,7 +2345,7 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context,
}
context->bw_ctx.dml.soc.dram_clock_change_latency_us =
dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.pstate_latency_us;
- dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, false);
+ dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, DC_VALIDATE_MODE_AND_PROGRAMMING);
maxMpcComb = context->bw_ctx.dml.vba.maxMpcComb;
if (is_subvp_p_drr) {
context->bw_ctx.dml.vba.DRAMClockChangeSupport[vlevel][maxMpcComb] = dm_dram_clock_change_vblank_w_mall_sub_vp;
@@ -2387,7 +2391,8 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context,
context->bw_ctx.dml.soc.fclk_change_latency_us =
dc->clk_mgr->bw_params->dummy_pstate_table[dummy_latency_index].dummy_pstate_latency_us;
}
- dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel_temp, false);
+ dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel_temp,
+ DC_VALIDATE_MODE_AND_PROGRAMMING);
if (vlevel_temp < vlevel) {
vlevel = vlevel_temp;
maxMpcComb = context->bw_ctx.dml.vba.maxMpcComb;
@@ -2408,7 +2413,8 @@ void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context,
stream_status->fpo_in_use = false;
}
context->bw_ctx.dml.soc.fclk_change_latency_us = dc->clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.fclk_change_latency_us;
- dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, false);
+ dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel,
+ DC_VALIDATE_MODE_AND_PROGRAMMING);
}
}
}
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h
index 276e90e4e0ce..273d2bd79d85 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.h
@@ -49,7 +49,7 @@ bool dcn32_internal_validate_bw(struct dc *dc,
display_e2e_pipe_params_st *pipes,
int *pipe_cnt_out,
int *vlevel_out,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void dcn32_calculate_wm_and_dlg_fpu(struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c
index 86ac7d59fd32..0748ef36a16a 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/display_mode_vba_util_32.c
@@ -1595,6 +1595,7 @@ double dml32_TruncToValidBPP(
unsigned int NonDSCBPP0;
unsigned int NonDSCBPP1;
unsigned int NonDSCBPP2;
+ unsigned int NonDSCBPP3 = BPP_INVALID;
if (Format == dm_420) {
NonDSCBPP0 = 12;
@@ -1603,6 +1604,7 @@ double dml32_TruncToValidBPP(
MinDSCBPP = 6;
MaxDSCBPP = 1.5 * DSCInputBitPerComponent - 1.0 / 16;
} else if (Format == dm_444) {
+ NonDSCBPP3 = 18;
NonDSCBPP0 = 24;
NonDSCBPP1 = 30;
NonDSCBPP2 = 36;
@@ -1667,6 +1669,8 @@ double dml32_TruncToValidBPP(
return NonDSCBPP1;
else if (MaxLinkBPP >= NonDSCBPP0)
return 16.0;
+ else if ((Output == dm_dp2p0 || Output == dm_dp) && NonDSCBPP3 != BPP_INVALID && MaxLinkBPP >= NonDSCBPP3)
+ return NonDSCBPP3; // Special case to allow 6bpc RGB for DP connections.
else
return BPP_INVALID;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c
index beed7adbbd43..5d73efa2f0c9 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c
@@ -159,7 +159,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = {
.dppclk_mhz = 1200.0,
.phyclk_mhz = 810.0,
.phyclk_d18_mhz = 667.0,
- .dscclk_mhz = 417.0,
+ .dscclk_mhz = 400.0,
.dtbclk_mhz = 600.0,
},
},
@@ -195,9 +195,9 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = {
.dcn_downspread_percent = 0.5,
.gpuvm_min_page_size_bytes = 4096,
.hostvm_min_page_size_bytes = 4096,
- .do_urgent_latency_adjustment = 1,
+ .do_urgent_latency_adjustment = 0,
.urgent_latency_adjustment_fabric_clock_component_us = 0,
- .urgent_latency_adjustment_fabric_clock_reference_mhz = 3000,
+ .urgent_latency_adjustment_fabric_clock_reference_mhz = 0,
};
void dcn35_build_wm_range_table_fpu(struct clk_mgr *clk_mgr)
@@ -367,6 +367,8 @@ void dcn35_update_bw_bounding_box_fpu(struct dc *dc,
clock_limits[i].socclk_mhz;
dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].memclk_mhz =
clk_table->entries[i].memclk_mhz * clk_table->entries[i].wck_ratio;
+
+ dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dram_speed_mts = clock_limits[i].dram_speed_mts;
dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dtbclk_mhz =
clock_limits[i].dtbclk_mhz;
dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dcfclk_levels =
@@ -435,7 +437,7 @@ static unsigned int get_vertical_back_porch(struct dc_crtc_timing *timing)
int dcn35_populate_dml_pipes_from_context_fpu(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i, pipe_cnt;
struct resource_context *res_ctx = &context->res_ctx;
@@ -444,7 +446,7 @@ int dcn35_populate_dml_pipes_from_context_fpu(struct dc *dc,
const unsigned int max_allowed_vblank_nom = 1023;
dcn31_populate_dml_pipes_from_context(dc, context, pipes,
- fast_validate);
+ validate_mode);
for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) {
struct dc_crtc_timing *timing;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.h
index 067480fc3691..d121c5afce71 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.h
@@ -37,7 +37,7 @@ void dcn35_update_bw_bounding_box_fpu(struct dc *dc,
int dcn35_populate_dml_pipes_from_context_fpu(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void dcn35_decide_zstate_support(struct dc *dc, struct dc_state *context);
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c
index d9e63c4fdd95..6f516af82956 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.c
@@ -401,6 +401,7 @@ void dcn351_update_bw_bounding_box_fpu(struct dc *dc,
clock_limits[i].socclk_mhz;
dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].memclk_mhz =
clk_table->entries[i].memclk_mhz * clk_table->entries[i].wck_ratio;
+ dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dram_speed_mts = clock_limits[i].dram_speed_mts;
dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dtbclk_mhz =
clock_limits[i].dtbclk_mhz;
dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dcfclk_levels =
@@ -469,7 +470,7 @@ static unsigned int get_vertical_back_porch(struct dc_crtc_timing *timing)
int dcn351_populate_dml_pipes_from_context_fpu(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i, pipe_cnt;
struct resource_context *res_ctx = &context->res_ctx;
@@ -478,7 +479,7 @@ int dcn351_populate_dml_pipes_from_context_fpu(struct dc *dc,
const unsigned int max_allowed_vblank_nom = 1023;
dcn31_populate_dml_pipes_from_context(dc, context, pipes,
- fast_validate);
+ validate_mode);
for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) {
struct dc_crtc_timing *timing;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.h
index f93efab9a668..f71d9d8d0759 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn351/dcn351_fpu.h
@@ -12,7 +12,7 @@ void dcn351_update_bw_bounding_box_fpu(struct dc *dc,
int dcn351_populate_dml_pipes_from_context_fpu(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void dcn351_decide_zstate_support(struct dc *dc, struct dc_state *context);
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn401/dcn401_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn401/dcn401_fpu.c
deleted file mode 100644
index 4fbecb5ff349..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn401/dcn401_fpu.c
+++ /dev/null
@@ -1,239 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-#include "dcn401_fpu.h"
-#include "dcn401/dcn401_resource.h"
-// We need this includes for WATERMARKS_* defines
-#include "clk_mgr/dcn401/dcn401_smu14_driver_if.h"
-#include "link.h"
-
-#define DC_LOGGER_INIT(logger)
-
-void dcn401_build_wm_range_table_fpu(struct clk_mgr *clk_mgr)
-{
- /* defaults */
- double pstate_latency_us = clk_mgr->ctx->dc->dml.soc.dram_clock_change_latency_us;
- double fclk_change_latency_us = clk_mgr->ctx->dc->dml.soc.fclk_change_latency_us;
- double sr_exit_time_us = clk_mgr->ctx->dc->dml.soc.sr_exit_time_us;
- double sr_enter_plus_exit_time_us = clk_mgr->ctx->dc->dml.soc.sr_enter_plus_exit_time_us;
- /* For min clocks use as reported by PM FW and report those as min */
- uint16_t min_uclk_mhz = clk_mgr->bw_params->clk_table.entries[0].memclk_mhz;
- uint16_t min_dcfclk_mhz = clk_mgr->bw_params->clk_table.entries[0].dcfclk_mhz;
- uint16_t setb_min_uclk_mhz = min_uclk_mhz;
- uint16_t dcfclk_mhz_for_the_second_state = clk_mgr->ctx->dc->dml.soc.clock_limits[2].dcfclk_mhz;
-
- dc_assert_fp_enabled();
-
- /* For Set B ranges use min clocks state 2 when available, and report those to PM FW */
- if (dcfclk_mhz_for_the_second_state)
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].pmfw_breakdown.min_dcfclk = dcfclk_mhz_for_the_second_state;
- else
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].pmfw_breakdown.min_dcfclk = clk_mgr->bw_params->clk_table.entries[0].dcfclk_mhz;
-
- if (clk_mgr->bw_params->clk_table.entries[2].memclk_mhz)
- setb_min_uclk_mhz = clk_mgr->bw_params->clk_table.entries[2].memclk_mhz;
-
- /* Set A - Normal - default values */
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].valid = true;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.pstate_latency_us = pstate_latency_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.fclk_change_latency_us = fclk_change_latency_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.sr_exit_time_us = sr_exit_time_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].dml_input.sr_enter_plus_exit_time_us = sr_enter_plus_exit_time_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.wm_type = WATERMARKS_CLOCK_RANGE;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.min_dcfclk = min_dcfclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.max_dcfclk = 0xFFFF;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.min_uclk = min_uclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_A].pmfw_breakdown.max_uclk = 0xFFFF;
-
- /* Set B - Performance - higher clocks, using DPM[2] DCFCLK and UCLK */
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].valid = true;
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.pstate_latency_us = pstate_latency_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.fclk_change_latency_us = fclk_change_latency_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.sr_exit_time_us = sr_exit_time_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].dml_input.sr_enter_plus_exit_time_us = sr_enter_plus_exit_time_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].pmfw_breakdown.wm_type = WATERMARKS_CLOCK_RANGE;
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].pmfw_breakdown.max_dcfclk = 0xFFFF;
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].pmfw_breakdown.min_uclk = setb_min_uclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_B].pmfw_breakdown.max_uclk = 0xFFFF;
-
- /* Set C - Dummy P-State - P-State latency set to "dummy p-state" value */
- /* 'DalDummyClockChangeLatencyNs' registry key option set to 0x7FFFFFFF can be used to disable Set C for dummy p-state */
- if (clk_mgr->ctx->dc->bb_overrides.dummy_clock_change_latency_ns != 0x7FFFFFFF) {
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].valid = true;
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.pstate_latency_us = 50;
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.fclk_change_latency_us = fclk_change_latency_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.sr_exit_time_us = sr_exit_time_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].dml_input.sr_enter_plus_exit_time_us = sr_enter_plus_exit_time_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].pmfw_breakdown.wm_type = WATERMARKS_DUMMY_PSTATE;
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].pmfw_breakdown.min_dcfclk = min_dcfclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].pmfw_breakdown.max_dcfclk = 0xFFFF;
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].pmfw_breakdown.min_uclk = min_uclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_C].pmfw_breakdown.max_uclk = 0xFFFF;
- clk_mgr->bw_params->dummy_pstate_table[0].dram_speed_mts = clk_mgr->bw_params->clk_table.entries[0].memclk_mhz * 16;
- clk_mgr->bw_params->dummy_pstate_table[0].dummy_pstate_latency_us = 50;
- clk_mgr->bw_params->dummy_pstate_table[1].dram_speed_mts = clk_mgr->bw_params->clk_table.entries[1].memclk_mhz * 16;
- clk_mgr->bw_params->dummy_pstate_table[1].dummy_pstate_latency_us = 9;
- clk_mgr->bw_params->dummy_pstate_table[2].dram_speed_mts = clk_mgr->bw_params->clk_table.entries[2].memclk_mhz * 16;
- clk_mgr->bw_params->dummy_pstate_table[2].dummy_pstate_latency_us = 8;
- clk_mgr->bw_params->dummy_pstate_table[3].dram_speed_mts = clk_mgr->bw_params->clk_table.entries[3].memclk_mhz * 16;
- clk_mgr->bw_params->dummy_pstate_table[3].dummy_pstate_latency_us = 5;
- }
- /* Set D - MALL - SR enter and exit time specific to MALL, TBD after bringup or later phase for now use DRAM values / 2 */
- /* For MALL DRAM clock change latency is N/A, for watermak calculations use lowest value dummy P state latency */
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].valid = true;
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.pstate_latency_us = clk_mgr->bw_params->dummy_pstate_table[3].dummy_pstate_latency_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.fclk_change_latency_us = fclk_change_latency_us;
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.sr_exit_time_us = sr_exit_time_us / 2; // TBD
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].dml_input.sr_enter_plus_exit_time_us = sr_enter_plus_exit_time_us / 2; // TBD
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.wm_type = WATERMARKS_MALL;
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.min_dcfclk = min_dcfclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.max_dcfclk = 0xFFFF;
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.min_uclk = min_uclk_mhz;
- clk_mgr->bw_params->wm_table.nv_entries[WM_D].pmfw_breakdown.max_uclk = 0xFFFF;
-}
-
-/*
- * dcn401_update_bw_bounding_box
- *
- * This would override some dcn4_01 ip_or_soc initial parameters hardcoded from
- * spreadsheet with actual values as per dGPU SKU:
- * - with passed few options from dc->config
- * - with dentist_vco_frequency from Clk Mgr (currently hardcoded, but might
- * need to get it from PM FW)
- * - with passed latency values (passed in ns units) in dc-> bb override for
- * debugging purposes
- * - with passed latencies from VBIOS (in 100_ns units) if available for
- * certain dGPU SKU
- * - with number of DRAM channels from VBIOS (which differ for certain dGPU SKU
- * of the same ASIC)
- * - clocks levels with passed clk_table entries from Clk Mgr as reported by PM
- * FW for different clocks (which might differ for certain dGPU SKU of the
- * same ASIC)
- */
-void dcn401_update_bw_bounding_box_fpu(struct dc *dc, struct clk_bw_params *bw_params)
-{
- dc_assert_fp_enabled();
-
- /* Override from passed dc->bb_overrides if available*/
- if (dc->bb_overrides.sr_exit_time_ns)
- dc->dml2_options.bbox_overrides.sr_exit_latency_us =
- dc->bb_overrides.sr_exit_time_ns / 1000.0;
-
- if (dc->bb_overrides.sr_enter_plus_exit_time_ns)
- dc->dml2_options.bbox_overrides.sr_enter_plus_exit_latency_us =
- dc->bb_overrides.sr_enter_plus_exit_time_ns / 1000.0;
-
- if (dc->bb_overrides.urgent_latency_ns)
- dc->dml2_options.bbox_overrides.urgent_latency_us =
- dc->bb_overrides.urgent_latency_ns / 1000.0;
-
- if (dc->bb_overrides.dram_clock_change_latency_ns)
- dc->dml2_options.bbox_overrides.dram_clock_change_latency_us =
- dc->bb_overrides.dram_clock_change_latency_ns / 1000.0;
-
- if (dc->bb_overrides.fclk_clock_change_latency_ns)
- dc->dml2_options.bbox_overrides.fclk_change_latency_us =
- dc->bb_overrides.fclk_clock_change_latency_ns / 1000;
-
- /* Override from VBIOS if VBIOS bb_info available */
- if (dc->ctx->dc_bios->funcs->get_soc_bb_info) {
- struct bp_soc_bb_info bb_info = {0};
- if (dc->ctx->dc_bios->funcs->get_soc_bb_info(dc->ctx->dc_bios, &bb_info) == BP_RESULT_OK) {
- if (bb_info.dram_clock_change_latency_100ns > 0)
- dc->dml2_options.bbox_overrides.dram_clock_change_latency_us =
- bb_info.dram_clock_change_latency_100ns * 10;
-
- if (bb_info.dram_sr_enter_exit_latency_100ns > 0)
- dc->dml2_options.bbox_overrides.sr_enter_plus_exit_latency_us =
- bb_info.dram_sr_enter_exit_latency_100ns * 10;
-
- if (bb_info.dram_sr_exit_latency_100ns > 0)
- dc->dml2_options.bbox_overrides.sr_exit_latency_us =
- bb_info.dram_sr_exit_latency_100ns * 10;
- }
- }
-
- /* Override from VBIOS for num_chan */
- if (dc->ctx->dc_bios->vram_info.num_chans) {
- dc->dml2_options.bbox_overrides.dram_num_chan =
- dc->ctx->dc_bios->vram_info.num_chans;
-
- }
-
- if (dc->ctx->dc_bios->vram_info.dram_channel_width_bytes)
- dc->dml2_options.bbox_overrides.dram_chanel_width_bytes =
- dc->ctx->dc_bios->vram_info.dram_channel_width_bytes;
-
- dc->dml2_options.bbox_overrides.disp_pll_vco_speed_mhz = dc->clk_mgr->dentist_vco_freq_khz / 1000.0;
- dc->dml2_options.bbox_overrides.xtalclk_mhz = dc->ctx->dc_bios->fw_info.pll_info.crystal_frequency / 1000.0;
- dc->dml2_options.bbox_overrides.dchub_refclk_mhz = dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000.0;
- dc->dml2_options.bbox_overrides.dprefclk_mhz = dc->clk_mgr->dprefclk_khz / 1000.0;
-
- if (dc->clk_mgr->bw_params->clk_table.num_entries > 1) {
- unsigned int i = 0;
-
- dc->dml2_options.bbox_overrides.clks_table.num_states = dc->clk_mgr->bw_params->clk_table.num_entries;
-
- dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dcfclk_levels =
- dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dcfclk_levels;
-
- dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_fclk_levels =
- dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_fclk_levels;
-
- dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_memclk_levels =
- dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_memclk_levels;
-
- dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_socclk_levels =
- dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_socclk_levels;
-
- dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dtbclk_levels =
- dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dtbclk_levels;
-
- dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dispclk_levels =
- dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dispclk_levels;
-
- dc->dml2_options.bbox_overrides.clks_table.num_entries_per_clk.num_dppclk_levels =
- dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dppclk_levels;
-
- for (i = 0; i < dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dcfclk_levels; i++) {
- if (dc->clk_mgr->bw_params->clk_table.entries[i].dcfclk_mhz)
- dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dcfclk_mhz =
- dc->clk_mgr->bw_params->clk_table.entries[i].dcfclk_mhz;
- }
-
- for (i = 0; i < dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_fclk_levels; i++) {
- if (dc->clk_mgr->bw_params->clk_table.entries[i].fclk_mhz)
- dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].fclk_mhz =
- dc->clk_mgr->bw_params->clk_table.entries[i].fclk_mhz;
- }
-
- for (i = 0; i < dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_memclk_levels; i++) {
- if (dc->clk_mgr->bw_params->clk_table.entries[i].memclk_mhz)
- dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].memclk_mhz =
- dc->clk_mgr->bw_params->clk_table.entries[i].memclk_mhz;
- }
-
- for (i = 0; i < dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_socclk_levels; i++) {
- if (dc->clk_mgr->bw_params->clk_table.entries[i].socclk_mhz)
- dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].socclk_mhz =
- dc->clk_mgr->bw_params->clk_table.entries[i].socclk_mhz;
- }
-
- for (i = 0; i < dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dtbclk_levels; i++) {
- if (dc->clk_mgr->bw_params->clk_table.entries[i].dtbclk_mhz)
- dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dtbclk_mhz =
- dc->clk_mgr->bw_params->clk_table.entries[i].dtbclk_mhz;
- }
-
- for (i = 0; i < dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dispclk_levels; i++) {
- if (dc->clk_mgr->bw_params->clk_table.entries[i].dispclk_mhz) {
- dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dispclk_mhz =
- dc->clk_mgr->bw_params->clk_table.entries[i].dispclk_mhz;
- dc->dml2_options.bbox_overrides.clks_table.clk_entries[i].dppclk_mhz =
- dc->clk_mgr->bw_params->clk_table.entries[i].dispclk_mhz;
- }
- }
- }
-}
-
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn401/dcn401_fpu.h b/drivers/gpu/drm/amd/display/dc/dml/dcn401/dcn401_fpu.h
deleted file mode 100644
index 329f1788843c..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn401/dcn401_fpu.h
+++ /dev/null
@@ -1,14 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-#ifndef __DCN401_FPU_H__
-#define __DCN401_FPU_H__
-
-#include "clk_mgr.h"
-
-void dcn401_build_wm_range_table_fpu(struct clk_mgr *clk_mgr);
-
-void dcn401_update_bw_bounding_box_fpu(struct dc *dc, struct clk_bw_params *bw_params);
-
-#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c
index 412e75eb4704..12ff65b6a7e5 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.c
@@ -122,17 +122,6 @@ void print__data_rq_misc_params_st(struct display_mode_lib *mode_lib, const stru
dml_print("DML_RQ_DLG_CALC: =====================================\n");
}
-void print__rq_dlg_params_st(struct display_mode_lib *mode_lib, const struct _vcs_dpi_display_rq_dlg_params_st *rq_dlg_param)
-{
- dml_print("DML_RQ_DLG_CALC: =====================================\n");
- dml_print("DML_RQ_DLG_CALC: DISPLAY_RQ_DLG_PARAM_ST\n");
- dml_print("DML_RQ_DLG_CALC: <LUMA>\n");
- print__data_rq_dlg_params_st(mode_lib, &rq_dlg_param->rq_l);
- dml_print("DML_RQ_DLG_CALC: <CHROMA>\n");
- print__data_rq_dlg_params_st(mode_lib, &rq_dlg_param->rq_c);
- dml_print("DML_RQ_DLG_CALC: =====================================\n");
-}
-
void print__dlg_sys_params_st(struct display_mode_lib *mode_lib, const struct _vcs_dpi_display_dlg_sys_params_st *dlg_sys_param)
{
dml_print("DML_RQ_DLG_CALC: =====================================\n");
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.h b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.h
index ebcd717744e5..2bc64c4081dc 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_rq_dlg_helpers.h
@@ -35,7 +35,6 @@ void print__rq_params_st(struct display_mode_lib *mode_lib, const struct _vcs_dp
void print__data_rq_sizing_params_st(struct display_mode_lib *mode_lib, const struct _vcs_dpi_display_data_rq_sizing_params_st *rq_sizing);
void print__data_rq_dlg_params_st(struct display_mode_lib *mode_lib, const struct _vcs_dpi_display_data_rq_dlg_params_st *rq_dlg_param);
void print__data_rq_misc_params_st(struct display_mode_lib *mode_lib, const struct _vcs_dpi_display_data_rq_misc_params_st *rq_misc_param);
-void print__rq_dlg_params_st(struct display_mode_lib *mode_lib, const struct _vcs_dpi_display_rq_dlg_params_st *rq_dlg_param);
void print__dlg_sys_params_st(struct display_mode_lib *mode_lib, const struct _vcs_dpi_display_dlg_sys_params_st *dlg_sys_param);
void print__data_rq_regs_st(struct display_mode_lib *mode_lib, const struct _vcs_dpi_display_data_rq_regs_st *rq_regs);
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
index 072bd0539605..6b2ab4ec2b5f 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h
@@ -66,11 +66,15 @@ static inline double dml_max5(double a, double b, double c, double d, double e)
static inline double dml_ceil(double a, double granularity)
{
+ if (granularity == 0)
+ return 0;
return (double) dcn_bw_ceil2(a, granularity);
}
static inline double dml_floor(double a, double granularity)
{
+ if (granularity == 0)
+ return 0;
return (double) dcn_bw_floor2(a, granularity);
}
@@ -114,11 +118,15 @@ static inline double dml_ceil_2(double f)
static inline double dml_ceil_ex(double x, double granularity)
{
+ if (granularity == 0)
+ return 0;
return (double) dcn_bw_ceil2(x, granularity);
}
static inline double dml_floor_ex(double x, double granularity)
{
+ if (granularity == 0)
+ return 0;
return (double) dcn_bw_floor2(x, granularity);
}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/Makefile b/drivers/gpu/drm/amd/display/dc/dml2/Makefile
index c4378e620cbf..4c21ce42054c 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dml2/Makefile
@@ -28,11 +28,19 @@ dml2_ccflags := $(CC_FLAGS_FPU)
dml2_rcflags := $(CC_FLAGS_NO_FPU)
ifneq ($(CONFIG_FRAME_WARN),0)
-ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)),y)
-frame_warn_flag := -Wframe-larger-than=3072
-else
-frame_warn_flag := -Wframe-larger-than=2048
-endif
+ ifeq ($(filter y,$(CONFIG_KASAN)$(CONFIG_KCSAN)),y)
+ ifeq ($(CONFIG_CC_IS_CLANG)$(CONFIG_COMPILE_TEST),yy)
+ frame_warn_limit := 4096
+ else
+ frame_warn_limit := 3072
+ endif
+ else
+ frame_warn_limit := 2048
+ endif
+
+ ifeq ($(call test-lt, $(CONFIG_FRAME_WARN), $(frame_warn_limit)),y)
+ frame_warn_flag := -Wframe-larger-than=$(frame_warn_limit)
+ endif
endif
subdir-ccflags-y += -I$(FULL_AMD_DISPLAY_PATH)/dc/dml2
@@ -73,11 +81,11 @@ AMD_DAL_DML2 = $(addprefix $(AMDDALPATH)/dc/dml2/,$(DML2))
AMD_DISPLAY_FILES += $(AMD_DAL_DML2)
-CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml_top.o := $(dml2_ccflags)
-CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml_top_mcache.o := $(dml2_ccflags)
-CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml2_top_optimization := $(dml2_ccflags)
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.o := $(dml2_ccflags)
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.o := $(dml2_ccflags) $(frame_warn_flag)
+CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_core/dml2_core_utils.o := $(dml2_ccflags) $(frame_warn_flag)
+CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml2_top_interfaces.o := $(dml2_ccflags)
+CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml2_top_soc15.o := $(dml2_ccflags)
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_core/dml2_core_factory.o := $(dml2_ccflags)
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.o := $(dml2_ccflags)
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_factory.o := $(dml2_ccflags)
@@ -87,19 +95,16 @@ CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn3.o := $(dml2_ccflag
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.o := $(dml2_ccflags)
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_factory.o := $(dml2_ccflags)
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_standalone_libraries/lib_float_math.o := $(dml2_ccflags)
-
-
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/src/dml21_wrapper.o := $(dml2_ccflags)
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/dml21_translation_helper.o := $(dml2_ccflags)
CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/dml21_utils.o := $(dml2_ccflags)
-CFLAGS_$(AMDDALPATH)/dc/dml2/dml21/inc/dml2_debug.o := $(dml2_ccflags)
-CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml_top.o := $(dml2_rcflags)
-CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml_top_mcache.o := $(dml2_rcflags)
-CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml2_top_optimization.o := $(dml2_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.o := $(dml2_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.o := $(dml2_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_core/dml2_core_factory.o := $(dml2_rcflags)
+CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_core/dml2_core_utils.o := $(dml2_rcflags)
+CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml2_top_interfaces.o := $(dml2_rcflags)
+CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_top/dml2_top_soc15.o := $(dml2_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.o := $(dml2_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_factory.o := $(dml2_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_dcn4.o := $(dml2_rcflags)
@@ -111,13 +116,11 @@ CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml2_standalone_libraries/lib_floa
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/src/dml21_wrapper.o := $(dml2_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/dml21_translation_helper.o := $(dml2_rcflags)
CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/dml21_utils.o := $(dml2_rcflags)
-CFLAGS_REMOVE_$(AMDDALPATH)/dc/dml2/dml21/inc/dml2_debug.o := $(dml2_rcflags)
-DML21 := src/dml2_top/dml_top.o
-DML21 += src/dml2_top/dml_top_mcache.o
-DML21 += src/dml2_top/dml2_top_optimization.o
-DML21 += src/inc/dml2_debug.o
+DML21 := src/dml2_top/dml2_top_interfaces.o
+DML21 += src/dml2_top/dml2_top_soc15.o
DML21 += src/dml2_core/dml2_core_dcn4.o
+DML21 += src/dml2_core/dml2_core_utils.o
DML21 += src/dml2_core/dml2_core_factory.o
DML21 += src/dml2_core/dml2_core_dcn4_calcs.o
DML21 += src/dml2_dpmm/dml2_dpmm_dcn4.o
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c
index d851c081e376..715f9019a33e 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core.c
@@ -32,6 +32,7 @@
#define DML2_MAX_FMT_420_BUFFER_WIDTH 4096
#define TB_BORROWED_MAX 400
+#define DML_MAX_VSTARTUP_START 1023
// ---------------------------
// Declaration Begins
@@ -1222,6 +1223,7 @@ static dml_bool_t CalculatePrefetchSchedule(struct display_mode_lib_scratch_st *
s->dst_y_prefetch_oto = s->Tvm_oto_lines + 2 * s->Tr0_oto_lines + s->Lsw_oto;
s->dst_y_prefetch_equ = p->VStartup - (*p->TSetup + dml_max(p->TWait + p->TCalc, *p->Tdmdl)) / s->LineTime - (*p->DSTYAfterScaler + (dml_float_t) *p->DSTXAfterScaler / (dml_float_t)p->myPipe->HTotal);
+ s->dst_y_prefetch_equ = dml_min(s->dst_y_prefetch_equ, 63.75); // limit to the reg limit of U6.2 for DST_Y_PREFETCH
#ifdef __DML_VBA_DEBUG__
dml_print("DML::%s: HTotal = %u\n", __func__, p->myPipe->HTotal);
@@ -1735,7 +1737,7 @@ static void CalculateBytePerPixelAndBlockSizes(
#endif
} // CalculateBytePerPixelAndBlockSizes
-static dml_float_t CalculateTWait(
+static noinline_for_stack dml_float_t CalculateTWait(
dml_uint_t PrefetchMode,
enum dml_use_mall_for_pstate_change_mode UseMALLForPStateChange,
dml_bool_t SynchronizeDRRDisplaysForUCLKPStateChangeFinal,
@@ -4457,7 +4459,7 @@ static void CalculateSwathWidth(
}
} // CalculateSwathWidth
-static dml_float_t CalculateExtraLatency(
+static noinline_for_stack dml_float_t CalculateExtraLatency(
dml_uint_t RoundTripPingLatencyCycles,
dml_uint_t ReorderingBytes,
dml_float_t DCFCLK,
@@ -5914,7 +5916,7 @@ static dml_uint_t DSCDelayRequirement(
return DSCDelayRequirement_val;
}
-static dml_bool_t CalculateVActiveBandwithSupport(dml_uint_t NumberOfActiveSurfaces,
+static noinline_for_stack dml_bool_t CalculateVActiveBandwithSupport(dml_uint_t NumberOfActiveSurfaces,
dml_float_t ReturnBW,
dml_bool_t NotUrgentLatencyHiding[],
dml_float_t ReadBandwidthLuma[],
@@ -6018,7 +6020,7 @@ static void CalculatePrefetchBandwithSupport(
#endif
}
-static dml_float_t CalculateBandwidthAvailableForImmediateFlip(
+static noinline_for_stack dml_float_t CalculateBandwidthAvailableForImmediateFlip(
dml_uint_t NumberOfActiveSurfaces,
dml_float_t ReturnBW,
dml_float_t ReadBandwidthLuma[],
@@ -6209,10 +6211,11 @@ static dml_uint_t CalculateMaxVStartup(
dml_print("DML::%s: vblank_avail = %u\n", __func__, vblank_avail);
dml_print("DML::%s: max_vstartup_lines = %u\n", __func__, max_vstartup_lines);
#endif
+ max_vstartup_lines = (dml_uint_t) dml_min(max_vstartup_lines, DML_MAX_VSTARTUP_START);
return max_vstartup_lines;
}
-static void set_calculate_prefetch_schedule_params(struct display_mode_lib_st *mode_lib,
+static noinline_for_stack void set_calculate_prefetch_schedule_params(struct display_mode_lib_st *mode_lib,
struct CalculatePrefetchSchedule_params_st *CalculatePrefetchSchedule_params,
dml_uint_t j,
dml_uint_t k)
@@ -6264,7 +6267,7 @@ static void set_calculate_prefetch_schedule_params(struct display_mode_lib_st *m
CalculatePrefetchSchedule_params->Tno_bw = &mode_lib->ms.Tno_bw[k];
}
-static void dml_prefetch_check(struct display_mode_lib_st *mode_lib)
+static noinline_for_stack void dml_prefetch_check(struct display_mode_lib_st *mode_lib)
{
struct dml_core_mode_support_locals_st *s = &mode_lib->scratch.dml_core_mode_support_locals;
struct CalculatePrefetchSchedule_params_st *CalculatePrefetchSchedule_params = &mode_lib->scratch.CalculatePrefetchSchedule_params;
@@ -6300,9 +6303,9 @@ static void dml_prefetch_check(struct display_mode_lib_st *mode_lib)
mode_lib->ms.meta_row_bandwidth_this_state,
mode_lib->ms.dpte_row_bandwidth_this_state,
mode_lib->ms.NoOfDPPThisState,
- mode_lib->ms.UrgentBurstFactorLuma,
- mode_lib->ms.UrgentBurstFactorChroma,
- mode_lib->ms.UrgentBurstFactorCursor);
+ mode_lib->ms.UrgentBurstFactorLuma[j],
+ mode_lib->ms.UrgentBurstFactorChroma[j],
+ mode_lib->ms.UrgentBurstFactorCursor[j]);
s->VMDataOnlyReturnBWPerState = dml_get_return_bw_mbps_vm_only(
&mode_lib->ms.soc,
@@ -6433,7 +6436,7 @@ static void dml_prefetch_check(struct display_mode_lib_st *mode_lib)
/* Output */
&mode_lib->ms.UrgentBurstFactorCursorPre[k],
&mode_lib->ms.UrgentBurstFactorLumaPre[k],
- &mode_lib->ms.UrgentBurstFactorChroma[k],
+ &mode_lib->ms.UrgentBurstFactorChromaPre[k],
&mode_lib->ms.NotUrgentLatencyHidingPre[k]);
mode_lib->ms.cursor_bw_pre[k] = mode_lib->ms.cache_display_cfg.plane.NumberOfCursors[k] * mode_lib->ms.cache_display_cfg.plane.CursorWidth[k] *
@@ -6457,9 +6460,9 @@ static void dml_prefetch_check(struct display_mode_lib_st *mode_lib)
mode_lib->ms.cursor_bw_pre,
mode_lib->ms.prefetch_vmrow_bw,
mode_lib->ms.NoOfDPPThisState,
- mode_lib->ms.UrgentBurstFactorLuma,
- mode_lib->ms.UrgentBurstFactorChroma,
- mode_lib->ms.UrgentBurstFactorCursor,
+ mode_lib->ms.UrgentBurstFactorLuma[j],
+ mode_lib->ms.UrgentBurstFactorChroma[j],
+ mode_lib->ms.UrgentBurstFactorCursor[j],
mode_lib->ms.UrgentBurstFactorLumaPre,
mode_lib->ms.UrgentBurstFactorChromaPre,
mode_lib->ms.UrgentBurstFactorCursorPre,
@@ -6516,9 +6519,9 @@ static void dml_prefetch_check(struct display_mode_lib_st *mode_lib)
mode_lib->ms.cursor_bw,
mode_lib->ms.cursor_bw_pre,
mode_lib->ms.NoOfDPPThisState,
- mode_lib->ms.UrgentBurstFactorLuma,
- mode_lib->ms.UrgentBurstFactorChroma,
- mode_lib->ms.UrgentBurstFactorCursor,
+ mode_lib->ms.UrgentBurstFactorLuma[j],
+ mode_lib->ms.UrgentBurstFactorChroma[j],
+ mode_lib->ms.UrgentBurstFactorCursor[j],
mode_lib->ms.UrgentBurstFactorLumaPre,
mode_lib->ms.UrgentBurstFactorChromaPre,
mode_lib->ms.UrgentBurstFactorCursorPre);
@@ -6585,9 +6588,9 @@ static void dml_prefetch_check(struct display_mode_lib_st *mode_lib)
mode_lib->ms.cursor_bw_pre,
mode_lib->ms.prefetch_vmrow_bw,
mode_lib->ms.NoOfDPP[j], // VBA_ERROR DPPPerSurface is not assigned at this point, should use NoOfDpp here
- mode_lib->ms.UrgentBurstFactorLuma,
- mode_lib->ms.UrgentBurstFactorChroma,
- mode_lib->ms.UrgentBurstFactorCursor,
+ mode_lib->ms.UrgentBurstFactorLuma[j],
+ mode_lib->ms.UrgentBurstFactorChroma[j],
+ mode_lib->ms.UrgentBurstFactorCursor[j],
mode_lib->ms.UrgentBurstFactorLumaPre,
mode_lib->ms.UrgentBurstFactorChromaPre,
mode_lib->ms.UrgentBurstFactorCursorPre,
@@ -7808,9 +7811,9 @@ dml_bool_t dml_core_mode_support(struct display_mode_lib_st *mode_lib)
mode_lib->ms.DETBufferSizeYThisState[k],
mode_lib->ms.DETBufferSizeCThisState[k],
/* Output */
- &mode_lib->ms.UrgentBurstFactorCursor[k],
- &mode_lib->ms.UrgentBurstFactorLuma[k],
- &mode_lib->ms.UrgentBurstFactorChroma[k],
+ &mode_lib->ms.UrgentBurstFactorCursor[j][k],
+ &mode_lib->ms.UrgentBurstFactorLuma[j][k],
+ &mode_lib->ms.UrgentBurstFactorChroma[j][k],
&mode_lib->ms.NotUrgentLatencyHiding[k]);
}
@@ -8317,7 +8320,7 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc
if (clk_cfg->dcfclk_option != dml_use_override_freq)
locals->Dcfclk = mode_lib->ms.DCFCLK;
else
- locals->Dcfclk = clk_cfg->dcfclk_freq_mhz;
+ locals->Dcfclk = clk_cfg->dcfclk_mhz;
#ifdef __DML_VBA_DEBUG__
dml_print_dml_policy(&mode_lib->ms.policy);
@@ -8370,7 +8373,7 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc
if (clk_cfg->dispclk_option == dml_use_required_freq)
locals->Dispclk = locals->Dispclk_calculated;
else if (clk_cfg->dispclk_option == dml_use_override_freq)
- locals->Dispclk = clk_cfg->dispclk_freq_mhz;
+ locals->Dispclk = clk_cfg->dispclk_mhz;
else
locals->Dispclk = mode_lib->ms.state.dispclk_mhz;
#ifdef __DML_VBA_DEBUG__
@@ -8411,7 +8414,7 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc
if (clk_cfg->dppclk_option[k] == dml_use_required_freq)
locals->Dppclk[k] = locals->Dppclk_calculated[k];
else if (clk_cfg->dppclk_option[k] == dml_use_override_freq)
- locals->Dppclk[k] = clk_cfg->dppclk_freq_mhz[k];
+ locals->Dppclk[k] = clk_cfg->dppclk_mhz[k];
else
locals->Dppclk[k] = mode_lib->ms.state.dppclk_mhz;
#ifdef __DML_VBA_DEBUG__
@@ -9189,6 +9192,8 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc
&locals->FractionOfUrgentBandwidth,
&s->dummy_boolean[0]); // dml_bool_t *PrefetchBandwidthSupport
+
+
if (s->VRatioPrefetchMoreThanMax != false || s->DestinationLineTimesForPrefetchLessThan2 != false) {
dml_print("DML::%s: VRatioPrefetchMoreThanMax = %u\n", __func__, s->VRatioPrefetchMoreThanMax);
dml_print("DML::%s: DestinationLineTimesForPrefetchLessThan2 = %u\n", __func__, s->DestinationLineTimesForPrefetchLessThan2);
@@ -9203,6 +9208,7 @@ void dml_core_mode_programming(struct display_mode_lib_st *mode_lib, const struc
}
}
+
if (locals->PrefetchModeSupported == true && mode_lib->ms.support.ImmediateFlipSupport == true) {
locals->BandwidthAvailableForImmediateFlip = CalculateBandwidthAvailableForImmediateFlip(
mode_lib->ms.num_active_planes,
@@ -10183,7 +10189,7 @@ dml_uint_t dml_mode_support_ex(struct dml_mode_support_ex_params_st *in_out_para
result = mode_support_pwr_states(&in_out_params->out_lowest_state_idx,
in_out_params->mode_lib,
in_out_params->in_display_cfg,
- 0,
+ in_out_params->in_start_state_idx,
in_out_params->mode_lib->states.num_states - 1);
if (result)
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core_structs.h b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core_structs.h
index f951936bb579..dbeb08466092 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core_structs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_core_structs.h
@@ -28,6 +28,7 @@
#define __DISPLAY_MODE_CORE_STRUCT_H__
#include "display_mode_lib_defines.h"
+#include "dml_top_display_cfg_types.h"
enum dml_project_id {
dml_project_invalid = 0,
@@ -37,6 +38,7 @@ enum dml_project_id {
dml_project_dcn35 = 3,
dml_project_dcn351 = 4,
dml_project_dcn401 = 5,
+ dml_project_dcn36 = 6,
};
enum dml_prefetch_modes {
dml_prefetch_support_uclk_fclk_and_stutter_if_possible = 0,
@@ -49,7 +51,9 @@ enum dml_use_mall_for_pstate_change_mode {
dml_use_mall_pstate_change_disable = 0,
dml_use_mall_pstate_change_full_frame = 1,
dml_use_mall_pstate_change_sub_viewport = 2,
- dml_use_mall_pstate_change_phantom_pipe = 3
+ dml_use_mall_pstate_change_phantom_pipe = 3,
+ dml_use_mall_pstate_change_phantom_pipe_no_data_return = 4,
+ dml_use_mall_pstate_change_imall = 5
};
enum dml_use_mall_for_static_screen_mode {
dml_use_mall_static_screen_disable = 0,
@@ -171,7 +175,11 @@ enum dml_swizzle_mode {
dml_sw_256kb_z_x = 28,
dml_sw_256kb_s_x = 29,
dml_sw_256kb_d_x = 30,
- dml_sw_256kb_r_x = 31
+ dml_sw_256kb_r_x = 31,
+ dml_sw_256b_2d = 32,
+ dml_sw_4kb_2d = 33,
+ dml_sw_64kb_2d = 34,
+ dml_sw_256kb_2d = 35
};
enum dml_lb_depth {
dml_lb_6 = 0,
@@ -223,24 +231,28 @@ enum dml_mpc_use_policy {
dml_mpc_disabled = 0,
dml_mpc_as_possible = 1,
dml_mpc_as_needed_for_voltage = 2,
- dml_mpc_as_needed_for_pstate_and_voltage = 3
+ dml_mpc_as_needed_for_pstate_and_voltage = 3,
+ dml_mpc_as_needed = 4,
+ dml_mpc_2to1 = 5
};
enum dml_odm_use_policy {
dml_odm_use_policy_bypass = 0,
dml_odm_use_policy_combine_as_needed = 1,
dml_odm_use_policy_combine_2to1 = 2,
- dml_odm_use_policy_combine_4to1 = 3,
- dml_odm_use_policy_split_1to2 = 4,
- dml_odm_use_policy_mso_1to2 = 5,
- dml_odm_use_policy_mso_1to4 = 6
+ dml_odm_use_policy_combine_3to1 = 3,
+ dml_odm_use_policy_combine_4to1 = 4,
+ dml_odm_use_policy_split_1to2 = 5,
+ dml_odm_use_policy_mso_1to2 = 6,
+ dml_odm_use_policy_mso_1to4 = 7
};
enum dml_odm_mode {
dml_odm_mode_bypass = 0,
dml_odm_mode_combine_2to1 = 1,
- dml_odm_mode_combine_4to1 = 2,
- dml_odm_mode_split_1to2 = 3,
- dml_odm_mode_mso_1to2 = 4,
- dml_odm_mode_mso_1to4 = 5
+ dml_odm_mode_combine_3to1 = 2,
+ dml_odm_mode_combine_4to1 = 3,
+ dml_odm_mode_split_1to2 = 4,
+ dml_odm_mode_mso_1to2 = 5,
+ dml_odm_mode_mso_1to4 = 6
};
enum dml_writeback_configuration {
dml_whole_buffer_for_single_stream_no_interleave = 0,
@@ -289,6 +301,17 @@ struct soc_state_bounding_box_st {
dml_float_t fclk_change_latency_us;
dml_float_t usr_retraining_latency_us;
dml_bool_t use_ideal_dram_bw_strobe;
+ dml_float_t g6_temp_read_blackout_us;
+
+ struct {
+ dml_uint_t urgent_ramp_uclk_cycles;
+ dml_uint_t trip_to_memory_uclk_cycles;
+ dml_uint_t meta_trip_to_memory_uclk_cycles;
+ dml_uint_t maximum_latency_when_urgent_uclk_cycles;
+ dml_uint_t average_latency_when_urgent_uclk_cycles;
+ dml_uint_t maximum_latency_when_non_urgent_uclk_cycles;
+ dml_uint_t average_latency_when_non_urgent_uclk_cycles;
+ } dml_dcn401_uclk_dpm_dependent_soc_qos_params;
};
struct soc_bounding_box_st {
@@ -297,7 +320,7 @@ struct soc_bounding_box_st {
dml_float_t pcierefclk_mhz;
dml_float_t refclk_mhz;
dml_float_t amclk_mhz;
- dml_float_t max_outstanding_reqs;
+ dml_uint_t max_outstanding_reqs;
dml_float_t pct_ideal_sdp_bw_after_urgent;
dml_float_t pct_ideal_fabric_bw_after_urgent;
dml_float_t pct_ideal_dram_bw_after_urgent_pixel_only;
@@ -308,6 +331,16 @@ struct soc_bounding_box_st {
dml_float_t max_avg_fabric_bw_use_normal_percent;
dml_float_t max_avg_dram_bw_use_normal_percent;
dml_float_t max_avg_dram_bw_use_normal_strobe_percent;
+
+ dml_float_t svp_prefetch_pct_ideal_sdp_bw_after_urgent;
+ dml_float_t svp_prefetch_pct_ideal_fabric_bw_after_urgent;
+ dml_float_t svp_prefetch_pct_ideal_dram_bw_after_urgent_pixel_only;
+ dml_float_t svp_prefetch_pct_ideal_dram_bw_after_urgent_pixel_and_vm;
+ dml_float_t svp_prefetch_pct_ideal_dram_bw_after_urgent_vm_only;
+ dml_float_t svp_prefetch_max_avg_sdp_bw_use_normal_percent;
+ dml_float_t svp_prefetch_max_avg_fabric_bw_use_normal_percent;
+ dml_float_t svp_prefetch_max_avg_dram_bw_use_normal_percent;
+
dml_uint_t round_trip_ping_latency_dcfclk_cycles;
dml_uint_t urgent_out_of_order_return_per_channel_pixel_only_bytes;
dml_uint_t urgent_out_of_order_return_per_channel_pixel_and_vm_bytes;
@@ -324,6 +357,26 @@ struct soc_bounding_box_st {
dml_uint_t mall_allocated_for_dcn_mbytes;
dml_float_t dispclk_dppclk_vco_speed_mhz;
dml_bool_t do_urgent_latency_adjustment;
+
+ dml_uint_t mem_word_bytes;
+ dml_uint_t num_dcc_mcaches;
+ dml_uint_t mcache_size_bytes;
+ dml_uint_t mcache_line_size_bytes;
+
+ struct {
+ dml_bool_t UseNewDCN401SOCParameters;
+ dml_uint_t df_qos_response_time_fclk_cycles;
+ dml_uint_t max_round_trip_to_furthest_cs_fclk_cycles;
+ dml_uint_t mall_overhead_fclk_cycles;
+ dml_uint_t meta_trip_adder_fclk_cycles;
+ dml_uint_t average_transport_distance_fclk_cycles;
+ dml_float_t umc_urgent_ramp_latency_margin;
+ dml_float_t umc_max_latency_margin;
+ dml_float_t umc_average_latency_margin;
+ dml_float_t fabric_max_transport_latency_margin;
+ dml_float_t fabric_average_transport_latency_margin;
+ } dml_dcn401_soc_qos_params;
+
};
struct ip_params_st {
@@ -515,6 +568,10 @@ struct dml_plane_cfg_st {
dml_uint_t CursorWidth[__DML_NUM_PLANES__];
dml_uint_t CursorBPP[__DML_NUM_PLANES__];
+ dml_bool_t setup_for_tdlut[__DML_NUM_PLANES__];
+ enum dml2_tdlut_addressing_mode tdlut_addressing_mode[__DML_NUM_PLANES__];
+ enum dml2_tdlut_width_mode tdlut_width_mode[__DML_NUM_PLANES__];
+
enum dml_use_mall_for_static_screen_mode UseMALLForStaticScreen[__DML_NUM_PLANES__];
enum dml_use_mall_for_pstate_change_mode UseMALLForPStateChange[__DML_NUM_PLANES__];
@@ -604,6 +661,17 @@ struct dml_hw_resource_st {
dml_float_t DLGRefClkFreqMHz; /// <brief DLG Global Reference timer
};
+/// @brief To control the clk usage for model programming
+struct dml_clk_cfg_st {
+ enum dml_clk_cfg_policy dcfclk_option; ///< brief Use for mode_program; user can select between use the min require clk req as calculated by DML or use the test-specific freq
+ enum dml_clk_cfg_policy dispclk_option; ///< brief Use for mode_program; user can select between use the min require clk req as calculated by DML or use the test-specific freq
+ enum dml_clk_cfg_policy dppclk_option[__DML_NUM_PLANES__];
+
+ dml_float_t dcfclk_mhz;
+ dml_float_t dispclk_mhz;
+ dml_float_t dppclk_mhz[__DML_NUM_PLANES__];
+}; // dml_clk_cfg_st
+
/// @brief DML display configuration.
/// Describe how to display a surface in multi-plane setup and output to different output and writeback using the specified timgin
struct dml_display_cfg_st {
@@ -616,19 +684,9 @@ struct dml_display_cfg_st {
unsigned int num_timings;
struct dml_hw_resource_st hw; //< brief for mode programming
+ struct dml_clk_cfg_st clk_overrides; //< brief for mode programming clk override
}; // dml_display_cfg_st
-/// @brief To control the clk usage for model programming
-struct dml_clk_cfg_st {
- enum dml_clk_cfg_policy dcfclk_option; ///< brief Use for mode_program; user can select between use the min require clk req as calculated by DML or use the test-specific freq
- enum dml_clk_cfg_policy dispclk_option; ///< brief Use for mode_program; user can select between use the min require clk req as calculated by DML or use the test-specific freq
- enum dml_clk_cfg_policy dppclk_option[__DML_NUM_PLANES__];
-
- dml_float_t dcfclk_freq_mhz;
- dml_float_t dispclk_freq_mhz;
- dml_float_t dppclk_freq_mhz[__DML_NUM_PLANES__];
-}; // dml_clk_cfg_st
-
/// @brief DML mode evaluation and programming policy
/// Those knobs that affect mode support and mode programming
struct dml_mode_eval_policy_st {
@@ -884,11 +942,11 @@ struct mode_support_st {
dml_uint_t meta_row_height[__DML_NUM_PLANES__];
dml_uint_t meta_row_height_chroma[__DML_NUM_PLANES__];
dml_float_t UrgLatency;
- dml_float_t UrgentBurstFactorCursor[__DML_NUM_PLANES__];
+ dml_float_t UrgentBurstFactorCursor[2][__DML_NUM_PLANES__];
dml_float_t UrgentBurstFactorCursorPre[__DML_NUM_PLANES__];
- dml_float_t UrgentBurstFactorLuma[__DML_NUM_PLANES__];
+ dml_float_t UrgentBurstFactorLuma[2][__DML_NUM_PLANES__];
dml_float_t UrgentBurstFactorLumaPre[__DML_NUM_PLANES__];
- dml_float_t UrgentBurstFactorChroma[__DML_NUM_PLANES__];
+ dml_float_t UrgentBurstFactorChroma[2][__DML_NUM_PLANES__];
dml_float_t UrgentBurstFactorChromaPre[__DML_NUM_PLANES__];
dml_float_t MaximumSwathWidthInLineBufferLuma;
dml_float_t MaximumSwathWidthInLineBufferChroma;
@@ -1859,6 +1917,7 @@ struct display_mode_lib_st {
struct dml_mode_support_ex_params_st {
struct display_mode_lib_st *mode_lib;
const struct dml_display_cfg_st *in_display_cfg;
+ dml_uint_t in_start_state_idx;
dml_uint_t out_lowest_state_idx;
struct dml_mode_support_info_st *out_evaluation_info;
};
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_util.c b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_util.c
index c247aee89caf..89890c88fd66 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/display_mode_util.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/display_mode_util.c
@@ -690,12 +690,12 @@ __DML_DLL_EXPORT__ void dml_print_clk_cfg(const struct dml_clk_cfg_st *clk_cfg)
dml_print("DML: clk_cfg: dcfclk_option = %d\n", clk_cfg->dcfclk_option);
dml_print("DML: clk_cfg: dispclk_option = %d\n", clk_cfg->dispclk_option);
- dml_print("DML: clk_cfg: dcfclk_freq_mhz = %f\n", clk_cfg->dcfclk_freq_mhz);
- dml_print("DML: clk_cfg: dispclk_freq_mhz = %f\n", clk_cfg->dispclk_freq_mhz);
+ dml_print("DML: clk_cfg: dcfclk_mhz = %f\n", clk_cfg->dcfclk_mhz);
+ dml_print("DML: clk_cfg: dispclk_mhz = %f\n", clk_cfg->dispclk_mhz);
for (dml_uint_t i = 0; i < DCN_DML__NUM_PLANE; i++) {
dml_print("DML: clk_cfg: i=%d, dppclk_option = %d\n", i, clk_cfg->dppclk_option[i]);
- dml_print("DML: clk_cfg: i=%d, dppclk_freq_mhz = %f\n", i, clk_cfg->dppclk_freq_mhz[i]);
+ dml_print("DML: clk_cfg: i=%d, dppclk_mhz = %f\n", i, clk_cfg->dppclk_mhz[i]);
}
}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c
index 138b4b1e42ed..a06217a9eef6 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.c
@@ -2,7 +2,6 @@
//
// Copyright 2024 Advanced Micro Devices, Inc.
-
#include "dml21_wrapper.h"
#include "dml2_core_dcn4_calcs.h"
#include "dml2_internal_shared_types.h"
@@ -10,286 +9,264 @@
#include "dml21_utils.h"
#include "dml21_translation_helper.h"
#include "bounding_boxes/dcn4_soc_bb.h"
-#include "bounding_boxes/dcn3_soc_bb.h"
-
-static void dml21_init_socbb_params(struct dml2_initialize_instance_in_out *dml_init,
- const struct dml2_configuration_options *config,
- const struct dc *in_dc)
-{
- const struct dml2_soc_bb *soc_bb;
- const struct dml2_soc_qos_parameters *qos_params;
- switch (in_dc->ctx->dce_version) {
- case DCN_VERSION_3_2: // TODO : Temporary for N-1 validation. Remove this after N-1 validation phase is complete.
- soc_bb = &dml2_socbb_dcn31;
- qos_params = &dml_dcn31_soc_qos_params;
- break;
- case DCN_VERSION_4_01:
- default:
- if (config->bb_from_dmub)
- soc_bb = config->bb_from_dmub;
- else
- soc_bb = &dml2_socbb_dcn401;
-
- qos_params = &dml_dcn4_variant_a_soc_qos_params;
- }
-
- /* patch soc bb */
- memcpy(&dml_init->soc_bb, soc_bb, sizeof(struct dml2_soc_bb));
-
- /* patch qos params */
- memcpy(&dml_init->soc_bb.qos_parameters, qos_params, sizeof(struct dml2_soc_qos_parameters));
-}
-
-static void dml21_external_socbb_params(struct dml2_initialize_instance_in_out *dml_init,
- const struct dml2_configuration_options *config)
-{
- memcpy(&dml_init->soc_bb, &config->external_socbb_ip_params->soc_bb, sizeof(struct dml2_soc_bb));
-}
-
-static void dml21_external_ip_params(struct dml2_initialize_instance_in_out *dml_init,
+static void dml21_populate_pmo_options(struct dml2_pmo_options *pmo_options,
+ const struct dc *in_dc,
const struct dml2_configuration_options *config)
{
- memcpy(&dml_init->ip_caps, &config->external_socbb_ip_params->ip_params, sizeof(struct dml2_ip_capabilities));
+ bool disable_fams2 = !in_dc->debug.fams2_config.bits.enable;
+
+ /* ODM options */
+ pmo_options->disable_dyn_odm = !config->minimize_dispclk_using_odm;
+ pmo_options->disable_dyn_odm_for_multi_stream = true;
+ pmo_options->disable_dyn_odm_for_stream_with_svp = true;
+
+ pmo_options->disable_vblank = ((in_dc->debug.dml21_disable_pstate_method_mask >> 1) & 1);
+
+ /* NOTE: DRR and SubVP Require FAMS2 */
+ pmo_options->disable_svp = ((in_dc->debug.dml21_disable_pstate_method_mask >> 2) & 1) ||
+ in_dc->debug.force_disable_subvp ||
+ disable_fams2;
+ pmo_options->disable_drr_clamped = ((in_dc->debug.dml21_disable_pstate_method_mask >> 3) & 1) ||
+ disable_fams2;
+ pmo_options->disable_drr_var = ((in_dc->debug.dml21_disable_pstate_method_mask >> 4) & 1) ||
+ disable_fams2;
+ pmo_options->disable_fams2 = disable_fams2;
+
+ pmo_options->disable_drr_var_when_var_active = in_dc->debug.disable_fams_gaming == INGAME_FAMS_DISABLE ||
+ in_dc->debug.disable_fams_gaming == INGAME_FAMS_MULTI_DISP_CLAMPED_ONLY;
+ pmo_options->disable_drr_clamped_when_var_active = in_dc->debug.disable_fams_gaming == INGAME_FAMS_DISABLE;
}
-static void dml21_init_ip_params(struct dml2_initialize_instance_in_out *dml_init,
+/*
+ * Populate dml_init based on default static values in soc bb. The default
+ * values are for reference and support at least minimal operation of current
+ * SoC and DCN hardware. The values could be modifed by subsequent override
+ * functions to reflect our true hardware capability.
+ */
+static void populate_default_dml_init_params(struct dml2_initialize_instance_in_out *dml_init,
const struct dml2_configuration_options *config,
const struct dc *in_dc)
{
- const struct dml2_ip_capabilities *ip_caps;
-
switch (in_dc->ctx->dce_version) {
- case DCN_VERSION_3_2: // TODO : Temporary for N-1 validation. Remove this after N-1 validation phase is complete.
- ip_caps = &dml2_dcn31_max_ip_caps;
- break;
case DCN_VERSION_4_01:
+ dml_init->options.project_id = dml2_project_dcn4x_stage2_auto_drr_svp;
+ dml21_populate_pmo_options(&dml_init->options.pmo_options, in_dc, config);
+ dml_init->soc_bb = dml2_socbb_dcn401;
+ dml_init->soc_bb.qos_parameters = dml_dcn4_variant_a_soc_qos_params;
+ dml_init->ip_caps = dml2_dcn401_max_ip_caps;
+ break;
default:
- ip_caps = &dml2_dcn401_max_ip_caps;
+ memset(dml_init, 0, sizeof(*dml_init));
+ DC_ERR("unsupported dcn version for DML21!");
+ return;
}
-
- memcpy(&dml_init->ip_caps, ip_caps, sizeof(struct dml2_ip_capabilities));
}
-void dml21_initialize_soc_bb_params(struct dml2_initialize_instance_in_out *dml_init,
+static void override_dml_init_with_values_from_hardware_default(struct dml2_initialize_instance_in_out *dml_init,
const struct dml2_configuration_options *config,
const struct dc *in_dc)
{
- if (config->use_native_soc_bb_construction)
- dml21_init_socbb_params(dml_init, config, in_dc);
- else
- dml21_external_socbb_params(dml_init, config);
+ dml_init->soc_bb.dchub_refclk_mhz = in_dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000;
+ dml_init->soc_bb.dprefclk_mhz = in_dc->clk_mgr->dprefclk_khz / 1000;
+ dml_init->soc_bb.dispclk_dppclk_vco_speed_mhz = in_dc->clk_mgr->dentist_vco_freq_khz / 1000.0;
}
-void dml21_initialize_ip_params(struct dml2_initialize_instance_in_out *dml_init,
+/*
+ * SMU stands for System Management Unit. It is a power management processor.
+ * It owns the initialization of dc's clock table and programming of clock values
+ * based on dc's requests.
+ * Our clock values in base soc bb is a dummy placeholder. The real clock values
+ * are retrieved from SMU firmware to dc clock table at runtime.
+ * This function overrides our dummy placeholder values with real values in dc
+ * clock table.
+ */
+static void override_dml_init_with_values_from_smu(
+ struct dml2_initialize_instance_in_out *dml_init,
const struct dml2_configuration_options *config,
const struct dc *in_dc)
{
- if (config->use_native_soc_bb_construction)
- dml21_init_ip_params(dml_init, config, in_dc);
- else
- dml21_external_ip_params(dml_init, config);
-}
-
-void dml21_apply_soc_bb_overrides(struct dml2_initialize_instance_in_out *dml_init,
- const struct dml2_configuration_options *config, const struct dc *in_dc)
-{
int i;
-
const struct clk_bw_params *dc_bw_params = in_dc->clk_mgr->bw_params;
const struct clk_limit_table *dc_clk_table = &dc_bw_params->clk_table;
- struct dml2_soc_bb *dml_soc_bb = &dml_init->soc_bb;
- struct dml2_soc_state_table *dml_clk_table = &dml_soc_bb->clk_table;
-
- /* override clocks if smu is present */
- if (in_dc->clk_mgr->funcs->is_smu_present && in_dc->clk_mgr->funcs->is_smu_present(in_dc->clk_mgr)) {
- /* dcfclk */
- if (dc_clk_table->num_entries_per_clk.num_dcfclk_levels) {
- dml_clk_table->dcfclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_dcfclk_levels;
- for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
- if (i < dml_clk_table->dcfclk.num_clk_values) {
- if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.dcfclk_mhz &&
- dc_clk_table->entries[i].dcfclk_mhz > dc_bw_params->dc_mode_limit.dcfclk_mhz) {
- if (i == 0 || dc_clk_table->entries[i-1].dcfclk_mhz < dc_bw_params->dc_mode_limit.dcfclk_mhz) {
- dml_clk_table->dcfclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.dcfclk_mhz * 1000;
- dml_clk_table->dcfclk.num_clk_values = i + 1;
- } else {
- dml_clk_table->dcfclk.clk_values_khz[i] = 0;
- dml_clk_table->dcfclk.num_clk_values = i;
- }
+ struct dml2_soc_state_table *dml_clk_table = &dml_init->soc_bb.clk_table;
+
+ if (!in_dc->clk_mgr->funcs->is_smu_present ||
+ !in_dc->clk_mgr->funcs->is_smu_present(in_dc->clk_mgr))
+ /* skip if smu is not present */
+ return;
+
+ /* dcfclk */
+ if (dc_clk_table->num_entries_per_clk.num_dcfclk_levels) {
+ dml_clk_table->dcfclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_dcfclk_levels;
+ for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
+ if (i < dml_clk_table->dcfclk.num_clk_values) {
+ if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.dcfclk_mhz &&
+ dc_clk_table->entries[i].dcfclk_mhz > dc_bw_params->dc_mode_limit.dcfclk_mhz) {
+ if (i == 0 || dc_clk_table->entries[i-1].dcfclk_mhz < dc_bw_params->dc_mode_limit.dcfclk_mhz) {
+ dml_clk_table->dcfclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.dcfclk_mhz * 1000;
+ dml_clk_table->dcfclk.num_clk_values = i + 1;
} else {
- dml_clk_table->dcfclk.clk_values_khz[i] = dc_clk_table->entries[i].dcfclk_mhz * 1000;
+ dml_clk_table->dcfclk.clk_values_khz[i] = 0;
+ dml_clk_table->dcfclk.num_clk_values = i;
}
} else {
- dml_clk_table->dcfclk.clk_values_khz[i] = 0;
+ dml_clk_table->dcfclk.clk_values_khz[i] = dc_clk_table->entries[i].dcfclk_mhz * 1000;
}
+ } else {
+ dml_clk_table->dcfclk.clk_values_khz[i] = 0;
}
}
+ }
- /* fclk */
- if (dc_clk_table->num_entries_per_clk.num_fclk_levels) {
- dml_clk_table->fclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_fclk_levels;
- for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
- if (i < dml_clk_table->fclk.num_clk_values) {
- if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.fclk_mhz &&
- dc_clk_table->entries[i].fclk_mhz > dc_bw_params->dc_mode_limit.fclk_mhz) {
- if (i == 0 || dc_clk_table->entries[i-1].fclk_mhz < dc_bw_params->dc_mode_limit.fclk_mhz) {
- dml_clk_table->fclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.fclk_mhz * 1000;
- dml_clk_table->fclk.num_clk_values = i + 1;
- } else {
- dml_clk_table->fclk.clk_values_khz[i] = 0;
- dml_clk_table->fclk.num_clk_values = i;
- }
+ /* fclk */
+ if (dc_clk_table->num_entries_per_clk.num_fclk_levels) {
+ dml_clk_table->fclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_fclk_levels;
+ for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
+ if (i < dml_clk_table->fclk.num_clk_values) {
+ if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.fclk_mhz &&
+ dc_clk_table->entries[i].fclk_mhz > dc_bw_params->dc_mode_limit.fclk_mhz) {
+ if (i == 0 || dc_clk_table->entries[i-1].fclk_mhz < dc_bw_params->dc_mode_limit.fclk_mhz) {
+ dml_clk_table->fclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.fclk_mhz * 1000;
+ dml_clk_table->fclk.num_clk_values = i + 1;
} else {
- dml_clk_table->fclk.clk_values_khz[i] = dc_clk_table->entries[i].fclk_mhz * 1000;
+ dml_clk_table->fclk.clk_values_khz[i] = 0;
+ dml_clk_table->fclk.num_clk_values = i;
}
} else {
- dml_clk_table->fclk.clk_values_khz[i] = 0;
+ dml_clk_table->fclk.clk_values_khz[i] = dc_clk_table->entries[i].fclk_mhz * 1000;
}
+ } else {
+ dml_clk_table->fclk.clk_values_khz[i] = 0;
}
}
+ }
- /* uclk */
- if (dc_clk_table->num_entries_per_clk.num_memclk_levels) {
- dml_clk_table->uclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_memclk_levels;
- for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
- if (i < dml_clk_table->uclk.num_clk_values) {
- if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.memclk_mhz &&
- dc_clk_table->entries[i].memclk_mhz > dc_bw_params->dc_mode_limit.memclk_mhz) {
- if (i == 0 || dc_clk_table->entries[i-1].memclk_mhz < dc_bw_params->dc_mode_limit.memclk_mhz) {
- dml_clk_table->uclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.memclk_mhz * 1000;
- dml_clk_table->uclk.num_clk_values = i + 1;
- } else {
- dml_clk_table->uclk.clk_values_khz[i] = 0;
- dml_clk_table->uclk.num_clk_values = i;
- }
+ /* uclk */
+ if (dc_clk_table->num_entries_per_clk.num_memclk_levels) {
+ dml_clk_table->uclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_memclk_levels;
+ for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
+ if (i < dml_clk_table->uclk.num_clk_values) {
+ if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.memclk_mhz &&
+ dc_clk_table->entries[i].memclk_mhz > dc_bw_params->dc_mode_limit.memclk_mhz) {
+ if (i == 0 || dc_clk_table->entries[i-1].memclk_mhz < dc_bw_params->dc_mode_limit.memclk_mhz) {
+ dml_clk_table->uclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.memclk_mhz * 1000;
+ dml_clk_table->uclk.num_clk_values = i + 1;
} else {
- dml_clk_table->uclk.clk_values_khz[i] = dc_clk_table->entries[i].memclk_mhz * 1000;
+ dml_clk_table->uclk.clk_values_khz[i] = 0;
+ dml_clk_table->uclk.num_clk_values = i;
}
} else {
- dml_clk_table->uclk.clk_values_khz[i] = 0;
+ dml_clk_table->uclk.clk_values_khz[i] = dc_clk_table->entries[i].memclk_mhz * 1000;
}
+ } else {
+ dml_clk_table->uclk.clk_values_khz[i] = 0;
}
}
+ }
- /* dispclk */
- if (dc_clk_table->num_entries_per_clk.num_dispclk_levels) {
- dml_clk_table->dispclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_dispclk_levels;
- for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
- if (i < dml_clk_table->dispclk.num_clk_values) {
- if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.dispclk_mhz &&
- dc_clk_table->entries[i].dispclk_mhz > dc_bw_params->dc_mode_limit.dispclk_mhz) {
- if (i == 0 || dc_clk_table->entries[i-1].dispclk_mhz < dc_bw_params->dc_mode_limit.dispclk_mhz) {
- dml_clk_table->dispclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.dispclk_mhz * 1000;
- dml_clk_table->dispclk.num_clk_values = i + 1;
- } else {
- dml_clk_table->dispclk.clk_values_khz[i] = 0;
- dml_clk_table->dispclk.num_clk_values = i;
- }
+ /* dispclk */
+ if (dc_clk_table->num_entries_per_clk.num_dispclk_levels) {
+ dml_clk_table->dispclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_dispclk_levels;
+ for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
+ if (i < dml_clk_table->dispclk.num_clk_values) {
+ if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.dispclk_mhz &&
+ dc_clk_table->entries[i].dispclk_mhz > dc_bw_params->dc_mode_limit.dispclk_mhz) {
+ if (i == 0 || dc_clk_table->entries[i-1].dispclk_mhz < dc_bw_params->dc_mode_limit.dispclk_mhz) {
+ dml_clk_table->dispclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.dispclk_mhz * 1000;
+ dml_clk_table->dispclk.num_clk_values = i + 1;
} else {
- dml_clk_table->dispclk.clk_values_khz[i] = dc_clk_table->entries[i].dispclk_mhz * 1000;
+ dml_clk_table->dispclk.clk_values_khz[i] = 0;
+ dml_clk_table->dispclk.num_clk_values = i;
}
} else {
- dml_clk_table->dispclk.clk_values_khz[i] = 0;
+ dml_clk_table->dispclk.clk_values_khz[i] = dc_clk_table->entries[i].dispclk_mhz * 1000;
}
+ } else {
+ dml_clk_table->dispclk.clk_values_khz[i] = 0;
}
}
+ }
- /* dppclk */
- if (dc_clk_table->num_entries_per_clk.num_dppclk_levels) {
- dml_clk_table->dppclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_dppclk_levels;
- for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
- if (i < dml_clk_table->dppclk.num_clk_values) {
- if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.dppclk_mhz &&
- dc_clk_table->entries[i].dppclk_mhz > dc_bw_params->dc_mode_limit.dppclk_mhz) {
- if (i == 0 || dc_clk_table->entries[i-1].dppclk_mhz < dc_bw_params->dc_mode_limit.dppclk_mhz) {
- dml_clk_table->dppclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.dppclk_mhz * 1000;
- dml_clk_table->dppclk.num_clk_values = i + 1;
- } else {
- dml_clk_table->dppclk.clk_values_khz[i] = 0;
- dml_clk_table->dppclk.num_clk_values = i;
- }
+ /* dppclk */
+ if (dc_clk_table->num_entries_per_clk.num_dppclk_levels) {
+ dml_clk_table->dppclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_dppclk_levels;
+ for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
+ if (i < dml_clk_table->dppclk.num_clk_values) {
+ if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.dppclk_mhz &&
+ dc_clk_table->entries[i].dppclk_mhz > dc_bw_params->dc_mode_limit.dppclk_mhz) {
+ if (i == 0 || dc_clk_table->entries[i-1].dppclk_mhz < dc_bw_params->dc_mode_limit.dppclk_mhz) {
+ dml_clk_table->dppclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.dppclk_mhz * 1000;
+ dml_clk_table->dppclk.num_clk_values = i + 1;
} else {
- dml_clk_table->dppclk.clk_values_khz[i] = dc_clk_table->entries[i].dppclk_mhz * 1000;
+ dml_clk_table->dppclk.clk_values_khz[i] = 0;
+ dml_clk_table->dppclk.num_clk_values = i;
}
} else {
- dml_clk_table->dppclk.clk_values_khz[i] = 0;
+ dml_clk_table->dppclk.clk_values_khz[i] = dc_clk_table->entries[i].dppclk_mhz * 1000;
}
+ } else {
+ dml_clk_table->dppclk.clk_values_khz[i] = 0;
}
}
+ }
- /* dtbclk */
- if (dc_clk_table->num_entries_per_clk.num_dtbclk_levels) {
- dml_clk_table->dtbclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_dtbclk_levels;
- for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
- if (i < dml_clk_table->dtbclk.num_clk_values) {
- if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.dtbclk_mhz &&
- dc_clk_table->entries[i].dtbclk_mhz > dc_bw_params->dc_mode_limit.dtbclk_mhz) {
- if (i == 0 || dc_clk_table->entries[i-1].dtbclk_mhz < dc_bw_params->dc_mode_limit.dtbclk_mhz) {
- dml_clk_table->dtbclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.dtbclk_mhz * 1000;
- dml_clk_table->dtbclk.num_clk_values = i + 1;
- } else {
- dml_clk_table->dtbclk.clk_values_khz[i] = 0;
- dml_clk_table->dtbclk.num_clk_values = i;
- }
+ /* dtbclk */
+ if (dc_clk_table->num_entries_per_clk.num_dtbclk_levels) {
+ dml_clk_table->dtbclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_dtbclk_levels;
+ for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
+ if (i < dml_clk_table->dtbclk.num_clk_values) {
+ if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.dtbclk_mhz &&
+ dc_clk_table->entries[i].dtbclk_mhz > dc_bw_params->dc_mode_limit.dtbclk_mhz) {
+ if (i == 0 || dc_clk_table->entries[i-1].dtbclk_mhz < dc_bw_params->dc_mode_limit.dtbclk_mhz) {
+ dml_clk_table->dtbclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.dtbclk_mhz * 1000;
+ dml_clk_table->dtbclk.num_clk_values = i + 1;
} else {
- dml_clk_table->dtbclk.clk_values_khz[i] = dc_clk_table->entries[i].dtbclk_mhz * 1000;
+ dml_clk_table->dtbclk.clk_values_khz[i] = 0;
+ dml_clk_table->dtbclk.num_clk_values = i;
}
} else {
- dml_clk_table->dtbclk.clk_values_khz[i] = 0;
+ dml_clk_table->dtbclk.clk_values_khz[i] = dc_clk_table->entries[i].dtbclk_mhz * 1000;
}
+ } else {
+ dml_clk_table->dtbclk.clk_values_khz[i] = 0;
}
}
+ }
- /* socclk */
- if (dc_clk_table->num_entries_per_clk.num_socclk_levels) {
- dml_clk_table->socclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_socclk_levels;
- for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
- if (i < dml_clk_table->socclk.num_clk_values) {
- if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.socclk_mhz &&
- dc_clk_table->entries[i].socclk_mhz > dc_bw_params->dc_mode_limit.socclk_mhz) {
- if (i == 0 || dc_clk_table->entries[i-1].socclk_mhz < dc_bw_params->dc_mode_limit.socclk_mhz) {
- dml_clk_table->socclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.socclk_mhz * 1000;
- dml_clk_table->socclk.num_clk_values = i + 1;
- } else {
- dml_clk_table->socclk.clk_values_khz[i] = 0;
- dml_clk_table->socclk.num_clk_values = i;
- }
+ /* socclk */
+ if (dc_clk_table->num_entries_per_clk.num_socclk_levels) {
+ dml_clk_table->socclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_socclk_levels;
+ for (i = 0; i < min(DML_MAX_CLK_TABLE_SIZE, MAX_NUM_DPM_LVL); i++) {
+ if (i < dml_clk_table->socclk.num_clk_values) {
+ if (config->use_clock_dc_limits && dc_bw_params->dc_mode_limit.socclk_mhz &&
+ dc_clk_table->entries[i].socclk_mhz > dc_bw_params->dc_mode_limit.socclk_mhz) {
+ if (i == 0 || dc_clk_table->entries[i-1].socclk_mhz < dc_bw_params->dc_mode_limit.socclk_mhz) {
+ dml_clk_table->socclk.clk_values_khz[i] = dc_bw_params->dc_mode_limit.socclk_mhz * 1000;
+ dml_clk_table->socclk.num_clk_values = i + 1;
} else {
- dml_clk_table->socclk.clk_values_khz[i] = dc_clk_table->entries[i].socclk_mhz * 1000;
+ dml_clk_table->socclk.clk_values_khz[i] = 0;
+ dml_clk_table->socclk.num_clk_values = i;
}
} else {
- dml_clk_table->socclk.clk_values_khz[i] = 0;
+ dml_clk_table->socclk.clk_values_khz[i] = dc_clk_table->entries[i].socclk_mhz * 1000;
}
+ } else {
+ dml_clk_table->socclk.clk_values_khz[i] = 0;
}
}
-
- /* do not override phyclks for now */
- /* phyclk */
- // dml_clk_table->phyclk.num_clk_values = dc_clk_table->num_entries_per_clk.num_phyclk_levels;
- // for (i = 0; i < DML_MAX_CLK_TABLE_SIZE; i++) {
- // dml_clk_table->phyclk.clk_values_khz[i] = dc_clk_table->entries[i].phyclk_mhz * 1000;
- // }
-
- /* phyclk_d18 */
- // dml_clk_table->phyclk_d18.num_clk_values = dc_clk_table->num_entries_per_clk.num_phyclk_d18_levels;
- // for (i = 0; i < DML_MAX_CLK_TABLE_SIZE; i++) {
- // dml_clk_table->phyclk_d18.clk_values_khz[i] = dc_clk_table->entries[i].phyclk_d18_mhz * 1000;
- // }
-
- /* phyclk_d32 */
- // dml_clk_table->phyclk_d32.num_clk_values = dc_clk_table->num_entries_per_clk.num_phyclk_d32_levels;
- // for (i = 0; i < DML_MAX_CLK_TABLE_SIZE; i++) {
- // dml_clk_table->phyclk_d32.clk_values_khz[i] = dc_clk_table->entries[i].phyclk_d32_mhz * 1000;
- // }
}
+}
- dml_soc_bb->dchub_refclk_mhz = in_dc->res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000;
- dml_soc_bb->dprefclk_mhz = in_dc->clk_mgr->dprefclk_khz / 1000;
- dml_soc_bb->xtalclk_mhz = in_dc->ctx->dc_bios->fw_info.pll_info.crystal_frequency / 1000;
- dml_soc_bb->dispclk_dppclk_vco_speed_mhz = in_dc->clk_mgr->dentist_vco_freq_khz / 1000.0;
+static void override_dml_init_with_values_from_vbios(
+ struct dml2_initialize_instance_in_out *dml_init,
+ const struct dml2_configuration_options *config,
+ const struct dc *in_dc)
+{
+ const struct clk_bw_params *dc_bw_params = in_dc->clk_mgr->bw_params;
+ struct dml2_soc_bb *dml_soc_bb = &dml_init->soc_bb;
+ struct dml2_soc_state_table *dml_clk_table = &dml_init->soc_bb.clk_table;
- /* override bounding box paramters from VBIOS */
if (in_dc->ctx->dc_bios->bb_info.dram_clock_change_latency_100ns > 0)
dml_soc_bb->power_management_parameters.dram_clk_change_blackout_us =
(in_dc->ctx->dc_bios->bb_info.dram_clock_change_latency_100ns + 9) / 10;
@@ -302,48 +279,152 @@ void dml21_apply_soc_bb_overrides(struct dml2_initialize_instance_in_out *dml_in
dml_soc_bb->power_management_parameters.stutter_exit_latency_us =
(in_dc->ctx->dc_bios->bb_info.dram_sr_exit_latency_100ns + 9) / 10;
- if (in_dc->ctx->dc_bios->vram_info.num_chans) {
+ if (dc_bw_params->num_channels) {
+ dml_clk_table->dram_config.channel_count = dc_bw_params->num_channels;
+ dml_soc_bb->mall_allocated_for_dcn_mbytes = in_dc->caps.mall_size_total / 1048576;
+ } else if (in_dc->ctx->dc_bios->vram_info.num_chans) {
dml_clk_table->dram_config.channel_count = in_dc->ctx->dc_bios->vram_info.num_chans;
dml_soc_bb->mall_allocated_for_dcn_mbytes = in_dc->caps.mall_size_total / 1048576;
}
- if (in_dc->ctx->dc_bios->vram_info.dram_channel_width_bytes) {
+ if (dc_bw_params->dram_channel_width_bytes) {
+ dml_clk_table->dram_config.channel_width_bytes = dc_bw_params->dram_channel_width_bytes;
+ } else if (in_dc->ctx->dc_bios->vram_info.dram_channel_width_bytes) {
dml_clk_table->dram_config.channel_width_bytes = in_dc->ctx->dc_bios->vram_info.dram_channel_width_bytes;
}
- /* override bounding box paramters from DC config */
- if (in_dc->bb_overrides.sr_exit_time_ns) {
- dml_soc_bb->power_management_parameters.stutter_exit_latency_us =
- in_dc->bb_overrides.sr_exit_time_ns / 1000.0;
+ dml_init->soc_bb.xtalclk_mhz = in_dc->ctx->dc_bios->fw_info.pll_info.crystal_frequency / 1000;
+}
+
+
+static void override_dml_init_with_values_from_dmub(struct dml2_initialize_instance_in_out *dml_init,
+ const struct dml2_configuration_options *config,
+ const struct dc *in_dc)
+{
+ /*
+ * TODO - There seems to be overlaps between the values overriden from
+ * dmub and vbios. Investigate and identify the values that DMUB needs
+ * to own.
+ */
+// const struct dmub_soc_bb_params *dmub_bb_params =
+// (const struct dmub_soc_bb_params *)config->bb_from_dmub;
+
+// if (dmub_bb_params == NULL)
+// return;
+
+// if (dmub_bb_params->dram_clk_change_blackout_ns > 0)
+// dml_init->soc_bb.power_management_parameters.dram_clk_change_blackout_us =
+// (double) dmub_bb_params->dram_clk_change_blackout_ns / 1000.0;
+// if (dmub_bb_params->dram_clk_change_read_only_ns > 0)
+// dml_init->soc_bb.power_management_parameters.dram_clk_change_read_only_us =
+// (double) dmub_bb_params->dram_clk_change_read_only_ns / 1000.0;
+// if (dmub_bb_params->dram_clk_change_write_only_ns > 0)
+// dml_init->soc_bb.power_management_parameters.dram_clk_change_write_only_us =
+// (double) dmub_bb_params->dram_clk_change_write_only_ns / 1000.0;
+// if (dmub_bb_params->fclk_change_blackout_ns > 0)
+// dml_init->soc_bb.power_management_parameters.fclk_change_blackout_us =
+// (double) dmub_bb_params->fclk_change_blackout_ns / 1000.0;
+// if (dmub_bb_params->g7_ppt_blackout_ns > 0)
+// dml_init->soc_bb.power_management_parameters.g7_ppt_blackout_us =
+// (double) dmub_bb_params->g7_ppt_blackout_ns / 1000.0;
+// if (dmub_bb_params->stutter_enter_plus_exit_latency_ns > 0)
+// dml_init->soc_bb.power_management_parameters.stutter_enter_plus_exit_latency_us =
+// (double) dmub_bb_params->stutter_enter_plus_exit_latency_ns / 1000.0;
+// if (dmub_bb_params->stutter_exit_latency_ns > 0)
+// dml_init->soc_bb.power_management_parameters.stutter_exit_latency_us =
+// (double) dmub_bb_params->stutter_exit_latency_ns / 1000.0;
+// if (dmub_bb_params->z8_stutter_enter_plus_exit_latency_ns > 0)
+// dml_init->soc_bb.power_management_parameters.z8_stutter_enter_plus_exit_latency_us =
+// (double) dmub_bb_params->z8_stutter_enter_plus_exit_latency_ns / 1000.0;
+// if (dmub_bb_params->z8_stutter_exit_latency_ns > 0)
+// dml_init->soc_bb.power_management_parameters.z8_stutter_exit_latency_us =
+// (double) dmub_bb_params->z8_stutter_exit_latency_ns / 1000.0;
+// if (dmub_bb_params->z8_min_idle_time_ns > 0)
+// dml_init->soc_bb.power_management_parameters.z8_min_idle_time =
+// (double) dmub_bb_params->z8_min_idle_time_ns / 1000.0;
+// #ifndef TRIM_DML2_DCN6B_IP_SENSITIVE
+// if (dmub_bb_params->type_b_dram_clk_change_blackout_ns > 0)
+// dml_init->soc_bb.power_management_parameters.lpddr5_dram_clk_change_blackout_us =
+// (double) dmub_bb_params->type_b_dram_clk_change_blackout_ns / 1000.0;
+// if (dmub_bb_params->type_b_ppt_blackout_ns > 0)
+// dml_init->soc_bb.power_management_parameters.lpddr5_ppt_blackout_us =
+// (double) dmub_bb_params->type_b_ppt_blackout_ns / 1000.0;
+// #else
+// if (dmub_bb_params->type_b_dram_clk_change_blackout_ns > 0)
+// dml_init->soc_bb.power_management_parameters.type_b_dram_clk_change_blackout_us =
+// (double) dmub_bb_params->type_b_dram_clk_change_blackout_ns / 1000.0;
+// if (dmub_bb_params->type_b_ppt_blackout_ns > 0)
+// dml_init->soc_bb.power_management_parameters.type_b_ppt_blackout_us =
+// (double) dmub_bb_params->type_b_ppt_blackout_ns / 1000.0;
+// #endif
+// if (dmub_bb_params->vmin_limit_dispclk_khz > 0)
+// dml_init->soc_bb.vmin_limit.dispclk_khz = dmub_bb_params->vmin_limit_dispclk_khz;
+// if (dmub_bb_params->vmin_limit_dcfclk_khz > 0)
+// dml_init->soc_bb.vmin_limit.dcfclk_khz = dmub_bb_params->vmin_limit_dcfclk_khz;
+// if (dmub_bb_params->g7_temperature_read_blackout_ns > 0)
+// dml_init->soc_bb.power_management_parameters.g7_temperature_read_blackout_us =
+// (double) dmub_bb_params->g7_temperature_read_blackout_ns / 1000.0;
+}
+
+static void override_dml_init_with_values_from_software_policy(struct dml2_initialize_instance_in_out *dml_init,
+ const struct dml2_configuration_options *config,
+ const struct dc *in_dc)
+{
+ if (!config->use_native_soc_bb_construction) {
+ dml_init->soc_bb = config->external_socbb_ip_params->soc_bb;
+ dml_init->ip_caps = config->external_socbb_ip_params->ip_params;
}
- if (in_dc->bb_overrides.sr_enter_plus_exit_time_ns) {
- dml_soc_bb->power_management_parameters.stutter_enter_plus_exit_latency_us =
+ if (in_dc->bb_overrides.sr_exit_time_ns)
+ dml_init->soc_bb.power_management_parameters.stutter_exit_latency_us =
+ in_dc->bb_overrides.sr_exit_time_ns / 1000.0;
+
+ if (in_dc->bb_overrides.sr_enter_plus_exit_time_ns)
+ dml_init->soc_bb.power_management_parameters.stutter_enter_plus_exit_latency_us =
in_dc->bb_overrides.sr_enter_plus_exit_time_ns / 1000.0;
- }
- if (in_dc->bb_overrides.dram_clock_change_latency_ns) {
- dml_soc_bb->power_management_parameters.dram_clk_change_blackout_us =
+ if (in_dc->bb_overrides.dram_clock_change_latency_ns)
+ dml_init->soc_bb.power_management_parameters.dram_clk_change_blackout_us =
in_dc->bb_overrides.dram_clock_change_latency_ns / 1000.0;
- }
- if (in_dc->bb_overrides.fclk_clock_change_latency_ns) {
- dml_soc_bb->power_management_parameters.fclk_change_blackout_us =
+ if (in_dc->bb_overrides.fclk_clock_change_latency_ns)
+ dml_init->soc_bb.power_management_parameters.fclk_change_blackout_us =
in_dc->bb_overrides.fclk_clock_change_latency_ns / 1000.0;
+}
+
+void dml21_populate_dml_init_params(struct dml2_initialize_instance_in_out *dml_init,
+ const struct dml2_configuration_options *config,
+ const struct dc *in_dc)
+{
+ populate_default_dml_init_params(dml_init, config, in_dc);
+
+ override_dml_init_with_values_from_hardware_default(dml_init, config, in_dc);
+
+ override_dml_init_with_values_from_smu(dml_init, config, in_dc);
+
+ override_dml_init_with_values_from_vbios(dml_init, config, in_dc);
+
+ override_dml_init_with_values_from_dmub(dml_init, config, in_dc);
+
+ override_dml_init_with_values_from_software_policy(dml_init, config, in_dc);
+}
+
+static unsigned int calc_max_hardware_v_total(const struct dc_stream_state *stream)
+{
+ unsigned int max_hw_v_total = stream->ctx->dc->caps.max_v_total;
+
+ if (stream->ctx->dc->caps.vtotal_limited_by_fp2) {
+ max_hw_v_total -= stream->timing.v_front_porch + 1;
}
- //TODO
- // if (in_dc->bb_overrides.dummy_clock_change_latency_ns) {
- // dml_soc_bb->power_management_parameters.dram_clk_change_blackout_us =
- // in_dc->bb_overrides.dram_clock_change_latency_ns / 1000.0;
- // }
+ return max_hw_v_total;
}
static void populate_dml21_timing_config_from_stream_state(struct dml2_timing_cfg *timing,
struct dc_stream_state *stream,
struct dml2_context *dml_ctx)
{
- unsigned int hblank_start, vblank_start;
+ unsigned int hblank_start, vblank_start, min_hardware_refresh_in_uhz;
timing->h_active = stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right;
timing->v_active = stream->timing.v_addressable + stream->timing.v_border_bottom + stream->timing.v_border_top;
@@ -371,11 +452,23 @@ static void populate_dml21_timing_config_from_stream_state(struct dml2_timing_cf
- stream->timing.v_border_top - stream->timing.v_border_bottom;
timing->drr_config.enabled = stream->ignore_msa_timing_param;
- timing->drr_config.min_refresh_uhz = stream->timing.min_refresh_in_uhz;
timing->drr_config.drr_active_variable = stream->vrr_active_variable;
timing->drr_config.drr_active_fixed = stream->vrr_active_fixed;
timing->drr_config.disallowed = !stream->allow_freesync;
+ /* limit min refresh rate to DC cap */
+ min_hardware_refresh_in_uhz = stream->timing.min_refresh_in_uhz;
+ if (stream->ctx->dc->caps.max_v_total != 0) {
+ min_hardware_refresh_in_uhz = div64_u64((stream->timing.pix_clk_100hz * 100000000ULL),
+ (stream->timing.h_total * (long long)calc_max_hardware_v_total(stream)));
+ }
+
+ if (stream->timing.min_refresh_in_uhz > min_hardware_refresh_in_uhz) {
+ timing->drr_config.min_refresh_uhz = stream->timing.min_refresh_in_uhz;
+ } else {
+ timing->drr_config.min_refresh_uhz = min_hardware_refresh_in_uhz;
+ }
+
if (dml_ctx->config.callbacks.get_max_flickerless_instant_vtotal_increase &&
stream->ctx->dc->config.enable_fpo_flicker_detection == 1)
timing->drr_config.max_instant_vtotal_delta = dml_ctx->config.callbacks.get_max_flickerless_instant_vtotal_increase(stream, false);
@@ -422,6 +515,21 @@ static void populate_dml21_timing_config_from_stream_state(struct dml2_timing_cf
timing->vblank_nom = timing->v_total - timing->v_active;
}
+/**
+ * adjust_dml21_hblank_timing_config_from_pipe_ctx - Adjusts the horizontal blanking timing configuration
+ * based on the pipe context.
+ * @timing: Pointer to the dml2_timing_cfg structure to be adjusted.
+ * @pipe: Pointer to the pipe_ctx structure containing the horizontal blanking borrow value.
+ *
+ * This function modifies the horizontal active and blank end timings by adding and subtracting
+ * the horizontal blanking borrow value from the pipe context, respectively.
+ */
+static void adjust_dml21_hblank_timing_config_from_pipe_ctx(struct dml2_timing_cfg *timing, struct pipe_ctx *pipe)
+{
+ timing->h_active += pipe->hblank_borrow;
+ timing->h_blank_end -= pipe->hblank_borrow;
+}
+
static void populate_dml21_output_config_from_stream_state(struct dml2_link_output_cfg *output,
struct dc_stream_state *stream, const struct pipe_ctx *pipe)
{
@@ -491,7 +599,8 @@ static void populate_dml21_output_config_from_stream_state(struct dml2_link_outp
static void populate_dml21_stream_overrides_from_stream_state(
struct dml2_stream_parameters *stream_desc,
- struct dc_stream_state *stream)
+ struct dc_stream_state *stream,
+ struct dc_stream_status *stream_status)
{
switch (stream->debug.force_odm_combine_segments) {
case 0:
@@ -516,7 +625,9 @@ static void populate_dml21_stream_overrides_from_stream_state(
if (!stream->ctx->dc->debug.enable_single_display_2to1_odm_policy ||
stream->debug.force_odm_combine_segments > 0)
stream_desc->overrides.disable_dynamic_odm = true;
- stream_desc->overrides.disable_subvp = stream->ctx->dc->debug.force_disable_subvp || stream->hw_cursor_req;
+ stream_desc->overrides.disable_subvp = stream->ctx->dc->debug.force_disable_subvp ||
+ stream->hw_cursor_req ||
+ stream_status->mall_stream_config.cursor_size_limit_subvp;
}
static enum dml2_swizzle_mode gfx_addr3_to_dml2_swizzle_mode(enum swizzle_mode_addr3_values addr3_mode)
@@ -683,11 +794,20 @@ static void populate_dml21_surface_config_from_plane_state(
surface->dcc.informative.fraction_of_zero_size_request_plane1 = plane_state->dcc.independent_64b_blks_c;
surface->dcc.plane0.pitch = plane_state->dcc.meta_pitch;
surface->dcc.plane1.pitch = plane_state->dcc.meta_pitch_c;
- if (in_dc->ctx->dce_version < DCN_VERSION_4_01) {
- /* needed for N-1 testing */
+
+ // Update swizzle / array mode based on the gfx_format
+ switch (plane_state->tiling_info.gfxversion) {
+ case DcGfxVersion7:
+ case DcGfxVersion8:
+ break;
+ case DcGfxVersion9:
+ case DcGfxVersion10:
+ case DcGfxVersion11:
surface->tiling = gfx9_to_dml2_swizzle_mode(plane_state->tiling_info.gfx9.swizzle);
- } else {
+ break;
+ case DcGfxAddr3:
surface->tiling = gfx_addr3_to_dml2_swizzle_mode(plane_state->tiling_info.gfx_addr3.swizzle);
+ break;
}
}
@@ -709,6 +829,7 @@ static const struct scaler_data *get_scaler_data_for_plane(
temp_pipe->plane_state = pipe->plane_state;
temp_pipe->plane_res.scl_data.taps = pipe->plane_res.scl_data.taps;
temp_pipe->stream_res = pipe->stream_res;
+ temp_pipe->hblank_borrow = pipe->hblank_borrow;
dml_ctx->config.callbacks.build_scaling_params(temp_pipe);
break;
}
@@ -739,6 +860,7 @@ static void populate_dml21_plane_config_from_plane_state(struct dml2_context *dm
plane->pixel_format = dml2_420_10;
break;
case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616:
+ case SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616:
case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616F:
case SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616F:
plane->pixel_format = dml2_444_64;
@@ -840,6 +962,7 @@ static void populate_dml21_plane_config_from_plane_state(struct dml2_context *dm
plane->tdlut.tdlut_width_mode = dml2_tdlut_width_17_cube;
break;
case DC_CM2_GPU_MEM_SIZE_TRANSFORMED:
+ default:
//plane->tdlut.tdlut_width_mode = dml2_tdlut_width_flatten; // dml2_tdlut_width_flatten undefined
break;
}
@@ -864,7 +987,7 @@ static void populate_dml21_plane_config_from_plane_state(struct dml2_context *dm
}
//TODO : Could be possibly moved to a common helper layer.
-static bool dml21_wrapper_get_plane_id(const struct dc_state *context, const struct dc_plane_state *plane, unsigned int *plane_id)
+static bool dml21_wrapper_get_plane_id(const struct dc_state *context, unsigned int stream_id, const struct dc_plane_state *plane, unsigned int *plane_id)
{
int i, j;
@@ -872,10 +995,12 @@ static bool dml21_wrapper_get_plane_id(const struct dc_state *context, const str
return false;
for (i = 0; i < context->stream_count; i++) {
- for (j = 0; j < context->stream_status[i].plane_count; j++) {
- if (context->stream_status[i].plane_states[j] == plane) {
- *plane_id = (i << 16) | j;
- return true;
+ if (context->streams[i]->stream_id == stream_id) {
+ for (j = 0; j < context->stream_status[i].plane_count; j++) {
+ if (context->stream_status[i].plane_states[j] == plane) {
+ *plane_id = (i << 16) | j;
+ return true;
+ }
}
}
}
@@ -898,14 +1023,14 @@ static unsigned int map_stream_to_dml21_display_cfg(const struct dml2_context *d
return location;
}
-static unsigned int map_plane_to_dml21_display_cfg(const struct dml2_context *dml_ctx,
+unsigned int map_plane_to_dml21_display_cfg(const struct dml2_context *dml_ctx, unsigned int stream_id,
const struct dc_plane_state *plane, const struct dc_state *context)
{
unsigned int plane_id;
int i = 0;
int location = -1;
- if (!dml21_wrapper_get_plane_id(context, plane, &plane_id)) {
+ if (!dml21_wrapper_get_plane_id(context, stream_id, plane, &plane_id)) {
ASSERT(false);
return -1;
}
@@ -971,10 +1096,11 @@ bool dml21_map_dc_state_into_dml_display_cfg(const struct dc *in_dc, struct dc_s
if (disp_cfg_stream_location < 0)
disp_cfg_stream_location = dml_dispcfg->num_streams++;
- ASSERT(disp_cfg_stream_location >= 0 && disp_cfg_stream_location <= __DML2_WRAPPER_MAX_STREAMS_PLANES__);
+ ASSERT(disp_cfg_stream_location >= 0 && disp_cfg_stream_location < __DML2_WRAPPER_MAX_STREAMS_PLANES__);
populate_dml21_timing_config_from_stream_state(&dml_dispcfg->stream_descriptors[disp_cfg_stream_location].timing, context->streams[stream_index], dml_ctx);
+ adjust_dml21_hblank_timing_config_from_pipe_ctx(&dml_dispcfg->stream_descriptors[disp_cfg_stream_location].timing, &context->res_ctx.pipe_ctx[stream_index]);
populate_dml21_output_config_from_stream_state(&dml_dispcfg->stream_descriptors[disp_cfg_stream_location].output, context->streams[stream_index], &context->res_ctx.pipe_ctx[stream_index]);
- populate_dml21_stream_overrides_from_stream_state(&dml_dispcfg->stream_descriptors[disp_cfg_stream_location], context->streams[stream_index]);
+ populate_dml21_stream_overrides_from_stream_state(&dml_dispcfg->stream_descriptors[disp_cfg_stream_location], context->streams[stream_index], &context->stream_status[stream_index]);
dml_dispcfg->stream_descriptors[disp_cfg_stream_location].overrides.hw.twait_budgeting.fclk_pstate = dml2_twait_budgeting_setting_if_needed;
dml_dispcfg->stream_descriptors[disp_cfg_stream_location].overrides.hw.twait_budgeting.uclk_pstate = dml2_twait_budgeting_setting_if_needed;
@@ -990,18 +1116,18 @@ bool dml21_map_dc_state_into_dml_display_cfg(const struct dc *in_dc, struct dc_s
dml_dispcfg->plane_descriptors[disp_cfg_plane_location].stream_index = disp_cfg_stream_location;
} else {
for (plane_index = 0; plane_index < context->stream_status[stream_index].plane_count; plane_index++) {
- disp_cfg_plane_location = map_plane_to_dml21_display_cfg(dml_ctx, context->stream_status[stream_index].plane_states[plane_index], context);
+ disp_cfg_plane_location = map_plane_to_dml21_display_cfg(dml_ctx, context->streams[stream_index]->stream_id, context->stream_status[stream_index].plane_states[plane_index], context);
if (disp_cfg_plane_location < 0)
disp_cfg_plane_location = dml_dispcfg->num_planes++;
- ASSERT(disp_cfg_plane_location >= 0 && disp_cfg_plane_location <= __DML2_WRAPPER_MAX_STREAMS_PLANES__);
+ ASSERT(disp_cfg_plane_location >= 0 && disp_cfg_plane_location < __DML2_WRAPPER_MAX_STREAMS_PLANES__);
populate_dml21_surface_config_from_plane_state(in_dc, &dml_dispcfg->plane_descriptors[disp_cfg_plane_location].surface, context->stream_status[stream_index].plane_states[plane_index]);
populate_dml21_plane_config_from_plane_state(dml_ctx, &dml_dispcfg->plane_descriptors[disp_cfg_plane_location], context->stream_status[stream_index].plane_states[plane_index], context, stream_index);
dml_dispcfg->plane_descriptors[disp_cfg_plane_location].stream_index = disp_cfg_stream_location;
- if (dml21_wrapper_get_plane_id(context, context->stream_status[stream_index].plane_states[plane_index], &dml_ctx->v21.dml_to_dc_pipe_mapping.disp_cfg_to_plane_id[disp_cfg_plane_location]))
+ if (dml21_wrapper_get_plane_id(context, context->streams[stream_index]->stream_id, context->stream_status[stream_index].plane_states[plane_index], &dml_ctx->v21.dml_to_dc_pipe_mapping.disp_cfg_to_plane_id[disp_cfg_plane_location]))
dml_ctx->v21.dml_to_dc_pipe_mapping.disp_cfg_to_plane_id_valid[disp_cfg_plane_location] = true;
/* apply forced pstate policy */
@@ -1037,28 +1163,8 @@ void dml21_copy_clocks_to_dc_state(struct dml2_context *in_ctx, struct dc_state
context->bw_ctx.bw.dcn.clk.dtbclk_en = in_ctx->v21.mode_programming.programming->min_clocks.dcn4x.dtbrefclk_khz > 0;
context->bw_ctx.bw.dcn.clk.ref_dtbclk_khz = in_ctx->v21.mode_programming.programming->min_clocks.dcn4x.dtbrefclk_khz;
context->bw_ctx.bw.dcn.clk.socclk_khz = in_ctx->v21.mode_programming.programming->min_clocks.dcn4x.socclk_khz;
-}
-
-void dml21_extract_legacy_watermark_set(const struct dc *in_dc, struct dcn_watermarks *watermark, enum dml2_dchub_watermark_reg_set_index reg_set_idx, struct dml2_context *in_ctx)
-{
- struct dml2_core_internal_display_mode_lib *mode_lib = &in_ctx->v21.dml_init.dml2_instance->core_instance.clean_me_up.mode_lib;
- double refclk_freq_in_mhz = (in_ctx->v21.display_config.overrides.hw.dlg_ref_clk_mhz > 0) ? (double)in_ctx->v21.display_config.overrides.hw.dlg_ref_clk_mhz : mode_lib->soc.dchub_refclk_mhz;
-
- if (reg_set_idx >= DML2_DCHUB_WATERMARK_SET_NUM) {
- /* invalid register set index */
- return;
- }
-
- /* convert to legacy format (time in ns) */
- watermark->urgent_ns = ((double)in_ctx->v21.mode_programming.programming->global_regs.wm_regs[reg_set_idx].urgent / refclk_freq_in_mhz) * 1000.0;
- watermark->pte_meta_urgent_ns = ((double)in_ctx->v21.mode_programming.programming->global_regs.wm_regs[reg_set_idx].urgent / refclk_freq_in_mhz) * 1000.0;
- watermark->cstate_pstate.cstate_enter_plus_exit_ns = ((double)in_ctx->v21.mode_programming.programming->global_regs.wm_regs[reg_set_idx].sr_enter / refclk_freq_in_mhz) * 1000.0;
- watermark->cstate_pstate.cstate_exit_ns = ((double)in_ctx->v21.mode_programming.programming->global_regs.wm_regs[reg_set_idx].sr_exit / refclk_freq_in_mhz) * 1000.0;
- watermark->cstate_pstate.pstate_change_ns = ((double)in_ctx->v21.mode_programming.programming->global_regs.wm_regs[reg_set_idx].uclk_pstate / refclk_freq_in_mhz) * 1000.0;
- watermark->urgent_latency_ns = ((double)in_ctx->v21.mode_programming.programming->global_regs.wm_regs[reg_set_idx].urgent / refclk_freq_in_mhz) * 1000.0;
- watermark->cstate_pstate.fclk_pstate_change_ns = ((double)in_ctx->v21.mode_programming.programming->global_regs.wm_regs[reg_set_idx].fclk_pstate / refclk_freq_in_mhz) * 1000.0;
- watermark->frac_urg_bw_flip = in_ctx->v21.mode_programming.programming->global_regs.wm_regs[reg_set_idx].frac_urg_bw_flip;
- watermark->frac_urg_bw_nom = in_ctx->v21.mode_programming.programming->global_regs.wm_regs[reg_set_idx].frac_urg_bw_nom;
+ context->bw_ctx.bw.dcn.clk.subvp_prefetch_dramclk_khz = in_ctx->v21.mode_programming.programming->min_clocks.dcn4x.svp_prefetch_no_throttle.uclk_khz;
+ context->bw_ctx.bw.dcn.clk.subvp_prefetch_fclk_khz = in_ctx->v21.mode_programming.programming->min_clocks.dcn4x.svp_prefetch_no_throttle.fclk_khz;
}
static struct dml2_dchub_watermark_regs *wm_set_index_to_dc_wm_set(union dcn_watermark_set *watermarks, const enum dml2_dchub_watermark_reg_set_index wm_index)
@@ -1104,53 +1210,6 @@ void dml21_extract_watermark_sets(const struct dc *in_dc, union dcn_watermark_se
}
}
-
-void dml21_populate_pipe_ctx_dlg_params(struct dml2_context *dml_ctx, struct dc_state *context, struct pipe_ctx *pipe_ctx, struct dml2_per_stream_programming *stream_programming)
-{
- unsigned int hactive, vactive, hblank_start, vblank_start, hblank_end, vblank_end;
- struct dc_crtc_timing *timing = &pipe_ctx->stream->timing;
- union dml2_global_sync_programming *global_sync = &stream_programming->global_sync;
-
- hactive = timing->h_addressable + timing->h_border_left + timing->h_border_right;
- vactive = timing->v_addressable + timing->v_border_bottom + timing->v_border_top;
- hblank_start = pipe_ctx->stream->timing.h_total - pipe_ctx->stream->timing.h_front_porch;
- vblank_start = pipe_ctx->stream->timing.v_total - pipe_ctx->stream->timing.v_front_porch;
-
- hblank_end = hblank_start - timing->h_addressable - timing->h_border_left - timing->h_border_right;
- vblank_end = vblank_start - timing->v_addressable - timing->v_border_top - timing->v_border_bottom;
-
- if (dml_ctx->config.svp_pstate.callbacks.get_pipe_subvp_type(context, pipe_ctx) == SUBVP_PHANTOM) {
- /* phantom has its own global sync */
- global_sync = &stream_programming->phantom_stream.global_sync;
- }
-
- pipe_ctx->pipe_dlg_param.vstartup_start = global_sync->dcn4x.vstartup_lines;
- pipe_ctx->pipe_dlg_param.vupdate_offset = global_sync->dcn4x.vupdate_offset_pixels;
- pipe_ctx->pipe_dlg_param.vupdate_width = global_sync->dcn4x.vupdate_vupdate_width_pixels;
- pipe_ctx->pipe_dlg_param.vready_offset = global_sync->dcn4x.vready_offset_pixels;
- pipe_ctx->pipe_dlg_param.pstate_keepout = global_sync->dcn4x.pstate_keepout_start_lines;
-
- pipe_ctx->pipe_dlg_param.otg_inst = pipe_ctx->stream_res.tg->inst;
-
- pipe_ctx->pipe_dlg_param.hactive = hactive;
- pipe_ctx->pipe_dlg_param.vactive = vactive;
- pipe_ctx->pipe_dlg_param.htotal = pipe_ctx->stream->timing.h_total;
- pipe_ctx->pipe_dlg_param.vtotal = pipe_ctx->stream->timing.v_total;
- pipe_ctx->pipe_dlg_param.hblank_end = hblank_end;
- pipe_ctx->pipe_dlg_param.vblank_end = vblank_end;
- pipe_ctx->pipe_dlg_param.hblank_start = hblank_start;
- pipe_ctx->pipe_dlg_param.vblank_start = vblank_start;
- pipe_ctx->pipe_dlg_param.vfront_porch = pipe_ctx->stream->timing.v_front_porch;
- pipe_ctx->pipe_dlg_param.pixel_rate_mhz = pipe_ctx->stream->timing.pix_clk_100hz / 10000.00;
- pipe_ctx->pipe_dlg_param.refresh_rate = ((timing->pix_clk_100hz * 100) / timing->h_total) / timing->v_total;
- pipe_ctx->pipe_dlg_param.vtotal_max = pipe_ctx->stream->adjust.v_total_max;
- pipe_ctx->pipe_dlg_param.vtotal_min = pipe_ctx->stream->adjust.v_total_min;
- pipe_ctx->pipe_dlg_param.recout_height = pipe_ctx->plane_res.scl_data.recout.height;
- pipe_ctx->pipe_dlg_param.recout_width = pipe_ctx->plane_res.scl_data.recout.width;
- pipe_ctx->pipe_dlg_param.full_recout_height = pipe_ctx->plane_res.scl_data.recout.height;
- pipe_ctx->pipe_dlg_param.full_recout_width = pipe_ctx->plane_res.scl_data.recout.width;
-}
-
void dml21_map_hw_resources(struct dml2_context *dml_ctx)
{
unsigned int i = 0;
@@ -1186,22 +1245,22 @@ void dml21_set_dc_p_state_type(
bool sub_vp_enabled)
{
switch (stream_programming->uclk_pstate_method) {
- case dml2_uclk_pstate_support_method_vactive:
- case dml2_uclk_pstate_support_method_fw_vactive_drr:
+ case dml2_pstate_method_vactive:
+ case dml2_pstate_method_fw_vactive_drr:
pipe_ctx->p_state_type = P_STATE_V_ACTIVE;
break;
- case dml2_uclk_pstate_support_method_vblank:
- case dml2_uclk_pstate_support_method_fw_vblank_drr:
+ case dml2_pstate_method_vblank:
+ case dml2_pstate_method_fw_vblank_drr:
if (sub_vp_enabled)
pipe_ctx->p_state_type = P_STATE_V_BLANK_SUB_VP;
else
pipe_ctx->p_state_type = P_STATE_V_BLANK;
break;
- case dml2_uclk_pstate_support_method_fw_subvp_phantom:
- case dml2_uclk_pstate_support_method_fw_subvp_phantom_drr:
+ case dml2_pstate_method_fw_svp:
+ case dml2_pstate_method_fw_svp_drr:
pipe_ctx->p_state_type = P_STATE_SUB_VP;
break;
- case dml2_uclk_pstate_support_method_fw_drr:
+ case dml2_pstate_method_fw_drr:
if (sub_vp_enabled)
pipe_ctx->p_state_type = P_STATE_DRR_SUB_VP;
else
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.h
index 476a7f6e4875..9880d3e0398e 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_translation_helper.h
@@ -11,20 +11,18 @@ struct dc_state;
struct dcn_watermarks;
union dcn_watermark_set;
struct pipe_ctx;
+struct dc_plane_state;
struct dml2_context;
struct dml2_configuration_options;
struct dml2_initialize_instance_in_out;
-void dml21_apply_soc_bb_overrides(struct dml2_initialize_instance_in_out *dml_init, const struct dml2_configuration_options *config, const struct dc *in_dc);
-void dml21_initialize_soc_bb_params(struct dml2_initialize_instance_in_out *dml_init, const struct dml2_configuration_options *config, const struct dc *in_dc);
-void dml21_initialize_ip_params(struct dml2_initialize_instance_in_out *dml_init, const struct dml2_configuration_options *config, const struct dc *in_dc);
+void dml21_populate_dml_init_params(struct dml2_initialize_instance_in_out *dml_init, const struct dml2_configuration_options *config, const struct dc *in_dc);
bool dml21_map_dc_state_into_dml_display_cfg(const struct dc *in_dc, struct dc_state *context, struct dml2_context *dml_ctx);
void dml21_copy_clocks_to_dc_state(struct dml2_context *in_ctx, struct dc_state *context);
-void dml21_populate_pipe_ctx_dlg_params(struct dml2_context *dml_ctx, struct dc_state *context, struct pipe_ctx *pipe_ctx, struct dml2_per_stream_programming *stream_programming);
-void dml21_extract_legacy_watermark_set(const struct dc *in_dc, struct dcn_watermarks *watermark, enum dml2_dchub_watermark_reg_set_index reg_set_idx, struct dml2_context *in_ctx);
void dml21_extract_watermark_sets(const struct dc *in_dc, union dcn_watermark_set *watermarks, struct dml2_context *in_ctx);
void dml21_map_hw_resources(struct dml2_context *dml_ctx);
void dml21_get_pipe_mcache_config(struct dc_state *context, struct pipe_ctx *pipe_ctx, struct dml2_per_plane_programming *pln_prog, struct dml2_pipe_configuration_descriptor *mcache_pipe_config);
void dml21_set_dc_p_state_type(struct pipe_ctx *pipe_ctx, struct dml2_per_stream_programming *stream_programming, bool sub_vp_enabled);
+unsigned int map_plane_to_dml21_display_cfg(const struct dml2_context *dml_ctx, unsigned int stream_id, const struct dc_plane_state *plane, const struct dc_state *context);
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.c
index 51d491bffa32..ee721606b883 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.c
@@ -142,108 +142,21 @@ int dml21_find_dc_pipes_for_plane(const struct dc *in_dc,
return num_pipes;
}
-
-void dml21_update_pipe_ctx_dchub_regs(struct dml2_display_rq_regs *rq_regs,
- struct dml2_display_dlg_regs *disp_dlg_regs,
- struct dml2_display_ttu_regs *disp_ttu_regs,
- struct pipe_ctx *out)
+void dml21_pipe_populate_global_sync(struct dml2_context *dml_ctx,
+ struct dc_state *context,
+ struct pipe_ctx *pipe_ctx,
+ struct dml2_per_stream_programming *stream_programming)
{
- memset(&out->rq_regs, 0, sizeof(out->rq_regs));
- out->rq_regs.rq_regs_l.chunk_size = rq_regs->rq_regs_l.chunk_size;
- out->rq_regs.rq_regs_l.min_chunk_size = rq_regs->rq_regs_l.min_chunk_size;
- //out->rq_regs.rq_regs_l.meta_chunk_size = rq_regs->rq_regs_l.meta_chunk_size;
- //out->rq_regs.rq_regs_l.min_meta_chunk_size = rq_regs->rq_regs_l.min_meta_chunk_size;
- out->rq_regs.rq_regs_l.dpte_group_size = rq_regs->rq_regs_l.dpte_group_size;
- out->rq_regs.rq_regs_l.mpte_group_size = rq_regs->rq_regs_l.mpte_group_size;
- out->rq_regs.rq_regs_l.swath_height = rq_regs->rq_regs_l.swath_height;
- out->rq_regs.rq_regs_l.pte_row_height_linear = rq_regs->rq_regs_l.pte_row_height_linear;
-
- out->rq_regs.rq_regs_c.chunk_size = rq_regs->rq_regs_c.chunk_size;
- out->rq_regs.rq_regs_c.min_chunk_size = rq_regs->rq_regs_c.min_chunk_size;
- //out->rq_regs.rq_regs_c.meta_chunk_size = rq_regs->rq_regs_c.meta_chunk_size;
- //out->rq_regs.rq_regs_c.min_meta_chunk_size = rq_regs->rq_regs_c.min_meta_chunk_size;
- out->rq_regs.rq_regs_c.dpte_group_size = rq_regs->rq_regs_c.dpte_group_size;
- out->rq_regs.rq_regs_c.mpte_group_size = rq_regs->rq_regs_c.mpte_group_size;
- out->rq_regs.rq_regs_c.swath_height = rq_regs->rq_regs_c.swath_height;
- out->rq_regs.rq_regs_c.pte_row_height_linear = rq_regs->rq_regs_c.pte_row_height_linear;
-
- out->rq_regs.drq_expansion_mode = rq_regs->drq_expansion_mode;
- out->rq_regs.prq_expansion_mode = rq_regs->prq_expansion_mode;
- //out->rq_regs.mrq_expansion_mode = rq_regs->mrq_expansion_mode;
- out->rq_regs.crq_expansion_mode = rq_regs->crq_expansion_mode;
- out->rq_regs.plane1_base_address = rq_regs->plane1_base_address;
- out->unbounded_req = rq_regs->unbounded_request_enabled;
-
- memset(&out->dlg_regs, 0, sizeof(out->dlg_regs));
- out->dlg_regs.refcyc_h_blank_end = disp_dlg_regs->refcyc_h_blank_end;
- out->dlg_regs.dlg_vblank_end = disp_dlg_regs->dlg_vblank_end;
- out->dlg_regs.min_dst_y_next_start = disp_dlg_regs->min_dst_y_next_start;
- out->dlg_regs.refcyc_per_htotal = disp_dlg_regs->refcyc_per_htotal;
- out->dlg_regs.refcyc_x_after_scaler = disp_dlg_regs->refcyc_x_after_scaler;
- out->dlg_regs.dst_y_after_scaler = disp_dlg_regs->dst_y_after_scaler;
- out->dlg_regs.dst_y_prefetch = disp_dlg_regs->dst_y_prefetch;
- out->dlg_regs.dst_y_per_vm_vblank = disp_dlg_regs->dst_y_per_vm_vblank;
- out->dlg_regs.dst_y_per_row_vblank = disp_dlg_regs->dst_y_per_row_vblank;
- out->dlg_regs.dst_y_per_vm_flip = disp_dlg_regs->dst_y_per_vm_flip;
- out->dlg_regs.dst_y_per_row_flip = disp_dlg_regs->dst_y_per_row_flip;
- out->dlg_regs.ref_freq_to_pix_freq = disp_dlg_regs->ref_freq_to_pix_freq;
- out->dlg_regs.vratio_prefetch = disp_dlg_regs->vratio_prefetch;
- out->dlg_regs.vratio_prefetch_c = disp_dlg_regs->vratio_prefetch_c;
- out->dlg_regs.refcyc_per_tdlut_group = disp_dlg_regs->refcyc_per_tdlut_group;
- out->dlg_regs.refcyc_per_pte_group_vblank_l = disp_dlg_regs->refcyc_per_pte_group_vblank_l;
- out->dlg_regs.refcyc_per_pte_group_vblank_c = disp_dlg_regs->refcyc_per_pte_group_vblank_c;
- //out->dlg_regs.refcyc_per_meta_chunk_vblank_l = disp_dlg_regs->refcyc_per_meta_chunk_vblank_l;
- //out->dlg_regs.refcyc_per_meta_chunk_vblank_c = disp_dlg_regs->refcyc_per_meta_chunk_vblank_c;
- out->dlg_regs.refcyc_per_pte_group_flip_l = disp_dlg_regs->refcyc_per_pte_group_flip_l;
- out->dlg_regs.refcyc_per_pte_group_flip_c = disp_dlg_regs->refcyc_per_pte_group_flip_c;
- //out->dlg_regs.refcyc_per_meta_chunk_flip_l = disp_dlg_regs->refcyc_per_meta_chunk_flip_l;
- //out->dlg_regs.refcyc_per_meta_chunk_flip_c = disp_dlg_regs->refcyc_per_meta_chunk_flip_c;
- out->dlg_regs.dst_y_per_pte_row_nom_l = disp_dlg_regs->dst_y_per_pte_row_nom_l;
- out->dlg_regs.dst_y_per_pte_row_nom_c = disp_dlg_regs->dst_y_per_pte_row_nom_c;
- out->dlg_regs.refcyc_per_pte_group_nom_l = disp_dlg_regs->refcyc_per_pte_group_nom_l;
- out->dlg_regs.refcyc_per_pte_group_nom_c = disp_dlg_regs->refcyc_per_pte_group_nom_c;
- //out->dlg_regs.dst_y_per_meta_row_nom_l = disp_dlg_regs->dst_y_per_meta_row_nom_l;
- //out->dlg_regs.dst_y_per_meta_row_nom_c = disp_dlg_regs->dst_y_per_meta_row_nom_c;
- //out->dlg_regs.refcyc_per_meta_chunk_nom_l = disp_dlg_regs->refcyc_per_meta_chunk_nom_l;
- //out->dlg_regs.refcyc_per_meta_chunk_nom_c = disp_dlg_regs->refcyc_per_meta_chunk_nom_c;
- out->dlg_regs.refcyc_per_line_delivery_pre_l = disp_dlg_regs->refcyc_per_line_delivery_pre_l;
- out->dlg_regs.refcyc_per_line_delivery_pre_c = disp_dlg_regs->refcyc_per_line_delivery_pre_c;
- out->dlg_regs.refcyc_per_line_delivery_l = disp_dlg_regs->refcyc_per_line_delivery_l;
- out->dlg_regs.refcyc_per_line_delivery_c = disp_dlg_regs->refcyc_per_line_delivery_c;
- out->dlg_regs.refcyc_per_vm_group_vblank = disp_dlg_regs->refcyc_per_vm_group_vblank;
- out->dlg_regs.refcyc_per_vm_group_flip = disp_dlg_regs->refcyc_per_vm_group_flip;
- out->dlg_regs.refcyc_per_vm_req_vblank = disp_dlg_regs->refcyc_per_vm_req_vblank;
- out->dlg_regs.refcyc_per_vm_req_flip = disp_dlg_regs->refcyc_per_vm_req_flip;
- out->dlg_regs.dst_y_offset_cur0 = disp_dlg_regs->dst_y_offset_cur0;
- out->dlg_regs.chunk_hdl_adjust_cur0 = disp_dlg_regs->chunk_hdl_adjust_cur0;
- //out->dlg_regs.dst_y_offset_cur1 = disp_dlg_regs->dst_y_offset_cur1;
- //out->dlg_regs.chunk_hdl_adjust_cur1 = disp_dlg_regs->chunk_hdl_adjust_cur1;
- out->dlg_regs.vready_after_vcount0 = disp_dlg_regs->vready_after_vcount0;
- out->dlg_regs.dst_y_delta_drq_limit = disp_dlg_regs->dst_y_delta_drq_limit;
- out->dlg_regs.refcyc_per_vm_dmdata = disp_dlg_regs->refcyc_per_vm_dmdata;
- out->dlg_regs.dmdata_dl_delta = disp_dlg_regs->dmdata_dl_delta;
-
- memset(&out->ttu_regs, 0, sizeof(out->ttu_regs));
- out->ttu_regs.qos_level_low_wm = disp_ttu_regs->qos_level_low_wm;
- out->ttu_regs.qos_level_high_wm = disp_ttu_regs->qos_level_high_wm;
- out->ttu_regs.min_ttu_vblank = disp_ttu_regs->min_ttu_vblank;
- out->ttu_regs.qos_level_flip = disp_ttu_regs->qos_level_flip;
- out->ttu_regs.refcyc_per_req_delivery_l = disp_ttu_regs->refcyc_per_req_delivery_l;
- out->ttu_regs.refcyc_per_req_delivery_c = disp_ttu_regs->refcyc_per_req_delivery_c;
- out->ttu_regs.refcyc_per_req_delivery_cur0 = disp_ttu_regs->refcyc_per_req_delivery_cur0;
- //out->ttu_regs.refcyc_per_req_delivery_cur1 = disp_ttu_regs->refcyc_per_req_delivery_cur1;
- out->ttu_regs.refcyc_per_req_delivery_pre_l = disp_ttu_regs->refcyc_per_req_delivery_pre_l;
- out->ttu_regs.refcyc_per_req_delivery_pre_c = disp_ttu_regs->refcyc_per_req_delivery_pre_c;
- out->ttu_regs.refcyc_per_req_delivery_pre_cur0 = disp_ttu_regs->refcyc_per_req_delivery_pre_cur0;
- //out->ttu_regs.refcyc_per_req_delivery_pre_cur1 = disp_ttu_regs->refcyc_per_req_delivery_pre_cur1;
- out->ttu_regs.qos_level_fixed_l = disp_ttu_regs->qos_level_fixed_l;
- out->ttu_regs.qos_level_fixed_c = disp_ttu_regs->qos_level_fixed_c;
- out->ttu_regs.qos_level_fixed_cur0 = disp_ttu_regs->qos_level_fixed_cur0;
- //out->ttu_regs.qos_level_fixed_cur1 = disp_ttu_regs->qos_level_fixed_cur1;
- out->ttu_regs.qos_ramp_disable_l = disp_ttu_regs->qos_ramp_disable_l;
- out->ttu_regs.qos_ramp_disable_c = disp_ttu_regs->qos_ramp_disable_c;
- out->ttu_regs.qos_ramp_disable_cur0 = disp_ttu_regs->qos_ramp_disable_cur0;
- //out->ttu_regs.qos_ramp_disable_cur1 = disp_ttu_regs->qos_ramp_disable_cur1;
+ union dml2_global_sync_programming *global_sync = &stream_programming->global_sync;
+
+ if (dml_ctx->config.svp_pstate.callbacks.get_pipe_subvp_type(context, pipe_ctx) == SUBVP_PHANTOM) {
+ /* phantom has its own global sync */
+ global_sync = &stream_programming->phantom_stream.global_sync;
+ }
+
+ memcpy(&pipe_ctx->global_sync,
+ global_sync,
+ sizeof(union dml2_global_sync_programming));
}
void dml21_populate_mall_allocation_size(struct dc_state *context,
@@ -301,28 +214,16 @@ void dml21_program_dc_pipe(struct dml2_context *dml_ctx, struct dc_state *contex
{
unsigned int pipe_reg_index = 0;
- dml21_populate_pipe_ctx_dlg_params(dml_ctx, context, pipe_ctx, stream_prog);
+ dml21_pipe_populate_global_sync(dml_ctx, context, pipe_ctx, stream_prog);
find_pipe_regs_idx(dml_ctx, pipe_ctx, &pipe_reg_index);
if (dml_ctx->config.svp_pstate.callbacks.get_pipe_subvp_type(context, pipe_ctx) == SUBVP_PHANTOM) {
memcpy(&pipe_ctx->hubp_regs, pln_prog->phantom_plane.pipe_regs[pipe_reg_index], sizeof(struct dml2_dchub_per_pipe_register_set));
pipe_ctx->unbounded_req = false;
-
- /* legacy only, should be removed later */
- dml21_update_pipe_ctx_dchub_regs(&pln_prog->phantom_plane.pipe_regs[pipe_reg_index]->rq_regs,
- &pln_prog->phantom_plane.pipe_regs[pipe_reg_index]->dlg_regs,
- &pln_prog->phantom_plane.pipe_regs[pipe_reg_index]->ttu_regs, pipe_ctx);
-
pipe_ctx->det_buffer_size_kb = 0;
} else {
memcpy(&pipe_ctx->hubp_regs, pln_prog->pipe_regs[pipe_reg_index], sizeof(struct dml2_dchub_per_pipe_register_set));
pipe_ctx->unbounded_req = pln_prog->pipe_regs[pipe_reg_index]->rq_regs.unbounded_request_enabled;
-
- /* legacy only, should be removed later */
- dml21_update_pipe_ctx_dchub_regs(&pln_prog->pipe_regs[pipe_reg_index]->rq_regs,
- &pln_prog->pipe_regs[pipe_reg_index]->dlg_regs,
- &pln_prog->pipe_regs[pipe_reg_index]->ttu_regs, pipe_ctx);
-
pipe_ctx->det_buffer_size_kb = pln_prog->pipe_regs[pipe_reg_index]->det_size * 64;
}
@@ -331,7 +232,6 @@ void dml21_program_dc_pipe(struct dml2_context *dml_ctx, struct dc_state *contex
context->bw_ctx.bw.dcn.clk.dppclk_khz = pipe_ctx->plane_res.bw.dppclk_khz;
dml21_populate_mall_allocation_size(context, dml_ctx, pln_prog, pipe_ctx);
- memcpy(&context->bw_ctx.bw.dcn.mcache_allocations[pipe_ctx->pipe_idx], &pln_prog->mcache_allocation, sizeof(struct dml2_mcache_surface_allocation));
bool sub_vp_enabled = is_sub_vp_enabled(pipe_ctx->stream->ctx->dc, context);
@@ -482,7 +382,9 @@ void dml21_build_fams2_programming(const struct dc *dc,
unsigned int num_fams2_streams = 0;
/* reset fams2 data */
- memset(&context->bw_ctx.bw.dcn.fams2_stream_params, 0, sizeof(struct dmub_fams2_stream_static_state) * DML2_MAX_PLANES);
+ memset(&context->bw_ctx.bw.dcn.fams2_stream_base_params, 0, sizeof(union dmub_cmd_fams2_config) * DML2_MAX_PLANES);
+ memset(&context->bw_ctx.bw.dcn.fams2_stream_sub_params, 0, sizeof(union dmub_cmd_fams2_config) * DML2_MAX_PLANES);
+ memset(&context->bw_ctx.bw.dcn.fams2_stream_sub_params_v2, 0, sizeof(union dmub_fams2_stream_static_sub_state_v2) * DML2_MAX_PLANES);
memset(&context->bw_ctx.bw.dcn.fams2_global_config, 0, sizeof(struct dmub_cmd_fams2_global_config));
if (dml_ctx->v21.mode_programming.programming->fams2_required) {
@@ -490,8 +392,10 @@ void dml21_build_fams2_programming(const struct dc *dc,
int dml_stream_idx;
struct dc_stream_state *phantom_stream;
struct dc_stream_status *phantom_status;
+ enum fams2_stream_type type = 0;
- struct dmub_fams2_stream_static_state *static_state = &context->bw_ctx.bw.dcn.fams2_stream_params[num_fams2_streams];
+ union dmub_cmd_fams2_config *static_base_state = &context->bw_ctx.bw.dcn.fams2_stream_base_params[num_fams2_streams];
+ union dmub_cmd_fams2_config *static_sub_state = &context->bw_ctx.bw.dcn.fams2_stream_sub_params[num_fams2_streams];
struct dc_stream_state *stream = context->streams[i];
@@ -508,28 +412,45 @@ void dml21_build_fams2_programming(const struct dc *dc,
}
/* copy static state from PMO */
- memcpy(static_state,
- &dml_ctx->v21.mode_programming.programming->stream_programming[dml_stream_idx].fams2_params,
- sizeof(struct dmub_fams2_stream_static_state));
-
- /* get information from context */
- static_state->num_planes = context->stream_status[i].plane_count;
- static_state->otg_inst = context->stream_status[i].primary_otg_inst;
-
- /* populate pipe masks for planes */
- for (j = 0; j < context->stream_status[i].plane_count; j++) {
- for (k = 0; k < dc->res_pool->pipe_count; k++) {
- if (context->res_ctx.pipe_ctx[k].stream &&
- context->res_ctx.pipe_ctx[k].stream->stream_id == stream->stream_id &&
- context->res_ctx.pipe_ctx[k].plane_state == context->stream_status[i].plane_states[j]) {
- static_state->pipe_mask |= (1 << k);
- static_state->plane_pipe_masks[j] |= (1 << k);
+ memcpy(static_base_state,
+ &dml_ctx->v21.mode_programming.programming->stream_programming[dml_stream_idx].fams2_base_params,
+ sizeof(union dmub_cmd_fams2_config));
+
+ if (dc->debug.fams_version.major == 3) {
+ memcpy(&context->bw_ctx.bw.dcn.fams2_stream_sub_params_v2[num_fams2_streams],
+ &dml_ctx->v21.mode_programming.programming->stream_programming[dml_stream_idx].fams2_sub_params_v2,
+ sizeof(union dmub_fams2_stream_static_sub_state_v2));
+ } else {
+ memcpy(static_sub_state,
+ &dml_ctx->v21.mode_programming.programming->stream_programming[dml_stream_idx].fams2_sub_params,
+ sizeof(union dmub_cmd_fams2_config));
+ }
+
+ switch (dc->debug.fams_version.minor) {
+ case 1:
+ default:
+ type = static_base_state->stream_v1.base.type;
+
+ /* get information from context */
+ static_base_state->stream_v1.base.num_planes = context->stream_status[i].plane_count;
+ static_base_state->stream_v1.base.otg_inst = context->stream_status[i].primary_otg_inst;
+
+ /* populate pipe masks for planes */
+ for (j = 0; j < context->stream_status[i].plane_count; j++) {
+ for (k = 0; k < dc->res_pool->pipe_count; k++) {
+ if (context->res_ctx.pipe_ctx[k].stream &&
+ context->res_ctx.pipe_ctx[k].stream->stream_id == stream->stream_id &&
+ context->res_ctx.pipe_ctx[k].plane_state == context->stream_status[i].plane_states[j]) {
+ static_base_state->stream_v1.base.pipe_mask |= (1 << k);
+ static_base_state->stream_v1.base.plane_pipe_masks[j] |= (1 << k);
+ }
}
}
}
+
/* get per method programming */
- switch (static_state->type) {
+ switch (type) {
case FAMS2_STREAM_TYPE_VBLANK:
case FAMS2_STREAM_TYPE_VACTIVE:
case FAMS2_STREAM_TYPE_DRR:
@@ -543,16 +464,27 @@ void dml21_build_fams2_programming(const struct dc *dc,
/* phantom status should always be present */
ASSERT(phantom_status);
- static_state->sub_state.subvp.phantom_otg_inst = phantom_status->primary_otg_inst;
+ if (!phantom_status)
+ break;
- /* populate pipe masks for phantom planes */
- for (j = 0; j < phantom_status->plane_count; j++) {
- for (k = 0; k < dc->res_pool->pipe_count; k++) {
- if (context->res_ctx.pipe_ctx[k].stream &&
- context->res_ctx.pipe_ctx[k].stream->stream_id == phantom_stream->stream_id &&
- context->res_ctx.pipe_ctx[k].plane_state == phantom_status->plane_states[j]) {
- static_state->sub_state.subvp.phantom_pipe_mask |= (1 << k);
- static_state->sub_state.subvp.phantom_plane_pipe_masks[j] |= (1 << k);
+ switch (dc->debug.fams_version.minor) {
+ case 1:
+ default:
+ static_sub_state->stream_v1.sub_state.subvp.phantom_otg_inst = phantom_status->primary_otg_inst;
+
+ /* populate pipe masks for phantom planes */
+ for (j = 0; j < phantom_status->plane_count; j++) {
+ for (k = 0; k < dc->res_pool->pipe_count; k++) {
+ if (context->res_ctx.pipe_ctx[k].stream &&
+ context->res_ctx.pipe_ctx[k].stream->stream_id == phantom_stream->stream_id &&
+ context->res_ctx.pipe_ctx[k].plane_state == phantom_status->plane_states[j]) {
+ switch (dc->debug.fams_version.minor) {
+ case 1:
+ default:
+ static_sub_state->stream_v1.sub_state.subvp.phantom_pipe_mask |= (1 << k);
+ static_sub_state->stream_v1.sub_state.subvp.phantom_plane_pipe_masks[j] |= (1 << k);
+ }
+ }
}
}
}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.h
index d5153fbac921..4bff52eaaef8 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_utils.h
@@ -18,10 +18,10 @@ struct dml2_display_ttu_regs;
int dml21_helper_find_dml_pipe_idx_by_stream_id(struct dml2_context *ctx, unsigned int stream_id);
int dml21_find_dml_pipe_idx_by_plane_id(struct dml2_context *ctx, unsigned int plane_id);
bool dml21_get_plane_id(const struct dc_state *state, const struct dc_plane_state *plane, unsigned int *plane_id);
-void dml21_update_pipe_ctx_dchub_regs(struct dml2_display_rq_regs *rq_regs,
- struct dml2_display_dlg_regs *disp_dlg_regs,
- struct dml2_display_ttu_regs *disp_ttu_regs,
- struct pipe_ctx *out);
+void dml21_pipe_populate_global_sync(struct dml2_context *dml_ctx,
+ struct dc_state *context,
+ struct pipe_ctx *pipe_ctx,
+ struct dml2_per_stream_programming *stream_programming);
void dml21_populate_mall_allocation_size(struct dc_state *context,
struct dml2_context *in_ctx,
struct dml2_per_plane_programming *pln_prog,
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.c
index bbc28b9a15a3..03de3cf06ae5 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.c
@@ -2,7 +2,6 @@
//
// Copyright 2024 Advanced Micro Devices, Inc.
-
#include "dml2_internal_types.h"
#include "dml_top.h"
#include "dml2_core_dcn4_calcs.h"
@@ -11,13 +10,15 @@
#include "dml21_translation_helper.h"
#include "dml2_dc_resource_mgmt.h"
+#define INVALID -1
+
static bool dml21_allocate_memory(struct dml2_context **dml_ctx)
{
- *dml_ctx = kzalloc(sizeof(struct dml2_context), GFP_KERNEL);
+ *dml_ctx = vzalloc(sizeof(struct dml2_context));
if (!(*dml_ctx))
return false;
- (*dml_ctx)->v21.dml_init.dml2_instance = kzalloc(sizeof(struct dml2_instance), GFP_KERNEL);
+ (*dml_ctx)->v21.dml_init.dml2_instance = vzalloc(sizeof(struct dml2_instance));
if (!((*dml_ctx)->v21.dml_init.dml2_instance))
return false;
@@ -27,22 +28,18 @@ static bool dml21_allocate_memory(struct dml2_context **dml_ctx)
(*dml_ctx)->v21.mode_support.display_config = &(*dml_ctx)->v21.display_config;
(*dml_ctx)->v21.mode_programming.display_config = (*dml_ctx)->v21.mode_support.display_config;
- (*dml_ctx)->v21.mode_programming.programming = kzalloc(sizeof(struct dml2_display_cfg_programming), GFP_KERNEL);
+ (*dml_ctx)->v21.mode_programming.programming = vzalloc(sizeof(struct dml2_display_cfg_programming));
if (!((*dml_ctx)->v21.mode_programming.programming))
return false;
return true;
}
-static void dml21_apply_debug_options(const struct dc *in_dc, struct dml2_context *dml_ctx, const struct dml2_configuration_options *config)
+static void dml21_populate_configuration_options(const struct dc *in_dc,
+ struct dml2_context *dml_ctx,
+ const struct dml2_configuration_options *config)
{
- bool disable_fams2;
- struct dml2_pmo_options *pmo_options = &dml_ctx->v21.dml_init.options.pmo_options;
-
- /* ODM options */
- pmo_options->disable_dyn_odm = !config->minimize_dispclk_using_odm;
- pmo_options->disable_dyn_odm_for_multi_stream = true;
- pmo_options->disable_dyn_odm_for_stream_with_svp = true;
+ dml_ctx->config = *config;
/* UCLK P-State options */
if (in_dc->debug.dml21_force_pstate_method) {
@@ -52,51 +49,22 @@ static void dml21_apply_debug_options(const struct dc *in_dc, struct dml2_contex
} else {
dml_ctx->config.pmo.force_pstate_method_enable = false;
}
-
- pmo_options->disable_vblank = ((in_dc->debug.dml21_disable_pstate_method_mask >> 1) & 1);
-
- /* NOTE: DRR and SubVP Require FAMS2 */
- disable_fams2 = !in_dc->debug.fams2_config.bits.enable;
- pmo_options->disable_svp = ((in_dc->debug.dml21_disable_pstate_method_mask >> 2) & 1) ||
- in_dc->debug.force_disable_subvp ||
- disable_fams2;
- pmo_options->disable_drr_clamped = ((in_dc->debug.dml21_disable_pstate_method_mask >> 3) & 1) ||
- disable_fams2;
- pmo_options->disable_drr_var = ((in_dc->debug.dml21_disable_pstate_method_mask >> 4) & 1) ||
- disable_fams2;
- pmo_options->disable_fams2 = disable_fams2;
-
- pmo_options->disable_drr_var_when_var_active = in_dc->debug.disable_fams_gaming == INGAME_FAMS_DISABLE ||
- in_dc->debug.disable_fams_gaming == INGAME_FAMS_MULTI_DISP_CLAMPED_ONLY;
- pmo_options->disable_drr_clamped_when_var_active = in_dc->debug.disable_fams_gaming == INGAME_FAMS_DISABLE;
}
-static void dml21_init(const struct dc *in_dc, struct dml2_context **dml_ctx, const struct dml2_configuration_options *config)
+static void dml21_init(const struct dc *in_dc, struct dml2_context *dml_ctx, const struct dml2_configuration_options *config)
{
- switch (in_dc->ctx->dce_version) {
- case DCN_VERSION_4_01:
- case DCN_VERSION_3_2: // TODO : Temporary for N-1 validation. Remove this after N-1 validation phase is complete.
- (*dml_ctx)->v21.dml_init.options.project_id = dml2_project_dcn4x_stage2_auto_drr_svp;
- break;
- default:
- (*dml_ctx)->v21.dml_init.options.project_id = dml2_project_invalid;
- }
- (*dml_ctx)->architecture = dml2_architecture_21;
+ dml_ctx->architecture = dml2_architecture_21;
+
+ dml21_populate_configuration_options(in_dc, dml_ctx, config);
- /* Store configuration options */
- (*dml_ctx)->config = *config;
+ DC_FP_START();
- /*Initialize SOCBB and DCNIP params */
- dml21_initialize_soc_bb_params(&(*dml_ctx)->v21.dml_init, config, in_dc);
- dml21_initialize_ip_params(&(*dml_ctx)->v21.dml_init, config, in_dc);
- dml21_apply_soc_bb_overrides(&(*dml_ctx)->v21.dml_init, config, in_dc);
+ dml21_populate_dml_init_params(&dml_ctx->v21.dml_init, config, in_dc);
- /* apply debug overrides */
- dml21_apply_debug_options(in_dc, *dml_ctx, config);
+ dml2_initialize_instance(&dml_ctx->v21.dml_init);
- /*Initialize DML21 instance */
- dml2_initialize_instance(&(*dml_ctx)->v21.dml_init);
+ DC_FP_END();
}
bool dml21_create(const struct dc *in_dc, struct dml2_context **dml_ctx, const struct dml2_configuration_options *config)
@@ -105,15 +73,15 @@ bool dml21_create(const struct dc *in_dc, struct dml2_context **dml_ctx, const s
if (!dml21_allocate_memory(dml_ctx))
return false;
- dml21_init(in_dc, dml_ctx, config);
+ dml21_init(in_dc, *dml_ctx, config);
return true;
}
void dml21_destroy(struct dml2_context *dml2)
{
- kfree(dml2->v21.dml_init.dml2_instance);
- kfree(dml2->v21.mode_programming.programming);
+ vfree(dml2->v21.dml_init.dml2_instance);
+ vfree(dml2->v21.mode_programming.programming);
}
static void dml21_calculate_rq_and_dlg_params(const struct dc *dc, struct dc_state *context, struct resource_context *out_new_hw_state,
@@ -125,6 +93,7 @@ static void dml21_calculate_rq_and_dlg_params(const struct dc *dc, struct dc_sta
struct pipe_ctx *dc_main_pipes[__DML2_WRAPPER_MAX_STREAMS_PLANES__];
struct pipe_ctx *dc_phantom_pipes[__DML2_WRAPPER_MAX_STREAMS_PLANES__] = {0};
int num_pipes;
+ unsigned int dml_phantom_prog_idx;
context->bw_ctx.bw.dcn.clk.dppclk_khz = 0;
@@ -138,6 +107,9 @@ static void dml21_calculate_rq_and_dlg_params(const struct dc *dc, struct dc_sta
context->bw_ctx.bw.dcn.mall_ss_psr_active_size_bytes = 0;
context->bw_ctx.bw.dcn.mall_subvp_size_bytes = 0;
+ /* phantom's start after main planes */
+ dml_phantom_prog_idx = in_ctx->v21.mode_programming.programming->display_config.num_planes;
+
for (dml_prog_idx = 0; dml_prog_idx < DML2_MAX_PLANES; dml_prog_idx++) {
pln_prog = &in_ctx->v21.mode_programming.programming->plane_programming[dml_prog_idx];
@@ -163,6 +135,16 @@ static void dml21_calculate_rq_and_dlg_params(const struct dc *dc, struct dc_sta
dml21_program_dc_pipe(in_ctx, context, dc_phantom_pipes[dc_pipe_index], pln_prog, stream_prog);
}
}
+
+ /* copy per plane mcache allocation */
+ memcpy(&context->bw_ctx.bw.dcn.mcache_allocations[dml_prog_idx], &pln_prog->mcache_allocation, sizeof(struct dml2_mcache_surface_allocation));
+ if (pln_prog->phantom_plane.valid) {
+ memcpy(&context->bw_ctx.bw.dcn.mcache_allocations[dml_phantom_prog_idx],
+ &pln_prog->phantom_plane.mcache_allocation,
+ sizeof(struct dml2_mcache_surface_allocation));
+
+ dml_phantom_prog_idx++;
+ }
}
/* assign global clocks */
@@ -190,10 +172,40 @@ static void dml21_calculate_rq_and_dlg_params(const struct dc *dc, struct dc_sta
}
}
+static void dml21_prepare_mcache_params(struct dml2_context *dml_ctx, struct dc_state *context, struct dc_mcache_params *mcache_params)
+{
+ int dc_plane_idx = 0;
+ int dml_prog_idx, stream_idx, plane_idx;
+ struct dml2_per_plane_programming *pln_prog = NULL;
+
+ for (stream_idx = 0; stream_idx < context->stream_count; stream_idx++) {
+ for (plane_idx = 0; plane_idx < context->stream_status[stream_idx].plane_count; plane_idx++) {
+ dml_prog_idx = map_plane_to_dml21_display_cfg(dml_ctx, context->streams[stream_idx]->stream_id, context->stream_status[stream_idx].plane_states[plane_idx], context);
+ if (dml_prog_idx == INVALID) {
+ continue;
+ }
+ pln_prog = &dml_ctx->v21.mode_programming.programming->plane_programming[dml_prog_idx];
+ mcache_params[dc_plane_idx].valid = pln_prog->mcache_allocation.valid;
+ mcache_params[dc_plane_idx].num_mcaches_plane0 = pln_prog->mcache_allocation.num_mcaches_plane0;
+ mcache_params[dc_plane_idx].num_mcaches_plane1 = pln_prog->mcache_allocation.num_mcaches_plane1;
+ mcache_params[dc_plane_idx].requires_dedicated_mall_mcache = pln_prog->mcache_allocation.requires_dedicated_mall_mcache;
+ mcache_params[dc_plane_idx].last_slice_sharing.plane0_plane1 = pln_prog->mcache_allocation.last_slice_sharing.plane0_plane1;
+ memcpy(mcache_params[dc_plane_idx].mcache_x_offsets_plane0,
+ pln_prog->mcache_allocation.mcache_x_offsets_plane0,
+ sizeof(int) * (DML2_MAX_MCACHES + 1));
+ memcpy(mcache_params[dc_plane_idx].mcache_x_offsets_plane1,
+ pln_prog->mcache_allocation.mcache_x_offsets_plane1,
+ sizeof(int) * (DML2_MAX_MCACHES + 1));
+ dc_plane_idx++;
+ }
+ }
+}
+
static bool dml21_mode_check_and_programming(const struct dc *in_dc, struct dc_state *context, struct dml2_context *dml_ctx)
{
bool result = false;
struct dml2_build_mode_programming_in_out *mode_programming = &dml_ctx->v21.mode_programming;
+ struct dc_mcache_params mcache_params[MAX_PLANES] = {0};
memset(&dml_ctx->v21.display_config, 0, sizeof(struct dml2_display_cfg));
memset(&dml_ctx->v21.dml_to_dc_pipe_mapping, 0, sizeof(struct dml2_dml_to_dc_pipe_mapping));
@@ -216,7 +228,9 @@ static bool dml21_mode_check_and_programming(const struct dc *in_dc, struct dc_s
if (!result)
return false;
+ DC_FP_START();
result = dml2_build_mode_programming(mode_programming);
+ DC_FP_END();
if (!result)
return false;
@@ -226,6 +240,14 @@ static bool dml21_mode_check_and_programming(const struct dc *in_dc, struct dc_s
dml2_map_dc_pipes(dml_ctx, context, NULL, &dml_ctx->v21.dml_to_dc_pipe_mapping, in_dc->current_state);
/* if subvp phantoms are present, expand them into dc context */
dml21_handle_phantom_streams_planes(in_dc, context, dml_ctx);
+
+ if (in_dc->res_pool->funcs->program_mcache_pipe_config) {
+ //Prepare mcache params for each plane based on mcache output from DML
+ dml21_prepare_mcache_params(dml_ctx, context, mcache_params);
+
+ //populate mcache regs to each pipe
+ dml_ctx->config.callbacks.allocate_mcache(context, mcache_params);
+ }
}
/* Copy DML CLK, WM and REG outputs to bandwidth context */
@@ -233,13 +255,6 @@ static bool dml21_mode_check_and_programming(const struct dc *in_dc, struct dc_s
dml21_calculate_rq_and_dlg_params(in_dc, context, &context->res_ctx, dml_ctx, in_dc->res_pool->pipe_count);
dml21_copy_clocks_to_dc_state(dml_ctx, context);
dml21_extract_watermark_sets(in_dc, &context->bw_ctx.bw.dcn.watermarks, dml_ctx);
- if (in_dc->ctx->dce_version == DCN_VERSION_3_2) {
- dml21_extract_legacy_watermark_set(in_dc, &context->bw_ctx.bw.dcn.watermarks.a, DML2_DCHUB_WATERMARK_SET_A, dml_ctx);
- dml21_extract_legacy_watermark_set(in_dc, &context->bw_ctx.bw.dcn.watermarks.b, DML2_DCHUB_WATERMARK_SET_A, dml_ctx);
- dml21_extract_legacy_watermark_set(in_dc, &context->bw_ctx.bw.dcn.watermarks.c, DML2_DCHUB_WATERMARK_SET_A, dml_ctx);
- dml21_extract_legacy_watermark_set(in_dc, &context->bw_ctx.bw.dcn.watermarks.d, DML2_DCHUB_WATERMARK_SET_A, dml_ctx);
- }
-
dml21_build_fams2_programming(in_dc, context, dml_ctx);
}
@@ -266,22 +281,26 @@ static bool dml21_check_mode_support(const struct dc *in_dc, struct dc_state *co
mode_support->dml2_instance = dml_init->dml2_instance;
dml21_map_dc_state_into_dml_display_cfg(in_dc, context, dml_ctx);
dml_ctx->v21.mode_programming.dml2_instance->scratch.build_mode_programming_locals.mode_programming_params.programming = dml_ctx->v21.mode_programming.programming;
+ DC_FP_START();
is_supported = dml2_check_mode_supported(mode_support);
+ DC_FP_END();
if (!is_supported)
return false;
return true;
}
-bool dml21_validate(const struct dc *in_dc, struct dc_state *context, struct dml2_context *dml_ctx, bool fast_validate)
+bool dml21_validate(const struct dc *in_dc, struct dc_state *context, struct dml2_context *dml_ctx,
+ enum dc_validate_mode validate_mode)
{
bool out = false;
- /* Use dml_validate_only for fast_validate path */
- if (fast_validate) {
+ /* Use dml21_check_mode_support for DC_VALIDATE_MODE_ONLY and DC_VALIDATE_MODE_AND_STATE_INDEX path */
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING)
out = dml21_check_mode_support(in_dc, context, dml_ctx);
- } else
+ else
out = dml21_mode_check_and_programming(in_dc, context, dml_ctx);
+
return out;
}
@@ -420,8 +439,12 @@ void dml21_copy(struct dml2_context *dst_dml_ctx,
dst_dml_ctx->v21.mode_programming.programming = dst_dml2_programming;
+ DC_FP_START();
+
/* need to initialize copied instance for internal references to be correct */
dml2_initialize_instance(&dst_dml_ctx->v21.dml_init);
+
+ DC_FP_END();
}
bool dml21_create_copy(struct dml2_context **dst_dml_ctx,
@@ -436,7 +459,7 @@ bool dml21_create_copy(struct dml2_context **dst_dml_ctx,
return true;
}
-void dml21_reinit(const struct dc *in_dc, struct dml2_context **dml_ctx, const struct dml2_configuration_options *config)
+void dml21_reinit(const struct dc *in_dc, struct dml2_context *dml_ctx, const struct dml2_configuration_options *config)
{
dml21_init(in_dc, dml_ctx, config);
}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.h
index b2075b8c363b..15f92029d2e5 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/dml21_wrapper.h
@@ -8,11 +8,13 @@
#include "os_types.h"
#include "dml_top_soc_parameter_types.h"
+#include "dml_top_display_cfg_types.h"
struct dc;
struct dc_state;
struct dml2_configuration_options;
struct dml2_context;
+enum dc_validate_mode;
/**
* dml2_create - Creates dml21_context.
@@ -32,22 +34,23 @@ void dml21_copy(struct dml2_context *dst_dml_ctx,
struct dml2_context *src_dml_ctx);
bool dml21_create_copy(struct dml2_context **dst_dml_ctx,
struct dml2_context *src_dml_ctx);
-void dml21_reinit(const struct dc *in_dc, struct dml2_context **dml_ctx, const struct dml2_configuration_options *config);
+void dml21_reinit(const struct dc *in_dc, struct dml2_context *dml_ctx, const struct dml2_configuration_options *config);
/**
* dml21_validate - Determines if a display configuration is supported or not.
* @in_dc: dc.
* @context: dc_state to be validated.
- * @fast_validate: Fast validate will not populate context.res_ctx.
+ * @validate_mode: DC_VALIDATE_MODE_ONLY and DC_VALIDATE_MODE_AND_STATE_INDEX
+ * will not populate context.res_ctx.
*
* Based on fast_validate option internally would call:
*
- * -dml21_mode_check_and_programming - for non fast_validate option
+ * -dml21_mode_check_and_programming - for DC_VALIDATE_MODE_AND_PROGRAMMING option
* Calculates if dc_state can be supported on the input display
* configuration. If supported, generates the necessary HW
* programming for the new dc_state.
*
- * -dml21_check_mode_support - for fast_validate option
+ * -dml21_check_mode_support - for DC_VALIDATE_MODE_ONLY and DC_VALIDATE_MODE_AND_STATE_INDEX option
* Calculates if dc_state can be supported for the input display
* config.
@@ -55,7 +58,8 @@ void dml21_reinit(const struct dc *in_dc, struct dml2_context **dml_ctx, const s
* separate dc_states for validation.
* Return: True if mode is supported, false otherwise.
*/
-bool dml21_validate(const struct dc *in_dc, struct dc_state *context, struct dml2_context *dml_ctx, bool fast_validate);
+bool dml21_validate(const struct dc *in_dc, struct dc_state *context, struct dml2_context *dml_ctx,
+ enum dc_validate_mode validate_mode);
/* Prepare hubp mcache_regs for hubp mcache ID and split coordinate programming */
void dml21_prepare_mcache_programming(struct dc *in_dc, struct dc_state *context, struct dml2_context *dml_ctx);
@@ -65,4 +69,67 @@ struct socbb_ip_params_external {
struct dml2_ip_capabilities ip_params;
struct dml2_soc_bb soc_bb;
};
+
+/*mcache parameters decided by dml*/
+struct dc_mcache_params {
+ bool valid;
+ /*
+ * For iMALL, dedicated mall mcaches are required (sharing of last
+ * slice possible), for legacy phantom or phantom without return
+ * the only mall mcaches need to be valid.
+ */
+ bool requires_dedicated_mall_mcache;
+ unsigned int num_mcaches_plane0;
+ unsigned int num_mcaches_plane1;
+ /*
+ * Generally, plane0/1 slices must use a disjoint set of caches
+ * but in some cases the final segement of the two planes can
+ * use the same cache. If plane0_plane1 is set, then this is
+ * allowed.
+ *
+ * Similarly, the caches allocated to MALL prefetcher are generally
+ * disjoint, but if mall_prefetch is set, then the final segment
+ * between the main and the mall pixel requestor can use the same
+ * cache.
+ *
+ * Note that both bits may be set at the same time.
+ */
+ struct {
+ bool mall_comb_mcache_p0;
+ bool mall_comb_mcache_p1;
+ bool plane0_plane1;
+ } last_slice_sharing;
+ /*
+ * A plane is divided into vertical slices of mcaches,
+ * which wrap on the surface width.
+ *
+ * For example, if the surface width is 7680, and split into
+ * three slices of equal width, the boundary array would contain
+ * [2560, 5120, 7680]
+ *
+ * The assignments are
+ * 0 = [0 .. 2559]
+ * 1 = [2560 .. 5119]
+ * 2 = [5120 .. 7679]
+ * 0 = [7680 .. INF]
+ * The final element implicitly is the same as the first, and
+ * at first seems invalid since it is never referenced (since)
+ * it is outside the surface. However, its useful when shifting
+ * (see below).
+ *
+ * For any given valid mcache assignment, a shifted version, wrapped
+ * on the surface width boundary is also assumed to be valid.
+ *
+ * For example, shifting [2560, 5120, 7680] by -50 results in
+ * [2510, 5170, 7630].
+ *
+ * The assignments are now:
+ * 0 = [0 .. 2509]
+ * 1 = [2510 .. 5169]
+ * 2 = [5170 .. 7629]
+ * 0 = [7630 .. INF]
+ */
+ int mcache_x_offsets_plane0[DML2_MAX_MCACHES + 1];
+ int mcache_x_offsets_plane1[DML2_MAX_MCACHES + 1];
+};
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/bounding_boxes/dcn3_soc_bb.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/bounding_boxes/dcn3_soc_bb.h
deleted file mode 100644
index d82c681a5402..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/bounding_boxes/dcn3_soc_bb.h
+++ /dev/null
@@ -1,401 +0,0 @@
-/*
- * Copyright 2022 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- * Authors: AMD
- *
- */
-
-#ifndef __DML_DML_DCN3_SOC_BB__
-#define __DML_DML_DCN3_SOC_BB__
-
-#include "dml_top_soc_parameter_types.h"
-
-static const struct dml2_soc_qos_parameters dml_dcn31_soc_qos_params = {
- .derate_table = {
- .system_active_urgent = {
- .dram_derate_percent_pixel = 22,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 76,
- .dcfclk_derate_percent = 100,
- },
- .system_active_average = {
- .dram_derate_percent_pixel = 17,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 57,
- .dcfclk_derate_percent = 75,
- },
- .dcn_mall_prefetch_urgent = {
- .dram_derate_percent_pixel = 22,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 76,
- .dcfclk_derate_percent = 100,
- },
- .dcn_mall_prefetch_average = {
- .dram_derate_percent_pixel = 17,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 57,
- .dcfclk_derate_percent = 75,
- },
- .system_idle_average = {
- .dram_derate_percent_pixel = 17,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 57,
- .dcfclk_derate_percent = 100,
- },
- },
- .writeback = {
- .base_latency_us = 12,
- .scaling_factor_us = 0,
- .scaling_factor_mhz = 0,
- },
- .qos_params = {
- .dcn4x = {
- .df_qos_response_time_fclk_cycles = 300,
- .max_round_trip_to_furthest_cs_fclk_cycles = 350,
- .mall_overhead_fclk_cycles = 50,
- .meta_trip_adder_fclk_cycles = 36,
- .average_transport_distance_fclk_cycles = 257,
- .umc_urgent_ramp_latency_margin = 50,
- .umc_max_latency_margin = 30,
- .umc_average_latency_margin = 20,
- .fabric_max_transport_latency_margin = 20,
- .fabric_average_transport_latency_margin = 10,
-
- .per_uclk_dpm_params = {
- {
- .minimum_uclk_khz = 97,
- .urgent_ramp_uclk_cycles = 472,
- .trip_to_memory_uclk_cycles = 827,
- .meta_trip_to_memory_uclk_cycles = 827,
- .maximum_latency_when_urgent_uclk_cycles = 72,
- .average_latency_when_urgent_uclk_cycles = 61,
- .maximum_latency_when_non_urgent_uclk_cycles = 827,
- .average_latency_when_non_urgent_uclk_cycles = 118,
- },
- {
- .minimum_uclk_khz = 435,
- .urgent_ramp_uclk_cycles = 546,
- .trip_to_memory_uclk_cycles = 848,
- .meta_trip_to_memory_uclk_cycles = 848,
- .maximum_latency_when_urgent_uclk_cycles = 146,
- .average_latency_when_urgent_uclk_cycles = 90,
- .maximum_latency_when_non_urgent_uclk_cycles = 848,
- .average_latency_when_non_urgent_uclk_cycles = 135,
- },
- {
- .minimum_uclk_khz = 731,
- .urgent_ramp_uclk_cycles = 632,
- .trip_to_memory_uclk_cycles = 874,
- .meta_trip_to_memory_uclk_cycles = 874,
- .maximum_latency_when_urgent_uclk_cycles = 232,
- .average_latency_when_urgent_uclk_cycles = 124,
- .maximum_latency_when_non_urgent_uclk_cycles = 874,
- .average_latency_when_non_urgent_uclk_cycles = 155,
- },
- {
- .minimum_uclk_khz = 1187,
- .urgent_ramp_uclk_cycles = 716,
- .trip_to_memory_uclk_cycles = 902,
- .meta_trip_to_memory_uclk_cycles = 902,
- .maximum_latency_when_urgent_uclk_cycles = 316,
- .average_latency_when_urgent_uclk_cycles = 160,
- .maximum_latency_when_non_urgent_uclk_cycles = 902,
- .average_latency_when_non_urgent_uclk_cycles = 177,
- },
- },
- },
- },
- .qos_type = dml2_qos_param_type_dcn4x,
-};
-
-static const struct dml2_soc_bb dml2_socbb_dcn31 = {
- .clk_table = {
- .uclk = {
- .clk_values_khz = {97000, 435000, 731000, 1187000},
- .num_clk_values = 4,
- },
- .fclk = {
- .clk_values_khz = {300000, 2500000},
- .num_clk_values = 2,
- },
- .dcfclk = {
- .clk_values_khz = {200000, 1800000},
- .num_clk_values = 2,
- },
- .dispclk = {
- .clk_values_khz = {100000, 2000000},
- .num_clk_values = 2,
- },
- .dppclk = {
- .clk_values_khz = {100000, 2000000},
- .num_clk_values = 2,
- },
- .dtbclk = {
- .clk_values_khz = {100000, 2000000},
- .num_clk_values = 2,
- },
- .phyclk = {
- .clk_values_khz = {810000, 810000},
- .num_clk_values = 2,
- },
- .socclk = {
- .clk_values_khz = {300000, 1600000},
- .num_clk_values = 2,
- },
- .dscclk = {
- .clk_values_khz = {666667, 666667},
- .num_clk_values = 2,
- },
- .phyclk_d18 = {
- .clk_values_khz = {625000, 625000},
- .num_clk_values = 2,
- },
- .phyclk_d32 = {
- .clk_values_khz = {2000000, 2000000},
- .num_clk_values = 2,
- },
- .dram_config = {
- .channel_width_bytes = 2,
- .channel_count = 16,
- .transactions_per_clock = 16,
- },
- },
-
- .qos_parameters = {
- .derate_table = {
- .system_active_urgent = {
- .dram_derate_percent_pixel = 22,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 76,
- .dcfclk_derate_percent = 100,
- },
- .system_active_average = {
- .dram_derate_percent_pixel = 17,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 57,
- .dcfclk_derate_percent = 75,
- },
- .dcn_mall_prefetch_urgent = {
- .dram_derate_percent_pixel = 22,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 76,
- .dcfclk_derate_percent = 100,
- },
- .dcn_mall_prefetch_average = {
- .dram_derate_percent_pixel = 17,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 57,
- .dcfclk_derate_percent = 75,
- },
- .system_idle_average = {
- .dram_derate_percent_pixel = 17,
- .dram_derate_percent_vm = 0,
- .dram_derate_percent_pixel_and_vm = 0,
- .fclk_derate_percent = 57,
- .dcfclk_derate_percent = 100,
- },
- },
- .writeback = {
- .base_latency_us = 0,
- .scaling_factor_us = 0,
- .scaling_factor_mhz = 0,
- },
- .qos_params = {
- .dcn4x = {
- .df_qos_response_time_fclk_cycles = 300,
- .max_round_trip_to_furthest_cs_fclk_cycles = 350,
- .mall_overhead_fclk_cycles = 50,
- .meta_trip_adder_fclk_cycles = 36,
- .average_transport_distance_fclk_cycles = 260,
- .umc_urgent_ramp_latency_margin = 50,
- .umc_max_latency_margin = 30,
- .umc_average_latency_margin = 20,
- .fabric_max_transport_latency_margin = 20,
- .fabric_average_transport_latency_margin = 10,
-
- .per_uclk_dpm_params = {
- {
- // State 1
- .minimum_uclk_khz = 0,
- .urgent_ramp_uclk_cycles = 472,
- .trip_to_memory_uclk_cycles = 827,
- .meta_trip_to_memory_uclk_cycles = 827,
- .maximum_latency_when_urgent_uclk_cycles = 72,
- .average_latency_when_urgent_uclk_cycles = 72,
- .maximum_latency_when_non_urgent_uclk_cycles = 827,
- .average_latency_when_non_urgent_uclk_cycles = 117,
- },
- {
- // State 2
- .minimum_uclk_khz = 0,
- .urgent_ramp_uclk_cycles = 546,
- .trip_to_memory_uclk_cycles = 848,
- .meta_trip_to_memory_uclk_cycles = 848,
- .maximum_latency_when_urgent_uclk_cycles = 146,
- .average_latency_when_urgent_uclk_cycles = 146,
- .maximum_latency_when_non_urgent_uclk_cycles = 848,
- .average_latency_when_non_urgent_uclk_cycles = 133,
- },
- {
- // State 3
- .minimum_uclk_khz = 0,
- .urgent_ramp_uclk_cycles = 564,
- .trip_to_memory_uclk_cycles = 853,
- .meta_trip_to_memory_uclk_cycles = 853,
- .maximum_latency_when_urgent_uclk_cycles = 164,
- .average_latency_when_urgent_uclk_cycles = 164,
- .maximum_latency_when_non_urgent_uclk_cycles = 853,
- .average_latency_when_non_urgent_uclk_cycles = 136,
- },
- {
- // State 4
- .minimum_uclk_khz = 0,
- .urgent_ramp_uclk_cycles = 613,
- .trip_to_memory_uclk_cycles = 869,
- .meta_trip_to_memory_uclk_cycles = 869,
- .maximum_latency_when_urgent_uclk_cycles = 213,
- .average_latency_when_urgent_uclk_cycles = 213,
- .maximum_latency_when_non_urgent_uclk_cycles = 869,
- .average_latency_when_non_urgent_uclk_cycles = 149,
- },
- {
- // State 5
- .minimum_uclk_khz = 0,
- .urgent_ramp_uclk_cycles = 632,
- .trip_to_memory_uclk_cycles = 874,
- .meta_trip_to_memory_uclk_cycles = 874,
- .maximum_latency_when_urgent_uclk_cycles = 232,
- .average_latency_when_urgent_uclk_cycles = 232,
- .maximum_latency_when_non_urgent_uclk_cycles = 874,
- .average_latency_when_non_urgent_uclk_cycles = 153,
- },
- {
- // State 6
- .minimum_uclk_khz = 0,
- .urgent_ramp_uclk_cycles = 665,
- .trip_to_memory_uclk_cycles = 885,
- .meta_trip_to_memory_uclk_cycles = 885,
- .maximum_latency_when_urgent_uclk_cycles = 265,
- .average_latency_when_urgent_uclk_cycles = 265,
- .maximum_latency_when_non_urgent_uclk_cycles = 885,
- .average_latency_when_non_urgent_uclk_cycles = 161,
- },
- {
- // State 7
- .minimum_uclk_khz = 0,
- .urgent_ramp_uclk_cycles = 689,
- .trip_to_memory_uclk_cycles = 895,
- .meta_trip_to_memory_uclk_cycles = 895,
- .maximum_latency_when_urgent_uclk_cycles = 289,
- .average_latency_when_urgent_uclk_cycles = 289,
- .maximum_latency_when_non_urgent_uclk_cycles = 895,
- .average_latency_when_non_urgent_uclk_cycles = 167,
- },
- {
- // State 8
- .minimum_uclk_khz = 0,
- .urgent_ramp_uclk_cycles = 716,
- .trip_to_memory_uclk_cycles = 902,
- .meta_trip_to_memory_uclk_cycles = 902,
- .maximum_latency_when_urgent_uclk_cycles = 316,
- .average_latency_when_urgent_uclk_cycles = 316,
- .maximum_latency_when_non_urgent_uclk_cycles = 902,
- .average_latency_when_non_urgent_uclk_cycles = 174,
- },
- },
- },
- },
- .qos_type = dml2_qos_param_type_dcn4x,
- },
-
- .power_management_parameters = {
- .dram_clk_change_blackout_us = 400,
- .fclk_change_blackout_us = 0,
- .g7_ppt_blackout_us = 0,
- .stutter_enter_plus_exit_latency_us = 50,
- .stutter_exit_latency_us = 43,
- .z8_stutter_enter_plus_exit_latency_us = 0,
- .z8_stutter_exit_latency_us = 0,
- },
-
- .vmin_limit = {
- .dispclk_khz = 600 * 1000,
- },
-
- .dprefclk_mhz = 700,
- .xtalclk_mhz = 100,
- .pcie_refclk_mhz = 100,
- .dchub_refclk_mhz = 50,
- .mall_allocated_for_dcn_mbytes = 64,
- .max_outstanding_reqs = 512,
- .fabric_datapath_to_dcn_data_return_bytes = 64,
- .return_bus_width_bytes = 64,
- .hostvm_min_page_size_kbytes = 0,
- .gpuvm_min_page_size_kbytes = 256,
- .phy_downspread_percent = 0,
- .dcn_downspread_percent = 0,
- .dispclk_dppclk_vco_speed_mhz = 4500,
- .do_urgent_latency_adjustment = 0,
- .mem_word_bytes = 32,
- .num_dcc_mcaches = 8,
- .mcache_size_bytes = 2048,
- .mcache_line_size_bytes = 32,
- .max_fclk_for_uclk_dpm_khz = 1250 * 1000,
-};
-
-static const struct dml2_ip_capabilities dml2_dcn31_max_ip_caps = {
- .pipe_count = 4,
- .otg_count = 4,
- .num_dsc = 4,
- .max_num_dp2p0_streams = 4,
- .max_num_hdmi_frl_outputs = 1,
- .max_num_dp2p0_outputs = 4,
- .rob_buffer_size_kbytes = 192,
- .config_return_buffer_size_in_kbytes = 1152,
- .meta_fifo_size_in_kentries = 22,
- .compressed_buffer_segment_size_in_kbytes = 64,
- .subvp_drr_scheduling_margin_us = 100,
- .subvp_prefetch_end_to_mall_start_us = 15,
- .subvp_fw_processing_delay = 15,
-
- .fams2 = {
- .max_allow_delay_us = 100 * 1000,
- .scheduling_delay_us = 50,
- .vertical_interrupt_ack_delay_us = 18,
- .allow_programming_delay_us = 18,
- .min_allow_width_us = 20,
- .subvp_df_throttle_delay_us = 100,
- .subvp_programming_delay_us = 18,
- .subvp_prefetch_to_mall_delay_us = 18,
- .drr_programming_delay_us = 18,
- },
-};
-
-#endif /* __DML_DML_DCN3_SOC_BB__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/bounding_boxes/dcn4_soc_bb.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/bounding_boxes/dcn4_soc_bb.h
index 8ef7977841de..793e1c038efd 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/bounding_boxes/dcn4_soc_bb.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/bounding_boxes/dcn4_soc_bb.h
@@ -344,6 +344,7 @@ static const struct dml2_ip_capabilities dml2_dcn401_max_ip_caps = {
.config_return_buffer_segment_size_in_kbytes = 64,
.meta_fifo_size_in_kentries = 22,
.compressed_buffer_segment_size_in_kbytes = 64,
+ .cursor_buffer_size = 24,
.max_flip_time_us = 80,
.max_flip_time_lines = 32,
.hostvm_mode = 0,
@@ -354,7 +355,7 @@ static const struct dml2_ip_capabilities dml2_dcn401_max_ip_caps = {
.fams2 = {
.max_allow_delay_us = 100 * 1000,
- .scheduling_delay_us = 125,
+ .scheduling_delay_us = 550,
.vertical_interrupt_ack_delay_us = 40,
.allow_programming_delay_us = 18,
.min_allow_width_us = 20,
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_dchub_registers.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_dchub_registers.h
index 25b607e7b726..b05030926ce8 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_dchub_registers.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_dchub_registers.h
@@ -46,6 +46,7 @@ struct dml2_display_dlg_regs {
uint32_t dst_y_delta_drq_limit;
uint32_t refcyc_per_vm_dmdata;
uint32_t dmdata_dl_delta;
+ uint32_t dst_y_svp_drq_limit;
// MRQ
uint32_t refcyc_per_meta_chunk_vblank_l;
@@ -156,6 +157,8 @@ struct dml2_dchub_watermark_regs {
uint32_t urgent;
uint32_t sr_enter;
uint32_t sr_exit;
+ uint32_t sr_enter_z8;
+ uint32_t sr_exit_z8;
uint32_t uclk_pstate;
uint32_t fclk_pstate;
uint32_t temp_read_or_ppt;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_display_cfg_types.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_display_cfg_types.h
index b132f676a68d..e8dc6471c0be 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_display_cfg_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_display_cfg_types.h
@@ -10,9 +10,10 @@
#define DML2_MAX_PLANES 8
#define DML2_MAX_DCN_PIPES 8
#define DML2_MAX_MCACHES 8 // assume plane is going to be supported by a max of 8 mcaches
+#define DML2_MAX_WRITEBACK 3
enum dml2_swizzle_mode {
- dml2_sw_linear,
+ dml2_sw_linear, // SW_LINEAR accepts 256 byte aligned pitch and also 128 byte aligned pitch if DCC is not enabled
dml2_sw_256b_2d,
dml2_sw_4kb_2d,
dml2_sw_64kb_2d,
@@ -24,7 +25,8 @@ enum dml2_swizzle_mode {
dml2_gfx11_sw_64kb_d_x,
dml2_gfx11_sw_64kb_r_x,
dml2_gfx11_sw_256kb_d_x,
- dml2_gfx11_sw_256kb_r_x
+ dml2_gfx11_sw_256kb_r_x,
+
};
enum dml2_source_format_class {
@@ -38,7 +40,13 @@ enum dml2_source_format_class {
dml2_rgbe_alpha = 9,
dml2_rgbe = 10,
dml2_mono_8 = 11,
- dml2_mono_16 = 12
+ dml2_mono_16 = 12,
+ dml2_422_planar_8 = 13,
+ dml2_422_planar_10 = 14,
+ dml2_422_planar_12 = 15,
+ dml2_422_packed_8 = 16,
+ dml2_422_packed_10 = 17,
+ dml2_422_packed_12 = 18
};
enum dml2_rotation_angle {
@@ -121,15 +129,6 @@ enum dml2_dsc_enable_option {
dml2_dsc_enable_if_necessary = 2
};
-enum dml2_pstate_support_method {
- dml2_pstate_method_uninitialized,
- dml2_pstate_method_not_supported,
- dml2_pstate_method_vactive,
- dml2_pstate_method_vblank,
- dml2_pstate_method_svp,
- dml2_pstate_method_drr
-};
-
enum dml2_tdlut_addressing_mode {
dml2_tdlut_sw_linear = 0,
dml2_tdlut_simple_linear = 1
@@ -167,7 +166,7 @@ struct dml2_surface_cfg {
enum dml2_swizzle_mode tiling;
struct {
- unsigned long pitch;
+ unsigned long pitch; // In elements, two pixels per element in 422 packed format
unsigned long width;
unsigned long height;
} plane0;
@@ -223,6 +222,7 @@ struct dml2_composition_cfg {
struct {
bool enabled;
+ bool upsp_enabled;
struct {
double h_ratio;
double v_ratio;
@@ -287,22 +287,23 @@ struct dml2_link_output_cfg {
bool validate_output; // Do not validate the link configuration for this display stream.
};
-struct dml2_writeback_cfg {
- bool enable;
+struct dml2_writeback_info {
enum dml2_source_format_class pixel_format;
- unsigned int active_writebacks_per_surface;
+ unsigned long input_width;
+ unsigned long input_height;
+ unsigned long output_width;
+ unsigned long output_height;
+ unsigned long v_taps;
+ unsigned long h_taps;
+ unsigned long v_taps_chroma;
+ unsigned long h_taps_chroma;
+ double h_ratio;
+ double v_ratio;
+};
- struct {
- bool enabled;
- unsigned long input_width;
- unsigned long input_height;
- unsigned long output_width;
- unsigned long output_height;
- unsigned long v_taps;
- unsigned long h_taps;
- double h_ratio;
- double v_ratio;
- } scaling_info;
+struct dml2_writeback_cfg {
+ unsigned int active_writebacks_per_stream;
+ struct dml2_writeback_info writeback_stream[DML2_MAX_WRITEBACK];
};
struct dml2_plane_parameters {
@@ -385,6 +386,7 @@ struct dml2_plane_parameters {
long reserved_vblank_time_ns;
unsigned int max_vactive_det_fill_delay_us; // 0 = no reserved time, +ve = explicit max delay
unsigned int gpuvm_min_page_size_kbytes;
+ unsigned int hostvm_min_page_size_kbytes;
enum dml2_svp_mode_override legacy_svp_config; //TODO remove in favor of svp_config
@@ -425,6 +427,7 @@ struct dml2_stream_parameters {
struct dml2_display_cfg {
bool gpuvm_enable;
+ bool ffbm_enable;
bool hostvm_enable;
// Allocate DET proportionally between streams based on pixel rate
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_soc_parameter_types.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_soc_parameter_types.h
index ebd8abe894a9..8c9f414aa6bf 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_soc_parameter_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_soc_parameter_types.h
@@ -93,12 +93,15 @@ struct dml2_soc_power_management_parameters {
double dram_clk_change_write_only_us;
double fclk_change_blackout_us;
double g7_ppt_blackout_us;
+ double g7_temperature_read_blackout_us;
double stutter_enter_plus_exit_latency_us;
double stutter_exit_latency_us;
double z8_stutter_enter_plus_exit_latency_us;
double z8_stutter_exit_latency_us;
double z8_min_idle_time;
double g6_temp_read_blackout_us[DML_MAX_CLK_TABLE_SIZE];
+ double type_b_dram_clk_change_blackout_us;
+ double type_b_ppt_blackout_us;
};
struct dml2_clk_table {
@@ -130,6 +133,7 @@ struct dml2_soc_state_table {
struct dml2_soc_vmin_clock_limits {
unsigned long dispclk_khz;
+ unsigned long dcfclk_khz;
};
struct dml2_soc_bb {
@@ -138,6 +142,7 @@ struct dml2_soc_bb {
struct dml2_soc_power_management_parameters power_management_parameters;
struct dml2_soc_vmin_clock_limits vmin_limit;
+ double lower_bound_bandwidth_dchub;
unsigned int dprefclk_mhz;
unsigned int xtalclk_mhz;
unsigned int pcie_refclk_mhz;
@@ -167,11 +172,13 @@ struct dml2_ip_capabilities {
unsigned int max_num_dp2p0_streams;
unsigned int max_num_hdmi_frl_outputs;
unsigned int max_num_dp2p0_outputs;
+ unsigned int max_num_wb;
unsigned int rob_buffer_size_kbytes;
unsigned int config_return_buffer_size_in_kbytes;
unsigned int config_return_buffer_segment_size_in_kbytes;
unsigned int meta_fifo_size_in_kentries;
unsigned int compressed_buffer_segment_size_in_kbytes;
+ unsigned int cursor_buffer_size;
unsigned int max_flip_time_us;
unsigned int max_flip_time_lines;
unsigned int hostvm_mode;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_types.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_types.h
index eeb96c455658..98c0234e2f47 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/inc/dml_top_types.h
@@ -14,11 +14,6 @@
struct dml2_instance;
-enum dml2_status {
- dml2_success = 0,
- dml2_error_generic = 1
-};
-
enum dml2_project_id {
dml2_project_invalid = 0,
dml2_project_dcn4x_stage1 = 1,
@@ -26,20 +21,14 @@ enum dml2_project_id {
dml2_project_dcn4x_stage2_auto_drr_svp = 3,
};
-enum dml2_dram_clock_change_support {
- dml2_dram_clock_change_vactive = 0,
- dml2_dram_clock_change_vblank = 1,
- dml2_dram_clock_change_vblank_and_vactive = 2,
- dml2_dram_clock_change_drr = 3,
- dml2_dram_clock_change_mall_svp = 4,
- dml2_dram_clock_change_mall_full_frame = 6,
- dml2_dram_clock_change_unsupported = 7
-};
-
-enum dml2_fclock_change_support {
- dml2_fclock_change_vactive = 0,
- dml2_fclock_change_vblank = 1,
- dml2_fclock_change_unsupported = 2
+enum dml2_pstate_change_support {
+ dml2_pstate_change_vactive = 0,
+ dml2_pstate_change_vblank = 1,
+ dml2_pstate_change_vblank_and_vactive = 2,
+ dml2_pstate_change_drr = 3,
+ dml2_pstate_change_mall_svp = 4,
+ dml2_pstate_change_mall_full_frame = 6,
+ dml2_pstate_change_unsupported = 7
};
enum dml2_output_type_and_rate__type {
@@ -64,7 +53,9 @@ enum dml2_output_type_and_rate__rate {
dml2_output_rate_hdmi_rate_6x4 = 9,
dml2_output_rate_hdmi_rate_8x4 = 10,
dml2_output_rate_hdmi_rate_10x4 = 11,
- dml2_output_rate_hdmi_rate_12x4 = 12
+ dml2_output_rate_hdmi_rate_12x4 = 12,
+ dml2_output_rate_hdmi_rate_16x4 = 13,
+ dml2_output_rate_hdmi_rate_20x4 = 14
};
struct dml2_pmo_options {
@@ -202,24 +193,23 @@ struct dml2_mcache_surface_allocation {
} informative;
};
-enum dml2_uclk_pstate_support_method {
- dml2_uclk_pstate_support_method_not_supported = 0,
- /* hw */
- dml2_uclk_pstate_support_method_vactive = 1,
- dml2_uclk_pstate_support_method_vblank = 2,
- dml2_uclk_pstate_support_method_reserved_hw = 5,
- /* fw */
- dml2_uclk_pstate_support_method_fw_subvp_phantom = 6,
- dml2_uclk_pstate_support_method_reserved_fw = 10,
- /* fw w/drr */
- dml2_uclk_pstate_support_method_fw_vactive_drr = 11,
- dml2_uclk_pstate_support_method_fw_vblank_drr = 12,
- dml2_uclk_pstate_support_method_fw_subvp_phantom_drr = 13,
- dml2_uclk_pstate_support_method_reserved_fw_drr_fixed = 20,
- dml2_uclk_pstate_support_method_fw_drr = 21,
- dml2_uclk_pstate_support_method_reserved_fw_drr_var = 22,
-
- dml2_uclk_pstate_support_method_count
+enum dml2_pstate_method {
+ dml2_pstate_method_na = 0,
+ /* hw exclusive modes */
+ dml2_pstate_method_vactive = 1,
+ dml2_pstate_method_vblank = 2,
+ dml2_pstate_method_reserved_hw = 5,
+ /* fw assisted exclusive modes */
+ dml2_pstate_method_fw_svp = 6,
+ dml2_pstate_method_reserved_fw = 10,
+ /* fw assisted modes requiring drr modulation */
+ dml2_pstate_method_fw_vactive_drr = 11,
+ dml2_pstate_method_fw_vblank_drr = 12,
+ dml2_pstate_method_fw_svp_drr = 13,
+ dml2_pstate_method_reserved_fw_drr_clamped = 20,
+ dml2_pstate_method_fw_drr = 21,
+ dml2_pstate_method_reserved_fw_drr_var = 22,
+ dml2_pstate_method_count
};
struct dml2_per_plane_programming {
@@ -241,7 +231,7 @@ struct dml2_per_plane_programming {
// If a stream is using odm split, then this value is always 1
unsigned int num_dpps_required;
- enum dml2_uclk_pstate_support_method uclk_pstate_support_method;
+ enum dml2_pstate_method uclk_pstate_support_method;
// MALL size requirements for MALL SS and SubVP
unsigned int surface_size_mall_bytes;
@@ -252,6 +242,7 @@ struct dml2_per_plane_programming {
struct {
bool valid;
struct dml2_plane_parameters descriptor;
+ struct dml2_mcache_surface_allocation mcache_allocation;
struct dml2_dchub_per_pipe_register_set *pipe_regs[DML2_MAX_PLANES];
} phantom_plane;
};
@@ -281,7 +272,7 @@ struct dml2_per_stream_programming {
unsigned int num_odms_required;
- enum dml2_uclk_pstate_support_method uclk_pstate_method;
+ enum dml2_pstate_method uclk_pstate_method;
struct {
bool enabled;
@@ -289,7 +280,11 @@ struct dml2_per_stream_programming {
union dml2_global_sync_programming global_sync;
} phantom_stream;
- struct dmub_fams2_stream_static_state fams2_params;
+ union dmub_cmd_fams2_config fams2_base_params;
+ union {
+ union dmub_cmd_fams2_config fams2_sub_params;
+ union dmub_fams2_stream_static_sub_state_v2 fams2_sub_params_v2;
+ };
};
//-----------------
@@ -339,7 +334,7 @@ struct dml2_mode_support_info {
bool DCCMetaBufferSizeNotExceeded;
bool TotalVerticalActiveBandwidthSupport;
bool VActiveBandwidthSupport;
- enum dml2_fclock_change_support FCLKChangeSupport[DML2_MAX_PLANES];
+ enum dml2_pstate_change_support FCLKChangeSupport[DML2_MAX_PLANES];
bool USRRetrainingSupport;
bool PrefetchSupported;
bool DynamicMetadataSupported;
@@ -361,6 +356,7 @@ struct dml2_mode_support_info {
unsigned int AlignedYPitch[DML2_MAX_PLANES];
unsigned int AlignedCPitch[DML2_MAX_PLANES];
bool g6_temp_read_support;
+ bool temp_read_or_ppt_support;
}; // dml2_mode_support_info
struct dml2_display_cfg_programming {
@@ -392,6 +388,11 @@ struct dml2_display_cfg_programming {
unsigned long fclk_khz;
unsigned long dcfclk_khz;
} svp_prefetch;
+ struct {
+ unsigned long uclk_khz;
+ unsigned long fclk_khz;
+ unsigned long dcfclk_khz;
+ } svp_prefetch_no_throttle;
unsigned long deepsleep_dcfclk_khz;
unsigned long dispclk_khz;
@@ -444,7 +445,7 @@ struct dml2_display_cfg_programming {
double pstate_change_us;
double fclk_pstate_change_us;
double usr_retraining_us;
- double g6_temp_read_watermark_us;
+ double temp_read_or_ppt_watermark_us;
} watermarks;
struct {
@@ -458,6 +459,10 @@ struct dml2_display_cfg_programming {
} plane_info[DML2_MAX_PLANES];
struct {
+ unsigned int total_num_dpps_required;
+ } dpp;
+
+ struct {
unsigned long long total_surface_size_in_mall_bytes;
unsigned int subviewport_lines_needed_in_mall[DML2_MAX_PLANES];
} mall;
@@ -653,6 +658,7 @@ struct dml2_display_cfg_programming {
double DisplayPipeLineDeliveryTimeLumaPrefetch[DML2_MAX_PLANES];
double DisplayPipeLineDeliveryTimeChromaPrefetch[DML2_MAX_PLANES];
+ double WritebackRequiredBandwidth;
double WritebackAllowDRAMClockChangeEndPosition[DML2_MAX_PLANES];
double WritebackAllowFCLKChangeEndPosition[DML2_MAX_PLANES];
double DSCCLK_calculated[DML2_MAX_PLANES];
@@ -662,6 +668,7 @@ struct dml2_display_cfg_programming {
double MaxActiveDRAMClockChangeLatencySupported[DML2_MAX_PLANES];
unsigned int PrefetchMode[DML2_MAX_PLANES]; // LEGACY_ONLY
bool ROBUrgencyAvoidance;
+ double LowestPrefetchMargin;
} misc;
struct dml2_mode_support_info mode_support_info;
@@ -672,9 +679,15 @@ struct dml2_display_cfg_programming {
// unlimited # of mcache
struct dml2_mcache_surface_allocation non_optimized_mcache_allocation[DML2_MAX_PLANES];
+ bool failed_prefetch;
+ bool failed_uclk_pstate;
bool failed_mcache_validation;
bool failed_dpmm;
bool failed_mode_programming;
+ bool failed_mode_programming_dcfclk;
+ bool failed_mode_programming_prefetch;
+ bool failed_mode_programming_flip;
+ bool failed_map_watermarks;
} informative;
};
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.c
index 3d41ffde91c1..6ee37386f672 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.c
@@ -9,7 +9,7 @@
#include "dml2_debug.h"
#include "lib_float_math.h"
-static const struct dml2_core_ip_params core_dcn4_ip_caps_base = {
+struct dml2_core_ip_params core_dcn4_ip_caps_base = {
// Hardcoded values for DCN3x
.vblank_nom_default_us = 668,
.remote_iommu_outstanding_translations = 256,
@@ -90,6 +90,7 @@ static void patch_ip_caps_with_explicit_ip_params(struct dml2_ip_capabilities *i
ip_caps->config_return_buffer_segment_size_in_kbytes = ip_params->config_return_buffer_segment_size_in_kbytes;
ip_caps->meta_fifo_size_in_kentries = ip_params->meta_fifo_size_in_kentries;
ip_caps->compressed_buffer_segment_size_in_kbytes = ip_params->compressed_buffer_segment_size_in_kbytes;
+ ip_caps->cursor_buffer_size = ip_params->cursor_buffer_size;
ip_caps->max_flip_time_us = ip_params->max_flip_time_us;
ip_caps->max_flip_time_lines = ip_params->max_flip_time_lines;
ip_caps->hostvm_mode = ip_params->hostvm_mode;
@@ -114,6 +115,7 @@ static void patch_ip_params_with_ip_caps(struct dml2_core_ip_params *ip_params,
ip_params->config_return_buffer_segment_size_in_kbytes = ip_caps->config_return_buffer_segment_size_in_kbytes;
ip_params->meta_fifo_size_in_kentries = ip_caps->meta_fifo_size_in_kentries;
ip_params->compressed_buffer_segment_size_in_kbytes = ip_caps->compressed_buffer_segment_size_in_kbytes;
+ ip_params->cursor_buffer_size = ip_caps->cursor_buffer_size;
ip_params->max_flip_time_us = ip_caps->max_flip_time_us;
ip_params->max_flip_time_lines = ip_caps->max_flip_time_lines;
ip_params->hostvm_mode = ip_caps->hostvm_mode;
@@ -139,9 +141,8 @@ bool core_dcn4_initialize(struct dml2_core_initialize_in_out *in_out)
core->clean_me_up.mode_lib.ip.subvp_fw_processing_delay_us = core_dcn4_ip_caps_base.subvp_pstate_allow_width_us;
core->clean_me_up.mode_lib.ip.subvp_swath_height_margin_lines = core_dcn4_ip_caps_base.subvp_swath_height_margin_lines;
} else {
- memcpy(&core->clean_me_up.mode_lib.ip, &core_dcn4_ip_caps_base, sizeof(struct dml2_core_ip_params));
+ memcpy(&core->clean_me_up.mode_lib.ip, &core_dcn4_ip_caps_base, sizeof(struct dml2_core_ip_params));
patch_ip_params_with_ip_caps(&core->clean_me_up.mode_lib.ip, in_out->ip_caps);
-
core->clean_me_up.mode_lib.ip.imall_supported = false;
}
@@ -252,7 +253,8 @@ static void expand_implict_subvp(const struct display_configuation_with_meta *di
static void pack_mode_programming_params_with_implicit_subvp(struct dml2_core_instance *core, const struct display_configuation_with_meta *display_cfg,
const struct dml2_display_cfg *svp_expanded_display_cfg, struct dml2_display_cfg_programming *programming, struct dml2_core_scratch *scratch)
{
- unsigned int stream_index, plane_index, pipe_offset, stream_already_populated_mask, main_plane_index;
+ unsigned int stream_index, plane_index, pipe_offset, stream_already_populated_mask, main_plane_index, mcache_index;
+ unsigned int total_main_mcaches_required = 0;
int total_pipe_regs_copied = 0;
int dml_internal_pipe_index = 0;
const struct dml2_plane_parameters *main_plane;
@@ -316,32 +318,20 @@ static void pack_mode_programming_params_with_implicit_subvp(struct dml2_core_in
// Setup the appropriate p-state strategy
if (display_cfg->stage3.performed && display_cfg->stage3.success) {
- switch (display_cfg->stage3.pstate_switch_modes[plane_index]) {
- case dml2_uclk_pstate_support_method_vactive:
- case dml2_uclk_pstate_support_method_vblank:
- case dml2_uclk_pstate_support_method_fw_subvp_phantom:
- case dml2_uclk_pstate_support_method_fw_drr:
- case dml2_uclk_pstate_support_method_fw_vactive_drr:
- case dml2_uclk_pstate_support_method_fw_vblank_drr:
- case dml2_uclk_pstate_support_method_fw_subvp_phantom_drr:
- programming->plane_programming[plane_index].uclk_pstate_support_method = display_cfg->stage3.pstate_switch_modes[plane_index];
- break;
- case dml2_uclk_pstate_support_method_reserved_hw:
- case dml2_uclk_pstate_support_method_reserved_fw:
- case dml2_uclk_pstate_support_method_reserved_fw_drr_fixed:
- case dml2_uclk_pstate_support_method_reserved_fw_drr_var:
- case dml2_uclk_pstate_support_method_not_supported:
- case dml2_uclk_pstate_support_method_count:
- default:
- programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_uclk_pstate_support_method_not_supported;
- break;
- }
+ programming->plane_programming[plane_index].uclk_pstate_support_method = display_cfg->stage3.pstate_switch_modes[plane_index];
} else {
- programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_uclk_pstate_support_method_not_supported;
+ programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_pstate_method_na;
}
dml2_core_calcs_get_mall_allocation(&core->clean_me_up.mode_lib, &programming->plane_programming[plane_index].surface_size_mall_bytes, dml_internal_pipe_index);
+ memcpy(&programming->plane_programming[plane_index].mcache_allocation,
+ &display_cfg->stage2.mcache_allocations[plane_index],
+ sizeof(struct dml2_mcache_surface_allocation));
+ total_main_mcaches_required += programming->plane_programming[plane_index].mcache_allocation.num_mcaches_plane0 +
+ programming->plane_programming[plane_index].mcache_allocation.num_mcaches_plane1 -
+ (programming->plane_programming[plane_index].mcache_allocation.last_slice_sharing.plane0_plane1 ? 1 : 0);
+
for (pipe_offset = 0; pipe_offset < programming->plane_programming[plane_index].num_dpps_required; pipe_offset++) {
// Assign storage for this pipe's register values
programming->plane_programming[plane_index].pipe_regs[pipe_offset] = &programming->pipe_regs[total_pipe_regs_copied];
@@ -360,7 +350,8 @@ static void pack_mode_programming_params_with_implicit_subvp(struct dml2_core_in
/* unconditionally populate fams2 params */
dml2_core_calcs_get_stream_fams2_programming(&core->clean_me_up.mode_lib,
display_cfg,
- &programming->stream_programming[main_plane->stream_index].fams2_params,
+ &programming->stream_programming[main_plane->stream_index].fams2_base_params,
+ &programming->stream_programming[main_plane->stream_index].fams2_sub_params,
programming->stream_programming[main_plane->stream_index].uclk_pstate_method,
plane_index);
@@ -379,6 +370,22 @@ static void pack_mode_programming_params_with_implicit_subvp(struct dml2_core_in
memcpy(&programming->plane_programming[main_plane_index].phantom_plane.descriptor, phantom_plane, sizeof(struct dml2_plane_parameters));
dml2_core_calcs_get_mall_allocation(&core->clean_me_up.mode_lib, &programming->plane_programming[main_plane_index].svp_size_mall_bytes, dml_internal_pipe_index);
+
+ /* generate mcache allocation, phantoms use identical mcache configuration, but in the MALL set and unique mcache ID's beginning after all main ID's */
+ memcpy(&programming->plane_programming[main_plane_index].phantom_plane.mcache_allocation,
+ &programming->plane_programming[main_plane_index].mcache_allocation,
+ sizeof(struct dml2_mcache_surface_allocation));
+ for (mcache_index = 0; mcache_index < programming->plane_programming[main_plane_index].phantom_plane.mcache_allocation.num_mcaches_plane0; mcache_index++) {
+ programming->plane_programming[main_plane_index].phantom_plane.mcache_allocation.global_mcache_ids_plane0[mcache_index] += total_main_mcaches_required;
+ programming->plane_programming[main_plane_index].phantom_plane.mcache_allocation.global_mcache_ids_mall_plane0[mcache_index] =
+ programming->plane_programming[main_plane_index].phantom_plane.mcache_allocation.global_mcache_ids_plane0[mcache_index];
+ }
+ for (mcache_index = 0; mcache_index < programming->plane_programming[main_plane_index].phantom_plane.mcache_allocation.num_mcaches_plane1; mcache_index++) {
+ programming->plane_programming[main_plane_index].phantom_plane.mcache_allocation.global_mcache_ids_plane1[mcache_index] += total_main_mcaches_required;
+ programming->plane_programming[main_plane_index].phantom_plane.mcache_allocation.global_mcache_ids_mall_plane1[mcache_index] =
+ programming->plane_programming[main_plane_index].phantom_plane.mcache_allocation.global_mcache_ids_plane1[mcache_index];
+ }
+
for (pipe_offset = 0; pipe_offset < programming->plane_programming[main_plane_index].num_dpps_required; pipe_offset++) {
// Assign storage for this pipe's register values
programming->plane_programming[main_plane_index].phantom_plane.pipe_regs[pipe_offset] = &programming->pipe_regs[total_pipe_regs_copied];
@@ -449,10 +456,10 @@ bool core_dcn4_mode_support(struct dml2_core_mode_support_in_out *in_out)
in_out->mode_support_result.global.active.urgent_bw_dram_kbps = (unsigned long)math_ceil2((l->mode_support_ex_params.out_evaluation_info->urg_bandwidth_required_flip[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram] * 1000), 1.0);
in_out->mode_support_result.global.svp_prefetch.average_bw_dram_kbps = (unsigned long)math_ceil2((l->mode_support_ex_params.out_evaluation_info->avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram] * 1000), 1.0);
in_out->mode_support_result.global.svp_prefetch.urgent_bw_dram_kbps = (unsigned long)math_ceil2((l->mode_support_ex_params.out_evaluation_info->urg_bandwidth_required_flip[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram] * 1000), 1.0);
- dml2_printf("DML::%s: in_out->mode_support_result.global.active.urgent_bw_sdp_kbps = %ld\n", __func__, in_out->mode_support_result.global.active.urgent_bw_sdp_kbps);
- dml2_printf("DML::%s: in_out->mode_support_result.global.svp_prefetch.urgent_bw_sdp_kbps = %ld\n", __func__, in_out->mode_support_result.global.svp_prefetch.urgent_bw_sdp_kbps);
- dml2_printf("DML::%s: in_out->mode_support_result.global.active.urgent_bw_dram_kbps = %ld\n", __func__, in_out->mode_support_result.global.active.urgent_bw_dram_kbps);
- dml2_printf("DML::%s: in_out->mode_support_result.global.svp_prefetch.urgent_bw_dram_kbps = %ld\n", __func__, in_out->mode_support_result.global.svp_prefetch.urgent_bw_dram_kbps);
+ DML_LOG_VERBOSE("DML::%s: in_out->mode_support_result.global.active.urgent_bw_sdp_kbps = %ld\n", __func__, in_out->mode_support_result.global.active.urgent_bw_sdp_kbps);
+ DML_LOG_VERBOSE("DML::%s: in_out->mode_support_result.global.svp_prefetch.urgent_bw_sdp_kbps = %ld\n", __func__, in_out->mode_support_result.global.svp_prefetch.urgent_bw_sdp_kbps);
+ DML_LOG_VERBOSE("DML::%s: in_out->mode_support_result.global.active.urgent_bw_dram_kbps = %ld\n", __func__, in_out->mode_support_result.global.active.urgent_bw_dram_kbps);
+ DML_LOG_VERBOSE("DML::%s: in_out->mode_support_result.global.svp_prefetch.urgent_bw_dram_kbps = %ld\n", __func__, in_out->mode_support_result.global.svp_prefetch.urgent_bw_dram_kbps);
for (i = 0; i < l->svp_expanded_display_cfg.num_planes; i++) {
in_out->mode_support_result.per_plane[i].dppclk_khz = (unsigned int)(core->clean_me_up.mode_lib.ms.RequiredDPPCLK[i] * 1000);
@@ -502,7 +509,7 @@ bool core_dcn4_mode_support(struct dml2_core_mode_support_in_out *in_out)
stream_index = l->svp_expanded_display_cfg.plane_descriptors[i].stream_index;
in_out->mode_support_result.per_stream[stream_index].dscclk_khz = (unsigned int)core->clean_me_up.mode_lib.ms.required_dscclk_freq_mhz[i] * 1000;
- dml2_printf("CORE_DCN4::%s: i=%d stream_index=%d, in_out->mode_support_result.per_stream[stream_index].dscclk_khz = %u\n", __func__, i, stream_index, in_out->mode_support_result.per_stream[stream_index].dscclk_khz);
+ DML_LOG_VERBOSE("CORE_DCN4::%s: i=%d stream_index=%d, in_out->mode_support_result.per_stream[stream_index].dscclk_khz = %u\n", __func__, i, stream_index, in_out->mode_support_result.per_stream[stream_index].dscclk_khz);
if (!((stream_bitmask >> stream_index) & 0x1)) {
in_out->mode_support_result.cfg_support_info.stream_support_info[stream_index].odms_used = odm_count;
@@ -572,22 +579,26 @@ bool core_dcn4_mode_programming(struct dml2_core_mode_programming_in_out *in_out
in_out->programming->plane_programming[plane_index].num_dpps_required = core->clean_me_up.mode_lib.mp.NoOfDPP[plane_index];
if (in_out->programming->display_config.plane_descriptors[plane_index].overrides.legacy_svp_config == dml2_svp_mode_override_main_pipe)
- in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_uclk_pstate_support_method_fw_subvp_phantom;
+ in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_pstate_method_fw_svp;
else if (in_out->programming->display_config.plane_descriptors[plane_index].overrides.legacy_svp_config == dml2_svp_mode_override_phantom_pipe)
- in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_uclk_pstate_support_method_fw_subvp_phantom;
+ in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_pstate_method_fw_svp;
else if (in_out->programming->display_config.plane_descriptors[plane_index].overrides.legacy_svp_config == dml2_svp_mode_override_phantom_pipe_no_data_return)
- in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_uclk_pstate_support_method_fw_subvp_phantom;
+ in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_pstate_method_fw_svp;
else {
if (core->clean_me_up.mode_lib.mp.MaxActiveDRAMClockChangeLatencySupported[plane_index] >= core->clean_me_up.mode_lib.soc.power_management_parameters.dram_clk_change_blackout_us)
- in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_uclk_pstate_support_method_vactive;
+ in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_pstate_method_vactive;
else if (core->clean_me_up.mode_lib.mp.TWait[plane_index] >= core->clean_me_up.mode_lib.soc.power_management_parameters.dram_clk_change_blackout_us)
- in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_uclk_pstate_support_method_vblank;
+ in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_pstate_method_vblank;
else
- in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_uclk_pstate_support_method_not_supported;
+ in_out->programming->plane_programming[plane_index].uclk_pstate_support_method = dml2_pstate_method_na;
}
dml2_core_calcs_get_mall_allocation(&core->clean_me_up.mode_lib, &in_out->programming->plane_programming[plane_index].surface_size_mall_bytes, dml_internal_pipe_index);
+ memcpy(&in_out->programming->plane_programming[plane_index].mcache_allocation,
+ &in_out->display_cfg->stage2.mcache_allocations[plane_index],
+ sizeof(struct dml2_mcache_surface_allocation));
+
for (pipe_offset = 0; pipe_offset < in_out->programming->plane_programming[plane_index].num_dpps_required; pipe_offset++) {
in_out->programming->plane_programming[plane_index].plane_descriptor = &in_out->programming->display_config.plane_descriptors[plane_index];
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.h
index e62b2d3eeee6..a68bb001a346 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4.h
@@ -9,7 +9,4 @@ bool core_dcn4_mode_support(struct dml2_core_mode_support_in_out *in_out);
bool core_dcn4_mode_programming(struct dml2_core_mode_programming_in_out *in_out);
bool core_dcn4_populate_informative(struct dml2_core_populate_informative_in_out *in_out);
bool core_dcn4_calculate_mcache_allocation(struct dml2_calculate_mcache_allocation_in_out *in_out);
-
-bool core_dcn4_unit_test(void);
-
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c
index 601320b1be81..b9cff2198511 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.c
@@ -11,7 +11,11 @@
#define DML2_MAX_FMT_420_BUFFER_WIDTH 4096
#define DML_MAX_NUM_OF_SLICES_PER_DSC 4
+#define DML_MAX_COMPRESSION_RATIO 4
+//#define DML_MODE_SUPPORT_USE_DPM_DRAM_BW
+//#define DML_GLOBAL_PREFETCH_CHECK
#define ALLOW_SDPIF_RATE_LIMIT_PRE_CSTATE
+#define DML_MAX_VSTARTUP_START 1023
const char *dml2_core_internal_bw_type_str(enum dml2_core_internal_bw_type bw_type)
{
@@ -50,104 +54,104 @@ static double dml2_core_div_rem(double dividend, unsigned int divisor, unsigned
static void dml2_print_mode_support_info(const struct dml2_core_internal_mode_support_info *support, bool fail_only)
{
- dml2_printf("DML: ===================================== \n");
- dml2_printf("DML: DML_MODE_SUPPORT_INFO_ST\n");
+ DML_LOG_VERBOSE("DML: ===================================== \n");
+ DML_LOG_VERBOSE("DML: DML_MODE_SUPPORT_INFO_ST\n");
if (!fail_only || support->ScaleRatioAndTapsSupport == 0)
- dml2_printf("DML: support: ScaleRatioAndTapsSupport = %d\n", support->ScaleRatioAndTapsSupport);
+ DML_LOG_VERBOSE("DML: support: ScaleRatioAndTapsSupport = %d\n", support->ScaleRatioAndTapsSupport);
if (!fail_only || support->SourceFormatPixelAndScanSupport == 0)
- dml2_printf("DML: support: SourceFormatPixelAndScanSupport = %d\n", support->SourceFormatPixelAndScanSupport);
+ DML_LOG_VERBOSE("DML: support: SourceFormatPixelAndScanSupport = %d\n", support->SourceFormatPixelAndScanSupport);
if (!fail_only || support->ViewportSizeSupport == 0)
- dml2_printf("DML: support: ViewportSizeSupport = %d\n", support->ViewportSizeSupport);
+ DML_LOG_VERBOSE("DML: support: ViewportSizeSupport = %d\n", support->ViewportSizeSupport);
if (!fail_only || support->LinkRateDoesNotMatchDPVersion == 1)
- dml2_printf("DML: support: LinkRateDoesNotMatchDPVersion = %d\n", support->LinkRateDoesNotMatchDPVersion);
+ DML_LOG_VERBOSE("DML: support: LinkRateDoesNotMatchDPVersion = %d\n", support->LinkRateDoesNotMatchDPVersion);
if (!fail_only || support->LinkRateForMultistreamNotIndicated == 1)
- dml2_printf("DML: support: LinkRateForMultistreamNotIndicated = %d\n", support->LinkRateForMultistreamNotIndicated);
+ DML_LOG_VERBOSE("DML: support: LinkRateForMultistreamNotIndicated = %d\n", support->LinkRateForMultistreamNotIndicated);
if (!fail_only || support->BPPForMultistreamNotIndicated == 1)
- dml2_printf("DML: support: BPPForMultistreamNotIndicated = %d\n", support->BPPForMultistreamNotIndicated);
+ DML_LOG_VERBOSE("DML: support: BPPForMultistreamNotIndicated = %d\n", support->BPPForMultistreamNotIndicated);
if (!fail_only || support->MultistreamWithHDMIOreDP == 1)
- dml2_printf("DML: support: MultistreamWithHDMIOreDP = %d\n", support->MultistreamWithHDMIOreDP);
+ DML_LOG_VERBOSE("DML: support: MultistreamWithHDMIOreDP = %d\n", support->MultistreamWithHDMIOreDP);
if (!fail_only || support->ExceededMultistreamSlots == 1)
- dml2_printf("DML: support: ExceededMultistreamSlots = %d\n", support->ExceededMultistreamSlots);
+ DML_LOG_VERBOSE("DML: support: ExceededMultistreamSlots = %d\n", support->ExceededMultistreamSlots);
if (!fail_only || support->MSOOrODMSplitWithNonDPLink == 1)
- dml2_printf("DML: support: MSOOrODMSplitWithNonDPLink = %d\n", support->MSOOrODMSplitWithNonDPLink);
+ DML_LOG_VERBOSE("DML: support: MSOOrODMSplitWithNonDPLink = %d\n", support->MSOOrODMSplitWithNonDPLink);
if (!fail_only || support->NotEnoughLanesForMSO == 1)
- dml2_printf("DML: support: NotEnoughLanesForMSO = %d\n", support->NotEnoughLanesForMSO);
+ DML_LOG_VERBOSE("DML: support: NotEnoughLanesForMSO = %d\n", support->NotEnoughLanesForMSO);
if (!fail_only || support->P2IWith420 == 1)
- dml2_printf("DML: support: P2IWith420 = %d\n", support->P2IWith420);
+ DML_LOG_VERBOSE("DML: support: P2IWith420 = %d\n", support->P2IWith420);
if (!fail_only || support->DSC422NativeNotSupported == 1)
- dml2_printf("DML: support: DSC422NativeNotSupported = %d\n", support->DSC422NativeNotSupported);
+ DML_LOG_VERBOSE("DML: support: DSC422NativeNotSupported = %d\n", support->DSC422NativeNotSupported);
if (!fail_only || support->DSCSlicesODMModeSupported == 0)
- dml2_printf("DML: support: DSCSlicesODMModeSupported = %d\n", support->DSCSlicesODMModeSupported);
+ DML_LOG_VERBOSE("DML: support: DSCSlicesODMModeSupported = %d\n", support->DSCSlicesODMModeSupported);
if (!fail_only || support->NotEnoughDSCUnits == 1)
- dml2_printf("DML: support: NotEnoughDSCUnits = %d\n", support->NotEnoughDSCUnits);
+ DML_LOG_VERBOSE("DML: support: NotEnoughDSCUnits = %d\n", support->NotEnoughDSCUnits);
if (!fail_only || support->NotEnoughDSCSlices == 1)
- dml2_printf("DML: support: NotEnoughDSCSlices = %d\n", support->NotEnoughDSCSlices);
+ DML_LOG_VERBOSE("DML: support: NotEnoughDSCSlices = %d\n", support->NotEnoughDSCSlices);
if (!fail_only || support->ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe == 1)
- dml2_printf("DML: support: ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe = %d\n", support->ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe);
+ DML_LOG_VERBOSE("DML: support: ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe = %d\n", support->ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe);
if (!fail_only || support->InvalidCombinationOfMALLUseForPStateAndStaticScreen == 1)
- dml2_printf("DML: support: InvalidCombinationOfMALLUseForPStateAndStaticScreen = %d\n", support->InvalidCombinationOfMALLUseForPStateAndStaticScreen);
+ DML_LOG_VERBOSE("DML: support: InvalidCombinationOfMALLUseForPStateAndStaticScreen = %d\n", support->InvalidCombinationOfMALLUseForPStateAndStaticScreen);
if (!fail_only || support->DSCCLKRequiredMoreThanSupported == 1)
- dml2_printf("DML: support: DSCCLKRequiredMoreThanSupported = %d\n", support->DSCCLKRequiredMoreThanSupported);
+ DML_LOG_VERBOSE("DML: support: DSCCLKRequiredMoreThanSupported = %d\n", support->DSCCLKRequiredMoreThanSupported);
if (!fail_only || support->PixelsPerLinePerDSCUnitSupport == 0)
- dml2_printf("DML: support: PixelsPerLinePerDSCUnitSupport = %d\n", support->PixelsPerLinePerDSCUnitSupport);
+ DML_LOG_VERBOSE("DML: support: PixelsPerLinePerDSCUnitSupport = %d\n", support->PixelsPerLinePerDSCUnitSupport);
if (!fail_only || support->DTBCLKRequiredMoreThanSupported == 1)
- dml2_printf("DML: support: DTBCLKRequiredMoreThanSupported = %d\n", support->DTBCLKRequiredMoreThanSupported);
+ DML_LOG_VERBOSE("DML: support: DTBCLKRequiredMoreThanSupported = %d\n", support->DTBCLKRequiredMoreThanSupported);
if (!fail_only || support->InvalidCombinationOfMALLUseForPState == 1)
- dml2_printf("DML: support: InvalidCombinationOfMALLUseForPState = %d\n", support->InvalidCombinationOfMALLUseForPState);
+ DML_LOG_VERBOSE("DML: support: InvalidCombinationOfMALLUseForPState = %d\n", support->InvalidCombinationOfMALLUseForPState);
if (!fail_only || support->ROBSupport == 0)
- dml2_printf("DML: support: ROBSupport = %d\n", support->ROBSupport);
+ DML_LOG_VERBOSE("DML: support: ROBSupport = %d\n", support->ROBSupport);
if (!fail_only || support->OutstandingRequestsSupport == 0)
- dml2_printf("DML: support: OutstandingRequestsSupport = %d\n", support->OutstandingRequestsSupport);
+ DML_LOG_VERBOSE("DML: support: OutstandingRequestsSupport = %d\n", support->OutstandingRequestsSupport);
if (!fail_only || support->OutstandingRequestsUrgencyAvoidance == 0)
- dml2_printf("DML: support: OutstandingRequestsUrgencyAvoidance = %d\n", support->OutstandingRequestsUrgencyAvoidance);
+ DML_LOG_VERBOSE("DML: support: OutstandingRequestsUrgencyAvoidance = %d\n", support->OutstandingRequestsUrgencyAvoidance);
if (!fail_only || support->DISPCLK_DPPCLK_Support == 0)
- dml2_printf("DML: support: DISPCLK_DPPCLK_Support = %d\n", support->DISPCLK_DPPCLK_Support);
+ DML_LOG_VERBOSE("DML: support: DISPCLK_DPPCLK_Support = %d\n", support->DISPCLK_DPPCLK_Support);
if (!fail_only || support->TotalAvailablePipesSupport == 0)
- dml2_printf("DML: support: TotalAvailablePipesSupport = %d\n", support->TotalAvailablePipesSupport);
+ DML_LOG_VERBOSE("DML: support: TotalAvailablePipesSupport = %d\n", support->TotalAvailablePipesSupport);
if (!fail_only || support->NumberOfOTGSupport == 0)
- dml2_printf("DML: support: NumberOfOTGSupport = %d\n", support->NumberOfOTGSupport);
+ DML_LOG_VERBOSE("DML: support: NumberOfOTGSupport = %d\n", support->NumberOfOTGSupport);
if (!fail_only || support->NumberOfHDMIFRLSupport == 0)
- dml2_printf("DML: support: NumberOfHDMIFRLSupport = %d\n", support->NumberOfHDMIFRLSupport);
+ DML_LOG_VERBOSE("DML: support: NumberOfHDMIFRLSupport = %d\n", support->NumberOfHDMIFRLSupport);
if (!fail_only || support->NumberOfDP2p0Support == 0)
- dml2_printf("DML: support: NumberOfDP2p0Support = %d\n", support->NumberOfDP2p0Support);
+ DML_LOG_VERBOSE("DML: support: NumberOfDP2p0Support = %d\n", support->NumberOfDP2p0Support);
if (!fail_only || support->EnoughWritebackUnits == 0)
- dml2_printf("DML: support: EnoughWritebackUnits = %d\n", support->EnoughWritebackUnits);
+ DML_LOG_VERBOSE("DML: support: EnoughWritebackUnits = %d\n", support->EnoughWritebackUnits);
if (!fail_only || support->WritebackScaleRatioAndTapsSupport == 0)
- dml2_printf("DML: support: WritebackScaleRatioAndTapsSupport = %d\n", support->WritebackScaleRatioAndTapsSupport);
+ DML_LOG_VERBOSE("DML: support: WritebackScaleRatioAndTapsSupport = %d\n", support->WritebackScaleRatioAndTapsSupport);
if (!fail_only || support->WritebackLatencySupport == 0)
- dml2_printf("DML: support: WritebackLatencySupport = %d\n", support->WritebackLatencySupport);
+ DML_LOG_VERBOSE("DML: support: WritebackLatencySupport = %d\n", support->WritebackLatencySupport);
if (!fail_only || support->CursorSupport == 0)
- dml2_printf("DML: support: CursorSupport = %d\n", support->CursorSupport);
+ DML_LOG_VERBOSE("DML: support: CursorSupport = %d\n", support->CursorSupport);
if (!fail_only || support->PitchSupport == 0)
- dml2_printf("DML: support: PitchSupport = %d\n", support->PitchSupport);
+ DML_LOG_VERBOSE("DML: support: PitchSupport = %d\n", support->PitchSupport);
if (!fail_only || support->ViewportExceedsSurface == 1)
- dml2_printf("DML: support: ViewportExceedsSurface = %d\n", support->ViewportExceedsSurface);
+ DML_LOG_VERBOSE("DML: support: ViewportExceedsSurface = %d\n", support->ViewportExceedsSurface);
if (!fail_only || support->PrefetchSupported == 0)
- dml2_printf("DML: support: PrefetchSupported = %d\n", support->PrefetchSupported);
+ DML_LOG_VERBOSE("DML: support: PrefetchSupported = %d\n", support->PrefetchSupported);
if (!fail_only || support->EnoughUrgentLatencyHidingSupport == 0)
- dml2_printf("DML: support: EnoughUrgentLatencyHidingSupport = %d\n", support->EnoughUrgentLatencyHidingSupport);
+ DML_LOG_VERBOSE("DML: support: EnoughUrgentLatencyHidingSupport = %d\n", support->EnoughUrgentLatencyHidingSupport);
if (!fail_only || support->AvgBandwidthSupport == 0)
- dml2_printf("DML: support: AvgBandwidthSupport = %d\n", support->AvgBandwidthSupport);
+ DML_LOG_VERBOSE("DML: support: AvgBandwidthSupport = %d\n", support->AvgBandwidthSupport);
if (!fail_only || support->DynamicMetadataSupported == 0)
- dml2_printf("DML: support: DynamicMetadataSupported = %d\n", support->DynamicMetadataSupported);
+ DML_LOG_VERBOSE("DML: support: DynamicMetadataSupported = %d\n", support->DynamicMetadataSupported);
if (!fail_only || support->VRatioInPrefetchSupported == 0)
- dml2_printf("DML: support: VRatioInPrefetchSupported = %d\n", support->VRatioInPrefetchSupported);
- if (!fail_only || support->PTEBufferSizeNotExceeded == 1)
- dml2_printf("DML: support: PTEBufferSizeNotExceeded = %d\n", support->PTEBufferSizeNotExceeded);
- if (!fail_only || support->DCCMetaBufferSizeNotExceeded == 1)
- dml2_printf("DML: support: DCCMetaBufferSizeNotExceeded = %d\n", support->DCCMetaBufferSizeNotExceeded);
+ DML_LOG_VERBOSE("DML: support: VRatioInPrefetchSupported = %d\n", support->VRatioInPrefetchSupported);
+ if (!fail_only || support->PTEBufferSizeNotExceeded == 0)
+ DML_LOG_VERBOSE("DML: support: PTEBufferSizeNotExceeded = %d\n", support->PTEBufferSizeNotExceeded);
+ if (!fail_only || support->DCCMetaBufferSizeNotExceeded == 0)
+ DML_LOG_VERBOSE("DML: support: DCCMetaBufferSizeNotExceeded = %d\n", support->DCCMetaBufferSizeNotExceeded);
if (!fail_only || support->ExceededMALLSize == 1)
- dml2_printf("DML: support: ExceededMALLSize = %d\n", support->ExceededMALLSize);
+ DML_LOG_VERBOSE("DML: support: ExceededMALLSize = %d\n", support->ExceededMALLSize);
if (!fail_only || support->g6_temp_read_support == 0)
- dml2_printf("DML: support: g6_temp_read_support = %d\n", support->g6_temp_read_support);
+ DML_LOG_VERBOSE("DML: support: g6_temp_read_support = %d\n", support->g6_temp_read_support);
if (!fail_only || support->ImmediateFlipSupport == 0)
- dml2_printf("DML: support: ImmediateFlipSupport = %d\n", support->ImmediateFlipSupport);
+ DML_LOG_VERBOSE("DML: support: ImmediateFlipSupport = %d\n", support->ImmediateFlipSupport);
if (!fail_only || support->LinkCapacitySupport == 0)
- dml2_printf("DML: support: LinkCapacitySupport = %d\n", support->LinkCapacitySupport);
+ DML_LOG_VERBOSE("DML: support: LinkCapacitySupport = %d\n", support->LinkCapacitySupport);
if (!fail_only || support->ModeSupport == 0)
- dml2_printf("DML: support: ModeSupport = %d\n", support->ModeSupport);
- dml2_printf("DML: ===================================== \n");
+ DML_LOG_VERBOSE("DML: support: ModeSupport = %d\n", support->ModeSupport);
+ DML_LOG_VERBOSE("DML: ===================================== \n");
}
static void get_stream_output_bpp(double *out_bpp, const struct dml2_display_cfg *display_cfg)
@@ -175,11 +179,9 @@ static void get_stream_output_bpp(double *out_bpp, const struct dml2_display_cfg
} else {
out_bpp[k] = 0;
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d bpc=%f\n", __func__, k, bpc);
- dml2_printf("DML::%s: k=%d dsc.enable=%d\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable);
- dml2_printf("DML::%s: k=%d out_bpp=%f\n", __func__, k, out_bpp[k]);
-#endif
+ DML_LOG_VERBOSE("DML::%s: k=%d bpc=%f\n", __func__, k, bpc);
+ DML_LOG_VERBOSE("DML::%s: k=%d dsc.enable=%d\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable);
+ DML_LOG_VERBOSE("DML::%s: k=%d out_bpp=%f\n", __func__, k, out_bpp[k]);
}
}
@@ -208,9 +210,7 @@ static unsigned int dml_get_num_active_pipes(int unsigned num_planes, const stru
num_active_pipes = num_active_pipes + (unsigned int)cfg_support_info->plane_support_info[k].dpps_used;
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: num_active_pipes = %d\n", __func__, num_active_pipes);
-#endif
+ DML_LOG_VERBOSE("DML::%s: num_active_pipes = %d\n", __func__, num_active_pipes);
return num_active_pipes;
}
@@ -247,7 +247,7 @@ static bool dml_get_is_phantom_pipe(const struct dml2_display_cfg *display_cfg,
unsigned int plane_idx = mode_lib->mp.pipe_plane[pipe_idx];
bool is_phantom = dml_is_phantom_pipe(&display_cfg->plane_descriptors[plane_idx]);
- dml2_printf("DML::%s: pipe_idx=%d legacy_svp_config=%0d is_phantom=%d\n", __func__, pipe_idx, display_cfg->plane_descriptors[plane_idx].overrides.legacy_svp_config, is_phantom);
+ DML_LOG_VERBOSE("DML::%s: pipe_idx=%d legacy_svp_config=%0d is_phantom=%d\n", __func__, pipe_idx, display_cfg->plane_descriptors[plane_idx].overrides.legacy_svp_config, is_phantom);
return is_phantom;
}
@@ -315,12 +315,11 @@ dml_get_var_func(meta_trip_memory_us, double, mode_lib->mp.MetaTripToMemory);
dml_get_var_func(wm_fclk_change, double, mode_lib->mp.Watermark.FCLKChangeWatermark);
dml_get_var_func(wm_usr_retraining, double, mode_lib->mp.Watermark.USRRetrainingWatermark);
-dml_get_var_func(wm_g6_temp_read, double, mode_lib->mp.Watermark.g6_temp_read_watermark_us);
+dml_get_var_func(wm_temp_read_or_ppt, double, mode_lib->mp.Watermark.temp_read_or_ppt_watermark_us);
dml_get_var_func(wm_dram_clock_change, double, mode_lib->mp.Watermark.DRAMClockChangeWatermark);
dml_get_var_func(fraction_of_urgent_bandwidth, double, mode_lib->mp.FractionOfUrgentBandwidth);
dml_get_var_func(fraction_of_urgent_bandwidth_imm_flip, double, mode_lib->mp.FractionOfUrgentBandwidthImmediateFlip);
dml_get_var_func(fraction_of_urgent_bandwidth_mall, double, mode_lib->mp.FractionOfUrgentBandwidthMALL);
-dml_get_var_func(urgent_latency, double, mode_lib->mp.UrgentLatency);
dml_get_var_func(wm_writeback_dram_clock_change, double, mode_lib->mp.Watermark.WritebackDRAMClockChangeWatermark);
dml_get_var_func(wm_writeback_fclk_change, double, mode_lib->mp.Watermark.WritebackFCLKChangeWatermark);
dml_get_var_func(stutter_efficiency, double, mode_lib->mp.StutterEfficiency);
@@ -355,7 +354,9 @@ dml_get_var_func(svp_prefetch_urg_bw_available_sdp, double, mode_lib->mp.urg_ban
dml_get_var_func(svp_prefetch_urg_bw_available_dram, double, mode_lib->mp.urg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram]);
dml_get_var_func(svp_prefetch_urg_bw_available_dram_vm_only, double, mode_lib->mp.urg_bandwidth_available_vm_only[dml2_core_internal_soc_state_svp_prefetch]);
+dml_get_var_func(urgent_latency, double, mode_lib->mp.UrgentLatency);
dml_get_var_func(max_urgent_latency_us, double, mode_lib->ms.support.max_urgent_latency_us);
+dml_get_var_func(max_non_urgent_latency_us, double, mode_lib->ms.support.max_non_urgent_latency_us);
dml_get_var_func(avg_non_urgent_latency_us, double, mode_lib->ms.support.avg_non_urgent_latency_us);
dml_get_var_func(avg_urgent_latency_us, double, mode_lib->ms.support.avg_urgent_latency_us);
@@ -410,19 +411,17 @@ static void CalculateMaxDETAndMinCompressedBufferSize(
*nomDETInKByte = (unsigned int)(math_floor2((double)*MaxTotalDETInKByte / (double)MaxNumDPP, ConfigReturnBufferSegmentSizeInKByte));
*MinCompressedBufferSizeInKByte = ConfigReturnBufferSizeInKByte - *MaxTotalDETInKByte;
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: is_mrq_present = %u\n", __func__, is_mrq_present);
- dml2_printf("DML::%s: ConfigReturnBufferSizeInKByte = %u\n", __func__, ConfigReturnBufferSizeInKByte);
- dml2_printf("DML::%s: ROBBufferSizeInKByte = %u\n", __func__, ROBBufferSizeInKByte);
- dml2_printf("DML::%s: MaxNumDPP = %u\n", __func__, MaxNumDPP);
- dml2_printf("DML::%s: MaxTotalDETInKByte = %u\n", __func__, *MaxTotalDETInKByte);
- dml2_printf("DML::%s: nomDETInKByte = %u\n", __func__, *nomDETInKByte);
- dml2_printf("DML::%s: MinCompressedBufferSizeInKByte = %u\n", __func__, *MinCompressedBufferSizeInKByte);
-#endif
+ DML_LOG_VERBOSE("DML::%s: is_mrq_present = %u\n", __func__, is_mrq_present);
+ DML_LOG_VERBOSE("DML::%s: ConfigReturnBufferSizeInKByte = %u\n", __func__, ConfigReturnBufferSizeInKByte);
+ DML_LOG_VERBOSE("DML::%s: ROBBufferSizeInKByte = %u\n", __func__, ROBBufferSizeInKByte);
+ DML_LOG_VERBOSE("DML::%s: MaxNumDPP = %u\n", __func__, MaxNumDPP);
+ DML_LOG_VERBOSE("DML::%s: MaxTotalDETInKByte = %u\n", __func__, *MaxTotalDETInKByte);
+ DML_LOG_VERBOSE("DML::%s: nomDETInKByte = %u\n", __func__, *nomDETInKByte);
+ DML_LOG_VERBOSE("DML::%s: MinCompressedBufferSizeInKByte = %u\n", __func__, *MinCompressedBufferSizeInKByte);
if (nomDETInKByteOverrideEnable) {
*nomDETInKByte = nomDETInKByteOverrideValue;
- dml2_printf("DML::%s: nomDETInKByte = %u (overrided)\n", __func__, *nomDETInKByte);
+ DML_LOG_VERBOSE("DML::%s: nomDETInKByte = %u (overrided)\n", __func__, *nomDETInKByte);
}
}
@@ -466,6 +465,24 @@ static bool dml_is_420(enum dml2_source_format_class source_format)
case dml2_420_12:
val = 1;
break;
+ case dml2_422_planar_8:
+ val = 0;
+ break;
+ case dml2_422_planar_10:
+ val = 0;
+ break;
+ case dml2_422_planar_12:
+ val = 0;
+ break;
+ case dml2_422_packed_8:
+ val = 0;
+ break;
+ case dml2_422_packed_10:
+ val = 0;
+ break;
+ case dml2_422_packed_12:
+ val = 0;
+ break;
case dml2_rgbe_alpha:
val = 0;
break;
@@ -479,7 +496,7 @@ static bool dml_is_420(enum dml2_source_format_class source_format)
val = 0;
break;
default:
- DML2_ASSERT(0);
+ DML_ASSERT(0);
break;
}
return val;
@@ -487,33 +504,32 @@ static bool dml_is_420(enum dml2_source_format_class source_format)
static unsigned int dml_get_tile_block_size_bytes(enum dml2_swizzle_mode sw_mode)
{
- switch (sw_mode) {
- case (dml2_sw_linear):
- return 256; break;
- case (dml2_sw_256b_2d):
- return 256; break;
- case (dml2_sw_4kb_2d):
- return 4096; break;
- case (dml2_sw_64kb_2d):
- return 65536; break;
- case (dml2_sw_256kb_2d):
- return 262144; break;
- case (dml2_gfx11_sw_linear):
- return 256; break;
- case (dml2_gfx11_sw_64kb_d):
- return 65536; break;
- case (dml2_gfx11_sw_64kb_d_t):
- return 65536; break;
- case (dml2_gfx11_sw_64kb_d_x):
- return 65536; break;
- case (dml2_gfx11_sw_64kb_r_x):
- return 65536; break;
- case (dml2_gfx11_sw_256kb_d_x):
- return 262144; break;
- case (dml2_gfx11_sw_256kb_r_x):
- return 262144; break;
- default:
- DML2_ASSERT(0);
+ if (sw_mode == dml2_sw_linear)
+ return 256;
+ else if (sw_mode == dml2_sw_256b_2d)
+ return 256;
+ else if (sw_mode == dml2_sw_4kb_2d)
+ return 4096;
+ else if (sw_mode == dml2_sw_64kb_2d)
+ return 65536;
+ else if (sw_mode == dml2_sw_256kb_2d)
+ return 262144;
+ else if (sw_mode == dml2_gfx11_sw_linear)
+ return 256;
+ else if (sw_mode == dml2_gfx11_sw_64kb_d)
+ return 65536;
+ else if (sw_mode == dml2_gfx11_sw_64kb_d_t)
+ return 65536;
+ else if (sw_mode == dml2_gfx11_sw_64kb_d_x)
+ return 65536;
+ else if (sw_mode == dml2_gfx11_sw_64kb_r_x)
+ return 65536;
+ else if (sw_mode == dml2_gfx11_sw_256kb_d_x)
+ return 262144;
+ else if (sw_mode == dml2_gfx11_sw_256kb_r_x)
+ return 262144;
+ else {
+ DML_ASSERT(0);
return 256;
}
}
@@ -548,8 +564,8 @@ static int unsigned dml_get_gfx_version(enum dml2_swizzle_mode sw_mode)
sw_mode == dml2_gfx11_sw_256kb_r_x) {
version = 11;
} else {
- dml2_printf("ERROR: Invalid sw_mode setting! val=%u\n", sw_mode);
- DML2_ASSERT(0);
+ DML_LOG_VERBOSE("ERROR: Invalid sw_mode setting! val=%u\n", sw_mode);
+ DML_ASSERT(0);
}
return version;
@@ -579,8 +595,8 @@ static void CalculateBytePerPixelAndBlockSizes(
{
*BytePerPixelDETY = 0;
*BytePerPixelDETC = 0;
- *BytePerPixelY = 0;
- *BytePerPixelC = 0;
+ *BytePerPixelY = 1;
+ *BytePerPixelC = 1;
if (SourcePixelFormat == dml2_444_64) {
*BytePerPixelDETY = 8;
@@ -623,21 +639,19 @@ static void CalculateBytePerPixelAndBlockSizes(
*BytePerPixelY = 2;
*BytePerPixelC = 4;
} else {
- dml2_printf("ERROR: DML::%s: SourcePixelFormat = %u not supported!\n", __func__, SourcePixelFormat);
- DML2_ASSERT(0);
+ DML_LOG_VERBOSE("ERROR: DML::%s: SourcePixelFormat = %u not supported!\n", __func__, SourcePixelFormat);
+ DML_ASSERT(0);
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: SourcePixelFormat = %u\n", __func__, SourcePixelFormat);
- dml2_printf("DML::%s: BytePerPixelDETY = %f\n", __func__, *BytePerPixelDETY);
- dml2_printf("DML::%s: BytePerPixelDETC = %f\n", __func__, *BytePerPixelDETC);
- dml2_printf("DML::%s: BytePerPixelY = %u\n", __func__, *BytePerPixelY);
- dml2_printf("DML::%s: BytePerPixelC = %u\n", __func__, *BytePerPixelC);
- dml2_printf("DML::%s: pitch_y = %u\n", __func__, pitch_y);
- dml2_printf("DML::%s: pitch_c = %u\n", __func__, pitch_c);
- dml2_printf("DML::%s: surf_linear128_l = %u\n", __func__, *surf_linear128_l);
- dml2_printf("DML::%s: surf_linear128_c = %u\n", __func__, *surf_linear128_c);
-#endif
+ DML_LOG_VERBOSE("DML::%s: SourcePixelFormat = %u\n", __func__, SourcePixelFormat);
+ DML_LOG_VERBOSE("DML::%s: BytePerPixelDETY = %f\n", __func__, *BytePerPixelDETY);
+ DML_LOG_VERBOSE("DML::%s: BytePerPixelDETC = %f\n", __func__, *BytePerPixelDETC);
+ DML_LOG_VERBOSE("DML::%s: BytePerPixelY = %u\n", __func__, *BytePerPixelY);
+ DML_LOG_VERBOSE("DML::%s: BytePerPixelC = %u\n", __func__, *BytePerPixelC);
+ DML_LOG_VERBOSE("DML::%s: pitch_y = %u\n", __func__, pitch_y);
+ DML_LOG_VERBOSE("DML::%s: pitch_c = %u\n", __func__, pitch_c);
+ DML_LOG_VERBOSE("DML::%s: surf_linear128_l = %u\n", __func__, *surf_linear128_l);
+ DML_LOG_VERBOSE("DML::%s: surf_linear128_c = %u\n", __func__, *surf_linear128_c);
if (dml_get_gfx_version(SurfaceTiling) == 11) {
*surf_linear128_l = 0;
@@ -681,12 +695,10 @@ static void CalculateBytePerPixelAndBlockSizes(
*BlockWidth256BytesY = 256U / *BytePerPixelY / *BlockHeight256BytesY;
*BlockWidth256BytesC = 256U / *BytePerPixelC / *BlockHeight256BytesC;
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: BlockWidth256BytesY = %u\n", __func__, *BlockWidth256BytesY);
- dml2_printf("DML::%s: BlockHeight256BytesY = %u\n", __func__, *BlockHeight256BytesY);
- dml2_printf("DML::%s: BlockWidth256BytesC = %u\n", __func__, *BlockWidth256BytesC);
- dml2_printf("DML::%s: BlockHeight256BytesC = %u\n", __func__, *BlockHeight256BytesC);
-#endif
+ DML_LOG_VERBOSE("DML::%s: BlockWidth256BytesY = %u\n", __func__, *BlockWidth256BytesY);
+ DML_LOG_VERBOSE("DML::%s: BlockHeight256BytesY = %u\n", __func__, *BlockHeight256BytesY);
+ DML_LOG_VERBOSE("DML::%s: BlockWidth256BytesC = %u\n", __func__, *BlockWidth256BytesC);
+ DML_LOG_VERBOSE("DML::%s: BlockHeight256BytesC = %u\n", __func__, *BlockHeight256BytesC);
if (dml_get_gfx_version(SurfaceTiling) == 11) {
if (SurfaceTiling == dml2_gfx11_sw_linear) {
@@ -730,8 +742,8 @@ static void CalculateBytePerPixelAndBlockSizes(
} else if (SurfaceTiling == dml2_sw_256kb_2d) {
macro_tile_scale = 32;
} else {
- dml2_printf("ERROR: Invalid SurfaceTiling setting! val=%u\n", SurfaceTiling);
- DML2_ASSERT(0);
+ DML_LOG_VERBOSE("ERROR: Invalid SurfaceTiling setting! val=%u\n", SurfaceTiling);
+ DML_ASSERT(0);
}
*MacroTileHeightY = macro_tile_scale * *BlockHeight256BytesY;
@@ -744,12 +756,10 @@ static void CalculateBytePerPixelAndBlockSizes(
}
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: MacroTileWidthY = %u\n", __func__, *MacroTileWidthY);
- dml2_printf("DML::%s: MacroTileHeightY = %u\n", __func__, *MacroTileHeightY);
- dml2_printf("DML::%s: MacroTileWidthC = %u\n", __func__, *MacroTileWidthC);
- dml2_printf("DML::%s: MacroTileHeightC = %u\n", __func__, *MacroTileHeightC);
-#endif
+ DML_LOG_VERBOSE("DML::%s: MacroTileWidthY = %u\n", __func__, *MacroTileWidthY);
+ DML_LOG_VERBOSE("DML::%s: MacroTileHeightY = %u\n", __func__, *MacroTileHeightY);
+ DML_LOG_VERBOSE("DML::%s: MacroTileWidthC = %u\n", __func__, *MacroTileWidthC);
+ DML_LOG_VERBOSE("DML::%s: MacroTileHeightC = %u\n", __func__, *MacroTileHeightC);
}
static void CalculateSinglePipeDPPCLKAndSCLThroughput(
@@ -820,7 +830,7 @@ static void CalculateSwathWidth(
// Output
unsigned int req_per_swath_ub_l[],
unsigned int req_per_swath_ub_c[],
- unsigned int SwathWidthSingleDPPY[],
+ unsigned int SwathWidthSingleDPPY[], // post-rotated plane width
unsigned int SwathWidthSingleDPPC[],
unsigned int SwathWidthY[], // per-pipe
unsigned int SwathWidthC[], // per-pipe
@@ -838,10 +848,8 @@ static void CalculateSwathWidth(
unsigned int surface_width_ub_c;
unsigned int surface_height_ub_c;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ForceSingleDPP = %u\n", __func__, ForceSingleDPP);
- dml2_printf("DML::%s: NumberOfActiveSurfaces = %u\n", __func__, NumberOfActiveSurfaces);
-#endif
+ DML_LOG_VERBOSE("DML::%s: ForceSingleDPP = %u\n", __func__, ForceSingleDPP);
+ DML_LOG_VERBOSE("DML::%s: NumberOfActiveSurfaces = %u\n", __func__, NumberOfActiveSurfaces);
for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
if (!dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle)) {
@@ -850,11 +858,9 @@ static void CalculateSwathWidth(
SwathWidthSingleDPPY[k] = (unsigned int)display_cfg->plane_descriptors[k].composition.viewport.plane0.height;
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u ViewportWidth=%u\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.width);
- dml2_printf("DML::%s: k=%u ViewportHeight=%u\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.height);
- dml2_printf("DML::%s: k=%u DPPPerSurface=%u\n", __func__, k, DPPPerSurface[k]);
-#endif
+ DML_LOG_VERBOSE("DML::%s: k=%u ViewportWidth=%lu\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.width);
+ DML_LOG_VERBOSE("DML::%s: k=%u ViewportHeight=%lu\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.height);
+ DML_LOG_VERBOSE("DML::%s: k=%u DPPPerSurface=%u\n", __func__, k, DPPPerSurface[k]);
MainSurfaceODMMode = ODMMode[k];
@@ -877,13 +883,11 @@ static void CalculateSwathWidth(
}
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u HActive=%u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active);
- dml2_printf("DML::%s: k=%u HRatio=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
- dml2_printf("DML::%s: k=%u MainSurfaceODMMode=%u\n", __func__, k, MainSurfaceODMMode);
- dml2_printf("DML::%s: k=%u SwathWidthSingleDPPY=%u\n", __func__, k, SwathWidthSingleDPPY[k]);
- dml2_printf("DML::%s: k=%u SwathWidthY=%u\n", __func__, k, SwathWidthY[k]);
-#endif
+ DML_LOG_VERBOSE("DML::%s: k=%u HActive=%lu\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active);
+ DML_LOG_VERBOSE("DML::%s: k=%u HRatio=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%u MainSurfaceODMMode=%u\n", __func__, k, MainSurfaceODMMode);
+ DML_LOG_VERBOSE("DML::%s: k=%u SwathWidthSingleDPPY=%u\n", __func__, k, SwathWidthSingleDPPY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u SwathWidthY=%u\n", __func__, k, SwathWidthY[k]);
if (dml_is_420(display_cfg->plane_descriptors[k].pixel_format)) {
SwathWidthC[k] = SwathWidthY[k] / 2;
@@ -912,22 +916,20 @@ static void CalculateSwathWidth(
surface_width_ub_c = (unsigned int)math_ceil2((double)display_cfg->plane_descriptors[k].surface.plane1.width, req_width_horz_c);
surface_height_ub_c = (unsigned int)math_ceil2((double)display_cfg->plane_descriptors[k].surface.plane1.height, Read256BytesBlockHeightC[k]);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u surface_width_ub_l=%u\n", __func__, k, surface_width_ub_l);
- dml2_printf("DML::%s: k=%u surface_height_ub_l=%u\n", __func__, k, surface_height_ub_l);
- dml2_printf("DML::%s: k=%u surface_width_ub_c=%u\n", __func__, k, surface_width_ub_c);
- dml2_printf("DML::%s: k=%u surface_height_ub_c=%u\n", __func__, k, surface_height_ub_c);
- dml2_printf("DML::%s: k=%u req_width_horz_y=%u\n", __func__, k, req_width_horz_y);
- dml2_printf("DML::%s: k=%u req_width_horz_c=%u\n", __func__, k, req_width_horz_c);
- dml2_printf("DML::%s: k=%u Read256BytesBlockWidthY=%u\n", __func__, k, Read256BytesBlockWidthY[k]);
- dml2_printf("DML::%s: k=%u Read256BytesBlockHeightY=%u\n", __func__, k, Read256BytesBlockHeightY[k]);
- dml2_printf("DML::%s: k=%u Read256BytesBlockWidthC=%u\n", __func__, k, Read256BytesBlockWidthC[k]);
- dml2_printf("DML::%s: k=%u Read256BytesBlockHeightC=%u\n", __func__, k, Read256BytesBlockHeightC[k]);
- dml2_printf("DML::%s: k=%u req_width_horz_y=%u\n", __func__, k, req_width_horz_y);
- dml2_printf("DML::%s: k=%u req_width_horz_c=%u\n", __func__, k, req_width_horz_c);
- dml2_printf("DML::%s: k=%u ViewportStationary=%u\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.stationary);
- dml2_printf("DML::%s: k=%u DPPPerSurface=%u\n", __func__, k, DPPPerSurface[k]);
-#endif
+ DML_LOG_VERBOSE("DML::%s: k=%u surface_width_ub_l=%u\n", __func__, k, surface_width_ub_l);
+ DML_LOG_VERBOSE("DML::%s: k=%u surface_height_ub_l=%u\n", __func__, k, surface_height_ub_l);
+ DML_LOG_VERBOSE("DML::%s: k=%u surface_width_ub_c=%u\n", __func__, k, surface_width_ub_c);
+ DML_LOG_VERBOSE("DML::%s: k=%u surface_height_ub_c=%u\n", __func__, k, surface_height_ub_c);
+ DML_LOG_VERBOSE("DML::%s: k=%u req_width_horz_y=%u\n", __func__, k, req_width_horz_y);
+ DML_LOG_VERBOSE("DML::%s: k=%u req_width_horz_c=%u\n", __func__, k, req_width_horz_c);
+ DML_LOG_VERBOSE("DML::%s: k=%u Read256BytesBlockWidthY=%u\n", __func__, k, Read256BytesBlockWidthY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u Read256BytesBlockHeightY=%u\n", __func__, k, Read256BytesBlockHeightY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u Read256BytesBlockWidthC=%u\n", __func__, k, Read256BytesBlockWidthC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u Read256BytesBlockHeightC=%u\n", __func__, k, Read256BytesBlockHeightC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u req_width_horz_y=%u\n", __func__, k, req_width_horz_y);
+ DML_LOG_VERBOSE("DML::%s: k=%u req_width_horz_c=%u\n", __func__, k, req_width_horz_c);
+ DML_LOG_VERBOSE("DML::%s: k=%u ViewportStationary=%u\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.stationary);
+ DML_LOG_VERBOSE("DML::%s: k=%u DPPPerSurface=%u\n", __func__, k, DPPPerSurface[k]);
req_per_swath_ub_l[k] = 0;
req_per_swath_ub_c[k] = 0;
@@ -973,15 +975,12 @@ static void CalculateSwathWidth(
}
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u swath_width_luma_ub=%u\n", __func__, k, swath_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u swath_width_chroma_ub=%u\n", __func__, k, swath_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathHeightY=%u\n", __func__, k, MaximumSwathHeightY[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathHeightC=%u\n", __func__, k, MaximumSwathHeightC[k]);
- dml2_printf("DML::%s: k=%u req_per_swath_ub_l=%u\n", __func__, k, req_per_swath_ub_l[k]);
- dml2_printf("DML::%s: k=%u req_per_swath_ub_c=%u\n", __func__, k, req_per_swath_ub_c[k]);
-#endif
-
+ DML_LOG_VERBOSE("DML::%s: k=%u swath_width_luma_ub=%u\n", __func__, k, swath_width_luma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u swath_width_chroma_ub=%u\n", __func__, k, swath_width_chroma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathHeightY=%u\n", __func__, k, MaximumSwathHeightY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathHeightC=%u\n", __func__, k, MaximumSwathHeightC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u req_per_swath_ub_l=%u\n", __func__, k, req_per_swath_ub_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u req_per_swath_ub_c=%u\n", __func__, k, req_per_swath_ub_c[k]);
}
}
@@ -996,13 +995,11 @@ static bool UnboundedRequest(bool unb_req_force_en, bool unb_req_force_val, unsi
if (unb_req_force_en) {
unb_req_en = unb_req_force_val && unb_req_ok;
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: unb_req_force_en = %u\n", __func__, unb_req_force_en);
- dml2_printf("DML::%s: unb_req_force_val = %u\n", __func__, unb_req_force_val);
- dml2_printf("DML::%s: unb_req_ok = %u\n", __func__, unb_req_ok);
- dml2_printf("DML::%s: unb_req_en = %u\n", __func__, unb_req_en);
-#endif
- return (unb_req_en);
+ DML_LOG_VERBOSE("DML::%s: unb_req_force_en = %u\n", __func__, unb_req_force_en);
+ DML_LOG_VERBOSE("DML::%s: unb_req_force_val = %u\n", __func__, unb_req_force_val);
+ DML_LOG_VERBOSE("DML::%s: unb_req_ok = %u\n", __func__, unb_req_ok);
+ DML_LOG_VERBOSE("DML::%s: unb_req_en = %u\n", __func__, unb_req_en);
+ return unb_req_en;
}
static void CalculateDETBufferSize(
@@ -1032,16 +1029,14 @@ static void CalculateDETBufferSize(
bool NextPotentialSurfaceToAssignDETPieceFound;
bool MinimizeReallocationSuccess = false;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ForceSingleDPP = %u\n", __func__, ForceSingleDPP);
- dml2_printf("DML::%s: nomDETInKByte = %u\n", __func__, nomDETInKByte);
- dml2_printf("DML::%s: NumberOfActiveSurfaces = %u\n", __func__, NumberOfActiveSurfaces);
- dml2_printf("DML::%s: UnboundedRequestEnabled = %u\n", __func__, UnboundedRequestEnabled);
- dml2_printf("DML::%s: MaxTotalDETInKByte = %u\n", __func__, MaxTotalDETInKByte);
- dml2_printf("DML::%s: ConfigReturnBufferSizeInKByte = %u\n", __func__, ConfigReturnBufferSizeInKByte);
- dml2_printf("DML::%s: MinCompressedBufferSizeInKByte = %u\n", __func__, MinCompressedBufferSizeInKByte);
- dml2_printf("DML::%s: CompressedBufferSegmentSizeInkByte = %u\n", __func__, CompressedBufferSegmentSizeInkByte);
-#endif
+ DML_LOG_VERBOSE("DML::%s: ForceSingleDPP = %u\n", __func__, ForceSingleDPP);
+ DML_LOG_VERBOSE("DML::%s: nomDETInKByte = %u\n", __func__, nomDETInKByte);
+ DML_LOG_VERBOSE("DML::%s: NumberOfActiveSurfaces = %u\n", __func__, NumberOfActiveSurfaces);
+ DML_LOG_VERBOSE("DML::%s: UnboundedRequestEnabled = %u\n", __func__, UnboundedRequestEnabled);
+ DML_LOG_VERBOSE("DML::%s: MaxTotalDETInKByte = %u\n", __func__, MaxTotalDETInKByte);
+ DML_LOG_VERBOSE("DML::%s: ConfigReturnBufferSizeInKByte = %u\n", __func__, ConfigReturnBufferSizeInKByte);
+ DML_LOG_VERBOSE("DML::%s: MinCompressedBufferSizeInKByte = %u\n", __func__, MinCompressedBufferSizeInKByte);
+ DML_LOG_VERBOSE("DML::%s: CompressedBufferSegmentSizeInkByte = %u\n", __func__, CompressedBufferSegmentSizeInkByte);
// Note: Will use default det size if that fits 2 swaths
if (UnboundedRequestEnabled) {
@@ -1070,19 +1065,15 @@ static void CalculateDETBufferSize(
l->minDET = l->minDET + ConfigReturnBufferSegmentSizeInkByte;
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u minDET = %u\n", __func__, k, l->minDET);
- dml2_printf("DML::%s: k=%u max_minDET = %u\n", __func__, k, l->max_minDET);
- dml2_printf("DML::%s: k=%u minDET_pipe = %u\n", __func__, k, l->minDET_pipe);
- dml2_printf("DML::%s: k=%u full_swath_bytes_l = %u\n", __func__, k, full_swath_bytes_l[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_c = %u\n", __func__, k, full_swath_bytes_c[k]);
-#endif
+ DML_LOG_VERBOSE("DML::%s: k=%u minDET = %u\n", __func__, k, l->minDET);
+ DML_LOG_VERBOSE("DML::%s: k=%u max_minDET = %u\n", __func__, k, l->max_minDET);
+ DML_LOG_VERBOSE("DML::%s: k=%u minDET_pipe = %u\n", __func__, k, l->minDET_pipe);
+ DML_LOG_VERBOSE("DML::%s: k=%u full_swath_bytes_l = %u\n", __func__, k, full_swath_bytes_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u full_swath_bytes_c = %u\n", __func__, k, full_swath_bytes_c[k]);
if (l->minDET_pipe == 0) {
l->minDET_pipe = (unsigned int)(math_max2(128, math_ceil2(((double)full_swath_bytes_l[k] + (double)full_swath_bytes_c[k]) / 1024.0, ConfigReturnBufferSegmentSizeInkByte)));
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u minDET_pipe = %u (assume each plane take half DET)\n", __func__, k, l->minDET_pipe);
-#endif
+ DML_LOG_VERBOSE("DML::%s: k=%u minDET_pipe = %u (assume each plane take half DET)\n", __func__, k, l->minDET_pipe);
}
if (dml_is_phantom_pipe(&display_cfg->plane_descriptors[k])) {
@@ -1095,12 +1086,10 @@ static void CalculateDETBufferSize(
l->DETBufferSizePoolInKByte = l->DETBufferSizePoolInKByte - (ForceSingleDPP ? 1 : DPPPerSurface[k]) * l->minDET_pipe;
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u DPPPerSurface = %u\n", __func__, k, DPPPerSurface[k]);
- dml2_printf("DML::%s: k=%u DETSizeOverride = %u\n", __func__, k, display_cfg->plane_descriptors[k].overrides.det_size_override_kb);
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByte = %u\n", __func__, k, DETBufferSizeInKByte[k]);
- dml2_printf("DML::%s: DETBufferSizePoolInKByte = %u\n", __func__, l->DETBufferSizePoolInKByte);
-#endif
+ DML_LOG_VERBOSE("DML::%s: k=%u DPPPerSurface = %u\n", __func__, k, DPPPerSurface[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u DETSizeOverride = %u\n", __func__, k, display_cfg->plane_descriptors[k].overrides.det_size_override_kb);
+ DML_LOG_VERBOSE("DML::%s: k=%u DETBufferSizeInKByte = %u\n", __func__, k, DETBufferSizeInKByte[k]);
+ DML_LOG_VERBOSE("DML::%s: DETBufferSizePoolInKByte = %u\n", __func__, l->DETBufferSizePoolInKByte);
}
if (display_cfg->minimize_det_reallocation) {
@@ -1172,14 +1161,12 @@ static void CalculateDETBufferSize(
l->TotalBandwidth = l->TotalBandwidth + ReadBandwidthLuma[k] + ReadBandwidthChroma[k];
}
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: --- Before bandwidth adjustment ---\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: --- Before bandwidth adjustment ---\n", __func__);
for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByte = %u\n", __func__, k, DETBufferSizeInKByte[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u DETBufferSizeInKByte = %u\n", __func__, k, DETBufferSizeInKByte[k]);
}
- dml2_printf("DML::%s: --- DET allocation with bandwidth ---\n", __func__);
-#endif
- dml2_printf("DML::%s: TotalBandwidth = %f\n", __func__, l->TotalBandwidth);
+ DML_LOG_VERBOSE("DML::%s: --- DET allocation with bandwidth ---\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: TotalBandwidth = %f\n", __func__, l->TotalBandwidth);
l->BandwidthOfSurfacesNotAssignedDETPiece = l->TotalBandwidth;
for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
@@ -1191,10 +1178,8 @@ static void CalculateDETBufferSize(
} else {
DETPieceAssignedToThisSurfaceAlready[k] = false;
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u DETPieceAssignedToThisSurfaceAlready = %u\n", __func__, k, DETPieceAssignedToThisSurfaceAlready[k]);
- dml2_printf("DML::%s: k=%u BandwidthOfSurfacesNotAssignedDETPiece = %f\n", __func__, k, l->BandwidthOfSurfacesNotAssignedDETPiece);
-#endif
+ DML_LOG_VERBOSE("DML::%s: k=%u DETPieceAssignedToThisSurfaceAlready = %u\n", __func__, k, DETPieceAssignedToThisSurfaceAlready[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u BandwidthOfSurfacesNotAssignedDETPiece = %f\n", __func__, k, l->BandwidthOfSurfacesNotAssignedDETPiece);
}
for (unsigned int j = 0; j < NumberOfActiveSurfaces; ++j) {
@@ -1202,22 +1187,18 @@ static void CalculateDETBufferSize(
l->NextSurfaceToAssignDETPiece = 0;
for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: j=%u k=%u, ReadBandwidthLuma[k] = %f\n", __func__, j, k, ReadBandwidthLuma[k]);
- dml2_printf("DML::%s: j=%u k=%u, ReadBandwidthChroma[k] = %f\n", __func__, j, k, ReadBandwidthChroma[k]);
- dml2_printf("DML::%s: j=%u k=%u, ReadBandwidthLuma[Next] = %f\n", __func__, j, k, ReadBandwidthLuma[l->NextSurfaceToAssignDETPiece]);
- dml2_printf("DML::%s: j=%u k=%u, ReadBandwidthChroma[Next] = %f\n", __func__, j, k, ReadBandwidthChroma[l->NextSurfaceToAssignDETPiece]);
- dml2_printf("DML::%s: j=%u k=%u, NextSurfaceToAssignDETPiece = %u\n", __func__, j, k, l->NextSurfaceToAssignDETPiece);
-#endif
+ DML_LOG_VERBOSE("DML::%s: j=%u k=%u, ReadBandwidthLuma[k] = %f\n", __func__, j, k, ReadBandwidthLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: j=%u k=%u, ReadBandwidthChroma[k] = %f\n", __func__, j, k, ReadBandwidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: j=%u k=%u, ReadBandwidthLuma[Next] = %f\n", __func__, j, k, ReadBandwidthLuma[l->NextSurfaceToAssignDETPiece]);
+ DML_LOG_VERBOSE("DML::%s: j=%u k=%u, ReadBandwidthChroma[Next] = %f\n", __func__, j, k, ReadBandwidthChroma[l->NextSurfaceToAssignDETPiece]);
+ DML_LOG_VERBOSE("DML::%s: j=%u k=%u, NextSurfaceToAssignDETPiece = %u\n", __func__, j, k, l->NextSurfaceToAssignDETPiece);
if (!DETPieceAssignedToThisSurfaceAlready[k] && (!NextPotentialSurfaceToAssignDETPieceFound ||
ReadBandwidthLuma[k] + ReadBandwidthChroma[k] < ReadBandwidthLuma[l->NextSurfaceToAssignDETPiece] + ReadBandwidthChroma[l->NextSurfaceToAssignDETPiece])) {
l->NextSurfaceToAssignDETPiece = k;
NextPotentialSurfaceToAssignDETPieceFound = true;
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: j=%u k=%u, DETPieceAssignedToThisSurfaceAlready = %u\n", __func__, j, k, DETPieceAssignedToThisSurfaceAlready[k]);
- dml2_printf("DML::%s: j=%u k=%u, NextPotentialSurfaceToAssignDETPieceFound = %u\n", __func__, j, k, NextPotentialSurfaceToAssignDETPieceFound);
-#endif
+ DML_LOG_VERBOSE("DML::%s: j=%u k=%u, DETPieceAssignedToThisSurfaceAlready = %u\n", __func__, j, k, DETPieceAssignedToThisSurfaceAlready[k]);
+ DML_LOG_VERBOSE("DML::%s: j=%u k=%u, NextPotentialSurfaceToAssignDETPieceFound = %u\n", __func__, j, k, NextPotentialSurfaceToAssignDETPieceFound);
}
if (NextPotentialSurfaceToAssignDETPieceFound) {
@@ -1227,20 +1208,16 @@ static void CalculateDETBufferSize(
* (ForceSingleDPP ? 1 : DPPPerSurface[l->NextSurfaceToAssignDETPiece]) * ConfigReturnBufferSegmentSizeInkByte,
math_floor2((double)l->DETBufferSizePoolInKByte, (ForceSingleDPP ? 1 : DPPPerSurface[l->NextSurfaceToAssignDETPiece]) * ConfigReturnBufferSegmentSizeInkByte)));
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: j=%u, DETBufferSizePoolInKByte = %u\n", __func__, j, l->DETBufferSizePoolInKByte);
- dml2_printf("DML::%s: j=%u, NextSurfaceToAssignDETPiece = %u\n", __func__, j, l->NextSurfaceToAssignDETPiece);
- dml2_printf("DML::%s: j=%u, ReadBandwidthLuma[%u] = %f\n", __func__, j, l->NextSurfaceToAssignDETPiece, ReadBandwidthLuma[l->NextSurfaceToAssignDETPiece]);
- dml2_printf("DML::%s: j=%u, ReadBandwidthChroma[%u] = %f\n", __func__, j, l->NextSurfaceToAssignDETPiece, ReadBandwidthChroma[l->NextSurfaceToAssignDETPiece]);
- dml2_printf("DML::%s: j=%u, BandwidthOfSurfacesNotAssignedDETPiece = %f\n", __func__, j, l->BandwidthOfSurfacesNotAssignedDETPiece);
- dml2_printf("DML::%s: j=%u, NextDETBufferPieceInKByte = %u\n", __func__, j, l->NextDETBufferPieceInKByte);
- dml2_printf("DML::%s: j=%u, DETBufferSizeInKByte[%u] increases from %u ", __func__, j, l->NextSurfaceToAssignDETPiece, DETBufferSizeInKByte[l->NextSurfaceToAssignDETPiece]);
-#endif
+ DML_LOG_VERBOSE("DML::%s: j=%u, DETBufferSizePoolInKByte = %u\n", __func__, j, l->DETBufferSizePoolInKByte);
+ DML_LOG_VERBOSE("DML::%s: j=%u, NextSurfaceToAssignDETPiece = %u\n", __func__, j, l->NextSurfaceToAssignDETPiece);
+ DML_LOG_VERBOSE("DML::%s: j=%u, ReadBandwidthLuma[%u] = %f\n", __func__, j, l->NextSurfaceToAssignDETPiece, ReadBandwidthLuma[l->NextSurfaceToAssignDETPiece]);
+ DML_LOG_VERBOSE("DML::%s: j=%u, ReadBandwidthChroma[%u] = %f\n", __func__, j, l->NextSurfaceToAssignDETPiece, ReadBandwidthChroma[l->NextSurfaceToAssignDETPiece]);
+ DML_LOG_VERBOSE("DML::%s: j=%u, BandwidthOfSurfacesNotAssignedDETPiece = %f\n", __func__, j, l->BandwidthOfSurfacesNotAssignedDETPiece);
+ DML_LOG_VERBOSE("DML::%s: j=%u, NextDETBufferPieceInKByte = %u\n", __func__, j, l->NextDETBufferPieceInKByte);
+ DML_LOG_VERBOSE("DML::%s: j=%u, DETBufferSizeInKByte[%u] increases from %u ", __func__, j, l->NextSurfaceToAssignDETPiece, DETBufferSizeInKByte[l->NextSurfaceToAssignDETPiece]);
DETBufferSizeInKByte[l->NextSurfaceToAssignDETPiece] = DETBufferSizeInKByte[l->NextSurfaceToAssignDETPiece] + l->NextDETBufferPieceInKByte / (ForceSingleDPP ? 1 : DPPPerSurface[l->NextSurfaceToAssignDETPiece]);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("to %u\n", DETBufferSizeInKByte[l->NextSurfaceToAssignDETPiece]);
-#endif
+ DML_LOG_VERBOSE("to %u\n", DETBufferSizeInKByte[l->NextSurfaceToAssignDETPiece]);
l->DETBufferSizePoolInKByte = l->DETBufferSizePoolInKByte - l->NextDETBufferPieceInKByte;
DETPieceAssignedToThisSurfaceAlready[l->NextSurfaceToAssignDETPiece] = true;
@@ -1252,13 +1229,11 @@ static void CalculateDETBufferSize(
}
*CompressedBufferSizeInkByte = *CompressedBufferSizeInkByte * CompressedBufferSegmentSizeInkByte / ConfigReturnBufferSegmentSizeInkByte;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: --- After bandwidth adjustment ---\n", __func__);
- dml2_printf("DML::%s: CompressedBufferSizeInkByte = %u\n", __func__, *CompressedBufferSizeInkByte);
+ DML_LOG_VERBOSE("DML::%s: --- After bandwidth adjustment ---\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: CompressedBufferSizeInkByte = %u\n", __func__, *CompressedBufferSizeInkByte);
for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByte = %u (TotalReadBandWidth=%f)\n", __func__, k, DETBufferSizeInKByte[k], ReadBandwidthLuma[k] + ReadBandwidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u DETBufferSizeInKByte = %u (TotalReadBandWidth=%f)\n", __func__, k, DETBufferSizeInKByte[k], ReadBandwidthLuma[k] + ReadBandwidthChroma[k]);
}
-#endif
}
static double CalculateRequiredDispclk(
@@ -1403,7 +1378,6 @@ static unsigned int dscceComputeDelay(
// N422/N420 operate at 2 pixels per clock
unsigned int pixelsPerClock, padding_pixels, ssm_group_priming_delay, ssm_pipeline_delay, obsm_pipeline_delay, slice_padded_pixels, ixd_plus_padding, ixd_plus_padding_groups, cycles_per_group, group_delay, pipeline_delay, pixels, additional_group_delay, lines_to_reach_ixd, groups_to_reach_ixd, slice_width_groups, initial_xmit_delay, number_of_lines_to_reach_ixd, slice_width_modified;
-
if (pixelFormat == dml2_420)
pixelsPerClock = 2;
// #all other modes operate at 1 pixel per clock
@@ -1428,7 +1402,6 @@ static unsigned int dscceComputeDelay(
}
}
-
//sub-stream multiplexer balance fifo priming delay in groups as per dsc standard
if (bpc == 8)
ssm_group_priming_delay = 83;
@@ -1447,9 +1420,6 @@ static unsigned int dscceComputeDelay(
//determine number of padded pixels in the last group of a slice line, computed as
slice_padded_pixels = 3 * slice_width_groups - slice_width_modified;
-
-
-
//determine integer number of complete slice lines required to reach initial transmit delay without ssm delay considered
number_of_lines_to_reach_ixd = initial_xmit_delay / slice_width_modified;
@@ -1463,7 +1433,6 @@ static unsigned int dscceComputeDelay(
//number of groups required for a slice to reach initial transmit delay is the sum of the padded initial transmit delay plus the ssm group priming delay
groups_to_reach_ixd = ixd_plus_padding_groups + ssm_group_priming_delay;
-
//number of lines required to reach padded initial transmit delay in groups in slices to the left of the last horizontal slice
//needs to be rounded up as a complete slice lines are buffered prior to initial transmit delay being reached in the last horizontal slice
lines_to_reach_ixd = (groups_to_reach_ixd + slice_width_groups - 1) / slice_width_groups; //round up lines to reach ixd to next
@@ -1494,19 +1463,16 @@ static unsigned int dscceComputeDelay(
//pixel delay is group_delay (converted to pixels) + pipeline, however, first group is a special case since it is processed as soon as it arrives (i.e., in 3 cycles regardless of pixel format)
pixels = (group_delay - 1) * cycles_per_group + 3 + pipeline_delay;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: bpc: %u\n", __func__, bpc);
- dml2_printf("DML::%s: BPP: %f\n", __func__, BPP);
- dml2_printf("DML::%s: sliceWidth: %u\n", __func__, sliceWidth);
- dml2_printf("DML::%s: numSlices: %u\n", __func__, numSlices);
- dml2_printf("DML::%s: pixelFormat: %u\n", __func__, pixelFormat);
- dml2_printf("DML::%s: Output: %u\n", __func__, Output);
- dml2_printf("DML::%s: pixels: %u\n", __func__, pixels);
-#endif
+ DML_LOG_VERBOSE("DML::%s: bpc: %u\n", __func__, bpc);
+ DML_LOG_VERBOSE("DML::%s: BPP: %f\n", __func__, BPP);
+ DML_LOG_VERBOSE("DML::%s: sliceWidth: %u\n", __func__, sliceWidth);
+ DML_LOG_VERBOSE("DML::%s: numSlices: %u\n", __func__, numSlices);
+ DML_LOG_VERBOSE("DML::%s: pixelFormat: %u\n", __func__, pixelFormat);
+ DML_LOG_VERBOSE("DML::%s: Output: %u\n", __func__, Output);
+ DML_LOG_VERBOSE("DML::%s: pixels: %u\n", __func__, pixels);
return pixels;
}
-
//updated in dcn4
static unsigned int dscComputeDelay(enum dml2_output_format_class pixelFormat, enum dml2_output_encoder_class Output)
{
@@ -1578,10 +1544,8 @@ static unsigned int dscComputeDelay(enum dml2_output_format_class pixelFormat, e
// sft
Delay = Delay + 1;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: pixelFormat = %u\n", __func__, pixelFormat);
- dml2_printf("DML::%s: Delay = %u\n", __func__, Delay);
-#endif
+ DML_LOG_VERBOSE("DML::%s: pixelFormat = %u\n", __func__, pixelFormat);
+ DML_LOG_VERBOSE("DML::%s: Delay = %u\n", __func__, Delay);
return Delay;
}
@@ -1652,10 +1616,8 @@ static unsigned int CalculateVMAndRowBytes(struct dml2_core_shared_calculate_vm_
}
meta_surface_bytes = (unsigned int)(p->DCCMetaPitch * vp_height_meta_ub * p->BytePerPixel / 256.0);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DCCMetaPitch = %u\n", __func__, p->DCCMetaPitch);
- dml2_printf("DML::%s: meta_surface_bytes = %u\n", __func__, meta_surface_bytes);
-#endif
+ DML_LOG_VERBOSE("DML::%s: DCCMetaPitch = %u\n", __func__, p->DCCMetaPitch);
+ DML_LOG_VERBOSE("DML::%s: meta_surface_bytes = %u\n", __func__, meta_surface_bytes);
if (p->GPUVMEnable == true) {
double meta_vmpg_bytes = 4.0 * 1024.0;
*p->meta_pte_bytes_per_frame_ub = (unsigned int)((math_ceil2((double) (meta_surface_bytes - meta_vmpg_bytes) / (8 * meta_vmpg_bytes), 1) + 1) * 64);
@@ -1709,25 +1671,23 @@ static unsigned int CalculateVMAndRowBytes(struct dml2_core_shared_calculate_vm_
vm_bytes = *p->meta_pte_bytes_per_frame_ub + extra_mpde_bytes + *p->dpde0_bytes_per_frame_ub + extra_dpde_bytes;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DCCEnable = %u\n", __func__, p->DCCEnable);
- dml2_printf("DML::%s: GPUVMEnable = %u\n", __func__, p->GPUVMEnable);
- dml2_printf("DML::%s: SwModeLinear = %u\n", __func__, p->SurfaceTiling == dml2_sw_linear);
- dml2_printf("DML::%s: BytePerPixel = %u\n", __func__, p->BytePerPixel);
- dml2_printf("DML::%s: GPUVMMaxPageTableLevels = %u\n", __func__, p->GPUVMMaxPageTableLevels);
- dml2_printf("DML::%s: BlockHeight256Bytes = %u\n", __func__, p->BlockHeight256Bytes);
- dml2_printf("DML::%s: BlockWidth256Bytes = %u\n", __func__, p->BlockWidth256Bytes);
- dml2_printf("DML::%s: MacroTileHeight = %u\n", __func__, p->MacroTileHeight);
- dml2_printf("DML::%s: MacroTileWidth = %u\n", __func__, p->MacroTileWidth);
- dml2_printf("DML::%s: meta_pte_bytes_per_frame_ub = %u\n", __func__, *p->meta_pte_bytes_per_frame_ub);
- dml2_printf("DML::%s: dpde0_bytes_per_frame_ub = %u\n", __func__, *p->dpde0_bytes_per_frame_ub);
- dml2_printf("DML::%s: extra_mpde_bytes = %u\n", __func__, extra_mpde_bytes);
- dml2_printf("DML::%s: extra_dpde_bytes = %u\n", __func__, extra_dpde_bytes);
- dml2_printf("DML::%s: vm_bytes = %u\n", __func__, vm_bytes);
- dml2_printf("DML::%s: ViewportHeight = %u\n", __func__, p->ViewportHeight);
- dml2_printf("DML::%s: SwathWidth = %u\n", __func__, p->SwathWidth);
- dml2_printf("DML::%s: vp_height_dpte_ub = %u\n", __func__, vp_height_dpte_ub);
-#endif
+ DML_LOG_VERBOSE("DML::%s: DCCEnable = %u\n", __func__, p->DCCEnable);
+ DML_LOG_VERBOSE("DML::%s: GPUVMEnable = %u\n", __func__, p->GPUVMEnable);
+ DML_LOG_VERBOSE("DML::%s: SwModeLinear = %u\n", __func__, p->SurfaceTiling == dml2_sw_linear);
+ DML_LOG_VERBOSE("DML::%s: BytePerPixel = %u\n", __func__, p->BytePerPixel);
+ DML_LOG_VERBOSE("DML::%s: GPUVMMaxPageTableLevels = %u\n", __func__, p->GPUVMMaxPageTableLevels);
+ DML_LOG_VERBOSE("DML::%s: BlockHeight256Bytes = %u\n", __func__, p->BlockHeight256Bytes);
+ DML_LOG_VERBOSE("DML::%s: BlockWidth256Bytes = %u\n", __func__, p->BlockWidth256Bytes);
+ DML_LOG_VERBOSE("DML::%s: MacroTileHeight = %u\n", __func__, p->MacroTileHeight);
+ DML_LOG_VERBOSE("DML::%s: MacroTileWidth = %u\n", __func__, p->MacroTileWidth);
+ DML_LOG_VERBOSE("DML::%s: meta_pte_bytes_per_frame_ub = %u\n", __func__, *p->meta_pte_bytes_per_frame_ub);
+ DML_LOG_VERBOSE("DML::%s: dpde0_bytes_per_frame_ub = %u\n", __func__, *p->dpde0_bytes_per_frame_ub);
+ DML_LOG_VERBOSE("DML::%s: extra_mpde_bytes = %u\n", __func__, extra_mpde_bytes);
+ DML_LOG_VERBOSE("DML::%s: extra_dpde_bytes = %u\n", __func__, extra_dpde_bytes);
+ DML_LOG_VERBOSE("DML::%s: vm_bytes = %u\n", __func__, vm_bytes);
+ DML_LOG_VERBOSE("DML::%s: ViewportHeight = %u\n", __func__, p->ViewportHeight);
+ DML_LOG_VERBOSE("DML::%s: SwathWidth = %u\n", __func__, p->SwathWidth);
+ DML_LOG_VERBOSE("DML::%s: vp_height_dpte_ub = %u\n", __func__, vp_height_dpte_ub);
if (p->SurfaceTiling == dml2_sw_linear) {
*p->PixelPTEReqHeight = 1;
@@ -1763,22 +1723,20 @@ static unsigned int CalculateVMAndRowBytes(struct dml2_core_shared_calculate_vm_
*p->vmpg_width = 1024 * p->GPUVMMinPageSizeKBytes / (p->MacroTileHeight * p->BytePerPixel);
if (p->GPUVMEnable == true) {
- dml2_printf("DML::%s: GPUVMMinPageSizeKBytes=%u and sw_mode=%u (tile_size=%d) not supported!\n",
+ DML_LOG_VERBOSE("DML::%s: GPUVMMinPageSizeKBytes=%u and sw_mode=%u (tile_size=%d) not supported!\n",
__func__, p->GPUVMMinPageSizeKBytes, p->SurfaceTiling, dml_get_tile_block_size_bytes(p->SurfaceTiling));
- DML2_ASSERT(0);
+ DML_ASSERT(0);
}
}
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: GPUVMMinPageSizeKBytes = %u\n", __func__, p->GPUVMMinPageSizeKBytes);
- dml2_printf("DML::%s: PixelPTEReqHeight = %u\n", __func__, *p->PixelPTEReqHeight);
- dml2_printf("DML::%s: PixelPTEReqWidth = %u\n", __func__, *p->PixelPTEReqWidth);
- dml2_printf("DML::%s: PixelPTEReqWidth_linear = %u\n", __func__, PixelPTEReqWidth_linear);
- dml2_printf("DML::%s: PTERequestSize = %u\n", __func__, *p->PTERequestSize);
- dml2_printf("DML::%s: Pitch = %u\n", __func__, p->Pitch);
- dml2_printf("DML::%s: vmpg_width = %u\n", __func__, *p->vmpg_width);
- dml2_printf("DML::%s: vmpg_height = %u\n", __func__, *p->vmpg_height);
-#endif
+ DML_LOG_VERBOSE("DML::%s: GPUVMMinPageSizeKBytes = %u\n", __func__, p->GPUVMMinPageSizeKBytes);
+ DML_LOG_VERBOSE("DML::%s: PixelPTEReqHeight = %u\n", __func__, *p->PixelPTEReqHeight);
+ DML_LOG_VERBOSE("DML::%s: PixelPTEReqWidth = %u\n", __func__, *p->PixelPTEReqWidth);
+ DML_LOG_VERBOSE("DML::%s: PixelPTEReqWidth_linear = %u\n", __func__, PixelPTEReqWidth_linear);
+ DML_LOG_VERBOSE("DML::%s: PTERequestSize = %u\n", __func__, *p->PTERequestSize);
+ DML_LOG_VERBOSE("DML::%s: Pitch = %u\n", __func__, p->Pitch);
+ DML_LOG_VERBOSE("DML::%s: vmpg_width = %u\n", __func__, *p->vmpg_width);
+ DML_LOG_VERBOSE("DML::%s: vmpg_height = %u\n", __func__, *p->vmpg_height);
*p->dpte_row_height_one_row_per_frame = vp_height_dpte_ub;
*p->dpte_row_width_ub_one_row_per_frame = (unsigned int)((math_ceil2(((double)p->Pitch * (double)*p->dpte_row_height_one_row_per_frame / (double)*p->PixelPTEReqHeight - 1) / (double)*p->PixelPTEReqWidth, 1) + 1) * (double)*p->PixelPTEReqWidth);
@@ -1796,7 +1754,7 @@ static unsigned int CalculateVMAndRowBytes(struct dml2_core_shared_calculate_vm_
*p->dpte_row_height_linear = 128;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: dpte_row_width_ub = %u (linear)\n", __func__, *p->dpte_row_width_ub);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_width_ub = %u (linear)\n", __func__, *p->dpte_row_width_ub);
#endif
} else if (!dml_is_vertical_rotation(p->RotationAngle)) {
@@ -1810,7 +1768,7 @@ static unsigned int CalculateVMAndRowBytes(struct dml2_core_shared_calculate_vm_
*p->dpte_row_width_ub = (unsigned int)((math_ceil2((double)(p->SwathWidth - 1) / (double)*p->PixelPTEReqWidth, 1) + 1.0) * *p->PixelPTEReqWidth);
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: dpte_row_width_ub = %u (tiled horz)\n", __func__, *p->dpte_row_width_ub);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_width_ub = %u (tiled horz)\n", __func__, *p->dpte_row_width_ub);
#endif
*p->PixelPTEBytesPerRow = *p->dpte_row_width_ub / *p->PixelPTEReqWidth * *p->PTERequestSize;
@@ -1825,7 +1783,7 @@ static unsigned int CalculateVMAndRowBytes(struct dml2_core_shared_calculate_vm_
*p->PixelPTEBytesPerRow = (unsigned int)((double)*p->dpte_row_width_ub / (double)*p->PixelPTEReqHeight * *p->PTERequestSize);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: dpte_row_width_ub = %u (tiled vert)\n", __func__, *p->dpte_row_width_ub);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_width_ub = %u (tiled vert)\n", __func__, *p->dpte_row_width_ub);
#endif
}
@@ -1837,18 +1795,18 @@ static unsigned int CalculateVMAndRowBytes(struct dml2_core_shared_calculate_vm_
*p->PixelPTEBytesPerRowStorage = *p->PixelPTEBytesPerRow;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: GPUVMMinPageSizeKBytes = %u\n", __func__, p->GPUVMMinPageSizeKBytes);
- dml2_printf("DML::%s: GPUVMEnable = %u\n", __func__, p->GPUVMEnable);
- dml2_printf("DML::%s: meta_row_height = %u\n", __func__, *p->meta_row_height);
- dml2_printf("DML::%s: dpte_row_height = %u\n", __func__, *p->dpte_row_height);
- dml2_printf("DML::%s: dpte_row_height_linear = %u\n", __func__, *p->dpte_row_height_linear);
- dml2_printf("DML::%s: dpte_row_width_ub = %u\n", __func__, *p->dpte_row_width_ub);
- dml2_printf("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, *p->PixelPTEBytesPerRow);
- dml2_printf("DML::%s: PixelPTEBytesPerRowStorage = %u\n", __func__, *p->PixelPTEBytesPerRowStorage);
- dml2_printf("DML::%s: PTEBufferSizeInRequests = %u\n", __func__, p->PTEBufferSizeInRequests);
- dml2_printf("DML::%s: dpte_row_height_one_row_per_frame = %u\n", __func__, *p->dpte_row_height_one_row_per_frame);
- dml2_printf("DML::%s: dpte_row_width_ub_one_row_per_frame = %u\n", __func__, *p->dpte_row_width_ub_one_row_per_frame);
- dml2_printf("DML::%s: PixelPTEBytesPerRow_one_row_per_frame = %u\n", __func__, *p->PixelPTEBytesPerRow_one_row_per_frame);
+ DML_LOG_VERBOSE("DML::%s: GPUVMMinPageSizeKBytes = %u\n", __func__, p->GPUVMMinPageSizeKBytes);
+ DML_LOG_VERBOSE("DML::%s: GPUVMEnable = %u\n", __func__, p->GPUVMEnable);
+ DML_LOG_VERBOSE("DML::%s: meta_row_height = %u\n", __func__, *p->meta_row_height);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_height = %u\n", __func__, *p->dpte_row_height);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_height_linear = %u\n", __func__, *p->dpte_row_height_linear);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_width_ub = %u\n", __func__, *p->dpte_row_width_ub);
+ DML_LOG_VERBOSE("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, *p->PixelPTEBytesPerRow);
+ DML_LOG_VERBOSE("DML::%s: PixelPTEBytesPerRowStorage = %u\n", __func__, *p->PixelPTEBytesPerRowStorage);
+ DML_LOG_VERBOSE("DML::%s: PTEBufferSizeInRequests = %u\n", __func__, p->PTEBufferSizeInRequests);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_height_one_row_per_frame = %u\n", __func__, *p->dpte_row_height_one_row_per_frame);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_width_ub_one_row_per_frame = %u\n", __func__, *p->dpte_row_width_ub_one_row_per_frame);
+ DML_LOG_VERBOSE("DML::%s: PixelPTEBytesPerRow_one_row_per_frame = %u\n", __func__, *p->PixelPTEBytesPerRow_one_row_per_frame);
#endif
return vm_bytes;
@@ -1879,12 +1837,12 @@ static unsigned int CalculatePrefetchSourceLines(
double numLines = 0;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatio = %f\n", __func__, VRatio);
- dml2_printf("DML::%s: VTaps = %u\n", __func__, VTaps);
- dml2_printf("DML::%s: ViewportXStart = %u\n", __func__, ViewportXStart);
- dml2_printf("DML::%s: ViewportYStart = %u\n", __func__, ViewportYStart);
- dml2_printf("DML::%s: ViewportStationary = %u\n", __func__, ViewportStationary);
- dml2_printf("DML::%s: SwathHeight = %u\n", __func__, SwathHeight);
+ DML_LOG_VERBOSE("DML::%s: VRatio = %f\n", __func__, VRatio);
+ DML_LOG_VERBOSE("DML::%s: VTaps = %u\n", __func__, VTaps);
+ DML_LOG_VERBOSE("DML::%s: ViewportXStart = %u\n", __func__, ViewportXStart);
+ DML_LOG_VERBOSE("DML::%s: ViewportYStart = %u\n", __func__, ViewportYStart);
+ DML_LOG_VERBOSE("DML::%s: ViewportStationary = %u\n", __func__, ViewportStationary);
+ DML_LOG_VERBOSE("DML::%s: SwathHeight = %u\n", __func__, SwathHeight);
#endif
if (ProgressiveToInterlaceUnitInOPP)
*VInitPreFill = (unsigned int)(math_floor2((VRatio + (double)VTaps + 1) / 2.0, 1));
@@ -1919,11 +1877,11 @@ static unsigned int CalculatePrefetchSourceLines(
numLines = *MaxNumSwath * SwathHeight + MaxPartialSwath;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: vp_start_rot = %u\n", __func__, vp_start_rot);
- dml2_printf("DML::%s: VInitPreFill = %u\n", __func__, *VInitPreFill);
- dml2_printf("DML::%s: MaxPartialSwath = %u\n", __func__, MaxPartialSwath);
- dml2_printf("DML::%s: MaxNumSwath = %u\n", __func__, *MaxNumSwath);
- dml2_printf("DML::%s: Prefetch source lines = %3.2f\n", __func__, numLines);
+ DML_LOG_VERBOSE("DML::%s: vp_start_rot = %u\n", __func__, vp_start_rot);
+ DML_LOG_VERBOSE("DML::%s: VInitPreFill = %u\n", __func__, *VInitPreFill);
+ DML_LOG_VERBOSE("DML::%s: MaxPartialSwath = %u\n", __func__, MaxPartialSwath);
+ DML_LOG_VERBOSE("DML::%s: MaxNumSwath = %u\n", __func__, *MaxNumSwath);
+ DML_LOG_VERBOSE("DML::%s: Prefetch source lines = %3.2f\n", __func__, numLines);
#endif
return (unsigned int)(numLines);
@@ -1992,8 +1950,8 @@ static void CalculateMALLUseForStaticScreen(
if (is_using_mall_for_ss[k])
TotalSurfaceSizeInMALL = TotalSurfaceSizeInMALL + SurfaceSizeInMALL[k];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, is_using_mall_for_ss = %u\n", __func__, k, is_using_mall_for_ss[k]);
- dml2_printf("DML::%s: k=%u, TotalSurfaceSizeInMALL = %u\n", __func__, k, TotalSurfaceSizeInMALL);
+ DML_LOG_VERBOSE("DML::%s: k=%u, is_using_mall_for_ss = %u\n", __func__, k, is_using_mall_for_ss[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, TotalSurfaceSizeInMALL = %u\n", __func__, k, TotalSurfaceSizeInMALL);
#endif
}
@@ -2007,7 +1965,7 @@ static void CalculateMALLUseForStaticScreen(
(!CanAddAnotherSurfaceToMALL || SurfaceSizeInMALL[k] < SurfaceSizeInMALL[SurfaceToAddToMALL])) {
CanAddAnotherSurfaceToMALL = true;
SurfaceToAddToMALL = k;
- dml2_printf("DML::%s: k=%u, UseMALLForStaticScreen = %u (dis, en, optimize)\n", __func__, k, display_cfg->plane_descriptors[k].overrides.refresh_from_mall);
+ DML_LOG_VERBOSE("DML::%s: k=%u, UseMALLForStaticScreen = %u (dis, en, optimize)\n", __func__, k, display_cfg->plane_descriptors[k].overrides.refresh_from_mall);
}
}
if (CanAddAnotherSurfaceToMALL) {
@@ -2015,8 +1973,8 @@ static void CalculateMALLUseForStaticScreen(
TotalSurfaceSizeInMALL = TotalSurfaceSizeInMALL + SurfaceSizeInMALL[SurfaceToAddToMALL];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: SurfaceToAddToMALL = %u\n", __func__, SurfaceToAddToMALL);
- dml2_printf("DML::%s: TotalSurfaceSizeInMALL = %u\n", __func__, TotalSurfaceSizeInMALL);
+ DML_LOG_VERBOSE("DML::%s: SurfaceToAddToMALL = %u\n", __func__, SurfaceToAddToMALL);
+ DML_LOG_VERBOSE("DML::%s: TotalSurfaceSizeInMALL = %u\n", __func__, TotalSurfaceSizeInMALL);
#endif
}
}
@@ -2090,7 +2048,6 @@ static void CalculateDCCConfiguration(
yuv420 = 1;
else
yuv420 = 0;
-
horz_div_l = 1;
horz_div_c = 1;
vert_div_l = 1;
@@ -2189,15 +2146,15 @@ static void CalculateDCCConfiguration(
segment_order_vert_contiguous_chroma = 0;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DCCEnabled = %u\n", __func__, DCCEnabled);
- dml2_printf("DML::%s: nomDETInKByte = %u\n", __func__, nomDETInKByte);
- dml2_printf("DML::%s: DETBufferSizeForDCC = %u\n", __func__, DETBufferSizeForDCC);
- dml2_printf("DML::%s: req128_horz_wc_l = %u\n", __func__, req128_horz_wc_l);
- dml2_printf("DML::%s: req128_horz_wc_c = %u\n", __func__, req128_horz_wc_c);
- dml2_printf("DML::%s: full_swath_bytes_horz_wc_l = %u\n", __func__, full_swath_bytes_horz_wc_l);
- dml2_printf("DML::%s: full_swath_bytes_vert_wc_c = %u\n", __func__, full_swath_bytes_vert_wc_c);
- dml2_printf("DML::%s: segment_order_horz_contiguous_luma = %u\n", __func__, segment_order_horz_contiguous_luma);
- dml2_printf("DML::%s: segment_order_horz_contiguous_chroma = %u\n", __func__, segment_order_horz_contiguous_chroma);
+ DML_LOG_VERBOSE("DML::%s: DCCEnabled = %u\n", __func__, DCCEnabled);
+ DML_LOG_VERBOSE("DML::%s: nomDETInKByte = %u\n", __func__, nomDETInKByte);
+ DML_LOG_VERBOSE("DML::%s: DETBufferSizeForDCC = %u\n", __func__, DETBufferSizeForDCC);
+ DML_LOG_VERBOSE("DML::%s: req128_horz_wc_l = %u\n", __func__, req128_horz_wc_l);
+ DML_LOG_VERBOSE("DML::%s: req128_horz_wc_c = %u\n", __func__, req128_horz_wc_c);
+ DML_LOG_VERBOSE("DML::%s: full_swath_bytes_horz_wc_l = %u\n", __func__, full_swath_bytes_horz_wc_l);
+ DML_LOG_VERBOSE("DML::%s: full_swath_bytes_vert_wc_c = %u\n", __func__, full_swath_bytes_vert_wc_c);
+ DML_LOG_VERBOSE("DML::%s: segment_order_horz_contiguous_luma = %u\n", __func__, segment_order_horz_contiguous_luma);
+ DML_LOG_VERBOSE("DML::%s: segment_order_horz_contiguous_chroma = %u\n", __func__, segment_order_horz_contiguous_chroma);
#endif
if (DCCProgrammingAssumesScanDirectionUnknown == true) {
if (req128_horz_wc_l == 0 && req128_vert_wc_l == 0) {
@@ -2287,12 +2244,12 @@ static void CalculateDCCConfiguration(
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: MaxUncompressedBlockLuma = %u\n", __func__, *MaxUncompressedBlockLuma);
- dml2_printf("DML::%s: MaxCompressedBlockLuma = %u\n", __func__, *MaxCompressedBlockLuma);
- dml2_printf("DML::%s: IndependentBlockLuma = %u\n", __func__, *IndependentBlockLuma);
- dml2_printf("DML::%s: MaxUncompressedBlockChroma = %u\n", __func__, *MaxUncompressedBlockChroma);
- dml2_printf("DML::%s: MaxCompressedBlockChroma = %u\n", __func__, *MaxCompressedBlockChroma);
- dml2_printf("DML::%s: IndependentBlockChroma = %u\n", __func__, *IndependentBlockChroma);
+ DML_LOG_VERBOSE("DML::%s: MaxUncompressedBlockLuma = %u\n", __func__, *MaxUncompressedBlockLuma);
+ DML_LOG_VERBOSE("DML::%s: MaxCompressedBlockLuma = %u\n", __func__, *MaxCompressedBlockLuma);
+ DML_LOG_VERBOSE("DML::%s: IndependentBlockLuma = %u\n", __func__, *IndependentBlockLuma);
+ DML_LOG_VERBOSE("DML::%s: MaxUncompressedBlockChroma = %u\n", __func__, *MaxUncompressedBlockChroma);
+ DML_LOG_VERBOSE("DML::%s: MaxCompressedBlockChroma = %u\n", __func__, *MaxCompressedBlockChroma);
+ DML_LOG_VERBOSE("DML::%s: IndependentBlockChroma = %u\n", __func__, *IndependentBlockChroma);
#endif
}
@@ -2312,26 +2269,26 @@ static void calculate_mcache_row_bytes(
unsigned int mvmpg_per_mcache;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: num_chans = %u\n", __func__, p->num_chans);
- dml2_printf("DML::%s: mem_word_bytes = %u\n", __func__, p->mem_word_bytes);
- dml2_printf("DML::%s: mcache_line_size_bytes = %u\n", __func__, p->mcache_line_size_bytes);
- dml2_printf("DML::%s: mcache_size_bytes = %u\n", __func__, p->mcache_size_bytes);
- dml2_printf("DML::%s: gpuvm_enable = %u\n", __func__, p->gpuvm_enable);
- dml2_printf("DML::%s: gpuvm_page_size_kbytes = %u\n", __func__, p->gpuvm_page_size_kbytes);
- dml2_printf("DML::%s: vp_stationary = %u\n", __func__, p->vp_stationary);
- dml2_printf("DML::%s: tiling_mode = %u\n", __func__, p->tiling_mode);
- dml2_printf("DML::%s: vp_start_x = %u\n", __func__, p->vp_start_x);
- dml2_printf("DML::%s: vp_start_y = %u\n", __func__, p->vp_start_y);
- dml2_printf("DML::%s: full_vp_width = %u\n", __func__, p->full_vp_width);
- dml2_printf("DML::%s: full_vp_height = %u\n", __func__, p->full_vp_height);
- dml2_printf("DML::%s: blk_width = %u\n", __func__, p->blk_width);
- dml2_printf("DML::%s: blk_height = %u\n", __func__, p->blk_height);
- dml2_printf("DML::%s: vmpg_width = %u\n", __func__, p->vmpg_width);
- dml2_printf("DML::%s: vmpg_height = %u\n", __func__, p->vmpg_height);
- dml2_printf("DML::%s: full_swath_bytes = %u\n", __func__, p->full_swath_bytes);
-#endif
- DML2_ASSERT(p->mcache_line_size_bytes != 0);
- DML2_ASSERT(p->mcache_size_bytes != 0);
+ DML_LOG_VERBOSE("DML::%s: num_chans = %u\n", __func__, p->num_chans);
+ DML_LOG_VERBOSE("DML::%s: mem_word_bytes = %u\n", __func__, p->mem_word_bytes);
+ DML_LOG_VERBOSE("DML::%s: mcache_line_size_bytes = %u\n", __func__, p->mcache_line_size_bytes);
+ DML_LOG_VERBOSE("DML::%s: mcache_size_bytes = %u\n", __func__, p->mcache_size_bytes);
+ DML_LOG_VERBOSE("DML::%s: gpuvm_enable = %u\n", __func__, p->gpuvm_enable);
+ DML_LOG_VERBOSE("DML::%s: gpuvm_page_size_kbytes = %u\n", __func__, p->gpuvm_page_size_kbytes);
+ DML_LOG_VERBOSE("DML::%s: vp_stationary = %u\n", __func__, p->vp_stationary);
+ DML_LOG_VERBOSE("DML::%s: tiling_mode = %u\n", __func__, p->tiling_mode);
+ DML_LOG_VERBOSE("DML::%s: vp_start_x = %u\n", __func__, p->vp_start_x);
+ DML_LOG_VERBOSE("DML::%s: vp_start_y = %u\n", __func__, p->vp_start_y);
+ DML_LOG_VERBOSE("DML::%s: full_vp_width = %u\n", __func__, p->full_vp_width);
+ DML_LOG_VERBOSE("DML::%s: full_vp_height = %u\n", __func__, p->full_vp_height);
+ DML_LOG_VERBOSE("DML::%s: blk_width = %u\n", __func__, p->blk_width);
+ DML_LOG_VERBOSE("DML::%s: blk_height = %u\n", __func__, p->blk_height);
+ DML_LOG_VERBOSE("DML::%s: vmpg_width = %u\n", __func__, p->vmpg_width);
+ DML_LOG_VERBOSE("DML::%s: vmpg_height = %u\n", __func__, p->vmpg_height);
+ DML_LOG_VERBOSE("DML::%s: full_swath_bytes = %u\n", __func__, p->full_swath_bytes);
+#endif
+ DML_ASSERT(p->mcache_line_size_bytes != 0);
+ DML_ASSERT(p->mcache_size_bytes != 0);
*p->mvmpg_width = 0;
*p->mvmpg_height = 0;
@@ -2339,6 +2296,7 @@ static void calculate_mcache_row_bytes(
if (p->full_vp_height == 0 && p->full_vp_width == 0) {
*p->num_mcaches = 0;
*p->mcache_row_bytes = 0;
+ *p->mcache_row_bytes_per_channel = 0;
} else {
blk_bytes = dml_get_tile_block_size_bytes(p->tiling_mode);
@@ -2355,8 +2313,8 @@ static void calculate_mcache_row_bytes(
*p->mvmpg_width = p->vmpg_width;
*p->mvmpg_height = p->vmpg_height;
} else if (!((blk_bytes == 65536) && (vmpg_bytes == 4096))) {
- dml2_printf("ERROR: DML::%s: Tiling size and vm page size combination not supported\n", __func__);
- DML2_ASSERT(0);
+ DML_LOG_VERBOSE("ERROR: DML::%s: Tiling size and vm page size combination not supported\n", __func__);
+ DML_ASSERT(0);
}
}
@@ -2407,38 +2365,42 @@ static void calculate_mcache_row_bytes(
// If this mcache_row_bytes for the full viewport of the surface is less than or equal to mcache_bytes,
// then one mcache can be used for this request stream. If not, it is useful to know the width of the viewport that can be supported in the mcache_bytes.
- if (p->gpuvm_enable || !p->surf_vert) {
- *p->mcache_row_bytes = mvmpg_per_row_ub * meta_per_mvmpg_per_channel_ub;
+ if (p->gpuvm_enable || p->surf_vert) {
+ *p->mcache_row_bytes_per_channel = mvmpg_per_row_ub * meta_per_mvmpg_per_channel_ub;
+ *p->mcache_row_bytes = *p->mcache_row_bytes_per_channel * p->num_chans;
} else { // horizontal and gpuvm disable
*p->mcache_row_bytes = *p->meta_row_width_ub * p->blk_height * p->bytes_per_pixel / 256;
- *p->mcache_row_bytes = (unsigned int)math_ceil2((double)*p->mcache_row_bytes / p->num_chans, p->mcache_line_size_bytes);
+ if (p->mcache_line_size_bytes != 0)
+ *p->mcache_row_bytes_per_channel = (unsigned int)math_ceil2((double)*p->mcache_row_bytes / p->num_chans, p->mcache_line_size_bytes);
}
*p->dcc_dram_bw_pref_overhead_factor = 1 + math_max2(1.0 / 256.0, *p->mcache_row_bytes / p->full_swath_bytes); // dcc_dr_oh_pref
- *p->num_mcaches = (unsigned int)math_ceil2((double)*p->mcache_row_bytes / p->mcache_size_bytes, 1);
+ if (p->mcache_size_bytes != 0)
+ *p->num_mcaches = (unsigned int)math_ceil2((double)*p->mcache_row_bytes_per_channel / p->mcache_size_bytes, 1);
mvmpg_per_mcache = p->mcache_size_bytes / meta_per_mvmpg_per_channel_ub;
*p->mvmpg_per_mcache_lb = (unsigned int)math_floor2(mvmpg_per_mcache, 1);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: gpuvm_enable = %u\n", __func__, p->gpuvm_enable);
- dml2_printf("DML::%s: vmpg_bytes = %u\n", __func__, vmpg_bytes);
- dml2_printf("DML::%s: blk_bytes = %u\n", __func__, blk_bytes);
- dml2_printf("DML::%s: meta_per_mvmpg_per_channel = %f\n", __func__, meta_per_mvmpg_per_channel);
- dml2_printf("DML::%s: mvmpg_per_row_ub = %u\n", __func__, mvmpg_per_row_ub);
- dml2_printf("DML::%s: meta_row_width_ub = %u\n", __func__, *p->meta_row_width_ub);
- dml2_printf("DML::%s: mvmpg_width = %u\n", __func__, *p->mvmpg_width);
- dml2_printf("DML::%s: mvmpg_height = %u\n", __func__, *p->mvmpg_height);
- dml2_printf("DML::%s: dcc_dram_bw_nom_overhead_factor = %f\n", __func__, *p->dcc_dram_bw_nom_overhead_factor);
- dml2_printf("DML::%s: dcc_dram_bw_pref_overhead_factor = %f\n", __func__, *p->dcc_dram_bw_pref_overhead_factor);
+ DML_LOG_VERBOSE("DML::%s: gpuvm_enable = %u\n", __func__, p->gpuvm_enable);
+ DML_LOG_VERBOSE("DML::%s: vmpg_bytes = %u\n", __func__, vmpg_bytes);
+ DML_LOG_VERBOSE("DML::%s: blk_bytes = %u\n", __func__, blk_bytes);
+ DML_LOG_VERBOSE("DML::%s: meta_per_mvmpg_per_channel = %f\n", __func__, meta_per_mvmpg_per_channel);
+ DML_LOG_VERBOSE("DML::%s: mvmpg_per_row_ub = %u\n", __func__, mvmpg_per_row_ub);
+ DML_LOG_VERBOSE("DML::%s: meta_row_width_ub = %u\n", __func__, *p->meta_row_width_ub);
+ DML_LOG_VERBOSE("DML::%s: mvmpg_width = %u\n", __func__, *p->mvmpg_width);
+ DML_LOG_VERBOSE("DML::%s: mvmpg_height = %u\n", __func__, *p->mvmpg_height);
+ DML_LOG_VERBOSE("DML::%s: dcc_dram_bw_nom_overhead_factor = %f\n", __func__, *p->dcc_dram_bw_nom_overhead_factor);
+ DML_LOG_VERBOSE("DML::%s: dcc_dram_bw_pref_overhead_factor = %f\n", __func__, *p->dcc_dram_bw_pref_overhead_factor);
#endif
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: mcache_row_bytes = %u\n", __func__, *p->mcache_row_bytes);
- dml2_printf("DML::%s: num_mcaches = %u\n", __func__, *p->num_mcaches);
+ DML_LOG_VERBOSE("DML::%s: mcache_row_bytes = %u\n", __func__, *p->mcache_row_bytes);
+ DML_LOG_VERBOSE("DML::%s: mcache_row_bytes_per_channel = %u\n", __func__, *p->mcache_row_bytes_per_channel);
+ DML_LOG_VERBOSE("DML::%s: num_mcaches = %u\n", __func__, *p->num_mcaches);
#endif
- DML2_ASSERT(*p->num_mcaches > 0);
+ DML_ASSERT(*p->num_mcaches > 0);
}
static void calculate_mcache_setting(
@@ -2452,11 +2414,13 @@ static void calculate_mcache_setting(
*p->num_mcaches_l = 0;
*p->mcache_row_bytes_l = 0;
+ *p->mcache_row_bytes_per_channel_l = 0;
*p->dcc_dram_bw_nom_overhead_factor_l = 1.0;
*p->dcc_dram_bw_pref_overhead_factor_l = 1.0;
*p->num_mcaches_c = 0;
*p->mcache_row_bytes_c = 0;
+ *p->mcache_row_bytes_per_channel_c = 0;
*p->dcc_dram_bw_nom_overhead_factor_c = 1.0;
*p->dcc_dram_bw_pref_overhead_factor_c = 1.0;
@@ -2492,6 +2456,7 @@ static void calculate_mcache_setting(
// output
l->l_p.num_mcaches = p->num_mcaches_l;
l->l_p.mcache_row_bytes = p->mcache_row_bytes_l;
+ l->l_p.mcache_row_bytes_per_channel = p->mcache_row_bytes_per_channel_l;
l->l_p.dcc_dram_bw_nom_overhead_factor = p->dcc_dram_bw_nom_overhead_factor_l;
l->l_p.dcc_dram_bw_pref_overhead_factor = p->dcc_dram_bw_pref_overhead_factor_l;
l->l_p.mvmpg_width = &l->mvmpg_width_l;
@@ -2501,7 +2466,7 @@ static void calculate_mcache_setting(
l->l_p.mvmpg_per_mcache_lb = &l->mvmpg_per_mcache_lb_l;
calculate_mcache_row_bytes(scratch, &l->l_p);
- dml2_assert(*p->num_mcaches_l > 0);
+ DML_ASSERT(*p->num_mcaches_l > 0);
if (l->is_dual_plane) {
l->c_p.num_chans = p->num_chans;
@@ -2527,6 +2492,7 @@ static void calculate_mcache_setting(
// output
l->c_p.num_mcaches = p->num_mcaches_c;
l->c_p.mcache_row_bytes = p->mcache_row_bytes_c;
+ l->c_p.mcache_row_bytes_per_channel = p->mcache_row_bytes_per_channel_c;
l->c_p.dcc_dram_bw_nom_overhead_factor = p->dcc_dram_bw_nom_overhead_factor_c;
l->c_p.dcc_dram_bw_pref_overhead_factor = p->dcc_dram_bw_pref_overhead_factor_c;
l->c_p.mvmpg_width = &l->mvmpg_width_c;
@@ -2536,12 +2502,12 @@ static void calculate_mcache_setting(
l->c_p.mvmpg_per_mcache_lb = &l->mvmpg_per_mcache_lb_c;
calculate_mcache_row_bytes(scratch, &l->c_p);
- dml2_assert(*p->num_mcaches_c > 0);
+ DML_ASSERT(*p->num_mcaches_c > 0);
}
// Sharing for iMALL access
- l->mcache_remainder_l = *p->mcache_row_bytes_l % p->mcache_size_bytes;
- l->mcache_remainder_c = *p->mcache_row_bytes_c % p->mcache_size_bytes;
+ l->mcache_remainder_l = *p->mcache_row_bytes_per_channel_l % p->mcache_size_bytes;
+ l->mcache_remainder_c = *p->mcache_row_bytes_per_channel_c % p->mcache_size_bytes;
l->mvmpg_access_width_l = p->surf_vert ? l->mvmpg_height_l : l->mvmpg_width_l;
l->mvmpg_access_width_c = p->surf_vert ? l->mvmpg_height_c : l->mvmpg_width_c;
@@ -2561,40 +2527,42 @@ static void calculate_mcache_setting(
if (*p->num_mcaches_l) {
l->avg_mcache_element_size_l = l->meta_row_width_l / *p->num_mcaches_l;
}
-
- if (l->is_dual_plane && *p->num_mcaches_c) {
+ if (l->is_dual_plane) {
l->avg_mcache_element_size_c = l->meta_row_width_c / *p->num_mcaches_c;
- if (!p->imall_enable || (*p->mall_comb_mcache_l == *p->mall_comb_mcache_c)) {
- l->lc_comb_last_mcache_size = (unsigned int)((l->mcache_remainder_l * (*p->mall_comb_mcache_l ? 2 : 1) * l->luma_time_factor) +
- (l->mcache_remainder_c * (*p->mall_comb_mcache_c ? 2 : 1)));
+ /* if either remainder is 0, then mcache sharing is not needed or not possible due to full utilization */
+ if (l->mcache_remainder_l && l->mcache_remainder_c) {
+ if (!p->imall_enable || (*p->mall_comb_mcache_l == *p->mall_comb_mcache_c)) {
+ l->lc_comb_last_mcache_size = (unsigned int)((l->mcache_remainder_l * (*p->mall_comb_mcache_l ? 2 : 1) * l->luma_time_factor) +
+ (l->mcache_remainder_c * (*p->mall_comb_mcache_c ? 2 : 1)));
+ }
+ *p->lc_comb_mcache = (l->lc_comb_last_mcache_size <= p->mcache_size_bytes) && (*p->mall_comb_mcache_l == *p->mall_comb_mcache_c);
}
- *p->lc_comb_mcache = (l->lc_comb_last_mcache_size <= p->mcache_size_bytes) && (*p->mall_comb_mcache_l == *p->mall_comb_mcache_c);
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: imall_enable = %u\n", __func__, p->imall_enable);
- dml2_printf("DML::%s: is_dual_plane = %u\n", __func__, l->is_dual_plane);
- dml2_printf("DML::%s: surf_vert = %u\n", __func__, p->surf_vert);
- dml2_printf("DML::%s: mvmpg_width_l = %u\n", __func__, l->mvmpg_width_l);
- dml2_printf("DML::%s: mvmpg_height_l = %u\n", __func__, l->mvmpg_height_l);
- dml2_printf("DML::%s: mcache_remainder_l = %f\n", __func__, l->mcache_remainder_l);
- dml2_printf("DML::%s: num_mcaches_l = %u\n", __func__, *p->num_mcaches_l);
- dml2_printf("DML::%s: avg_mcache_element_size_l = %u\n", __func__, l->avg_mcache_element_size_l);
- dml2_printf("DML::%s: mvmpg_access_width_l = %u\n", __func__, l->mvmpg_access_width_l);
- dml2_printf("DML::%s: mall_comb_mcache_l = %u\n", __func__, *p->mall_comb_mcache_l);
+ DML_LOG_VERBOSE("DML::%s: imall_enable = %u\n", __func__, p->imall_enable);
+ DML_LOG_VERBOSE("DML::%s: is_dual_plane = %u\n", __func__, l->is_dual_plane);
+ DML_LOG_VERBOSE("DML::%s: surf_vert = %u\n", __func__, p->surf_vert);
+ DML_LOG_VERBOSE("DML::%s: mvmpg_width_l = %u\n", __func__, l->mvmpg_width_l);
+ DML_LOG_VERBOSE("DML::%s: mvmpg_height_l = %u\n", __func__, l->mvmpg_height_l);
+ DML_LOG_VERBOSE("DML::%s: mcache_remainder_l = %f\n", __func__, l->mcache_remainder_l);
+ DML_LOG_VERBOSE("DML::%s: num_mcaches_l = %u\n", __func__, *p->num_mcaches_l);
+ DML_LOG_VERBOSE("DML::%s: avg_mcache_element_size_l = %u\n", __func__, l->avg_mcache_element_size_l);
+ DML_LOG_VERBOSE("DML::%s: mvmpg_access_width_l = %u\n", __func__, l->mvmpg_access_width_l);
+ DML_LOG_VERBOSE("DML::%s: mall_comb_mcache_l = %u\n", __func__, *p->mall_comb_mcache_l);
if (l->is_dual_plane) {
- dml2_printf("DML::%s: mvmpg_width_c = %u\n", __func__, l->mvmpg_width_c);
- dml2_printf("DML::%s: mvmpg_height_c = %u\n", __func__, l->mvmpg_height_c);
- dml2_printf("DML::%s: mcache_remainder_c = %f\n", __func__, l->mcache_remainder_c);
- dml2_printf("DML::%s: luma_time_factor = %f\n", __func__, l->luma_time_factor);
- dml2_printf("DML::%s: num_mcaches_c = %u\n", __func__, *p->num_mcaches_c);
- dml2_printf("DML::%s: avg_mcache_element_size_c = %u\n", __func__, l->avg_mcache_element_size_c);
- dml2_printf("DML::%s: mvmpg_access_width_c = %u\n", __func__, l->mvmpg_access_width_c);
- dml2_printf("DML::%s: mall_comb_mcache_c = %u\n", __func__, *p->mall_comb_mcache_c);
- dml2_printf("DML::%s: lc_comb_last_mcache_size = %u\n", __func__, l->lc_comb_last_mcache_size);
- dml2_printf("DML::%s: lc_comb_mcache = %u\n", __func__, *p->lc_comb_mcache);
+ DML_LOG_VERBOSE("DML::%s: mvmpg_width_c = %u\n", __func__, l->mvmpg_width_c);
+ DML_LOG_VERBOSE("DML::%s: mvmpg_height_c = %u\n", __func__, l->mvmpg_height_c);
+ DML_LOG_VERBOSE("DML::%s: mcache_remainder_c = %f\n", __func__, l->mcache_remainder_c);
+ DML_LOG_VERBOSE("DML::%s: luma_time_factor = %f\n", __func__, l->luma_time_factor);
+ DML_LOG_VERBOSE("DML::%s: num_mcaches_c = %u\n", __func__, *p->num_mcaches_c);
+ DML_LOG_VERBOSE("DML::%s: avg_mcache_element_size_c = %u\n", __func__, l->avg_mcache_element_size_c);
+ DML_LOG_VERBOSE("DML::%s: mvmpg_access_width_c = %u\n", __func__, l->mvmpg_access_width_c);
+ DML_LOG_VERBOSE("DML::%s: mall_comb_mcache_c = %u\n", __func__, *p->mall_comb_mcache_c);
+ DML_LOG_VERBOSE("DML::%s: lc_comb_last_mcache_size = %u\n", __func__, l->lc_comb_last_mcache_size);
+ DML_LOG_VERBOSE("DML::%s: lc_comb_mcache = %u\n", __func__, *p->lc_comb_mcache);
}
#endif
// calculate split_coordinate
@@ -2614,11 +2582,11 @@ static void calculate_mcache_setting(
}
#ifdef __DML_VBA_DEBUG__
for (n = 0; n < *p->num_mcaches_l; n++)
- dml2_printf("DML::%s: mcache_offsets_l[%u] = %u\n", __func__, n, p->mcache_offsets_l[n]);
+ DML_LOG_VERBOSE("DML::%s: mcache_offsets_l[%u] = %u\n", __func__, n, p->mcache_offsets_l[n]);
if (l->is_dual_plane) {
for (n = 0; n < *p->num_mcaches_c; n++)
- dml2_printf("DML::%s: mcache_offsets_c[%u] = %u\n", __func__, n, p->mcache_offsets_c[n]);
+ DML_LOG_VERBOSE("DML::%s: mcache_offsets_c[%u] = %u\n", __func__, n, p->mcache_offsets_c[n]);
}
#endif
@@ -2635,10 +2603,10 @@ static void calculate_mcache_setting(
#ifdef __DML_VBA_DEBUG__
for (n = 0; n < *p->num_mcaches_l; n++)
- dml2_printf("DML::%s: mcache_offsets_l[%u] = %u\n", __func__, n, p->mcache_offsets_l[n]);
+ DML_LOG_VERBOSE("DML::%s: mcache_offsets_l[%u] = %u\n", __func__, n, p->mcache_offsets_l[n]);
for (n = 0; n < *p->num_mcaches_c; n++)
- dml2_printf("DML::%s: mcache_offsets_c[%u] = %u\n", __func__, n, p->mcache_offsets_c[n]);
+ DML_LOG_VERBOSE("DML::%s: mcache_offsets_c[%u] = %u\n", __func__, n, p->mcache_offsets_c[n]);
#endif
}
@@ -2669,8 +2637,8 @@ static void calculate_mall_bw_overhead_factor(
mall_prefetch_dram_overhead_factor[k] = 2.0;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, mall_prefetch_sdp_overhead_factor = %f\n", __func__, k, mall_prefetch_sdp_overhead_factor[k]);
- dml2_printf("DML::%s: k=%u, mall_prefetch_dram_overhead_factor = %f\n", __func__, k, mall_prefetch_dram_overhead_factor[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, mall_prefetch_sdp_overhead_factor = %f\n", __func__, k, mall_prefetch_sdp_overhead_factor[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, mall_prefetch_dram_overhead_factor = %f\n", __func__, k, mall_prefetch_dram_overhead_factor[k]);
#endif
}
}
@@ -2682,12 +2650,12 @@ static double dml_get_return_bandwidth_available(
bool is_avg_bw,
bool is_hvm_en,
bool is_hvm_only,
- double dcflk_mhz,
+ double dcfclk_mhz,
double fclk_mhz,
double dram_bw_mbps)
{
double return_bw_mbps = 0.;
- double ideal_sdp_bandwidth = (double)soc->return_bus_width_bytes * dcflk_mhz;
+ double ideal_sdp_bandwidth = (double)soc->return_bus_width_bytes * dcfclk_mhz;
double ideal_fabric_bandwidth = fclk_mhz * (double)soc->fabric_datapath_to_dcn_data_return_bytes;
double ideal_dram_bandwidth = dram_bw_mbps; //dram_speed_mts * soc->clk_table.dram_config.channel_count * soc->clk_table.dram_config.channel_width_bytes;
@@ -2747,26 +2715,24 @@ static double dml_get_return_bandwidth_available(
else // dml2_core_internal_bw_dram
return_bw_mbps = derate_dram_bandwidth;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: is_avg_bw = %u\n", __func__, is_avg_bw);
- dml2_printf("DML::%s: is_hvm_en = %u\n", __func__, is_hvm_en);
- dml2_printf("DML::%s: is_hvm_only = %u\n", __func__, is_hvm_only);
- dml2_printf("DML::%s: state_type = %s\n", __func__, dml2_core_internal_soc_state_type_str(state_type));
- dml2_printf("DML::%s: bw_type = %s\n", __func__, dml2_core_internal_bw_type_str(bw_type));
- dml2_printf("DML::%s: dcflk_mhz = %f\n", __func__, dcflk_mhz);
- dml2_printf("DML::%s: fclk_mhz = %f\n", __func__, fclk_mhz);
- dml2_printf("DML::%s: ideal_sdp_bandwidth = %f\n", __func__, ideal_sdp_bandwidth);
- dml2_printf("DML::%s: ideal_fabric_bandwidth = %f\n", __func__, ideal_fabric_bandwidth);
- dml2_printf("DML::%s: ideal_dram_bandwidth = %f\n", __func__, ideal_dram_bandwidth);
- dml2_printf("DML::%s: derate_sdp_bandwidth = %f (derate %f)\n", __func__, derate_sdp_bandwidth, derate_sdp_factor);
- dml2_printf("DML::%s: derate_fabric_bandwidth = %f (derate %f)\n", __func__, derate_fabric_bandwidth, derate_fabric_factor);
- dml2_printf("DML::%s: derate_dram_bandwidth = %f (derate %f)\n", __func__, derate_dram_bandwidth, derate_dram_factor);
- dml2_printf("DML::%s: return_bw_mbps = %f\n", __func__, return_bw_mbps);
-#endif
+ DML_LOG_VERBOSE("DML::%s: is_avg_bw = %u\n", __func__, is_avg_bw);
+ DML_LOG_VERBOSE("DML::%s: is_hvm_en = %u\n", __func__, is_hvm_en);
+ DML_LOG_VERBOSE("DML::%s: is_hvm_only = %u\n", __func__, is_hvm_only);
+ DML_LOG_VERBOSE("DML::%s: state_type = %s\n", __func__, dml2_core_internal_soc_state_type_str(state_type));
+ DML_LOG_VERBOSE("DML::%s: bw_type = %s\n", __func__, dml2_core_internal_bw_type_str(bw_type));
+ DML_LOG_VERBOSE("DML::%s: dcfclk_mhz = %f\n", __func__, dcfclk_mhz);
+ DML_LOG_VERBOSE("DML::%s: fclk_mhz = %f\n", __func__, fclk_mhz);
+ DML_LOG_VERBOSE("DML::%s: ideal_sdp_bandwidth = %f\n", __func__, ideal_sdp_bandwidth);
+ DML_LOG_VERBOSE("DML::%s: ideal_fabric_bandwidth = %f\n", __func__, ideal_fabric_bandwidth);
+ DML_LOG_VERBOSE("DML::%s: ideal_dram_bandwidth = %f\n", __func__, ideal_dram_bandwidth);
+ DML_LOG_VERBOSE("DML::%s: derate_sdp_bandwidth = %f (derate %f)\n", __func__, derate_sdp_bandwidth, derate_sdp_factor);
+ DML_LOG_VERBOSE("DML::%s: derate_fabric_bandwidth = %f (derate %f)\n", __func__, derate_fabric_bandwidth, derate_fabric_factor);
+ DML_LOG_VERBOSE("DML::%s: derate_dram_bandwidth = %f (derate %f)\n", __func__, derate_dram_bandwidth, derate_dram_factor);
+ DML_LOG_VERBOSE("DML::%s: return_bw_mbps = %f\n", __func__, return_bw_mbps);
return return_bw_mbps;
}
-static void calculate_bandwidth_available(
+static noinline_for_stack void calculate_bandwidth_available(
double avg_bandwidth_available_min[dml2_core_internal_soc_state_max],
double avg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
double urg_bandwidth_available_min[dml2_core_internal_soc_state_max], // min between SDP and DRAM
@@ -2782,9 +2748,9 @@ static void calculate_bandwidth_available(
{
unsigned int n, m;
- dml2_printf("DML::%s: dcfclk_mhz = %f\n", __func__, dcfclk_mhz);
- dml2_printf("DML::%s: fclk_mhz = %f\n", __func__, fclk_mhz);
- dml2_printf("DML::%s: dram_bw_mbps = %f\n", __func__, dram_bw_mbps);
+ DML_LOG_VERBOSE("DML::%s: dcfclk_mhz = %f\n", __func__, dcfclk_mhz);
+ DML_LOG_VERBOSE("DML::%s: fclk_mhz = %f\n", __func__, fclk_mhz);
+ DML_LOG_VERBOSE("DML::%s: dram_bw_mbps = %f\n", __func__, dram_bw_mbps);
// Calculate all the bandwidth availabe
for (m = 0; m < dml2_core_internal_soc_state_max; m++) {
@@ -2803,8 +2769,8 @@ static void calculate_bandwidth_available(
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: avg_bandwidth_available[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), avg_bandwidth_available[m][n]);
- dml2_printf("DML::%s: urg_bandwidth_available[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), urg_bandwidth_available[m][n]);
+ DML_LOG_VERBOSE("DML::%s: avg_bandwidth_available[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), avg_bandwidth_available[m][n]);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_available[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), urg_bandwidth_available[m][n]);
#endif
// urg_bandwidth_available_vm_only is indexed by soc_state
@@ -2818,9 +2784,9 @@ static void calculate_bandwidth_available(
urg_bandwidth_available_min[m] = math_min2(urg_bandwidth_available[m][dml2_core_internal_bw_dram], urg_bandwidth_available[m][dml2_core_internal_bw_sdp]);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: avg_bandwidth_available_min[%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), avg_bandwidth_available_min[m]);
- dml2_printf("DML::%s: urg_bandwidth_available_min[%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), urg_bandwidth_available_min[m]);
- dml2_printf("DML::%s: urg_bandwidth_available_vm_only[%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), urg_bandwidth_available_vm_only[n]);
+ DML_LOG_VERBOSE("DML::%s: avg_bandwidth_available_min[%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), avg_bandwidth_available_min[m]);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_available_min[%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), urg_bandwidth_available_min[m]);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_available_vm_only[%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), urg_bandwidth_available_vm_only[n]);
#endif
}
}
@@ -2854,13 +2820,13 @@ static void calculate_avg_bandwidth_required(
// SysActive and SVP Prefetch AVG bandwidth Check
for (k = 0; k < num_active_planes; ++k) {
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: plane %0d\n", __func__, k);
- dml2_printf("DML::%s: ReadBandwidthLuma=%f\n", __func__, ReadBandwidthLuma[k]);
- dml2_printf("DML::%s: ReadBandwidthChroma=%f\n", __func__, ReadBandwidthChroma[k]);
- dml2_printf("DML::%s: dcc_dram_bw_nom_overhead_factor_p0=%f\n", __func__, dcc_dram_bw_nom_overhead_factor_p0[k]);
- dml2_printf("DML::%s: dcc_dram_bw_nom_overhead_factor_p1=%f\n", __func__, dcc_dram_bw_nom_overhead_factor_p1[k]);
- dml2_printf("DML::%s: mall_prefetch_dram_overhead_factor=%f\n", __func__, mall_prefetch_dram_overhead_factor[k]);
- dml2_printf("DML::%s: mall_prefetch_sdp_overhead_factor=%f\n", __func__, mall_prefetch_sdp_overhead_factor[k]);
+ DML_LOG_VERBOSE("DML::%s: plane %0d\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: ReadBandwidthLuma=%f\n", __func__, ReadBandwidthLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: ReadBandwidthChroma=%f\n", __func__, ReadBandwidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: dcc_dram_bw_nom_overhead_factor_p0=%f\n", __func__, dcc_dram_bw_nom_overhead_factor_p0[k]);
+ DML_LOG_VERBOSE("DML::%s: dcc_dram_bw_nom_overhead_factor_p1=%f\n", __func__, dcc_dram_bw_nom_overhead_factor_p1[k]);
+ DML_LOG_VERBOSE("DML::%s: mall_prefetch_dram_overhead_factor=%f\n", __func__, mall_prefetch_dram_overhead_factor[k]);
+ DML_LOG_VERBOSE("DML::%s: mall_prefetch_sdp_overhead_factor=%f\n", __func__, mall_prefetch_sdp_overhead_factor[k]);
#endif
sdp_overhead_factor = mall_prefetch_sdp_overhead_factor[k];
@@ -2877,10 +2843,10 @@ static void calculate_avg_bandwidth_required(
avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram] += dram_overhead_factor_p0 * ReadBandwidthLuma[k] + dram_overhead_factor_p1 * ReadBandwidthChroma[k] + cursor_bw[k];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_sys_active), dml2_core_internal_bw_type_str(dml2_core_internal_bw_sdp), avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp]);
- dml2_printf("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_sys_active), dml2_core_internal_bw_type_str(dml2_core_internal_bw_dram), avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram]);
- dml2_printf("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_svp_prefetch), dml2_core_internal_bw_type_str(dml2_core_internal_bw_sdp), avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp]);
- dml2_printf("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_svp_prefetch), dml2_core_internal_bw_type_str(dml2_core_internal_bw_dram), avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram]);
+ DML_LOG_VERBOSE("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_sys_active), dml2_core_internal_bw_type_str(dml2_core_internal_bw_sdp), avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp]);
+ DML_LOG_VERBOSE("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_sys_active), dml2_core_internal_bw_type_str(dml2_core_internal_bw_dram), avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram]);
+ DML_LOG_VERBOSE("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_svp_prefetch), dml2_core_internal_bw_type_str(dml2_core_internal_bw_sdp), avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp]);
+ DML_LOG_VERBOSE("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_svp_prefetch), dml2_core_internal_bw_type_str(dml2_core_internal_bw_dram), avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram]);
#endif
}
}
@@ -3055,10 +3021,10 @@ static void CalculateVMRowAndSwath(struct dml2_core_internal_scratch *scratch,
&p->MaxNumSwathY[k]);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, vm_bytes_l = %u (before hvm level)\n", __func__, k, s->vm_bytes_l);
- dml2_printf("DML::%s: k=%u, vm_bytes_c = %u (before hvm level)\n", __func__, k, s->vm_bytes_c);
- dml2_printf("DML::%s: k=%u, meta_row_bytes_per_row_ub_l = %u\n", __func__, k, s->meta_row_bytes_per_row_ub_l[k]);
- dml2_printf("DML::%s: k=%u, meta_row_bytes_per_row_ub_c = %u\n", __func__, k, s->meta_row_bytes_per_row_ub_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, vm_bytes_l = %u (before hvm level)\n", __func__, k, s->vm_bytes_l);
+ DML_LOG_VERBOSE("DML::%s: k=%u, vm_bytes_c = %u (before hvm level)\n", __func__, k, s->vm_bytes_c);
+ DML_LOG_VERBOSE("DML::%s: k=%u, meta_row_bytes_per_row_ub_l = %u\n", __func__, k, s->meta_row_bytes_per_row_ub_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, meta_row_bytes_per_row_ub_c = %u\n", __func__, k, s->meta_row_bytes_per_row_ub_c[k]);
#endif
p->vm_bytes[k] = (s->vm_bytes_l + s->vm_bytes_c) * (1 + 8 * s->HostVMDynamicLevels);
p->meta_row_bytes[k] = s->meta_row_bytes_per_row_ub_l[k] + s->meta_row_bytes_per_row_ub_c[k];
@@ -3066,8 +3032,8 @@ static void CalculateVMRowAndSwath(struct dml2_core_internal_scratch *scratch,
p->meta_row_bytes_per_row_ub_c[k] = s->meta_row_bytes_per_row_ub_c[k];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, meta_row_bytes = %u\n", __func__, k, p->meta_row_bytes[k]);
- dml2_printf("DML::%s: k=%u, vm_bytes = %u (after hvm level)\n", __func__, k, p->vm_bytes[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, meta_row_bytes = %u\n", __func__, k, p->meta_row_bytes[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, vm_bytes = %u (after hvm level)\n", __func__, k, p->vm_bytes[k]);
#endif
if (s->PixelPTEBytesPerRowStorageY[k] <= 64 * s->PTEBufferSizeInRequestsForLuma[k] && s->PixelPTEBytesPerRowStorageC[k] <= 64 * s->PTEBufferSizeInRequestsForChroma[k]) {
p->PTEBufferSizeNotExceeded[k] = true;
@@ -3079,18 +3045,18 @@ static void CalculateVMRowAndSwath(struct dml2_core_internal_scratch *scratch,
s->PixelPTEBytesPerRowC_one_row_per_frame[k] <= 64 * 2 * s->PTEBufferSizeInRequestsForChroma[k]);
#ifdef __DML_VBA_DEBUG__
if (p->PTEBufferSizeNotExceeded[k] == 0 || s->one_row_per_frame_fits_in_buffer[k] == 0) {
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowY = %u (before hvm level)\n", __func__, k, s->PixelPTEBytesPerRowY[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowC = %u (before hvm level)\n", __func__, k, s->PixelPTEBytesPerRowC[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowStorageY = %u\n", __func__, k, s->PixelPTEBytesPerRowStorageY[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowStorageC = %u\n", __func__, k, s->PixelPTEBytesPerRowStorageC[k]);
- dml2_printf("DML::%s: k=%u, PTEBufferSizeInRequestsForLuma = %u\n", __func__, k, s->PTEBufferSizeInRequestsForLuma[k]);
- dml2_printf("DML::%s: k=%u, PTEBufferSizeInRequestsForChroma = %u\n", __func__, k, s->PTEBufferSizeInRequestsForChroma[k]);
- dml2_printf("DML::%s: k=%u, PTEBufferSizeNotExceeded (not one_row_per_frame) = %u\n", __func__, k, p->PTEBufferSizeNotExceeded[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEBytesPerRowY = %u (before hvm level)\n", __func__, k, s->PixelPTEBytesPerRowY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEBytesPerRowC = %u (before hvm level)\n", __func__, k, s->PixelPTEBytesPerRowC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEBytesPerRowStorageY = %u\n", __func__, k, s->PixelPTEBytesPerRowStorageY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEBytesPerRowStorageC = %u\n", __func__, k, s->PixelPTEBytesPerRowStorageC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PTEBufferSizeInRequestsForLuma = %u\n", __func__, k, s->PTEBufferSizeInRequestsForLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PTEBufferSizeInRequestsForChroma = %u\n", __func__, k, s->PTEBufferSizeInRequestsForChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PTEBufferSizeNotExceeded (not one_row_per_frame) = %u\n", __func__, k, p->PTEBufferSizeNotExceeded[k]);
- dml2_printf("DML::%s: k=%u, HostVMDynamicLevels = %u\n", __func__, k, s->HostVMDynamicLevels);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowY_one_row_per_frame = %u\n", __func__, k, s->PixelPTEBytesPerRowY_one_row_per_frame[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowC_one_row_per_frame = %u\n", __func__, k, s->PixelPTEBytesPerRowC_one_row_per_frame[k]);
- dml2_printf("DML::%s: k=%u, one_row_per_frame_fits_in_buffer = %u\n", __func__, k, s->one_row_per_frame_fits_in_buffer[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, HostVMDynamicLevels = %u\n", __func__, k, s->HostVMDynamicLevels);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEBytesPerRowY_one_row_per_frame = %u\n", __func__, k, s->PixelPTEBytesPerRowY_one_row_per_frame[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEBytesPerRowC_one_row_per_frame = %u\n", __func__, k, s->PixelPTEBytesPerRowC_one_row_per_frame[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, one_row_per_frame_fits_in_buffer = %u\n", __func__, k, s->one_row_per_frame_fits_in_buffer[k]);
}
#endif
}
@@ -3121,8 +3087,8 @@ static void CalculateVMRowAndSwath(struct dml2_core_internal_scratch *scratch,
for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
p->DCCMetaBufferSizeNotExceeded[k] = true;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, SurfaceSizeInMALL = %u\n", __func__, k, p->SurfaceSizeInMALL[k]);
- dml2_printf("DML::%s: k=%u, is_using_mall_for_ss = %u\n", __func__, k, p->is_using_mall_for_ss[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, SurfaceSizeInMALL = %u\n", __func__, k, p->SurfaceSizeInMALL[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, is_using_mall_for_ss = %u\n", __func__, k, p->is_using_mall_for_ss[k]);
#endif
p->use_one_row_for_frame[k] = p->myPipe[k].FORCE_ONE_ROW_FOR_FRAME || p->is_using_mall_for_ss[k] || (p->display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_main_pipe) ||
(dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) || (p->display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes > 64 && dml_is_vertical_rotation(p->myPipe[k].RotationAngle));
@@ -3145,9 +3111,9 @@ static void CalculateVMRowAndSwath(struct dml2_core_internal_scratch *scratch,
p->DCCMetaBufferSizeNotExceeded[k] = false;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, meta_row_bytes = %d\n", __func__, k, p->meta_row_bytes[k]);
- dml2_printf("DML::%s: k=%d, DCCMetaBufferSizeBytes = %d\n", __func__, k, p->DCCMetaBufferSizeBytes);
- dml2_printf("DML::%s: k=%d, DCCMetaBufferSizeNotExceeded = %d\n", __func__, k, p->DCCMetaBufferSizeNotExceeded[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, meta_row_bytes = %d\n", __func__, k, p->meta_row_bytes[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, DCCMetaBufferSizeBytes = %d\n", __func__, k, p->DCCMetaBufferSizeBytes);
+ DML_LOG_VERBOSE("DML::%s: k=%d, DCCMetaBufferSizeNotExceeded = %d\n", __func__, k, p->DCCMetaBufferSizeNotExceeded[k]);
#endif
}
@@ -3184,20 +3150,20 @@ static void CalculateVMRowAndSwath(struct dml2_core_internal_scratch *scratch,
&p->dpte_row_bw[k],
&p->meta_row_bw[k]);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, use_one_row_for_frame = %u\n", __func__, k, p->use_one_row_for_frame[k]);
- dml2_printf("DML::%s: k=%u, use_one_row_for_frame_flip = %u\n", __func__, k, p->use_one_row_for_frame_flip[k]);
- dml2_printf("DML::%s: k=%u, UseMALLForPStateChange = %u\n", __func__, k, p->display_cfg->plane_descriptors[k].overrides.legacy_svp_config);
- dml2_printf("DML::%s: k=%u, dpte_row_height_luma = %u\n", __func__, k, p->dpte_row_height_luma[k]);
- dml2_printf("DML::%s: k=%u, dpte_row_width_luma_ub = %u\n", __func__, k, p->dpte_row_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowY = %u (after hvm level)\n", __func__, k, s->PixelPTEBytesPerRowY[k]);
- dml2_printf("DML::%s: k=%u, dpte_row_height_chroma = %u\n", __func__, k, p->dpte_row_height_chroma[k]);
- dml2_printf("DML::%s: k=%u, dpte_row_width_chroma_ub = %u\n", __func__, k, p->dpte_row_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowC = %u (after hvm level)\n", __func__, k, s->PixelPTEBytesPerRowC[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRow = %u\n", __func__, k, p->PixelPTEBytesPerRow[k]);
- dml2_printf("DML::%s: k=%u, PTEBufferSizeNotExceeded = %u\n", __func__, k, p->PTEBufferSizeNotExceeded[k]);
- dml2_printf("DML::%s: k=%u, gpuvm_enable = %u\n", __func__, k, p->display_cfg->gpuvm_enable);
- dml2_printf("DML::%s: k=%u, PTE_BUFFER_MODE = %u\n", __func__, k, p->PTE_BUFFER_MODE[k]);
- dml2_printf("DML::%s: k=%u, BIGK_FRAGMENT_SIZE = %u\n", __func__, k, p->BIGK_FRAGMENT_SIZE[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, use_one_row_for_frame = %u\n", __func__, k, p->use_one_row_for_frame[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, use_one_row_for_frame_flip = %u\n", __func__, k, p->use_one_row_for_frame_flip[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, UseMALLForPStateChange = %u\n", __func__, k, p->display_cfg->plane_descriptors[k].overrides.legacy_svp_config);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_row_height_luma = %u\n", __func__, k, p->dpte_row_height_luma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_row_width_luma_ub = %u\n", __func__, k, p->dpte_row_width_luma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEBytesPerRowY = %u (after hvm level)\n", __func__, k, s->PixelPTEBytesPerRowY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_row_height_chroma = %u\n", __func__, k, p->dpte_row_height_chroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_row_width_chroma_ub = %u\n", __func__, k, p->dpte_row_width_chroma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEBytesPerRowC = %u (after hvm level)\n", __func__, k, s->PixelPTEBytesPerRowC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEBytesPerRow = %u\n", __func__, k, p->PixelPTEBytesPerRow[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PTEBufferSizeNotExceeded = %u\n", __func__, k, p->PTEBufferSizeNotExceeded[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, gpuvm_enable = %u\n", __func__, k, p->display_cfg->gpuvm_enable);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PTE_BUFFER_MODE = %u\n", __func__, k, p->PTE_BUFFER_MODE[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, BIGK_FRAGMENT_SIZE = %u\n", __func__, k, p->BIGK_FRAGMENT_SIZE[k]);
#endif
}
}
@@ -3232,19 +3198,19 @@ static double CalculateUrgentLatency(
}
#ifdef __DML_VBA_DEBUG__
if (qos_type == dml2_qos_param_type_dcn4x) {
- dml2_printf("DML::%s: qos_type = %d\n", __func__, qos_type);
- dml2_printf("DML::%s: urgent_ramp_uclk_cycles = %d\n", __func__, urgent_ramp_uclk_cycles);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, uclk_freq_mhz);
- dml2_printf("DML::%s: umc_urgent_ramp_latency_margin = %f\n", __func__, umc_urgent_ramp_latency_margin);
+ DML_LOG_VERBOSE("DML::%s: qos_type = %d\n", __func__, qos_type);
+ DML_LOG_VERBOSE("DML::%s: urgent_ramp_uclk_cycles = %d\n", __func__, urgent_ramp_uclk_cycles);
+ DML_LOG_VERBOSE("DML::%s: uclk_freq_mhz = %f\n", __func__, uclk_freq_mhz);
+ DML_LOG_VERBOSE("DML::%s: umc_urgent_ramp_latency_margin = %f\n", __func__, umc_urgent_ramp_latency_margin);
} else {
- dml2_printf("DML::%s: UrgentLatencyPixelDataOnly = %f\n", __func__, UrgentLatencyPixelDataOnly);
- dml2_printf("DML::%s: UrgentLatencyPixelMixedWithVMData = %f\n", __func__, UrgentLatencyPixelMixedWithVMData);
- dml2_printf("DML::%s: UrgentLatencyVMDataOnly = %f\n", __func__, UrgentLatencyVMDataOnly);
- dml2_printf("DML::%s: UrgentLatencyAdjustmentFabricClockComponent = %f\n", __func__, UrgentLatencyAdjustmentFabricClockComponent);
- dml2_printf("DML::%s: UrgentLatencyAdjustmentFabricClockReference = %f\n", __func__, UrgentLatencyAdjustmentFabricClockReference);
+ DML_LOG_VERBOSE("DML::%s: UrgentLatencyPixelDataOnly = %f\n", __func__, UrgentLatencyPixelDataOnly);
+ DML_LOG_VERBOSE("DML::%s: UrgentLatencyPixelMixedWithVMData = %f\n", __func__, UrgentLatencyPixelMixedWithVMData);
+ DML_LOG_VERBOSE("DML::%s: UrgentLatencyVMDataOnly = %f\n", __func__, UrgentLatencyVMDataOnly);
+ DML_LOG_VERBOSE("DML::%s: UrgentLatencyAdjustmentFabricClockComponent = %f\n", __func__, UrgentLatencyAdjustmentFabricClockComponent);
+ DML_LOG_VERBOSE("DML::%s: UrgentLatencyAdjustmentFabricClockReference = %f\n", __func__, UrgentLatencyAdjustmentFabricClockReference);
}
- dml2_printf("DML::%s: FabricClock = %f\n", __func__, FabricClock);
- dml2_printf("DML::%s: UrgentLatency = %f\n", __func__, urgent_latency);
+ DML_LOG_VERBOSE("DML::%s: FabricClock = %f\n", __func__, FabricClock);
+ DML_LOG_VERBOSE("DML::%s: UrgentLatency = %f\n", __func__, urgent_latency);
#endif
return urgent_latency;
}
@@ -3271,18 +3237,18 @@ static double CalculateTripToMemory(
#ifdef __DML_VBA_DEBUG__
if (qos_type == dml2_qos_param_type_dcn4x) {
- dml2_printf("DML::%s: qos_type = %d\n", __func__, qos_type);
- dml2_printf("DML::%s: max_round_trip_to_furthest_cs_fclk_cycles = %d\n", __func__, max_round_trip_to_furthest_cs_fclk_cycles);
- dml2_printf("DML::%s: mall_overhead_fclk_cycles = %d\n", __func__, mall_overhead_fclk_cycles);
- dml2_printf("DML::%s: trip_to_memory_uclk_cycles = %d\n", __func__, trip_to_memory_uclk_cycles);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, uclk_freq_mhz);
- dml2_printf("DML::%s: FabricClock = %f\n", __func__, FabricClock);
- dml2_printf("DML::%s: fabric_max_transport_latency_margin = %f\n", __func__, fabric_max_transport_latency_margin);
- dml2_printf("DML::%s: umc_max_latency_margin = %f\n", __func__, umc_max_latency_margin);
+ DML_LOG_VERBOSE("DML::%s: qos_type = %d\n", __func__, qos_type);
+ DML_LOG_VERBOSE("DML::%s: max_round_trip_to_furthest_cs_fclk_cycles = %d\n", __func__, max_round_trip_to_furthest_cs_fclk_cycles);
+ DML_LOG_VERBOSE("DML::%s: mall_overhead_fclk_cycles = %d\n", __func__, mall_overhead_fclk_cycles);
+ DML_LOG_VERBOSE("DML::%s: trip_to_memory_uclk_cycles = %d\n", __func__, trip_to_memory_uclk_cycles);
+ DML_LOG_VERBOSE("DML::%s: uclk_freq_mhz = %f\n", __func__, uclk_freq_mhz);
+ DML_LOG_VERBOSE("DML::%s: FabricClock = %f\n", __func__, FabricClock);
+ DML_LOG_VERBOSE("DML::%s: fabric_max_transport_latency_margin = %f\n", __func__, fabric_max_transport_latency_margin);
+ DML_LOG_VERBOSE("DML::%s: umc_max_latency_margin = %f\n", __func__, umc_max_latency_margin);
} else {
- dml2_printf("DML::%s: UrgLatency = %f\n", __func__, UrgLatency);
+ DML_LOG_VERBOSE("DML::%s: UrgLatency = %f\n", __func__, UrgLatency);
}
- dml2_printf("DML::%s: trip_to_memory_us = %f\n", __func__, trip_to_memory_us);
+ DML_LOG_VERBOSE("DML::%s: trip_to_memory_us = %f\n", __func__, trip_to_memory_us);
#endif
@@ -3309,14 +3275,14 @@ static double CalculateMetaTripToMemory(
#ifdef __DML_VBA_DEBUG__
if (qos_type == dml2_qos_param_type_dcn4x) {
- dml2_printf("DML::%s: qos_type = %d\n", __func__, qos_type);
- dml2_printf("DML::%s: meta_trip_to_memory_fclk_cycles = %d\n", __func__, meta_trip_to_memory_fclk_cycles);
- dml2_printf("DML::%s: meta_trip_to_memory_uclk_cycles = %d\n", __func__, meta_trip_to_memory_uclk_cycles);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, uclk_freq_mhz);
+ DML_LOG_VERBOSE("DML::%s: qos_type = %d\n", __func__, qos_type);
+ DML_LOG_VERBOSE("DML::%s: meta_trip_to_memory_fclk_cycles = %d\n", __func__, meta_trip_to_memory_fclk_cycles);
+ DML_LOG_VERBOSE("DML::%s: meta_trip_to_memory_uclk_cycles = %d\n", __func__, meta_trip_to_memory_uclk_cycles);
+ DML_LOG_VERBOSE("DML::%s: uclk_freq_mhz = %f\n", __func__, uclk_freq_mhz);
} else {
- dml2_printf("DML::%s: UrgLatency = %f\n", __func__, UrgLatency);
+ DML_LOG_VERBOSE("DML::%s: UrgLatency = %f\n", __func__, UrgLatency);
}
- dml2_printf("DML::%s: meta_trip_to_memory_us = %f\n", __func__, meta_trip_to_memory_us);
+ DML_LOG_VERBOSE("DML::%s: meta_trip_to_memory_us = %f\n", __func__, meta_trip_to_memory_us);
#endif
@@ -3333,7 +3299,6 @@ static void calculate_cursor_req_attributes(
unsigned int *cursor_bytes_per_chunk,
unsigned int *cursor_bytes)
{
- unsigned int cursor_pitch = 0;
unsigned int cursor_bytes_per_req = 0;
unsigned int cursor_width_bytes = 0;
unsigned int cursor_height = 0;
@@ -3341,10 +3306,6 @@ static void calculate_cursor_req_attributes(
//SW determines the cursor pitch to support the maximum cursor_width that will be used but the following restrictions apply.
//- For 2bpp, cursor_pitch = 256 pixels due to min cursor request size of 64B
//- For 32 or 64 bpp, cursor_pitch = 64, 128 or 256 pixels depending on the cursor width
- if (cursor_bpp == 2)
- cursor_pitch = 256;
- else
- cursor_pitch = (unsigned int)1 << (unsigned int)math_ceil2(math_log((float)cursor_width, 2), 1);
//The cursor requestor uses a cursor request size of 64B, 128B, or 256B depending on the cursor_width and cursor_bpp as follows.
@@ -3384,8 +3345,8 @@ static void calculate_cursor_req_attributes(
*cursor_lines_per_chunk = 1;
} else {
if (cursor_width > 0) {
- dml2_printf("DML::%s: Invalid cursor_bpp = %d\n", __func__, cursor_bpp);
- dml2_assert(0);
+ DML_LOG_VERBOSE("DML::%s: Invalid cursor_bpp = %d\n", __func__, cursor_bpp);
+ DML_ASSERT(0);
}
}
@@ -3396,15 +3357,15 @@ static void calculate_cursor_req_attributes(
cursor_height = cursor_width;
*cursor_bytes = *cursor_bytes_per_line * cursor_height;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: cursor_bpp = %d\n", __func__, cursor_bpp);
- dml2_printf("DML::%s: cursor_width = %d\n", __func__, cursor_width);
- dml2_printf("DML::%s: cursor_width_bytes = %d\n", __func__, cursor_width_bytes);
- dml2_printf("DML::%s: cursor_bytes_per_req = %d\n", __func__, cursor_bytes_per_req);
- dml2_printf("DML::%s: cursor_lines_per_chunk = %d\n", __func__, *cursor_lines_per_chunk);
- dml2_printf("DML::%s: cursor_bytes_per_line = %d\n", __func__, *cursor_bytes_per_line);
- dml2_printf("DML::%s: cursor_bytes_per_chunk = %d\n", __func__, *cursor_bytes_per_chunk);
- dml2_printf("DML::%s: cursor_bytes = %d\n", __func__, *cursor_bytes);
- dml2_printf("DML::%s: cursor_pitch = %d\n", __func__, cursor_pitch);
+ DML_LOG_VERBOSE("DML::%s: cursor_bpp = %d\n", __func__, cursor_bpp);
+ DML_LOG_VERBOSE("DML::%s: cursor_width = %d\n", __func__, cursor_width);
+ DML_LOG_VERBOSE("DML::%s: cursor_width_bytes = %d\n", __func__, cursor_width_bytes);
+ DML_LOG_VERBOSE("DML::%s: cursor_bytes_per_req = %d\n", __func__, cursor_bytes_per_req);
+ DML_LOG_VERBOSE("DML::%s: cursor_lines_per_chunk = %d\n", __func__, *cursor_lines_per_chunk);
+ DML_LOG_VERBOSE("DML::%s: cursor_bytes_per_line = %d\n", __func__, *cursor_bytes_per_line);
+ DML_LOG_VERBOSE("DML::%s: cursor_bytes_per_chunk = %d\n", __func__, *cursor_bytes_per_chunk);
+ DML_LOG_VERBOSE("DML::%s: cursor_bytes = %d\n", __func__, *cursor_bytes);
+ DML_LOG_VERBOSE("DML::%s: cursor_pitch = %d\n", __func__, cursor_bpp == 2 ? 256 : (unsigned int)1 << (unsigned int)math_ceil2(math_log((float)cursor_width, 2), 1));
#endif
}
@@ -3428,20 +3389,20 @@ static void calculate_cursor_urgent_burst_factor(
CursorBufferSizeInTime = LinesInCursorBuffer * LineTime;
if (CursorBufferSizeInTime - UrgentLatency <= 0) {
*NotEnoughUrgentLatencyHiding = 1;
- *UrgentBurstFactorCursor = 0;
+ *UrgentBurstFactorCursor = 1;
} else {
*NotEnoughUrgentLatencyHiding = 0;
*UrgentBurstFactorCursor = CursorBufferSizeInTime / (CursorBufferSizeInTime - UrgentLatency);
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: LinesInCursorBuffer = %u\n", __func__, LinesInCursorBuffer);
- dml2_printf("DML::%s: CursorBufferSizeInTime = %f\n", __func__, CursorBufferSizeInTime);
- dml2_printf("DML::%s: CursorBufferSize = %u (kbytes)\n", __func__, CursorBufferSize);
- dml2_printf("DML::%s: cursor_bytes_per_chunk = %u\n", __func__, cursor_bytes_per_chunk);
- dml2_printf("DML::%s: cursor_lines_per_chunk = %u\n", __func__, cursor_lines_per_chunk);
- dml2_printf("DML::%s: UrgentBurstFactorCursor = %f\n", __func__, *UrgentBurstFactorCursor);
- dml2_printf("DML::%s: NotEnoughUrgentLatencyHiding = %d\n", __func__, *NotEnoughUrgentLatencyHiding);
+ DML_LOG_VERBOSE("DML::%s: LinesInCursorBuffer = %u\n", __func__, LinesInCursorBuffer);
+ DML_LOG_VERBOSE("DML::%s: CursorBufferSizeInTime = %f\n", __func__, CursorBufferSizeInTime);
+ DML_LOG_VERBOSE("DML::%s: CursorBufferSize = %u (kbytes)\n", __func__, CursorBufferSize);
+ DML_LOG_VERBOSE("DML::%s: cursor_bytes_per_chunk = %u\n", __func__, cursor_bytes_per_chunk);
+ DML_LOG_VERBOSE("DML::%s: cursor_lines_per_chunk = %u\n", __func__, cursor_lines_per_chunk);
+ DML_LOG_VERBOSE("DML::%s: UrgentBurstFactorCursor = %f\n", __func__, *UrgentBurstFactorCursor);
+ DML_LOG_VERBOSE("DML::%s: NotEnoughUrgentLatencyHiding = %d\n", __func__, *NotEnoughUrgentLatencyHiding);
#endif
}
@@ -3476,22 +3437,22 @@ static void CalculateUrgentBurstFactor(
*UrgentBurstFactorChroma = 0;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatio = %f\n", __func__, VRatio);
- dml2_printf("DML::%s: VRatioC = %f\n", __func__, VRatioC);
- dml2_printf("DML::%s: DETBufferSizeY = %d\n", __func__, DETBufferSizeY);
- dml2_printf("DML::%s: DETBufferSizeC = %d\n", __func__, DETBufferSizeC);
- dml2_printf("DML::%s: BytePerPixelInDETY = %f\n", __func__, BytePerPixelInDETY);
- dml2_printf("DML::%s: swath_width_luma_ub = %d\n", __func__, swath_width_luma_ub);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, LineTime);
+ DML_LOG_VERBOSE("DML::%s: VRatio = %f\n", __func__, VRatio);
+ DML_LOG_VERBOSE("DML::%s: VRatioC = %f\n", __func__, VRatioC);
+ DML_LOG_VERBOSE("DML::%s: DETBufferSizeY = %d\n", __func__, DETBufferSizeY);
+ DML_LOG_VERBOSE("DML::%s: DETBufferSizeC = %d\n", __func__, DETBufferSizeC);
+ DML_LOG_VERBOSE("DML::%s: BytePerPixelInDETY = %f\n", __func__, BytePerPixelInDETY);
+ DML_LOG_VERBOSE("DML::%s: swath_width_luma_ub = %d\n", __func__, swath_width_luma_ub);
+ DML_LOG_VERBOSE("DML::%s: LineTime = %f\n", __func__, LineTime);
#endif
- DML2_ASSERT(VRatio > 0);
+ DML_ASSERT(VRatio > 0);
LinesInDETLuma = (dml_is_phantom_pipe(plane_cfg) ? 1024 * 1024 : DETBufferSizeY) / BytePerPixelInDETY / swath_width_luma_ub;
DETBufferSizeInTimeLuma = math_floor2(LinesInDETLuma, SwathHeightY) * LineTime / VRatio;
if (DETBufferSizeInTimeLuma - UrgentLatency <= 0) {
*NotEnoughUrgentLatencyHiding = 1;
- *UrgentBurstFactorLuma = 0;
+ *UrgentBurstFactorLuma = 1;
} else {
*UrgentBurstFactorLuma = DETBufferSizeInTimeLuma / (DETBufferSizeInTimeLuma - UrgentLatency);
}
@@ -3502,24 +3463,23 @@ static void CalculateUrgentBurstFactor(
DETBufferSizeInTimeChroma = math_floor2(LinesInDETChroma, SwathHeightC) * LineTime / VRatioC;
if (DETBufferSizeInTimeChroma - UrgentLatency <= 0) {
*NotEnoughUrgentLatencyHiding = 1;
- *UrgentBurstFactorChroma = 0;
+ *UrgentBurstFactorChroma = 1;
} else {
*UrgentBurstFactorChroma = DETBufferSizeInTimeChroma / (DETBufferSizeInTimeChroma - UrgentLatency);
}
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: LinesInDETLuma = %f\n", __func__, LinesInDETLuma);
- dml2_printf("DML::%s: UrgentLatency = %f\n", __func__, UrgentLatency);
- dml2_printf("DML::%s: DETBufferSizeInTimeLuma = %f\n", __func__, DETBufferSizeInTimeLuma);
- dml2_printf("DML::%s: UrgentBurstFactorLuma = %f\n", __func__, *UrgentBurstFactorLuma);
- dml2_printf("DML::%s: UrgentBurstFactorChroma = %f\n", __func__, *UrgentBurstFactorChroma);
- dml2_printf("DML::%s: NotEnoughUrgentLatencyHiding = %d\n", __func__, *NotEnoughUrgentLatencyHiding);
+ DML_LOG_VERBOSE("DML::%s: LinesInDETLuma = %f\n", __func__, LinesInDETLuma);
+ DML_LOG_VERBOSE("DML::%s: UrgentLatency = %f\n", __func__, UrgentLatency);
+ DML_LOG_VERBOSE("DML::%s: DETBufferSizeInTimeLuma = %f\n", __func__, DETBufferSizeInTimeLuma);
+ DML_LOG_VERBOSE("DML::%s: UrgentBurstFactorLuma = %f\n", __func__, *UrgentBurstFactorLuma);
+ DML_LOG_VERBOSE("DML::%s: UrgentBurstFactorChroma = %f\n", __func__, *UrgentBurstFactorChroma);
+ DML_LOG_VERBOSE("DML::%s: NotEnoughUrgentLatencyHiding = %d\n", __func__, *NotEnoughUrgentLatencyHiding);
#endif
-
}
-static void CalculateDCFCLKDeepSleep(
+static void CalculateDCFCLKDeepSleepTdlut(
const struct dml2_display_cfg *display_cfg,
unsigned int NumberOfActiveSurfaces,
unsigned int BytePerPixelY[],
@@ -3534,6 +3494,10 @@ static void CalculateDCFCLKDeepSleep(
double ReadBandwidthChroma[],
unsigned int ReturnBusWidth,
+ double dispclk,
+ unsigned int tdlut_bytes_to_deliver[],
+ double prefetch_swath_time_us[],
+
// Output
double *DCFClkDeepSleep)
{
@@ -3568,9 +3532,25 @@ static void CalculateDCFCLKDeepSleep(
}
DCFClkDeepSleepPerSurface[k] = math_max2(DCFClkDeepSleepPerSurface[k], pixel_rate_mhz / 16);
+ // adjust for 3dlut delivery time
+ if (display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut && tdlut_bytes_to_deliver[k] > 0) {
+ double tdlut_required_deepsleep_dcfclk = (double) tdlut_bytes_to_deliver[k] / 64.0 / prefetch_swath_time_us[k];
+
+ DML_LOG_VERBOSE("DML::%s: k=%d, DCFClkDeepSleepPerSurface = %f\n", __func__, k, DCFClkDeepSleepPerSurface[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, tdlut_bytes_to_deliver = %d\n", __func__, k, tdlut_bytes_to_deliver[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, prefetch_swath_time_us = %f\n", __func__, k, prefetch_swath_time_us[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, tdlut_required_deepsleep_dcfclk = %f\n", __func__, k, tdlut_required_deepsleep_dcfclk);
+
+ // increase the deepsleep dcfclk to match the original dispclk throughput rate
+ if (tdlut_required_deepsleep_dcfclk > DCFClkDeepSleepPerSurface[k]) {
+ DCFClkDeepSleepPerSurface[k] = math_max2(DCFClkDeepSleepPerSurface[k], tdlut_required_deepsleep_dcfclk);
+ DCFClkDeepSleepPerSurface[k] = math_max2(DCFClkDeepSleepPerSurface[k], dispclk / 4.0);
+ }
+ }
+
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, PixelClock = %f\n", __func__, k, pixel_rate_mhz);
- dml2_printf("DML::%s: k=%u, DCFClkDeepSleepPerSurface = %f\n", __func__, k, DCFClkDeepSleepPerSurface[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelClock = %f\n", __func__, k, pixel_rate_mhz);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DCFClkDeepSleepPerSurface = %f\n", __func__, k, DCFClkDeepSleepPerSurface[k]);
#endif
}
@@ -3581,16 +3561,63 @@ static void CalculateDCFCLKDeepSleep(
*DCFClkDeepSleep = math_max2(8.0, __DML2_CALCS_DCFCLK_FACTOR__ * ReadBandwidth / (double)ReturnBusWidth);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: __DML2_CALCS_DCFCLK_FACTOR__ = %f\n", __func__, __DML2_CALCS_DCFCLK_FACTOR__);
- dml2_printf("DML::%s: ReadBandwidth = %f\n", __func__, ReadBandwidth);
- dml2_printf("DML::%s: ReturnBusWidth = %u\n", __func__, ReturnBusWidth);
- dml2_printf("DML::%s: DCFClkDeepSleep = %f\n", __func__, *DCFClkDeepSleep);
+ DML_LOG_VERBOSE("DML::%s: __DML2_CALCS_DCFCLK_FACTOR__ = %f\n", __func__, __DML2_CALCS_DCFCLK_FACTOR__);
+ DML_LOG_VERBOSE("DML::%s: ReadBandwidth = %f\n", __func__, ReadBandwidth);
+ DML_LOG_VERBOSE("DML::%s: ReturnBusWidth = %u\n", __func__, ReturnBusWidth);
+ DML_LOG_VERBOSE("DML::%s: DCFClkDeepSleep = %f\n", __func__, *DCFClkDeepSleep);
#endif
for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
*DCFClkDeepSleep = math_max2(*DCFClkDeepSleep, DCFClkDeepSleepPerSurface[k]);
}
- dml2_printf("DML::%s: DCFClkDeepSleep = %f (final)\n", __func__, *DCFClkDeepSleep);
+
+ DML_LOG_VERBOSE("DML::%s: DCFClkDeepSleep = %f (final)\n", __func__, *DCFClkDeepSleep);
+}
+
+static noinline_for_stack void CalculateDCFCLKDeepSleep(
+ const struct dml2_display_cfg *display_cfg,
+ unsigned int NumberOfActiveSurfaces,
+ unsigned int BytePerPixelY[],
+ unsigned int BytePerPixelC[],
+ unsigned int SwathWidthY[],
+ unsigned int SwathWidthC[],
+ unsigned int DPPPerSurface[],
+ double PSCL_THROUGHPUT[],
+ double PSCL_THROUGHPUT_CHROMA[],
+ double Dppclk[],
+ double ReadBandwidthLuma[],
+ double ReadBandwidthChroma[],
+ unsigned int ReturnBusWidth,
+
+ // Output
+ double *DCFClkDeepSleep)
+{
+ double zero_double[DML2_MAX_PLANES];
+ unsigned int zero_integer[DML2_MAX_PLANES];
+
+ memset(zero_double, 0, DML2_MAX_PLANES * sizeof(double));
+ memset(zero_integer, 0, DML2_MAX_PLANES * sizeof(unsigned int));
+
+ CalculateDCFCLKDeepSleepTdlut(
+ display_cfg,
+ NumberOfActiveSurfaces,
+ BytePerPixelY,
+ BytePerPixelC,
+ SwathWidthY,
+ SwathWidthC,
+ DPPPerSurface,
+ PSCL_THROUGHPUT,
+ PSCL_THROUGHPUT_CHROMA,
+ Dppclk,
+ ReadBandwidthLuma,
+ ReadBandwidthChroma,
+ ReturnBusWidth,
+ 0,
+ zero_integer, //tdlut_bytes_to_deliver,
+ zero_double, //prefetch_swath_time_us,
+
+ // Output
+ DCFClkDeepSleep);
}
static double CalculateWriteBackDelay(
@@ -3631,23 +3658,23 @@ static unsigned int CalculateMaxVStartup(
double line_time_us = (double)timing->h_total / ((double)timing->pixel_clock_khz / 1000);
unsigned int vblank_actual = timing->v_total - timing->v_active;
unsigned int vblank_nom_default_in_line = (unsigned int)math_floor2((double)vblank_nom_default_us / line_time_us, 1.0);
- unsigned int vblank_nom_input = (unsigned int)math_min2(timing->vblank_nom, vblank_nom_default_in_line);
- unsigned int vblank_avail = (vblank_nom_input == 0) ? vblank_nom_default_in_line : vblank_nom_input;
+ unsigned int vblank_avail = (timing->vblank_nom == 0) ? vblank_nom_default_in_line : (unsigned int)timing->vblank_nom;
vblank_size = (unsigned int)math_min2(vblank_actual, vblank_avail);
if (timing->interlaced && !ptoi_supported)
- max_vstartup_lines = (unsigned int)(math_floor2(vblank_size / 2.0, 1.0));
+ max_vstartup_lines = (unsigned int)(math_floor2((vblank_size - 1) / 2.0, 1.0));
else
max_vstartup_lines = vblank_size - (unsigned int)math_max2(1.0, math_ceil2(write_back_delay_us / line_time_us, 1.0));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VBlankNom = %u\n", __func__, timing->vblank_nom);
- dml2_printf("DML::%s: vblank_nom_default_us = %u\n", __func__, vblank_nom_default_us);
- dml2_printf("DML::%s: line_time_us = %f\n", __func__, line_time_us);
- dml2_printf("DML::%s: vblank_actual = %u\n", __func__, vblank_actual);
- dml2_printf("DML::%s: vblank_avail = %u\n", __func__, vblank_avail);
- dml2_printf("DML::%s: max_vstartup_lines = %u\n", __func__, max_vstartup_lines);
+ DML_LOG_VERBOSE("DML::%s: VBlankNom = %lu\n", __func__, timing->vblank_nom);
+ DML_LOG_VERBOSE("DML::%s: vblank_nom_default_us = %u\n", __func__, vblank_nom_default_us);
+ DML_LOG_VERBOSE("DML::%s: line_time_us = %f\n", __func__, line_time_us);
+ DML_LOG_VERBOSE("DML::%s: vblank_actual = %u\n", __func__, vblank_actual);
+ DML_LOG_VERBOSE("DML::%s: vblank_avail = %u\n", __func__, vblank_avail);
+ DML_LOG_VERBOSE("DML::%s: max_vstartup_lines = %u\n", __func__, max_vstartup_lines);
#endif
+ max_vstartup_lines = (unsigned int)math_min2(max_vstartup_lines, DML_MAX_VSTARTUP_START);
return max_vstartup_lines;
}
@@ -3670,9 +3697,9 @@ static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch
const long MAXIMUMCOMPRESSION = 4;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ForceSingleDPP = %u\n", __func__, p->ForceSingleDPP);
+ DML_LOG_VERBOSE("DML::%s: ForceSingleDPP = %u\n", __func__, p->ForceSingleDPP);
for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- dml2_printf("DML::%s: DPPPerSurface[%u] = %u\n", __func__, k, p->DPPPerSurface[k]);
+ DML_LOG_VERBOSE("DML::%s: DPPPerSurface[%u] = %u\n", __func__, k, p->DPPPerSurface[k]);
}
#endif
CalculateSwathWidth(
@@ -3706,15 +3733,15 @@ static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch
p->full_swath_bytes_l[k] = (unsigned int)(p->swath_width_luma_ub[k] * p->BytePerPixDETY[k] * MaximumSwathHeightY[k]);
p->full_swath_bytes_c[k] = (unsigned int)(p->swath_width_chroma_ub[k] * p->BytePerPixDETC[k] * MaximumSwathHeightC[k]);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u DPPPerSurface = %u\n", __func__, k, p->DPPPerSurface[k]);
- dml2_printf("DML::%s: k=%u swath_width_luma_ub = %u\n", __func__, k, p->swath_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u BytePerPixDETY = %f\n", __func__, k, p->BytePerPixDETY[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathHeightY = %u\n", __func__, k, MaximumSwathHeightY[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_l = %u\n", __func__, k, p->full_swath_bytes_l[k]);
- dml2_printf("DML::%s: k=%u swath_width_chroma_ub = %u\n", __func__, k, p->swath_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u BytePerPixDETC = %f\n", __func__, k, p->BytePerPixDETC[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathHeightC = %u\n", __func__, k, MaximumSwathHeightC[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_c = %u\n", __func__, k, p->full_swath_bytes_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u DPPPerSurface = %u\n", __func__, k, p->DPPPerSurface[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u swath_width_luma_ub = %u\n", __func__, k, p->swath_width_luma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u BytePerPixDETY = %f\n", __func__, k, p->BytePerPixDETY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathHeightY = %u\n", __func__, k, MaximumSwathHeightY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u full_swath_bytes_l = %u\n", __func__, k, p->full_swath_bytes_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u swath_width_chroma_ub = %u\n", __func__, k, p->swath_width_chroma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u BytePerPixDETC = %f\n", __func__, k, p->BytePerPixDETC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathHeightC = %u\n", __func__, k, MaximumSwathHeightC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u full_swath_bytes_c = %u\n", __func__, k, p->full_swath_bytes_c[k]);
#endif
if (p->display_cfg->plane_descriptors[k].pixel_format == dml2_420_10) {
p->full_swath_bytes_l[k] = (unsigned int)(math_ceil2((double)p->full_swath_bytes_l[k], 256));
@@ -3757,11 +3784,11 @@ static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch
p->CompressedBufferSizeInkByte);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: TotalActiveDPP = %u\n", __func__, TotalActiveDPP);
- dml2_printf("DML::%s: nomDETInKByte = %u\n", __func__, p->nomDETInKByte);
- dml2_printf("DML::%s: ConfigReturnBufferSizeInKByte = %u\n", __func__, p->ConfigReturnBufferSizeInKByte);
- dml2_printf("DML::%s: UnboundedRequestEnabled = %u\n", __func__, *p->UnboundedRequestEnabled);
- dml2_printf("DML::%s: CompressedBufferSizeInkByte = %u\n", __func__, *p->CompressedBufferSizeInkByte);
+ DML_LOG_VERBOSE("DML::%s: TotalActiveDPP = %u\n", __func__, TotalActiveDPP);
+ DML_LOG_VERBOSE("DML::%s: nomDETInKByte = %u\n", __func__, p->nomDETInKByte);
+ DML_LOG_VERBOSE("DML::%s: ConfigReturnBufferSizeInKByte = %u\n", __func__, p->ConfigReturnBufferSizeInKByte);
+ DML_LOG_VERBOSE("DML::%s: UnboundedRequestEnabled = %u\n", __func__, *p->UnboundedRequestEnabled);
+ DML_LOG_VERBOSE("DML::%s: CompressedBufferSizeInkByte = %u\n", __func__, *p->CompressedBufferSizeInkByte);
#endif
*p->ViewportSizeSupport = true;
@@ -3769,7 +3796,7 @@ static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch
DETBufferSizeInKByteForSwathCalculation = (dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k]) ? 1024 : p->DETBufferSizeInKByte[k]);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByteForSwathCalculation = %u\n", __func__, k, DETBufferSizeInKByteForSwathCalculation);
+ DML_LOG_VERBOSE("DML::%s: k=%u DETBufferSizeInKByteForSwathCalculation = %u\n", __func__, k, DETBufferSizeInKByteForSwathCalculation);
#endif
if (p->display_cfg->plane_descriptors[k].surface.tiling == dml2_sw_linear) {
p->SwathHeightY[k] = MaximumSwathHeightY[k];
@@ -3826,13 +3853,13 @@ static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch
if ((p->full_swath_bytes_l[k] / 2 + p->full_swath_bytes_c[k] / 2 > DETBufferSizeInKByteForSwathCalculation * 1024 / 2) ||
p->SwathWidth[k] > p->MaximumSwathWidthLuma[k] || (p->SwathHeightC[k] > 0 && p->SwathWidthChroma[k] > p->MaximumSwathWidthChroma[k])) {
*p->ViewportSizeSupport = false;
- dml2_printf("DML::%s: k=%u full_swath_bytes_l=%u\n", __func__, k, p->full_swath_bytes_l[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_c=%u\n", __func__, k, p->full_swath_bytes_c[k]);
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByteForSwathCalculation=%u\n", __func__, k, DETBufferSizeInKByteForSwathCalculation);
- dml2_printf("DML::%s: k=%u SwathWidth=%u\n", __func__, k, p->SwathWidth[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathWidthLuma=%f\n", __func__, k, p->MaximumSwathWidthLuma[k]);
- dml2_printf("DML::%s: k=%u SwathWidthChroma=%d\n", __func__, k, p->SwathWidthChroma[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathWidthChroma=%f\n", __func__, k, p->MaximumSwathWidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u full_swath_bytes_l=%u\n", __func__, k, p->full_swath_bytes_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u full_swath_bytes_c=%u\n", __func__, k, p->full_swath_bytes_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u DETBufferSizeInKByteForSwathCalculation=%u\n", __func__, k, DETBufferSizeInKByteForSwathCalculation);
+ DML_LOG_VERBOSE("DML::%s: k=%u SwathWidth=%u\n", __func__, k, p->SwathWidth[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathWidthLuma=%f\n", __func__, k, p->MaximumSwathWidthLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u SwathWidthChroma=%d\n", __func__, k, p->SwathWidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathWidthChroma=%f\n", __func__, k, p->MaximumSwathWidthChroma[k]);
p->ViewportSizeSupportPerSurface[k] = false;
} else {
p->ViewportSizeSupportPerSurface[k] = true;
@@ -3840,35 +3867,35 @@ static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch
if (p->SwathHeightC[k] == 0) {
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, All DET will be used for plane0\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: k=%u, All DET will be used for plane0\n", __func__, k);
#endif
p->DETBufferSizeY[k] = p->DETBufferSizeInKByte[k] * 1024;
p->DETBufferSizeC[k] = 0;
} else if (RoundedUpSwathSizeBytesY[k] <= 1.5 * RoundedUpSwathSizeBytesC[k]) {
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, Half DET will be used for plane0, and half for plane1\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: k=%u, Half DET will be used for plane0, and half for plane1\n", __func__, k);
#endif
p->DETBufferSizeY[k] = p->DETBufferSizeInKByte[k] * 1024 / 2;
p->DETBufferSizeC[k] = p->DETBufferSizeInKByte[k] * 1024 / 2;
} else {
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, 2/3 DET will be used for plane0, and 1/3 for plane1\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: k=%u, 2/3 DET will be used for plane0, and 1/3 for plane1\n", __func__, k);
#endif
p->DETBufferSizeY[k] = (unsigned int)(math_floor2(p->DETBufferSizeInKByte[k] * 1024 * 2 / 3, 1024));
p->DETBufferSizeC[k] = p->DETBufferSizeInKByte[k] * 1024 - p->DETBufferSizeY[k];
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u SwathHeightY = %u\n", __func__, k, p->SwathHeightY[k]);
- dml2_printf("DML::%s: k=%u SwathHeightC = %u\n", __func__, k, p->SwathHeightC[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_l = %u\n", __func__, k, p->full_swath_bytes_l[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_c = %u\n", __func__, k, p->full_swath_bytes_c[k]);
- dml2_printf("DML::%s: k=%u RoundedUpSwathSizeBytesY = %u\n", __func__, k, RoundedUpSwathSizeBytesY[k]);
- dml2_printf("DML::%s: k=%u RoundedUpSwathSizeBytesC = %u\n", __func__, k, RoundedUpSwathSizeBytesC[k]);
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByte = %u\n", __func__, k, p->DETBufferSizeInKByte[k]);
- dml2_printf("DML::%s: k=%u DETBufferSizeY = %u\n", __func__, k, p->DETBufferSizeY[k]);
- dml2_printf("DML::%s: k=%u DETBufferSizeC = %u\n", __func__, k, p->DETBufferSizeC[k]);
- dml2_printf("DML::%s: k=%u ViewportSizeSupportPerSurface = %u\n", __func__, k, p->ViewportSizeSupportPerSurface[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u SwathHeightY = %u\n", __func__, k, p->SwathHeightY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u SwathHeightC = %u\n", __func__, k, p->SwathHeightC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u full_swath_bytes_l = %u\n", __func__, k, p->full_swath_bytes_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u full_swath_bytes_c = %u\n", __func__, k, p->full_swath_bytes_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u RoundedUpSwathSizeBytesY = %u\n", __func__, k, RoundedUpSwathSizeBytesY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u RoundedUpSwathSizeBytesC = %u\n", __func__, k, RoundedUpSwathSizeBytesC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u DETBufferSizeInKByte = %u\n", __func__, k, p->DETBufferSizeInKByte[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u DETBufferSizeY = %u\n", __func__, k, p->DETBufferSizeY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u DETBufferSizeC = %u\n", __func__, k, p->DETBufferSizeC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u ViewportSizeSupportPerSurface = %u\n", __func__, k, p->ViewportSizeSupportPerSurface[k]);
#endif
}
@@ -3878,12 +3905,12 @@ static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch
*p->compbuf_reserved_space_64b = (unsigned int)math_ceil2(math_max2(*p->compbuf_reserved_space_64b,
(double)(p->rob_buffer_size_kbytes * 1024 / 64) - (double)(RoundedUpSwathSizeBytesY[SurfaceDoingUnboundedRequest] * TTUFIFODEPTH / (p->mrq_present ? MAXIMUMCOMPRESSION : 1) / 64)), 1.0);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: RoundedUpSwathSizeBytesY[%d] = %u\n", __func__, SurfaceDoingUnboundedRequest, RoundedUpSwathSizeBytesY[SurfaceDoingUnboundedRequest]);
- dml2_printf("DML::%s: rob_buffer_size_kbytes = %u\n", __func__, p->rob_buffer_size_kbytes);
+ DML_LOG_VERBOSE("DML::%s: RoundedUpSwathSizeBytesY[%d] = %u\n", __func__, SurfaceDoingUnboundedRequest, RoundedUpSwathSizeBytesY[SurfaceDoingUnboundedRequest]);
+ DML_LOG_VERBOSE("DML::%s: rob_buffer_size_kbytes = %u\n", __func__, p->rob_buffer_size_kbytes);
#endif
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: compbuf_reserved_space_64b = %u\n", __func__, *p->compbuf_reserved_space_64b);
+ DML_LOG_VERBOSE("DML::%s: compbuf_reserved_space_64b = %u\n", __func__, *p->compbuf_reserved_space_64b);
#endif
*p->hw_debug5 = false;
@@ -3898,12 +3925,12 @@ static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch
+ *p->CompressedBufferSizeInkByte * MAXIMUMCOMPRESSION * 1024) > TTUFIFODEPTH * (RoundedUpSwathSizeBytesY[k] + RoundedUpSwathSizeBytesC[k])))
*p->hw_debug5 = true;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u UnboundedRequestEnabled = %u\n", __func__, k, *p->UnboundedRequestEnabled);
- dml2_printf("DML::%s: k=%u MAXIMUMCOMPRESSION = %lu\n", __func__, k, MAXIMUMCOMPRESSION);
- dml2_printf("DML::%s: k=%u TTUFIFODEPTH = %lu\n", __func__, k, TTUFIFODEPTH);
- dml2_printf("DML::%s: k=%u CompressedBufferSizeInkByte = %u\n", __func__, k, *p->CompressedBufferSizeInkByte);
- dml2_printf("DML::%s: k=%u RoundedUpSwathSizeBytesC = %u\n", __func__, k, RoundedUpSwathSizeBytesC[k]);
- dml2_printf("DML::%s: k=%u hw_debug5 = %u\n", __func__, k, *p->hw_debug5);
+ DML_LOG_VERBOSE("DML::%s: k=%u UnboundedRequestEnabled = %u\n", __func__, k, *p->UnboundedRequestEnabled);
+ DML_LOG_VERBOSE("DML::%s: k=%u MAXIMUMCOMPRESSION = %lu\n", __func__, k, MAXIMUMCOMPRESSION);
+ DML_LOG_VERBOSE("DML::%s: k=%u TTUFIFODEPTH = %lu\n", __func__, k, TTUFIFODEPTH);
+ DML_LOG_VERBOSE("DML::%s: k=%u CompressedBufferSizeInkByte = %u\n", __func__, k, *p->CompressedBufferSizeInkByte);
+ DML_LOG_VERBOSE("DML::%s: k=%u RoundedUpSwathSizeBytesC = %u\n", __func__, k, RoundedUpSwathSizeBytesC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u hw_debug5 = %u\n", __func__, k, *p->hw_debug5);
#endif
}
#endif
@@ -4064,7 +4091,7 @@ static bool ValidateODMMode(enum dml2_odm_mode ODMMode,
return true;
}
-static void CalculateODMMode(
+static noinline_for_stack void CalculateODMMode(
unsigned int MaximumPixelsPerLinePerDSCUnit,
unsigned int HActive,
enum dml2_output_format_class OutFormat,
@@ -4101,15 +4128,15 @@ static void CalculateODMMode(
SurfaceRequiredDISPCLKWithODMCombineThreeToOne = CalculateRequiredDispclk(dml2_odm_mode_combine_3to1, PixelClock);
SurfaceRequiredDISPCLKWithODMCombineFourToOne = CalculateRequiredDispclk(dml2_odm_mode_combine_4to1, PixelClock);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ODMUse = %d\n", __func__, ODMUse);
- dml2_printf("DML::%s: Output = %d\n", __func__, Output);
- dml2_printf("DML::%s: DSCEnable = %d\n", __func__, DSCEnable);
- dml2_printf("DML::%s: MaxDispclk = %f\n", __func__, MaxDispclk);
- dml2_printf("DML::%s: MaximumPixelsPerLinePerDSCUnit = %d\n", __func__, MaximumPixelsPerLinePerDSCUnit);
- dml2_printf("DML::%s: SurfaceRequiredDISPCLKWithoutODMCombine = %f\n", __func__, SurfaceRequiredDISPCLKWithoutODMCombine);
- dml2_printf("DML::%s: SurfaceRequiredDISPCLKWithODMCombineTwoToOne = %f\n", __func__, SurfaceRequiredDISPCLKWithODMCombineTwoToOne);
- dml2_printf("DML::%s: SurfaceRequiredDISPCLKWithODMCombineThreeToOne = %f\n", __func__, SurfaceRequiredDISPCLKWithODMCombineThreeToOne);
- dml2_printf("DML::%s: SurfaceRequiredDISPCLKWithODMCombineFourToOne = %f\n", __func__, SurfaceRequiredDISPCLKWithODMCombineFourToOne);
+ DML_LOG_VERBOSE("DML::%s: ODMUse = %d\n", __func__, ODMUse);
+ DML_LOG_VERBOSE("DML::%s: Output = %d\n", __func__, Output);
+ DML_LOG_VERBOSE("DML::%s: DSCEnable = %d\n", __func__, DSCEnable);
+ DML_LOG_VERBOSE("DML::%s: MaxDispclk = %f\n", __func__, MaxDispclk);
+ DML_LOG_VERBOSE("DML::%s: MaximumPixelsPerLinePerDSCUnit = %d\n", __func__, MaximumPixelsPerLinePerDSCUnit);
+ DML_LOG_VERBOSE("DML::%s: SurfaceRequiredDISPCLKWithoutODMCombine = %f\n", __func__, SurfaceRequiredDISPCLKWithoutODMCombine);
+ DML_LOG_VERBOSE("DML::%s: SurfaceRequiredDISPCLKWithODMCombineTwoToOne = %f\n", __func__, SurfaceRequiredDISPCLKWithODMCombineTwoToOne);
+ DML_LOG_VERBOSE("DML::%s: SurfaceRequiredDISPCLKWithODMCombineThreeToOne = %f\n", __func__, SurfaceRequiredDISPCLKWithODMCombineThreeToOne);
+ DML_LOG_VERBOSE("DML::%s: SurfaceRequiredDISPCLKWithODMCombineFourToOne = %f\n", __func__, SurfaceRequiredDISPCLKWithODMCombineFourToOne);
#endif
if (ODMUse == dml2_odm_mode_auto)
DecidedODMMode = DecideODMMode(HActive,
@@ -4154,14 +4181,14 @@ static void CalculateODMMode(
*NumberOfDPP = NumberOfDPPRequired;
*RequiredDISPCLKPerSurface = success ? DISPCLKRequired : 0;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ODMMode = %d\n", __func__, *ODMMode);
- dml2_printf("DML::%s: NumberOfDPP = %d\n", __func__, *NumberOfDPP);
- dml2_printf("DML::%s: TotalAvailablePipesSupport = %d\n", __func__, *TotalAvailablePipesSupport);
- dml2_printf("DML::%s: RequiredDISPCLKPerSurface = %f\n", __func__, *RequiredDISPCLKPerSurface);
+ DML_LOG_VERBOSE("DML::%s: ODMMode = %d\n", __func__, *ODMMode);
+ DML_LOG_VERBOSE("DML::%s: NumberOfDPP = %d\n", __func__, *NumberOfDPP);
+ DML_LOG_VERBOSE("DML::%s: TotalAvailablePipesSupport = %d\n", __func__, *TotalAvailablePipesSupport);
+ DML_LOG_VERBOSE("DML::%s: RequiredDISPCLKPerSurface = %f\n", __func__, *RequiredDISPCLKPerSurface);
#endif
}
-static void CalculateOutputLink(
+static noinline_for_stack void CalculateOutputLink(
struct dml2_core_internal_scratch *s,
double PHYCLK,
double PHYCLKD18,
@@ -4201,17 +4228,17 @@ static void CalculateOutputLink(
*OutputRate = dml2_core_internal_output_rate_unknown;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DSCEnable = %u (dis, en, en_if_necessary)\n", __func__, DSCEnable);
- dml2_printf("DML::%s: PHYCLK = %f\n", __func__, PHYCLK);
- dml2_printf("DML::%s: PixelClockBackEnd = %f\n", __func__, PixelClockBackEnd);
- dml2_printf("DML::%s: AudioSampleRate = %f\n", __func__, AudioSampleRate);
- dml2_printf("DML::%s: HActive = %u\n", __func__, HActive);
- dml2_printf("DML::%s: HTotal = %u\n", __func__, HTotal);
- dml2_printf("DML::%s: ODMModeNoDSC = %u\n", __func__, ODMModeNoDSC);
- dml2_printf("DML::%s: ODMModeDSC = %u\n", __func__, ODMModeDSC);
- dml2_printf("DML::%s: ForcedOutputLinkBPP = %f\n", __func__, ForcedOutputLinkBPP);
- dml2_printf("DML::%s: Output (encoder) = %u\n", __func__, Output);
- dml2_printf("DML::%s: OutputLinkDPRate = %u\n", __func__, OutputLinkDPRate);
+ DML_LOG_VERBOSE("DML::%s: DSCEnable = %u (dis, en, en_if_necessary)\n", __func__, DSCEnable);
+ DML_LOG_VERBOSE("DML::%s: PHYCLK = %f\n", __func__, PHYCLK);
+ DML_LOG_VERBOSE("DML::%s: PixelClockBackEnd = %f\n", __func__, PixelClockBackEnd);
+ DML_LOG_VERBOSE("DML::%s: AudioSampleRate = %f\n", __func__, AudioSampleRate);
+ DML_LOG_VERBOSE("DML::%s: HActive = %u\n", __func__, HActive);
+ DML_LOG_VERBOSE("DML::%s: HTotal = %u\n", __func__, HTotal);
+ DML_LOG_VERBOSE("DML::%s: ODMModeNoDSC = %u\n", __func__, ODMModeNoDSC);
+ DML_LOG_VERBOSE("DML::%s: ODMModeDSC = %u\n", __func__, ODMModeDSC);
+ DML_LOG_VERBOSE("DML::%s: ForcedOutputLinkBPP = %f\n", __func__, ForcedOutputLinkBPP);
+ DML_LOG_VERBOSE("DML::%s: Output (encoder) = %u\n", __func__, Output);
+ DML_LOG_VERBOSE("DML::%s: OutputLinkDPRate = %u\n", __func__, OutputLinkDPRate);
#endif
{
if (Output == dml2_hdmi) {
@@ -4396,9 +4423,9 @@ static void CalculateOutputLink(
}
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: RequiresDSC = %u\n", __func__, *RequiresDSC);
- dml2_printf("DML::%s: RequiresFEC = %u\n", __func__, *RequiresFEC);
- dml2_printf("DML::%s: OutBpp = %f\n", __func__, *OutBpp);
+ DML_LOG_VERBOSE("DML::%s: RequiresDSC = %u\n", __func__, *RequiresDSC);
+ DML_LOG_VERBOSE("DML::%s: RequiresFEC = %u\n", __func__, *RequiresFEC);
+ DML_LOG_VERBOSE("DML::%s: OutBpp = %f\n", __func__, *OutBpp);
#endif
}
@@ -4480,17 +4507,17 @@ static unsigned int DSCDelayRequirement(
DSCDelayRequirement_val = 0;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DSCEnabled= %u\n", __func__, DSCEnabled);
- dml2_printf("DML::%s: ODMMode = %u\n", __func__, ODMMode);
- dml2_printf("DML::%s: OutputBpp = %f\n", __func__, OutputBpp);
- dml2_printf("DML::%s: HActive = %u\n", __func__, HActive);
- dml2_printf("DML::%s: HTotal = %u\n", __func__, HTotal);
- dml2_printf("DML::%s: PixelClock = %f\n", __func__, PixelClock);
- dml2_printf("DML::%s: PixelClockBackEnd = %f\n", __func__, PixelClockBackEnd);
- dml2_printf("DML::%s: OutputFormat = %u\n", __func__, OutputFormat);
- dml2_printf("DML::%s: DSCInputBitPerComponent = %u\n", __func__, DSCInputBitPerComponent);
- dml2_printf("DML::%s: NumberOfDSCSlices = %u\n", __func__, NumberOfDSCSlices);
- dml2_printf("DML::%s: DSCDelayRequirement_val = %u\n", __func__, DSCDelayRequirement_val);
+ DML_LOG_VERBOSE("DML::%s: DSCEnabled= %u\n", __func__, DSCEnabled);
+ DML_LOG_VERBOSE("DML::%s: ODMMode = %u\n", __func__, ODMMode);
+ DML_LOG_VERBOSE("DML::%s: OutputBpp = %f\n", __func__, OutputBpp);
+ DML_LOG_VERBOSE("DML::%s: HActive = %u\n", __func__, HActive);
+ DML_LOG_VERBOSE("DML::%s: HTotal = %u\n", __func__, HTotal);
+ DML_LOG_VERBOSE("DML::%s: PixelClock = %f\n", __func__, PixelClock);
+ DML_LOG_VERBOSE("DML::%s: PixelClockBackEnd = %f\n", __func__, PixelClockBackEnd);
+ DML_LOG_VERBOSE("DML::%s: OutputFormat = %u\n", __func__, OutputFormat);
+ DML_LOG_VERBOSE("DML::%s: DSCInputBitPerComponent = %u\n", __func__, DSCInputBitPerComponent);
+ DML_LOG_VERBOSE("DML::%s: NumberOfDSCSlices = %u\n", __func__, NumberOfDSCSlices);
+ DML_LOG_VERBOSE("DML::%s: DSCDelayRequirement_val = %u\n", __func__, DSCDelayRequirement_val);
#endif
return DSCDelayRequirement_val;
@@ -4563,10 +4590,10 @@ static void CalculateSurfaceSizeInMall(
(TotalSurfaceSizeInMALLForSubVP > MALLAllocatedForDCNInBytes);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: MALLAllocatedForDCN = %u\n", __func__, MALLAllocatedForDCN * 1024 * 1024);
- dml2_printf("DML::%s: TotalSurfaceSizeInMALLForSubVP = %u\n", __func__, TotalSurfaceSizeInMALLForSubVP);
- dml2_printf("DML::%s: TotalSurfaceSizeInMALLForSS = %u\n", __func__, TotalSurfaceSizeInMALLForSS);
- dml2_printf("DML::%s: ExceededMALLSize = %u\n", __func__, *ExceededMALLSize);
+ DML_LOG_VERBOSE("DML::%s: MALLAllocatedForDCN = %u\n", __func__, MALLAllocatedForDCN * 1024 * 1024);
+ DML_LOG_VERBOSE("DML::%s: TotalSurfaceSizeInMALLForSubVP = %u\n", __func__, TotalSurfaceSizeInMALLForSubVP);
+ DML_LOG_VERBOSE("DML::%s: TotalSurfaceSizeInMALLForSS = %u\n", __func__, TotalSurfaceSizeInMALLForSS);
+ DML_LOG_VERBOSE("DML::%s: ExceededMALLSize = %u\n", __func__, *ExceededMALLSize);
#endif
}
@@ -4583,7 +4610,6 @@ static void calculate_tdlut_setting(
unsigned int tdlut_vmpg_per_frame;
unsigned int tdlut_pte_req_per_frame;
unsigned int tdlut_bytes_per_line;
- unsigned int tdlut_delivery_cycles;
double tdlut_drain_rate;
unsigned int tdlut_mpc_width;
unsigned int tdlut_bytes_per_group_simple;
@@ -4592,6 +4618,7 @@ static void calculate_tdlut_setting(
*p->tdlut_groups_per_2row_ub = 0;
*p->tdlut_opt_time = 0;
*p->tdlut_drain_time = 0;
+ *p->tdlut_bytes_to_deliver = 0;
*p->tdlut_bytes_per_group = 0;
*p->tdlut_pte_bytes_per_frame = 0;
*p->tdlut_bytes_per_frame = 0;
@@ -4645,42 +4672,47 @@ static void calculate_tdlut_setting(
*p->tdlut_bytes_per_frame = tdlut_bytes_per_line * tdlut_mpc_width * tdlut_mpc_width;
*p->tdlut_bytes_per_group = tdlut_bytes_per_line * tdlut_mpc_width;
//the delivery cycles is DispClk cycles per line * number of lines * number of slices
- tdlut_delivery_cycles = (unsigned int)math_ceil2(tdlut_mpc_width/2.0, 1) * tdlut_mpc_width * tdlut_mpc_width;
+ //tdlut_delivery_cycles = (unsigned int)math_ceil2(tdlut_mpc_width/2.0, 1) * tdlut_mpc_width * tdlut_mpc_width;
tdlut_drain_rate = tdlut_bytes_per_line * p->dispclk_mhz / math_ceil2(tdlut_mpc_width/2.0, 1);
} else {
//tdlut_addressing_mode = tdlut_simple_linear, 3dlut width should be 4*1229=4916 elements
*p->tdlut_bytes_per_frame = (unsigned int)math_ceil2(tdlut_width * tdlut_bpe, 256);
*p->tdlut_bytes_per_group = tdlut_bytes_per_group_simple;
- tdlut_delivery_cycles = (unsigned int)math_ceil2(tdlut_width/2.0, 1);
+ //tdlut_delivery_cycles = (unsigned int)math_ceil2(tdlut_width/2.0, 1);
tdlut_drain_rate = 2 * tdlut_bpe * p->dispclk_mhz;
}
//the tdlut is fetched during the 2 row times of prefetch.
if (p->setup_for_tdlut) {
*p->tdlut_groups_per_2row_ub = (unsigned int)math_ceil2((double) *p->tdlut_bytes_per_frame / *p->tdlut_bytes_per_group, 1);
- *p->tdlut_opt_time = (*p->tdlut_bytes_per_frame - p->cursor_buffer_size * 1024) / tdlut_drain_rate;
+ if (*p->tdlut_bytes_per_frame > p->cursor_buffer_size * 1024)
+ *p->tdlut_opt_time = (*p->tdlut_bytes_per_frame - p->cursor_buffer_size * 1024) / tdlut_drain_rate;
+ else
+ *p->tdlut_opt_time = 0;
*p->tdlut_drain_time = p->cursor_buffer_size * 1024 / tdlut_drain_rate;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: gpuvm_enable = %d\n", __func__, p->gpuvm_enable);
- dml2_printf("DML::%s: vmpg_bytes = %d\n", __func__, vmpg_bytes);
- dml2_printf("DML::%s: tdlut_vmpg_per_frame = %d\n", __func__, tdlut_vmpg_per_frame);
- dml2_printf("DML::%s: tdlut_pte_req_per_frame = %d\n", __func__, tdlut_pte_req_per_frame);
-
- dml2_printf("DML::%s: dispclk_mhz = %f\n", __func__, p->dispclk_mhz);
- dml2_printf("DML::%s: tdlut_width = %u\n", __func__, tdlut_width);
- dml2_printf("DML::%s: tdlut_addressing_mode = %s\n", __func__, (p->tdlut_addressing_mode == dml2_tdlut_sw_linear) ? "sw_linear" : "simple_linear");
- dml2_printf("DML::%s: tdlut_pitch_bytes = %u\n", __func__, tdlut_pitch_bytes);
- dml2_printf("DML::%s: tdlut_footprint_bytes = %u\n", __func__, tdlut_footprint_bytes);
- dml2_printf("DML::%s: tdlut_bytes_per_frame = %u\n", __func__, *p->tdlut_bytes_per_frame);
- dml2_printf("DML::%s: tdlut_bytes_per_line = %u\n", __func__, tdlut_bytes_per_line);
- dml2_printf("DML::%s: tdlut_bytes_per_group = %u\n", __func__, *p->tdlut_bytes_per_group);
- dml2_printf("DML::%s: tdlut_drain_rate = %f\n", __func__, tdlut_drain_rate);
- dml2_printf("DML::%s: tdlut_delivery_cycles = %u\n", __func__, tdlut_delivery_cycles);
- dml2_printf("DML::%s: tdlut_opt_time = %f\n", __func__, *p->tdlut_opt_time);
- dml2_printf("DML::%s: tdlut_drain_time = %f\n", __func__, *p->tdlut_drain_time);
- dml2_printf("DML::%s: tdlut_groups_per_2row_ub = %d\n", __func__, *p->tdlut_groups_per_2row_ub);
+ *p->tdlut_bytes_to_deliver = (unsigned int) (p->cursor_buffer_size * 1024.0);
+ }
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: gpuvm_enable = %d\n", __func__, p->gpuvm_enable);
+ DML_LOG_VERBOSE("DML::%s: vmpg_bytes = %d\n", __func__, vmpg_bytes);
+ DML_LOG_VERBOSE("DML::%s: tdlut_vmpg_per_frame = %d\n", __func__, tdlut_vmpg_per_frame);
+ DML_LOG_VERBOSE("DML::%s: tdlut_pte_req_per_frame = %d\n", __func__, tdlut_pte_req_per_frame);
+
+ DML_LOG_VERBOSE("DML::%s: dispclk_mhz = %f\n", __func__, p->dispclk_mhz);
+ DML_LOG_VERBOSE("DML::%s: tdlut_width = %u\n", __func__, tdlut_width);
+ DML_LOG_VERBOSE("DML::%s: tdlut_addressing_mode = %s\n", __func__, (p->tdlut_addressing_mode == dml2_tdlut_sw_linear) ? "sw_linear" : "simple_linear");
+ DML_LOG_VERBOSE("DML::%s: tdlut_pitch_bytes = %u\n", __func__, tdlut_pitch_bytes);
+ DML_LOG_VERBOSE("DML::%s: tdlut_footprint_bytes = %u\n", __func__, tdlut_footprint_bytes);
+ DML_LOG_VERBOSE("DML::%s: tdlut_bytes_per_frame = %u\n", __func__, *p->tdlut_bytes_per_frame);
+ DML_LOG_VERBOSE("DML::%s: tdlut_bytes_per_line = %u\n", __func__, tdlut_bytes_per_line);
+ DML_LOG_VERBOSE("DML::%s: tdlut_bytes_per_group = %u\n", __func__, *p->tdlut_bytes_per_group);
+ DML_LOG_VERBOSE("DML::%s: tdlut_drain_rate = %f\n", __func__, tdlut_drain_rate);
+ DML_LOG_VERBOSE("DML::%s: tdlut_delivery_cycles = %u\n", __func__, p->tdlut_addressing_mode == dml2_tdlut_sw_linear ? (unsigned int)math_ceil2(tdlut_mpc_width/2.0, 1) * tdlut_mpc_width * tdlut_mpc_width : (unsigned int)math_ceil2(tdlut_width/2.0, 1));
+ DML_LOG_VERBOSE("DML::%s: tdlut_opt_time = %f\n", __func__, *p->tdlut_opt_time);
+ DML_LOG_VERBOSE("DML::%s: tdlut_drain_time = %f\n", __func__, *p->tdlut_drain_time);
+ DML_LOG_VERBOSE("DML::%s: tdlut_bytes_to_deliver = %d\n", __func__, *p->tdlut_bytes_to_deliver);
+ DML_LOG_VERBOSE("DML::%s: tdlut_groups_per_2row_ub = %d\n", __func__, *p->tdlut_groups_per_2row_ub);
#endif
}
@@ -4726,10 +4758,10 @@ static void CalculateTarb(
*Tarb = extra_bytes / ReturnBW;
*Tarb_prefetch = extra_bytes_prefetch / ReturnBW;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: PixelChunkSizeInKByte = %d\n", __func__, PixelChunkSizeInKByte);
- dml2_printf("DML::%s: MetaChunkSize = %d\n", __func__, MetaChunkSize);
- dml2_printf("DML::%s: extra_bytes = %f\n", __func__, extra_bytes);
- dml2_printf("DML::%s: extra_bytes_prefetch = %f\n", __func__, extra_bytes_prefetch);
+ DML_LOG_VERBOSE("DML::%s: PixelChunkSizeInKByte = %d\n", __func__, PixelChunkSizeInKByte);
+ DML_LOG_VERBOSE("DML::%s: MetaChunkSize = %d\n", __func__, MetaChunkSize);
+ DML_LOG_VERBOSE("DML::%s: extra_bytes = %f\n", __func__, extra_bytes);
+ DML_LOG_VERBOSE("DML::%s: extra_bytes_prefetch = %f\n", __func__, extra_bytes_prefetch);
#endif
}
@@ -4744,10 +4776,10 @@ static double CalculateTWait(
TWait = math_max2(reserved_vblank_time_ns/1000.0, g6_temp_read_blackout_us) + t_urg_trip;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: reserved_vblank_time_ns = %d\n", __func__, reserved_vblank_time_ns);
- dml2_printf("DML::%s: UrgentLatency = %f\n", __func__, UrgentLatency);
- dml2_printf("DML::%s: Ttrip = %f\n", __func__, Ttrip);
- dml2_printf("DML::%s: TWait = %f\n", __func__, TWait);
+ DML_LOG_VERBOSE("DML::%s: reserved_vblank_time_ns = %ld\n", __func__, reserved_vblank_time_ns);
+ DML_LOG_VERBOSE("DML::%s: UrgentLatency = %f\n", __func__, UrgentLatency);
+ DML_LOG_VERBOSE("DML::%s: Ttrip = %f\n", __func__, Ttrip);
+ DML_LOG_VERBOSE("DML::%s: TWait = %f\n", __func__, TWait);
#endif
return TWait;
}
@@ -4793,20 +4825,20 @@ static void CalculateVUpdateAndDynamicMetadataParameters(
*Tdmsks = *Tdmsks / 2;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DynamicMetadataLinesBeforeActiveRequired = %u\n", __func__, DynamicMetadataLinesBeforeActiveRequired);
- dml2_printf("DML::%s: VBlank = %u\n", __func__, VBlank);
- dml2_printf("DML::%s: HTotal = %u\n", __func__, HTotal);
- dml2_printf("DML::%s: PixelClock = %f\n", __func__, PixelClock);
- dml2_printf("DML::%s: Dppclk = %f\n", __func__, Dppclk);
- dml2_printf("DML::%s: DCFClkDeepSleep = %f\n", __func__, DCFClkDeepSleep);
- dml2_printf("DML::%s: MaxInterDCNTileRepeaters = %u\n", __func__, MaxInterDCNTileRepeaters);
- dml2_printf("DML::%s: TotalRepeaterDelayTime = %f\n", __func__, TotalRepeaterDelayTime);
+ DML_LOG_VERBOSE("DML::%s: DynamicMetadataLinesBeforeActiveRequired = %u\n", __func__, DynamicMetadataLinesBeforeActiveRequired);
+ DML_LOG_VERBOSE("DML::%s: VBlank = %u\n", __func__, VBlank);
+ DML_LOG_VERBOSE("DML::%s: HTotal = %u\n", __func__, HTotal);
+ DML_LOG_VERBOSE("DML::%s: PixelClock = %f\n", __func__, PixelClock);
+ DML_LOG_VERBOSE("DML::%s: Dppclk = %f\n", __func__, Dppclk);
+ DML_LOG_VERBOSE("DML::%s: DCFClkDeepSleep = %f\n", __func__, DCFClkDeepSleep);
+ DML_LOG_VERBOSE("DML::%s: MaxInterDCNTileRepeaters = %u\n", __func__, MaxInterDCNTileRepeaters);
+ DML_LOG_VERBOSE("DML::%s: TotalRepeaterDelayTime = %f\n", __func__, TotalRepeaterDelayTime);
- dml2_printf("DML::%s: VUpdateWidthPix = %u\n", __func__, *VUpdateWidthPix);
- dml2_printf("DML::%s: VReadyOffsetPix = %u\n", __func__, *VReadyOffsetPix);
- dml2_printf("DML::%s: VUpdateOffsetPix = %u\n", __func__, *VUpdateOffsetPix);
+ DML_LOG_VERBOSE("DML::%s: VUpdateWidthPix = %u\n", __func__, *VUpdateWidthPix);
+ DML_LOG_VERBOSE("DML::%s: VReadyOffsetPix = %u\n", __func__, *VReadyOffsetPix);
+ DML_LOG_VERBOSE("DML::%s: VUpdateOffsetPix = %u\n", __func__, *VUpdateOffsetPix);
- dml2_printf("DML::%s: Tdmsks = %f\n", __func__, *Tdmsks);
+ DML_LOG_VERBOSE("DML::%s: Tdmsks = %f\n", __func__, *Tdmsks);
#endif
}
@@ -4829,6 +4861,7 @@ static double get_urgent_bandwidth_required(
double ReadBandwidthChroma[],
double PrefetchBandwidthLuma[],
double PrefetchBandwidthChroma[],
+ double PrefetchBandwidthMax[],
double excess_vactive_fill_bw_l[],
double excess_vactive_fill_bw_c[],
double cursor_bw[],
@@ -4867,11 +4900,11 @@ static double get_urgent_bandwidth_required(
l->adj_factor_cur_pre = UrgentBurstFactorCursorPre[k];
bool is_phantom = dml_is_phantom_pipe(&display_cfg->plane_descriptors[k]);
- bool exclude_this_plane = 0;
+ bool exclude_this_plane = false;
// Exclude phantom pipe in bw calculation for non svp prefetch state
if (state_type != dml2_core_internal_soc_state_svp_prefetch && is_phantom)
- exclude_this_plane = 1;
+ exclude_this_plane = true;
// The qualified row bandwidth, qual_row_bw, accounts for the regular non-flip row bandwidth when there is no possible immediate flip or HostVM invalidation flip.
// The qual_row_bw is zero if HostVM is possible and only non-zero and equal to row_bw(i) if immediate flip is not allowed for that pipe.
@@ -4892,19 +4925,20 @@ static double get_urgent_bandwidth_required(
l->vm_row_bw = NumberOfDPP[k] * prefetch_vmrow_bw[k];
l->flip_and_active_bw = l->per_plane_flip_bw[k] + ReadBandwidthLuma[k] * l->adj_factor_p0 + ReadBandwidthChroma[k] * l->adj_factor_p1 + cursor_bw[k] * l->adj_factor_cur;
l->flip_and_prefetch_bw = l->per_plane_flip_bw[k] + NumberOfDPP[k] * (PrefetchBandwidthLuma[k] * l->adj_factor_p0_pre + PrefetchBandwidthChroma[k] * l->adj_factor_p1_pre) + prefetch_cursor_bw[k] * l->adj_factor_cur_pre;
+ l->flip_and_prefetch_bw_max = l->per_plane_flip_bw[k] + NumberOfDPP[k] * (PrefetchBandwidthMax[k] * l->adj_factor_p0_pre + PrefetchBandwidthChroma[k] * l->adj_factor_p1_pre) + prefetch_cursor_bw[k] * l->adj_factor_cur_pre;
l->active_and_excess_bw = (ReadBandwidthLuma[k] + excess_vactive_fill_bw_l[k]) * l->tmp_nom_adj_factor_p0 + (ReadBandwidthChroma[k] + excess_vactive_fill_bw_c[k]) * l->tmp_nom_adj_factor_p1 + dpte_row_bw[k] + meta_row_bw[k];
- surface_required_bw[k] = math_max4(l->vm_row_bw, l->flip_and_active_bw, l->flip_and_prefetch_bw, l->active_and_excess_bw);
+ surface_required_bw[k] = math_max5(l->vm_row_bw, l->flip_and_active_bw, l->flip_and_prefetch_bw, l->active_and_excess_bw, l->flip_and_prefetch_bw_max);
/* export peak required bandwidth for the surface */
surface_peak_required_bw[k] = math_max2(surface_required_bw[k], surface_peak_required_bw[k]);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, max1: vm_row_bw=%f\n", __func__, k, l->vm_row_bw);
- dml2_printf("DML::%s: k=%d, max2: flip_and_active_bw=%f\n", __func__, k, l->flip_and_active_bw);
- dml2_printf("DML::%s: k=%d, max3: flip_and_prefetch_bw=%f\n", __func__, k, l->flip_and_prefetch_bw);
- dml2_printf("DML::%s: k=%d, max4: active_and_excess_bw=%f\n", __func__, k, l->active_and_excess_bw);
- dml2_printf("DML::%s: k=%d, surface_required_bw=%f\n", __func__, k, surface_required_bw[k]);
- dml2_printf("DML::%s: k=%d, surface_peak_required_bw=%f\n", __func__, k, surface_peak_required_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, max1: vm_row_bw=%f\n", __func__, k, l->vm_row_bw);
+ DML_LOG_VERBOSE("DML::%s: k=%d, max2: flip_and_active_bw=%f\n", __func__, k, l->flip_and_active_bw);
+ DML_LOG_VERBOSE("DML::%s: k=%d, max3: flip_and_prefetch_bw=%f\n", __func__, k, l->flip_and_prefetch_bw);
+ DML_LOG_VERBOSE("DML::%s: k=%d, max4: active_and_excess_bw=%f\n", __func__, k, l->active_and_excess_bw);
+ DML_LOG_VERBOSE("DML::%s: k=%d, surface_required_bw=%f\n", __func__, k, surface_required_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, surface_peak_required_bw=%f\n", __func__, k, surface_peak_required_bw[k]);
#endif
} else {
surface_required_bw[k] = 0.0;
@@ -4913,34 +4947,34 @@ static double get_urgent_bandwidth_required(
l->required_bandwidth_mbps += surface_required_bw[k];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, NumberOfDPP=%d\n", __func__, k, NumberOfDPP[k]);
- dml2_printf("DML::%s: k=%d, use_qual_row_bw=%d\n", __func__, k, use_qual_row_bw);
- dml2_printf("DML::%s: k=%d, immediate_flip=%d\n", __func__, k, display_cfg->plane_descriptors[k].immediate_flip);
- dml2_printf("DML::%s: k=%d, mall_svp_prefetch_factor=%f\n", __func__, k, l->mall_svp_prefetch_factor);
- dml2_printf("DML::%s: k=%d, adj_factor_p0=%f\n", __func__, k, l->adj_factor_p0);
- dml2_printf("DML::%s: k=%d, adj_factor_p1=%f\n", __func__, k, l->adj_factor_p1);
- dml2_printf("DML::%s: k=%d, adj_factor_cur=%f\n", __func__, k, l->adj_factor_cur);
+ DML_LOG_VERBOSE("DML::%s: k=%d, NumberOfDPP=%d\n", __func__, k, NumberOfDPP[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, use_qual_row_bw=%d\n", __func__, k, use_qual_row_bw);
+ DML_LOG_VERBOSE("DML::%s: k=%d, immediate_flip=%d\n", __func__, k, display_cfg->plane_descriptors[k].immediate_flip);
+ DML_LOG_VERBOSE("DML::%s: k=%d, mall_svp_prefetch_factor=%f\n", __func__, k, l->mall_svp_prefetch_factor);
+ DML_LOG_VERBOSE("DML::%s: k=%d, adj_factor_p0=%f\n", __func__, k, l->adj_factor_p0);
+ DML_LOG_VERBOSE("DML::%s: k=%d, adj_factor_p1=%f\n", __func__, k, l->adj_factor_p1);
+ DML_LOG_VERBOSE("DML::%s: k=%d, adj_factor_cur=%f\n", __func__, k, l->adj_factor_cur);
- dml2_printf("DML::%s: k=%d, adj_factor_p0_pre=%f\n", __func__, k, l->adj_factor_p0_pre);
- dml2_printf("DML::%s: k=%d, adj_factor_p1_pre=%f\n", __func__, k, l->adj_factor_p1_pre);
- dml2_printf("DML::%s: k=%d, adj_factor_cur_pre=%f\n", __func__, k, l->adj_factor_cur_pre);
+ DML_LOG_VERBOSE("DML::%s: k=%d, adj_factor_p0_pre=%f\n", __func__, k, l->adj_factor_p0_pre);
+ DML_LOG_VERBOSE("DML::%s: k=%d, adj_factor_p1_pre=%f\n", __func__, k, l->adj_factor_p1_pre);
+ DML_LOG_VERBOSE("DML::%s: k=%d, adj_factor_cur_pre=%f\n", __func__, k, l->adj_factor_cur_pre);
- dml2_printf("DML::%s: k=%d, per_plane_flip_bw=%f\n", __func__, k, l->per_plane_flip_bw[k]);
- dml2_printf("DML::%s: k=%d, prefetch_vmrow_bw=%f\n", __func__, k, prefetch_vmrow_bw[k]);
- dml2_printf("DML::%s: k=%d, ReadBandwidthLuma=%f\n", __func__, k, ReadBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%d, ReadBandwidthChroma=%f\n", __func__, k, ReadBandwidthChroma[k]);
- dml2_printf("DML::%s: k=%d, excess_vactive_fill_bw_l=%f\n", __func__, k, excess_vactive_fill_bw_l[k]);
- dml2_printf("DML::%s: k=%d, excess_vactive_fill_bw_c=%f\n", __func__, k, excess_vactive_fill_bw_c[k]);
- dml2_printf("DML::%s: k=%d, cursor_bw=%f\n", __func__, k, cursor_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, per_plane_flip_bw=%f\n", __func__, k, l->per_plane_flip_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, prefetch_vmrow_bw=%f\n", __func__, k, prefetch_vmrow_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, ReadBandwidthLuma=%f\n", __func__, k, ReadBandwidthLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, ReadBandwidthChroma=%f\n", __func__, k, ReadBandwidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, excess_vactive_fill_bw_l=%f\n", __func__, k, excess_vactive_fill_bw_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, excess_vactive_fill_bw_c=%f\n", __func__, k, excess_vactive_fill_bw_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, cursor_bw=%f\n", __func__, k, cursor_bw[k]);
- dml2_printf("DML::%s: k=%d, meta_row_bw=%f\n", __func__, k, meta_row_bw[k]);
- dml2_printf("DML::%s: k=%d, dpte_row_bw=%f\n", __func__, k, dpte_row_bw[k]);
- dml2_printf("DML::%s: k=%d, PrefetchBandwidthLuma=%f\n", __func__, k, PrefetchBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%d, PrefetchBandwidthChroma=%f\n", __func__, k, PrefetchBandwidthChroma[k]);
- dml2_printf("DML::%s: k=%d, prefetch_cursor_bw=%f\n", __func__, k, prefetch_cursor_bw[k]);
- dml2_printf("DML::%s: k=%d, required_bandwidth_mbps=%f (total), inc_flip_bw=%d, is_phantom=%d exclude_this_plane=%d\n", __func__, k, l->required_bandwidth_mbps, inc_flip_bw, is_phantom, exclude_this_plane);
- dml2_printf("DML::%s: k=%d, required_bandwidth_mbps=%f (total), soc_state=%s, inc_flip_bw=%d, is_phantom=%d exclude_this_plane=%d\n", __func__, k, l->required_bandwidth_mbps, dml2_core_internal_soc_state_type_str(state_type), inc_flip_bw, is_phantom, exclude_this_plane);
- dml2_printf("DML::%s: k=%d, required_bandwidth_mbps=%f (total), inc_flip_bw=%d, is_phantom=%d exclude_this_plane=%d\n", __func__, k, l->required_bandwidth_mbps, inc_flip_bw, is_phantom, exclude_this_plane);
+ DML_LOG_VERBOSE("DML::%s: k=%d, meta_row_bw=%f\n", __func__, k, meta_row_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, dpte_row_bw=%f\n", __func__, k, dpte_row_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, PrefetchBandwidthLuma=%f\n", __func__, k, PrefetchBandwidthLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, PrefetchBandwidthChroma=%f\n", __func__, k, PrefetchBandwidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, prefetch_cursor_bw=%f\n", __func__, k, prefetch_cursor_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, required_bandwidth_mbps=%f (total), inc_flip_bw=%d, is_phantom=%d exclude_this_plane=%d\n", __func__, k, l->required_bandwidth_mbps, inc_flip_bw, is_phantom, exclude_this_plane);
+ DML_LOG_VERBOSE("DML::%s: k=%d, required_bandwidth_mbps=%f (total), soc_state=%s, inc_flip_bw=%d, is_phantom=%d exclude_this_plane=%d\n", __func__, k, l->required_bandwidth_mbps, dml2_core_internal_soc_state_type_str(state_type), inc_flip_bw, is_phantom, exclude_this_plane);
+ DML_LOG_VERBOSE("DML::%s: k=%d, required_bandwidth_mbps=%f (total), inc_flip_bw=%d, is_phantom=%d exclude_this_plane=%d\n", __func__, k, l->required_bandwidth_mbps, inc_flip_bw, is_phantom, exclude_this_plane);
#endif
}
@@ -4964,7 +4998,7 @@ static void CalculateExtraLatency(
double HostVMInefficiencyFactorPrefetch,
unsigned int HostVMMinPageSize,
enum dml2_qos_param_type qos_type,
- bool max_oustanding_when_urgent_expected,
+ bool max_outstanding_when_urgent_expected,
unsigned int max_outstanding_requests,
unsigned int request_size_bytes_luma[],
unsigned int request_size_bytes_chroma[],
@@ -5012,7 +5046,7 @@ static void CalculateExtraLatency(
if (qos_type == dml2_qos_param_type_dcn4x) {
*ExtraLatency_sr = dchub_arb_to_ret_delay / DCFCLK;
*ExtraLatency = *ExtraLatency_sr;
- if (max_oustanding_when_urgent_expected)
+ if (max_outstanding_when_urgent_expected)
*ExtraLatency = *ExtraLatency + (ROBBufferSizeInKByte * 1024 - max_outstanding_requests * max_request_size_bytes) / ReturnBW;
} else {
*ExtraLatency_sr = dchub_arb_to_ret_delay / DCFCLK + RoundTripPingLatencyCycles / FabricClock + ReorderingBytes / ReturnBW;
@@ -5024,19 +5058,19 @@ static void CalculateExtraLatency(
*ExtraLatency_sr = *ExtraLatency_sr + Tarb;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: qos_type=%u\n", __func__, qos_type);
- dml2_printf("DML::%s: hostvm_mode=%u\n", __func__, hostvm_mode);
- dml2_printf("DML::%s: Tex_trips=%u\n", __func__, Tex_trips);
- dml2_printf("DML::%s: max_oustanding_when_urgent_expected=%u\n", __func__, max_oustanding_when_urgent_expected);
- dml2_printf("DML::%s: FabricClock=%f\n", __func__, FabricClock);
- dml2_printf("DML::%s: DCFCLK=%f\n", __func__, DCFCLK);
- dml2_printf("DML::%s: ReturnBW=%f\n", __func__, ReturnBW);
- dml2_printf("DML::%s: RoundTripPingLatencyCycles=%u\n", __func__, RoundTripPingLatencyCycles);
- dml2_printf("DML::%s: ReorderingBytes=%u\n", __func__, ReorderingBytes);
- dml2_printf("DML::%s: Tarb=%f\n", __func__, Tarb);
- dml2_printf("DML::%s: ExtraLatency=%f\n", __func__, *ExtraLatency);
- dml2_printf("DML::%s: ExtraLatency_sr=%f\n", __func__, *ExtraLatency_sr);
- dml2_printf("DML::%s: ExtraLatencyPrefetch=%f\n", __func__, *ExtraLatencyPrefetch);
+ DML_LOG_VERBOSE("DML::%s: qos_type=%u\n", __func__, qos_type);
+ DML_LOG_VERBOSE("DML::%s: hostvm_mode=%u\n", __func__, hostvm_mode);
+ DML_LOG_VERBOSE("DML::%s: Tex_trips=%f\n", __func__, Tex_trips);
+ DML_LOG_VERBOSE("DML::%s: max_outstanding_when_urgent_expected=%u\n", __func__, max_outstanding_when_urgent_expected);
+ DML_LOG_VERBOSE("DML::%s: FabricClock=%f\n", __func__, FabricClock);
+ DML_LOG_VERBOSE("DML::%s: DCFCLK=%f\n", __func__, DCFCLK);
+ DML_LOG_VERBOSE("DML::%s: ReturnBW=%f\n", __func__, ReturnBW);
+ DML_LOG_VERBOSE("DML::%s: RoundTripPingLatencyCycles=%u\n", __func__, RoundTripPingLatencyCycles);
+ DML_LOG_VERBOSE("DML::%s: ReorderingBytes=%u\n", __func__, ReorderingBytes);
+ DML_LOG_VERBOSE("DML::%s: Tarb=%f\n", __func__, Tarb);
+ DML_LOG_VERBOSE("DML::%s: ExtraLatency=%f\n", __func__, *ExtraLatency);
+ DML_LOG_VERBOSE("DML::%s: ExtraLatency_sr=%f\n", __func__, *ExtraLatency_sr);
+ DML_LOG_VERBOSE("DML::%s: ExtraLatencyPrefetch=%f\n", __func__, *ExtraLatencyPrefetch);
#endif
}
@@ -5069,20 +5103,18 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
s->trip_to_mem = 0.0;
*p->Tvm_trips = 0.0;
*p->Tr0_trips = 0.0;
- s->Tvm_no_trip_oto = 0.0;
- s->Tr0_no_trip_oto = 0.0;
s->Tvm_trips_rounded = 0.0;
s->Tr0_trips_rounded = 0.0;
s->max_Tsw = 0.0;
s->Lsw_oto = 0.0;
- s->Tpre_rounded = 0.0;
+ *p->Tpre_rounded = 0.0;
s->prefetch_bw_equ = 0.0;
s->Tvm_equ = 0.0;
s->Tr0_equ = 0.0;
s->Tdmbf = 0.0;
s->Tdmec = 0.0;
s->Tdmsks = 0.0;
- s->prefetch_sw_bytes = 0.0;
+ *p->prefetch_sw_bytes = 0.0;
s->prefetch_bw_pr = 0.0;
s->bytes_pp = 0.0;
s->dep_bytes = 0.0;
@@ -5093,6 +5125,7 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
s->Tsw_est3 = 0.0;
s->cursor_prefetch_bytes = 0;
*p->prefetch_cursor_bw = 0;
+ *p->RequiredPrefetchBWMax = 0.0;
dcc_mrq_enable = (p->dcc_enable && p->mrq_present);
@@ -5104,20 +5137,20 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
s->HostVMDynamicLevelsTrips = 0;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: dcc_enable = %u\n", __func__, p->dcc_enable);
- dml2_printf("DML::%s: mrq_present = %u\n", __func__, p->mrq_present);
- dml2_printf("DML::%s: dcc_mrq_enable = %u\n", __func__, dcc_mrq_enable);
- dml2_printf("DML::%s: GPUVMEnable = %u\n", __func__, p->display_cfg->gpuvm_enable);
- dml2_printf("DML::%s: GPUVMPageTableLevels = %u\n", __func__, p->display_cfg->gpuvm_max_page_table_levels);
- dml2_printf("DML::%s: DCCEnable = %u\n", __func__, p->myPipe->DCCEnable);
- dml2_printf("DML::%s: VStartup = %u\n", __func__, p->VStartup);
- dml2_printf("DML::%s: HostVMEnable = %u\n", __func__, p->display_cfg->hostvm_enable);
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, p->HostVMInefficiencyFactor);
- dml2_printf("DML::%s: TWait = %f\n", __func__, p->TWait);
- dml2_printf("DML::%s: TWait_p = %f\n", __func__, s->TWait_p);
- dml2_printf("DML::%s: Ttrip = %f\n", __func__, p->Ttrip);
- dml2_printf("DML::%s: myPipe->Dppclk = %f\n", __func__, p->myPipe->Dppclk);
- dml2_printf("DML::%s: myPipe->Dispclk = %f\n", __func__, p->myPipe->Dispclk);
+ DML_LOG_VERBOSE("DML::%s: dcc_enable = %u\n", __func__, p->dcc_enable);
+ DML_LOG_VERBOSE("DML::%s: mrq_present = %u\n", __func__, p->mrq_present);
+ DML_LOG_VERBOSE("DML::%s: dcc_mrq_enable = %u\n", __func__, dcc_mrq_enable);
+ DML_LOG_VERBOSE("DML::%s: GPUVMEnable = %u\n", __func__, p->display_cfg->gpuvm_enable);
+ DML_LOG_VERBOSE("DML::%s: GPUVMPageTableLevels = %u\n", __func__, p->display_cfg->gpuvm_max_page_table_levels);
+ DML_LOG_VERBOSE("DML::%s: DCCEnable = %u\n", __func__, p->myPipe->DCCEnable);
+ DML_LOG_VERBOSE("DML::%s: VStartup = %u\n", __func__, p->VStartup);
+ DML_LOG_VERBOSE("DML::%s: HostVMEnable = %u\n", __func__, p->display_cfg->hostvm_enable);
+ DML_LOG_VERBOSE("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, p->HostVMInefficiencyFactor);
+ DML_LOG_VERBOSE("DML::%s: TWait = %f\n", __func__, p->TWait);
+ DML_LOG_VERBOSE("DML::%s: TWait_p = %f\n", __func__, s->TWait_p);
+ DML_LOG_VERBOSE("DML::%s: Ttrip = %f\n", __func__, p->Ttrip);
+ DML_LOG_VERBOSE("DML::%s: myPipe->Dppclk = %f\n", __func__, p->myPipe->Dppclk);
+ DML_LOG_VERBOSE("DML::%s: myPipe->Dispclk = %f\n", __func__, p->myPipe->Dispclk);
#endif
CalculateVUpdateAndDynamicMetadataParameters(
p->MaxInterDCNTileRepeaters,
@@ -5163,11 +5196,11 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
if (p->DynamicMetadataEnable == true) {
if (p->VStartup * s->LineTime < *p->TSetup + *p->Tdmdl + s->Tdmbf + s->Tdmec + s->Tdmsks) {
*p->NotEnoughTimeForDynamicMetadata = true;
- dml2_printf("DML::%s: Not Enough Time for Dynamic Meta!\n", __func__);
- dml2_printf("DML::%s: Tdmbf: %fus - time for dmd transfer from dchub to dio output buffer\n", __func__, s->Tdmbf);
- dml2_printf("DML::%s: Tdmec: %fus - time dio takes to transfer dmd\n", __func__, s->Tdmec);
- dml2_printf("DML::%s: Tdmsks: %fus - time before active dmd must complete transmission at dio\n", __func__, s->Tdmsks);
- dml2_printf("DML::%s: Tdmdl: %fus - time for fabric to become ready and fetch dmd \n", __func__, *p->Tdmdl);
+ DML_LOG_VERBOSE("DML::%s: Not Enough Time for Dynamic Meta!\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: Tdmbf: %fus - time for dmd transfer from dchub to dio output buffer\n", __func__, s->Tdmbf);
+ DML_LOG_VERBOSE("DML::%s: Tdmec: %fus - time dio takes to transfer dmd\n", __func__, s->Tdmec);
+ DML_LOG_VERBOSE("DML::%s: Tdmsks: %fus - time before active dmd must complete transmission at dio\n", __func__, s->Tdmsks);
+ DML_LOG_VERBOSE("DML::%s: Tdmdl: %fus - time for fabric to become ready and fetch dmd \n", __func__, *p->Tdmdl);
} else {
*p->NotEnoughTimeForDynamicMetadata = false;
}
@@ -5193,20 +5226,21 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
((p->myPipe->ODMMode == dml2_odm_mode_mso_1to4) ? (double)p->myPipe->HActive * 3.0 / 4.0 : 0));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DynamicMetadataVMEnabled = %u\n", __func__, p->DynamicMetadataVMEnabled);
- dml2_printf("DML::%s: DPPCycles = %u\n", __func__, s->DPPCycles);
- dml2_printf("DML::%s: PixelClock = %f\n", __func__, p->myPipe->PixelClock);
- dml2_printf("DML::%s: Dppclk = %f\n", __func__, p->myPipe->Dppclk);
- dml2_printf("DML::%s: DISPCLKCycles = %u\n", __func__, s->DISPCLKCycles);
- dml2_printf("DML::%s: DISPCLK = %f\n", __func__, p->myPipe->Dispclk);
- dml2_printf("DML::%s: DSCDelay = %u\n", __func__, p->DSCDelay);
- dml2_printf("DML::%s: ODMMode = %u\n", __func__, p->myPipe->ODMMode);
- dml2_printf("DML::%s: DPP_RECOUT_WIDTH = %u\n", __func__, p->DPP_RECOUT_WIDTH);
- dml2_printf("DML::%s: DSTXAfterScaler = %u\n", __func__, *p->DSTXAfterScaler);
+ DML_LOG_VERBOSE("DML::%s: DynamicMetadataVMEnabled = %u\n", __func__, p->DynamicMetadataVMEnabled);
+ DML_LOG_VERBOSE("DML::%s: DPPCycles = %u\n", __func__, s->DPPCycles);
+ DML_LOG_VERBOSE("DML::%s: PixelClock = %f\n", __func__, p->myPipe->PixelClock);
+ DML_LOG_VERBOSE("DML::%s: Dppclk = %f\n", __func__, p->myPipe->Dppclk);
+ DML_LOG_VERBOSE("DML::%s: DISPCLKCycles = %u\n", __func__, s->DISPCLKCycles);
+ DML_LOG_VERBOSE("DML::%s: DISPCLK = %f\n", __func__, p->myPipe->Dispclk);
+ DML_LOG_VERBOSE("DML::%s: DSCDelay = %u\n", __func__, p->DSCDelay);
+ DML_LOG_VERBOSE("DML::%s: ODMMode = %u\n", __func__, p->myPipe->ODMMode);
+ DML_LOG_VERBOSE("DML::%s: DPP_RECOUT_WIDTH = %u\n", __func__, p->DPP_RECOUT_WIDTH);
+ DML_LOG_VERBOSE("DML::%s: DSTXAfterScaler = %u\n", __func__, *p->DSTXAfterScaler);
- dml2_printf("DML::%s: setup_for_tdlut = %u\n", __func__, p->setup_for_tdlut);
- dml2_printf("DML::%s: tdlut_opt_time = %f\n", __func__, p->tdlut_opt_time);
- dml2_printf("DML::%s: tdlut_pte_bytes_per_frame = %u\n", __func__, p->tdlut_pte_bytes_per_frame);
+ DML_LOG_VERBOSE("DML::%s: setup_for_tdlut = %u\n", __func__, p->setup_for_tdlut);
+ DML_LOG_VERBOSE("DML::%s: tdlut_opt_time = %f\n", __func__, p->tdlut_opt_time);
+ DML_LOG_VERBOSE("DML::%s: tdlut_pte_bytes_per_frame = %u\n", __func__, p->tdlut_pte_bytes_per_frame);
+ DML_LOG_VERBOSE("DML::%s: tdlut_drain_time = %f\n", __func__, p->tdlut_drain_time);
#endif
if (p->OutputFormat == dml2_420 || (p->myPipe->InterlaceEnable && p->myPipe->ProgressiveToInterlaceUnitInOPP))
@@ -5218,17 +5252,17 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
*p->DSTYAfterScaler = (unsigned int)(math_floor2(s->DSTTotalPixelsAfterScaler / p->myPipe->HTotal, 1));
*p->DSTXAfterScaler = (unsigned int)(s->DSTTotalPixelsAfterScaler - ((double)(*p->DSTYAfterScaler * p->myPipe->HTotal)));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DSTXAfterScaler = %u (final)\n", __func__, *p->DSTXAfterScaler);
- dml2_printf("DML::%s: DSTYAfterScaler = %u (final)\n", __func__, *p->DSTYAfterScaler);
+ DML_LOG_VERBOSE("DML::%s: DSTXAfterScaler = %u (final)\n", __func__, *p->DSTXAfterScaler);
+ DML_LOG_VERBOSE("DML::%s: DSTYAfterScaler = %u (final)\n", __func__, *p->DSTYAfterScaler);
#endif
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tr0_trips = %f\n", __func__, *p->Tr0_trips);
- dml2_printf("DML::%s: Tvm_trips = %f\n", __func__, *p->Tvm_trips);
- dml2_printf("DML::%s: trip_to_mem = %f\n", __func__, s->trip_to_mem);
- dml2_printf("DML::%s: ExtraLatencyPrefetch = %f\n", __func__, p->ExtraLatencyPrefetch);
- dml2_printf("DML::%s: GPUVMPageTableLevels = %u\n", __func__, p->display_cfg->gpuvm_max_page_table_levels);
- dml2_printf("DML::%s: HostVMDynamicLevelsTrips = %u\n", __func__, s->HostVMDynamicLevelsTrips);
+ DML_LOG_VERBOSE("DML::%s: Tr0_trips = %f\n", __func__, *p->Tr0_trips);
+ DML_LOG_VERBOSE("DML::%s: Tvm_trips = %f\n", __func__, *p->Tvm_trips);
+ DML_LOG_VERBOSE("DML::%s: trip_to_mem = %f\n", __func__, s->trip_to_mem);
+ DML_LOG_VERBOSE("DML::%s: ExtraLatencyPrefetch = %f\n", __func__, p->ExtraLatencyPrefetch);
+ DML_LOG_VERBOSE("DML::%s: GPUVMPageTableLevels = %u\n", __func__, p->display_cfg->gpuvm_max_page_table_levels);
+ DML_LOG_VERBOSE("DML::%s: HostVMDynamicLevelsTrips = %u\n", __func__, s->HostVMDynamicLevelsTrips);
#endif
if (p->display_cfg->gpuvm_enable) {
s->Tvm_trips_rounded = math_ceil2(4.0 * *p->Tvm_trips / s->LineTime, 1.0) / 4.0 * s->LineTime;
@@ -5277,23 +5311,8 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
s->bytes_pp = p->myPipe->BytePerPixelY + p->myPipe->BytePerPixelC;
}
- s->prefetch_bw_pr = s->bytes_pp * p->myPipe->PixelClock / (double)p->myPipe->DPPPerSurface;
- if (p->myPipe->VRatio < 1.0)
- s->prefetch_bw_pr = p->myPipe->VRatio * s->prefetch_bw_pr;
- s->max_Tsw = (math_max2(p->PrefetchSourceLinesY, p->PrefetchSourceLinesC) * s->LineTime);
-
- s->prefetch_sw_bytes = p->PrefetchSourceLinesY * p->swath_width_luma_ub * p->myPipe->BytePerPixelY + p->PrefetchSourceLinesC * p->swath_width_chroma_ub * p->myPipe->BytePerPixelC;
- s->prefetch_bw_pr = s->prefetch_bw_pr * p->mall_prefetch_sdp_overhead_factor;
- s->prefetch_sw_bytes = s->prefetch_sw_bytes * p->mall_prefetch_sdp_overhead_factor;
- s->prefetch_bw_oto = math_max2(s->prefetch_bw_pr, s->prefetch_sw_bytes / s->max_Tsw);
-
- s->min_Lsw_oto = math_max2(p->PrefetchSourceLinesY, p->PrefetchSourceLinesC) / __DML2_CALCS_MAX_VRATIO_PRE_OTO__;
- s->min_Lsw_oto = math_max2(s->min_Lsw_oto, 2.0);
- s->min_Lsw_oto = math_max2(s->min_Lsw_oto, p->tdlut_drain_time / s->LineTime);
-
- s->min_Lsw_equ = math_max2(p->PrefetchSourceLinesY, p->PrefetchSourceLinesC) / __DML2_CALCS_MAX_VRATIO_PRE_EQU__;
- s->min_Lsw_equ = math_max2(s->min_Lsw_equ, 2.0);
- s->min_Lsw_equ = math_max2(s->min_Lsw_equ, p->tdlut_drain_time / s->LineTime);
+ *p->prefetch_sw_bytes = p->PrefetchSourceLinesY * p->swath_width_luma_ub * p->myPipe->BytePerPixelY + p->PrefetchSourceLinesC * p->swath_width_chroma_ub * p->myPipe->BytePerPixelC;
+ *p->prefetch_sw_bytes = *p->prefetch_sw_bytes * p->mall_prefetch_sdp_overhead_factor;
vm_bytes = p->vm_bytes; // vm_bytes is dpde0_bytes_per_frame_ub_l + dpde0_bytes_per_frame_ub_c + 2*extra_dpde_bytes;
extra_tdpe_bytes = (unsigned int)math_max2(0, (p->display_cfg->gpuvm_max_page_table_levels - 1) * 128);
@@ -5302,126 +5321,180 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
vm_bytes = vm_bytes + p->tdlut_pte_bytes_per_frame + (p->display_cfg->gpuvm_enable ? extra_tdpe_bytes : 0);
tdlut_row_bytes = (unsigned long) math_ceil2(p->tdlut_bytes_per_frame/2.0, 1.0);
+
+ s->min_Lsw_oto = math_max2(p->PrefetchSourceLinesY, p->PrefetchSourceLinesC) / __DML2_CALCS_MAX_VRATIO_PRE_OTO__;
+ s->min_Lsw_oto = math_max2(s->min_Lsw_oto, p->tdlut_drain_time / s->LineTime);
+ s->min_Lsw_oto = math_max2(s->min_Lsw_oto, 2.0);
+
+ // use vactive swath bw for prefetch oto and also cap prefetch_bw_oto to max_vratio_oto
+ // Note: in prefetch calculation, acounting is done mostly per-pipe.
+ // vactive swath bw represents the per-surface (aka per dml plane) bw to move vratio_l/c lines of bytes_l/c per line time
+ s->per_pipe_vactive_sw_bw = p->vactive_sw_bw_l / (double)p->myPipe->DPPPerSurface;
+
+ // one-to-one prefetch bw as one line of bytes per line time (as per vratio_pre_l/c = 1)
+ s->prefetch_bw_oto = (p->swath_width_luma_ub * p->myPipe->BytePerPixelY) / s->LineTime;
+
+ if (p->myPipe->BytePerPixelC > 0) {
+ s->per_pipe_vactive_sw_bw += p->vactive_sw_bw_c / (double)p->myPipe->DPPPerSurface;
+ s->prefetch_bw_oto += (p->swath_width_chroma_ub * p->myPipe->BytePerPixelC) / s->LineTime;
+ }
+
+ /* oto prefetch bw should be always be less than total vactive bw */
+ //DML_ASSERT(s->prefetch_bw_oto < s->per_pipe_vactive_sw_bw * p->myPipe->DPPPerSurface);
+
+ s->prefetch_bw_oto = math_max2(s->per_pipe_vactive_sw_bw, s->prefetch_bw_oto) * p->mall_prefetch_sdp_overhead_factor;
+
+ s->prefetch_bw_oto = math_min2(s->prefetch_bw_oto, *p->prefetch_sw_bytes/(s->min_Lsw_oto*s->LineTime));
+
+ s->Lsw_oto = math_ceil2(4.0 * *p->prefetch_sw_bytes / s->prefetch_bw_oto / s->LineTime, 1.0) / 4.0;
+
s->prefetch_bw_oto = math_max3(s->prefetch_bw_oto,
p->vm_bytes * p->HostVMInefficiencyFactor / (31 * s->LineTime) - *p->Tno_bw,
(p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / (15 * s->LineTime));
- s->Lsw_oto = math_ceil2(4.0 * math_max2(s->prefetch_sw_bytes / s->prefetch_bw_oto / s->LineTime, s->min_Lsw_oto), 1.0) / 4.0;
+
+ /* oto bw needs to be outputted even if the oto schedule isn't being used to avoid ms/mp mismatch.
+ * mp will fail if ms decides to use equ schedule and mp decides to use oto schedule
+ * and the required bandwidth increases when going from ms to mp
+ */
+ *p->RequiredPrefetchBWMax = s->prefetch_bw_oto;
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: vactive_sw_bw_l = %f\n", __func__, p->vactive_sw_bw_l);
+ DML_LOG_VERBOSE("DML::%s: vactive_sw_bw_c = %f\n", __func__, p->vactive_sw_bw_c);
+ DML_LOG_VERBOSE("DML::%s: per_pipe_vactive_sw_bw = %f\n", __func__, s->per_pipe_vactive_sw_bw);
+#endif
if (p->display_cfg->gpuvm_enable == true) {
- s->Tvm_no_trip_oto = math_max2(
+ s->Tvm_oto = math_max3(
+ *p->Tvm_trips,
*p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw_oto,
s->LineTime / 4.0);
- s->Tvm_oto = math_max2(
- *p->Tvm_trips,
- s->Tvm_no_trip_oto);
+
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tvm_oto max0 = %f\n", __func__, *p->Tvm_trips);
- dml2_printf("DML::%s: Tvm_oto max1 = %f\n", __func__, *p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw_oto);
- dml2_printf("DML::%s: Tvm_oto max2 = %f\n", __func__, s->LineTime / 4.0);
+ DML_LOG_VERBOSE("DML::%s: Tvm_oto max0 = %f\n", __func__, *p->Tvm_trips);
+ DML_LOG_VERBOSE("DML::%s: Tvm_oto max1 = %f\n", __func__, *p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw_oto);
+ DML_LOG_VERBOSE("DML::%s: Tvm_oto max2 = %f\n", __func__, s->LineTime / 4.0);
#endif
} else {
- s->Tvm_no_trip_oto = s->Tvm_trips_rounded;
s->Tvm_oto = s->Tvm_trips_rounded;
}
if ((p->display_cfg->gpuvm_enable == true || p->setup_for_tdlut || dcc_mrq_enable)) {
- s->Tr0_no_trip_oto = math_max2(
+ s->Tr0_oto = math_max3(
+ *p->Tr0_trips,
(p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / s->prefetch_bw_oto,
s->LineTime / 4.0);
- s->Tr0_oto = math_max2(
- *p->Tr0_trips,
- s->Tr0_no_trip_oto);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tr0_oto max0 = %f\n", __func__, *p->Tr0_trips);
- dml2_printf("DML::%s: Tr0_oto max1 = %f\n", __func__, (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / s->prefetch_bw_oto);
- dml2_printf("DML::%s: Tr0_oto max2 = %f\n", __func__, s->LineTime / 4);
+ DML_LOG_VERBOSE("DML::%s: Tr0_oto max0 = %f\n", __func__, *p->Tr0_trips);
+ DML_LOG_VERBOSE("DML::%s: Tr0_oto max1 = %f\n", __func__, (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / s->prefetch_bw_oto);
+ DML_LOG_VERBOSE("DML::%s: Tr0_oto max2 = %f\n", __func__, s->LineTime / 4);
#endif
- } else {
- s->Tr0_no_trip_oto = (s->LineTime - s->Tvm_oto) / 4.0;
- s->Tr0_oto = s->Tr0_no_trip_oto;
- }
+ } else
+ s->Tr0_oto = s->LineTime / 4.0;
s->Tvm_oto_lines = math_ceil2(4.0 * s->Tvm_oto / s->LineTime, 1) / 4.0;
s->Tr0_oto_lines = math_ceil2(4.0 * s->Tr0_oto / s->LineTime, 1) / 4.0;
s->dst_y_prefetch_oto = s->Tvm_oto_lines + 2 * s->Tr0_oto_lines + s->Lsw_oto;
+#ifdef DML_GLOBAL_PREFETCH_CHECK
+ DML_LOG_VERBOSE("DML::%s: impacted_Tpre = %f\n", __func__, p->impacted_dst_y_pre);
+ if (p->impacted_dst_y_pre > 0) {
+ DML_LOG_VERBOSE("DML::%s: dst_y_prefetch_oto = %f\n", __func__, s->dst_y_prefetch_oto);
+ s->dst_y_prefetch_oto = math_max2(s->dst_y_prefetch_oto, p->impacted_dst_y_pre);
+ DML_LOG_VERBOSE("DML::%s: dst_y_prefetch_oto = %f (impacted)\n", __func__, s->dst_y_prefetch_oto);
+ }
+#endif
+ *p->Tpre_oto = s->dst_y_prefetch_oto * s->LineTime;
+
//To (time for delay after scaler) in line time
Lo = (unsigned int)(*p->DSTYAfterScaler + (double)*p->DSTXAfterScaler / (double)p->myPipe->HTotal);
+ s->min_Lsw_equ = math_max2(p->PrefetchSourceLinesY, p->PrefetchSourceLinesC) / __DML2_CALCS_MAX_VRATIO_PRE_EQU__;
+ s->min_Lsw_equ = math_max2(s->min_Lsw_equ, p->tdlut_drain_time / s->LineTime);
+ s->min_Lsw_equ = math_max2(s->min_Lsw_equ, 2.0);
//Tpre_equ in line time
if (p->DynamicMetadataVMEnabled && p->DynamicMetadataEnable)
s->dst_y_prefetch_equ = p->VStartup - (*p->TSetup + math_max2(p->TCalc, *p->Tvm_trips) + s->TWait_p) / s->LineTime - Lo;
else
s->dst_y_prefetch_equ = p->VStartup - (*p->TSetup + math_max2(p->TCalc, p->ExtraLatencyPrefetch) + s->TWait_p) / s->LineTime - Lo;
+
+#ifdef DML_GLOBAL_PREFETCH_CHECK
+ s->dst_y_prefetch_equ_impacted = math_max2(p->impacted_dst_y_pre, s->dst_y_prefetch_equ);
+
+ s->dst_y_prefetch_equ_impacted = math_min2(s->dst_y_prefetch_equ_impacted, 63.75); // limit to the reg limit of U6.2 for DST_Y_PREFETCH
+
+ if (s->dst_y_prefetch_equ_impacted > s->dst_y_prefetch_equ)
+ s->dst_y_prefetch_equ -= s->dst_y_prefetch_equ_impacted - s->dst_y_prefetch_equ;
+#endif
+
s->dst_y_prefetch_equ = math_min2(s->dst_y_prefetch_equ, 63.75); // limit to the reg limit of U6.2 for DST_Y_PREFETCH
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: HTotal = %u\n", __func__, p->myPipe->HTotal);
- dml2_printf("DML::%s: min_Lsw_oto = %f\n", __func__, s->min_Lsw_oto);
- dml2_printf("DML::%s: min_Lsw_equ = %f\n", __func__, s->min_Lsw_equ);
- dml2_printf("DML::%s: Tno_bw = %f\n", __func__, *p->Tno_bw);
- dml2_printf("DML::%s: Tno_bw_flip = %f\n", __func__, *p->Tno_bw_flip);
- dml2_printf("DML::%s: ExtraLatencyPrefetch = %f\n", __func__, p->ExtraLatencyPrefetch);
- dml2_printf("DML::%s: trip_to_mem = %f\n", __func__, s->trip_to_mem);
- dml2_printf("DML::%s: mall_prefetch_sdp_overhead_factor = %f\n", __func__, p->mall_prefetch_sdp_overhead_factor);
- dml2_printf("DML::%s: BytePerPixelY = %u\n", __func__, p->myPipe->BytePerPixelY);
- dml2_printf("DML::%s: PrefetchSourceLinesY = %f\n", __func__, p->PrefetchSourceLinesY);
- dml2_printf("DML::%s: swath_width_luma_ub = %u\n", __func__, p->swath_width_luma_ub);
- dml2_printf("DML::%s: BytePerPixelC = %u\n", __func__, p->myPipe->BytePerPixelC);
- dml2_printf("DML::%s: PrefetchSourceLinesC = %f\n", __func__, p->PrefetchSourceLinesC);
- dml2_printf("DML::%s: swath_width_chroma_ub = %u\n", __func__, p->swath_width_chroma_ub);
- dml2_printf("DML::%s: prefetch_sw_bytes = %f\n", __func__, s->prefetch_sw_bytes);
- dml2_printf("DML::%s: max_Tsw = %f\n", __func__, s->max_Tsw);
- dml2_printf("DML::%s: bytes_pp = %f\n", __func__, s->bytes_pp);
- dml2_printf("DML::%s: vm_bytes = %u\n", __func__, vm_bytes);
- dml2_printf("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, p->PixelPTEBytesPerRow);
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, p->HostVMInefficiencyFactor);
- dml2_printf("DML::%s: Tvm_trips = %f\n", __func__, *p->Tvm_trips);
- dml2_printf("DML::%s: Tr0_trips = %f\n", __func__, *p->Tr0_trips);
- dml2_printf("DML::%s: Tvm_trips_flip = %f\n", __func__, *p->Tvm_trips_flip);
- dml2_printf("DML::%s: Tr0_trips_flip = %f\n", __func__, *p->Tr0_trips_flip);
- dml2_printf("DML::%s: prefetch_bw_pr = %f\n", __func__, s->prefetch_bw_pr);
- dml2_printf("DML::%s: prefetch_bw_oto = %f\n", __func__, s->prefetch_bw_oto);
- dml2_printf("DML::%s: Tr0_oto = %f\n", __func__, s->Tr0_oto);
- dml2_printf("DML::%s: Tvm_oto = %f\n", __func__, s->Tvm_oto);
- dml2_printf("DML::%s: Tvm_oto_lines = %f\n", __func__, s->Tvm_oto_lines);
- dml2_printf("DML::%s: Tr0_oto_lines = %f\n", __func__, s->Tr0_oto_lines);
- dml2_printf("DML::%s: Lsw_oto = %f\n", __func__, s->Lsw_oto);
- dml2_printf("DML::%s: dst_y_prefetch_oto = %f\n", __func__, s->dst_y_prefetch_oto);
- dml2_printf("DML::%s: dst_y_prefetch_equ = %f\n", __func__, s->dst_y_prefetch_equ);
- dml2_printf("DML::%s: tdlut_row_bytes = %d\n", __func__, tdlut_row_bytes);
- dml2_printf("DML::%s: meta_row_bytes = %d\n", __func__, p->meta_row_bytes);
-#endif
- double Tpre = s->dst_y_prefetch_equ * s->LineTime;
+ DML_LOG_VERBOSE("DML::%s: HTotal = %u\n", __func__, p->myPipe->HTotal);
+ DML_LOG_VERBOSE("DML::%s: min_Lsw_oto = %f\n", __func__, s->min_Lsw_oto);
+ DML_LOG_VERBOSE("DML::%s: min_Lsw_equ = %f\n", __func__, s->min_Lsw_equ);
+ DML_LOG_VERBOSE("DML::%s: Tno_bw = %f\n", __func__, *p->Tno_bw);
+ DML_LOG_VERBOSE("DML::%s: Tno_bw_flip = %f\n", __func__, *p->Tno_bw_flip);
+ DML_LOG_VERBOSE("DML::%s: ExtraLatencyPrefetch = %f\n", __func__, p->ExtraLatencyPrefetch);
+ DML_LOG_VERBOSE("DML::%s: trip_to_mem = %f\n", __func__, s->trip_to_mem);
+ DML_LOG_VERBOSE("DML::%s: mall_prefetch_sdp_overhead_factor = %f\n", __func__, p->mall_prefetch_sdp_overhead_factor);
+ DML_LOG_VERBOSE("DML::%s: BytePerPixelY = %u\n", __func__, p->myPipe->BytePerPixelY);
+ DML_LOG_VERBOSE("DML::%s: PrefetchSourceLinesY = %f\n", __func__, p->PrefetchSourceLinesY);
+ DML_LOG_VERBOSE("DML::%s: swath_width_luma_ub = %u\n", __func__, p->swath_width_luma_ub);
+ DML_LOG_VERBOSE("DML::%s: BytePerPixelC = %u\n", __func__, p->myPipe->BytePerPixelC);
+ DML_LOG_VERBOSE("DML::%s: PrefetchSourceLinesC = %f\n", __func__, p->PrefetchSourceLinesC);
+ DML_LOG_VERBOSE("DML::%s: swath_width_chroma_ub = %u\n", __func__, p->swath_width_chroma_ub);
+ DML_LOG_VERBOSE("DML::%s: prefetch_sw_bytes = %f\n", __func__, *p->prefetch_sw_bytes);
+ DML_LOG_VERBOSE("DML::%s: max_Tsw = %f\n", __func__, s->max_Tsw);
+ DML_LOG_VERBOSE("DML::%s: bytes_pp = %f\n", __func__, s->bytes_pp);
+ DML_LOG_VERBOSE("DML::%s: vm_bytes = %u\n", __func__, vm_bytes);
+ DML_LOG_VERBOSE("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, p->PixelPTEBytesPerRow);
+ DML_LOG_VERBOSE("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, p->HostVMInefficiencyFactor);
+ DML_LOG_VERBOSE("DML::%s: Tvm_trips = %f\n", __func__, *p->Tvm_trips);
+ DML_LOG_VERBOSE("DML::%s: Tr0_trips = %f\n", __func__, *p->Tr0_trips);
+ DML_LOG_VERBOSE("DML::%s: Tvm_trips_flip = %f\n", __func__, *p->Tvm_trips_flip);
+ DML_LOG_VERBOSE("DML::%s: Tr0_trips_flip = %f\n", __func__, *p->Tr0_trips_flip);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw_pr = %f\n", __func__, s->prefetch_bw_pr);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw_oto = %f\n", __func__, s->prefetch_bw_oto);
+ DML_LOG_VERBOSE("DML::%s: Tr0_oto = %f\n", __func__, s->Tr0_oto);
+ DML_LOG_VERBOSE("DML::%s: Tvm_oto = %f\n", __func__, s->Tvm_oto);
+ DML_LOG_VERBOSE("DML::%s: Tvm_oto_lines = %f\n", __func__, s->Tvm_oto_lines);
+ DML_LOG_VERBOSE("DML::%s: Tr0_oto_lines = %f\n", __func__, s->Tr0_oto_lines);
+ DML_LOG_VERBOSE("DML::%s: Lsw_oto = %f\n", __func__, s->Lsw_oto);
+ DML_LOG_VERBOSE("DML::%s: dst_y_prefetch_oto = %f\n", __func__, s->dst_y_prefetch_oto);
+ DML_LOG_VERBOSE("DML::%s: dst_y_prefetch_equ = %f\n", __func__, s->dst_y_prefetch_equ);
+ DML_LOG_VERBOSE("DML::%s: tdlut_row_bytes = %d\n", __func__, tdlut_row_bytes);
+ DML_LOG_VERBOSE("DML::%s: meta_row_bytes = %d\n", __func__, p->meta_row_bytes);
+#endif
s->dst_y_prefetch_equ = math_floor2(4.0 * (s->dst_y_prefetch_equ + 0.125), 1) / 4.0;
- s->Tpre_rounded = s->dst_y_prefetch_equ * s->LineTime;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: dst_y_prefetch_equ: %f (after round)\n", __func__, s->dst_y_prefetch_equ);
- dml2_printf("DML::%s: LineTime: %f\n", __func__, s->LineTime);
- dml2_printf("DML::%s: VStartup: %u\n", __func__, p->VStartup);
- dml2_printf("DML::%s: Tvstartup: %fus - time between vstartup and first pixel of active\n", __func__, p->VStartup * s->LineTime);
- dml2_printf("DML::%s: TSetup: %fus - time from vstartup to vready\n", __func__, *p->TSetup);
- dml2_printf("DML::%s: TCalc: %fus - time for calculations in dchub starting at vready\n", __func__, p->TCalc);
- dml2_printf("DML::%s: TWait: %fus - time for fabric to become ready max(pstate exit,cstate enter/exit, urgent latency) after TCalc\n", __func__, p->TWait);
- dml2_printf("DML::%s: Tdmbf: %fus - time for dmd transfer from dchub to dio output buffer\n", __func__, s->Tdmbf);
- dml2_printf("DML::%s: Tdmec: %fus - time dio takes to transfer dmd\n", __func__, s->Tdmec);
- dml2_printf("DML::%s: Tdmsks: %fus - time before active dmd must complete transmission at dio\n", __func__, s->Tdmsks);
- dml2_printf("DML::%s: TWait = %f\n", __func__, p->TWait);
- dml2_printf("DML::%s: TWait_p = %f\n", __func__, s->TWait_p);
- dml2_printf("DML::%s: Ttrip = %f\n", __func__, p->Ttrip);
- dml2_printf("DML::%s: Tex = %f\n", __func__, p->ExtraLatencyPrefetch);
- dml2_printf("DML::%s: Tdmdl_vm: %fus - time for vm stages of dmd \n", __func__, *p->Tdmdl_vm);
- dml2_printf("DML::%s: Tdmdl: %fus - time for fabric to become ready and fetch dmd \n", __func__, *p->Tdmdl);
- dml2_printf("DML::%s: TWait_p: %fus\n", __func__, s->TWait_p);
- dml2_printf("DML::%s: Ttrip: %fus\n", __func__, p->Ttrip);
- dml2_printf("DML::%s: DSTXAfterScaler: %u pixels - number of pixel clocks pipeline and buffer delay after scaler \n", __func__, *p->DSTXAfterScaler);
- dml2_printf("DML::%s: DSTYAfterScaler: %u lines - number of lines of pipeline and buffer delay after scaler \n", __func__, *p->DSTYAfterScaler);
- dml2_printf("DML::%s: vm_bytes: %f (hvm inefficiency scaled)\n", __func__, vm_bytes*p->HostVMInefficiencyFactor);
- dml2_printf("DML::%s: row_bytes: %f (hvm inefficiency scaled, 1 row)\n", __func__, p->PixelPTEBytesPerRow*p->HostVMInefficiencyFactor+p->meta_row_bytes+tdlut_row_bytes);
- dml2_printf("DML::%s: Tno_bw: %f\n", __func__, *p->Tno_bw);
- dml2_printf("DML::%s: Tpre=%f Tpre_rounded: %f, delta=%f\n", __func__, Tpre, s->Tpre_rounded, (s->Tpre_rounded - Tpre));
- dml2_printf("DML::%s: Tvm_trips=%f Tvm_trips_rounded: %f, delta=%f\n", __func__, *p->Tvm_trips, s->Tvm_trips_rounded, (s->Tvm_trips_rounded - *p->Tvm_trips));
+ *p->Tpre_rounded = s->dst_y_prefetch_equ * s->LineTime;
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: dst_y_prefetch_equ: %f (after round)\n", __func__, s->dst_y_prefetch_equ);
+ DML_LOG_VERBOSE("DML::%s: LineTime: %f\n", __func__, s->LineTime);
+ DML_LOG_VERBOSE("DML::%s: VStartup: %u\n", __func__, p->VStartup);
+ DML_LOG_VERBOSE("DML::%s: Tvstartup: %fus - time between vstartup and first pixel of active\n", __func__, p->VStartup * s->LineTime);
+ DML_LOG_VERBOSE("DML::%s: TSetup: %fus - time from vstartup to vready\n", __func__, *p->TSetup);
+ DML_LOG_VERBOSE("DML::%s: TCalc: %fus - time for calculations in dchub starting at vready\n", __func__, p->TCalc);
+ DML_LOG_VERBOSE("DML::%s: TWait: %fus - time for fabric to become ready max(pstate exit,cstate enter/exit, urgent latency) after TCalc\n", __func__, p->TWait);
+ DML_LOG_VERBOSE("DML::%s: Tdmbf: %fus - time for dmd transfer from dchub to dio output buffer\n", __func__, s->Tdmbf);
+ DML_LOG_VERBOSE("DML::%s: Tdmec: %fus - time dio takes to transfer dmd\n", __func__, s->Tdmec);
+ DML_LOG_VERBOSE("DML::%s: Tdmsks: %fus - time before active dmd must complete transmission at dio\n", __func__, s->Tdmsks);
+ DML_LOG_VERBOSE("DML::%s: TWait = %f\n", __func__, p->TWait);
+ DML_LOG_VERBOSE("DML::%s: TWait_p = %f\n", __func__, s->TWait_p);
+ DML_LOG_VERBOSE("DML::%s: Ttrip = %f\n", __func__, p->Ttrip);
+ DML_LOG_VERBOSE("DML::%s: Tex = %f\n", __func__, p->ExtraLatencyPrefetch);
+ DML_LOG_VERBOSE("DML::%s: Tdmdl_vm: %fus - time for vm stages of dmd \n", __func__, *p->Tdmdl_vm);
+ DML_LOG_VERBOSE("DML::%s: Tdmdl: %fus - time for fabric to become ready and fetch dmd \n", __func__, *p->Tdmdl);
+ DML_LOG_VERBOSE("DML::%s: TWait_p: %fus\n", __func__, s->TWait_p);
+ DML_LOG_VERBOSE("DML::%s: Ttrip: %fus\n", __func__, p->Ttrip);
+ DML_LOG_VERBOSE("DML::%s: DSTXAfterScaler: %u pixels - number of pixel clocks pipeline and buffer delay after scaler \n", __func__, *p->DSTXAfterScaler);
+ DML_LOG_VERBOSE("DML::%s: DSTYAfterScaler: %u lines - number of lines of pipeline and buffer delay after scaler \n", __func__, *p->DSTYAfterScaler);
+ DML_LOG_VERBOSE("DML::%s: vm_bytes: %f (hvm inefficiency scaled)\n", __func__, vm_bytes*p->HostVMInefficiencyFactor);
+ DML_LOG_VERBOSE("DML::%s: row_bytes: %f (hvm inefficiency scaled, 1 row)\n", __func__, p->PixelPTEBytesPerRow*p->HostVMInefficiencyFactor+p->meta_row_bytes+tdlut_row_bytes);
+ DML_LOG_VERBOSE("DML::%s: Tno_bw: %f\n", __func__, *p->Tno_bw);
+ DML_LOG_VERBOSE("DML::%s: Tpre=%f Tpre_rounded: %f, delta=%f\n", __func__, (s->dst_y_prefetch_equ * s->LineTime), *p->Tpre_rounded, (*p->Tpre_rounded - (s->dst_y_prefetch_equ * s->LineTime)));
+ DML_LOG_VERBOSE("DML::%s: Tvm_trips=%f Tvm_trips_rounded: %f, delta=%f\n", __func__, *p->Tvm_trips, s->Tvm_trips_rounded, (s->Tvm_trips_rounded - *p->Tvm_trips));
#endif
*p->dst_y_per_vm_vblank = 0;
@@ -5434,87 +5507,94 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
// Tpre_rounded is Tpre rounding to 2-bit fraction
// Tvm_trips_rounded is Tvm_trips ceiling to 1/4 line time
// Tr0_trips_rounded is Tr0_trips ceiling to 1/4 line time
- // So that means prefetch bw calculated can be higher since the total time availabe for prefetch is less
- bool min_Lsw_equ_ok = s->Tpre_rounded >= s->Tvm_trips_rounded + 2.0*s->Tr0_trips_rounded + s->min_Lsw_equ*s->LineTime;
+ // So that means prefetch bw calculated can be higher since the total time available for prefetch is less
+ bool min_Lsw_equ_ok = *p->Tpre_rounded >= s->Tvm_trips_rounded + 2.0*s->Tr0_trips_rounded + s->min_Lsw_equ*s->LineTime;
+ bool tpre_gt_req_latency = true;
+#if 0
+ // Check that Tpre_rounded is big enough if all of the stages of the prefetch are time constrained.
+ // The terms Tvm_trips_rounded and Tr0_trips_rounded represent the min time constraints for the VM and row stages.
+ // Normally, these terms cover the overall time constraint for Tpre >= (Tex + max{Ttrip, Turg}), but if these terms are at their minimum, an explicit check is necessary.
+ tpre_gt_req_latency = *p->Tpre_rounded > (math_max2(p->Turg, s->trip_to_mem) + p->ExtraLatencyPrefetch);
+#endif
- if (s->dst_y_prefetch_equ > 1 && min_Lsw_equ_ok) {
+ if (s->dst_y_prefetch_equ > 1 && min_Lsw_equ_ok && tpre_gt_req_latency) {
s->prefetch_bw1 = 0.;
s->prefetch_bw2 = 0.;
s->prefetch_bw3 = 0.;
s->prefetch_bw4 = 0.;
// prefetch_bw1: VM + 2*R0 + SW
- if (s->Tpre_rounded - *p->Tno_bw > 0) {
+ if (*p->Tpre_rounded - *p->Tno_bw > 0) {
s->prefetch_bw1 = (vm_bytes * p->HostVMInefficiencyFactor
+ 2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes)
- + s->prefetch_sw_bytes)
- / (s->Tpre_rounded - *p->Tno_bw);
- s->Tsw_est1 = s->prefetch_sw_bytes / s->prefetch_bw1;
+ + *p->prefetch_sw_bytes)
+ / (*p->Tpre_rounded - *p->Tno_bw);
+ s->Tsw_est1 = *p->prefetch_sw_bytes / s->prefetch_bw1;
} else
s->prefetch_bw1 = 0;
- dml2_printf("DML::%s: prefetch_bw1: %f\n", __func__, s->prefetch_bw1);
- if ((s->Tsw_est1 < s->min_Lsw_equ * s->LineTime) && (s->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.75 * s->LineTime - *p->Tno_bw > 0)) {
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw1: %f\n", __func__, s->prefetch_bw1);
+ if ((s->Tsw_est1 < s->min_Lsw_equ * s->LineTime) && (*p->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.75 * s->LineTime - *p->Tno_bw > 0)) {
s->prefetch_bw1 = (vm_bytes * p->HostVMInefficiencyFactor + 2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes)) /
- (s->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.75 * s->LineTime - *p->Tno_bw);
+ (*p->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.75 * s->LineTime - *p->Tno_bw);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: vm and 2 rows bytes = %f\n", __func__, (vm_bytes * p->HostVMInefficiencyFactor + 2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes)));
- dml2_printf("DML::%s: Tpre_rounded = %f\n", __func__, s->Tpre_rounded);
- dml2_printf("DML::%s: minus term = %f\n", __func__, s->min_Lsw_equ * s->LineTime + 0.75 * s->LineTime + *p->Tno_bw);
- dml2_printf("DML::%s: min_Lsw_equ = %f\n", __func__, s->min_Lsw_equ);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, s->LineTime);
- dml2_printf("DML::%s: Tno_bw = %f\n", __func__, *p->Tno_bw);
- dml2_printf("DML::%s: Time to fetch vm and 2 rows = %f\n", __func__, (s->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.75 * s->LineTime - *p->Tno_bw));
- dml2_printf("DML::%s: prefetch_bw1: %f (updated)\n", __func__, s->prefetch_bw1);
+ DML_LOG_VERBOSE("DML::%s: vm and 2 rows bytes = %f\n", __func__, (vm_bytes * p->HostVMInefficiencyFactor + 2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes)));
+ DML_LOG_VERBOSE("DML::%s: Tpre_rounded = %f\n", __func__, *p->Tpre_rounded);
+ DML_LOG_VERBOSE("DML::%s: minus term = %f\n", __func__, s->min_Lsw_equ * s->LineTime + 0.75 * s->LineTime + *p->Tno_bw);
+ DML_LOG_VERBOSE("DML::%s: min_Lsw_equ = %f\n", __func__, s->min_Lsw_equ);
+ DML_LOG_VERBOSE("DML::%s: LineTime = %f\n", __func__, s->LineTime);
+ DML_LOG_VERBOSE("DML::%s: Tno_bw = %f\n", __func__, *p->Tno_bw);
+ DML_LOG_VERBOSE("DML::%s: Time to fetch vm and 2 rows = %f\n", __func__, (*p->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.75 * s->LineTime - *p->Tno_bw));
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw1: %f (updated)\n", __func__, s->prefetch_bw1);
#endif
}
// prefetch_bw2: VM + SW
- if (s->Tpre_rounded - *p->Tno_bw - 2.0 * s->Tr0_trips_rounded > 0) {
- s->prefetch_bw2 = (vm_bytes * p->HostVMInefficiencyFactor + s->prefetch_sw_bytes) /
- (s->Tpre_rounded - *p->Tno_bw - 2.0 * s->Tr0_trips_rounded);
- s->Tsw_est2 = s->prefetch_sw_bytes / s->prefetch_bw2;
+ if (*p->Tpre_rounded - *p->Tno_bw - 2.0 * s->Tr0_trips_rounded > 0) {
+ s->prefetch_bw2 = (vm_bytes * p->HostVMInefficiencyFactor + *p->prefetch_sw_bytes) /
+ (*p->Tpre_rounded - *p->Tno_bw - 2.0 * s->Tr0_trips_rounded);
+ s->Tsw_est2 = *p->prefetch_sw_bytes / s->prefetch_bw2;
} else
s->prefetch_bw2 = 0;
- dml2_printf("DML::%s: prefetch_bw2: %f\n", __func__, s->prefetch_bw2);
- if ((s->Tsw_est2 < s->min_Lsw_equ * s->LineTime) && ((s->Tpre_rounded - *p->Tno_bw - 2.0 * s->Tr0_trips_rounded - s->min_Lsw_equ * s->LineTime - 0.25 * s->LineTime) > 0)) {
- s->prefetch_bw2 = vm_bytes * p->HostVMInefficiencyFactor / (s->Tpre_rounded - *p->Tno_bw - 2.0 * s->Tr0_trips_rounded - s->min_Lsw_equ * s->LineTime - 0.25 * s->LineTime);
- dml2_printf("DML::%s: prefetch_bw2: %f (updated)\n", __func__, s->prefetch_bw2);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw2: %f\n", __func__, s->prefetch_bw2);
+ if ((s->Tsw_est2 < s->min_Lsw_equ * s->LineTime) && ((*p->Tpre_rounded - *p->Tno_bw - 2.0 * s->Tr0_trips_rounded - s->min_Lsw_equ * s->LineTime - 0.25 * s->LineTime) > 0)) {
+ s->prefetch_bw2 = vm_bytes * p->HostVMInefficiencyFactor / (*p->Tpre_rounded - *p->Tno_bw - 2.0 * s->Tr0_trips_rounded - s->min_Lsw_equ * s->LineTime - 0.25 * s->LineTime);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw2: %f (updated)\n", __func__, s->prefetch_bw2);
}
// prefetch_bw3: 2*R0 + SW
- if (s->Tpre_rounded - s->Tvm_trips_rounded > 0) {
- s->prefetch_bw3 = (2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) + s->prefetch_sw_bytes) /
- (s->Tpre_rounded - s->Tvm_trips_rounded);
- s->Tsw_est3 = s->prefetch_sw_bytes / s->prefetch_bw3;
+ if (*p->Tpre_rounded - s->Tvm_trips_rounded > 0) {
+ s->prefetch_bw3 = (2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) + *p->prefetch_sw_bytes) /
+ (*p->Tpre_rounded - s->Tvm_trips_rounded);
+ s->Tsw_est3 = *p->prefetch_sw_bytes / s->prefetch_bw3;
} else
s->prefetch_bw3 = 0;
- dml2_printf("DML::%s: prefetch_bw3: %f\n", __func__, s->prefetch_bw3);
- if ((s->Tsw_est3 < s->min_Lsw_equ * s->LineTime) && ((s->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.5 * s->LineTime - s->Tvm_trips_rounded) > 0)) {
- s->prefetch_bw3 = (2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes)) / (s->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.5 * s->LineTime - s->Tvm_trips_rounded);
- dml2_printf("DML::%s: prefetch_bw3: %f (updated)\n", __func__, s->prefetch_bw3);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw3: %f\n", __func__, s->prefetch_bw3);
+ if ((s->Tsw_est3 < s->min_Lsw_equ * s->LineTime) && ((*p->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.5 * s->LineTime - s->Tvm_trips_rounded) > 0)) {
+ s->prefetch_bw3 = (2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes)) / (*p->Tpre_rounded - s->min_Lsw_equ * s->LineTime - 0.5 * s->LineTime - s->Tvm_trips_rounded);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw3: %f (updated)\n", __func__, s->prefetch_bw3);
}
// prefetch_bw4: SW
- if (s->Tpre_rounded - s->Tvm_trips_rounded - 2 * s->Tr0_trips_rounded > 0)
- s->prefetch_bw4 = s->prefetch_sw_bytes / (s->Tpre_rounded - s->Tvm_trips_rounded - 2 * s->Tr0_trips_rounded);
+ if (*p->Tpre_rounded - s->Tvm_trips_rounded - 2 * s->Tr0_trips_rounded > 0)
+ s->prefetch_bw4 = *p->prefetch_sw_bytes / (*p->Tpre_rounded - s->Tvm_trips_rounded - 2 * s->Tr0_trips_rounded);
else
s->prefetch_bw4 = 0;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tno_bw: %f\n", __func__, *p->Tno_bw);
- dml2_printf("DML::%s: Tpre=%f Tpre_rounded: %f, delta=%f\n", __func__, Tpre, s->Tpre_rounded, (s->Tpre_rounded - Tpre));
- dml2_printf("DML::%s: Tvm_trips=%f Tvm_trips_rounded: %f, delta=%f\n", __func__, *p->Tvm_trips, s->Tvm_trips_rounded, (s->Tvm_trips_rounded - *p->Tvm_trips));
- dml2_printf("DML::%s: Tr0_trips=%f Tr0_trips_rounded: %f, delta=%f\n", __func__, *p->Tr0_trips, s->Tr0_trips_rounded, (s->Tr0_trips_rounded - *p->Tr0_trips));
- dml2_printf("DML::%s: Tsw_est1: %f\n", __func__, s->Tsw_est1);
- dml2_printf("DML::%s: Tsw_est2: %f\n", __func__, s->Tsw_est2);
- dml2_printf("DML::%s: Tsw_est3: %f\n", __func__, s->Tsw_est3);
- dml2_printf("DML::%s: prefetch_bw1: %f (final)\n", __func__, s->prefetch_bw1);
- dml2_printf("DML::%s: prefetch_bw2: %f (final)\n", __func__, s->prefetch_bw2);
- dml2_printf("DML::%s: prefetch_bw3: %f (final)\n", __func__, s->prefetch_bw3);
- dml2_printf("DML::%s: prefetch_bw4: %f (final)\n", __func__, s->prefetch_bw4);
+ DML_LOG_VERBOSE("DML::%s: Tno_bw: %f\n", __func__, *p->Tno_bw);
+ DML_LOG_VERBOSE("DML::%s: Tpre=%f Tpre_rounded: %f, delta=%f\n", __func__, s->dst_y_prefetch_equ * s->LineTime, *p->Tpre_rounded, (*p->Tpre_rounded - (s->dst_y_prefetch_equ * s->LineTime)));
+ DML_LOG_VERBOSE("DML::%s: Tvm_trips=%f Tvm_trips_rounded: %f, delta=%f\n", __func__, *p->Tvm_trips, s->Tvm_trips_rounded, (s->Tvm_trips_rounded - *p->Tvm_trips));
+ DML_LOG_VERBOSE("DML::%s: Tr0_trips=%f Tr0_trips_rounded: %f, delta=%f\n", __func__, *p->Tr0_trips, s->Tr0_trips_rounded, (s->Tr0_trips_rounded - *p->Tr0_trips));
+ DML_LOG_VERBOSE("DML::%s: Tsw_est1: %f\n", __func__, s->Tsw_est1);
+ DML_LOG_VERBOSE("DML::%s: Tsw_est2: %f\n", __func__, s->Tsw_est2);
+ DML_LOG_VERBOSE("DML::%s: Tsw_est3: %f\n", __func__, s->Tsw_est3);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw1: %f (final)\n", __func__, s->prefetch_bw1);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw2: %f (final)\n", __func__, s->prefetch_bw2);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw3: %f (final)\n", __func__, s->prefetch_bw3);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw4: %f (final)\n", __func__, s->prefetch_bw4);
#endif
{
bool Case1OK = false;
@@ -5533,14 +5613,14 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
double total_row_bytes = (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes);
- dml2_printf("DML::%s: Tvm_trips_rounded = %f\n", __func__, s->Tvm_trips_rounded);
- dml2_printf("DML::%s: Tr0_trips_rounded = %f\n", __func__, s->Tr0_trips_rounded);
+ DML_LOG_VERBOSE("DML::%s: Tvm_trips_rounded = %f\n", __func__, s->Tvm_trips_rounded);
+ DML_LOG_VERBOSE("DML::%s: Tr0_trips_rounded = %f\n", __func__, s->Tr0_trips_rounded);
if (s->prefetch_bw1 > 0) {
double vm_transfer_time = *p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw1;
double row_transfer_time = total_row_bytes / s->prefetch_bw1;
- dml2_printf("DML::%s: Case1: vm_transfer_time = %f\n", __func__, vm_transfer_time);
- dml2_printf("DML::%s: Case1: row_transfer_time = %f\n", __func__, row_transfer_time);
+ DML_LOG_VERBOSE("DML::%s: Case1: vm_transfer_time = %f\n", __func__, vm_transfer_time);
+ DML_LOG_VERBOSE("DML::%s: Case1: row_transfer_time = %f\n", __func__, row_transfer_time);
if (vm_transfer_time >= s->Tvm_trips_rounded && row_transfer_time >= s->Tr0_trips_rounded) {
Case1OK = true;
}
@@ -5553,8 +5633,8 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
if (s->prefetch_bw2 > 0) {
double vm_transfer_time = *p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw2;
double row_transfer_time = total_row_bytes / s->prefetch_bw2;
- dml2_printf("DML::%s: Case2: vm_transfer_time = %f\n", __func__, vm_transfer_time);
- dml2_printf("DML::%s: Case2: row_transfer_time = %f\n", __func__, row_transfer_time);
+ DML_LOG_VERBOSE("DML::%s: Case2: vm_transfer_time = %f\n", __func__, vm_transfer_time);
+ DML_LOG_VERBOSE("DML::%s: Case2: row_transfer_time = %f\n", __func__, row_transfer_time);
if (vm_transfer_time >= s->Tvm_trips_rounded && row_transfer_time < s->Tr0_trips_rounded) {
Case2OK = true;
}
@@ -5566,8 +5646,8 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
if (s->prefetch_bw3 > 0) {
double vm_transfer_time = *p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw3;
double row_transfer_time = total_row_bytes / s->prefetch_bw3;
- dml2_printf("DML::%s: Case3: vm_transfer_time = %f\n", __func__, vm_transfer_time);
- dml2_printf("DML::%s: Case3: row_transfer_time = %f\n", __func__, row_transfer_time);
+ DML_LOG_VERBOSE("DML::%s: Case3: vm_transfer_time = %f\n", __func__, vm_transfer_time);
+ DML_LOG_VERBOSE("DML::%s: Case3: row_transfer_time = %f\n", __func__, row_transfer_time);
if (vm_transfer_time < s->Tvm_trips_rounded && row_transfer_time >= s->Tr0_trips_rounded) {
Case3OK = true;
}
@@ -5587,10 +5667,10 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
p->vm_bytes * p->HostVMInefficiencyFactor / (31 * s->LineTime) - *p->Tno_bw,
(p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / (15 * s->LineTime));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Case1OK: %u\n", __func__, Case1OK);
- dml2_printf("DML::%s: Case2OK: %u\n", __func__, Case2OK);
- dml2_printf("DML::%s: Case3OK: %u\n", __func__, Case3OK);
- dml2_printf("DML::%s: prefetch_bw_equ: %f\n", __func__, s->prefetch_bw_equ);
+ DML_LOG_VERBOSE("DML::%s: Case1OK: %u\n", __func__, Case1OK);
+ DML_LOG_VERBOSE("DML::%s: Case2OK: %u\n", __func__, Case2OK);
+ DML_LOG_VERBOSE("DML::%s: Case3OK: %u\n", __func__, Case3OK);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw_equ: %f\n", __func__, s->prefetch_bw_equ);
#endif
if (s->prefetch_bw_equ > 0) {
@@ -5610,16 +5690,13 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
} else {
s->Tvm_equ = 0;
s->Tr0_equ = 0;
- dml2_printf("DML::%s: prefetch_bw_equ equals 0!\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: prefetch_bw_equ equals 0!\n", __func__);
}
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tvm_equ = %f\n", __func__, s->Tvm_equ);
- dml2_printf("DML::%s: Tr0_equ = %f\n", __func__, s->Tr0_equ);
+ DML_LOG_VERBOSE("DML::%s: Tvm_equ = %f\n", __func__, s->Tvm_equ);
+ DML_LOG_VERBOSE("DML::%s: Tr0_equ = %f\n", __func__, s->Tr0_equ);
#endif
- // Lsw = dst_y_prefetch - (dst_y_per_vm_vblank + 2*dst_y_per_row_vblank)
- s->Lsw_equ = s->dst_y_prefetch_equ - math_ceil2(4.0 * (s->Tvm_equ + 2 * s->Tr0_equ) / s->LineTime, 1.0) / 4.0;
-
// Use the more stressful prefetch schedule
if (s->dst_y_prefetch_oto < s->dst_y_prefetch_equ) {
*p->dst_y_prefetch = s->dst_y_prefetch_oto;
@@ -5628,57 +5705,66 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
*p->dst_y_per_vm_vblank = math_ceil2(4.0 * s->TimeForFetchingVM / s->LineTime, 1.0) / 4.0;
*p->dst_y_per_row_vblank = math_ceil2(4.0 * s->TimeForFetchingRowInVBlank / s->LineTime, 1.0) / 4.0;
- s->dst_y_per_vm_no_trip_vblank = math_ceil2(4.0 * s->Tvm_no_trip_oto / s->LineTime, 1.0) / 4.0;
- s->dst_y_per_row_no_trip_vblank = math_ceil2(4.0 * s->Tr0_no_trip_oto / s->LineTime, 1.0) / 4.0;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Using oto scheduling for prefetch\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: Using oto scheduling for prefetch\n", __func__);
#endif
+
} else {
*p->dst_y_prefetch = s->dst_y_prefetch_equ;
+
+ if (s->dst_y_prefetch_equ < s->dst_y_prefetch_equ_impacted)
+ *p->dst_y_prefetch = s->dst_y_prefetch_equ_impacted;
+
s->TimeForFetchingVM = s->Tvm_equ;
s->TimeForFetchingRowInVBlank = s->Tr0_equ;
*p->dst_y_per_vm_vblank = math_ceil2(4.0 * s->TimeForFetchingVM / s->LineTime, 1.0) / 4.0;
*p->dst_y_per_row_vblank = math_ceil2(4.0 * s->TimeForFetchingRowInVBlank / s->LineTime, 1.0) / 4.0;
- s->dst_y_per_vm_no_trip_vblank = *p->dst_y_per_vm_vblank;
- s->dst_y_per_row_no_trip_vblank = *p->dst_y_per_row_vblank;
+
+ /* equ bw should be propagated so a ceiling of the equ bw is accounted for prior to mode programming.
+ * Overall bandwidth may be lower when going from mode support to mode programming but final pixel data
+ * bandwidth may end up higher than what was calculated in mode support.
+ */
+ *p->RequiredPrefetchBWMax = math_max2(s->prefetch_bw_equ, *p->RequiredPrefetchBWMax);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Using equ bw scheduling for prefetch\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: Using equ bw scheduling for prefetch\n", __func__);
#endif
}
- /* take worst case Lsw to calculate bandwidth requirement regardless of schedule */
- s->LinesToRequestPrefetchPixelData = math_min2(s->Lsw_equ, s->Lsw_oto); // Lsw
+ // Lsw = dst_y_prefetch - (dst_y_per_vm_vblank + 2*dst_y_per_row_vblank)
+ s->LinesToRequestPrefetchPixelData = *p->dst_y_prefetch - *p->dst_y_per_vm_vblank - 2 * *p->dst_y_per_row_vblank; // Lsw
s->cursor_prefetch_bytes = (unsigned int)math_max2(p->cursor_bytes_per_chunk, 4 * p->cursor_bytes_per_line);
*p->prefetch_cursor_bw = p->num_cursors * s->cursor_prefetch_bytes / (s->LinesToRequestPrefetchPixelData * s->LineTime);
+ *p->prefetch_swath_time_us = (s->LinesToRequestPrefetchPixelData * s->LineTime);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: TimeForFetchingVM = %f\n", __func__, s->TimeForFetchingVM);
- dml2_printf("DML::%s: TimeForFetchingRowInVBlank = %f\n", __func__, s->TimeForFetchingRowInVBlank);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, s->LineTime);
- dml2_printf("DML::%s: dst_y_prefetch = %f\n", __func__, *p->dst_y_prefetch);
- dml2_printf("DML::%s: dst_y_per_vm_vblank = %f\n", __func__, *p->dst_y_per_vm_vblank);
- dml2_printf("DML::%s: dst_y_per_row_vblank = %f\n", __func__, *p->dst_y_per_row_vblank);
- dml2_printf("DML::%s: LinesToRequestPrefetchPixelData = %f\n", __func__, s->LinesToRequestPrefetchPixelData);
- dml2_printf("DML::%s: PrefetchSourceLinesY = %f\n", __func__, p->PrefetchSourceLinesY);
+ DML_LOG_VERBOSE("DML::%s: TimeForFetchingVM = %f\n", __func__, s->TimeForFetchingVM);
+ DML_LOG_VERBOSE("DML::%s: TimeForFetchingRowInVBlank = %f\n", __func__, s->TimeForFetchingRowInVBlank);
+ DML_LOG_VERBOSE("DML::%s: LineTime = %f\n", __func__, s->LineTime);
+ DML_LOG_VERBOSE("DML::%s: dst_y_prefetch = %f\n", __func__, *p->dst_y_prefetch);
+ DML_LOG_VERBOSE("DML::%s: dst_y_per_vm_vblank = %f\n", __func__, *p->dst_y_per_vm_vblank);
+ DML_LOG_VERBOSE("DML::%s: dst_y_per_row_vblank = %f\n", __func__, *p->dst_y_per_row_vblank);
+ DML_LOG_VERBOSE("DML::%s: LinesToRequestPrefetchPixelData = %f\n", __func__, s->LinesToRequestPrefetchPixelData);
+ DML_LOG_VERBOSE("DML::%s: PrefetchSourceLinesY = %f\n", __func__, p->PrefetchSourceLinesY);
+ DML_LOG_VERBOSE("DML::%s: prefetch_swath_time_us = %f\n", __func__, *p->prefetch_swath_time_us);
- dml2_printf("DML::%s: cursor_bytes_per_chunk = %d\n", __func__, p->cursor_bytes_per_chunk);
- dml2_printf("DML::%s: cursor_bytes_per_line = %d\n", __func__, p->cursor_bytes_per_line);
- dml2_printf("DML::%s: cursor_prefetch_bytes = %d\n", __func__, s->cursor_prefetch_bytes);
- dml2_printf("DML::%s: prefetch_cursor_bw = %f\n", __func__, *p->prefetch_cursor_bw);
+ DML_LOG_VERBOSE("DML::%s: cursor_bytes_per_chunk = %d\n", __func__, p->cursor_bytes_per_chunk);
+ DML_LOG_VERBOSE("DML::%s: cursor_bytes_per_line = %d\n", __func__, p->cursor_bytes_per_line);
+ DML_LOG_VERBOSE("DML::%s: cursor_prefetch_bytes = %d\n", __func__, s->cursor_prefetch_bytes);
+ DML_LOG_VERBOSE("DML::%s: prefetch_cursor_bw = %f\n", __func__, *p->prefetch_cursor_bw);
#endif
- dml2_assert(*p->dst_y_prefetch < 64);
+ DML_ASSERT(*p->dst_y_prefetch < 64);
unsigned int min_lsw_required = (unsigned int)math_max2(2, p->tdlut_drain_time / s->LineTime);
if (s->LinesToRequestPrefetchPixelData >= min_lsw_required && s->prefetch_bw_equ > 0) {
*p->VRatioPrefetchY = (double)p->PrefetchSourceLinesY / s->LinesToRequestPrefetchPixelData;
*p->VRatioPrefetchY = math_max2(*p->VRatioPrefetchY, 1.0);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatioPrefetchY = %f\n", __func__, *p->VRatioPrefetchY);
- dml2_printf("DML::%s: SwathHeightY = %u\n", __func__, p->SwathHeightY);
- dml2_printf("DML::%s: VInitPreFillY = %u\n", __func__, p->VInitPreFillY);
+ DML_LOG_VERBOSE("DML::%s: VRatioPrefetchY = %f\n", __func__, *p->VRatioPrefetchY);
+ DML_LOG_VERBOSE("DML::%s: SwathHeightY = %u\n", __func__, p->SwathHeightY);
+ DML_LOG_VERBOSE("DML::%s: VInitPreFillY = %u\n", __func__, p->VInitPreFillY);
#endif
if ((p->SwathHeightY > 4) && (p->VInitPreFillY > 3)) {
if (s->LinesToRequestPrefetchPixelData > (p->VInitPreFillY - 3.0) / 2.0) {
@@ -5686,13 +5772,13 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
(double)p->MaxNumSwathY * p->SwathHeightY / (s->LinesToRequestPrefetchPixelData - (p->VInitPreFillY - 3.0) / 2.0));
} else {
s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: No time to prefetch!. LinesToRequestPrefetchPixelData=%f VinitPreFillY=%u\n", __func__, s->LinesToRequestPrefetchPixelData, p->VInitPreFillY);
+ DML_LOG_VERBOSE("DML::%s: No time to prefetch!. LinesToRequestPrefetchPixelData=%f VinitPreFillY=%u\n", __func__, s->LinesToRequestPrefetchPixelData, p->VInitPreFillY);
*p->VRatioPrefetchY = 0;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatioPrefetchY = %f\n", __func__, *p->VRatioPrefetchY);
- dml2_printf("DML::%s: PrefetchSourceLinesY = %f\n", __func__, p->PrefetchSourceLinesY);
- dml2_printf("DML::%s: MaxNumSwathY = %u\n", __func__, p->MaxNumSwathY);
+ DML_LOG_VERBOSE("DML::%s: VRatioPrefetchY = %f\n", __func__, *p->VRatioPrefetchY);
+ DML_LOG_VERBOSE("DML::%s: PrefetchSourceLinesY = %f\n", __func__, p->PrefetchSourceLinesY);
+ DML_LOG_VERBOSE("DML::%s: MaxNumSwathY = %u\n", __func__, p->MaxNumSwathY);
#endif
}
@@ -5700,22 +5786,22 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
*p->VRatioPrefetchC = math_max2(*p->VRatioPrefetchC, 1.0);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatioPrefetchC = %f\n", __func__, *p->VRatioPrefetchC);
- dml2_printf("DML::%s: SwathHeightC = %u\n", __func__, p->SwathHeightC);
- dml2_printf("DML::%s: VInitPreFillC = %u\n", __func__, p->VInitPreFillC);
+ DML_LOG_VERBOSE("DML::%s: VRatioPrefetchC = %f\n", __func__, *p->VRatioPrefetchC);
+ DML_LOG_VERBOSE("DML::%s: SwathHeightC = %u\n", __func__, p->SwathHeightC);
+ DML_LOG_VERBOSE("DML::%s: VInitPreFillC = %u\n", __func__, p->VInitPreFillC);
#endif
if ((p->SwathHeightC > 4) && (p->VInitPreFillC > 3)) {
if (s->LinesToRequestPrefetchPixelData > (p->VInitPreFillC - 3.0) / 2.0) {
*p->VRatioPrefetchC = math_max2(*p->VRatioPrefetchC, (double)p->MaxNumSwathC * p->SwathHeightC / (s->LinesToRequestPrefetchPixelData - (p->VInitPreFillC - 3.0) / 2.0));
} else {
s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: No time to prefetch!. LinesToRequestPrefetchPixelData=%f VInitPreFillC=%u\n", __func__, s->LinesToRequestPrefetchPixelData, p->VInitPreFillC);
+ DML_LOG_VERBOSE("DML::%s: No time to prefetch!. LinesToRequestPrefetchPixelData=%f VInitPreFillC=%u\n", __func__, s->LinesToRequestPrefetchPixelData, p->VInitPreFillC);
*p->VRatioPrefetchC = 0;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatioPrefetchC = %f\n", __func__, *p->VRatioPrefetchC);
- dml2_printf("DML::%s: PrefetchSourceLinesC = %f\n", __func__, p->PrefetchSourceLinesC);
- dml2_printf("DML::%s: MaxNumSwathC = %u\n", __func__, p->MaxNumSwathC);
+ DML_LOG_VERBOSE("DML::%s: VRatioPrefetchC = %f\n", __func__, *p->VRatioPrefetchC);
+ DML_LOG_VERBOSE("DML::%s: PrefetchSourceLinesC = %f\n", __func__, p->PrefetchSourceLinesC);
+ DML_LOG_VERBOSE("DML::%s: MaxNumSwathC = %u\n", __func__, p->MaxNumSwathC);
#endif
}
@@ -5723,34 +5809,34 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
*p->RequiredPrefetchPixelDataBWChroma = (double)p->PrefetchSourceLinesC / s->LinesToRequestPrefetchPixelData * p->myPipe->BytePerPixelC * p->swath_width_chroma_ub / s->LineTime;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: BytePerPixelY = %u\n", __func__, p->myPipe->BytePerPixelY);
- dml2_printf("DML::%s: swath_width_luma_ub = %u\n", __func__, p->swath_width_luma_ub);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, s->LineTime);
- dml2_printf("DML::%s: RequiredPrefetchPixelDataBWLuma = %f\n", __func__, *p->RequiredPrefetchPixelDataBWLuma);
- dml2_printf("DML::%s: RequiredPrefetchPixelDataBWChroma = %f\n", __func__, *p->RequiredPrefetchPixelDataBWChroma);
+ DML_LOG_VERBOSE("DML::%s: BytePerPixelY = %u\n", __func__, p->myPipe->BytePerPixelY);
+ DML_LOG_VERBOSE("DML::%s: swath_width_luma_ub = %u\n", __func__, p->swath_width_luma_ub);
+ DML_LOG_VERBOSE("DML::%s: LineTime = %f\n", __func__, s->LineTime);
+ DML_LOG_VERBOSE("DML::%s: RequiredPrefetchPixelDataBWLuma = %f\n", __func__, *p->RequiredPrefetchPixelDataBWLuma);
+ DML_LOG_VERBOSE("DML::%s: RequiredPrefetchPixelDataBWChroma = %f\n", __func__, *p->RequiredPrefetchPixelDataBWChroma);
#endif
} else {
s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: No time to prefetch!, LinesToRequestPrefetchPixelData: %f, should be >= %d\n", __func__, s->LinesToRequestPrefetchPixelData, min_lsw_required);
- dml2_printf("DML::%s: No time to prefetch!, prefetch_bw_equ: %f, should be > 0\n", __func__, s->prefetch_bw_equ);
+ DML_LOG_VERBOSE("DML::%s: No time to prefetch!, LinesToRequestPrefetchPixelData: %f, should be >= %d\n", __func__, s->LinesToRequestPrefetchPixelData, min_lsw_required);
+ DML_LOG_VERBOSE("DML::%s: No time to prefetch!, prefetch_bw_equ: %f, should be > 0\n", __func__, s->prefetch_bw_equ);
*p->VRatioPrefetchY = 0;
*p->VRatioPrefetchC = 0;
*p->RequiredPrefetchPixelDataBWLuma = 0;
*p->RequiredPrefetchPixelDataBWChroma = 0;
}
- dml2_printf("DML: Tpre: %fus - sum of time to request 2 x data pte, swaths\n", (double)s->LinesToRequestPrefetchPixelData * s->LineTime + 2.0 * s->TimeForFetchingRowInVBlank + s->TimeForFetchingVM);
- dml2_printf("DML: Tvm: %fus - time to fetch vm\n", s->TimeForFetchingVM);
- dml2_printf("DML: Tr0: %fus - time to fetch first row of data pagetables\n", s->TimeForFetchingRowInVBlank);
- dml2_printf("DML: Tsw: %fus = time to fetch enough pixel data and cursor data to feed the scalers init position and detile\n", (double)s->LinesToRequestPrefetchPixelData * s->LineTime);
- dml2_printf("DML: To: %fus - time for propagation from scaler to optc\n", (*p->DSTYAfterScaler + ((double)(*p->DSTXAfterScaler) / (double)p->myPipe->HTotal)) * s->LineTime);
- dml2_printf("DML: Tvstartup - TSetup - Tcalc - TWait - Tpre - To > 0\n");
- dml2_printf("DML: Tslack(pre): %fus - time left over in schedule\n", p->VStartup * s->LineTime - s->TimeForFetchingVM - 2 * s->TimeForFetchingRowInVBlank - (*p->DSTYAfterScaler + ((double)(*p->DSTXAfterScaler) / (double)p->myPipe->HTotal)) * s->LineTime - p->TWait - p->TCalc - *p->TSetup);
- dml2_printf("DML: row_bytes = dpte_row_bytes (per_pipe) = PixelPTEBytesPerRow = : %u\n", p->PixelPTEBytesPerRow);
+ DML_LOG_VERBOSE("DML: Tpre: %fus - sum of time to request 2 x data pte, swaths\n", (double)s->LinesToRequestPrefetchPixelData * s->LineTime + 2.0 * s->TimeForFetchingRowInVBlank + s->TimeForFetchingVM);
+ DML_LOG_VERBOSE("DML: Tvm: %fus - time to fetch vm\n", s->TimeForFetchingVM);
+ DML_LOG_VERBOSE("DML: Tr0: %fus - time to fetch first row of data pagetables\n", s->TimeForFetchingRowInVBlank);
+ DML_LOG_VERBOSE("DML: Tsw: %fus = time to fetch enough pixel data and cursor data to feed the scalers init position and detile\n", (double)s->LinesToRequestPrefetchPixelData * s->LineTime);
+ DML_LOG_VERBOSE("DML: To: %fus - time for propagation from scaler to optc\n", (*p->DSTYAfterScaler + ((double)(*p->DSTXAfterScaler) / (double)p->myPipe->HTotal)) * s->LineTime);
+ DML_LOG_VERBOSE("DML: Tvstartup - TSetup - Tcalc - TWait - Tpre - To > 0\n");
+ DML_LOG_VERBOSE("DML: Tslack(pre): %fus - time left over in schedule\n", p->VStartup * s->LineTime - s->TimeForFetchingVM - 2 * s->TimeForFetchingRowInVBlank - (*p->DSTYAfterScaler + ((double)(*p->DSTXAfterScaler) / (double)p->myPipe->HTotal)) * s->LineTime - p->TWait - p->TCalc - *p->TSetup);
+ DML_LOG_VERBOSE("DML: row_bytes = dpte_row_bytes (per_pipe) = PixelPTEBytesPerRow = : %u\n", p->PixelPTEBytesPerRow);
} else {
- dml2_printf("DML::%s: No time to prefetch! dst_y_prefetch_equ = %f (should be > 1)\n", __func__, s->dst_y_prefetch_equ);
- dml2_printf("DML::%s: No time to prefetch! min_Lsw_equ_ok = %d, Tpre_rounded (%f) should be >= Tvm_trips_rounded (%f) + 2.0*Tr0_trips_rounded (%f) + min_Tsw_equ (%f)\n",
- __func__, min_Lsw_equ_ok, s->Tpre_rounded, s->Tvm_trips_rounded, 2.0*s->Tr0_trips_rounded, s->min_Lsw_equ*s->LineTime);
+ DML_LOG_VERBOSE("DML::%s: No time to prefetch! dst_y_prefetch_equ = %f (should be > 1)\n", __func__, s->dst_y_prefetch_equ);
+ DML_LOG_VERBOSE("DML::%s: No time to prefetch! min_Lsw_equ_ok = %d, Tpre_rounded (%f) should be >= Tvm_trips_rounded (%f) + 2.0*Tr0_trips_rounded (%f) + min_Tsw_equ (%f)\n",
+ __func__, min_Lsw_equ_ok, *p->Tpre_rounded, s->Tvm_trips_rounded, 2.0*s->Tr0_trips_rounded, s->min_Lsw_equ*s->LineTime);
s->NoTimeToPrefetch = true;
s->TimeForFetchingVM = 0;
s->TimeForFetchingRowInVBlank = 0;
@@ -5769,36 +5855,36 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
if (vm_bytes == 0) {
prefetch_vm_bw = 0;
- } else if (s->dst_y_per_vm_no_trip_vblank > 0) {
+ } else if (*p->dst_y_per_vm_vblank > 0) {
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, p->HostVMInefficiencyFactor);
- dml2_printf("DML::%s: dst_y_per_vm_vblank = %f\n", __func__, *p->dst_y_per_vm_vblank);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, s->LineTime);
+ DML_LOG_VERBOSE("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, p->HostVMInefficiencyFactor);
+ DML_LOG_VERBOSE("DML::%s: dst_y_per_vm_vblank = %f\n", __func__, *p->dst_y_per_vm_vblank);
+ DML_LOG_VERBOSE("DML::%s: LineTime = %f\n", __func__, s->LineTime);
#endif
- prefetch_vm_bw = vm_bytes * p->HostVMInefficiencyFactor / (s->dst_y_per_vm_no_trip_vblank * s->LineTime);
+ prefetch_vm_bw = vm_bytes * p->HostVMInefficiencyFactor / (*p->dst_y_per_vm_vblank * s->LineTime);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: prefetch_vm_bw = %f\n", __func__, prefetch_vm_bw);
+ DML_LOG_VERBOSE("DML::%s: prefetch_vm_bw = %f\n", __func__, prefetch_vm_bw);
#endif
} else {
prefetch_vm_bw = 0;
s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: No time to prefetch!. dst_y_per_vm_vblank=%f (should be > 0)\n", __func__, *p->dst_y_per_vm_vblank);
+ DML_LOG_VERBOSE("DML::%s: No time to prefetch!. dst_y_per_vm_vblank=%f (should be > 0)\n", __func__, *p->dst_y_per_vm_vblank);
}
if (p->PixelPTEBytesPerRow == 0 && tdlut_row_bytes == 0) {
prefetch_row_bw = 0;
- } else if (s->dst_y_per_row_no_trip_vblank > 0) {
- prefetch_row_bw = (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + tdlut_row_bytes) / (s->dst_y_per_row_no_trip_vblank * s->LineTime);
+ } else if (*p->dst_y_per_row_vblank > 0) {
+ prefetch_row_bw = (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + tdlut_row_bytes) / (*p->dst_y_per_row_vblank * s->LineTime);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, p->PixelPTEBytesPerRow);
- dml2_printf("DML::%s: dst_y_per_row_vblank = %f\n", __func__, *p->dst_y_per_row_vblank);
- dml2_printf("DML::%s: prefetch_row_bw = %f\n", __func__, prefetch_row_bw);
+ DML_LOG_VERBOSE("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, p->PixelPTEBytesPerRow);
+ DML_LOG_VERBOSE("DML::%s: dst_y_per_row_vblank = %f\n", __func__, *p->dst_y_per_row_vblank);
+ DML_LOG_VERBOSE("DML::%s: prefetch_row_bw = %f\n", __func__, prefetch_row_bw);
#endif
} else {
prefetch_row_bw = 0;
s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: No time to prefetch!. dst_y_per_row_vblank=%f (should be > 0)\n", __func__, *p->dst_y_per_row_vblank);
+ DML_LOG_VERBOSE("DML::%s: No time to prefetch!. dst_y_per_row_vblank=%f (should be > 0)\n", __func__, *p->dst_y_per_row_vblank);
}
*p->prefetch_vmrow_bw = math_max2(prefetch_vm_bw, prefetch_row_bw);
@@ -5818,16 +5904,181 @@ static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch
*p->prefetch_vmrow_bw = 0;
}
- dml2_printf("DML::%s: dst_y_per_vm_vblank = %f (final)\n", __func__, *p->dst_y_per_vm_vblank);
- dml2_printf("DML::%s: dst_y_per_row_vblank = %f (final)\n", __func__, *p->dst_y_per_row_vblank);
- dml2_printf("DML::%s: prefetch_vmrow_bw = %f (final)\n", __func__, *p->prefetch_vmrow_bw);
- dml2_printf("DML::%s: RequiredPrefetchPixelDataBWLuma = %f (final)\n", __func__, *p->RequiredPrefetchPixelDataBWLuma);
- dml2_printf("DML::%s: RequiredPrefetchPixelDataBWChroma = %f (final)\n", __func__, *p->RequiredPrefetchPixelDataBWChroma);
- dml2_printf("DML::%s: NoTimeToPrefetch=%d\n", __func__, s->NoTimeToPrefetch);
+ DML_LOG_VERBOSE("DML::%s: dst_y_per_vm_vblank = %f (final)\n", __func__, *p->dst_y_per_vm_vblank);
+ DML_LOG_VERBOSE("DML::%s: dst_y_per_row_vblank = %f (final)\n", __func__, *p->dst_y_per_row_vblank);
+ DML_LOG_VERBOSE("DML::%s: prefetch_vmrow_bw = %f (final)\n", __func__, *p->prefetch_vmrow_bw);
+ DML_LOG_VERBOSE("DML::%s: RequiredPrefetchPixelDataBWLuma = %f (final)\n", __func__, *p->RequiredPrefetchPixelDataBWLuma);
+ DML_LOG_VERBOSE("DML::%s: RequiredPrefetchPixelDataBWChroma = %f (final)\n", __func__, *p->RequiredPrefetchPixelDataBWChroma);
+ DML_LOG_VERBOSE("DML::%s: NoTimeToPrefetch=%d\n", __func__, s->NoTimeToPrefetch);
return s->NoTimeToPrefetch;
}
+static unsigned int get_num_lb_source_lines(unsigned int max_line_buffer_lines,
+ unsigned int line_buffer_size_bits,
+ unsigned int num_pipes,
+ unsigned int vp_width,
+ unsigned int vp_height,
+ double h_ratio,
+ enum dml2_rotation_angle rotation_angle)
+{
+ unsigned int num_lb_source_lines = 0;
+ double lb_bit_per_pixel = 57.0;
+ unsigned recin_width = vp_width/num_pipes;
+
+ if (dml_is_vertical_rotation(rotation_angle))
+ recin_width = vp_height/num_pipes;
+
+ num_lb_source_lines = (unsigned int) math_min2((double) max_line_buffer_lines,
+ math_floor2(line_buffer_size_bits / lb_bit_per_pixel / (recin_width / math_max2(h_ratio, 1.0)), 1.0));
+
+ return num_lb_source_lines;
+}
+
+static unsigned int find_max_impact_plane(unsigned int this_plane_idx, unsigned int num_planes, unsigned int Trpd_dcfclk_cycles[])
+{
+ int max_value = -1;
+ int max_idx = -1;
+ for (unsigned int i = 0; i < num_planes; i++) {
+ if (i != this_plane_idx && (int) Trpd_dcfclk_cycles[i] > max_value) {
+ max_value = Trpd_dcfclk_cycles[i];
+ max_idx = i;
+ }
+ }
+ if (max_idx <= 0) {
+ DML_ASSERT(max_idx >= 0);
+ max_idx = this_plane_idx;
+ }
+
+ return max_idx;
+}
+
+static double calculate_impacted_Tsw(unsigned int exclude_plane_idx, unsigned int num_planes, double *prefetch_swath_bytes, double bw_mbps)
+{
+ double sum = 0.;
+ for (unsigned int i = 0; i < num_planes; i++) {
+ if (i != exclude_plane_idx) {
+ sum += prefetch_swath_bytes[i];
+ }
+ }
+ return sum / bw_mbps;
+}
+
+// a global check against the aggregate effect of the per plane prefetch schedule
+static noinline_for_stack bool CheckGlobalPrefetchAdmissibility(struct dml2_core_internal_scratch *scratch,
+ struct dml2_core_calcs_CheckGlobalPrefetchAdmissibility_params *p)
+{
+ struct dml2_core_calcs_CheckGlobalPrefetchAdmissibility_locals *s = &scratch->CheckGlobalPrefetchAdmissibility_locals;
+ unsigned int i, k;
+
+ memset(s, 0, sizeof(struct dml2_core_calcs_CheckGlobalPrefetchAdmissibility_locals));
+
+ *p->recalc_prefetch_schedule = 0;
+ s->prefetch_global_check_passed = 1;
+ // worst case if the rob and cdb is fully hogged
+ s->max_Trpd_dcfclk_cycles = (unsigned int) math_ceil2((p->rob_buffer_size_kbytes*1024 + p->compressed_buffer_size_kbytes*DML_MAX_COMPRESSION_RATIO*1024)/64.0, 1.0);
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: num_active_planes = %d\n", __func__, p->num_active_planes);
+ DML_LOG_VERBOSE("DML::%s: rob_buffer_size_kbytes = %d\n", __func__, p->rob_buffer_size_kbytes);
+ DML_LOG_VERBOSE("DML::%s: compressed_buffer_size_kbytes = %d\n", __func__, p->compressed_buffer_size_kbytes);
+ DML_LOG_VERBOSE("DML::%s: estimated_urg_bandwidth_required_mbps = %f\n", __func__, p->estimated_urg_bandwidth_required_mbps);
+ DML_LOG_VERBOSE("DML::%s: estimated_dcfclk_mhz = %f\n", __func__, p->estimated_dcfclk_mhz);
+ DML_LOG_VERBOSE("DML::%s: max_Trpd_dcfclk_cycles = %u\n", __func__, s->max_Trpd_dcfclk_cycles);
+#endif
+
+ // calculate the return impact from each plane, request is 256B per dcfclk
+ for (i = 0; i < p->num_active_planes; i++) {
+ s->src_detile_buf_size_bytes_l[i] = p->detile_buffer_size_bytes_l[i];
+ s->src_detile_buf_size_bytes_c[i] = p->detile_buffer_size_bytes_c[i];
+ s->src_swath_bytes_l[i] = p->full_swath_bytes_l[i];
+ s->src_swath_bytes_c[i] = p->full_swath_bytes_c[i];
+
+ if (p->pixel_format[i] == dml2_420_10) {
+ s->src_detile_buf_size_bytes_l[i] = (unsigned int) (s->src_detile_buf_size_bytes_l[i] * 1.5);
+ s->src_detile_buf_size_bytes_c[i] = (unsigned int) (s->src_detile_buf_size_bytes_c[i] * 1.5);
+ s->src_swath_bytes_l[i] = (unsigned int) (s->src_swath_bytes_l[i] * 1.5);
+ s->src_swath_bytes_c[i] = (unsigned int) (s->src_swath_bytes_c[i] * 1.5);
+ }
+
+ s->burst_bytes_to_fill_det = (unsigned int) (math_floor2(s->src_detile_buf_size_bytes_l[i] / p->chunk_bytes_l, 1) * p->chunk_bytes_l);
+ s->burst_bytes_to_fill_det += (unsigned int) (math_floor2(p->lb_source_lines_l[i] / p->swath_height_l[i], 1) * s->src_swath_bytes_l[i]);
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: i=%u pixel_format = %d\n", __func__, i, p->pixel_format[i]);
+ DML_LOG_VERBOSE("DML::%s: i=%u chunk_bytes_l = %d\n", __func__, i, p->chunk_bytes_l);
+ DML_LOG_VERBOSE("DML::%s: i=%u lb_source_lines_l = %d\n", __func__, i, p->lb_source_lines_l[i]);
+ DML_LOG_VERBOSE("DML::%s: i=%u src_detile_buf_size_bytes_l=%d\n", __func__, i, s->src_detile_buf_size_bytes_l[i]);
+ DML_LOG_VERBOSE("DML::%s: i=%u src_swath_bytes_l=%d\n", __func__, i, s->src_swath_bytes_l[i]);
+ DML_LOG_VERBOSE("DML::%s: i=%u burst_bytes_to_fill_det=%d (luma)\n", __func__, i, s->burst_bytes_to_fill_det);
+#endif
+
+ if (s->src_swath_bytes_c[i] > 0) { // dual_plane
+ s->burst_bytes_to_fill_det += (unsigned int) (math_floor2(s->src_detile_buf_size_bytes_c[i] / p->chunk_bytes_c, 1) * p->chunk_bytes_c);
+
+ if (p->pixel_format[i] == dml2_422_planar_8 || p->pixel_format[i] == dml2_422_planar_10 || p->pixel_format[i] == dml2_422_planar_12) {
+ s->burst_bytes_to_fill_det += (unsigned int) (math_floor2(p->lb_source_lines_c[i] / p->swath_height_c[i], 1) * s->src_swath_bytes_c[i]);
+ }
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: i=%u chunk_bytes_c = %d\n", __func__, i, p->chunk_bytes_c);
+ DML_LOG_VERBOSE("DML::%s: i=%u lb_source_lines_c = %d\n", __func__, i, p->lb_source_lines_c[i]);
+ DML_LOG_VERBOSE("DML::%s: i=%u src_detile_buf_size_bytes_c=%d\n", __func__, i, s->src_detile_buf_size_bytes_c[i]);
+ DML_LOG_VERBOSE("DML::%s: i=%u src_swath_bytes_c=%d\n", __func__, i, s->src_swath_bytes_c[i]);
+#endif
+ }
+
+ s->time_to_fill_det_us = (double) s->burst_bytes_to_fill_det / (256 * p->estimated_dcfclk_mhz); // fill time assume full burst at request rate
+ s->accumulated_return_path_dcfclk_cycles[i] = (unsigned int) math_ceil2(((DML_MAX_COMPRESSION_RATIO-1) * 64 * p->estimated_dcfclk_mhz) * s->time_to_fill_det_us / 64.0, 1.0); //for 64B per DCFClk
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: i=%u burst_bytes_to_fill_det=%d\n", __func__, i, s->burst_bytes_to_fill_det);
+ DML_LOG_VERBOSE("DML::%s: i=%u time_to_fill_det_us=%f\n", __func__, i, s->time_to_fill_det_us);
+ DML_LOG_VERBOSE("DML::%s: i=%u accumulated_return_path_dcfclk_cycles=%u\n", __func__, i, s->accumulated_return_path_dcfclk_cycles[i]);
+#endif
+ // clamping to worst case delay which is one which occupy the full rob+cdb
+ if (s->accumulated_return_path_dcfclk_cycles[i] > s->max_Trpd_dcfclk_cycles)
+ s->accumulated_return_path_dcfclk_cycles[i] = s->max_Trpd_dcfclk_cycles;
+ }
+
+ // Figure out the impacted prefetch time for each plane
+ // if impacted_Tre is > equ bw Tpre, we need to fail the prefetch schedule as we need a higher state to support the bw
+ for (i = 0; i < p->num_active_planes; i++) {
+ k = find_max_impact_plane(i, p->num_active_planes, s->accumulated_return_path_dcfclk_cycles); // plane k causes most impact to plane i
+ // the rest of planes (except for k) complete for bw
+ p->impacted_dst_y_pre[i] = s->accumulated_return_path_dcfclk_cycles[k]/p->estimated_dcfclk_mhz;
+ p->impacted_dst_y_pre[i] += calculate_impacted_Tsw(k, p->num_active_planes, p->prefetch_sw_bytes, p->estimated_urg_bandwidth_required_mbps);
+ p->impacted_dst_y_pre[i] = math_ceil2(p->impacted_dst_y_pre[i] / p->line_time[i], 0.25);
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: i=%u impacted_Tpre=%f (k=%u)\n", __func__, i, p->impacted_dst_y_pre[i], k);
+#endif
+ }
+
+ if (p->Tpre_rounded != NULL && p->Tpre_oto != NULL) {
+ for (i = 0; i < p->num_active_planes; i++) {
+ if (p->impacted_dst_y_pre[i] > p->dst_y_prefetch[i]) {
+ s->prefetch_global_check_passed = 0;
+ *p->recalc_prefetch_schedule = 1;
+ }
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: i=%u Tpre_rounded=%f\n", __func__, i, p->Tpre_rounded[i]);
+ DML_LOG_VERBOSE("DML::%s: i=%u Tpre_oto=%f\n", __func__, i, p->Tpre_oto[i]);
+#endif
+ }
+ } else {
+ // likely a mode programming calls, assume support, and no recalc - not used anyways
+ s->prefetch_global_check_passed = 1;
+ *p->recalc_prefetch_schedule = 0;
+ }
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: prefetch_global_check_passed=%u\n", __func__, s->prefetch_global_check_passed);
+ DML_LOG_VERBOSE("DML::%s: recalc_prefetch_schedule=%u\n", __func__, *p->recalc_prefetch_schedule);
+#endif
+
+ return s->prefetch_global_check_passed;
+}
+
static void calculate_peak_bandwidth_required(
struct dml2_core_internal_scratch *s,
struct dml2_core_calcs_calculate_peak_bandwidth_required_params *p)
@@ -5840,8 +6091,8 @@ static void calculate_peak_bandwidth_required(
memset(l, 0, sizeof(struct dml2_core_shared_calculate_peak_bandwidth_required_locals));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: inc_flip_bw = %d\n", __func__, p->inc_flip_bw);
- dml2_printf("DML::%s: NumberOfActiveSurfaces = %d\n", __func__, p->num_active_planes);
+ DML_LOG_VERBOSE("DML::%s: inc_flip_bw = %d\n", __func__, p->inc_flip_bw);
+ DML_LOG_VERBOSE("DML::%s: NumberOfActiveSurfaces = %d\n", __func__, p->num_active_planes);
#endif
for (unsigned int k = 0; k < p->num_active_planes; ++k) {
@@ -5870,6 +6121,7 @@ static void calculate_peak_bandwidth_required(
p->surface_read_bandwidth_c,
l->zero_array, //PrefetchBandwidthLuma,
l->zero_array, //PrefetchBandwidthChroma,
+ l->zero_array, //PrefetchBWMax
l->zero_array,
l->zero_array,
l->zero_array,
@@ -5906,6 +6158,7 @@ static void calculate_peak_bandwidth_required(
p->surface_read_bandwidth_c,
l->zero_array, //PrefetchBandwidthLuma,
l->zero_array, //PrefetchBandwidthChroma,
+ l->zero_array, //PrefetchBWMax
p->excess_vactive_fill_bw_l,
p->excess_vactive_fill_bw_c,
p->cursor_bw,
@@ -5942,6 +6195,7 @@ static void calculate_peak_bandwidth_required(
p->surface_read_bandwidth_c,
p->prefetch_bandwidth_l,
p->prefetch_bandwidth_c,
+ p->prefetch_bandwidth_max, // to prevent ms/mp mismatches where mp prefetch bw > ms prefetch bw
p->excess_vactive_fill_bw_l,
p->excess_vactive_fill_bw_c,
p->cursor_bw,
@@ -5978,6 +6232,7 @@ static void calculate_peak_bandwidth_required(
p->surface_read_bandwidth_c,
p->prefetch_bandwidth_l,
p->prefetch_bandwidth_c,
+ p->prefetch_bandwidth_max, // to prevent ms/mp mismatch where mp prefetch bw > ms prefetch bw
p->excess_vactive_fill_bw_l,
p->excess_vactive_fill_bw_c,
p->cursor_bw,
@@ -6014,6 +6269,7 @@ static void calculate_peak_bandwidth_required(
p->surface_read_bandwidth_c,
p->prefetch_bandwidth_l,
p->prefetch_bandwidth_c,
+ p->prefetch_bandwidth_max, // to prevent ms/mp mismatches where mp prefetch bw > ms prefetch bw
p->excess_vactive_fill_bw_l,
p->excess_vactive_fill_bw_c,
p->cursor_bw,
@@ -6032,12 +6288,12 @@ static void calculate_peak_bandwidth_required(
p->surface_peak_required_bw[m][n]);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: urg_vactive_bandwidth_required%s[%s][%s]=%f\n", __func__, (p->inc_flip_bw ? "_flip" : ""), dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), p->urg_vactive_bandwidth_required[m][n]);
- dml2_printf("DML::%s: urg_bandwidth_required%s[%s][%s]=%f\n", __func__, (p->inc_flip_bw ? "_flip" : ""), dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), p->urg_bandwidth_required[m][n]);
- dml2_printf("DML::%s: urg_bandwidth_required_qual[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), p->urg_bandwidth_required[m][n]);
- dml2_printf("DML::%s: non_urg_bandwidth_required%s[%s][%s]=%f\n", __func__, (p->inc_flip_bw ? "_flip" : ""), dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), p->non_urg_bandwidth_required[m][n]);
+ DML_LOG_VERBOSE("DML::%s: urg_vactive_bandwidth_required%s[%s][%s]=%f\n", __func__, (p->inc_flip_bw ? "_flip" : ""), dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), p->urg_vactive_bandwidth_required[m][n]);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_required%s[%s][%s]=%f\n", __func__, (p->inc_flip_bw ? "_flip" : ""), dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), p->urg_bandwidth_required[m][n]);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_required_qual[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), p->urg_bandwidth_required[m][n]);
+ DML_LOG_VERBOSE("DML::%s: non_urg_bandwidth_required%s[%s][%s]=%f\n", __func__, (p->inc_flip_bw ? "_flip" : ""), dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), p->non_urg_bandwidth_required[m][n]);
#endif
- dml2_assert(p->urg_bandwidth_required[m][n] >= p->non_urg_bandwidth_required[m][n]);
+ DML_ASSERT(p->urg_bandwidth_required[m][n] >= p->non_urg_bandwidth_required[m][n]);
}
}
}
@@ -6046,7 +6302,7 @@ static void check_urgent_bandwidth_support(
double *frac_urg_bandwidth_nom,
double *frac_urg_bandwidth_mall,
bool *vactive_bandwidth_support_ok, // vactive ok
- bool *bandwidth_support_ok, // max of vm, prefetch, vactive all ok
+ bool *bandwidth_support_ok,// max of vm, prefetch, vactive all ok
unsigned int mall_allocated_for_dcn_mbytes,
double non_urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
@@ -6099,24 +6355,23 @@ static void check_urgent_bandwidth_support(
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: frac_urg_bandwidth_nom_sdp = %f\n", __func__, frac_urg_bandwidth_nom_sdp);
- dml2_printf("DML::%s: frac_urg_bandwidth_nom_dram = %f\n", __func__, frac_urg_bandwidth_nom_dram);
- dml2_printf("DML::%s: frac_urg_bandwidth_nom = %f\n", __func__, *frac_urg_bandwidth_nom);
+ DML_LOG_VERBOSE("DML::%s: frac_urg_bandwidth_nom_sdp = %f\n", __func__, frac_urg_bandwidth_nom_sdp);
+ DML_LOG_VERBOSE("DML::%s: frac_urg_bandwidth_nom_dram = %f\n", __func__, frac_urg_bandwidth_nom_dram);
+ DML_LOG_VERBOSE("DML::%s: frac_urg_bandwidth_nom = %f\n", __func__, *frac_urg_bandwidth_nom);
- dml2_printf("DML::%s: frac_urg_bandwidth_mall_sdp = %f\n", __func__, frac_urg_bandwidth_mall_sdp);
- dml2_printf("DML::%s: frac_urg_bandwidth_mall_dram = %f\n", __func__, frac_urg_bandwidth_mall_dram);
- dml2_printf("DML::%s: frac_urg_bandwidth_mall = %f\n", __func__, *frac_urg_bandwidth_mall);
- dml2_printf("DML::%s: bandwidth_support_ok = %d\n", __func__, *bandwidth_support_ok);
+ DML_LOG_VERBOSE("DML::%s: frac_urg_bandwidth_mall_sdp = %f\n", __func__, frac_urg_bandwidth_mall_sdp);
+ DML_LOG_VERBOSE("DML::%s: frac_urg_bandwidth_mall_dram = %f\n", __func__, frac_urg_bandwidth_mall_dram);
+ DML_LOG_VERBOSE("DML::%s: frac_urg_bandwidth_mall = %f\n", __func__, *frac_urg_bandwidth_mall);
+ DML_LOG_VERBOSE("DML::%s: bandwidth_support_ok = %d\n", __func__, *bandwidth_support_ok);
for (unsigned int m = 0; m < dml2_core_internal_soc_state_max; m++) {
for (unsigned int n = 0; n < dml2_core_internal_bw_max; n++) {
- dml2_printf("DML::%s: state:%s bw_type:%s urg_bandwidth_available=%f %s urg_bandwidth_required=%f\n",
+ DML_LOG_VERBOSE("DML::%s: state:%s bw_type:%s urg_bandwidth_available=%f %s urg_bandwidth_required=%f\n",
__func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n),
urg_bandwidth_available[m][n], (urg_bandwidth_available[m][n] < urg_bandwidth_required[m][n]) ? "<" : ">=", urg_bandwidth_required[m][n]);
}
}
#endif
-
}
static double get_bandwidth_available_for_immediate_flip(enum dml2_core_internal_soc_state_type eval_state,
@@ -6132,14 +6387,14 @@ static double get_bandwidth_available_for_immediate_flip(enum dml2_core_internal
flip_bw_available_mbps = flip_bw_available_sdp_mbps < flip_bw_available_dram_mbps ? flip_bw_available_sdp_mbps : flip_bw_available_dram_mbps;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: eval_state = %s\n", __func__, dml2_core_internal_soc_state_type_str(eval_state));
- dml2_printf("DML::%s: urg_bandwidth_available_sdp_mbps = %f\n", __func__, urg_bandwidth_available[eval_state][dml2_core_internal_bw_sdp]);
- dml2_printf("DML::%s: urg_bandwidth_available_dram_mbps = %f\n", __func__, urg_bandwidth_available[eval_state][dml2_core_internal_bw_dram]);
- dml2_printf("DML::%s: urg_bandwidth_required_sdp_mbps = %f\n", __func__, urg_bandwidth_required[eval_state][dml2_core_internal_bw_sdp]);
- dml2_printf("DML::%s: urg_bandwidth_required_dram_mbps = %f\n", __func__, urg_bandwidth_required[eval_state][dml2_core_internal_bw_dram]);
- dml2_printf("DML::%s: flip_bw_available_sdp_mbps = %f\n", __func__, flip_bw_available_sdp_mbps);
- dml2_printf("DML::%s: flip_bw_available_dram_mbps = %f\n", __func__, flip_bw_available_dram_mbps);
- dml2_printf("DML::%s: flip_bw_available_mbps = %f\n", __func__, flip_bw_available_mbps);
+ DML_LOG_VERBOSE("DML::%s: eval_state = %s\n", __func__, dml2_core_internal_soc_state_type_str(eval_state));
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_available_sdp_mbps = %f\n", __func__, urg_bandwidth_available[eval_state][dml2_core_internal_bw_sdp]);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_available_dram_mbps = %f\n", __func__, urg_bandwidth_available[eval_state][dml2_core_internal_bw_dram]);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_required_sdp_mbps = %f\n", __func__, urg_bandwidth_required[eval_state][dml2_core_internal_bw_sdp]);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_required_dram_mbps = %f\n", __func__, urg_bandwidth_required[eval_state][dml2_core_internal_bw_dram]);
+ DML_LOG_VERBOSE("DML::%s: flip_bw_available_sdp_mbps = %f\n", __func__, flip_bw_available_sdp_mbps);
+ DML_LOG_VERBOSE("DML::%s: flip_bw_available_dram_mbps = %f\n", __func__, flip_bw_available_dram_mbps);
+ DML_LOG_VERBOSE("DML::%s: flip_bw_available_mbps = %f\n", __func__, flip_bw_available_mbps);
#endif
return flip_bw_available_mbps;
@@ -6164,28 +6419,28 @@ static void calculate_immediate_flip_bandwidth_support(
*flip_bandwidth_support_ok &= urg_bandwidth_available[eval_state][n] >= urg_bandwidth_required_flip[eval_state][n];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: n = %s\n", __func__, dml2_core_internal_bw_type_str(n));
- dml2_printf("DML::%s: urg_bandwidth_available = %f\n", __func__, urg_bandwidth_available[eval_state][n]);
- dml2_printf("DML::%s: non_urg_bandwidth_required_flip = %f\n", __func__, non_urg_bandwidth_required_flip[eval_state][n]);
- dml2_printf("DML::%s: urg_bandwidth_required_flip = %f\n", __func__, urg_bandwidth_required_flip[eval_state][n]);
- dml2_printf("DML::%s: flip_bandwidth_support_ok = %d\n", __func__, *flip_bandwidth_support_ok);
+ DML_LOG_VERBOSE("DML::%s: n = %s\n", __func__, dml2_core_internal_bw_type_str(n));
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_available = %f\n", __func__, urg_bandwidth_available[eval_state][n]);
+ DML_LOG_VERBOSE("DML::%s: non_urg_bandwidth_required_flip = %f\n", __func__, non_urg_bandwidth_required_flip[eval_state][n]);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_required_flip = %f\n", __func__, urg_bandwidth_required_flip[eval_state][n]);
+ DML_LOG_VERBOSE("DML::%s: flip_bandwidth_support_ok = %d\n", __func__, *flip_bandwidth_support_ok);
#endif
- dml2_assert(urg_bandwidth_required_flip[eval_state][n] >= non_urg_bandwidth_required_flip[eval_state][n]);
+ DML_ASSERT(urg_bandwidth_required_flip[eval_state][n] >= non_urg_bandwidth_required_flip[eval_state][n]);
}
*frac_urg_bandwidth_flip = (frac_urg_bw_flip_sdp > frac_urg_bw_flip_dram) ? frac_urg_bw_flip_sdp : frac_urg_bw_flip_dram;
*flip_bandwidth_support_ok &= (*frac_urg_bandwidth_flip <= 1.0);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: eval_state = %s\n", __func__, dml2_core_internal_soc_state_type_str(eval_state));
- dml2_printf("DML::%s: frac_urg_bw_flip_sdp = %f\n", __func__, frac_urg_bw_flip_sdp);
- dml2_printf("DML::%s: frac_urg_bw_flip_dram = %f\n", __func__, frac_urg_bw_flip_dram);
- dml2_printf("DML::%s: frac_urg_bandwidth_flip = %f\n", __func__, *frac_urg_bandwidth_flip);
- dml2_printf("DML::%s: flip_bandwidth_support_ok = %d\n", __func__, *flip_bandwidth_support_ok);
+ DML_LOG_VERBOSE("DML::%s: eval_state = %s\n", __func__, dml2_core_internal_soc_state_type_str(eval_state));
+ DML_LOG_VERBOSE("DML::%s: frac_urg_bw_flip_sdp = %f\n", __func__, frac_urg_bw_flip_sdp);
+ DML_LOG_VERBOSE("DML::%s: frac_urg_bw_flip_dram = %f\n", __func__, frac_urg_bw_flip_dram);
+ DML_LOG_VERBOSE("DML::%s: frac_urg_bandwidth_flip = %f\n", __func__, *frac_urg_bandwidth_flip);
+ DML_LOG_VERBOSE("DML::%s: flip_bandwidth_support_ok = %d\n", __func__, *flip_bandwidth_support_ok);
for (unsigned int m = 0; m < dml2_core_internal_soc_state_max; m++) {
for (unsigned int n = 0; n < dml2_core_internal_bw_max; n++) {
- dml2_printf("DML::%s: state:%s bw_type:%s, urg_bandwidth_available=%f %s urg_bandwidth_required=%f\n",
+ DML_LOG_VERBOSE("DML::%s: state:%s bw_type:%s, urg_bandwidth_available=%f %s urg_bandwidth_required=%f\n",
__func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n),
urg_bandwidth_available[m][n], (urg_bandwidth_available[m][n] < urg_bandwidth_required_flip[m][n]) ? "<" : ">=", urg_bandwidth_required_flip[m][n]);
}
@@ -6235,27 +6490,27 @@ static void CalculateFlipSchedule(
l->dpte_row_bytes = DPTEBytesPerRow;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: GPUVMEnable = %u\n", __func__, GPUVMEnable);
- dml2_printf("DML::%s: ip.max_flip_time_us = %d\n", __func__, max_flip_time_us);
- dml2_printf("DML::%s: ip.max_flip_time_lines = %d\n", __func__, max_flip_time_lines);
- dml2_printf("DML::%s: BandwidthAvailableForImmediateFlip = %f\n", __func__, BandwidthAvailableForImmediateFlip);
- dml2_printf("DML::%s: TotImmediateFlipBytes = %u\n", __func__, TotImmediateFlipBytes);
- dml2_printf("DML::%s: use_lb_flip_bw = %u\n", __func__, use_lb_flip_bw);
- dml2_printf("DML::%s: iflip_enable = %u\n", __func__, iflip_enable);
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, HostVMInefficiencyFactor);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, LineTime);
- dml2_printf("DML::%s: Tno_bw_flip = %f\n", __func__, Tno_bw_flip);
- dml2_printf("DML::%s: Tvm_trips_flip = %f\n", __func__, Tvm_trips_flip);
- dml2_printf("DML::%s: Tr0_trips_flip = %f\n", __func__, Tr0_trips_flip);
- dml2_printf("DML::%s: Tvm_trips_flip_rounded = %f\n", __func__, Tvm_trips_flip_rounded);
- dml2_printf("DML::%s: Tr0_trips_flip_rounded = %f\n", __func__, Tr0_trips_flip_rounded);
- dml2_printf("DML::%s: vm_bytes = %f\n", __func__, vm_bytes);
- dml2_printf("DML::%s: DPTEBytesPerRow = %f\n", __func__, DPTEBytesPerRow);
- dml2_printf("DML::%s: meta_row_bytes = %d\n", __func__, meta_row_bytes);
- dml2_printf("DML::%s: dpte_row_bytes = %f\n", __func__, l->dpte_row_bytes);
- dml2_printf("DML::%s: dpte_row_height = %d\n", __func__, dpte_row_height);
- dml2_printf("DML::%s: meta_row_height = %d\n", __func__, meta_row_height);
- dml2_printf("DML::%s: VRatio = %f\n", __func__, VRatio);
+ DML_LOG_VERBOSE("DML::%s: GPUVMEnable = %u\n", __func__, GPUVMEnable);
+ DML_LOG_VERBOSE("DML::%s: ip.max_flip_time_us = %d\n", __func__, max_flip_time_us);
+ DML_LOG_VERBOSE("DML::%s: ip.max_flip_time_lines = %d\n", __func__, max_flip_time_lines);
+ DML_LOG_VERBOSE("DML::%s: BandwidthAvailableForImmediateFlip = %f\n", __func__, BandwidthAvailableForImmediateFlip);
+ DML_LOG_VERBOSE("DML::%s: TotImmediateFlipBytes = %u\n", __func__, TotImmediateFlipBytes);
+ DML_LOG_VERBOSE("DML::%s: use_lb_flip_bw = %u\n", __func__, use_lb_flip_bw);
+ DML_LOG_VERBOSE("DML::%s: iflip_enable = %u\n", __func__, iflip_enable);
+ DML_LOG_VERBOSE("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, HostVMInefficiencyFactor);
+ DML_LOG_VERBOSE("DML::%s: LineTime = %f\n", __func__, LineTime);
+ DML_LOG_VERBOSE("DML::%s: Tno_bw_flip = %f\n", __func__, Tno_bw_flip);
+ DML_LOG_VERBOSE("DML::%s: Tvm_trips_flip = %f\n", __func__, Tvm_trips_flip);
+ DML_LOG_VERBOSE("DML::%s: Tr0_trips_flip = %f\n", __func__, Tr0_trips_flip);
+ DML_LOG_VERBOSE("DML::%s: Tvm_trips_flip_rounded = %f\n", __func__, Tvm_trips_flip_rounded);
+ DML_LOG_VERBOSE("DML::%s: Tr0_trips_flip_rounded = %f\n", __func__, Tr0_trips_flip_rounded);
+ DML_LOG_VERBOSE("DML::%s: vm_bytes = %f\n", __func__, vm_bytes);
+ DML_LOG_VERBOSE("DML::%s: DPTEBytesPerRow = %f\n", __func__, DPTEBytesPerRow);
+ DML_LOG_VERBOSE("DML::%s: meta_row_bytes = %d\n", __func__, meta_row_bytes);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_bytes = %f\n", __func__, l->dpte_row_bytes);
+ DML_LOG_VERBOSE("DML::%s: dpte_row_height = %d\n", __func__, dpte_row_height);
+ DML_LOG_VERBOSE("DML::%s: meta_row_height = %d\n", __func__, meta_row_height);
+ DML_LOG_VERBOSE("DML::%s: VRatio = %f\n", __func__, VRatio);
#endif
if (TotImmediateFlipBytes > 0 && (GPUVMEnable || dcc_mrq_enable)) {
@@ -6282,9 +6537,9 @@ static void CalculateFlipSchedule(
l->min_row_time = l->min_row_height * LineTime / VRatio;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: min_row_time = %f\n", __func__, l->min_row_time);
+ DML_LOG_VERBOSE("DML::%s: min_row_time = %f\n", __func__, l->min_row_time);
#endif
- dml2_assert(l->min_row_time > 0);
+ DML_ASSERT(l->min_row_time > 0);
if (use_lb_flip_bw) {
// For mode check, calculation the flip bw requirement with worst case flip time
@@ -6305,20 +6560,20 @@ static void CalculateFlipSchedule(
l->hvm_scaled_vm_bytes / (l->max_flip_time - Tno_bw_flip - 2 * Tr0_trips_flip_rounded),
l->hvm_scaled_row_bytes / (l->max_flip_time - Tvm_trips_flip_rounded));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: max_flip_time = %f\n", __func__, l->max_flip_time);
- dml2_printf("DML::%s: total vm bytes (hvm ineff scaled) = %f\n", __func__, l->hvm_scaled_vm_bytes);
- dml2_printf("DML::%s: total row bytes (%d row, hvm ineff scaled) = %f\n", __func__, l->num_rows, l->hvm_scaled_row_bytes);
- dml2_printf("DML::%s: total vm+row bytes (hvm ineff scaled) = %f\n", __func__, l->hvm_scaled_vm_row_bytes);
- dml2_printf("DML::%s: lb_flip_bw for vm and row = %f\n", __func__, l->hvm_scaled_vm_row_bytes / (l->max_flip_time - Tno_bw_flip));
- dml2_printf("DML::%s: lb_flip_bw for vm = %f\n", __func__, l->hvm_scaled_vm_bytes / (l->max_flip_time - Tno_bw_flip - 2 * Tr0_trips_flip_rounded));
- dml2_printf("DML::%s: lb_flip_bw for row = %f\n", __func__, l->hvm_scaled_row_bytes / (l->max_flip_time - Tvm_trips_flip_rounded));
+ DML_LOG_VERBOSE("DML::%s: max_flip_time = %f\n", __func__, l->max_flip_time);
+ DML_LOG_VERBOSE("DML::%s: total vm bytes (hvm ineff scaled) = %f\n", __func__, l->hvm_scaled_vm_bytes);
+ DML_LOG_VERBOSE("DML::%s: total row bytes (%f row, hvm ineff scaled) = %f\n", __func__, l->num_rows, l->hvm_scaled_row_bytes);
+ DML_LOG_VERBOSE("DML::%s: total vm+row bytes (hvm ineff scaled) = %f\n", __func__, l->hvm_scaled_vm_row_bytes);
+ DML_LOG_VERBOSE("DML::%s: lb_flip_bw for vm and row = %f\n", __func__, l->hvm_scaled_vm_row_bytes / (l->max_flip_time - Tno_bw_flip));
+ DML_LOG_VERBOSE("DML::%s: lb_flip_bw for vm = %f\n", __func__, l->hvm_scaled_vm_bytes / (l->max_flip_time - Tno_bw_flip - 2 * Tr0_trips_flip_rounded));
+ DML_LOG_VERBOSE("DML::%s: lb_flip_bw for row = %f\n", __func__, l->hvm_scaled_row_bytes / (l->max_flip_time - Tvm_trips_flip_rounded));
if (l->lb_flip_bw > 0) {
- dml2_printf("DML::%s: mode_support est Tvm_flip = %f (bw-based)\n", __func__, Tno_bw_flip + l->hvm_scaled_vm_bytes / l->lb_flip_bw);
- dml2_printf("DML::%s: mode_support est Tr0_flip = %f (bw-based)\n", __func__, l->hvm_scaled_row_bytes / l->lb_flip_bw / l->num_rows);
- dml2_printf("DML::%s: mode_support est dst_y_per_vm_flip = %f (bw-based)\n", __func__, Tno_bw_flip + l->hvm_scaled_vm_bytes / l->lb_flip_bw / LineTime);
- dml2_printf("DML::%s: mode_support est dst_y_per_row_flip = %f (bw-based)\n", __func__, l->hvm_scaled_row_bytes / l->lb_flip_bw / LineTime / l->num_rows);
- dml2_printf("DML::%s: Tvm_trips_flip_rounded + 2*Tr0_trips_flip_rounded = %f\n", __func__, (Tvm_trips_flip_rounded + 2 * Tr0_trips_flip_rounded));
+ DML_LOG_VERBOSE("DML::%s: mode_support est Tvm_flip = %f (bw-based)\n", __func__, Tno_bw_flip + l->hvm_scaled_vm_bytes / l->lb_flip_bw);
+ DML_LOG_VERBOSE("DML::%s: mode_support est Tr0_flip = %f (bw-based)\n", __func__, l->hvm_scaled_row_bytes / l->lb_flip_bw / l->num_rows);
+ DML_LOG_VERBOSE("DML::%s: mode_support est dst_y_per_vm_flip = %f (bw-based)\n", __func__, Tno_bw_flip + l->hvm_scaled_vm_bytes / l->lb_flip_bw / LineTime);
+ DML_LOG_VERBOSE("DML::%s: mode_support est dst_y_per_row_flip = %f (bw-based)\n", __func__, l->hvm_scaled_row_bytes / l->lb_flip_bw / LineTime / l->num_rows);
+ DML_LOG_VERBOSE("DML::%s: Tvm_trips_flip_rounded + 2*Tr0_trips_flip_rounded = %f\n", __func__, (Tvm_trips_flip_rounded + 2 * Tr0_trips_flip_rounded));
}
#endif
l->lb_flip_bw = math_max3(l->lb_flip_bw,
@@ -6326,8 +6581,8 @@ static void CalculateFlipSchedule(
(l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / (15 * LineTime));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: lb_flip_bw for vm reg limit = %f\n", __func__, l->hvm_scaled_vm_bytes / (31 * LineTime) - Tno_bw_flip);
- dml2_printf("DML::%s: lb_flip_bw for row reg limit = %f\n", __func__, (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / (15 * LineTime));
+ DML_LOG_VERBOSE("DML::%s: lb_flip_bw for vm reg limit = %f\n", __func__, l->hvm_scaled_vm_bytes / (31 * LineTime) - Tno_bw_flip);
+ DML_LOG_VERBOSE("DML::%s: lb_flip_bw for row reg limit = %f\n", __func__, (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / (15 * LineTime));
#endif
}
@@ -6339,13 +6594,12 @@ static void CalculateFlipSchedule(
} else {
if (iflip_enable) {
l->ImmediateFlipBW = (double)per_pipe_flip_bytes * BandwidthAvailableForImmediateFlip / (double)TotImmediateFlipBytes; // flip_bw(i)
- double portion = (double)per_pipe_flip_bytes / (double)TotImmediateFlipBytes;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: per_pipe_flip_bytes = %d\n", __func__, per_pipe_flip_bytes);
- dml2_printf("DML::%s: BandwidthAvailableForImmediateFlip = %f\n", __func__, BandwidthAvailableForImmediateFlip);
- dml2_printf("DML::%s: ImmediateFlipBW = %f\n", __func__, l->ImmediateFlipBW);
- dml2_printf("DML::%s: portion of flip bw = %f\n", __func__, portion);
+ DML_LOG_VERBOSE("DML::%s: per_pipe_flip_bytes = %d\n", __func__, per_pipe_flip_bytes);
+ DML_LOG_VERBOSE("DML::%s: BandwidthAvailableForImmediateFlip = %f\n", __func__, BandwidthAvailableForImmediateFlip);
+ DML_LOG_VERBOSE("DML::%s: ImmediateFlipBW = %f\n", __func__, l->ImmediateFlipBW);
+ DML_LOG_VERBOSE("DML::%s: portion of flip bw = %f\n", __func__, (double)per_pipe_flip_bytes / (double)TotImmediateFlipBytes);
#endif
if (l->ImmediateFlipBW == 0) {
l->Tvm_flip = 0;
@@ -6360,11 +6614,11 @@ static void CalculateFlipSchedule(
LineTime / 4.0);
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: total vm bytes (hvm ineff scaled) = %f\n", __func__, vm_bytes * HostVMInefficiencyFactor);
- dml2_printf("DML::%s: total row bytes (hvm ineff scaled, one row) = %f\n", __func__, (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes));
+ DML_LOG_VERBOSE("DML::%s: total vm bytes (hvm ineff scaled) = %f\n", __func__, vm_bytes * HostVMInefficiencyFactor);
+ DML_LOG_VERBOSE("DML::%s: total row bytes (hvm ineff scaled, one row) = %f\n", __func__, (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes));
- dml2_printf("DML::%s: Tvm_flip = %f (bw-based), Tvm_trips_flip = %f (latency-based)\n", __func__, Tno_bw_flip + vm_bytes * HostVMInefficiencyFactor / l->ImmediateFlipBW, Tvm_trips_flip);
- dml2_printf("DML::%s: Tr0_flip = %f (bw-based), Tr0_trips_flip = %f (latency-based)\n", __func__, (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / l->ImmediateFlipBW, Tr0_trips_flip);
+ DML_LOG_VERBOSE("DML::%s: Tvm_flip = %f (bw-based), Tvm_trips_flip = %f (latency-based)\n", __func__, Tno_bw_flip + vm_bytes * HostVMInefficiencyFactor / l->ImmediateFlipBW, Tvm_trips_flip);
+ DML_LOG_VERBOSE("DML::%s: Tr0_flip = %f (bw-based), Tr0_trips_flip = %f (latency-based)\n", __func__, (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / l->ImmediateFlipBW, Tr0_trips_flip);
#endif
*dst_y_per_vm_flip = math_ceil2(4.0 * (l->Tvm_flip / LineTime), 1.0) / 4.0;
*dst_y_per_row_flip = math_ceil2(4.0 * (l->Tr0_flip / LineTime), 1.0) / 4.0;
@@ -6397,14 +6651,14 @@ static void CalculateFlipSchedule(
#ifdef __DML_VBA_DEBUG__
if (!use_lb_flip_bw) {
- dml2_printf("DML::%s: dst_y_per_vm_flip = %f (should be < 32)\n", __func__, *dst_y_per_vm_flip);
- dml2_printf("DML::%s: dst_y_per_row_flip = %f (should be < 16)\n", __func__, *dst_y_per_row_flip);
- dml2_printf("DML::%s: Tvm_flip = %f (final)\n", __func__, l->Tvm_flip);
- dml2_printf("DML::%s: Tr0_flip = %f (final)\n", __func__, l->Tr0_flip);
- dml2_printf("DML::%s: Tvm_flip + 2*Tr0_flip = %f (should be <= min_row_time=%f)\n", __func__, l->Tvm_flip + 2 * l->Tr0_flip, l->min_row_time);
+ DML_LOG_VERBOSE("DML::%s: dst_y_per_vm_flip = %f (should be < 32)\n", __func__, *dst_y_per_vm_flip);
+ DML_LOG_VERBOSE("DML::%s: dst_y_per_row_flip = %f (should be < 16)\n", __func__, *dst_y_per_row_flip);
+ DML_LOG_VERBOSE("DML::%s: Tvm_flip = %f (final)\n", __func__, l->Tvm_flip);
+ DML_LOG_VERBOSE("DML::%s: Tr0_flip = %f (final)\n", __func__, l->Tr0_flip);
+ DML_LOG_VERBOSE("DML::%s: Tvm_flip + 2*Tr0_flip = %f (should be <= min_row_time=%f)\n", __func__, l->Tvm_flip + 2 * l->Tr0_flip, l->min_row_time);
}
- dml2_printf("DML::%s: final_flip_bw = %f\n", __func__, *final_flip_bw);
- dml2_printf("DML::%s: ImmediateFlipSupportedForPipe = %u\n", __func__, *ImmediateFlipSupportedForPipe);
+ DML_LOG_VERBOSE("DML::%s: final_flip_bw = %f\n", __func__, *final_flip_bw);
+ DML_LOG_VERBOSE("DML::%s: ImmediateFlipSupportedForPipe = %u\n", __func__, *ImmediateFlipSupportedForPipe);
#endif
}
@@ -6422,7 +6676,7 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
p->Watermark->UrgentWatermark = p->mmSOCParameters.UrgentLatency + p->mmSOCParameters.ExtraLatency;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: UrgentWatermark = %f\n", __func__, p->Watermark->UrgentWatermark);
+ DML_LOG_VERBOSE("DML::%s: UrgentWatermark = %f\n", __func__, p->Watermark->UrgentWatermark);
#endif
p->Watermark->USRRetrainingWatermark = p->mmSOCParameters.UrgentLatency + p->mmSOCParameters.ExtraLatency + p->mmSOCParameters.USRRetrainingLatency + p->mmSOCParameters.SMNLatency;
@@ -6438,28 +6692,28 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
p->Watermark->Z8StutterExitWatermark += p->mmSOCParameters.max_urgent_latency_us + p->mmSOCParameters.df_response_time_us;
p->Watermark->Z8StutterEnterPlusExitWatermark += p->mmSOCParameters.max_urgent_latency_us + p->mmSOCParameters.df_response_time_us;
}
- p->Watermark->g6_temp_read_watermark_us = p->mmSOCParameters.g6_temp_read_blackout_us + p->Watermark->UrgentWatermark;
+ p->Watermark->temp_read_or_ppt_watermark_us = p->mmSOCParameters.g6_temp_read_blackout_us + p->Watermark->UrgentWatermark;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: UrgentLatency = %f\n", __func__, p->mmSOCParameters.UrgentLatency);
- dml2_printf("DML::%s: ExtraLatency = %f\n", __func__, p->mmSOCParameters.ExtraLatency);
- dml2_printf("DML::%s: DRAMClockChangeLatency = %f\n", __func__, p->mmSOCParameters.DRAMClockChangeLatency);
- dml2_printf("DML::%s: SREnterPlusExitZ8Time = %f\n", __func__, p->mmSOCParameters.SREnterPlusExitZ8Time);
- dml2_printf("DML::%s: SREnterPlusExitTime = %f\n", __func__, p->mmSOCParameters.SREnterPlusExitTime);
- dml2_printf("DML::%s: UrgentWatermark = %f\n", __func__, p->Watermark->UrgentWatermark);
- dml2_printf("DML::%s: USRRetrainingWatermark = %f\n", __func__, p->Watermark->USRRetrainingWatermark);
- dml2_printf("DML::%s: DRAMClockChangeWatermark = %f\n", __func__, p->Watermark->DRAMClockChangeWatermark);
- dml2_printf("DML::%s: FCLKChangeWatermark = %f\n", __func__, p->Watermark->FCLKChangeWatermark);
- dml2_printf("DML::%s: StutterExitWatermark = %f\n", __func__, p->Watermark->StutterExitWatermark);
- dml2_printf("DML::%s: StutterEnterPlusExitWatermark = %f\n", __func__, p->Watermark->StutterEnterPlusExitWatermark);
- dml2_printf("DML::%s: Z8StutterExitWatermark = %f\n", __func__, p->Watermark->Z8StutterExitWatermark);
- dml2_printf("DML::%s: Z8StutterEnterPlusExitWatermark = %f\n", __func__, p->Watermark->Z8StutterEnterPlusExitWatermark);
- dml2_printf("DML::%s: g6_temp_read_watermark_us = %f\n", __func__, p->Watermark->g6_temp_read_watermark_us);
+ DML_LOG_VERBOSE("DML::%s: UrgentLatency = %f\n", __func__, p->mmSOCParameters.UrgentLatency);
+ DML_LOG_VERBOSE("DML::%s: ExtraLatency = %f\n", __func__, p->mmSOCParameters.ExtraLatency);
+ DML_LOG_VERBOSE("DML::%s: DRAMClockChangeLatency = %f\n", __func__, p->mmSOCParameters.DRAMClockChangeLatency);
+ DML_LOG_VERBOSE("DML::%s: SREnterPlusExitZ8Time = %f\n", __func__, p->mmSOCParameters.SREnterPlusExitZ8Time);
+ DML_LOG_VERBOSE("DML::%s: SREnterPlusExitTime = %f\n", __func__, p->mmSOCParameters.SREnterPlusExitTime);
+ DML_LOG_VERBOSE("DML::%s: UrgentWatermark = %f\n", __func__, p->Watermark->UrgentWatermark);
+ DML_LOG_VERBOSE("DML::%s: USRRetrainingWatermark = %f\n", __func__, p->Watermark->USRRetrainingWatermark);
+ DML_LOG_VERBOSE("DML::%s: DRAMClockChangeWatermark = %f\n", __func__, p->Watermark->DRAMClockChangeWatermark);
+ DML_LOG_VERBOSE("DML::%s: FCLKChangeWatermark = %f\n", __func__, p->Watermark->FCLKChangeWatermark);
+ DML_LOG_VERBOSE("DML::%s: StutterExitWatermark = %f\n", __func__, p->Watermark->StutterExitWatermark);
+ DML_LOG_VERBOSE("DML::%s: StutterEnterPlusExitWatermark = %f\n", __func__, p->Watermark->StutterEnterPlusExitWatermark);
+ DML_LOG_VERBOSE("DML::%s: Z8StutterExitWatermark = %f\n", __func__, p->Watermark->Z8StutterExitWatermark);
+ DML_LOG_VERBOSE("DML::%s: Z8StutterEnterPlusExitWatermark = %f\n", __func__, p->Watermark->Z8StutterEnterPlusExitWatermark);
+ DML_LOG_VERBOSE("DML::%s: temp_read_or_ppt_watermark_us = %f\n", __func__, p->Watermark->temp_read_or_ppt_watermark_us);
#endif
s->TotalActiveWriteback = 0;
for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
+ if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0) {
s->TotalActiveWriteback = s->TotalActiveWriteback + 1;
}
}
@@ -6487,11 +6741,11 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
p->Watermark->WritebackFCLKChangeWatermark = p->Watermark->WritebackFCLKChangeWatermark + p->mmSOCParameters.USRRetrainingLatency;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: WritebackDRAMClockChangeWatermark = %f\n", __func__, p->Watermark->WritebackDRAMClockChangeWatermark);
- dml2_printf("DML::%s: WritebackFCLKChangeWatermark = %f\n", __func__, p->Watermark->WritebackFCLKChangeWatermark);
- dml2_printf("DML::%s: WritebackUrgentWatermark = %f\n", __func__, p->Watermark->WritebackUrgentWatermark);
- dml2_printf("DML::%s: USRRetrainingRequired = %u\n", __func__, p->USRRetrainingRequired);
- dml2_printf("DML::%s: USRRetrainingLatency = %f\n", __func__, p->mmSOCParameters.USRRetrainingLatency);
+ DML_LOG_VERBOSE("DML::%s: WritebackDRAMClockChangeWatermark = %f\n", __func__, p->Watermark->WritebackDRAMClockChangeWatermark);
+ DML_LOG_VERBOSE("DML::%s: WritebackFCLKChangeWatermark = %f\n", __func__, p->Watermark->WritebackFCLKChangeWatermark);
+ DML_LOG_VERBOSE("DML::%s: WritebackUrgentWatermark = %f\n", __func__, p->Watermark->WritebackUrgentWatermark);
+ DML_LOG_VERBOSE("DML::%s: USRRetrainingRequired = %u\n", __func__, p->USRRetrainingRequired);
+ DML_LOG_VERBOSE("DML::%s: USRRetrainingLatency = %f\n", __func__, p->mmSOCParameters.USRRetrainingLatency);
#endif
s->TotalPixelBW = 0.0;
@@ -6522,11 +6776,11 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
s->LBLatencyHidingSourceLinesC[k] = (unsigned int)(math_min2((double)p->MaxLineBufferLines, math_floor2((double)p->LineBufferSize / LBBitPerPixel / ((double)p->SwathWidthC[k] / math_max2(h_ratio_c, 1.0)), 1)) - (v_taps_c - 1));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, MaxLineBufferLines= %u\n", __func__, k, p->MaxLineBufferLines);
- dml2_printf("DML::%s: k=%u, LineBufferSize = %u\n", __func__, k, p->LineBufferSize);
- dml2_printf("DML::%s: k=%u, LBBitPerPixel = %u\n", __func__, k, LBBitPerPixel);
- dml2_printf("DML::%s: k=%u, HRatio = %f\n", __func__, k, h_ratio);
- dml2_printf("DML::%s: k=%u, VTaps = %f\n", __func__, k, v_taps);
+ DML_LOG_VERBOSE("DML::%s: k=%u, MaxLineBufferLines = %u\n", __func__, k, p->MaxLineBufferLines);
+ DML_LOG_VERBOSE("DML::%s: k=%u, LineBufferSize = %u\n", __func__, k, p->LineBufferSize);
+ DML_LOG_VERBOSE("DML::%s: k=%u, LBBitPerPixel = %f\n", __func__, k, LBBitPerPixel);
+ DML_LOG_VERBOSE("DML::%s: k=%u, HRatio = %f\n", __func__, k, h_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VTaps = %f\n", __func__, k, v_taps);
#endif
s->EffectiveLBLatencyHidingY = s->LBLatencyHidingSourceLinesY[k] / v_ratio * (h_total / pixel_clock_mhz);
@@ -6563,7 +6817,7 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
s->ActiveDRAMClockChangeLatencyMargin[k] = s->ActiveClockChangeLatencyHiding - p->Watermark->DRAMClockChangeWatermark;
s->ActiveFCLKChangeLatencyMargin[k] = s->ActiveClockChangeLatencyHiding - p->Watermark->FCLKChangeWatermark;
s->USRRetrainingLatencyMargin[k] = s->ActiveClockChangeLatencyHiding - p->Watermark->USRRetrainingWatermark;
- s->g6_temp_read_latency_margin[k] = s->ActiveClockChangeLatencyHiding - p->Watermark->g6_temp_read_watermark_us;
+ s->g6_temp_read_latency_margin[k] = s->ActiveClockChangeLatencyHiding - p->Watermark->temp_read_or_ppt_watermark_us;
if (p->VActiveLatencyHidingMargin)
p->VActiveLatencyHidingMargin[k] = s->ActiveDRAMClockChangeLatencyMargin[k];
@@ -6571,9 +6825,12 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
if (p->VActiveLatencyHidingUs)
p->VActiveLatencyHidingUs[k] = s->ActiveClockChangeLatencyHiding;
- if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.enable) {
- s->WritebackLatencyHiding = (double)p->WritebackInterfaceBufferSize * 1024.0 / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height * (double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height * (double)h_total / pixel_clock_mhz) * 4.0);
- if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format == dml2_444_64) {
+ if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0) {
+ s->WritebackLatencyHiding = (double)p->WritebackInterfaceBufferSize * 1024.0
+ / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_height
+ * (double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_width
+ / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].input_height * (double)h_total / pixel_clock_mhz) * 4.0);
+ if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].pixel_format == dml2_444_64) {
s->WritebackLatencyHiding = s->WritebackLatencyHiding / 2;
}
s->WritebackDRAMClockChangeLatencyMargin = s->WritebackLatencyHiding - p->Watermark->WritebackDRAMClockChangeWatermark;
@@ -6588,36 +6845,36 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
uclk_pstate_change_strategy = p->display_cfg->plane_descriptors[k].overrides.uclk_pstate_change_strategy;
reserved_vblank_time_us = (double)p->display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns / 1000;
- p->FCLKChangeSupport[k] = dml2_fclock_change_unsupported;
+ p->FCLKChangeSupport[k] = dml2_pstate_change_unsupported;
if (s->ActiveFCLKChangeLatencyMargin[k] > 0)
- p->FCLKChangeSupport[k] = dml2_fclock_change_vactive;
+ p->FCLKChangeSupport[k] = dml2_pstate_change_vactive;
else if (reserved_vblank_time_us >= p->mmSOCParameters.FCLKChangeLatency)
- p->FCLKChangeSupport[k] = dml2_fclock_change_vblank;
+ p->FCLKChangeSupport[k] = dml2_pstate_change_vblank;
- if (p->FCLKChangeSupport[k] == dml2_fclock_change_unsupported)
+ if (p->FCLKChangeSupport[k] == dml2_pstate_change_unsupported)
*p->global_fclk_change_supported = false;
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_unsupported;
+ p->DRAMClockChangeSupport[k] = dml2_pstate_change_unsupported;
if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_auto) {
if (p->display_cfg->overrides.all_streams_blanked ||
(s->ActiveDRAMClockChangeLatencyMargin[k] > 0 && reserved_vblank_time_us >= p->mmSOCParameters.DRAMClockChangeLatency))
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vblank_and_vactive;
+ p->DRAMClockChangeSupport[k] = dml2_pstate_change_vblank_and_vactive;
else if (s->ActiveDRAMClockChangeLatencyMargin[k] > 0)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vactive;
+ p->DRAMClockChangeSupport[k] = dml2_pstate_change_vactive;
else if (reserved_vblank_time_us >= p->mmSOCParameters.DRAMClockChangeLatency)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vblank;
+ p->DRAMClockChangeSupport[k] = dml2_pstate_change_vblank;
} else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_vactive && s->ActiveDRAMClockChangeLatencyMargin[k] > 0)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vactive;
+ p->DRAMClockChangeSupport[k] = dml2_pstate_change_vactive;
else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_vblank && reserved_vblank_time_us >= p->mmSOCParameters.DRAMClockChangeLatency)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vblank;
+ p->DRAMClockChangeSupport[k] = dml2_pstate_change_vblank;
else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_drr)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_drr;
+ p->DRAMClockChangeSupport[k] = dml2_pstate_change_drr;
else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_mall_svp)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_mall_svp;
+ p->DRAMClockChangeSupport[k] = dml2_pstate_change_mall_svp;
else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_mall_full_frame)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_mall_full_frame;
+ p->DRAMClockChangeSupport[k] = dml2_pstate_change_mall_full_frame;
- if (p->DRAMClockChangeSupport[k] == dml2_dram_clock_change_unsupported)
+ if (p->DRAMClockChangeSupport[k] == dml2_pstate_change_unsupported)
*p->global_dram_clock_change_supported = false;
s->dst_y_pstate = (unsigned int)(math_ceil2((p->mmSOCParameters.DRAMClockChangeLatency + p->mmSOCParameters.UrgentLatency) / (h_total / pixel_clock_mhz), 1));
@@ -6626,16 +6883,16 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
s->sub_vp_lines_l = s->src_y_pstate_l + s->src_y_ahead_l + p->meta_row_height_l[k];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, DETBufferSizeY = %u\n", __func__, k, p->DETBufferSizeY[k]);
- dml2_printf("DML::%s: k=%u, BytePerPixelDETY = %f\n", __func__, k, p->BytePerPixelDETY[k]);
- dml2_printf("DML::%s: k=%u, SwathWidthY = %u\n", __func__, k, p->SwathWidthY[k]);
- dml2_printf("DML::%s: k=%u, SwathHeightY = %u\n", __func__, k, p->SwathHeightY[k]);
- dml2_printf("DML::%s: k=%u, LBLatencyHidingSourceLinesY = %u\n", __func__, k, s->LBLatencyHidingSourceLinesY[k]);
- dml2_printf("DML::%s: k=%u, dst_y_pstate = %u\n", __func__, k, s->dst_y_pstate);
- dml2_printf("DML::%s: k=%u, src_y_pstate_l = %u\n", __func__, k, s->src_y_pstate_l);
- dml2_printf("DML::%s: k=%u, src_y_ahead_l = %u\n", __func__, k, s->src_y_ahead_l);
- dml2_printf("DML::%s: k=%u, meta_row_height_l = %u\n", __func__, k, p->meta_row_height_l[k]);
- dml2_printf("DML::%s: k=%u, sub_vp_lines_l = %u\n", __func__, k, s->sub_vp_lines_l);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DETBufferSizeY = %u\n", __func__, k, p->DETBufferSizeY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, BytePerPixelDETY = %f\n", __func__, k, p->BytePerPixelDETY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, SwathWidthY = %u\n", __func__, k, p->SwathWidthY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, SwathHeightY = %u\n", __func__, k, p->SwathHeightY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, LBLatencyHidingSourceLinesY = %u\n", __func__, k, s->LBLatencyHidingSourceLinesY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dst_y_pstate = %u\n", __func__, k, s->dst_y_pstate);
+ DML_LOG_VERBOSE("DML::%s: k=%u, src_y_pstate_l = %u\n", __func__, k, s->src_y_pstate_l);
+ DML_LOG_VERBOSE("DML::%s: k=%u, src_y_ahead_l = %u\n", __func__, k, s->src_y_ahead_l);
+ DML_LOG_VERBOSE("DML::%s: k=%u, meta_row_height_l = %u\n", __func__, k, p->meta_row_height_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, sub_vp_lines_l = %u\n", __func__, k, s->sub_vp_lines_l);
#endif
p->SubViewportLinesNeededInMALL[k] = s->sub_vp_lines_l;
@@ -6650,10 +6907,10 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
p->SubViewportLinesNeededInMALL[k] = (unsigned int)(math_max2(s->sub_vp_lines_l, s->sub_vp_lines_c));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, meta_row_height_c = %u\n", __func__, k, p->meta_row_height_c[k]);
- dml2_printf("DML::%s: k=%u, src_y_pstate_c = %u\n", __func__, k, s->src_y_pstate_c);
- dml2_printf("DML::%s: k=%u, src_y_ahead_c = %u\n", __func__, k, s->src_y_ahead_c);
- dml2_printf("DML::%s: k=%u, sub_vp_lines_c = %u\n", __func__, k, s->sub_vp_lines_c);
+ DML_LOG_VERBOSE("DML::%s: k=%u, meta_row_height_c = %u\n", __func__, k, p->meta_row_height_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, src_y_pstate_c = %u\n", __func__, k, s->src_y_pstate_c);
+ DML_LOG_VERBOSE("DML::%s: k=%u, src_y_ahead_c = %u\n", __func__, k, s->src_y_ahead_c);
+ DML_LOG_VERBOSE("DML::%s: k=%u, sub_vp_lines_c = %u\n", __func__, k, s->sub_vp_lines_c);
#endif
}
}
@@ -6675,10 +6932,10 @@ static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DRAMClockChangeSupport = %u\n", __func__, *p->global_dram_clock_change_supported);
- dml2_printf("DML::%s: FCLKChangeSupport = %u\n", __func__, *p->global_fclk_change_supported);
- dml2_printf("DML::%s: MaxActiveFCLKChangeLatencySupported = %f\n", __func__, *p->MaxActiveFCLKChangeLatencySupported);
- dml2_printf("DML::%s: USRRetrainingSupport = %u\n", __func__, *p->USRRetrainingSupport);
+ DML_LOG_VERBOSE("DML::%s: DRAMClockChangeSupport = %u\n", __func__, *p->global_dram_clock_change_supported);
+ DML_LOG_VERBOSE("DML::%s: FCLKChangeSupport = %u\n", __func__, *p->global_fclk_change_supported);
+ DML_LOG_VERBOSE("DML::%s: MaxActiveFCLKChangeLatencySupported = %f\n", __func__, *p->MaxActiveFCLKChangeLatencySupported);
+ DML_LOG_VERBOSE("DML::%s: USRRetrainingSupport = %u\n", __func__, *p->USRRetrainingSupport);
#endif
}
@@ -6725,7 +6982,7 @@ static void calculate_bytes_to_fetch_required_to_hide_latency(
}
}
-static void calculate_vactive_det_fill_latency(
+static noinline_for_stack void calculate_vactive_det_fill_latency(
const struct dml2_display_cfg *display_cfg,
unsigned int num_active_planes,
unsigned int bytes_required_l[],
@@ -6824,7 +7081,7 @@ static unsigned int get_qos_param_index(unsigned long uclk_freq_khz, const struc
unsigned int index = 0;
for (i = 0; i < DML_MAX_CLK_TABLE_SIZE; i++) {
- dml2_printf("DML::%s: per_uclk_dpm_params[%d].minimum_uclk_khz = %d\n", __func__, i, per_uclk_dpm_params[i].minimum_uclk_khz);
+ DML_LOG_VERBOSE("DML::%s: per_uclk_dpm_params[%d].minimum_uclk_khz = %ld\n", __func__, i, per_uclk_dpm_params[i].minimum_uclk_khz);
if (i == 0)
index = 0;
@@ -6836,31 +7093,30 @@ static unsigned int get_qos_param_index(unsigned long uclk_freq_khz, const struc
break;
}
}
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: uclk_freq_khz = %d\n", __func__, uclk_freq_khz);
- dml2_printf("DML::%s: index = %d\n", __func__, index);
-#endif
+ DML_LOG_VERBOSE("DML::%s: uclk_freq_khz = %ld\n", __func__, uclk_freq_khz);
+ DML_LOG_VERBOSE("DML::%s: index = %d\n", __func__, index);
return index;
}
static unsigned int get_active_min_uclk_dpm_index(unsigned long uclk_freq_khz, const struct dml2_soc_state_table *clk_table)
{
unsigned int i;
- bool clk_entry_found = 0;
+ bool clk_entry_found = false;
for (i = 0; i < clk_table->uclk.num_clk_values; i++) {
- dml2_printf("DML::%s: clk_table.uclk.clk_values_khz[%d] = %d\n", __func__, i, clk_table->uclk.clk_values_khz[i]);
+ DML_LOG_VERBOSE("DML::%s: clk_table.uclk.clk_values_khz[%d] = %ld\n", __func__, i, clk_table->uclk.clk_values_khz[i]);
if (uclk_freq_khz == clk_table->uclk.clk_values_khz[i]) {
- clk_entry_found = 1;
+ clk_entry_found = true;
break;
}
}
- dml2_assert(clk_entry_found);
+ if (!clk_entry_found)
+ DML_ASSERT(clk_entry_found);
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: uclk_freq_khz = %ld\n", __func__, uclk_freq_khz);
- dml2_printf("DML::%s: index = %d\n", __func__, i);
+ DML_LOG_VERBOSE("DML::%s: uclk_freq_khz = %ld\n", __func__, uclk_freq_khz);
+ DML_LOG_VERBOSE("DML::%s: index = %d\n", __func__, i);
#endif
return i;
}
@@ -6900,10 +7156,10 @@ static void calculate_hostvm_inefficiency_factor(
if ((*HostVMInefficiencyFactorPrefetch < 4) && (remote_iommu_outstanding_translations < max_outstanding_reqs))
*HostVMInefficiencyFactorPrefetch = 4;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: urg_bandwidth_avail_active_pixel_and_vm = %f\n", __func__, urg_bandwidth_avail_active_pixel_and_vm);
- dml2_printf("DML::%s: urg_bandwidth_avail_active_vm_only = %f\n", __func__, urg_bandwidth_avail_active_vm_only);
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, *HostVMInefficiencyFactor);
- dml2_printf("DML::%s: HostVMInefficiencyFactorPrefetch = %f\n", __func__, *HostVMInefficiencyFactorPrefetch);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_avail_active_pixel_and_vm = %f\n", __func__, urg_bandwidth_avail_active_pixel_and_vm);
+ DML_LOG_VERBOSE("DML::%s: urg_bandwidth_avail_active_vm_only = %f\n", __func__, urg_bandwidth_avail_active_vm_only);
+ DML_LOG_VERBOSE("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, *HostVMInefficiencyFactor);
+ DML_LOG_VERBOSE("DML::%s: HostVMInefficiencyFactorPrefetch = %f\n", __func__, *HostVMInefficiencyFactorPrefetch);
#endif
}
}
@@ -6915,8 +7171,7 @@ struct dml2_core_internal_g6_temp_read_blackouts_table {
} entries[DML_MAX_CLK_TABLE_SIZE];
};
-static const struct dml2_core_internal_g6_temp_read_blackouts_table
- core_dcn4_g6_temp_read_blackout_table = {
+struct dml2_core_internal_g6_temp_read_blackouts_table core_dcn4_g6_temp_read_blackout_table = {
.entries = {
{
.uclk_khz = 96000,
@@ -7018,27 +7273,660 @@ static void calculate_pstate_keepout_dst_lines(
}
}
+static noinline_for_stack void dml_core_ms_prefetch_check(struct dml2_core_internal_display_mode_lib *mode_lib,
+ const struct dml2_display_cfg *display_cfg)
+{
+ struct dml2_core_calcs_mode_support_locals *s = &mode_lib->scratch.dml_core_mode_support_locals;
+ struct dml2_core_calcs_calculate_tdlut_setting_params *calculate_tdlut_setting_params = &mode_lib->scratch.calculate_tdlut_setting_params;
+ struct dml2_core_calcs_CalculatePrefetchSchedule_params *CalculatePrefetchSchedule_params = &mode_lib->scratch.CalculatePrefetchSchedule_params;
+ struct dml2_core_calcs_calculate_peak_bandwidth_required_params *calculate_peak_bandwidth_params = &mode_lib->scratch.calculate_peak_bandwidth_params;
+#ifdef DML_GLOBAL_PREFETCH_CHECK
+ struct dml2_core_calcs_CheckGlobalPrefetchAdmissibility_params *CheckGlobalPrefetchAdmissibility_params = &mode_lib->scratch.CheckGlobalPrefetchAdmissibility_params;
+#endif
+ struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params *CalculateWatermarks_params = &mode_lib->scratch.CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params;
+
+ double min_return_bw_for_latency;
+ unsigned int k;
+
+ mode_lib->ms.TimeCalc = 24 / mode_lib->ms.dcfclk_deepsleep;
+
+ calculate_hostvm_inefficiency_factor(
+ &s->HostVMInefficiencyFactor,
+ &s->HostVMInefficiencyFactorPrefetch,
+
+ display_cfg->gpuvm_enable,
+ display_cfg->hostvm_enable,
+ mode_lib->ip.remote_iommu_outstanding_translations,
+ mode_lib->soc.max_outstanding_reqs,
+ mode_lib->ms.support.urg_bandwidth_available_pixel_and_vm[dml2_core_internal_soc_state_sys_active],
+ mode_lib->ms.support.urg_bandwidth_available_vm_only[dml2_core_internal_soc_state_sys_active]);
+
+ mode_lib->ms.Total3dlutActive = 0;
+ for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
+ if (display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut)
+ mode_lib->ms.Total3dlutActive = mode_lib->ms.Total3dlutActive + 1;
+
+ // Calculate tdlut schedule related terms
+ calculate_tdlut_setting_params->dispclk_mhz = mode_lib->ms.RequiredDISPCLK;
+ calculate_tdlut_setting_params->setup_for_tdlut = display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut;
+ calculate_tdlut_setting_params->tdlut_width_mode = display_cfg->plane_descriptors[k].tdlut.tdlut_width_mode;
+ calculate_tdlut_setting_params->tdlut_addressing_mode = display_cfg->plane_descriptors[k].tdlut.tdlut_addressing_mode;
+ calculate_tdlut_setting_params->cursor_buffer_size = mode_lib->ip.cursor_buffer_size;
+ calculate_tdlut_setting_params->gpuvm_enable = display_cfg->gpuvm_enable;
+ calculate_tdlut_setting_params->gpuvm_page_size_kbytes = display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes;
+ calculate_tdlut_setting_params->tdlut_mpc_width_flag = display_cfg->plane_descriptors[k].tdlut.tdlut_mpc_width_flag;
+ calculate_tdlut_setting_params->is_gfx11 = dml_get_gfx_version(display_cfg->plane_descriptors[k].surface.tiling);
+
+ // output
+ calculate_tdlut_setting_params->tdlut_pte_bytes_per_frame = &s->tdlut_pte_bytes_per_frame[k];
+ calculate_tdlut_setting_params->tdlut_bytes_per_frame = &s->tdlut_bytes_per_frame[k];
+ calculate_tdlut_setting_params->tdlut_groups_per_2row_ub = &s->tdlut_groups_per_2row_ub[k];
+ calculate_tdlut_setting_params->tdlut_opt_time = &s->tdlut_opt_time[k];
+ calculate_tdlut_setting_params->tdlut_drain_time = &s->tdlut_drain_time[k];
+ calculate_tdlut_setting_params->tdlut_bytes_to_deliver = &s->tdlut_bytes_to_deliver[k];
+ calculate_tdlut_setting_params->tdlut_bytes_per_group = &s->tdlut_bytes_per_group[k];
+
+ calculate_tdlut_setting(&mode_lib->scratch, calculate_tdlut_setting_params);
+ }
+
+ min_return_bw_for_latency = mode_lib->ms.support.urg_bandwidth_available_min_latency[dml2_core_internal_soc_state_sys_active];
+
+ if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn3)
+ s->ReorderingBytes = (unsigned int)(mode_lib->soc.clk_table.dram_config.channel_count * math_max3(mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_pixel_only_bytes,
+ mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_pixel_and_vm_bytes,
+ mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_vm_only_bytes));
+
+ CalculateExtraLatency(
+ display_cfg,
+ mode_lib->ip.rob_buffer_size_kbytes,
+ mode_lib->soc.qos_parameters.qos_params.dcn32x.loaded_round_trip_latency_fclk_cycles,
+ s->ReorderingBytes,
+ mode_lib->ms.DCFCLK,
+ mode_lib->ms.FabricClock,
+ mode_lib->ip.pixel_chunk_size_kbytes,
+ min_return_bw_for_latency,
+ mode_lib->ms.num_active_planes,
+ mode_lib->ms.NoOfDPP,
+ mode_lib->ms.dpte_group_bytes,
+ s->tdlut_bytes_per_group,
+ s->HostVMInefficiencyFactor,
+ s->HostVMInefficiencyFactorPrefetch,
+ mode_lib->soc.hostvm_min_page_size_kbytes,
+ mode_lib->soc.qos_parameters.qos_type,
+ !(display_cfg->overrides.max_outstanding_when_urgent_expected_disable),
+ mode_lib->soc.max_outstanding_reqs,
+ mode_lib->ms.support.request_size_bytes_luma,
+ mode_lib->ms.support.request_size_bytes_chroma,
+ mode_lib->ip.meta_chunk_size_kbytes,
+ mode_lib->ip.dchub_arb_to_ret_delay,
+ mode_lib->ms.TripToMemory,
+ mode_lib->ip.hostvm_mode,
+
+ // output
+ &mode_lib->ms.ExtraLatency,
+ &mode_lib->ms.ExtraLatency_sr,
+ &mode_lib->ms.ExtraLatencyPrefetch);
+
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++)
+ s->impacted_dst_y_pre[k] = 0;
+
+ s->recalc_prefetch_schedule = 0;
+ s->recalc_prefetch_done = 0;
+ do {
+ mode_lib->ms.support.PrefetchSupported = true;
+
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
+ s->line_times[k] = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
+ s->pixel_format[k] = display_cfg->plane_descriptors[k].pixel_format;
+
+ s->lb_source_lines_l[k] = get_num_lb_source_lines(mode_lib->ip.max_line_buffer_lines, mode_lib->ip.line_buffer_size_bits,
+ mode_lib->ms.NoOfDPP[k],
+ display_cfg->plane_descriptors[k].composition.viewport.plane0.width,
+ display_cfg->plane_descriptors[k].composition.viewport.plane0.height,
+ display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio,
+ display_cfg->plane_descriptors[k].composition.rotation_angle);
+
+ s->lb_source_lines_c[k] = get_num_lb_source_lines(mode_lib->ip.max_line_buffer_lines, mode_lib->ip.line_buffer_size_bits,
+ mode_lib->ms.NoOfDPP[k],
+ display_cfg->plane_descriptors[k].composition.viewport.plane1.width,
+ display_cfg->plane_descriptors[k].composition.viewport.plane1.height,
+ display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio,
+ display_cfg->plane_descriptors[k].composition.rotation_angle);
+
+ struct dml2_core_internal_DmlPipe *myPipe = &s->myPipe;
+
+ mode_lib->ms.TWait[k] = CalculateTWait(
+ display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns,
+ mode_lib->ms.UrgLatency,
+ mode_lib->ms.TripToMemory,
+ !dml_is_phantom_pipe(&display_cfg->plane_descriptors[k]) && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.drr_config.enabled ?
+ get_g6_temp_read_blackout_us(&mode_lib->soc, (unsigned int)(mode_lib->ms.uclk_freq_mhz * 1000), mode_lib->ms.state_idx) : 0.0);
+
+ myPipe->Dppclk = mode_lib->ms.RequiredDPPCLK[k];
+ myPipe->Dispclk = mode_lib->ms.RequiredDISPCLK;
+ myPipe->PixelClock = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
+ myPipe->DCFClkDeepSleep = mode_lib->ms.dcfclk_deepsleep;
+ myPipe->DPPPerSurface = mode_lib->ms.NoOfDPP[k];
+ myPipe->ScalerEnabled = display_cfg->plane_descriptors[k].composition.scaler_info.enabled;
+ myPipe->VRatio = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
+ myPipe->VRatioChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
+ myPipe->VTaps = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps;
+ myPipe->VTapsChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps;
+ myPipe->RotationAngle = display_cfg->plane_descriptors[k].composition.rotation_angle;
+ myPipe->mirrored = display_cfg->plane_descriptors[k].composition.mirrored;
+ myPipe->BlockWidth256BytesY = mode_lib->ms.Read256BlockWidthY[k];
+ myPipe->BlockHeight256BytesY = mode_lib->ms.Read256BlockHeightY[k];
+ myPipe->BlockWidth256BytesC = mode_lib->ms.Read256BlockWidthC[k];
+ myPipe->BlockHeight256BytesC = mode_lib->ms.Read256BlockHeightC[k];
+ myPipe->InterlaceEnable = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced;
+ myPipe->NumberOfCursors = display_cfg->plane_descriptors[k].cursor.num_cursors;
+ myPipe->VBlank = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active;
+ myPipe->HTotal = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total;
+ myPipe->HActive = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active;
+ myPipe->DCCEnable = display_cfg->plane_descriptors[k].surface.dcc.enable;
+ myPipe->ODMMode = mode_lib->ms.ODMMode[k];
+ myPipe->SourcePixelFormat = display_cfg->plane_descriptors[k].pixel_format;
+ myPipe->BytePerPixelY = mode_lib->ms.BytePerPixelY[k];
+ myPipe->BytePerPixelC = mode_lib->ms.BytePerPixelC[k];
+ myPipe->ProgressiveToInterlaceUnitInOPP = mode_lib->ip.ptoi_supported;
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: Calling CalculatePrefetchSchedule for k=%u\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: MaximumVStartup = %u\n", __func__, s->MaximumVStartup[k]);
+#endif
+ CalculatePrefetchSchedule_params->display_cfg = display_cfg;
+ CalculatePrefetchSchedule_params->HostVMInefficiencyFactor = s->HostVMInefficiencyFactorPrefetch;
+ CalculatePrefetchSchedule_params->myPipe = myPipe;
+ CalculatePrefetchSchedule_params->DSCDelay = mode_lib->ms.DSCDelay[k];
+ CalculatePrefetchSchedule_params->DPPCLKDelaySubtotalPlusCNVCFormater = mode_lib->ip.dppclk_delay_subtotal + mode_lib->ip.dppclk_delay_cnvc_formatter;
+ CalculatePrefetchSchedule_params->DPPCLKDelaySCL = mode_lib->ip.dppclk_delay_scl;
+ CalculatePrefetchSchedule_params->DPPCLKDelaySCLLBOnly = mode_lib->ip.dppclk_delay_scl_lb_only;
+ CalculatePrefetchSchedule_params->DPPCLKDelayCNVCCursor = mode_lib->ip.dppclk_delay_cnvc_cursor;
+ CalculatePrefetchSchedule_params->DISPCLKDelaySubtotal = mode_lib->ip.dispclk_delay_subtotal;
+ CalculatePrefetchSchedule_params->DPP_RECOUT_WIDTH = (unsigned int)(mode_lib->ms.SwathWidthY[k] / display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
+ CalculatePrefetchSchedule_params->OutputFormat = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format;
+ CalculatePrefetchSchedule_params->MaxInterDCNTileRepeaters = mode_lib->ip.max_inter_dcn_tile_repeaters;
+ CalculatePrefetchSchedule_params->VStartup = s->MaximumVStartup[k];
+ CalculatePrefetchSchedule_params->HostVMMinPageSize = mode_lib->soc.hostvm_min_page_size_kbytes;
+ CalculatePrefetchSchedule_params->DynamicMetadataEnable = display_cfg->plane_descriptors[k].dynamic_meta_data.enable;
+ CalculatePrefetchSchedule_params->DynamicMetadataVMEnabled = mode_lib->ip.dynamic_metadata_vm_enabled;
+ CalculatePrefetchSchedule_params->DynamicMetadataLinesBeforeActiveRequired = display_cfg->plane_descriptors[k].dynamic_meta_data.lines_before_active_required;
+ CalculatePrefetchSchedule_params->DynamicMetadataTransmittedBytes = display_cfg->plane_descriptors[k].dynamic_meta_data.transmitted_bytes;
+ CalculatePrefetchSchedule_params->UrgentLatency = mode_lib->ms.UrgLatency;
+ CalculatePrefetchSchedule_params->ExtraLatencyPrefetch = mode_lib->ms.ExtraLatencyPrefetch;
+ CalculatePrefetchSchedule_params->TCalc = mode_lib->ms.TimeCalc;
+ CalculatePrefetchSchedule_params->vm_bytes = mode_lib->ms.vm_bytes[k];
+ CalculatePrefetchSchedule_params->PixelPTEBytesPerRow = mode_lib->ms.DPTEBytesPerRow[k];
+ CalculatePrefetchSchedule_params->PrefetchSourceLinesY = mode_lib->ms.PrefetchLinesY[k];
+ CalculatePrefetchSchedule_params->VInitPreFillY = mode_lib->ms.PrefillY[k];
+ CalculatePrefetchSchedule_params->MaxNumSwathY = mode_lib->ms.MaxNumSwathY[k];
+ CalculatePrefetchSchedule_params->PrefetchSourceLinesC = mode_lib->ms.PrefetchLinesC[k];
+ CalculatePrefetchSchedule_params->VInitPreFillC = mode_lib->ms.PrefillC[k];
+ CalculatePrefetchSchedule_params->MaxNumSwathC = mode_lib->ms.MaxNumSwathC[k];
+ CalculatePrefetchSchedule_params->swath_width_luma_ub = mode_lib->ms.swath_width_luma_ub[k];
+ CalculatePrefetchSchedule_params->swath_width_chroma_ub = mode_lib->ms.swath_width_chroma_ub[k];
+ CalculatePrefetchSchedule_params->SwathHeightY = mode_lib->ms.SwathHeightY[k];
+ CalculatePrefetchSchedule_params->SwathHeightC = mode_lib->ms.SwathHeightC[k];
+ CalculatePrefetchSchedule_params->TWait = mode_lib->ms.TWait[k];
+ CalculatePrefetchSchedule_params->Ttrip = mode_lib->ms.TripToMemory;
+ CalculatePrefetchSchedule_params->Turg = mode_lib->ms.UrgLatency;
+ CalculatePrefetchSchedule_params->setup_for_tdlut = display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut;
+ CalculatePrefetchSchedule_params->tdlut_pte_bytes_per_frame = s->tdlut_pte_bytes_per_frame[k];
+ CalculatePrefetchSchedule_params->tdlut_bytes_per_frame = s->tdlut_bytes_per_frame[k];
+ CalculatePrefetchSchedule_params->tdlut_opt_time = s->tdlut_opt_time[k];
+ CalculatePrefetchSchedule_params->tdlut_drain_time = s->tdlut_drain_time[k];
+ CalculatePrefetchSchedule_params->num_cursors = (display_cfg->plane_descriptors[k].cursor.cursor_width > 0);
+ CalculatePrefetchSchedule_params->cursor_bytes_per_chunk = s->cursor_bytes_per_chunk[k];
+ CalculatePrefetchSchedule_params->cursor_bytes_per_line = s->cursor_bytes_per_line[k];
+ CalculatePrefetchSchedule_params->dcc_enable = display_cfg->plane_descriptors[k].surface.dcc.enable;
+ CalculatePrefetchSchedule_params->mrq_present = mode_lib->ip.dcn_mrq_present;
+ CalculatePrefetchSchedule_params->meta_row_bytes = mode_lib->ms.meta_row_bytes[k];
+ CalculatePrefetchSchedule_params->mall_prefetch_sdp_overhead_factor = mode_lib->ms.mall_prefetch_sdp_overhead_factor[k];
+ CalculatePrefetchSchedule_params->impacted_dst_y_pre = s->impacted_dst_y_pre[k];
+ CalculatePrefetchSchedule_params->vactive_sw_bw_l = mode_lib->ms.vactive_sw_bw_l[k];
+ CalculatePrefetchSchedule_params->vactive_sw_bw_c = mode_lib->ms.vactive_sw_bw_c[k];
+
+ // output
+ CalculatePrefetchSchedule_params->DSTXAfterScaler = &s->DSTXAfterScaler[k];
+ CalculatePrefetchSchedule_params->DSTYAfterScaler = &s->DSTYAfterScaler[k];
+ CalculatePrefetchSchedule_params->dst_y_prefetch = &mode_lib->ms.dst_y_prefetch[k];
+ CalculatePrefetchSchedule_params->dst_y_per_vm_vblank = &mode_lib->ms.LinesForVM[k];
+ CalculatePrefetchSchedule_params->dst_y_per_row_vblank = &mode_lib->ms.LinesForDPTERow[k];
+ CalculatePrefetchSchedule_params->VRatioPrefetchY = &mode_lib->ms.VRatioPreY[k];
+ CalculatePrefetchSchedule_params->VRatioPrefetchC = &mode_lib->ms.VRatioPreC[k];
+ CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWLuma = &mode_lib->ms.RequiredPrefetchPixelDataBWLuma[k]; // prefetch_sw_bw_l
+ CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWChroma = &mode_lib->ms.RequiredPrefetchPixelDataBWChroma[k]; // prefetch_sw_bw_c
+ CalculatePrefetchSchedule_params->RequiredPrefetchBWMax = &mode_lib->ms.RequiredPrefetchBWMax[k];
+ CalculatePrefetchSchedule_params->NotEnoughTimeForDynamicMetadata = &mode_lib->ms.NoTimeForDynamicMetadata[k];
+ CalculatePrefetchSchedule_params->Tno_bw = &mode_lib->ms.Tno_bw[k];
+ CalculatePrefetchSchedule_params->Tno_bw_flip = &mode_lib->ms.Tno_bw_flip[k];
+ CalculatePrefetchSchedule_params->prefetch_vmrow_bw = &mode_lib->ms.prefetch_vmrow_bw[k];
+ CalculatePrefetchSchedule_params->Tdmdl_vm = &s->dummy_single[0];
+ CalculatePrefetchSchedule_params->Tdmdl = &s->dummy_single[1];
+ CalculatePrefetchSchedule_params->TSetup = &s->dummy_single[2];
+ CalculatePrefetchSchedule_params->Tvm_trips = &s->Tvm_trips[k];
+ CalculatePrefetchSchedule_params->Tr0_trips = &s->Tr0_trips[k];
+ CalculatePrefetchSchedule_params->Tvm_trips_flip = &s->Tvm_trips_flip[k];
+ CalculatePrefetchSchedule_params->Tr0_trips_flip = &s->Tr0_trips_flip[k];
+ CalculatePrefetchSchedule_params->Tvm_trips_flip_rounded = &s->Tvm_trips_flip_rounded[k];
+ CalculatePrefetchSchedule_params->Tr0_trips_flip_rounded = &s->Tr0_trips_flip_rounded[k];
+ CalculatePrefetchSchedule_params->VUpdateOffsetPix = &s->dummy_integer[0];
+ CalculatePrefetchSchedule_params->VUpdateWidthPix = &s->dummy_integer[1];
+ CalculatePrefetchSchedule_params->VReadyOffsetPix = &s->dummy_integer[2];
+ CalculatePrefetchSchedule_params->prefetch_cursor_bw = &mode_lib->ms.prefetch_cursor_bw[k];
+ CalculatePrefetchSchedule_params->prefetch_sw_bytes = &s->prefetch_sw_bytes[k];
+ CalculatePrefetchSchedule_params->Tpre_rounded = &s->Tpre_rounded[k];
+ CalculatePrefetchSchedule_params->Tpre_oto = &s->Tpre_oto[k];
+ CalculatePrefetchSchedule_params->prefetch_swath_time_us = &s->prefetch_swath_time_us[k];
+
+ mode_lib->ms.NoTimeForPrefetch[k] = CalculatePrefetchSchedule(&mode_lib->scratch, CalculatePrefetchSchedule_params);
+
+ mode_lib->ms.support.PrefetchSupported &= !mode_lib->ms.NoTimeForPrefetch[k];
+ DML_LOG_VERBOSE("DML::%s: k=%d, dst_y_per_vm_vblank = %f\n", __func__, k, *CalculatePrefetchSchedule_params->dst_y_per_vm_vblank);
+ DML_LOG_VERBOSE("DML::%s: k=%d, dst_y_per_row_vblank = %f\n", __func__, k, *CalculatePrefetchSchedule_params->dst_y_per_row_vblank);
+ } // for k num_planes
+
+ CalculateDCFCLKDeepSleepTdlut(
+ display_cfg,
+ mode_lib->ms.num_active_planes,
+ mode_lib->ms.BytePerPixelY,
+ mode_lib->ms.BytePerPixelC,
+ mode_lib->ms.SwathWidthY,
+ mode_lib->ms.SwathWidthC,
+ mode_lib->ms.NoOfDPP,
+ mode_lib->ms.PSCL_FACTOR,
+ mode_lib->ms.PSCL_FACTOR_CHROMA,
+ mode_lib->ms.RequiredDPPCLK,
+ mode_lib->ms.vactive_sw_bw_l,
+ mode_lib->ms.vactive_sw_bw_c,
+ mode_lib->soc.return_bus_width_bytes,
+ mode_lib->ms.RequiredDISPCLK,
+ s->tdlut_bytes_to_deliver,
+ s->prefetch_swath_time_us,
+
+ /* Output */
+ &mode_lib->ms.dcfclk_deepsleep);
+
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
+ if (mode_lib->ms.dst_y_prefetch[k] < 2.0
+ || mode_lib->ms.LinesForVM[k] >= 32.0
+ || mode_lib->ms.LinesForDPTERow[k] >= 16.0
+ || mode_lib->ms.NoTimeForPrefetch[k] == true
+ || s->DSTYAfterScaler[k] > 8) {
+ mode_lib->ms.support.PrefetchSupported = false;
+ DML_LOG_VERBOSE("DML::%s: k=%d, dst_y_prefetch=%f (should not be < 2)\n", __func__, k, mode_lib->ms.dst_y_prefetch[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, LinesForVM=%f (should not be >= 32)\n", __func__, k, mode_lib->ms.LinesForVM[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, LinesForDPTERow=%f (should not be >= 16)\n", __func__, k, mode_lib->ms.LinesForDPTERow[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, DSTYAfterScaler=%d (should be <= 8)\n", __func__, k, s->DSTYAfterScaler[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, NoTimeForPrefetch=%d\n", __func__, k, mode_lib->ms.NoTimeForPrefetch[k]);
+ }
+ }
+
+ mode_lib->ms.support.DynamicMetadataSupported = true;
+ for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
+ if (mode_lib->ms.NoTimeForDynamicMetadata[k] == true) {
+ mode_lib->ms.support.DynamicMetadataSupported = false;
+ }
+ }
+
+ mode_lib->ms.support.VRatioInPrefetchSupported = true;
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
+ if (mode_lib->ms.VRatioPreY[k] > __DML2_CALCS_MAX_VRATIO_PRE__ ||
+ mode_lib->ms.VRatioPreC[k] > __DML2_CALCS_MAX_VRATIO_PRE__) {
+ mode_lib->ms.support.VRatioInPrefetchSupported = false;
+ DML_LOG_VERBOSE("DML::%s: k=%d VRatioPreY = %f (should be <= %f)\n", __func__, k, mode_lib->ms.VRatioPreY[k], __DML2_CALCS_MAX_VRATIO_PRE__);
+ DML_LOG_VERBOSE("DML::%s: k=%d VRatioPreC = %f (should be <= %f)\n", __func__, k, mode_lib->ms.VRatioPreC[k], __DML2_CALCS_MAX_VRATIO_PRE__);
+ DML_LOG_VERBOSE("DML::%s: VRatioInPrefetchSupported = %u\n", __func__, mode_lib->ms.support.VRatioInPrefetchSupported);
+ }
+ }
+
+ mode_lib->ms.support.PrefetchSupported &= mode_lib->ms.support.VRatioInPrefetchSupported;
+
+ // By default, do not recalc prefetch schedule
+ s->recalc_prefetch_schedule = 0;
+
+ // Only do urg vs prefetch bandwidth check, flip schedule check, power saving feature support check IF the Prefetch Schedule Check is ok
+ if (mode_lib->ms.support.PrefetchSupported) {
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
+ // Calculate Urgent burst factor for prefetch
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: k=%d, Calling CalculateUrgentBurstFactor (for prefetch)\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: k=%d, VRatioPreY=%f\n", __func__, k, mode_lib->ms.VRatioPreY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, VRatioPreC=%f\n", __func__, k, mode_lib->ms.VRatioPreC[k]);
+#endif
+ CalculateUrgentBurstFactor(
+ &display_cfg->plane_descriptors[k],
+ mode_lib->ms.swath_width_luma_ub[k],
+ mode_lib->ms.swath_width_chroma_ub[k],
+ mode_lib->ms.SwathHeightY[k],
+ mode_lib->ms.SwathHeightC[k],
+ s->line_times[k],
+ mode_lib->ms.UrgLatency,
+ mode_lib->ms.VRatioPreY[k],
+ mode_lib->ms.VRatioPreC[k],
+ mode_lib->ms.BytePerPixelInDETY[k],
+ mode_lib->ms.BytePerPixelInDETC[k],
+ mode_lib->ms.DETBufferSizeY[k],
+ mode_lib->ms.DETBufferSizeC[k],
+ /* Output */
+ &mode_lib->ms.UrgentBurstFactorLumaPre[k],
+ &mode_lib->ms.UrgentBurstFactorChromaPre[k],
+ &mode_lib->ms.NotEnoughUrgentLatencyHidingPre[k]);
+ }
+
+ // Calculate urgent bandwidth required, both urg and non urg peak bandwidth
+ // assume flip bw is 0 at this point
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++)
+ mode_lib->ms.final_flip_bw[k] = 0;
+
+ calculate_peak_bandwidth_params->urg_vactive_bandwidth_required = mode_lib->ms.support.urg_vactive_bandwidth_required;
+ calculate_peak_bandwidth_params->urg_bandwidth_required = mode_lib->ms.support.urg_bandwidth_required;
+ calculate_peak_bandwidth_params->urg_bandwidth_required_qual = mode_lib->ms.support.urg_bandwidth_required_qual;
+ calculate_peak_bandwidth_params->non_urg_bandwidth_required = mode_lib->ms.support.non_urg_bandwidth_required;
+ calculate_peak_bandwidth_params->surface_avg_vactive_required_bw = mode_lib->ms.surface_avg_vactive_required_bw;
+ calculate_peak_bandwidth_params->surface_peak_required_bw = mode_lib->ms.surface_peak_required_bw;
+
+ calculate_peak_bandwidth_params->display_cfg = display_cfg;
+ calculate_peak_bandwidth_params->inc_flip_bw = 0;
+ calculate_peak_bandwidth_params->num_active_planes = mode_lib->ms.num_active_planes;
+ calculate_peak_bandwidth_params->num_of_dpp = mode_lib->ms.NoOfDPP;
+ calculate_peak_bandwidth_params->dcc_dram_bw_nom_overhead_factor_p0 = mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0;
+ calculate_peak_bandwidth_params->dcc_dram_bw_nom_overhead_factor_p1 = mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1;
+ calculate_peak_bandwidth_params->dcc_dram_bw_pref_overhead_factor_p0 = mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p0;
+ calculate_peak_bandwidth_params->dcc_dram_bw_pref_overhead_factor_p1 = mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p1;
+ calculate_peak_bandwidth_params->mall_prefetch_sdp_overhead_factor = mode_lib->ms.mall_prefetch_sdp_overhead_factor;
+ calculate_peak_bandwidth_params->mall_prefetch_dram_overhead_factor = mode_lib->ms.mall_prefetch_dram_overhead_factor;
+
+ calculate_peak_bandwidth_params->surface_read_bandwidth_l = mode_lib->ms.vactive_sw_bw_l;
+ calculate_peak_bandwidth_params->surface_read_bandwidth_c = mode_lib->ms.vactive_sw_bw_c;
+ calculate_peak_bandwidth_params->prefetch_bandwidth_l = mode_lib->ms.RequiredPrefetchPixelDataBWLuma;
+ calculate_peak_bandwidth_params->prefetch_bandwidth_c = mode_lib->ms.RequiredPrefetchPixelDataBWChroma;
+ calculate_peak_bandwidth_params->prefetch_bandwidth_max = mode_lib->ms.RequiredPrefetchBWMax;
+ calculate_peak_bandwidth_params->excess_vactive_fill_bw_l = mode_lib->ms.excess_vactive_fill_bw_l;
+ calculate_peak_bandwidth_params->excess_vactive_fill_bw_c = mode_lib->ms.excess_vactive_fill_bw_c;
+ calculate_peak_bandwidth_params->cursor_bw = mode_lib->ms.cursor_bw;
+ calculate_peak_bandwidth_params->dpte_row_bw = mode_lib->ms.dpte_row_bw;
+ calculate_peak_bandwidth_params->meta_row_bw = mode_lib->ms.meta_row_bw;
+ calculate_peak_bandwidth_params->prefetch_cursor_bw = mode_lib->ms.prefetch_cursor_bw;
+ calculate_peak_bandwidth_params->prefetch_vmrow_bw = mode_lib->ms.prefetch_vmrow_bw;
+ calculate_peak_bandwidth_params->flip_bw = mode_lib->ms.final_flip_bw;
+ calculate_peak_bandwidth_params->urgent_burst_factor_l = mode_lib->ms.UrgentBurstFactorLuma;
+ calculate_peak_bandwidth_params->urgent_burst_factor_c = mode_lib->ms.UrgentBurstFactorChroma;
+ calculate_peak_bandwidth_params->urgent_burst_factor_cursor = mode_lib->ms.UrgentBurstFactorCursor;
+ calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_l = mode_lib->ms.UrgentBurstFactorLumaPre;
+ calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_c = mode_lib->ms.UrgentBurstFactorChromaPre;
+ calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_cursor = mode_lib->ms.UrgentBurstFactorCursorPre;
+
+ calculate_peak_bandwidth_required(
+ &mode_lib->scratch,
+ calculate_peak_bandwidth_params);
+
+ // Check urg peak bandwidth against available urg bw
+ // check at SDP and DRAM, for all soc states (SVP prefetch an Sys Active)
+ check_urgent_bandwidth_support(
+ &s->dummy_single[0], // double* frac_urg_bandwidth
+ &s->dummy_single[1], // double* frac_urg_bandwidth_mall
+ &mode_lib->ms.support.UrgVactiveBandwidthSupport,
+ &mode_lib->ms.support.PrefetchBandwidthSupported,
+
+ mode_lib->soc.mall_allocated_for_dcn_mbytes,
+ mode_lib->ms.support.non_urg_bandwidth_required,
+ mode_lib->ms.support.urg_vactive_bandwidth_required,
+ mode_lib->ms.support.urg_bandwidth_required,
+ mode_lib->ms.support.urg_bandwidth_available);
+
+ mode_lib->ms.support.PrefetchSupported &= mode_lib->ms.support.PrefetchBandwidthSupported;
+ DML_LOG_VERBOSE("DML::%s: PrefetchBandwidthSupported=%0d\n", __func__, mode_lib->ms.support.PrefetchBandwidthSupported);
+
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
+ if (mode_lib->ms.NotEnoughUrgentLatencyHidingPre[k]) {
+ mode_lib->ms.support.PrefetchSupported = false;
+ DML_LOG_VERBOSE("DML::%s: k=%d, NotEnoughUrgentLatencyHidingPre=%d\n", __func__, k, mode_lib->ms.NotEnoughUrgentLatencyHidingPre[k]);
+ }
+ }
+
+#ifdef DML_GLOBAL_PREFETCH_CHECK
+ if (mode_lib->ms.support.PrefetchSupported && mode_lib->ms.num_active_planes > 1 && s->recalc_prefetch_done == 0) {
+ CheckGlobalPrefetchAdmissibility_params->num_active_planes = mode_lib->ms.num_active_planes;
+ CheckGlobalPrefetchAdmissibility_params->pixel_format = s->pixel_format;
+ CheckGlobalPrefetchAdmissibility_params->chunk_bytes_l = mode_lib->ip.pixel_chunk_size_kbytes * 1024;
+ CheckGlobalPrefetchAdmissibility_params->chunk_bytes_c = mode_lib->ip.pixel_chunk_size_kbytes * 1024;
+ CheckGlobalPrefetchAdmissibility_params->lb_source_lines_l = s->lb_source_lines_l;
+ CheckGlobalPrefetchAdmissibility_params->lb_source_lines_c = s->lb_source_lines_c;
+ CheckGlobalPrefetchAdmissibility_params->swath_height_l = mode_lib->ms.SwathHeightY;
+ CheckGlobalPrefetchAdmissibility_params->swath_height_c = mode_lib->ms.SwathHeightC;
+ CheckGlobalPrefetchAdmissibility_params->rob_buffer_size_kbytes = mode_lib->ip.rob_buffer_size_kbytes;
+ CheckGlobalPrefetchAdmissibility_params->compressed_buffer_size_kbytes = mode_lib->ms.CompressedBufferSizeInkByte;
+ CheckGlobalPrefetchAdmissibility_params->detile_buffer_size_bytes_l = mode_lib->ms.DETBufferSizeY;
+ CheckGlobalPrefetchAdmissibility_params->detile_buffer_size_bytes_c = mode_lib->ms.DETBufferSizeC;
+ CheckGlobalPrefetchAdmissibility_params->full_swath_bytes_l = s->full_swath_bytes_l;
+ CheckGlobalPrefetchAdmissibility_params->full_swath_bytes_c = s->full_swath_bytes_c;
+ CheckGlobalPrefetchAdmissibility_params->prefetch_sw_bytes = s->prefetch_sw_bytes;
+ CheckGlobalPrefetchAdmissibility_params->Tpre_rounded = s->Tpre_rounded;
+ CheckGlobalPrefetchAdmissibility_params->Tpre_oto = s->Tpre_oto;
+ CheckGlobalPrefetchAdmissibility_params->estimated_urg_bandwidth_required_mbps = mode_lib->ms.support.urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
+ CheckGlobalPrefetchAdmissibility_params->line_time = s->line_times;
+ CheckGlobalPrefetchAdmissibility_params->dst_y_prefetch = mode_lib->ms.dst_y_prefetch;
+ if (CheckGlobalPrefetchAdmissibility_params->estimated_urg_bandwidth_required_mbps < 10 * 1024)
+ CheckGlobalPrefetchAdmissibility_params->estimated_urg_bandwidth_required_mbps = 10 * 1024;
+
+ CheckGlobalPrefetchAdmissibility_params->estimated_dcfclk_mhz = (CheckGlobalPrefetchAdmissibility_params->estimated_urg_bandwidth_required_mbps / (double) mode_lib->soc.return_bus_width_bytes) /
+ ((double)mode_lib->soc.qos_parameters.derate_table.system_active_urgent.dcfclk_derate_percent / 100.0);
+
+ // if recalc_prefetch_schedule is set, recalculate the prefetch schedule with the new impacted_Tpre, prefetch should be possible
+ CheckGlobalPrefetchAdmissibility_params->recalc_prefetch_schedule = &s->recalc_prefetch_schedule;
+ CheckGlobalPrefetchAdmissibility_params->impacted_dst_y_pre = s->impacted_dst_y_pre;
+ mode_lib->ms.support.PrefetchSupported = CheckGlobalPrefetchAdmissibility(&mode_lib->scratch, CheckGlobalPrefetchAdmissibility_params);
+ s->recalc_prefetch_done = 1;
+ s->recalc_prefetch_schedule = 1;
+ }
+#endif
+ } // prefetch schedule ok, do urg bw and flip schedule
+ } while (s->recalc_prefetch_schedule);
+
+ // Flip Schedule
+ // Both prefetch schedule and BW okay
+ if (mode_lib->ms.support.PrefetchSupported == true) {
+ mode_lib->ms.BandwidthAvailableForImmediateFlip =
+ get_bandwidth_available_for_immediate_flip(
+ dml2_core_internal_soc_state_sys_active,
+ mode_lib->ms.support.urg_bandwidth_required_qual, // no flip
+ mode_lib->ms.support.urg_bandwidth_available);
+
+ mode_lib->ms.TotImmediateFlipBytes = 0;
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
+ if (display_cfg->plane_descriptors[k].immediate_flip) {
+ s->per_pipe_flip_bytes[k] = get_pipe_flip_bytes(
+ s->HostVMInefficiencyFactor,
+ mode_lib->ms.vm_bytes[k],
+ mode_lib->ms.DPTEBytesPerRow[k],
+ mode_lib->ms.meta_row_bytes[k]);
+ } else {
+ s->per_pipe_flip_bytes[k] = 0;
+ }
+ mode_lib->ms.TotImmediateFlipBytes += s->per_pipe_flip_bytes[k] * mode_lib->ms.NoOfDPP[k];
+
+ }
+
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
+ CalculateFlipSchedule(
+ &mode_lib->scratch,
+ display_cfg->plane_descriptors[k].immediate_flip,
+ 1, // use_lb_flip_bw
+ s->HostVMInefficiencyFactor,
+ s->Tvm_trips_flip[k],
+ s->Tr0_trips_flip[k],
+ s->Tvm_trips_flip_rounded[k],
+ s->Tr0_trips_flip_rounded[k],
+ display_cfg->gpuvm_enable,
+ mode_lib->ms.vm_bytes[k],
+ mode_lib->ms.DPTEBytesPerRow[k],
+ mode_lib->ms.BandwidthAvailableForImmediateFlip,
+ mode_lib->ms.TotImmediateFlipBytes,
+ display_cfg->plane_descriptors[k].pixel_format,
+ (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)),
+ display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio,
+ display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio,
+ mode_lib->ms.Tno_bw_flip[k],
+ mode_lib->ms.dpte_row_height[k],
+ mode_lib->ms.dpte_row_height_chroma[k],
+ mode_lib->ms.use_one_row_for_frame_flip[k],
+ mode_lib->ip.max_flip_time_us,
+ mode_lib->ip.max_flip_time_lines,
+ s->per_pipe_flip_bytes[k],
+ mode_lib->ms.meta_row_bytes[k],
+ s->meta_row_height_luma[k],
+ s->meta_row_height_chroma[k],
+ mode_lib->ip.dcn_mrq_present && display_cfg->plane_descriptors[k].surface.dcc.enable,
+
+ /* Output */
+ &mode_lib->ms.dst_y_per_vm_flip[k],
+ &mode_lib->ms.dst_y_per_row_flip[k],
+ &mode_lib->ms.final_flip_bw[k],
+ &mode_lib->ms.ImmediateFlipSupportedForPipe[k]);
+ }
+
+ calculate_peak_bandwidth_params->urg_vactive_bandwidth_required = s->dummy_bw;
+ calculate_peak_bandwidth_params->urg_bandwidth_required = mode_lib->ms.support.urg_bandwidth_required_flip;
+ calculate_peak_bandwidth_params->urg_bandwidth_required_qual = s->dummy_bw;
+ calculate_peak_bandwidth_params->non_urg_bandwidth_required = mode_lib->ms.support.non_urg_bandwidth_required_flip;
+ calculate_peak_bandwidth_params->surface_avg_vactive_required_bw = s->surface_dummy_bw;
+ calculate_peak_bandwidth_params->surface_peak_required_bw = mode_lib->ms.surface_peak_required_bw;
+
+ calculate_peak_bandwidth_params->display_cfg = display_cfg;
+ calculate_peak_bandwidth_params->inc_flip_bw = 1;
+ calculate_peak_bandwidth_params->num_active_planes = mode_lib->ms.num_active_planes;
+ calculate_peak_bandwidth_params->num_of_dpp = mode_lib->ms.NoOfDPP;
+ calculate_peak_bandwidth_params->dcc_dram_bw_nom_overhead_factor_p0 = mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0;
+ calculate_peak_bandwidth_params->dcc_dram_bw_nom_overhead_factor_p1 = mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1;
+ calculate_peak_bandwidth_params->dcc_dram_bw_pref_overhead_factor_p0 = mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p0;
+ calculate_peak_bandwidth_params->dcc_dram_bw_pref_overhead_factor_p1 = mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p1;
+ calculate_peak_bandwidth_params->mall_prefetch_sdp_overhead_factor = mode_lib->ms.mall_prefetch_sdp_overhead_factor;
+ calculate_peak_bandwidth_params->mall_prefetch_dram_overhead_factor = mode_lib->ms.mall_prefetch_dram_overhead_factor;
+
+ calculate_peak_bandwidth_params->surface_read_bandwidth_l = mode_lib->ms.vactive_sw_bw_l;
+ calculate_peak_bandwidth_params->surface_read_bandwidth_c = mode_lib->ms.vactive_sw_bw_c;
+ calculate_peak_bandwidth_params->prefetch_bandwidth_l = mode_lib->ms.RequiredPrefetchPixelDataBWLuma;
+ calculate_peak_bandwidth_params->prefetch_bandwidth_c = mode_lib->ms.RequiredPrefetchPixelDataBWChroma;
+ calculate_peak_bandwidth_params->prefetch_bandwidth_max = mode_lib->ms.RequiredPrefetchBWMax;
+ calculate_peak_bandwidth_params->excess_vactive_fill_bw_l = mode_lib->ms.excess_vactive_fill_bw_l;
+ calculate_peak_bandwidth_params->excess_vactive_fill_bw_c = mode_lib->ms.excess_vactive_fill_bw_c;
+ calculate_peak_bandwidth_params->cursor_bw = mode_lib->ms.cursor_bw;
+ calculate_peak_bandwidth_params->dpte_row_bw = mode_lib->ms.dpte_row_bw;
+ calculate_peak_bandwidth_params->meta_row_bw = mode_lib->ms.meta_row_bw;
+ calculate_peak_bandwidth_params->prefetch_cursor_bw = mode_lib->ms.prefetch_cursor_bw;
+ calculate_peak_bandwidth_params->prefetch_vmrow_bw = mode_lib->ms.prefetch_vmrow_bw;
+ calculate_peak_bandwidth_params->flip_bw = mode_lib->ms.final_flip_bw;
+ calculate_peak_bandwidth_params->urgent_burst_factor_l = mode_lib->ms.UrgentBurstFactorLuma;
+ calculate_peak_bandwidth_params->urgent_burst_factor_c = mode_lib->ms.UrgentBurstFactorChroma;
+ calculate_peak_bandwidth_params->urgent_burst_factor_cursor = mode_lib->ms.UrgentBurstFactorCursor;
+ calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_l = mode_lib->ms.UrgentBurstFactorLumaPre;
+ calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_c = mode_lib->ms.UrgentBurstFactorChromaPre;
+ calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_cursor = mode_lib->ms.UrgentBurstFactorCursorPre;
+
+ calculate_peak_bandwidth_required(
+ &mode_lib->scratch,
+ calculate_peak_bandwidth_params);
+
+ calculate_immediate_flip_bandwidth_support(
+ &s->dummy_single[0], // double* frac_urg_bandwidth_flip
+ &mode_lib->ms.support.ImmediateFlipSupport,
+
+ dml2_core_internal_soc_state_sys_active,
+ mode_lib->ms.support.urg_bandwidth_required_flip,
+ mode_lib->ms.support.non_urg_bandwidth_required_flip,
+ mode_lib->ms.support.urg_bandwidth_available);
+
+ for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
+ if (display_cfg->plane_descriptors[k].immediate_flip == true && mode_lib->ms.ImmediateFlipSupportedForPipe[k] == false)
+ mode_lib->ms.support.ImmediateFlipSupport = false;
+ }
+
+ } else { // if prefetch not support, assume iflip is not supported too
+ mode_lib->ms.support.ImmediateFlipSupport = false;
+ }
+
+ s->mSOCParameters.UrgentLatency = mode_lib->ms.UrgLatency;
+ s->mSOCParameters.ExtraLatency = mode_lib->ms.ExtraLatency;
+ s->mSOCParameters.ExtraLatency_sr = mode_lib->ms.ExtraLatency_sr;
+ s->mSOCParameters.WritebackLatency = mode_lib->soc.qos_parameters.writeback.base_latency_us;
+ s->mSOCParameters.DRAMClockChangeLatency = mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us;
+ s->mSOCParameters.FCLKChangeLatency = mode_lib->soc.power_management_parameters.fclk_change_blackout_us;
+ s->mSOCParameters.SRExitTime = mode_lib->soc.power_management_parameters.stutter_exit_latency_us;
+ s->mSOCParameters.SREnterPlusExitTime = mode_lib->soc.power_management_parameters.stutter_enter_plus_exit_latency_us;
+ s->mSOCParameters.SRExitZ8Time = mode_lib->soc.power_management_parameters.z8_stutter_exit_latency_us;
+ s->mSOCParameters.SREnterPlusExitZ8Time = mode_lib->soc.power_management_parameters.z8_stutter_enter_plus_exit_latency_us;
+ s->mSOCParameters.USRRetrainingLatency = 0;
+ s->mSOCParameters.SMNLatency = 0;
+ s->mSOCParameters.g6_temp_read_blackout_us = get_g6_temp_read_blackout_us(&mode_lib->soc, (unsigned int)(mode_lib->ms.uclk_freq_mhz * 1000), mode_lib->ms.state_idx);
+ s->mSOCParameters.max_urgent_latency_us = get_max_urgent_latency_us(&mode_lib->soc.qos_parameters.qos_params.dcn4x, mode_lib->ms.uclk_freq_mhz, mode_lib->ms.FabricClock, mode_lib->ms.state_idx);
+ s->mSOCParameters.df_response_time_us = mode_lib->soc.qos_parameters.qos_params.dcn4x.df_qos_response_time_fclk_cycles / mode_lib->ms.FabricClock;
+ s->mSOCParameters.qos_type = mode_lib->soc.qos_parameters.qos_type;
+
+ CalculateWatermarks_params->display_cfg = display_cfg;
+ CalculateWatermarks_params->USRRetrainingRequired = false;
+ CalculateWatermarks_params->NumberOfActiveSurfaces = mode_lib->ms.num_active_planes;
+ CalculateWatermarks_params->MaxLineBufferLines = mode_lib->ip.max_line_buffer_lines;
+ CalculateWatermarks_params->LineBufferSize = mode_lib->ip.line_buffer_size_bits;
+ CalculateWatermarks_params->WritebackInterfaceBufferSize = mode_lib->ip.writeback_interface_buffer_size_kbytes;
+ CalculateWatermarks_params->DCFCLK = mode_lib->ms.DCFCLK;
+ CalculateWatermarks_params->SynchronizeTimings = display_cfg->overrides.synchronize_timings;
+ CalculateWatermarks_params->SynchronizeDRRDisplaysForUCLKPStateChange = display_cfg->overrides.synchronize_ddr_displays_for_uclk_pstate_change;
+ CalculateWatermarks_params->dpte_group_bytes = mode_lib->ms.dpte_group_bytes;
+ CalculateWatermarks_params->mmSOCParameters = s->mSOCParameters;
+ CalculateWatermarks_params->WritebackChunkSize = mode_lib->ip.writeback_chunk_size_kbytes;
+ CalculateWatermarks_params->SOCCLK = mode_lib->ms.SOCCLK;
+ CalculateWatermarks_params->DCFClkDeepSleep = mode_lib->ms.dcfclk_deepsleep;
+ CalculateWatermarks_params->DETBufferSizeY = mode_lib->ms.DETBufferSizeY;
+ CalculateWatermarks_params->DETBufferSizeC = mode_lib->ms.DETBufferSizeC;
+ CalculateWatermarks_params->SwathHeightY = mode_lib->ms.SwathHeightY;
+ CalculateWatermarks_params->SwathHeightC = mode_lib->ms.SwathHeightC;
+ CalculateWatermarks_params->SwathWidthY = mode_lib->ms.SwathWidthY;
+ CalculateWatermarks_params->SwathWidthC = mode_lib->ms.SwathWidthC;
+ CalculateWatermarks_params->DPPPerSurface = mode_lib->ms.NoOfDPP;
+ CalculateWatermarks_params->BytePerPixelDETY = mode_lib->ms.BytePerPixelInDETY;
+ CalculateWatermarks_params->BytePerPixelDETC = mode_lib->ms.BytePerPixelInDETC;
+ CalculateWatermarks_params->DSTXAfterScaler = s->DSTXAfterScaler;
+ CalculateWatermarks_params->DSTYAfterScaler = s->DSTYAfterScaler;
+ CalculateWatermarks_params->UnboundedRequestEnabled = mode_lib->ms.UnboundedRequestEnabled;
+ CalculateWatermarks_params->CompressedBufferSizeInkByte = mode_lib->ms.CompressedBufferSizeInkByte;
+ CalculateWatermarks_params->meta_row_height_l = s->meta_row_height_luma;
+ CalculateWatermarks_params->meta_row_height_c = s->meta_row_height_chroma;
+
+ // Output
+ CalculateWatermarks_params->Watermark = &mode_lib->ms.support.watermarks; // Watermarks *Watermark
+ CalculateWatermarks_params->DRAMClockChangeSupport = mode_lib->ms.support.DRAMClockChangeSupport;
+ CalculateWatermarks_params->global_dram_clock_change_supported = &mode_lib->ms.support.global_dram_clock_change_supported;
+ CalculateWatermarks_params->MaxActiveDRAMClockChangeLatencySupported = &s->dummy_single_array[0]; // double *MaxActiveDRAMClockChangeLatencySupported[]
+ CalculateWatermarks_params->SubViewportLinesNeededInMALL = mode_lib->ms.SubViewportLinesNeededInMALL; // unsigned int SubViewportLinesNeededInMALL[]
+ CalculateWatermarks_params->FCLKChangeSupport = mode_lib->ms.support.FCLKChangeSupport;
+ CalculateWatermarks_params->global_fclk_change_supported = &mode_lib->ms.support.global_fclk_change_supported;
+ CalculateWatermarks_params->MaxActiveFCLKChangeLatencySupported = &s->dummy_single[0]; // double *MaxActiveFCLKChangeLatencySupported
+ CalculateWatermarks_params->USRRetrainingSupport = &mode_lib->ms.support.USRRetrainingSupport;
+ CalculateWatermarks_params->g6_temp_read_support = &mode_lib->ms.support.g6_temp_read_support;
+ CalculateWatermarks_params->VActiveLatencyHidingMargin = mode_lib->ms.VActiveLatencyHidingMargin;
+ CalculateWatermarks_params->VActiveLatencyHidingUs = mode_lib->ms.VActiveLatencyHidingUs;
+
+ CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(&mode_lib->scratch, CalculateWatermarks_params);
+
+ calculate_pstate_keepout_dst_lines(display_cfg, &mode_lib->ms.support.watermarks, s->dummy_integer_array[0]);
+ DML_LOG_VERBOSE("DML::%s: Done prefetch calculation\n", __func__);
+
+}
+
+
static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out_params)
{
struct dml2_core_internal_display_mode_lib *mode_lib = in_out_params->mode_lib;
const struct dml2_display_cfg *display_cfg = in_out_params->in_display_cfg;
const struct dml2_mcg_min_clock_table *min_clk_table = in_out_params->min_clk_table;
-#if defined(__DML_VBA_DEBUG__)
- double old_ReadBandwidthLuma;
- double old_ReadBandwidthChroma;
-#endif
double outstanding_latency_us = 0;
- double min_return_bw_for_latency;
struct dml2_core_calcs_mode_support_locals *s = &mode_lib->scratch.dml_core_mode_support_locals;
- struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params *CalculateWatermarks_params = &mode_lib->scratch.CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params;
struct dml2_core_calcs_CalculateVMRowAndSwath_params *CalculateVMRowAndSwath_params = &mode_lib->scratch.CalculateVMRowAndSwath_params;
struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params *CalculateSwathAndDETConfiguration_params = &mode_lib->scratch.CalculateSwathAndDETConfiguration_params;
- struct dml2_core_calcs_CalculatePrefetchSchedule_params *CalculatePrefetchSchedule_params = &mode_lib->scratch.CalculatePrefetchSchedule_params;
- struct dml2_core_calcs_calculate_tdlut_setting_params *calculate_tdlut_setting_params = &mode_lib->scratch.calculate_tdlut_setting_params;
struct dml2_core_calcs_calculate_mcache_setting_params *calculate_mcache_setting_params = &mode_lib->scratch.calculate_mcache_setting_params;
- struct dml2_core_calcs_calculate_peak_bandwidth_required_params *calculate_peak_bandwidth_params = &mode_lib->scratch.calculate_peak_bandwidth_params;
struct dml2_core_calcs_calculate_bytes_to_fetch_required_to_hide_latency_params *calculate_bytes_to_fetch_required_to_hide_latency_params = &mode_lib->scratch.calculate_bytes_to_fetch_required_to_hide_latency_params;
unsigned int k, m, n;
@@ -7054,9 +7942,9 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.FabricClock = ((double)min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].min_fclk_khz / 1000);
mode_lib->ms.MaxDCFCLK = (double)min_clk_table->max_clocks_khz.dcfclk / 1000;
mode_lib->ms.MaxFabricClock = (double)min_clk_table->max_clocks_khz.fclk / 1000;
- mode_lib->ms.max_dispclk_freq_mhz = (double)min_clk_table->max_clocks_khz.dispclk / 1000;
+ mode_lib->ms.max_dispclk_freq_mhz = (double)min_clk_table->max_ss_clocks_khz.dispclk / 1000;
mode_lib->ms.max_dscclk_freq_mhz = (double)min_clk_table->max_clocks_khz.dscclk / 1000;
- mode_lib->ms.max_dppclk_freq_mhz = (double)min_clk_table->max_clocks_khz.dppclk / 1000;
+ mode_lib->ms.max_dppclk_freq_mhz = (double)min_clk_table->max_ss_clocks_khz.dppclk / 1000;
mode_lib->ms.uclk_freq_mhz = dram_bw_kbps_to_uclk_mhz(min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].pre_derate_dram_bw_kbps, &mode_lib->soc.clk_table.dram_config);
mode_lib->ms.dram_bw_mbps = ((double)min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].pre_derate_dram_bw_kbps / 1000);
mode_lib->ms.max_dram_bw_mbps = ((double)min_clk_table->dram_bw_table.entries[min_clk_table->dram_bw_table.num_entries - 1].pre_derate_dram_bw_kbps / 1000);
@@ -7064,31 +7952,25 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.active_min_uclk_dpm_index = get_active_min_uclk_dpm_index((unsigned int) (mode_lib->ms.uclk_freq_mhz * 1000.0), &mode_lib->soc.clk_table);
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: --- START --- \n", __func__);
- dml2_printf("DML::%s: num_active_planes = %u\n", __func__, mode_lib->ms.num_active_planes);
- dml2_printf("DML::%s: min_clk_index = %0d\n", __func__, in_out_params->min_clk_index);
- dml2_printf("DML::%s: qos_param_index = %0d\n", __func__, mode_lib->ms.qos_param_index);
- dml2_printf("DML::%s: SOCCLK = %f\n", __func__, mode_lib->ms.SOCCLK);
- dml2_printf("DML::%s: dram_bw_mbps = %f\n", __func__, mode_lib->ms.dram_bw_mbps);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, mode_lib->ms.uclk_freq_mhz);
- dml2_printf("DML::%s: DCFCLK = %f\n", __func__, mode_lib->ms.DCFCLK);
- dml2_printf("DML::%s: FabricClock = %f\n", __func__, mode_lib->ms.FabricClock);
- dml2_printf("DML::%s: MaxDCFCLK = %f\n", __func__, mode_lib->ms.MaxDCFCLK);
- dml2_printf("DML::%s: max_dispclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dispclk_freq_mhz);
- dml2_printf("DML::%s: max_dscclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dscclk_freq_mhz);
- dml2_printf("DML::%s: max_dppclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dppclk_freq_mhz);
- dml2_printf("DML::%s: MaxFabricClock = %f\n", __func__, mode_lib->ms.MaxFabricClock);
- dml2_printf("DML::%s: ip.compressed_buffer_segment_size_in_kbytes = %u\n", __func__, mode_lib->ip.compressed_buffer_segment_size_in_kbytes);
- dml2_printf("DML::%s: ip.dcn_mrq_present = %u\n", __func__, mode_lib->ip.dcn_mrq_present);
+ DML_LOG_VERBOSE("DML::%s: --- START --- \n", __func__);
+ DML_LOG_VERBOSE("DML::%s: num_active_planes = %u\n", __func__, mode_lib->ms.num_active_planes);
+ DML_LOG_VERBOSE("DML::%s: min_clk_index = %0d\n", __func__, in_out_params->min_clk_index);
+ DML_LOG_VERBOSE("DML::%s: qos_param_index = %0d\n", __func__, mode_lib->ms.qos_param_index);
+ DML_LOG_VERBOSE("DML::%s: SOCCLK = %f\n", __func__, mode_lib->ms.SOCCLK);
+ DML_LOG_VERBOSE("DML::%s: dram_bw_mbps = %f\n", __func__, mode_lib->ms.dram_bw_mbps);
+ DML_LOG_VERBOSE("DML::%s: uclk_freq_mhz = %f\n", __func__, mode_lib->ms.uclk_freq_mhz);
+ DML_LOG_VERBOSE("DML::%s: DCFCLK = %f\n", __func__, mode_lib->ms.DCFCLK);
+ DML_LOG_VERBOSE("DML::%s: FabricClock = %f\n", __func__, mode_lib->ms.FabricClock);
+ DML_LOG_VERBOSE("DML::%s: MaxDCFCLK = %f\n", __func__, mode_lib->ms.MaxDCFCLK);
+ DML_LOG_VERBOSE("DML::%s: max_dispclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dispclk_freq_mhz);
+ DML_LOG_VERBOSE("DML::%s: max_dscclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dscclk_freq_mhz);
+ DML_LOG_VERBOSE("DML::%s: max_dppclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dppclk_freq_mhz);
+ DML_LOG_VERBOSE("DML::%s: MaxFabricClock = %f\n", __func__, mode_lib->ms.MaxFabricClock);
+ DML_LOG_VERBOSE("DML::%s: ip.compressed_buffer_segment_size_in_kbytes = %u\n", __func__, mode_lib->ip.compressed_buffer_segment_size_in_kbytes);
+ DML_LOG_VERBOSE("DML::%s: ip.dcn_mrq_present = %u\n", __func__, mode_lib->ip.dcn_mrq_present);
for (k = 0; k < mode_lib->ms.num_active_planes; k++)
- dml2_printf("DML::%s: plane_%d: reserved_vblank_time_ns = %u\n", __func__, k, display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns);
-
- // dml2_printf_dml_policy(&mode_lib->ms.policy);
- // dml2_printf_dml_display_cfg_timing(&display_cfg->timing, mode_lib->ms.num_active_planes);
- // dml2_printf_dml_display_cfg_plane(&display_cfg->plane, mode_lib->ms.num_active_planes);
- // dml2_printf_dml_display_cfg_surface(&display_cfg->surface, mode_lib->ms.num_active_planes);
- // dml2_printf_dml_display_cfg_output(&display_cfg->output, mode_lib->ms.num_active_planes);
+ DML_LOG_VERBOSE("DML::%s: plane_%d: reserved_vblank_time_ns = %lu\n", __func__, k, display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns);
#endif
CalculateMaxDETAndMinCompressedBufferSize(
@@ -7183,46 +8065,44 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
}
for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- mode_lib->ms.SurfaceReadBandwidthLuma[k] = mode_lib->ms.SwathWidthYSingleDPP[k] * math_ceil2(mode_lib->ms.BytePerPixelY[k], 1.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- mode_lib->ms.SurfaceReadBandwidthChroma[k] = mode_lib->ms.SwathWidthCSingleDPP[k] * math_ceil2(mode_lib->ms.BytePerPixelC[k], 2.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
+ mode_lib->ms.vactive_sw_bw_l[k] = mode_lib->ms.SwathWidthYSingleDPP[k] * math_ceil2(mode_lib->ms.BytePerPixelY[k], 1.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
+ mode_lib->ms.vactive_sw_bw_c[k] = mode_lib->ms.SwathWidthCSingleDPP[k] * math_ceil2(mode_lib->ms.BytePerPixelC[k], 2.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
mode_lib->ms.cursor_bw[k] = display_cfg->plane_descriptors[k].cursor.num_cursors * display_cfg->plane_descriptors[k].cursor.cursor_width *
display_cfg->plane_descriptors[k].cursor.cursor_bpp / 8.0 / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000));
#ifdef __DML_VBA_DEBUG__
- old_ReadBandwidthLuma = mode_lib->ms.SwathWidthYSingleDPP[k] * math_ceil2(mode_lib->ms.BytePerPixelInDETY[k], 1.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- old_ReadBandwidthChroma = mode_lib->ms.SwathWidthYSingleDPP[k] / 2 * math_ceil2(mode_lib->ms.BytePerPixelInDETC[k], 2.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio / 2.0;
- dml2_printf("DML::%s: k=%u, old_ReadBandwidthLuma = %f\n", __func__, k, old_ReadBandwidthLuma);
- dml2_printf("DML::%s: k=%u, old_ReadBandwidthChroma = %f\n", __func__, k, old_ReadBandwidthChroma);
- dml2_printf("DML::%s: k=%u, ReadBandwidthLuma = %f\n", __func__, k, mode_lib->ms.SurfaceReadBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%u, ReadBandwidthChroma = %f\n", __func__, k, mode_lib->ms.SurfaceReadBandwidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, old_ReadBandwidthLuma = %f\n", __func__, k, mode_lib->ms.SwathWidthYSingleDPP[k] * math_ceil2(mode_lib->ms.BytePerPixelInDETY[k], 1.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%u, old_ReadBandwidthChroma = %f\n", __func__, k, mode_lib->ms.SwathWidthYSingleDPP[k] / 2 * math_ceil2(mode_lib->ms.BytePerPixelInDETC[k], 2.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio / 2.0);
+ DML_LOG_VERBOSE("DML::%s: k=%u, vactive_sw_bw_l = %f\n", __func__, k, mode_lib->ms.vactive_sw_bw_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, vactive_sw_bw_c = %f\n", __func__, k, mode_lib->ms.vactive_sw_bw_c[k]);
#endif
}
// Writeback bandwidth
for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format == dml2_444_64) {
- mode_lib->ms.WriteBandwidth[k] = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height
- * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width
- / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height
+ if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0 && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].pixel_format == dml2_444_64) {
+ mode_lib->ms.WriteBandwidth[k][0] = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_height
+ * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_width
+ / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].input_height
* display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total
/ ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * 8.0;
- } else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- mode_lib->ms.WriteBandwidth[k] = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height
- * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width
- / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height
+ } else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0) {
+ mode_lib->ms.WriteBandwidth[k][0] = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_height
+ * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_width
+ / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].input_height
* display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total
/ ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * 4.0;
} else {
- mode_lib->ms.WriteBandwidth[k] = 0.0;
+ mode_lib->ms.WriteBandwidth[k][0] = 0.0;
}
}
/*Writeback Latency support check*/
mode_lib->ms.support.WritebackLatencySupport = true;
for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true &&
- (mode_lib->ms.WriteBandwidth[k] > mode_lib->ip.writeback_interface_buffer_size_kbytes * 1024 / ((double)mode_lib->soc.qos_parameters.writeback.base_latency_us))) {
+ if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0 &&
+ (mode_lib->ms.WriteBandwidth[k][0] > mode_lib->ip.writeback_interface_buffer_size_kbytes * 1024 / ((double)mode_lib->soc.qos_parameters.writeback.base_latency_us))) {
mode_lib->ms.support.WritebackLatencySupport = false;
}
}
@@ -7231,19 +8111,19 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
/* Writeback Scale Ratio and Taps Support Check */
mode_lib->ms.support.WritebackScaleRatioAndTapsSupport = true;
for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio > mode_lib->ip.writeback_max_hscl_ratio
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio > mode_lib->ip.writeback_max_vscl_ratio
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio < mode_lib->ip.writeback_min_hscl_ratio
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio < mode_lib->ip.writeback_min_vscl_ratio
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps > (unsigned int) mode_lib->ip.writeback_max_hscl_taps
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps > (unsigned int) mode_lib->ip.writeback_max_vscl_taps
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio > (unsigned int)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio > (unsigned int)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps
- || (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps > 2.0 && ((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps % 2) == 1))) {
+ if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0) {
+ if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_ratio > mode_lib->ip.writeback_max_hscl_ratio
+ || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_ratio > mode_lib->ip.writeback_max_vscl_ratio
+ || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_ratio < mode_lib->ip.writeback_min_hscl_ratio
+ || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_ratio < mode_lib->ip.writeback_min_vscl_ratio
+ || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_taps > (unsigned int) mode_lib->ip.writeback_max_hscl_taps
+ || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_taps > (unsigned int) mode_lib->ip.writeback_max_vscl_taps
+ || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_ratio > (unsigned int)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_taps
+ || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_ratio > (unsigned int)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_taps
+ || (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_taps > 2.0 && ((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_taps % 2) == 1))) {
mode_lib->ms.support.WritebackScaleRatioAndTapsSupport = false;
}
- if (2.0 * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height * (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps - 1) * 57 > mode_lib->ip.writeback_line_buffer_buffer_size) {
+ if (2.0 * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_height * (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_taps - 1) * 57 > mode_lib->ip.writeback_line_buffer_buffer_size) {
mode_lib->ms.support.WritebackScaleRatioAndTapsSupport = false;
}
}
@@ -7315,13 +8195,13 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.MaximumSwathWidthLuma[k] = math_min2(s->MaximumSwathWidthSupportLuma, mode_lib->ms.MaximumSwathWidthInLineBufferLuma);
mode_lib->ms.MaximumSwathWidthChroma[k] = math_min2(s->MaximumSwathWidthSupportChroma, mode_lib->ms.MaximumSwathWidthInLineBufferChroma);
- dml2_printf("DML::%s: k=%u MaximumSwathWidthLuma=%f\n", __func__, k, mode_lib->ms.MaximumSwathWidthLuma[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathWidthSupportLuma=%u\n", __func__, k, s->MaximumSwathWidthSupportLuma);
- dml2_printf("DML::%s: k=%u MaximumSwathWidthInLineBufferLuma=%f\n", __func__, k, mode_lib->ms.MaximumSwathWidthInLineBufferLuma);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathWidthLuma=%f\n", __func__, k, mode_lib->ms.MaximumSwathWidthLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathWidthSupportLuma=%u\n", __func__, k, s->MaximumSwathWidthSupportLuma);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathWidthInLineBufferLuma=%f\n", __func__, k, mode_lib->ms.MaximumSwathWidthInLineBufferLuma);
- dml2_printf("DML::%s: k=%u MaximumSwathWidthChroma=%f\n", __func__, k, mode_lib->ms.MaximumSwathWidthChroma[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathWidthSupportChroma=%u\n", __func__, k, s->MaximumSwathWidthSupportChroma);
- dml2_printf("DML::%s: k=%u MaximumSwathWidthInLineBufferChroma=%f\n", __func__, k, mode_lib->ms.MaximumSwathWidthInLineBufferChroma);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathWidthChroma=%f\n", __func__, k, mode_lib->ms.MaximumSwathWidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathWidthSupportChroma=%u\n", __func__, k, s->MaximumSwathWidthSupportChroma);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaximumSwathWidthInLineBufferChroma=%f\n", __func__, k, mode_lib->ms.MaximumSwathWidthInLineBufferChroma);
}
/* Cursor Support Check */
@@ -7358,11 +8238,11 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.support.AlignedCPitch[k] > display_cfg->plane_descriptors[k].surface.plane1.pitch) {
mode_lib->ms.support.PitchSupport = false;
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: k=%u AlignedYPitch = %d\n", __func__, k, mode_lib->ms.support.AlignedYPitch[k]);
- dml2_printf("DML::%s: k=%u PitchY = %d\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane0.pitch);
- dml2_printf("DML::%s: k=%u AlignedCPitch = %d\n", __func__, k, mode_lib->ms.support.AlignedCPitch[k]);
- dml2_printf("DML::%s: k=%u PitchC = %d\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane1.pitch);
- dml2_printf("DML::%s: k=%u PitchSupport = %d\n", __func__, k, mode_lib->ms.support.PitchSupport);
+ DML_LOG_VERBOSE("DML::%s: k=%u AlignedYPitch = %d\n", __func__, k, mode_lib->ms.support.AlignedYPitch[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u PitchY = %ld\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane0.pitch);
+ DML_LOG_VERBOSE("DML::%s: k=%u AlignedCPitch = %d\n", __func__, k, mode_lib->ms.support.AlignedCPitch[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u PitchC = %ld\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane1.pitch);
+ DML_LOG_VERBOSE("DML::%s: k=%u PitchSupport = %d\n", __func__, k, mode_lib->ms.support.PitchSupport);
#endif
}
@@ -7394,11 +8274,11 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
display_cfg->plane_descriptors[k].composition.viewport.plane0.height > display_cfg->plane_descriptors[k].surface.plane0.height) {
mode_lib->ms.support.ViewportExceedsSurface = true;
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: k=%u ViewportWidth = %d\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.width);
- dml2_printf("DML::%s: k=%u SurfaceWidthY = %d\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane0.width);
- dml2_printf("DML::%s: k=%u ViewportHeight = %d\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.height);
- dml2_printf("DML::%s: k=%u SurfaceHeightY = %d\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane0.height);
- dml2_printf("DML::%s: k=%u ViewportExceedsSurface = %d\n", __func__, k, mode_lib->ms.support.ViewportExceedsSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u ViewportWidth = %ld\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.width);
+ DML_LOG_VERBOSE("DML::%s: k=%u SurfaceWidthY = %ld\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane0.width);
+ DML_LOG_VERBOSE("DML::%s: k=%u ViewportHeight = %ld\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.height);
+ DML_LOG_VERBOSE("DML::%s: k=%u SurfaceHeightY = %ld\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane0.height);
+ DML_LOG_VERBOSE("DML::%s: k=%u ViewportExceedsSurface = %d\n", __func__, k, mode_lib->ms.support.ViewportExceedsSurface);
#endif
}
if (dml_is_420(display_cfg->plane_descriptors[k].pixel_format) || display_cfg->plane_descriptors[k].pixel_format == dml2_rgbe_alpha) {
@@ -7423,8 +8303,8 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
CalculateSwathAndDETConfiguration_params->nomDETInKByte = mode_lib->ms.NomDETInKByte;
CalculateSwathAndDETConfiguration_params->ConfigReturnBufferSegmentSizeInkByte = mode_lib->ip.config_return_buffer_segment_size_in_kbytes;
CalculateSwathAndDETConfiguration_params->CompressedBufferSegmentSizeInkByte = mode_lib->ip.compressed_buffer_segment_size_in_kbytes;
- CalculateSwathAndDETConfiguration_params->ReadBandwidthLuma = mode_lib->ms.SurfaceReadBandwidthLuma;
- CalculateSwathAndDETConfiguration_params->ReadBandwidthChroma = mode_lib->ms.SurfaceReadBandwidthChroma;
+ CalculateSwathAndDETConfiguration_params->ReadBandwidthLuma = mode_lib->ms.vactive_sw_bw_l;
+ CalculateSwathAndDETConfiguration_params->ReadBandwidthChroma = mode_lib->ms.vactive_sw_bw_c;
CalculateSwathAndDETConfiguration_params->MaximumSwathWidthLuma = mode_lib->ms.MaximumSwathWidthLuma;
CalculateSwathAndDETConfiguration_params->MaximumSwathWidthChroma = mode_lib->ms.MaximumSwathWidthChroma;
CalculateSwathAndDETConfiguration_params->Read256BytesBlockHeightY = mode_lib->ms.Read256BlockHeightY;
@@ -7580,8 +8460,8 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.TotalNumberOfActiveDPP = mode_lib->ms.TotalNumberOfActiveDPP + s->NumberOfDPPDSC;
}
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: k=%d RequiresDSC = %d\n", __func__, k, mode_lib->ms.RequiresDSC[k]);
- dml2_printf("DML::%s: k=%d ODMMode = %d\n", __func__, k, mode_lib->ms.ODMMode[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d RequiresDSC = %d\n", __func__, k, mode_lib->ms.RequiresDSC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d ODMMode = %d\n", __func__, k, mode_lib->ms.ODMMode[k]);
#endif
// ensure the number dsc slices is integer multiple based on ODM mode
@@ -7597,9 +8477,9 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.support.DSCSlicesODMModeSupported = ((mode_lib->ms.support.NumberOfDSCSlices[k] % 4) == 0);
#if defined(__DML_VBA_DEBUG__)
if (!mode_lib->ms.support.DSCSlicesODMModeSupported) {
- dml2_printf("DML::%s: k=%d Invalid dsc num_slices and ODM mode setting\n", __func__, k);
- dml2_printf("DML::%s: k=%d num_slices = %d\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.overrides.num_slices);
- dml2_printf("DML::%s: k=%d ODMMode = %d\n", __func__, k, mode_lib->ms.ODMMode[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d Invalid dsc num_slices and ODM mode setting\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: k=%d num_slices = %d\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.overrides.num_slices);
+ DML_LOG_VERBOSE("DML::%s: k=%d ODMMode = %d\n", __func__, k, mode_lib->ms.ODMMode[k]);
}
#endif
} else {
@@ -7644,7 +8524,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.MPCCombine[k] = false;
mode_lib->ms.NoOfDPP[k] = 1;
if (!mode_lib->ms.SingleDPPViewportSizeSupportPerSurface[k]) {
- dml2_printf("WARNING: DML::%s: MPCC is override to disable but viewport is too large to be supported with single pipe!\n", __func__);
+ DML_LOG_VERBOSE("WARNING: DML::%s: MPCC is override to disable but viewport is too large to be supported with single pipe!\n", __func__);
}
} else {
if ((mode_lib->ms.MinDPPCLKUsingSingleDPP[k] > mode_lib->ms.max_dppclk_freq_mhz) || !mode_lib->ms.SingleDPPViewportSizeSupportPerSurface[k]) {
@@ -7654,7 +8534,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
}
}
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: k=%d, NoOfDPP = %d\n", __func__, k, mode_lib->ms.NoOfDPP[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, NoOfDPP = %d\n", __func__, k, mode_lib->ms.NoOfDPP[k]);
#endif
}
@@ -7671,16 +8551,16 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
//DISPCLK/DPPCLK
mode_lib->ms.WritebackRequiredDISPCLK = 0;
for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable) {
+ if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0) {
mode_lib->ms.WritebackRequiredDISPCLK = math_max2(mode_lib->ms.WritebackRequiredDISPCLK,
- CalculateWriteBackDISPCLK(display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format,
+ CalculateWriteBackDISPCLK(display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].pixel_format,
((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000),
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_width,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_ratio,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_ratio,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_taps,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_taps,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].input_width,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_width,
display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total,
mode_lib->ip.writeback_line_buffer_buffer_size));
}
@@ -7712,7 +8592,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
if (!s->stream_visited[display_cfg->plane_descriptors[k].stream_index]) {
s->stream_visited[display_cfg->plane_descriptors[k].stream_index] = 1;
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true)
+ if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0)
s->TotalNumberOfActiveWriteback = s->TotalNumberOfActiveWriteback + 1;
s->TotalNumberOfActiveOTG = s->TotalNumberOfActiveOTG + 1;
@@ -7824,7 +8704,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.audio_sample_rate,
display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.audio_sample_layout);
- if (mode_lib->ms.RequiredDTBCLK[k] > ((double)min_clk_table->max_clocks_khz.dtbclk / 1000)) {
+ if (mode_lib->ms.RequiredDTBCLK[k] > ((double)min_clk_table->max_ss_clocks_khz.dtbclk / 1000)) {
mode_lib->ms.support.DTBCLKRequiredMoreThanSupported = true;
}
} else {
@@ -7853,7 +8733,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
s->DSCFormatFactor = 1;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, RequiresDSC = %u\n", __func__, k, mode_lib->ms.RequiresDSC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, RequiresDSC = %u\n", __func__, k, mode_lib->ms.RequiresDSC[k]);
#endif
if (mode_lib->ms.RequiresDSC[k] == true) {
s->PixelClockBackEndFactor = 3.0;
@@ -7871,10 +8751,10 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, PixelClockBackEnd = %f\n", __func__, k, s->PixelClockBackEnd[k]);
- dml2_printf("DML::%s: k=%u, required_dscclk_freq_mhz = %f\n", __func__, k, mode_lib->ms.required_dscclk_freq_mhz[k]);
- dml2_printf("DML::%s: k=%u, DSCFormatFactor = %u\n", __func__, k, s->DSCFormatFactor);
- dml2_printf("DML::%s: k=%u, DSCCLKRequiredMoreThanSupported = %u\n", __func__, k, mode_lib->ms.support.DSCCLKRequiredMoreThanSupported);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelClockBackEnd = %f\n", __func__, k, s->PixelClockBackEnd[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, required_dscclk_freq_mhz = %f\n", __func__, k, mode_lib->ms.required_dscclk_freq_mhz[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DSCFormatFactor = %u\n", __func__, k, s->DSCFormatFactor);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DSCCLKRequiredMoreThanSupported = %u\n", __func__, k, mode_lib->ms.support.DSCCLKRequiredMoreThanSupported);
#endif
}
}
@@ -8109,13 +8989,13 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.support.DCCMetaBufferSizeNotExceeded = false;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, PTEBufferSizeNotExceeded = %u\n", __func__, k, mode_lib->ms.PTEBufferSizeNotExceeded[k]);
- dml2_printf("DML::%s: k=%u, DCCMetaBufferSizeNotExceeded = %u\n", __func__, k, mode_lib->ms.DCCMetaBufferSizeNotExceeded[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PTEBufferSizeNotExceeded = %u\n", __func__, k, mode_lib->ms.PTEBufferSizeNotExceeded[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DCCMetaBufferSizeNotExceeded = %u\n", __func__, k, mode_lib->ms.DCCMetaBufferSizeNotExceeded[k]);
#endif
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: PTEBufferSizeNotExceeded = %u\n", __func__, mode_lib->ms.support.PTEBufferSizeNotExceeded);
- dml2_printf("DML::%s: DCCMetaBufferSizeNotExceeded = %u\n", __func__, mode_lib->ms.support.DCCMetaBufferSizeNotExceeded);
+ DML_LOG_VERBOSE("DML::%s: PTEBufferSizeNotExceeded = %u\n", __func__, mode_lib->ms.support.PTEBufferSizeNotExceeded);
+ DML_LOG_VERBOSE("DML::%s: DCCMetaBufferSizeNotExceeded = %u\n", __func__, mode_lib->ms.support.DCCMetaBufferSizeNotExceeded);
#endif
/* VActive bytes to fetch for UCLK P-State */
@@ -8188,7 +9068,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
double line_time_us = (double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- bool cursor_not_enough_urgent_latency_hiding = 0;
+ bool cursor_not_enough_urgent_latency_hiding = false;
if (display_cfg->plane_descriptors[k].cursor.num_cursors > 0) {
calculate_cursor_req_attributes(
@@ -8217,9 +9097,9 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.UrgentBurstFactorCursorPre[k] = mode_lib->ms.UrgentBurstFactorCursor[k];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, Calling CalculateUrgentBurstFactor\n", __func__, k);
- dml2_printf("DML::%s: k=%d, VRatio=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
- dml2_printf("DML::%s: k=%d, VRatioChroma=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%d, Calling CalculateUrgentBurstFactor\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: k=%d, VRatio=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%d, VRatioChroma=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio);
#endif
CalculateUrgentBurstFactor(
@@ -8256,23 +9136,23 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.PSCL_FACTOR,
mode_lib->ms.PSCL_FACTOR_CHROMA,
mode_lib->ms.RequiredDPPCLK,
- mode_lib->ms.SurfaceReadBandwidthLuma,
- mode_lib->ms.SurfaceReadBandwidthChroma,
+ mode_lib->ms.vactive_sw_bw_l,
+ mode_lib->ms.vactive_sw_bw_c,
mode_lib->soc.return_bus_width_bytes,
/* Output */
&mode_lib->ms.dcfclk_deepsleep);
for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
+ if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0) {
mode_lib->ms.WritebackDelayTime[k] = mode_lib->soc.qos_parameters.writeback.base_latency_us + CalculateWriteBackDelay(
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].pixel_format,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_ratio,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_ratio,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_taps,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_width,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_height,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].input_height,
display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total) / mode_lib->ms.RequiredDISPCLK;
} else {
mode_lib->ms.WritebackDelayTime[k] = 0.0;
@@ -8291,7 +9171,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, MaximumVStartup = %u\n", __func__, k, s->MaximumVStartup[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, MaximumVStartup = %u\n", __func__, k, s->MaximumVStartup[k]);
#endif
/* Immediate Flip and MALL parameters */
@@ -8340,16 +9220,15 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
(s->SubViewportMALLPStateMethod && s->FullFrameMALLPStateMethod) || s->SubViewportMALLRefreshGreaterThan120Hz;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: SubViewportMALLPStateMethod = %u\n", __func__, s->SubViewportMALLPStateMethod);
- dml2_printf("DML::%s: PhantomPipeMALLPStateMethod = %u\n", __func__, s->PhantomPipeMALLPStateMethod);
- dml2_printf("DML::%s: FullFrameMALLPStateMethod = %u\n", __func__, s->FullFrameMALLPStateMethod);
- dml2_printf("DML::%s: SubViewportMALLRefreshGreaterThan120Hz = %u\n", __func__, s->SubViewportMALLRefreshGreaterThan120Hz);
- dml2_printf("DML::%s: InvalidCombinationOfMALLUseForPState = %u\n", __func__, mode_lib->ms.support.InvalidCombinationOfMALLUseForPState);
- dml2_printf("DML::%s: in_out_params->min_clk_index = %u\n", __func__, in_out_params->min_clk_index);
- dml2_printf("DML::%s: mode_lib->ms.DCFCLK = %f\n", __func__, mode_lib->ms.DCFCLK);
- dml2_printf("DML::%s: mode_lib->ms.FabricClock = %f\n", __func__, mode_lib->ms.FabricClock);
- dml2_printf("DML::%s: mode_lib->ms.uclk_freq_mhz = %f\n", __func__, mode_lib->ms.uclk_freq_mhz);
- dml2_printf("DML::%s: urgent latency tolerance = %f\n", __func__, ((mode_lib->ip.rob_buffer_size_kbytes - mode_lib->ip.pixel_chunk_size_kbytes) * 1024 / (mode_lib->ms.DCFCLK * mode_lib->soc.return_bus_width_bytes)));
+ DML_LOG_VERBOSE("DML::%s: SubViewportMALLPStateMethod = %u\n", __func__, s->SubViewportMALLPStateMethod);
+ DML_LOG_VERBOSE("DML::%s: PhantomPipeMALLPStateMethod = %u\n", __func__, s->PhantomPipeMALLPStateMethod);
+ DML_LOG_VERBOSE("DML::%s: FullFrameMALLPStateMethod = %u\n", __func__, s->FullFrameMALLPStateMethod);
+ DML_LOG_VERBOSE("DML::%s: SubViewportMALLRefreshGreaterThan120Hz = %u\n", __func__, s->SubViewportMALLRefreshGreaterThan120Hz);
+ DML_LOG_VERBOSE("DML::%s: InvalidCombinationOfMALLUseForPState = %u\n", __func__, mode_lib->ms.support.InvalidCombinationOfMALLUseForPState);
+ DML_LOG_VERBOSE("DML::%s: in_out_params->min_clk_index = %u\n", __func__, in_out_params->min_clk_index);
+ DML_LOG_VERBOSE("DML::%s: mode_lib->ms.DCFCLK = %f\n", __func__, mode_lib->ms.DCFCLK);
+ DML_LOG_VERBOSE("DML::%s: mode_lib->ms.FabricClock = %f\n", __func__, mode_lib->ms.FabricClock);
+ DML_LOG_VERBOSE("DML::%s: mode_lib->ms.uclk_freq_mhz = %f\n", __func__, mode_lib->ms.uclk_freq_mhz);
#endif
mode_lib->ms.support.OutstandingRequestsSupport = true;
@@ -8367,6 +9246,13 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
+ mode_lib->soc.qos_parameters.qos_params.dcn4x.average_transport_distance_fclk_cycles / mode_lib->ms.FabricClock)
* (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_average_transport_latency_margin / 100.0);
+ mode_lib->ms.support.max_non_urgent_latency_us
+ = mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->ms.qos_param_index].maximum_latency_when_non_urgent_uclk_cycles
+ / mode_lib->ms.uclk_freq_mhz * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_max_latency_margin / 100.0)
+ + mode_lib->soc.qos_parameters.qos_params.dcn4x.mall_overhead_fclk_cycles / mode_lib->ms.FabricClock
+ + mode_lib->soc.qos_parameters.qos_params.dcn4x.max_round_trip_to_furthest_cs_fclk_cycles / mode_lib->ms.FabricClock
+ * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin / 100.0);
+
for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn4x) {
@@ -8382,10 +9268,10 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: avg_urgent_latency_us = %f\n", __func__, mode_lib->ms.support.avg_urgent_latency_us);
- dml2_printf("DML::%s: avg_non_urgent_latency_us = %f\n", __func__, mode_lib->ms.support.avg_non_urgent_latency_us);
- dml2_printf("DML::%s: k=%d, request_size_bytes_luma = %d\n", __func__, k, mode_lib->ms.support.request_size_bytes_luma[k]);
- dml2_printf("DML::%s: k=%d, outstanding_latency_us = %f (luma)\n", __func__, k, outstanding_latency_us);
+ DML_LOG_VERBOSE("DML::%s: avg_urgent_latency_us = %f\n", __func__, mode_lib->ms.support.avg_urgent_latency_us);
+ DML_LOG_VERBOSE("DML::%s: avg_non_urgent_latency_us = %f\n", __func__, mode_lib->ms.support.avg_non_urgent_latency_us);
+ DML_LOG_VERBOSE("DML::%s: k=%d, request_size_bytes_luma = %d\n", __func__, k, mode_lib->ms.support.request_size_bytes_luma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, outstanding_latency_us = %f (luma)\n", __func__, k, outstanding_latency_us);
#endif
}
@@ -8401,14 +9287,14 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.support.OutstandingRequestsUrgencyAvoidance = false;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, request_size_bytes_chroma = %d\n", __func__, k, mode_lib->ms.support.request_size_bytes_chroma[k]);
- dml2_printf("DML::%s: k=%d, outstanding_latency_us = %f (chroma)\n", __func__, k, outstanding_latency_us);
+ DML_LOG_VERBOSE("DML::%s: k=%d, request_size_bytes_chroma = %d\n", __func__, k, mode_lib->ms.support.request_size_bytes_chroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, outstanding_latency_us = %f (chroma)\n", __func__, k, outstanding_latency_us);
#endif
}
}
memset(calculate_mcache_setting_params, 0, sizeof(struct dml2_core_calcs_calculate_mcache_setting_params));
- if (mode_lib->soc.mall_allocated_for_dcn_mbytes == 0 || mode_lib->ip.dcn_mrq_present) {
+ if (mode_lib->soc.mcache_size_bytes == 0 || mode_lib->ip.dcn_mrq_present) {
for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
mode_lib->ms.mall_prefetch_sdp_overhead_factor[k] = 1.0;
mode_lib->ms.mall_prefetch_dram_overhead_factor[k] = 1.0;
@@ -8463,11 +9349,13 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
calculate_mcache_setting_params->num_mcaches_l = &mode_lib->ms.num_mcaches_l[k];
calculate_mcache_setting_params->mcache_row_bytes_l = &mode_lib->ms.mcache_row_bytes_l[k];
+ calculate_mcache_setting_params->mcache_row_bytes_per_channel_l = &mode_lib->ms.mcache_row_bytes_per_channel_l[k];
calculate_mcache_setting_params->mcache_offsets_l = mode_lib->ms.mcache_offsets_l[k];
calculate_mcache_setting_params->mcache_shift_granularity_l = &mode_lib->ms.mcache_shift_granularity_l[k];
calculate_mcache_setting_params->num_mcaches_c = &mode_lib->ms.num_mcaches_c[k];
calculate_mcache_setting_params->mcache_row_bytes_c = &mode_lib->ms.mcache_row_bytes_c[k];
+ calculate_mcache_setting_params->mcache_row_bytes_per_channel_c = &mode_lib->ms.mcache_row_bytes_per_channel_c[k];
calculate_mcache_setting_params->mcache_offsets_c = mode_lib->ms.mcache_offsets_c[k];
calculate_mcache_setting_params->mcache_shift_granularity_c = &mode_lib->ms.mcache_shift_granularity_c[k];
@@ -8515,8 +9403,11 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
display_cfg->hostvm_enable,
mode_lib->ms.MaxDCFCLK,
mode_lib->ms.MaxFabricClock,
+#ifdef DML_MODE_SUPPORT_USE_DPM_DRAM_BW
+ mode_lib->ms.dram_bw_mbps);
+#else
mode_lib->ms.max_dram_bw_mbps);
-
+#endif
// Average BW support check
calculate_avg_bandwidth_required(
@@ -8524,8 +9415,8 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
// input
display_cfg,
mode_lib->ms.num_active_planes,
- mode_lib->ms.SurfaceReadBandwidthLuma,
- mode_lib->ms.SurfaceReadBandwidthChroma,
+ mode_lib->ms.vactive_sw_bw_l,
+ mode_lib->ms.vactive_sw_bw_c,
mode_lib->ms.cursor_bw,
mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0,
mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1,
@@ -8543,7 +9434,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
if (mode_lib->ms.NotEnoughUrgentLatencyHiding[k]) {
mode_lib->ms.support.EnoughUrgentLatencyHidingSupport = false;
- dml2_printf("DML::%s: k=%u NotEnoughUrgentLatencyHiding set\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: k=%u NotEnoughUrgentLatencyHiding set\n", __func__, k);
}
}
@@ -8552,543 +9443,13 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
if (!mode_lib->ms.support.avg_bandwidth_support_ok[m][n] && (m == dml2_core_internal_soc_state_sys_active || mode_lib->soc.mall_allocated_for_dcn_mbytes > 0)) {
mode_lib->ms.support.AvgBandwidthSupport = false;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: avg_bandwidth_support_ok[%s][%s] not ok\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n));
+ DML_LOG_VERBOSE("DML::%s: avg_bandwidth_support_ok[%s][%s] not ok\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n));
#endif
}
}
}
- /* Prefetch Check */
- {
- mode_lib->ms.TimeCalc = 24 / mode_lib->ms.dcfclk_deepsleep;
-
- calculate_hostvm_inefficiency_factor(
- &s->HostVMInefficiencyFactor,
- &s->HostVMInefficiencyFactorPrefetch,
-
- display_cfg->gpuvm_enable,
- display_cfg->hostvm_enable,
- mode_lib->ip.remote_iommu_outstanding_translations,
- mode_lib->soc.max_outstanding_reqs,
- mode_lib->ms.support.urg_bandwidth_available_pixel_and_vm[dml2_core_internal_soc_state_sys_active],
- mode_lib->ms.support.urg_bandwidth_available_vm_only[dml2_core_internal_soc_state_sys_active]);
-
- mode_lib->ms.Total3dlutActive = 0;
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut)
- mode_lib->ms.Total3dlutActive = mode_lib->ms.Total3dlutActive + 1;
-
- // Calculate tdlut schedule related terms
- calculate_tdlut_setting_params->dispclk_mhz = mode_lib->ms.RequiredDISPCLK;
- calculate_tdlut_setting_params->setup_for_tdlut = display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut;
- calculate_tdlut_setting_params->tdlut_width_mode = display_cfg->plane_descriptors[k].tdlut.tdlut_width_mode;
- calculate_tdlut_setting_params->tdlut_addressing_mode = display_cfg->plane_descriptors[k].tdlut.tdlut_addressing_mode;
- calculate_tdlut_setting_params->cursor_buffer_size = mode_lib->ip.cursor_buffer_size;
- calculate_tdlut_setting_params->gpuvm_enable = display_cfg->gpuvm_enable;
- calculate_tdlut_setting_params->gpuvm_page_size_kbytes = display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes;
- calculate_tdlut_setting_params->tdlut_mpc_width_flag = display_cfg->plane_descriptors[k].tdlut.tdlut_mpc_width_flag;
- calculate_tdlut_setting_params->is_gfx11 = dml_get_gfx_version(display_cfg->plane_descriptors[k].surface.tiling);
-
- // output
- calculate_tdlut_setting_params->tdlut_pte_bytes_per_frame = &s->tdlut_pte_bytes_per_frame[k];
- calculate_tdlut_setting_params->tdlut_bytes_per_frame = &s->tdlut_bytes_per_frame[k];
- calculate_tdlut_setting_params->tdlut_groups_per_2row_ub = &s->tdlut_groups_per_2row_ub[k];
- calculate_tdlut_setting_params->tdlut_opt_time = &s->tdlut_opt_time[k];
- calculate_tdlut_setting_params->tdlut_drain_time = &s->tdlut_drain_time[k];
- calculate_tdlut_setting_params->tdlut_bytes_per_group = &s->tdlut_bytes_per_group[k];
-
- calculate_tdlut_setting(&mode_lib->scratch, calculate_tdlut_setting_params);
- }
-
- min_return_bw_for_latency = mode_lib->ms.support.urg_bandwidth_available_min_latency[dml2_core_internal_soc_state_sys_active];
-
- if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn3)
- s->ReorderingBytes = (unsigned int)(mode_lib->soc.clk_table.dram_config.channel_count * math_max3(mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_pixel_only_bytes,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_pixel_and_vm_bytes,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_vm_only_bytes));
-
- CalculateExtraLatency(
- display_cfg,
- mode_lib->ip.rob_buffer_size_kbytes,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.loaded_round_trip_latency_fclk_cycles,
- s->ReorderingBytes,
- mode_lib->ms.DCFCLK,
- mode_lib->ms.FabricClock,
- mode_lib->ip.pixel_chunk_size_kbytes,
- min_return_bw_for_latency,
- mode_lib->ms.num_active_planes,
- mode_lib->ms.NoOfDPP,
- mode_lib->ms.dpte_group_bytes,
- s->tdlut_bytes_per_group,
- s->HostVMInefficiencyFactor,
- s->HostVMInefficiencyFactorPrefetch,
- mode_lib->soc.hostvm_min_page_size_kbytes,
- mode_lib->soc.qos_parameters.qos_type,
- !(display_cfg->overrides.max_outstanding_when_urgent_expected_disable),
- mode_lib->soc.max_outstanding_reqs,
- mode_lib->ms.support.request_size_bytes_luma,
- mode_lib->ms.support.request_size_bytes_chroma,
- mode_lib->ip.meta_chunk_size_kbytes,
- mode_lib->ip.dchub_arb_to_ret_delay,
- mode_lib->ms.TripToMemory,
- mode_lib->ip.hostvm_mode,
-
- // output
- &mode_lib->ms.ExtraLatency,
- &mode_lib->ms.ExtraLatency_sr,
- &mode_lib->ms.ExtraLatencyPrefetch);
-
- {
- mode_lib->ms.support.PrefetchSupported = true;
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- struct dml2_core_internal_DmlPipe *myPipe = &s->myPipe;
-
- mode_lib->ms.TWait[k] = CalculateTWait(
- display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns,
- mode_lib->ms.UrgLatency,
- mode_lib->ms.TripToMemory,
- !dml_is_phantom_pipe(&display_cfg->plane_descriptors[k]) && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.drr_config.enabled ?
- get_g6_temp_read_blackout_us(&mode_lib->soc, (unsigned int)(mode_lib->ms.uclk_freq_mhz * 1000), in_out_params->min_clk_index) : 0.0);
-
- myPipe->Dppclk = mode_lib->ms.RequiredDPPCLK[k];
- myPipe->Dispclk = mode_lib->ms.RequiredDISPCLK;
- myPipe->PixelClock = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- myPipe->DCFClkDeepSleep = mode_lib->ms.dcfclk_deepsleep;
- myPipe->DPPPerSurface = mode_lib->ms.NoOfDPP[k];
- myPipe->ScalerEnabled = display_cfg->plane_descriptors[k].composition.scaler_info.enabled;
- myPipe->VRatio = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- myPipe->VRatioChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- myPipe->VTaps = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps;
- myPipe->VTapsChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps;
- myPipe->RotationAngle = display_cfg->plane_descriptors[k].composition.rotation_angle;
- myPipe->mirrored = display_cfg->plane_descriptors[k].composition.mirrored;
- myPipe->BlockWidth256BytesY = mode_lib->ms.Read256BlockWidthY[k];
- myPipe->BlockHeight256BytesY = mode_lib->ms.Read256BlockHeightY[k];
- myPipe->BlockWidth256BytesC = mode_lib->ms.Read256BlockWidthC[k];
- myPipe->BlockHeight256BytesC = mode_lib->ms.Read256BlockHeightC[k];
- myPipe->InterlaceEnable = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced;
- myPipe->NumberOfCursors = display_cfg->plane_descriptors[k].cursor.num_cursors;
- myPipe->VBlank = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active;
- myPipe->HTotal = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total;
- myPipe->HActive = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active;
- myPipe->DCCEnable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- myPipe->ODMMode = mode_lib->ms.ODMMode[k];
- myPipe->SourcePixelFormat = display_cfg->plane_descriptors[k].pixel_format;
- myPipe->BytePerPixelY = mode_lib->ms.BytePerPixelY[k];
- myPipe->BytePerPixelC = mode_lib->ms.BytePerPixelC[k];
- myPipe->ProgressiveToInterlaceUnitInOPP = mode_lib->ip.ptoi_supported;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Calling CalculatePrefetchSchedule for k=%u\n", __func__, k);
- dml2_printf("DML::%s: MaximumVStartup = %u\n", __func__, s->MaximumVStartup[k]);
-#endif
- CalculatePrefetchSchedule_params->display_cfg = display_cfg;
- CalculatePrefetchSchedule_params->HostVMInefficiencyFactor = s->HostVMInefficiencyFactorPrefetch;
- CalculatePrefetchSchedule_params->myPipe = myPipe;
- CalculatePrefetchSchedule_params->DSCDelay = mode_lib->ms.DSCDelay[k];
- CalculatePrefetchSchedule_params->DPPCLKDelaySubtotalPlusCNVCFormater = mode_lib->ip.dppclk_delay_subtotal + mode_lib->ip.dppclk_delay_cnvc_formatter;
- CalculatePrefetchSchedule_params->DPPCLKDelaySCL = mode_lib->ip.dppclk_delay_scl;
- CalculatePrefetchSchedule_params->DPPCLKDelaySCLLBOnly = mode_lib->ip.dppclk_delay_scl_lb_only;
- CalculatePrefetchSchedule_params->DPPCLKDelayCNVCCursor = mode_lib->ip.dppclk_delay_cnvc_cursor;
- CalculatePrefetchSchedule_params->DISPCLKDelaySubtotal = mode_lib->ip.dispclk_delay_subtotal;
- CalculatePrefetchSchedule_params->DPP_RECOUT_WIDTH = (unsigned int)(mode_lib->ms.SwathWidthY[k] / display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
- CalculatePrefetchSchedule_params->OutputFormat = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format;
- CalculatePrefetchSchedule_params->MaxInterDCNTileRepeaters = mode_lib->ip.max_inter_dcn_tile_repeaters;
- CalculatePrefetchSchedule_params->VStartup = s->MaximumVStartup[k];
- CalculatePrefetchSchedule_params->HostVMMinPageSize = mode_lib->soc.hostvm_min_page_size_kbytes;
- CalculatePrefetchSchedule_params->DynamicMetadataEnable = display_cfg->plane_descriptors[k].dynamic_meta_data.enable;
- CalculatePrefetchSchedule_params->DynamicMetadataVMEnabled = mode_lib->ip.dynamic_metadata_vm_enabled;
- CalculatePrefetchSchedule_params->DynamicMetadataLinesBeforeActiveRequired = display_cfg->plane_descriptors[k].dynamic_meta_data.lines_before_active_required;
- CalculatePrefetchSchedule_params->DynamicMetadataTransmittedBytes = display_cfg->plane_descriptors[k].dynamic_meta_data.transmitted_bytes;
- CalculatePrefetchSchedule_params->UrgentLatency = mode_lib->ms.UrgLatency;
- CalculatePrefetchSchedule_params->ExtraLatencyPrefetch = mode_lib->ms.ExtraLatencyPrefetch;
- CalculatePrefetchSchedule_params->TCalc = mode_lib->ms.TimeCalc;
- CalculatePrefetchSchedule_params->vm_bytes = mode_lib->ms.vm_bytes[k];
- CalculatePrefetchSchedule_params->PixelPTEBytesPerRow = mode_lib->ms.DPTEBytesPerRow[k];
- CalculatePrefetchSchedule_params->PrefetchSourceLinesY = mode_lib->ms.PrefetchLinesY[k];
- CalculatePrefetchSchedule_params->VInitPreFillY = mode_lib->ms.PrefillY[k];
- CalculatePrefetchSchedule_params->MaxNumSwathY = mode_lib->ms.MaxNumSwathY[k];
- CalculatePrefetchSchedule_params->PrefetchSourceLinesC = mode_lib->ms.PrefetchLinesC[k];
- CalculatePrefetchSchedule_params->VInitPreFillC = mode_lib->ms.PrefillC[k];
- CalculatePrefetchSchedule_params->MaxNumSwathC = mode_lib->ms.MaxNumSwathC[k];
- CalculatePrefetchSchedule_params->swath_width_luma_ub = mode_lib->ms.swath_width_luma_ub[k];
- CalculatePrefetchSchedule_params->swath_width_chroma_ub = mode_lib->ms.swath_width_chroma_ub[k];
- CalculatePrefetchSchedule_params->SwathHeightY = mode_lib->ms.SwathHeightY[k];
- CalculatePrefetchSchedule_params->SwathHeightC = mode_lib->ms.SwathHeightC[k];
- CalculatePrefetchSchedule_params->TWait = mode_lib->ms.TWait[k];
- CalculatePrefetchSchedule_params->Ttrip = mode_lib->ms.TripToMemory;
- CalculatePrefetchSchedule_params->Turg = mode_lib->ms.UrgLatency;
- CalculatePrefetchSchedule_params->setup_for_tdlut = display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut;
- CalculatePrefetchSchedule_params->tdlut_pte_bytes_per_frame = s->tdlut_pte_bytes_per_frame[k];
- CalculatePrefetchSchedule_params->tdlut_bytes_per_frame = s->tdlut_bytes_per_frame[k];
- CalculatePrefetchSchedule_params->tdlut_opt_time = s->tdlut_opt_time[k];
- CalculatePrefetchSchedule_params->tdlut_drain_time = s->tdlut_drain_time[k];
- CalculatePrefetchSchedule_params->num_cursors = (display_cfg->plane_descriptors[k].cursor.cursor_width > 0);
- CalculatePrefetchSchedule_params->cursor_bytes_per_chunk = s->cursor_bytes_per_chunk[k];
- CalculatePrefetchSchedule_params->cursor_bytes_per_line = s->cursor_bytes_per_line[k];
- CalculatePrefetchSchedule_params->dcc_enable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- CalculatePrefetchSchedule_params->mrq_present = mode_lib->ip.dcn_mrq_present;
- CalculatePrefetchSchedule_params->meta_row_bytes = mode_lib->ms.meta_row_bytes[k];
- CalculatePrefetchSchedule_params->mall_prefetch_sdp_overhead_factor = mode_lib->ms.mall_prefetch_sdp_overhead_factor[k];
-
- // output
- CalculatePrefetchSchedule_params->DSTXAfterScaler = &s->DSTXAfterScaler[k];
- CalculatePrefetchSchedule_params->DSTYAfterScaler = &s->DSTYAfterScaler[k];
- CalculatePrefetchSchedule_params->dst_y_prefetch = &mode_lib->ms.dst_y_prefetch[k];
- CalculatePrefetchSchedule_params->dst_y_per_vm_vblank = &mode_lib->ms.LinesForVM[k];
- CalculatePrefetchSchedule_params->dst_y_per_row_vblank = &mode_lib->ms.LinesForDPTERow[k];
- CalculatePrefetchSchedule_params->VRatioPrefetchY = &mode_lib->ms.VRatioPreY[k];
- CalculatePrefetchSchedule_params->VRatioPrefetchC = &mode_lib->ms.VRatioPreC[k];
- CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWLuma = &mode_lib->ms.RequiredPrefetchPixelDataBWLuma[k]; // prefetch_sw_bw_l
- CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWChroma = &mode_lib->ms.RequiredPrefetchPixelDataBWChroma[k]; // prefetch_sw_bw_c
- CalculatePrefetchSchedule_params->NotEnoughTimeForDynamicMetadata = &mode_lib->ms.NoTimeForDynamicMetadata[k];
- CalculatePrefetchSchedule_params->Tno_bw = &mode_lib->ms.Tno_bw[k];
- CalculatePrefetchSchedule_params->Tno_bw_flip = &mode_lib->ms.Tno_bw_flip[k];
- CalculatePrefetchSchedule_params->prefetch_vmrow_bw = &mode_lib->ms.prefetch_vmrow_bw[k];
- CalculatePrefetchSchedule_params->Tdmdl_vm = &s->dummy_single[0];
- CalculatePrefetchSchedule_params->Tdmdl = &s->dummy_single[1];
- CalculatePrefetchSchedule_params->TSetup = &s->dummy_single[2];
- CalculatePrefetchSchedule_params->Tvm_trips = &s->Tvm_trips[k];
- CalculatePrefetchSchedule_params->Tr0_trips = &s->Tr0_trips[k];
- CalculatePrefetchSchedule_params->Tvm_trips_flip = &s->Tvm_trips_flip[k];
- CalculatePrefetchSchedule_params->Tr0_trips_flip = &s->Tr0_trips_flip[k];
- CalculatePrefetchSchedule_params->Tvm_trips_flip_rounded = &s->Tvm_trips_flip_rounded[k];
- CalculatePrefetchSchedule_params->Tr0_trips_flip_rounded = &s->Tr0_trips_flip_rounded[k];
- CalculatePrefetchSchedule_params->VUpdateOffsetPix = &s->dummy_integer[0];
- CalculatePrefetchSchedule_params->VUpdateWidthPix = &s->dummy_integer[1];
- CalculatePrefetchSchedule_params->VReadyOffsetPix = &s->dummy_integer[2];
- CalculatePrefetchSchedule_params->prefetch_cursor_bw = &mode_lib->ms.prefetch_cursor_bw[k];
-
- mode_lib->ms.NoTimeForPrefetch[k] = CalculatePrefetchSchedule(&mode_lib->scratch, CalculatePrefetchSchedule_params);
-
- mode_lib->ms.support.PrefetchSupported &= !mode_lib->ms.NoTimeForPrefetch[k];
- dml2_printf("DML::%s: k=%d, dst_y_per_vm_vblank = %f\n", __func__, k, *CalculatePrefetchSchedule_params->dst_y_per_vm_vblank);
- dml2_printf("DML::%s: k=%d, dst_y_per_row_vblank = %f\n", __func__, k, *CalculatePrefetchSchedule_params->dst_y_per_row_vblank);
- } // for k num_planes
-
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (mode_lib->ms.dst_y_prefetch[k] < 2.0
- || mode_lib->ms.LinesForVM[k] >= 32.0
- || mode_lib->ms.LinesForDPTERow[k] >= 16.0
- || mode_lib->ms.NoTimeForPrefetch[k] == true
- || s->DSTYAfterScaler[k] > 8) {
- mode_lib->ms.support.PrefetchSupported = false;
- dml2_printf("DML::%s: k=%d, dst_y_prefetch=%f (should not be < 2)\n", __func__, k, mode_lib->ms.dst_y_prefetch[k]);
- dml2_printf("DML::%s: k=%d, LinesForVM=%f (should not be >= 32)\n", __func__, k, mode_lib->ms.LinesForVM[k]);
- dml2_printf("DML::%s: k=%d, LinesForDPTERow=%f (should not be >= 16)\n", __func__, k, mode_lib->ms.LinesForDPTERow[k]);
- dml2_printf("DML::%s: k=%d, DSTYAfterScaler=%d (should be <= 8)\n", __func__, k, s->DSTYAfterScaler[k]);
- dml2_printf("DML::%s: k=%d, NoTimeForPrefetch=%d\n", __func__, k, mode_lib->ms.NoTimeForPrefetch[k]);
- }
- }
-
- mode_lib->ms.support.DynamicMetadataSupported = true;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (mode_lib->ms.NoTimeForDynamicMetadata[k] == true) {
- mode_lib->ms.support.DynamicMetadataSupported = false;
- }
- }
-
- mode_lib->ms.support.VRatioInPrefetchSupported = true;
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (mode_lib->ms.VRatioPreY[k] > __DML2_CALCS_MAX_VRATIO_PRE__ ||
- mode_lib->ms.VRatioPreC[k] > __DML2_CALCS_MAX_VRATIO_PRE__) {
- mode_lib->ms.support.VRatioInPrefetchSupported = false;
- dml2_printf("DML::%s: k=%d VRatioPreY = %f (should be <= %f)\n", __func__, k, mode_lib->ms.VRatioPreY[k], __DML2_CALCS_MAX_VRATIO_PRE__);
- dml2_printf("DML::%s: k=%d VRatioPreC = %f (should be <= %f)\n", __func__, k, mode_lib->ms.VRatioPreC[k], __DML2_CALCS_MAX_VRATIO_PRE__);
- dml2_printf("DML::%s: VRatioInPrefetchSupported = %u\n", __func__, mode_lib->ms.support.VRatioInPrefetchSupported);
- }
- }
-
- // Only do urg vs prefetch bandwidth check, flip schedule check, power saving feature support check IF the Prefetch Schedule Check is ok
- if (mode_lib->ms.support.PrefetchSupported) {
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- double line_time_us = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- // Calculate Urgent burst factor for prefetch
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, Calling CalculateUrgentBurstFactor (for prefetch)\n", __func__, k);
- dml2_printf("DML::%s: k=%d, VRatioPreY=%f\n", __func__, k, mode_lib->ms.VRatioPreY[k]);
- dml2_printf("DML::%s: k=%d, VRatioPreC=%f\n", __func__, k, mode_lib->ms.VRatioPreC[k]);
-#endif
- CalculateUrgentBurstFactor(
- &display_cfg->plane_descriptors[k],
- mode_lib->ms.swath_width_luma_ub[k],
- mode_lib->ms.swath_width_chroma_ub[k],
- mode_lib->ms.SwathHeightY[k],
- mode_lib->ms.SwathHeightC[k],
- line_time_us,
- mode_lib->ms.UrgLatency,
- mode_lib->ms.VRatioPreY[k],
- mode_lib->ms.VRatioPreC[k],
- mode_lib->ms.BytePerPixelInDETY[k],
- mode_lib->ms.BytePerPixelInDETC[k],
- mode_lib->ms.DETBufferSizeY[k],
- mode_lib->ms.DETBufferSizeC[k],
- /* Output */
- &mode_lib->ms.UrgentBurstFactorLumaPre[k],
- &mode_lib->ms.UrgentBurstFactorChromaPre[k],
- &mode_lib->ms.NotEnoughUrgentLatencyHidingPre[k]);
- }
-
- // Calculate urgent bandwidth required, both urg and non urg peak bandwidth
- // assume flip bw is 0 at this point
- for (k = 0; k < mode_lib->ms.num_active_planes; k++)
- mode_lib->ms.final_flip_bw[k] = 0;
-
- calculate_peak_bandwidth_params->urg_vactive_bandwidth_required = mode_lib->ms.support.urg_vactive_bandwidth_required;
- calculate_peak_bandwidth_params->urg_bandwidth_required = mode_lib->ms.support.urg_bandwidth_required;
- calculate_peak_bandwidth_params->urg_bandwidth_required_qual = mode_lib->ms.support.urg_bandwidth_required_qual;
- calculate_peak_bandwidth_params->non_urg_bandwidth_required = mode_lib->ms.support.non_urg_bandwidth_required;
- calculate_peak_bandwidth_params->surface_avg_vactive_required_bw = mode_lib->ms.surface_avg_vactive_required_bw;
- calculate_peak_bandwidth_params->surface_peak_required_bw = mode_lib->ms.surface_peak_required_bw;
-
- calculate_peak_bandwidth_params->display_cfg = display_cfg;
- calculate_peak_bandwidth_params->inc_flip_bw = 0;
- calculate_peak_bandwidth_params->num_active_planes = mode_lib->ms.num_active_planes;
- calculate_peak_bandwidth_params->num_of_dpp = mode_lib->ms.NoOfDPP;
- calculate_peak_bandwidth_params->dcc_dram_bw_nom_overhead_factor_p0 = mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0;
- calculate_peak_bandwidth_params->dcc_dram_bw_nom_overhead_factor_p1 = mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1;
- calculate_peak_bandwidth_params->dcc_dram_bw_pref_overhead_factor_p0 = mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p0;
- calculate_peak_bandwidth_params->dcc_dram_bw_pref_overhead_factor_p1 = mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p1;
- calculate_peak_bandwidth_params->mall_prefetch_sdp_overhead_factor = mode_lib->ms.mall_prefetch_sdp_overhead_factor;
- calculate_peak_bandwidth_params->mall_prefetch_dram_overhead_factor = mode_lib->ms.mall_prefetch_dram_overhead_factor;
-
- calculate_peak_bandwidth_params->surface_read_bandwidth_l = mode_lib->ms.SurfaceReadBandwidthLuma;
- calculate_peak_bandwidth_params->surface_read_bandwidth_c = mode_lib->ms.SurfaceReadBandwidthChroma;
- calculate_peak_bandwidth_params->prefetch_bandwidth_l = mode_lib->ms.RequiredPrefetchPixelDataBWLuma;
- calculate_peak_bandwidth_params->prefetch_bandwidth_c = mode_lib->ms.RequiredPrefetchPixelDataBWChroma;
- calculate_peak_bandwidth_params->excess_vactive_fill_bw_l = mode_lib->ms.excess_vactive_fill_bw_l;
- calculate_peak_bandwidth_params->excess_vactive_fill_bw_c = mode_lib->ms.excess_vactive_fill_bw_c;
- calculate_peak_bandwidth_params->cursor_bw = mode_lib->ms.cursor_bw;
- calculate_peak_bandwidth_params->dpte_row_bw = mode_lib->ms.dpte_row_bw;
- calculate_peak_bandwidth_params->meta_row_bw = mode_lib->ms.meta_row_bw;
- calculate_peak_bandwidth_params->prefetch_cursor_bw = mode_lib->ms.prefetch_cursor_bw;
- calculate_peak_bandwidth_params->prefetch_vmrow_bw = mode_lib->ms.prefetch_vmrow_bw;
- calculate_peak_bandwidth_params->flip_bw = mode_lib->ms.final_flip_bw;
- calculate_peak_bandwidth_params->urgent_burst_factor_l = mode_lib->ms.UrgentBurstFactorLuma;
- calculate_peak_bandwidth_params->urgent_burst_factor_c = mode_lib->ms.UrgentBurstFactorChroma;
- calculate_peak_bandwidth_params->urgent_burst_factor_cursor = mode_lib->ms.UrgentBurstFactorCursor;
- calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_l = mode_lib->ms.UrgentBurstFactorLumaPre;
- calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_c = mode_lib->ms.UrgentBurstFactorChromaPre;
- calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_cursor = mode_lib->ms.UrgentBurstFactorCursorPre;
-
- calculate_peak_bandwidth_required(
- &mode_lib->scratch,
- calculate_peak_bandwidth_params);
-
- // Check urg peak bandwidth against available urg bw
- // check at SDP and DRAM, for all soc states (SVP prefetch an Sys Active)
- check_urgent_bandwidth_support(
- &s->dummy_single[0], // double* frac_urg_bandwidth
- &s->dummy_single[1], // double* frac_urg_bandwidth_mall
- &mode_lib->ms.support.UrgVactiveBandwidthSupport,
- &mode_lib->ms.support.PrefetchBandwidthSupported,
-
- mode_lib->soc.mall_allocated_for_dcn_mbytes,
- mode_lib->ms.support.non_urg_bandwidth_required,
- mode_lib->ms.support.urg_vactive_bandwidth_required,
- mode_lib->ms.support.urg_bandwidth_required,
- mode_lib->ms.support.urg_bandwidth_available);
-
- mode_lib->ms.support.PrefetchSupported &= mode_lib->ms.support.PrefetchBandwidthSupported;
- dml2_printf("DML::%s: PrefetchBandwidthSupported=%0d\n", __func__, mode_lib->ms.support.PrefetchBandwidthSupported);
-
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (mode_lib->ms.NotEnoughUrgentLatencyHidingPre[k]) {
- mode_lib->ms.support.PrefetchSupported = false;
- dml2_printf("DML::%s: k=%d, NotEnoughUrgentLatencyHidingPre=%d\n", __func__, k, mode_lib->ms.NotEnoughUrgentLatencyHidingPre[k]);
- }
- }
-
-
- // Both prefetch schedule and BW okay
- if (mode_lib->ms.support.PrefetchSupported == true && mode_lib->ms.support.VRatioInPrefetchSupported == true) {
- mode_lib->ms.BandwidthAvailableForImmediateFlip =
- get_bandwidth_available_for_immediate_flip(
- dml2_core_internal_soc_state_sys_active,
- mode_lib->ms.support.urg_bandwidth_required_qual, // no flip
- mode_lib->ms.support.urg_bandwidth_available);
-
- mode_lib->ms.TotImmediateFlipBytes = 0;
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (display_cfg->plane_descriptors[k].immediate_flip) {
- s->per_pipe_flip_bytes[k] = get_pipe_flip_bytes(
- s->HostVMInefficiencyFactor,
- mode_lib->ms.vm_bytes[k],
- mode_lib->ms.DPTEBytesPerRow[k],
- mode_lib->ms.meta_row_bytes[k]);
- } else {
- s->per_pipe_flip_bytes[k] = 0;
- }
- mode_lib->ms.TotImmediateFlipBytes += s->per_pipe_flip_bytes[k] * mode_lib->ms.NoOfDPP[k];
-
- }
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- CalculateFlipSchedule(
- &mode_lib->scratch,
- display_cfg->plane_descriptors[k].immediate_flip,
- 1, // use_lb_flip_bw
- s->HostVMInefficiencyFactor,
- s->Tvm_trips_flip[k],
- s->Tr0_trips_flip[k],
- s->Tvm_trips_flip_rounded[k],
- s->Tr0_trips_flip_rounded[k],
- display_cfg->gpuvm_enable,
- mode_lib->ms.vm_bytes[k],
- mode_lib->ms.DPTEBytesPerRow[k],
- mode_lib->ms.BandwidthAvailableForImmediateFlip,
- mode_lib->ms.TotImmediateFlipBytes,
- display_cfg->plane_descriptors[k].pixel_format,
- (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)),
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio,
- mode_lib->ms.Tno_bw_flip[k],
- mode_lib->ms.dpte_row_height[k],
- mode_lib->ms.dpte_row_height_chroma[k],
- mode_lib->ms.use_one_row_for_frame_flip[k],
- mode_lib->ip.max_flip_time_us,
- mode_lib->ip.max_flip_time_lines,
- s->per_pipe_flip_bytes[k],
- mode_lib->ms.meta_row_bytes[k],
- s->meta_row_height_luma[k],
- s->meta_row_height_chroma[k],
- mode_lib->ip.dcn_mrq_present && display_cfg->plane_descriptors[k].surface.dcc.enable,
-
- /* Output */
- &mode_lib->ms.dst_y_per_vm_flip[k],
- &mode_lib->ms.dst_y_per_row_flip[k],
- &mode_lib->ms.final_flip_bw[k],
- &mode_lib->ms.ImmediateFlipSupportedForPipe[k]);
- }
-
- calculate_peak_bandwidth_params->urg_vactive_bandwidth_required = s->dummy_bw;
- calculate_peak_bandwidth_params->urg_bandwidth_required = mode_lib->ms.support.urg_bandwidth_required_flip;
- calculate_peak_bandwidth_params->urg_bandwidth_required_qual = s->dummy_bw;
- calculate_peak_bandwidth_params->non_urg_bandwidth_required = mode_lib->ms.support.non_urg_bandwidth_required_flip;
- calculate_peak_bandwidth_params->surface_avg_vactive_required_bw = s->surface_dummy_bw;
- calculate_peak_bandwidth_params->surface_peak_required_bw = mode_lib->ms.surface_peak_required_bw;
-
- calculate_peak_bandwidth_params->display_cfg = display_cfg;
- calculate_peak_bandwidth_params->inc_flip_bw = 1;
- calculate_peak_bandwidth_params->num_active_planes = mode_lib->ms.num_active_planes;
- calculate_peak_bandwidth_params->num_of_dpp = mode_lib->ms.NoOfDPP;
- calculate_peak_bandwidth_params->dcc_dram_bw_nom_overhead_factor_p0 = mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0;
- calculate_peak_bandwidth_params->dcc_dram_bw_nom_overhead_factor_p1 = mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1;
- calculate_peak_bandwidth_params->dcc_dram_bw_pref_overhead_factor_p0 = mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p0;
- calculate_peak_bandwidth_params->dcc_dram_bw_pref_overhead_factor_p1 = mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p1;
- calculate_peak_bandwidth_params->mall_prefetch_sdp_overhead_factor = mode_lib->ms.mall_prefetch_sdp_overhead_factor;
- calculate_peak_bandwidth_params->mall_prefetch_dram_overhead_factor = mode_lib->ms.mall_prefetch_dram_overhead_factor;
-
- calculate_peak_bandwidth_params->surface_read_bandwidth_l = mode_lib->ms.SurfaceReadBandwidthLuma;
- calculate_peak_bandwidth_params->surface_read_bandwidth_c = mode_lib->ms.SurfaceReadBandwidthChroma;
- calculate_peak_bandwidth_params->prefetch_bandwidth_l = mode_lib->ms.RequiredPrefetchPixelDataBWLuma;
- calculate_peak_bandwidth_params->prefetch_bandwidth_c = mode_lib->ms.RequiredPrefetchPixelDataBWChroma;
- calculate_peak_bandwidth_params->excess_vactive_fill_bw_l = mode_lib->ms.excess_vactive_fill_bw_l;
- calculate_peak_bandwidth_params->excess_vactive_fill_bw_c = mode_lib->ms.excess_vactive_fill_bw_c;
- calculate_peak_bandwidth_params->cursor_bw = mode_lib->ms.cursor_bw;
- calculate_peak_bandwidth_params->dpte_row_bw = mode_lib->ms.dpte_row_bw;
- calculate_peak_bandwidth_params->meta_row_bw = mode_lib->ms.meta_row_bw;
- calculate_peak_bandwidth_params->prefetch_cursor_bw = mode_lib->ms.prefetch_cursor_bw;
- calculate_peak_bandwidth_params->prefetch_vmrow_bw = mode_lib->ms.prefetch_vmrow_bw;
- calculate_peak_bandwidth_params->flip_bw = mode_lib->ms.final_flip_bw;
- calculate_peak_bandwidth_params->urgent_burst_factor_l = mode_lib->ms.UrgentBurstFactorLuma;
- calculate_peak_bandwidth_params->urgent_burst_factor_c = mode_lib->ms.UrgentBurstFactorChroma;
- calculate_peak_bandwidth_params->urgent_burst_factor_cursor = mode_lib->ms.UrgentBurstFactorCursor;
- calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_l = mode_lib->ms.UrgentBurstFactorLumaPre;
- calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_c = mode_lib->ms.UrgentBurstFactorChromaPre;
- calculate_peak_bandwidth_params->urgent_burst_factor_prefetch_cursor = mode_lib->ms.UrgentBurstFactorCursorPre;
-
- calculate_peak_bandwidth_required(
- &mode_lib->scratch,
- calculate_peak_bandwidth_params);
-
- calculate_immediate_flip_bandwidth_support(
- &s->dummy_single[0], // double* frac_urg_bandwidth_flip
- &mode_lib->ms.support.ImmediateFlipSupport,
-
- dml2_core_internal_soc_state_sys_active,
- mode_lib->ms.support.urg_bandwidth_required_flip,
- mode_lib->ms.support.non_urg_bandwidth_required_flip,
- mode_lib->ms.support.urg_bandwidth_available);
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->plane_descriptors[k].immediate_flip == true && mode_lib->ms.ImmediateFlipSupportedForPipe[k] == false)
- mode_lib->ms.support.ImmediateFlipSupport = false;
- }
-
- } else { // if prefetch not support, assume iflip is not supported too
- mode_lib->ms.support.ImmediateFlipSupport = false;
- }
- } // prefetch schedule
- }
-
- s->mSOCParameters.UrgentLatency = mode_lib->ms.UrgLatency;
- s->mSOCParameters.ExtraLatency = mode_lib->ms.ExtraLatency;
- s->mSOCParameters.ExtraLatency_sr = mode_lib->ms.ExtraLatency_sr;
- s->mSOCParameters.WritebackLatency = mode_lib->soc.qos_parameters.writeback.base_latency_us;
- s->mSOCParameters.DRAMClockChangeLatency = mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us;
- s->mSOCParameters.FCLKChangeLatency = mode_lib->soc.power_management_parameters.fclk_change_blackout_us;
- s->mSOCParameters.SRExitTime = mode_lib->soc.power_management_parameters.stutter_exit_latency_us;
- s->mSOCParameters.SREnterPlusExitTime = mode_lib->soc.power_management_parameters.stutter_enter_plus_exit_latency_us;
- s->mSOCParameters.SRExitZ8Time = mode_lib->soc.power_management_parameters.z8_stutter_exit_latency_us;
- s->mSOCParameters.SREnterPlusExitZ8Time = mode_lib->soc.power_management_parameters.z8_stutter_enter_plus_exit_latency_us;
- s->mSOCParameters.USRRetrainingLatency = 0;
- s->mSOCParameters.SMNLatency = 0;
- s->mSOCParameters.g6_temp_read_blackout_us = get_g6_temp_read_blackout_us(&mode_lib->soc, (unsigned int)(mode_lib->ms.uclk_freq_mhz * 1000), in_out_params->min_clk_index);
- s->mSOCParameters.max_urgent_latency_us = get_max_urgent_latency_us(&mode_lib->soc.qos_parameters.qos_params.dcn4x, mode_lib->ms.uclk_freq_mhz, mode_lib->ms.FabricClock, in_out_params->min_clk_index);
- s->mSOCParameters.df_response_time_us = mode_lib->soc.qos_parameters.qos_params.dcn4x.df_qos_response_time_fclk_cycles / mode_lib->ms.FabricClock;
- s->mSOCParameters.qos_type = mode_lib->soc.qos_parameters.qos_type;
-
- CalculateWatermarks_params->display_cfg = display_cfg;
- CalculateWatermarks_params->USRRetrainingRequired = false;
- CalculateWatermarks_params->NumberOfActiveSurfaces = mode_lib->ms.num_active_planes;
- CalculateWatermarks_params->MaxLineBufferLines = mode_lib->ip.max_line_buffer_lines;
- CalculateWatermarks_params->LineBufferSize = mode_lib->ip.line_buffer_size_bits;
- CalculateWatermarks_params->WritebackInterfaceBufferSize = mode_lib->ip.writeback_interface_buffer_size_kbytes;
- CalculateWatermarks_params->DCFCLK = mode_lib->ms.DCFCLK;
- CalculateWatermarks_params->SynchronizeTimings = display_cfg->overrides.synchronize_timings;
- CalculateWatermarks_params->SynchronizeDRRDisplaysForUCLKPStateChange = display_cfg->overrides.synchronize_ddr_displays_for_uclk_pstate_change;
- CalculateWatermarks_params->dpte_group_bytes = mode_lib->ms.dpte_group_bytes;
- CalculateWatermarks_params->mmSOCParameters = s->mSOCParameters;
- CalculateWatermarks_params->WritebackChunkSize = mode_lib->ip.writeback_chunk_size_kbytes;
- CalculateWatermarks_params->SOCCLK = mode_lib->ms.SOCCLK;
- CalculateWatermarks_params->DCFClkDeepSleep = mode_lib->ms.dcfclk_deepsleep;
- CalculateWatermarks_params->DETBufferSizeY = mode_lib->ms.DETBufferSizeY;
- CalculateWatermarks_params->DETBufferSizeC = mode_lib->ms.DETBufferSizeC;
- CalculateWatermarks_params->SwathHeightY = mode_lib->ms.SwathHeightY;
- CalculateWatermarks_params->SwathHeightC = mode_lib->ms.SwathHeightC;
- CalculateWatermarks_params->SwathWidthY = mode_lib->ms.SwathWidthY;
- CalculateWatermarks_params->SwathWidthC = mode_lib->ms.SwathWidthC;
- CalculateWatermarks_params->DPPPerSurface = mode_lib->ms.NoOfDPP;
- CalculateWatermarks_params->BytePerPixelDETY = mode_lib->ms.BytePerPixelInDETY;
- CalculateWatermarks_params->BytePerPixelDETC = mode_lib->ms.BytePerPixelInDETC;
- CalculateWatermarks_params->DSTXAfterScaler = s->DSTXAfterScaler;
- CalculateWatermarks_params->DSTYAfterScaler = s->DSTYAfterScaler;
- CalculateWatermarks_params->UnboundedRequestEnabled = mode_lib->ms.UnboundedRequestEnabled;
- CalculateWatermarks_params->CompressedBufferSizeInkByte = mode_lib->ms.CompressedBufferSizeInkByte;
- CalculateWatermarks_params->meta_row_height_l = s->meta_row_height_luma;
- CalculateWatermarks_params->meta_row_height_c = s->meta_row_height_chroma;
-
- // Output
- CalculateWatermarks_params->Watermark = &mode_lib->ms.support.watermarks; // Watermarks *Watermark
- CalculateWatermarks_params->DRAMClockChangeSupport = mode_lib->ms.support.DRAMClockChangeSupport;
- CalculateWatermarks_params->global_dram_clock_change_supported = &mode_lib->ms.support.global_dram_clock_change_supported;
- CalculateWatermarks_params->MaxActiveDRAMClockChangeLatencySupported = &s->dummy_single_array[0]; // double *MaxActiveDRAMClockChangeLatencySupported[]
- CalculateWatermarks_params->SubViewportLinesNeededInMALL = mode_lib->ms.SubViewportLinesNeededInMALL; // unsigned int SubViewportLinesNeededInMALL[]
- CalculateWatermarks_params->FCLKChangeSupport = mode_lib->ms.support.FCLKChangeSupport;
- CalculateWatermarks_params->global_fclk_change_supported = &mode_lib->ms.support.global_fclk_change_supported;
- CalculateWatermarks_params->MaxActiveFCLKChangeLatencySupported = &s->dummy_single[0]; // double *MaxActiveFCLKChangeLatencySupported
- CalculateWatermarks_params->USRRetrainingSupport = &mode_lib->ms.support.USRRetrainingSupport;
- CalculateWatermarks_params->g6_temp_read_support = &mode_lib->ms.support.g6_temp_read_support;
- CalculateWatermarks_params->VActiveLatencyHidingMargin = mode_lib->ms.VActiveLatencyHidingMargin;
- CalculateWatermarks_params->VActiveLatencyHidingUs = mode_lib->ms.VActiveLatencyHidingUs;
-
- CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(&mode_lib->scratch, CalculateWatermarks_params);
-
- calculate_pstate_keepout_dst_lines(display_cfg, &mode_lib->ms.support.watermarks, s->dummy_integer_array[0]);
- }
- dml2_printf("DML::%s: Done prefetch calculation\n", __func__);
- // End of Prefetch Check
+ dml_core_ms_prefetch_check(mode_lib, display_cfg);
mode_lib->ms.support.max_urgent_latency_us = s->mSOCParameters.max_urgent_latency_us;
@@ -9116,16 +9477,16 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
s->pstate_bytes_required_c,
mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0,
mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1,
- mode_lib->ms.SurfaceReadBandwidthLuma,
- mode_lib->ms.SurfaceReadBandwidthChroma,
+ mode_lib->ms.vactive_sw_bw_l,
+ mode_lib->ms.vactive_sw_bw_c,
mode_lib->ms.surface_avg_vactive_required_bw,
mode_lib->ms.surface_peak_required_bw,
/* outputs */
mode_lib->ms.dram_change_vactive_det_fill_delay_us);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: max_urgent_latency_us = %f\n", __func__, s->mSOCParameters.max_urgent_latency_us);
- dml2_printf("DML::%s: ROBSupport = %u\n", __func__, mode_lib->ms.support.ROBSupport);
+ DML_LOG_VERBOSE("DML::%s: max_urgent_latency_us = %f\n", __func__, s->mSOCParameters.max_urgent_latency_us);
+ DML_LOG_VERBOSE("DML::%s: ROBSupport = %u\n", __func__, mode_lib->ms.support.ROBSupport);
#endif
/*Mode Support, Voltage State and SOC Configuration*/
@@ -9175,24 +9536,24 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
&& !mode_lib->ms.support.ExceededMALLSize
&& mode_lib->ms.support.g6_temp_read_support
&& ((!display_cfg->hostvm_enable && !s->ImmediateFlipRequired) || mode_lib->ms.support.ImmediateFlipSupport)) {
- dml2_printf("DML::%s: mode is supported\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: mode is supported\n", __func__);
mode_lib->ms.support.ModeSupport = true;
} else {
- dml2_printf("DML::%s: mode is NOT supported\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: mode is NOT supported\n", __func__);
mode_lib->ms.support.ModeSupport = false;
}
}
// Since now the mode_support work on 1 particular power state, so there is only 1 state idx (index 0).
- dml2_printf("DML::%s: ModeSupport = %u\n", __func__, mode_lib->ms.support.ModeSupport);
- dml2_printf("DML::%s: ImmediateFlipSupport = %u\n", __func__, mode_lib->ms.support.ImmediateFlipSupport);
+ DML_LOG_VERBOSE("DML::%s: ModeSupport = %u\n", __func__, mode_lib->ms.support.ModeSupport);
+ DML_LOG_VERBOSE("DML::%s: ImmediateFlipSupport = %u\n", __func__, mode_lib->ms.support.ImmediateFlipSupport);
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
mode_lib->ms.support.MPCCombineEnable[k] = mode_lib->ms.MPCCombine[k];
mode_lib->ms.support.DPPPerSurface[k] = mode_lib->ms.NoOfDPP[k];
}
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
+ for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
mode_lib->ms.support.ODMMode[k] = mode_lib->ms.ODMMode[k];
mode_lib->ms.support.DSCEnabled[k] = mode_lib->ms.RequiresDSC[k];
mode_lib->ms.support.FECEnabled[k] = mode_lib->ms.RequiresFEC[k];
@@ -9201,8 +9562,8 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
mode_lib->ms.support.OutputRate[k] = mode_lib->ms.OutputRate[k];
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: k=%d, ODMMode = %u\n", __func__, k, mode_lib->ms.support.ODMMode[k]);
- dml2_printf("DML::%s: k=%d, DSCEnabled = %u\n", __func__, k, mode_lib->ms.support.DSCEnabled[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, ODMMode = %u\n", __func__, k, mode_lib->ms.support.ODMMode[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, DSCEnabled = %u\n", __func__, k, mode_lib->ms.support.DSCEnabled[k]);
#endif
}
@@ -9210,7 +9571,7 @@ static bool dml_core_mode_support(struct dml2_core_calcs_mode_support_ex *in_out
if (!mode_lib->ms.support.ModeSupport)
dml2_print_mode_support_info(&mode_lib->ms.support, true);
- dml2_printf("DML::%s: --- DONE --- \n", __func__);
+ DML_LOG_VERBOSE("DML::%s: --- DONE --- \n", __func__);
#endif
return mode_lib->ms.support.ModeSupport;
@@ -9220,18 +9581,18 @@ unsigned int dml2_core_calcs_mode_support_ex(struct dml2_core_calcs_mode_support
{
unsigned int result;
- dml2_printf("DML::%s: ------------- START ----------\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: ------------- START ----------\n", __func__);
result = dml_core_mode_support(in_out_params);
if (result)
*in_out_params->out_evaluation_info = in_out_params->mode_lib->ms.support;
- dml2_printf("DML::%s: is_mode_support = %u (min_clk_index=%d)\n", __func__, result, in_out_params->min_clk_index);
+ DML_LOG_VERBOSE("DML::%s: is_mode_support = %u (min_clk_index=%d)\n", __func__, result, in_out_params->min_clk_index);
for (unsigned int k = 0; k < in_out_params->in_display_cfg->num_planes; k++)
- dml2_printf("DML::%s: plane_%d: reserved_vblank_time_ns = %u\n", __func__, k, in_out_params->in_display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns);
+ DML_LOG_VERBOSE("DML::%s: plane_%d: reserved_vblank_time_ns = %lu\n", __func__, k, in_out_params->in_display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns);
- dml2_printf("DML::%s: ------------- DONE ----------\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: ------------- DONE ----------\n", __func__);
return result;
}
@@ -9265,19 +9626,19 @@ static void CalculatePixelDeliveryTimes(
double pixel_clock_mhz = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u : HRatio = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
- dml2_printf("DML::%s: k=%u : VRatio = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
- dml2_printf("DML::%s: k=%u : HRatioChroma = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio);
- dml2_printf("DML::%s: k=%u : VRatioChroma = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio);
- dml2_printf("DML::%s: k=%u : VRatioPrefetchY = %f\n", __func__, k, VRatioPrefetchY[k]);
- dml2_printf("DML::%s: k=%u : VRatioPrefetchC = %f\n", __func__, k, VRatioPrefetchC[k]);
- dml2_printf("DML::%s: k=%u : swath_width_luma_ub = %u\n", __func__, k, swath_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u : swath_width_chroma_ub = %u\n", __func__, k, swath_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u : PSCL_THROUGHPUT = %f\n", __func__, k, PSCL_THROUGHPUT[k]);
- dml2_printf("DML::%s: k=%u : PSCL_THROUGHPUT_CHROMA = %f\n", __func__, k, PSCL_THROUGHPUT_CHROMA[k]);
- dml2_printf("DML::%s: k=%u : DPPPerSurface = %u\n", __func__, k, cfg_support_info->plane_support_info[k].dpps_used);
- dml2_printf("DML::%s: k=%u : pixel_clock_mhz = %f\n", __func__, k, pixel_clock_mhz);
- dml2_printf("DML::%s: k=%u : Dppclk = %f\n", __func__, k, Dppclk[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : HRatio = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%u : VRatio = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%u : HRatioChroma = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%u : VRatioChroma = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%u : VRatioPrefetchY = %f\n", __func__, k, VRatioPrefetchY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : VRatioPrefetchC = %f\n", __func__, k, VRatioPrefetchC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : swath_width_luma_ub = %u\n", __func__, k, swath_width_luma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : swath_width_chroma_ub = %u\n", __func__, k, swath_width_chroma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : PSCL_THROUGHPUT = %f\n", __func__, k, PSCL_THROUGHPUT[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : PSCL_THROUGHPUT_CHROMA = %f\n", __func__, k, PSCL_THROUGHPUT_CHROMA[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : DPPPerSurface = %u\n", __func__, k, cfg_support_info->plane_support_info[k].dpps_used);
+ DML_LOG_VERBOSE("DML::%s: k=%u : pixel_clock_mhz = %f\n", __func__, k, pixel_clock_mhz);
+ DML_LOG_VERBOSE("DML::%s: k=%u : Dppclk = %f\n", __func__, k, Dppclk[k]);
#endif
if (display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio <= 1) {
DisplayPipeLineDeliveryTimeLuma[k] = swath_width_luma_ub[k] * cfg_support_info->plane_support_info[k].dpps_used / display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio / pixel_clock_mhz;
@@ -9311,10 +9672,10 @@ static void CalculatePixelDeliveryTimes(
}
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u : DisplayPipeLineDeliveryTimeLuma = %f\n", __func__, k, DisplayPipeLineDeliveryTimeLuma[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeLineDeliveryTimeLumaPrefetch = %f\n", __func__, k, DisplayPipeLineDeliveryTimeLumaPrefetch[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeLineDeliveryTimeChroma = %f\n", __func__, k, DisplayPipeLineDeliveryTimeChroma[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeLineDeliveryTimeChromaPrefetch = %f\n", __func__, k, DisplayPipeLineDeliveryTimeChromaPrefetch[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : DisplayPipeLineDeliveryTimeLuma = %f\n", __func__, k, DisplayPipeLineDeliveryTimeLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : DisplayPipeLineDeliveryTimeLumaPrefetch = %f\n", __func__, k, DisplayPipeLineDeliveryTimeLumaPrefetch[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : DisplayPipeLineDeliveryTimeChroma = %f\n", __func__, k, DisplayPipeLineDeliveryTimeChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : DisplayPipeLineDeliveryTimeChromaPrefetch = %f\n", __func__, k, DisplayPipeLineDeliveryTimeChromaPrefetch[k]);
#endif
}
@@ -9330,12 +9691,12 @@ static void CalculatePixelDeliveryTimes(
DisplayPipeRequestDeliveryTimeChromaPrefetch[k] = DisplayPipeLineDeliveryTimeChromaPrefetch[k] / req_per_swath_ub_c[k];
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeLuma = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeLuma[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeLumaPrefetch = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeLumaPrefetch[k]);
- dml2_printf("DML::%s: k=%u : req_per_swath_ub_l = %d\n", __func__, k, req_per_swath_ub_l[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeChroma = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeChroma[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeChromaPrefetch = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeChromaPrefetch[k]);
- dml2_printf("DML::%s: k=%u : req_per_swath_ub_c = %d\n", __func__, k, req_per_swath_ub_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeLuma = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeLumaPrefetch = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeLumaPrefetch[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : req_per_swath_ub_l = %d\n", __func__, k, req_per_swath_ub_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeChroma = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeChromaPrefetch = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeChromaPrefetch[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u : req_per_swath_ub_c = %d\n", __func__, k, req_per_swath_ub_c[k]);
#endif
}
}
@@ -9431,14 +9792,14 @@ static void CalculateMetaAndPTETimes(struct dml2_core_shared_CalculateMetaAndPTE
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, DST_Y_PER_META_ROW_NOM_L = %f\n", __func__, k, p->DST_Y_PER_META_ROW_NOM_L[k]);
- dml2_printf("DML::%s: k=%d, DST_Y_PER_META_ROW_NOM_C = %f\n", __func__, k, p->DST_Y_PER_META_ROW_NOM_C[k]);
- dml2_printf("DML::%s: k=%d, TimePerMetaChunkNominal = %f\n", __func__, k, p->TimePerMetaChunkNominal[k]);
- dml2_printf("DML::%s: k=%d, TimePerMetaChunkVBlank = %f\n", __func__, k, p->TimePerMetaChunkVBlank[k]);
- dml2_printf("DML::%s: k=%d, TimePerMetaChunkFlip = %f\n", __func__, k, p->TimePerMetaChunkFlip[k]);
- dml2_printf("DML::%s: k=%d, TimePerChromaMetaChunkNominal = %f\n", __func__, k, p->TimePerChromaMetaChunkNominal[k]);
- dml2_printf("DML::%s: k=%d, TimePerChromaMetaChunkVBlank = %f\n", __func__, k, p->TimePerChromaMetaChunkVBlank[k]);
- dml2_printf("DML::%s: k=%d, TimePerChromaMetaChunkFlip = %f\n", __func__, k, p->TimePerChromaMetaChunkFlip[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, DST_Y_PER_META_ROW_NOM_L = %f\n", __func__, k, p->DST_Y_PER_META_ROW_NOM_L[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, DST_Y_PER_META_ROW_NOM_C = %f\n", __func__, k, p->DST_Y_PER_META_ROW_NOM_C[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, TimePerMetaChunkNominal = %f\n", __func__, k, p->TimePerMetaChunkNominal[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, TimePerMetaChunkVBlank = %f\n", __func__, k, p->TimePerMetaChunkVBlank[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, TimePerMetaChunkFlip = %f\n", __func__, k, p->TimePerMetaChunkFlip[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, TimePerChromaMetaChunkNominal = %f\n", __func__, k, p->TimePerChromaMetaChunkNominal[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, TimePerChromaMetaChunkVBlank = %f\n", __func__, k, p->TimePerChromaMetaChunkVBlank[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, TimePerChromaMetaChunkFlip = %f\n", __func__, k, p->TimePerChromaMetaChunkFlip[k]);
#endif
}
@@ -9459,7 +9820,7 @@ static void CalculateMetaAndPTETimes(struct dml2_core_shared_CalculateMetaAndPTE
else
p->time_per_tdlut_group[k] = 0;
- dml2_printf("DML::%s: k=%u, time_per_tdlut_group = %f\n", __func__, k, p->time_per_tdlut_group[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, time_per_tdlut_group = %f\n", __func__, k, p->time_per_tdlut_group[k]);
if (p->display_cfg->gpuvm_enable == true) {
if (!dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle)) {
@@ -9475,14 +9836,14 @@ static void CalculateMetaAndPTETimes(struct dml2_core_shared_CalculateMetaAndPTE
if (dpte_groups_per_row_luma_ub <= 2) {
dpte_groups_per_row_luma_ub = dpte_groups_per_row_luma_ub + 1;
}
- dml2_printf("DML::%s: k=%u, use_one_row_for_frame = %u\n", __func__, k, p->use_one_row_for_frame[k]);
- dml2_printf("DML::%s: k=%u, dpte_group_bytes = %u\n", __func__, k, p->dpte_group_bytes[k]);
- dml2_printf("DML::%s: k=%u, PTERequestSizeY = %u\n", __func__, k, p->PTERequestSizeY[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEReqWidthY = %u\n", __func__, k, p->PixelPTEReqWidthY[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEReqHeightY = %u\n", __func__, k, p->PixelPTEReqHeightY[k]);
- dml2_printf("DML::%s: k=%u, dpte_row_width_luma_ub = %u\n", __func__, k, p->dpte_row_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u, dpte_group_width_luma = %u\n", __func__, k, dpte_group_width_luma);
- dml2_printf("DML::%s: k=%u, dpte_groups_per_row_luma_ub = %u\n", __func__, k, dpte_groups_per_row_luma_ub);
+ DML_LOG_VERBOSE("DML::%s: k=%u, use_one_row_for_frame = %u\n", __func__, k, p->use_one_row_for_frame[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_group_bytes = %u\n", __func__, k, p->dpte_group_bytes[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PTERequestSizeY = %u\n", __func__, k, p->PTERequestSizeY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEReqWidthY = %u\n", __func__, k, p->PixelPTEReqWidthY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, PixelPTEReqHeightY = %u\n", __func__, k, p->PixelPTEReqHeightY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_row_width_luma_ub = %u\n", __func__, k, p->dpte_row_width_luma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_group_width_luma = %u\n", __func__, k, dpte_group_width_luma);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_groups_per_row_luma_ub = %u\n", __func__, k, dpte_groups_per_row_luma_ub);
p->time_per_pte_group_nom_luma[k] = p->DST_Y_PER_PTE_ROW_NOM_L[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_luma_ub;
p->time_per_pte_group_vblank_luma[k] = p->dst_y_per_row_vblank[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_luma_ub;
@@ -9506,9 +9867,9 @@ static void CalculateMetaAndPTETimes(struct dml2_core_shared_CalculateMetaAndPTE
if (dpte_groups_per_row_chroma_ub <= 2) {
dpte_groups_per_row_chroma_ub = dpte_groups_per_row_chroma_ub + 1;
}
- dml2_printf("DML::%s: k=%u, dpte_row_width_chroma_ub = %u\n", __func__, k, p->dpte_row_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u, dpte_group_width_chroma = %u\n", __func__, k, dpte_group_width_chroma);
- dml2_printf("DML::%s: k=%u, dpte_groups_per_row_chroma_ub = %u\n", __func__, k, dpte_groups_per_row_chroma_ub);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_row_width_chroma_ub = %u\n", __func__, k, p->dpte_row_width_chroma_ub[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_group_width_chroma = %u\n", __func__, k, dpte_group_width_chroma);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpte_groups_per_row_chroma_ub = %u\n", __func__, k, dpte_groups_per_row_chroma_ub);
p->time_per_pte_group_nom_chroma[k] = p->DST_Y_PER_PTE_ROW_NOM_C[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_chroma_ub;
p->time_per_pte_group_vblank_chroma[k] = p->dst_y_per_row_vblank[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_chroma_ub;
@@ -9523,17 +9884,17 @@ static void CalculateMetaAndPTETimes(struct dml2_core_shared_CalculateMetaAndPTE
p->time_per_pte_group_flip_chroma[k] = 0;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, dst_y_per_row_vblank = %f\n", __func__, k, p->dst_y_per_row_vblank[k]);
- dml2_printf("DML::%s: k=%u, dst_y_per_row_flip = %f\n", __func__, k, p->dst_y_per_row_flip[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dst_y_per_row_vblank = %f\n", __func__, k, p->dst_y_per_row_vblank[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dst_y_per_row_flip = %f\n", __func__, k, p->dst_y_per_row_flip[k]);
- dml2_printf("DML::%s: k=%u, DST_Y_PER_PTE_ROW_NOM_L = %f\n", __func__, k, p->DST_Y_PER_PTE_ROW_NOM_L[k]);
- dml2_printf("DML::%s: k=%u, DST_Y_PER_PTE_ROW_NOM_C = %f\n", __func__, k, p->DST_Y_PER_PTE_ROW_NOM_C[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_nom_luma = %f\n", __func__, k, p->time_per_pte_group_nom_luma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_vblank_luma = %f\n", __func__, k, p->time_per_pte_group_vblank_luma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_flip_luma = %f\n", __func__, k, p->time_per_pte_group_flip_luma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_nom_chroma = %f\n", __func__, k, p->time_per_pte_group_nom_chroma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_vblank_chroma = %f\n", __func__, k, p->time_per_pte_group_vblank_chroma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_flip_chroma = %f\n", __func__, k, p->time_per_pte_group_flip_chroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DST_Y_PER_PTE_ROW_NOM_L = %f\n", __func__, k, p->DST_Y_PER_PTE_ROW_NOM_L[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DST_Y_PER_PTE_ROW_NOM_C = %f\n", __func__, k, p->DST_Y_PER_PTE_ROW_NOM_C[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, time_per_pte_group_nom_luma = %f\n", __func__, k, p->time_per_pte_group_nom_luma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, time_per_pte_group_vblank_luma = %f\n", __func__, k, p->time_per_pte_group_vblank_luma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, time_per_pte_group_flip_luma = %f\n", __func__, k, p->time_per_pte_group_flip_luma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, time_per_pte_group_nom_chroma = %f\n", __func__, k, p->time_per_pte_group_nom_chroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, time_per_pte_group_vblank_chroma = %f\n", __func__, k, p->time_per_pte_group_vblank_chroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, time_per_pte_group_flip_chroma = %f\n", __func__, k, p->time_per_pte_group_flip_chroma[k]);
#endif
}
} // CalculateMetaAndPTETimes
@@ -9569,18 +9930,18 @@ static void CalculateVMGroupAndRequestTimes(
double line_time;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: NumberOfActiveSurfaces = %u\n", __func__, NumberOfActiveSurfaces);
+ DML_LOG_VERBOSE("DML::%s: NumberOfActiveSurfaces = %u\n", __func__, NumberOfActiveSurfaces);
#endif
for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
double pixel_clock_mhz = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
bool dcc_mrq_enable = display_cfg->plane_descriptors[k].surface.dcc.enable && mrq_present;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, dcc_mrq_enable = %u\n", __func__, k, dcc_mrq_enable);
- dml2_printf("DML::%s: k=%u, vm_group_bytes = %u\n", __func__, k, vm_group_bytes[k]);
- dml2_printf("DML::%s: k=%u, dpde0_bytes_per_frame_ub_l = %u\n", __func__, k, dpde0_bytes_per_frame_ub_l[k]);
- dml2_printf("DML::%s: k=%u, dpde0_bytes_per_frame_ub_c = %u\n", __func__, k, dpde0_bytes_per_frame_ub_c[k]);
- dml2_printf("DML::%s: k=%d, meta_pte_bytes_per_frame_ub_l = %d\n", __func__, k, meta_pte_bytes_per_frame_ub_l[k]);
- dml2_printf("DML::%s: k=%d, meta_pte_bytes_per_frame_ub_c = %d\n", __func__, k, meta_pte_bytes_per_frame_ub_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dcc_mrq_enable = %u\n", __func__, k, dcc_mrq_enable);
+ DML_LOG_VERBOSE("DML::%s: k=%u, vm_group_bytes = %u\n", __func__, k, vm_group_bytes[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpde0_bytes_per_frame_ub_l = %u\n", __func__, k, dpde0_bytes_per_frame_ub_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dpde0_bytes_per_frame_ub_c = %u\n", __func__, k, dpde0_bytes_per_frame_ub_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, meta_pte_bytes_per_frame_ub_l = %d\n", __func__, k, meta_pte_bytes_per_frame_ub_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d, meta_pte_bytes_per_frame_ub_c = %d\n", __func__, k, meta_pte_bytes_per_frame_ub_c[k]);
#endif
if (display_cfg->gpuvm_enable) {
@@ -9649,13 +10010,13 @@ static void CalculateVMGroupAndRequestTimes(
else
TimePerVMRequestFlip[k] = 0.0;
- dml2_printf("DML::%s: k=%u, dst_y_per_vm_vblank = %f\n", __func__, k, dst_y_per_vm_vblank[k]);
- dml2_printf("DML::%s: k=%u, dst_y_per_vm_flip = %f\n", __func__, k, dst_y_per_vm_flip[k]);
- dml2_printf("DML::%s: k=%u, line_time = %f\n", __func__, k, line_time);
- dml2_printf("DML::%s: k=%u, num_group_per_lower_vm_stage_pref = %f\n", __func__, k, num_group_per_lower_vm_stage_pref);
- dml2_printf("DML::%s: k=%u, num_group_per_lower_vm_stage_flip = %f\n", __func__, k, num_group_per_lower_vm_stage_flip);
- dml2_printf("DML::%s: k=%u, num_req_per_lower_vm_stage_pref = %f\n", __func__, k, num_req_per_lower_vm_stage_pref);
- dml2_printf("DML::%s: k=%u, num_req_per_lower_vm_stage_flip = %f\n", __func__, k, num_req_per_lower_vm_stage_flip);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dst_y_per_vm_vblank = %f\n", __func__, k, dst_y_per_vm_vblank[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, dst_y_per_vm_flip = %f\n", __func__, k, dst_y_per_vm_flip[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, line_time = %f\n", __func__, k, line_time);
+ DML_LOG_VERBOSE("DML::%s: k=%u, num_group_per_lower_vm_stage_pref = %d\n", __func__, k, num_group_per_lower_vm_stage_pref);
+ DML_LOG_VERBOSE("DML::%s: k=%u, num_group_per_lower_vm_stage_flip = %d\n", __func__, k, num_group_per_lower_vm_stage_flip);
+ DML_LOG_VERBOSE("DML::%s: k=%u, num_req_per_lower_vm_stage_pref = %d\n", __func__, k, num_req_per_lower_vm_stage_pref);
+ DML_LOG_VERBOSE("DML::%s: k=%u, num_req_per_lower_vm_stage_flip = %d\n", __func__, k, num_req_per_lower_vm_stage_flip);
if (display_cfg->gpuvm_max_page_table_levels > 2) {
TimePerVMGroupVBlank[k] = TimePerVMGroupVBlank[k] / 2;
@@ -9672,10 +10033,10 @@ static void CalculateVMGroupAndRequestTimes(
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, TimePerVMGroupVBlank = %f\n", __func__, k, TimePerVMGroupVBlank[k]);
- dml2_printf("DML::%s: k=%u, TimePerVMGroupFlip = %f\n", __func__, k, TimePerVMGroupFlip[k]);
- dml2_printf("DML::%s: k=%u, TimePerVMRequestVBlank = %f\n", __func__, k, TimePerVMRequestVBlank[k]);
- dml2_printf("DML::%s: k=%u, TimePerVMRequestFlip = %f\n", __func__, k, TimePerVMRequestFlip[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, TimePerVMGroupVBlank = %f\n", __func__, k, TimePerVMGroupVBlank[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, TimePerVMGroupFlip = %f\n", __func__, k, TimePerVMGroupFlip[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, TimePerVMRequestVBlank = %f\n", __func__, k, TimePerVMRequestVBlank[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, TimePerVMRequestFlip = %f\n", __func__, k, TimePerVMRequestFlip[k]);
#endif
}
}
@@ -9691,7 +10052,6 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
unsigned int SingleVTotal = 0;
bool SameTiming = true;
bool FoundCriticalSurface = false;
- double LastZ8StutterPeriod = 0;
memset(l, 0, sizeof(struct dml2_core_calcs_CalculateStutterEfficiency_locals));
@@ -9705,9 +10065,9 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
}
l->TotalCompressedReadBandwidth = l->TotalCompressedReadBandwidth + p->ReadBandwidthSurfaceLuma[k] / math_min2(p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane0, l->MaximumEffectiveCompressionLuma);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceLuma = %f\n", __func__, k, p->ReadBandwidthSurfaceLuma[k]);
- dml2_printf("DML::%s: k=%u, NetDCCRateLuma = %f\n", __func__, k, p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane0);
- dml2_printf("DML::%s: k=%u, MaximumEffectiveCompressionLuma = %f\n", __func__, k, l->MaximumEffectiveCompressionLuma);
+ DML_LOG_VERBOSE("DML::%s: k=%u, ReadBandwidthSurfaceLuma = %f\n", __func__, k, p->ReadBandwidthSurfaceLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, NetDCCRateLuma = %f\n", __func__, k, p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane0);
+ DML_LOG_VERBOSE("DML::%s: k=%u, MaximumEffectiveCompressionLuma = %f\n", __func__, k, l->MaximumEffectiveCompressionLuma);
#endif
l->TotalZeroSizeRequestReadBandwidth = l->TotalZeroSizeRequestReadBandwidth + p->ReadBandwidthSurfaceLuma[k] * p->display_cfg->plane_descriptors[k].surface.dcc.informative.fraction_of_zero_size_request_plane0;
l->TotalZeroSizeCompressedReadBandwidth = l->TotalZeroSizeCompressedReadBandwidth + p->ReadBandwidthSurfaceLuma[k] * p->display_cfg->plane_descriptors[k].surface.dcc.informative.fraction_of_zero_size_request_plane0 / l->MaximumEffectiveCompressionLuma;
@@ -9720,9 +10080,9 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
}
l->TotalCompressedReadBandwidth = l->TotalCompressedReadBandwidth + p->ReadBandwidthSurfaceChroma[k] / math_min2(p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane1, l->MaximumEffectiveCompressionChroma);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceChroma = %f\n", __func__, k, p->ReadBandwidthSurfaceChroma[k]);
- dml2_printf("DML::%s: k=%u, NetDCCRateChroma = %f\n", __func__, k, p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane1);
- dml2_printf("DML::%s: k=%u, MaximumEffectiveCompressionChroma = %f\n", __func__, k, l->MaximumEffectiveCompressionChroma);
+ DML_LOG_VERBOSE("DML::%s: k=%u, ReadBandwidthSurfaceChroma = %f\n", __func__, k, p->ReadBandwidthSurfaceChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, NetDCCRateChroma = %f\n", __func__, k, p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane1);
+ DML_LOG_VERBOSE("DML::%s: k=%u, MaximumEffectiveCompressionChroma = %f\n", __func__, k, l->MaximumEffectiveCompressionChroma);
#endif
l->TotalZeroSizeRequestReadBandwidth = l->TotalZeroSizeRequestReadBandwidth + p->ReadBandwidthSurfaceChroma[k] * p->display_cfg->plane_descriptors[k].surface.dcc.informative.fraction_of_zero_size_request_plane1;
l->TotalZeroSizeCompressedReadBandwidth = l->TotalZeroSizeCompressedReadBandwidth + p->ReadBandwidthSurfaceChroma[k] * p->display_cfg->plane_descriptors[k].surface.dcc.informative.fraction_of_zero_size_request_plane1 / l->MaximumEffectiveCompressionChroma;
@@ -9738,19 +10098,19 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
l->AverageDCCZeroSizeFraction = l->TotalZeroSizeRequestReadBandwidth / p->TotalDataReadBandwidth;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: UnboundedRequestEnabled = %u\n", __func__, p->UnboundedRequestEnabled);
- dml2_printf("DML::%s: TotalCompressedReadBandwidth = %f\n", __func__, l->TotalCompressedReadBandwidth);
- dml2_printf("DML::%s: TotalZeroSizeRequestReadBandwidth = %f\n", __func__, l->TotalZeroSizeRequestReadBandwidth);
- dml2_printf("DML::%s: TotalZeroSizeCompressedReadBandwidth = %f\n", __func__, l->TotalZeroSizeCompressedReadBandwidth);
- dml2_printf("DML::%s: MaximumEffectiveCompressionLuma = %f\n", __func__, l->MaximumEffectiveCompressionLuma);
- dml2_printf("DML::%s: MaximumEffectiveCompressionChroma = %f\n", __func__, l->MaximumEffectiveCompressionChroma);
- dml2_printf("DML::%s: AverageDCCCompressionRate = %f\n", __func__, l->AverageDCCCompressionRate);
- dml2_printf("DML::%s: AverageDCCZeroSizeFraction = %f\n", __func__, l->AverageDCCZeroSizeFraction);
+ DML_LOG_VERBOSE("DML::%s: UnboundedRequestEnabled = %u\n", __func__, p->UnboundedRequestEnabled);
+ DML_LOG_VERBOSE("DML::%s: TotalCompressedReadBandwidth = %f\n", __func__, l->TotalCompressedReadBandwidth);
+ DML_LOG_VERBOSE("DML::%s: TotalZeroSizeRequestReadBandwidth = %f\n", __func__, l->TotalZeroSizeRequestReadBandwidth);
+ DML_LOG_VERBOSE("DML::%s: TotalZeroSizeCompressedReadBandwidth = %f\n", __func__, l->TotalZeroSizeCompressedReadBandwidth);
+ DML_LOG_VERBOSE("DML::%s: MaximumEffectiveCompressionLuma = %f\n", __func__, l->MaximumEffectiveCompressionLuma);
+ DML_LOG_VERBOSE("DML::%s: MaximumEffectiveCompressionChroma = %f\n", __func__, l->MaximumEffectiveCompressionChroma);
+ DML_LOG_VERBOSE("DML::%s: AverageDCCCompressionRate = %f\n", __func__, l->AverageDCCCompressionRate);
+ DML_LOG_VERBOSE("DML::%s: AverageDCCZeroSizeFraction = %f\n", __func__, l->AverageDCCZeroSizeFraction);
- dml2_printf("DML::%s: CompbufReservedSpace64B = %u (%f kbytes)\n", __func__, p->CompbufReservedSpace64B, p->CompbufReservedSpace64B * 64 / 1024.0);
- dml2_printf("DML::%s: CompbufReservedSpaceZs = %u\n", __func__, p->CompbufReservedSpaceZs);
- dml2_printf("DML::%s: CompressedBufferSizeInkByte = %u kbytes\n", __func__, p->CompressedBufferSizeInkByte);
- dml2_printf("DML::%s: ROBBufferSizeInKByte = %u kbytes\n", __func__, p->ROBBufferSizeInKByte);
+ DML_LOG_VERBOSE("DML::%s: CompbufReservedSpace64B = %u (%f kbytes)\n", __func__, p->CompbufReservedSpace64B, p->CompbufReservedSpace64B * 64 / 1024.0);
+ DML_LOG_VERBOSE("DML::%s: CompbufReservedSpaceZs = %u\n", __func__, p->CompbufReservedSpaceZs);
+ DML_LOG_VERBOSE("DML::%s: CompressedBufferSizeInkByte = %u kbytes\n", __func__, p->CompressedBufferSizeInkByte);
+ DML_LOG_VERBOSE("DML::%s: ROBBufferSizeInKByte = %u kbytes\n", __func__, p->ROBBufferSizeInKByte);
#endif
if (l->AverageDCCZeroSizeFraction == 1) {
l->AverageZeroSizeCompressionRate = l->TotalZeroSizeRequestReadBandwidth / l->TotalZeroSizeCompressedReadBandwidth;
@@ -9767,10 +10127,10 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: min 1 = %f\n", __func__, p->CompressedBufferSizeInkByte * 1024 * l->AverageDCCCompressionRate);
- dml2_printf("DML::%s: min 2 = %f\n", __func__, p->MetaFIFOSizeInKEntries * 1024 * 64 / (l->AverageDCCZeroSizeFraction / l->AverageZeroSizeCompressionRate + 1 / l->AverageDCCCompressionRate));
- dml2_printf("DML::%s: min 3 = %d\n", __func__, (p->ROBBufferSizeInKByte * 1024 - p->CompbufReservedSpace64B * 64));
- dml2_printf("DML::%s: min 4 = %f\n", __func__, (p->ZeroSizeBufferEntries - p->CompbufReservedSpaceZs) * 64 / (l->AverageDCCZeroSizeFraction / l->AverageZeroSizeCompressionRate));
+ DML_LOG_VERBOSE("DML::%s: min 1 = %f\n", __func__, p->CompressedBufferSizeInkByte * 1024 * l->AverageDCCCompressionRate);
+ DML_LOG_VERBOSE("DML::%s: min 2 = %f\n", __func__, p->MetaFIFOSizeInKEntries * 1024 * 64 / (l->AverageDCCZeroSizeFraction / l->AverageZeroSizeCompressionRate + 1 / l->AverageDCCCompressionRate));
+ DML_LOG_VERBOSE("DML::%s: min 3 = %d\n", __func__, (p->ROBBufferSizeInKByte * 1024 - p->CompbufReservedSpace64B * 64));
+ DML_LOG_VERBOSE("DML::%s: min 4 = %f\n", __func__, (p->ZeroSizeBufferEntries - p->CompbufReservedSpaceZs) * 64 / (l->AverageDCCZeroSizeFraction / l->AverageZeroSizeCompressionRate));
#endif
} else {
l->EffectiveCompressedBufferSize = math_min2((double)p->CompressedBufferSizeInkByte * 1024 * l->AverageDCCCompressionRate,
@@ -9778,16 +10138,16 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
((double)p->ROBBufferSizeInKByte * 1024 - p->CompbufReservedSpace64B * 64) * (p->rob_alloc_compressed ? l->AverageDCCCompressionRate : 1.0);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: min 1 = %f\n", __func__, p->CompressedBufferSizeInkByte * 1024 * l->AverageDCCCompressionRate);
- dml2_printf("DML::%s: min 2 = %f\n", __func__, p->MetaFIFOSizeInKEntries * 1024 * 64 * l->AverageDCCCompressionRate);
+ DML_LOG_VERBOSE("DML::%s: min 1 = %f\n", __func__, p->CompressedBufferSizeInkByte * 1024 * l->AverageDCCCompressionRate);
+ DML_LOG_VERBOSE("DML::%s: min 2 = %f\n", __func__, p->MetaFIFOSizeInKEntries * 1024 * 64 * l->AverageDCCCompressionRate);
#endif
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: MetaFIFOSizeInKEntries = %u\n", __func__, p->MetaFIFOSizeInKEntries);
- dml2_printf("DML::%s: ZeroSizeBufferEntries = %u\n", __func__, p->ZeroSizeBufferEntries);
- dml2_printf("DML::%s: AverageZeroSizeCompressionRate = %f\n", __func__, l->AverageZeroSizeCompressionRate);
- dml2_printf("DML::%s: EffectiveCompressedBufferSize = %f (%f kbytes)\n", __func__, l->EffectiveCompressedBufferSize, l->EffectiveCompressedBufferSize / 1024.0);
+ DML_LOG_VERBOSE("DML::%s: MetaFIFOSizeInKEntries = %u\n", __func__, p->MetaFIFOSizeInKEntries);
+ DML_LOG_VERBOSE("DML::%s: ZeroSizeBufferEntries = %u\n", __func__, p->ZeroSizeBufferEntries);
+ DML_LOG_VERBOSE("DML::%s: AverageZeroSizeCompressionRate = %f\n", __func__, l->AverageZeroSizeCompressionRate);
+ DML_LOG_VERBOSE("DML::%s: EffectiveCompressedBufferSize = %f (%f kbytes)\n", __func__, l->EffectiveCompressedBufferSize, l->EffectiveCompressedBufferSize / 1024.0);
#endif
*p->StutterPeriod = 0;
@@ -9798,15 +10158,15 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
l->LinesInDETYRoundedDownToSwath = math_floor2(l->LinesInDETY, p->SwathHeightY[k]);
l->DETBufferingTimeY = l->LinesInDETYRoundedDownToSwath * ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, DETBufferSizeY = %u (%u kbytes)\n", __func__, k, p->DETBufferSizeY[k], p->DETBufferSizeY[k] / 1024);
- dml2_printf("DML::%s: k=%u, BytePerPixelDETY = %f\n", __func__, k, p->BytePerPixelDETY[k]);
- dml2_printf("DML::%s: k=%u, SwathWidthY = %u\n", __func__, k, p->SwathWidthY[k]);
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceLuma = %f\n", __func__, k, p->ReadBandwidthSurfaceLuma[k]);
- dml2_printf("DML::%s: k=%u, TotalDataReadBandwidth = %f\n", __func__, k, p->TotalDataReadBandwidth);
- dml2_printf("DML::%s: k=%u, LinesInDETY = %f\n", __func__, k, l->LinesInDETY);
- dml2_printf("DML::%s: k=%u, LinesInDETYRoundedDownToSwath = %f\n", __func__, k, l->LinesInDETYRoundedDownToSwath);
- dml2_printf("DML::%s: k=%u, VRatio = %f\n", __func__, k, p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
- dml2_printf("DML::%s: k=%u, DETBufferingTimeY = %f\n", __func__, k, l->DETBufferingTimeY);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DETBufferSizeY = %u (%u kbytes)\n", __func__, k, p->DETBufferSizeY[k], p->DETBufferSizeY[k] / 1024);
+ DML_LOG_VERBOSE("DML::%s: k=%u, BytePerPixelDETY = %f\n", __func__, k, p->BytePerPixelDETY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, SwathWidthY = %u\n", __func__, k, p->SwathWidthY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, ReadBandwidthSurfaceLuma = %f\n", __func__, k, p->ReadBandwidthSurfaceLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, TotalDataReadBandwidth = %f\n", __func__, k, p->TotalDataReadBandwidth);
+ DML_LOG_VERBOSE("DML::%s: k=%u, LinesInDETY = %f\n", __func__, k, l->LinesInDETY);
+ DML_LOG_VERBOSE("DML::%s: k=%u, LinesInDETYRoundedDownToSwath = %f\n", __func__, k, l->LinesInDETYRoundedDownToSwath);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VRatio = %f\n", __func__, k, p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DETBufferingTimeY = %f\n", __func__, k, l->DETBufferingTimeY);
#endif
if (!FoundCriticalSurface || l->DETBufferingTimeY < *p->StutterPeriod) {
@@ -9826,17 +10186,17 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
l->SinglePipeCriticalSurface = (p->DPPPerSurface[k] == 1);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, FoundCriticalSurface = %u\n", __func__, k, FoundCriticalSurface);
- dml2_printf("DML::%s: k=%u, StutterPeriod = %f\n", __func__, k, *p->StutterPeriod);
- dml2_printf("DML::%s: k=%u, MinTTUVBlankCriticalSurface = %f\n", __func__, k, l->MinTTUVBlankCriticalSurface);
- dml2_printf("DML::%s: k=%u, FrameTimeCriticalSurface= %f\n", __func__, k, l->FrameTimeCriticalSurface);
- dml2_printf("DML::%s: k=%u, VActiveTimeCriticalSurface = %f\n", __func__, k, l->VActiveTimeCriticalSurface);
- dml2_printf("DML::%s: k=%u, BytePerPixelYCriticalSurface = %u\n", __func__, k, l->BytePerPixelYCriticalSurface);
- dml2_printf("DML::%s: k=%u, SwathWidthYCriticalSurface = %f\n", __func__, k, l->SwathWidthYCriticalSurface);
- dml2_printf("DML::%s: k=%u, SwathHeightYCriticalSurface = %f\n", __func__, k, l->SwathHeightYCriticalSurface);
- dml2_printf("DML::%s: k=%u, BlockWidth256BytesYCriticalSurface = %u\n", __func__, k, l->BlockWidth256BytesYCriticalSurface);
- dml2_printf("DML::%s: k=%u, SinglePlaneCriticalSurface = %u\n", __func__, k, l->SinglePlaneCriticalSurface);
- dml2_printf("DML::%s: k=%u, SinglePipeCriticalSurface = %u\n", __func__, k, l->SinglePipeCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, FoundCriticalSurface = %u\n", __func__, k, FoundCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, StutterPeriod = %f\n", __func__, k, *p->StutterPeriod);
+ DML_LOG_VERBOSE("DML::%s: k=%u, MinTTUVBlankCriticalSurface = %f\n", __func__, k, l->MinTTUVBlankCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, FrameTimeCriticalSurface= %f\n", __func__, k, l->FrameTimeCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VActiveTimeCriticalSurface = %f\n", __func__, k, l->VActiveTimeCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, BytePerPixelYCriticalSurface = %u\n", __func__, k, l->BytePerPixelYCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, SwathWidthYCriticalSurface = %f\n", __func__, k, l->SwathWidthYCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, SwathHeightYCriticalSurface = %f\n", __func__, k, l->SwathHeightYCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, BlockWidth256BytesYCriticalSurface = %u\n", __func__, k, l->BlockWidth256BytesYCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, SinglePlaneCriticalSurface = %u\n", __func__, k, l->SinglePlaneCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: k=%u, SinglePipeCriticalSurface = %u\n", __func__, k, l->SinglePipeCriticalSurface);
#endif
}
}
@@ -9854,14 +10214,14 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer = math_min2(*p->StutterPeriod * p->TotalDataReadBandwidth, l->EffectiveCompressedBufferSize);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: AverageDCCCompressionRate = %f\n", __func__, l->AverageDCCCompressionRate);
- dml2_printf("DML::%s: StutterPeriod*TotalDataReadBandwidth = %f (%f kbytes)\n", __func__, *p->StutterPeriod * p->TotalDataReadBandwidth, (*p->StutterPeriod * p->TotalDataReadBandwidth) / 1024.0);
- dml2_printf("DML::%s: EffectiveCompressedBufferSize = %f (%f kbytes)\n", __func__, l->EffectiveCompressedBufferSize, l->EffectiveCompressedBufferSize / 1024.0);
- dml2_printf("DML::%s: PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer = %f (%f kbytes)\n", __func__, l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer, l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer / 1024);
- dml2_printf("DML::%s: ReturnBW = %f\n", __func__, p->ReturnBW);
- dml2_printf("DML::%s: TotalDataReadBandwidth = %f\n", __func__, p->TotalDataReadBandwidth);
- dml2_printf("DML::%s: TotalRowReadBandwidth = %f\n", __func__, l->TotalRowReadBandwidth);
- dml2_printf("DML::%s: DCFCLK = %f\n", __func__, p->DCFCLK);
+ DML_LOG_VERBOSE("DML::%s: AverageDCCCompressionRate = %f\n", __func__, l->AverageDCCCompressionRate);
+ DML_LOG_VERBOSE("DML::%s: StutterPeriod*TotalDataReadBandwidth = %f (%f kbytes)\n", __func__, *p->StutterPeriod * p->TotalDataReadBandwidth, (*p->StutterPeriod * p->TotalDataReadBandwidth) / 1024.0);
+ DML_LOG_VERBOSE("DML::%s: EffectiveCompressedBufferSize = %f (%f kbytes)\n", __func__, l->EffectiveCompressedBufferSize, l->EffectiveCompressedBufferSize / 1024.0);
+ DML_LOG_VERBOSE("DML::%s: PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer = %f (%f kbytes)\n", __func__, l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer, l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer / 1024);
+ DML_LOG_VERBOSE("DML::%s: ReturnBW = %f\n", __func__, p->ReturnBW);
+ DML_LOG_VERBOSE("DML::%s: TotalDataReadBandwidth = %f\n", __func__, p->TotalDataReadBandwidth);
+ DML_LOG_VERBOSE("DML::%s: TotalRowReadBandwidth = %f\n", __func__, l->TotalRowReadBandwidth);
+ DML_LOG_VERBOSE("DML::%s: DCFCLK = %f\n", __func__, p->DCFCLK);
#endif
l->StutterBurstTime = l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer
@@ -9870,10 +10230,10 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
/ math_min2(p->DCFCLK * 64, p->ReturnBW * (p->hw_debug5 ? 1 : l->AverageDCCCompressionRate)) +
*p->StutterPeriod * l->TotalRowReadBandwidth / p->ReturnBW;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Part 1 = %f\n", __func__, l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer / p->ReturnBW / (p->hw_debug5 ? 1 : l->AverageDCCCompressionRate));
- dml2_printf("DML::%s: Part 2 = %f\n", __func__, (*p->StutterPeriod * p->TotalDataReadBandwidth - l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer) / (p->DCFCLK * 64));
- dml2_printf("DML::%s: Part 3 = %f\n", __func__, *p->StutterPeriod * l->TotalRowReadBandwidth / p->ReturnBW);
- dml2_printf("DML::%s: StutterBurstTime = %f\n", __func__, l->StutterBurstTime);
+ DML_LOG_VERBOSE("DML::%s: Part 1 = %f\n", __func__, l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer / p->ReturnBW / (p->hw_debug5 ? 1 : l->AverageDCCCompressionRate));
+ DML_LOG_VERBOSE("DML::%s: Part 2 = %f\n", __func__, (*p->StutterPeriod * p->TotalDataReadBandwidth - l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer) / (p->DCFCLK * 64));
+ DML_LOG_VERBOSE("DML::%s: Part 3 = %f\n", __func__, *p->StutterPeriod * l->TotalRowReadBandwidth / p->ReturnBW);
+ DML_LOG_VERBOSE("DML::%s: StutterBurstTime = %f\n", __func__, l->StutterBurstTime);
#endif
l->TotalActiveWriteback = 0;
memset(l->stream_visited, 0, DML2_MAX_PLANES * sizeof(bool));
@@ -9882,7 +10242,7 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
if (!dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) {
if (!l->stream_visited[p->display_cfg->plane_descriptors[k].stream_index]) {
- if (p->display_cfg->stream_descriptors[k].writeback.enable)
+ if (p->display_cfg->stream_descriptors[k].writeback.active_writebacks_per_stream > 0)
l->TotalActiveWriteback = l->TotalActiveWriteback + 1;
if (TotalNumberOfActiveOTG == 0) { // first otg
@@ -9902,9 +10262,9 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
if (l->TotalActiveWriteback == 0) {
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: SRExitTime = %f\n", __func__, p->SRExitTime);
- dml2_printf("DML::%s: SRExitZ8Time = %f\n", __func__, p->SRExitZ8Time);
- dml2_printf("DML::%s: StutterPeriod = %f\n", __func__, *p->StutterPeriod);
+ DML_LOG_VERBOSE("DML::%s: SRExitTime = %f\n", __func__, p->SRExitTime);
+ DML_LOG_VERBOSE("DML::%s: SRExitZ8Time = %f\n", __func__, p->SRExitZ8Time);
+ DML_LOG_VERBOSE("DML::%s: StutterPeriod = %f\n", __func__, *p->StutterPeriod);
#endif
*p->StutterEfficiencyNotIncludingVBlank = math_max2(0., 1 - (p->SRExitTime + l->StutterBurstTime) / *p->StutterPeriod) * 100;
*p->Z8StutterEfficiencyNotIncludingVBlank = math_max2(0., 1 - (p->SRExitZ8Time + l->StutterBurstTime) / *p->StutterPeriod) * 100;
@@ -9917,11 +10277,11 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
*p->Z8NumberOfStutterBurstsPerFrame = 0;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VActiveTimeCriticalSurface = %f\n", __func__, l->VActiveTimeCriticalSurface);
- dml2_printf("DML::%s: StutterEfficiencyNotIncludingVBlank = %f\n", __func__, *p->StutterEfficiencyNotIncludingVBlank);
- dml2_printf("DML::%s: Z8StutterEfficiencyNotIncludingVBlank = %f\n", __func__, *p->Z8StutterEfficiencyNotIncludingVBlank);
- dml2_printf("DML::%s: NumberOfStutterBurstsPerFrame = %u\n", __func__, *p->NumberOfStutterBurstsPerFrame);
- dml2_printf("DML::%s: Z8NumberOfStutterBurstsPerFrame = %u\n", __func__, *p->Z8NumberOfStutterBurstsPerFrame);
+ DML_LOG_VERBOSE("DML::%s: VActiveTimeCriticalSurface = %f\n", __func__, l->VActiveTimeCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: StutterEfficiencyNotIncludingVBlank = %f\n", __func__, *p->StutterEfficiencyNotIncludingVBlank);
+ DML_LOG_VERBOSE("DML::%s: Z8StutterEfficiencyNotIncludingVBlank = %f\n", __func__, *p->Z8StutterEfficiencyNotIncludingVBlank);
+ DML_LOG_VERBOSE("DML::%s: NumberOfStutterBurstsPerFrame = %u\n", __func__, *p->NumberOfStutterBurstsPerFrame);
+ DML_LOG_VERBOSE("DML::%s: Z8NumberOfStutterBurstsPerFrame = %u\n", __func__, *p->Z8NumberOfStutterBurstsPerFrame);
#endif
if (*p->StutterEfficiencyNotIncludingVBlank > 0) {
@@ -9936,7 +10296,7 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
}
if (*p->Z8StutterEfficiencyNotIncludingVBlank > 0) {
- LastZ8StutterPeriod = l->VActiveTimeCriticalSurface - (*p->Z8NumberOfStutterBurstsPerFrame - 1) * *p->StutterPeriod;
+ //LastZ8StutterPeriod = l->VActiveTimeCriticalSurface - (*p->Z8NumberOfStutterBurstsPerFrame - 1) * *p->StutterPeriod;
if (!((p->SynchronizeTimings || TotalNumberOfActiveOTG == 1) && SameTiming)) {
*p->Z8StutterEfficiency = *p->Z8StutterEfficiencyNotIncludingVBlank;
} else {
@@ -9948,25 +10308,25 @@ static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratc
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: TotalNumberOfActiveOTG = %u\n", __func__, TotalNumberOfActiveOTG);
- dml2_printf("DML::%s: SameTiming = %u\n", __func__, SameTiming);
- dml2_printf("DML::%s: SynchronizeTimings = %u\n", __func__, p->SynchronizeTimings);
- dml2_printf("DML::%s: LastZ8StutterPeriod = %f\n", __func__, LastZ8StutterPeriod);
- dml2_printf("DML::%s: Z8StutterEnterPlusExitWatermark = %f\n", __func__, p->Z8StutterEnterPlusExitWatermark);
- dml2_printf("DML::%s: StutterBurstTime = %f\n", __func__, l->StutterBurstTime);
- dml2_printf("DML::%s: StutterPeriod = %f\n", __func__, *p->StutterPeriod);
- dml2_printf("DML::%s: StutterEfficiency = %f\n", __func__, *p->StutterEfficiency);
- dml2_printf("DML::%s: Z8StutterEfficiency = %f\n", __func__, *p->Z8StutterEfficiency);
- dml2_printf("DML::%s: StutterEfficiencyNotIncludingVBlank = %f\n", __func__, *p->StutterEfficiencyNotIncludingVBlank);
- dml2_printf("DML::%s: Z8NumberOfStutterBurstsPerFrame = %u\n", __func__, *p->Z8NumberOfStutterBurstsPerFrame);
+ DML_LOG_VERBOSE("DML::%s: TotalNumberOfActiveOTG = %u\n", __func__, TotalNumberOfActiveOTG);
+ DML_LOG_VERBOSE("DML::%s: SameTiming = %u\n", __func__, SameTiming);
+ DML_LOG_VERBOSE("DML::%s: SynchronizeTimings = %u\n", __func__, p->SynchronizeTimings);
+ DML_LOG_VERBOSE("DML::%s: LastZ8StutterPeriod = %f\n", __func__, *p->Z8StutterEfficiencyNotIncludingVBlank > 0 ? l->VActiveTimeCriticalSurface - (*p->Z8NumberOfStutterBurstsPerFrame - 1) * *p->StutterPeriod : 0);
+ DML_LOG_VERBOSE("DML::%s: Z8StutterEnterPlusExitWatermark = %f\n", __func__, p->Z8StutterEnterPlusExitWatermark);
+ DML_LOG_VERBOSE("DML::%s: StutterBurstTime = %f\n", __func__, l->StutterBurstTime);
+ DML_LOG_VERBOSE("DML::%s: StutterPeriod = %f\n", __func__, *p->StutterPeriod);
+ DML_LOG_VERBOSE("DML::%s: StutterEfficiency = %f\n", __func__, *p->StutterEfficiency);
+ DML_LOG_VERBOSE("DML::%s: Z8StutterEfficiency = %f\n", __func__, *p->Z8StutterEfficiency);
+ DML_LOG_VERBOSE("DML::%s: StutterEfficiencyNotIncludingVBlank = %f\n", __func__, *p->StutterEfficiencyNotIncludingVBlank);
+ DML_LOG_VERBOSE("DML::%s: Z8NumberOfStutterBurstsPerFrame = %u\n", __func__, *p->Z8NumberOfStutterBurstsPerFrame);
#endif
*p->DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE = !(!p->UnboundedRequestEnabled && (p->NumberOfActiveSurfaces == 1) && l->SinglePlaneCriticalSurface && l->SinglePipeCriticalSurface);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DETBufferSizeYCriticalSurface = %u\n", __func__, l->DETBufferSizeYCriticalSurface);
- dml2_printf("DML::%s: PixelChunkSizeInKByte = %u\n", __func__, p->PixelChunkSizeInKByte);
- dml2_printf("DML::%s: DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE = %u\n", __func__, *p->DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE);
+ DML_LOG_VERBOSE("DML::%s: DETBufferSizeYCriticalSurface = %u\n", __func__, l->DETBufferSizeYCriticalSurface);
+ DML_LOG_VERBOSE("DML::%s: PixelChunkSizeInKByte = %u\n", __func__, p->PixelChunkSizeInKByte);
+ DML_LOG_VERBOSE("DML::%s: DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE = %u\n", __func__, *p->DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE);
#endif
}
@@ -9984,6 +10344,7 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params *CalculateSwathAndDETConfiguration_params = &mode_lib->scratch.CalculateSwathAndDETConfiguration_params;
struct dml2_core_calcs_CalculateStutterEfficiency_params *CalculateStutterEfficiency_params = &mode_lib->scratch.CalculateStutterEfficiency_params;
struct dml2_core_calcs_CalculatePrefetchSchedule_params *CalculatePrefetchSchedule_params = &mode_lib->scratch.CalculatePrefetchSchedule_params;
+ struct dml2_core_calcs_CheckGlobalPrefetchAdmissibility_params *CheckGlobalPrefetchAdmissibility_params = &mode_lib->scratch.CheckGlobalPrefetchAdmissibility_params;
struct dml2_core_calcs_calculate_mcache_setting_params *calculate_mcache_setting_params = &mode_lib->scratch.calculate_mcache_setting_params;
struct dml2_core_calcs_calculate_tdlut_setting_params *calculate_tdlut_setting_params = &mode_lib->scratch.calculate_tdlut_setting_params;
struct dml2_core_shared_CalculateMetaAndPTETimes_params *CalculateMetaAndPTETimes_params = &mode_lib->scratch.CalculateMetaAndPTETimes_params;
@@ -9999,7 +10360,7 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
double max_uclk_mhz = 0;
double min_return_latency_in_DCFCLK_cycles = 0;
- dml2_printf("DML::%s: --- START --- \n", __func__);
+ DML_LOG_VERBOSE("DML::%s: --- START --- \n", __func__);
memset(&mode_lib->scratch, 0, sizeof(struct dml2_core_internal_scratch));
memset(&mode_lib->mp, 0, sizeof(struct dml2_core_internal_mode_program));
@@ -10021,13 +10382,13 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
for (k = 0; k < s->num_active_planes; ++k) {
unsigned int stream_index = display_cfg->plane_descriptors[k].stream_index;
- dml2_assert(cfg_support_info->stream_support_info[stream_index].odms_used <= 4);
- dml2_assert(cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 4 ||
+ DML_ASSERT(cfg_support_info->stream_support_info[stream_index].odms_used <= 4);
+ DML_ASSERT(cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 4 ||
cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 2 ||
cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 1);
if (cfg_support_info->stream_support_info[stream_index].odms_used > 1)
- dml2_assert(cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 1);
+ DML_ASSERT(cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 1);
switch (cfg_support_info->stream_support_info[stream_index].odms_used) {
case (4):
@@ -10053,57 +10414,51 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
for (k = 0; k < s->num_active_planes; ++k) {
mode_lib->mp.NoOfDPP[k] = cfg_support_info->plane_support_info[k].dpps_used;
mode_lib->mp.Dppclk[k] = programming->plane_programming[k].min_clocks.dcn4x.dppclk_khz / 1000.0;
- dml2_assert(mode_lib->mp.Dppclk[k] > 0);
+ DML_ASSERT(mode_lib->mp.Dppclk[k] > 0);
}
for (k = 0; k < s->num_active_planes; ++k) {
unsigned int stream_index = display_cfg->plane_descriptors[k].stream_index;
mode_lib->mp.DSCCLK[k] = programming->stream_programming[stream_index].min_clocks.dcn4x.dscclk_khz / 1000.0;
- dml2_printf("DML::%s: k=%d stream_index=%d, mode_lib->mp.DSCCLK = %f\n", __func__, k, stream_index, mode_lib->mp.DSCCLK[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d stream_index=%d, mode_lib->mp.DSCCLK = %f\n", __func__, k, stream_index, mode_lib->mp.DSCCLK[k]);
}
mode_lib->mp.Dispclk = programming->min_clocks.dcn4x.dispclk_khz / 1000.0;
mode_lib->mp.DCFCLKDeepSleep = programming->min_clocks.dcn4x.deepsleep_dcfclk_khz / 1000.0;
- dml2_assert(mode_lib->mp.Dcfclk > 0);
- dml2_assert(mode_lib->mp.FabricClock > 0);
- dml2_assert(mode_lib->mp.dram_bw_mbps > 0);
- dml2_assert(mode_lib->mp.uclk_freq_mhz > 0);
- dml2_assert(mode_lib->mp.GlobalDPPCLK > 0);
- dml2_assert(mode_lib->mp.Dispclk > 0);
- dml2_assert(mode_lib->mp.DCFCLKDeepSleep > 0);
- dml2_assert(s->SOCCLK > 0);
-
-#ifdef __DML_VBA_DEBUG__
- // dml2_printf_dml_display_cfg_timing(&display_cfg->timing, s->num_active_planes);
- // dml2_printf_dml_display_cfg_plane(&display_cfg->plane, s->num_active_planes);
- // dml2_printf_dml_display_cfg_surface(&display_cfg->surface, s->num_active_planes);
- // dml2_printf_dml_display_cfg_output(&display_cfg->output, s->num_active_planes);
- // dml2_printf_dml_display_cfg_hw_resource(&display_cfg->hw, s->num_active_planes);
-
- dml2_printf("DML::%s: num_active_planes = %u\n", __func__, s->num_active_planes);
- dml2_printf("DML::%s: num_active_pipes = %u\n", __func__, mode_lib->mp.num_active_pipes);
- dml2_printf("DML::%s: Dcfclk = %f\n", __func__, mode_lib->mp.Dcfclk);
- dml2_printf("DML::%s: FabricClock = %f\n", __func__, mode_lib->mp.FabricClock);
- dml2_printf("DML::%s: dram_bw_mbps = %f\n", __func__, mode_lib->mp.dram_bw_mbps);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, mode_lib->mp.uclk_freq_mhz);
- dml2_printf("DML::%s: Dispclk = %f\n", __func__, mode_lib->mp.Dispclk);
+ DML_ASSERT(mode_lib->mp.Dcfclk > 0);
+ DML_ASSERT(mode_lib->mp.FabricClock > 0);
+ DML_ASSERT(mode_lib->mp.dram_bw_mbps > 0);
+ DML_ASSERT(mode_lib->mp.uclk_freq_mhz > 0);
+ DML_ASSERT(mode_lib->mp.GlobalDPPCLK > 0);
+ DML_ASSERT(mode_lib->mp.Dispclk > 0);
+ DML_ASSERT(mode_lib->mp.DCFCLKDeepSleep > 0);
+ DML_ASSERT(s->SOCCLK > 0);
+
+#ifdef __DML_VBA_DEBUG__
+ DML_LOG_VERBOSE("DML::%s: num_active_planes = %u\n", __func__, s->num_active_planes);
+ DML_LOG_VERBOSE("DML::%s: num_active_pipes = %u\n", __func__, mode_lib->mp.num_active_pipes);
+ DML_LOG_VERBOSE("DML::%s: Dcfclk = %f\n", __func__, mode_lib->mp.Dcfclk);
+ DML_LOG_VERBOSE("DML::%s: FabricClock = %f\n", __func__, mode_lib->mp.FabricClock);
+ DML_LOG_VERBOSE("DML::%s: dram_bw_mbps = %f\n", __func__, mode_lib->mp.dram_bw_mbps);
+ DML_LOG_VERBOSE("DML::%s: uclk_freq_mhz = %f\n", __func__, mode_lib->mp.uclk_freq_mhz);
+ DML_LOG_VERBOSE("DML::%s: Dispclk = %f\n", __func__, mode_lib->mp.Dispclk);
for (k = 0; k < s->num_active_planes; ++k) {
- dml2_printf("DML::%s: Dppclk[%0d] = %f\n", __func__, k, mode_lib->mp.Dppclk[k]);
- }
- dml2_printf("DML::%s: GlobalDPPCLK = %f\n", __func__, mode_lib->mp.GlobalDPPCLK);
- dml2_printf("DML::%s: DCFCLKDeepSleep = %f\n", __func__, mode_lib->mp.DCFCLKDeepSleep);
- dml2_printf("DML::%s: SOCCLK = %f\n", __func__, s->SOCCLK);
- dml2_printf("DML::%s: min_clk_index = %0d\n", __func__, in_out_params->min_clk_index);
- dml2_printf("DML::%s: min_clk_table min_fclk_khz = %d\n", __func__, min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].min_fclk_khz);
- dml2_printf("DML::%s: min_clk_table uclk_mhz = %f\n", __func__, dram_bw_kbps_to_uclk_mhz(min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].pre_derate_dram_bw_kbps, &mode_lib->soc.clk_table.dram_config));
+ DML_LOG_VERBOSE("DML::%s: Dppclk[%0d] = %f\n", __func__, k, mode_lib->mp.Dppclk[k]);
+ }
+ DML_LOG_VERBOSE("DML::%s: GlobalDPPCLK = %f\n", __func__, mode_lib->mp.GlobalDPPCLK);
+ DML_LOG_VERBOSE("DML::%s: DCFCLKDeepSleep = %f\n", __func__, mode_lib->mp.DCFCLKDeepSleep);
+ DML_LOG_VERBOSE("DML::%s: SOCCLK = %f\n", __func__, s->SOCCLK);
+ DML_LOG_VERBOSE("DML::%s: min_clk_index = %0d\n", __func__, in_out_params->min_clk_index);
+ DML_LOG_VERBOSE("DML::%s: min_clk_table min_fclk_khz = %ld\n", __func__, min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].min_fclk_khz);
+ DML_LOG_VERBOSE("DML::%s: min_clk_table uclk_mhz = %f\n", __func__, dram_bw_kbps_to_uclk_mhz(min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].pre_derate_dram_bw_kbps, &mode_lib->soc.clk_table.dram_config));
for (k = 0; k < mode_lib->mp.num_active_pipes; ++k) {
- dml2_printf("DML::%s: pipe=%d is in plane=%d\n", __func__, k, mode_lib->mp.pipe_plane[k]);
- dml2_printf("DML::%s: Per-plane DPPPerSurface[%0d] = %d\n", __func__, k, mode_lib->mp.NoOfDPP[k]);
+ DML_LOG_VERBOSE("DML::%s: pipe=%d is in plane=%d\n", __func__, k, mode_lib->mp.pipe_plane[k]);
+ DML_LOG_VERBOSE("DML::%s: Per-plane DPPPerSurface[%0d] = %d\n", __func__, k, mode_lib->mp.NoOfDPP[k]);
}
for (k = 0; k < s->num_active_planes; k++)
- dml2_printf("DML::%s: plane_%d: reserved_vblank_time_ns = %u\n", __func__, k, display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns);
+ DML_LOG_VERBOSE("DML::%s: plane_%d: reserved_vblank_time_ns = %lu\n", __func__, k, display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns);
#endif
CalculateMaxDETAndMinCompressedBufferSize(
@@ -10198,10 +10553,10 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
for (k = 0; k < s->num_active_planes; ++k) {
mode_lib->mp.cursor_bw[k] = display_cfg->plane_descriptors[k].cursor.num_cursors * display_cfg->plane_descriptors[k].cursor.cursor_width * display_cfg->plane_descriptors[k].cursor.cursor_bpp / 8.0 /
((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000));
- mode_lib->mp.SurfaceReadBandwidthLuma[k] = mode_lib->mp.SwathWidthSingleDPPY[k] * mode_lib->mp.BytePerPixelY[k] / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- mode_lib->mp.SurfaceReadBandwidthChroma[k] = mode_lib->mp.SwathWidthSingleDPPC[k] * mode_lib->mp.BytePerPixelC[k] / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- dml2_printf("DML::%s: ReadBandwidthSurfaceLuma[%i] = %fBps\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthLuma[k]);
- dml2_printf("DML::%s: ReadBandwidthSurfaceChroma[%i] = %fBps\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthChroma[k]);
+ mode_lib->mp.vactive_sw_bw_l[k] = mode_lib->mp.SwathWidthSingleDPPY[k] * mode_lib->mp.BytePerPixelY[k] / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
+ mode_lib->mp.vactive_sw_bw_c[k] = mode_lib->mp.SwathWidthSingleDPPC[k] * mode_lib->mp.BytePerPixelC[k] / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
+ DML_LOG_VERBOSE("DML::%s: vactive_sw_bw_l[%i] = %fBps\n", __func__, k, mode_lib->mp.vactive_sw_bw_l[k]);
+ DML_LOG_VERBOSE("DML::%s: vactive_sw_bw_c[%i] = %fBps\n", __func__, k, mode_lib->mp.vactive_sw_bw_c[k]);
}
CalculateSwathAndDETConfiguration_params->display_cfg = display_cfg;
@@ -10217,8 +10572,8 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
CalculateSwathAndDETConfiguration_params->nomDETInKByte = s->NomDETInKByte;
CalculateSwathAndDETConfiguration_params->ConfigReturnBufferSegmentSizeInkByte = mode_lib->ip.config_return_buffer_segment_size_in_kbytes;
CalculateSwathAndDETConfiguration_params->CompressedBufferSegmentSizeInkByte = mode_lib->ip.compressed_buffer_segment_size_in_kbytes;
- CalculateSwathAndDETConfiguration_params->ReadBandwidthLuma = mode_lib->mp.SurfaceReadBandwidthLuma;
- CalculateSwathAndDETConfiguration_params->ReadBandwidthChroma = mode_lib->mp.SurfaceReadBandwidthChroma;
+ CalculateSwathAndDETConfiguration_params->ReadBandwidthLuma = mode_lib->mp.vactive_sw_bw_l;
+ CalculateSwathAndDETConfiguration_params->ReadBandwidthChroma = mode_lib->mp.vactive_sw_bw_c;
CalculateSwathAndDETConfiguration_params->MaximumSwathWidthLuma = s->dummy_single_array[0];
CalculateSwathAndDETConfiguration_params->MaximumSwathWidthChroma = s->dummy_single_array[1];
CalculateSwathAndDETConfiguration_params->Read256BytesBlockHeightY = mode_lib->mp.Read256BlockHeightY;
@@ -10466,11 +10821,13 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
calculate_mcache_setting_params->num_mcaches_l = &mode_lib->mp.num_mcaches_l[k];
calculate_mcache_setting_params->mcache_row_bytes_l = &mode_lib->mp.mcache_row_bytes_l[k];
+ calculate_mcache_setting_params->mcache_row_bytes_per_channel_l = &mode_lib->mp.mcache_row_bytes_per_channel_l[k];
calculate_mcache_setting_params->mcache_offsets_l = mode_lib->mp.mcache_offsets_l[k];
calculate_mcache_setting_params->mcache_shift_granularity_l = &mode_lib->mp.mcache_shift_granularity_l[k];
calculate_mcache_setting_params->num_mcaches_c = &mode_lib->mp.num_mcaches_c[k];
calculate_mcache_setting_params->mcache_row_bytes_c = &mode_lib->mp.mcache_row_bytes_c[k];
+ calculate_mcache_setting_params->mcache_row_bytes_per_channel_c = &mode_lib->mp.mcache_row_bytes_per_channel_c[k];
calculate_mcache_setting_params->mcache_offsets_c = mode_lib->mp.mcache_offsets_c[k];
calculate_mcache_setting_params->mcache_shift_granularity_c = &mode_lib->mp.mcache_shift_granularity_c[k];
@@ -10539,8 +10896,8 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
calculate_tdlut_setting_params->tdlut_groups_per_2row_ub = &s->tdlut_groups_per_2row_ub[k];
calculate_tdlut_setting_params->tdlut_opt_time = &s->tdlut_opt_time[k];
calculate_tdlut_setting_params->tdlut_drain_time = &s->tdlut_drain_time[k];
+ calculate_tdlut_setting_params->tdlut_bytes_to_deliver = &s->tdlut_bytes_to_deliver[k];
calculate_tdlut_setting_params->tdlut_bytes_per_group = &s->tdlut_bytes_per_group[k];
-
calculate_tdlut_setting(&mode_lib->scratch, calculate_tdlut_setting_params);
}
@@ -10583,17 +10940,17 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
mode_lib->mp.TCalc = 24.0 / mode_lib->mp.DCFCLKDeepSleep;
for (k = 0; k < s->num_active_planes; ++k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
+ if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0) {
mode_lib->mp.WritebackDelay[k] =
mode_lib->soc.qos_parameters.writeback.base_latency_us
+ CalculateWriteBackDelay(
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].pixel_format,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].h_ratio,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_ratio,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].v_taps,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_width,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].output_height,
+ display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.writeback_stream[0].input_height,
display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total) / mode_lib->mp.Dispclk;
} else
mode_lib->mp.WritebackDelay[k] = 0;
@@ -10678,11 +11035,26 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin);
for (k = 0; k < s->num_active_planes; ++k) {
- bool cursor_not_enough_urgent_latency_hiding = 0;
- double line_time_us = 0.0;
-
- line_time_us = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total /
+ bool cursor_not_enough_urgent_latency_hiding = false;
+ s->line_times[k] = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total /
((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
+
+ s->pixel_format[k] = display_cfg->plane_descriptors[k].pixel_format;
+
+ s->lb_source_lines_l[k] = get_num_lb_source_lines(mode_lib->ip.max_line_buffer_lines, mode_lib->ip.line_buffer_size_bits,
+ mode_lib->mp.NoOfDPP[k],
+ display_cfg->plane_descriptors[k].composition.viewport.plane0.width,
+ display_cfg->plane_descriptors[k].composition.viewport.plane0.height,
+ display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio,
+ display_cfg->plane_descriptors[k].composition.rotation_angle);
+
+ s->lb_source_lines_c[k] = get_num_lb_source_lines(mode_lib->ip.max_line_buffer_lines, mode_lib->ip.line_buffer_size_bits,
+ mode_lib->mp.NoOfDPP[k],
+ display_cfg->plane_descriptors[k].composition.viewport.plane1.width,
+ display_cfg->plane_descriptors[k].composition.viewport.plane1.height,
+ display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio,
+ display_cfg->plane_descriptors[k].composition.rotation_angle);
+
if (display_cfg->plane_descriptors[k].cursor.num_cursors > 0) {
calculate_cursor_req_attributes(
display_cfg->plane_descriptors[k].cursor.cursor_width,
@@ -10699,7 +11071,7 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
display_cfg->plane_descriptors[k].cursor.cursor_width,
s->cursor_bytes_per_chunk[k],
s->cursor_lines_per_chunk[k],
- line_time_us,
+ s->line_times[k],
mode_lib->mp.UrgentLatency,
// output
@@ -10714,7 +11086,7 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
mode_lib->mp.swath_width_chroma_ub[k],
mode_lib->mp.SwathHeightY[k],
mode_lib->mp.SwathHeightC[k],
- line_time_us,
+ s->line_times[k],
mode_lib->mp.UrgentLatency,
display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio,
display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio,
@@ -10739,8 +11111,8 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
mode_lib->mp.WritebackDelay[k]);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
- dml2_printf("DML::%s: k=%u WritebackDelay = %f\n", __func__, k, mode_lib->mp.WritebackDelay[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u WritebackDelay = %f\n", __func__, k, mode_lib->mp.WritebackDelay[k]);
#endif
}
@@ -10749,25 +11121,54 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
s->immediate_flip_required = s->immediate_flip_required || display_cfg->plane_descriptors[k].immediate_flip;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: immediate_flip_required = %u\n", __func__, s->immediate_flip_required);
-#endif
+ DML_LOG_VERBOSE("DML::%s: immediate_flip_required = %u\n", __func__, s->immediate_flip_required);
+#endif
+
+ if (s->num_active_planes > 1) {
+ CheckGlobalPrefetchAdmissibility_params->num_active_planes = s->num_active_planes;
+ CheckGlobalPrefetchAdmissibility_params->pixel_format = s->pixel_format;
+ CheckGlobalPrefetchAdmissibility_params->chunk_bytes_l = mode_lib->ip.pixel_chunk_size_kbytes * 1024;
+ CheckGlobalPrefetchAdmissibility_params->chunk_bytes_c = mode_lib->ip.pixel_chunk_size_kbytes * 1024;
+ CheckGlobalPrefetchAdmissibility_params->lb_source_lines_l = s->lb_source_lines_l;
+ CheckGlobalPrefetchAdmissibility_params->lb_source_lines_c = s->lb_source_lines_c;
+ CheckGlobalPrefetchAdmissibility_params->swath_height_l = mode_lib->mp.SwathHeightY;
+ CheckGlobalPrefetchAdmissibility_params->swath_height_c = mode_lib->mp.SwathHeightC;
+ CheckGlobalPrefetchAdmissibility_params->rob_buffer_size_kbytes = mode_lib->ip.rob_buffer_size_kbytes;
+ CheckGlobalPrefetchAdmissibility_params->compressed_buffer_size_kbytes = mode_lib->mp.CompressedBufferSizeInkByte;
+ CheckGlobalPrefetchAdmissibility_params->detile_buffer_size_bytes_l = mode_lib->mp.DETBufferSizeY;
+ CheckGlobalPrefetchAdmissibility_params->detile_buffer_size_bytes_c = mode_lib->mp.DETBufferSizeC;
+ CheckGlobalPrefetchAdmissibility_params->full_swath_bytes_l = s->full_swath_bytes_l;
+ CheckGlobalPrefetchAdmissibility_params->full_swath_bytes_c = s->full_swath_bytes_c;
+ CheckGlobalPrefetchAdmissibility_params->prefetch_sw_bytes = s->prefetch_sw_bytes;
+ CheckGlobalPrefetchAdmissibility_params->Tpre_rounded = 0; // don't care
+ CheckGlobalPrefetchAdmissibility_params->Tpre_oto = 0; // don't care
+ CheckGlobalPrefetchAdmissibility_params->estimated_urg_bandwidth_required_mbps = mode_lib->mp.urg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
+ CheckGlobalPrefetchAdmissibility_params->estimated_dcfclk_mhz = mode_lib->mp.Dcfclk;
+ CheckGlobalPrefetchAdmissibility_params->line_time = s->line_times;
+ CheckGlobalPrefetchAdmissibility_params->dst_y_prefetch = mode_lib->mp.dst_y_prefetch;
+
+ // if recalc_prefetch_schedule is set, recalculate the prefetch schedule with the new impacted_Tpre, prefetch should be possible
+ CheckGlobalPrefetchAdmissibility_params->recalc_prefetch_schedule = &s->dummy_boolean[0];
+ CheckGlobalPrefetchAdmissibility_params->impacted_dst_y_pre = s->impacted_dst_y_pre;
+ CheckGlobalPrefetchAdmissibility(&mode_lib->scratch, CheckGlobalPrefetchAdmissibility_params); // dont care about the check output for mode programming
+ }
{
s->DestinationLineTimesForPrefetchLessThan2 = false;
s->VRatioPrefetchMoreThanMax = false;
- dml2_printf("DML::%s: Start one iteration of prefetch schedule evaluation\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: Start one iteration of prefetch schedule evaluation\n", __func__);
for (k = 0; k < s->num_active_planes; ++k) {
struct dml2_core_internal_DmlPipe *myPipe = &s->myPipe;
- dml2_printf("DML::%s: k=%d MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
mode_lib->mp.TWait[k] = CalculateTWait(
- display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns,
- mode_lib->mp.UrgentLatency,
- mode_lib->mp.TripToMemory,
- !dml_is_phantom_pipe(&display_cfg->plane_descriptors[k]) && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.drr_config.enabled ?
- get_g6_temp_read_blackout_us(&mode_lib->soc, (unsigned int)(mode_lib->mp.uclk_freq_mhz * 1000), in_out_params->min_clk_index) : 0.0);
+ display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns,
+ mode_lib->mp.UrgentLatency,
+ mode_lib->mp.TripToMemory,
+ !dml_is_phantom_pipe(&display_cfg->plane_descriptors[k]) && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.drr_config.enabled ?
+ get_g6_temp_read_blackout_us(&mode_lib->soc, (unsigned int)(mode_lib->mp.uclk_freq_mhz * 1000), in_out_params->min_clk_index) : 0.0);
myPipe->Dppclk = mode_lib->mp.Dppclk[k];
myPipe->Dispclk = mode_lib->mp.Dispclk;
@@ -10798,7 +11199,7 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
myPipe->ProgressiveToInterlaceUnitInOPP = mode_lib->ip.ptoi_supported;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Calling CalculatePrefetchSchedule for k=%u\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: Calling CalculatePrefetchSchedule for k=%u\n", __func__, k);
#endif
CalculatePrefetchSchedule_params->display_cfg = display_cfg;
CalculatePrefetchSchedule_params->HostVMInefficiencyFactor = s->HostVMInefficiencyFactorPrefetch;
@@ -10848,6 +11249,9 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
CalculatePrefetchSchedule_params->mrq_present = mode_lib->ip.dcn_mrq_present;
CalculatePrefetchSchedule_params->meta_row_bytes = mode_lib->mp.meta_row_bytes[k];
CalculatePrefetchSchedule_params->mall_prefetch_sdp_overhead_factor = mode_lib->mp.mall_prefetch_sdp_overhead_factor[k];
+ CalculatePrefetchSchedule_params->impacted_dst_y_pre = s->impacted_dst_y_pre[k];
+ CalculatePrefetchSchedule_params->vactive_sw_bw_l = mode_lib->mp.vactive_sw_bw_l[k];
+ CalculatePrefetchSchedule_params->vactive_sw_bw_c = mode_lib->mp.vactive_sw_bw_c[k];
// output
CalculatePrefetchSchedule_params->DSTXAfterScaler = &mode_lib->mp.DSTXAfterScaler[k];
@@ -10859,6 +11263,7 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
CalculatePrefetchSchedule_params->VRatioPrefetchC = &mode_lib->mp.VRatioPrefetchC[k];
CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWLuma = &mode_lib->mp.RequiredPrefetchPixelDataBWLuma[k];
CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWChroma = &mode_lib->mp.RequiredPrefetchPixelDataBWChroma[k];
+ CalculatePrefetchSchedule_params->RequiredPrefetchBWMax = &s->dummy_single_array[0][k];
CalculatePrefetchSchedule_params->NotEnoughTimeForDynamicMetadata = &mode_lib->mp.NotEnoughTimeForDynamicMetadata[k];
CalculatePrefetchSchedule_params->Tno_bw = &mode_lib->mp.Tno_bw[k];
CalculatePrefetchSchedule_params->Tno_bw_flip = &mode_lib->mp.Tno_bw_flip[k];
@@ -10876,11 +11281,20 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
CalculatePrefetchSchedule_params->VUpdateWidthPix = &mode_lib->mp.VUpdateWidthPix[k];
CalculatePrefetchSchedule_params->VReadyOffsetPix = &mode_lib->mp.VReadyOffsetPix[k];
CalculatePrefetchSchedule_params->prefetch_cursor_bw = &mode_lib->mp.prefetch_cursor_bw[k];
+ CalculatePrefetchSchedule_params->prefetch_sw_bytes = &s->prefetch_sw_bytes[k];
+ CalculatePrefetchSchedule_params->Tpre_rounded = &s->Tpre_rounded[k];
+ CalculatePrefetchSchedule_params->Tpre_oto = &s->Tpre_oto[k];
+ CalculatePrefetchSchedule_params->prefetch_swath_time_us = &s->dummy_single[0];
mode_lib->mp.NoTimeToPrefetch[k] = CalculatePrefetchSchedule(&mode_lib->scratch, CalculatePrefetchSchedule_params);
+ if (s->impacted_dst_y_pre[k] > 0)
+ mode_lib->mp.impacted_prefetch_margin_us[k] = (mode_lib->mp.dst_y_prefetch[k] - s->impacted_dst_y_pre[k]) * s->line_times[k];
+ else
+ mode_lib->mp.impacted_prefetch_margin_us[k] = 0;
+
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%0u NoTimeToPrefetch=%0d\n", __func__, k, mode_lib->mp.NoTimeToPrefetch[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u NoTimeToPrefetch=%0d\n", __func__, k, mode_lib->mp.NoTimeToPrefetch[k]);
#endif
mode_lib->mp.VStartupMin[k] = s->MaxVStartupLines[k];
} // for k
@@ -10890,9 +11304,9 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
if (mode_lib->mp.NoTimeToPrefetch[k] == true ||
mode_lib->mp.NotEnoughTimeForDynamicMetadata[k] ||
mode_lib->mp.DSTYAfterScaler[k] > 8) {
- dml2_printf("DML::%s: k=%u, NoTimeToPrefetch = %0d\n", __func__, k, mode_lib->mp.NoTimeToPrefetch[k]);
- dml2_printf("DML::%s: k=%u, NotEnoughTimeForDynamicMetadata=%u\n", __func__, k, mode_lib->mp.NotEnoughTimeForDynamicMetadata[k]);
- dml2_printf("DML::%s: k=%u, DSTYAfterScaler=%u (should be <= 0)\n", __func__, k, mode_lib->mp.DSTYAfterScaler[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, NoTimeToPrefetch = %0d\n", __func__, k, mode_lib->mp.NoTimeToPrefetch[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, NotEnoughTimeForDynamicMetadata=%u\n", __func__, k, mode_lib->mp.NotEnoughTimeForDynamicMetadata[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, DSTYAfterScaler=%u (should be <= 0)\n", __func__, k, mode_lib->mp.DSTYAfterScaler[k]);
mode_lib->mp.PrefetchModeSupported = false;
}
if (mode_lib->mp.dst_y_prefetch[k] < 2)
@@ -10901,24 +11315,24 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
if (mode_lib->mp.VRatioPrefetchY[k] > __DML2_CALCS_MAX_VRATIO_PRE__ ||
mode_lib->mp.VRatioPrefetchC[k] > __DML2_CALCS_MAX_VRATIO_PRE__) {
s->VRatioPrefetchMoreThanMax = true;
- dml2_printf("DML::%s: k=%d, VRatioPrefetchY=%f (should not be < %f)\n", __func__, k, mode_lib->mp.VRatioPrefetchY[k], __DML2_CALCS_MAX_VRATIO_PRE__);
- dml2_printf("DML::%s: k=%d, VRatioPrefetchC=%f (should not be < %f)\n", __func__, k, mode_lib->mp.VRatioPrefetchC[k], __DML2_CALCS_MAX_VRATIO_PRE__);
- dml2_printf("DML::%s: VRatioPrefetchMoreThanMax = %u\n", __func__, s->VRatioPrefetchMoreThanMax);
+ DML_LOG_VERBOSE("DML::%s: k=%d, VRatioPrefetchY=%f (should not be < %f)\n", __func__, k, mode_lib->mp.VRatioPrefetchY[k], __DML2_CALCS_MAX_VRATIO_PRE__);
+ DML_LOG_VERBOSE("DML::%s: k=%d, VRatioPrefetchC=%f (should not be < %f)\n", __func__, k, mode_lib->mp.VRatioPrefetchC[k], __DML2_CALCS_MAX_VRATIO_PRE__);
+ DML_LOG_VERBOSE("DML::%s: VRatioPrefetchMoreThanMax = %u\n", __func__, s->VRatioPrefetchMoreThanMax);
}
if (mode_lib->mp.NotEnoughUrgentLatencyHiding[k]) {
- dml2_printf("DML::%s: k=%u, NotEnoughUrgentLatencyHiding = %u\n", __func__, k, mode_lib->mp.NotEnoughUrgentLatencyHiding[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, NotEnoughUrgentLatencyHiding = %u\n", __func__, k, mode_lib->mp.NotEnoughUrgentLatencyHiding[k]);
mode_lib->mp.PrefetchModeSupported = false;
}
}
if (s->VRatioPrefetchMoreThanMax == true || s->DestinationLineTimesForPrefetchLessThan2 == true) {
- dml2_printf("DML::%s: VRatioPrefetchMoreThanMax = %u\n", __func__, s->VRatioPrefetchMoreThanMax);
- dml2_printf("DML::%s: DestinationLineTimesForPrefetchLessThan2 = %u\n", __func__, s->DestinationLineTimesForPrefetchLessThan2);
+ DML_LOG_VERBOSE("DML::%s: VRatioPrefetchMoreThanMax = %u\n", __func__, s->VRatioPrefetchMoreThanMax);
+ DML_LOG_VERBOSE("DML::%s: DestinationLineTimesForPrefetchLessThan2 = %u\n", __func__, s->DestinationLineTimesForPrefetchLessThan2);
mode_lib->mp.PrefetchModeSupported = false;
}
- dml2_printf("DML::%s: Prefetch schedule is %sOK at vstartup = %u\n", __func__,
+ DML_LOG_VERBOSE("DML::%s: Prefetch schedule is %sOK at vstartup = %u\n", __func__,
mode_lib->mp.PrefetchModeSupported ? "" : "NOT ", CalculatePrefetchSchedule_params->VStartup);
// Prefetch schedule OK, now check prefetch bw
@@ -10946,24 +11360,24 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
&mode_lib->mp.NotEnoughUrgentLatencyHidingPre[k]);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%0u DPPPerSurface=%u\n", __func__, k, mode_lib->mp.NoOfDPP[k]);
- dml2_printf("DML::%s: k=%0u UrgentBurstFactorLuma=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorLuma[k]);
- dml2_printf("DML::%s: k=%0u UrgentBurstFactorChroma=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorChroma[k]);
- dml2_printf("DML::%s: k=%0u UrgentBurstFactorLumaPre=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorLumaPre[k]);
- dml2_printf("DML::%s: k=%0u UrgentBurstFactorChromaPre=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorChromaPre[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u DPPPerSurface=%u\n", __func__, k, mode_lib->mp.NoOfDPP[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u UrgentBurstFactorLuma=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u UrgentBurstFactorChroma=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u UrgentBurstFactorLumaPre=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorLumaPre[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u UrgentBurstFactorChromaPre=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorChromaPre[k]);
- dml2_printf("DML::%s: k=%0u VRatioPrefetchY=%f\n", __func__, k, mode_lib->mp.VRatioPrefetchY[k]);
- dml2_printf("DML::%s: k=%0u VRatioY=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
+ DML_LOG_VERBOSE("DML::%s: k=%0u VRatioPrefetchY=%f\n", __func__, k, mode_lib->mp.VRatioPrefetchY[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u VRatioY=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
- dml2_printf("DML::%s: k=%0u prefetch_vmrow_bw=%f\n", __func__, k, mode_lib->mp.prefetch_vmrow_bw[k]);
- dml2_printf("DML::%s: k=%0u ReadBandwidthSurfaceLuma=%f\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%0u ReadBandwidthSurfaceChroma=%f\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthChroma[k]);
- dml2_printf("DML::%s: k=%0u cursor_bw=%f\n", __func__, k, mode_lib->mp.cursor_bw[k]);
- dml2_printf("DML::%s: k=%0u dpte_row_bw=%f\n", __func__, k, mode_lib->mp.dpte_row_bw[k]);
- dml2_printf("DML::%s: k=%0u meta_row_bw=%f\n", __func__, k, mode_lib->mp.meta_row_bw[k]);
- dml2_printf("DML::%s: k=%0u RequiredPrefetchPixelDataBWLuma=%f\n", __func__, k, mode_lib->mp.RequiredPrefetchPixelDataBWLuma[k]);
- dml2_printf("DML::%s: k=%0u RequiredPrefetchPixelDataBWChroma=%f\n", __func__, k, mode_lib->mp.RequiredPrefetchPixelDataBWChroma[k]);
- dml2_printf("DML::%s: k=%0u prefetch_cursor_bw=%f\n", __func__, k, mode_lib->mp.prefetch_cursor_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u prefetch_vmrow_bw=%f\n", __func__, k, mode_lib->mp.prefetch_vmrow_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u vactive_sw_bw_l=%f\n", __func__, k, mode_lib->mp.vactive_sw_bw_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u vactive_sw_bw_c=%f\n", __func__, k, mode_lib->mp.vactive_sw_bw_c[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u cursor_bw=%f\n", __func__, k, mode_lib->mp.cursor_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u dpte_row_bw=%f\n", __func__, k, mode_lib->mp.dpte_row_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u meta_row_bw=%f\n", __func__, k, mode_lib->mp.meta_row_bw[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u RequiredPrefetchPixelDataBWLuma=%f\n", __func__, k, mode_lib->mp.RequiredPrefetchPixelDataBWLuma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u RequiredPrefetchPixelDataBWChroma=%f\n", __func__, k, mode_lib->mp.RequiredPrefetchPixelDataBWChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%0u prefetch_cursor_bw=%f\n", __func__, k, mode_lib->mp.prefetch_cursor_bw[k]);
#endif
}
@@ -10988,10 +11402,11 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
calculate_peak_bandwidth_params->mall_prefetch_sdp_overhead_factor = mode_lib->mp.mall_prefetch_sdp_overhead_factor;
calculate_peak_bandwidth_params->mall_prefetch_dram_overhead_factor = mode_lib->mp.mall_prefetch_dram_overhead_factor;
- calculate_peak_bandwidth_params->surface_read_bandwidth_l = mode_lib->mp.SurfaceReadBandwidthLuma;
- calculate_peak_bandwidth_params->surface_read_bandwidth_c = mode_lib->mp.SurfaceReadBandwidthChroma;
+ calculate_peak_bandwidth_params->surface_read_bandwidth_l = mode_lib->mp.vactive_sw_bw_l;
+ calculate_peak_bandwidth_params->surface_read_bandwidth_c = mode_lib->mp.vactive_sw_bw_c;
calculate_peak_bandwidth_params->prefetch_bandwidth_l = mode_lib->mp.RequiredPrefetchPixelDataBWLuma;
calculate_peak_bandwidth_params->prefetch_bandwidth_c = mode_lib->mp.RequiredPrefetchPixelDataBWChroma;
+ calculate_peak_bandwidth_params->prefetch_bandwidth_max = s->dummy_single_array[0];
calculate_peak_bandwidth_params->excess_vactive_fill_bw_l = mode_lib->mp.excess_vactive_fill_bw_l;
calculate_peak_bandwidth_params->excess_vactive_fill_bw_c = mode_lib->mp.excess_vactive_fill_bw_c;
calculate_peak_bandwidth_params->cursor_bw = mode_lib->mp.cursor_bw;
@@ -11026,11 +11441,11 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
mode_lib->mp.urg_bandwidth_available);
if (!mode_lib->mp.PrefetchModeSupported)
- dml2_printf("DML::%s: Bandwidth not sufficient for prefetch!\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: Bandwidth not sufficient for prefetch!\n", __func__);
for (k = 0; k < s->num_active_planes; ++k) {
if (mode_lib->mp.NotEnoughUrgentLatencyHidingPre[k]) {
- dml2_printf("DML::%s: k=%u, NotEnoughUrgentLatencyHidingPre = %u\n", __func__, k, mode_lib->mp.NotEnoughUrgentLatencyHidingPre[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, NotEnoughUrgentLatencyHidingPre = %u\n", __func__, k, mode_lib->mp.NotEnoughUrgentLatencyHidingPre[k]);
mode_lib->mp.PrefetchModeSupported = false;
}
}
@@ -11056,12 +11471,12 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
}
mode_lib->mp.TotImmediateFlipBytes += s->per_pipe_flip_bytes[k] * mode_lib->mp.NoOfDPP[k];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k = %u\n", __func__, k);
- dml2_printf("DML::%s: DPPPerSurface = %u\n", __func__, mode_lib->mp.NoOfDPP[k]);
- dml2_printf("DML::%s: vm_bytes = %u\n", __func__, mode_lib->mp.vm_bytes[k]);
- dml2_printf("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, mode_lib->mp.PixelPTEBytesPerRow[k]);
- dml2_printf("DML::%s: meta_row_bytes = %u\n", __func__, mode_lib->mp.meta_row_bytes[k]);
- dml2_printf("DML::%s: TotImmediateFlipBytes = %u\n", __func__, mode_lib->mp.TotImmediateFlipBytes);
+ DML_LOG_VERBOSE("DML::%s: k = %u\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: DPPPerSurface = %u\n", __func__, mode_lib->mp.NoOfDPP[k]);
+ DML_LOG_VERBOSE("DML::%s: vm_bytes = %u\n", __func__, mode_lib->mp.vm_bytes[k]);
+ DML_LOG_VERBOSE("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, mode_lib->mp.PixelPTEBytesPerRow[k]);
+ DML_LOG_VERBOSE("DML::%s: meta_row_bytes = %u\n", __func__, mode_lib->mp.meta_row_bytes[k]);
+ DML_LOG_VERBOSE("DML::%s: TotImmediateFlipBytes = %u\n", __func__, mode_lib->mp.TotImmediateFlipBytes);
#endif
}
for (k = 0; k < s->num_active_planes; ++k) {
@@ -11120,8 +11535,8 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
calculate_peak_bandwidth_params->mall_prefetch_sdp_overhead_factor = mode_lib->mp.mall_prefetch_sdp_overhead_factor;
calculate_peak_bandwidth_params->mall_prefetch_dram_overhead_factor = mode_lib->mp.mall_prefetch_dram_overhead_factor;
- calculate_peak_bandwidth_params->surface_read_bandwidth_l = mode_lib->mp.SurfaceReadBandwidthLuma;
- calculate_peak_bandwidth_params->surface_read_bandwidth_c = mode_lib->mp.SurfaceReadBandwidthChroma;
+ calculate_peak_bandwidth_params->surface_read_bandwidth_l = mode_lib->mp.vactive_sw_bw_l;
+ calculate_peak_bandwidth_params->surface_read_bandwidth_c = mode_lib->mp.vactive_sw_bw_c;
calculate_peak_bandwidth_params->prefetch_bandwidth_l = mode_lib->mp.RequiredPrefetchPixelDataBWLuma;
calculate_peak_bandwidth_params->prefetch_bandwidth_c = mode_lib->mp.RequiredPrefetchPixelDataBWChroma;
calculate_peak_bandwidth_params->excess_vactive_fill_bw_l = mode_lib->mp.excess_vactive_fill_bw_l;
@@ -11131,6 +11546,7 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
calculate_peak_bandwidth_params->meta_row_bw = mode_lib->mp.meta_row_bw;
calculate_peak_bandwidth_params->prefetch_cursor_bw = mode_lib->mp.prefetch_cursor_bw;
calculate_peak_bandwidth_params->prefetch_vmrow_bw = mode_lib->mp.prefetch_vmrow_bw;
+ calculate_peak_bandwidth_params->prefetch_bandwidth_max = s->dummy_single_array[0];
calculate_peak_bandwidth_params->flip_bw = mode_lib->mp.final_flip_bw;
calculate_peak_bandwidth_params->urgent_burst_factor_l = mode_lib->mp.UrgentBurstFactorLuma;
calculate_peak_bandwidth_params->urgent_burst_factor_c = mode_lib->mp.UrgentBurstFactorChroma;
@@ -11153,13 +11569,13 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
mode_lib->mp.urg_bandwidth_available);
if (!mode_lib->mp.ImmediateFlipSupported)
- dml2_printf("DML::%s: Bandwidth not sufficient for flip!", __func__);
+ DML_LOG_VERBOSE("DML::%s: Bandwidth not sufficient for flip!", __func__);
for (k = 0; k < s->num_active_planes; ++k) {
if (display_cfg->plane_descriptors[k].immediate_flip && mode_lib->mp.ImmediateFlipSupportedForPipe[k] == false) {
mode_lib->mp.ImmediateFlipSupported = false;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Pipe %0d not supporting iflip!\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: Pipe %0d not supporting iflip!\n", __func__, k);
#endif
}
}
@@ -11172,28 +11588,28 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
mode_lib->mp.PrefetchAndImmediateFlipSupported = (mode_lib->mp.PrefetchModeSupported == true && (!must_support_iflip || mode_lib->mp.ImmediateFlipSupported));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: PrefetchModeSupported = %u\n", __func__, mode_lib->mp.PrefetchModeSupported);
+ DML_LOG_VERBOSE("DML::%s: PrefetchModeSupported = %u\n", __func__, mode_lib->mp.PrefetchModeSupported);
for (k = 0; k < s->num_active_planes; ++k)
- dml2_printf("DML::%s: immediate_flip_required[%u] = %u\n", __func__, k, display_cfg->plane_descriptors[k].immediate_flip);
- dml2_printf("DML::%s: HostVMEnable = %u\n", __func__, display_cfg->hostvm_enable);
- dml2_printf("DML::%s: ImmediateFlipSupported = %u\n", __func__, mode_lib->mp.ImmediateFlipSupported);
- dml2_printf("DML::%s: PrefetchAndImmediateFlipSupported = %u\n", __func__, mode_lib->mp.PrefetchAndImmediateFlipSupported);
+ DML_LOG_VERBOSE("DML::%s: immediate_flip_required[%u] = %u\n", __func__, k, display_cfg->plane_descriptors[k].immediate_flip);
+ DML_LOG_VERBOSE("DML::%s: HostVMEnable = %u\n", __func__, display_cfg->hostvm_enable);
+ DML_LOG_VERBOSE("DML::%s: ImmediateFlipSupported = %u\n", __func__, mode_lib->mp.ImmediateFlipSupported);
+ DML_LOG_VERBOSE("DML::%s: PrefetchAndImmediateFlipSupported = %u\n", __func__, mode_lib->mp.PrefetchAndImmediateFlipSupported);
#endif
- dml2_printf("DML::%s: Done one iteration: k=%d, MaxVStartupLines=%u\n", __func__, k, s->MaxVStartupLines[k]);
+ DML_LOG_VERBOSE("DML::%s: Done one iteration: k=%d, MaxVStartupLines=%u\n", __func__, k, s->MaxVStartupLines[k]);
}
for (k = 0; k < s->num_active_planes; ++k)
- dml2_printf("DML::%s: k=%d MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
if (!mode_lib->mp.PrefetchAndImmediateFlipSupported) {
- dml2_printf("DML::%s: Bad, Prefetch and flip scheduling solution NOT found!\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: Bad, Prefetch and flip scheduling solution NOT found!\n", __func__);
} else {
- dml2_printf("DML::%s: Good, Prefetch and flip scheduling solution found\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: Good, Prefetch and flip scheduling solution found\n", __func__);
// DCC Configuration
for (k = 0; k < s->num_active_planes; ++k) {
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Calculate DCC configuration for surface k=%u\n", __func__, k);
+ DML_LOG_VERBOSE("DML::%s: Calculate DCC configuration for surface k=%u\n", __func__, k);
#endif
CalculateDCCConfiguration(
display_cfg->plane_descriptors[k].surface.dcc.enable,
@@ -11238,8 +11654,8 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
s->mmSOCParameters.USRRetrainingLatency = 0;
s->mmSOCParameters.SMNLatency = 0;
s->mmSOCParameters.g6_temp_read_blackout_us = get_g6_temp_read_blackout_us(&mode_lib->soc, (unsigned int)(mode_lib->mp.uclk_freq_mhz * 1000), in_out_params->min_clk_index);
- s->mmSOCParameters.max_urgent_latency_us = get_max_urgent_latency_us(&mode_lib->soc.qos_parameters.qos_params.dcn4x, mode_lib->ms.uclk_freq_mhz, mode_lib->ms.FabricClock, in_out_params->min_clk_index);
- s->mmSOCParameters.df_response_time_us = mode_lib->soc.qos_parameters.qos_params.dcn4x.df_qos_response_time_fclk_cycles / mode_lib->ms.FabricClock;
+ s->mmSOCParameters.max_urgent_latency_us = get_max_urgent_latency_us(&mode_lib->soc.qos_parameters.qos_params.dcn4x, mode_lib->mp.uclk_freq_mhz, mode_lib->mp.FabricClock, in_out_params->min_clk_index);
+ s->mmSOCParameters.df_response_time_us = mode_lib->soc.qos_parameters.qos_params.dcn4x.df_qos_response_time_fclk_cycles / mode_lib->mp.FabricClock;
s->mmSOCParameters.qos_type = mode_lib->soc.qos_parameters.qos_type;
CalculateWatermarks_params->display_cfg = display_cfg;
@@ -11289,7 +11705,7 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(&mode_lib->scratch, CalculateWatermarks_params);
for (k = 0; k < s->num_active_planes; ++k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
+ if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.active_writebacks_per_stream > 0) {
mode_lib->mp.WritebackAllowDRAMClockChangeEndPosition[k] = math_max2(0, mode_lib->mp.VStartupMin[k] * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total /
((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) - mode_lib->mp.Watermark.WritebackDRAMClockChangeWatermark);
mode_lib->mp.WritebackAllowFCLKChangeEndPosition[k] = math_max2(0, mode_lib->mp.VStartupMin[k] * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total /
@@ -11302,8 +11718,8 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
calculate_pstate_keepout_dst_lines(display_cfg, &mode_lib->mp.Watermark, mode_lib->mp.pstate_keepout_dst_lines);
- dml2_printf("DML::%s: DEBUG stream_index = %0d\n", __func__, display_cfg->plane_descriptors[0].stream_index);
- dml2_printf("DML::%s: DEBUG PixelClock = %d kHz\n", __func__, (display_cfg->stream_descriptors[display_cfg->plane_descriptors[0].stream_index].timing.pixel_clock_khz));
+ DML_LOG_VERBOSE("DML::%s: DEBUG stream_index = %0d\n", __func__, display_cfg->plane_descriptors[0].stream_index);
+ DML_LOG_VERBOSE("DML::%s: DEBUG PixelClock = %ld kHz\n", __func__, (display_cfg->stream_descriptors[display_cfg->plane_descriptors[0].stream_index].timing.pixel_clock_khz));
//Display Pipeline Delivery Time in Prefetch, Groups
CalculatePixelDeliveryTimes(
@@ -11415,15 +11831,15 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
mode_lib->mp.MinTTUVBlank[k] = mode_lib->mp.TCalc + mode_lib->mp.MinTTUVBlank[k];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, MinTTUVBlank = %f (before vstartup margin)\n", __func__, k, mode_lib->mp.MinTTUVBlank[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, MinTTUVBlank = %f (before vstartup margin)\n", __func__, k, mode_lib->mp.MinTTUVBlank[k]);
#endif
s->Tvstartup_margin = (s->MaxVStartupLines[k] - mode_lib->mp.VStartupMin[k]) * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
mode_lib->mp.MinTTUVBlank[k] = mode_lib->mp.MinTTUVBlank[k] + s->Tvstartup_margin;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, Tvstartup_margin = %f\n", __func__, k, s->Tvstartup_margin);
- dml2_printf("DML::%s: k=%u, MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
- dml2_printf("DML::%s: k=%u, MinTTUVBlank = %f\n", __func__, k, mode_lib->mp.MinTTUVBlank[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, Tvstartup_margin = %f\n", __func__, k, s->Tvstartup_margin);
+ DML_LOG_VERBOSE("DML::%s: k=%u, MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, MinTTUVBlank = %f\n", __func__, k, mode_lib->mp.MinTTUVBlank[k]);
#endif
mode_lib->mp.Tdmdl[k] = mode_lib->mp.Tdmdl[k] + s->Tvstartup_margin;
@@ -11442,9 +11858,9 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
s->blank_lines_remaining = (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active) - mode_lib->mp.VStartup[k];
if (s->blank_lines_remaining < 0) {
- dml2_printf("ERROR: Vstartup is larger than vblank!?\n");
+ DML_LOG_VERBOSE("ERROR: Vstartup is larger than vblank!?\n");
s->blank_lines_remaining = 0;
- DML2_ASSERT(0);
+ DML_ASSERT(0);
}
mode_lib->mp.MIN_DST_Y_NEXT_START[k] = s->dlg_vblank_start + s->blank_lines_remaining + s->LSetup;
@@ -11458,42 +11874,42 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
mode_lib->mp.VREADY_AT_OR_AFTER_VSYNC[k] = false;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, VStartup = %u (max)\n", __func__, k, mode_lib->mp.VStartup[k]);
- dml2_printf("DML::%s: k=%u, VStartupMin = %u (max)\n", __func__, k, mode_lib->mp.VStartupMin[k]);
- dml2_printf("DML::%s: k=%u, VUpdateOffsetPix = %u\n", __func__, k, mode_lib->mp.VUpdateOffsetPix[k]);
- dml2_printf("DML::%s: k=%u, VUpdateWidthPix = %u\n", __func__, k, mode_lib->mp.VUpdateWidthPix[k]);
- dml2_printf("DML::%s: k=%u, VReadyOffsetPix = %u\n", __func__, k, mode_lib->mp.VReadyOffsetPix[k]);
- dml2_printf("DML::%s: k=%u, HTotal = %u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total);
- dml2_printf("DML::%s: k=%u, VTotal = %u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total);
- dml2_printf("DML::%s: k=%u, VActive = %u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active);
- dml2_printf("DML::%s: k=%u, VFrontPorch = %u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_front_porch);
- dml2_printf("DML::%s: k=%u, TSetup = %f\n", __func__, k, mode_lib->mp.TSetup[k]);
- dml2_printf("DML::%s: k=%u, MIN_DST_Y_NEXT_START = %f\n", __func__, k, mode_lib->mp.MIN_DST_Y_NEXT_START[k]);
- dml2_printf("DML::%s: k=%u, VREADY_AT_OR_AFTER_VSYNC = %u\n", __func__, k, mode_lib->mp.VREADY_AT_OR_AFTER_VSYNC[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VStartup = %u (max)\n", __func__, k, mode_lib->mp.VStartup[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VStartupMin = %u (max)\n", __func__, k, mode_lib->mp.VStartupMin[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VUpdateOffsetPix = %u\n", __func__, k, mode_lib->mp.VUpdateOffsetPix[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VUpdateWidthPix = %u\n", __func__, k, mode_lib->mp.VUpdateWidthPix[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VReadyOffsetPix = %u\n", __func__, k, mode_lib->mp.VReadyOffsetPix[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, HTotal = %lu\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VTotal = %lu\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VActive = %lu\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VFrontPorch = %lu\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_front_porch);
+ DML_LOG_VERBOSE("DML::%s: k=%u, TSetup = %f\n", __func__, k, mode_lib->mp.TSetup[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, MIN_DST_Y_NEXT_START = %f\n", __func__, k, mode_lib->mp.MIN_DST_Y_NEXT_START[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, VREADY_AT_OR_AFTER_VSYNC = %u\n", __func__, k, mode_lib->mp.VREADY_AT_OR_AFTER_VSYNC[k]);
#endif
}
//Maximum Bandwidth Used
- s->TotalWRBandwidth = 0;
- s->WRBandwidth = 0;
- for (k = 0; k < s->num_active_planes; ++k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format == dml2_444_32) {
- s->WRBandwidth = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width /
- (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * 4;
- } else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- s->WRBandwidth = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width /
- (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * 8;
+ mode_lib->mp.TotalWRBandwidth = 0;
+ for (k = 0; k < display_cfg->num_streams; ++k) {
+ s->WRBandwidth = 0;
+ if (display_cfg->stream_descriptors[k].writeback.active_writebacks_per_stream > 0) {
+ s->WRBandwidth = display_cfg->stream_descriptors[k].writeback.writeback_stream[0].output_height
+ * display_cfg->stream_descriptors[k].writeback.writeback_stream[0].output_width /
+ (display_cfg->stream_descriptors[k].timing.h_total * display_cfg->stream_descriptors[k].writeback.writeback_stream[0].input_height
+ / ((double)display_cfg->stream_descriptors[k].timing.pixel_clock_khz / 1000))
+ * (display_cfg->stream_descriptors[k].writeback.writeback_stream[0].pixel_format == dml2_444_32 ? 4.0 : 8.0);
+ mode_lib->mp.TotalWRBandwidth = mode_lib->mp.TotalWRBandwidth + s->WRBandwidth;
}
- s->TotalWRBandwidth = s->TotalWRBandwidth + s->WRBandwidth;
}
mode_lib->mp.TotalDataReadBandwidth = 0;
for (k = 0; k < s->num_active_planes; ++k) {
- mode_lib->mp.TotalDataReadBandwidth = mode_lib->mp.TotalDataReadBandwidth + mode_lib->mp.SurfaceReadBandwidthLuma[k] + mode_lib->mp.SurfaceReadBandwidthChroma[k];
+ mode_lib->mp.TotalDataReadBandwidth = mode_lib->mp.TotalDataReadBandwidth + mode_lib->mp.vactive_sw_bw_l[k] + mode_lib->mp.vactive_sw_bw_c[k];
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, TotalDataReadBandwidth = %f\n", __func__, k, mode_lib->mp.TotalDataReadBandwidth);
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceLuma = %f\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceChroma = %f\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthChroma[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, TotalDataReadBandwidth = %f\n", __func__, k, mode_lib->mp.TotalDataReadBandwidth);
+ DML_LOG_VERBOSE("DML::%s: k=%u, vactive_sw_bw_l = %f\n", __func__, k, mode_lib->mp.vactive_sw_bw_l[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%u, vactive_sw_bw_c = %f\n", __func__, k, mode_lib->mp.vactive_sw_bw_c[k]);
#endif
}
@@ -11530,8 +11946,8 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
CalculateStutterEfficiency_params->BlockWidth256BytesC = mode_lib->mp.Read256BlockWidthC;
CalculateStutterEfficiency_params->DCCYMaxUncompressedBlock = mode_lib->mp.DCCYMaxUncompressedBlock;
CalculateStutterEfficiency_params->DCCCMaxUncompressedBlock = mode_lib->mp.DCCCMaxUncompressedBlock;
- CalculateStutterEfficiency_params->ReadBandwidthSurfaceLuma = mode_lib->mp.SurfaceReadBandwidthLuma;
- CalculateStutterEfficiency_params->ReadBandwidthSurfaceChroma = mode_lib->mp.SurfaceReadBandwidthChroma;
+ CalculateStutterEfficiency_params->ReadBandwidthSurfaceLuma = mode_lib->mp.vactive_sw_bw_l;
+ CalculateStutterEfficiency_params->ReadBandwidthSurfaceChroma = mode_lib->mp.vactive_sw_bw_c;
CalculateStutterEfficiency_params->dpte_row_bw = mode_lib->mp.dpte_row_bw;
CalculateStutterEfficiency_params->meta_row_bw = mode_lib->mp.meta_row_bw;
CalculateStutterEfficiency_params->rob_alloc_compressed = mode_lib->ip.dcn_mrq_present;
@@ -11573,28 +11989,28 @@ static bool dml_core_mode_programming(struct dml2_core_calcs_mode_programming_ex
min_return_latency_in_DCFCLK_cycles = (min_return_uclk_cycles / max_uclk_mhz + min_return_fclk_cycles / max_fclk_mhz) * hard_minimum_dcfclk_mhz;
mode_lib->mp.min_return_latency_in_dcfclk = (unsigned int)min_return_latency_in_DCFCLK_cycles;
mode_lib->mp.dcfclk_deep_sleep_hysteresis = (unsigned int)math_max2(32, (double)mode_lib->ip.pixel_chunk_size_kbytes * 1024 * 3 / 4 / 64 - min_return_latency_in_DCFCLK_cycles);
- DML2_ASSERT(mode_lib->mp.dcfclk_deep_sleep_hysteresis < 256);
+ DML_ASSERT(mode_lib->mp.dcfclk_deep_sleep_hysteresis < 256);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: max_fclk_mhz = %f\n", __func__, max_fclk_mhz);
- dml2_printf("DML::%s: max_uclk_mhz = %f\n", __func__, max_uclk_mhz);
- dml2_printf("DML::%s: hard_minimum_dcfclk_mhz = %f\n", __func__, hard_minimum_dcfclk_mhz);
- dml2_printf("DML::%s: min_return_uclk_cycles = %d\n", __func__, min_return_uclk_cycles);
- dml2_printf("DML::%s: min_return_fclk_cycles = %d\n", __func__, min_return_fclk_cycles);
- dml2_printf("DML::%s: min_return_latency_in_DCFCLK_cycles = %f\n", __func__, min_return_latency_in_DCFCLK_cycles);
- dml2_printf("DML::%s: dcfclk_deep_sleep_hysteresis = %d \n", __func__, mode_lib->mp.dcfclk_deep_sleep_hysteresis);
- dml2_printf("DML::%s: --- END --- \n", __func__);
+ DML_LOG_VERBOSE("DML::%s: max_fclk_mhz = %f\n", __func__, max_fclk_mhz);
+ DML_LOG_VERBOSE("DML::%s: max_uclk_mhz = %f\n", __func__, max_uclk_mhz);
+ DML_LOG_VERBOSE("DML::%s: hard_minimum_dcfclk_mhz = %f\n", __func__, hard_minimum_dcfclk_mhz);
+ DML_LOG_VERBOSE("DML::%s: min_return_uclk_cycles = %ld\n", __func__, min_return_uclk_cycles);
+ DML_LOG_VERBOSE("DML::%s: min_return_fclk_cycles = %ld\n", __func__, min_return_fclk_cycles);
+ DML_LOG_VERBOSE("DML::%s: min_return_latency_in_DCFCLK_cycles = %f\n", __func__, min_return_latency_in_DCFCLK_cycles);
+ DML_LOG_VERBOSE("DML::%s: dcfclk_deep_sleep_hysteresis = %d \n", __func__, mode_lib->mp.dcfclk_deep_sleep_hysteresis);
+ DML_LOG_VERBOSE("DML::%s: --- END --- \n", __func__);
#endif
return (in_out_params->mode_lib->mp.PrefetchAndImmediateFlipSupported);
}
bool dml2_core_calcs_mode_programming_ex(struct dml2_core_calcs_mode_programming_ex *in_out_params)
{
- dml2_printf("DML::%s: ------------- START ----------\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: ------------- START ----------\n", __func__);
bool result = dml_core_mode_programming(in_out_params);
- dml2_printf("DML::%s: result = %0d\n", __func__, result);
- dml2_printf("DML::%s: ------------- DONE ----------\n", __func__);
+ DML_LOG_VERBOSE("DML::%s: result = %0d\n", __func__, result);
+ DML_LOG_VERBOSE("DML::%s: ------------- DONE ----------\n", __func__);
return result;
}
@@ -11652,16 +12068,16 @@ void dml2_core_calcs_get_dpte_row_height(
unsigned int MacroTileHeight = is_plane1 ? MacroTileHeightC : MacroTileHeightY;
unsigned int PTEBufferSizeInRequests = is_plane1 ? mode_lib->ip.dpte_buffer_size_in_pte_reqs_chroma : mode_lib->ip.dpte_buffer_size_in_pte_reqs_luma;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML: %s: is_plane1 = %u\n", __func__, is_plane1);
- dml2_printf("DML: %s: BytePerPixel = %u\n", __func__, BytePerPixel);
- dml2_printf("DML: %s: BlockHeight256Bytes = %u\n", __func__, BlockHeight256Bytes);
- dml2_printf("DML: %s: BlockWidth256Bytes = %u\n", __func__, BlockWidth256Bytes);
- dml2_printf("DML: %s: MacroTileWidth = %u\n", __func__, MacroTileWidth);
- dml2_printf("DML: %s: MacroTileHeight = %u\n", __func__, MacroTileHeight);
- dml2_printf("DML: %s: PTEBufferSizeInRequests = %u\n", __func__, PTEBufferSizeInRequests);
- dml2_printf("DML: %s: dpte_buffer_size_in_pte_reqs_luma = %u\n", __func__, mode_lib->ip.dpte_buffer_size_in_pte_reqs_luma);
- dml2_printf("DML: %s: dpte_buffer_size_in_pte_reqs_chroma = %u\n", __func__, mode_lib->ip.dpte_buffer_size_in_pte_reqs_chroma);
- dml2_printf("DML: %s: GPUVMMinPageSizeKBytes = %u\n", __func__, GPUVMMinPageSizeKBytes);
+ DML_LOG_VERBOSE("DML: %s: is_plane1 = %u\n", __func__, is_plane1);
+ DML_LOG_VERBOSE("DML: %s: BytePerPixel = %u\n", __func__, BytePerPixel);
+ DML_LOG_VERBOSE("DML: %s: BlockHeight256Bytes = %u\n", __func__, BlockHeight256Bytes);
+ DML_LOG_VERBOSE("DML: %s: BlockWidth256Bytes = %u\n", __func__, BlockWidth256Bytes);
+ DML_LOG_VERBOSE("DML: %s: MacroTileWidth = %u\n", __func__, MacroTileWidth);
+ DML_LOG_VERBOSE("DML: %s: MacroTileHeight = %u\n", __func__, MacroTileHeight);
+ DML_LOG_VERBOSE("DML: %s: PTEBufferSizeInRequests = %u\n", __func__, PTEBufferSizeInRequests);
+ DML_LOG_VERBOSE("DML: %s: dpte_buffer_size_in_pte_reqs_luma = %u\n", __func__, mode_lib->ip.dpte_buffer_size_in_pte_reqs_luma);
+ DML_LOG_VERBOSE("DML: %s: dpte_buffer_size_in_pte_reqs_chroma = %u\n", __func__, mode_lib->ip.dpte_buffer_size_in_pte_reqs_chroma);
+ DML_LOG_VERBOSE("DML: %s: GPUVMMinPageSizeKBytes = %u\n", __func__, GPUVMMinPageSizeKBytes);
#endif
unsigned int dummy_integer[21];
@@ -11715,16 +12131,16 @@ void dml2_core_calcs_get_dpte_row_height(
CalculateVMAndRowBytes(&mode_lib->scratch.calculate_vm_and_row_bytes_params);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML: %s: dpte_row_height = %u\n", __func__, *dpte_row_height);
+ DML_LOG_VERBOSE("DML: %s: dpte_row_height = %u\n", __func__, *dpte_row_height);
#endif
}
static bool is_dual_plane(enum dml2_source_format_class source_format)
{
- bool ret_val = 0;
+ bool ret_val = false;
if ((source_format == dml2_420_12) || (source_format == dml2_420_8) || (source_format == dml2_420_10) || (source_format == dml2_rgbe_alpha))
- ret_val = 1;
+ ret_val = true;
return ret_val;
}
@@ -11742,7 +12158,9 @@ static void rq_dlg_get_wm_regs(const struct dml2_display_cfg *display_cfg, const
wm_regs->fclk_pstate = (int unsigned)(mode_lib->mp.Watermark.FCLKChangeWatermark * refclk_freq_in_mhz);
wm_regs->sr_enter = (int unsigned)(mode_lib->mp.Watermark.StutterEnterPlusExitWatermark * refclk_freq_in_mhz);
wm_regs->sr_exit = (int unsigned)(mode_lib->mp.Watermark.StutterExitWatermark * refclk_freq_in_mhz);
- wm_regs->temp_read_or_ppt = (int unsigned)(mode_lib->mp.Watermark.g6_temp_read_watermark_us * refclk_freq_in_mhz);
+ wm_regs->sr_enter_z8 = (int unsigned)(mode_lib->mp.Watermark.Z8StutterEnterPlusExitWatermark * refclk_freq_in_mhz);
+ wm_regs->sr_exit_z8 = (int unsigned)(mode_lib->mp.Watermark.Z8StutterExitWatermark * refclk_freq_in_mhz);
+ wm_regs->temp_read_or_ppt = (int unsigned)(mode_lib->mp.Watermark.temp_read_or_ppt_watermark_us * refclk_freq_in_mhz);
wm_regs->uclk_pstate = (int unsigned)(mode_lib->mp.Watermark.DRAMClockChangeWatermark * refclk_freq_in_mhz);
wm_regs->urgent = (int unsigned)(mode_lib->mp.Watermark.UrgentWatermark * refclk_freq_in_mhz);
wm_regs->usr = (int unsigned)(mode_lib->mp.Watermark.USRRetrainingWatermark * refclk_freq_in_mhz);
@@ -11768,11 +12186,11 @@ void dml2_core_calcs_cursor_dlg_reg(struct dml2_cursor_dlg_regs *cursor_dlg_regs
cursor_dlg_regs->dst_x_offset = (unsigned int) ((dst_x_offset > 0) ? dst_x_offset : 0);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG::%s: cursor_x_position=%d\n", __func__, p->cursor_x_position);
- dml2_printf("DML_DLG::%s: dlg_refclk_mhz=%f\n", __func__, p->dlg_refclk_mhz);
- dml2_printf("DML_DLG::%s: pixel_rate_mhz=%f\n", __func__, p->pixel_rate_mhz);
- dml2_printf("DML_DLG::%s: dst_x_offset=%d\n", __func__, dst_x_offset);
- dml2_printf("DML_DLG::%s: dst_x_offset=%d (reg)\n", __func__, cursor_dlg_regs->dst_x_offset);
+ DML_LOG_VERBOSE("DML_DLG::%s: cursor_x_position=%d\n", __func__, p->cursor_x_position);
+ DML_LOG_VERBOSE("DML_DLG::%s: dlg_refclk_mhz=%f\n", __func__, p->dlg_refclk_mhz);
+ DML_LOG_VERBOSE("DML_DLG::%s: pixel_rate_mhz=%f\n", __func__, p->pixel_rate_mhz);
+ DML_LOG_VERBOSE("DML_DLG::%s: dst_x_offset=%d\n", __func__, dst_x_offset);
+ DML_LOG_VERBOSE("DML_DLG::%s: dst_x_offset=%d (reg)\n", __func__, cursor_dlg_regs->dst_x_offset);
#endif
cursor_dlg_regs->chunk_hdl_adjust = 3;
@@ -11808,7 +12226,7 @@ static void rq_dlg_get_rq_reg(struct dml2_display_rq_regs *rq_regs,
double stored_swath_c_bytes;
bool is_phantom_pipe;
- dml2_printf("DML_DLG::%s: Calculation for pipe[%d] start\n", __func__, pipe_idx);
+ DML_LOG_VERBOSE("DML_DLG::%s: Calculation for pipe[%d] start\n", __func__, pipe_idx);
pixel_chunk_bytes = (unsigned int)(mode_lib->ip.pixel_chunk_size_kbytes * 1024);
min_pixel_chunk_bytes = (unsigned int)(mode_lib->ip.min_pixel_chunk_size_bytes);
@@ -11851,19 +12269,19 @@ static void rq_dlg_get_rq_reg(struct dml2_display_rq_regs *rq_regs,
if (sw_mode == dml2_sw_linear && display_cfg->gpuvm_enable) {
unsigned int p0_pte_row_height_linear = (unsigned int)(dml_get_dpte_row_height_linear_l(mode_lib, pipe_idx));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: p0_pte_row_height_linear = %u\n", __func__, p0_pte_row_height_linear);
+ DML_LOG_VERBOSE("DML_DLG: %s: p0_pte_row_height_linear = %u\n", __func__, p0_pte_row_height_linear);
#endif
- DML2_ASSERT(p0_pte_row_height_linear >= 8);
+ DML_ASSERT(p0_pte_row_height_linear >= 8);
rq_regs->rq_regs_l.pte_row_height_linear = math_log2_approx(p0_pte_row_height_linear) - 3;
if (dual_plane) {
unsigned int p1_pte_row_height_linear = (unsigned int)(dml_get_dpte_row_height_linear_c(mode_lib, pipe_idx));
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: p1_pte_row_height_linear = %u\n", __func__, p1_pte_row_height_linear);
+ DML_LOG_VERBOSE("DML_DLG: %s: p1_pte_row_height_linear = %u\n", __func__, p1_pte_row_height_linear);
#endif
if (sw_mode == dml2_sw_linear) {
- DML2_ASSERT(p1_pte_row_height_linear >= 8);
+ DML_ASSERT(p1_pte_row_height_linear >= 8);
}
rq_regs->rq_regs_c.pte_row_height_linear = math_log2_approx(p1_pte_row_height_linear) - 3;
}
@@ -11897,12 +12315,12 @@ static void rq_dlg_get_rq_reg(struct dml2_display_rq_regs *rq_regs,
if (stored_swath_l_bytes / stored_swath_c_bytes <= 1.5) {
detile_buf_plane1_addr = (unsigned int)(detile_buf_size_in_bytes / 2.0 / 1024.0); // half to chroma
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: detile_buf_plane1_addr = %d (1/2 to chroma)\n", __func__, detile_buf_plane1_addr);
+ DML_LOG_VERBOSE("DML_DLG: %s: detile_buf_plane1_addr = %d (1/2 to chroma)\n", __func__, detile_buf_plane1_addr);
#endif
} else {
detile_buf_plane1_addr = (unsigned int)(dml_round_to_multiple((unsigned int)((2.0 * detile_buf_size_in_bytes) / 3.0), 1024, 0) / 1024.0); // 2/3 to luma
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: detile_buf_plane1_addr = %d (1/3 chroma)\n", __func__, detile_buf_plane1_addr);
+ DML_LOG_VERBOSE("DML_DLG: %s: detile_buf_plane1_addr = %d (1/3 chroma)\n", __func__, detile_buf_plane1_addr);
#endif
}
}
@@ -11910,15 +12328,15 @@ static void rq_dlg_get_rq_reg(struct dml2_display_rq_regs *rq_regs,
rq_regs->plane1_base_address = detile_buf_plane1_addr;
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: is_phantom_pipe = %d\n", __func__, is_phantom_pipe);
- dml2_printf("DML_DLG: %s: stored_swath_l_bytes = %f\n", __func__, stored_swath_l_bytes);
- dml2_printf("DML_DLG: %s: stored_swath_c_bytes = %f\n", __func__, stored_swath_c_bytes);
- dml2_printf("DML_DLG: %s: detile_buf_size_in_bytes = %d\n", __func__, detile_buf_size_in_bytes);
- dml2_printf("DML_DLG: %s: detile_buf_plane1_addr = %d\n", __func__, detile_buf_plane1_addr);
- dml2_printf("DML_DLG: %s: plane1_base_address = %d\n", __func__, rq_regs->plane1_base_address);
+ DML_LOG_VERBOSE("DML_DLG: %s: is_phantom_pipe = %d\n", __func__, is_phantom_pipe);
+ DML_LOG_VERBOSE("DML_DLG: %s: stored_swath_l_bytes = %f\n", __func__, stored_swath_l_bytes);
+ DML_LOG_VERBOSE("DML_DLG: %s: stored_swath_c_bytes = %f\n", __func__, stored_swath_c_bytes);
+ DML_LOG_VERBOSE("DML_DLG: %s: detile_buf_size_in_bytes = %d\n", __func__, detile_buf_size_in_bytes);
+ DML_LOG_VERBOSE("DML_DLG: %s: detile_buf_plane1_addr = %d\n", __func__, detile_buf_plane1_addr);
+ DML_LOG_VERBOSE("DML_DLG: %s: plane1_base_address = %d\n", __func__, rq_regs->plane1_base_address);
#endif
- //dml2_printf_rq_regs_st(rq_regs);
- dml2_printf("DML_DLG::%s: Calculation for pipe[%d] done\n", __func__, pipe_idx);
+ //DML_LOG_VERBOSE_rq_regs_st(rq_regs);
+ DML_LOG_VERBOSE("DML_DLG::%s: Calculation for pipe[%d] done\n", __func__, pipe_idx);
}
static void rq_dlg_get_dlg_reg(
@@ -11933,10 +12351,10 @@ static void rq_dlg_get_dlg_reg(
memset(l, 0, sizeof(struct dml2_core_shared_rq_dlg_get_dlg_reg_locals));
- dml2_printf("DML_DLG::%s: Calculation for pipe_idx=%d\n", __func__, pipe_idx);
+ DML_LOG_VERBOSE("DML_DLG::%s: Calculation for pipe_idx=%d\n", __func__, pipe_idx);
l->plane_idx = dml_get_plane_idx(mode_lib, pipe_idx);
- dml2_assert(l->plane_idx < DML2_MAX_PLANES);
+ DML_ASSERT(l->plane_idx < DML2_MAX_PLANES);
l->source_format = dml2_444_8;
l->odm_mode = dml2_odm_mode_bypass;
@@ -11966,18 +12384,18 @@ static void rq_dlg_get_dlg_reg(
l->pclk_freq_in_mhz = (double)l->timing->pixel_clock_khz / 1000;
l->ref_freq_to_pix_freq = l->refclk_freq_in_mhz / l->pclk_freq_in_mhz;
- dml2_printf("DML_DLG::%s: plane_idx = %d\n", __func__, l->plane_idx);
- dml2_printf("DML_DLG: %s: htotal = %d\n", __func__, l->htotal);
- dml2_printf("DML_DLG: %s: refclk_freq_in_mhz = %3.2f\n", __func__, l->refclk_freq_in_mhz);
- dml2_printf("DML_DLG: %s: dlg_ref_clk_mhz = %3.2f\n", __func__, display_cfg->overrides.hw.dlg_ref_clk_mhz);
- dml2_printf("DML_DLG: %s: soc.refclk_mhz = %3.2f\n", __func__, mode_lib->soc.dchub_refclk_mhz);
- dml2_printf("DML_DLG: %s: pclk_freq_in_mhz = %3.2f\n", __func__, l->pclk_freq_in_mhz);
- dml2_printf("DML_DLG: %s: ref_freq_to_pix_freq = %3.2f\n", __func__, l->ref_freq_to_pix_freq);
- dml2_printf("DML_DLG: %s: interlaced = %d\n", __func__, l->interlaced);
+ DML_LOG_VERBOSE("DML_DLG::%s: plane_idx = %d\n", __func__, l->plane_idx);
+ DML_LOG_VERBOSE("DML_DLG: %s: htotal = %d\n", __func__, l->htotal);
+ DML_LOG_VERBOSE("DML_DLG: %s: refclk_freq_in_mhz = %3.2f\n", __func__, l->refclk_freq_in_mhz);
+ DML_LOG_VERBOSE("DML_DLG: %s: dlg_ref_clk_mhz = %3.2f\n", __func__, display_cfg->overrides.hw.dlg_ref_clk_mhz);
+ DML_LOG_VERBOSE("DML_DLG: %s: soc.refclk_mhz = %d\n", __func__, mode_lib->soc.dchub_refclk_mhz);
+ DML_LOG_VERBOSE("DML_DLG: %s: pclk_freq_in_mhz = %3.2f\n", __func__, l->pclk_freq_in_mhz);
+ DML_LOG_VERBOSE("DML_DLG: %s: ref_freq_to_pix_freq = %3.2f\n", __func__, l->ref_freq_to_pix_freq);
+ DML_LOG_VERBOSE("DML_DLG: %s: interlaced = %d\n", __func__, l->interlaced);
- DML2_ASSERT(l->refclk_freq_in_mhz != 0);
- DML2_ASSERT(l->pclk_freq_in_mhz != 0);
- DML2_ASSERT(l->ref_freq_to_pix_freq < 4.0);
+ DML_ASSERT(l->refclk_freq_in_mhz != 0);
+ DML_ASSERT(l->pclk_freq_in_mhz != 0);
+ DML_ASSERT(l->ref_freq_to_pix_freq < 4.0);
// Need to figure out which side of odm combine we're in
// Assume the pipe instance under the same plane is in order
@@ -12006,14 +12424,14 @@ static void rq_dlg_get_dlg_reg(
l->pipe_idx_in_combine = pipe_idx - l->first_pipe_idx_in_plane; // DML assumes the pipes in the same plane will have continuous indexing (i.e. plane 0 use pipe 0, 1, and plane 1 uses pipe 2, 3, etc.)
disp_dlg_regs->refcyc_h_blank_end = (unsigned int)(((double)l->hblank_end + (double)l->pipe_idx_in_combine * (double)l->hactive / (double)l->odm_combine_factor) * l->ref_freq_to_pix_freq);
- dml2_printf("DML_DLG: %s: pipe_idx = %d\n", __func__, pipe_idx);
- dml2_printf("DML_DLG: %s: first_pipe_idx_in_plane = %d\n", __func__, l->first_pipe_idx_in_plane);
- dml2_printf("DML_DLG: %s: pipe_idx_in_combine = %d\n", __func__, l->pipe_idx_in_combine);
- dml2_printf("DML_DLG: %s: odm_combine_factor = %d\n", __func__, l->odm_combine_factor);
+ DML_LOG_VERBOSE("DML_DLG: %s: pipe_idx = %d\n", __func__, pipe_idx);
+ DML_LOG_VERBOSE("DML_DLG: %s: first_pipe_idx_in_plane = %d\n", __func__, l->first_pipe_idx_in_plane);
+ DML_LOG_VERBOSE("DML_DLG: %s: pipe_idx_in_combine = %d\n", __func__, l->pipe_idx_in_combine);
+ DML_LOG_VERBOSE("DML_DLG: %s: odm_combine_factor = %d\n", __func__, l->odm_combine_factor);
}
- dml2_printf("DML_DLG: %s: refcyc_h_blank_end = %d\n", __func__, disp_dlg_regs->refcyc_h_blank_end);
+ DML_LOG_VERBOSE("DML_DLG: %s: refcyc_h_blank_end = %d\n", __func__, disp_dlg_regs->refcyc_h_blank_end);
- DML2_ASSERT(disp_dlg_regs->refcyc_h_blank_end < (unsigned int)math_pow(2, 13));
+ DML_ASSERT(disp_dlg_regs->refcyc_h_blank_end < (unsigned int)math_pow(2, 13));
disp_dlg_regs->ref_freq_to_pix_freq = (unsigned int)(l->ref_freq_to_pix_freq * math_pow(2, 19));
disp_dlg_regs->refcyc_per_htotal = (unsigned int)(l->ref_freq_to_pix_freq * (double)l->htotal * math_pow(2, 8));
@@ -12022,20 +12440,20 @@ static void rq_dlg_get_dlg_reg(
l->min_ttu_vblank = mode_lib->mp.MinTTUVBlank[mode_lib->mp.pipe_plane[pipe_idx]];
l->min_dst_y_next_start = (unsigned int)(mode_lib->mp.MIN_DST_Y_NEXT_START[mode_lib->mp.pipe_plane[pipe_idx]]);
- dml2_printf("DML_DLG: %s: min_ttu_vblank (us) = %3.2f\n", __func__, l->min_ttu_vblank);
- dml2_printf("DML_DLG: %s: min_dst_y_next_start = %d\n", __func__, l->min_dst_y_next_start);
- dml2_printf("DML_DLG: %s: ref_freq_to_pix_freq = %3.2f\n", __func__, l->ref_freq_to_pix_freq);
+ DML_LOG_VERBOSE("DML_DLG: %s: min_ttu_vblank (us) = %3.2f\n", __func__, l->min_ttu_vblank);
+ DML_LOG_VERBOSE("DML_DLG: %s: min_dst_y_next_start = %d\n", __func__, l->min_dst_y_next_start);
+ DML_LOG_VERBOSE("DML_DLG: %s: ref_freq_to_pix_freq = %3.2f\n", __func__, l->ref_freq_to_pix_freq);
l->vready_after_vcount0 = (unsigned int)(mode_lib->mp.VREADY_AT_OR_AFTER_VSYNC[mode_lib->mp.pipe_plane[pipe_idx]]);
disp_dlg_regs->vready_after_vcount0 = l->vready_after_vcount0;
- dml2_printf("DML_DLG: %s: vready_after_vcount0 = %d\n", __func__, disp_dlg_regs->vready_after_vcount0);
+ DML_LOG_VERBOSE("DML_DLG: %s: vready_after_vcount0 = %d\n", __func__, disp_dlg_regs->vready_after_vcount0);
l->dst_x_after_scaler = (unsigned int)(mode_lib->mp.DSTXAfterScaler[mode_lib->mp.pipe_plane[pipe_idx]]);
l->dst_y_after_scaler = (unsigned int)(mode_lib->mp.DSTYAfterScaler[mode_lib->mp.pipe_plane[pipe_idx]]);
- dml2_printf("DML_DLG: %s: dst_x_after_scaler = %d\n", __func__, l->dst_x_after_scaler);
- dml2_printf("DML_DLG: %s: dst_y_after_scaler = %d\n", __func__, l->dst_y_after_scaler);
+ DML_LOG_VERBOSE("DML_DLG: %s: dst_x_after_scaler = %d\n", __func__, l->dst_x_after_scaler);
+ DML_LOG_VERBOSE("DML_DLG: %s: dst_y_after_scaler = %d\n", __func__, l->dst_y_after_scaler);
l->dst_y_prefetch = mode_lib->mp.dst_y_prefetch[mode_lib->mp.pipe_plane[pipe_idx]];
l->dst_y_per_vm_vblank = mode_lib->mp.dst_y_per_vm_vblank[mode_lib->mp.pipe_plane[pipe_idx]];
@@ -12043,28 +12461,28 @@ static void rq_dlg_get_dlg_reg(
l->dst_y_per_vm_flip = mode_lib->mp.dst_y_per_vm_flip[mode_lib->mp.pipe_plane[pipe_idx]];
l->dst_y_per_row_flip = mode_lib->mp.dst_y_per_row_flip[mode_lib->mp.pipe_plane[pipe_idx]];
- dml2_printf("DML_DLG: %s: dst_y_prefetch (after rnd) = %3.2f\n", __func__, l->dst_y_prefetch);
- dml2_printf("DML_DLG: %s: dst_y_per_vm_flip = %3.2f\n", __func__, l->dst_y_per_vm_flip);
- dml2_printf("DML_DLG: %s: dst_y_per_row_flip = %3.2f\n", __func__, l->dst_y_per_row_flip);
- dml2_printf("DML_DLG: %s: dst_y_per_vm_vblank = %3.2f\n", __func__, l->dst_y_per_vm_vblank);
- dml2_printf("DML_DLG: %s: dst_y_per_row_vblank = %3.2f\n", __func__, l->dst_y_per_row_vblank);
+ DML_LOG_VERBOSE("DML_DLG: %s: dst_y_prefetch (after rnd) = %3.2f\n", __func__, l->dst_y_prefetch);
+ DML_LOG_VERBOSE("DML_DLG: %s: dst_y_per_vm_flip = %3.2f\n", __func__, l->dst_y_per_vm_flip);
+ DML_LOG_VERBOSE("DML_DLG: %s: dst_y_per_row_flip = %3.2f\n", __func__, l->dst_y_per_row_flip);
+ DML_LOG_VERBOSE("DML_DLG: %s: dst_y_per_vm_vblank = %3.2f\n", __func__, l->dst_y_per_vm_vblank);
+ DML_LOG_VERBOSE("DML_DLG: %s: dst_y_per_row_vblank = %3.2f\n", __func__, l->dst_y_per_row_vblank);
if (l->dst_y_prefetch > 0 && l->dst_y_per_vm_vblank > 0 && l->dst_y_per_row_vblank > 0) {
- DML2_ASSERT(l->dst_y_prefetch > (l->dst_y_per_vm_vblank + l->dst_y_per_row_vblank));
+ DML_ASSERT(l->dst_y_prefetch > (l->dst_y_per_vm_vblank + l->dst_y_per_row_vblank));
}
l->vratio_pre_l = mode_lib->mp.VRatioPrefetchY[mode_lib->mp.pipe_plane[pipe_idx]];
l->vratio_pre_c = mode_lib->mp.VRatioPrefetchC[mode_lib->mp.pipe_plane[pipe_idx]];
- dml2_printf("DML_DLG: %s: vratio_pre_l = %3.2f\n", __func__, l->vratio_pre_l);
- dml2_printf("DML_DLG: %s: vratio_pre_c = %3.2f\n", __func__, l->vratio_pre_c);
+ DML_LOG_VERBOSE("DML_DLG: %s: vratio_pre_l = %3.2f\n", __func__, l->vratio_pre_l);
+ DML_LOG_VERBOSE("DML_DLG: %s: vratio_pre_c = %3.2f\n", __func__, l->vratio_pre_c);
// Active
l->refcyc_per_line_delivery_pre_l = mode_lib->mp.DisplayPipeLineDeliveryTimeLumaPrefetch[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
l->refcyc_per_line_delivery_l = mode_lib->mp.DisplayPipeLineDeliveryTimeLuma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- dml2_printf("DML_DLG: %s: refcyc_per_line_delivery_pre_l = %3.2f\n", __func__, l->refcyc_per_line_delivery_pre_l);
- dml2_printf("DML_DLG: %s: refcyc_per_line_delivery_l = %3.2f\n", __func__, l->refcyc_per_line_delivery_l);
+ DML_LOG_VERBOSE("DML_DLG: %s: refcyc_per_line_delivery_pre_l = %3.2f\n", __func__, l->refcyc_per_line_delivery_pre_l);
+ DML_LOG_VERBOSE("DML_DLG: %s: refcyc_per_line_delivery_l = %3.2f\n", __func__, l->refcyc_per_line_delivery_l);
l->refcyc_per_line_delivery_pre_c = 0.0;
l->refcyc_per_line_delivery_c = 0.0;
@@ -12073,8 +12491,8 @@ static void rq_dlg_get_dlg_reg(
l->refcyc_per_line_delivery_pre_c = mode_lib->mp.DisplayPipeLineDeliveryTimeChromaPrefetch[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
l->refcyc_per_line_delivery_c = mode_lib->mp.DisplayPipeLineDeliveryTimeChroma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- dml2_printf("DML_DLG: %s: refcyc_per_line_delivery_pre_c = %3.2f\n", __func__, l->refcyc_per_line_delivery_pre_c);
- dml2_printf("DML_DLG: %s: refcyc_per_line_delivery_c = %3.2f\n", __func__, l->refcyc_per_line_delivery_c);
+ DML_LOG_VERBOSE("DML_DLG: %s: refcyc_per_line_delivery_pre_c = %3.2f\n", __func__, l->refcyc_per_line_delivery_pre_c);
+ DML_LOG_VERBOSE("DML_DLG: %s: refcyc_per_line_delivery_c = %3.2f\n", __func__, l->refcyc_per_line_delivery_c);
}
disp_dlg_regs->refcyc_per_vm_dmdata = (unsigned int)(mode_lib->mp.Tdmdl_vm[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz);
@@ -12083,8 +12501,8 @@ static void rq_dlg_get_dlg_reg(
l->refcyc_per_req_delivery_pre_l = mode_lib->mp.DisplayPipeRequestDeliveryTimeLumaPrefetch[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
l->refcyc_per_req_delivery_l = mode_lib->mp.DisplayPipeRequestDeliveryTimeLuma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- dml2_printf("DML_DLG: %s: refcyc_per_req_delivery_pre_l = %3.2f\n", __func__, l->refcyc_per_req_delivery_pre_l);
- dml2_printf("DML_DLG: %s: refcyc_per_req_delivery_l = %3.2f\n", __func__, l->refcyc_per_req_delivery_l);
+ DML_LOG_VERBOSE("DML_DLG: %s: refcyc_per_req_delivery_pre_l = %3.2f\n", __func__, l->refcyc_per_req_delivery_pre_l);
+ DML_LOG_VERBOSE("DML_DLG: %s: refcyc_per_req_delivery_l = %3.2f\n", __func__, l->refcyc_per_req_delivery_l);
l->refcyc_per_req_delivery_pre_c = 0.0;
l->refcyc_per_req_delivery_c = 0.0;
@@ -12092,16 +12510,16 @@ static void rq_dlg_get_dlg_reg(
l->refcyc_per_req_delivery_pre_c = mode_lib->mp.DisplayPipeRequestDeliveryTimeChromaPrefetch[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
l->refcyc_per_req_delivery_c = mode_lib->mp.DisplayPipeRequestDeliveryTimeChroma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- dml2_printf("DML_DLG: %s: refcyc_per_req_delivery_pre_c = %3.2f\n", __func__, l->refcyc_per_req_delivery_pre_c);
- dml2_printf("DML_DLG: %s: refcyc_per_req_delivery_c = %3.2f\n", __func__, l->refcyc_per_req_delivery_c);
+ DML_LOG_VERBOSE("DML_DLG: %s: refcyc_per_req_delivery_pre_c = %3.2f\n", __func__, l->refcyc_per_req_delivery_pre_c);
+ DML_LOG_VERBOSE("DML_DLG: %s: refcyc_per_req_delivery_c = %3.2f\n", __func__, l->refcyc_per_req_delivery_c);
}
// TTU - Cursor
- DML2_ASSERT(display_cfg->plane_descriptors[l->plane_idx].cursor.num_cursors <= 1);
+ DML_ASSERT(display_cfg->plane_descriptors[l->plane_idx].cursor.num_cursors <= 1);
// Assign to register structures
disp_dlg_regs->min_dst_y_next_start = (unsigned int)((double)l->min_dst_y_next_start * math_pow(2, 2));
- DML2_ASSERT(disp_dlg_regs->min_dst_y_next_start < (unsigned int)math_pow(2, 18));
+ DML_ASSERT(disp_dlg_regs->min_dst_y_next_start < (unsigned int)math_pow(2, 18));
disp_dlg_regs->dst_y_after_scaler = l->dst_y_after_scaler; // in terms of line
disp_dlg_regs->refcyc_x_after_scaler = (unsigned int)((double)l->dst_x_after_scaler * l->ref_freq_to_pix_freq); // in terms of refclk
@@ -12114,10 +12532,10 @@ static void rq_dlg_get_dlg_reg(
disp_dlg_regs->vratio_prefetch = (unsigned int)(l->vratio_pre_l * math_pow(2, 19));
disp_dlg_regs->vratio_prefetch_c = (unsigned int)(l->vratio_pre_c * math_pow(2, 19));
- dml2_printf("DML_DLG: %s: disp_dlg_regs->dst_y_per_vm_vblank = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_vm_vblank);
- dml2_printf("DML_DLG: %s: disp_dlg_regs->dst_y_per_row_vblank = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_row_vblank);
- dml2_printf("DML_DLG: %s: disp_dlg_regs->dst_y_per_vm_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_vm_flip);
- dml2_printf("DML_DLG: %s: disp_dlg_regs->dst_y_per_row_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_row_flip);
+ DML_LOG_VERBOSE("DML_DLG: %s: disp_dlg_regs->dst_y_per_vm_vblank = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_vm_vblank);
+ DML_LOG_VERBOSE("DML_DLG: %s: disp_dlg_regs->dst_y_per_row_vblank = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_row_vblank);
+ DML_LOG_VERBOSE("DML_DLG: %s: disp_dlg_regs->dst_y_per_vm_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_vm_flip);
+ DML_LOG_VERBOSE("DML_DLG: %s: disp_dlg_regs->dst_y_per_row_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_row_flip);
disp_dlg_regs->refcyc_per_vm_group_vblank = (unsigned int)(mode_lib->mp.TimePerVMGroupVBlank[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz);
disp_dlg_regs->refcyc_per_vm_group_flip = (unsigned int)(mode_lib->mp.TimePerVMGroupFlip[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz);
@@ -12184,11 +12602,11 @@ static void rq_dlg_get_dlg_reg(
disp_ttu_regs->qos_ramp_disable_c = 0;
disp_ttu_regs->min_ttu_vblank = (unsigned int)(l->min_ttu_vblank * l->refclk_freq_in_mhz);
- // CHECK for HW registers' range, DML2_ASSERT or clamp
- DML2_ASSERT(l->refcyc_per_req_delivery_pre_l < math_pow(2, 13));
- DML2_ASSERT(l->refcyc_per_req_delivery_l < math_pow(2, 13));
- DML2_ASSERT(l->refcyc_per_req_delivery_pre_c < math_pow(2, 13));
- DML2_ASSERT(l->refcyc_per_req_delivery_c < math_pow(2, 13));
+ // CHECK for HW registers' range, DML_ASSERT or clamp
+ DML_ASSERT(l->refcyc_per_req_delivery_pre_l < math_pow(2, 13));
+ DML_ASSERT(l->refcyc_per_req_delivery_l < math_pow(2, 13));
+ DML_ASSERT(l->refcyc_per_req_delivery_pre_c < math_pow(2, 13));
+ DML_ASSERT(l->refcyc_per_req_delivery_c < math_pow(2, 13));
if (disp_dlg_regs->refcyc_per_vm_group_vblank >= (unsigned int)math_pow(2, 23))
disp_dlg_regs->refcyc_per_vm_group_vblank = (unsigned int)(math_pow(2, 23) - 1);
@@ -12202,16 +12620,16 @@ static void rq_dlg_get_dlg_reg(
disp_dlg_regs->refcyc_per_vm_req_flip = (unsigned int)(math_pow(2, 23) - 1);
- DML2_ASSERT(disp_dlg_regs->dst_y_after_scaler < (unsigned int)8);
- DML2_ASSERT(disp_dlg_regs->refcyc_x_after_scaler < (unsigned int)math_pow(2, 13));
+ DML_ASSERT(disp_dlg_regs->dst_y_after_scaler < (unsigned int)8);
+ DML_ASSERT(disp_dlg_regs->refcyc_x_after_scaler < (unsigned int)math_pow(2, 13));
if (disp_dlg_regs->dst_y_per_pte_row_nom_l >= (unsigned int)math_pow(2, 17)) {
- dml2_printf("DML_DLG: %s: Warning DST_Y_PER_PTE_ROW_NOM_L %u > register max U15.2 %u, clamp to max\n", __func__, disp_dlg_regs->dst_y_per_pte_row_nom_l, (unsigned int)math_pow(2, 17) - 1);
+ DML_LOG_VERBOSE("DML_DLG: %s: Warning DST_Y_PER_PTE_ROW_NOM_L %u > register max U15.2 %u, clamp to max\n", __func__, disp_dlg_regs->dst_y_per_pte_row_nom_l, (unsigned int)math_pow(2, 17) - 1);
l->dst_y_per_pte_row_nom_l = (unsigned int)math_pow(2, 17) - 1;
}
if (l->dual_plane) {
if (disp_dlg_regs->dst_y_per_pte_row_nom_c >= (unsigned int)math_pow(2, 17)) {
- dml2_printf("DML_DLG: %s: Warning DST_Y_PER_PTE_ROW_NOM_C %u > register max U15.2 %u, clamp to max\n", __func__, disp_dlg_regs->dst_y_per_pte_row_nom_c, (unsigned int)math_pow(2, 17) - 1);
+ DML_LOG_VERBOSE("DML_DLG: %s: Warning DST_Y_PER_PTE_ROW_NOM_C %u > register max U15.2 %u, clamp to max\n", __func__, disp_dlg_regs->dst_y_per_pte_row_nom_c, (unsigned int)math_pow(2, 17) - 1);
l->dst_y_per_pte_row_nom_c = (unsigned int)math_pow(2, 17) - 1;
}
}
@@ -12222,20 +12640,20 @@ static void rq_dlg_get_dlg_reg(
if (disp_dlg_regs->refcyc_per_pte_group_nom_c >= (unsigned int)math_pow(2, 23))
disp_dlg_regs->refcyc_per_pte_group_nom_c = (unsigned int)(math_pow(2, 23) - 1);
}
- DML2_ASSERT(disp_dlg_regs->refcyc_per_pte_group_vblank_l < (unsigned int)math_pow(2, 13));
+ DML_ASSERT(disp_dlg_regs->refcyc_per_pte_group_vblank_l < (unsigned int)math_pow(2, 13));
if (l->dual_plane) {
- DML2_ASSERT(disp_dlg_regs->refcyc_per_pte_group_vblank_c < (unsigned int)math_pow(2, 13));
+ DML_ASSERT(disp_dlg_regs->refcyc_per_pte_group_vblank_c < (unsigned int)math_pow(2, 13));
}
- DML2_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_pre_l < (unsigned int)math_pow(2, 13));
- DML2_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_l < (unsigned int)math_pow(2, 13));
- DML2_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_pre_c < (unsigned int)math_pow(2, 13));
- DML2_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_c < (unsigned int)math_pow(2, 13));
- DML2_ASSERT(disp_ttu_regs->qos_level_low_wm < (unsigned int)math_pow(2, 14));
- DML2_ASSERT(disp_ttu_regs->qos_level_high_wm < (unsigned int)math_pow(2, 14));
- DML2_ASSERT(disp_ttu_regs->min_ttu_vblank < (unsigned int)math_pow(2, 24));
+ DML_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_pre_l < (unsigned int)math_pow(2, 13));
+ DML_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_l < (unsigned int)math_pow(2, 13));
+ DML_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_pre_c < (unsigned int)math_pow(2, 13));
+ DML_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_c < (unsigned int)math_pow(2, 13));
+ DML_ASSERT(disp_ttu_regs->qos_level_low_wm < (unsigned int)math_pow(2, 14));
+ DML_ASSERT(disp_ttu_regs->qos_level_high_wm < (unsigned int)math_pow(2, 14));
+ DML_ASSERT(disp_ttu_regs->min_ttu_vblank < (unsigned int)math_pow(2, 24));
- dml2_printf("DML_DLG::%s: Calculation for pipe[%d] done\n", __func__, pipe_idx);
+ DML_LOG_VERBOSE("DML_DLG::%s: Calculation for pipe[%d] done\n", __func__, pipe_idx);
}
}
@@ -12258,11 +12676,11 @@ static void rq_dlg_get_arb_params(const struct dml2_display_cfg *display_cfg, co
arb_param->pstate_stall_threshold = (unsigned int)(mode_lib->ip_caps.fams2.max_allow_delay_us * refclk_freq_in_mhz);
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: max_req_outstanding = %d\n", __func__, arb_param->max_req_outstanding);
- dml2_printf("DML::%s: sdpif_request_rate_limit = %d\n", __func__, arb_param->sdpif_request_rate_limit);
- dml2_printf("DML::%s: compbuf_reserved_space_kbytes = %d\n", __func__, arb_param->compbuf_reserved_space_kbytes);
- dml2_printf("DML::%s: allow_sdpif_rate_limit_when_cstate_req = %d\n", __func__, arb_param->allow_sdpif_rate_limit_when_cstate_req);
- dml2_printf("DML::%s: dcfclk_deep_sleep_hysteresis = %d\n", __func__, arb_param->dcfclk_deep_sleep_hysteresis);
+ DML_LOG_VERBOSE("DML::%s: max_req_outstanding = %d\n", __func__, arb_param->max_req_outstanding);
+ DML_LOG_VERBOSE("DML::%s: sdpif_request_rate_limit = %d\n", __func__, arb_param->sdpif_request_rate_limit);
+ DML_LOG_VERBOSE("DML::%s: compbuf_reserved_space_kbytes = %d\n", __func__, arb_param->compbuf_reserved_space_kbytes);
+ DML_LOG_VERBOSE("DML::%s: allow_sdpif_rate_limit_when_cstate_req = %d\n", __func__, arb_param->allow_sdpif_rate_limit_when_cstate_req);
+ DML_LOG_VERBOSE("DML::%s: dcfclk_deep_sleep_hysteresis = %d\n", __func__, arb_param->dcfclk_deep_sleep_hysteresis);
#endif
}
@@ -12321,14 +12739,18 @@ void dml2_core_calcs_get_global_fams2_programming(const struct dml2_core_interna
void dml2_core_calcs_get_stream_fams2_programming(const struct dml2_core_internal_display_mode_lib *mode_lib,
const struct display_configuation_with_meta *display_cfg,
- struct dmub_fams2_stream_static_state *fams2_programming,
- enum dml2_uclk_pstate_support_method pstate_method,
+ union dmub_cmd_fams2_config *fams2_base_programming,
+ union dmub_cmd_fams2_config *fams2_sub_programming,
+ enum dml2_pstate_method pstate_method,
int plane_index)
{
const struct dml2_plane_parameters *plane_descriptor = &display_cfg->display_config.plane_descriptors[plane_index];
const struct dml2_stream_parameters *stream_descriptor = &display_cfg->display_config.stream_descriptors[plane_descriptor->stream_index];
const struct dml2_fams2_meta *stream_fams2_meta = &display_cfg->stage3.stream_fams2_meta[plane_descriptor->stream_index];
+ struct dmub_fams2_cmd_stream_static_base_state *base_programming = &fams2_base_programming->stream_v1.base;
+ union dmub_fams2_cmd_stream_static_sub_state *sub_programming = &fams2_sub_programming->stream_v1.sub_state;
+
unsigned int i;
if (display_cfg->display_config.overrides.all_streams_blanked) {
@@ -12337,110 +12759,110 @@ void dml2_core_calcs_get_stream_fams2_programming(const struct dml2_core_interna
}
/* from display configuration */
- fams2_programming->htotal = (uint16_t)stream_descriptor->timing.h_total;
- fams2_programming->vtotal = (uint16_t)stream_descriptor->timing.v_total;
- fams2_programming->vblank_start = (uint16_t)(stream_fams2_meta->nom_vtotal -
+ base_programming->htotal = (uint16_t)stream_descriptor->timing.h_total;
+ base_programming->vtotal = (uint16_t)stream_descriptor->timing.v_total;
+ base_programming->vblank_start = (uint16_t)(stream_fams2_meta->nom_vtotal -
stream_descriptor->timing.v_front_porch);
- fams2_programming->vblank_end = (uint16_t)(stream_fams2_meta->nom_vtotal -
+ base_programming->vblank_end = (uint16_t)(stream_fams2_meta->nom_vtotal -
stream_descriptor->timing.v_front_porch -
stream_descriptor->timing.v_active);
- fams2_programming->config.bits.is_drr = stream_descriptor->timing.drr_config.enabled;
+ base_programming->config.bits.is_drr = stream_descriptor->timing.drr_config.enabled;
/* from meta */
- fams2_programming->otg_vline_time_ns =
+ base_programming->otg_vline_time_ns =
(unsigned int)(stream_fams2_meta->otg_vline_time_us * 1000.0);
- fams2_programming->scheduling_delay_otg_vlines = (uint8_t)stream_fams2_meta->scheduling_delay_otg_vlines;
- fams2_programming->contention_delay_otg_vlines = (uint8_t)stream_fams2_meta->contention_delay_otg_vlines;
- fams2_programming->vline_int_ack_delay_otg_vlines = (uint8_t)stream_fams2_meta->vertical_interrupt_ack_delay_otg_vlines;
- fams2_programming->drr_keepout_otg_vline = (uint16_t)(stream_fams2_meta->nom_vtotal -
+ base_programming->scheduling_delay_otg_vlines = (uint8_t)stream_fams2_meta->scheduling_delay_otg_vlines;
+ base_programming->contention_delay_otg_vlines = (uint8_t)stream_fams2_meta->contention_delay_otg_vlines;
+ base_programming->vline_int_ack_delay_otg_vlines = (uint8_t)stream_fams2_meta->vertical_interrupt_ack_delay_otg_vlines;
+ base_programming->drr_keepout_otg_vline = (uint16_t)(stream_fams2_meta->nom_vtotal -
stream_descriptor->timing.v_front_porch -
stream_fams2_meta->method_drr.programming_delay_otg_vlines);
- fams2_programming->allow_to_target_delay_otg_vlines = (uint8_t)stream_fams2_meta->allow_to_target_delay_otg_vlines;
- fams2_programming->max_vtotal = (uint16_t)stream_fams2_meta->max_vtotal;
+ base_programming->allow_to_target_delay_otg_vlines = (uint8_t)stream_fams2_meta->allow_to_target_delay_otg_vlines;
+ base_programming->max_vtotal = (uint16_t)stream_fams2_meta->max_vtotal;
/* from core */
- fams2_programming->config.bits.min_ttu_vblank_usable = true;
+ base_programming->config.bits.min_ttu_vblank_usable = true;
for (i = 0; i < display_cfg->display_config.num_planes; i++) {
/* check if all planes support p-state in blank */
if (display_cfg->display_config.plane_descriptors[i].stream_index == plane_descriptor->stream_index &&
mode_lib->mp.MinTTUVBlank[i] <= mode_lib->mp.Watermark.DRAMClockChangeWatermark) {
- fams2_programming->config.bits.min_ttu_vblank_usable = false;
+ base_programming->config.bits.min_ttu_vblank_usable = false;
break;
}
}
switch (pstate_method) {
- case dml2_uclk_pstate_support_method_vactive:
- case dml2_uclk_pstate_support_method_fw_vactive_drr:
+ case dml2_pstate_method_vactive:
+ case dml2_pstate_method_fw_vactive_drr:
/* legacy vactive */
- fams2_programming->type = FAMS2_STREAM_TYPE_VACTIVE;
- fams2_programming->sub_state.legacy.vactive_det_fill_delay_otg_vlines =
- (uint8_t)stream_fams2_meta->method_vactive.max_vactive_det_fill_delay_otg_vlines;
- fams2_programming->allow_start_otg_vline =
- (uint16_t)stream_fams2_meta->method_vactive.common.allow_start_otg_vline;
- fams2_programming->allow_end_otg_vline =
- (uint16_t)stream_fams2_meta->method_vactive.common.allow_end_otg_vline;
- fams2_programming->config.bits.clamp_vtotal_min = true;
+ base_programming->type = FAMS2_STREAM_TYPE_VACTIVE;
+ sub_programming->legacy.vactive_det_fill_delay_otg_vlines =
+ (uint8_t)stream_fams2_meta->method_vactive.max_vactive_det_fill_delay_otg_vlines;
+ base_programming->allow_start_otg_vline =
+ (uint16_t)stream_fams2_meta->method_vactive.common.allow_start_otg_vline;
+ base_programming->allow_end_otg_vline =
+ (uint16_t)stream_fams2_meta->method_vactive.common.allow_end_otg_vline;
+ base_programming->config.bits.clamp_vtotal_min = true;
break;
- case dml2_uclk_pstate_support_method_vblank:
- case dml2_uclk_pstate_support_method_fw_vblank_drr:
+ case dml2_pstate_method_vblank:
+ case dml2_pstate_method_fw_vblank_drr:
/* legacy vblank */
- fams2_programming->type = FAMS2_STREAM_TYPE_VBLANK;
- fams2_programming->allow_start_otg_vline =
- (uint16_t)stream_fams2_meta->method_vblank.common.allow_start_otg_vline;
- fams2_programming->allow_end_otg_vline =
- (uint16_t)stream_fams2_meta->method_vblank.common.allow_end_otg_vline;
- fams2_programming->config.bits.clamp_vtotal_min = true;
+ base_programming->type = FAMS2_STREAM_TYPE_VBLANK;
+ base_programming->allow_start_otg_vline =
+ (uint16_t)stream_fams2_meta->method_vblank.common.allow_start_otg_vline;
+ base_programming->allow_end_otg_vline =
+ (uint16_t)stream_fams2_meta->method_vblank.common.allow_end_otg_vline;
+ base_programming->config.bits.clamp_vtotal_min = true;
break;
- case dml2_uclk_pstate_support_method_fw_drr:
+ case dml2_pstate_method_fw_drr:
/* drr */
- fams2_programming->type = FAMS2_STREAM_TYPE_DRR;
- fams2_programming->sub_state.drr.programming_delay_otg_vlines =
- (uint8_t)stream_fams2_meta->method_drr.programming_delay_otg_vlines;
- fams2_programming->sub_state.drr.nom_stretched_vtotal =
- (uint16_t)stream_fams2_meta->method_drr.stretched_vtotal;
- fams2_programming->allow_start_otg_vline =
- (uint16_t)stream_fams2_meta->method_drr.common.allow_start_otg_vline;
- fams2_programming->allow_end_otg_vline =
- (uint16_t)stream_fams2_meta->method_drr.common.allow_end_otg_vline;
+ base_programming->type = FAMS2_STREAM_TYPE_DRR;
+ sub_programming->drr.programming_delay_otg_vlines =
+ (uint8_t)stream_fams2_meta->method_drr.programming_delay_otg_vlines;
+ sub_programming->drr.nom_stretched_vtotal =
+ (uint16_t)stream_fams2_meta->method_drr.stretched_vtotal;
+ base_programming->allow_start_otg_vline =
+ (uint16_t)stream_fams2_meta->method_drr.common.allow_start_otg_vline;
+ base_programming->allow_end_otg_vline =
+ (uint16_t)stream_fams2_meta->method_drr.common.allow_end_otg_vline;
/* drr only clamps to vtotal min for single display */
- fams2_programming->config.bits.clamp_vtotal_min = display_cfg->display_config.num_streams == 1;
- fams2_programming->sub_state.drr.only_stretch_if_required = true;
+ base_programming->config.bits.clamp_vtotal_min = display_cfg->display_config.num_streams == 1;
+ sub_programming->drr.only_stretch_if_required = true;
break;
- case dml2_uclk_pstate_support_method_fw_subvp_phantom:
- case dml2_uclk_pstate_support_method_fw_subvp_phantom_drr:
+ case dml2_pstate_method_fw_svp:
+ case dml2_pstate_method_fw_svp_drr:
/* subvp */
- fams2_programming->type = FAMS2_STREAM_TYPE_SUBVP;
- fams2_programming->sub_state.subvp.vratio_numerator =
- (uint16_t)(plane_descriptor->composition.scaler_info.plane0.v_ratio * 1000.0);
- fams2_programming->sub_state.subvp.vratio_denominator = 1000;
- fams2_programming->sub_state.subvp.programming_delay_otg_vlines =
- (uint8_t)stream_fams2_meta->method_subvp.programming_delay_otg_vlines;
- fams2_programming->sub_state.subvp.prefetch_to_mall_otg_vlines =
- (uint8_t)stream_fams2_meta->method_subvp.prefetch_to_mall_delay_otg_vlines;
- fams2_programming->sub_state.subvp.phantom_vtotal =
- (uint16_t)stream_fams2_meta->method_subvp.phantom_vtotal;
- fams2_programming->sub_state.subvp.phantom_vactive =
- (uint16_t)stream_fams2_meta->method_subvp.phantom_vactive;
- fams2_programming->sub_state.subvp.config.bits.is_multi_planar =
- plane_descriptor->surface.plane1.height > 0;
- fams2_programming->sub_state.subvp.config.bits.is_yuv420 =
- plane_descriptor->pixel_format == dml2_420_8 ||
- plane_descriptor->pixel_format == dml2_420_10 ||
- plane_descriptor->pixel_format == dml2_420_12;
-
- fams2_programming->allow_start_otg_vline =
- (uint16_t)stream_fams2_meta->method_subvp.common.allow_start_otg_vline;
- fams2_programming->allow_end_otg_vline =
- (uint16_t)stream_fams2_meta->method_subvp.common.allow_end_otg_vline;
- fams2_programming->config.bits.clamp_vtotal_min = true;
+ base_programming->type = FAMS2_STREAM_TYPE_SUBVP;
+ sub_programming->subvp.vratio_numerator =
+ (uint16_t)(plane_descriptor->composition.scaler_info.plane0.v_ratio * 1000.0);
+ sub_programming->subvp.vratio_denominator = 1000;
+ sub_programming->subvp.programming_delay_otg_vlines =
+ (uint8_t)stream_fams2_meta->method_subvp.programming_delay_otg_vlines;
+ sub_programming->subvp.prefetch_to_mall_otg_vlines =
+ (uint8_t)stream_fams2_meta->method_subvp.prefetch_to_mall_delay_otg_vlines;
+ sub_programming->subvp.phantom_vtotal =
+ (uint16_t)stream_fams2_meta->method_subvp.phantom_vtotal;
+ sub_programming->subvp.phantom_vactive =
+ (uint16_t)stream_fams2_meta->method_subvp.phantom_vactive;
+ sub_programming->subvp.config.bits.is_multi_planar =
+ plane_descriptor->surface.plane1.height > 0;
+ sub_programming->subvp.config.bits.is_yuv420 =
+ plane_descriptor->pixel_format == dml2_420_8 ||
+ plane_descriptor->pixel_format == dml2_420_10 ||
+ plane_descriptor->pixel_format == dml2_420_12;
+
+ base_programming->allow_start_otg_vline =
+ (uint16_t)stream_fams2_meta->method_subvp.common.allow_start_otg_vline;
+ base_programming->allow_end_otg_vline =
+ (uint16_t)stream_fams2_meta->method_subvp.common.allow_end_otg_vline;
+ base_programming->config.bits.clamp_vtotal_min = true;
break;
- case dml2_uclk_pstate_support_method_reserved_hw:
- case dml2_uclk_pstate_support_method_reserved_fw:
- case dml2_uclk_pstate_support_method_reserved_fw_drr_fixed:
- case dml2_uclk_pstate_support_method_reserved_fw_drr_var:
- case dml2_uclk_pstate_support_method_not_supported:
- case dml2_uclk_pstate_support_method_count:
+ case dml2_pstate_method_reserved_hw:
+ case dml2_pstate_method_reserved_fw:
+ case dml2_pstate_method_reserved_fw_drr_clamped:
+ case dml2_pstate_method_reserved_fw_drr_var:
+ case dml2_pstate_method_na:
+ case dml2_pstate_method_count:
default:
/* this should never happen */
break;
@@ -12531,10 +12953,10 @@ void dml2_core_calcs_get_stream_support_info(const struct dml2_display_cfg *disp
out->vblank_reserved_time_us = display_cfg->plane_descriptors[plane_index].overrides.reserved_vblank_time_ns / 1000;
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: subvp_fw_processing_delay_us = %d\n", __func__, mode_lib->ip.subvp_fw_processing_delay_us);
- dml2_printf("DML::%s: subvp_pstate_allow_width_us = %d\n", __func__, mode_lib->ip.subvp_pstate_allow_width_us);
- dml2_printf("DML::%s: subvp_swath_height_margin_lines = %d\n", __func__, mode_lib->ip.subvp_swath_height_margin_lines);
- dml2_printf("DML::%s: vblank_reserved_time_us = %f\n", __func__, out->vblank_reserved_time_us);
+ DML_LOG_VERBOSE("DML::%s: subvp_fw_processing_delay_us = %d\n", __func__, mode_lib->ip.subvp_fw_processing_delay_us);
+ DML_LOG_VERBOSE("DML::%s: subvp_pstate_allow_width_us = %d\n", __func__, mode_lib->ip.subvp_pstate_allow_width_us);
+ DML_LOG_VERBOSE("DML::%s: subvp_swath_height_margin_lines = %d\n", __func__, mode_lib->ip.subvp_swath_height_margin_lines);
+ DML_LOG_VERBOSE("DML::%s: vblank_reserved_time_us = %u\n", __func__, out->vblank_reserved_time_us);
#endif
}
@@ -12569,6 +12991,8 @@ void dml2_core_calcs_get_informative(const struct dml2_core_internal_display_mod
out->informative.mode_support_info.InvalidCombinationOfMALLUseForPState = mode_lib->ms.support.InvalidCombinationOfMALLUseForPState;
out->informative.mode_support_info.ExceededMALLSize = mode_lib->ms.support.ExceededMALLSize;
out->informative.mode_support_info.EnoughWritebackUnits = mode_lib->ms.support.EnoughWritebackUnits;
+ out->informative.mode_support_info.temp_read_or_ppt_support = mode_lib->ms.support.temp_read_or_ppt_support;
+ out->informative.mode_support_info.g6_temp_read_support = mode_lib->ms.support.g6_temp_read_support;
out->informative.mode_support_info.ExceededMultistreamSlots = mode_lib->ms.support.ExceededMultistreamSlots;
out->informative.mode_support_info.NotEnoughDSCUnits = mode_lib->ms.support.NotEnoughDSCUnits;
@@ -12645,6 +13069,10 @@ void dml2_core_calcs_get_informative(const struct dml2_core_internal_display_mod
out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_10x4;
else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_hdmi_rate_12x4)
out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_12x4;
+ else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_hdmi_rate_16x4)
+ out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_16x4;
+ else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_hdmi_rate_20x4)
+ out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_20x4;
out->informative.mode_support_info.AlignedYPitch[k] = mode_lib->ms.support.AlignedYPitch[k];
out->informative.mode_support_info.AlignedCPitch[k] = mode_lib->ms.support.AlignedCPitch[k];
@@ -12662,11 +13090,14 @@ void dml2_core_calcs_get_informative(const struct dml2_core_internal_display_mod
out->informative.watermarks.pstate_change_us = dml_get_wm_dram_clock_change(mode_lib);
out->informative.watermarks.fclk_pstate_change_us = dml_get_wm_fclk_change(mode_lib);
out->informative.watermarks.usr_retraining_us = dml_get_wm_usr_retraining(mode_lib);
- out->informative.watermarks.g6_temp_read_watermark_us = dml_get_wm_g6_temp_read(mode_lib);
+ out->informative.watermarks.temp_read_or_ppt_watermark_us = dml_get_wm_temp_read_or_ppt(mode_lib);
out->informative.mall.total_surface_size_in_mall_bytes = 0;
- for (k = 0; k < out->display_config.num_planes; ++k)
+ out->informative.dpp.total_num_dpps_required = 0;
+ for (k = 0; k < out->display_config.num_planes; ++k) {
out->informative.mall.total_surface_size_in_mall_bytes += mode_lib->mp.SurfaceSizeInTheMALL[k];
+ out->informative.dpp.total_num_dpps_required += mode_lib->mp.NoOfDPP[k];
+ }
out->informative.qos.min_return_latency_in_dcfclk = mode_lib->mp.min_return_latency_in_dcfclk;
out->informative.qos.urgent_latency_us = dml_get_urgent_latency(mode_lib);
@@ -12745,6 +13176,8 @@ void dml2_core_calcs_get_informative(const struct dml2_core_internal_display_mod
out->informative.qos.max_active_fclk_change_latency_supported = dml_get_fclk_change_latency(mode_lib);
+ out->informative.misc.LowestPrefetchMargin = 10 * 1000 * 1000;
+
for (k = 0; k < out->display_config.num_planes; k++) {
if ((out->display_config.plane_descriptors->overrides.reserved_vblank_time_ns >= 1000.0 * mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us)
@@ -12824,6 +13257,7 @@ void dml2_core_calcs_get_informative(const struct dml2_core_internal_display_mod
out->informative.misc.DisplayPipeLineDeliveryTimeLumaPrefetch[k] = mode_lib->mp.DisplayPipeLineDeliveryTimeLumaPrefetch[k];
out->informative.misc.DisplayPipeLineDeliveryTimeChromaPrefetch[k] = mode_lib->mp.DisplayPipeLineDeliveryTimeChromaPrefetch[k];
+ out->informative.misc.WritebackRequiredBandwidth = mode_lib->mp.TotalWRBandwidth / 1000.0;
out->informative.misc.WritebackAllowDRAMClockChangeEndPosition[k] = mode_lib->mp.WritebackAllowDRAMClockChangeEndPosition[k];
out->informative.misc.WritebackAllowFCLKChangeEndPosition[k] = mode_lib->mp.WritebackAllowFCLKChangeEndPosition[k];
out->informative.misc.DSCCLK_calculated[k] = mode_lib->mp.DSCCLK[k];
@@ -12831,6 +13265,9 @@ void dml2_core_calcs_get_informative(const struct dml2_core_internal_display_mod
out->informative.misc.PTE_BUFFER_MODE[k] = mode_lib->mp.PTE_BUFFER_MODE[k];
out->informative.misc.DSCDelay[k] = mode_lib->mp.DSCDelay[k];
out->informative.misc.MaxActiveDRAMClockChangeLatencySupported[k] = mode_lib->mp.MaxActiveDRAMClockChangeLatencySupported[k];
+
+ if (mode_lib->mp.impacted_prefetch_margin_us[k] < out->informative.misc.LowestPrefetchMargin)
+ out->informative.misc.LowestPrefetchMargin = mode_lib->mp.impacted_prefetch_margin_us[k];
}
// For this DV informative layer, all pipes in the same planes will just use the same id
@@ -12853,16 +13290,11 @@ void dml2_core_calcs_get_informative(const struct dml2_core_internal_display_mod
out->informative.non_optimized_mcache_allocation[k].global_mcache_ids_plane1[n] = k;
}
}
-
- out->informative.qos.max_non_urgent_latency_us = mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->mp.qos_param_index].maximum_latency_when_non_urgent_uclk_cycles
- / mode_lib->mp.uclk_freq_mhz * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_max_latency_margin / 100.0)
- + mode_lib->soc.qos_parameters.qos_params.dcn4x.mall_overhead_fclk_cycles / mode_lib->mp.FabricClock
- + mode_lib->soc.qos_parameters.qos_params.dcn4x.max_round_trip_to_furthest_cs_fclk_cycles / mode_lib->mp.FabricClock
- * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin / 100.0);
+ out->informative.qos.max_non_urgent_latency_us = dml_get_max_non_urgent_latency_us(mode_lib);
if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn4x) {
if (((mode_lib->ip.rob_buffer_size_kbytes - mode_lib->ip.pixel_chunk_size_kbytes) * 1024
- / mode_lib->mp.non_urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp]) >= out->informative.qos.max_non_urgent_latency_us) {
+ / mode_lib->ms.support.non_urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp]) >= out->informative.qos.max_non_urgent_latency_us) {
out->informative.misc.ROBUrgencyAvoidance = true;
} else {
out->informative.misc.ROBUrgencyAvoidance = false;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.h
index df2d1550a14b..27ef0e096b25 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_dcn4_calcs.h
@@ -28,7 +28,7 @@ void dml2_core_calcs_get_plane_support_info(const struct dml2_display_cfg *displ
void dml2_core_calcs_get_informative(const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_display_cfg_programming *out);
void dml2_core_calcs_get_stream_support_info(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, struct core_stream_support_info *out, int plane_index);
void dml2_core_calcs_get_mall_allocation(struct dml2_core_internal_display_mode_lib *mode_lib, unsigned int *out, int pipe_index);
-void dml2_core_calcs_get_stream_fams2_programming(const struct dml2_core_internal_display_mode_lib *mode_lib, const struct display_configuation_with_meta *display_cfg, struct dmub_fams2_stream_static_state *fams2_programming, enum dml2_uclk_pstate_support_method pstate_method, int plane_index);
+void dml2_core_calcs_get_stream_fams2_programming(const struct dml2_core_internal_display_mode_lib *mode_lib, const struct display_configuation_with_meta *display_cfg, union dmub_cmd_fams2_config *fams2_base_programming, union dmub_cmd_fams2_config *fams2_sub_programming, enum dml2_pstate_method pstate_method, int plane_index);
void dml2_core_calcs_get_global_fams2_programming(const struct dml2_core_internal_display_mode_lib *mode_lib, const struct display_configuation_with_meta *display_cfg, struct dmub_cmd_fams2_global_config *fams2_global_config);
void dml2_core_calcs_get_dpte_row_height(unsigned int *dpte_row_height, struct dml2_core_internal_display_mode_lib *mode_lib, bool is_plane1, enum dml2_source_format_class SourcePixelFormat, enum dml2_swizzle_mode SurfaceTiling, enum dml2_rotation_angle ScanDirection, unsigned int pitch, unsigned int GPUVMMinPageSizeKBytes);
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_shared.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_shared.c
deleted file mode 100644
index 8f3c1c0b1cc1..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_shared.c
+++ /dev/null
@@ -1,12413 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-
-#include "dml2_internal_shared_types.h"
-#include "dml2_core_shared.h"
-#include "dml2_debug.h"
-#include "lib_float_math.h"
-
-double dml2_core_shared_div_rem(double dividend, unsigned int divisor, unsigned int *remainder)
-{
- *remainder = ((dividend / divisor) - (int)(dividend / divisor) > 0);
- return dividend / divisor;
-
-}
-
-/*
- * START OF STATIC HELPERS
- * These static methods are baseline implemenations from DCN4. These should NEVER
- * be modified when developing new DCNs. New DCN code should replace the static helpers
- * using the function pointer pattern.
- */
-
-static void dml2_print_dml_mode_support_info(const struct dml2_core_internal_mode_support_info *support, bool fail_only);
-static void get_stream_output_bpp(double *out_bpp, const struct dml2_display_cfg *display_cfg);
-static unsigned int dml_round_to_multiple(unsigned int num, unsigned int multiple, bool up);
-static unsigned int dml_get_num_active_pipes(int unsigned num_planes, const struct core_display_cfg_support_info *cfg_support_info);
-static void dml_calc_pipe_plane_mapping(const struct core_display_cfg_support_info *cfg_support_info, unsigned int *pipe_plane);
-static bool dml_is_phantom_pipe(const struct dml2_plane_parameters *plane_cfg);
-static bool dml_get_is_phantom_pipe(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, unsigned int pipe_idx);
-static void CalculateMaxDETAndMinCompressedBufferSize(unsigned int ConfigReturnBufferSizeInKByte,
- unsigned int ConfigReturnBufferSegmentSizeInKByte,
- unsigned int ROBBufferSizeInKByte,
- unsigned int MaxNumDPP,
- unsigned int nomDETInKByteOverrideEnable, // VBA_DELTA, allow DV to override default DET size
- unsigned int nomDETInKByteOverrideValue, // VBA_DELTA
- bool is_mrq_present,
-
- // Output
- unsigned int *MaxTotalDETInKByte,
- unsigned int *nomDETInKByte,
- unsigned int *MinCompressedBufferSizeInKByte);
- static void PixelClockAdjustmentForProgressiveToInterlaceUnit(const struct dml2_display_cfg *display_cfg, bool ptoi_supported, double *PixelClockBackEnd);
-static unsigned int dml_get_tile_block_size_bytes(enum dml2_swizzle_mode sw_mode);
-static bool dml_is_vertical_rotation(enum dml2_rotation_angle Scan);
-static int unsigned dml_get_gfx_version(enum dml2_swizzle_mode sw_mode);
-static void CalculateBytePerPixelAndBlockSizes(enum dml2_source_format_class SourcePixelFormat,
- enum dml2_swizzle_mode SurfaceTiling,
- unsigned int pitch_y,
- unsigned int pitch_c,
-
- // Output
- unsigned int *BytePerPixelY,
- unsigned int *BytePerPixelC,
- double *BytePerPixelDETY,
- double *BytePerPixelDETC,
- unsigned int *BlockHeight256BytesY,
- unsigned int *BlockHeight256BytesC,
- unsigned int *BlockWidth256BytesY,
- unsigned int *BlockWidth256BytesC,
- unsigned int *MacroTileHeightY,
- unsigned int *MacroTileHeightC,
- unsigned int *MacroTileWidthY,
- unsigned int *MacroTileWidthC,
- bool *surf_linear128_l,
- bool *surf_linear128_c);
-static void CalculateSinglePipeDPPCLKAndSCLThroughput(
- double HRatio,
- double HRatioChroma,
- double VRatio,
- double VRatioChroma,
- double MaxDCHUBToPSCLThroughput,
- double MaxPSCLToLBThroughput,
- double PixelClock,
- enum dml2_source_format_class SourcePixelFormat,
- unsigned int HTaps,
- unsigned int HTapsChroma,
- unsigned int VTaps,
- unsigned int VTapsChroma,
-
- // Output
- double *PSCL_THROUGHPUT,
- double *PSCL_THROUGHPUT_CHROMA,
- double *DPPCLKUsingSingleDPP);
-static void CalculateSwathWidth(
- const struct dml2_display_cfg *display_cfg,
- bool ForceSingleDPP,
- unsigned int NumberOfActiveSurfaces,
- enum dml2_odm_mode ODMMode[],
- unsigned int BytePerPixY[],
- unsigned int BytePerPixC[],
- unsigned int Read256BytesBlockHeightY[],
- unsigned int Read256BytesBlockHeightC[],
- unsigned int Read256BytesBlockWidthY[],
- unsigned int Read256BytesBlockWidthC[],
- bool surf_linear128_l[],
- bool surf_linear128_c[],
- unsigned int DPPPerSurface[],
-
- // Output
- unsigned int req_per_swath_ub_l[],
- unsigned int req_per_swath_ub_c[],
- unsigned int SwathWidthSingleDPPY[],
- unsigned int SwathWidthSingleDPPC[],
- unsigned int SwathWidthY[], // per-pipe
- unsigned int SwathWidthC[], // per-pipe
- unsigned int MaximumSwathHeightY[],
- unsigned int MaximumSwathHeightC[],
- unsigned int swath_width_luma_ub[], // per-pipe
- unsigned int swath_width_chroma_ub[]); // per-pipe
-static bool UnboundedRequest(bool unb_req_force_en, bool unb_req_force_val, unsigned int TotalNumberOfActiveDPP, bool NoChromaOrLinear);
-static void CalculateDETBufferSize(struct dml2_core_shared_calculate_det_buffer_size_params *p);
-static double CalculateRequiredDispclk(enum dml2_odm_mode ODMMode, double PixelClock);
-static double TruncToValidBPP(
- struct dml2_core_shared_TruncToValidBPP_locals *l,
- double LinkBitRate,
- unsigned int Lanes,
- unsigned int HTotal,
- unsigned int HActive,
- double PixelClock,
- double DesiredBPP,
- bool DSCEnable,
- enum dml2_output_encoder_class Output,
- enum dml2_output_format_class Format,
- unsigned int DSCInputBitPerComponent,
- unsigned int DSCSlices,
- unsigned int AudioRate,
- unsigned int AudioLayout,
- enum dml2_odm_mode ODMModeNoDSC,
- enum dml2_odm_mode ODMModeDSC,
-
- // Output
- unsigned int *RequiredSlots);
-static unsigned int dscceComputeDelay(
- unsigned int bpc,
- double BPP,
- unsigned int sliceWidth,
- unsigned int numSlices,
- enum dml2_output_format_class pixelFormat,
- enum dml2_output_encoder_class Output);
-static unsigned int dscComputeDelay(enum dml2_output_format_class pixelFormat, enum dml2_output_encoder_class Output);
-static unsigned int CalculateHostVMDynamicLevels(bool GPUVMEnable, bool HostVMEnable, unsigned int HostVMMinPageSize, unsigned int HostVMMaxNonCachedPageTableLevels);
-static unsigned int CalculateVMAndRowBytes(struct dml2_core_shared_calculate_vm_and_row_bytes_params *p);
-static unsigned int CalculatePrefetchSourceLines(
- double VRatio,
- unsigned int VTaps,
- bool Interlace,
- bool ProgressiveToInterlaceUnitInOPP,
- unsigned int SwathHeight,
- enum dml2_rotation_angle RotationAngle,
- bool mirrored,
- bool ViewportStationary,
- unsigned int SwathWidth,
- unsigned int ViewportHeight,
- unsigned int ViewportXStart,
- unsigned int ViewportYStart,
-
- // Output
- unsigned int *VInitPreFill,
- unsigned int *MaxNumSwath);
-static void CalculateRowBandwidth(
- bool GPUVMEnable,
- bool use_one_row_for_frame,
- enum dml2_source_format_class SourcePixelFormat,
- double VRatio,
- double VRatioChroma,
- bool DCCEnable,
- double LineTime,
- unsigned int PixelPTEBytesPerRowLuma,
- unsigned int PixelPTEBytesPerRowChroma,
- unsigned int dpte_row_height_luma,
- unsigned int dpte_row_height_chroma,
-
- bool mrq_present,
- unsigned int meta_row_bytes_per_row_ub_l,
- unsigned int meta_row_bytes_per_row_ub_c,
- unsigned int meta_row_height_luma,
- unsigned int meta_row_height_chroma,
-
- // Output
- double *dpte_row_bw,
- double *meta_row_bw);
-static void CalculateMALLUseForStaticScreen(
- const struct dml2_display_cfg *display_cfg,
- unsigned int NumberOfActiveSurfaces,
- unsigned int MALLAllocatedForDCN,
- unsigned int SurfaceSizeInMALL[],
- bool one_row_per_frame_fits_in_buffer[],
-
- // Output
- bool is_using_mall_for_ss[]);
-static void CalculateDCCConfiguration(
- bool DCCEnabled,
- bool DCCProgrammingAssumesScanDirectionUnknown,
- enum dml2_source_format_class SourcePixelFormat,
- unsigned int SurfaceWidthLuma,
- unsigned int SurfaceWidthChroma,
- unsigned int SurfaceHeightLuma,
- unsigned int SurfaceHeightChroma,
- unsigned int nomDETInKByte,
- unsigned int RequestHeight256ByteLuma,
- unsigned int RequestHeight256ByteChroma,
- enum dml2_swizzle_mode TilingFormat,
- unsigned int BytePerPixelY,
- unsigned int BytePerPixelC,
- double BytePerPixelDETY,
- double BytePerPixelDETC,
- enum dml2_rotation_angle RotationAngle,
-
- // Output
- enum dml2_core_internal_request_type *RequestLuma,
- enum dml2_core_internal_request_type *RequestChroma,
- unsigned int *MaxUncompressedBlockLuma,
- unsigned int *MaxUncompressedBlockChroma,
- unsigned int *MaxCompressedBlockLuma,
- unsigned int *MaxCompressedBlockChroma,
- unsigned int *IndependentBlockLuma,
- unsigned int *IndependentBlockChroma);
-static void calculate_mcache_row_bytes(struct dml2_core_internal_scratch *scratch, struct dml2_core_calcs_calculate_mcache_row_bytes_params *p);
-static void calculate_mcache_setting(struct dml2_core_internal_scratch *scratch, struct dml2_core_calcs_calculate_mcache_setting_params *p);
-static void calculate_mall_bw_overhead_factor(
- double mall_prefetch_sdp_overhead_factor[],
- double mall_prefetch_dram_overhead_factor[],
-
- // input
- const struct dml2_display_cfg *display_cfg,
- unsigned int num_active_planes);
-static double dml_get_return_bandwidth_available(
- const struct dml2_soc_bb *soc,
- enum dml2_core_internal_soc_state_type state_type,
- enum dml2_core_internal_bw_type bw_type,
- bool is_avg_bw,
- bool is_hvm_en,
- bool is_hvm_only,
- double dcflk_mhz,
- double fclk_mhz,
- double dram_bw_mbps);
-static void calculate_bandwidth_available(
- double avg_bandwidth_available_min[dml2_core_internal_soc_state_max],
- double avg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_available_min[dml2_core_internal_soc_state_max], // min between SDP and DRAM
- double urg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_available_vm_only[dml2_core_internal_soc_state_max],
- double urg_bandwidth_available_pixel_and_vm[dml2_core_internal_soc_state_max],
-
- const struct dml2_soc_bb *soc,
- bool HostVMEnable,
- double dcfclk_mhz,
- double fclk_mhz,
- double dram_bw_mbps);
-static void calculate_avg_bandwidth_required(
- double avg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
-
- // input
- const struct dml2_display_cfg *display_cfg,
- unsigned int num_active_planes,
- double ReadBandwidthLuma[],
- double ReadBandwidthChroma[],
- double cursor_bw[],
- double dcc_dram_bw_nom_overhead_factor_p0[],
- double dcc_dram_bw_nom_overhead_factor_p1[],
- double mall_prefetch_dram_overhead_factor[],
- double mall_prefetch_sdp_overhead_factor[]);
-static void CalculateVMRowAndSwath(struct dml2_core_internal_scratch *scratch, struct dml2_core_calcs_CalculateVMRowAndSwath_params *p);
-static double CalculateUrgentLatency(
- double UrgentLatencyPixelDataOnly,
- double UrgentLatencyPixelMixedWithVMData,
- double UrgentLatencyVMDataOnly,
- bool DoUrgentLatencyAdjustment,
- double UrgentLatencyAdjustmentFabricClockComponent,
- double UrgentLatencyAdjustmentFabricClockReference,
- double FabricClock,
- double uclk_freq_mhz,
- enum dml2_qos_param_type qos_type,
- unsigned int urgent_ramp_uclk_cycles,
- unsigned int df_qos_response_time_fclk_cycles,
- unsigned int max_round_trip_to_furthest_cs_fclk_cycles,
- unsigned int mall_overhead_fclk_cycles,
- double umc_urgent_ramp_latency_margin,
- double fabric_max_transport_latency_margin);
-static double CalculateTripToMemory(
- double UrgLatency,
- double FabricClock,
- double uclk_freq_mhz,
- enum dml2_qos_param_type qos_type,
- unsigned int trip_to_memory_uclk_cycles,
- unsigned int max_round_trip_to_furthest_cs_fclk_cycles,
- unsigned int mall_overhead_fclk_cycles,
- double umc_max_latency_margin,
- double fabric_max_transport_latency_margin);
-static double CalculateMetaTripToMemory(
- double UrgLatency,
- double FabricClock,
- double uclk_freq_mhz,
- enum dml2_qos_param_type qos_type,
- unsigned int meta_trip_to_memory_uclk_cycles,
- unsigned int meta_trip_to_memory_fclk_cycles,
- double umc_max_latency_margin,
- double fabric_max_transport_latency_margin);
-static void calculate_cursor_req_attributes(
- unsigned int cursor_width,
- unsigned int cursor_bpp,
-
- // output
- unsigned int *cursor_lines_per_chunk,
- unsigned int *cursor_bytes_per_line,
- unsigned int *cursor_bytes_per_chunk,
- unsigned int *cursor_bytes);
-static void calculate_cursor_urgent_burst_factor(
- unsigned int CursorBufferSize,
- unsigned int CursorWidth,
- unsigned int cursor_bytes_per_chunk,
- unsigned int cursor_lines_per_chunk,
- double LineTime,
- double UrgentLatency,
-
- double *UrgentBurstFactorCursor,
- bool *NotEnoughUrgentLatencyHiding);
-static void CalculateUrgentBurstFactor(
- const struct dml2_plane_parameters *plane_cfg,
- unsigned int swath_width_luma_ub,
- unsigned int swath_width_chroma_ub,
- unsigned int SwathHeightY,
- unsigned int SwathHeightC,
- double LineTime,
- double UrgentLatency,
- double VRatio,
- double VRatioC,
- double BytePerPixelInDETY,
- double BytePerPixelInDETC,
- unsigned int DETBufferSizeY,
- unsigned int DETBufferSizeC,
- // Output
- double *UrgentBurstFactorLuma,
- double *UrgentBurstFactorChroma,
- bool *NotEnoughUrgentLatencyHiding);
-static void CalculateDCFCLKDeepSleep(
- const struct dml2_display_cfg *display_cfg,
- unsigned int NumberOfActiveSurfaces,
- unsigned int BytePerPixelY[],
- unsigned int BytePerPixelC[],
- unsigned int SwathWidthY[],
- unsigned int SwathWidthC[],
- unsigned int DPPPerSurface[],
- double PSCL_THROUGHPUT[],
- double PSCL_THROUGHPUT_CHROMA[],
- double Dppclk[],
- double ReadBandwidthLuma[],
- double ReadBandwidthChroma[],
- unsigned int ReturnBusWidth,
-
- // Output
- double *DCFClkDeepSleep);
-static double CalculateWriteBackDelay(
- enum dml2_source_format_class WritebackPixelFormat,
- double WritebackHRatio,
- double WritebackVRatio,
- unsigned int WritebackVTaps,
- unsigned int WritebackDestinationWidth,
- unsigned int WritebackDestinationHeight,
- unsigned int WritebackSourceHeight,
- unsigned int HTotal);
-static unsigned int CalculateMaxVStartup(
- bool ptoi_supported,
- unsigned int vblank_nom_default_us,
- const struct dml2_timing_cfg *timing,
- double write_back_delay_us);
-static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch *scratch, struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params *p);
-static void CalculateODMMode(
- unsigned int MaximumPixelsPerLinePerDSCUnit,
- unsigned int HActive,
- enum dml2_output_encoder_class Output,
- enum dml2_odm_mode ODMUse,
- double MaxDispclk,
- bool DSCEnable,
- unsigned int TotalNumberOfActiveDPP,
- unsigned int MaxNumDPP,
- double PixelClock,
-
- // Output
- bool *TotalAvailablePipesSupport,
- unsigned int *NumberOfDPP,
- enum dml2_odm_mode *ODMMode,
- double *RequiredDISPCLKPerSurface);
-static void CalculateOutputLink(
- struct dml2_core_internal_scratch *s,
- double PHYCLK,
- double PHYCLKD18,
- double PHYCLKD32,
- double Downspreading,
- bool IsMainSurfaceUsingTheIndicatedTiming,
- enum dml2_output_encoder_class Output,
- enum dml2_output_format_class OutputFormat,
- unsigned int HTotal,
- unsigned int HActive,
- double PixelClockBackEnd,
- double ForcedOutputLinkBPP,
- unsigned int DSCInputBitPerComponent,
- unsigned int NumberOfDSCSlices,
- double AudioSampleRate,
- unsigned int AudioSampleLayout,
- enum dml2_odm_mode ODMModeNoDSC,
- enum dml2_odm_mode ODMModeDSC,
- enum dml2_dsc_enable_option DSCEnable,
- unsigned int OutputLinkDPLanes,
- enum dml2_output_link_dp_rate OutputLinkDPRate,
-
- // Output
- bool *RequiresDSC,
- bool *RequiresFEC,
- double *OutBpp,
- enum dml2_core_internal_output_type *OutputType,
- enum dml2_core_internal_output_type_rate *OutputRate,
- unsigned int *RequiredSlots);
-static double CalculateWriteBackDISPCLK(
- enum dml2_source_format_class WritebackPixelFormat,
- double PixelClock,
- double WritebackHRatio,
- double WritebackVRatio,
- unsigned int WritebackHTaps,
- unsigned int WritebackVTaps,
- unsigned int WritebackSourceWidth,
- unsigned int WritebackDestinationWidth,
- unsigned int HTotal,
- unsigned int WritebackLineBufferSize);
-static double RequiredDTBCLK(
- bool DSCEnable,
- double PixelClock,
- enum dml2_output_format_class OutputFormat,
- double OutputBpp,
- unsigned int DSCSlices,
- unsigned int HTotal,
- unsigned int HActive,
- unsigned int AudioRate,
- unsigned int AudioLayout);
-static unsigned int DSCDelayRequirement(
- bool DSCEnabled,
- enum dml2_odm_mode ODMMode,
- unsigned int DSCInputBitPerComponent,
- double OutputBpp,
- unsigned int HActive,
- unsigned int HTotal,
- unsigned int NumberOfDSCSlices,
- enum dml2_output_format_class OutputFormat,
- enum dml2_output_encoder_class Output,
- double PixelClock,
- double PixelClockBackEnd);
-static void CalculateSurfaceSizeInMall(
- const struct dml2_display_cfg *display_cfg,
- unsigned int NumberOfActiveSurfaces,
- unsigned int MALLAllocatedForDCN,
- unsigned int BytesPerPixelY[],
- unsigned int BytesPerPixelC[],
- unsigned int Read256BytesBlockWidthY[],
- unsigned int Read256BytesBlockWidthC[],
- unsigned int Read256BytesBlockHeightY[],
- unsigned int Read256BytesBlockHeightC[],
- unsigned int ReadBlockWidthY[],
- unsigned int ReadBlockWidthC[],
- unsigned int ReadBlockHeightY[],
- unsigned int ReadBlockHeightC[],
-
- // Output
- unsigned int SurfaceSizeInMALL[],
- bool *ExceededMALLSize);
-static void calculate_tdlut_setting(struct dml2_core_internal_scratch *scratch, struct dml2_core_calcs_calculate_tdlut_setting_params *p);
-static void CalculateTarb(
- const struct dml2_display_cfg *display_cfg,
- unsigned int PixelChunkSizeInKByte,
- unsigned int NumberOfActiveSurfaces,
- unsigned int NumberOfDPP[],
- unsigned int dpte_group_bytes[],
- unsigned int tdlut_bytes_per_group[],
- double HostVMInefficiencyFactor,
- double HostVMInefficiencyFactorPrefetch,
- unsigned int HostVMMinPageSize,
- double ReturnBW,
-
- unsigned int MetaChunkSize,
-
- // output
- double *Tarb,
- double *Tarb_prefetch);
-static double CalculateTWait(long reserved_vblank_time_ns, double UrgentLatency, double Ttrip);
-static void CalculateVUpdateAndDynamicMetadataParameters(
- unsigned int MaxInterDCNTileRepeaters,
- double Dppclk,
- double Dispclk,
- double DCFClkDeepSleep,
- double PixelClock,
- unsigned int HTotal,
- unsigned int VBlank,
- unsigned int DynamicMetadataTransmittedBytes,
- unsigned int DynamicMetadataLinesBeforeActiveRequired,
- unsigned int InterlaceEnable,
- bool ProgressiveToInterlaceUnitInOPP,
-
- // Output
- double *TSetup,
- double *Tdmbf,
- double *Tdmec,
- double *Tdmsks,
- unsigned int *VUpdateOffsetPix,
- unsigned int *VUpdateWidthPix,
- unsigned int *VReadyOffsetPix);
-static double get_urgent_bandwidth_required(
- struct dml2_core_shared_get_urgent_bandwidth_required_locals *l,
- const struct dml2_display_cfg *display_cfg,
- enum dml2_core_internal_soc_state_type state_type,
- enum dml2_core_internal_bw_type bw_type,
- bool inc_flip_bw, // including flip bw
- unsigned int NumberOfActiveSurfaces,
- unsigned int NumberOfDPP[],
- double dcc_dram_bw_nom_overhead_factor_p0[],
- double dcc_dram_bw_nom_overhead_factor_p1[],
- double dcc_dram_bw_pref_overhead_factor_p0[],
- double dcc_dram_bw_pref_overhead_factor_p1[],
- double mall_prefetch_sdp_overhead_factor[],
- double mall_prefetch_dram_overhead_factor[],
- double ReadBandwidthLuma[],
- double ReadBandwidthChroma[],
- double PrefetchBandwidthLuma[],
- double PrefetchBandwidthChroma[],
- double cursor_bw[],
- double dpte_row_bw[],
- double meta_row_bw[],
- double prefetch_cursor_bw[],
- double prefetch_vmrow_bw[],
- double flip_bw[],
- double UrgentBurstFactorLuma[],
- double UrgentBurstFactorChroma[],
- double UrgentBurstFactorCursor[],
- double UrgentBurstFactorLumaPre[],
- double UrgentBurstFactorChromaPre[],
- double UrgentBurstFactorCursorPre[]);
-static void CalculateExtraLatency(
- const struct dml2_display_cfg *display_cfg,
- unsigned int ROBBufferSizeInKByte,
- unsigned int RoundTripPingLatencyCycles,
- unsigned int ReorderingBytes,
- double DCFCLK,
- double FabricClock,
- unsigned int PixelChunkSizeInKByte,
- double ReturnBW,
- unsigned int NumberOfActiveSurfaces,
- unsigned int NumberOfDPP[],
- unsigned int dpte_group_bytes[],
- unsigned int tdlut_bytes_per_group[],
- double HostVMInefficiencyFactor,
- double HostVMInefficiencyFactorPrefetch,
- unsigned int HostVMMinPageSize,
- enum dml2_qos_param_type qos_type,
- bool max_oustanding_when_urgent_expected,
- unsigned int max_outstanding_requests,
- unsigned int request_size_bytes_luma[],
- unsigned int request_size_bytes_chroma[],
- unsigned int MetaChunkSize,
- unsigned int dchub_arb_to_ret_delay,
- double Ttrip,
- unsigned int hostvm_mode,
-
- // output
- double *ExtraLatency, // Tex
- double *ExtraLatency_sr, // Tex_sr
- double *ExtraLatencyPrefetch);
-static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch, struct dml2_core_calcs_CalculatePrefetchSchedule_params *p);
-static void calculate_peak_bandwidth_required(
- struct dml2_core_internal_scratch *s,
-
- // output
- double urg_vactive_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double non_urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
-
- // input
- const struct dml2_display_cfg *display_cfg,
- unsigned int inc_flip_bw,
- unsigned int NumberOfActiveSurfaces,
- unsigned int NumberOfDPP[],
- double dcc_dram_bw_nom_overhead_factor_p0[],
- double dcc_dram_bw_nom_overhead_factor_p1[],
- double dcc_dram_bw_pref_overhead_factor_p0[],
- double dcc_dram_bw_pref_overhead_factor_p1[],
- double mall_prefetch_sdp_overhead_factor[],
- double mall_prefetch_dram_overhead_factor[],
- double ReadBandwidthLuma[],
- double ReadBandwidthChroma[],
- double PrefetchBandwidthLuma[],
- double PrefetchBandwidthChroma[],
- double cursor_bw[],
- double dpte_row_bw[],
- double meta_row_bw[],
- double prefetch_cursor_bw[],
- double prefetch_vmrow_bw[],
- double flip_bw[],
- double UrgentBurstFactorLuma[],
- double UrgentBurstFactorChroma[],
- double UrgentBurstFactorCursor[],
- double UrgentBurstFactorLumaPre[],
- double UrgentBurstFactorChromaPre[],
- double UrgentBurstFactorCursorPre[]);
-static void check_urgent_bandwidth_support(
- double *frac_urg_bandwidth_nom,
- double *frac_urg_bandwidth_mall,
- bool *vactive_bandwidth_support_ok, // vactive ok
- bool *bandwidth_support_ok, // max of vm, prefetch, vactive all ok
-
- unsigned int mall_allocated_for_dcn_mbytes,
- double non_urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_vactive_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max]);
-static double get_bandwidth_available_for_immediate_flip(
- enum dml2_core_internal_soc_state_type eval_state,
- double urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max], // no flip
- double urg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max]);
-static void calculate_immediate_flip_bandwidth_support(
- // Output
- double *frac_urg_bandwidth_flip,
- bool *flip_bandwidth_support_ok,
-
- // Input
- enum dml2_core_internal_soc_state_type eval_state,
- double urg_bandwidth_required_flip[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double non_urg_bandwidth_required_flip[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max]);
-static void CalculateFlipSchedule(
- struct dml2_core_internal_scratch *s,
- bool iflip_enable,
- bool use_lb_flip_bw,
- double HostVMInefficiencyFactor,
- double Tvm_trips_flip,
- double Tr0_trips_flip,
- double Tvm_trips_flip_rounded,
- double Tr0_trips_flip_rounded,
- bool GPUVMEnable,
- double vm_bytes, // vm_bytes
- double DPTEBytesPerRow, // dpte_row_bytes
- double BandwidthAvailableForImmediateFlip,
- unsigned int TotImmediateFlipBytes,
- enum dml2_source_format_class SourcePixelFormat,
- double LineTime,
- double VRatio,
- double VRatioChroma,
- double Tno_bw_flip,
- unsigned int dpte_row_height,
- unsigned int dpte_row_height_chroma,
- bool use_one_row_for_frame_flip,
- unsigned int max_flip_time_us,
- unsigned int per_pipe_flip_bytes,
- unsigned int meta_row_bytes,
- unsigned int meta_row_height,
- unsigned int meta_row_height_chroma,
- bool dcc_mrq_enable,
-
- // Output
- double *dst_y_per_vm_flip,
- double *dst_y_per_row_flip,
- double *final_flip_bw,
- bool *ImmediateFlipSupportedForPipe);
-static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
- struct dml2_core_internal_scratch *scratch,
- struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params *p);
-static double uclk_khz_to_dram_bw_mbps(unsigned long uclk_khz, const struct dml2_dram_params *dram_config);
-static double dram_bw_kbps_to_uclk_mhz(unsigned long long bw_kbps, const struct dml2_dram_params *dram_config);
-static unsigned int get_qos_param_index(unsigned long uclk_freq_khz, const struct dml2_dcn4_uclk_dpm_dependent_qos_params *per_uclk_dpm_params);
-static unsigned int get_active_min_uclk_dpm_index(unsigned long uclk_freq_khz, const struct dml2_soc_state_table *clk_table);
-static unsigned int get_pipe_flip_bytes(
- double hostvm_inefficiency_factor,
- unsigned int vm_bytes,
- unsigned int dpte_row_bytes,
- unsigned int meta_row_bytes);
-static void calculate_hostvm_inefficiency_factor(
- double *HostVMInefficiencyFactor,
- double *HostVMInefficiencyFactorPrefetch,
-
- bool gpuvm_enable,
- bool hostvm_enable,
- unsigned int remote_iommu_outstanding_translations,
- unsigned int max_outstanding_reqs,
- double urg_bandwidth_avail_active_pixel_and_vm,
- double urg_bandwidth_avail_active_vm_only);
-static void CalculatePixelDeliveryTimes(
- const struct dml2_display_cfg *display_cfg,
- const struct core_display_cfg_support_info *cfg_support_info,
- unsigned int NumberOfActiveSurfaces,
- double VRatioPrefetchY[],
- double VRatioPrefetchC[],
- unsigned int swath_width_luma_ub[],
- unsigned int swath_width_chroma_ub[],
- double PSCL_THROUGHPUT[],
- double PSCL_THROUGHPUT_CHROMA[],
- double Dppclk[],
- unsigned int BytePerPixelC[],
- unsigned int req_per_swath_ub_l[],
- unsigned int req_per_swath_ub_c[],
-
- // Output
- double DisplayPipeLineDeliveryTimeLuma[],
- double DisplayPipeLineDeliveryTimeChroma[],
- double DisplayPipeLineDeliveryTimeLumaPrefetch[],
- double DisplayPipeLineDeliveryTimeChromaPrefetch[],
- double DisplayPipeRequestDeliveryTimeLuma[],
- double DisplayPipeRequestDeliveryTimeChroma[],
- double DisplayPipeRequestDeliveryTimeLumaPrefetch[],
- double DisplayPipeRequestDeliveryTimeChromaPrefetch[]);
-static void CalculateMetaAndPTETimes(struct dml2_core_shared_CalculateMetaAndPTETimes_params *p);
-static void CalculateVMGroupAndRequestTimes(
- const struct dml2_display_cfg *display_cfg,
- unsigned int NumberOfActiveSurfaces,
- unsigned int BytePerPixelC[],
- double dst_y_per_vm_vblank[],
- double dst_y_per_vm_flip[],
- unsigned int dpte_row_width_luma_ub[],
- unsigned int dpte_row_width_chroma_ub[],
- unsigned int vm_group_bytes[],
- unsigned int dpde0_bytes_per_frame_ub_l[],
- unsigned int dpde0_bytes_per_frame_ub_c[],
- unsigned int tdlut_pte_bytes_per_frame[],
- unsigned int meta_pte_bytes_per_frame_ub_l[],
- unsigned int meta_pte_bytes_per_frame_ub_c[],
- bool mrq_present,
-
- // Output
- double TimePerVMGroupVBlank[],
- double TimePerVMGroupFlip[],
- double TimePerVMRequestVBlank[],
- double TimePerVMRequestFlip[]);
-static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratch, struct dml2_core_calcs_CalculateStutterEfficiency_params *p);
-static bool dml_is_dual_plane(enum dml2_source_format_class source_format);
-static unsigned int dml_get_plane_idx(const struct dml2_core_internal_display_mode_lib *mode_lib, unsigned int pipe_idx);
-static void rq_dlg_get_wm_regs(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_dchub_watermark_regs *wm_regs);
-static unsigned int log_and_substract_if_non_zero(unsigned int a, unsigned int subtrahend);
-static void rq_dlg_get_rq_reg(struct dml2_display_rq_regs *rq_regs,
- const struct dml2_display_cfg *display_cfg,
- const struct dml2_core_internal_display_mode_lib *mode_lib,
- unsigned int pipe_idx);
-static void rq_dlg_get_dlg_reg(struct dml2_core_internal_scratch *s,
- struct dml2_display_dlg_regs *disp_dlg_regs,
- struct dml2_display_ttu_regs *disp_ttu_regs,
- const struct dml2_display_cfg *display_cfg,
- const struct dml2_core_internal_display_mode_lib *mode_lib,
- const unsigned int pipe_idx);
-static void rq_dlg_get_arb_params(const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_display_arb_regs *arb_param);
-
-/*
- * END OF STATIC HELPERS
- */
-
-bool dml2_core_shared_mode_support(struct dml2_core_calcs_mode_support_ex *in_out_params)
-{
- struct dml2_core_internal_display_mode_lib *mode_lib = in_out_params->mode_lib;
- const struct dml2_display_cfg *display_cfg = in_out_params->in_display_cfg;
- const struct dml2_mcg_min_clock_table *min_clk_table = in_out_params->min_clk_table;
-
- struct dml2_core_calcs_mode_support_locals *s = &mode_lib->scratch.dml_core_mode_support_locals;
- struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params *CalculateWatermarks_params = &mode_lib->scratch.CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params;
- struct dml2_core_calcs_CalculateVMRowAndSwath_params *CalculateVMRowAndSwath_params = &mode_lib->scratch.CalculateVMRowAndSwath_params;
- struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params *CalculateSwathAndDETConfiguration_params = &mode_lib->scratch.CalculateSwathAndDETConfiguration_params;
- struct dml2_core_calcs_CalculatePrefetchSchedule_params *CalculatePrefetchSchedule_params = &mode_lib->scratch.CalculatePrefetchSchedule_params;
- struct dml2_core_calcs_calculate_tdlut_setting_params *calculate_tdlut_setting_params = &mode_lib->scratch.calculate_tdlut_setting_params;
- struct dml2_core_calcs_calculate_mcache_setting_params *calculate_mcache_setting_params = &mode_lib->scratch.calculate_mcache_setting_params;
- unsigned int k, m, n;
-
- memset(&mode_lib->ms, 0, sizeof(struct dml2_core_internal_mode_support));
-
- mode_lib->ms.num_active_planes = display_cfg->num_planes;
- get_stream_output_bpp(s->OutputBpp, display_cfg);
-
- mode_lib->ms.state_idx = in_out_params->min_clk_index;
- mode_lib->ms.SOCCLK = ((double)mode_lib->soc.clk_table.socclk.clk_values_khz[0] / 1000);
- mode_lib->ms.DCFCLK = ((double)min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].min_dcfclk_khz / 1000);
- mode_lib->ms.FabricClock = ((double)min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].min_fclk_khz / 1000);
- mode_lib->ms.MaxDCFCLK = (double)min_clk_table->max_clocks_khz.dcfclk / 1000;
- mode_lib->ms.MaxFabricClock = (double)min_clk_table->max_clocks_khz.fclk / 1000;
- mode_lib->ms.max_dispclk_freq_mhz = (double)min_clk_table->max_clocks_khz.dispclk / 1000;
- mode_lib->ms.max_dscclk_freq_mhz = (double)min_clk_table->max_clocks_khz.dscclk / 1000;
- mode_lib->ms.max_dppclk_freq_mhz = (double)min_clk_table->max_clocks_khz.dppclk / 1000;
- mode_lib->ms.uclk_freq_mhz = dram_bw_kbps_to_uclk_mhz(min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].pre_derate_dram_bw_kbps, &mode_lib->soc.clk_table.dram_config);
- mode_lib->ms.dram_bw_mbps = ((double)min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].pre_derate_dram_bw_kbps / 1000);
- mode_lib->ms.qos_param_index = get_qos_param_index((unsigned int)(mode_lib->ms.uclk_freq_mhz * 1000.0), mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params);
- mode_lib->ms.active_min_uclk_dpm_index = get_active_min_uclk_dpm_index((unsigned int)(mode_lib->ms.uclk_freq_mhz * 1000.0), &mode_lib->soc.clk_table);
-
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: --- START --- \n", __func__);
- dml2_printf("DML::%s: num_active_planes = %u\n", __func__, mode_lib->ms.num_active_planes);
- dml2_printf("DML::%s: min_clk_index = %0d\n", __func__, in_out_params->min_clk_index);
- dml2_printf("DML::%s: qos_param_index = %0d\n", __func__, mode_lib->ms.qos_param_index);
- dml2_printf("DML::%s: SOCCLK = %f\n", __func__, mode_lib->ms.SOCCLK);
- dml2_printf("DML::%s: dram_bw_mbps = %f\n", __func__, mode_lib->ms.dram_bw_mbps);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, mode_lib->ms.uclk_freq_mhz);
- dml2_printf("DML::%s: DCFCLK = %f\n", __func__, mode_lib->ms.DCFCLK);
- dml2_printf("DML::%s: FabricClock = %f\n", __func__, mode_lib->ms.FabricClock);
- dml2_printf("DML::%s: MaxDCFCLK = %f\n", __func__, mode_lib->ms.MaxDCFCLK);
- dml2_printf("DML::%s: max_dispclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dispclk_freq_mhz);
- dml2_printf("DML::%s: max_dscclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dscclk_freq_mhz);
- dml2_printf("DML::%s: max_dppclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dppclk_freq_mhz);
- dml2_printf("DML::%s: MaxFabricClock = %f\n", __func__, mode_lib->ms.MaxFabricClock);
- dml2_printf("DML::%s: max_dscclk_freq_mhz = %f\n", __func__, mode_lib->ms.max_dscclk_freq_mhz);
- dml2_printf("DML::%s: ip.compressed_buffer_segment_size_in_kbytes = %u\n", __func__, mode_lib->ip.compressed_buffer_segment_size_in_kbytes);
- dml2_printf("DML::%s: ip.dcn_mrq_present = %u\n", __func__, mode_lib->ip.dcn_mrq_present);
-
- for (k = 0; k < mode_lib->ms.num_active_planes; k++)
- dml2_printf("DML::%s: plane_%d: reserved_vblank_time_ns = %u\n", __func__, k, display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns);
-
- // dml2_printf_dml_policy(&mode_lib->ms.policy);
- // dml2_printf_dml_display_cfg_timing(&display_cfg->timing, mode_lib->ms.num_active_planes);
- // dml2_printf_dml_display_cfg_plane(&display_cfg->plane, mode_lib->ms.num_active_planes);
- // dml2_printf_dml_display_cfg_surface(&display_cfg->surface, mode_lib->ms.num_active_planes);
- // dml2_printf_dml_display_cfg_output(&display_cfg->output, mode_lib->ms.num_active_planes);
-#endif
-
- CalculateMaxDETAndMinCompressedBufferSize(
- mode_lib->ip.config_return_buffer_size_in_kbytes,
- mode_lib->ip.config_return_buffer_segment_size_in_kbytes,
- mode_lib->ip.rob_buffer_size_kbytes,
- mode_lib->ip.max_num_dpp,
- display_cfg->overrides.hw.force_nom_det_size_kbytes.enable,
- display_cfg->overrides.hw.force_nom_det_size_kbytes.value,
- mode_lib->ip.dcn_mrq_present,
-
- /* Output */
- &mode_lib->ms.MaxTotalDETInKByte,
- &mode_lib->ms.NomDETInKByte,
- &mode_lib->ms.MinCompressedBufferSizeInKByte);
-
- PixelClockAdjustmentForProgressiveToInterlaceUnit(display_cfg, mode_lib->ip.ptoi_supported, s->PixelClockBackEnd);
-
- /*MODE SUPPORT, VOLTAGE STATE AND SOC CONFIGURATION*/
-
- /*Scale Ratio, taps Support Check*/
- mode_lib->ms.support.ScaleRatioAndTapsSupport = true;
- // Many core tests are still setting scaling parameters "incorrectly"
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->plane_descriptors[k].composition.scaler_info.enabled == false
- && (dml2_core_shared_is_420(display_cfg->plane_descriptors[k].pixel_format)
- || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio != 1.0
- || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_taps != 1.0
- || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio != 1.0
- || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps != 1.0)) {
- mode_lib->ms.support.ScaleRatioAndTapsSupport = false;
- } else if (display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps < 1.0 || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps > 8.0
- || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_taps < 1.0 || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_taps > 8.0
- || (display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_taps > 1.0 && (display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_taps % 2) == 1)
- || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio > mode_lib->ip.max_hscl_ratio
- || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio > mode_lib->ip.max_vscl_ratio
- || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio > display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_taps
- || display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio > display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps
- || (dml2_core_shared_is_420(display_cfg->plane_descriptors[k].pixel_format)
- && (display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps < 1 || display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps > 8 ||
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_taps < 1 || display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_taps > 8 ||
- (display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_taps > 1 && display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_taps % 2 == 1) ||
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio > mode_lib->ip.max_hscl_ratio ||
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio > mode_lib->ip.max_vscl_ratio ||
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio > display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_taps ||
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio > display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps))) {
- mode_lib->ms.support.ScaleRatioAndTapsSupport = false;
- }
- }
-
- /*Source Format, Pixel Format and Scan Support Check*/
- mode_lib->ms.support.SourceFormatPixelAndScanSupport = true;
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->plane_descriptors[k].surface.tiling == dml2_sw_linear && dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle)) {
- mode_lib->ms.support.SourceFormatPixelAndScanSupport = false;
- }
- }
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- CalculateBytePerPixelAndBlockSizes(
- display_cfg->plane_descriptors[k].pixel_format,
- display_cfg->plane_descriptors[k].surface.tiling,
- display_cfg->plane_descriptors[k].surface.plane0.pitch,
- display_cfg->plane_descriptors[k].surface.plane1.pitch,
-
- /* Output */
- &mode_lib->ms.BytePerPixelY[k],
- &mode_lib->ms.BytePerPixelC[k],
- &mode_lib->ms.BytePerPixelInDETY[k],
- &mode_lib->ms.BytePerPixelInDETC[k],
- &mode_lib->ms.Read256BlockHeightY[k],
- &mode_lib->ms.Read256BlockHeightC[k],
- &mode_lib->ms.Read256BlockWidthY[k],
- &mode_lib->ms.Read256BlockWidthC[k],
- &mode_lib->ms.MacroTileHeightY[k],
- &mode_lib->ms.MacroTileHeightC[k],
- &mode_lib->ms.MacroTileWidthY[k],
- &mode_lib->ms.MacroTileWidthC[k],
- &mode_lib->ms.surf_linear128_l[k],
- &mode_lib->ms.surf_linear128_c[k]);
- }
-
- /*Bandwidth Support Check*/
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (!dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle)) {
- mode_lib->ms.SwathWidthYSingleDPP[k] = display_cfg->plane_descriptors[k].composition.viewport.plane0.width;
- mode_lib->ms.SwathWidthCSingleDPP[k] = display_cfg->plane_descriptors[k].composition.viewport.plane1.width;
- } else {
- mode_lib->ms.SwathWidthYSingleDPP[k] = display_cfg->plane_descriptors[k].composition.viewport.plane0.height;
- mode_lib->ms.SwathWidthCSingleDPP[k] = display_cfg->plane_descriptors[k].composition.viewport.plane1.height;
- }
- }
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- mode_lib->ms.SurfaceReadBandwidthLuma[k] = mode_lib->ms.SwathWidthYSingleDPP[k] * math_ceil2(mode_lib->ms.BytePerPixelY[k], 1.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- mode_lib->ms.SurfaceReadBandwidthChroma[k] = mode_lib->ms.SwathWidthCSingleDPP[k] * math_ceil2(mode_lib->ms.BytePerPixelC[k], 2.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
-
- mode_lib->ms.cursor_bw[k] = display_cfg->plane_descriptors[k].cursor.num_cursors * display_cfg->plane_descriptors[k].cursor.cursor_width *
- display_cfg->plane_descriptors[k].cursor.cursor_bpp / 8.0 / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000));
-
-#ifdef __DML_VBA_DEBUG__
- double old_ReadBandwidthLuma = mode_lib->ms.SwathWidthYSingleDPP[k] * math_ceil2(mode_lib->ms.BytePerPixelInDETY[k], 1.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- double old_ReadBandwidthChroma = mode_lib->ms.SwathWidthYSingleDPP[k] / 2 * math_ceil2(mode_lib->ms.BytePerPixelInDETC[k], 2.0) / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio / 2.0;
- dml2_printf("DML::%s: k=%u, old_ReadBandwidthLuma = %f\n", __func__, k, old_ReadBandwidthLuma);
- dml2_printf("DML::%s: k=%u, old_ReadBandwidthChroma = %f\n", __func__, k, old_ReadBandwidthChroma);
- dml2_printf("DML::%s: k=%u, ReadBandwidthLuma = %f\n", __func__, k, mode_lib->ms.SurfaceReadBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%u, ReadBandwidthChroma = %f\n", __func__, k, mode_lib->ms.SurfaceReadBandwidthChroma[k]);
-#endif
- }
-
- // Writeback bandwidth
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format == dml2_444_64) {
- mode_lib->ms.WriteBandwidth[k] = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height
- * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width
- / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height
- * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total
- / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * 8.0;
- } else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- mode_lib->ms.WriteBandwidth[k] = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height
- * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width
- / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height
- * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total
- / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * 4.0;
- } else {
- mode_lib->ms.WriteBandwidth[k] = 0.0;
- }
- }
-
- /*Writeback Latency support check*/
- mode_lib->ms.support.WritebackLatencySupport = true;
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true &&
- (mode_lib->ms.WriteBandwidth[k] > mode_lib->ip.writeback_interface_buffer_size_kbytes * 1024.0 / mode_lib->soc.qos_parameters.writeback.base_latency_us)) {
- mode_lib->ms.support.WritebackLatencySupport = false;
- }
- }
-
- /* Writeback Mode Support Check */
- s->TotalNumberOfActiveWriteback = 0;
- for (k = 0; k <= (unsigned int)mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true
- && (display_cfg->plane_descriptors[k].stream_index == k)) {
- s->TotalNumberOfActiveWriteback = s->TotalNumberOfActiveWriteback + 1;
- }
- }
-
- mode_lib->ms.support.EnoughWritebackUnits = 1;
- if (s->TotalNumberOfActiveWriteback > (unsigned int)mode_lib->ip.max_num_wb) {
- mode_lib->ms.support.EnoughWritebackUnits = false;
- }
-
- /* Writeback Scale Ratio and Taps Support Check */
- mode_lib->ms.support.WritebackScaleRatioAndTapsSupport = true;
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio > mode_lib->ip.writeback_max_hscl_ratio
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio > mode_lib->ip.writeback_max_vscl_ratio
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio < mode_lib->ip.writeback_min_hscl_ratio
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio < mode_lib->ip.writeback_min_vscl_ratio
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps > (unsigned int) mode_lib->ip.writeback_max_hscl_taps
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps > (unsigned int) mode_lib->ip.writeback_max_vscl_taps
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio > (unsigned int)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps
- || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio > (unsigned int)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps
- || (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps > 2.0 && ((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps % 2) == 1))) {
- mode_lib->ms.support.WritebackScaleRatioAndTapsSupport = false;
- }
- if (2.0 * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height * (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps - 1) * 57 > mode_lib->ip.writeback_line_buffer_buffer_size) {
- mode_lib->ms.support.WritebackScaleRatioAndTapsSupport = false;
- }
- }
- }
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- CalculateSinglePipeDPPCLKAndSCLThroughput(
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio,
- mode_lib->ip.max_dchub_pscl_bw_pix_per_clk,
- mode_lib->ip.max_pscl_lb_bw_pix_per_clk,
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000),
- display_cfg->plane_descriptors[k].pixel_format,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_taps,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_taps,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps,
- /* Output */
- &mode_lib->ms.PSCL_FACTOR[k],
- &mode_lib->ms.PSCL_FACTOR_CHROMA[k],
- &mode_lib->ms.MinDPPCLKUsingSingleDPP[k]);
- }
-
- // Max Viewport Size support
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (display_cfg->plane_descriptors[k].surface.tiling == dml2_sw_linear) {
- s->MaximumSwathWidthSupportLuma = 15360;
- } else if (!dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle) && mode_lib->ms.BytePerPixelC[k] > 0 && display_cfg->plane_descriptors[k].pixel_format != dml2_rgbe_alpha) { // horz video
- s->MaximumSwathWidthSupportLuma = 7680 + 16;
- } else if (dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle) && mode_lib->ms.BytePerPixelC[k] > 0 && display_cfg->plane_descriptors[k].pixel_format != dml2_rgbe_alpha) { // vert video
- s->MaximumSwathWidthSupportLuma = 4320 + 16;
- } else if (display_cfg->plane_descriptors[k].pixel_format == dml2_rgbe_alpha) { // rgbe + alpha
- s->MaximumSwathWidthSupportLuma = 5120 + 16;
- } else if (dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle) && mode_lib->ms.BytePerPixelY[k] == 8 && display_cfg->plane_descriptors[k].surface.dcc.enable == true) { // vert 64bpp
- s->MaximumSwathWidthSupportLuma = 3072 + 16;
- } else {
- s->MaximumSwathWidthSupportLuma = 6144 + 16;
- }
-
- if (dml2_core_shared_is_420(display_cfg->plane_descriptors[k].pixel_format)) {
- s->MaximumSwathWidthSupportChroma = (unsigned int)(s->MaximumSwathWidthSupportLuma / 2.0);
- } else {
- s->MaximumSwathWidthSupportChroma = s->MaximumSwathWidthSupportLuma;
- }
- mode_lib->ms.MaximumSwathWidthInLineBufferLuma = mode_lib->ip.line_buffer_size_bits * math_max2(display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio, 1.0) / 57 /*FIXME_STAGE2 was: LBBitPerPixel*/ /
- (display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps + math_max2(math_ceil2(display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio, 1.0) - 2, 0.0));
- if (mode_lib->ms.BytePerPixelC[k] == 0.0) {
- mode_lib->ms.MaximumSwathWidthInLineBufferChroma = 0;
- } else {
- mode_lib->ms.MaximumSwathWidthInLineBufferChroma = mode_lib->ip.line_buffer_size_bits * math_max2(display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio, 1.0) / 57 /*FIXME_STAGE2 was: LBBitPerPixel*/ /
- (display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps + math_max2(math_ceil2(display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio, 1.0) - 2, 0.0));
- }
- mode_lib->ms.MaximumSwathWidthLuma[k] = math_min2(s->MaximumSwathWidthSupportLuma, mode_lib->ms.MaximumSwathWidthInLineBufferLuma);
- mode_lib->ms.MaximumSwathWidthChroma[k] = math_min2(s->MaximumSwathWidthSupportChroma, mode_lib->ms.MaximumSwathWidthInLineBufferChroma);
- }
-
- /* Cursor Support Check */
- mode_lib->ms.support.CursorSupport = true;
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (display_cfg->plane_descriptors[k].cursor.cursor_width > 0.0) {
- if (display_cfg->plane_descriptors[k].cursor.cursor_bpp == 64 && mode_lib->ip.cursor_64bpp_support == false) {
- mode_lib->ms.support.CursorSupport = false;
- }
- }
- }
-
- /* Valid Pitch Check */
- mode_lib->ms.support.PitchSupport = true;
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
-
- // data pitch
- unsigned int alignment_l = mode_lib->ms.MacroTileWidthY[k];
-
- if (mode_lib->ms.surf_linear128_l[k])
- alignment_l = alignment_l / 2;
-
- mode_lib->ms.support.AlignedYPitch[k] = (unsigned int)math_ceil2(math_max2(display_cfg->plane_descriptors[k].surface.plane0.pitch, display_cfg->plane_descriptors[k].surface.plane0.width), alignment_l);
- if (dml2_core_shared_is_420(display_cfg->plane_descriptors[k].pixel_format) || display_cfg->plane_descriptors[k].pixel_format == dml2_rgbe_alpha) {
- unsigned int alignment_c = mode_lib->ms.MacroTileWidthC[k];
-
- if (mode_lib->ms.surf_linear128_c[k])
- alignment_c = alignment_c / 2;
- mode_lib->ms.support.AlignedCPitch[k] = (unsigned int)math_ceil2(math_max2(display_cfg->plane_descriptors[k].surface.plane1.pitch, display_cfg->plane_descriptors[k].surface.plane1.width), alignment_c);
- } else {
- mode_lib->ms.support.AlignedCPitch[k] = display_cfg->plane_descriptors[k].surface.plane1.pitch;
- }
-
- if (mode_lib->ms.support.AlignedYPitch[k] > display_cfg->plane_descriptors[k].surface.plane0.pitch ||
- mode_lib->ms.support.AlignedCPitch[k] > display_cfg->plane_descriptors[k].surface.plane1.pitch) {
- mode_lib->ms.support.PitchSupport = false;
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: k=%u AlignedYPitch = %d\n", __func__, k, mode_lib->ms.support.AlignedYPitch[k]);
- dml2_printf("DML::%s: k=%u PitchY = %d\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane0.pitch);
- dml2_printf("DML::%s: k=%u AlignedCPitch = %d\n", __func__, k, mode_lib->ms.support.AlignedCPitch[k]);
- dml2_printf("DML::%s: k=%u PitchC = %d\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane1.pitch);
- dml2_printf("DML::%s: k=%u PitchSupport = %d\n", __func__, k, mode_lib->ms.support.PitchSupport);
-#endif
- }
-
- // meta pitch
- if (mode_lib->ip.dcn_mrq_present && display_cfg->plane_descriptors[k].surface.dcc.enable) {
- mode_lib->ms.support.AlignedDCCMetaPitchY[k] = (unsigned int)math_ceil2(math_max2(display_cfg->plane_descriptors[k].surface.dcc.plane0.pitch,
- display_cfg->plane_descriptors[k].surface.plane0.width), 64.0 * mode_lib->ms.Read256BlockWidthY[k]);
-
- if (mode_lib->ms.support.AlignedDCCMetaPitchY[k] > display_cfg->plane_descriptors[k].surface.dcc.plane0.pitch)
- mode_lib->ms.support.PitchSupport = false;
-
- if (dml2_core_shared_is_420(display_cfg->plane_descriptors[k].pixel_format) || display_cfg->plane_descriptors[k].pixel_format == dml2_rgbe_alpha) {
- mode_lib->ms.support.AlignedDCCMetaPitchC[k] = (unsigned int)math_ceil2(math_max2(display_cfg->plane_descriptors[k].surface.dcc.plane1.pitch,
- display_cfg->plane_descriptors[k].surface.plane1.width), 64.0 * mode_lib->ms.Read256BlockWidthC[k]);
-
- if (mode_lib->ms.support.AlignedDCCMetaPitchC[k] > display_cfg->plane_descriptors[k].surface.dcc.plane1.pitch)
- mode_lib->ms.support.PitchSupport = false;
- }
- } else {
- mode_lib->ms.support.AlignedDCCMetaPitchY[k] = 0;
- mode_lib->ms.support.AlignedDCCMetaPitchC[k] = 0;
- }
- }
-
- mode_lib->ms.support.ViewportExceedsSurface = false;
- if (!display_cfg->overrides.hw.surface_viewport_size_check_disable) {
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (display_cfg->plane_descriptors[k].composition.viewport.plane0.width > display_cfg->plane_descriptors[k].surface.plane0.width || display_cfg->plane_descriptors[k].composition.viewport.plane0.height > display_cfg->plane_descriptors[k].surface.plane0.height) {
- mode_lib->ms.support.ViewportExceedsSurface = true;
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: k=%u ViewportWidth = %d\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.width);
- dml2_printf("DML::%s: k=%u SurfaceWidthY = %d\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane0.width);
- dml2_printf("DML::%s: k=%u ViewportHeight = %d\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.height);
- dml2_printf("DML::%s: k=%u SurfaceHeightY = %d\n", __func__, k, display_cfg->plane_descriptors[k].surface.plane0.height);
- dml2_printf("DML::%s: k=%u ViewportExceedsSurface = %d\n", __func__, k, mode_lib->ms.support.ViewportExceedsSurface);
-#endif
- if (dml2_core_shared_is_420(display_cfg->plane_descriptors[k].pixel_format) || display_cfg->plane_descriptors[k].pixel_format == dml2_rgbe_alpha) {
- if (display_cfg->plane_descriptors[k].composition.viewport.plane1.width > display_cfg->plane_descriptors[k].surface.plane1.width ||
- display_cfg->plane_descriptors[k].composition.viewport.plane1.height > display_cfg->plane_descriptors[k].surface.plane1.height) {
- mode_lib->ms.support.ViewportExceedsSurface = true;
- }
- }
- }
- }
- }
-
- CalculateSwathAndDETConfiguration_params->display_cfg = display_cfg;
- CalculateSwathAndDETConfiguration_params->ConfigReturnBufferSizeInKByte = mode_lib->ip.config_return_buffer_size_in_kbytes;
- CalculateSwathAndDETConfiguration_params->MaxTotalDETInKByte = mode_lib->ms.MaxTotalDETInKByte;
- CalculateSwathAndDETConfiguration_params->MinCompressedBufferSizeInKByte = mode_lib->ms.MinCompressedBufferSizeInKByte;
- CalculateSwathAndDETConfiguration_params->rob_buffer_size_kbytes = mode_lib->ip.rob_buffer_size_kbytes;
- CalculateSwathAndDETConfiguration_params->pixel_chunk_size_kbytes = mode_lib->ip.pixel_chunk_size_kbytes;
- CalculateSwathAndDETConfiguration_params->rob_buffer_size_kbytes = mode_lib->ip.rob_buffer_size_kbytes;
- CalculateSwathAndDETConfiguration_params->pixel_chunk_size_kbytes = mode_lib->ip.pixel_chunk_size_kbytes;
- CalculateSwathAndDETConfiguration_params->ForceSingleDPP = 1;
- CalculateSwathAndDETConfiguration_params->NumberOfActiveSurfaces = mode_lib->ms.num_active_planes;
- CalculateSwathAndDETConfiguration_params->nomDETInKByte = mode_lib->ms.NomDETInKByte;
- CalculateSwathAndDETConfiguration_params->ConfigReturnBufferSegmentSizeInkByte = mode_lib->ip.config_return_buffer_segment_size_in_kbytes;
- CalculateSwathAndDETConfiguration_params->CompressedBufferSegmentSizeInkByte = mode_lib->ip.compressed_buffer_segment_size_in_kbytes;
- CalculateSwathAndDETConfiguration_params->ReadBandwidthLuma = mode_lib->ms.SurfaceReadBandwidthLuma;
- CalculateSwathAndDETConfiguration_params->ReadBandwidthChroma = mode_lib->ms.SurfaceReadBandwidthChroma;
- CalculateSwathAndDETConfiguration_params->MaximumSwathWidthLuma = mode_lib->ms.MaximumSwathWidthLuma;
- CalculateSwathAndDETConfiguration_params->MaximumSwathWidthChroma = mode_lib->ms.MaximumSwathWidthChroma;
- CalculateSwathAndDETConfiguration_params->Read256BytesBlockHeightY = mode_lib->ms.Read256BlockHeightY;
- CalculateSwathAndDETConfiguration_params->Read256BytesBlockHeightC = mode_lib->ms.Read256BlockHeightC;
- CalculateSwathAndDETConfiguration_params->Read256BytesBlockWidthY = mode_lib->ms.Read256BlockWidthY;
- CalculateSwathAndDETConfiguration_params->Read256BytesBlockWidthC = mode_lib->ms.Read256BlockWidthC;
- CalculateSwathAndDETConfiguration_params->surf_linear128_l = mode_lib->ms.surf_linear128_l;
- CalculateSwathAndDETConfiguration_params->surf_linear128_c = mode_lib->ms.surf_linear128_c;
- CalculateSwathAndDETConfiguration_params->ODMMode = s->dummy_odm_mode;
- CalculateSwathAndDETConfiguration_params->BytePerPixY = mode_lib->ms.BytePerPixelY;
- CalculateSwathAndDETConfiguration_params->BytePerPixC = mode_lib->ms.BytePerPixelC;
- CalculateSwathAndDETConfiguration_params->BytePerPixDETY = mode_lib->ms.BytePerPixelInDETY;
- CalculateSwathAndDETConfiguration_params->BytePerPixDETC = mode_lib->ms.BytePerPixelInDETC;
- CalculateSwathAndDETConfiguration_params->DPPPerSurface = s->dummy_integer_array[2];
- CalculateSwathAndDETConfiguration_params->mrq_present = mode_lib->ip.dcn_mrq_present;
-
- // output
- CalculateSwathAndDETConfiguration_params->req_per_swath_ub_l = s->dummy_integer_array[0];
- CalculateSwathAndDETConfiguration_params->req_per_swath_ub_c = s->dummy_integer_array[1];
- CalculateSwathAndDETConfiguration_params->swath_width_luma_ub = s->dummy_integer_array[3];
- CalculateSwathAndDETConfiguration_params->swath_width_chroma_ub = s->dummy_integer_array[4];
- CalculateSwathAndDETConfiguration_params->SwathWidth = s->dummy_integer_array[5];
- CalculateSwathAndDETConfiguration_params->SwathWidthChroma = s->dummy_integer_array[6];
- CalculateSwathAndDETConfiguration_params->SwathHeightY = s->dummy_integer_array[7];
- CalculateSwathAndDETConfiguration_params->SwathHeightC = s->dummy_integer_array[8];
- CalculateSwathAndDETConfiguration_params->request_size_bytes_luma = s->dummy_integer_array[26];
- CalculateSwathAndDETConfiguration_params->request_size_bytes_chroma = s->dummy_integer_array[27];
- CalculateSwathAndDETConfiguration_params->DETBufferSizeInKByte = s->dummy_integer_array[9];
- CalculateSwathAndDETConfiguration_params->DETBufferSizeY = s->dummy_integer_array[10];
- CalculateSwathAndDETConfiguration_params->DETBufferSizeC = s->dummy_integer_array[11];
- CalculateSwathAndDETConfiguration_params->full_swath_bytes_l = s->full_swath_bytes_l;
- CalculateSwathAndDETConfiguration_params->full_swath_bytes_c = s->full_swath_bytes_c;
- CalculateSwathAndDETConfiguration_params->UnboundedRequestEnabled = &s->dummy_boolean[0];
- CalculateSwathAndDETConfiguration_params->compbuf_reserved_space_64b = &s->dummy_integer[1];
- CalculateSwathAndDETConfiguration_params->hw_debug5 = &s->dummy_boolean[2];
- CalculateSwathAndDETConfiguration_params->CompressedBufferSizeInkByte = &s->dummy_integer[0];
- CalculateSwathAndDETConfiguration_params->ViewportSizeSupportPerSurface = mode_lib->ms.SingleDPPViewportSizeSupportPerSurface;
- CalculateSwathAndDETConfiguration_params->ViewportSizeSupport = &s->dummy_boolean[1];
- CalculateSwathAndDETConfiguration_params->funcs = &mode_lib->funcs;
-
- // This calls is just to find out if there is enough DET space to support full vp in 1 pipe.
- CalculateSwathAndDETConfiguration(&mode_lib->scratch, CalculateSwathAndDETConfiguration_params);
-
- {
- mode_lib->ms.TotalNumberOfActiveDPP = 0;
- mode_lib->ms.support.TotalAvailablePipesSupport = true;
-
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- CalculateODMMode(
- mode_lib->ip.maximum_pixels_per_line_per_dsc_unit,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].overrides.odm_mode,
- mode_lib->ms.max_dispclk_freq_mhz,
- false, // DSCEnable
- mode_lib->ms.TotalNumberOfActiveDPP,
- mode_lib->ip.max_num_dpp,
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000),
-
- /* Output */
- &s->TotalAvailablePipesSupportNoDSC,
- &s->NumberOfDPPNoDSC,
- &s->ODMModeNoDSC,
- &s->RequiredDISPCLKPerSurfaceNoDSC);
-
- CalculateODMMode(
- mode_lib->ip.maximum_pixels_per_line_per_dsc_unit,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].overrides.odm_mode,
- mode_lib->ms.max_dispclk_freq_mhz,
- true, // DSCEnable
- mode_lib->ms.TotalNumberOfActiveDPP,
- mode_lib->ip.max_num_dpp,
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000),
-
- /* Output */
- &s->TotalAvailablePipesSupportDSC,
- &s->NumberOfDPPDSC,
- &s->ODMModeDSC,
- &s->RequiredDISPCLKPerSurfaceDSC);
-
- /*Number Of DSC Slices*/
- if (display_cfg->plane_descriptors[k].stream_index == k) {
- if (s->PixelClockBackEnd[k] > 4800) {
- mode_lib->ms.support.NumberOfDSCSlices[k] = (unsigned int)(math_ceil2(s->PixelClockBackEnd[k] / 600, 4));
- } else if (s->PixelClockBackEnd[k] > 2400) {
- mode_lib->ms.support.NumberOfDSCSlices[k] = 8;
- } else if (s->PixelClockBackEnd[k] > 1200) {
- mode_lib->ms.support.NumberOfDSCSlices[k] = 4;
- } else if (s->PixelClockBackEnd[k] > 340) {
- mode_lib->ms.support.NumberOfDSCSlices[k] = 2;
- } else {
- mode_lib->ms.support.NumberOfDSCSlices[k] = 1;
- }
- } else {
- mode_lib->ms.support.NumberOfDSCSlices[k] = 0;
- }
-
- if (s->ODMModeDSC == dml2_odm_mode_combine_2to1)
- mode_lib->ms.support.NumberOfDSCSlices[k] = 2 * (unsigned int)math_ceil2(mode_lib->ms.support.NumberOfDSCSlices[k] / 2.0, 1.0);
- else if (s->ODMModeDSC == dml2_odm_mode_combine_3to1)
- mode_lib->ms.support.NumberOfDSCSlices[k] = 12;
- else if (s->ODMModeDSC == dml2_odm_mode_combine_4to1)
- mode_lib->ms.support.NumberOfDSCSlices[k] = 4 * (unsigned int)math_ceil2(mode_lib->ms.support.NumberOfDSCSlices[k] / 4.0, 1.0);
-
- CalculateOutputLink(
- &mode_lib->scratch,
- ((double)mode_lib->soc.clk_table.phyclk.clk_values_khz[0] / 1000),
- ((double)mode_lib->soc.clk_table.phyclk_d18.clk_values_khz[0] / 1000),
- ((double)mode_lib->soc.clk_table.phyclk_d32.clk_values_khz[0] / 1000),
- mode_lib->soc.phy_downspread_percent,
- (display_cfg->plane_descriptors[k].stream_index == k),
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active,
- s->PixelClockBackEnd[k],
- s->OutputBpp[k],
- mode_lib->ip.maximum_dsc_bits_per_component,
- mode_lib->ms.support.NumberOfDSCSlices[k],
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.audio_sample_rate,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.audio_sample_layout,
- s->ODMModeNoDSC,
- s->ODMModeDSC,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_lane_count,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_link_rate,
-
- /* Output */
- &mode_lib->ms.RequiresDSC[k],
- &mode_lib->ms.RequiresFEC[k],
- &mode_lib->ms.OutputBpp[k],
- &mode_lib->ms.OutputType[k], // VBA_DELTA, VBA uses a string to represent type and rate, but DML uses enum, don't want to rely on strng
- &mode_lib->ms.OutputRate[k],
- &mode_lib->ms.RequiredSlots[k]);
-
- if (mode_lib->ms.RequiresDSC[k] == false) {
- mode_lib->ms.ODMMode[k] = s->ODMModeNoDSC;
- mode_lib->ms.RequiredDISPCLKPerSurface[k] = s->RequiredDISPCLKPerSurfaceNoDSC;
- if (!s->TotalAvailablePipesSupportNoDSC)
- mode_lib->ms.support.TotalAvailablePipesSupport = false;
- mode_lib->ms.TotalNumberOfActiveDPP = mode_lib->ms.TotalNumberOfActiveDPP + s->NumberOfDPPNoDSC;
- } else {
- mode_lib->ms.ODMMode[k] = s->ODMModeDSC;
- mode_lib->ms.RequiredDISPCLKPerSurface[k] = s->RequiredDISPCLKPerSurfaceDSC;
- if (!s->TotalAvailablePipesSupportDSC)
- mode_lib->ms.support.TotalAvailablePipesSupport = false;
- mode_lib->ms.TotalNumberOfActiveDPP = mode_lib->ms.TotalNumberOfActiveDPP + s->NumberOfDPPDSC;
- }
- dml2_printf("DML::%s: k=%d RequiresDSC = %d\n", __func__, k, mode_lib->ms.RequiresDSC[k]);
- dml2_printf("DML::%s: k=%d ODMMode = %d\n", __func__, k, mode_lib->ms.ODMMode[k]);
- }
-
- // FIXME_DCN4 - add odm vs mpc use check
-
- // FIXME_DCN4 - add imall cap check
- mode_lib->ms.support.incorrect_imall_usage = 0;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (mode_lib->ip.imall_supported && display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_imall)
- mode_lib->ms.support.incorrect_imall_usage = 1;
- }
-
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- mode_lib->ms.MPCCombine[k] = false;
- mode_lib->ms.NoOfDPP[k] = 1;
-
- if (mode_lib->ms.ODMMode[k] == dml2_odm_mode_combine_4to1) {
- mode_lib->ms.MPCCombine[k] = false;
- mode_lib->ms.NoOfDPP[k] = 4;
- } else if (mode_lib->ms.ODMMode[k] == dml2_odm_mode_combine_3to1) {
- mode_lib->ms.MPCCombine[k] = false;
- mode_lib->ms.NoOfDPP[k] = 3;
- } else if (mode_lib->ms.ODMMode[k] == dml2_odm_mode_combine_2to1) {
- mode_lib->ms.MPCCombine[k] = false;
- mode_lib->ms.NoOfDPP[k] = 2;
- } else if (display_cfg->plane_descriptors[k].overrides.mpcc_combine_factor == 2) {
- mode_lib->ms.MPCCombine[k] = true;
- mode_lib->ms.NoOfDPP[k] = 2;
- mode_lib->ms.TotalNumberOfActiveDPP++;
- } else if (display_cfg->plane_descriptors[k].overrides.mpcc_combine_factor == 1) {
- mode_lib->ms.MPCCombine[k] = false;
- mode_lib->ms.NoOfDPP[k] = 1;
- if (!mode_lib->ms.SingleDPPViewportSizeSupportPerSurface[k]) {
- dml2_printf("ERROR: DML::%s: MPCC is override to disable but viewport is too large to be supported with single pipe!\n", __func__);
- }
- } else {
- if ((mode_lib->ms.MinDPPCLKUsingSingleDPP[k] > mode_lib->ms.max_dppclk_freq_mhz) || !mode_lib->ms.SingleDPPViewportSizeSupportPerSurface[k]) {
- mode_lib->ms.MPCCombine[k] = true;
- mode_lib->ms.NoOfDPP[k] = 2;
- mode_lib->ms.TotalNumberOfActiveDPP++;
- }
- }
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: k=%d, NoOfDPP = %d\n", __func__, k, mode_lib->ms.NoOfDPP[k]);
-#endif
- }
-
- if (mode_lib->ms.TotalNumberOfActiveDPP > (unsigned int)mode_lib->ip.max_num_dpp)
- mode_lib->ms.support.TotalAvailablePipesSupport = false;
-
-
- mode_lib->ms.TotalNumberOfSingleDPPSurfaces = 0;
- for (k = 0; k < (unsigned int)mode_lib->ms.num_active_planes; ++k) {
- if (mode_lib->ms.NoOfDPP[k] == 1)
- mode_lib->ms.TotalNumberOfSingleDPPSurfaces = mode_lib->ms.TotalNumberOfSingleDPPSurfaces + 1;
- }
-
- //DISPCLK/DPPCLK
- mode_lib->ms.WritebackRequiredDISPCLK = 0;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable) {
- mode_lib->ms.WritebackRequiredDISPCLK = math_max2(mode_lib->ms.WritebackRequiredDISPCLK,
- CalculateWriteBackDISPCLK(display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format,
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000),
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_width,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total,
- mode_lib->ip.writeback_line_buffer_buffer_size));
- }
- }
-
- mode_lib->ms.RequiredDISPCLK = mode_lib->ms.WritebackRequiredDISPCLK;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- mode_lib->ms.RequiredDISPCLK = math_max2(mode_lib->ms.RequiredDISPCLK, mode_lib->ms.RequiredDISPCLKPerSurface[k]);
- }
-
- mode_lib->ms.GlobalDPPCLK = 0;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- mode_lib->ms.RequiredDPPCLK[k] = mode_lib->ms.MinDPPCLKUsingSingleDPP[k] / mode_lib->ms.NoOfDPP[k];
- mode_lib->ms.GlobalDPPCLK = math_max2(mode_lib->ms.GlobalDPPCLK, mode_lib->ms.RequiredDPPCLK[k]);
- }
-
- mode_lib->ms.support.DISPCLK_DPPCLK_Support = !((mode_lib->ms.RequiredDISPCLK > mode_lib->ms.max_dispclk_freq_mhz) || (mode_lib->ms.GlobalDPPCLK > mode_lib->ms.max_dppclk_freq_mhz));
- }
-
- /* Total Available OTG, HDMIFRL, DP Support Check */
- s->TotalNumberOfActiveOTG = 0;
- s->TotalNumberOfActiveHDMIFRL = 0;
- s->TotalNumberOfActiveDP2p0 = 0;
- s->TotalNumberOfActiveDP2p0Outputs = 0;
-
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (display_cfg->plane_descriptors[k].stream_index == k) {
- s->TotalNumberOfActiveOTG = s->TotalNumberOfActiveOTG + 1;
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmifrl)
- s->TotalNumberOfActiveHDMIFRL = s->TotalNumberOfActiveHDMIFRL + 1;
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_dp2p0) {
- s->TotalNumberOfActiveDP2p0 = s->TotalNumberOfActiveDP2p0 + 1;
- // FIXME_STAGE2: SW not using backend related stuff, need mapping for mst setup
- //if (display_cfg->output.OutputMultistreamId[k] == k || display_cfg->output.OutputMultistreamEn[k] == false) {
- s->TotalNumberOfActiveDP2p0Outputs = s->TotalNumberOfActiveDP2p0Outputs + 1;
- //}
- }
- }
- }
-
- mode_lib->ms.support.NumberOfOTGSupport = (s->TotalNumberOfActiveOTG <= (unsigned int)mode_lib->ip.max_num_otg);
- mode_lib->ms.support.NumberOfHDMIFRLSupport = (s->TotalNumberOfActiveHDMIFRL <= (unsigned int)mode_lib->ip.max_num_hdmi_frl_outputs);
- mode_lib->ms.support.NumberOfDP2p0Support = (s->TotalNumberOfActiveDP2p0 <= (unsigned int)mode_lib->ip.max_num_dp2p0_streams && s->TotalNumberOfActiveDP2p0Outputs <= (unsigned int)mode_lib->ip.max_num_dp2p0_outputs);
-
- mode_lib->ms.support.ExceededMultistreamSlots = false;
- mode_lib->ms.support.LinkCapacitySupport = true;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_disabled == false &&
- display_cfg->plane_descriptors[k].stream_index == k && (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_dp || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_dp2p0 || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_edp ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmi || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmifrl) && mode_lib->ms.OutputBpp[k] == 0) {
- mode_lib->ms.support.LinkCapacitySupport = false;
- }
- }
-
- mode_lib->ms.support.P2IWith420 = false;
- mode_lib->ms.support.DSCOnlyIfNecessaryWithBPP = false;
- mode_lib->ms.support.DSC422NativeNotSupported = false;
- mode_lib->ms.support.LinkRateDoesNotMatchDPVersion = false;
- mode_lib->ms.support.LinkRateForMultistreamNotIndicated = false;
- mode_lib->ms.support.BPPForMultistreamNotIndicated = false;
- mode_lib->ms.support.MultistreamWithHDMIOreDP = false;
- mode_lib->ms.support.MSOOrODMSplitWithNonDPLink = false;
- mode_lib->ms.support.NotEnoughLanesForMSO = false;
-
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (display_cfg->plane_descriptors[k].stream_index == k && (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_dp || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_dp2p0 || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_edp ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmi || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmifrl)) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format == dml2_420 && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced == 1 && mode_lib->ip.ptoi_supported == true)
- mode_lib->ms.support.P2IWith420 = true;
-
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable == dml2_dsc_enable_if_necessary && s->OutputBpp[k] != 0)
- mode_lib->ms.support.DSCOnlyIfNecessaryWithBPP = true;
- if ((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable == dml2_dsc_enable || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable == dml2_dsc_enable_if_necessary) && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format == dml2_n422 && !mode_lib->ip.dsc422_native_support)
- mode_lib->ms.support.DSC422NativeNotSupported = true;
-
- if (((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_link_rate == dml2_dp_rate_hbr || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_link_rate == dml2_dp_rate_hbr2 ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_link_rate == dml2_dp_rate_hbr3) &&
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder != dml2_dp && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder != dml2_edp) ||
- ((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_link_rate == dml2_dp_rate_uhbr10 || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_link_rate == dml2_dp_rate_uhbr13p5 ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_link_rate == dml2_dp_rate_uhbr20) &&
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder != dml2_dp2p0))
- mode_lib->ms.support.LinkRateDoesNotMatchDPVersion = true;
-
- // FIXME_STAGE2
- //if (display_cfg->output.OutputMultistreamEn[k] == 1) {
- // if (display_cfg->output.OutputMultistreamId[k] == k && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_link_rate == dml2_dp_rate_na)
- // mode_lib->ms.support.LinkRateForMultistreamNotIndicated = true;
- // if (display_cfg->output.OutputMultistreamId[k] == k && s->OutputBpp[k] == 0)
- // mode_lib->ms.support.BPPForMultistreamNotIndicated = true;
- // for (n = 0; n < mode_lib->ms.num_active_planes; ++n) {
- // if (display_cfg->output.OutputMultistreamId[k] == n && s->OutputBpp[k] == 0)
- // mode_lib->ms.support.BPPForMultistreamNotIndicated = true;
- // }
- //}
-
- if ((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_edp ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmi ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmifrl)) {
- // FIXME_STAGE2
- //if (display_cfg->output.OutputMultistreamEn[k] == 1 && display_cfg->output.OutputMultistreamId[k] == k)
- // mode_lib->ms.support.MultistreamWithHDMIOreDP = true;
- //for (n = 0; n < mode_lib->ms.num_active_planes; ++n) {
- // if (display_cfg->output.OutputMultistreamEn[k] == 1 && display_cfg->output.OutputMultistreamId[k] == n)
- // mode_lib->ms.support.MultistreamWithHDMIOreDP = true;
- //}
- }
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder != dml2_dp && (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].overrides.odm_mode == dml2_odm_mode_split_1to2 ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].overrides.odm_mode == dml2_odm_mode_mso_1to2 || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].overrides.odm_mode == dml2_odm_mode_mso_1to4))
- mode_lib->ms.support.MSOOrODMSplitWithNonDPLink = true;
-
- if ((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].overrides.odm_mode == dml2_odm_mode_mso_1to2 && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_lane_count < 2) ||
- (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].overrides.odm_mode == dml2_odm_mode_mso_1to4 && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_dp_lane_count < 4))
- mode_lib->ms.support.NotEnoughLanesForMSO = true;
- }
- }
-
- mode_lib->ms.support.DTBCLKRequiredMoreThanSupported = false;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (display_cfg->plane_descriptors[k].stream_index == k &&
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmifrl &&
- RequiredDTBCLK(
- mode_lib->ms.RequiresDSC[k],
- s->PixelClockBackEnd[k],
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format,
- mode_lib->ms.OutputBpp[k],
- mode_lib->ms.support.NumberOfDSCSlices[k],
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.audio_sample_rate,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.audio_sample_layout) > ((double)mode_lib->soc.clk_table.dtbclk.clk_values_khz[0] / 1000)) {
- mode_lib->ms.support.DTBCLKRequiredMoreThanSupported = true;
- }
- }
-
- mode_lib->ms.support.DSCCLKRequiredMoreThanSupported = false;
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->plane_descriptors[k].stream_index == k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_dp ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_dp2p0 ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_edp ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmifrl) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format == dml2_420) {
- s->DSCFormatFactor = 2;
- } else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format == dml2_444) {
- s->DSCFormatFactor = 1;
- } else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format == dml2_n422 || display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmifrl) {
- s->DSCFormatFactor = 2;
- } else {
- s->DSCFormatFactor = 1;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, RequiresDSC = %u\n", __func__, k, mode_lib->ms.RequiresDSC[k]);
-#endif
- if (mode_lib->ms.RequiresDSC[k] == true) {
- s->PixelClockBackEndFactor = 3.0;
-
- if (mode_lib->ms.ODMMode[k] == dml2_odm_mode_combine_4to1)
- s->PixelClockBackEndFactor = 12.0;
- else if (mode_lib->ms.ODMMode[k] == dml2_odm_mode_combine_3to1)
- s->PixelClockBackEndFactor = 9.0;
- else if (mode_lib->ms.ODMMode[k] == dml2_odm_mode_combine_2to1)
- s->PixelClockBackEndFactor = 6.0;
-
- mode_lib->ms.required_dscclk_freq_mhz[k] = s->PixelClockBackEnd[k] / s->PixelClockBackEndFactor / (double)s->DSCFormatFactor;
- if (mode_lib->ms.required_dscclk_freq_mhz[k] > mode_lib->ms.max_dscclk_freq_mhz) {
- mode_lib->ms.support.DSCCLKRequiredMoreThanSupported = true;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, PixelClockBackEnd = %f\n", __func__, k, s->PixelClockBackEnd[k]);
- dml2_printf("DML::%s: k=%u, required_dscclk_freq_mhz = %f\n", __func__, k, mode_lib->ms.required_dscclk_freq_mhz[k]);
- dml2_printf("DML::%s: k=%u, DSCFormatFactor = %u\n", __func__, k, s->DSCFormatFactor);
- dml2_printf("DML::%s: k=%u, DSCCLKRequiredMoreThanSupported = %u\n", __func__, k, mode_lib->ms.support.DSCCLKRequiredMoreThanSupported);
-#endif
- }
- }
- }
- }
-
- /* Check DSC Unit and Slices Support */
- mode_lib->ms.support.NotEnoughDSCSlices = false;
- s->TotalDSCUnitsRequired = 0;
- mode_lib->ms.support.PixelsPerLinePerDSCUnitSupport = true;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (mode_lib->ms.RequiresDSC[k] == true) {
- s->NumDSCUnitRequired = 1;
-
- if (mode_lib->ms.ODMMode[k] == dml2_odm_mode_combine_4to1)
- s->NumDSCUnitRequired = 4;
- else if (mode_lib->ms.ODMMode[k] == dml2_odm_mode_combine_3to1)
- s->NumDSCUnitRequired = 3;
- else if (mode_lib->ms.ODMMode[k] == dml2_odm_mode_combine_2to1)
- s->NumDSCUnitRequired = 2;
-
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active > s->NumDSCUnitRequired * (unsigned int)mode_lib->ip.maximum_pixels_per_line_per_dsc_unit)
- mode_lib->ms.support.PixelsPerLinePerDSCUnitSupport = false;
- s->TotalDSCUnitsRequired = s->TotalDSCUnitsRequired + s->NumDSCUnitRequired;
- if (mode_lib->ms.support.NumberOfDSCSlices[k] > 4 * s->NumDSCUnitRequired)
- mode_lib->ms.support.NotEnoughDSCSlices = true;
- }
- }
-
- mode_lib->ms.support.NotEnoughDSCUnits = false;
- if (s->TotalDSCUnitsRequired > (unsigned int)mode_lib->ip.num_dsc) {
- mode_lib->ms.support.NotEnoughDSCUnits = true;
- }
-
- /*DSC Delay per state*/
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- mode_lib->ms.DSCDelay[k] = DSCDelayRequirement(mode_lib->ms.RequiresDSC[k],
- mode_lib->ms.ODMMode[k],
- mode_lib->ip.maximum_dsc_bits_per_component,
- mode_lib->ms.OutputBpp[k],
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total,
- mode_lib->ms.support.NumberOfDSCSlices[k],
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder,
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000),
- s->PixelClockBackEnd[k]);
- }
-
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- for (m = 0; m < mode_lib->ms.num_active_planes; m++) {
- if (display_cfg->plane_descriptors[k].stream_index == m && mode_lib->ms.RequiresDSC[m] == true) {
- mode_lib->ms.DSCDelay[k] = mode_lib->ms.DSCDelay[m];
- }
- }
- }
-
- // Figure out the swath and DET configuration after the num dpp per plane is figured out
- CalculateSwathAndDETConfiguration_params->ForceSingleDPP = false;
- CalculateSwathAndDETConfiguration_params->ODMMode = mode_lib->ms.ODMMode;
- CalculateSwathAndDETConfiguration_params->DPPPerSurface = mode_lib->ms.NoOfDPP;
-
- // output
- CalculateSwathAndDETConfiguration_params->req_per_swath_ub_l = s->dummy_integer_array[0];
- CalculateSwathAndDETConfiguration_params->req_per_swath_ub_c = s->dummy_integer_array[1];
- CalculateSwathAndDETConfiguration_params->swath_width_luma_ub = mode_lib->ms.swath_width_luma_ub;
- CalculateSwathAndDETConfiguration_params->swath_width_chroma_ub = mode_lib->ms.swath_width_chroma_ub;
- CalculateSwathAndDETConfiguration_params->SwathWidth = mode_lib->ms.SwathWidthY;
- CalculateSwathAndDETConfiguration_params->SwathWidthChroma = mode_lib->ms.SwathWidthC;
- CalculateSwathAndDETConfiguration_params->SwathHeightY = mode_lib->ms.SwathHeightY;
- CalculateSwathAndDETConfiguration_params->SwathHeightC = mode_lib->ms.SwathHeightC;
- CalculateSwathAndDETConfiguration_params->request_size_bytes_luma = mode_lib->ms.support.request_size_bytes_luma;
- CalculateSwathAndDETConfiguration_params->request_size_bytes_chroma = mode_lib->ms.support.request_size_bytes_chroma;
- CalculateSwathAndDETConfiguration_params->DETBufferSizeInKByte = mode_lib->ms.DETBufferSizeInKByte; // FIXME: This is per pipe but the pipes in plane will use that
- CalculateSwathAndDETConfiguration_params->DETBufferSizeY = mode_lib->ms.DETBufferSizeY;
- CalculateSwathAndDETConfiguration_params->DETBufferSizeC = mode_lib->ms.DETBufferSizeC;
- CalculateSwathAndDETConfiguration_params->UnboundedRequestEnabled = &mode_lib->ms.UnboundedRequestEnabled;
- CalculateSwathAndDETConfiguration_params->compbuf_reserved_space_64b = s->dummy_integer_array[3];
- CalculateSwathAndDETConfiguration_params->hw_debug5 = s->dummy_boolean_array[1];
- CalculateSwathAndDETConfiguration_params->CompressedBufferSizeInkByte = &mode_lib->ms.CompressedBufferSizeInkByte;
- CalculateSwathAndDETConfiguration_params->ViewportSizeSupportPerSurface = s->dummy_boolean_array[0];
- CalculateSwathAndDETConfiguration_params->ViewportSizeSupport = &mode_lib->ms.support.ViewportSizeSupport;
- CalculateSwathAndDETConfiguration_params->funcs = &mode_lib->funcs;
-
- CalculateSwathAndDETConfiguration(&mode_lib->scratch, CalculateSwathAndDETConfiguration_params);
-
- if (mode_lib->soc.mall_allocated_for_dcn_mbytes == 0) {
- for (k = 0; k < mode_lib->ms.num_active_planes; k++)
- mode_lib->ms.SurfaceSizeInMALL[k] = 0;
- mode_lib->ms.support.ExceededMALLSize = 0;
- } else {
- CalculateSurfaceSizeInMall(
- display_cfg,
- mode_lib->ms.num_active_planes,
- mode_lib->soc.mall_allocated_for_dcn_mbytes,
-
- mode_lib->ms.BytePerPixelY,
- mode_lib->ms.BytePerPixelC,
- mode_lib->ms.Read256BlockWidthY,
- mode_lib->ms.Read256BlockWidthC,
- mode_lib->ms.Read256BlockHeightY,
- mode_lib->ms.Read256BlockHeightC,
- mode_lib->ms.MacroTileWidthY,
- mode_lib->ms.MacroTileWidthC,
- mode_lib->ms.MacroTileHeightY,
- mode_lib->ms.MacroTileHeightC,
-
- /* Output */
- mode_lib->ms.SurfaceSizeInMALL,
- &mode_lib->ms.support.ExceededMALLSize);
- }
-
- mode_lib->ms.TotalNumberOfDCCActiveDPP = 0;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (display_cfg->plane_descriptors[k].surface.dcc.enable == true) {
- mode_lib->ms.TotalNumberOfDCCActiveDPP = mode_lib->ms.TotalNumberOfDCCActiveDPP + mode_lib->ms.NoOfDPP[k];
- }
- }
-
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- s->SurfParameters[k].PixelClock = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- s->SurfParameters[k].DPPPerSurface = mode_lib->ms.NoOfDPP[k];
- s->SurfParameters[k].RotationAngle = display_cfg->plane_descriptors[k].composition.rotation_angle;
- s->SurfParameters[k].ViewportHeight = display_cfg->plane_descriptors[k].composition.viewport.plane0.height;
- s->SurfParameters[k].ViewportHeightC = display_cfg->plane_descriptors[k].composition.viewport.plane1.height;
- s->SurfParameters[k].BlockWidth256BytesY = mode_lib->ms.Read256BlockWidthY[k];
- s->SurfParameters[k].BlockHeight256BytesY = mode_lib->ms.Read256BlockHeightY[k];
- s->SurfParameters[k].BlockWidth256BytesC = mode_lib->ms.Read256BlockWidthC[k];
- s->SurfParameters[k].BlockHeight256BytesC = mode_lib->ms.Read256BlockHeightC[k];
- s->SurfParameters[k].BlockWidthY = mode_lib->ms.MacroTileWidthY[k];
- s->SurfParameters[k].BlockHeightY = mode_lib->ms.MacroTileHeightY[k];
- s->SurfParameters[k].BlockWidthC = mode_lib->ms.MacroTileWidthC[k];
- s->SurfParameters[k].BlockHeightC = mode_lib->ms.MacroTileHeightC[k];
- s->SurfParameters[k].InterlaceEnable = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced;
- s->SurfParameters[k].HTotal = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total;
- s->SurfParameters[k].DCCEnable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- s->SurfParameters[k].SourcePixelFormat = display_cfg->plane_descriptors[k].pixel_format;
- s->SurfParameters[k].SurfaceTiling = display_cfg->plane_descriptors[k].surface.tiling;
- s->SurfParameters[k].BytePerPixelY = mode_lib->ms.BytePerPixelY[k];
- s->SurfParameters[k].BytePerPixelC = mode_lib->ms.BytePerPixelC[k];
- s->SurfParameters[k].ProgressiveToInterlaceUnitInOPP = mode_lib->ip.ptoi_supported;
- s->SurfParameters[k].VRatio = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- s->SurfParameters[k].VRatioChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- s->SurfParameters[k].VTaps = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps;
- s->SurfParameters[k].VTapsChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps;
- s->SurfParameters[k].PitchY = display_cfg->plane_descriptors[k].surface.plane0.pitch;
- s->SurfParameters[k].PitchC = display_cfg->plane_descriptors[k].surface.plane1.pitch;
- s->SurfParameters[k].ViewportStationary = display_cfg->plane_descriptors[k].composition.viewport.stationary;
- s->SurfParameters[k].ViewportXStart = display_cfg->plane_descriptors[k].composition.viewport.plane0.x_start;
- s->SurfParameters[k].ViewportYStart = display_cfg->plane_descriptors[k].composition.viewport.plane0.y_start;
- s->SurfParameters[k].ViewportXStartC = display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start;
- s->SurfParameters[k].ViewportYStartC = display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start;
- s->SurfParameters[k].FORCE_ONE_ROW_FOR_FRAME = display_cfg->plane_descriptors[k].overrides.hw.force_one_row_for_frame;
- s->SurfParameters[k].SwathHeightY = mode_lib->ms.SwathHeightY[k];
- s->SurfParameters[k].SwathHeightC = mode_lib->ms.SwathHeightC[k];
-
- s->SurfParameters[k].DCCMetaPitchY = display_cfg->plane_descriptors[k].surface.dcc.plane0.pitch;
- s->SurfParameters[k].DCCMetaPitchC = display_cfg->plane_descriptors[k].surface.dcc.plane1.pitch;
- }
-
- CalculateVMRowAndSwath_params->display_cfg = display_cfg;
- CalculateVMRowAndSwath_params->NumberOfActiveSurfaces = mode_lib->ms.num_active_planes;
- CalculateVMRowAndSwath_params->myPipe = s->SurfParameters;
- CalculateVMRowAndSwath_params->SurfaceSizeInMALL = mode_lib->ms.SurfaceSizeInMALL;
- CalculateVMRowAndSwath_params->PTEBufferSizeInRequestsLuma = mode_lib->ip.dpte_buffer_size_in_pte_reqs_luma;
- CalculateVMRowAndSwath_params->PTEBufferSizeInRequestsChroma = mode_lib->ip.dpte_buffer_size_in_pte_reqs_chroma;
- CalculateVMRowAndSwath_params->MALLAllocatedForDCN = mode_lib->soc.mall_allocated_for_dcn_mbytes;
- CalculateVMRowAndSwath_params->SwathWidthY = mode_lib->ms.SwathWidthY;
- CalculateVMRowAndSwath_params->SwathWidthC = mode_lib->ms.SwathWidthC;
- CalculateVMRowAndSwath_params->HostVMMinPageSize = mode_lib->soc.hostvm_min_page_size_kbytes;
- CalculateVMRowAndSwath_params->DCCMetaBufferSizeBytes = mode_lib->ip.dcc_meta_buffer_size_bytes;
- CalculateVMRowAndSwath_params->mrq_present = mode_lib->ip.dcn_mrq_present;
-
- // output
- CalculateVMRowAndSwath_params->PTEBufferSizeNotExceeded = mode_lib->ms.PTEBufferSizeNotExceeded;
- CalculateVMRowAndSwath_params->dpte_row_width_luma_ub = s->dummy_integer_array[12];
- CalculateVMRowAndSwath_params->dpte_row_width_chroma_ub = s->dummy_integer_array[13];
- CalculateVMRowAndSwath_params->dpte_row_height_luma = mode_lib->ms.dpte_row_height;
- CalculateVMRowAndSwath_params->dpte_row_height_chroma = mode_lib->ms.dpte_row_height_chroma;
- CalculateVMRowAndSwath_params->dpte_row_height_linear_luma = s->dummy_integer_array[14]; // VBA_DELTA
- CalculateVMRowAndSwath_params->dpte_row_height_linear_chroma = s->dummy_integer_array[15]; // VBA_DELTA
- CalculateVMRowAndSwath_params->vm_group_bytes = s->dummy_integer_array[16];
- CalculateVMRowAndSwath_params->dpte_group_bytes = mode_lib->ms.dpte_group_bytes;
- CalculateVMRowAndSwath_params->PixelPTEReqWidthY = s->dummy_integer_array[17];
- CalculateVMRowAndSwath_params->PixelPTEReqHeightY = s->dummy_integer_array[18];
- CalculateVMRowAndSwath_params->PTERequestSizeY = s->dummy_integer_array[19];
- CalculateVMRowAndSwath_params->PixelPTEReqWidthC = s->dummy_integer_array[20];
- CalculateVMRowAndSwath_params->PixelPTEReqHeightC = s->dummy_integer_array[21];
- CalculateVMRowAndSwath_params->PTERequestSizeC = s->dummy_integer_array[22];
- CalculateVMRowAndSwath_params->vmpg_width_y = s->vmpg_width_y;
- CalculateVMRowAndSwath_params->vmpg_height_y = s->vmpg_height_y;
- CalculateVMRowAndSwath_params->vmpg_width_c = s->vmpg_width_c;
- CalculateVMRowAndSwath_params->vmpg_height_c = s->vmpg_height_c;
- CalculateVMRowAndSwath_params->dpde0_bytes_per_frame_ub_l = s->dummy_integer_array[23];
- CalculateVMRowAndSwath_params->dpde0_bytes_per_frame_ub_c = s->dummy_integer_array[24];
- CalculateVMRowAndSwath_params->PrefetchSourceLinesY = mode_lib->ms.PrefetchLinesY;
- CalculateVMRowAndSwath_params->PrefetchSourceLinesC = mode_lib->ms.PrefetchLinesC;
- CalculateVMRowAndSwath_params->VInitPreFillY = mode_lib->ms.PrefillY;
- CalculateVMRowAndSwath_params->VInitPreFillC = mode_lib->ms.PrefillC;
- CalculateVMRowAndSwath_params->MaxNumSwathY = mode_lib->ms.MaxNumSwathY;
- CalculateVMRowAndSwath_params->MaxNumSwathC = mode_lib->ms.MaxNumSwathC;
- CalculateVMRowAndSwath_params->dpte_row_bw = mode_lib->ms.dpte_row_bw;
- CalculateVMRowAndSwath_params->PixelPTEBytesPerRow = mode_lib->ms.DPTEBytesPerRow;
- CalculateVMRowAndSwath_params->vm_bytes = mode_lib->ms.vm_bytes;
- CalculateVMRowAndSwath_params->use_one_row_for_frame = mode_lib->ms.use_one_row_for_frame;
- CalculateVMRowAndSwath_params->use_one_row_for_frame_flip = mode_lib->ms.use_one_row_for_frame_flip;
- CalculateVMRowAndSwath_params->is_using_mall_for_ss = s->dummy_boolean_array[0];
- CalculateVMRowAndSwath_params->PTE_BUFFER_MODE = s->dummy_boolean_array[1];
- CalculateVMRowAndSwath_params->BIGK_FRAGMENT_SIZE = s->dummy_integer_array[25];
- CalculateVMRowAndSwath_params->DCCMetaBufferSizeNotExceeded = mode_lib->ms.DCCMetaBufferSizeNotExceeded;
- CalculateVMRowAndSwath_params->meta_row_bw = mode_lib->ms.meta_row_bw;
- CalculateVMRowAndSwath_params->meta_row_bytes = mode_lib->ms.meta_row_bytes;
- CalculateVMRowAndSwath_params->meta_req_width_luma = s->dummy_integer_array[26];
- CalculateVMRowAndSwath_params->meta_req_height_luma = s->dummy_integer_array[27];
- CalculateVMRowAndSwath_params->meta_row_width_luma = s->dummy_integer_array[28];
- CalculateVMRowAndSwath_params->meta_row_height_luma = s->meta_row_height_luma;
- CalculateVMRowAndSwath_params->meta_pte_bytes_per_frame_ub_l = s->dummy_integer_array[29];
- CalculateVMRowAndSwath_params->meta_req_width_chroma = s->dummy_integer_array[30];
- CalculateVMRowAndSwath_params->meta_req_height_chroma = s->dummy_integer_array[31];
- CalculateVMRowAndSwath_params->meta_row_width_chroma = s->dummy_integer_array[32];
- CalculateVMRowAndSwath_params->meta_row_height_chroma = s->meta_row_height_chroma;
- CalculateVMRowAndSwath_params->meta_pte_bytes_per_frame_ub_c = s->dummy_integer_array[33];
-
- CalculateVMRowAndSwath(&mode_lib->scratch, CalculateVMRowAndSwath_params);
-
- mode_lib->ms.support.PTEBufferSizeNotExceeded = true;
- mode_lib->ms.support.DCCMetaBufferSizeNotExceeded = true;
-
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (mode_lib->ms.PTEBufferSizeNotExceeded[k] == false)
- mode_lib->ms.support.PTEBufferSizeNotExceeded = false;
-
- if (mode_lib->ms.DCCMetaBufferSizeNotExceeded[k] == false)
- mode_lib->ms.support.DCCMetaBufferSizeNotExceeded = false;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, PTEBufferSizeNotExceeded = %u\n", __func__, k, mode_lib->ms.PTEBufferSizeNotExceeded[k]);
- dml2_printf("DML::%s: k=%u, DCCMetaBufferSizeNotExceeded = %u\n", __func__, k, mode_lib->ms.DCCMetaBufferSizeNotExceeded[k]);
-#endif
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: PTEBufferSizeNotExceeded = %u\n", __func__, mode_lib->ms.support.PTEBufferSizeNotExceeded);
- dml2_printf("DML::%s: DCCMetaBufferSizeNotExceeded = %u\n", __func__, mode_lib->ms.support.DCCMetaBufferSizeNotExceeded);
-#endif
-
- mode_lib->ms.UrgLatency = CalculateUrgentLatency(
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.base_latency_us,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.base_latency_pixel_vm_us,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.base_latency_vm_us,
- mode_lib->soc.do_urgent_latency_adjustment,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.scaling_factor_fclk_us,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.scaling_factor_mhz,
- mode_lib->ms.FabricClock,
- mode_lib->ms.uclk_freq_mhz,
- mode_lib->soc.qos_parameters.qos_type,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->ms.qos_param_index].urgent_ramp_uclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.df_qos_response_time_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.max_round_trip_to_furthest_cs_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.mall_overhead_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_urgent_ramp_latency_margin,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin);
-
- mode_lib->ms.TripToMemory = CalculateTripToMemory(
- mode_lib->ms.UrgLatency,
- mode_lib->ms.FabricClock,
- mode_lib->ms.uclk_freq_mhz,
- mode_lib->soc.qos_parameters.qos_type,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->ms.qos_param_index].trip_to_memory_uclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.max_round_trip_to_furthest_cs_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.mall_overhead_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_max_latency_margin,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin);
-
- mode_lib->ms.TripToMemory = math_max2(mode_lib->ms.UrgLatency, mode_lib->ms.TripToMemory);
-
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- double line_time_us = (double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- calculate_cursor_req_attributes(
- display_cfg->plane_descriptors[k].cursor.cursor_width,
- display_cfg->plane_descriptors[k].cursor.cursor_bpp,
-
- // output
- &s->cursor_lines_per_chunk[k],
- &s->cursor_bytes_per_line[k],
- &s->cursor_bytes_per_chunk[k],
- &s->cursor_bytes[k]);
-
- bool cursor_not_enough_urgent_latency_hiding = 0;
- calculate_cursor_urgent_burst_factor(
- mode_lib->ip.cursor_buffer_size,
- display_cfg->plane_descriptors[k].cursor.cursor_width,
- s->cursor_bytes_per_chunk[k],
- s->cursor_lines_per_chunk[k],
- line_time_us,
- mode_lib->ms.UrgLatency,
-
- // output
- &mode_lib->ms.UrgentBurstFactorCursor[k],
- &cursor_not_enough_urgent_latency_hiding);
- mode_lib->ms.UrgentBurstFactorCursorPre[k] = mode_lib->ms.UrgentBurstFactorCursor[k];
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, Calling CalculateUrgentBurstFactor\n", __func__, k);
- dml2_printf("DML::%s: k=%d, VRatio=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
- dml2_printf("DML::%s: k=%d, VRatioChroma=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio);
-#endif
-
- CalculateUrgentBurstFactor(
- &display_cfg->plane_descriptors[k],
- mode_lib->ms.swath_width_luma_ub[k],
- mode_lib->ms.swath_width_chroma_ub[k],
- mode_lib->ms.SwathHeightY[k],
- mode_lib->ms.SwathHeightC[k],
- line_time_us,
- mode_lib->ms.UrgLatency,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio,
- mode_lib->ms.BytePerPixelInDETY[k],
- mode_lib->ms.BytePerPixelInDETC[k],
- mode_lib->ms.DETBufferSizeY[k],
- mode_lib->ms.DETBufferSizeC[k],
-
- // Output
- &mode_lib->ms.UrgentBurstFactorLuma[k],
- &mode_lib->ms.UrgentBurstFactorChroma[k],
- &mode_lib->ms.NotEnoughUrgentLatencyHiding[k]);
-
- mode_lib->ms.NotEnoughUrgentLatencyHiding[k] = mode_lib->ms.NotEnoughUrgentLatencyHiding[k] || cursor_not_enough_urgent_latency_hiding;
- }
-
- CalculateDCFCLKDeepSleep(
- display_cfg,
- mode_lib->ms.num_active_planes,
- mode_lib->ms.BytePerPixelY,
- mode_lib->ms.BytePerPixelC,
- mode_lib->ms.SwathWidthY,
- mode_lib->ms.SwathWidthC,
- mode_lib->ms.NoOfDPP,
- mode_lib->ms.PSCL_FACTOR,
- mode_lib->ms.PSCL_FACTOR_CHROMA,
- mode_lib->ms.RequiredDPPCLK,
- mode_lib->ms.SurfaceReadBandwidthLuma,
- mode_lib->ms.SurfaceReadBandwidthChroma,
- mode_lib->soc.return_bus_width_bytes,
-
- /* Output */
- &mode_lib->ms.dcfclk_deepsleep);
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->plane_descriptors[k].stream_index == k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- mode_lib->ms.WritebackDelayTime[k] = mode_lib->soc.qos_parameters.writeback.base_latency_us + CalculateWriteBackDelay(
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total) / mode_lib->ms.RequiredDISPCLK;
- } else {
- mode_lib->ms.WritebackDelayTime[k] = 0.0;
- }
- for (m = 0; m <= mode_lib->ms.num_active_planes - 1; m++) {
- if (display_cfg->plane_descriptors[m].stream_index == k && display_cfg->stream_descriptors[display_cfg->plane_descriptors[m].stream_index].writeback.enable == true) {
- mode_lib->ms.WritebackDelayTime[k] = math_max2(mode_lib->ms.WritebackDelayTime[k],
- mode_lib->soc.qos_parameters.writeback.base_latency_us + CalculateWriteBackDelay(
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[m].stream_index].writeback.pixel_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[m].stream_index].writeback.scaling_info.h_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[m].stream_index].writeback.scaling_info.v_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[m].stream_index].writeback.scaling_info.v_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[m].stream_index].writeback.scaling_info.output_width,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[m].stream_index].writeback.scaling_info.output_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[m].stream_index].writeback.scaling_info.input_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[m].stream_index].timing.h_total) / mode_lib->ms.RequiredDISPCLK);
- }
- }
- }
- }
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- for (m = 0; m <= mode_lib->ms.num_active_planes - 1; m++) {
- if (display_cfg->plane_descriptors[k].stream_index == m) {
- mode_lib->ms.WritebackDelayTime[k] = mode_lib->ms.WritebackDelayTime[m];
- }
- }
- }
-
- // MaximumVStartup is actually Tvstartup_min in DCN4 programming guide
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- bool isInterlaceTiming = (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced && !mode_lib->ip.ptoi_supported);
- s->MaximumVStartup[k] = CalculateMaxVStartup(
- mode_lib->ip.ptoi_supported,
- mode_lib->ip.vblank_nom_default_us,
- &display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing,
- mode_lib->ms.WritebackDelayTime[k]);
- mode_lib->ms.MaxVStartupLines[k] = (isInterlaceTiming ? (2 * s->MaximumVStartup[k]) : s->MaximumVStartup[k]);
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, MaximumVStartup = %u\n", __func__, k, s->MaximumVStartup[k]);
-#endif
-
- /* Immediate Flip and MALL parameters */
- s->ImmediateFlipRequired = false;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- s->ImmediateFlipRequired = s->ImmediateFlipRequired || display_cfg->plane_descriptors[k].immediate_flip;
- }
-
- mode_lib->ms.support.ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe = false;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- mode_lib->ms.support.ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe =
- mode_lib->ms.support.ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe ||
- ((display_cfg->hostvm_enable == true || display_cfg->plane_descriptors[k].immediate_flip == true) &&
- (display_cfg->plane_descriptors[k].overrides.uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_mall_full_frame || dml_is_phantom_pipe(&display_cfg->plane_descriptors[k])));
- }
-
- mode_lib->ms.support.InvalidCombinationOfMALLUseForPStateAndStaticScreen = false;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- mode_lib->ms.support.InvalidCombinationOfMALLUseForPStateAndStaticScreen = mode_lib->ms.support.InvalidCombinationOfMALLUseForPStateAndStaticScreen ||
- ((display_cfg->plane_descriptors[k].overrides.refresh_from_mall == dml2_refresh_from_mall_mode_override_force_enable || display_cfg->plane_descriptors[k].overrides.refresh_from_mall == dml2_refresh_from_mall_mode_override_auto) && (dml_is_phantom_pipe(&display_cfg->plane_descriptors[k]))) ||
- ((display_cfg->plane_descriptors[k].overrides.refresh_from_mall == dml2_refresh_from_mall_mode_override_force_disable || display_cfg->plane_descriptors[k].overrides.refresh_from_mall == dml2_refresh_from_mall_mode_override_auto) && (display_cfg->plane_descriptors[k].overrides.uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_mall_full_frame));
- }
-
- s->FullFrameMALLPStateMethod = false;
- s->SubViewportMALLPStateMethod = false;
- s->PhantomPipeMALLPStateMethod = false;
- s->SubViewportMALLRefreshGreaterThan120Hz = false;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (display_cfg->plane_descriptors[k].overrides.uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_mall_full_frame)
- s->FullFrameMALLPStateMethod = true;
- if (display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_main_pipe) {
- s->SubViewportMALLPStateMethod = true;
- if (!display_cfg->overrides.enable_subvp_implicit_pmo) {
- // For dv, small frame tests will have very high refresh rate
- unsigned long long refresh_rate = (unsigned long long) ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz * 1000 /
- (double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total /
- (double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total);
- if (refresh_rate > 120)
- s->SubViewportMALLRefreshGreaterThan120Hz = true;
- }
- }
- if (dml_is_phantom_pipe(&display_cfg->plane_descriptors[k]))
- s->PhantomPipeMALLPStateMethod = true;
- }
- mode_lib->ms.support.InvalidCombinationOfMALLUseForPState = (s->SubViewportMALLPStateMethod != s->PhantomPipeMALLPStateMethod) ||
- (s->SubViewportMALLPStateMethod && s->FullFrameMALLPStateMethod) || s->SubViewportMALLRefreshGreaterThan120Hz;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: SubViewportMALLPStateMethod = %u\n", __func__, s->SubViewportMALLPStateMethod);
- dml2_printf("DML::%s: PhantomPipeMALLPStateMethod = %u\n", __func__, s->PhantomPipeMALLPStateMethod);
- dml2_printf("DML::%s: FullFrameMALLPStateMethod = %u\n", __func__, s->FullFrameMALLPStateMethod);
- dml2_printf("DML::%s: SubViewportMALLRefreshGreaterThan120Hz = %u\n", __func__, s->SubViewportMALLRefreshGreaterThan120Hz);
- dml2_printf("DML::%s: InvalidCombinationOfMALLUseForPState = %u\n", __func__, mode_lib->ms.support.InvalidCombinationOfMALLUseForPState);
- dml2_printf("DML::%s: in_out_params->min_clk_index = %u\n", __func__, in_out_params->min_clk_index);
- dml2_printf("DML::%s: mode_lib->ms.DCFCLK = %f\n", __func__, mode_lib->ms.DCFCLK);
- dml2_printf("DML::%s: mode_lib->ms.FabricClock = %f\n", __func__, mode_lib->ms.FabricClock);
- dml2_printf("DML::%s: mode_lib->ms.uclk_freq_mhz = %f\n", __func__, mode_lib->ms.uclk_freq_mhz);
- dml2_printf("DML::%s: max_urgent_latency_us = %f\n", __func__, mode_lib->ms.support.max_urgent_latency_us);
- dml2_printf("DML::%s: urgent latency tolerance = %f\n", __func__, ((mode_lib->ip.rob_buffer_size_kbytes - mode_lib->ip.pixel_chunk_size_kbytes) * 1024 / (mode_lib->ms.DCFCLK * mode_lib->soc.return_bus_width_bytes)));
-#endif
-
- mode_lib->ms.support.OutstandingRequestsSupport = true;
- mode_lib->ms.support.OutstandingRequestsUrgencyAvoidance = true;
-
- mode_lib->ms.support.avg_urgent_latency_us
- = (mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->ms.qos_param_index].average_latency_when_urgent_uclk_cycles / mode_lib->ms.uclk_freq_mhz
- * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_average_latency_margin / 100.0)
- + mode_lib->soc.qos_parameters.qos_params.dcn4x.average_transport_distance_fclk_cycles / mode_lib->ms.FabricClock)
- * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_average_transport_latency_margin / 100.0);
-
- mode_lib->ms.support.avg_non_urgent_latency_us
- = (mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->ms.qos_param_index].average_latency_when_non_urgent_uclk_cycles / mode_lib->ms.uclk_freq_mhz
- * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_average_latency_margin / 100.0)
- + mode_lib->soc.qos_parameters.qos_params.dcn4x.average_transport_distance_fclk_cycles / mode_lib->ms.FabricClock)
- * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_average_transport_latency_margin / 100.0);
-
- double outstanding_latency_us = 0;
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
-
- if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn4x) {
- outstanding_latency_us = (mode_lib->soc.max_outstanding_reqs * mode_lib->ms.support.request_size_bytes_luma[k]
- / (mode_lib->ms.DCFCLK * mode_lib->soc.return_bus_width_bytes));
-
- if (outstanding_latency_us < mode_lib->ms.support.avg_urgent_latency_us) {
- mode_lib->ms.support.OutstandingRequestsSupport = false;
- }
-
- if (outstanding_latency_us < mode_lib->ms.support.avg_non_urgent_latency_us) {
- mode_lib->ms.support.OutstandingRequestsUrgencyAvoidance = false;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: avg_urgent_latency_us = %f\n", __func__, mode_lib->ms.support.avg_urgent_latency_us);
- dml2_printf("DML::%s: avg_non_urgent_latency_us = %f\n", __func__, mode_lib->ms.support.avg_non_urgent_latency_us);
- dml2_printf("DML::%s: k=%d, request_size_bytes_luma = %d\n", __func__, k, mode_lib->ms.support.request_size_bytes_luma[k]);
- dml2_printf("DML::%s: k=%d, outstanding_latency_us = %f (luma)\n", __func__, k, outstanding_latency_us);
-#endif
- }
-
- if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn4x && mode_lib->ms.BytePerPixelC[k] > 0) {
- outstanding_latency_us = (mode_lib->soc.max_outstanding_reqs * mode_lib->ms.support.request_size_bytes_chroma[k]
- / (mode_lib->ms.DCFCLK * mode_lib->soc.return_bus_width_bytes));
-
- if (outstanding_latency_us < mode_lib->ms.support.avg_urgent_latency_us) {
- mode_lib->ms.support.OutstandingRequestsSupport = false;
- }
-
- if (outstanding_latency_us < mode_lib->ms.support.avg_non_urgent_latency_us) {
- mode_lib->ms.support.OutstandingRequestsUrgencyAvoidance = false;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, request_size_bytes_chroma = %d\n", __func__, k, mode_lib->ms.support.request_size_bytes_chroma[k]);
- dml2_printf("DML::%s: k=%d, outstanding_latency_us = %f (chroma)\n", __func__, k, outstanding_latency_us);
-#endif
- }
- }
-
- memset(calculate_mcache_setting_params, 0, sizeof(struct dml2_core_calcs_calculate_mcache_setting_params));
- if (mode_lib->soc.mall_allocated_for_dcn_mbytes == 0 || mode_lib->ip.dcn_mrq_present) {
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- mode_lib->ms.mall_prefetch_sdp_overhead_factor[k] = 1.0;
- mode_lib->ms.mall_prefetch_dram_overhead_factor[k] = 1.0;
- mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0[k] = 1.0;
- mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p0[k] = 1.0;
- mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1[k] = 1.0;
- mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p1[k] = 1.0;
- }
- } else {
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- calculate_mcache_setting_params->dcc_enable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- calculate_mcache_setting_params->num_chans = mode_lib->soc.clk_table.dram_config.channel_count;
- calculate_mcache_setting_params->mem_word_bytes = mode_lib->soc.mem_word_bytes;
- calculate_mcache_setting_params->mcache_size_bytes = mode_lib->soc.mcache_size_bytes;
- calculate_mcache_setting_params->mcache_line_size_bytes = mode_lib->soc.mcache_line_size_bytes;
- calculate_mcache_setting_params->gpuvm_enable = display_cfg->gpuvm_enable;
- calculate_mcache_setting_params->gpuvm_page_size_kbytes = display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes;
-
- calculate_mcache_setting_params->source_format = display_cfg->plane_descriptors[k].pixel_format;
- calculate_mcache_setting_params->surf_vert = dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle);
- calculate_mcache_setting_params->vp_stationary = display_cfg->plane_descriptors[k].composition.viewport.stationary;
- calculate_mcache_setting_params->tiling_mode = display_cfg->plane_descriptors[k].surface.tiling;
- calculate_mcache_setting_params->imall_enable = mode_lib->ip.imall_supported && display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_imall;
-
- calculate_mcache_setting_params->vp_start_x_l = display_cfg->plane_descriptors[k].composition.viewport.plane0.x_start;
- calculate_mcache_setting_params->vp_start_y_l = display_cfg->plane_descriptors[k].composition.viewport.plane0.y_start;
- calculate_mcache_setting_params->full_vp_width_l = display_cfg->plane_descriptors[k].composition.viewport.plane0.width;
- calculate_mcache_setting_params->full_vp_height_l = display_cfg->plane_descriptors[k].composition.viewport.plane0.height;
- calculate_mcache_setting_params->blk_width_l = mode_lib->ms.MacroTileWidthY[k];
- calculate_mcache_setting_params->blk_height_l = mode_lib->ms.MacroTileHeightY[k];
- calculate_mcache_setting_params->vmpg_width_l = s->vmpg_width_y[k];
- calculate_mcache_setting_params->vmpg_height_l = s->vmpg_height_y[k];
- calculate_mcache_setting_params->full_swath_bytes_l = s->full_swath_bytes_l[k];
- calculate_mcache_setting_params->bytes_per_pixel_l = mode_lib->ms.BytePerPixelY[k];
-
- calculate_mcache_setting_params->vp_start_x_c = display_cfg->plane_descriptors[k].composition.viewport.plane1.x_start;
- calculate_mcache_setting_params->vp_start_y_c = display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start;
- calculate_mcache_setting_params->full_vp_width_c = display_cfg->plane_descriptors[k].composition.viewport.plane1.width;
- calculate_mcache_setting_params->full_vp_height_c = display_cfg->plane_descriptors[k].composition.viewport.plane1.height;
- calculate_mcache_setting_params->blk_width_c = mode_lib->ms.MacroTileWidthC[k];
- calculate_mcache_setting_params->blk_height_c = mode_lib->ms.MacroTileHeightC[k];
- calculate_mcache_setting_params->vmpg_width_c = s->vmpg_width_c[k];
- calculate_mcache_setting_params->vmpg_height_c = s->vmpg_height_c[k];
- calculate_mcache_setting_params->full_swath_bytes_c = s->full_swath_bytes_c[k];
- calculate_mcache_setting_params->bytes_per_pixel_c = mode_lib->ms.BytePerPixelC[k];
-
- // output
- calculate_mcache_setting_params->dcc_dram_bw_nom_overhead_factor_l = &mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0[k];
- calculate_mcache_setting_params->dcc_dram_bw_pref_overhead_factor_l = &mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p0[k];
- calculate_mcache_setting_params->dcc_dram_bw_nom_overhead_factor_c = &mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1[k];
- calculate_mcache_setting_params->dcc_dram_bw_pref_overhead_factor_c = &mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p1[k];
-
- calculate_mcache_setting_params->num_mcaches_l = &mode_lib->ms.num_mcaches_l[k];
- calculate_mcache_setting_params->mcache_row_bytes_l = &mode_lib->ms.mcache_row_bytes_l[k];
- calculate_mcache_setting_params->mcache_offsets_l = mode_lib->ms.mcache_offsets_l[k];
- calculate_mcache_setting_params->mcache_shift_granularity_l = &mode_lib->ms.mcache_shift_granularity_l[k];
-
- calculate_mcache_setting_params->num_mcaches_c = &mode_lib->ms.num_mcaches_c[k];
- calculate_mcache_setting_params->mcache_row_bytes_c = &mode_lib->ms.mcache_row_bytes_c[k];
- calculate_mcache_setting_params->mcache_offsets_c = mode_lib->ms.mcache_offsets_c[k];
- calculate_mcache_setting_params->mcache_shift_granularity_c = &mode_lib->ms.mcache_shift_granularity_c[k];
-
- calculate_mcache_setting_params->mall_comb_mcache_l = &mode_lib->ms.mall_comb_mcache_l[k];
- calculate_mcache_setting_params->mall_comb_mcache_c = &mode_lib->ms.mall_comb_mcache_c[k];
- calculate_mcache_setting_params->lc_comb_mcache = &mode_lib->ms.lc_comb_mcache[k];
-
- calculate_mcache_setting(&mode_lib->scratch, calculate_mcache_setting_params);
- }
-
- calculate_mall_bw_overhead_factor(
- mode_lib->ms.mall_prefetch_sdp_overhead_factor,
- mode_lib->ms.mall_prefetch_dram_overhead_factor,
-
- // input
- display_cfg,
- mode_lib->ms.num_active_planes);
- }
-
- // Calculate all the bandwidth available
- // Need anothe bw for latency evaluation
- calculate_bandwidth_available(
- mode_lib->ms.support.avg_bandwidth_available_min, // not used
- mode_lib->ms.support.avg_bandwidth_available, // not used
- mode_lib->ms.support.urg_bandwidth_available_min_latency,
- mode_lib->ms.support.urg_bandwidth_available, // not used
- mode_lib->ms.support.urg_bandwidth_available_vm_only, // not used
- mode_lib->ms.support.urg_bandwidth_available_pixel_and_vm, // not used
-
- &mode_lib->soc,
- display_cfg->hostvm_enable,
- mode_lib->ms.DCFCLK,
- mode_lib->ms.FabricClock,
- mode_lib->ms.dram_bw_mbps);
-
- calculate_bandwidth_available(
- mode_lib->ms.support.avg_bandwidth_available_min,
- mode_lib->ms.support.avg_bandwidth_available,
- mode_lib->ms.support.urg_bandwidth_available_min,
- mode_lib->ms.support.urg_bandwidth_available,
- mode_lib->ms.support.urg_bandwidth_available_vm_only,
- mode_lib->ms.support.urg_bandwidth_available_pixel_and_vm,
-
- &mode_lib->soc,
- display_cfg->hostvm_enable,
- mode_lib->ms.MaxDCFCLK,
- mode_lib->ms.MaxFabricClock,
- mode_lib->ms.dram_bw_mbps);
-
-
- // Average BW support check
- calculate_avg_bandwidth_required(
- mode_lib->ms.support.avg_bandwidth_required,
- // input
- display_cfg,
- mode_lib->ms.num_active_planes,
- mode_lib->ms.SurfaceReadBandwidthLuma,
- mode_lib->ms.SurfaceReadBandwidthChroma,
- mode_lib->ms.cursor_bw,
- mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0,
- mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1,
- mode_lib->ms.mall_prefetch_dram_overhead_factor,
- mode_lib->ms.mall_prefetch_sdp_overhead_factor);
-
- for (m = 0; m < dml2_core_internal_bw_max; m++) { // check sdp and dram
- mode_lib->ms.support.avg_bandwidth_support_ok[dml2_core_internal_soc_state_sys_idle][m] = 1;
- mode_lib->ms.support.avg_bandwidth_support_ok[dml2_core_internal_soc_state_sys_active][m] = (mode_lib->ms.support.avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][m] <= mode_lib->ms.support.avg_bandwidth_available[dml2_core_internal_soc_state_sys_active][m]);
- mode_lib->ms.support.avg_bandwidth_support_ok[dml2_core_internal_soc_state_svp_prefetch][m] = (mode_lib->ms.support.avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][m] <= mode_lib->ms.support.avg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][m]);
- }
-
- mode_lib->ms.support.AvgBandwidthSupport = true;
- mode_lib->ms.support.EnoughUrgentLatencyHidingSupport = true;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (mode_lib->ms.NotEnoughUrgentLatencyHiding[k]) {
- mode_lib->ms.support.EnoughUrgentLatencyHidingSupport = false;
- dml2_printf("DML::%s: k=%u NotEnoughUrgentLatencyHiding set\n", __func__, k);
-
- }
- }
- for (m = 0; m < dml2_core_internal_soc_state_max; m++) {
- for (n = 0; n < dml2_core_internal_bw_max; n++) { // check sdp and dram
- if (!mode_lib->ms.support.avg_bandwidth_support_ok[m][n] && (m == dml2_core_internal_soc_state_sys_active || mode_lib->soc.mall_allocated_for_dcn_mbytes > 0)) {
- mode_lib->ms.support.AvgBandwidthSupport = false;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: avg_bandwidth_support_ok[%s][%s] not ok\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n));
-#endif
- }
- }
- }
-
- /* Prefetch Check */
- {
- mode_lib->ms.TimeCalc = 24 / mode_lib->ms.dcfclk_deepsleep;
-
-
- calculate_hostvm_inefficiency_factor(
- &s->HostVMInefficiencyFactor,
- &s->HostVMInefficiencyFactorPrefetch,
-
- display_cfg->gpuvm_enable,
- display_cfg->hostvm_enable,
- mode_lib->ip.remote_iommu_outstanding_translations,
- mode_lib->soc.max_outstanding_reqs,
- mode_lib->ms.support.urg_bandwidth_available_pixel_and_vm[dml2_core_internal_soc_state_sys_active],
- mode_lib->ms.support.urg_bandwidth_available_vm_only[dml2_core_internal_soc_state_sys_active]);
-
- mode_lib->ms.Total3dlutActive = 0;
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut)
- mode_lib->ms.Total3dlutActive = mode_lib->ms.Total3dlutActive + 1;
-
- // Calculate tdlut schedule related terms
- calculate_tdlut_setting_params->dispclk_mhz = mode_lib->ms.RequiredDISPCLK;
- calculate_tdlut_setting_params->setup_for_tdlut = display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut;
- calculate_tdlut_setting_params->tdlut_width_mode = display_cfg->plane_descriptors[k].tdlut.tdlut_width_mode;
- calculate_tdlut_setting_params->tdlut_addressing_mode = display_cfg->plane_descriptors[k].tdlut.tdlut_addressing_mode;
- calculate_tdlut_setting_params->cursor_buffer_size = mode_lib->ip.cursor_buffer_size;
- calculate_tdlut_setting_params->gpuvm_enable = display_cfg->gpuvm_enable;
- calculate_tdlut_setting_params->gpuvm_page_size_kbytes = display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes;
- calculate_tdlut_setting_params->tdlut_mpc_width_flag = display_cfg->plane_descriptors[k].tdlut.tdlut_mpc_width_flag;
- calculate_tdlut_setting_params->is_gfx11 = dml_get_gfx_version(display_cfg->plane_descriptors[k].surface.tiling);
-
- // output
- calculate_tdlut_setting_params->tdlut_pte_bytes_per_frame = &s->tdlut_pte_bytes_per_frame[k];
- calculate_tdlut_setting_params->tdlut_bytes_per_frame = &s->tdlut_bytes_per_frame[k];
- calculate_tdlut_setting_params->tdlut_groups_per_2row_ub = &s->tdlut_groups_per_2row_ub[k];
- calculate_tdlut_setting_params->tdlut_opt_time = &s->tdlut_opt_time[k];
- calculate_tdlut_setting_params->tdlut_drain_time = &s->tdlut_drain_time[k];
- calculate_tdlut_setting_params->tdlut_bytes_per_group = &s->tdlut_bytes_per_group[k];
-
- calculate_tdlut_setting(&mode_lib->scratch, calculate_tdlut_setting_params);
- }
-
- double min_return_bw_for_latency = mode_lib->ms.support.urg_bandwidth_available_min_latency[dml2_core_internal_soc_state_sys_active];
- if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn3)
- s->ReorderingBytes = (unsigned int)(mode_lib->soc.clk_table.dram_config.channel_count * math_max3(mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_pixel_only_bytes,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_pixel_and_vm_bytes,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_vm_only_bytes));
-
- CalculateExtraLatency(
- display_cfg,
- mode_lib->ip.rob_buffer_size_kbytes,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.loaded_round_trip_latency_fclk_cycles,
- s->ReorderingBytes,
- mode_lib->ms.DCFCLK,
- mode_lib->ms.FabricClock,
- mode_lib->ip.pixel_chunk_size_kbytes,
- min_return_bw_for_latency,
- mode_lib->ms.num_active_planes,
- mode_lib->ms.NoOfDPP,
- mode_lib->ms.dpte_group_bytes,
- s->tdlut_bytes_per_group,
- s->HostVMInefficiencyFactor,
- s->HostVMInefficiencyFactorPrefetch,
- mode_lib->soc.hostvm_min_page_size_kbytes,
- mode_lib->soc.qos_parameters.qos_type,
- !(display_cfg->overrides.max_outstanding_when_urgent_expected_disable),
- mode_lib->soc.max_outstanding_reqs,
- mode_lib->ms.support.request_size_bytes_luma,
- mode_lib->ms.support.request_size_bytes_chroma,
- mode_lib->ip.meta_chunk_size_kbytes,
- mode_lib->ip.dchub_arb_to_ret_delay,
- mode_lib->ms.TripToMemory,
- mode_lib->ip.hostvm_mode,
-
- // output
- &mode_lib->ms.ExtraLatency,
- &mode_lib->ms.ExtraLatency_sr,
- &mode_lib->ms.ExtraLatencyPrefetch);
-
- {
- mode_lib->ms.support.PrefetchSupported = true;
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
-
- mode_lib->ms.TWait[k] = CalculateTWait(
- display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns,
- mode_lib->ms.UrgLatency,
- mode_lib->ms.TripToMemory);
-
- struct dml2_core_internal_DmlPipe *myPipe = &s->myPipe;
- myPipe->Dppclk = mode_lib->ms.RequiredDPPCLK[k];
- myPipe->Dispclk = mode_lib->ms.RequiredDISPCLK;
- myPipe->PixelClock = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- myPipe->DCFClkDeepSleep = mode_lib->ms.dcfclk_deepsleep;
- myPipe->DPPPerSurface = mode_lib->ms.NoOfDPP[k];
- myPipe->ScalerEnabled = display_cfg->plane_descriptors[k].composition.scaler_info.enabled;
- myPipe->VRatio = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- myPipe->VRatioChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- myPipe->VTaps = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps;
- myPipe->VTapsChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps;
- myPipe->RotationAngle = display_cfg->plane_descriptors[k].composition.rotation_angle;
- myPipe->mirrored = display_cfg->plane_descriptors[k].composition.mirrored;
- myPipe->BlockWidth256BytesY = mode_lib->ms.Read256BlockWidthY[k];
- myPipe->BlockHeight256BytesY = mode_lib->ms.Read256BlockHeightY[k];
- myPipe->BlockWidth256BytesC = mode_lib->ms.Read256BlockWidthC[k];
- myPipe->BlockHeight256BytesC = mode_lib->ms.Read256BlockHeightC[k];
- myPipe->InterlaceEnable = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced;
- myPipe->NumberOfCursors = display_cfg->plane_descriptors[k].cursor.num_cursors;
- myPipe->VBlank = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active;
- myPipe->HTotal = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total;
- myPipe->HActive = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active;
- myPipe->DCCEnable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- myPipe->ODMMode = mode_lib->ms.ODMMode[k];
- myPipe->SourcePixelFormat = display_cfg->plane_descriptors[k].pixel_format;
- myPipe->BytePerPixelY = mode_lib->ms.BytePerPixelY[k];
- myPipe->BytePerPixelC = mode_lib->ms.BytePerPixelC[k];
- myPipe->ProgressiveToInterlaceUnitInOPP = mode_lib->ip.ptoi_supported;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Calling CalculatePrefetchSchedule for k=%u\n", __func__, k);
- dml2_printf("DML::%s: MaximumVStartup = %u\n", __func__, s->MaximumVStartup[k]);
-#endif
- CalculatePrefetchSchedule_params->display_cfg = display_cfg;
- CalculatePrefetchSchedule_params->HostVMInefficiencyFactor = s->HostVMInefficiencyFactorPrefetch;
- CalculatePrefetchSchedule_params->myPipe = myPipe;
- CalculatePrefetchSchedule_params->DSCDelay = mode_lib->ms.DSCDelay[k];
- CalculatePrefetchSchedule_params->DPPCLKDelaySubtotalPlusCNVCFormater = mode_lib->ip.dppclk_delay_subtotal + mode_lib->ip.dppclk_delay_cnvc_formatter;
- CalculatePrefetchSchedule_params->DPPCLKDelaySCL = mode_lib->ip.dppclk_delay_scl;
- CalculatePrefetchSchedule_params->DPPCLKDelaySCLLBOnly = mode_lib->ip.dppclk_delay_scl_lb_only;
- CalculatePrefetchSchedule_params->DPPCLKDelayCNVCCursor = mode_lib->ip.dppclk_delay_cnvc_cursor;
- CalculatePrefetchSchedule_params->DISPCLKDelaySubtotal = mode_lib->ip.dispclk_delay_subtotal;
- CalculatePrefetchSchedule_params->DPP_RECOUT_WIDTH = (unsigned int)(mode_lib->ms.SwathWidthY[k] / display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
- CalculatePrefetchSchedule_params->OutputFormat = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format;
- CalculatePrefetchSchedule_params->MaxInterDCNTileRepeaters = mode_lib->ip.max_inter_dcn_tile_repeaters;
- CalculatePrefetchSchedule_params->VStartup = s->MaximumVStartup[k];
- CalculatePrefetchSchedule_params->MaxVStartup = s->MaximumVStartup[k];
- CalculatePrefetchSchedule_params->HostVMMinPageSize = mode_lib->soc.hostvm_min_page_size_kbytes;
- CalculatePrefetchSchedule_params->DynamicMetadataEnable = display_cfg->plane_descriptors[k].dynamic_meta_data.enable;
- CalculatePrefetchSchedule_params->DynamicMetadataVMEnabled = mode_lib->ip.dynamic_metadata_vm_enabled;
- CalculatePrefetchSchedule_params->DynamicMetadataLinesBeforeActiveRequired = display_cfg->plane_descriptors[k].dynamic_meta_data.lines_before_active_required;
- CalculatePrefetchSchedule_params->DynamicMetadataTransmittedBytes = display_cfg->plane_descriptors[k].dynamic_meta_data.transmitted_bytes;
- CalculatePrefetchSchedule_params->UrgentLatency = mode_lib->ms.UrgLatency;
- CalculatePrefetchSchedule_params->ExtraLatencyPrefetch = mode_lib->ms.ExtraLatencyPrefetch;
- CalculatePrefetchSchedule_params->TCalc = mode_lib->ms.TimeCalc;
- CalculatePrefetchSchedule_params->vm_bytes = mode_lib->ms.vm_bytes[k];
- CalculatePrefetchSchedule_params->PixelPTEBytesPerRow = mode_lib->ms.DPTEBytesPerRow[k];
- CalculatePrefetchSchedule_params->PrefetchSourceLinesY = mode_lib->ms.PrefetchLinesY[k];
- CalculatePrefetchSchedule_params->VInitPreFillY = mode_lib->ms.PrefillY[k];
- CalculatePrefetchSchedule_params->MaxNumSwathY = mode_lib->ms.MaxNumSwathY[k];
- CalculatePrefetchSchedule_params->PrefetchSourceLinesC = mode_lib->ms.PrefetchLinesC[k];
- CalculatePrefetchSchedule_params->VInitPreFillC = mode_lib->ms.PrefillC[k];
- CalculatePrefetchSchedule_params->MaxNumSwathC = mode_lib->ms.MaxNumSwathC[k];
- CalculatePrefetchSchedule_params->swath_width_luma_ub = mode_lib->ms.swath_width_luma_ub[k];
- CalculatePrefetchSchedule_params->swath_width_chroma_ub = mode_lib->ms.swath_width_chroma_ub[k];
- CalculatePrefetchSchedule_params->SwathHeightY = mode_lib->ms.SwathHeightY[k];
- CalculatePrefetchSchedule_params->SwathHeightC = mode_lib->ms.SwathHeightC[k];
- CalculatePrefetchSchedule_params->TWait = mode_lib->ms.TWait[k];
- CalculatePrefetchSchedule_params->Ttrip = mode_lib->ms.TripToMemory;
- CalculatePrefetchSchedule_params->setup_for_tdlut = display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut;
- CalculatePrefetchSchedule_params->tdlut_pte_bytes_per_frame = s->tdlut_pte_bytes_per_frame[k];
- CalculatePrefetchSchedule_params->tdlut_bytes_per_frame = s->tdlut_bytes_per_frame[k];
- CalculatePrefetchSchedule_params->tdlut_opt_time = s->tdlut_opt_time[k];
- CalculatePrefetchSchedule_params->tdlut_drain_time = s->tdlut_drain_time[k];
- CalculatePrefetchSchedule_params->num_cursors = (display_cfg->plane_descriptors[k].cursor.cursor_width > 0);
- CalculatePrefetchSchedule_params->cursor_bytes_per_chunk = s->cursor_bytes_per_chunk[k];
- CalculatePrefetchSchedule_params->cursor_bytes_per_line = s->cursor_bytes_per_line[k];
- CalculatePrefetchSchedule_params->dcc_enable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- CalculatePrefetchSchedule_params->mrq_present = mode_lib->ip.dcn_mrq_present;
- CalculatePrefetchSchedule_params->meta_row_bytes = mode_lib->ms.meta_row_bytes[k];
- CalculatePrefetchSchedule_params->mall_prefetch_sdp_overhead_factor = mode_lib->ms.mall_prefetch_sdp_overhead_factor[k];
-
- // output
- CalculatePrefetchSchedule_params->DSTXAfterScaler = &s->DSTXAfterScaler[k];
- CalculatePrefetchSchedule_params->DSTYAfterScaler = &s->DSTYAfterScaler[k];
- CalculatePrefetchSchedule_params->dst_y_prefetch = &mode_lib->ms.dst_y_prefetch[k];
- CalculatePrefetchSchedule_params->dst_y_per_vm_vblank = &mode_lib->ms.LinesForVM[k];
- CalculatePrefetchSchedule_params->dst_y_per_row_vblank = &mode_lib->ms.LinesForDPTERow[k];
- CalculatePrefetchSchedule_params->VRatioPrefetchY = &mode_lib->ms.VRatioPreY[k];
- CalculatePrefetchSchedule_params->VRatioPrefetchC = &mode_lib->ms.VRatioPreC[k];
- CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWLuma = &mode_lib->ms.RequiredPrefetchPixelDataBWLuma[k]; // prefetch_sw_bw_l
- CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWChroma = &mode_lib->ms.RequiredPrefetchPixelDataBWChroma[k]; // prefetch_sw_bw_c
- CalculatePrefetchSchedule_params->NotEnoughTimeForDynamicMetadata = &mode_lib->ms.NoTimeForDynamicMetadata[k];
- CalculatePrefetchSchedule_params->Tno_bw = &mode_lib->ms.Tno_bw[k];
- CalculatePrefetchSchedule_params->Tno_bw_flip = &mode_lib->ms.Tno_bw_flip[k];
- CalculatePrefetchSchedule_params->prefetch_vmrow_bw = &mode_lib->ms.prefetch_vmrow_bw[k];
- CalculatePrefetchSchedule_params->Tdmdl_vm = &s->dummy_single[0];
- CalculatePrefetchSchedule_params->Tdmdl = &s->dummy_single[1];
- CalculatePrefetchSchedule_params->TSetup = &s->dummy_single[2];
- CalculatePrefetchSchedule_params->Tvm_trips = &s->Tvm_trips[k];
- CalculatePrefetchSchedule_params->Tr0_trips = &s->Tr0_trips[k];
- CalculatePrefetchSchedule_params->Tvm_trips_flip = &s->Tvm_trips_flip[k];
- CalculatePrefetchSchedule_params->Tr0_trips_flip = &s->Tr0_trips_flip[k];
- CalculatePrefetchSchedule_params->Tvm_trips_flip_rounded = &s->Tvm_trips_flip_rounded[k];
- CalculatePrefetchSchedule_params->Tr0_trips_flip_rounded = &s->Tr0_trips_flip_rounded[k];
- CalculatePrefetchSchedule_params->VUpdateOffsetPix = &s->dummy_integer[0];
- CalculatePrefetchSchedule_params->VUpdateWidthPix = &s->dummy_integer[1];
- CalculatePrefetchSchedule_params->VReadyOffsetPix = &s->dummy_integer[2];
- CalculatePrefetchSchedule_params->prefetch_cursor_bw = &mode_lib->ms.prefetch_cursor_bw[k];
-
- mode_lib->ms.NoTimeForPrefetch[k] = CalculatePrefetchSchedule(&mode_lib->scratch, CalculatePrefetchSchedule_params);
-
- mode_lib->ms.support.PrefetchSupported &= !mode_lib->ms.NoTimeForPrefetch[k];
- dml2_printf("DML::%s: k=%d, dst_y_per_vm_vblank = %f\n", __func__, k, *CalculatePrefetchSchedule_params->dst_y_per_vm_vblank);
- dml2_printf("DML::%s: k=%d, dst_y_per_row_vblank = %f\n", __func__, k, *CalculatePrefetchSchedule_params->dst_y_per_row_vblank);
- } // for k num_planes
-
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (mode_lib->ms.dst_y_prefetch[k] < 2.0
- || mode_lib->ms.LinesForVM[k] >= 32.0
- || mode_lib->ms.LinesForDPTERow[k] >= 16.0
- || mode_lib->ms.NoTimeForPrefetch[k] == true
- || s->DSTYAfterScaler[k] > 8) {
- mode_lib->ms.support.PrefetchSupported = false;
- dml2_printf("DML::%s: k=%d, dst_y_prefetch=%f (should not be < 2)\n", __func__, k, mode_lib->ms.dst_y_prefetch[k]);
- dml2_printf("DML::%s: k=%d, LinesForVM=%f (should not be >= 32)\n", __func__, k, mode_lib->ms.LinesForVM[k]);
- dml2_printf("DML::%s: k=%d, LinesForDPTERow=%f (should not be >= 16)\n", __func__, k, mode_lib->ms.LinesForDPTERow[k]);
- dml2_printf("DML::%s: k=%d, NoTimeForPrefetch=%d\n", __func__, k, mode_lib->ms.NoTimeForPrefetch[k]);
- dml2_printf("DML::%s: k=%d, DSTYAfterScaler=%d (should be <= 8)\n", __func__, k, s->DSTYAfterScaler[k]);
- }
- }
-
- mode_lib->ms.support.DynamicMetadataSupported = true;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (mode_lib->ms.NoTimeForDynamicMetadata[k] == true) {
- mode_lib->ms.support.DynamicMetadataSupported = false;
- }
- }
-
- mode_lib->ms.support.VRatioInPrefetchSupported = true;
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (mode_lib->ms.VRatioPreY[k] > __DML2_CALCS_MAX_VRATIO_PRE_ENHANCE_PREFETCH_ACC__ ||
- mode_lib->ms.VRatioPreC[k] > __DML2_CALCS_MAX_VRATIO_PRE_ENHANCE_PREFETCH_ACC__) {
- mode_lib->ms.support.VRatioInPrefetchSupported = false;
- dml2_printf("DML::%s: VRatioInPrefetchSupported = %u\n", __func__, mode_lib->ms.support.VRatioInPrefetchSupported);
- }
- }
-
- s->AnyLinesForVMOrRowTooLarge = false;
- for (k = 0; k < mode_lib->ms.num_active_planes; ++k) {
- if (mode_lib->ms.LinesForDPTERow[k] >= 16 || mode_lib->ms.LinesForVM[k] >= 32) {
- s->AnyLinesForVMOrRowTooLarge = true;
- }
- }
-
- // Only do urg vs prefetch bandwidth check, flip schedule check, power saving feature support check IF the Prefetch Schedule Check is ok
- if (mode_lib->ms.support.PrefetchSupported) {
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- // Calculate Urgent burst factor for prefetch
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, Calling CalculateUrgentBurstFactor (for prefetch)\n", __func__, k);
- dml2_printf("DML::%s: k=%d, VRatioPreY=%f\n", __func__, k, mode_lib->ms.VRatioPreY[k]);
- dml2_printf("DML::%s: k=%d, VRatioPreC=%f\n", __func__, k, mode_lib->ms.VRatioPreC[k]);
-#endif
- double line_time_us = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- CalculateUrgentBurstFactor(
- &display_cfg->plane_descriptors[k],
- mode_lib->ms.swath_width_luma_ub[k],
- mode_lib->ms.swath_width_chroma_ub[k],
- mode_lib->ms.SwathHeightY[k],
- mode_lib->ms.SwathHeightC[k],
- line_time_us,
- mode_lib->ms.UrgLatency,
- mode_lib->ms.VRatioPreY[k],
- mode_lib->ms.VRatioPreC[k],
- mode_lib->ms.BytePerPixelInDETY[k],
- mode_lib->ms.BytePerPixelInDETC[k],
- mode_lib->ms.DETBufferSizeY[k],
- mode_lib->ms.DETBufferSizeC[k],
- /* Output */
- &mode_lib->ms.UrgentBurstFactorLumaPre[k],
- &mode_lib->ms.UrgentBurstFactorChromaPre[k],
- &mode_lib->ms.NotEnoughUrgentLatencyHidingPre[k]);
- }
-
- // Calculate urgent bandwidth required, both urg and non urg peak bandwidth
- // assume flip bw is 0 at this point
- for (k = 0; k < mode_lib->ms.num_active_planes; k++)
- mode_lib->ms.final_flip_bw[k] = 0;
-
- calculate_peak_bandwidth_required(
- &mode_lib->scratch,
- mode_lib->ms.support.urg_vactive_bandwidth_required,
- mode_lib->ms.support.urg_bandwidth_required,
- mode_lib->ms.support.non_urg_bandwidth_required,
-
- display_cfg,
- 0, // inc_flip_bw
- mode_lib->ms.num_active_planes,
- mode_lib->ms.NoOfDPP,
- mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0,
- mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1,
- mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p0,
- mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p1,
- mode_lib->ms.mall_prefetch_sdp_overhead_factor,
- mode_lib->ms.mall_prefetch_dram_overhead_factor,
-
- mode_lib->ms.SurfaceReadBandwidthLuma,
- mode_lib->ms.SurfaceReadBandwidthChroma,
- mode_lib->ms.RequiredPrefetchPixelDataBWLuma,
- mode_lib->ms.RequiredPrefetchPixelDataBWChroma,
- mode_lib->ms.cursor_bw,
- mode_lib->ms.dpte_row_bw,
- mode_lib->ms.meta_row_bw,
- mode_lib->ms.prefetch_cursor_bw,
- mode_lib->ms.prefetch_vmrow_bw,
- mode_lib->ms.final_flip_bw,
- mode_lib->ms.UrgentBurstFactorLuma,
- mode_lib->ms.UrgentBurstFactorChroma,
- mode_lib->ms.UrgentBurstFactorCursor,
- mode_lib->ms.UrgentBurstFactorLumaPre,
- mode_lib->ms.UrgentBurstFactorChromaPre,
- mode_lib->ms.UrgentBurstFactorCursorPre);
-
- // Check urg peak bandwidth against available urg bw
- // check at SDP and DRAM, for all soc states (SVP prefetch an Sys Active)
- check_urgent_bandwidth_support(
- &s->dummy_single[0], // double* frac_urg_bandwidth
- &s->dummy_single[1], // double* frac_urg_bandwidth_mall
- &mode_lib->ms.support.UrgVactiveBandwidthSupport,
- &mode_lib->ms.support.PrefetchBandwidthSupported,
-
- mode_lib->soc.mall_allocated_for_dcn_mbytes,
- mode_lib->ms.support.non_urg_bandwidth_required,
- mode_lib->ms.support.urg_vactive_bandwidth_required,
- mode_lib->ms.support.urg_bandwidth_required,
- mode_lib->ms.support.urg_bandwidth_available);
-
- mode_lib->ms.support.PrefetchSupported &= mode_lib->ms.support.PrefetchBandwidthSupported;
- dml2_printf("DML::%s: PrefetchBandwidthSupported=%0d\n", __func__, mode_lib->ms.support.PrefetchBandwidthSupported);
-
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (mode_lib->ms.NotEnoughUrgentLatencyHidingPre[k]) {
- mode_lib->ms.support.PrefetchSupported = false;
- dml2_printf("DML::%s: k=%d, NotEnoughUrgentLatencyHidingPre=%d\n", __func__, k, mode_lib->ms.NotEnoughUrgentLatencyHidingPre[k]);
- }
- }
-
-
- // Both prefetch schedule and BW okay
- if (mode_lib->ms.support.PrefetchSupported == true && mode_lib->ms.support.VRatioInPrefetchSupported == true) {
- mode_lib->ms.BandwidthAvailableForImmediateFlip =
- get_bandwidth_available_for_immediate_flip(dml2_core_internal_soc_state_sys_active,
- mode_lib->ms.support.urg_bandwidth_required, // no flip
- mode_lib->ms.support.urg_bandwidth_available);
-
- mode_lib->ms.TotImmediateFlipBytes = 0;
- for (k = 0; k < mode_lib->ms.num_active_planes; k++) {
- if (display_cfg->plane_descriptors[k].immediate_flip) {
- s->per_pipe_flip_bytes[k] = get_pipe_flip_bytes(
- s->HostVMInefficiencyFactor,
- mode_lib->ms.vm_bytes[k],
- mode_lib->ms.DPTEBytesPerRow[k],
- mode_lib->ms.meta_row_bytes[k]);
- } else {
- s->per_pipe_flip_bytes[k] = 0;
- }
- mode_lib->ms.TotImmediateFlipBytes += s->per_pipe_flip_bytes[k] * mode_lib->ms.NoOfDPP[k];
-
- }
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- CalculateFlipSchedule(
- &mode_lib->scratch,
- display_cfg->plane_descriptors[k].immediate_flip,
- 1, // use_lb_flip_bw
- s->HostVMInefficiencyFactor,
- s->Tvm_trips_flip[k],
- s->Tr0_trips_flip[k],
- s->Tvm_trips_flip_rounded[k],
- s->Tr0_trips_flip_rounded[k],
- display_cfg->gpuvm_enable,
- mode_lib->ms.vm_bytes[k],
- mode_lib->ms.DPTEBytesPerRow[k],
- mode_lib->ms.BandwidthAvailableForImmediateFlip,
- mode_lib->ms.TotImmediateFlipBytes,
- display_cfg->plane_descriptors[k].pixel_format,
- (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)),
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio,
- mode_lib->ms.Tno_bw_flip[k],
- mode_lib->ms.dpte_row_height[k],
- mode_lib->ms.dpte_row_height_chroma[k],
- mode_lib->ms.use_one_row_for_frame_flip[k],
- mode_lib->ip.max_flip_time_us,
- s->per_pipe_flip_bytes[k],
- mode_lib->ms.meta_row_bytes[k],
- s->meta_row_height_luma[k],
- s->meta_row_height_chroma[k],
- mode_lib->ip.dcn_mrq_present && display_cfg->plane_descriptors[k].surface.dcc.enable,
-
- /* Output */
- &mode_lib->ms.dst_y_per_vm_flip[k],
- &mode_lib->ms.dst_y_per_row_flip[k],
- &mode_lib->ms.final_flip_bw[k],
- &mode_lib->ms.ImmediateFlipSupportedForPipe[k]);
- }
-
- calculate_peak_bandwidth_required(
- &mode_lib->scratch,
- s->dummy_bw,
- mode_lib->ms.support.urg_bandwidth_required_flip,
- mode_lib->ms.support.non_urg_bandwidth_required_flip,
-
- // Input
- display_cfg,
- 1, // inc_flip_bw
- mode_lib->ms.num_active_planes,
- mode_lib->ms.NoOfDPP,
-
- mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p0,
- mode_lib->ms.dcc_dram_bw_nom_overhead_factor_p1,
- mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p0,
- mode_lib->ms.dcc_dram_bw_pref_overhead_factor_p1,
- mode_lib->ms.mall_prefetch_sdp_overhead_factor,
- mode_lib->ms.mall_prefetch_dram_overhead_factor,
-
- mode_lib->ms.SurfaceReadBandwidthLuma,
- mode_lib->ms.SurfaceReadBandwidthChroma,
- mode_lib->ms.RequiredPrefetchPixelDataBWLuma,
- mode_lib->ms.RequiredPrefetchPixelDataBWChroma,
- mode_lib->ms.cursor_bw,
- mode_lib->ms.dpte_row_bw,
- mode_lib->ms.meta_row_bw,
- mode_lib->ms.prefetch_cursor_bw,
- mode_lib->ms.prefetch_vmrow_bw,
- mode_lib->ms.final_flip_bw,
- mode_lib->ms.UrgentBurstFactorLuma,
- mode_lib->ms.UrgentBurstFactorChroma,
- mode_lib->ms.UrgentBurstFactorCursor,
- mode_lib->ms.UrgentBurstFactorLumaPre,
- mode_lib->ms.UrgentBurstFactorChromaPre,
- mode_lib->ms.UrgentBurstFactorCursorPre);
-
- calculate_immediate_flip_bandwidth_support(
- &s->dummy_single[0], // double* frac_urg_bandwidth_flip
- &mode_lib->ms.support.ImmediateFlipSupport,
-
- dml2_core_internal_soc_state_sys_active,
- mode_lib->ms.support.urg_bandwidth_required_flip,
- mode_lib->ms.support.non_urg_bandwidth_required_flip,
- mode_lib->ms.support.urg_bandwidth_available);
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->plane_descriptors[k].immediate_flip == true && mode_lib->ms.ImmediateFlipSupportedForPipe[k] == false)
- mode_lib->ms.support.ImmediateFlipSupport = false;
- }
-
- } else { // if prefetch not support, assume iflip is not supported too
- mode_lib->ms.support.ImmediateFlipSupport = false;
- }
- } // prefetch schedule
- }
-
- s->mSOCParameters.UrgentLatency = mode_lib->ms.UrgLatency;
- s->mSOCParameters.ExtraLatency = mode_lib->ms.ExtraLatency;
- s->mSOCParameters.ExtraLatency_sr = mode_lib->ms.ExtraLatency_sr;
- s->mSOCParameters.WritebackLatency = mode_lib->soc.qos_parameters.writeback.base_latency_us;
- s->mSOCParameters.DRAMClockChangeLatency = mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us;
- s->mSOCParameters.FCLKChangeLatency = mode_lib->soc.power_management_parameters.fclk_change_blackout_us;
- s->mSOCParameters.SRExitTime = mode_lib->soc.power_management_parameters.stutter_exit_latency_us;
- s->mSOCParameters.SREnterPlusExitTime = mode_lib->soc.power_management_parameters.stutter_enter_plus_exit_latency_us;
- s->mSOCParameters.SRExitZ8Time = mode_lib->soc.power_management_parameters.z8_stutter_exit_latency_us;
- s->mSOCParameters.SREnterPlusExitZ8Time = mode_lib->soc.power_management_parameters.z8_stutter_enter_plus_exit_latency_us;
- s->mSOCParameters.USRRetrainingLatency = 0; // FIXME_STAGE2: no USR related bbox value
- s->mSOCParameters.SMNLatency = 0; // FIXME_STAGE2
-
- CalculateWatermarks_params->display_cfg = display_cfg;
- CalculateWatermarks_params->USRRetrainingRequired = false /*FIXME_STAGE2 was: mode_lib->ms.policy.USRRetrainingRequired, no new dml2 replacement*/;
- CalculateWatermarks_params->NumberOfActiveSurfaces = mode_lib->ms.num_active_planes;
- CalculateWatermarks_params->MaxLineBufferLines = mode_lib->ip.max_line_buffer_lines;
- CalculateWatermarks_params->LineBufferSize = mode_lib->ip.line_buffer_size_bits;
- CalculateWatermarks_params->WritebackInterfaceBufferSize = mode_lib->ip.writeback_interface_buffer_size_kbytes;
- CalculateWatermarks_params->DCFCLK = mode_lib->ms.DCFCLK;
- CalculateWatermarks_params->SynchronizeTimings = display_cfg->overrides.synchronize_timings;
- CalculateWatermarks_params->SynchronizeDRRDisplaysForUCLKPStateChange = display_cfg->overrides.synchronize_ddr_displays_for_uclk_pstate_change;
- CalculateWatermarks_params->dpte_group_bytes = mode_lib->ms.dpte_group_bytes;
- CalculateWatermarks_params->mmSOCParameters = s->mSOCParameters;
- CalculateWatermarks_params->WritebackChunkSize = mode_lib->ip.writeback_chunk_size_kbytes;
- CalculateWatermarks_params->SOCCLK = mode_lib->ms.SOCCLK;
- CalculateWatermarks_params->DCFClkDeepSleep = mode_lib->ms.dcfclk_deepsleep;
- CalculateWatermarks_params->DETBufferSizeY = mode_lib->ms.DETBufferSizeY;
- CalculateWatermarks_params->DETBufferSizeC = mode_lib->ms.DETBufferSizeC;
- CalculateWatermarks_params->SwathHeightY = mode_lib->ms.SwathHeightY;
- CalculateWatermarks_params->SwathHeightC = mode_lib->ms.SwathHeightC;
- //CalculateWatermarks_params->LBBitPerPixel = 57; // FIXME_STAGE2, need a new ip param?
- CalculateWatermarks_params->SwathWidthY = mode_lib->ms.SwathWidthY;
- CalculateWatermarks_params->SwathWidthC = mode_lib->ms.SwathWidthC;
- CalculateWatermarks_params->DPPPerSurface = mode_lib->ms.NoOfDPP;
- CalculateWatermarks_params->BytePerPixelDETY = mode_lib->ms.BytePerPixelInDETY;
- CalculateWatermarks_params->BytePerPixelDETC = mode_lib->ms.BytePerPixelInDETC;
- CalculateWatermarks_params->DSTXAfterScaler = s->DSTXAfterScaler;
- CalculateWatermarks_params->DSTYAfterScaler = s->DSTYAfterScaler;
- CalculateWatermarks_params->UnboundedRequestEnabled = mode_lib->ms.UnboundedRequestEnabled;
- CalculateWatermarks_params->CompressedBufferSizeInkByte = mode_lib->ms.CompressedBufferSizeInkByte;
- CalculateWatermarks_params->meta_row_height_l = s->meta_row_height_luma;
- CalculateWatermarks_params->meta_row_height_c = s->meta_row_height_chroma;
-
- // Output
- CalculateWatermarks_params->Watermark = &s->dummy_watermark; // Watermarks *Watermark
- CalculateWatermarks_params->DRAMClockChangeSupport = mode_lib->ms.support.DRAMClockChangeSupport;
- CalculateWatermarks_params->global_dram_clock_change_supported = &mode_lib->ms.support.global_dram_clock_change_supported;
- CalculateWatermarks_params->MaxActiveDRAMClockChangeLatencySupported = &s->dummy_single_array[0]; // double *MaxActiveDRAMClockChangeLatencySupported[]
- CalculateWatermarks_params->SubViewportLinesNeededInMALL = mode_lib->ms.SubViewportLinesNeededInMALL; // unsigned int SubViewportLinesNeededInMALL[]
- CalculateWatermarks_params->FCLKChangeSupport = mode_lib->ms.support.FCLKChangeSupport;
- CalculateWatermarks_params->global_fclk_change_supported = &mode_lib->ms.support.global_fclk_change_supported;
- CalculateWatermarks_params->MaxActiveFCLKChangeLatencySupported = &s->dummy_single[0]; // double *MaxActiveFCLKChangeLatencySupported
- CalculateWatermarks_params->USRRetrainingSupport = &mode_lib->ms.support.USRRetrainingSupport;
- CalculateWatermarks_params->VActiveLatencyHidingMargin = mode_lib->ms.VActiveLatencyHidingMargin;
- CalculateWatermarks_params->VActiveLatencyHidingUs = mode_lib->ms.VActiveLatencyHidingUs;
-
- CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(&mode_lib->scratch, CalculateWatermarks_params);
- }
-
- // End of Prefetch Check
-
- dml2_printf("DML::%s: Done prefetch calculation\n", __func__);
-
- //Re-ordering Buffer Support Check
- mode_lib->ms.support.max_urgent_latency_us
- = mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->ms.qos_param_index].maximum_latency_when_urgent_uclk_cycles / mode_lib->ms.uclk_freq_mhz
- * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_max_latency_margin / 100.0)
- + mode_lib->soc.qos_parameters.qos_params.dcn4x.mall_overhead_fclk_cycles / mode_lib->ms.FabricClock
- + mode_lib->soc.qos_parameters.qos_params.dcn4x.max_round_trip_to_furthest_cs_fclk_cycles / mode_lib->ms.FabricClock
- * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin / 100.0);
-
- if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn4x) {
- if (((mode_lib->ip.rob_buffer_size_kbytes - mode_lib->ip.pixel_chunk_size_kbytes) * 1024
- / mode_lib->ms.support.non_urg_bandwidth_required_flip[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp]) >= mode_lib->ms.support.max_urgent_latency_us) {
- mode_lib->ms.support.ROBSupport = true;
- } else {
- mode_lib->ms.support.ROBSupport = false;
- }
- } else {
- if (mode_lib->ip.rob_buffer_size_kbytes * 1024 >= mode_lib->soc.qos_parameters.qos_params.dcn32x.loaded_round_trip_latency_fclk_cycles * mode_lib->soc.fabric_datapath_to_dcn_data_return_bytes) {
- mode_lib->ms.support.ROBSupport = true;
- } else {
- mode_lib->ms.support.ROBSupport = false;
- }
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: max_urgent_latency_us = %f\n", __func__, mode_lib->ms.support.max_urgent_latency_us);
- dml2_printf("DML::%s: ROBSupport = %u\n", __func__, mode_lib->ms.support.ROBSupport);
-#endif
-
- /*Mode Support, Voltage State and SOC Configuration*/
- {
- // s->dram_clock_change_support = 1;
- // s->f_clock_change_support = 1;
-
- if (mode_lib->ms.support.ScaleRatioAndTapsSupport
- && mode_lib->ms.support.SourceFormatPixelAndScanSupport
- && mode_lib->ms.support.ViewportSizeSupport
- && !mode_lib->ms.support.LinkRateDoesNotMatchDPVersion
- && !mode_lib->ms.support.LinkRateForMultistreamNotIndicated
- && !mode_lib->ms.support.BPPForMultistreamNotIndicated
- && !mode_lib->ms.support.MultistreamWithHDMIOreDP
- && !mode_lib->ms.support.ExceededMultistreamSlots
- && !mode_lib->ms.support.MSOOrODMSplitWithNonDPLink
- && !mode_lib->ms.support.NotEnoughLanesForMSO
- //&& mode_lib->ms.support.LinkCapacitySupport == true // FIXME_STAGE2
- && !mode_lib->ms.support.P2IWith420
- && !mode_lib->ms.support.DSCOnlyIfNecessaryWithBPP
- && !mode_lib->ms.support.DSC422NativeNotSupported
- && !mode_lib->ms.support.NotEnoughDSCUnits
- && !mode_lib->ms.support.NotEnoughDSCSlices
- && !mode_lib->ms.support.ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe
- && !mode_lib->ms.support.InvalidCombinationOfMALLUseForPStateAndStaticScreen
- && !mode_lib->ms.support.DSCCLKRequiredMoreThanSupported
- && mode_lib->ms.support.PixelsPerLinePerDSCUnitSupport
- && !mode_lib->ms.support.DTBCLKRequiredMoreThanSupported
- && !mode_lib->ms.support.InvalidCombinationOfMALLUseForPState
- && mode_lib->ms.support.ROBSupport
- && mode_lib->ms.support.OutstandingRequestsSupport
- && mode_lib->ms.support.OutstandingRequestsUrgencyAvoidance
- && mode_lib->ms.support.DISPCLK_DPPCLK_Support
- && mode_lib->ms.support.TotalAvailablePipesSupport
- && mode_lib->ms.support.NumberOfOTGSupport
- && mode_lib->ms.support.NumberOfHDMIFRLSupport
- && mode_lib->ms.support.NumberOfDP2p0Support
- && mode_lib->ms.support.EnoughWritebackUnits
- && mode_lib->ms.support.WritebackLatencySupport
- && mode_lib->ms.support.WritebackScaleRatioAndTapsSupport
- && mode_lib->ms.support.CursorSupport
- && mode_lib->ms.support.PitchSupport
- && !mode_lib->ms.support.ViewportExceedsSurface
- && mode_lib->ms.support.PrefetchSupported
- && mode_lib->ms.support.EnoughUrgentLatencyHidingSupport
- && mode_lib->ms.support.AvgBandwidthSupport
- && mode_lib->ms.support.DynamicMetadataSupported
- && mode_lib->ms.support.VRatioInPrefetchSupported
- && mode_lib->ms.support.PTEBufferSizeNotExceeded
- && mode_lib->ms.support.DCCMetaBufferSizeNotExceeded
- && !mode_lib->ms.support.ExceededMALLSize
- && ((!display_cfg->hostvm_enable && !s->ImmediateFlipRequired) || mode_lib->ms.support.ImmediateFlipSupport)) {
- // && s->dram_clock_change_support == true
- // && s->f_clock_change_support == true
- // && (/*FIXME_STAGE2 was: mode_lib->ms.policy.USRRetrainingRequired, no new dml2 replacement || */ mode_lib->ms.support.USRRetrainingSupport)) {
- dml2_printf("DML::%s: mode is supported\n", __func__);
- mode_lib->ms.support.ModeSupport = true;
- } else {
- dml2_printf("DML::%s: mode is NOT supported\n", __func__);
- mode_lib->ms.support.ModeSupport = false;
- }
- }
-
- // Since now the mode_support work on 1 particular power state, so there is only 1 state idx (index 0).
- dml2_printf("DML::%s: ModeSupport = %u\n", __func__, mode_lib->ms.support.ModeSupport);
- dml2_printf("DML::%s: ImmediateFlipSupport = %u\n", __func__, mode_lib->ms.support.ImmediateFlipSupport);
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- mode_lib->ms.support.MPCCombineEnable[k] = mode_lib->ms.MPCCombine[k];
- mode_lib->ms.support.DPPPerSurface[k] = mode_lib->ms.NoOfDPP[k];
- }
-
- for (k = 0; k <= mode_lib->ms.num_active_planes - 1; k++) {
- if (display_cfg->plane_descriptors[k].stream_index == k) {
- mode_lib->ms.support.ODMMode[k] = mode_lib->ms.ODMMode[k];
- } else {
- mode_lib->ms.support.ODMMode[k] = dml2_odm_mode_bypass;
- }
-
- mode_lib->ms.support.DSCEnabled[k] = mode_lib->ms.RequiresDSC[k];
- mode_lib->ms.support.FECEnabled[k] = mode_lib->ms.RequiresFEC[k];
- mode_lib->ms.support.OutputBpp[k] = mode_lib->ms.OutputBpp[k];
- mode_lib->ms.support.OutputType[k] = mode_lib->ms.OutputType[k];
- mode_lib->ms.support.OutputRate[k] = mode_lib->ms.OutputRate[k];
-
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: k=%d, ODMMode = %u\n", __func__, k, mode_lib->ms.support.ODMMode[k]);
- dml2_printf("DML::%s: k=%d, DSCEnabled = %u\n", __func__, k, mode_lib->ms.support.DSCEnabled[k]);
-#endif
- }
-
-#if defined(__DML_VBA_DEBUG__)
- if (!mode_lib->ms.support.ModeSupport)
- dml2_print_dml_mode_support_info(&mode_lib->ms.support, true);
- dml2_printf("DML::%s: is_mode_support = %u (min_clk_index=%d)\n", __func__, mode_lib->ms.support.ModeSupport, in_out_params->min_clk_index);
- dml2_printf("DML::%s: --- DONE --- \n", __func__);
-#endif
-
- if (mode_lib->ms.support.ModeSupport) {
- *in_out_params->out_evaluation_info = in_out_params->mode_lib->ms.support;
- return true;
- } else {
- return false;
- }
-}
-
-static void dml2_print_dml_mode_support_info(const struct dml2_core_internal_mode_support_info *support, bool fail_only)
-{
- dml2_printf("DML: ===================================== \n");
- dml2_printf("DML: DML_MODE_SUPPORT_INFO_ST\n");
- if (!fail_only || support->ImmediateFlipSupport == 0)
- dml2_printf("DML: support: ImmediateFlipSupport = 0x%x\n", support->ImmediateFlipSupport);
- if (!fail_only || support->WritebackLatencySupport == 0)
- dml2_printf("DML: support: WritebackLatencySupport = 0x%x\n", support->WritebackLatencySupport);
- if (!fail_only || support->ScaleRatioAndTapsSupport == 0)
- dml2_printf("DML: support: ScaleRatioAndTapsSupport = 0x%x\n", support->ScaleRatioAndTapsSupport);
- if (!fail_only || support->SourceFormatPixelAndScanSupport == 0)
- dml2_printf("DML: support: SourceFormatPixelAndScanSupport = 0x%x\n", support->SourceFormatPixelAndScanSupport);
- if (!fail_only || support->P2IWith420 == 1)
- dml2_printf("DML: support: P2IWith420 = 0x%x\n", support->P2IWith420);
- if (!fail_only || support->DSCOnlyIfNecessaryWithBPP == 1)
- dml2_printf("DML: support: DSCOnlyIfNecessaryWithBPP = 0x%x\n", support->DSCOnlyIfNecessaryWithBPP);
- if (!fail_only || support->DSC422NativeNotSupported == 1)
- dml2_printf("DML: support: DSC422NativeNotSupported = 0x%x\n", support->DSC422NativeNotSupported);
- if (!fail_only || support->LinkRateDoesNotMatchDPVersion == 1)
- dml2_printf("DML: support: LinkRateDoesNotMatchDPVersion = 0x%x\n", support->LinkRateDoesNotMatchDPVersion);
- if (!fail_only || support->LinkRateForMultistreamNotIndicated == 1)
- dml2_printf("DML: support: LinkRateForMultistreamNotIndicated = 0x%x\n", support->LinkRateForMultistreamNotIndicated);
- if (!fail_only || support->BPPForMultistreamNotIndicated == 1)
- dml2_printf("DML: support: BPPForMultistreamNotIndicated = 0x%x\n", support->BPPForMultistreamNotIndicated);
- if (!fail_only || support->MultistreamWithHDMIOreDP == 1)
- dml2_printf("DML: support: MultistreamWithHDMIOreDP = 0x%x\n", support->MultistreamWithHDMIOreDP);
- if (!fail_only || support->MSOOrODMSplitWithNonDPLink == 1)
- dml2_printf("DML: support: MSOOrODMSplitWithNonDPLink = 0x%x\n", support->MSOOrODMSplitWithNonDPLink);
- if (!fail_only || support->NotEnoughLanesForMSO == 1)
- dml2_printf("DML: support: NotEnoughLanesForMSO = 0x%x\n", support->NotEnoughLanesForMSO);
- if (!fail_only || support->NumberOfOTGSupport == 0)
- dml2_printf("DML: support: NumberOfOTGSupport = 0x%x\n", support->NumberOfOTGSupport);
- if (!fail_only || support->NumberOfHDMIFRLSupport == 0)
- dml2_printf("DML: support: NumberOfHDMIFRLSupport = 0x%x\n", support->NumberOfHDMIFRLSupport);
- if (!fail_only || support->NumberOfDP2p0Support == 0)
- dml2_printf("DML: support: NumberOfDP2p0Support = 0x%x\n", support->NumberOfDP2p0Support);
- if (!fail_only || support->WritebackScaleRatioAndTapsSupport == 0)
- dml2_printf("DML: support: WritebackScaleRatioAndTapsSupport = 0x%x\n", support->WritebackScaleRatioAndTapsSupport);
- if (!fail_only || support->CursorSupport == 0)
- dml2_printf("DML: support: CursorSupport = 0x%x\n", support->CursorSupport);
- if (!fail_only || support->PitchSupport == 0)
- dml2_printf("DML: support: PitchSupport = 0x%x\n", support->PitchSupport);
- if (!fail_only || support->ViewportExceedsSurface == 1)
- dml2_printf("DML: support: ViewportExceedsSurface = 0x%x\n", support->ViewportExceedsSurface);
- if (!fail_only || support->ExceededMALLSize == 1)
- dml2_printf("DML: support: ExceededMALLSize = 0x%x\n", support->ExceededMALLSize);
- if (!fail_only || support->EnoughWritebackUnits == 0)
- dml2_printf("DML: support: EnoughWritebackUnits = 0x%x\n", support->EnoughWritebackUnits);
- if (!fail_only || support->ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe == 1)
- dml2_printf("DML: support: ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe = 0x%x\n", support->ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe);
- if (!fail_only || support->InvalidCombinationOfMALLUseForPStateAndStaticScreen == 1)
- dml2_printf("DML: support: InvalidCombinationOfMALLUseForPStateAndStaticScreen = 0x%x\n", support->InvalidCombinationOfMALLUseForPStateAndStaticScreen);
- if (!fail_only || support->InvalidCombinationOfMALLUseForPState == 1)
- dml2_printf("DML: support: InvalidCombinationOfMALLUseForPState = 0x%x\n", support->InvalidCombinationOfMALLUseForPState);
- if (!fail_only || support->ExceededMultistreamSlots == 1)
- dml2_printf("DML: support: ExceededMultistreamSlots = 0x%x\n", support->ExceededMultistreamSlots);
- if (!fail_only || support->NotEnoughDSCUnits == 1)
- dml2_printf("DML: support: NotEnoughDSCUnits = 0x%x\n", support->NotEnoughDSCUnits);
- if (!fail_only || support->NotEnoughDSCSlices == 1)
- dml2_printf("DML: support: NotEnoughDSCSlices = 0x%x\n", support->NotEnoughDSCSlices);
- if (!fail_only || support->PixelsPerLinePerDSCUnitSupport == 0)
- dml2_printf("DML: support: PixelsPerLinePerDSCUnitSupport = 0x%x\n", support->PixelsPerLinePerDSCUnitSupport);
- if (!fail_only || support->DSCCLKRequiredMoreThanSupported == 1)
- dml2_printf("DML: support: DSCCLKRequiredMoreThanSupported = 0x%x\n", support->DSCCLKRequiredMoreThanSupported);
- if (!fail_only || support->DTBCLKRequiredMoreThanSupported == 1)
- dml2_printf("DML: support: DTBCLKRequiredMoreThanSupported = 0x%x\n", support->DTBCLKRequiredMoreThanSupported);
- if (!fail_only || support->LinkCapacitySupport == 0)
- dml2_printf("DML: support: LinkCapacitySupport = 0x%x\n", support->LinkCapacitySupport);
- if (!fail_only || support->ROBSupport == 0)
- dml2_printf("DML: support: ROBSupport = %d\n", support->ROBSupport);
- if (!fail_only || support->OutstandingRequestsSupport == 0)
- dml2_printf("DML: support: OutstandingRequestsSupport = %d\n", support->OutstandingRequestsSupport);
- if (!fail_only || support->OutstandingRequestsUrgencyAvoidance == 0)
- dml2_printf("DML: support: OutstandingRequestsUrgencyAvoidance = %d\n", support->OutstandingRequestsUrgencyAvoidance);
- if (!fail_only || support->PTEBufferSizeNotExceeded == 0)
- dml2_printf("DML: support: PTEBufferSizeNotExceeded = %d\n", support->PTEBufferSizeNotExceeded);
- if (!fail_only || support->AvgBandwidthSupport == 0)
- dml2_printf("DML: support: AvgBandwidthSupport = %d\n", support->AvgBandwidthSupport);
- if (!fail_only || support->EnoughUrgentLatencyHidingSupport == 0)
- dml2_printf("DML: support: EnoughUrgentLatencyHidingSupport = %d\n", support->EnoughUrgentLatencyHidingSupport);
- if (!fail_only || support->PrefetchSupported == 0)
- dml2_printf("DML: support: PrefetchSupported = %d\n", support->PrefetchSupported);
- if (!fail_only || support->DynamicMetadataSupported == 0)
- dml2_printf("DML: support: DynamicMetadataSupported = %d\n", support->DynamicMetadataSupported);
- if (!fail_only || support->VRatioInPrefetchSupported == 0)
- dml2_printf("DML: support: VRatioInPrefetchSupported = %d\n", support->VRatioInPrefetchSupported);
- if (!fail_only || support->DISPCLK_DPPCLK_Support == 0)
- dml2_printf("DML: support: DISPCLK_DPPCLK_Support = %d\n", support->DISPCLK_DPPCLK_Support);
- if (!fail_only || support->TotalAvailablePipesSupport == 0)
- dml2_printf("DML: support: TotalAvailablePipesSupport = %d\n", support->TotalAvailablePipesSupport);
- if (!fail_only || support->ModeSupport == 0)
- dml2_printf("DML: support: ModeSupport = %d\n", support->ModeSupport);
- if (!fail_only || support->ViewportSizeSupport == 0)
- dml2_printf("DML: support: ViewportSizeSupport = %d\n", support->ViewportSizeSupport);
- dml2_printf("DML: ===================================== \n");
-}
-
-static void get_stream_output_bpp(double *out_bpp, const struct dml2_display_cfg *display_cfg)
-{
- for (unsigned int k = 0; k < display_cfg->num_planes; k++) {
- double bpc = (double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.bpc;
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable == dml2_dsc_disable) {
- switch (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format) {
- case dml2_444:
- out_bpp[k] = bpc * 3;
- break;
- case dml2_s422:
- out_bpp[k] = bpc * 2;
- break;
- case dml2_n422:
- out_bpp[k] = bpc * 2;
- break;
- case dml2_420:
- default:
- out_bpp[k] = bpc * 1.5;
- break;
- }
- } else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable == dml2_dsc_enable) {
- out_bpp[k] = (double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.dsc_compressed_bpp_x16 / 16;
- } else {
- out_bpp[k] = 0;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d bpc=%f\n", __func__, k, bpc);
- dml2_printf("DML::%s: k=%d dsc.enable=%d\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable);
- dml2_printf("DML::%s: k=%d out_bpp=%f\n", __func__, k, out_bpp[k]);
-#endif
- }
-}
-
-static unsigned int dml_round_to_multiple(unsigned int num, unsigned int multiple, bool up)
-{
- unsigned int remainder;
-
- if (multiple == 0)
- return num;
-
- remainder = num % multiple;
- if (remainder == 0)
- return num;
-
- if (up)
- return (num + multiple - remainder);
- else
- return (num - remainder);
-}
-
-static unsigned int dml_get_num_active_pipes(int unsigned num_planes, const struct core_display_cfg_support_info *cfg_support_info)
-{
- unsigned int num_active_pipes = 0;
-
- for (unsigned int k = 0; k < num_planes; k++) {
- num_active_pipes = num_active_pipes + (unsigned int)cfg_support_info->plane_support_info[k].dpps_used;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: num_active_pipes = %d\n", __func__, num_active_pipes);
-#endif
- return num_active_pipes;
-}
-
-static void dml_calc_pipe_plane_mapping(const struct core_display_cfg_support_info *cfg_support_info, unsigned int *pipe_plane)
-{
- unsigned int pipe_idx = 0;
-
- for (unsigned int k = 0; k < DML2_MAX_PLANES; ++k) {
- pipe_plane[k] = __DML2_CALCS_PIPE_NO_PLANE__;
- }
-
- for (unsigned int plane_idx = 0; plane_idx < DML2_MAX_PLANES; plane_idx++) {
- for (int i = 0; i < cfg_support_info->plane_support_info[plane_idx].dpps_used; i++) {
- pipe_plane[pipe_idx] = plane_idx;
- pipe_idx++;
- }
- }
-}
-
-static bool dml_is_phantom_pipe(const struct dml2_plane_parameters *plane_cfg)
-{
- bool is_phantom = false;
-
- if (plane_cfg->overrides.legacy_svp_config == dml2_svp_mode_override_phantom_pipe ||
- plane_cfg->overrides.legacy_svp_config == dml2_svp_mode_override_phantom_pipe_no_data_return) {
- is_phantom = true;
- }
-
- return is_phantom;
-}
-
-static bool dml_get_is_phantom_pipe(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, unsigned int pipe_idx)
-{
- unsigned int plane_idx = mode_lib->mp.pipe_plane[pipe_idx];
-
- bool is_phantom = dml_is_phantom_pipe(&display_cfg->plane_descriptors[plane_idx]);
- dml2_printf("DML::%s: pipe_idx=%d legacy_svp_config=%0d is_phantom=%d\n", __func__, pipe_idx, display_cfg->plane_descriptors[plane_idx].overrides.legacy_svp_config, is_phantom);
- return is_phantom;
-}
-
-static void CalculateMaxDETAndMinCompressedBufferSize(
- unsigned int ConfigReturnBufferSizeInKByte,
- unsigned int ConfigReturnBufferSegmentSizeInKByte,
- unsigned int ROBBufferSizeInKByte,
- unsigned int MaxNumDPP,
- unsigned int nomDETInKByteOverrideEnable, // VBA_DELTA, allow DV to override default DET size
- unsigned int nomDETInKByteOverrideValue, // VBA_DELTA
- bool is_mrq_present,
-
- // Output
- unsigned int *MaxTotalDETInKByte,
- unsigned int *nomDETInKByte,
- unsigned int *MinCompressedBufferSizeInKByte)
-{
- if (is_mrq_present)
- *MaxTotalDETInKByte = (unsigned int)math_ceil2((double)(ConfigReturnBufferSizeInKByte + ROBBufferSizeInKByte) * 4 / 5, 64);
- else
- *MaxTotalDETInKByte = ConfigReturnBufferSizeInKByte - ConfigReturnBufferSegmentSizeInKByte;
-
- *nomDETInKByte = (unsigned int)(math_floor2((double)*MaxTotalDETInKByte / (double)MaxNumDPP, ConfigReturnBufferSegmentSizeInKByte));
- *MinCompressedBufferSizeInKByte = ConfigReturnBufferSizeInKByte - *MaxTotalDETInKByte;
-
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: is_mrq_present = %u\n", __func__, is_mrq_present);
- dml2_printf("DML::%s: ConfigReturnBufferSizeInKByte = %u\n", __func__, ConfigReturnBufferSizeInKByte);
- dml2_printf("DML::%s: ROBBufferSizeInKByte = %u\n", __func__, ROBBufferSizeInKByte);
- dml2_printf("DML::%s: MaxNumDPP = %u\n", __func__, MaxNumDPP);
- dml2_printf("DML::%s: MaxTotalDETInKByte = %u\n", __func__, *MaxTotalDETInKByte);
- dml2_printf("DML::%s: nomDETInKByte = %u\n", __func__, *nomDETInKByte);
- dml2_printf("DML::%s: MinCompressedBufferSizeInKByte = %u\n", __func__, *MinCompressedBufferSizeInKByte);
-#endif
-
- if (nomDETInKByteOverrideEnable) {
- *nomDETInKByte = nomDETInKByteOverrideValue;
- dml2_printf("DML::%s: nomDETInKByte = %u (overrided)\n", __func__, *nomDETInKByte);
- }
-}
-
-static void PixelClockAdjustmentForProgressiveToInterlaceUnit(const struct dml2_display_cfg *display_cfg, bool ptoi_supported, double *PixelClockBackEnd)
-{
- //unsigned int num_active_planes = display_cfg->num_planes;
-
- //Progressive To Interlace Unit Effect
- for (unsigned int k = 0; k < display_cfg->num_planes; ++k) {
- PixelClockBackEnd[k] = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced == 1 && ptoi_supported == true) {
- // FIXME_STAGE2... can sw pass the pixel rate for interlaced directly
- //display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz = 2 * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz;
- }
- }
-}
-
-bool dml2_core_shared_is_420(enum dml2_source_format_class source_format)
-{
- bool val = false;
-
- switch (source_format) {
- case dml2_444_8:
- val = 0;
- break;
- case dml2_444_16:
- val = 0;
- break;
- case dml2_444_32:
- val = 0;
- break;
- case dml2_444_64:
- val = 0;
- break;
- case dml2_420_8:
- val = 1;
- break;
- case dml2_420_10:
- val = 1;
- break;
- case dml2_420_12:
- val = 1;
- break;
- case dml2_rgbe_alpha:
- val = 0;
- break;
- case dml2_rgbe:
- val = 0;
- break;
- case dml2_mono_8:
- val = 0;
- break;
- case dml2_mono_16:
- val = 0;
- break;
- default:
- DML2_ASSERT(0);
- break;
- }
- return val;
-}
-
-static unsigned int dml_get_tile_block_size_bytes(enum dml2_swizzle_mode sw_mode)
-{
- switch (sw_mode) {
- case (dml2_sw_linear):
- return 256;
- case (dml2_sw_256b_2d):
- return 256;
- case (dml2_sw_4kb_2d):
- return 4096;
- case (dml2_sw_64kb_2d):
- return 65536;
- case (dml2_sw_256kb_2d):
- return 262144;
- case (dml2_gfx11_sw_linear):
- return 256;
- case (dml2_gfx11_sw_64kb_d):
- return 65536;
- case (dml2_gfx11_sw_64kb_d_t):
- return 65536;
- case (dml2_gfx11_sw_64kb_d_x):
- return 65536;
- case (dml2_gfx11_sw_64kb_r_x):
- return 65536;
- case (dml2_gfx11_sw_256kb_d_x):
- return 262144;
- case (dml2_gfx11_sw_256kb_r_x):
- return 262144;
- default:
- DML2_ASSERT(0);
- return 256;
- }
-}
-
-const char *dml2_core_internal_bw_type_str(enum dml2_core_internal_bw_type bw_type)
-{
- switch (bw_type) {
- case (dml2_core_internal_bw_sdp):
- return("dml2_core_internal_bw_sdp");
- case (dml2_core_internal_bw_dram):
- return("dml2_core_internal_bw_dram");
- case (dml2_core_internal_bw_max):
- return("dml2_core_internal_bw_max");
- default:
- return("dml2_core_internal_bw_unknown");
- }
-}
-
-const char *dml2_core_internal_soc_state_type_str(enum dml2_core_internal_soc_state_type dml2_core_internal_soc_state_type)
-{
- switch (dml2_core_internal_soc_state_type) {
- case (dml2_core_internal_soc_state_sys_idle):
- return("dml2_core_internal_soc_state_sys_idle");
- case (dml2_core_internal_soc_state_sys_active):
- return("dml2_core_internal_soc_state_sys_active");
- case (dml2_core_internal_soc_state_svp_prefetch):
- return("dml2_core_internal_soc_state_svp_prefetch");
- case dml2_core_internal_soc_state_max:
- default:
- return("dml2_core_internal_soc_state_unknown");
- }
-}
-
-static bool dml_is_vertical_rotation(enum dml2_rotation_angle Scan)
-{
- bool is_vert = false;
- if (Scan == dml2_rotation_90 || Scan == dml2_rotation_270) {
- is_vert = true;
- } else {
- is_vert = false;
- }
- return is_vert;
-}
-
-static int unsigned dml_get_gfx_version(enum dml2_swizzle_mode sw_mode)
-{
- int unsigned version = 0;
-
- if (sw_mode == dml2_sw_linear ||
- sw_mode == dml2_sw_256b_2d ||
- sw_mode == dml2_sw_4kb_2d ||
- sw_mode == dml2_sw_64kb_2d ||
- sw_mode == dml2_sw_256kb_2d) {
- version = 12;
- } else if (sw_mode == dml2_gfx11_sw_linear ||
- sw_mode == dml2_gfx11_sw_64kb_d ||
- sw_mode == dml2_gfx11_sw_64kb_d_t ||
- sw_mode == dml2_gfx11_sw_64kb_d_x ||
- sw_mode == dml2_gfx11_sw_64kb_r_x ||
- sw_mode == dml2_gfx11_sw_256kb_d_x ||
- sw_mode == dml2_gfx11_sw_256kb_r_x) {
- version = 11;
- } else {
- dml2_printf("ERROR: Invalid sw_mode setting! val=%u\n", sw_mode);
- DML2_ASSERT(0);
- }
-
- return version;
-}
-
-static void CalculateBytePerPixelAndBlockSizes(
- enum dml2_source_format_class SourcePixelFormat,
- enum dml2_swizzle_mode SurfaceTiling,
- unsigned int pitch_y,
- unsigned int pitch_c,
-
- // Output
- unsigned int *BytePerPixelY,
- unsigned int *BytePerPixelC,
- double *BytePerPixelDETY,
- double *BytePerPixelDETC,
- unsigned int *BlockHeight256BytesY,
- unsigned int *BlockHeight256BytesC,
- unsigned int *BlockWidth256BytesY,
- unsigned int *BlockWidth256BytesC,
- unsigned int *MacroTileHeightY,
- unsigned int *MacroTileHeightC,
- unsigned int *MacroTileWidthY,
- unsigned int *MacroTileWidthC,
- bool *surf_linear128_l,
- bool *surf_linear128_c)
-{
- *BytePerPixelDETY = 0;
- *BytePerPixelDETC = 0;
- *BytePerPixelY = 0;
- *BytePerPixelC = 0;
-
- if (SourcePixelFormat == dml2_444_64) {
- *BytePerPixelDETY = 8;
- *BytePerPixelDETC = 0;
- *BytePerPixelY = 8;
- *BytePerPixelC = 0;
- } else if (SourcePixelFormat == dml2_444_32 || SourcePixelFormat == dml2_rgbe) {
- *BytePerPixelDETY = 4;
- *BytePerPixelDETC = 0;
- *BytePerPixelY = 4;
- *BytePerPixelC = 0;
- } else if (SourcePixelFormat == dml2_444_16 || SourcePixelFormat == dml2_mono_16) {
- *BytePerPixelDETY = 2;
- *BytePerPixelDETC = 0;
- *BytePerPixelY = 2;
- *BytePerPixelC = 0;
- } else if (SourcePixelFormat == dml2_444_8 || SourcePixelFormat == dml2_mono_8) {
- *BytePerPixelDETY = 1;
- *BytePerPixelDETC = 0;
- *BytePerPixelY = 1;
- *BytePerPixelC = 0;
- } else if (SourcePixelFormat == dml2_rgbe_alpha) {
- *BytePerPixelDETY = 4;
- *BytePerPixelDETC = 1;
- *BytePerPixelY = 4;
- *BytePerPixelC = 1;
- } else if (SourcePixelFormat == dml2_420_8) {
- *BytePerPixelDETY = 1;
- *BytePerPixelDETC = 2;
- *BytePerPixelY = 1;
- *BytePerPixelC = 2;
- } else if (SourcePixelFormat == dml2_420_12) {
- *BytePerPixelDETY = 2;
- *BytePerPixelDETC = 4;
- *BytePerPixelY = 2;
- *BytePerPixelC = 4;
- } else if (SourcePixelFormat == dml2_420_10) {
- *BytePerPixelDETY = (double)(4.0 / 3);
- *BytePerPixelDETC = (double)(8.0 / 3);
- *BytePerPixelY = 2;
- *BytePerPixelC = 4;
- } else {
- dml2_printf("ERROR: DML::%s: SourcePixelFormat = %u not supported!\n", __func__, SourcePixelFormat);
- DML2_ASSERT(0);
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: SourcePixelFormat = %u\n", __func__, SourcePixelFormat);
- dml2_printf("DML::%s: BytePerPixelDETY = %f\n", __func__, *BytePerPixelDETY);
- dml2_printf("DML::%s: BytePerPixelDETC = %f\n", __func__, *BytePerPixelDETC);
- dml2_printf("DML::%s: BytePerPixelY = %u\n", __func__, *BytePerPixelY);
- dml2_printf("DML::%s: BytePerPixelC = %u\n", __func__, *BytePerPixelC);
- dml2_printf("DML::%s: pitch_y = %u\n", __func__, pitch_y);
- dml2_printf("DML::%s: pitch_c = %u\n", __func__, pitch_c);
- dml2_printf("DML::%s: surf_linear128_l = %u\n", __func__, *surf_linear128_l);
- dml2_printf("DML::%s: surf_linear128_c = %u\n", __func__, *surf_linear128_c);
-#endif
-
- if (dml_get_gfx_version(SurfaceTiling) == 11) {
- *surf_linear128_l = 0;
- *surf_linear128_c = 0;
- } else {
- if (SurfaceTiling == dml2_sw_linear) {
- *surf_linear128_l = (((pitch_y * *BytePerPixelY) % 256) != 0);
-
- if (dml2_core_shared_is_420(SourcePixelFormat) || SourcePixelFormat == dml2_rgbe_alpha)
- *surf_linear128_c = (((pitch_c * *BytePerPixelC) % 256) != 0);
- }
- }
-
- if (!(dml2_core_shared_is_420(SourcePixelFormat) || SourcePixelFormat == dml2_rgbe_alpha)) {
- if (SurfaceTiling == dml2_sw_linear) {
- *BlockHeight256BytesY = 1;
- } else if (SourcePixelFormat == dml2_444_64) {
- *BlockHeight256BytesY = 4;
- } else if (SourcePixelFormat == dml2_444_8) {
- *BlockHeight256BytesY = 16;
- } else {
- *BlockHeight256BytesY = 8;
- }
- *BlockWidth256BytesY = 256U / *BytePerPixelY / *BlockHeight256BytesY;
- *BlockHeight256BytesC = 0;
- *BlockWidth256BytesC = 0;
- } else { // dual plane
- if (SurfaceTiling == dml2_sw_linear) {
- *BlockHeight256BytesY = 1;
- *BlockHeight256BytesC = 1;
- } else if (SourcePixelFormat == dml2_rgbe_alpha) {
- *BlockHeight256BytesY = 8;
- *BlockHeight256BytesC = 16;
- } else if (SourcePixelFormat == dml2_420_8) {
- *BlockHeight256BytesY = 16;
- *BlockHeight256BytesC = 8;
- } else {
- *BlockHeight256BytesY = 8;
- *BlockHeight256BytesC = 8;
- }
- *BlockWidth256BytesY = 256U / *BytePerPixelY / *BlockHeight256BytesY;
- *BlockWidth256BytesC = 256U / *BytePerPixelC / *BlockHeight256BytesC;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: BlockWidth256BytesY = %u\n", __func__, *BlockWidth256BytesY);
- dml2_printf("DML::%s: BlockHeight256BytesY = %u\n", __func__, *BlockHeight256BytesY);
- dml2_printf("DML::%s: BlockWidth256BytesC = %u\n", __func__, *BlockWidth256BytesC);
- dml2_printf("DML::%s: BlockHeight256BytesC = %u\n", __func__, *BlockHeight256BytesC);
-#endif
-
- if (dml_get_gfx_version(SurfaceTiling) == 11) {
- if (SurfaceTiling == dml2_gfx11_sw_linear) {
- *MacroTileHeightY = *BlockHeight256BytesY;
- *MacroTileWidthY = 256 / *BytePerPixelY / *MacroTileHeightY;
- *MacroTileHeightC = *BlockHeight256BytesC;
- if (*MacroTileHeightC == 0) {
- *MacroTileWidthC = 0;
- } else {
- *MacroTileWidthC = 256 / *BytePerPixelC / *MacroTileHeightC;
- }
- } else if (SurfaceTiling == dml2_gfx11_sw_64kb_d || SurfaceTiling == dml2_gfx11_sw_64kb_d_t || SurfaceTiling == dml2_gfx11_sw_64kb_d_x || SurfaceTiling == dml2_gfx11_sw_64kb_r_x) {
- *MacroTileHeightY = 16 * *BlockHeight256BytesY;
- *MacroTileWidthY = 65536 / *BytePerPixelY / *MacroTileHeightY;
- *MacroTileHeightC = 16 * *BlockHeight256BytesC;
- if (*MacroTileHeightC == 0) {
- *MacroTileWidthC = 0;
- } else {
- *MacroTileWidthC = 65536 / *BytePerPixelC / *MacroTileHeightC;
- }
- } else {
- *MacroTileHeightY = 32 * *BlockHeight256BytesY;
- *MacroTileWidthY = 65536 * 4 / *BytePerPixelY / *MacroTileHeightY;
- *MacroTileHeightC = 32 * *BlockHeight256BytesC;
- if (*MacroTileHeightC == 0) {
- *MacroTileWidthC = 0;
- } else {
- *MacroTileWidthC = 65536 * 4 / *BytePerPixelC / *MacroTileHeightC;
- }
- }
- } else {
- unsigned int macro_tile_size_bytes = dml_get_tile_block_size_bytes(SurfaceTiling);
- unsigned int macro_tile_scale = 1; // macro tile to 256B req scaling
-
- if (SurfaceTiling == dml2_sw_linear) {
- macro_tile_scale = 1;
- } else if (SurfaceTiling == dml2_sw_4kb_2d) {
- macro_tile_scale = 4;
- } else if (SurfaceTiling == dml2_sw_64kb_2d) {
- macro_tile_scale = 16;
- } else if (SurfaceTiling == dml2_sw_256kb_2d) {
- macro_tile_scale = 32;
- } else {
- dml2_printf("ERROR: Invalid SurfaceTiling setting! val=%u\n", SurfaceTiling);
- DML2_ASSERT(0);
- }
-
- *MacroTileHeightY = macro_tile_scale * *BlockHeight256BytesY;
- *MacroTileWidthY = macro_tile_size_bytes / *BytePerPixelY / *MacroTileHeightY;
- *MacroTileHeightC = macro_tile_scale * *BlockHeight256BytesC;
- if (*MacroTileHeightC == 0) {
- *MacroTileWidthC = 0;
- } else {
- *MacroTileWidthC = macro_tile_size_bytes / *BytePerPixelC / *MacroTileHeightC;
- }
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: MacroTileWidthY = %u\n", __func__, *MacroTileWidthY);
- dml2_printf("DML::%s: MacroTileHeightY = %u\n", __func__, *MacroTileHeightY);
- dml2_printf("DML::%s: MacroTileWidthC = %u\n", __func__, *MacroTileWidthC);
- dml2_printf("DML::%s: MacroTileHeightC = %u\n", __func__, *MacroTileHeightC);
-#endif
-}
-
-static void CalculateSinglePipeDPPCLKAndSCLThroughput(
- double HRatio,
- double HRatioChroma,
- double VRatio,
- double VRatioChroma,
- double MaxDCHUBToPSCLThroughput,
- double MaxPSCLToLBThroughput,
- double PixelClock,
- enum dml2_source_format_class SourcePixelFormat,
- unsigned int HTaps,
- unsigned int HTapsChroma,
- unsigned int VTaps,
- unsigned int VTapsChroma,
-
- // Output
- double *PSCL_THROUGHPUT,
- double *PSCL_THROUGHPUT_CHROMA,
- double *DPPCLKUsingSingleDPP)
-{
- if (HRatio > 1) {
- *PSCL_THROUGHPUT = math_min2(MaxDCHUBToPSCLThroughput, MaxPSCLToLBThroughput * HRatio / math_ceil2((double)HTaps / 6.0, 1.0));
- } else {
- *PSCL_THROUGHPUT = math_min2(MaxDCHUBToPSCLThroughput, MaxPSCLToLBThroughput);
- }
-
- double DPPCLKUsingSingleDPPLuma;
- double DPPCLKUsingSingleDPPChroma;
-
- DPPCLKUsingSingleDPPLuma = PixelClock * math_max3(VTaps / 6 * math_min2(1, HRatio), HRatio * VRatio / *PSCL_THROUGHPUT, 1);
-
- if ((HTaps > 6 || VTaps > 6) && DPPCLKUsingSingleDPPLuma < 2 * PixelClock)
- DPPCLKUsingSingleDPPLuma = 2 * PixelClock;
-
- if (!dml2_core_shared_is_420(SourcePixelFormat) && SourcePixelFormat != dml2_rgbe_alpha) {
- *PSCL_THROUGHPUT_CHROMA = 0;
- *DPPCLKUsingSingleDPP = DPPCLKUsingSingleDPPLuma;
- } else {
- if (HRatioChroma > 1) {
- *PSCL_THROUGHPUT_CHROMA = math_min2(MaxDCHUBToPSCLThroughput, MaxPSCLToLBThroughput * HRatioChroma / math_ceil2((double)HTapsChroma / 6.0, 1.0));
- } else {
- *PSCL_THROUGHPUT_CHROMA = math_min2(MaxDCHUBToPSCLThroughput, MaxPSCLToLBThroughput);
- }
- DPPCLKUsingSingleDPPChroma = PixelClock * math_max3(VTapsChroma / 6 * math_min2(1, HRatioChroma),
- HRatioChroma * VRatioChroma / *PSCL_THROUGHPUT_CHROMA, 1);
- if ((HTapsChroma > 6 || VTapsChroma > 6) && DPPCLKUsingSingleDPPChroma < 2 * PixelClock)
- DPPCLKUsingSingleDPPChroma = 2 * PixelClock;
- *DPPCLKUsingSingleDPP = math_max2(DPPCLKUsingSingleDPPLuma, DPPCLKUsingSingleDPPChroma);
- }
-}
-
-static void CalculateSwathWidth(
- const struct dml2_display_cfg *display_cfg,
- bool ForceSingleDPP,
- unsigned int NumberOfActiveSurfaces,
- enum dml2_odm_mode ODMMode[],
- unsigned int BytePerPixY[],
- unsigned int BytePerPixC[],
- unsigned int Read256BytesBlockHeightY[],
- unsigned int Read256BytesBlockHeightC[],
- unsigned int Read256BytesBlockWidthY[],
- unsigned int Read256BytesBlockWidthC[],
- bool surf_linear128_l[],
- bool surf_linear128_c[],
- unsigned int DPPPerSurface[],
-
- // Output
- unsigned int req_per_swath_ub_l[],
- unsigned int req_per_swath_ub_c[],
- unsigned int SwathWidthSingleDPPY[],
- unsigned int SwathWidthSingleDPPC[],
- unsigned int SwathWidthY[], // per-pipe
- unsigned int SwathWidthC[], // per-pipe
- unsigned int MaximumSwathHeightY[],
- unsigned int MaximumSwathHeightC[],
- unsigned int swath_width_luma_ub[], // per-pipe
- unsigned int swath_width_chroma_ub[]) // per-pipe
-{
- enum dml2_odm_mode MainSurfaceODMMode;
- double odm_hactive_factor = 1.0;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ForceSingleDPP = %u\n", __func__, ForceSingleDPP);
- dml2_printf("DML::%s: NumberOfActiveSurfaces = %u\n", __func__, NumberOfActiveSurfaces);
-#endif
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- if (!dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle)) {
- SwathWidthSingleDPPY[k] = (unsigned int)display_cfg->plane_descriptors[k].composition.viewport.plane0.width;
- } else {
- SwathWidthSingleDPPY[k] = (unsigned int)display_cfg->plane_descriptors[k].composition.viewport.plane0.height;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u ViewportWidth=%u\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.width);
- dml2_printf("DML::%s: k=%u ViewportHeight=%u\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.plane0.height);
- dml2_printf("DML::%s: k=%u DPPPerSurface=%u\n", __func__, k, DPPPerSurface[k]);
-#endif
-
- MainSurfaceODMMode = ODMMode[k];
- for (unsigned int j = 0; j < NumberOfActiveSurfaces; ++j) {
- if (display_cfg->plane_descriptors[k].stream_index == j) {
- MainSurfaceODMMode = ODMMode[j];
- }
- }
-
- if (ForceSingleDPP) {
- SwathWidthY[k] = SwathWidthSingleDPPY[k];
- } else {
- if (MainSurfaceODMMode == dml2_odm_mode_combine_4to1)
- odm_hactive_factor = 4.0;
- else if (MainSurfaceODMMode == dml2_odm_mode_combine_3to1)
- odm_hactive_factor = 3.0;
- else if (MainSurfaceODMMode == dml2_odm_mode_combine_2to1)
- odm_hactive_factor = 2.0;
-
- if (MainSurfaceODMMode == dml2_odm_mode_combine_4to1 || MainSurfaceODMMode == dml2_odm_mode_combine_3to1 || MainSurfaceODMMode == dml2_odm_mode_combine_2to1) {
- SwathWidthY[k] = (unsigned int)(math_min2((double)SwathWidthSingleDPPY[k], math_round((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active / odm_hactive_factor * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio)));
- } else if (DPPPerSurface[k] == 2) {
- SwathWidthY[k] = SwathWidthSingleDPPY[k] / 2;
- } else {
- SwathWidthY[k] = SwathWidthSingleDPPY[k];
- }
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u HActive=%u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active);
- dml2_printf("DML::%s: k=%u HRatio=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
- dml2_printf("DML::%s: k=%u MainSurfaceODMMode=%u\n", __func__, k, MainSurfaceODMMode);
- dml2_printf("DML::%s: k=%u SwathWidthSingleDPPY=%u\n", __func__, k, SwathWidthSingleDPPY[k]);
- dml2_printf("DML::%s: k=%u SwathWidthY=%u\n", __func__, k, SwathWidthY[k]);
-#endif
-
- if (dml2_core_shared_is_420(display_cfg->plane_descriptors[k].pixel_format)) {
- SwathWidthC[k] = SwathWidthY[k] / 2;
- SwathWidthSingleDPPC[k] = SwathWidthSingleDPPY[k] / 2;
- } else {
- SwathWidthC[k] = SwathWidthY[k];
- SwathWidthSingleDPPC[k] = SwathWidthSingleDPPY[k];
- }
-
- if (ForceSingleDPP == true) {
- SwathWidthY[k] = SwathWidthSingleDPPY[k];
- SwathWidthC[k] = SwathWidthSingleDPPC[k];
- }
-
- unsigned int req_width_horz_y = Read256BytesBlockWidthY[k];
- unsigned int req_width_horz_c = Read256BytesBlockWidthC[k];
-
- if (surf_linear128_l[k])
- req_width_horz_y = req_width_horz_y / 2;
-
- if (surf_linear128_c[k])
- req_width_horz_c = req_width_horz_c / 2;
-
- unsigned int surface_width_ub_l = (unsigned int)math_ceil2((double)display_cfg->plane_descriptors[k].surface.plane0.width, req_width_horz_y);
- unsigned int surface_height_ub_l = (unsigned int)math_ceil2((double)display_cfg->plane_descriptors[k].surface.plane0.height, Read256BytesBlockHeightY[k]);
- unsigned int surface_width_ub_c = (unsigned int)math_ceil2((double)display_cfg->plane_descriptors[k].surface.plane1.width, req_width_horz_c);
- unsigned int surface_height_ub_c = (unsigned int)math_ceil2((double)display_cfg->plane_descriptors[k].surface.plane1.height, Read256BytesBlockHeightC[k]);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u surface_width_ub_l=%u\n", __func__, k, surface_width_ub_l);
- dml2_printf("DML::%s: k=%u surface_height_ub_l=%u\n", __func__, k, surface_height_ub_l);
- dml2_printf("DML::%s: k=%u surface_width_ub_c=%u\n", __func__, k, surface_width_ub_c);
- dml2_printf("DML::%s: k=%u surface_height_ub_c=%u\n", __func__, k, surface_height_ub_c);
- dml2_printf("DML::%s: k=%u req_width_horz_y=%u\n", __func__, k, req_width_horz_y);
- dml2_printf("DML::%s: k=%u req_width_horz_c=%u\n", __func__, k, req_width_horz_c);
- dml2_printf("DML::%s: k=%u Read256BytesBlockWidthY=%u\n", __func__, k, Read256BytesBlockWidthY[k]);
- dml2_printf("DML::%s: k=%u Read256BytesBlockHeightY=%u\n", __func__, k, Read256BytesBlockHeightY[k]);
- dml2_printf("DML::%s: k=%u Read256BytesBlockWidthC=%u\n", __func__, k, Read256BytesBlockWidthC[k]);
- dml2_printf("DML::%s: k=%u Read256BytesBlockHeightC=%u\n", __func__, k, Read256BytesBlockHeightC[k]);
- dml2_printf("DML::%s: k=%u req_width_horz_y=%u\n", __func__, k, req_width_horz_y);
- dml2_printf("DML::%s: k=%u req_width_horz_c=%u\n", __func__, k, req_width_horz_c);
- dml2_printf("DML::%s: k=%u ViewportStationary=%u\n", __func__, k, display_cfg->plane_descriptors[k].composition.viewport.stationary);
- dml2_printf("DML::%s: k=%u DPPPerSurface=%u\n", __func__, k, DPPPerSurface[k]);
-#endif
-
- req_per_swath_ub_l[k] = 0;
- req_per_swath_ub_c[k] = 0;
- if (!dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle)) {
- MaximumSwathHeightY[k] = Read256BytesBlockHeightY[k];
- MaximumSwathHeightC[k] = Read256BytesBlockHeightC[k];
- if (display_cfg->plane_descriptors[k].composition.viewport.stationary && DPPPerSurface[k] == 1) {
- swath_width_luma_ub[k] = (unsigned int)(math_min2(surface_width_ub_l, math_floor2(display_cfg->plane_descriptors[k].composition.viewport.plane0.x_start + SwathWidthY[k] + req_width_horz_y - 1, req_width_horz_y) - math_floor2(display_cfg->plane_descriptors[k].composition.viewport.plane0.x_start, req_width_horz_y)));
- } else {
- swath_width_luma_ub[k] = (unsigned int)(math_min2(surface_width_ub_l, math_ceil2((double)SwathWidthY[k] - 1, req_width_horz_y) + req_width_horz_y));
- }
- req_per_swath_ub_l[k] = swath_width_luma_ub[k] / req_width_horz_y;
-
- if (BytePerPixC[k] > 0) {
- if (display_cfg->plane_descriptors[k].composition.viewport.stationary && DPPPerSurface[k] == 1) {
- swath_width_chroma_ub[k] = (unsigned int)(math_min2(surface_width_ub_c, math_floor2(display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start + SwathWidthC[k] + req_width_horz_c - 1, req_width_horz_c) - math_floor2(display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start, req_width_horz_c)));
- } else {
- swath_width_chroma_ub[k] = (unsigned int)(math_min2(surface_width_ub_c, math_ceil2((double)SwathWidthC[k] - 1, req_width_horz_c) + req_width_horz_c));
- }
- req_per_swath_ub_c[k] = swath_width_chroma_ub[k] / req_width_horz_c;
- } else {
- swath_width_chroma_ub[k] = 0;
- }
- } else {
- MaximumSwathHeightY[k] = Read256BytesBlockWidthY[k];
- MaximumSwathHeightC[k] = Read256BytesBlockWidthC[k];
-
- if (display_cfg->plane_descriptors[k].composition.viewport.stationary && DPPPerSurface[k] == 1) {
- swath_width_luma_ub[k] = (unsigned int)(math_min2(surface_height_ub_l, math_floor2(display_cfg->plane_descriptors[k].composition.viewport.plane0.y_start + SwathWidthY[k] + Read256BytesBlockHeightY[k] - 1, Read256BytesBlockHeightY[k]) - math_floor2(display_cfg->plane_descriptors[k].composition.viewport.plane0.y_start, Read256BytesBlockHeightY[k])));
- } else {
- swath_width_luma_ub[k] = (unsigned int)(math_min2(surface_height_ub_l, math_ceil2((double)SwathWidthY[k] - 1, Read256BytesBlockHeightY[k]) + Read256BytesBlockHeightY[k]));
- }
- req_per_swath_ub_l[k] = swath_width_luma_ub[k] / Read256BytesBlockHeightY[k];
- if (BytePerPixC[k] > 0) {
- if (display_cfg->plane_descriptors[k].composition.viewport.stationary && DPPPerSurface[k] == 1) {
- swath_width_chroma_ub[k] = (unsigned int)(math_min2(surface_height_ub_c, math_floor2(display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start + SwathWidthC[k] + Read256BytesBlockHeightC[k] - 1, Read256BytesBlockHeightC[k]) - math_floor2(display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start, Read256BytesBlockHeightC[k])));
- } else {
- swath_width_chroma_ub[k] = (unsigned int)(math_min2(surface_height_ub_c, math_ceil2((double)SwathWidthC[k] - 1, Read256BytesBlockHeightC[k]) + Read256BytesBlockHeightC[k]));
- }
- req_per_swath_ub_c[k] = swath_width_chroma_ub[k] / Read256BytesBlockHeightC[k];
- } else {
- swath_width_chroma_ub[k] = 0;
- }
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u swath_width_luma_ub=%u\n", __func__, k, swath_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u swath_width_chroma_ub=%u\n", __func__, k, swath_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathHeightY=%u\n", __func__, k, MaximumSwathHeightY[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathHeightC=%u\n", __func__, k, MaximumSwathHeightC[k]);
- dml2_printf("DML::%s: k=%u req_per_swath_ub_l=%u\n", __func__, k, req_per_swath_ub_l[k]);
- dml2_printf("DML::%s: k=%u req_per_swath_ub_c=%u\n", __func__, k, req_per_swath_ub_c[k]);
-#endif
-
- }
-}
-
-static bool UnboundedRequest(bool unb_req_force_en, bool unb_req_force_val, unsigned int TotalNumberOfActiveDPP, bool NoChromaOrLinear)
-{
- bool unb_req_ok = false;
- bool unb_req_en = false;
-
- unb_req_ok = (TotalNumberOfActiveDPP == 1 && NoChromaOrLinear);
- unb_req_en = unb_req_ok;
-
- if (unb_req_force_en) {
- unb_req_en = unb_req_force_val && unb_req_ok;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: unb_req_force_en = %u\n", __func__, unb_req_force_en);
- dml2_printf("DML::%s: unb_req_force_val = %u\n", __func__, unb_req_force_val);
- dml2_printf("DML::%s: unb_req_ok = %u\n", __func__, unb_req_ok);
- dml2_printf("DML::%s: unb_req_en = %u\n", __func__, unb_req_en);
-#endif
- return (unb_req_en);
-}
-
-static void CalculateDETBufferSize(struct dml2_core_shared_calculate_det_buffer_size_params *p)
-{
- unsigned int DETBufferSizePoolInKByte;
- unsigned int NextDETBufferPieceInKByte;
- bool DETPieceAssignedToThisSurfaceAlready[DML2_MAX_PLANES];
- bool NextPotentialSurfaceToAssignDETPieceFound;
- unsigned int NextSurfaceToAssignDETPiece;
- double TotalBandwidth;
- double BandwidthOfSurfacesNotAssignedDETPiece;
- unsigned int max_minDET;
- unsigned int minDET;
- unsigned int minDET_pipe;
- unsigned int TotalBandwidthPerStream[DML2_MAX_PLANES] = { 0 };
- unsigned int TotalPixelRate = 0;
- unsigned int DETBudgetPerStream[DML2_MAX_PLANES] = { 0 };
- unsigned int RemainingDETBudgetPerStream[DML2_MAX_PLANES] = { 0 };
- unsigned int IdealDETBudget, DeltaDETBudget;
- bool MinimizeReallocationSuccess = false;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ForceSingleDPP = %u\n", __func__, p->ForceSingleDPP);
- dml2_printf("DML::%s: nomDETInKByte = %u\n", __func__, p->nomDETInKByte);
- dml2_printf("DML::%s: NumberOfActiveSurfaces = %u\n", __func__, p->NumberOfActiveSurfaces);
- dml2_printf("DML::%s: UnboundedRequestEnabled = %u\n", __func__, p->UnboundedRequestEnabled);
- dml2_printf("DML::%s: MaxTotalDETInKByte = %u\n", __func__, p->MaxTotalDETInKByte);
- dml2_printf("DML::%s: ConfigReturnBufferSizeInKByte = %u\n", __func__, p->ConfigReturnBufferSizeInKByte);
- dml2_printf("DML::%s: MinCompressedBufferSizeInKByte = %u\n", __func__, p->MinCompressedBufferSizeInKByte);
- dml2_printf("DML::%s: CompressedBufferSegmentSizeInkByte = %u\n", __func__, p->CompressedBufferSegmentSizeInkByte);
-#endif
-
- // Note: Will use default det size if that fits 2 swaths
- if (p->UnboundedRequestEnabled) {
- if (p->display_cfg->plane_descriptors[0].overrides.det_size_override_kb > 0) {
- p->DETBufferSizeInKByte[0] = p->display_cfg->plane_descriptors[0].overrides.det_size_override_kb;
- } else {
- p->DETBufferSizeInKByte[0] = (unsigned int)math_max2(128.0, math_ceil2(2.0 * ((double)p->full_swath_bytes_l[0] + (double)p->full_swath_bytes_c[0]) / 1024.0, p->ConfigReturnBufferSegmentSizeInkByte));
- }
- *p->CompressedBufferSizeInkByte = p->ConfigReturnBufferSizeInKByte - p->DETBufferSizeInKByte[0];
- } else {
- DETBufferSizePoolInKByte = p->MaxTotalDETInKByte;
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- p->DETBufferSizeInKByte[k] = 0;
- if (dml2_core_shared_is_420(p->display_cfg->plane_descriptors[k].pixel_format)) {
- max_minDET = p->nomDETInKByte - p->ConfigReturnBufferSegmentSizeInkByte;
- } else {
- max_minDET = p->nomDETInKByte;
- }
- minDET = 128;
- minDET_pipe = 0;
-
- // add DET resource until can hold 2 full swaths
- while (minDET <= max_minDET && minDET_pipe == 0) {
- if (2.0 * ((double)p->full_swath_bytes_l[k] + (double)p->full_swath_bytes_c[k]) / 1024.0 <= minDET)
- minDET_pipe = minDET;
- minDET = minDET + p->ConfigReturnBufferSegmentSizeInkByte;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u minDET = %u\n", __func__, k, minDET);
- dml2_printf("DML::%s: k=%u max_minDET = %u\n", __func__, k, max_minDET);
- dml2_printf("DML::%s: k=%u minDET_pipe = %u\n", __func__, k, minDET_pipe);
- dml2_printf("DML::%s: k=%u full_swath_bytes_l = %u\n", __func__, k, p->full_swath_bytes_l[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_c = %u\n", __func__, k, p->full_swath_bytes_c[k]);
-#endif
-
- if (minDET_pipe == 0) {
- minDET_pipe = (unsigned int)(math_max2(128, math_ceil2(((double)p->full_swath_bytes_l[k] + (double)p->full_swath_bytes_c[k]) / 1024.0, p->ConfigReturnBufferSegmentSizeInkByte)));
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u minDET_pipe = %u (assume each plane take half DET)\n", __func__, k, minDET_pipe);
-#endif
- }
-
- if (dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) {
- p->DETBufferSizeInKByte[k] = 0;
- } else if (p->display_cfg->plane_descriptors[k].overrides.det_size_override_kb > 0) {
- p->DETBufferSizeInKByte[k] = p->display_cfg->plane_descriptors[k].overrides.det_size_override_kb;
- DETBufferSizePoolInKByte = DETBufferSizePoolInKByte - (p->ForceSingleDPP ? 1 : p->DPPPerSurface[k]) * p->display_cfg->plane_descriptors[k].overrides.det_size_override_kb;
- } else if ((p->ForceSingleDPP ? 1 : p->DPPPerSurface[k]) * minDET_pipe <= DETBufferSizePoolInKByte) {
- p->DETBufferSizeInKByte[k] = minDET_pipe;
- DETBufferSizePoolInKByte = DETBufferSizePoolInKByte - (p->ForceSingleDPP ? 1 : p->DPPPerSurface[k]) * minDET_pipe;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u DPPPerSurface = %u\n", __func__, k, p->DPPPerSurface[k]);
- dml2_printf("DML::%s: k=%u DETSizeOverride = %u\n", __func__, k, p->display_cfg->plane_descriptors[k].overrides.det_size_override_kb);
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByte = %u\n", __func__, k, p->DETBufferSizeInKByte[k]);
- dml2_printf("DML::%s: DETBufferSizePoolInKByte = %u\n", __func__, DETBufferSizePoolInKByte);
-#endif
- }
-
- if (p->display_cfg->minimize_det_reallocation) {
- MinimizeReallocationSuccess = true;
- // To minimize det reallocation, we don't distribute based on each surfaces bandwidth proportional to the global
- // but rather distribute DET across streams proportionally based on pixel rate, and only distribute based on
- // bandwidth between the planes on the same stream. This ensures that large scale re-distribution only on a
- // stream count and/or pixel rate change, which is must less likely then general bandwidth changes per plane.
-
- // Calculate total pixel rate
- for (unsigned int k = 0; k < p->display_cfg->num_streams; ++k) {
- TotalPixelRate += p->display_cfg->stream_descriptors[k].timing.pixel_clock_khz;
- }
-
- // Calculate per stream DET budget
- for (unsigned int k = 0; k < p->display_cfg->num_streams; ++k) {
- DETBudgetPerStream[k] = (unsigned int)((double)p->display_cfg->stream_descriptors[k].timing.pixel_clock_khz * p->MaxTotalDETInKByte / TotalPixelRate);
- RemainingDETBudgetPerStream[k] = DETBudgetPerStream[k];
- }
-
- // Calculate the per stream total bandwidth
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (!dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) {
- TotalBandwidthPerStream[p->display_cfg->plane_descriptors[k].stream_index] += (unsigned int)(p->ReadBandwidthLuma[k] + p->ReadBandwidthChroma[k]);
-
- // Check the minimum can be satisfied by budget
- if (RemainingDETBudgetPerStream[p->display_cfg->plane_descriptors[k].stream_index] >= p->DETBufferSizeInKByte[k]) {
- RemainingDETBudgetPerStream[p->display_cfg->plane_descriptors[k].stream_index] -= p->DETBufferSizeInKByte[k];
- } else {
- MinimizeReallocationSuccess = false;
- break;
- }
- }
- }
-
- if (MinimizeReallocationSuccess) {
- // Since a fixed budget per stream is sufficient to satisfy the minimums, just re-distribute each streams
- // budget proportionally across its planes
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (!dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) {
- IdealDETBudget = (unsigned int)(((p->ReadBandwidthLuma[k] + p->ReadBandwidthChroma[k]) / TotalBandwidthPerStream[p->display_cfg->plane_descriptors[k].stream_index])
- * DETBudgetPerStream[p->display_cfg->plane_descriptors[k].stream_index]);
-
- if (IdealDETBudget > p->DETBufferSizeInKByte[k]) {
- DeltaDETBudget = IdealDETBudget - p->DETBufferSizeInKByte[k];
- if (DeltaDETBudget > RemainingDETBudgetPerStream[p->display_cfg->plane_descriptors[k].stream_index])
- DeltaDETBudget = RemainingDETBudgetPerStream[p->display_cfg->plane_descriptors[k].stream_index];
-
- p->DETBufferSizeInKByte[k] += DeltaDETBudget;
- RemainingDETBudgetPerStream[p->display_cfg->plane_descriptors[k].stream_index] -= DeltaDETBudget;
- }
-
- // Split among the pipes per the plane
- p->DETBufferSizeInKByte[k] = (unsigned int)((double)p->DETBufferSizeInKByte[k] / (p->ForceSingleDPP ? 1 : p->DPPPerSurface[k]));
-
- // Round down to segment size
- p->DETBufferSizeInKByte[k] = (p->DETBufferSizeInKByte[k] / p->CompressedBufferSegmentSizeInkByte) * p->CompressedBufferSegmentSizeInkByte;
- }
- }
- }
- }
-
- if (!MinimizeReallocationSuccess) {
- TotalBandwidth = 0;
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (!dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) {
- TotalBandwidth = TotalBandwidth + p->ReadBandwidthLuma[k] + p->ReadBandwidthChroma[k];
- }
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: --- Before bandwidth adjustment ---\n", __func__);
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByte = %u\n", __func__, k, p->DETBufferSizeInKByte[k]);
- }
- dml2_printf("DML::%s: --- DET allocation with bandwidth ---\n", __func__);
-#endif
- dml2_printf("DML::%s: TotalBandwidth = %f\n", __func__, TotalBandwidth);
- BandwidthOfSurfacesNotAssignedDETPiece = TotalBandwidth;
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
-
- if (dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) {
- DETPieceAssignedToThisSurfaceAlready[k] = true;
- } else if (p->display_cfg->plane_descriptors[k].overrides.det_size_override_kb > 0 || (((double)(p->ForceSingleDPP ? 1 : p->DPPPerSurface[k]) * (double)p->DETBufferSizeInKByte[k] / (double)p->MaxTotalDETInKByte) >= ((p->ReadBandwidthLuma[k] + p->ReadBandwidthChroma[k]) / TotalBandwidth))) {
- DETPieceAssignedToThisSurfaceAlready[k] = true;
- BandwidthOfSurfacesNotAssignedDETPiece = BandwidthOfSurfacesNotAssignedDETPiece - p->ReadBandwidthLuma[k] - p->ReadBandwidthChroma[k];
- } else {
- DETPieceAssignedToThisSurfaceAlready[k] = false;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u DETPieceAssignedToThisSurfaceAlready = %u\n", __func__, k, DETPieceAssignedToThisSurfaceAlready[k]);
- dml2_printf("DML::%s: k=%u BandwidthOfSurfacesNotAssignedDETPiece = %f\n", __func__, k, BandwidthOfSurfacesNotAssignedDETPiece);
-#endif
- }
-
- for (unsigned int j = 0; j < p->NumberOfActiveSurfaces; ++j) {
- NextPotentialSurfaceToAssignDETPieceFound = false;
- NextSurfaceToAssignDETPiece = 0;
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: j=%u k=%u, ReadBandwidthLuma[k] = %f\n", __func__, j, k, p->ReadBandwidthLuma[k]);
- dml2_printf("DML::%s: j=%u k=%u, ReadBandwidthChroma[k] = %f\n", __func__, j, k, p->ReadBandwidthChroma[k]);
- dml2_printf("DML::%s: j=%u k=%u, ReadBandwidthLuma[Next] = %f\n", __func__, j, k, p->ReadBandwidthLuma[NextSurfaceToAssignDETPiece]);
- dml2_printf("DML::%s: j=%u k=%u, ReadBandwidthChroma[Next] = %f\n", __func__, j, k, p->ReadBandwidthChroma[NextSurfaceToAssignDETPiece]);
- dml2_printf("DML::%s: j=%u k=%u, NextSurfaceToAssignDETPiece = %u\n", __func__, j, k, NextSurfaceToAssignDETPiece);
-#endif
- if (!DETPieceAssignedToThisSurfaceAlready[k] && (!NextPotentialSurfaceToAssignDETPieceFound ||
- p->ReadBandwidthLuma[k] + p->ReadBandwidthChroma[k] < p->ReadBandwidthLuma[NextSurfaceToAssignDETPiece] + p->ReadBandwidthChroma[NextSurfaceToAssignDETPiece])) {
- NextSurfaceToAssignDETPiece = k;
- NextPotentialSurfaceToAssignDETPieceFound = true;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: j=%u k=%u, DETPieceAssignedToThisSurfaceAlready = %u\n", __func__, j, k, DETPieceAssignedToThisSurfaceAlready[k]);
- dml2_printf("DML::%s: j=%u k=%u, NextPotentialSurfaceToAssignDETPieceFound = %u\n", __func__, j, k, NextPotentialSurfaceToAssignDETPieceFound);
-#endif
- }
-
- if (NextPotentialSurfaceToAssignDETPieceFound) {
- NextDETBufferPieceInKByte = (unsigned int)(math_min2(
- math_round((double)DETBufferSizePoolInKByte * (p->ReadBandwidthLuma[NextSurfaceToAssignDETPiece] + p->ReadBandwidthChroma[NextSurfaceToAssignDETPiece]) / BandwidthOfSurfacesNotAssignedDETPiece /
- ((p->ForceSingleDPP ? 1 : p->DPPPerSurface[NextSurfaceToAssignDETPiece]) * p->ConfigReturnBufferSegmentSizeInkByte))
- * (p->ForceSingleDPP ? 1 : p->DPPPerSurface[NextSurfaceToAssignDETPiece]) * p->ConfigReturnBufferSegmentSizeInkByte,
- math_floor2((double)DETBufferSizePoolInKByte, (p->ForceSingleDPP ? 1 : p->DPPPerSurface[NextSurfaceToAssignDETPiece]) * p->ConfigReturnBufferSegmentSizeInkByte)));
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: j=%u, DETBufferSizePoolInKByte = %u\n", __func__, j, DETBufferSizePoolInKByte);
- dml2_printf("DML::%s: j=%u, NextSurfaceToAssignDETPiece = %u\n", __func__, j, NextSurfaceToAssignDETPiece);
- dml2_printf("DML::%s: j=%u, ReadBandwidthLuma[%u] = %f\n", __func__, j, NextSurfaceToAssignDETPiece, p->ReadBandwidthLuma[NextSurfaceToAssignDETPiece]);
- dml2_printf("DML::%s: j=%u, ReadBandwidthChroma[%u] = %f\n", __func__, j, NextSurfaceToAssignDETPiece, p->ReadBandwidthChroma[NextSurfaceToAssignDETPiece]);
- dml2_printf("DML::%s: j=%u, BandwidthOfSurfacesNotAssignedDETPiece = %f\n", __func__, j, BandwidthOfSurfacesNotAssignedDETPiece);
- dml2_printf("DML::%s: j=%u, NextDETBufferPieceInKByte = %u\n", __func__, j, NextDETBufferPieceInKByte);
- dml2_printf("DML::%s: j=%u, DETBufferSizeInKByte[%u] increases from %u ", __func__, j, NextSurfaceToAssignDETPiece, p->DETBufferSizeInKByte[NextSurfaceToAssignDETPiece]);
-#endif
-
- p->DETBufferSizeInKByte[NextSurfaceToAssignDETPiece] = p->DETBufferSizeInKByte[NextSurfaceToAssignDETPiece] + NextDETBufferPieceInKByte / (p->ForceSingleDPP ? 1 : p->DPPPerSurface[NextSurfaceToAssignDETPiece]);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("to %u\n", p->DETBufferSizeInKByte[NextSurfaceToAssignDETPiece]);
-#endif
-
- DETBufferSizePoolInKByte = DETBufferSizePoolInKByte - NextDETBufferPieceInKByte;
- DETPieceAssignedToThisSurfaceAlready[NextSurfaceToAssignDETPiece] = true;
- BandwidthOfSurfacesNotAssignedDETPiece = BandwidthOfSurfacesNotAssignedDETPiece - (p->ReadBandwidthLuma[NextSurfaceToAssignDETPiece] + p->ReadBandwidthChroma[NextSurfaceToAssignDETPiece]);
- }
- }
- }
- *p->CompressedBufferSizeInkByte = p->MinCompressedBufferSizeInKByte;
- }
- *p->CompressedBufferSizeInkByte = *p->CompressedBufferSizeInkByte * p->CompressedBufferSegmentSizeInkByte / p->ConfigReturnBufferSegmentSizeInkByte;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: --- After bandwidth adjustment ---\n", __func__);
- dml2_printf("DML::%s: CompressedBufferSizeInkByte = %u\n", __func__, *p->CompressedBufferSizeInkByte);
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByte = %u (TotalReadBandWidth=%f)\n", __func__, k, p->DETBufferSizeInKByte[k], p->ReadBandwidthLuma[k] + p->ReadBandwidthChroma[k]);
- }
-#endif
-}
-
-static double CalculateRequiredDispclk(
- enum dml2_odm_mode ODMMode,
- double PixelClock)
-{
-
- if (ODMMode == dml2_odm_mode_combine_4to1) {
- return PixelClock / 4.0;
- } else if (ODMMode == dml2_odm_mode_combine_3to1) {
- return PixelClock / 3.0;
- } else if (ODMMode == dml2_odm_mode_combine_2to1) {
- return PixelClock / 2.0;
- } else {
- return PixelClock;
- }
-}
-
-static double TruncToValidBPP(
- struct dml2_core_shared_TruncToValidBPP_locals *l,
- double LinkBitRate,
- unsigned int Lanes,
- unsigned int HTotal,
- unsigned int HActive,
- double PixelClock,
- double DesiredBPP,
- bool DSCEnable,
- enum dml2_output_encoder_class Output,
- enum dml2_output_format_class Format,
- unsigned int DSCInputBitPerComponent,
- unsigned int DSCSlices,
- unsigned int AudioRate,
- unsigned int AudioLayout,
- enum dml2_odm_mode ODMModeNoDSC,
- enum dml2_odm_mode ODMModeDSC,
-
- // Output
- unsigned int *RequiredSlots)
-{
- double MaxLinkBPP;
- unsigned int MinDSCBPP;
- double MaxDSCBPP;
- unsigned int NonDSCBPP0;
- unsigned int NonDSCBPP1;
- unsigned int NonDSCBPP2;
- enum dml2_odm_mode ODMMode;
-
- if (Format == dml2_420) {
- NonDSCBPP0 = 12;
- NonDSCBPP1 = 15;
- NonDSCBPP2 = 18;
- MinDSCBPP = 6;
- MaxDSCBPP = 16;
- } else if (Format == dml2_444) {
- NonDSCBPP0 = 24;
- NonDSCBPP1 = 30;
- NonDSCBPP2 = 36;
- MinDSCBPP = 8;
- MaxDSCBPP = 16;
- } else {
- if (Output == dml2_hdmi || Output == dml2_hdmifrl) {
- NonDSCBPP0 = 24;
- NonDSCBPP1 = 24;
- NonDSCBPP2 = 24;
- } else {
- NonDSCBPP0 = 16;
- NonDSCBPP1 = 20;
- NonDSCBPP2 = 24;
- }
- if (Format == dml2_n422 || Output == dml2_hdmifrl) {
- MinDSCBPP = 7;
- MaxDSCBPP = 16;
- } else {
- MinDSCBPP = 8;
- MaxDSCBPP = 16;
- }
- }
- if (Output == dml2_dp2p0) {
- MaxLinkBPP = LinkBitRate * Lanes / PixelClock * 128.0 / 132.0 * 383.0 / 384.0 * 65536.0 / 65540.0;
- } else if (DSCEnable && Output == dml2_dp) {
- MaxLinkBPP = LinkBitRate / 10.0 * 8.0 * Lanes / PixelClock * (1 - 2.4 / 100);
- } else {
- MaxLinkBPP = LinkBitRate / 10.0 * 8.0 * Lanes / PixelClock;
- }
-
- ODMMode = DSCEnable ? ODMModeDSC : ODMModeNoDSC;
-
- if (ODMMode == dml2_odm_mode_split_1to2) {
- MaxLinkBPP = 2 * MaxLinkBPP;
- }
-
- if (DesiredBPP == 0) {
- if (DSCEnable) {
- if (MaxLinkBPP < MinDSCBPP) {
- return __DML2_CALCS_DPP_INVALID__;
- } else if (MaxLinkBPP >= MaxDSCBPP) {
- return MaxDSCBPP;
- } else {
- return math_floor2(16.0 * MaxLinkBPP, 1.0) / 16.0;
- }
- } else {
- if (MaxLinkBPP >= NonDSCBPP2) {
- return NonDSCBPP2;
- } else if (MaxLinkBPP >= NonDSCBPP1) {
- return NonDSCBPP1;
- } else if (MaxLinkBPP >= NonDSCBPP0) {
- return NonDSCBPP0;
- } else {
- return __DML2_CALCS_DPP_INVALID__;
- }
- }
- } else {
- if (!((DSCEnable == false && (DesiredBPP == NonDSCBPP2 || DesiredBPP == NonDSCBPP1 || DesiredBPP == NonDSCBPP0)) ||
- (DSCEnable && DesiredBPP >= MinDSCBPP && DesiredBPP <= MaxDSCBPP))) {
- return __DML2_CALCS_DPP_INVALID__;
- } else {
- return DesiredBPP;
- }
- }
-}
-
-// updated for dcn4
-static unsigned int dscceComputeDelay(
- unsigned int bpc,
- double BPP,
- unsigned int sliceWidth,
- unsigned int numSlices,
- enum dml2_output_format_class pixelFormat,
- enum dml2_output_encoder_class Output)
-{
- // valid bpc = source bits per component in the set of {8, 10, 12}
- // valid bpp = increments of 1/16 of a bit
- // min = 6/7/8 in N420/N422/444, respectively
- // max = such that compression is 1:1
- //valid sliceWidth = number of pixels per slice line, must be less than or equal to 5184/numSlices (or 4096/numSlices in 420 mode)
- //valid numSlices = number of slices in the horiziontal direction per DSC engine in the set of {1, 2, 3, 4}
- //valid pixelFormat = pixel/color format in the set of {:N444_RGB, :S422, :N422, :N420}
-
- // fixed value
- unsigned int rcModelSize = 8192;
-
- // N422/N420 operate at 2 pixels per clock
- unsigned int pixelsPerClock, padding_pixels, ssm_group_priming_delay, ssm_pipeline_delay, obsm_pipeline_delay, slice_padded_pixels, ixd_plus_padding, ixd_plus_padding_groups, cycles_per_group, group_delay, pipeline_delay, pixels, additional_group_delay, lines_to_reach_ixd, groups_to_reach_ixd, slice_width_groups, initial_xmit_delay, number_of_lines_to_reach_ixd, slice_width_modified;
-
-
- if (pixelFormat == dml2_420)
- pixelsPerClock = 2;
- // #all other modes operate at 1 pixel per clock
- else if (pixelFormat == dml2_444)
- pixelsPerClock = 1;
- else if (pixelFormat == dml2_n422 || Output == dml2_hdmifrl)
- pixelsPerClock = 2;
- else
- pixelsPerClock = 1;
-
- //initial transmit delay as per PPS
- initial_xmit_delay = (unsigned int)(math_round(rcModelSize / 2.0 / BPP / pixelsPerClock));
-
- //slice width as seen by dscc_bcl in pixels or pixels pairs (depending on number of pixels per pixel container based on pixel format)
- slice_width_modified = (pixelFormat == dml2_444 || pixelFormat == dml2_420 || Output == dml2_hdmifrl) ? sliceWidth / 2 : sliceWidth;
-
- padding_pixels = ((slice_width_modified % 3) != 0) ? (3 - (slice_width_modified % 3)) * (initial_xmit_delay / slice_width_modified) : 0;
-
- if ((3.0 * pixelsPerClock * BPP) >= ((double)((initial_xmit_delay + 2) / 3) * (double)(3 + (pixelFormat == dml2_n422)))) {
- if ((initial_xmit_delay + padding_pixels) % 3 == 1) {
- initial_xmit_delay++;
- }
- }
-
-
- //sub-stream multiplexer balance fifo priming delay in groups as per dsc standard
- if (bpc == 8)
- ssm_group_priming_delay = 83;
- else if (bpc == 10)
- ssm_group_priming_delay = 91;
- else if (bpc == 12)
- ssm_group_priming_delay = 115;
- else if (bpc == 14)
- ssm_group_priming_delay = 123;
- else
- ssm_group_priming_delay = 128;
-
- //slice width in groups is rounded up to the nearest group as DSC adds padded pixels such that there are an integer number of groups per slice
- slice_width_groups = (slice_width_modified + 2) / 3;
-
- //determine number of padded pixels in the last group of a slice line, computed as
- slice_padded_pixels = 3 * slice_width_groups - slice_width_modified;
-
-
-
-
- //determine integer number of complete slice lines required to reach initial transmit delay without ssm delay considered
- number_of_lines_to_reach_ixd = initial_xmit_delay / slice_width_modified;
-
- //increase initial transmit delay by the number of padded pixels added to a slice line multipled by the integer number of complete lines to reach initial transmit delay
- //this step is necessary as each padded pixel added takes up a clock cycle and, therefore, adds to the overall delay
- ixd_plus_padding = initial_xmit_delay + slice_padded_pixels * number_of_lines_to_reach_ixd;
-
- //convert the padded initial transmit delay from pixels to groups by rounding up to the nearest group as DSC processes in groups of pixels
- ixd_plus_padding_groups = (ixd_plus_padding + 2) / 3;
-
- //number of groups required for a slice to reach initial transmit delay is the sum of the padded initial transmit delay plus the ssm group priming delay
- groups_to_reach_ixd = ixd_plus_padding_groups + ssm_group_priming_delay;
-
-
- //number of lines required to reach padded initial transmit delay in groups in slices to the left of the last horizontal slice
- //needs to be rounded up as a complete slice lines are buffered prior to initial transmit delay being reached in the last horizontal slice
- lines_to_reach_ixd = (groups_to_reach_ixd + slice_width_groups - 1) / slice_width_groups; //round up lines to reach ixd to next
-
- //determine if there are non-zero number of pixels reached in the group where initial transmit delay is reached
- //an additional group time (i.e., 3 pixel times) is required before the first output if there are no additional pixels beyond initial transmit delay
- additional_group_delay = ((initial_xmit_delay - number_of_lines_to_reach_ixd * slice_width_modified) % 3) == 0 ? 1 : 0;
-
- //number of pipeline delay cycles in the ssm block (can be determined empirically or analytically by inspecting the ssm block)
- ssm_pipeline_delay = 2;
-
- //number of pipe delay cycles in the obsm block (can be determined empirically or analytically by inspecting the obsm block)
- obsm_pipeline_delay = 1;
-
- //a group of pixels is worth 6 pixels in N422/N420 mode or 3 pixels in all other modes
- if (pixelFormat == dml2_420 || pixelFormat == dml2_444 || pixelFormat == dml2_n422 || Output == dml2_hdmifrl)
- cycles_per_group = 6;
- else
- cycles_per_group = 3;
- //delay of the bit stream contruction layer in pixels is the sum of:
- //1. number of pixel containers in a slice line multipled by the number of lines required to reach initial transmit delay multipled by number of slices to the left of the last horizontal slice
- //2. number of pixel containers required to reach initial transmit delay (specifically, in the last horizontal slice)
- //3. additional group of delay if initial transmit delay is reached exactly in a group
- //4. ssm and obsm pipeline delay (i.e., clock cycles of delay)
- group_delay = (lines_to_reach_ixd * slice_width_groups * (numSlices - 1)) + groups_to_reach_ixd + additional_group_delay;
- pipeline_delay = ssm_pipeline_delay + obsm_pipeline_delay;
-
- //pixel delay is group_delay (converted to pixels) + pipeline, however, first group is a special case since it is processed as soon as it arrives (i.e., in 3 cycles regardless of pixel format)
- pixels = (group_delay - 1) * cycles_per_group + 3 + pipeline_delay;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: bpc: %u\n", __func__, bpc);
- dml2_printf("DML::%s: BPP: %f\n", __func__, BPP);
- dml2_printf("DML::%s: sliceWidth: %u\n", __func__, sliceWidth);
- dml2_printf("DML::%s: numSlices: %u\n", __func__, numSlices);
- dml2_printf("DML::%s: pixelFormat: %u\n", __func__, pixelFormat);
- dml2_printf("DML::%s: Output: %u\n", __func__, Output);
- dml2_printf("DML::%s: pixels: %u\n", __func__, pixels);
-#endif
- return pixels;
-}
-
-
-//updated in dcn4
-static unsigned int dscComputeDelay(enum dml2_output_format_class pixelFormat, enum dml2_output_encoder_class Output)
-{
- unsigned int Delay = 0;
- unsigned int dispclk_per_dscclk = 3;
-
- // sfr
- Delay = Delay + 2;
-
- if (pixelFormat == dml2_420 || pixelFormat == dml2_n422 || (Output == dml2_hdmifrl && pixelFormat != dml2_444)) {
- dispclk_per_dscclk = 3 * 2;
- }
-
- if (pixelFormat == dml2_420) {
- //dscc top delay for pixel compression layer
- Delay = Delay + 16 * dispclk_per_dscclk;
-
- // dscc - input deserializer
- Delay = Delay + 5;
-
- // dscc - input cdc fifo
- Delay = Delay + 1 + 4 * dispclk_per_dscclk;
-
- // dscc - output cdc fifo
- Delay = Delay + 3 + 1 * dispclk_per_dscclk;
-
- // dscc - cdc uncertainty
- Delay = Delay + 3 + 3 * dispclk_per_dscclk;
- } else if (pixelFormat == dml2_n422 || (Output == dml2_hdmifrl && pixelFormat != dml2_444)) {
- //dscc top delay for pixel compression layer
- Delay = Delay + 16 * dispclk_per_dscclk;
- // dsccif
- Delay = Delay + 1;
- // dscc - input deserializer
- Delay = Delay + 5;
- // dscc - input cdc fifo
- Delay = Delay + 1 + 4 * dispclk_per_dscclk;
-
-
- // dscc - output cdc fifo
- Delay = Delay + 3 + 1 * dispclk_per_dscclk;
- // dscc - cdc uncertainty
- Delay = Delay + 3 + 3 * dispclk_per_dscclk;
- } else if (pixelFormat == dml2_s422) {
- //dscc top delay for pixel compression layer
- Delay = Delay + 17 * dispclk_per_dscclk;
-
- // dscc - input deserializer
- Delay = Delay + 3;
- // dscc - input cdc fifo
- Delay = Delay + 1 + 4 * dispclk_per_dscclk;
- // dscc - output cdc fifo
- Delay = Delay + 3 + 1 * dispclk_per_dscclk;
- // dscc - cdc uncertainty
- Delay = Delay + 3 + 3 * dispclk_per_dscclk;
- } else {
- //dscc top delay for pixel compression layer
- Delay = Delay + 16 * dispclk_per_dscclk;
- // dscc - input deserializer
- Delay = Delay + 3;
- // dscc - input cdc fifo
- Delay = Delay + 1 + 4 * dispclk_per_dscclk;
- // dscc - output cdc fifo
- Delay = Delay + 3 + 1 * dispclk_per_dscclk;
-
- // dscc - cdc uncertainty
- Delay = Delay + 3 + 3 * dispclk_per_dscclk;
- }
-
- // sft
- Delay = Delay + 1;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: pixelFormat = %u\n", __func__, pixelFormat);
- dml2_printf("DML::%s: Delay = %u\n", __func__, Delay);
-#endif
-
- return Delay;
-}
-
-static unsigned int CalculateHostVMDynamicLevels(
- bool GPUVMEnable,
- bool HostVMEnable,
- unsigned int HostVMMinPageSize,
- unsigned int HostVMMaxNonCachedPageTableLevels)
-{
- unsigned int HostVMDynamicLevels = 0;
-
- if (GPUVMEnable && HostVMEnable) {
- if (HostVMMinPageSize < 2048)
- HostVMDynamicLevels = HostVMMaxNonCachedPageTableLevels;
- else if (HostVMMinPageSize >= 2048 && HostVMMinPageSize < 1048576)
- HostVMDynamicLevels = (unsigned int)math_max2(0, (double)HostVMMaxNonCachedPageTableLevels - 1);
- else
- HostVMDynamicLevels = (unsigned int)math_max2(0, (double)HostVMMaxNonCachedPageTableLevels - 2);
- } else {
- HostVMDynamicLevels = 0;
- }
- return HostVMDynamicLevels;
-}
-
-static unsigned int CalculateVMAndRowBytes(struct dml2_core_shared_calculate_vm_and_row_bytes_params *p)
-{
- unsigned int extra_dpde_bytes;
- unsigned int extra_mpde_bytes;
- unsigned int MacroTileSizeBytes;
- unsigned int vp_height_dpte_ub;
-
- unsigned int meta_surface_bytes;
- unsigned int vm_bytes;
- unsigned int vp_height_meta_ub;
-
- *p->MetaRequestHeight = 8 * p->BlockHeight256Bytes;
- *p->MetaRequestWidth = 8 * p->BlockWidth256Bytes;
- if (p->SurfaceTiling == dml2_sw_linear) {
- *p->meta_row_height = 32;
- *p->meta_row_width = (unsigned int)(math_floor2(p->ViewportXStart + p->SwathWidth + *p->MetaRequestWidth - 1, *p->MetaRequestWidth) - math_floor2(p->ViewportXStart, *p->MetaRequestWidth));
- *p->meta_row_bytes = (unsigned int)(*p->meta_row_width * *p->MetaRequestHeight * p->BytePerPixel / 256.0); // FIXME_DCN4SW missing in old code but no dcc for linear anyways?
- } else if (!dml_is_vertical_rotation(p->RotationAngle)) {
- *p->meta_row_height = *p->MetaRequestHeight;
- if (p->ViewportStationary && p->NumberOfDPPs == 1) {
- *p->meta_row_width = (unsigned int)(math_floor2(p->ViewportXStart + p->SwathWidth + *p->MetaRequestWidth - 1, *p->MetaRequestWidth) - math_floor2(p->ViewportXStart, *p->MetaRequestWidth));
- } else {
- *p->meta_row_width = (unsigned int)(math_ceil2(p->SwathWidth - 1, *p->MetaRequestWidth) + *p->MetaRequestWidth);
- }
- *p->meta_row_bytes = (unsigned int)(*p->meta_row_width * *p->MetaRequestHeight * p->BytePerPixel / 256.0);
- } else {
- *p->meta_row_height = *p->MetaRequestWidth;
- if (p->ViewportStationary && p->NumberOfDPPs == 1) {
- *p->meta_row_width = (unsigned int)(math_floor2(p->ViewportYStart + p->ViewportHeight + *p->MetaRequestHeight - 1, *p->MetaRequestHeight) - math_floor2(p->ViewportYStart, *p->MetaRequestHeight));
- } else {
- *p->meta_row_width = (unsigned int)(math_ceil2(p->SwathWidth - 1, *p->MetaRequestHeight) + *p->MetaRequestHeight);
- }
- *p->meta_row_bytes = (unsigned int)(*p->meta_row_width * *p->MetaRequestWidth * p->BytePerPixel / 256.0);
- }
-
- if (p->ViewportStationary && p->is_phantom && (p->NumberOfDPPs == 1 || !dml_is_vertical_rotation(p->RotationAngle))) {
- vp_height_meta_ub = (unsigned int)(math_floor2(p->ViewportYStart + p->ViewportHeight + 64 * p->BlockHeight256Bytes - 1, 64 * p->BlockHeight256Bytes) - math_floor2(p->ViewportYStart, 64 * p->BlockHeight256Bytes));
- } else if (!dml_is_vertical_rotation(p->RotationAngle)) {
- vp_height_meta_ub = (unsigned int)(math_ceil2(p->ViewportHeight - 1, 64 * p->BlockHeight256Bytes) + 64 * p->BlockHeight256Bytes);
- } else {
- vp_height_meta_ub = (unsigned int)(math_ceil2(p->SwathWidth - 1, 64 * p->BlockHeight256Bytes) + 64 * p->BlockHeight256Bytes);
- }
-
- meta_surface_bytes = (unsigned int)(p->DCCMetaPitch * vp_height_meta_ub * p->BytePerPixel / 256.0);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DCCMetaPitch = %u\n", __func__, p->DCCMetaPitch);
- dml2_printf("DML::%s: meta_surface_bytes = %u\n", __func__, meta_surface_bytes);
-#endif
- if (p->GPUVMEnable == true) {
- double meta_vmpg_bytes = 4.0 * 1024.0;
- *p->meta_pte_bytes_per_frame_ub = (unsigned int)((math_ceil2((double)(meta_surface_bytes - meta_vmpg_bytes) / (8 * meta_vmpg_bytes), 1) + 1) * 64);
- extra_mpde_bytes = 128 * (p->GPUVMMaxPageTableLevels - 1);
- } else {
- *p->meta_pte_bytes_per_frame_ub = 0;
- extra_mpde_bytes = 0;
- }
-
- if (!p->DCCEnable || !p->mrq_present) {
- *p->meta_pte_bytes_per_frame_ub = 0;
- extra_mpde_bytes = 0;
- *p->meta_row_bytes = 0;
- }
-
- if (!p->GPUVMEnable) {
- *p->PixelPTEBytesPerRow = 0;
- *p->PixelPTEBytesPerRowStorage = 0;
- *p->dpte_row_width_ub = 0;
- *p->dpte_row_height = 0;
- *p->dpte_row_height_linear = 0;
- *p->PixelPTEBytesPerRow_one_row_per_frame = 0;
- *p->dpte_row_width_ub_one_row_per_frame = 0;
- *p->dpte_row_height_one_row_per_frame = 0;
- *p->vmpg_width = 0;
- *p->vmpg_height = 0;
- *p->PixelPTEReqWidth = 0;
- *p->PixelPTEReqHeight = 0;
- *p->PTERequestSize = 0;
- *p->dpde0_bytes_per_frame_ub = 0;
- return 0;
- }
-
- MacroTileSizeBytes = p->MacroTileWidth * p->BytePerPixel * p->MacroTileHeight;
-
- if (p->ViewportStationary && p->is_phantom && (p->NumberOfDPPs == 1 || !dml_is_vertical_rotation(p->RotationAngle))) {
- vp_height_dpte_ub = (unsigned int)(math_floor2(p->ViewportYStart + p->ViewportHeight + p->MacroTileHeight - 1, p->MacroTileHeight) - math_floor2(p->ViewportYStart, p->MacroTileHeight));
- } else if (!dml_is_vertical_rotation(p->RotationAngle)) {
- vp_height_dpte_ub = (unsigned int)(math_ceil2((double)p->ViewportHeight - 1, p->MacroTileHeight) + p->MacroTileHeight);
- } else {
- vp_height_dpte_ub = (unsigned int)(math_ceil2((double)p->SwathWidth - 1, p->MacroTileHeight) + p->MacroTileHeight);
- }
-
- if (p->GPUVMEnable == true && p->GPUVMMaxPageTableLevels > 1) {
- *p->dpde0_bytes_per_frame_ub = (unsigned int)(64 * (math_ceil2((double)(p->Pitch * vp_height_dpte_ub * p->BytePerPixel - MacroTileSizeBytes) / (double)(8 * 2097152), 1) + 1));
- extra_dpde_bytes = 128 * (p->GPUVMMaxPageTableLevels - 2);
- } else {
- *p->dpde0_bytes_per_frame_ub = 0;
- extra_dpde_bytes = 0;
- }
-
- vm_bytes = *p->meta_pte_bytes_per_frame_ub + extra_mpde_bytes + *p->dpde0_bytes_per_frame_ub + extra_dpde_bytes;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DCCEnable = %u\n", __func__, p->DCCEnable);
- dml2_printf("DML::%s: GPUVMEnable = %u\n", __func__, p->GPUVMEnable);
- dml2_printf("DML::%s: SwModeLinear = %u\n", __func__, p->SurfaceTiling == dml2_sw_linear);
- dml2_printf("DML::%s: BytePerPixel = %u\n", __func__, p->BytePerPixel);
- dml2_printf("DML::%s: GPUVMMaxPageTableLevels = %u\n", __func__, p->GPUVMMaxPageTableLevels);
- dml2_printf("DML::%s: BlockHeight256Bytes = %u\n", __func__, p->BlockHeight256Bytes);
- dml2_printf("DML::%s: BlockWidth256Bytes = %u\n", __func__, p->BlockWidth256Bytes);
- dml2_printf("DML::%s: MacroTileHeight = %u\n", __func__, p->MacroTileHeight);
- dml2_printf("DML::%s: MacroTileWidth = %u\n", __func__, p->MacroTileWidth);
- dml2_printf("DML::%s: meta_pte_bytes_per_frame_ub = %u\n", __func__, *p->meta_pte_bytes_per_frame_ub);
- dml2_printf("DML::%s: dpde0_bytes_per_frame_ub = %u\n", __func__, *p->dpde0_bytes_per_frame_ub);
- dml2_printf("DML::%s: extra_mpde_bytes = %u\n", __func__, extra_mpde_bytes);
- dml2_printf("DML::%s: extra_dpde_bytes = %u\n", __func__, extra_dpde_bytes);
- dml2_printf("DML::%s: vm_bytes = %u\n", __func__, vm_bytes);
- dml2_printf("DML::%s: ViewportHeight = %u\n", __func__, p->ViewportHeight);
- dml2_printf("DML::%s: SwathWidth = %u\n", __func__, p->SwathWidth);
- dml2_printf("DML::%s: vp_height_dpte_ub = %u\n", __func__, vp_height_dpte_ub);
-#endif
-
- unsigned int PixelPTEReqWidth_linear = 0; // VBA_DELTA. VBA doesn't calculate this
-
- if (p->SurfaceTiling == dml2_sw_linear) {
- *p->PixelPTEReqHeight = 1;
- *p->PixelPTEReqWidth = p->GPUVMMinPageSizeKBytes * 1024 * 8 / p->BytePerPixel;
- PixelPTEReqWidth_linear = p->GPUVMMinPageSizeKBytes * 1024 * 8 / p->BytePerPixel;
- *p->PTERequestSize = 64;
-
- *p->vmpg_height = 1;
- *p->vmpg_width = p->GPUVMMinPageSizeKBytes * 1024 / p->BytePerPixel;
- } else if (p->GPUVMMinPageSizeKBytes * 1024 >= dml_get_tile_block_size_bytes(p->SurfaceTiling)) { // 1 64B 8x1 PTE
- *p->PixelPTEReqHeight = p->MacroTileHeight;
- *p->PixelPTEReqWidth = 8 * 1024 * p->GPUVMMinPageSizeKBytes / (p->MacroTileHeight * p->BytePerPixel);
- *p->PTERequestSize = 64;
-
- *p->vmpg_height = p->MacroTileHeight;
- *p->vmpg_width = 1024 * p->GPUVMMinPageSizeKBytes / (p->MacroTileHeight * p->BytePerPixel);
-
- } else if (p->GPUVMMinPageSizeKBytes == 4 && dml_get_tile_block_size_bytes(p->SurfaceTiling) == 65536) { // 2 64B PTE requests to get 16 PTEs to cover the 64K tile
- // one 64KB tile, is 16x16x256B req
- *p->PixelPTEReqHeight = 16 * p->BlockHeight256Bytes;
- *p->PixelPTEReqWidth = 16 * p->BlockWidth256Bytes;
- *p->PTERequestSize = 128;
-
- *p->vmpg_height = *p->PixelPTEReqHeight;
- *p->vmpg_width = *p->PixelPTEReqWidth;
- } else {
- // default for rest of calculation to go through, when vm is disable, the calulated pte related values shouldnt be used anyways
- *p->PixelPTEReqHeight = p->MacroTileHeight;
- *p->PixelPTEReqWidth = 8 * 1024 * p->GPUVMMinPageSizeKBytes / (p->MacroTileHeight * p->BytePerPixel);
- *p->PTERequestSize = 64;
-
- *p->vmpg_height = p->MacroTileHeight;
- *p->vmpg_width = 1024 * p->GPUVMMinPageSizeKBytes / (p->MacroTileHeight * p->BytePerPixel);
-
- if (p->GPUVMEnable == true) {
- dml2_printf("DML::%s: GPUVMMinPageSizeKBytes=%u and sw_mode=%u (tile_size=%d) not supported!\n",
- __func__, p->GPUVMMinPageSizeKBytes, p->SurfaceTiling, dml_get_tile_block_size_bytes(p->SurfaceTiling));
- DML2_ASSERT(0);
- }
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: GPUVMMinPageSizeKBytes = %u\n", __func__, p->GPUVMMinPageSizeKBytes);
- dml2_printf("DML::%s: PixelPTEReqHeight = %u\n", __func__, *p->PixelPTEReqHeight);
- dml2_printf("DML::%s: PixelPTEReqWidth = %u\n", __func__, *p->PixelPTEReqWidth);
- dml2_printf("DML::%s: PixelPTEReqWidth_linear = %u\n", __func__, PixelPTEReqWidth_linear);
- dml2_printf("DML::%s: PTERequestSize = %u\n", __func__, *p->PTERequestSize);
- dml2_printf("DML::%s: Pitch = %u\n", __func__, p->Pitch);
- dml2_printf("DML::%s: vmpg_width = %u\n", __func__, *p->vmpg_width);
- dml2_printf("DML::%s: vmpg_height = %u\n", __func__, *p->vmpg_height);
-#endif
-
- *p->dpte_row_height_one_row_per_frame = vp_height_dpte_ub;
- *p->dpte_row_width_ub_one_row_per_frame = (unsigned int)((math_ceil2(((double)p->Pitch * (double)*p->dpte_row_height_one_row_per_frame / (double)*p->PixelPTEReqHeight - 1) / (double)*p->PixelPTEReqWidth, 1) + 1) * (double)*p->PixelPTEReqWidth);
- *p->PixelPTEBytesPerRow_one_row_per_frame = (unsigned int)((double)*p->dpte_row_width_ub_one_row_per_frame / (double)*p->PixelPTEReqWidth * *p->PTERequestSize);
-
- if (p->SurfaceTiling == dml2_sw_linear) {
- *p->dpte_row_height = (unsigned int)(math_min2(128, (double)(1ULL << (unsigned int)math_floor2(math_log((float)(p->PTEBufferSizeInRequests * *p->PixelPTEReqWidth / p->Pitch), 2.0), 1))));
- *p->dpte_row_width_ub = (unsigned int)(math_ceil2(((double)p->Pitch * (double)*p->dpte_row_height - 1), (double)*p->PixelPTEReqWidth) + *p->PixelPTEReqWidth);
- *p->PixelPTEBytesPerRow = (unsigned int)((double)*p->dpte_row_width_ub / (double)*p->PixelPTEReqWidth * *p->PTERequestSize);
- *p->dpte_row_height_linear = 0;
-
- // VBA_DELTA, VBA doesn't have programming value for pte row height linear.
- *p->dpte_row_height_linear = (unsigned int)1 << (unsigned int)math_floor2(math_log((float)(p->PTEBufferSizeInRequests * PixelPTEReqWidth_linear / p->Pitch), 2.0), 1);
- if (*p->dpte_row_height_linear > 128)
- *p->dpte_row_height_linear = 128;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: dpte_row_width_ub = %u (linear)\n", __func__, *p->dpte_row_width_ub);
-#endif
-
- } else if (!dml_is_vertical_rotation(p->RotationAngle)) {
- *p->dpte_row_height = *p->PixelPTEReqHeight;
-
- if (p->GPUVMMinPageSizeKBytes > 64) {
- *p->dpte_row_width_ub = (unsigned int)((math_ceil2(((double)p->Pitch * (double)*p->dpte_row_height / (double)*p->PixelPTEReqHeight - 1) / (double)*p->PixelPTEReqWidth, 1) + 1) * *p->PixelPTEReqWidth);
- } else if (p->ViewportStationary && (p->NumberOfDPPs == 1)) {
- *p->dpte_row_width_ub = (unsigned int)(math_floor2(p->ViewportXStart + p->SwathWidth + *p->PixelPTEReqWidth - 1, *p->PixelPTEReqWidth) - math_floor2(p->ViewportXStart, *p->PixelPTEReqWidth));
- } else {
- *p->dpte_row_width_ub = (unsigned int)((math_ceil2((double)(p->SwathWidth - 1) / (double)*p->PixelPTEReqWidth, 1) + 1.0) * *p->PixelPTEReqWidth);
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: dpte_row_width_ub = %u (tiled horz)\n", __func__, *p->dpte_row_width_ub);
-#endif
-
- *p->PixelPTEBytesPerRow = *p->dpte_row_width_ub / *p->PixelPTEReqWidth * *p->PTERequestSize;
- } else {
- *p->dpte_row_height = (unsigned int)(math_min2(*p->PixelPTEReqWidth, p->MacroTileWidth));
-
- if (p->ViewportStationary && (p->NumberOfDPPs == 1)) {
- *p->dpte_row_width_ub = (unsigned int)(math_floor2(p->ViewportYStart + p->ViewportHeight + *p->PixelPTEReqHeight - 1, *p->PixelPTEReqHeight) - math_floor2(p->ViewportYStart, *p->PixelPTEReqHeight));
- } else {
- *p->dpte_row_width_ub = (unsigned int)((math_ceil2((double)(p->SwathWidth - 1) / (double)*p->PixelPTEReqHeight, 1) + 1) * *p->PixelPTEReqHeight);
- }
-
- *p->PixelPTEBytesPerRow = (unsigned int)((double)*p->dpte_row_width_ub / (double)*p->PixelPTEReqHeight * *p->PTERequestSize);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: dpte_row_width_ub = %u (tiled vert)\n", __func__, *p->dpte_row_width_ub);
-#endif
- }
-
- if (p->GPUVMEnable != true) {
- *p->PixelPTEBytesPerRow = 0;
- *p->PixelPTEBytesPerRow_one_row_per_frame = 0;
- }
-
- *p->PixelPTEBytesPerRowStorage = *p->PixelPTEBytesPerRow;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: GPUVMMinPageSizeKBytes = %u\n", __func__, p->GPUVMMinPageSizeKBytes);
- dml2_printf("DML::%s: GPUVMEnable = %u\n", __func__, p->GPUVMEnable);
- dml2_printf("DML::%s: dpte_row_height = %u\n", __func__, *p->dpte_row_height);
- dml2_printf("DML::%s: dpte_row_height_linear = %u\n", __func__, *p->dpte_row_height_linear);
- dml2_printf("DML::%s: dpte_row_width_ub = %u\n", __func__, *p->dpte_row_width_ub);
- dml2_printf("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, *p->PixelPTEBytesPerRow);
- dml2_printf("DML::%s: PixelPTEBytesPerRowStorage = %u\n", __func__, *p->PixelPTEBytesPerRowStorage);
- dml2_printf("DML::%s: PTEBufferSizeInRequests = %u\n", __func__, p->PTEBufferSizeInRequests);
- dml2_printf("DML::%s: dpte_row_height_one_row_per_frame = %u\n", __func__, *p->dpte_row_height_one_row_per_frame);
- dml2_printf("DML::%s: dpte_row_width_ub_one_row_per_frame = %u\n", __func__, *p->dpte_row_width_ub_one_row_per_frame);
- dml2_printf("DML::%s: PixelPTEBytesPerRow_one_row_per_frame = %u\n", __func__, *p->PixelPTEBytesPerRow_one_row_per_frame);
-#endif
-
- return vm_bytes;
-} // CalculateVMAndRowBytes
-
-static unsigned int CalculatePrefetchSourceLines(
- double VRatio,
- unsigned int VTaps,
- bool Interlace,
- bool ProgressiveToInterlaceUnitInOPP,
- unsigned int SwathHeight,
- enum dml2_rotation_angle RotationAngle,
- bool mirrored,
- bool ViewportStationary,
- unsigned int SwathWidth,
- unsigned int ViewportHeight,
- unsigned int ViewportXStart,
- unsigned int ViewportYStart,
-
- // Output
- unsigned int *VInitPreFill,
- unsigned int *MaxNumSwath)
-{
-
- unsigned int vp_start_rot = 0;
- unsigned int sw0_tmp = 0;
- unsigned int MaxPartialSwath = 0;
- double numLines = 0;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatio = %f\n", __func__, VRatio);
- dml2_printf("DML::%s: VTaps = %u\n", __func__, VTaps);
- dml2_printf("DML::%s: ViewportXStart = %u\n", __func__, ViewportXStart);
- dml2_printf("DML::%s: ViewportYStart = %u\n", __func__, ViewportYStart);
- dml2_printf("DML::%s: ViewportStationary = %u\n", __func__, ViewportStationary);
- dml2_printf("DML::%s: SwathHeight = %u\n", __func__, SwathHeight);
-#endif
- if (ProgressiveToInterlaceUnitInOPP)
- *VInitPreFill = (unsigned int)(math_floor2((VRatio + (double)VTaps + 1) / 2.0, 1));
- else
- *VInitPreFill = (unsigned int)(math_floor2((VRatio + (double)VTaps + 1 + (Interlace ? 1 : 0) * 0.5 * VRatio) / 2.0, 1));
-
- if (ViewportStationary) {
- if (RotationAngle == dml2_rotation_180) {
- vp_start_rot = SwathHeight - (((unsigned int)(ViewportYStart + ViewportHeight - 1) % SwathHeight) + 1);
- } else if ((RotationAngle == dml2_rotation_270 && !mirrored) || (RotationAngle == dml2_rotation_90 && mirrored)) {
- vp_start_rot = ViewportXStart;
- } else if ((RotationAngle == dml2_rotation_90 && !mirrored) || (RotationAngle == dml2_rotation_270 && mirrored)) {
- vp_start_rot = SwathHeight - (((unsigned int)(ViewportYStart + SwathWidth - 1) % SwathHeight) + 1);
- } else {
- vp_start_rot = ViewportYStart;
- }
- sw0_tmp = SwathHeight - (vp_start_rot % SwathHeight);
- if (sw0_tmp < *VInitPreFill) {
- *MaxNumSwath = (unsigned int)(math_ceil2((*VInitPreFill - sw0_tmp) / (double)SwathHeight, 1) + 1);
- } else {
- *MaxNumSwath = 1;
- }
- MaxPartialSwath = (unsigned int)(math_max2(1, (unsigned int)(vp_start_rot + *VInitPreFill - 1) % SwathHeight));
- } else {
- *MaxNumSwath = (unsigned int)(math_ceil2((*VInitPreFill - 1.0) / (double)SwathHeight, 1) + 1);
- if (*VInitPreFill > 1) {
- MaxPartialSwath = (unsigned int)(math_max2(1, (unsigned int)(*VInitPreFill - 2) % SwathHeight));
- } else {
- MaxPartialSwath = (unsigned int)(math_max2(1, (unsigned int)(*VInitPreFill + SwathHeight - 2) % SwathHeight));
- }
- }
- numLines = *MaxNumSwath * SwathHeight + MaxPartialSwath;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: vp_start_rot = %u\n", __func__, vp_start_rot);
- dml2_printf("DML::%s: VInitPreFill = %u\n", __func__, *VInitPreFill);
- dml2_printf("DML::%s: MaxPartialSwath = %u\n", __func__, MaxPartialSwath);
- dml2_printf("DML::%s: MaxNumSwath = %u\n", __func__, *MaxNumSwath);
- dml2_printf("DML::%s: Prefetch source lines = %3.2f\n", __func__, numLines);
-#endif
- return (unsigned int)(numLines);
-
-}
-
-static void CalculateRowBandwidth(
- bool GPUVMEnable,
- bool use_one_row_for_frame,
- enum dml2_source_format_class SourcePixelFormat,
- double VRatio,
- double VRatioChroma,
- bool DCCEnable,
- double LineTime,
- unsigned int PixelPTEBytesPerRowLuma,
- unsigned int PixelPTEBytesPerRowChroma,
- unsigned int dpte_row_height_luma,
- unsigned int dpte_row_height_chroma,
-
- bool mrq_present,
- unsigned int meta_row_bytes_per_row_ub_l,
- unsigned int meta_row_bytes_per_row_ub_c,
- unsigned int meta_row_height_luma,
- unsigned int meta_row_height_chroma,
-
- // Output
- double *dpte_row_bw,
- double *meta_row_bw)
-{
- if (!DCCEnable || !mrq_present) {
- *meta_row_bw = 0;
- } else if (dml2_core_shared_is_420(SourcePixelFormat) || SourcePixelFormat == dml2_rgbe_alpha) {
- *meta_row_bw = VRatio * meta_row_bytes_per_row_ub_l / (meta_row_height_luma * LineTime)
- + VRatioChroma * meta_row_bytes_per_row_ub_c / (meta_row_height_chroma * LineTime);
- } else {
- *meta_row_bw = VRatio * meta_row_bytes_per_row_ub_l / (meta_row_height_luma * LineTime);
- }
-
- if (GPUVMEnable != true) {
- *dpte_row_bw = 0;
- } else if (dml2_core_shared_is_420(SourcePixelFormat) || SourcePixelFormat == dml2_rgbe_alpha) {
- *dpte_row_bw = VRatio * PixelPTEBytesPerRowLuma / (dpte_row_height_luma * LineTime)
- + VRatioChroma * PixelPTEBytesPerRowChroma / (dpte_row_height_chroma * LineTime);
- } else {
- *dpte_row_bw = VRatio * PixelPTEBytesPerRowLuma / (dpte_row_height_luma * LineTime);
- }
-}
-
-static void CalculateMALLUseForStaticScreen(
- const struct dml2_display_cfg *display_cfg,
- unsigned int NumberOfActiveSurfaces,
- unsigned int MALLAllocatedForDCN,
- unsigned int SurfaceSizeInMALL[],
- bool one_row_per_frame_fits_in_buffer[],
-
- // Output
- bool is_using_mall_for_ss[])
-{
-
- unsigned int SurfaceToAddToMALL;
- bool CanAddAnotherSurfaceToMALL;
- unsigned int TotalSurfaceSizeInMALL;
-
- TotalSurfaceSizeInMALL = 0;
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- is_using_mall_for_ss[k] = (display_cfg->plane_descriptors[k].overrides.refresh_from_mall == dml2_refresh_from_mall_mode_override_force_enable);
- if (is_using_mall_for_ss[k])
- TotalSurfaceSizeInMALL = TotalSurfaceSizeInMALL + SurfaceSizeInMALL[k];
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, is_using_mall_for_ss = %u\n", __func__, k, is_using_mall_for_ss[k]);
- dml2_printf("DML::%s: k=%u, TotalSurfaceSizeInMALL = %u\n", __func__, k, TotalSurfaceSizeInMALL);
-#endif
- }
-
- SurfaceToAddToMALL = 0;
- CanAddAnotherSurfaceToMALL = true;
- while (CanAddAnotherSurfaceToMALL) {
- CanAddAnotherSurfaceToMALL = false;
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- if (TotalSurfaceSizeInMALL + SurfaceSizeInMALL[k] <= MALLAllocatedForDCN * 1024 * 1024 &&
- !is_using_mall_for_ss[k] && display_cfg->plane_descriptors[k].overrides.refresh_from_mall != dml2_refresh_from_mall_mode_override_force_disable && one_row_per_frame_fits_in_buffer[k] &&
- (!CanAddAnotherSurfaceToMALL || SurfaceSizeInMALL[k] < SurfaceSizeInMALL[SurfaceToAddToMALL])) {
- CanAddAnotherSurfaceToMALL = true;
- SurfaceToAddToMALL = k;
- dml2_printf("DML::%s: k=%u, UseMALLForStaticScreen = %u (dis, en, optimize)\n", __func__, k, display_cfg->plane_descriptors[k].overrides.refresh_from_mall);
- }
- }
- if (CanAddAnotherSurfaceToMALL) {
- is_using_mall_for_ss[SurfaceToAddToMALL] = true;
- TotalSurfaceSizeInMALL = TotalSurfaceSizeInMALL + SurfaceSizeInMALL[SurfaceToAddToMALL];
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: SurfaceToAddToMALL = %u\n", __func__, SurfaceToAddToMALL);
- dml2_printf("DML::%s: TotalSurfaceSizeInMALL = %u\n", __func__, TotalSurfaceSizeInMALL);
-#endif
- }
- }
-}
-
-static void CalculateDCCConfiguration(
- bool DCCEnabled,
- bool DCCProgrammingAssumesScanDirectionUnknown,
- enum dml2_source_format_class SourcePixelFormat,
- unsigned int SurfaceWidthLuma,
- unsigned int SurfaceWidthChroma,
- unsigned int SurfaceHeightLuma,
- unsigned int SurfaceHeightChroma,
- unsigned int nomDETInKByte,
- unsigned int RequestHeight256ByteLuma,
- unsigned int RequestHeight256ByteChroma,
- enum dml2_swizzle_mode TilingFormat,
- unsigned int BytePerPixelY,
- unsigned int BytePerPixelC,
- double BytePerPixelDETY,
- double BytePerPixelDETC,
- enum dml2_rotation_angle RotationAngle,
-
- // Output
- enum dml2_core_internal_request_type *RequestLuma,
- enum dml2_core_internal_request_type *RequestChroma,
- unsigned int *MaxUncompressedBlockLuma,
- unsigned int *MaxUncompressedBlockChroma,
- unsigned int *MaxCompressedBlockLuma,
- unsigned int *MaxCompressedBlockChroma,
- unsigned int *IndependentBlockLuma,
- unsigned int *IndependentBlockChroma)
-{
- unsigned int DETBufferSizeForDCC = nomDETInKByte * 1024;
-
- unsigned int yuv420;
- unsigned int horz_div_l;
- unsigned int horz_div_c;
- unsigned int vert_div_l;
- unsigned int vert_div_c;
-
- unsigned int swath_buf_size;
- double detile_buf_vp_horz_limit;
- double detile_buf_vp_vert_limit;
-
- yuv420 = dml2_core_shared_is_420(SourcePixelFormat);
- horz_div_l = 1;
- horz_div_c = 1;
- vert_div_l = 1;
- vert_div_c = 1;
-
- if (BytePerPixelY == 1)
- vert_div_l = 0;
- if (BytePerPixelC == 1)
- vert_div_c = 0;
-
- if (BytePerPixelC == 0) {
- swath_buf_size = DETBufferSizeForDCC / 2 - 2 * 256;
- detile_buf_vp_horz_limit = (double)swath_buf_size / ((double)RequestHeight256ByteLuma * BytePerPixelY / (1 + horz_div_l));
- detile_buf_vp_vert_limit = (double)swath_buf_size / (256.0 / RequestHeight256ByteLuma / (1 + vert_div_l));
- } else {
- swath_buf_size = DETBufferSizeForDCC / 2 - 2 * 2 * 256;
- detile_buf_vp_horz_limit = (double)swath_buf_size / ((double)RequestHeight256ByteLuma * BytePerPixelY / (1 + horz_div_l) + (double)RequestHeight256ByteChroma * BytePerPixelC / (1 + horz_div_c) / (1 + yuv420));
- detile_buf_vp_vert_limit = (double)swath_buf_size / (256.0 / RequestHeight256ByteLuma / (1 + vert_div_l) + 256.0 / RequestHeight256ByteChroma / (1 + vert_div_c) / (1 + yuv420));
- }
-
- if (SourcePixelFormat == dml2_420_10) {
- detile_buf_vp_horz_limit = 1.5 * detile_buf_vp_horz_limit;
- detile_buf_vp_vert_limit = 1.5 * detile_buf_vp_vert_limit;
- }
-
- detile_buf_vp_horz_limit = math_floor2(detile_buf_vp_horz_limit - 1, 16);
- detile_buf_vp_vert_limit = math_floor2(detile_buf_vp_vert_limit - 1, 16);
-
- unsigned int MAS_vp_horz_limit;
- unsigned int MAS_vp_vert_limit;
- unsigned int max_vp_horz_width;
- unsigned int max_vp_vert_height;
- unsigned int eff_surf_width_l;
- unsigned int eff_surf_width_c;
- unsigned int eff_surf_height_l;
- unsigned int eff_surf_height_c;
-
- unsigned int full_swath_bytes_horz_wc_l;
- unsigned int full_swath_bytes_horz_wc_c;
- unsigned int full_swath_bytes_vert_wc_l;
- unsigned int full_swath_bytes_vert_wc_c;
-
- MAS_vp_horz_limit = SourcePixelFormat == dml2_rgbe_alpha ? 3840 : 6144;
- MAS_vp_vert_limit = SourcePixelFormat == dml2_rgbe_alpha ? 3840 : (BytePerPixelY == 8 ? 3072 : 6144);
- max_vp_horz_width = (unsigned int)(math_min2((double)MAS_vp_horz_limit, detile_buf_vp_horz_limit));
- max_vp_vert_height = (unsigned int)(math_min2((double)MAS_vp_vert_limit, detile_buf_vp_vert_limit));
- eff_surf_width_l = (SurfaceWidthLuma > max_vp_horz_width ? max_vp_horz_width : SurfaceWidthLuma);
- eff_surf_width_c = eff_surf_width_l / (1 + yuv420);
- eff_surf_height_l = (SurfaceHeightLuma > max_vp_vert_height ? max_vp_vert_height : SurfaceHeightLuma);
- eff_surf_height_c = eff_surf_height_l / (1 + yuv420);
-
- full_swath_bytes_horz_wc_l = eff_surf_width_l * RequestHeight256ByteLuma * BytePerPixelY;
- full_swath_bytes_vert_wc_l = eff_surf_height_l * 256 / RequestHeight256ByteLuma;
- if (BytePerPixelC > 0) {
- full_swath_bytes_horz_wc_c = eff_surf_width_c * RequestHeight256ByteChroma * BytePerPixelC;
- full_swath_bytes_vert_wc_c = eff_surf_height_c * 256 / RequestHeight256ByteChroma;
- } else {
- full_swath_bytes_horz_wc_c = 0;
- full_swath_bytes_vert_wc_c = 0;
- }
-
- if (SourcePixelFormat == dml2_420_10) {
- full_swath_bytes_horz_wc_l = (unsigned int)(math_ceil2((double)full_swath_bytes_horz_wc_l * 2.0 / 3.0, 256.0));
- full_swath_bytes_horz_wc_c = (unsigned int)(math_ceil2((double)full_swath_bytes_horz_wc_c * 2.0 / 3.0, 256.0));
- full_swath_bytes_vert_wc_l = (unsigned int)(math_ceil2((double)full_swath_bytes_vert_wc_l * 2.0 / 3.0, 256.0));
- full_swath_bytes_vert_wc_c = (unsigned int)(math_ceil2((double)full_swath_bytes_vert_wc_c * 2.0 / 3.0, 256.0));
- }
-
- unsigned int req128_horz_wc_l;
- unsigned int req128_horz_wc_c;
- unsigned int req128_vert_wc_l;
- unsigned int req128_vert_wc_c;
-
- if (2 * full_swath_bytes_horz_wc_l + 2 * full_swath_bytes_horz_wc_c <= DETBufferSizeForDCC) {
- req128_horz_wc_l = 0;
- req128_horz_wc_c = 0;
- } else if (full_swath_bytes_horz_wc_l < 1.5 * full_swath_bytes_horz_wc_c && 2 * full_swath_bytes_horz_wc_l + full_swath_bytes_horz_wc_c <= DETBufferSizeForDCC) {
- req128_horz_wc_l = 0;
- req128_horz_wc_c = 1;
- } else if (full_swath_bytes_horz_wc_l >= 1.5 * full_swath_bytes_horz_wc_c && full_swath_bytes_horz_wc_l + 2 * full_swath_bytes_horz_wc_c <= DETBufferSizeForDCC) {
- req128_horz_wc_l = 1;
- req128_horz_wc_c = 0;
- } else {
- req128_horz_wc_l = 1;
- req128_horz_wc_c = 1;
- }
-
- if (2 * full_swath_bytes_vert_wc_l + 2 * full_swath_bytes_vert_wc_c <= DETBufferSizeForDCC) {
- req128_vert_wc_l = 0;
- req128_vert_wc_c = 0;
- } else if (full_swath_bytes_vert_wc_l < 1.5 * full_swath_bytes_vert_wc_c && 2 * full_swath_bytes_vert_wc_l + full_swath_bytes_vert_wc_c <= DETBufferSizeForDCC) {
- req128_vert_wc_l = 0;
- req128_vert_wc_c = 1;
- } else if (full_swath_bytes_vert_wc_l >= 1.5 * full_swath_bytes_vert_wc_c && full_swath_bytes_vert_wc_l + 2 * full_swath_bytes_vert_wc_c <= DETBufferSizeForDCC) {
- req128_vert_wc_l = 1;
- req128_vert_wc_c = 0;
- } else {
- req128_vert_wc_l = 1;
- req128_vert_wc_c = 1;
- }
-
- unsigned int segment_order_horz_contiguous_luma;
- unsigned int segment_order_horz_contiguous_chroma;
- unsigned int segment_order_vert_contiguous_luma;
- unsigned int segment_order_vert_contiguous_chroma;
-
- if (BytePerPixelY == 2) {
- segment_order_horz_contiguous_luma = 0;
- segment_order_vert_contiguous_luma = 1;
- } else {
- segment_order_horz_contiguous_luma = 1;
- segment_order_vert_contiguous_luma = 0;
- }
-
- if (BytePerPixelC == 2) {
- segment_order_horz_contiguous_chroma = 0;
- segment_order_vert_contiguous_chroma = 1;
- } else {
- segment_order_horz_contiguous_chroma = 1;
- segment_order_vert_contiguous_chroma = 0;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DCCEnabled = %u\n", __func__, DCCEnabled);
- dml2_printf("DML::%s: nomDETInKByte = %u\n", __func__, nomDETInKByte);
- dml2_printf("DML::%s: DETBufferSizeForDCC = %u\n", __func__, DETBufferSizeForDCC);
- dml2_printf("DML::%s: req128_horz_wc_l = %u\n", __func__, req128_horz_wc_l);
- dml2_printf("DML::%s: req128_horz_wc_c = %u\n", __func__, req128_horz_wc_c);
- dml2_printf("DML::%s: full_swath_bytes_horz_wc_l = %u\n", __func__, full_swath_bytes_horz_wc_l);
- dml2_printf("DML::%s: full_swath_bytes_vert_wc_c = %u\n", __func__, full_swath_bytes_vert_wc_c);
- dml2_printf("DML::%s: segment_order_horz_contiguous_luma = %u\n", __func__, segment_order_horz_contiguous_luma);
- dml2_printf("DML::%s: segment_order_horz_contiguous_chroma = %u\n", __func__, segment_order_horz_contiguous_chroma);
-#endif
- if (DCCProgrammingAssumesScanDirectionUnknown == true) {
- if (req128_horz_wc_l == 0 && req128_vert_wc_l == 0) {
- *RequestLuma = dml2_core_internal_request_type_256_bytes;
- } else if ((req128_horz_wc_l == 1 && segment_order_horz_contiguous_luma == 0) || (req128_vert_wc_l == 1 && segment_order_vert_contiguous_luma == 0)) {
- *RequestLuma = dml2_core_internal_request_type_128_bytes_non_contiguous;
- } else {
- *RequestLuma = dml2_core_internal_request_type_128_bytes_contiguous;
- }
- if (req128_horz_wc_c == 0 && req128_vert_wc_c == 0) {
- *RequestChroma = dml2_core_internal_request_type_256_bytes;
- } else if ((req128_horz_wc_c == 1 && segment_order_horz_contiguous_chroma == 0) || (req128_vert_wc_c == 1 && segment_order_vert_contiguous_chroma == 0)) {
- *RequestChroma = dml2_core_internal_request_type_128_bytes_non_contiguous;
- } else {
- *RequestChroma = dml2_core_internal_request_type_128_bytes_contiguous;
- }
- } else if (!dml_is_vertical_rotation(RotationAngle)) {
- if (req128_horz_wc_l == 0) {
- *RequestLuma = dml2_core_internal_request_type_256_bytes;
- } else if (segment_order_horz_contiguous_luma == 0) {
- *RequestLuma = dml2_core_internal_request_type_128_bytes_non_contiguous;
- } else {
- *RequestLuma = dml2_core_internal_request_type_128_bytes_contiguous;
- }
- if (req128_horz_wc_c == 0) {
- *RequestChroma = dml2_core_internal_request_type_256_bytes;
- } else if (segment_order_horz_contiguous_chroma == 0) {
- *RequestChroma = dml2_core_internal_request_type_128_bytes_non_contiguous;
- } else {
- *RequestChroma = dml2_core_internal_request_type_128_bytes_contiguous;
- }
- } else {
- if (req128_vert_wc_l == 0) {
- *RequestLuma = dml2_core_internal_request_type_256_bytes;
- } else if (segment_order_vert_contiguous_luma == 0) {
- *RequestLuma = dml2_core_internal_request_type_128_bytes_non_contiguous;
- } else {
- *RequestLuma = dml2_core_internal_request_type_128_bytes_contiguous;
- }
- if (req128_vert_wc_c == 0) {
- *RequestChroma = dml2_core_internal_request_type_256_bytes;
- } else if (segment_order_vert_contiguous_chroma == 0) {
- *RequestChroma = dml2_core_internal_request_type_128_bytes_non_contiguous;
- } else {
- *RequestChroma = dml2_core_internal_request_type_128_bytes_contiguous;
- }
- }
-
- if (*RequestLuma == dml2_core_internal_request_type_256_bytes) {
- *MaxUncompressedBlockLuma = 256;
- *MaxCompressedBlockLuma = 256;
- *IndependentBlockLuma = 0;
- } else if (*RequestLuma == dml2_core_internal_request_type_128_bytes_contiguous) {
- *MaxUncompressedBlockLuma = 256;
- *MaxCompressedBlockLuma = 128;
- *IndependentBlockLuma = 128;
- } else {
- *MaxUncompressedBlockLuma = 256;
- *MaxCompressedBlockLuma = 64;
- *IndependentBlockLuma = 64;
- }
-
- if (*RequestChroma == dml2_core_internal_request_type_256_bytes) {
- *MaxUncompressedBlockChroma = 256;
- *MaxCompressedBlockChroma = 256;
- *IndependentBlockChroma = 0;
- } else if (*RequestChroma == dml2_core_internal_request_type_128_bytes_contiguous) {
- *MaxUncompressedBlockChroma = 256;
- *MaxCompressedBlockChroma = 128;
- *IndependentBlockChroma = 128;
- } else {
- *MaxUncompressedBlockChroma = 256;
- *MaxCompressedBlockChroma = 64;
- *IndependentBlockChroma = 64;
- }
-
- if (DCCEnabled != true || BytePerPixelC == 0) {
- *MaxUncompressedBlockChroma = 0;
- *MaxCompressedBlockChroma = 0;
- *IndependentBlockChroma = 0;
- }
-
- if (DCCEnabled != true) {
- *MaxUncompressedBlockLuma = 0;
- *MaxCompressedBlockLuma = 0;
- *IndependentBlockLuma = 0;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: MaxUncompressedBlockLuma = %u\n", __func__, *MaxUncompressedBlockLuma);
- dml2_printf("DML::%s: MaxCompressedBlockLuma = %u\n", __func__, *MaxCompressedBlockLuma);
- dml2_printf("DML::%s: IndependentBlockLuma = %u\n", __func__, *IndependentBlockLuma);
- dml2_printf("DML::%s: MaxUncompressedBlockChroma = %u\n", __func__, *MaxUncompressedBlockChroma);
- dml2_printf("DML::%s: MaxCompressedBlockChroma = %u\n", __func__, *MaxCompressedBlockChroma);
- dml2_printf("DML::%s: IndependentBlockChroma = %u\n", __func__, *IndependentBlockChroma);
-#endif
-
-}
-
-static void calculate_mcache_row_bytes(
- struct dml2_core_internal_scratch *scratch,
- struct dml2_core_calcs_calculate_mcache_row_bytes_params *p)
-{
- unsigned int vmpg_bytes = 0;
- unsigned int blk_bytes = 0;
- float meta_per_mvmpg_per_channel = 0;
- unsigned int est_blk_per_vmpg = 2;
- unsigned int mvmpg_per_row_ub = 0;
- unsigned int full_vp_width_mvmpg_aligned = 0;
- unsigned int full_vp_height_mvmpg_aligned = 0;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: num_chans = %u\n", __func__, p->num_chans);
- dml2_printf("DML::%s: mem_word_bytes = %u\n", __func__, p->mem_word_bytes);
- dml2_printf("DML::%s: mcache_line_size_bytes = %u\n", __func__, p->mcache_line_size_bytes);
- dml2_printf("DML::%s: mcache_size_bytes = %u\n", __func__, p->mcache_size_bytes);
- dml2_printf("DML::%s: gpuvm_enable = %u\n", __func__, p->gpuvm_enable);
- dml2_printf("DML::%s: gpuvm_page_size_kbytes = %u\n", __func__, p->gpuvm_page_size_kbytes);
- dml2_printf("DML::%s: vp_stationary = %u\n", __func__, p->vp_stationary);
- dml2_printf("DML::%s: tiling_mode = %u\n", __func__, p->tiling_mode);
- dml2_printf("DML::%s: vp_start_x = %u\n", __func__, p->vp_start_x);
- dml2_printf("DML::%s: vp_start_y = %u\n", __func__, p->vp_start_y);
- dml2_printf("DML::%s: full_vp_width = %u\n", __func__, p->full_vp_width);
- dml2_printf("DML::%s: full_vp_height = %u\n", __func__, p->full_vp_height);
- dml2_printf("DML::%s: blk_width = %u\n", __func__, p->blk_width);
- dml2_printf("DML::%s: blk_height = %u\n", __func__, p->blk_height);
- dml2_printf("DML::%s: vmpg_width = %u\n", __func__, p->vmpg_width);
- dml2_printf("DML::%s: vmpg_height = %u\n", __func__, p->vmpg_height);
- dml2_printf("DML::%s: full_swath_bytes = %u\n", __func__, p->full_swath_bytes);
-#endif
- DML2_ASSERT(p->mcache_line_size_bytes != 0);
- DML2_ASSERT(p->mcache_size_bytes != 0);
-
- *p->mvmpg_width = 0;
- *p->mvmpg_height = 0;
-
- if (p->full_vp_height == 0 && p->full_vp_width == 0) {
- *p->num_mcaches = 0;
- *p->mcache_row_bytes = 0;
- } else {
- blk_bytes = dml_get_tile_block_size_bytes(p->tiling_mode);
-
- // if gpuvm is not enable, the alignment boundary should be in terms of tiling block size
- vmpg_bytes = p->gpuvm_page_size_kbytes * 1024;
-
- //With vmpg_bytes >= tile blk_bytes, the meta_row_width alignment equations are relative to the vmpg_width/height.
- // But for 4KB page with 64KB tile block, we need the meta for all pages in the tile block.
- // Therefore, the alignment is relative to the blk_width/height. The factor of 16 vmpg per 64KB tile block is applied at the end.
- *p->mvmpg_width = p->blk_width;
- *p->mvmpg_height = p->blk_height;
- if (p->gpuvm_enable) {
- if (vmpg_bytes >= blk_bytes) {
- *p->mvmpg_width = p->vmpg_width;
- *p->mvmpg_height = p->vmpg_height;
- } else if (!((blk_bytes == 65536) && (vmpg_bytes == 4096))) {
- dml2_printf("ERROR: DML::%s: Tiling size and vm page size combination not supported\n", __func__);
- DML2_ASSERT(0);
- }
- }
-
- //For plane0 & 1, first calculate full_vp_width/height_l/c aligned to vmpg_width/height_l/c
- full_vp_width_mvmpg_aligned = (unsigned int)(math_floor2((p->vp_start_x + p->full_vp_width) + *p->mvmpg_width - 1, *p->mvmpg_width) - math_floor2(p->vp_start_x, *p->mvmpg_width));
- full_vp_height_mvmpg_aligned = (unsigned int)(math_floor2((p->vp_start_y + p->full_vp_height) + *p->mvmpg_height - 1, *p->mvmpg_height) - math_floor2(p->vp_start_y, *p->mvmpg_height));
-
- *p->full_vp_access_width_mvmpg_aligned = p->surf_vert ? full_vp_height_mvmpg_aligned : full_vp_width_mvmpg_aligned;
-
- //Use the equation for the exact alignment when possible. Note that the exact alignment cannot be used for horizontal access if vmpg_bytes > blk_bytes.
- if (!p->surf_vert) { //horizontal access
- if (p->vp_stationary == 1 && vmpg_bytes <= blk_bytes)
- *p->meta_row_width_ub = full_vp_width_mvmpg_aligned;
- else
- *p->meta_row_width_ub = (unsigned int)math_ceil2((double)p->full_vp_width - 1, *p->mvmpg_width) + *p->mvmpg_width;
- mvmpg_per_row_ub = *p->meta_row_width_ub / *p->mvmpg_width;
- } else { //vertical access
- if (p->vp_stationary == 1)
- *p->meta_row_width_ub = full_vp_height_mvmpg_aligned;
- else
- *p->meta_row_width_ub = (unsigned int)math_ceil2((double)p->full_vp_height - 1, *p->mvmpg_height) + *p->mvmpg_height;
- mvmpg_per_row_ub = *p->meta_row_width_ub / *p->mvmpg_height;
- }
-
- unsigned int meta_per_mvmpg_per_channel_ub = 0;
-
- if (p->gpuvm_enable) {
- meta_per_mvmpg_per_channel = (float)vmpg_bytes / (float)256 / p->num_chans;
-
- //but using the est_blk_per_vmpg between 2 and 4, to be not as pessimestic
- if (p->surf_vert && vmpg_bytes > blk_bytes) {
- meta_per_mvmpg_per_channel = (float)est_blk_per_vmpg * blk_bytes / 256 / p->num_chans;
- }
-
- *p->dcc_dram_bw_nom_overhead_factor = 1 + math_max2(1.0 / 256.0, math_ceil2(meta_per_mvmpg_per_channel, p->mem_word_bytes) / (256 * meta_per_mvmpg_per_channel)); // dcc_dr_oh_nom
- } else {
- meta_per_mvmpg_per_channel = (float)blk_bytes / (float)256 / p->num_chans;
-
- if (!p->surf_vert)
- *p->dcc_dram_bw_nom_overhead_factor = 1 + 1.0 / 256.0;
- else
- *p->dcc_dram_bw_nom_overhead_factor = 1 + math_max2(1.0 / 256.0, math_ceil2(meta_per_mvmpg_per_channel, p->mem_word_bytes) / (256 * meta_per_mvmpg_per_channel));
- }
-
- meta_per_mvmpg_per_channel_ub = (unsigned int)math_ceil2((double)meta_per_mvmpg_per_channel, p->mcache_line_size_bytes);
-
- //but for 4KB vmpg with 64KB tile blk
- if (p->gpuvm_enable && (blk_bytes == 65536) && (vmpg_bytes == 4096))
- meta_per_mvmpg_per_channel_ub = 16 * meta_per_mvmpg_per_channel_ub;
-
- // If this mcache_row_bytes for the full viewport of the surface is less than or equal to mcache_bytes,
- // then one mcache can be used for this request stream. If not, it is useful to know the width of the viewport that can be supported in the mcache_bytes.
- if (p->gpuvm_enable || !p->surf_vert) {
- *p->mcache_row_bytes = mvmpg_per_row_ub * meta_per_mvmpg_per_channel_ub;
- } else { // horizontal and gpuvm disable
- *p->mcache_row_bytes = *p->meta_row_width_ub * p->blk_height * p->bytes_per_pixel / 256;
- *p->mcache_row_bytes = (unsigned int)math_ceil2((double)*p->mcache_row_bytes / p->num_chans, p->mcache_line_size_bytes);
- }
-
- *p->dcc_dram_bw_pref_overhead_factor = 1 + math_max2(1.0 / 256.0, *p->mcache_row_bytes / p->full_swath_bytes); // dcc_dr_oh_pref
- *p->num_mcaches = (unsigned int)math_ceil2((double)*p->mcache_row_bytes / p->mcache_size_bytes, 1);
-
- unsigned int mvmpg_per_mcache = p->mcache_size_bytes / meta_per_mvmpg_per_channel_ub;
- *p->mvmpg_per_mcache_lb = (unsigned int)math_floor2(mvmpg_per_mcache, 1);
-
- DML2_ASSERT(*p->num_mcaches > 0);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: gpuvm_enable = %u\n", __func__, p->gpuvm_enable);
- dml2_printf("DML::%s: vmpg_bytes = %u\n", __func__, vmpg_bytes);
- dml2_printf("DML::%s: blk_bytes = %u\n", __func__, blk_bytes);
- dml2_printf("DML::%s: meta_per_mvmpg_per_channel = %f\n", __func__, meta_per_mvmpg_per_channel);
- dml2_printf("DML::%s: mvmpg_per_row_ub = %u\n", __func__, mvmpg_per_row_ub);
- dml2_printf("DML::%s: meta_row_width_ub = %u\n", __func__, *p->meta_row_width_ub);
- dml2_printf("DML::%s: mvmpg_width = %u\n", __func__, *p->mvmpg_width);
- dml2_printf("DML::%s: mvmpg_height = %u\n", __func__, *p->mvmpg_height);
- dml2_printf("DML::%s: num_mcaches = %u\n", __func__, *p->num_mcaches);
- dml2_printf("DML::%s: dcc_dram_bw_nom_overhead_factor = %f\n", __func__, *p->dcc_dram_bw_nom_overhead_factor);
- dml2_printf("DML::%s: dcc_dram_bw_pref_overhead_factor = %f\n", __func__, *p->dcc_dram_bw_pref_overhead_factor);
-#endif
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: mcache_row_bytes = %u\n", __func__, *p->mcache_row_bytes);
- dml2_printf("DML::%s: num_mcaches = %u\n", __func__, *p->num_mcaches);
-#endif
-}
-
-static void calculate_mcache_setting(
- struct dml2_core_internal_scratch *scratch,
- struct dml2_core_calcs_calculate_mcache_setting_params *p)
-{
- unsigned int n;
-
- struct dml2_core_shared_calculate_mcache_setting_locals *l = &scratch->calculate_mcache_setting_locals;
- memset(l, 0, sizeof(struct dml2_core_shared_calculate_mcache_setting_locals));
-
- *p->num_mcaches_l = 0;
- *p->mcache_row_bytes_l = 0;
- *p->dcc_dram_bw_nom_overhead_factor_l = 1.0;
- *p->dcc_dram_bw_pref_overhead_factor_l = 1.0;
-
- *p->num_mcaches_c = 0;
- *p->mcache_row_bytes_c = 0;
- *p->dcc_dram_bw_nom_overhead_factor_c = 1.0;
- *p->dcc_dram_bw_pref_overhead_factor_c = 1.0;
-
- *p->mall_comb_mcache_l = 0;
- *p->mall_comb_mcache_c = 0;
- *p->lc_comb_mcache = 0;
-
- if (!p->dcc_enable)
- return;
-
- l->is_dual_plane = dml2_core_shared_is_420(p->source_format) || p->source_format == dml2_rgbe_alpha;
-
- l->l_p.num_chans = p->num_chans;
- l->l_p.mem_word_bytes = p->mem_word_bytes;
- l->l_p.mcache_size_bytes = p->mcache_size_bytes;
- l->l_p.mcache_line_size_bytes = p->mcache_line_size_bytes;
- l->l_p.gpuvm_enable = p->gpuvm_enable;
- l->l_p.gpuvm_page_size_kbytes = p->gpuvm_page_size_kbytes;
- l->l_p.surf_vert = p->surf_vert;
- l->l_p.vp_stationary = p->vp_stationary;
- l->l_p.tiling_mode = p->tiling_mode;
- l->l_p.vp_start_x = p->vp_start_x_l;
- l->l_p.vp_start_y = p->vp_start_y_l;
- l->l_p.full_vp_width = p->full_vp_width_l;
- l->l_p.full_vp_height = p->full_vp_height_l;
- l->l_p.blk_width = p->blk_width_l;
- l->l_p.blk_height = p->blk_height_l;
- l->l_p.vmpg_width = p->vmpg_width_l;
- l->l_p.vmpg_height = p->vmpg_height_l;
- l->l_p.full_swath_bytes = p->full_swath_bytes_l;
- l->l_p.bytes_per_pixel = p->bytes_per_pixel_l;
-
- // output
- l->l_p.num_mcaches = p->num_mcaches_l;
- l->l_p.mcache_row_bytes = p->mcache_row_bytes_l;
- l->l_p.dcc_dram_bw_nom_overhead_factor = p->dcc_dram_bw_nom_overhead_factor_l;
- l->l_p.dcc_dram_bw_pref_overhead_factor = p->dcc_dram_bw_pref_overhead_factor_l;
- l->l_p.mvmpg_width = &l->mvmpg_width_l;
- l->l_p.mvmpg_height = &l->mvmpg_height_l;
- l->l_p.full_vp_access_width_mvmpg_aligned = &l->full_vp_access_width_mvmpg_aligned_l;
- l->l_p.meta_row_width_ub = &l->meta_row_width_l;
- l->l_p.mvmpg_per_mcache_lb = &l->mvmpg_per_mcache_lb_l;
-
- calculate_mcache_row_bytes(scratch, &l->l_p);
- dml2_assert(*p->num_mcaches_l > 0);
-
- if (l->is_dual_plane) {
- l->c_p.num_chans = p->num_chans;
- l->c_p.mem_word_bytes = p->mem_word_bytes;
- l->c_p.mcache_size_bytes = p->mcache_size_bytes;
- l->c_p.mcache_line_size_bytes = p->mcache_line_size_bytes;
- l->c_p.gpuvm_enable = p->gpuvm_enable;
- l->c_p.gpuvm_page_size_kbytes = p->gpuvm_page_size_kbytes;
- l->c_p.surf_vert = p->surf_vert;
- l->c_p.vp_stationary = p->vp_stationary;
- l->c_p.tiling_mode = p->tiling_mode;
- l->c_p.vp_start_x = p->vp_start_x_c;
- l->c_p.vp_start_y = p->vp_start_y_c;
- l->c_p.full_vp_width = p->full_vp_width_c;
- l->c_p.full_vp_height = p->full_vp_height_c;
- l->c_p.blk_width = p->blk_width_c;
- l->c_p.blk_height = p->blk_height_c;
- l->c_p.vmpg_width = p->vmpg_width_c;
- l->c_p.vmpg_height = p->vmpg_height_c;
- l->c_p.full_swath_bytes = p->full_swath_bytes_c;
- l->c_p.bytes_per_pixel = p->bytes_per_pixel_c;
-
- // output
- l->c_p.num_mcaches = p->num_mcaches_c;
- l->c_p.mcache_row_bytes = p->mcache_row_bytes_c;
- l->c_p.dcc_dram_bw_nom_overhead_factor = p->dcc_dram_bw_nom_overhead_factor_c;
- l->c_p.dcc_dram_bw_pref_overhead_factor = p->dcc_dram_bw_pref_overhead_factor_c;
- l->c_p.mvmpg_width = &l->mvmpg_width_c;
- l->c_p.mvmpg_height = &l->mvmpg_height_c;
- l->c_p.full_vp_access_width_mvmpg_aligned = &l->full_vp_access_width_mvmpg_aligned_c;
- l->c_p.meta_row_width_ub = &l->meta_row_width_c;
- l->c_p.mvmpg_per_mcache_lb = &l->mvmpg_per_mcache_lb_c;
-
- calculate_mcache_row_bytes(scratch, &l->c_p);
- dml2_assert(*p->num_mcaches_c > 0);
- }
-
- // Sharing for iMALL access
- l->mcache_remainder_l = *p->mcache_row_bytes_l % p->mcache_size_bytes;
- l->mcache_remainder_c = *p->mcache_row_bytes_c % p->mcache_size_bytes;
- l->mvmpg_access_width_l = p->surf_vert ? l->mvmpg_height_l : l->mvmpg_width_l;
- l->mvmpg_access_width_c = p->surf_vert ? l->mvmpg_height_c : l->mvmpg_width_c;
-
- if (p->imall_enable) {
- *p->mall_comb_mcache_l = (2 * l->mcache_remainder_l <= p->mcache_size_bytes);
-
- if (l->is_dual_plane)
- *p->mall_comb_mcache_c = (2 * l->mcache_remainder_c <= p->mcache_size_bytes);
- }
-
- if (!p->surf_vert) // horizonatal access
- l->luma_time_factor = (double)l->mvmpg_height_c / l->mvmpg_height_l * 2;
- else // vertical access
- l->luma_time_factor = (double)l->mvmpg_width_c / l->mvmpg_width_l * 2;
-
- // The algorithm starts with computing a non-integer, avg_mcache_element_size_l/c:
- l->avg_mcache_element_size_l = l->meta_row_width_l / *p->num_mcaches_l;
- if (l->is_dual_plane) {
- l->avg_mcache_element_size_c = l->meta_row_width_c / *p->num_mcaches_c;
-
- if (!p->imall_enable || (*p->mall_comb_mcache_l == *p->mall_comb_mcache_c)) {
- l->lc_comb_last_mcache_size = (unsigned int)((l->mcache_remainder_l * (*p->mall_comb_mcache_l ? 2 : 1) * l->luma_time_factor) +
- (l->mcache_remainder_c * (*p->mall_comb_mcache_c ? 2 : 1)));
- }
- *p->lc_comb_mcache = (l->lc_comb_last_mcache_size <= p->mcache_size_bytes) && (*p->mall_comb_mcache_l == *p->mall_comb_mcache_c);
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: imall_enable = %u\n", __func__, p->imall_enable);
- dml2_printf("DML::%s: is_dual_plane = %u\n", __func__, l->is_dual_plane);
- dml2_printf("DML::%s: surf_vert = %u\n", __func__, p->surf_vert);
- dml2_printf("DML::%s: mvmpg_width_l = %u\n", __func__, l->mvmpg_width_l);
- dml2_printf("DML::%s: mvmpg_height_l = %u\n", __func__, l->mvmpg_height_l);
- dml2_printf("DML::%s: mcache_remainder_l = %f\n", __func__, l->mcache_remainder_l);
- dml2_printf("DML::%s: num_mcaches_l = %u\n", __func__, *p->num_mcaches_l);
- dml2_printf("DML::%s: avg_mcache_element_size_l = %u\n", __func__, l->avg_mcache_element_size_l);
- dml2_printf("DML::%s: mvmpg_access_width_l = %u\n", __func__, l->mvmpg_access_width_l);
- dml2_printf("DML::%s: mall_comb_mcache_l = %u\n", __func__, *p->mall_comb_mcache_l);
-
- if (l->is_dual_plane) {
- dml2_printf("DML::%s: mvmpg_width_c = %u\n", __func__, l->mvmpg_width_c);
- dml2_printf("DML::%s: mvmpg_height_c = %u\n", __func__, l->mvmpg_height_c);
- dml2_printf("DML::%s: mcache_remainder_c = %f\n", __func__, l->mcache_remainder_c);
- dml2_printf("DML::%s: luma_time_factor = %f\n", __func__, l->luma_time_factor);
- dml2_printf("DML::%s: num_mcaches_c = %u\n", __func__, *p->num_mcaches_c);
- dml2_printf("DML::%s: avg_mcache_element_size_c = %u\n", __func__, l->avg_mcache_element_size_c);
- dml2_printf("DML::%s: mvmpg_access_width_c = %u\n", __func__, l->mvmpg_access_width_c);
- dml2_printf("DML::%s: mall_comb_mcache_c = %u\n", __func__, *p->mall_comb_mcache_c);
- dml2_printf("DML::%s: lc_comb_last_mcache_size = %u\n", __func__, l->lc_comb_last_mcache_size);
- dml2_printf("DML::%s: lc_comb_mcache = %u\n", __func__, *p->lc_comb_mcache);
- }
-#endif
- // calculate split_coordinate
- l->full_vp_access_width_l = p->surf_vert ? p->full_vp_height_l : p->full_vp_width_l;
- l->full_vp_access_width_c = p->surf_vert ? p->full_vp_height_c : p->full_vp_width_c;
-
- for (n = 0; n < *p->num_mcaches_l - 1; n++) {
- p->mcache_offsets_l[n] = (unsigned int)(math_floor2((n + 1) * l->avg_mcache_element_size_l / l->mvmpg_access_width_l, 1)) * l->mvmpg_access_width_l;
- }
- p->mcache_offsets_l[*p->num_mcaches_l - 1] = l->full_vp_access_width_l;
-
- if (l->is_dual_plane) {
- for (n = 0; n < *p->num_mcaches_c - 1; n++) {
- p->mcache_offsets_c[n] = (unsigned int)(math_floor2((n + 1) * l->avg_mcache_element_size_c / l->mvmpg_access_width_c, 1)) * l->mvmpg_access_width_c;
- }
- p->mcache_offsets_c[*p->num_mcaches_c - 1] = l->full_vp_access_width_c;
- }
-#ifdef __DML_VBA_DEBUG__
- for (n = 0; n < *p->num_mcaches_l; n++)
- dml2_printf("DML::%s: mcache_offsets_l[%u] = %u\n", __func__, n, p->mcache_offsets_l[n]);
-
- if (l->is_dual_plane) {
- for (n = 0; n < *p->num_mcaches_c; n++)
- dml2_printf("DML::%s: mcache_offsets_c[%u] = %u\n", __func__, n, p->mcache_offsets_c[n]);
- }
-#endif
-
- // Luma/Chroma combine in the last mcache
- // In the case of Luma/Chroma combine-mCache (with lc_comb_mcache==1), all mCaches except the last segment are filled as much as possible, when stay aligned to mvmpg boundary
- if (*p->lc_comb_mcache && l->is_dual_plane) {
- for (n = 0; n < *p->num_mcaches_l - 1; n++)
- p->mcache_offsets_l[n] = (n + 1) * l->mvmpg_per_mcache_lb_l * l->mvmpg_access_width_l;
- p->mcache_offsets_l[*p->num_mcaches_l - 1] = l->full_vp_access_width_l;
-
- for (n = 0; n < *p->num_mcaches_c - 1; n++)
- p->mcache_offsets_c[n] = (n + 1) * l->mvmpg_per_mcache_lb_c * l->mvmpg_access_width_c;
- p->mcache_offsets_c[*p->num_mcaches_c - 1] = l->full_vp_access_width_c;
-
-#ifdef __DML_VBA_DEBUG__
- for (n = 0; n < *p->num_mcaches_l; n++)
- dml2_printf("DML::%s: mcache_offsets_l[%u] = %u\n", __func__, n, p->mcache_offsets_l[n]);
-
- for (n = 0; n < *p->num_mcaches_c; n++)
- dml2_printf("DML::%s: mcache_offsets_c[%u] = %u\n", __func__, n, p->mcache_offsets_c[n]);
-#endif
- }
-
- *p->mcache_shift_granularity_l = l->mvmpg_access_width_l;
- *p->mcache_shift_granularity_c = l->mvmpg_access_width_c;
-}
-
-static void calculate_mall_bw_overhead_factor(
- double mall_prefetch_sdp_overhead_factor[], //mall_sdp_oh_nom/pref
- double mall_prefetch_dram_overhead_factor[], //mall_dram_oh_nom/pref
-
- // input
- const struct dml2_display_cfg *display_cfg,
- unsigned int num_active_planes)
-{
- for (unsigned int k = 0; k < num_active_planes; ++k) {
- mall_prefetch_sdp_overhead_factor[k] = 1.0;
- mall_prefetch_dram_overhead_factor[k] = 1.0;
-
- // SDP - on the return side
- if (display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_imall) // always no data return
- mall_prefetch_sdp_overhead_factor[k] = 1.25;
- else if (display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_phantom_pipe_no_data_return)
- mall_prefetch_sdp_overhead_factor[k] = 0.25;
-
- // DRAM
- if (display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_imall)
- mall_prefetch_dram_overhead_factor[k] = 2.0;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, mall_prefetch_sdp_overhead_factor = %f\n", __func__, k, mall_prefetch_sdp_overhead_factor[k]);
- dml2_printf("DML::%s: k=%u, mall_prefetch_dram_overhead_factor = %f\n", __func__, k, mall_prefetch_dram_overhead_factor[k]);
-#endif
- }
-}
-
-static double dml_get_return_bandwidth_available(
- const struct dml2_soc_bb *soc,
- enum dml2_core_internal_soc_state_type state_type,
- enum dml2_core_internal_bw_type bw_type,
- bool is_avg_bw,
- bool is_hvm_en,
- bool is_hvm_only,
- double dcflk_mhz,
- double fclk_mhz,
- double dram_bw_mbps)
-{
- double return_bw_mbps = 0.;
- double ideal_sdp_bandwidth = (double)soc->return_bus_width_bytes * dcflk_mhz;
- double ideal_fabric_bandwidth = fclk_mhz * (double)soc->fabric_datapath_to_dcn_data_return_bytes;
- double ideal_dram_bandwidth = dram_bw_mbps; //dram_speed_mts * soc->clk_table.dram_config.channel_count * soc->clk_table.dram_config.channel_width_bytes;
-
- double derate_sdp_factor = 1;
- double derate_fabric_factor = 1;
- double derate_dram_factor = 1;
-
- if (is_avg_bw) {
- if (state_type == dml2_core_internal_soc_state_svp_prefetch) {
- derate_sdp_factor = soc->qos_parameters.derate_table.dcn_mall_prefetch_average.dcfclk_derate_percent / 100.0;
- derate_fabric_factor = soc->qos_parameters.derate_table.dcn_mall_prefetch_average.fclk_derate_percent / 100.0;
- derate_dram_factor = soc->qos_parameters.derate_table.dcn_mall_prefetch_average.dram_derate_percent_pixel / 100.0;
- } else { // just assume sys_active
- derate_sdp_factor = soc->qos_parameters.derate_table.system_active_average.dcfclk_derate_percent / 100.0;
- derate_fabric_factor = soc->qos_parameters.derate_table.system_active_average.fclk_derate_percent / 100.0;
- derate_dram_factor = soc->qos_parameters.derate_table.system_active_average.dram_derate_percent_pixel / 100.0;
- }
- } else { // urgent bw
- if (state_type == dml2_core_internal_soc_state_svp_prefetch) {
- derate_sdp_factor = soc->qos_parameters.derate_table.dcn_mall_prefetch_urgent.dcfclk_derate_percent / 100.0;
- derate_fabric_factor = soc->qos_parameters.derate_table.dcn_mall_prefetch_urgent.fclk_derate_percent / 100.0;
- derate_dram_factor = soc->qos_parameters.derate_table.dcn_mall_prefetch_urgent.dram_derate_percent_pixel / 100.0;
-
- if (is_hvm_en) {
- if (is_hvm_only)
- derate_dram_factor = soc->qos_parameters.derate_table.dcn_mall_prefetch_urgent.dram_derate_percent_vm / 100.0;
- else
- derate_dram_factor = soc->qos_parameters.derate_table.dcn_mall_prefetch_urgent.dram_derate_percent_pixel_and_vm / 100.0;
- } else {
- derate_dram_factor = soc->qos_parameters.derate_table.dcn_mall_prefetch_urgent.dram_derate_percent_pixel / 100.0;
- }
- } else { // just assume sys_active
- derate_sdp_factor = soc->qos_parameters.derate_table.system_active_urgent.dcfclk_derate_percent / 100.0;
- derate_fabric_factor = soc->qos_parameters.derate_table.system_active_urgent.fclk_derate_percent / 100.0;
-
- if (is_hvm_en) {
- if (is_hvm_only)
- derate_dram_factor = soc->qos_parameters.derate_table.system_active_urgent.dram_derate_percent_vm / 100.0;
- else
- derate_dram_factor = soc->qos_parameters.derate_table.system_active_urgent.dram_derate_percent_pixel_and_vm / 100.0;
- } else {
- derate_dram_factor = soc->qos_parameters.derate_table.system_active_urgent.dram_derate_percent_pixel / 100.0;
- }
- }
- }
-
- double derate_sdp_bandwidth = ideal_sdp_bandwidth * derate_sdp_factor;
- double derate_fabric_bandwidth = ideal_fabric_bandwidth * derate_fabric_factor;
- double derate_dram_bandwidth = ideal_dram_bandwidth * derate_dram_factor;
-
- if (bw_type == dml2_core_internal_bw_sdp)
- return_bw_mbps = math_min2(derate_sdp_bandwidth, derate_fabric_bandwidth);
- else // dml2_core_internal_bw_dram
- return_bw_mbps = derate_dram_bandwidth;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: is_avg_bw = %u\n", __func__, is_avg_bw);
- dml2_printf("DML::%s: is_hvm_en = %u\n", __func__, is_hvm_en);
- dml2_printf("DML::%s: is_hvm_only = %u\n", __func__, is_hvm_only);
- dml2_printf("DML::%s: state_type = %s\n", __func__, dml2_core_internal_soc_state_type_str(state_type));
- dml2_printf("DML::%s: bw_type = %s\n", __func__, dml2_core_internal_bw_type_str(bw_type));
- dml2_printf("DML::%s: dcflk_mhz = %f\n", __func__, dcflk_mhz);
- dml2_printf("DML::%s: fclk_mhz = %f\n", __func__, fclk_mhz);
- dml2_printf("DML::%s: ideal_sdp_bandwidth = %f\n", __func__, ideal_sdp_bandwidth);
- dml2_printf("DML::%s: ideal_fabric_bandwidth = %f\n", __func__, ideal_fabric_bandwidth);
- dml2_printf("DML::%s: ideal_dram_bandwidth = %f\n", __func__, ideal_dram_bandwidth);
- dml2_printf("DML::%s: derate_sdp_bandwidth = %f (derate %f)\n", __func__, derate_sdp_bandwidth, derate_sdp_factor);
- dml2_printf("DML::%s: derate_fabric_bandwidth = %f (derate %f)\n", __func__, derate_fabric_bandwidth, derate_fabric_factor);
- dml2_printf("DML::%s: derate_dram_bandwidth = %f (derate %f)\n", __func__, derate_dram_bandwidth, derate_dram_factor);
- dml2_printf("DML::%s: return_bw_mbps = %f\n", __func__, return_bw_mbps);
-#endif
- return return_bw_mbps;
-}
-
-static void calculate_bandwidth_available(
- double avg_bandwidth_available_min[dml2_core_internal_soc_state_max],
- double avg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_available_min[dml2_core_internal_soc_state_max], // min between SDP and DRAM
- double urg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_available_vm_only[dml2_core_internal_soc_state_max],
- double urg_bandwidth_available_pixel_and_vm[dml2_core_internal_soc_state_max],
-
- const struct dml2_soc_bb *soc,
- bool HostVMEnable,
- double dcfclk_mhz,
- double fclk_mhz,
- double dram_bw_mbps)
-{
- unsigned int n, m;
-
- dml2_printf("DML::%s: dcfclk_mhz = %f\n", __func__, dcfclk_mhz);
- dml2_printf("DML::%s: fclk_mhz = %f\n", __func__, fclk_mhz);
- dml2_printf("DML::%s: dram_bw_mbps = %f\n", __func__, dram_bw_mbps);
-
- // Calculate all the bandwidth availabe
- for (m = 0; m < dml2_core_internal_soc_state_max; m++) {
- for (n = 0; n < dml2_core_internal_bw_max; n++) {
- avg_bandwidth_available[m][n] = dml_get_return_bandwidth_available(soc,
- m, // soc_state
- n, // bw_type
- 1, // avg_bw
- HostVMEnable,
- 0, // hvm_only
- dcfclk_mhz,
- fclk_mhz,
- dram_bw_mbps);
-
- urg_bandwidth_available[m][n] = dml_get_return_bandwidth_available(soc, m, n, 0, HostVMEnable, 0, dcfclk_mhz, fclk_mhz, dram_bw_mbps);
-
-
- dml2_printf("DML::%s: avg_bandwidth_available[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), avg_bandwidth_available[m][n]);
- dml2_printf("DML::%s: urg_bandwidth_available[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), urg_bandwidth_available[m][n]);
-
- // urg_bandwidth_available_vm_only is indexed by soc_state
- if (n == dml2_core_internal_bw_dram) {
- urg_bandwidth_available_vm_only[m] = dml_get_return_bandwidth_available(soc, m, n, 0, HostVMEnable, 1, dcfclk_mhz, fclk_mhz, dram_bw_mbps);
- urg_bandwidth_available_pixel_and_vm[m] = dml_get_return_bandwidth_available(soc, m, n, 0, HostVMEnable, 0, dcfclk_mhz, fclk_mhz, dram_bw_mbps);
- }
- }
-
- avg_bandwidth_available_min[m] = math_min2(avg_bandwidth_available[m][dml2_core_internal_bw_dram], avg_bandwidth_available[m][dml2_core_internal_bw_sdp]);
- urg_bandwidth_available_min[m] = math_min2(urg_bandwidth_available[m][dml2_core_internal_bw_dram], urg_bandwidth_available[m][dml2_core_internal_bw_sdp]);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: avg_bandwidth_available_min[%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), avg_bandwidth_available_min[m]);
- dml2_printf("DML::%s: urg_bandwidth_available_min[%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), urg_bandwidth_available_min[m]);
- dml2_printf("DML::%s: urg_bandwidth_available_vm_only[%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(m), urg_bandwidth_available_vm_only[n]);
-#endif
- }
-}
-
-static void calculate_avg_bandwidth_required(
- double avg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
-
- // input
- const struct dml2_display_cfg *display_cfg,
- unsigned int num_active_planes,
- double ReadBandwidthLuma[],
- double ReadBandwidthChroma[],
- double cursor_bw[],
- double dcc_dram_bw_nom_overhead_factor_p0[],
- double dcc_dram_bw_nom_overhead_factor_p1[],
- double mall_prefetch_dram_overhead_factor[],
- double mall_prefetch_sdp_overhead_factor[])
-{
- unsigned int n, m, k;
-
- // Average BW support check
- for (m = 0; m < dml2_core_internal_soc_state_max; m++) {
- for (n = 0; n < dml2_core_internal_bw_max; n++) { // sdp, dram
- avg_bandwidth_required[m][n] = 0;
- }
- }
-
- // SysActive and SVP Prefetch AVG bandwidth Check
- for (k = 0; k < num_active_planes; ++k) {
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: plane %0d\n", __func__, k);
- dml2_printf("DML::%s: ReadBandwidthLuma=%f\n", __func__, ReadBandwidthLuma[k]);
- dml2_printf("DML::%s: ReadBandwidthChroma=%f\n", __func__, ReadBandwidthChroma[k]);
- dml2_printf("DML::%s: dcc_dram_bw_nom_overhead_factor_p0=%f\n", __func__, dcc_dram_bw_nom_overhead_factor_p0[k]);
- dml2_printf("DML::%s: dcc_dram_bw_nom_overhead_factor_p1=%f\n", __func__, dcc_dram_bw_nom_overhead_factor_p1[k]);
- dml2_printf("DML::%s: mall_prefetch_dram_overhead_factor=%f\n", __func__, mall_prefetch_dram_overhead_factor[k]);
- dml2_printf("DML::%s: mall_prefetch_sdp_overhead_factor=%f\n", __func__, mall_prefetch_sdp_overhead_factor[k]);
-#endif
-
- double sdp_overhead_factor = mall_prefetch_sdp_overhead_factor[k];
- double dram_overhead_factor_p0 = dcc_dram_bw_nom_overhead_factor_p0[k] * mall_prefetch_dram_overhead_factor[k];
- double dram_overhead_factor_p1 = dcc_dram_bw_nom_overhead_factor_p1[k] * mall_prefetch_dram_overhead_factor[k];
-
- // FIXME_DCN4, was missing cursor_bw in here, but do I actually need that and tdlut bw for average bandwidth calculation?
- // active avg bw not include phantom, but svp_prefetch avg bw should include phantom pipes
- if (!dml_is_phantom_pipe(&display_cfg->plane_descriptors[k])) {
- avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp] += sdp_overhead_factor * (ReadBandwidthLuma[k] + ReadBandwidthChroma[k]) + cursor_bw[k];
- avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram] += dram_overhead_factor_p0 * ReadBandwidthLuma[k] + dram_overhead_factor_p1 * ReadBandwidthChroma[k] + cursor_bw[k];
- }
- avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp] += sdp_overhead_factor * (ReadBandwidthLuma[k] + ReadBandwidthChroma[k]) + cursor_bw[k];
- avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram] += dram_overhead_factor_p0 * ReadBandwidthLuma[k] + dram_overhead_factor_p1 * ReadBandwidthChroma[k] + cursor_bw[k];
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_sys_active), dml2_core_internal_bw_type_str(dml2_core_internal_bw_sdp), avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp]);
- dml2_printf("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_sys_active), dml2_core_internal_bw_type_str(dml2_core_internal_bw_dram), avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram]);
- dml2_printf("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_svp_prefetch), dml2_core_internal_bw_type_str(dml2_core_internal_bw_sdp), avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp]);
- dml2_printf("DML::%s: avg_bandwidth_required[%s][%s]=%f\n", __func__, dml2_core_internal_soc_state_type_str(dml2_core_internal_soc_state_svp_prefetch), dml2_core_internal_bw_type_str(dml2_core_internal_bw_dram), avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram]);
-#endif
- }
-}
-
-static void CalculateVMRowAndSwath(struct dml2_core_internal_scratch *scratch,
- struct dml2_core_calcs_CalculateVMRowAndSwath_params *p)
-{
- struct dml2_core_calcs_CalculateVMRowAndSwath_locals *s = &scratch->CalculateVMRowAndSwath_locals;
-
- s->HostVMDynamicLevels = CalculateHostVMDynamicLevels(p->display_cfg->gpuvm_enable, p->display_cfg->hostvm_enable, p->HostVMMinPageSize, p->display_cfg->hostvm_max_non_cached_page_table_levels);
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (p->display_cfg->hostvm_enable == true) {
- p->vm_group_bytes[k] = 512;
- p->dpte_group_bytes[k] = 512;
- } else if (p->display_cfg->gpuvm_enable == true) {
- p->vm_group_bytes[k] = 2048;
- if (p->display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes >= 64 && dml_is_vertical_rotation(p->myPipe[k].RotationAngle)) {
- p->dpte_group_bytes[k] = 512;
- } else {
- p->dpte_group_bytes[k] = 2048;
- }
- } else {
- p->vm_group_bytes[k] = 0;
- p->dpte_group_bytes[k] = 0;
- }
-
- if (dml2_core_shared_is_420(p->myPipe[k].SourcePixelFormat) || p->myPipe[k].SourcePixelFormat == dml2_rgbe_alpha) {
- if ((p->myPipe[k].SourcePixelFormat == dml2_420_10 || p->myPipe[k].SourcePixelFormat == dml2_420_12) && !dml_is_vertical_rotation(p->myPipe[k].RotationAngle)) {
- s->PTEBufferSizeInRequestsForLuma[k] = (p->PTEBufferSizeInRequestsLuma + p->PTEBufferSizeInRequestsChroma) / 2;
- s->PTEBufferSizeInRequestsForChroma[k] = s->PTEBufferSizeInRequestsForLuma[k];
- } else {
- s->PTEBufferSizeInRequestsForLuma[k] = p->PTEBufferSizeInRequestsLuma;
- s->PTEBufferSizeInRequestsForChroma[k] = p->PTEBufferSizeInRequestsChroma;
- }
-
- scratch->calculate_vm_and_row_bytes_params.ViewportStationary = p->myPipe[k].ViewportStationary;
- scratch->calculate_vm_and_row_bytes_params.DCCEnable = p->myPipe[k].DCCEnable;
- scratch->calculate_vm_and_row_bytes_params.NumberOfDPPs = p->myPipe[k].DPPPerSurface;
- scratch->calculate_vm_and_row_bytes_params.BlockHeight256Bytes = p->myPipe[k].BlockHeight256BytesC;
- scratch->calculate_vm_and_row_bytes_params.BlockWidth256Bytes = p->myPipe[k].BlockWidth256BytesC;
- scratch->calculate_vm_and_row_bytes_params.SourcePixelFormat = p->myPipe[k].SourcePixelFormat;
- scratch->calculate_vm_and_row_bytes_params.SurfaceTiling = p->myPipe[k].SurfaceTiling;
- scratch->calculate_vm_and_row_bytes_params.BytePerPixel = p->myPipe[k].BytePerPixelC;
- scratch->calculate_vm_and_row_bytes_params.RotationAngle = p->myPipe[k].RotationAngle;
- scratch->calculate_vm_and_row_bytes_params.SwathWidth = p->SwathWidthC[k];
- scratch->calculate_vm_and_row_bytes_params.ViewportHeight = p->myPipe[k].ViewportHeightC;
- scratch->calculate_vm_and_row_bytes_params.ViewportXStart = p->myPipe[k].ViewportXStartC;
- scratch->calculate_vm_and_row_bytes_params.ViewportYStart = p->myPipe[k].ViewportYStartC;
- scratch->calculate_vm_and_row_bytes_params.GPUVMEnable = p->display_cfg->gpuvm_enable;
- scratch->calculate_vm_and_row_bytes_params.GPUVMMaxPageTableLevels = p->display_cfg->gpuvm_max_page_table_levels;
- scratch->calculate_vm_and_row_bytes_params.GPUVMMinPageSizeKBytes = p->display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes;
- scratch->calculate_vm_and_row_bytes_params.PTEBufferSizeInRequests = s->PTEBufferSizeInRequestsForChroma[k];
- scratch->calculate_vm_and_row_bytes_params.Pitch = p->myPipe[k].PitchC;
- scratch->calculate_vm_and_row_bytes_params.MacroTileWidth = p->myPipe[k].BlockWidthC;
- scratch->calculate_vm_and_row_bytes_params.MacroTileHeight = p->myPipe[k].BlockHeightC;
- scratch->calculate_vm_and_row_bytes_params.is_phantom = dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k]);
- scratch->calculate_vm_and_row_bytes_params.DCCMetaPitch = p->myPipe[k].DCCMetaPitchC;
- scratch->calculate_vm_and_row_bytes_params.mrq_present = p->mrq_present;
-
- scratch->calculate_vm_and_row_bytes_params.PixelPTEBytesPerRow = &s->PixelPTEBytesPerRowC[k];
- scratch->calculate_vm_and_row_bytes_params.PixelPTEBytesPerRowStorage = &s->PixelPTEBytesPerRowStorageC[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_width_ub = &p->dpte_row_width_chroma_ub[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_height = &p->dpte_row_height_chroma[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_height_linear = &p->dpte_row_height_linear_chroma[k];
- scratch->calculate_vm_and_row_bytes_params.PixelPTEBytesPerRow_one_row_per_frame = &s->PixelPTEBytesPerRowC_one_row_per_frame[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_width_ub_one_row_per_frame = &s->dpte_row_width_chroma_ub_one_row_per_frame[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_height_one_row_per_frame = &s->dpte_row_height_chroma_one_row_per_frame[k];
- scratch->calculate_vm_and_row_bytes_params.vmpg_width = &p->vmpg_width_c[k];
- scratch->calculate_vm_and_row_bytes_params.vmpg_height = &p->vmpg_height_c[k];
- scratch->calculate_vm_and_row_bytes_params.PixelPTEReqWidth = &p->PixelPTEReqWidthC[k];
- scratch->calculate_vm_and_row_bytes_params.PixelPTEReqHeight = &p->PixelPTEReqHeightC[k];
- scratch->calculate_vm_and_row_bytes_params.PTERequestSize = &p->PTERequestSizeC[k];
- scratch->calculate_vm_and_row_bytes_params.dpde0_bytes_per_frame_ub = &p->dpde0_bytes_per_frame_ub_c[k];
-
- scratch->calculate_vm_and_row_bytes_params.meta_row_bytes = &s->meta_row_bytes_per_row_ub_c[k];
- scratch->calculate_vm_and_row_bytes_params.MetaRequestWidth = &p->meta_req_width_chroma[k];
- scratch->calculate_vm_and_row_bytes_params.MetaRequestHeight = &p->meta_req_height_chroma[k];
- scratch->calculate_vm_and_row_bytes_params.meta_row_width = &p->meta_row_width_chroma[k];
- scratch->calculate_vm_and_row_bytes_params.meta_row_height = &p->meta_row_height_chroma[k];
- scratch->calculate_vm_and_row_bytes_params.meta_pte_bytes_per_frame_ub = &p->meta_pte_bytes_per_frame_ub_c[k];
-
- s->vm_bytes_c = CalculateVMAndRowBytes(&scratch->calculate_vm_and_row_bytes_params);
-
- p->PrefetchSourceLinesC[k] = CalculatePrefetchSourceLines(
- p->myPipe[k].VRatioChroma,
- p->myPipe[k].VTapsChroma,
- p->myPipe[k].InterlaceEnable,
- p->myPipe[k].ProgressiveToInterlaceUnitInOPP,
- p->myPipe[k].SwathHeightC,
- p->myPipe[k].RotationAngle,
- p->myPipe[k].mirrored,
- p->myPipe[k].ViewportStationary,
- p->SwathWidthC[k],
- p->myPipe[k].ViewportHeightC,
- p->myPipe[k].ViewportXStartC,
- p->myPipe[k].ViewportYStartC,
-
- // Output
- &p->VInitPreFillC[k],
- &p->MaxNumSwathC[k]);
- } else {
- s->PTEBufferSizeInRequestsForLuma[k] = p->PTEBufferSizeInRequestsLuma + p->PTEBufferSizeInRequestsChroma;
- s->PTEBufferSizeInRequestsForChroma[k] = 0;
- s->PixelPTEBytesPerRowC[k] = 0;
- s->PixelPTEBytesPerRowStorageC[k] = 0;
- s->vm_bytes_c = 0;
- p->MaxNumSwathC[k] = 0;
- p->PrefetchSourceLinesC[k] = 0;
- s->dpte_row_height_chroma_one_row_per_frame[k] = 0;
- s->dpte_row_width_chroma_ub_one_row_per_frame[k] = 0;
- s->PixelPTEBytesPerRowC_one_row_per_frame[k] = 0;
- }
-
- scratch->calculate_vm_and_row_bytes_params.ViewportStationary = p->myPipe[k].ViewportStationary;
- scratch->calculate_vm_and_row_bytes_params.DCCEnable = p->myPipe[k].DCCEnable;
- scratch->calculate_vm_and_row_bytes_params.NumberOfDPPs = p->myPipe[k].DPPPerSurface;
- scratch->calculate_vm_and_row_bytes_params.BlockHeight256Bytes = p->myPipe[k].BlockHeight256BytesY;
- scratch->calculate_vm_and_row_bytes_params.BlockWidth256Bytes = p->myPipe[k].BlockWidth256BytesY;
- scratch->calculate_vm_and_row_bytes_params.SourcePixelFormat = p->myPipe[k].SourcePixelFormat;
- scratch->calculate_vm_and_row_bytes_params.SurfaceTiling = p->myPipe[k].SurfaceTiling;
- scratch->calculate_vm_and_row_bytes_params.BytePerPixel = p->myPipe[k].BytePerPixelY;
- scratch->calculate_vm_and_row_bytes_params.RotationAngle = p->myPipe[k].RotationAngle;
- scratch->calculate_vm_and_row_bytes_params.SwathWidth = p->SwathWidthY[k];
- scratch->calculate_vm_and_row_bytes_params.ViewportHeight = p->myPipe[k].ViewportHeight;
- scratch->calculate_vm_and_row_bytes_params.ViewportXStart = p->myPipe[k].ViewportXStart;
- scratch->calculate_vm_and_row_bytes_params.ViewportYStart = p->myPipe[k].ViewportYStart;
- scratch->calculate_vm_and_row_bytes_params.GPUVMEnable = p->display_cfg->gpuvm_enable;
- scratch->calculate_vm_and_row_bytes_params.GPUVMMaxPageTableLevels = p->display_cfg->gpuvm_max_page_table_levels;
- scratch->calculate_vm_and_row_bytes_params.GPUVMMinPageSizeKBytes = p->display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes;
- scratch->calculate_vm_and_row_bytes_params.PTEBufferSizeInRequests = s->PTEBufferSizeInRequestsForLuma[k];
- scratch->calculate_vm_and_row_bytes_params.Pitch = p->myPipe[k].PitchY;
- scratch->calculate_vm_and_row_bytes_params.MacroTileWidth = p->myPipe[k].BlockWidthY;
- scratch->calculate_vm_and_row_bytes_params.MacroTileHeight = p->myPipe[k].BlockHeightY;
- scratch->calculate_vm_and_row_bytes_params.is_phantom = dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k]);
- scratch->calculate_vm_and_row_bytes_params.DCCMetaPitch = p->myPipe[k].DCCMetaPitchY;
- scratch->calculate_vm_and_row_bytes_params.mrq_present = p->mrq_present;
-
- scratch->calculate_vm_and_row_bytes_params.PixelPTEBytesPerRow = &s->PixelPTEBytesPerRowY[k];
- scratch->calculate_vm_and_row_bytes_params.PixelPTEBytesPerRowStorage = &s->PixelPTEBytesPerRowStorageY[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_width_ub = &p->dpte_row_width_luma_ub[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_height = &p->dpte_row_height_luma[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_height_linear = &p->dpte_row_height_linear_luma[k];
- scratch->calculate_vm_and_row_bytes_params.PixelPTEBytesPerRow_one_row_per_frame = &s->PixelPTEBytesPerRowY_one_row_per_frame[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_width_ub_one_row_per_frame = &s->dpte_row_width_luma_ub_one_row_per_frame[k];
- scratch->calculate_vm_and_row_bytes_params.dpte_row_height_one_row_per_frame = &s->dpte_row_height_luma_one_row_per_frame[k];
- scratch->calculate_vm_and_row_bytes_params.vmpg_width = &p->vmpg_width_y[k];
- scratch->calculate_vm_and_row_bytes_params.vmpg_height = &p->vmpg_height_y[k];
- scratch->calculate_vm_and_row_bytes_params.PixelPTEReqWidth = &p->PixelPTEReqWidthY[k];
- scratch->calculate_vm_and_row_bytes_params.PixelPTEReqHeight = &p->PixelPTEReqHeightY[k];
- scratch->calculate_vm_and_row_bytes_params.PTERequestSize = &p->PTERequestSizeY[k];
- scratch->calculate_vm_and_row_bytes_params.dpde0_bytes_per_frame_ub = &p->dpde0_bytes_per_frame_ub_l[k];
-
- scratch->calculate_vm_and_row_bytes_params.meta_row_bytes = &s->meta_row_bytes_per_row_ub_l[k];
- scratch->calculate_vm_and_row_bytes_params.MetaRequestWidth = &p->meta_req_width_luma[k];
- scratch->calculate_vm_and_row_bytes_params.MetaRequestHeight = &p->meta_req_height_luma[k];
- scratch->calculate_vm_and_row_bytes_params.meta_row_width = &p->meta_row_width_luma[k];
- scratch->calculate_vm_and_row_bytes_params.meta_row_height = &p->meta_row_height_luma[k];
- scratch->calculate_vm_and_row_bytes_params.meta_pte_bytes_per_frame_ub = &p->meta_pte_bytes_per_frame_ub_l[k];
-
- s->vm_bytes_l = CalculateVMAndRowBytes(&scratch->calculate_vm_and_row_bytes_params);
-
- p->PrefetchSourceLinesY[k] = CalculatePrefetchSourceLines(
- p->myPipe[k].VRatio,
- p->myPipe[k].VTaps,
- p->myPipe[k].InterlaceEnable,
- p->myPipe[k].ProgressiveToInterlaceUnitInOPP,
- p->myPipe[k].SwathHeightY,
- p->myPipe[k].RotationAngle,
- p->myPipe[k].mirrored,
- p->myPipe[k].ViewportStationary,
- p->SwathWidthY[k],
- p->myPipe[k].ViewportHeight,
- p->myPipe[k].ViewportXStart,
- p->myPipe[k].ViewportYStart,
-
- // Output
- &p->VInitPreFillY[k],
- &p->MaxNumSwathY[k]);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, vm_bytes_l = %u (before hvm level)\n", __func__, k, s->vm_bytes_l);
- dml2_printf("DML::%s: k=%u, vm_bytes_c = %u (before hvm level)\n", __func__, k, s->vm_bytes_c);
- dml2_printf("DML::%s: k=%u, meta_row_bytes_per_row_ub_l = %u\n", __func__, k, s->meta_row_bytes_per_row_ub_l[k]);
- dml2_printf("DML::%s: k=%u, meta_row_bytes_per_row_ub_c = %u\n", __func__, k, s->meta_row_bytes_per_row_ub_c[k]);
-#endif
- p->vm_bytes[k] = (s->vm_bytes_l + s->vm_bytes_c) * (1 + 8 * s->HostVMDynamicLevels);
- p->meta_row_bytes[k] = s->meta_row_bytes_per_row_ub_l[k] + s->meta_row_bytes_per_row_ub_c[k];
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, meta_row_bytes = %u\n", __func__, k, p->meta_row_bytes[k]);
- dml2_printf("DML::%s: k=%u, vm_bytes = %u (after hvm level)\n", __func__, k, p->vm_bytes[k]);
-#endif
- if (s->PixelPTEBytesPerRowStorageY[k] <= 64 * s->PTEBufferSizeInRequestsForLuma[k] && s->PixelPTEBytesPerRowStorageC[k] <= 64 * s->PTEBufferSizeInRequestsForChroma[k]) {
- p->PTEBufferSizeNotExceeded[k] = true;
- } else {
- p->PTEBufferSizeNotExceeded[k] = false;
- }
-
- s->one_row_per_frame_fits_in_buffer[k] = (s->PixelPTEBytesPerRowY_one_row_per_frame[k] <= 64 * 2 * s->PTEBufferSizeInRequestsForLuma[k] &&
- s->PixelPTEBytesPerRowC_one_row_per_frame[k] <= 64 * 2 * s->PTEBufferSizeInRequestsForChroma[k]);
-#ifdef __DML_VBA_DEBUG__
- if (p->PTEBufferSizeNotExceeded[k] == 0 || s->one_row_per_frame_fits_in_buffer[k] == 0) {
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowY = %u (before hvm level)\n", __func__, k, s->PixelPTEBytesPerRowY[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowC = %u (before hvm level)\n", __func__, k, s->PixelPTEBytesPerRowC[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowStorageY = %u\n", __func__, k, s->PixelPTEBytesPerRowStorageY[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowStorageC = %u\n", __func__, k, s->PixelPTEBytesPerRowStorageC[k]);
- dml2_printf("DML::%s: k=%u, PTEBufferSizeInRequestsForLuma = %u\n", __func__, k, s->PTEBufferSizeInRequestsForLuma[k]);
- dml2_printf("DML::%s: k=%u, PTEBufferSizeInRequestsForChroma = %u\n", __func__, k, s->PTEBufferSizeInRequestsForChroma[k]);
- dml2_printf("DML::%s: k=%u, PTEBufferSizeNotExceeded (not one_row_per_frame) = %u\n", __func__, k, p->PTEBufferSizeNotExceeded[k]);
-
- dml2_printf("DML::%s: k=%u, HostVMDynamicLevels = %u\n", __func__, k, s->HostVMDynamicLevels);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowY_one_row_per_frame = %u\n", __func__, k, s->PixelPTEBytesPerRowY_one_row_per_frame[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowC_one_row_per_frame = %u\n", __func__, k, s->PixelPTEBytesPerRowC_one_row_per_frame[k]);
- dml2_printf("DML::%s: k=%u, one_row_per_frame_fits_in_buffer = %u\n", __func__, k, s->one_row_per_frame_fits_in_buffer[k]);
- }
-#endif
- }
-
- CalculateMALLUseForStaticScreen(
- p->display_cfg,
- p->NumberOfActiveSurfaces,
- p->MALLAllocatedForDCN,
- p->SurfaceSizeInMALL,
- s->one_row_per_frame_fits_in_buffer,
- // Output
- p->is_using_mall_for_ss);
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (p->display_cfg->gpuvm_enable) {
- if (p->display_cfg->plane_descriptors[k].overrides.hw.force_pte_buffer_mode.enable == 1) {
- p->PTE_BUFFER_MODE[k] = p->display_cfg->plane_descriptors[k].overrides.hw.force_pte_buffer_mode.value;
- }
- p->PTE_BUFFER_MODE[k] = p->myPipe[k].FORCE_ONE_ROW_FOR_FRAME || p->is_using_mall_for_ss[k] || (p->display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_main_pipe) ||
- dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k]) || (p->display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes > 64);
- p->BIGK_FRAGMENT_SIZE[k] = (unsigned int)(math_log((float)p->display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes * 1024, 2) - 12);
- } else {
- p->PTE_BUFFER_MODE[k] = 0;
- p->BIGK_FRAGMENT_SIZE[k] = 0;
- }
- }
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- p->DCCMetaBufferSizeNotExceeded[k] = true;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, SurfaceSizeInMALL = %u\n", __func__, k, p->SurfaceSizeInMALL[k]);
- dml2_printf("DML::%s: k=%u, is_using_mall_for_ss = %u\n", __func__, k, p->is_using_mall_for_ss[k]);
-#endif
- p->use_one_row_for_frame[k] = p->myPipe[k].FORCE_ONE_ROW_FOR_FRAME || p->is_using_mall_for_ss[k] || (p->display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_main_pipe) ||
- (dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) || (p->display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes > 64 && dml_is_vertical_rotation(p->myPipe[k].RotationAngle));
-
- p->use_one_row_for_frame_flip[k] = p->use_one_row_for_frame[k] && !(p->display_cfg->plane_descriptors[k].overrides.uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_mall_full_frame);
-
- if (p->use_one_row_for_frame[k]) {
- p->dpte_row_height_luma[k] = s->dpte_row_height_luma_one_row_per_frame[k];
- p->dpte_row_width_luma_ub[k] = s->dpte_row_width_luma_ub_one_row_per_frame[k];
- s->PixelPTEBytesPerRowY[k] = s->PixelPTEBytesPerRowY_one_row_per_frame[k];
- p->dpte_row_height_chroma[k] = s->dpte_row_height_chroma_one_row_per_frame[k];
- p->dpte_row_width_chroma_ub[k] = s->dpte_row_width_chroma_ub_one_row_per_frame[k];
- s->PixelPTEBytesPerRowC[k] = s->PixelPTEBytesPerRowC_one_row_per_frame[k];
- p->PTEBufferSizeNotExceeded[k] = s->one_row_per_frame_fits_in_buffer[k];
- }
-
- if (p->meta_row_bytes[k] <= p->DCCMetaBufferSizeBytes) {
- p->DCCMetaBufferSizeNotExceeded[k] = true;
- } else {
- p->DCCMetaBufferSizeNotExceeded[k] = false;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, meta_row_bytes = %d\n", __func__, k, p->meta_row_bytes[k]);
- dml2_printf("DML::%s: k=%d, DCCMetaBufferSizeBytes = %d\n", __func__, k, p->DCCMetaBufferSizeBytes);
- dml2_printf("DML::%s: k=%d, DCCMetaBufferSizeNotExceeded = %d\n", __func__, k, p->DCCMetaBufferSizeNotExceeded[k]);
-#endif
- }
-
- s->PixelPTEBytesPerRowY[k] = s->PixelPTEBytesPerRowY[k] * (1 + 8 * s->HostVMDynamicLevels);
- s->PixelPTEBytesPerRowC[k] = s->PixelPTEBytesPerRowC[k] * (1 + 8 * s->HostVMDynamicLevels);
- p->PixelPTEBytesPerRow[k] = s->PixelPTEBytesPerRowY[k] + s->PixelPTEBytesPerRowC[k];
-
- // if one row of dPTEs is meant to span the entire frame, then for these calculations, we will pretend like that one big row is fetched in two halfs
- if (p->use_one_row_for_frame[k])
- p->PixelPTEBytesPerRow[k] = p->PixelPTEBytesPerRow[k] / 2;
-
- CalculateRowBandwidth(
- p->display_cfg->gpuvm_enable,
- p->use_one_row_for_frame[k],
- p->myPipe[k].SourcePixelFormat,
- p->myPipe[k].VRatio,
- p->myPipe[k].VRatioChroma,
- p->myPipe[k].DCCEnable,
- p->myPipe[k].HTotal / p->myPipe[k].PixelClock,
- s->PixelPTEBytesPerRowY[k],
- s->PixelPTEBytesPerRowC[k],
- p->dpte_row_height_luma[k],
- p->dpte_row_height_chroma[k],
-
- p->mrq_present,
- s->meta_row_bytes_per_row_ub_l[k],
- s->meta_row_bytes_per_row_ub_c[k],
- p->meta_row_height_luma[k],
- p->meta_row_height_chroma[k],
-
- // Output
- &p->dpte_row_bw[k],
- &p->meta_row_bw[k]);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, use_one_row_for_frame = %u\n", __func__, k, p->use_one_row_for_frame[k]);
- dml2_printf("DML::%s: k=%u, use_one_row_for_frame_flip = %u\n", __func__, k, p->use_one_row_for_frame_flip[k]);
- dml2_printf("DML::%s: k=%u, UseMALLForPStateChange = %u\n", __func__, k, p->display_cfg->plane_descriptors[k].overrides.legacy_svp_config);
- dml2_printf("DML::%s: k=%u, dpte_row_height_luma = %u\n", __func__, k, p->dpte_row_height_luma[k]);
- dml2_printf("DML::%s: k=%u, dpte_row_width_luma_ub = %u\n", __func__, k, p->dpte_row_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowY = %u (after hvm level)\n", __func__, k, s->PixelPTEBytesPerRowY[k]);
- dml2_printf("DML::%s: k=%u, dpte_row_height_chroma = %u\n", __func__, k, p->dpte_row_height_chroma[k]);
- dml2_printf("DML::%s: k=%u, dpte_row_width_chroma_ub = %u\n", __func__, k, p->dpte_row_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRowC = %u (after hvm level)\n", __func__, k, s->PixelPTEBytesPerRowC[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEBytesPerRow = %u\n", __func__, k, p->PixelPTEBytesPerRow[k]);
- dml2_printf("DML::%s: k=%u, PTEBufferSizeNotExceeded = %u\n", __func__, k, p->PTEBufferSizeNotExceeded[k]);
- dml2_printf("DML::%s: k=%u, gpuvm_enable = %u\n", __func__, k, p->display_cfg->gpuvm_enable);
- dml2_printf("DML::%s: k=%u, PTE_BUFFER_MODE = %u\n", __func__, k, p->PTE_BUFFER_MODE[k]);
- dml2_printf("DML::%s: k=%u, BIGK_FRAGMENT_SIZE = %u\n", __func__, k, p->BIGK_FRAGMENT_SIZE[k]);
-#endif
- }
-}
-
-static double CalculateUrgentLatency(
- double UrgentLatencyPixelDataOnly,
- double UrgentLatencyPixelMixedWithVMData,
- double UrgentLatencyVMDataOnly,
- bool DoUrgentLatencyAdjustment,
- double UrgentLatencyAdjustmentFabricClockComponent,
- double UrgentLatencyAdjustmentFabricClockReference,
- double FabricClock,
- double uclk_freq_mhz,
- enum dml2_qos_param_type qos_type,
- unsigned int urgent_ramp_uclk_cycles,
- unsigned int df_qos_response_time_fclk_cycles,
- unsigned int max_round_trip_to_furthest_cs_fclk_cycles,
- unsigned int mall_overhead_fclk_cycles,
- double umc_urgent_ramp_latency_margin,
- double fabric_max_transport_latency_margin)
-{
- double urgent_latency = 0;
- if (qos_type == dml2_qos_param_type_dcn4x) {
- urgent_latency = (df_qos_response_time_fclk_cycles + mall_overhead_fclk_cycles) / FabricClock
- + max_round_trip_to_furthest_cs_fclk_cycles / FabricClock * (1 + fabric_max_transport_latency_margin / 100.0)
- + urgent_ramp_uclk_cycles / uclk_freq_mhz * (1 + umc_urgent_ramp_latency_margin / 100.0);
- } else {
- urgent_latency = math_max3(UrgentLatencyPixelDataOnly, UrgentLatencyPixelMixedWithVMData, UrgentLatencyVMDataOnly);
- if (DoUrgentLatencyAdjustment == true) {
- urgent_latency = urgent_latency + UrgentLatencyAdjustmentFabricClockComponent * (UrgentLatencyAdjustmentFabricClockReference / FabricClock - 1);
- }
- }
-#ifdef __DML_VBA_DEBUG__
- if (qos_type == dml2_qos_param_type_dcn4x) {
- dml2_printf("DML::%s: qos_type = %d\n", __func__, qos_type);
- dml2_printf("DML::%s: urgent_ramp_uclk_cycles = %d\n", __func__, urgent_ramp_uclk_cycles);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, uclk_freq_mhz);
- dml2_printf("DML::%s: umc_urgent_ramp_latency_margin = %f\n", __func__, umc_urgent_ramp_latency_margin);
- } else {
- dml2_printf("DML::%s: UrgentLatencyPixelDataOnly = %f\n", __func__, UrgentLatencyPixelDataOnly);
- dml2_printf("DML::%s: UrgentLatencyPixelMixedWithVMData = %f\n", __func__, UrgentLatencyPixelMixedWithVMData);
- dml2_printf("DML::%s: UrgentLatencyVMDataOnly = %f\n", __func__, UrgentLatencyVMDataOnly);
- dml2_printf("DML::%s: UrgentLatencyAdjustmentFabricClockComponent = %f\n", __func__, UrgentLatencyAdjustmentFabricClockComponent);
- dml2_printf("DML::%s: UrgentLatencyAdjustmentFabricClockReference = %f\n", __func__, UrgentLatencyAdjustmentFabricClockReference);
- }
- dml2_printf("DML::%s: FabricClock = %f\n", __func__, FabricClock);
- dml2_printf("DML::%s: UrgentLatency = %f\n", __func__, urgent_latency);
-#endif
- return urgent_latency;
-}
-
-static double CalculateTripToMemory(
- double UrgLatency,
- double FabricClock,
- double uclk_freq_mhz,
- enum dml2_qos_param_type qos_type,
- unsigned int trip_to_memory_uclk_cycles,
- unsigned int max_round_trip_to_furthest_cs_fclk_cycles,
- unsigned int mall_overhead_fclk_cycles,
- double umc_max_latency_margin,
- double fabric_max_transport_latency_margin)
-{
- double trip_to_memory_us;
- if (qos_type == dml2_qos_param_type_dcn4x) {
- trip_to_memory_us = mall_overhead_fclk_cycles / FabricClock
- + max_round_trip_to_furthest_cs_fclk_cycles / FabricClock * (1.0 + fabric_max_transport_latency_margin / 100.0)
- + trip_to_memory_uclk_cycles / uclk_freq_mhz * (1.0 + umc_max_latency_margin / 100.0);
- } else {
- trip_to_memory_us = UrgLatency;
- }
-
-#ifdef __DML_VBA_DEBUG__
- if (qos_type == dml2_qos_param_type_dcn4x) {
- dml2_printf("DML::%s: qos_type = %d\n", __func__, qos_type);
- dml2_printf("DML::%s: max_round_trip_to_furthest_cs_fclk_cycles = %d\n", __func__, max_round_trip_to_furthest_cs_fclk_cycles);
- dml2_printf("DML::%s: mall_overhead_fclk_cycles = %d\n", __func__, mall_overhead_fclk_cycles);
- dml2_printf("DML::%s: trip_to_memory_uclk_cycles = %d\n", __func__, trip_to_memory_uclk_cycles);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, uclk_freq_mhz);
- dml2_printf("DML::%s: FabricClock = %f\n", __func__, FabricClock);
- dml2_printf("DML::%s: fabric_max_transport_latency_margin = %f\n", __func__, fabric_max_transport_latency_margin);
- dml2_printf("DML::%s: umc_max_latency_margin = %f\n", __func__, umc_max_latency_margin);
- } else {
- dml2_printf("DML::%s: UrgLatency = %f\n", __func__, UrgLatency);
- }
- dml2_printf("DML::%s: trip_to_memory_us = %f\n", __func__, trip_to_memory_us);
-#endif
-
-
- return trip_to_memory_us;
-}
-
-static double CalculateMetaTripToMemory(
- double UrgLatency,
- double FabricClock,
- double uclk_freq_mhz,
- enum dml2_qos_param_type qos_type,
- unsigned int meta_trip_to_memory_uclk_cycles,
- unsigned int meta_trip_to_memory_fclk_cycles,
- double umc_max_latency_margin,
- double fabric_max_transport_latency_margin)
-{
- double meta_trip_to_memory_us;
- if (qos_type == dml2_qos_param_type_dcn4x) {
- meta_trip_to_memory_us = meta_trip_to_memory_fclk_cycles / FabricClock * (1.0 + fabric_max_transport_latency_margin / 100.0)
- + meta_trip_to_memory_uclk_cycles / uclk_freq_mhz * (1.0 + umc_max_latency_margin / 100.0);
- } else {
- meta_trip_to_memory_us = UrgLatency;
- }
-
-#ifdef __DML_VBA_DEBUG__
- if (qos_type == dml2_qos_param_type_dcn4x) {
- dml2_printf("DML::%s: qos_type = %d\n", __func__, qos_type);
- dml2_printf("DML::%s: meta_trip_to_memory_fclk_cycles = %d\n", __func__, meta_trip_to_memory_fclk_cycles);
- dml2_printf("DML::%s: meta_trip_to_memory_uclk_cycles = %d\n", __func__, meta_trip_to_memory_uclk_cycles);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, uclk_freq_mhz);
- } else {
- dml2_printf("DML::%s: UrgLatency = %f\n", __func__, UrgLatency);
- }
- dml2_printf("DML::%s: meta_trip_to_memory_us = %f\n", __func__, meta_trip_to_memory_us);
-#endif
-
-
- return meta_trip_to_memory_us;
-}
-
-static void calculate_cursor_req_attributes(
- unsigned int cursor_width,
- unsigned int cursor_bpp,
-
- // output
- unsigned int *cursor_lines_per_chunk,
- unsigned int *cursor_bytes_per_line,
- unsigned int *cursor_bytes_per_chunk,
- unsigned int *cursor_bytes)
-{
- unsigned int cursor_pitch = 0;
- unsigned int cursor_bytes_per_req = 0;
- unsigned int cursor_width_bytes = 0;
- unsigned int cursor_height = 0;
-
- //SW determines the cursor pitch to support the maximum cursor_width that will be used but the following restrictions apply.
- //- For 2bpp, cursor_pitch = 256 pixels due to min cursor request size of 64B
- //- For 32 or 64 bpp, cursor_pitch = 64, 128 or 256 pixels depending on the cursor width
- if (cursor_bpp == 2)
- cursor_pitch = 256;
- else
- cursor_pitch = (unsigned int)1 << (unsigned int)math_ceil2(math_log((float)cursor_width, 2), 1);
-
- //The cursor requestor uses a cursor request size of 64B, 128B, or 256B depending on the cursor_width and cursor_bpp as follows.
-
- cursor_width_bytes = (unsigned int)math_ceil2((double)cursor_width * cursor_bpp / 8, 1);
- if (cursor_width_bytes <= 64)
- cursor_bytes_per_req = 64;
- else if (cursor_width_bytes <= 128)
- cursor_bytes_per_req = 128;
- else
- cursor_bytes_per_req = 256;
-
- //If cursor_width_bytes is greater than 256B, then multiple 256B requests are issued to fetch the entire cursor line.
- *cursor_bytes_per_line = (unsigned int)math_ceil2((double)cursor_width_bytes, cursor_bytes_per_req);
-
- //Nominally, the cursor chunk is 1KB or 2KB but it is restricted to a power of 2 number of lines with a maximum of 16 lines.
- if (cursor_bpp == 2) {
- *cursor_lines_per_chunk = 16;
- } else if (cursor_bpp == 32) {
- if (cursor_width <= 32)
- *cursor_lines_per_chunk = 16;
- else if (cursor_width <= 64)
- *cursor_lines_per_chunk = 8;
- else if (cursor_width <= 128)
- *cursor_lines_per_chunk = 4;
- else
- *cursor_lines_per_chunk = 2;
- } else if (cursor_bpp == 64) {
- if (cursor_width <= 16)
- *cursor_lines_per_chunk = 16;
- else if (cursor_width <= 32)
- *cursor_lines_per_chunk = 8;
- else if (cursor_width <= 64)
- *cursor_lines_per_chunk = 4;
- else if (cursor_width <= 128)
- *cursor_lines_per_chunk = 2;
- else
- *cursor_lines_per_chunk = 1;
- } else {
- if (cursor_width > 0) {
- dml2_printf("DML::%s: Invalid cursor_bpp = %d\n", __func__, cursor_bpp);
- dml2_assert(0);
- }
- }
-
- *cursor_bytes_per_chunk = *cursor_bytes_per_line * *cursor_lines_per_chunk;
-
- // For the cursor implementation, all requested data is stored in the return buffer. Given this fact, the cursor_bytes can be directly compared with the CursorBufferSize.
- // Only cursor_width is provided for worst case sizing so assume that the cursor is square
- cursor_height = cursor_width;
- *cursor_bytes = *cursor_bytes_per_line * cursor_height;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: cursor_bpp = %d\n", __func__, cursor_bpp);
- dml2_printf("DML::%s: cursor_width = %d\n", __func__, cursor_width);
- dml2_printf("DML::%s: cursor_width_bytes = %d\n", __func__, cursor_width_bytes);
- dml2_printf("DML::%s: cursor_bytes_per_req = %d\n", __func__, cursor_bytes_per_req);
- dml2_printf("DML::%s: cursor_lines_per_chunk = %d\n", __func__, *cursor_lines_per_chunk);
- dml2_printf("DML::%s: cursor_bytes_per_line = %d\n", __func__, *cursor_bytes_per_line);
- dml2_printf("DML::%s: cursor_bytes_per_chunk = %d\n", __func__, *cursor_bytes_per_chunk);
- dml2_printf("DML::%s: cursor_bytes = %d\n", __func__, *cursor_bytes);
- dml2_printf("DML::%s: cursor_pitch = %d\n", __func__, cursor_pitch);
-#endif
-}
-
-static void calculate_cursor_urgent_burst_factor(
- unsigned int CursorBufferSize,
- unsigned int CursorWidth,
- unsigned int cursor_bytes_per_chunk,
- unsigned int cursor_lines_per_chunk,
- double LineTime,
- double UrgentLatency,
-
- double *UrgentBurstFactorCursor,
- bool *NotEnoughUrgentLatencyHiding)
-{
- unsigned int LinesInCursorBuffer = 0;
- double CursorBufferSizeInTime = 0;
-
- if (CursorWidth > 0) {
- LinesInCursorBuffer = (unsigned int)math_floor2(CursorBufferSize * 1024.0 / (double)cursor_bytes_per_chunk, 1) * cursor_lines_per_chunk;
-
- CursorBufferSizeInTime = LinesInCursorBuffer * LineTime;
- if (CursorBufferSizeInTime - UrgentLatency <= 0) {
- *NotEnoughUrgentLatencyHiding = 1;
- *UrgentBurstFactorCursor = 0;
- } else {
- *NotEnoughUrgentLatencyHiding = 0;
- *UrgentBurstFactorCursor = CursorBufferSizeInTime / (CursorBufferSizeInTime - UrgentLatency);
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: LinesInCursorBuffer = %u\n", __func__, LinesInCursorBuffer);
- dml2_printf("DML::%s: CursorBufferSizeInTime = %f\n", __func__, CursorBufferSizeInTime);
- dml2_printf("DML::%s: CursorBufferSize = %u (kbytes)\n", __func__, CursorBufferSize);
- dml2_printf("DML::%s: cursor_bytes_per_chunk = %u\n", __func__, cursor_bytes_per_chunk);
- dml2_printf("DML::%s: cursor_lines_per_chunk = %u\n", __func__, cursor_lines_per_chunk);
- dml2_printf("DML::%s: UrgentBurstFactorCursor = %f\n", __func__, *UrgentBurstFactorCursor);
- dml2_printf("DML::%s: NotEnoughUrgentLatencyHiding = %d\n", __func__, *NotEnoughUrgentLatencyHiding);
-#endif
-
- }
-}
-
-static void CalculateUrgentBurstFactor(
- const struct dml2_plane_parameters *plane_cfg,
- unsigned int swath_width_luma_ub,
- unsigned int swath_width_chroma_ub,
- unsigned int SwathHeightY,
- unsigned int SwathHeightC,
- double LineTime,
- double UrgentLatency,
- double VRatio,
- double VRatioC,
- double BytePerPixelInDETY,
- double BytePerPixelInDETC,
- unsigned int DETBufferSizeY,
- unsigned int DETBufferSizeC,
- // Output
- double *UrgentBurstFactorLuma,
- double *UrgentBurstFactorChroma,
- bool *NotEnoughUrgentLatencyHiding)
-{
- double LinesInDETLuma;
- double LinesInDETChroma;
- double DETBufferSizeInTimeLuma;
- double DETBufferSizeInTimeChroma;
-
- *NotEnoughUrgentLatencyHiding = 0;
- *UrgentBurstFactorLuma = 0;
- *UrgentBurstFactorChroma = 0;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatio = %f\n", __func__, VRatio);
- dml2_printf("DML::%s: VRatioC = %f\n", __func__, VRatioC);
- dml2_printf("DML::%s: DETBufferSizeY = %d\n", __func__, DETBufferSizeY);
- dml2_printf("DML::%s: DETBufferSizeC = %d\n", __func__, DETBufferSizeC);
- dml2_printf("DML::%s: BytePerPixelInDETY = %f\n", __func__, BytePerPixelInDETY);
- dml2_printf("DML::%s: swath_width_luma_ub = %d\n", __func__, swath_width_luma_ub);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, LineTime);
-#endif
- DML2_ASSERT(VRatio > 0);
-
- LinesInDETLuma = (dml_is_phantom_pipe(plane_cfg) ? 1024 * 1024 : DETBufferSizeY) / BytePerPixelInDETY / swath_width_luma_ub;
-
- DETBufferSizeInTimeLuma = math_floor2(LinesInDETLuma, SwathHeightY) * LineTime / VRatio;
- if (DETBufferSizeInTimeLuma - UrgentLatency <= 0) {
- *NotEnoughUrgentLatencyHiding = 1;
- *UrgentBurstFactorLuma = 0;
- } else {
- *UrgentBurstFactorLuma = DETBufferSizeInTimeLuma / (DETBufferSizeInTimeLuma - UrgentLatency);
- }
-
- if (BytePerPixelInDETC > 0) {
- LinesInDETChroma = (dml_is_phantom_pipe(plane_cfg) ? 1024 * 1024 : DETBufferSizeC) / BytePerPixelInDETC / swath_width_chroma_ub;
-
- DETBufferSizeInTimeChroma = math_floor2(LinesInDETChroma, SwathHeightC) * LineTime / VRatioC;
- if (DETBufferSizeInTimeChroma - UrgentLatency <= 0) {
- *NotEnoughUrgentLatencyHiding = 1;
- *UrgentBurstFactorChroma = 0;
- } else {
- *UrgentBurstFactorChroma = DETBufferSizeInTimeChroma / (DETBufferSizeInTimeChroma - UrgentLatency);
- }
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: LinesInDETLuma = %f\n", __func__, LinesInDETLuma);
- dml2_printf("DML::%s: UrgentLatency = %f\n", __func__, UrgentLatency);
- dml2_printf("DML::%s: DETBufferSizeInTimeLuma = %f\n", __func__, DETBufferSizeInTimeLuma);
- dml2_printf("DML::%s: UrgentBurstFactorLuma = %f\n", __func__, *UrgentBurstFactorLuma);
- dml2_printf("DML::%s: UrgentBurstFactorChroma = %f\n", __func__, *UrgentBurstFactorChroma);
- dml2_printf("DML::%s: NotEnoughUrgentLatencyHiding = %d\n", __func__, *NotEnoughUrgentLatencyHiding);
-#endif
-
-}
-
-static void CalculateDCFCLKDeepSleep(
- const struct dml2_display_cfg *display_cfg,
- unsigned int NumberOfActiveSurfaces,
- unsigned int BytePerPixelY[],
- unsigned int BytePerPixelC[],
- unsigned int SwathWidthY[],
- unsigned int SwathWidthC[],
- unsigned int DPPPerSurface[],
- double PSCL_THROUGHPUT[],
- double PSCL_THROUGHPUT_CHROMA[],
- double Dppclk[],
- double ReadBandwidthLuma[],
- double ReadBandwidthChroma[],
- unsigned int ReturnBusWidth,
-
- // Output
- double *DCFClkDeepSleep)
-{
- double DisplayPipeLineDeliveryTimeLuma;
- double DisplayPipeLineDeliveryTimeChroma;
- double DCFClkDeepSleepPerSurface[DML2_MAX_PLANES];
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- double pixel_rate_mhz = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
-
- if (display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio <= 1) {
- DisplayPipeLineDeliveryTimeLuma = SwathWidthY[k] * DPPPerSurface[k] / display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio / pixel_rate_mhz;
- } else {
- DisplayPipeLineDeliveryTimeLuma = SwathWidthY[k] / PSCL_THROUGHPUT[k] / Dppclk[k];
- }
- if (BytePerPixelC[k] == 0) {
- DisplayPipeLineDeliveryTimeChroma = 0;
- } else {
- if (display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio <= 1) {
- DisplayPipeLineDeliveryTimeChroma = SwathWidthC[k] * DPPPerSurface[k] / display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio / pixel_rate_mhz;
- } else {
- DisplayPipeLineDeliveryTimeChroma = SwathWidthC[k] / PSCL_THROUGHPUT_CHROMA[k] / Dppclk[k];
- }
- }
-
- if (BytePerPixelC[k] > 0) {
- DCFClkDeepSleepPerSurface[k] = math_max2(__DML2_CALCS_DCFCLK_FACTOR__ * SwathWidthY[k] * BytePerPixelY[k] / 32.0 / DisplayPipeLineDeliveryTimeLuma,
- __DML2_CALCS_DCFCLK_FACTOR__ * SwathWidthC[k] * BytePerPixelC[k] / 32.0 / DisplayPipeLineDeliveryTimeChroma);
- } else {
- DCFClkDeepSleepPerSurface[k] = __DML2_CALCS_DCFCLK_FACTOR__ * SwathWidthY[k] * BytePerPixelY[k] / 64.0 / DisplayPipeLineDeliveryTimeLuma;
- }
- DCFClkDeepSleepPerSurface[k] = math_max2(DCFClkDeepSleepPerSurface[k], pixel_rate_mhz / 16);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, PixelClock = %f\n", __func__, k, pixel_rate_mhz);
- dml2_printf("DML::%s: k=%u, DCFClkDeepSleepPerSurface = %f\n", __func__, k, DCFClkDeepSleepPerSurface[k]);
-#endif
- }
-
- double ReadBandwidth = 0.0;
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- ReadBandwidth = ReadBandwidth + ReadBandwidthLuma[k] + ReadBandwidthChroma[k];
- }
-
- *DCFClkDeepSleep = math_max2(8.0, __DML2_CALCS_DCFCLK_FACTOR__ * ReadBandwidth / (double)ReturnBusWidth);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: __DML2_CALCS_DCFCLK_FACTOR__ = %f\n", __func__, __DML2_CALCS_DCFCLK_FACTOR__);
- dml2_printf("DML::%s: ReadBandwidth = %f\n", __func__, ReadBandwidth);
- dml2_printf("DML::%s: ReturnBusWidth = %u\n", __func__, ReturnBusWidth);
- dml2_printf("DML::%s: DCFClkDeepSleep = %f\n", __func__, *DCFClkDeepSleep);
-#endif
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- *DCFClkDeepSleep = math_max2(*DCFClkDeepSleep, DCFClkDeepSleepPerSurface[k]);
- }
- dml2_printf("DML::%s: DCFClkDeepSleep = %f (final)\n", __func__, *DCFClkDeepSleep);
-}
-
-static double CalculateWriteBackDelay(
- enum dml2_source_format_class WritebackPixelFormat,
- double WritebackHRatio,
- double WritebackVRatio,
- unsigned int WritebackVTaps,
- unsigned int WritebackDestinationWidth,
- unsigned int WritebackDestinationHeight,
- unsigned int WritebackSourceHeight,
- unsigned int HTotal)
-{
- double CalculateWriteBackDelay;
- double Line_length;
- double Output_lines_last_notclamped;
- double WritebackVInit;
-
- WritebackVInit = (WritebackVRatio + WritebackVTaps + 1) / 2;
- Line_length = math_max2((double)WritebackDestinationWidth, math_ceil2((double)WritebackDestinationWidth / 6.0, 1.0) * WritebackVTaps);
- Output_lines_last_notclamped = WritebackDestinationHeight - 1 - math_ceil2(((double)WritebackSourceHeight - (double)WritebackVInit) / (double)WritebackVRatio, 1.0);
- if (Output_lines_last_notclamped < 0) {
- CalculateWriteBackDelay = 0;
- } else {
- CalculateWriteBackDelay = Output_lines_last_notclamped * Line_length + (HTotal - WritebackDestinationWidth) + 80;
- }
- return CalculateWriteBackDelay;
-}
-
-static unsigned int CalculateMaxVStartup(
- bool ptoi_supported,
- unsigned int vblank_nom_default_us,
- const struct dml2_timing_cfg *timing,
- double write_back_delay_us)
-{
- unsigned int vblank_size = 0;
- unsigned int max_vstartup_lines = 0;
-
- double line_time_us = (double)timing->h_total / ((double)timing->pixel_clock_khz / 1000);
- unsigned int vblank_actual = timing->v_total - timing->v_active;
- unsigned int vblank_nom_default_in_line = (unsigned int)math_floor2((double)vblank_nom_default_us / line_time_us, 1.0);
- unsigned int vblank_nom_input = (unsigned int)math_min2(timing->vblank_nom, vblank_nom_default_in_line);
- unsigned int vblank_avail = (vblank_nom_input == 0) ? vblank_nom_default_in_line : vblank_nom_input;
-
- vblank_size = (unsigned int)math_min2(vblank_actual, vblank_avail);
-
- if (timing->interlaced && !ptoi_supported)
- max_vstartup_lines = (unsigned int)(math_floor2(vblank_size / 2.0, 1.0));
- else
- max_vstartup_lines = vblank_size - (unsigned int)math_max2(1.0, math_ceil2(write_back_delay_us / line_time_us, 1.0));
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VBlankNom = %u\n", __func__, timing->vblank_nom);
- dml2_printf("DML::%s: vblank_nom_default_us = %u\n", __func__, vblank_nom_default_us);
- dml2_printf("DML::%s: line_time_us = %f\n", __func__, line_time_us);
- dml2_printf("DML::%s: vblank_actual = %u\n", __func__, vblank_actual);
- dml2_printf("DML::%s: vblank_avail = %u\n", __func__, vblank_avail);
- dml2_printf("DML::%s: max_vstartup_lines = %u\n", __func__, max_vstartup_lines);
-#endif
- return max_vstartup_lines;
-}
-
-static void CalculateSwathAndDETConfiguration(struct dml2_core_internal_scratch *scratch,
- struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params *p)
-{
- struct dml2_core_shared_CalculateSwathAndDETConfiguration_locals *l = &scratch->CalculateSwathAndDETConfiguration_locals;
- memset(l, 0, sizeof(struct dml2_core_shared_CalculateSwathAndDETConfiguration_locals));
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ForceSingleDPP = %u\n", __func__, p->ForceSingleDPP);
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- dml2_printf("DML::%s: DPPPerSurface[%u] = %u\n", __func__, k, p->DPPPerSurface[k]);
- }
-#endif
- CalculateSwathWidth(
- p->display_cfg,
- p->ForceSingleDPP,
- p->NumberOfActiveSurfaces,
- p->ODMMode,
- p->BytePerPixY,
- p->BytePerPixC,
- p->Read256BytesBlockHeightY,
- p->Read256BytesBlockHeightC,
- p->Read256BytesBlockWidthY,
- p->Read256BytesBlockWidthC,
- p->surf_linear128_l,
- p->surf_linear128_c,
- p->DPPPerSurface,
-
- // Output
- p->req_per_swath_ub_l,
- p->req_per_swath_ub_c,
- l->SwathWidthSingleDPP,
- l->SwathWidthSingleDPPChroma,
- p->SwathWidth,
- p->SwathWidthChroma,
- l->MaximumSwathHeightY,
- l->MaximumSwathHeightC,
- p->swath_width_luma_ub,
- p->swath_width_chroma_ub);
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- p->full_swath_bytes_l[k] = (unsigned int)(p->swath_width_luma_ub[k] * p->BytePerPixDETY[k] * l->MaximumSwathHeightY[k]);
- p->full_swath_bytes_c[k] = (unsigned int)(p->swath_width_chroma_ub[k] * p->BytePerPixDETC[k] * l->MaximumSwathHeightC[k]);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u DPPPerSurface = %u\n", __func__, k, p->DPPPerSurface[k]);
- dml2_printf("DML::%s: k=%u swath_width_luma_ub = %u\n", __func__, k, p->swath_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u BytePerPixDETY = %f\n", __func__, k, p->BytePerPixDETY[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathHeightY = %u\n", __func__, k, l->MaximumSwathHeightY[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_l = %u\n", __func__, k, p->full_swath_bytes_l[k]);
- dml2_printf("DML::%s: k=%u swath_width_chroma_ub = %u\n", __func__, k, p->swath_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u BytePerPixDETC = %f\n", __func__, k, p->BytePerPixDETC[k]);
- dml2_printf("DML::%s: k=%u MaximumSwathHeightC = %u\n", __func__, k, l->MaximumSwathHeightC[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_c = %u\n", __func__, k, p->full_swath_bytes_c[k]);
-#endif
- if (p->display_cfg->plane_descriptors[k].pixel_format == dml2_420_10) {
- p->full_swath_bytes_l[k] = (unsigned int)(math_ceil2((double)p->full_swath_bytes_l[k], 256));
- p->full_swath_bytes_c[k] = (unsigned int)(math_ceil2((double)p->full_swath_bytes_c[k], 256));
- }
- }
-
- unsigned int TotalActiveDPP = 0;
- bool NoChromaOrLinear = true;
- unsigned int SurfaceDoingUnboundedRequest = 0;
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- TotalActiveDPP = TotalActiveDPP + (p->ForceSingleDPP ? 1 : p->DPPPerSurface[k]);
- if (p->DPPPerSurface[k] > 0)
- SurfaceDoingUnboundedRequest = k;
- if (dml2_core_shared_is_420(p->display_cfg->plane_descriptors[k].pixel_format) || p->display_cfg->plane_descriptors[k].pixel_format == dml2_rgbe_alpha
- || p->display_cfg->plane_descriptors[k].surface.tiling == dml2_sw_linear) {
- NoChromaOrLinear = false;
- }
- l->SwathTimeValueUs[k] = (unsigned int) ((double)l->MaximumSwathHeightY[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total
- / p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz * 1000);
- }
-
- *p->UnboundedRequestEnabled = UnboundedRequest(p->display_cfg->overrides.hw.force_unbounded_requesting.enable, p->display_cfg->overrides.hw.force_unbounded_requesting.value, TotalActiveDPP, NoChromaOrLinear);
-
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.display_cfg = p->display_cfg;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.ForceSingleDPP = p->ForceSingleDPP;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.NumberOfActiveSurfaces = p->NumberOfActiveSurfaces;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.UnboundedRequestEnabled = *p->UnboundedRequestEnabled;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.nomDETInKByte = p->nomDETInKByte;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.MaxTotalDETInKByte = p->MaxTotalDETInKByte;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.ConfigReturnBufferSizeInKByte = p->ConfigReturnBufferSizeInKByte;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.MinCompressedBufferSizeInKByte = p->MinCompressedBufferSizeInKByte;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.ConfigReturnBufferSegmentSizeInkByte = p->ConfigReturnBufferSegmentSizeInkByte;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.CompressedBufferSegmentSizeInkByte = p->CompressedBufferSegmentSizeInkByte;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.ReadBandwidthLuma = p->ReadBandwidthLuma;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.ReadBandwidthChroma = p->ReadBandwidthChroma;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.full_swath_bytes_l = p->full_swath_bytes_l;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.full_swath_bytes_c = p->full_swath_bytes_c;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.DPPPerSurface = p->DPPPerSurface;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.DETBufferSizeInKByte = p->DETBufferSizeInKByte;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.CompressedBufferSizeInkByte = p->CompressedBufferSizeInkByte;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.swath_time_value_us = l->SwathTimeValueUs;
- scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params.bestEffortMinActiveLatencyHidingUs = p->display_cfg->overrides.best_effort_min_active_latency_hiding_us;
- if (p->funcs->calculate_det_buffer_size) {
- p->funcs->calculate_det_buffer_size(&scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params);
- } else {
- CalculateDETBufferSize(&scratch->CalculateSwathAndDETConfiguration_locals.calculate_det_buffer_size_params);
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: TotalActiveDPP = %u\n", __func__, TotalActiveDPP);
- dml2_printf("DML::%s: nomDETInKByte = %u\n", __func__, p->nomDETInKByte);
- dml2_printf("DML::%s: ConfigReturnBufferSizeInKByte = %u\n", __func__, p->ConfigReturnBufferSizeInKByte);
- dml2_printf("DML::%s: UnboundedRequestEnabled = %u\n", __func__, *p->UnboundedRequestEnabled);
- dml2_printf("DML::%s: CompressedBufferSizeInkByte = %u\n", __func__, *p->CompressedBufferSizeInkByte);
-#endif
-
- unsigned int DETBufferSizeInKByteForSwathCalculation;
- *p->ViewportSizeSupport = true;
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
-
- DETBufferSizeInKByteForSwathCalculation = (dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k]) ? 1024 : p->DETBufferSizeInKByte[k]);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByteForSwathCalculation = %u\n", __func__, k, DETBufferSizeInKByteForSwathCalculation);
-#endif
-
- if (p->full_swath_bytes_l[k] + p->full_swath_bytes_c[k] <= DETBufferSizeInKByteForSwathCalculation * 1024 / 2) {
- p->SwathHeightY[k] = l->MaximumSwathHeightY[k];
- p->SwathHeightC[k] = l->MaximumSwathHeightC[k];
- l->RoundedUpSwathSizeBytesY[k] = p->full_swath_bytes_l[k];
- l->RoundedUpSwathSizeBytesC[k] = p->full_swath_bytes_c[k];
- p->request_size_bytes_luma[k] = 256;
- p->request_size_bytes_chroma[k] = 256;
-
- } else if (p->full_swath_bytes_l[k] >= 1.5 * p->full_swath_bytes_c[k] && p->full_swath_bytes_l[k] / 2 + p->full_swath_bytes_c[k] <= DETBufferSizeInKByteForSwathCalculation * 1024 / 2) {
- p->SwathHeightY[k] = l->MaximumSwathHeightY[k] / 2;
- p->SwathHeightC[k] = l->MaximumSwathHeightC[k];
- l->RoundedUpSwathSizeBytesY[k] = p->full_swath_bytes_l[k] / 2;
- l->RoundedUpSwathSizeBytesC[k] = p->full_swath_bytes_c[k];
- p->request_size_bytes_luma[k] = ((p->BytePerPixY[k] == 2) == dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle)) ? 128 : 64;
- p->request_size_bytes_chroma[k] = 256;
-
- } else if (p->full_swath_bytes_l[k] < 1.5 * p->full_swath_bytes_c[k] && p->full_swath_bytes_l[k] + p->full_swath_bytes_c[k] / 2 <= DETBufferSizeInKByteForSwathCalculation * 1024 / 2) {
- p->SwathHeightY[k] = l->MaximumSwathHeightY[k];
- p->SwathHeightC[k] = l->MaximumSwathHeightC[k] / 2;
- l->RoundedUpSwathSizeBytesY[k] = p->full_swath_bytes_l[k];
- l->RoundedUpSwathSizeBytesC[k] = p->full_swath_bytes_c[k] / 2;
- p->request_size_bytes_luma[k] = 256;
- p->request_size_bytes_chroma[k] = ((p->BytePerPixC[k] == 2) == dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle)) ? 128 : 64;
-
- } else {
- p->SwathHeightY[k] = l->MaximumSwathHeightY[k] / 2;
- p->SwathHeightC[k] = l->MaximumSwathHeightC[k] / 2;
- l->RoundedUpSwathSizeBytesY[k] = p->full_swath_bytes_l[k] / 2;
- l->RoundedUpSwathSizeBytesC[k] = p->full_swath_bytes_c[k] / 2;
- p->request_size_bytes_luma[k] = ((p->BytePerPixY[k] == 2) == dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle)) ? 128 : 64;
- p->request_size_bytes_chroma[k] = ((p->BytePerPixC[k] == 2) == dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle)) ? 128 : 64;
- }
-
- if (p->SwathHeightC[k] == 0)
- p->request_size_bytes_chroma[k] = 0;
-
- if ((p->full_swath_bytes_l[k] / 2 + p->full_swath_bytes_c[k] / 2 > DETBufferSizeInKByteForSwathCalculation * 1024 / 2) ||
- p->SwathWidth[k] > p->MaximumSwathWidthLuma[k] || (p->SwathHeightC[k] > 0 && p->SwathWidthChroma[k] > p->MaximumSwathWidthChroma[k])) {
- *p->ViewportSizeSupport = false;
- p->ViewportSizeSupportPerSurface[k] = false;
- } else {
- p->ViewportSizeSupportPerSurface[k] = true;
- }
-
- if (p->SwathHeightC[k] == 0) {
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, All DET will be used for plane0\n", __func__, k);
-#endif
- p->DETBufferSizeY[k] = p->DETBufferSizeInKByte[k] * 1024;
- p->DETBufferSizeC[k] = 0;
- } else if (l->RoundedUpSwathSizeBytesY[k] <= 1.5 * l->RoundedUpSwathSizeBytesC[k]) {
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, Half DET will be used for plane0, and half for plane1\n", __func__, k);
-#endif
- p->DETBufferSizeY[k] = p->DETBufferSizeInKByte[k] * 1024 / 2;
- p->DETBufferSizeC[k] = p->DETBufferSizeInKByte[k] * 1024 / 2;
- } else {
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, 2/3 DET will be used for plane0, and 1/3 for plane1\n", __func__, k);
-#endif
- p->DETBufferSizeY[k] = (unsigned int)(math_floor2(p->DETBufferSizeInKByte[k] * 1024 * 2 / 3, 1024));
- p->DETBufferSizeC[k] = p->DETBufferSizeInKByte[k] * 1024 - p->DETBufferSizeY[k];
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u SwathHeightY = %u\n", __func__, k, p->SwathHeightY[k]);
- dml2_printf("DML::%s: k=%u SwathHeightC = %u\n", __func__, k, p->SwathHeightC[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_l = %u\n", __func__, k, p->full_swath_bytes_l[k]);
- dml2_printf("DML::%s: k=%u full_swath_bytes_c = %u\n", __func__, k, p->full_swath_bytes_c[k]);
- dml2_printf("DML::%s: k=%u RoundedUpSwathSizeBytesY = %u\n", __func__, k, l->RoundedUpSwathSizeBytesY[k]);
- dml2_printf("DML::%s: k=%u RoundedUpSwathSizeBytesC = %u\n", __func__, k, l->RoundedUpSwathSizeBytesC[k]);
- dml2_printf("DML::%s: k=%u DETBufferSizeInKByte = %u\n", __func__, k, p->DETBufferSizeInKByte[k]);
- dml2_printf("DML::%s: k=%u DETBufferSizeY = %u\n", __func__, k, p->DETBufferSizeY[k]);
- dml2_printf("DML::%s: k=%u DETBufferSizeC = %u\n", __func__, k, p->DETBufferSizeC[k]);
- dml2_printf("DML::%s: k=%u ViewportSizeSupportPerSurface = %u\n", __func__, k, p->ViewportSizeSupportPerSurface[k]);
-#endif
-
- }
-
- const long TTUFIFODEPTH = 8;
- const long MAXIMUMCOMPRESSION = 4;
- *p->compbuf_reserved_space_64b = 2 * p->pixel_chunk_size_kbytes * 1024 / 64;
- if (*p->UnboundedRequestEnabled) {
- *p->compbuf_reserved_space_64b = (unsigned int)math_ceil2(math_max2(*p->compbuf_reserved_space_64b,
- (double)(p->rob_buffer_size_kbytes * 1024 / 64) - (double)(l->RoundedUpSwathSizeBytesY[SurfaceDoingUnboundedRequest] * TTUFIFODEPTH / 64)), 1.0);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: RoundedUpSwathSizeBytesY[%d] = %u\n", __func__, SurfaceDoingUnboundedRequest, l->RoundedUpSwathSizeBytesY[SurfaceDoingUnboundedRequest]);
- dml2_printf("DML::%s: rob_buffer_size_kbytes = %u\n", __func__, p->rob_buffer_size_kbytes);
-#endif
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: compbuf_reserved_space_64b = %u\n", __func__, *p->compbuf_reserved_space_64b);
-#endif
-
- *p->hw_debug5 = false;
- if (!p->mrq_present) {
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (!(*p->UnboundedRequestEnabled)
- && p->display_cfg->plane_descriptors[k].surface.dcc.enable
- && ((p->rob_buffer_size_kbytes * 1024 + *p->CompressedBufferSizeInkByte * MAXIMUMCOMPRESSION * 1024) > TTUFIFODEPTH * (l->RoundedUpSwathSizeBytesY[k] + l->RoundedUpSwathSizeBytesC[k])))
- *p->hw_debug5 = true;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u UnboundedRequestEnabled = %u\n", __func__, k, *p->UnboundedRequestEnabled);
- dml2_printf("DML::%s: k=%u MAXIMUMCOMPRESSION = %lu\n", __func__, k, MAXIMUMCOMPRESSION);
- dml2_printf("DML::%s: k=%u TTUFIFODEPTH = %lu\n", __func__, k, TTUFIFODEPTH);
- dml2_printf("DML::%s: k=%u CompressedBufferSizeInkByte = %u\n", __func__, k, *p->CompressedBufferSizeInkByte);
- dml2_printf("DML::%s: k=%u RoundedUpSwathSizeBytesC = %u\n", __func__, k, l->RoundedUpSwathSizeBytesC[k]);
- dml2_printf("DML::%s: k=%u hw_debug5 = %u\n", __func__, k, *p->hw_debug5);
-#endif
- }
- }
-}
-
-static void CalculateODMMode(
- unsigned int MaximumPixelsPerLinePerDSCUnit,
- unsigned int HActive,
- enum dml2_output_encoder_class Output,
- enum dml2_odm_mode ODMUse,
- double MaxDispclk,
- bool DSCEnable,
- unsigned int TotalNumberOfActiveDPP,
- unsigned int MaxNumDPP,
- double PixelClock,
-
- // Output
- bool *TotalAvailablePipesSupport,
- unsigned int *NumberOfDPP,
- enum dml2_odm_mode *ODMMode,
- double *RequiredDISPCLKPerSurface)
-{
- double SurfaceRequiredDISPCLKWithoutODMCombine;
- double SurfaceRequiredDISPCLKWithODMCombineTwoToOne;
- double SurfaceRequiredDISPCLKWithODMCombineThreeToOne;
- double SurfaceRequiredDISPCLKWithODMCombineFourToOne;
-
- SurfaceRequiredDISPCLKWithoutODMCombine = CalculateRequiredDispclk(dml2_odm_mode_bypass, PixelClock);
- SurfaceRequiredDISPCLKWithODMCombineTwoToOne = CalculateRequiredDispclk(dml2_odm_mode_combine_2to1, PixelClock);
- SurfaceRequiredDISPCLKWithODMCombineThreeToOne = CalculateRequiredDispclk(dml2_odm_mode_combine_3to1, PixelClock);
- SurfaceRequiredDISPCLKWithODMCombineFourToOne = CalculateRequiredDispclk(dml2_odm_mode_combine_4to1, PixelClock);
- *TotalAvailablePipesSupport = true;
-
- if (ODMUse == dml2_odm_mode_bypass || ODMUse == dml2_odm_mode_auto)
- *ODMMode = dml2_odm_mode_bypass;
- else if (ODMUse == dml2_odm_mode_combine_2to1)
- *ODMMode = dml2_odm_mode_combine_2to1;
- else if (ODMUse == dml2_odm_mode_combine_3to1)
- *ODMMode = dml2_odm_mode_combine_3to1;
- else if (ODMUse == dml2_odm_mode_combine_4to1)
- *ODMMode = dml2_odm_mode_combine_4to1;
- else if (ODMUse == dml2_odm_mode_split_1to2)
- *ODMMode = dml2_odm_mode_split_1to2;
- else if (ODMUse == dml2_odm_mode_mso_1to2)
- *ODMMode = dml2_odm_mode_mso_1to2;
- else if (ODMUse == dml2_odm_mode_mso_1to4)
- *ODMMode = dml2_odm_mode_mso_1to4;
-
- *RequiredDISPCLKPerSurface = SurfaceRequiredDISPCLKWithoutODMCombine;
- *NumberOfDPP = 0;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ODMUse = %d\n", __func__, ODMUse);
- dml2_printf("DML::%s: Output = %d\n", __func__, Output);
- dml2_printf("DML::%s: DSCEnable = %d\n", __func__, DSCEnable);
- dml2_printf("DML::%s: MaxDispclk = %f\n", __func__, MaxDispclk);
- dml2_printf("DML::%s: MaximumPixelsPerLinePerDSCUnit = %d\n", __func__, MaximumPixelsPerLinePerDSCUnit);
- dml2_printf("DML::%s: SurfaceRequiredDISPCLKWithoutODMCombine = %f\n", __func__, SurfaceRequiredDISPCLKWithoutODMCombine);
- dml2_printf("DML::%s: SurfaceRequiredDISPCLKWithODMCombineTwoToOne = %f\n", __func__, SurfaceRequiredDISPCLKWithODMCombineTwoToOne);
- dml2_printf("DML::%s: SurfaceRequiredDISPCLKWithODMCombineThreeToOne = %f\n", __func__, SurfaceRequiredDISPCLKWithODMCombineThreeToOne);
- dml2_printf("DML::%s: SurfaceRequiredDISPCLKWithODMCombineFourToOne = %f\n", __func__, SurfaceRequiredDISPCLKWithODMCombineFourToOne);
-#endif
-
- if (ODMUse == dml2_odm_mode_combine_4to1 || (ODMUse == dml2_odm_mode_auto &&
- (SurfaceRequiredDISPCLKWithODMCombineThreeToOne > MaxDispclk || (DSCEnable && (HActive > 3 * MaximumPixelsPerLinePerDSCUnit))))) {
- if (TotalNumberOfActiveDPP + 4 <= MaxNumDPP) {
- *ODMMode = dml2_odm_mode_combine_4to1;
- *RequiredDISPCLKPerSurface = SurfaceRequiredDISPCLKWithODMCombineFourToOne;
- *NumberOfDPP = 4;
- } else {
- *TotalAvailablePipesSupport = false;
- }
- } else if (ODMUse == dml2_odm_mode_combine_3to1 || (ODMUse == dml2_odm_mode_auto &&
- ((SurfaceRequiredDISPCLKWithODMCombineTwoToOne > MaxDispclk && SurfaceRequiredDISPCLKWithODMCombineThreeToOne <= MaxDispclk) ||
- (DSCEnable && (HActive > 2 * MaximumPixelsPerLinePerDSCUnit))))) {
- if (TotalNumberOfActiveDPP + 3 <= MaxNumDPP) {
- *ODMMode = dml2_odm_mode_combine_3to1;
- *RequiredDISPCLKPerSurface = SurfaceRequiredDISPCLKWithODMCombineThreeToOne;
- *NumberOfDPP = 3;
- } else {
- *TotalAvailablePipesSupport = false;
- }
-
- } else if (ODMUse == dml2_odm_mode_combine_2to1 || (ODMUse == dml2_odm_mode_auto &&
- ((SurfaceRequiredDISPCLKWithoutODMCombine > MaxDispclk && SurfaceRequiredDISPCLKWithODMCombineTwoToOne <= MaxDispclk) ||
- (DSCEnable && (HActive > MaximumPixelsPerLinePerDSCUnit))))) {
- if (TotalNumberOfActiveDPP + 2 <= MaxNumDPP) {
- *ODMMode = dml2_odm_mode_combine_2to1;
- *RequiredDISPCLKPerSurface = SurfaceRequiredDISPCLKWithODMCombineTwoToOne;
- *NumberOfDPP = 2;
- } else {
- *TotalAvailablePipesSupport = false;
- }
-
- } else {
- if (TotalNumberOfActiveDPP + 1 <= MaxNumDPP) {
- *NumberOfDPP = 1;
- } else {
- *TotalAvailablePipesSupport = false;
- }
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: ODMMode = %d\n", __func__, *ODMMode);
- dml2_printf("DML::%s: NumberOfDPP = %d\n", __func__, *NumberOfDPP);
- dml2_printf("DML::%s: TotalAvailablePipesSupport = %d\n", __func__, *TotalAvailablePipesSupport);
- dml2_printf("DML::%s: RequiredDISPCLKPerSurface = %f\n", __func__, *RequiredDISPCLKPerSurface);
-#endif
-
-}
-
-static void CalculateOutputLink(
- struct dml2_core_internal_scratch *s,
- double PHYCLK,
- double PHYCLKD18,
- double PHYCLKD32,
- double Downspreading,
- bool IsMainSurfaceUsingTheIndicatedTiming,
- enum dml2_output_encoder_class Output,
- enum dml2_output_format_class OutputFormat,
- unsigned int HTotal,
- unsigned int HActive,
- double PixelClockBackEnd,
- double ForcedOutputLinkBPP,
- unsigned int DSCInputBitPerComponent,
- unsigned int NumberOfDSCSlices,
- double AudioSampleRate,
- unsigned int AudioSampleLayout,
- enum dml2_odm_mode ODMModeNoDSC,
- enum dml2_odm_mode ODMModeDSC,
- enum dml2_dsc_enable_option DSCEnable,
- unsigned int OutputLinkDPLanes,
- enum dml2_output_link_dp_rate OutputLinkDPRate,
-
- // Output
- bool *RequiresDSC,
- bool *RequiresFEC,
- double *OutBpp,
- enum dml2_core_internal_output_type *OutputType,
- enum dml2_core_internal_output_type_rate *OutputRate,
- unsigned int *RequiredSlots)
-{
- bool LinkDSCEnable;
- unsigned int dummy;
- *RequiresDSC = false;
- *RequiresFEC = false;
- *OutBpp = 0;
-
- *OutputType = dml2_core_internal_output_type_unknown;
- *OutputRate = dml2_core_internal_output_rate_unknown;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DSCEnable = %u (dis, en, en_if_necessary)\n", __func__, DSCEnable);
- dml2_printf("DML::%s: IsMainSurfaceUsingTheIndicatedTiming = %u\n", __func__, IsMainSurfaceUsingTheIndicatedTiming);
- dml2_printf("DML::%s: PHYCLK = %f\n", __func__, PHYCLK);
- dml2_printf("DML::%s: PixelClockBackEnd = %f\n", __func__, PixelClockBackEnd);
- dml2_printf("DML::%s: AudioSampleRate = %f\n", __func__, AudioSampleRate);
- dml2_printf("DML::%s: HActive = %u\n", __func__, HActive);
- dml2_printf("DML::%s: HTotal = %u\n", __func__, HTotal);
- dml2_printf("DML::%s: ODMModeNoDSC = %u\n", __func__, ODMModeNoDSC);
- dml2_printf("DML::%s: ODMModeDSC = %u\n", __func__, ODMModeDSC);
- dml2_printf("DML::%s: ForcedOutputLinkBPP = %f\n", __func__, ForcedOutputLinkBPP);
- dml2_printf("DML::%s: Output (encoder) = %u\n", __func__, Output);
- dml2_printf("DML::%s: OutputLinkDPRate = %u\n", __func__, OutputLinkDPRate);
-#endif
- if (IsMainSurfaceUsingTheIndicatedTiming) {
- if (Output == dml2_hdmi) {
- *RequiresDSC = false;
- *RequiresFEC = false;
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, math_min2(600, PHYCLK) * 10, 3, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, false, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, &dummy);
- //OutputTypeAndRate = "HDMI";
- *OutputType = dml2_core_internal_output_type_hdmi;
- } else if (Output == dml2_dp || Output == dml2_dp2p0 || Output == dml2_edp) {
- if (DSCEnable == dml2_dsc_enable) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- if (Output == dml2_dp || Output == dml2_dp2p0) {
- *RequiresFEC = true;
- } else {
- *RequiresFEC = false;
- }
- } else {
- *RequiresDSC = false;
- LinkDSCEnable = false;
- if (Output == dml2_dp2p0) {
- *RequiresFEC = true;
- } else {
- *RequiresFEC = false;
- }
- }
- if (Output == dml2_dp2p0) {
- *OutBpp = 0;
- if ((OutputLinkDPRate == dml2_dp_rate_na || OutputLinkDPRate == dml2_dp_rate_uhbr10) && PHYCLKD32 >= 10000.0 / 32) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 10000, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
- if (*OutBpp == 0 && PHYCLKD32 < 13500.0 / 32 && DSCEnable == dml2_dsc_enable_if_necessary && ForcedOutputLinkBPP == 0) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 10000, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
- }
- //OutputTypeAndRate = Output & " UHBR10";
- *OutputType = dml2_core_internal_output_type_dp2p0;
- *OutputRate = dml2_core_internal_output_rate_dp_rate_uhbr10;
- }
- if ((OutputLinkDPRate == dml2_dp_rate_na || OutputLinkDPRate == dml2_dp_rate_uhbr13p5) && *OutBpp == 0 && PHYCLKD32 >= 13500.0 / 32) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 13500, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
-
- if (*OutBpp == 0 && PHYCLKD32 < 20000 / 32 && DSCEnable == dml2_dsc_enable_if_necessary && ForcedOutputLinkBPP == 0) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 13500, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
- }
- //OutputTypeAndRate = Output & " UHBR13p5";
- *OutputType = dml2_core_internal_output_type_dp2p0;
- *OutputRate = dml2_core_internal_output_rate_dp_rate_uhbr13p5;
- }
- if ((OutputLinkDPRate == dml2_dp_rate_na || OutputLinkDPRate == dml2_dp_rate_uhbr20) && *OutBpp == 0 && PHYCLKD32 >= 20000 / 32) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 20000, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
- if (*OutBpp == 0 && DSCEnable == dml2_dsc_enable_if_necessary && ForcedOutputLinkBPP == 0) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 20000, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
- }
- //OutputTypeAndRate = Output & " UHBR20";
- *OutputType = dml2_core_internal_output_type_dp2p0;
- *OutputRate = dml2_core_internal_output_rate_dp_rate_uhbr20;
- }
- } else { // output is dp or edp
- *OutBpp = 0;
- if ((OutputLinkDPRate == dml2_dp_rate_na || OutputLinkDPRate == dml2_dp_rate_hbr) && PHYCLK >= 270) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 2700, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
- if (*OutBpp == 0 && PHYCLK < 540 && DSCEnable == dml2_dsc_enable_if_necessary && ForcedOutputLinkBPP == 0) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- if (Output == dml2_dp) {
- *RequiresFEC = true;
- }
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 2700, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
- }
- //OutputTypeAndRate = Output & " HBR";
- *OutputType = (Output == dml2_dp) ? dml2_core_internal_output_type_dp : dml2_core_internal_output_type_edp;
- *OutputRate = dml2_core_internal_output_rate_dp_rate_hbr;
- }
- if ((OutputLinkDPRate == dml2_dp_rate_na || OutputLinkDPRate == dml2_dp_rate_hbr2) && *OutBpp == 0 && PHYCLK >= 540) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 5400, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
-
- if (*OutBpp == 0 && PHYCLK < 810 && DSCEnable == dml2_dsc_enable_if_necessary && ForcedOutputLinkBPP == 0) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- if (Output == dml2_dp) {
- *RequiresFEC = true;
- }
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 5400, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
- }
- //OutputTypeAndRate = Output & " HBR2";
- *OutputType = (Output == dml2_dp) ? dml2_core_internal_output_type_dp : dml2_core_internal_output_type_edp;
- *OutputRate = dml2_core_internal_output_rate_dp_rate_hbr2;
- }
- if ((OutputLinkDPRate == dml2_dp_rate_na || OutputLinkDPRate == dml2_dp_rate_hbr3) && *OutBpp == 0 && PHYCLK >= 810) { // VBA_ERROR, vba code doesn't have hbr3 check
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 8100, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
-
- if (*OutBpp == 0 && DSCEnable == dml2_dsc_enable_if_necessary && ForcedOutputLinkBPP == 0) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- if (Output == dml2_dp) {
- *RequiresFEC = true;
- }
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, (1 - Downspreading / 100) * 8100, OutputLinkDPLanes, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output,
- OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, RequiredSlots);
- }
- //OutputTypeAndRate = Output & " HBR3";
- *OutputType = (Output == dml2_dp) ? dml2_core_internal_output_type_dp : dml2_core_internal_output_type_edp;
- *OutputRate = dml2_core_internal_output_rate_dp_rate_hbr3;
- }
- }
- } else if (Output == dml2_hdmifrl) {
- if (DSCEnable == dml2_dsc_enable) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- *RequiresFEC = true;
- } else {
- *RequiresDSC = false;
- LinkDSCEnable = false;
- *RequiresFEC = false;
- }
- *OutBpp = 0;
- if (PHYCLKD18 >= 3000.0 / 18) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, 3000, 3, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output, OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, &dummy);
- //OutputTypeAndRate = Output & "3x3";
- *OutputType = dml2_core_internal_output_type_hdmifrl;
- *OutputRate = dml2_core_internal_output_rate_hdmi_rate_3x3;
- }
- if (*OutBpp == 0 && PHYCLKD18 >= 6000.0 / 18) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, 6000, 3, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output, OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, &dummy);
- //OutputTypeAndRate = Output & "6x3";
- *OutputType = dml2_core_internal_output_type_hdmifrl;
- *OutputRate = dml2_core_internal_output_rate_hdmi_rate_6x3;
- }
- if (*OutBpp == 0 && PHYCLKD18 >= 6000.0 / 18) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, 6000, 4, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output, OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, &dummy);
- //OutputTypeAndRate = Output & "6x4";
- *OutputType = dml2_core_internal_output_type_hdmifrl;
- *OutputRate = dml2_core_internal_output_rate_hdmi_rate_6x4;
- }
- if (*OutBpp == 0 && PHYCLKD18 >= 8000.0 / 18) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, 8000, 4, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output, OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, &dummy);
- //OutputTypeAndRate = Output & "8x4";
- *OutputType = dml2_core_internal_output_type_hdmifrl;
- *OutputRate = dml2_core_internal_output_rate_hdmi_rate_8x4;
- }
- if (*OutBpp == 0 && PHYCLKD18 >= 10000.0 / 18) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, 10000, 4, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output, OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, &dummy);
- if (*OutBpp == 0 && DSCEnable == dml2_dsc_enable_if_necessary && ForcedOutputLinkBPP == 0 && PHYCLKD18 < 12000.0 / 18) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- *RequiresFEC = true;
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, 10000, 4, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output, OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, &dummy);
- }
- //OutputTypeAndRate = Output & "10x4";
- *OutputType = dml2_core_internal_output_type_hdmifrl;
- *OutputRate = dml2_core_internal_output_rate_hdmi_rate_10x4;
- }
- if (*OutBpp == 0 && PHYCLKD18 >= 12000.0 / 18) {
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, 12000, 4, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output, OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, &dummy);
- if (*OutBpp == 0 && DSCEnable == dml2_dsc_enable_if_necessary && ForcedOutputLinkBPP == 0) {
- *RequiresDSC = true;
- LinkDSCEnable = true;
- *RequiresFEC = true;
- *OutBpp = TruncToValidBPP(&s->TruncToValidBPP_locals, 12000, 4, HTotal, HActive, PixelClockBackEnd, ForcedOutputLinkBPP, LinkDSCEnable, Output, OutputFormat, DSCInputBitPerComponent, NumberOfDSCSlices, (unsigned int)AudioSampleRate, AudioSampleLayout, ODMModeNoDSC, ODMModeDSC, &dummy);
- }
- //OutputTypeAndRate = Output & "12x4";
- *OutputType = dml2_core_internal_output_type_hdmifrl;
- *OutputRate = dml2_core_internal_output_rate_hdmi_rate_12x4;
- }
- }
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: RequiresDSC = %u\n", __func__, *RequiresDSC);
- dml2_printf("DML::%s: RequiresFEC = %u\n", __func__, *RequiresFEC);
- dml2_printf("DML::%s: OutBpp = %f\n", __func__, *OutBpp);
-#endif
-}
-
-static double CalculateWriteBackDISPCLK(
- enum dml2_source_format_class WritebackPixelFormat,
- double PixelClock,
- double WritebackHRatio,
- double WritebackVRatio,
- unsigned int WritebackHTaps,
- unsigned int WritebackVTaps,
- unsigned int WritebackSourceWidth,
- unsigned int WritebackDestinationWidth,
- unsigned int HTotal,
- unsigned int WritebackLineBufferSize)
-{
- double DISPCLK_H, DISPCLK_V, DISPCLK_HB;
-
- DISPCLK_H = PixelClock * math_ceil2((double)WritebackHTaps / 8.0, 1) / WritebackHRatio;
- DISPCLK_V = PixelClock * (WritebackVTaps * math_ceil2((double)WritebackDestinationWidth / 6.0, 1) + 8.0) / (double)HTotal;
- DISPCLK_HB = PixelClock * WritebackVTaps * (WritebackDestinationWidth * WritebackVTaps - WritebackLineBufferSize / 57.0) / 6.0 / (double)WritebackSourceWidth;
- return math_max3(DISPCLK_H, DISPCLK_V, DISPCLK_HB);
-}
-
-static double RequiredDTBCLK(
- bool DSCEnable,
- double PixelClock,
- enum dml2_output_format_class OutputFormat,
- double OutputBpp,
- unsigned int DSCSlices,
- unsigned int HTotal,
- unsigned int HActive,
- unsigned int AudioRate,
- unsigned int AudioLayout)
-{
- if (DSCEnable != true) {
- return math_max2(PixelClock / 4.0 * OutputBpp / 24.0, 25.0);
- } else {
- double PixelWordRate = PixelClock / (OutputFormat == dml2_444 ? 1 : 2);
- double HCActive = math_ceil2(DSCSlices * math_ceil2(OutputBpp * math_ceil2(HActive / DSCSlices, 1) / 8.0, 1) / 3.0, 1);
- double HCBlank = 64 + 32 * math_ceil2(AudioRate * (AudioLayout == 1 ? 1 : 0.25) * HTotal / (PixelClock * 1000), 1);
- double AverageTribyteRate = PixelWordRate * (HCActive + HCBlank) / HTotal;
- double HActiveTribyteRate = PixelWordRate * HCActive / HActive;
- return math_max4(PixelWordRate / 4.0, AverageTribyteRate / 4.0, HActiveTribyteRate / 4.0, 25.0) * 1.002;
- }
-}
-
-static unsigned int DSCDelayRequirement(
- bool DSCEnabled,
- enum dml2_odm_mode ODMMode,
- unsigned int DSCInputBitPerComponent,
- double OutputBpp,
- unsigned int HActive,
- unsigned int HTotal,
- unsigned int NumberOfDSCSlices,
- enum dml2_output_format_class OutputFormat,
- enum dml2_output_encoder_class Output,
- double PixelClock,
- double PixelClockBackEnd)
-{
- unsigned int DSCDelayRequirement_val = 0;
- unsigned int NumberOfDSCSlicesFactor = 1;
-
- if (DSCEnabled == true && OutputBpp != 0) {
-
- if (ODMMode == dml2_odm_mode_combine_4to1)
- NumberOfDSCSlicesFactor = 4;
- else if (ODMMode == dml2_odm_mode_combine_3to1)
- NumberOfDSCSlicesFactor = 3;
- else if (ODMMode == dml2_odm_mode_combine_2to1)
- NumberOfDSCSlicesFactor = 2;
-
- DSCDelayRequirement_val = NumberOfDSCSlicesFactor * (dscceComputeDelay(DSCInputBitPerComponent, OutputBpp, (unsigned int)(math_ceil2((double)HActive / (double)NumberOfDSCSlices, 1.0)),
- (NumberOfDSCSlices / NumberOfDSCSlicesFactor), OutputFormat, Output) + dscComputeDelay(OutputFormat, Output));
-
- DSCDelayRequirement_val = (unsigned int)(DSCDelayRequirement_val + (HTotal - HActive) * math_ceil2((double)DSCDelayRequirement_val / (double)HActive, 1.0));
- DSCDelayRequirement_val = (unsigned int)(DSCDelayRequirement_val * PixelClock / PixelClockBackEnd);
-
- } else {
- DSCDelayRequirement_val = 0;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DSCEnabled= %u\n", __func__, DSCEnabled);
- dml2_printf("DML::%s: ODMMode = %u\n", __func__, ODMMode);
- dml2_printf("DML::%s: OutputBpp = %f\n", __func__, OutputBpp);
- dml2_printf("DML::%s: HActive = %u\n", __func__, HActive);
- dml2_printf("DML::%s: HTotal = %u\n", __func__, HTotal);
- dml2_printf("DML::%s: PixelClock = %f\n", __func__, PixelClock);
- dml2_printf("DML::%s: PixelClockBackEnd = %f\n", __func__, PixelClockBackEnd);
- dml2_printf("DML::%s: OutputFormat = %u\n", __func__, OutputFormat);
- dml2_printf("DML::%s: DSCInputBitPerComponent = %u\n", __func__, DSCInputBitPerComponent);
- dml2_printf("DML::%s: NumberOfDSCSlices = %u\n", __func__, NumberOfDSCSlices);
- dml2_printf("DML::%s: DSCDelayRequirement_val = %u\n", __func__, DSCDelayRequirement_val);
-#endif
-
- return DSCDelayRequirement_val;
-}
-
-static void CalculateSurfaceSizeInMall(
- const struct dml2_display_cfg *display_cfg,
- unsigned int NumberOfActiveSurfaces,
- unsigned int MALLAllocatedForDCN,
- unsigned int BytesPerPixelY[],
- unsigned int BytesPerPixelC[],
- unsigned int Read256BytesBlockWidthY[],
- unsigned int Read256BytesBlockWidthC[],
- unsigned int Read256BytesBlockHeightY[],
- unsigned int Read256BytesBlockHeightC[],
- unsigned int ReadBlockWidthY[],
- unsigned int ReadBlockWidthC[],
- unsigned int ReadBlockHeightY[],
- unsigned int ReadBlockHeightC[],
-
- // Output
- unsigned int SurfaceSizeInMALL[],
- bool *ExceededMALLSize)
-{
- unsigned int TotalSurfaceSizeInMALLForSS = 0;
- unsigned int TotalSurfaceSizeInMALLForSubVP = 0;
- unsigned int MALLAllocatedForDCNInBytes = MALLAllocatedForDCN * 1024 * 1024;
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- const struct dml2_composition_cfg *composition = &display_cfg->plane_descriptors[k].composition;
- const struct dml2_surface_cfg *surface = &display_cfg->plane_descriptors[k].surface;
-
- if (composition->viewport.stationary) {
- SurfaceSizeInMALL[k] = (unsigned int)(math_min2(math_ceil2((double)surface->plane0.width, ReadBlockWidthY[k]),
- math_floor2(composition->viewport.plane0.x_start + composition->viewport.plane0.width + ReadBlockWidthY[k] - 1, ReadBlockWidthY[k]) -
- math_floor2((double)composition->viewport.plane0.x_start, ReadBlockWidthY[k])) *
- math_min2(math_ceil2((double)surface->plane0.height, ReadBlockHeightY[k]),
- math_floor2((double)composition->viewport.plane0.y_start + composition->viewport.plane0.height + ReadBlockHeightY[k] - 1, ReadBlockHeightY[k]) -
- math_floor2((double)composition->viewport.plane0.y_start, ReadBlockHeightY[k])) * BytesPerPixelY[k]);
-
- if (ReadBlockWidthC[k] > 0) {
- SurfaceSizeInMALL[k] = (unsigned int)(SurfaceSizeInMALL[k] +
- math_min2(math_ceil2((double)surface->plane1.width, ReadBlockWidthC[k]),
- math_floor2((double)composition->viewport.plane1.y_start + composition->viewport.plane1.width + ReadBlockWidthC[k] - 1, ReadBlockWidthC[k]) -
- math_floor2((double)composition->viewport.plane1.y_start, ReadBlockWidthC[k])) *
- math_min2(math_ceil2((double)surface->plane1.height, ReadBlockHeightC[k]),
- math_floor2((double)composition->viewport.plane1.y_start + composition->viewport.plane1.height + ReadBlockHeightC[k] - 1, ReadBlockHeightC[k]) -
- math_floor2(composition->viewport.plane1.y_start, ReadBlockHeightC[k])) * BytesPerPixelC[k]);
- }
- if (surface->dcc.enable) {
- SurfaceSizeInMALL[k] = (unsigned int)(SurfaceSizeInMALL[k] +
- math_min2(math_ceil2(surface->plane0.width, 8 * Read256BytesBlockWidthY[k]),
- math_floor2(composition->viewport.plane0.x_start + composition->viewport.plane0.width + 8 * Read256BytesBlockWidthY[k] - 1, 8 * Read256BytesBlockWidthY[k]) -
- math_floor2(composition->viewport.plane0.x_start, 8 * Read256BytesBlockWidthY[k])) *
- math_min2(math_ceil2(surface->plane0.height, 8 * Read256BytesBlockHeightY[k]),
- math_floor2(composition->viewport.plane0.y_start + composition->viewport.plane0.height + 8 * Read256BytesBlockHeightY[k] - 1, 8 * Read256BytesBlockHeightY[k]) -
- math_floor2(composition->viewport.plane0.y_start, 8 * Read256BytesBlockHeightY[k])) * BytesPerPixelY[k] / 256) + (64 * 1024);
- if (Read256BytesBlockWidthC[k] > 0) {
- SurfaceSizeInMALL[k] = (unsigned int)(SurfaceSizeInMALL[k] +
- math_min2(math_ceil2(surface->plane1.width, 8 * Read256BytesBlockWidthC[k]),
- math_floor2(composition->viewport.plane1.y_start + composition->viewport.plane1.width + 8 * Read256BytesBlockWidthC[k] - 1, 8 * Read256BytesBlockWidthC[k]) -
- math_floor2(composition->viewport.plane1.y_start, 8 * Read256BytesBlockWidthC[k])) *
- math_min2(math_ceil2(surface->plane1.height, 8 * Read256BytesBlockHeightC[k]),
- math_floor2(composition->viewport.plane1.y_start + composition->viewport.plane1.height + 8 * Read256BytesBlockHeightC[k] - 1, 8 * Read256BytesBlockHeightC[k]) -
- math_floor2(composition->viewport.plane1.y_start, 8 * Read256BytesBlockHeightC[k])) * BytesPerPixelC[k] / 256);
- }
- }
- } else {
- SurfaceSizeInMALL[k] = (unsigned int)(math_ceil2(math_min2(surface->plane0.width, composition->viewport.plane0.width + ReadBlockWidthY[k] - 1), ReadBlockWidthY[k]) *
- math_ceil2(math_min2(surface->plane0.height, composition->viewport.plane0.height + ReadBlockHeightY[k] - 1), ReadBlockHeightY[k]) * BytesPerPixelY[k]);
- if (ReadBlockWidthC[k] > 0) {
- SurfaceSizeInMALL[k] = (unsigned int)(SurfaceSizeInMALL[k] +
- math_ceil2(math_min2(surface->plane1.width, composition->viewport.plane1.width + ReadBlockWidthC[k] - 1), ReadBlockWidthC[k]) *
- math_ceil2(math_min2(surface->plane1.height, composition->viewport.plane1.height + ReadBlockHeightC[k] - 1), ReadBlockHeightC[k]) * BytesPerPixelC[k]);
- }
- if (surface->dcc.enable) {
- SurfaceSizeInMALL[k] = (unsigned int)(SurfaceSizeInMALL[k] +
- math_ceil2(math_min2(surface->plane0.width, composition->viewport.plane0.width + 8 * Read256BytesBlockWidthY[k] - 1), 8 * Read256BytesBlockWidthY[k]) *
- math_ceil2(math_min2(surface->plane0.height, composition->viewport.plane0.height + 8 * Read256BytesBlockHeightY[k] - 1), 8 * Read256BytesBlockHeightY[k]) * BytesPerPixelY[k] / 256) + (64 * 1024);
-
- if (Read256BytesBlockWidthC[k] > 0) {
- SurfaceSizeInMALL[k] = (unsigned int)(SurfaceSizeInMALL[k] +
- math_ceil2(math_min2(surface->plane1.width, composition->viewport.plane1.width + 8 * Read256BytesBlockWidthC[k] - 1), 8 * Read256BytesBlockWidthC[k]) *
- math_ceil2(math_min2(surface->plane1.height, composition->viewport.plane1.height + 8 * Read256BytesBlockHeightC[k] - 1), 8 * Read256BytesBlockHeightC[k]) * BytesPerPixelC[k] / 256);
- }
- }
- }
- }
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- /* SS and Subvp counted separate as they are never used at the same time */
- if (dml_is_phantom_pipe(&display_cfg->plane_descriptors[k]))
- TotalSurfaceSizeInMALLForSubVP += SurfaceSizeInMALL[k];
- else if (display_cfg->plane_descriptors[k].overrides.refresh_from_mall == dml2_refresh_from_mall_mode_override_force_enable)
- TotalSurfaceSizeInMALLForSS += SurfaceSizeInMALL[k];
- }
-
- *ExceededMALLSize = (TotalSurfaceSizeInMALLForSS > MALLAllocatedForDCNInBytes) ||
- (TotalSurfaceSizeInMALLForSubVP > MALLAllocatedForDCNInBytes);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: MALLAllocatedForDCN = %u\n", __func__, MALLAllocatedForDCN * 1024 * 1024);
- dml2_printf("DML::%s: TotalSurfaceSizeInMALLForSubVP = %u\n", __func__, TotalSurfaceSizeInMALLForSubVP);
- dml2_printf("DML::%s: TotalSurfaceSizeInMALLForSS = %u\n", __func__, TotalSurfaceSizeInMALLForSS);
- dml2_printf("DML::%s: ExceededMALLSize = %u\n", __func__, *ExceededMALLSize);
-#endif
-}
-
-static void calculate_tdlut_setting(
- struct dml2_core_internal_scratch *scratch,
- struct dml2_core_calcs_calculate_tdlut_setting_params *p)
-{
- if (!p->setup_for_tdlut) {
- *p->tdlut_groups_per_2row_ub = 0;
- *p->tdlut_opt_time = 0;
- *p->tdlut_drain_time = 0;
- *p->tdlut_bytes_per_group = 0;
- *p->tdlut_pte_bytes_per_frame = 0;
- *p->tdlut_bytes_per_frame = 0;
- return;
- }
-
- // locals
- unsigned int tdlut_bpe = 8;
- unsigned int tdlut_width;
- unsigned int tdlut_pitch_bytes;
- unsigned int tdlut_footprint_bytes;
- unsigned int vmpg_bytes;
- unsigned int tdlut_vmpg_per_frame;
- unsigned int tdlut_pte_req_per_frame;
- unsigned int tdlut_bytes_per_line;
- unsigned int tdlut_delivery_cycles;
- double tdlut_drain_rate;
- unsigned int tdlut_mpc_width;
- unsigned int tdlut_bytes_per_group_simple;
-
- if (p->tdlut_mpc_width_flag) {
- tdlut_mpc_width = 33;
- tdlut_bytes_per_group_simple = 39 * 256;
- } else {
- tdlut_mpc_width = 17;
- tdlut_bytes_per_group_simple = 10 * 256;
- }
-
- vmpg_bytes = p->gpuvm_page_size_kbytes * 1024;
-
- if (p->tdlut_addressing_mode == dml2_tdlut_simple_linear) {
- if (p->tdlut_width_mode == dml2_tdlut_width_17_cube)
- tdlut_width = 4916;
- else
- tdlut_width = 35940;
- } else {
- if (p->tdlut_width_mode == dml2_tdlut_width_17_cube)
- tdlut_width = 17;
- else // dml2_tdlut_width_33_cube
- tdlut_width = 33;
- }
-
- if (p->is_gfx11)
- tdlut_pitch_bytes = (unsigned int)math_ceil2(tdlut_width * tdlut_bpe, 256); //256B alignment
- else
- tdlut_pitch_bytes = (unsigned int)math_ceil2(tdlut_width * tdlut_bpe, 128); //128B alignment
-
- if (p->tdlut_addressing_mode == dml2_tdlut_sw_linear)
- tdlut_footprint_bytes = tdlut_pitch_bytes * tdlut_width * tdlut_width;
- else
- tdlut_footprint_bytes = tdlut_pitch_bytes;
-
- if (!p->gpuvm_enable) {
- tdlut_vmpg_per_frame = 0;
- tdlut_pte_req_per_frame = 0;
- } else {
- tdlut_vmpg_per_frame = (unsigned int)math_ceil2(tdlut_footprint_bytes - 1, vmpg_bytes) / vmpg_bytes + 1;
- tdlut_pte_req_per_frame = (unsigned int)math_ceil2(tdlut_vmpg_per_frame - 1, 8) / 8 + 1;
- }
- tdlut_bytes_per_line = (unsigned int)math_ceil2(tdlut_width * tdlut_bpe, 64); //64b request
- *p->tdlut_pte_bytes_per_frame = tdlut_pte_req_per_frame * 64;
-
- if (p->tdlut_addressing_mode == dml2_tdlut_sw_linear) {
- //the tdlut_width is either 17 or 33 but the 33x33x33 is subsampled every other line/slice
- *p->tdlut_bytes_per_frame = tdlut_bytes_per_line * tdlut_mpc_width * tdlut_mpc_width;
- *p->tdlut_bytes_per_group = tdlut_bytes_per_line * tdlut_mpc_width;
- //the delivery cycles is DispClk cycles per line * number of lines * number of slices
- tdlut_delivery_cycles = (unsigned int)math_ceil2(tdlut_mpc_width / 2.0, 1) * tdlut_mpc_width * tdlut_mpc_width;
- tdlut_drain_rate = tdlut_bytes_per_line * p->dispclk_mhz / math_ceil2(tdlut_mpc_width/2.0, 1);
- } else {
- //tdlut_addressing_mode = tdlut_simple_linear, 3dlut width should be 4*1229=4916 elements
- *p->tdlut_bytes_per_frame = (unsigned int)math_ceil2(tdlut_width * tdlut_bpe, 256);
- *p->tdlut_bytes_per_group = tdlut_bytes_per_group_simple;
- tdlut_delivery_cycles = (unsigned int)math_ceil2(tdlut_width / 2.0, 1);
- tdlut_drain_rate = 2 * tdlut_bpe * p->dispclk_mhz;
- }
-
- //the tdlut is fetched during the 2 row times of prefetch.
- if (p->setup_for_tdlut) {
- *p->tdlut_groups_per_2row_ub = (unsigned int)math_ceil2(*p->tdlut_bytes_per_frame / *p->tdlut_bytes_per_group, 1);
- *p->tdlut_opt_time = (*p->tdlut_bytes_per_frame - p->cursor_buffer_size * 1024) / tdlut_drain_rate;
- *p->tdlut_drain_time = p->cursor_buffer_size * 1024 / tdlut_drain_rate;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: gpuvm_enable = %d\n", __func__, p->gpuvm_enable);
- dml2_printf("DML::%s: vmpg_bytes = %d\n", __func__, vmpg_bytes);
- dml2_printf("DML::%s: tdlut_vmpg_per_frame = %d\n", __func__, tdlut_vmpg_per_frame);
- dml2_printf("DML::%s: tdlut_pte_req_per_frame = %d\n", __func__, tdlut_pte_req_per_frame);
- dml2_printf("DML::%s: dispclk_mhz = %f\n", __func__, p->dispclk_mhz);
- dml2_printf("DML::%s: tdlut_width = %u\n", __func__, tdlut_width);
- dml2_printf("DML::%s: tdlut_addressing_mode = %u\n", __func__, p->tdlut_addressing_mode);
- dml2_printf("DML::%s: tdlut_pitch_bytes = %u\n", __func__, tdlut_pitch_bytes);
- dml2_printf("DML::%s: tdlut_footprint_bytes = %u\n", __func__, tdlut_footprint_bytes);
- dml2_printf("DML::%s: tdlut_bytes_per_frame = %u\n", __func__, *p->tdlut_bytes_per_frame);
- dml2_printf("DML::%s: tdlut_bytes_per_line = %u\n", __func__, tdlut_bytes_per_line);
- dml2_printf("DML::%s: tdlut_bytes_per_group = %u\n", __func__, *p->tdlut_bytes_per_group);
- dml2_printf("DML::%s: tdlut_drain_rate = %f\n", __func__, tdlut_drain_rate);
- dml2_printf("DML::%s: tdlut_delivery_cycles = %u\n", __func__, tdlut_delivery_cycles);
- dml2_printf("DML::%s: tdlut_opt_time = %f\n", __func__, *p->tdlut_opt_time);
- dml2_printf("DML::%s: tdlut_drain_time = %f\n", __func__, *p->tdlut_drain_time);
- dml2_printf("DML::%s: tdlut_groups_per_2row_ub = %d\n", __func__, *p->tdlut_groups_per_2row_ub);
-#endif
-}
-
-static void CalculateTarb(
- const struct dml2_display_cfg *display_cfg,
- unsigned int PixelChunkSizeInKByte,
- unsigned int NumberOfActiveSurfaces,
- unsigned int NumberOfDPP[],
- unsigned int dpte_group_bytes[],
- unsigned int tdlut_bytes_per_group[],
- double HostVMInefficiencyFactor,
- double HostVMInefficiencyFactorPrefetch,
- unsigned int HostVMMinPageSize,
- double ReturnBW,
- unsigned int MetaChunkSize,
-
- // output
- double *Tarb,
- double *Tarb_prefetch)
-{
- double extra_bytes = 0;
- double extra_bytes_prefetch = 0;
- double HostVMDynamicLevels = CalculateHostVMDynamicLevels(display_cfg->gpuvm_enable, display_cfg->hostvm_enable, HostVMMinPageSize, display_cfg->hostvm_max_non_cached_page_table_levels);
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- extra_bytes = extra_bytes + (NumberOfDPP[k] * PixelChunkSizeInKByte * 1024);
-
- if (display_cfg->plane_descriptors[k].surface.dcc.enable)
- extra_bytes = extra_bytes + (MetaChunkSize * 1024);
-
- if (display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut)
- extra_bytes = extra_bytes + tdlut_bytes_per_group[k];
- }
-
- extra_bytes_prefetch = extra_bytes;
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- if (display_cfg->gpuvm_enable == true) {
- extra_bytes = extra_bytes + NumberOfDPP[k] * dpte_group_bytes[k] * (1 + 8 * HostVMDynamicLevels) * HostVMInefficiencyFactor;
- extra_bytes_prefetch = extra_bytes_prefetch + NumberOfDPP[k] * dpte_group_bytes[k] * (1 + 8 * HostVMDynamicLevels) * HostVMInefficiencyFactorPrefetch;
- }
- }
- *Tarb = extra_bytes / ReturnBW;
- *Tarb_prefetch = extra_bytes_prefetch / ReturnBW;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: PixelChunkSizeInKByte = %d\n", __func__, PixelChunkSizeInKByte);
- dml2_printf("DML::%s: MetaChunkSize = %d\n", __func__, MetaChunkSize);
- dml2_printf("DML::%s: extra_bytes = %f\n", __func__, extra_bytes);
- dml2_printf("DML::%s: extra_bytes_prefetch = %f\n", __func__, extra_bytes_prefetch);
-#endif
-}
-
-static double CalculateTWait(
- long reserved_vblank_time_ns,
- double UrgentLatency,
- double Ttrip)
-{
- double TWait;
- double t_urg_trip = math_max2(UrgentLatency, Ttrip);
- TWait = reserved_vblank_time_ns / 1000.0 + t_urg_trip;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: reserved_vblank_time_ns = %d\n", __func__, reserved_vblank_time_ns);
- dml2_printf("DML::%s: UrgentLatency = %f\n", __func__, UrgentLatency);
- dml2_printf("DML::%s: Ttrip = %f\n", __func__, Ttrip);
- dml2_printf("DML::%s: TWait = %f\n", __func__, TWait);
-#endif
- return TWait;
-}
-
-
-static void CalculateVUpdateAndDynamicMetadataParameters(
- unsigned int MaxInterDCNTileRepeaters,
- double Dppclk,
- double Dispclk,
- double DCFClkDeepSleep,
- double PixelClock,
- unsigned int HTotal,
- unsigned int VBlank,
- unsigned int DynamicMetadataTransmittedBytes,
- unsigned int DynamicMetadataLinesBeforeActiveRequired,
- unsigned int InterlaceEnable,
- bool ProgressiveToInterlaceUnitInOPP,
-
- // Output
- double *TSetup,
- double *Tdmbf,
- double *Tdmec,
- double *Tdmsks,
- unsigned int *VUpdateOffsetPix,
- unsigned int *VUpdateWidthPix,
- unsigned int *VReadyOffsetPix)
-{
- double TotalRepeaterDelayTime;
- TotalRepeaterDelayTime = MaxInterDCNTileRepeaters * (2 / Dppclk + 3 / Dispclk);
- *VUpdateWidthPix = (unsigned int)(math_ceil2((14.0 / DCFClkDeepSleep + 12.0 / Dppclk + TotalRepeaterDelayTime) * PixelClock, 1.0));
- *VReadyOffsetPix = (unsigned int)(math_ceil2(math_max2(150.0 / Dppclk, TotalRepeaterDelayTime + 20.0 / DCFClkDeepSleep + 10.0 / Dppclk) * PixelClock, 1.0));
- *VUpdateOffsetPix = (unsigned int)(math_ceil2(HTotal / 4.0, 1.0));
- *TSetup = (*VUpdateOffsetPix + *VUpdateWidthPix + *VReadyOffsetPix) / PixelClock;
- *Tdmbf = DynamicMetadataTransmittedBytes / 4.0 / Dispclk;
- *Tdmec = HTotal / PixelClock;
-
- if (DynamicMetadataLinesBeforeActiveRequired == 0) {
- *Tdmsks = VBlank * HTotal / PixelClock / 2.0;
- } else {
- *Tdmsks = DynamicMetadataLinesBeforeActiveRequired * HTotal / PixelClock;
- }
- if (InterlaceEnable == 1 && ProgressiveToInterlaceUnitInOPP == false) {
- *Tdmsks = *Tdmsks / 2;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DynamicMetadataLinesBeforeActiveRequired = %u\n", __func__, DynamicMetadataLinesBeforeActiveRequired);
- dml2_printf("DML::%s: VBlank = %u\n", __func__, VBlank);
- dml2_printf("DML::%s: HTotal = %u\n", __func__, HTotal);
- dml2_printf("DML::%s: PixelClock = %f\n", __func__, PixelClock);
- dml2_printf("DML::%s: Dppclk = %f\n", __func__, Dppclk);
- dml2_printf("DML::%s: DCFClkDeepSleep = %f\n", __func__, DCFClkDeepSleep);
- dml2_printf("DML::%s: MaxInterDCNTileRepeaters = %u\n", __func__, MaxInterDCNTileRepeaters);
- dml2_printf("DML::%s: TotalRepeaterDelayTime = %f\n", __func__, TotalRepeaterDelayTime);
-
- dml2_printf("DML::%s: VUpdateWidthPix = %u\n", __func__, *VUpdateWidthPix);
- dml2_printf("DML::%s: VReadyOffsetPix = %u\n", __func__, *VReadyOffsetPix);
- dml2_printf("DML::%s: VUpdateOffsetPix = %u\n", __func__, *VUpdateOffsetPix);
-
- dml2_printf("DML::%s: Tdmsks = %f\n", __func__, *Tdmsks);
-#endif
-}
-
-static double get_urgent_bandwidth_required(
- struct dml2_core_shared_get_urgent_bandwidth_required_locals *l,
- const struct dml2_display_cfg *display_cfg,
- enum dml2_core_internal_soc_state_type state_type,
- enum dml2_core_internal_bw_type bw_type,
- bool inc_flip_bw, // including flip bw
- unsigned int NumberOfActiveSurfaces,
- unsigned int NumberOfDPP[],
- double dcc_dram_bw_nom_overhead_factor_p0[],
- double dcc_dram_bw_nom_overhead_factor_p1[],
- double dcc_dram_bw_pref_overhead_factor_p0[],
- double dcc_dram_bw_pref_overhead_factor_p1[],
- double mall_prefetch_sdp_overhead_factor[],
- double mall_prefetch_dram_overhead_factor[],
- double ReadBandwidthLuma[],
- double ReadBandwidthChroma[],
- double PrefetchBandwidthLuma[],
- double PrefetchBandwidthChroma[],
- double cursor_bw[],
- double dpte_row_bw[],
- double meta_row_bw[],
- double prefetch_cursor_bw[],
- double prefetch_vmrow_bw[],
- double flip_bw[],
- double UrgentBurstFactorLuma[],
- double UrgentBurstFactorChroma[],
- double UrgentBurstFactorCursor[],
- double UrgentBurstFactorLumaPre[],
- double UrgentBurstFactorChromaPre[],
- double UrgentBurstFactorCursorPre[])
-{
- memset(l, 0, sizeof(struct dml2_core_shared_get_urgent_bandwidth_required_locals));
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- l->mall_svp_prefetch_factor = (state_type == dml2_core_internal_soc_state_svp_prefetch) ? (bw_type == dml2_core_internal_bw_dram ? mall_prefetch_dram_overhead_factor[k] : mall_prefetch_sdp_overhead_factor[k]) : 1.0;
- l->tmp_nom_adj_factor_p0 = (bw_type == dml2_core_internal_bw_dram ? dcc_dram_bw_nom_overhead_factor_p0[k] : 1.0) * l->mall_svp_prefetch_factor;
- l->tmp_nom_adj_factor_p1 = (bw_type == dml2_core_internal_bw_dram ? dcc_dram_bw_nom_overhead_factor_p1[k] : 1.0) * l->mall_svp_prefetch_factor;
- l->tmp_pref_adj_factor_p0 = (bw_type == dml2_core_internal_bw_dram ? dcc_dram_bw_pref_overhead_factor_p0[k] : 1.0) * l->mall_svp_prefetch_factor;
- l->tmp_pref_adj_factor_p1 = (bw_type == dml2_core_internal_bw_dram ? dcc_dram_bw_pref_overhead_factor_p1[k] : 1.0) * l->mall_svp_prefetch_factor;
-
- l->adj_factor_p0 = UrgentBurstFactorLuma[k] * l->tmp_nom_adj_factor_p0;
- l->adj_factor_p1 = UrgentBurstFactorChroma[k] * l->tmp_nom_adj_factor_p1;
- l->adj_factor_cur = UrgentBurstFactorCursor[k];
- l->adj_factor_p0_pre = UrgentBurstFactorLumaPre[k] * l->tmp_pref_adj_factor_p0;
- l->adj_factor_p1_pre = UrgentBurstFactorChromaPre[k] * l->tmp_pref_adj_factor_p1;
- l->adj_factor_cur_pre = UrgentBurstFactorCursorPre[k];
-
- // both dchub_urgent_bw_at_sdp_noflip and dchub_urgent_bw_at_dram_noflip don't include the phantom_pipe because iflips dont occur while phantom_pipe is active
- bool is_phantom = dml_is_phantom_pipe(&display_cfg->plane_descriptors[k]);
- bool exclude_this_plane = 0;
-
- // Exclude phantom pipe in bw calculation for non svp prefetch state
- if (state_type != dml2_core_internal_soc_state_svp_prefetch && is_phantom)
- exclude_this_plane = 1;
-
- if (display_cfg->plane_descriptors[k].immediate_flip == false || !inc_flip_bw)
- l->per_plane_flip_bw[k] = NumberOfDPP[k] * (dpte_row_bw[k] + meta_row_bw[k]);
- else
- l->per_plane_flip_bw[k] = NumberOfDPP[k] * flip_bw[k];
-
-
- if (!exclude_this_plane) {
- l->required_bandwidth_mbps_this_surface = math_max3(NumberOfDPP[k] * prefetch_vmrow_bw[k],
- l->per_plane_flip_bw[k] + ReadBandwidthLuma[k] * l->adj_factor_p0 + ReadBandwidthChroma[k] * l->adj_factor_p1 + cursor_bw[k] * l->adj_factor_cur,
- l->per_plane_flip_bw[k] + NumberOfDPP[k] * (PrefetchBandwidthLuma[k] * l->adj_factor_p0_pre + PrefetchBandwidthChroma[k] * l->adj_factor_p1_pre) + prefetch_cursor_bw[k] * l->adj_factor_cur_pre);
-
- l->required_bandwidth_mbps = l->required_bandwidth_mbps + l->required_bandwidth_mbps_this_surface;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, NumberOfDPP=%d\n", __func__, k, NumberOfDPP[k]);
- dml2_printf("DML::%s: k=%d, mall_svp_prefetch_factor=%f\n", __func__, k, l->mall_svp_prefetch_factor);
- dml2_printf("DML::%s: k=%d, adj_factor_p0=%f\n", __func__, k, l->adj_factor_p0);
- dml2_printf("DML::%s: k=%d, adj_factor_p1=%f\n", __func__, k, l->adj_factor_p1);
- dml2_printf("DML::%s: k=%d, adj_factor_cur=%f\n", __func__, k, l->adj_factor_cur);
-
- dml2_printf("DML::%s: k=%d, adj_factor_p0_pre=%f\n", __func__, k, l->adj_factor_p0_pre);
- dml2_printf("DML::%s: k=%d, adj_factor_p1_pre=%f\n", __func__, k, l->adj_factor_p1_pre);
- dml2_printf("DML::%s: k=%d, adj_factor_cur_pre=%f\n", __func__, k, l->adj_factor_cur_pre);
-
- dml2_printf("DML::%s: k=%d, per_plane_flip_bw=%f\n", __func__, k, l->per_plane_flip_bw[k]);
- dml2_printf("DML::%s: k=%d, prefetch_vmrow_bw=%f\n", __func__, k, prefetch_vmrow_bw[k]);
- dml2_printf("DML::%s: k=%d, ReadBandwidthLuma=%f\n", __func__, k, ReadBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%d, ReadBandwidthChroma=%f\n", __func__, k, ReadBandwidthChroma[k]);
- dml2_printf("DML::%s: k=%d, cursor_bw=%f\n", __func__, k, cursor_bw[k]);
-
- dml2_printf("DML::%s: k=%d, meta_row_bw=%f\n", __func__, k, meta_row_bw[k]);
- dml2_printf("DML::%s: k=%d, dpte_row_bw=%f\n", __func__, k, dpte_row_bw[k]);
- dml2_printf("DML::%s: k=%d, PrefetchBandwidthLuma=%f\n", __func__, k, PrefetchBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%d, PrefetchBandwidthChroma=%f\n", __func__, k, PrefetchBandwidthChroma[k]);
- dml2_printf("DML::%s: k=%d, prefetch_cursor_bw=%f\n", __func__, k, prefetch_cursor_bw[k]);
- dml2_printf("DML::%s: k=%d, required_bandwidth_mbps=%f (total), inc_flip_bw=%d, is_phantom=%d exclude_this_plane=%d\n", __func__, k, l->required_bandwidth_mbps, inc_flip_bw, is_phantom, exclude_this_plane);
-#endif
- }
-
- return l->required_bandwidth_mbps;
-}
-
-static void CalculateExtraLatency(
- const struct dml2_display_cfg *display_cfg,
- unsigned int ROBBufferSizeInKByte,
- unsigned int RoundTripPingLatencyCycles,
- unsigned int ReorderingBytes,
- double DCFCLK,
- double FabricClock,
- unsigned int PixelChunkSizeInKByte,
- double ReturnBW,
- unsigned int NumberOfActiveSurfaces,
- unsigned int NumberOfDPP[],
- unsigned int dpte_group_bytes[],
- unsigned int tdlut_bytes_per_group[],
- double HostVMInefficiencyFactor,
- double HostVMInefficiencyFactorPrefetch,
- unsigned int HostVMMinPageSize,
- enum dml2_qos_param_type qos_type,
- bool max_oustanding_when_urgent_expected,
- unsigned int max_outstanding_requests,
- unsigned int request_size_bytes_luma[],
- unsigned int request_size_bytes_chroma[],
- unsigned int MetaChunkSize,
- unsigned int dchub_arb_to_ret_delay,
- double Ttrip,
- unsigned int hostvm_mode,
-
- // output
- double *ExtraLatency,
- double *ExtraLatency_sr,
- double *ExtraLatencyPrefetch)
-{
- double Tarb;
- double Tarb_prefetch;
-
- CalculateTarb(
- display_cfg,
- PixelChunkSizeInKByte,
- NumberOfActiveSurfaces,
- NumberOfDPP,
- dpte_group_bytes,
- tdlut_bytes_per_group,
- HostVMInefficiencyFactor,
- HostVMInefficiencyFactorPrefetch,
- HostVMMinPageSize,
- ReturnBW,
- MetaChunkSize,
- // output
- &Tarb,
- &Tarb_prefetch);
-
- unsigned int max_request_size_bytes = 0;
- double Tex_trips = (display_cfg->hostvm_enable && hostvm_mode == 1) ? (2.0 * Ttrip) : 0.0;
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- if (request_size_bytes_luma[k] > max_request_size_bytes)
- max_request_size_bytes = request_size_bytes_luma[k];
- if (request_size_bytes_chroma[k] > max_request_size_bytes)
- max_request_size_bytes = request_size_bytes_chroma[k];
- }
-
- if (qos_type == dml2_qos_param_type_dcn4x) {
- *ExtraLatency_sr = dchub_arb_to_ret_delay / DCFCLK;
- *ExtraLatency = *ExtraLatency_sr;
- if (max_oustanding_when_urgent_expected)
- *ExtraLatency = *ExtraLatency + (ROBBufferSizeInKByte * 1024 - max_outstanding_requests * max_request_size_bytes) / ReturnBW;
- } else {
- *ExtraLatency_sr = dchub_arb_to_ret_delay / DCFCLK + RoundTripPingLatencyCycles / FabricClock + ReorderingBytes / ReturnBW;
- *ExtraLatency = *ExtraLatency_sr;
- }
- *ExtraLatency = *ExtraLatency + Tex_trips;
- *ExtraLatencyPrefetch = *ExtraLatency + Tarb_prefetch;
- *ExtraLatency = *ExtraLatency + Tarb;
- *ExtraLatency_sr = *ExtraLatency_sr + Tarb;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: qos_type=%u\n", __func__, qos_type);
- dml2_printf("DML::%s: hostvm_mode=%u\n", __func__, hostvm_mode);
- dml2_printf("DML::%s: Tex_trips=%u\n", __func__, Tex_trips);
- dml2_printf("DML::%s: max_oustanding_when_urgent_expected=%u\n", __func__, max_oustanding_when_urgent_expected);
- dml2_printf("DML::%s: FabricClock=%f\n", __func__, FabricClock);
- dml2_printf("DML::%s: DCFCLK=%f\n", __func__, DCFCLK);
- dml2_printf("DML::%s: ReturnBW=%f\n", __func__, ReturnBW);
- dml2_printf("DML::%s: RoundTripPingLatencyCycles=%u\n", __func__, RoundTripPingLatencyCycles);
- dml2_printf("DML::%s: ReorderingBytes=%u\n", __func__, ReorderingBytes);
- dml2_printf("DML::%s: Tarb=%f\n", __func__, Tarb);
- dml2_printf("DML::%s: ExtraLatency=%f\n", __func__, *ExtraLatency);
- dml2_printf("DML::%s: ExtraLatency_sr=%f\n", __func__, *ExtraLatency_sr);
- dml2_printf("DML::%s: ExtraLatencyPrefetch=%f\n", __func__, *ExtraLatencyPrefetch);
-#endif
-}
-
-static bool CalculatePrefetchSchedule(struct dml2_core_internal_scratch *scratch, struct dml2_core_calcs_CalculatePrefetchSchedule_params *p)
-{
- struct dml2_core_calcs_CalculatePrefetchSchedule_locals *s = &scratch->CalculatePrefetchSchedule_locals;
-
- s->NoTimeToPrefetch = false;
- s->DPPCycles = 0;
- s->DISPCLKCycles = 0;
- s->DSTTotalPixelsAfterScaler = 0.0;
- s->LineTime = 0.0;
- s->dst_y_prefetch_equ = 0.0;
- s->prefetch_bw_oto = 0.0;
- s->Tvm_oto = 0.0;
- s->Tr0_oto = 0.0;
- s->Tvm_oto_lines = 0.0;
- s->Tr0_oto_lines = 0.0;
- s->dst_y_prefetch_oto = 0.0;
- s->TimeForFetchingVM = 0.0;
- s->TimeForFetchingRowInVBlank = 0.0;
- s->LinesToRequestPrefetchPixelData = 0.0;
- s->HostVMDynamicLevelsTrips = 0;
- s->trip_to_mem = 0.0;
- *p->Tvm_trips = 0.0;
- *p->Tr0_trips = 0.0;
- s->Tvm_trips_rounded = 0.0;
- s->Tr0_trips_rounded = 0.0;
- s->max_Tsw = 0.0;
- s->Lsw_oto = 0.0;
- s->Tpre_rounded = 0.0;
- s->prefetch_bw_equ = 0.0;
- s->Tvm_equ = 0.0;
- s->Tr0_equ = 0.0;
- s->Tdmbf = 0.0;
- s->Tdmec = 0.0;
- s->Tdmsks = 0.0;
- s->prefetch_sw_bytes = 0.0;
- s->prefetch_bw_pr = 0.0;
- s->bytes_pp = 0.0;
- s->dep_bytes = 0.0;
- s->min_Lsw_oto = 0.0;
- s->Tsw_est1 = 0.0;
- s->Tsw_est3 = 0.0;
- s->cursor_prefetch_bytes = 0;
- *p->prefetch_cursor_bw = 0;
- bool dcc_mrq_enable = (p->dcc_enable && p->mrq_present);
-
- s->TWait_p = p->TWait - p->Ttrip; // TWait includes max(Turg, Ttrip)
-
- if (p->display_cfg->gpuvm_enable == true && p->display_cfg->hostvm_enable == true) {
- s->HostVMDynamicLevelsTrips = p->display_cfg->hostvm_max_non_cached_page_table_levels;
- } else {
- s->HostVMDynamicLevelsTrips = 0;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: dcc_enable = %u\n", __func__, p->dcc_enable);
- dml2_printf("DML::%s: mrq_present = %u\n", __func__, p->mrq_present);
- dml2_printf("DML::%s: dcc_mrq_enable = %u\n", __func__, dcc_mrq_enable);
- dml2_printf("DML::%s: GPUVMEnable = %u\n", __func__, p->display_cfg->gpuvm_enable);
- dml2_printf("DML::%s: GPUVMPageTableLevels = %u\n", __func__, p->display_cfg->gpuvm_max_page_table_levels);
- dml2_printf("DML::%s: DCCEnable = %u\n", __func__, p->myPipe->DCCEnable);
- dml2_printf("DML::%s: VStartup = %u\n", __func__, p->VStartup);
- dml2_printf("DML::%s: MaxVStartup = %u\n", __func__, p->MaxVStartup);
- dml2_printf("DML::%s: HostVMEnable = %u\n", __func__, p->display_cfg->hostvm_enable);
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, p->HostVMInefficiencyFactor);
- dml2_printf("DML::%s: TWait = %f\n", __func__, p->TWait);
- dml2_printf("DML::%s: TWait_p = %f\n", __func__, s->TWait_p);
- dml2_printf("DML::%s: Ttrip = %f\n", __func__, p->Ttrip);
- dml2_printf("DML::%s: myPipe->Dppclk = %f\n", __func__, p->myPipe->Dppclk);
- dml2_printf("DML::%s: myPipe->Dispclk = %f\n", __func__, p->myPipe->Dispclk);
-#endif
- CalculateVUpdateAndDynamicMetadataParameters(
- p->MaxInterDCNTileRepeaters,
- p->myPipe->Dppclk,
- p->myPipe->Dispclk,
- p->myPipe->DCFClkDeepSleep,
- p->myPipe->PixelClock,
- p->myPipe->HTotal,
- p->myPipe->VBlank,
- p->DynamicMetadataTransmittedBytes,
- p->DynamicMetadataLinesBeforeActiveRequired,
- p->myPipe->InterlaceEnable,
- p->myPipe->ProgressiveToInterlaceUnitInOPP,
- p->TSetup,
-
- // Output
- &s->Tdmbf,
- &s->Tdmec,
- &s->Tdmsks,
- p->VUpdateOffsetPix,
- p->VUpdateWidthPix,
- p->VReadyOffsetPix);
-
- s->LineTime = p->myPipe->HTotal / p->myPipe->PixelClock;
- s->trip_to_mem = p->Ttrip;
- *p->Tvm_trips = p->ExtraLatencyPrefetch + s->trip_to_mem * (p->display_cfg->gpuvm_max_page_table_levels * (s->HostVMDynamicLevelsTrips + 1));
- if (dcc_mrq_enable)
- *p->Tvm_trips_flip = *p->Tvm_trips;
- else
- *p->Tvm_trips_flip = *p->Tvm_trips - s->trip_to_mem;
- *p->Tr0_trips_flip = s->trip_to_mem * (s->HostVMDynamicLevelsTrips + 1);
- *p->Tr0_trips = math_max2(*p->Tr0_trips_flip, p->tdlut_opt_time / 2);
-
- if (p->DynamicMetadataVMEnabled == true) {
- *p->Tdmdl_vm = s->TWait_p + *p->Tvm_trips;
- *p->Tdmdl = *p->Tdmdl_vm + p->Ttrip;
- } else {
- *p->Tdmdl_vm = 0;
- *p->Tdmdl = p->TWait + p->ExtraLatencyPrefetch; // Tex
- }
-
- if (p->DynamicMetadataEnable == true) {
- if (p->VStartup * s->LineTime < *p->TSetup + *p->Tdmdl + s->Tdmbf + s->Tdmec + s->Tdmsks) {
- *p->NotEnoughTimeForDynamicMetadata = true;
- dml2_printf("DML::%s: Not Enough Time for Dynamic Meta!\n", __func__);
- dml2_printf("DML::%s: Tdmbf: %fus - time for dmd transfer from dchub to dio output buffer\n", __func__, s->Tdmbf);
- dml2_printf("DML::%s: Tdmec: %fus - time dio takes to transfer dmd\n", __func__, s->Tdmec);
- dml2_printf("DML::%s: Tdmsks: %fus - time before active dmd must complete transmission at dio\n", __func__, s->Tdmsks);
- dml2_printf("DML::%s: Tdmdl: %fus - time for fabric to become ready and fetch dmd \n", __func__, *p->Tdmdl);
- } else {
- *p->NotEnoughTimeForDynamicMetadata = false;
- }
- } else {
- *p->NotEnoughTimeForDynamicMetadata = false;
- }
-
- if (p->myPipe->ScalerEnabled)
- s->DPPCycles = (unsigned int)(p->DPPCLKDelaySubtotalPlusCNVCFormater + p->DPPCLKDelaySCL);
- else
- s->DPPCycles = (unsigned int)(p->DPPCLKDelaySubtotalPlusCNVCFormater + p->DPPCLKDelaySCLLBOnly);
-
- s->DPPCycles = (unsigned int)(s->DPPCycles + p->myPipe->NumberOfCursors * p->DPPCLKDelayCNVCCursor);
-
- s->DISPCLKCycles = (unsigned int)p->DISPCLKDelaySubtotal;
-
- if (p->myPipe->Dppclk == 0.0 || p->myPipe->Dispclk == 0.0)
- return true;
-
- *p->DSTXAfterScaler = (unsigned int)math_round(s->DPPCycles * p->myPipe->PixelClock / p->myPipe->Dppclk + s->DISPCLKCycles * p->myPipe->PixelClock / p->myPipe->Dispclk + p->DSCDelay);
- *p->DSTXAfterScaler = (unsigned int)math_round(*p->DSTXAfterScaler + (p->myPipe->ODMMode != dml2_odm_mode_bypass ? 18 : 0) + (p->myPipe->DPPPerSurface - 1) * p->DPP_RECOUT_WIDTH +
- ((p->myPipe->ODMMode == dml2_odm_mode_split_1to2 || p->myPipe->ODMMode == dml2_odm_mode_mso_1to2) ? (double)p->myPipe->HActive / 2.0 : 0) +
- ((p->myPipe->ODMMode == dml2_odm_mode_mso_1to4) ? (double)p->myPipe->HActive * 3.0 / 4.0 : 0));
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DynamicMetadataVMEnabled = %u\n", __func__, p->DynamicMetadataVMEnabled);
- dml2_printf("DML::%s: DPPCycles = %u\n", __func__, s->DPPCycles);
- dml2_printf("DML::%s: PixelClock = %f\n", __func__, p->myPipe->PixelClock);
- dml2_printf("DML::%s: Dppclk = %f\n", __func__, p->myPipe->Dppclk);
- dml2_printf("DML::%s: DISPCLKCycles = %u\n", __func__, s->DISPCLKCycles);
- dml2_printf("DML::%s: DISPCLK = %f\n", __func__, p->myPipe->Dispclk);
- dml2_printf("DML::%s: DSCDelay = %u\n", __func__, p->DSCDelay);
- dml2_printf("DML::%s: ODMMode = %u\n", __func__, p->myPipe->ODMMode);
- dml2_printf("DML::%s: DPP_RECOUT_WIDTH = %u\n", __func__, p->DPP_RECOUT_WIDTH);
- dml2_printf("DML::%s: DSTXAfterScaler = %u\n", __func__, *p->DSTXAfterScaler);
-
- dml2_printf("DML::%s: setup_for_tdlut = %u\n", __func__, p->setup_for_tdlut);
- dml2_printf("DML::%s: tdlut_opt_time = %f\n", __func__, p->tdlut_opt_time);
- dml2_printf("DML::%s: tdlut_pte_bytes_per_frame = %u\n", __func__, p->tdlut_pte_bytes_per_frame);
-#endif
-
- if (p->OutputFormat == dml2_420 || (p->myPipe->InterlaceEnable && p->myPipe->ProgressiveToInterlaceUnitInOPP))
- *p->DSTYAfterScaler = 1;
- else
- *p->DSTYAfterScaler = 0;
-
- s->DSTTotalPixelsAfterScaler = *p->DSTYAfterScaler * p->myPipe->HTotal + *p->DSTXAfterScaler;
- *p->DSTYAfterScaler = (unsigned int)(math_floor2(s->DSTTotalPixelsAfterScaler / p->myPipe->HTotal, 1));
- *p->DSTXAfterScaler = (unsigned int)(s->DSTTotalPixelsAfterScaler - ((double)(*p->DSTYAfterScaler * p->myPipe->HTotal)));
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DSTXAfterScaler = %u (final)\n", __func__, *p->DSTXAfterScaler);
- dml2_printf("DML::%s: DSTYAfterScaler = %u (final)\n", __func__, *p->DSTYAfterScaler);
-#endif
-
- s->NoTimeToPrefetch = false;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tr0_trips = %f\n", __func__, *p->Tr0_trips);
- dml2_printf("DML::%s: Tvm_trips = %f\n", __func__, *p->Tvm_trips);
- dml2_printf("DML::%s: trip_to_mem = %f\n", __func__, s->trip_to_mem);
- dml2_printf("DML::%s: ExtraLatencyPrefetch = %f\n", __func__, p->ExtraLatencyPrefetch);
- dml2_printf("DML::%s: GPUVMPageTableLevels = %u\n", __func__, p->display_cfg->gpuvm_max_page_table_levels);
- dml2_printf("DML::%s: HostVMDynamicLevelsTrips = %u\n", __func__, s->HostVMDynamicLevelsTrips);
-#endif
- if (p->display_cfg->gpuvm_enable) {
- s->Tvm_trips_rounded = math_ceil2(4.0 * *p->Tvm_trips / s->LineTime, 1.0) / 4.0 * s->LineTime;
- *p->Tvm_trips_flip_rounded = math_ceil2(4.0 * *p->Tvm_trips_flip / s->LineTime, 1.0) / 4.0 * s->LineTime;
- } else {
- s->Tvm_trips_rounded = s->LineTime / 4.0;
- *p->Tvm_trips_flip_rounded = s->LineTime / 4.0;
- }
- s->Tvm_trips_rounded = math_max2(s->Tvm_trips_rounded, s->LineTime / 4.0);
- *p->Tvm_trips_flip_rounded = math_max2(*p->Tvm_trips_flip_rounded, s->LineTime / 4.0);
-
- if (p->display_cfg->gpuvm_enable == true || p->setup_for_tdlut || dcc_mrq_enable) {
- s->Tr0_trips_rounded = math_ceil2(4.0 * *p->Tr0_trips / s->LineTime, 1.0) / 4.0 * s->LineTime;
- *p->Tr0_trips_flip_rounded = math_ceil2(4.0 * *p->Tr0_trips_flip / s->LineTime, 1.0) / 4.0 * s->LineTime;
- } else {
- s->Tr0_trips_rounded = s->LineTime / 4.0;
- *p->Tr0_trips_flip_rounded = s->LineTime / 4.0;
- }
- s->Tr0_trips_rounded = math_max2(s->Tr0_trips_rounded, s->LineTime / 4.0);
- *p->Tr0_trips_flip_rounded = math_max2(*p->Tr0_trips_flip_rounded, s->LineTime / 4.0);
-
- *p->Tno_bw_flip = 0;
- if (p->display_cfg->gpuvm_enable == true) {
- if (p->display_cfg->gpuvm_max_page_table_levels >= 3) {
- *p->Tno_bw = p->ExtraLatencyPrefetch + s->trip_to_mem * (double)((p->display_cfg->gpuvm_max_page_table_levels - 2) * (s->HostVMDynamicLevelsTrips + 1));
- } else if (p->display_cfg->gpuvm_max_page_table_levels == 1 && !dcc_mrq_enable && !p->setup_for_tdlut) {
- *p->Tno_bw = p->ExtraLatencyPrefetch;
- } else {
- *p->Tno_bw = 0;
- }
- *p->Tno_bw_flip = *p->Tno_bw;
- } else {
- *p->Tno_bw = 0;
- }
-
- if (dml2_core_shared_is_420(p->myPipe->SourcePixelFormat)) {
- s->bytes_pp = p->myPipe->BytePerPixelY + p->myPipe->BytePerPixelC / 4.0;
- } else {
- s->bytes_pp = p->myPipe->BytePerPixelY + p->myPipe->BytePerPixelC;
- }
-
- s->prefetch_bw_pr = s->bytes_pp * p->myPipe->PixelClock / (double)p->myPipe->DPPPerSurface;
- if (p->myPipe->VRatio < 1.0)
- s->prefetch_bw_pr = p->myPipe->VRatio * s->prefetch_bw_pr;
- s->max_Tsw = (math_max2(p->PrefetchSourceLinesY, p->PrefetchSourceLinesC) * s->LineTime);
-
- s->prefetch_sw_bytes = p->PrefetchSourceLinesY * p->swath_width_luma_ub * p->myPipe->BytePerPixelY + p->PrefetchSourceLinesC * p->swath_width_chroma_ub * p->myPipe->BytePerPixelC;
- s->prefetch_bw_pr = s->prefetch_bw_pr * p->mall_prefetch_sdp_overhead_factor;
- s->prefetch_sw_bytes = s->prefetch_sw_bytes * p->mall_prefetch_sdp_overhead_factor;
- s->prefetch_bw_oto = math_max2(s->prefetch_bw_pr, s->prefetch_sw_bytes / s->max_Tsw);
-
- s->min_Lsw_oto = math_max2(p->PrefetchSourceLinesY, p->PrefetchSourceLinesC) / __DML2_CALCS_MAX_VRATIO_PRE_OTO__;
- s->min_Lsw_oto = math_max2(s->min_Lsw_oto, 2.0);
- s->min_Lsw_oto = math_max2(s->min_Lsw_oto, p->tdlut_drain_time / s->LineTime);
-
- unsigned int vm_bytes = p->vm_bytes; // vm_bytes is dpde0_bytes_per_frame_ub_l + dpde0_bytes_per_frame_ub_c + 2*extra_dpde_bytes;
- unsigned int extra_tdpe_bytes = (unsigned int)math_max2(0, (p->display_cfg->gpuvm_max_page_table_levels - 1) * 128);
-
- if (p->setup_for_tdlut)
- vm_bytes = vm_bytes + p->tdlut_pte_bytes_per_frame + (p->display_cfg->gpuvm_enable ? extra_tdpe_bytes : 0);
-
- unsigned long tdlut_row_bytes = (unsigned long) math_ceil2(p->tdlut_bytes_per_frame/2.0, 1.0);
- s->prefetch_bw_oto = math_max3(s->prefetch_bw_oto,
- p->vm_bytes * p->HostVMInefficiencyFactor / (31 * s->LineTime) - *p->Tno_bw,
- (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / (15 * s->LineTime));
- s->Lsw_oto = math_ceil2(4.0 * math_max2(s->prefetch_sw_bytes / s->prefetch_bw_oto / s->LineTime, s->min_Lsw_oto), 1.0) / 4.0;
-
- if (p->display_cfg->gpuvm_enable == true) {
- s->Tvm_oto = math_max3(
- *p->Tvm_trips,
- *p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw_oto,
- s->LineTime / 4.0);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tvm_oto max0 = %f\n", __func__, *p->Tvm_trips);
- dml2_printf("DML::%s: Tvm_oto max1 = %f\n", __func__, *p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw_oto);
- dml2_printf("DML::%s: Tvm_oto max2 = %f\n", __func__, s->LineTime / 4);
-#endif
-
- } else
- s->Tvm_oto = s->LineTime / 4.0;
-
- if ((p->display_cfg->gpuvm_enable == true || p->setup_for_tdlut || dcc_mrq_enable)) {
- s->Tr0_oto = math_max3(
- *p->Tr0_trips,
- (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / s->prefetch_bw_oto,
- s->LineTime / 4.0);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tr0_oto max0 = %f\n", __func__, *p->Tr0_trips);
- dml2_printf("DML::%s: Tr0_oto max1 = %f\n", __func__, (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / s->prefetch_bw_oto);
- dml2_printf("DML::%s: Tr0_oto max2 = %f\n", __func__, s->LineTime / 4);
-#endif
- } else
- s->Tr0_oto = (s->LineTime - s->Tvm_oto) / 4.0;
-
- s->Tvm_oto_lines = math_ceil2(4.0 * s->Tvm_oto / s->LineTime, 1) / 4.0;
- s->Tr0_oto_lines = math_ceil2(4.0 * s->Tr0_oto / s->LineTime, 1) / 4.0;
- s->dst_y_prefetch_oto = s->Tvm_oto_lines + 2 * s->Tr0_oto_lines + s->Lsw_oto;
-
- //To (time for delay after scaler) in line time
- unsigned int Lo = (unsigned int)(*p->DSTYAfterScaler + (double)*p->DSTXAfterScaler / (double)p->myPipe->HTotal);
-
- //Tpre_equ in line time
- s->dst_y_prefetch_equ = p->VStartup - (*p->TSetup + math_max2(s->TWait_p + p->TCalc, *p->Tdmdl - p->Ttrip)) / s->LineTime - Lo;
- s->dst_y_prefetch_equ = math_min2(s->dst_y_prefetch_equ, 63.75); // limit to the reg limit of U6.2 for DST_Y_PREFETCH
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: HTotal = %u\n", __func__, p->myPipe->HTotal);
- dml2_printf("DML::%s: min_Lsw_oto = %f\n", __func__, s->min_Lsw_oto);
- dml2_printf("DML::%s: Tno_bw = %f\n", __func__, *p->Tno_bw);
- dml2_printf("DML::%s: Tno_bw_flip = %f\n", __func__, *p->Tno_bw_flip);
- dml2_printf("DML::%s: ExtraLatencyPrefetch = %f\n", __func__, p->ExtraLatencyPrefetch);
- dml2_printf("DML::%s: trip_to_mem = %f\n", __func__, s->trip_to_mem);
- dml2_printf("DML::%s: mall_prefetch_sdp_overhead_factor = %f\n", __func__, p->mall_prefetch_sdp_overhead_factor);
- dml2_printf("DML::%s: BytePerPixelY = %u\n", __func__, p->myPipe->BytePerPixelY);
- dml2_printf("DML::%s: PrefetchSourceLinesY = %f\n", __func__, p->PrefetchSourceLinesY);
- dml2_printf("DML::%s: swath_width_luma_ub = %u\n", __func__, p->swath_width_luma_ub);
- dml2_printf("DML::%s: BytePerPixelC = %u\n", __func__, p->myPipe->BytePerPixelC);
- dml2_printf("DML::%s: PrefetchSourceLinesC = %f\n", __func__, p->PrefetchSourceLinesC);
- dml2_printf("DML::%s: swath_width_chroma_ub = %u\n", __func__, p->swath_width_chroma_ub);
- dml2_printf("DML::%s: prefetch_sw_bytes = %f\n", __func__, s->prefetch_sw_bytes);
- dml2_printf("DML::%s: max_Tsw = %f\n", __func__, s->max_Tsw);
- dml2_printf("DML::%s: bytes_pp = %f\n", __func__, s->bytes_pp);
- dml2_printf("DML::%s: vm_bytes = %u\n", __func__, vm_bytes);
- dml2_printf("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, p->PixelPTEBytesPerRow);
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, p->HostVMInefficiencyFactor);
- dml2_printf("DML::%s: Tvm_trips = %f\n", __func__, *p->Tvm_trips);
- dml2_printf("DML::%s: Tr0_trips = %f\n", __func__, *p->Tr0_trips);
- dml2_printf("DML::%s: Tvm_trips_flip = %f\n", __func__, *p->Tvm_trips_flip);
- dml2_printf("DML::%s: Tr0_trips_flip = %f\n", __func__, *p->Tr0_trips_flip);
- dml2_printf("DML::%s: prefetch_bw_pr = %f\n", __func__, s->prefetch_bw_pr);
- dml2_printf("DML::%s: prefetch_bw_oto = %f\n", __func__, s->prefetch_bw_oto);
- dml2_printf("DML::%s: Tr0_oto = %f\n", __func__, s->Tr0_oto);
- dml2_printf("DML::%s: Tvm_oto = %f\n", __func__, s->Tvm_oto);
- dml2_printf("DML::%s: Tvm_oto_lines = %f\n", __func__, s->Tvm_oto_lines);
- dml2_printf("DML::%s: Tr0_oto_lines = %f\n", __func__, s->Tr0_oto_lines);
- dml2_printf("DML::%s: Lsw_oto = %f\n", __func__, s->Lsw_oto);
- dml2_printf("DML::%s: dst_y_prefetch_oto = %f\n", __func__, s->dst_y_prefetch_oto);
- dml2_printf("DML::%s: dst_y_prefetch_equ = %f\n", __func__, s->dst_y_prefetch_equ);
- dml2_printf("DML::%s: tdlut_row_bytes = %d\n", __func__, tdlut_row_bytes);
- dml2_printf("DML::%s: meta_row_bytes = %d\n", __func__, p->meta_row_bytes);
-#endif
-
- s->dst_y_prefetch_equ = math_floor2(4.0 * (s->dst_y_prefetch_equ + 0.125), 1) / 4.0;
- s->Tpre_rounded = s->dst_y_prefetch_equ * s->LineTime;
-
- dml2_printf("DML::%s: dst_y_prefetch_equ: %f (after round)\n", __func__, s->dst_y_prefetch_equ);
- dml2_printf("DML::%s: LineTime: %f\n", __func__, s->LineTime);
- dml2_printf("DML::%s: VStartup: %u\n", __func__, p->VStartup);
- dml2_printf("DML::%s: Tvstartup: %fus - time between vstartup and first pixel of active\n", __func__, p->VStartup * s->LineTime);
- dml2_printf("DML::%s: TSetup: %fus - time from vstartup to vready\n", __func__, *p->TSetup);
- dml2_printf("DML::%s: TCalc: %fus - time for calculations in dchub starting at vready\n", __func__, p->TCalc);
- dml2_printf("DML::%s: TWait: %fus - time for fabric to become ready max(pstate exit,cstate enter/exit, urgent latency) after TCalc\n", __func__, p->TWait);
- dml2_printf("DML::%s: Tdmbf: %fus - time for dmd transfer from dchub to dio output buffer\n", __func__, s->Tdmbf);
- dml2_printf("DML::%s: Tdmec: %fus - time dio takes to transfer dmd\n", __func__, s->Tdmec);
- dml2_printf("DML::%s: Tdmsks: %fus - time before active dmd must complete transmission at dio\n", __func__, s->Tdmsks);
- dml2_printf("DML::%s: Tdmdl_vm: %fus - time for vm stages of dmd \n", __func__, *p->Tdmdl_vm);
- dml2_printf("DML::%s: Tdmdl: %fus - time for fabric to become ready and fetch dmd \n", __func__, *p->Tdmdl);
- dml2_printf("DML::%s: TWait_p: %fus\n", __func__, s->TWait_p);
- dml2_printf("DML::%s: Ttrip: %fus\n", __func__, p->Ttrip);
- dml2_printf("DML::%s: DSTXAfterScaler: %u pixels - number of pixel clocks pipeline and buffer delay after scaler \n", __func__, *p->DSTXAfterScaler);
- dml2_printf("DML::%s: DSTYAfterScaler: %u lines - number of lines of pipeline and buffer delay after scaler \n", __func__, *p->DSTYAfterScaler);
-
- s->dep_bytes = math_max2(vm_bytes * p->HostVMInefficiencyFactor, p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes);
-
- dml2_printf("DML::%s: dep_bytes: %f\n", __func__, s->dep_bytes);
- dml2_printf("DML::%s: prefetch_sw_bytes: %f\n", __func__, s->prefetch_sw_bytes);
- dml2_printf("DML::%s: vm_bytes: %f (hvm inefficiency scaled)\n", __func__, vm_bytes * p->HostVMInefficiencyFactor);
- dml2_printf("DML::%s: row_bytes: %f (hvm inefficiency scaled, 1 row)\n", __func__, p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes);
-
- if (s->prefetch_sw_bytes < s->dep_bytes) {
- s->prefetch_sw_bytes = 2 * s->dep_bytes;
- dml2_printf("DML::%s: bump prefetch_sw_bytes to %f\n", __func__, s->prefetch_sw_bytes);
- }
-
- *p->dst_y_per_vm_vblank = 0;
- *p->dst_y_per_row_vblank = 0;
- *p->VRatioPrefetchY = 0;
- *p->VRatioPrefetchC = 0;
- *p->RequiredPrefetchPixelDataBWLuma = 0;
-
- if (s->dst_y_prefetch_equ > 1) {
- s->prefetch_bw1 = 0.;
- s->prefetch_bw2 = 0.;
- s->prefetch_bw3 = 0.;
- s->prefetch_bw4 = 0.;
-
- if (s->Tpre_rounded - *p->Tno_bw > 0) {
- s->prefetch_bw1 = (vm_bytes * p->HostVMInefficiencyFactor
- + 2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes)
- + s->prefetch_sw_bytes)
- / (s->Tpre_rounded - *p->Tno_bw);
- s->Tsw_est1 = s->prefetch_sw_bytes / s->prefetch_bw1;
- } else
- s->prefetch_bw1 = 0;
-
- dml2_printf("DML::%s: prefetch_bw1: %f\n", __func__, s->prefetch_bw1);
- if (p->VStartup == p->MaxVStartup && (s->Tsw_est1 / s->LineTime < s->min_Lsw_oto) && s->Tpre_rounded - s->min_Lsw_oto * s->LineTime - 0.75 * s->LineTime - *p->Tno_bw > 0) {
- s->prefetch_bw1 = (vm_bytes * p->HostVMInefficiencyFactor + 2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes)) /
- (s->Tpre_rounded - s->min_Lsw_oto * s->LineTime - 0.75 * s->LineTime - *p->Tno_bw);
- dml2_printf("DML::%s: prefetch_bw1: %f (updated)\n", __func__, s->prefetch_bw1);
- }
-
- if (s->Tpre_rounded - *p->Tno_bw - 2 * s->Tr0_trips_rounded > 0)
- s->prefetch_bw2 = (vm_bytes * p->HostVMInefficiencyFactor + s->prefetch_sw_bytes) /
- (s->Tpre_rounded - *p->Tno_bw - 2 * s->Tr0_trips_rounded);
- else
- s->prefetch_bw2 = 0;
-
- if (s->Tpre_rounded - s->Tvm_trips_rounded > 0) {
- s->prefetch_bw3 = (2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) + s->prefetch_sw_bytes) /
- (s->Tpre_rounded - s->Tvm_trips_rounded);
- s->Tsw_est3 = s->prefetch_sw_bytes / s->prefetch_bw3;
- } else
- s->prefetch_bw3 = 0;
-
-
- dml2_printf("DML::%s: prefetch_bw3: %f\n", __func__, s->prefetch_bw3);
- if (p->VStartup == p->MaxVStartup && (s->Tsw_est3 / s->LineTime < s->min_Lsw_oto) && s->Tpre_rounded - s->min_Lsw_oto * s->LineTime - 0.5 * s->LineTime - s->Tvm_trips_rounded > 0) {
- s->prefetch_bw3 = (2 * (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes)) / (s->Tpre_rounded - s->min_Lsw_oto * s->LineTime - 0.5 * s->LineTime - s->Tvm_trips_rounded);
- dml2_printf("DML::%s: prefetch_bw3: %f (updated)\n", __func__, s->prefetch_bw3);
- }
-
- if (s->Tpre_rounded - s->Tvm_trips_rounded - 2 * s->Tr0_trips_rounded > 0)
- s->prefetch_bw4 = s->prefetch_sw_bytes / (s->Tpre_rounded - s->Tvm_trips_rounded - 2 * s->Tr0_trips_rounded);
- else
- s->prefetch_bw4 = 0;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tpre_rounded: %f\n", __func__, s->Tpre_rounded);
- dml2_printf("DML::%s: Tno_bw: %f\n", __func__, *p->Tno_bw);
- dml2_printf("DML::%s: Tvm_trips_rounded: %f\n", __func__, s->Tvm_trips_rounded);
- dml2_printf("DML::%s: Tr0_trips_rounded: %f\n", __func__, 2 * s->Tr0_trips_rounded);
- dml2_printf("DML::%s: Tsw_est1: %f\n", __func__, s->Tsw_est1);
- dml2_printf("DML::%s: Tsw_est3: %f\n", __func__, s->Tsw_est3);
- dml2_printf("DML::%s: prefetch_bw1: %f (final)\n", __func__, s->prefetch_bw1);
- dml2_printf("DML::%s: prefetch_bw2: %f (final)\n", __func__, s->prefetch_bw2);
- dml2_printf("DML::%s: prefetch_bw3: %f (final)\n", __func__, s->prefetch_bw3);
- dml2_printf("DML::%s: prefetch_bw4: %f (final)\n", __func__, s->prefetch_bw4);
-#endif
-
- {
- bool Case1OK = false;
- bool Case2OK = false;
- bool Case3OK = false;
-
- if (s->prefetch_bw1 > 0) {
- if (*p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw1 >= s->Tvm_trips_rounded &&
- (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / s->prefetch_bw1 >= s->Tr0_trips_rounded) {
- Case1OK = true;
- }
- }
-
- if (s->prefetch_bw2 > 0) {
- if (*p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw2 >= s->Tvm_trips_rounded &&
- (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / s->prefetch_bw2 < s->Tr0_trips_rounded) {
- Case2OK = true;
- }
- }
-
- if (s->prefetch_bw3 > 0) {
- if (*p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw3 < s->Tvm_trips_rounded &&
- (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / s->prefetch_bw3 >= s->Tr0_trips_rounded) {
- Case3OK = true;
- }
- }
-
- if (Case1OK) {
- s->prefetch_bw_equ = s->prefetch_bw1;
- } else if (Case2OK) {
- s->prefetch_bw_equ = s->prefetch_bw2;
- } else if (Case3OK) {
- s->prefetch_bw_equ = s->prefetch_bw3;
- } else {
- s->prefetch_bw_equ = s->prefetch_bw4;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Case1OK: %u\n", __func__, Case1OK);
- dml2_printf("DML::%s: Case2OK: %u\n", __func__, Case2OK);
- dml2_printf("DML::%s: Case3OK: %u\n", __func__, Case3OK);
- dml2_printf("DML::%s: prefetch_bw_equ: %f\n", __func__, s->prefetch_bw_equ);
-#endif
- s->prefetch_bw_equ = math_max3(s->prefetch_bw_equ,
- p->vm_bytes * p->HostVMInefficiencyFactor / (31 * s->LineTime) - *p->Tno_bw,
- (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / (15 * s->LineTime));
-
- if (s->prefetch_bw_equ > 0) {
- if (p->display_cfg->gpuvm_enable == true) {
- s->Tvm_equ = math_max3(*p->Tno_bw + vm_bytes * p->HostVMInefficiencyFactor / s->prefetch_bw_equ, *p->Tvm_trips, s->LineTime / 4);
- } else {
- s->Tvm_equ = s->LineTime / 4;
- }
-
- if (p->display_cfg->gpuvm_enable == true || dcc_mrq_enable || p->setup_for_tdlut) {
- s->Tr0_equ = math_max3((p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + p->meta_row_bytes + tdlut_row_bytes) / s->prefetch_bw_equ, // PixelPTEBytesPerRow is dpte_row_bytes
- *p->Tr0_trips,
- s->LineTime / 4);
- } else {
- s->Tr0_equ = s->LineTime / 4;
- }
- } else {
- s->Tvm_equ = 0;
- s->Tr0_equ = 0;
- dml2_printf("DML::%s: prefetch_bw_equ equals 0!\n", __func__);
- }
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Tvm_equ = %f\n", __func__, s->Tvm_equ);
- dml2_printf("DML::%s: Tr0_equ = %f\n", __func__, s->Tr0_equ);
-#endif
-
- if (s->dst_y_prefetch_oto < s->dst_y_prefetch_equ) {
- *p->dst_y_prefetch = s->dst_y_prefetch_oto;
- s->TimeForFetchingVM = s->Tvm_oto;
- s->TimeForFetchingRowInVBlank = s->Tr0_oto;
-
- *p->dst_y_per_vm_vblank = math_ceil2(4.0 * s->TimeForFetchingVM / s->LineTime, 1.0) / 4.0;
- *p->dst_y_per_row_vblank = math_ceil2(4.0 * s->TimeForFetchingRowInVBlank / s->LineTime, 1.0) / 4.0;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Using oto bw scheduling for prefetch\n", __func__);
-#endif
-
- } else {
- *p->dst_y_prefetch = s->dst_y_prefetch_equ;
- s->TimeForFetchingVM = s->Tvm_equ;
- s->TimeForFetchingRowInVBlank = s->Tr0_equ;
-
- if (p->VStartup == p->MaxVStartup) {
- *p->dst_y_per_vm_vblank = math_floor2(4.0 * s->TimeForFetchingVM / s->LineTime, 1.0) / 4.0;
- *p->dst_y_per_row_vblank = math_floor2(4.0 * s->TimeForFetchingRowInVBlank / s->LineTime, 1.0) / 4.0;
- } else {
- *p->dst_y_per_vm_vblank = math_ceil2(4.0 * s->TimeForFetchingVM / s->LineTime, 1.0) / 4.0;
- *p->dst_y_per_row_vblank = math_ceil2(4.0 * s->TimeForFetchingRowInVBlank / s->LineTime, 1.0) / 4.0;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Using equ bw scheduling for prefetch\n", __func__);
-#endif
- }
- dml2_assert(*p->dst_y_prefetch < 64);
-
- // Lsw = dst_y_prefetch - (dst_y_per_vm_vblank + 2*dst_y_per_row_vblank)
- s->LinesToRequestPrefetchPixelData = *p->dst_y_prefetch - *p->dst_y_per_vm_vblank - 2 * *p->dst_y_per_row_vblank; // Lsw
-
- s->cursor_prefetch_bytes = (unsigned int)math_max2(p->cursor_bytes_per_chunk, 4 * p->cursor_bytes_per_line);
- *p->prefetch_cursor_bw = p->num_cursors * s->cursor_prefetch_bytes / (s->LinesToRequestPrefetchPixelData * s->LineTime);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: TimeForFetchingVM = %f\n", __func__, s->TimeForFetchingVM);
- dml2_printf("DML::%s: TimeForFetchingRowInVBlank = %f\n", __func__, s->TimeForFetchingRowInVBlank);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, s->LineTime);
- dml2_printf("DML::%s: dst_y_prefetch = %f\n", __func__, *p->dst_y_prefetch);
- dml2_printf("DML::%s: dst_y_per_vm_vblank = %f\n", __func__, *p->dst_y_per_vm_vblank);
- dml2_printf("DML::%s: dst_y_per_row_vblank = %f\n", __func__, *p->dst_y_per_row_vblank);
- dml2_printf("DML::%s: LinesToRequestPrefetchPixelData = %f\n", __func__, s->LinesToRequestPrefetchPixelData);
- dml2_printf("DML::%s: PrefetchSourceLinesY = %f\n", __func__, p->PrefetchSourceLinesY);
-
- dml2_printf("DML::%s: cursor_bytes_per_chunk = %d\n", __func__, p->cursor_bytes_per_chunk);
- dml2_printf("DML::%s: cursor_bytes_per_line = %d\n", __func__, p->cursor_bytes_per_line);
- dml2_printf("DML::%s: cursor_prefetch_bytes = %d\n", __func__, s->cursor_prefetch_bytes);
- dml2_printf("DML::%s: prefetch_cursor_bw = %f\n", __func__, *p->prefetch_cursor_bw);
-#endif
- unsigned int min_lsw_required = (unsigned int)math_max2(2, p->tdlut_drain_time / s->LineTime);
-
- if (s->LinesToRequestPrefetchPixelData >= min_lsw_required && s->prefetch_bw_equ > 0) {
- *p->VRatioPrefetchY = (double)p->PrefetchSourceLinesY / s->LinesToRequestPrefetchPixelData;
- *p->VRatioPrefetchY = math_max2(*p->VRatioPrefetchY, 1.0);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatioPrefetchY = %f\n", __func__, *p->VRatioPrefetchY);
- dml2_printf("DML::%s: SwathHeightY = %u\n", __func__, p->SwathHeightY);
- dml2_printf("DML::%s: VInitPreFillY = %u\n", __func__, p->VInitPreFillY);
-#endif
- if ((p->SwathHeightY > 4) && (p->VInitPreFillY > 3)) {
- if (s->LinesToRequestPrefetchPixelData > (p->VInitPreFillY - 3.0) / 2.0) {
- *p->VRatioPrefetchY = math_max2(*p->VRatioPrefetchY,
- (double)p->MaxNumSwathY * p->SwathHeightY / (s->LinesToRequestPrefetchPixelData - (p->VInitPreFillY - 3.0) / 2.0));
- } else {
- s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: MyErr set. LinesToRequestPrefetchPixelData=%f VinitPreFillY=%u\n", __func__, s->LinesToRequestPrefetchPixelData, p->VInitPreFillY);
- *p->VRatioPrefetchY = 0;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatioPrefetchY = %f\n", __func__, *p->VRatioPrefetchY);
- dml2_printf("DML::%s: PrefetchSourceLinesY = %f\n", __func__, p->PrefetchSourceLinesY);
- dml2_printf("DML::%s: MaxNumSwathY = %u\n", __func__, p->MaxNumSwathY);
-#endif
- }
-
- *p->VRatioPrefetchC = (double)p->PrefetchSourceLinesC / s->LinesToRequestPrefetchPixelData;
- *p->VRatioPrefetchC = math_max2(*p->VRatioPrefetchC, 1.0);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatioPrefetchC = %f\n", __func__, *p->VRatioPrefetchC);
- dml2_printf("DML::%s: SwathHeightC = %u\n", __func__, p->SwathHeightC);
- dml2_printf("DML::%s: VInitPreFillC = %u\n", __func__, p->VInitPreFillC);
-#endif
- if ((p->SwathHeightC > 4) && (p->VInitPreFillC > 3)) {
- if (s->LinesToRequestPrefetchPixelData > (p->VInitPreFillC - 3.0) / 2.0) {
- *p->VRatioPrefetchC = math_max2(*p->VRatioPrefetchC, (double)p->MaxNumSwathC * p->SwathHeightC / (s->LinesToRequestPrefetchPixelData - (p->VInitPreFillC - 3.0) / 2.0));
- } else {
- s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: MyErr set. LinesToRequestPrefetchPixelData=%f VInitPreFillC=%u\n", __func__, s->LinesToRequestPrefetchPixelData, p->VInitPreFillC);
- *p->VRatioPrefetchC = 0;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VRatioPrefetchC = %f\n", __func__, *p->VRatioPrefetchC);
- dml2_printf("DML::%s: PrefetchSourceLinesC = %f\n", __func__, p->PrefetchSourceLinesC);
- dml2_printf("DML::%s: MaxNumSwathC = %u\n", __func__, p->MaxNumSwathC);
-#endif
- }
-
- *p->RequiredPrefetchPixelDataBWLuma = (double)p->PrefetchSourceLinesY / s->LinesToRequestPrefetchPixelData * p->myPipe->BytePerPixelY * p->swath_width_luma_ub / s->LineTime;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: BytePerPixelY = %u\n", __func__, p->myPipe->BytePerPixelY);
- dml2_printf("DML::%s: swath_width_luma_ub = %u\n", __func__, p->swath_width_luma_ub);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, s->LineTime);
- dml2_printf("DML::%s: RequiredPrefetchPixelDataBWLuma = %f\n", __func__, *p->RequiredPrefetchPixelDataBWLuma);
-#endif
- *p->RequiredPrefetchPixelDataBWChroma = (double)p->PrefetchSourceLinesC / s->LinesToRequestPrefetchPixelData * p->myPipe->BytePerPixelC * p->swath_width_chroma_ub / s->LineTime;
- } else {
- s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: MyErr set, LinesToRequestPrefetchPixelData: %f, should be >= %d\n", __func__, s->LinesToRequestPrefetchPixelData, min_lsw_required);
- dml2_printf("DML::%s: MyErr set, prefetch_bw_equ: %f, should be > 0\n", __func__, s->prefetch_bw_equ);
- *p->VRatioPrefetchY = 0;
- *p->VRatioPrefetchC = 0;
- *p->RequiredPrefetchPixelDataBWLuma = 0;
- *p->RequiredPrefetchPixelDataBWChroma = 0;
- }
-
- dml2_printf("DML: Tpre: %fus - sum of time to request 2 x data pte, swaths\n", (double)s->LinesToRequestPrefetchPixelData * s->LineTime + 2.0 * s->TimeForFetchingRowInVBlank + s->TimeForFetchingVM);
- dml2_printf("DML: Tvm: %fus - time to fetch vm\n", s->TimeForFetchingVM);
- dml2_printf("DML: Tr0: %fus - time to fetch first row of data pagetables\n", s->TimeForFetchingRowInVBlank);
- dml2_printf("DML: Tsw: %fus = time to fetch enough pixel data and cursor data to feed the scalers init position and detile\n", (double)s->LinesToRequestPrefetchPixelData * s->LineTime);
- dml2_printf("DML: To: %fus - time for propagation from scaler to optc\n", (*p->DSTYAfterScaler + ((double)(*p->DSTXAfterScaler) / (double)p->myPipe->HTotal)) * s->LineTime);
- dml2_printf("DML: Tvstartup - TSetup - Tcalc - TWait - Tpre - To > 0\n");
- dml2_printf("DML: Tslack(pre): %fus - time left over in schedule\n", p->VStartup * s->LineTime - s->TimeForFetchingVM - 2 * s->TimeForFetchingRowInVBlank - (*p->DSTYAfterScaler + ((double)(*p->DSTXAfterScaler) / (double)p->myPipe->HTotal)) * s->LineTime - p->TWait - p->TCalc - *p->TSetup);
- dml2_printf("DML: row_bytes = dpte_row_bytes (per_pipe) = PixelPTEBytesPerRow = : %u\n", p->PixelPTEBytesPerRow);
-
- } else {
- dml2_printf("DML::%s: MyErr set, dst_y_prefetch_equ = %f (should be > 1)\n", __func__, s->dst_y_prefetch_equ);
- s->NoTimeToPrefetch = true;
- s->TimeForFetchingVM = 0;
- s->TimeForFetchingRowInVBlank = 0;
- *p->dst_y_per_vm_vblank = 0;
- *p->dst_y_per_row_vblank = 0;
- s->LinesToRequestPrefetchPixelData = 0;
- *p->VRatioPrefetchY = 0;
- *p->VRatioPrefetchC = 0;
- *p->RequiredPrefetchPixelDataBWLuma = 0;
- *p->RequiredPrefetchPixelDataBWChroma = 0;
- }
-
- {
- double prefetch_vm_bw;
- double prefetch_row_bw;
-
- if (vm_bytes == 0) {
- prefetch_vm_bw = 0;
- } else if (*p->dst_y_per_vm_vblank > 0) {
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, p->HostVMInefficiencyFactor);
- dml2_printf("DML::%s: dst_y_per_vm_vblank = %f\n", __func__, *p->dst_y_per_vm_vblank);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, s->LineTime);
-#endif
- prefetch_vm_bw = vm_bytes * p->HostVMInefficiencyFactor / (*p->dst_y_per_vm_vblank * s->LineTime);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: prefetch_vm_bw = %f\n", __func__, prefetch_vm_bw);
-#endif
- } else {
- prefetch_vm_bw = 0;
- s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: MyErr set. dst_y_per_vm_vblank=%f (should be > 0)\n", __func__, *p->dst_y_per_vm_vblank);
- }
-
- if (p->PixelPTEBytesPerRow == 0 && tdlut_row_bytes == 0) {
- prefetch_row_bw = 0;
- } else if (*p->dst_y_per_row_vblank > 0) {
- prefetch_row_bw = (p->PixelPTEBytesPerRow * p->HostVMInefficiencyFactor + tdlut_row_bytes) / (*p->dst_y_per_row_vblank * s->LineTime);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, p->PixelPTEBytesPerRow);
- dml2_printf("DML::%s: dst_y_per_row_vblank = %f\n", __func__, *p->dst_y_per_row_vblank);
- dml2_printf("DML::%s: prefetch_row_bw = %f\n", __func__, prefetch_row_bw);
-#endif
- } else {
- prefetch_row_bw = 0;
- s->NoTimeToPrefetch = true;
- dml2_printf("DML::%s: MyErr set. dst_y_per_row_vblank=%f (should be > 0)\n", __func__, *p->dst_y_per_row_vblank);
- }
-
- *p->prefetch_vmrow_bw = math_max2(prefetch_vm_bw, prefetch_row_bw);
- }
-
- if (s->NoTimeToPrefetch) {
- s->TimeForFetchingVM = 0;
- s->TimeForFetchingRowInVBlank = 0;
- *p->dst_y_per_vm_vblank = 0;
- *p->dst_y_per_row_vblank = 0;
- *p->dst_y_prefetch = 0;
- s->LinesToRequestPrefetchPixelData = 0;
- *p->VRatioPrefetchY = 0;
- *p->VRatioPrefetchC = 0;
- *p->RequiredPrefetchPixelDataBWLuma = 0;
- *p->RequiredPrefetchPixelDataBWChroma = 0;
- }
-
- dml2_printf("DML::%s: dst_y_per_vm_vblank = %f (final)\n", __func__, *p->dst_y_per_vm_vblank);
- dml2_printf("DML::%s: dst_y_per_row_vblank = %f (final)\n", __func__, *p->dst_y_per_row_vblank);
- dml2_printf("DML::%s: NoTimeToPrefetch=%d\n", __func__, s->NoTimeToPrefetch);
- return s->NoTimeToPrefetch;
-}
-
-static void calculate_peak_bandwidth_required(
- struct dml2_core_internal_scratch *s,
-
- // output
- double urg_vactive_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double non_urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
-
- // input
- const struct dml2_display_cfg *display_cfg,
- unsigned int inc_flip_bw,
- unsigned int NumberOfActiveSurfaces,
- unsigned int NumberOfDPP[],
- double dcc_dram_bw_nom_overhead_factor_p0[],
- double dcc_dram_bw_nom_overhead_factor_p1[],
- double dcc_dram_bw_pref_overhead_factor_p0[],
- double dcc_dram_bw_pref_overhead_factor_p1[],
- double mall_prefetch_sdp_overhead_factor[],
- double mall_prefetch_dram_overhead_factor[],
- double ReadBandwidthLuma[],
- double ReadBandwidthChroma[],
- double PrefetchBandwidthLuma[],
- double PrefetchBandwidthChroma[],
- double cursor_bw[],
- double dpte_row_bw[],
- double meta_row_bw[],
- double prefetch_cursor_bw[],
- double prefetch_vmrow_bw[],
- double flip_bw[],
- double UrgentBurstFactorLuma[],
- double UrgentBurstFactorChroma[],
- double UrgentBurstFactorCursor[],
- double UrgentBurstFactorLumaPre[],
- double UrgentBurstFactorChromaPre[],
- double UrgentBurstFactorCursorPre[])
-{
- unsigned int n;
- unsigned int m;
-
- struct dml2_core_shared_calculate_peak_bandwidth_required_locals *l = &s->calculate_peak_bandwidth_required_locals;
-
- memset(l, 0, sizeof(struct dml2_core_shared_calculate_peak_bandwidth_required_locals));
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: inc_flip_bw = %d\n", __func__, inc_flip_bw);
- dml2_printf("DML::%s: NumberOfActiveSurfaces = %d\n", __func__, NumberOfActiveSurfaces);
-#endif
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- l->unity_array[k] = 1.0;
- l->zero_array[k] = 0.0;
- }
-
- for (m = 0; m < dml2_core_internal_soc_state_max; m++) {
- for (n = 0; n < dml2_core_internal_bw_max; n++) {
- urg_vactive_bandwidth_required[m][n] = get_urgent_bandwidth_required(
- &s->get_urgent_bandwidth_required_locals,
- display_cfg,
- m,
- n,
- 0, //inc_flip_bw,
- NumberOfActiveSurfaces,
- NumberOfDPP,
- dcc_dram_bw_nom_overhead_factor_p0,
- dcc_dram_bw_nom_overhead_factor_p1,
- dcc_dram_bw_pref_overhead_factor_p0,
- dcc_dram_bw_pref_overhead_factor_p1,
- mall_prefetch_sdp_overhead_factor,
- mall_prefetch_dram_overhead_factor,
- ReadBandwidthLuma,
- ReadBandwidthChroma,
- l->zero_array, //PrefetchBandwidthLuma,
- l->zero_array, //PrefetchBandwidthChroma,
- cursor_bw,
- dpte_row_bw,
- meta_row_bw,
- l->zero_array, //prefetch_cursor_bw,
- l->zero_array, //prefetch_vmrow_bw,
- l->zero_array, //flip_bw,
- UrgentBurstFactorLuma,
- UrgentBurstFactorChroma,
- UrgentBurstFactorCursor,
- UrgentBurstFactorLumaPre,
- UrgentBurstFactorChromaPre,
- UrgentBurstFactorCursorPre);
-
-
- urg_bandwidth_required[m][n] = get_urgent_bandwidth_required(
- &s->get_urgent_bandwidth_required_locals,
- display_cfg,
- m,
- n,
- inc_flip_bw,
- NumberOfActiveSurfaces,
- NumberOfDPP,
- dcc_dram_bw_nom_overhead_factor_p0,
- dcc_dram_bw_nom_overhead_factor_p1,
- dcc_dram_bw_pref_overhead_factor_p0,
- dcc_dram_bw_pref_overhead_factor_p1,
- mall_prefetch_sdp_overhead_factor,
- mall_prefetch_dram_overhead_factor,
- ReadBandwidthLuma,
- ReadBandwidthChroma,
- PrefetchBandwidthLuma,
- PrefetchBandwidthChroma,
- cursor_bw,
- dpte_row_bw,
- meta_row_bw,
- prefetch_cursor_bw,
- prefetch_vmrow_bw,
- flip_bw,
- UrgentBurstFactorLuma,
- UrgentBurstFactorChroma,
- UrgentBurstFactorCursor,
- UrgentBurstFactorLumaPre,
- UrgentBurstFactorChromaPre,
- UrgentBurstFactorCursorPre);
-
- non_urg_bandwidth_required[m][n] = get_urgent_bandwidth_required(
- &s->get_urgent_bandwidth_required_locals,
- display_cfg,
- m,
- n,
- inc_flip_bw,
- NumberOfActiveSurfaces,
- NumberOfDPP,
- dcc_dram_bw_nom_overhead_factor_p0,
- dcc_dram_bw_nom_overhead_factor_p1,
- dcc_dram_bw_pref_overhead_factor_p0,
- dcc_dram_bw_pref_overhead_factor_p1,
- mall_prefetch_sdp_overhead_factor,
- mall_prefetch_dram_overhead_factor,
- ReadBandwidthLuma,
- ReadBandwidthChroma,
- PrefetchBandwidthLuma,
- PrefetchBandwidthChroma,
- cursor_bw,
- dpte_row_bw,
- meta_row_bw,
- prefetch_cursor_bw,
- prefetch_vmrow_bw,
- flip_bw,
- l->unity_array,
- l->unity_array,
- l->unity_array,
- l->unity_array,
- l->unity_array,
- l->unity_array);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: urg_vactive_bandwidth_required%s[%s][%s]=%f\n", __func__, (inc_flip_bw ? "_flip" : ""), dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), urg_vactive_bandwidth_required[m][n]);
- dml2_printf("DML::%s: urg_bandwidth_required%s[%s][%s]=%f\n", __func__, (inc_flip_bw ? "_flip" : ""), dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), urg_bandwidth_required[m][n]);
- dml2_printf("DML::%s: non_urg_bandwidth_required%s[%s][%s]=%f\n", __func__, (inc_flip_bw ? "_flip" : ""), dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n), non_urg_bandwidth_required[m][n]);
-#endif
- dml2_assert(urg_bandwidth_required[m][n] >= non_urg_bandwidth_required[m][n]);
- }
- }
-}
-
-static void check_urgent_bandwidth_support(
- double *frac_urg_bandwidth_nom,
- double *frac_urg_bandwidth_mall,
- bool *vactive_bandwidth_support_ok, // vactive ok
- bool *bandwidth_support_ok, // max of vm, prefetch, vactive all ok
-
- unsigned int mall_allocated_for_dcn_mbytes,
- double non_urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_vactive_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max])
-{
- *bandwidth_support_ok = 1;
- *vactive_bandwidth_support_ok = 1;
-
- double frac_urg_bandwidth_nom_sdp = non_urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp] / urg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- double frac_urg_bandwidth_nom_dram = non_urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram] / urg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
- double frac_urg_bandwidth_mall_sdp = non_urg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp] / urg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- double frac_urg_bandwidth_mall_dram = non_urg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram] / urg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
-
- // Check urgent bandwidth required at sdp vs urgent bandwidth avail at sdp -> FractionOfUrgentBandwidth
- // Check urgent bandwidth required at dram vs urgent bandwidth avail at dram
- // Check urgent bandwidth required at sdp vs urgent bandwidth avail at sdp, svp_prefetch -> FractionOfUrgentBandwidthMALL
- // Check urgent bandwidth required at dram vs urgent bandwidth avail at dram, svp_prefetch
-
- *bandwidth_support_ok &= urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp] <= urg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- *bandwidth_support_ok &= urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram] <= urg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
-
- if (mall_allocated_for_dcn_mbytes > 0) {
- *bandwidth_support_ok &= urg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp] <= urg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- *bandwidth_support_ok &= urg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram] <= urg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
- }
-
- *frac_urg_bandwidth_nom = math_max2(frac_urg_bandwidth_nom_sdp, frac_urg_bandwidth_nom_dram);
- *frac_urg_bandwidth_mall = math_max2(frac_urg_bandwidth_mall_sdp, frac_urg_bandwidth_mall_dram);
-
- *bandwidth_support_ok &= (*frac_urg_bandwidth_nom <= 1.0);
-
- if (mall_allocated_for_dcn_mbytes > 0)
- *bandwidth_support_ok &= (*frac_urg_bandwidth_mall <= 1.0);
-
- *vactive_bandwidth_support_ok &= urg_vactive_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp] <= urg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- *vactive_bandwidth_support_ok &= urg_vactive_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram] <= urg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
- if (mall_allocated_for_dcn_mbytes > 0) {
- *vactive_bandwidth_support_ok &= urg_vactive_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp] <= urg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- *vactive_bandwidth_support_ok &= urg_vactive_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram] <= urg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: frac_urg_bandwidth_nom_sdp = %f\n", __func__, frac_urg_bandwidth_nom_sdp);
- dml2_printf("DML::%s: frac_urg_bandwidth_nom_dram = %f\n", __func__, frac_urg_bandwidth_nom_dram);
- dml2_printf("DML::%s: frac_urg_bandwidth_nom = %f\n", __func__, *frac_urg_bandwidth_nom);
-
- dml2_printf("DML::%s: frac_urg_bandwidth_mall_sdp = %f\n", __func__, frac_urg_bandwidth_mall_sdp);
- dml2_printf("DML::%s: frac_urg_bandwidth_mall_dram = %f\n", __func__, frac_urg_bandwidth_mall_dram);
- dml2_printf("DML::%s: frac_urg_bandwidth_mall = %f\n", __func__, *frac_urg_bandwidth_mall);
- dml2_printf("DML::%s: bandwidth_support_ok = %d\n", __func__, *bandwidth_support_ok);
-#endif
-
-}
-
-static double get_bandwidth_available_for_immediate_flip(enum dml2_core_internal_soc_state_type eval_state,
- double urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max], // no flip
- double urg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max])
-{
- double flip_bw_available_mbps;
- double flip_bw_available_sdp_mbps;
- double flip_bw_available_dram_mbps;
-
- flip_bw_available_sdp_mbps = urg_bandwidth_available[eval_state][dml2_core_internal_bw_sdp] - urg_bandwidth_required[eval_state][dml2_core_internal_bw_sdp];
- flip_bw_available_dram_mbps = urg_bandwidth_available[eval_state][dml2_core_internal_bw_dram] - urg_bandwidth_required[eval_state][dml2_core_internal_bw_dram];
- flip_bw_available_mbps = flip_bw_available_sdp_mbps < flip_bw_available_dram_mbps ? flip_bw_available_sdp_mbps : flip_bw_available_dram_mbps;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: eval_state = %s\n", __func__, dml2_core_internal_soc_state_type_str(eval_state));
- dml2_printf("DML::%s: urg_bandwidth_available_sdp_mbps = %f\n", __func__, urg_bandwidth_available[eval_state][dml2_core_internal_bw_sdp]);
- dml2_printf("DML::%s: urg_bandwidth_available_dram_mbps = %f\n", __func__, urg_bandwidth_available[eval_state][dml2_core_internal_bw_dram]);
- dml2_printf("DML::%s: urg_bandwidth_required_sdp_mbps = %f\n", __func__, urg_bandwidth_required[eval_state][dml2_core_internal_bw_sdp]);
- dml2_printf("DML::%s: urg_bandwidth_required_dram_mbps = %f\n", __func__, urg_bandwidth_required[eval_state][dml2_core_internal_bw_dram]);
- dml2_printf("DML::%s: flip_bw_available_sdp_mbps = %f\n", __func__, flip_bw_available_sdp_mbps);
- dml2_printf("DML::%s: flip_bw_available_dram_mbps = %f\n", __func__, flip_bw_available_dram_mbps);
- dml2_printf("DML::%s: flip_bw_available_mbps = %f\n", __func__, flip_bw_available_mbps);
-#endif
-
- return flip_bw_available_mbps;
-}
-
-static void calculate_immediate_flip_bandwidth_support(
- // Output
- double *frac_urg_bandwidth_flip,
- bool *flip_bandwidth_support_ok,
-
- // Input
- enum dml2_core_internal_soc_state_type eval_state,
- double urg_bandwidth_required_flip[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double non_urg_bandwidth_required_flip[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max],
- double urg_bandwidth_available[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max])
-{
- double frac_urg_bw_flip_sdp = non_urg_bandwidth_required_flip[eval_state][dml2_core_internal_bw_sdp] / urg_bandwidth_available[eval_state][dml2_core_internal_bw_sdp];
- double frac_urg_bw_flip_dram = non_urg_bandwidth_required_flip[eval_state][dml2_core_internal_bw_dram] / urg_bandwidth_available[eval_state][dml2_core_internal_bw_dram];
-
- *flip_bandwidth_support_ok = true;
- for (unsigned int n = 0; n < dml2_core_internal_bw_max; n++) { // check sdp and dram
- *flip_bandwidth_support_ok &= urg_bandwidth_available[eval_state][n] >= urg_bandwidth_required_flip[eval_state][n];
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: n = %s\n", __func__, dml2_core_internal_bw_type_str((enum dml2_core_internal_bw_type) eval_state));
- dml2_printf("DML::%s: urg_bandwidth_available = %f\n", __func__, urg_bandwidth_available[eval_state][n]);
- dml2_printf("DML::%s: non_urg_bandwidth_required_flip = %f\n", __func__, non_urg_bandwidth_required_flip[eval_state][n]);
- dml2_printf("DML::%s: urg_bandwidth_required_flip = %f\n", __func__, urg_bandwidth_required_flip[eval_state][n]);
- dml2_printf("DML::%s: flip_bandwidth_support_ok = %d\n", __func__, *flip_bandwidth_support_ok);
-#endif
- dml2_assert(urg_bandwidth_required_flip[eval_state][n] > non_urg_bandwidth_required_flip[eval_state][n]);
- }
-
- *frac_urg_bandwidth_flip = (frac_urg_bw_flip_sdp > frac_urg_bw_flip_dram) ? frac_urg_bw_flip_sdp : frac_urg_bw_flip_dram;
- *flip_bandwidth_support_ok &= (*frac_urg_bandwidth_flip <= 1);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: eval_state = %s\n", __func__, dml2_core_internal_soc_state_type_str(eval_state));
- dml2_printf("DML::%s: frac_urg_bw_flip_sdp = %f\n", __func__, frac_urg_bw_flip_sdp);
- dml2_printf("DML::%s: frac_urg_bw_flip_dram = %f\n", __func__, frac_urg_bw_flip_dram);
- dml2_printf("DML::%s: frac_urg_bandwidth_flip = %f\n", __func__, *frac_urg_bandwidth_flip);
- dml2_printf("DML::%s: flip_bandwidth_support_ok = %d\n", __func__, *flip_bandwidth_support_ok);
-
- for (unsigned int m = 0; m < dml2_core_internal_soc_state_max; m++) {
- for (unsigned int n = 0; n < dml2_core_internal_bw_max; n++) {
- dml2_printf("DML::%s: state:%s bw_type:%s, urg_bandwidth_available=%f %s urg_bandwidth_required=%f\n",
- __func__, dml2_core_internal_soc_state_type_str(m), dml2_core_internal_bw_type_str(n),
- urg_bandwidth_available[m][n], (urg_bandwidth_available[m][n] < urg_bandwidth_required_flip[m][n]) ? "<" : ">=", urg_bandwidth_required_flip[m][n]);
- }
- }
-#endif
-}
-
-static void CalculateFlipSchedule(
- struct dml2_core_internal_scratch *s,
- bool iflip_enable,
- bool use_lb_flip_bw,
- double HostVMInefficiencyFactor,
- double Tvm_trips_flip,
- double Tr0_trips_flip,
- double Tvm_trips_flip_rounded,
- double Tr0_trips_flip_rounded,
- bool GPUVMEnable,
- double vm_bytes, // vm_bytes
- double DPTEBytesPerRow, // dpte_row_bytes
- double BandwidthAvailableForImmediateFlip,
- unsigned int TotImmediateFlipBytes,
- enum dml2_source_format_class SourcePixelFormat,
- double LineTime,
- double VRatio,
- double VRatioChroma,
- double Tno_bw_flip,
- unsigned int dpte_row_height,
- unsigned int dpte_row_height_chroma,
- bool use_one_row_for_frame_flip,
- unsigned int max_flip_time_us,
- unsigned int per_pipe_flip_bytes,
- unsigned int meta_row_bytes,
- unsigned int meta_row_height,
- unsigned int meta_row_height_chroma,
- bool dcc_mrq_enable,
-
- // Output
- double *dst_y_per_vm_flip,
- double *dst_y_per_row_flip,
- double *final_flip_bw,
- bool *ImmediateFlipSupportedForPipe)
-{
- struct dml2_core_shared_CalculateFlipSchedule_locals *l = &s->CalculateFlipSchedule_locals;
-
- l->dual_plane = dml2_core_shared_is_420(SourcePixelFormat) || SourcePixelFormat == dml2_rgbe_alpha;
- l->dpte_row_bytes = DPTEBytesPerRow;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: GPUVMEnable = %u\n", __func__, GPUVMEnable);
- dml2_printf("DML::%s: ip.max_flip_time_us = %d\n", __func__, max_flip_time_us);
- dml2_printf("DML::%s: BandwidthAvailableForImmediateFlip = %f\n", __func__, BandwidthAvailableForImmediateFlip);
- dml2_printf("DML::%s: TotImmediateFlipBytes = %u\n", __func__, TotImmediateFlipBytes);
- dml2_printf("DML::%s: use_lb_flip_bw = %u\n", __func__, use_lb_flip_bw);
- dml2_printf("DML::%s: iflip_enable = %u\n", __func__, iflip_enable);
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, HostVMInefficiencyFactor);
- dml2_printf("DML::%s: LineTime = %f\n", __func__, LineTime);
- dml2_printf("DML::%s: Tno_bw_flip = %f\n", __func__, Tno_bw_flip);
- dml2_printf("DML::%s: Tvm_trips_flip = %f\n", __func__, Tvm_trips_flip);
- dml2_printf("DML::%s: Tr0_trips_flip = %f\n", __func__, Tr0_trips_flip);
- dml2_printf("DML::%s: Tvm_trips_flip_rounded = %f\n", __func__, Tvm_trips_flip_rounded);
- dml2_printf("DML::%s: Tr0_trips_flip_rounded = %f\n", __func__, Tr0_trips_flip_rounded);
- dml2_printf("DML::%s: vm_bytes = %f\n", __func__, vm_bytes);
- dml2_printf("DML::%s: DPTEBytesPerRow = %f\n", __func__, DPTEBytesPerRow);
- dml2_printf("DML::%s: meta_row_bytes = %d\n", __func__, meta_row_bytes);
- dml2_printf("DML::%s: dpte_row_bytes = %f\n", __func__, l->dpte_row_bytes);
- dml2_printf("DML::%s: dpte_row_height = %d\n", __func__, dpte_row_height);
- dml2_printf("DML::%s: meta_row_height = %d\n", __func__, meta_row_height);
- dml2_printf("DML::%s: VRatio = %f\n", __func__, VRatio);
-#endif
-
- if (TotImmediateFlipBytes > 0 && (GPUVMEnable || dcc_mrq_enable)) {
- if (l->dual_plane) {
- if (dcc_mrq_enable & GPUVMEnable) {
- l->min_row_height = math_min2(dpte_row_height, meta_row_height);
- l->min_row_height_chroma = math_min2(dpte_row_height_chroma, meta_row_height_chroma);
- } else if (GPUVMEnable) {
- l->min_row_height = dpte_row_height;
- l->min_row_height_chroma = dpte_row_height_chroma;
- } else {
- l->min_row_height = meta_row_height;
- l->min_row_height_chroma = meta_row_height_chroma;
- }
- l->min_row_time = math_min2(l->min_row_height * LineTime / VRatio, l->min_row_height_chroma * LineTime / VRatioChroma);
- } else {
- if (dcc_mrq_enable & GPUVMEnable)
- l->min_row_height = math_min2(dpte_row_height, meta_row_height);
- else if (GPUVMEnable)
- l->min_row_height = dpte_row_height;
- else
- l->min_row_height = meta_row_height;
-
- l->min_row_time = l->min_row_height * LineTime / VRatio;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: min_row_time = %f\n", __func__, l->min_row_time);
-#endif
- dml2_assert(l->min_row_time > 0);
-
- if (use_lb_flip_bw) {
- // For mode check, calculation the flip bw requirement with worst case flip time
- l->max_flip_time = math_min2(l->min_row_time, math_max2(Tvm_trips_flip_rounded + 2 * Tr0_trips_flip_rounded, (double)max_flip_time_us));
-
- //The lower bound on flip bandwidth
- // Note: The get_urgent_bandwidth_required already consider dpte_row_bw and meta_row_bw in bandwidth calculation, so leave final_flip_bw = 0 if iflip not required
- l->lb_flip_bw = 0;
-
- if (iflip_enable) {
- l->hvm_scaled_vm_bytes = vm_bytes * HostVMInefficiencyFactor;
- l->num_rows = 2;
- l->hvm_scaled_row_bytes = (l->num_rows * l->dpte_row_bytes * HostVMInefficiencyFactor + l->num_rows * meta_row_bytes);
- l->hvm_scaled_vm_row_bytes = l->hvm_scaled_vm_bytes + l->hvm_scaled_row_bytes;
- l->lb_flip_bw = math_max3(
- l->hvm_scaled_vm_row_bytes / (l->max_flip_time - Tno_bw_flip),
- l->hvm_scaled_vm_bytes / (l->max_flip_time - Tno_bw_flip - 2 * Tr0_trips_flip_rounded),
- l->hvm_scaled_row_bytes / (l->max_flip_time - Tvm_trips_flip_rounded));
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: max_flip_time = %f\n", __func__, l->max_flip_time);
- dml2_printf("DML::%s: total vm bytes (hvm ineff scaled) = %f\n", __func__, l->hvm_scaled_vm_bytes);
- dml2_printf("DML::%s: total row bytes (hvm ineff scaled) = %f\n", __func__, l->hvm_scaled_row_bytes);
- dml2_printf("DML::%s: total vm+row bytes (hvm ineff scaled) = %f\n", __func__, l->hvm_scaled_vm_row_bytes);
- dml2_printf("DML::%s: lb_flip_bw for vm and row = %f\n", __func__, l->hvm_scaled_vm_row_bytes / (l->max_flip_time - Tno_bw_flip));
- dml2_printf("DML::%s: lb_flip_bw for vm = %f\n", __func__, l->hvm_scaled_vm_bytes / (l->max_flip_time - Tno_bw_flip - 2 * Tr0_trips_flip_rounded));
- dml2_printf("DML::%s: lb_flip_bw for row = %f\n", __func__, l->hvm_scaled_row_bytes / (l->max_flip_time - Tvm_trips_flip_rounded));
-
- if (l->lb_flip_bw > 0) {
- dml2_printf("DML::%s: mode_support est Tvm_flip = %f (bw-based)\n", __func__, Tno_bw_flip + l->hvm_scaled_vm_bytes / l->lb_flip_bw);
- dml2_printf("DML::%s: mode_support est Tr0_flip = %f (bw-based)\n", __func__, l->hvm_scaled_row_bytes / l->lb_flip_bw / l->num_rows);
- dml2_printf("DML::%s: mode_support est dst_y_per_vm_flip = %f (bw-based)\n", __func__, Tno_bw_flip + l->hvm_scaled_vm_bytes / l->lb_flip_bw / LineTime);
- dml2_printf("DML::%s: mode_support est dst_y_per_row_flip = %f (bw-based)\n", __func__, l->hvm_scaled_row_bytes / l->lb_flip_bw / LineTime / l->num_rows);
- }
-#endif
- l->lb_flip_bw = math_max3(l->lb_flip_bw,
- l->hvm_scaled_vm_bytes / (31 * LineTime) - Tno_bw_flip,
- (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / (15 * LineTime));
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: lb_flip_bw for vm reg limit = %f\n", __func__, l->hvm_scaled_vm_bytes / (31 * LineTime) - Tno_bw_flip);
- dml2_printf("DML::%s: lb_flip_bw for row reg limit = %f\n", __func__, (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / (15 * LineTime));
-#endif
- }
-
- *final_flip_bw = l->lb_flip_bw;
-
- *dst_y_per_vm_flip = 1; // not used
- *dst_y_per_row_flip = 1; // not used
- *ImmediateFlipSupportedForPipe = true;
- } else {
- if (iflip_enable) {
- l->ImmediateFlipBW = (double)per_pipe_flip_bytes * BandwidthAvailableForImmediateFlip / (double)TotImmediateFlipBytes; // flip_bw(i)
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: per_pipe_flip_bytes = %d\n", __func__, per_pipe_flip_bytes);
- dml2_printf("DML::%s: BandwidthAvailableForImmediateFlip = %f\n", __func__, BandwidthAvailableForImmediateFlip);
- dml2_printf("DML::%s: ImmediateFlipBW = %f\n", __func__, l->ImmediateFlipBW);
-#endif
- if (l->ImmediateFlipBW == 0) {
- l->Tvm_flip = 0;
- l->Tr0_flip = 0;
- } else {
- l->Tvm_flip = math_max3(Tvm_trips_flip,
- Tno_bw_flip + vm_bytes * HostVMInefficiencyFactor / l->ImmediateFlipBW,
- LineTime / 4.0);
-
- l->Tr0_flip = math_max3(Tr0_trips_flip,
- (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / l->ImmediateFlipBW,
- LineTime / 4.0);
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: total vm bytes (hvm ineff scaled) = %f\n", __func__, vm_bytes * HostVMInefficiencyFactor);
- dml2_printf("DML::%s: total row bytes (hvm ineff scaled, one row) = %f\n", __func__, (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes));
-
- dml2_printf("DML::%s: Tvm_flip = %f (bw-based), Tvm_trips_flip = %f (latency-based)\n", __func__, Tno_bw_flip + vm_bytes * HostVMInefficiencyFactor / l->ImmediateFlipBW, Tvm_trips_flip);
- dml2_printf("DML::%s: Tr0_flip = %f (bw-based), Tr0_trips_flip = %f (latency-based)\n", __func__, (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / l->ImmediateFlipBW, Tr0_trips_flip);
-#endif
- *dst_y_per_vm_flip = math_ceil2(4.0 * (l->Tvm_flip / LineTime), 1.0) / 4.0;
- *dst_y_per_row_flip = math_ceil2(4.0 * (l->Tr0_flip / LineTime), 1.0) / 4.0;
-
- *final_flip_bw = math_max2(vm_bytes * HostVMInefficiencyFactor / (*dst_y_per_vm_flip * LineTime),
- (l->dpte_row_bytes * HostVMInefficiencyFactor + meta_row_bytes) / (*dst_y_per_row_flip * LineTime));
-
- if (*dst_y_per_vm_flip >= 32 || *dst_y_per_row_flip >= 16 || l->Tvm_flip + 2 * l->Tr0_flip > l->min_row_time) {
- *ImmediateFlipSupportedForPipe = false;
- } else {
- *ImmediateFlipSupportedForPipe = iflip_enable;
- }
- } else {
- l->Tvm_flip = 0;
- l->Tr0_flip = 0;
- *dst_y_per_vm_flip = 0;
- *dst_y_per_row_flip = 0;
- *final_flip_bw = 0;
- *ImmediateFlipSupportedForPipe = iflip_enable;
- }
- }
- } else {
- l->Tvm_flip = 0;
- l->Tr0_flip = 0;
- *dst_y_per_vm_flip = 0;
- *dst_y_per_row_flip = 0;
- *final_flip_bw = 0;
- *ImmediateFlipSupportedForPipe = iflip_enable;
- }
-
-#ifdef __DML_VBA_DEBUG__
- if (!use_lb_flip_bw) {
- dml2_printf("DML::%s: dst_y_per_vm_flip = %f (should be < 32)\n", __func__, *dst_y_per_vm_flip);
- dml2_printf("DML::%s: dst_y_per_row_flip = %f (should be < 16)\n", __func__, *dst_y_per_row_flip);
- dml2_printf("DML::%s: Tvm_flip = %f (final)\n", __func__, l->Tvm_flip);
- dml2_printf("DML::%s: Tr0_flip = %f (final)\n", __func__, l->Tr0_flip);
- }
- dml2_printf("DML::%s: final_flip_bw = %f\n", __func__, *final_flip_bw);
- dml2_printf("DML::%s: ImmediateFlipSupportedForPipe = %u\n", __func__, *ImmediateFlipSupportedForPipe);
-#endif
-}
-
-static void CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(
- struct dml2_core_internal_scratch *scratch,
- struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params *p)
-{
- struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_locals *s = &scratch->CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_locals;
-
- s->TotalActiveWriteback = 0;
- p->Watermark->UrgentWatermark = p->mmSOCParameters.UrgentLatency + p->mmSOCParameters.ExtraLatency;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: UrgentWatermark = %f\n", __func__, p->Watermark->UrgentWatermark);
-#endif
-
- p->Watermark->USRRetrainingWatermark = p->mmSOCParameters.UrgentLatency + p->mmSOCParameters.ExtraLatency + p->mmSOCParameters.USRRetrainingLatency + p->mmSOCParameters.SMNLatency;
- p->Watermark->DRAMClockChangeWatermark = p->mmSOCParameters.DRAMClockChangeLatency + p->Watermark->UrgentWatermark;
- p->Watermark->FCLKChangeWatermark = p->mmSOCParameters.FCLKChangeLatency + p->Watermark->UrgentWatermark;
- p->Watermark->StutterExitWatermark = p->mmSOCParameters.SRExitTime + p->mmSOCParameters.ExtraLatency_sr + 10 / p->DCFClkDeepSleep;
- p->Watermark->StutterEnterPlusExitWatermark = p->mmSOCParameters.SREnterPlusExitTime + p->mmSOCParameters.ExtraLatency_sr + 10 / p->DCFClkDeepSleep;
- p->Watermark->Z8StutterExitWatermark = p->mmSOCParameters.SRExitZ8Time + p->mmSOCParameters.ExtraLatency_sr + 10 / p->DCFClkDeepSleep;
- p->Watermark->Z8StutterEnterPlusExitWatermark = p->mmSOCParameters.SREnterPlusExitZ8Time + p->mmSOCParameters.ExtraLatency_sr + 10 / p->DCFClkDeepSleep;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: UrgentLatency = %f\n", __func__, p->mmSOCParameters.UrgentLatency);
- dml2_printf("DML::%s: ExtraLatency = %f\n", __func__, p->mmSOCParameters.ExtraLatency);
- dml2_printf("DML::%s: DRAMClockChangeLatency = %f\n", __func__, p->mmSOCParameters.DRAMClockChangeLatency);
- dml2_printf("DML::%s: SREnterPlusExitZ8Time = %f\n", __func__, p->mmSOCParameters.SREnterPlusExitZ8Time);
- dml2_printf("DML::%s: SREnterPlusExitTime = %f\n", __func__, p->mmSOCParameters.SREnterPlusExitTime);
- dml2_printf("DML::%s: UrgentWatermark = %f\n", __func__, p->Watermark->UrgentWatermark);
- dml2_printf("DML::%s: USRRetrainingWatermark = %f\n", __func__, p->Watermark->USRRetrainingWatermark);
- dml2_printf("DML::%s: DRAMClockChangeWatermark = %f\n", __func__, p->Watermark->DRAMClockChangeWatermark);
- dml2_printf("DML::%s: FCLKChangeWatermark = %f\n", __func__, p->Watermark->FCLKChangeWatermark);
- dml2_printf("DML::%s: StutterExitWatermark = %f\n", __func__, p->Watermark->StutterExitWatermark);
- dml2_printf("DML::%s: StutterEnterPlusExitWatermark = %f\n", __func__, p->Watermark->StutterEnterPlusExitWatermark);
- dml2_printf("DML::%s: Z8StutterExitWatermark = %f\n", __func__, p->Watermark->Z8StutterExitWatermark);
- dml2_printf("DML::%s: Z8StutterEnterPlusExitWatermark = %f\n", __func__, p->Watermark->Z8StutterEnterPlusExitWatermark);
-#endif
-
- s->TotalActiveWriteback = 0;
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- s->TotalActiveWriteback = s->TotalActiveWriteback + 1;
- }
- }
-
- if (s->TotalActiveWriteback <= 1) {
- p->Watermark->WritebackUrgentWatermark = p->mmSOCParameters.WritebackLatency;
- } else {
- p->Watermark->WritebackUrgentWatermark = p->mmSOCParameters.WritebackLatency + p->WritebackChunkSize * 1024.0 / 32.0 / p->SOCCLK;
- }
- if (p->USRRetrainingRequired)
- p->Watermark->WritebackUrgentWatermark = p->Watermark->WritebackUrgentWatermark + p->mmSOCParameters.USRRetrainingLatency;
-
- if (s->TotalActiveWriteback <= 1) {
- p->Watermark->WritebackDRAMClockChangeWatermark = p->mmSOCParameters.DRAMClockChangeLatency + p->mmSOCParameters.WritebackLatency;
- p->Watermark->WritebackFCLKChangeWatermark = p->mmSOCParameters.FCLKChangeLatency + p->mmSOCParameters.WritebackLatency;
- } else {
- p->Watermark->WritebackDRAMClockChangeWatermark = p->mmSOCParameters.DRAMClockChangeLatency + p->mmSOCParameters.WritebackLatency + p->WritebackChunkSize * 1024.0 / 32.0 / p->SOCCLK;
- p->Watermark->WritebackFCLKChangeWatermark = p->mmSOCParameters.FCLKChangeLatency + p->mmSOCParameters.WritebackLatency + p->WritebackChunkSize * 1024 / 32 / p->SOCCLK;
- }
-
- if (p->USRRetrainingRequired)
- p->Watermark->WritebackDRAMClockChangeWatermark = p->Watermark->WritebackDRAMClockChangeWatermark + p->mmSOCParameters.USRRetrainingLatency;
-
- if (p->USRRetrainingRequired)
- p->Watermark->WritebackFCLKChangeWatermark = p->Watermark->WritebackFCLKChangeWatermark + p->mmSOCParameters.USRRetrainingLatency;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: WritebackDRAMClockChangeWatermark = %f\n", __func__, p->Watermark->WritebackDRAMClockChangeWatermark);
- dml2_printf("DML::%s: WritebackFCLKChangeWatermark = %f\n", __func__, p->Watermark->WritebackFCLKChangeWatermark);
- dml2_printf("DML::%s: WritebackUrgentWatermark = %f\n", __func__, p->Watermark->WritebackUrgentWatermark);
- dml2_printf("DML::%s: USRRetrainingRequired = %u\n", __func__, p->USRRetrainingRequired);
- dml2_printf("DML::%s: USRRetrainingLatency = %f\n", __func__, p->mmSOCParameters.USRRetrainingLatency);
-#endif
-
- s->TotalPixelBW = 0.0;
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- double h_total = (double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total;
- double pixel_clock_mhz = p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000.0;
- double v_ratio = p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- double v_ratio_c = p->display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
-
- s->TotalPixelBW = s->TotalPixelBW + p->DPPPerSurface[k]
- * (p->SwathWidthY[k] * p->BytePerPixelDETY[k] * v_ratio + p->SwathWidthC[k] * p->BytePerPixelDETC[k] * v_ratio_c) / (h_total / pixel_clock_mhz);
- }
-
- *p->global_fclk_change_supported = true;
- *p->global_dram_clock_change_supported = true;
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- double h_total = (double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total;
- double pixel_clock_mhz = p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000.0;
- double v_ratio = p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- double v_ratio_c = p->display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- double v_taps = p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps;
- double v_taps_c = p->display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps;
- double h_ratio = p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio;
- double h_ratio_c = p->display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio;
- double LBBitPerPixel = 57;
-
- s->LBLatencyHidingSourceLinesY[k] = (unsigned int)(math_min2((double)p->MaxLineBufferLines, math_floor2((double)p->LineBufferSize / LBBitPerPixel / ((double)p->SwathWidthY[k] / math_max2(h_ratio, 1.0)), 1)) - (v_taps - 1));
- s->LBLatencyHidingSourceLinesC[k] = (unsigned int)(math_min2((double)p->MaxLineBufferLines, math_floor2((double)p->LineBufferSize / LBBitPerPixel / ((double)p->SwathWidthC[k] / math_max2(h_ratio_c, 1.0)), 1)) - (v_taps_c - 1));
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, MaxLineBufferLines= %u\n", __func__, k, p->MaxLineBufferLines);
- dml2_printf("DML::%s: k=%u, LineBufferSize = %u\n", __func__, k, p->LineBufferSize);
- dml2_printf("DML::%s: k=%u, LBBitPerPixel = %u\n", __func__, k, LBBitPerPixel);
- dml2_printf("DML::%s: k=%u, HRatio = %f\n", __func__, k, h_ratio);
- dml2_printf("DML::%s: k=%u, VTaps = %f\n", __func__, k, v_taps);
-#endif
-
- s->EffectiveLBLatencyHidingY = s->LBLatencyHidingSourceLinesY[k] / v_ratio * (h_total / pixel_clock_mhz);
- s->EffectiveLBLatencyHidingC = s->LBLatencyHidingSourceLinesC[k] / v_ratio_c * (h_total / pixel_clock_mhz);
-
- s->EffectiveDETBufferSizeY = p->DETBufferSizeY[k];
- if (p->UnboundedRequestEnabled) {
- s->EffectiveDETBufferSizeY = s->EffectiveDETBufferSizeY + p->CompressedBufferSizeInkByte * 1024 * (p->SwathWidthY[k] * p->BytePerPixelDETY[k] * v_ratio) / (h_total / pixel_clock_mhz) / s->TotalPixelBW;
- }
-
- s->LinesInDETY[k] = (double)s->EffectiveDETBufferSizeY / p->BytePerPixelDETY[k] / p->SwathWidthY[k];
- s->LinesInDETYRoundedDownToSwath[k] = (unsigned int)(math_floor2(s->LinesInDETY[k], p->SwathHeightY[k]));
- s->FullDETBufferingTimeY = s->LinesInDETYRoundedDownToSwath[k] * (h_total / pixel_clock_mhz) / v_ratio;
-
- s->ActiveClockChangeLatencyHidingY = s->EffectiveLBLatencyHidingY + s->FullDETBufferingTimeY - ((double)p->DSTXAfterScaler[k] / h_total + (double)p->DSTYAfterScaler[k]) * h_total / pixel_clock_mhz;
-
- if (p->NumberOfActiveSurfaces > 1) {
- s->ActiveClockChangeLatencyHidingY = s->ActiveClockChangeLatencyHidingY - (1.0 - 1.0 / (double)p->NumberOfActiveSurfaces) * (double)p->SwathHeightY[k] * (double)h_total / pixel_clock_mhz / v_ratio;
- }
-
- if (p->BytePerPixelDETC[k] > 0) {
- s->LinesInDETC[k] = p->DETBufferSizeC[k] / p->BytePerPixelDETC[k] / p->SwathWidthC[k];
- s->LinesInDETCRoundedDownToSwath[k] = (unsigned int)(math_floor2(s->LinesInDETC[k], p->SwathHeightC[k]));
- s->FullDETBufferingTimeC = s->LinesInDETCRoundedDownToSwath[k] * (h_total / pixel_clock_mhz) / v_ratio_c;
- s->ActiveClockChangeLatencyHidingC = s->EffectiveLBLatencyHidingC + s->FullDETBufferingTimeC - ((double)p->DSTXAfterScaler[k] / (double)h_total + (double)p->DSTYAfterScaler[k]) * (double)h_total / pixel_clock_mhz;
- if (p->NumberOfActiveSurfaces > 1) {
- s->ActiveClockChangeLatencyHidingC = s->ActiveClockChangeLatencyHidingC - (1.0 - 1.0 / (double)p->NumberOfActiveSurfaces) * (double)p->SwathHeightC[k] * (double)h_total / pixel_clock_mhz / v_ratio_c;
- }
- s->ActiveClockChangeLatencyHiding = math_min2(s->ActiveClockChangeLatencyHidingY, s->ActiveClockChangeLatencyHidingC);
- } else {
- s->ActiveClockChangeLatencyHiding = s->ActiveClockChangeLatencyHidingY;
- }
-
- s->ActiveDRAMClockChangeLatencyMargin[k] = s->ActiveClockChangeLatencyHiding - p->Watermark->DRAMClockChangeWatermark;
- s->ActiveFCLKChangeLatencyMargin[k] = s->ActiveClockChangeLatencyHiding - p->Watermark->FCLKChangeWatermark;
- s->USRRetrainingLatencyMargin[k] = s->ActiveClockChangeLatencyHiding - p->Watermark->USRRetrainingWatermark;
-
- if (p->VActiveLatencyHidingMargin)
- p->VActiveLatencyHidingMargin[k] = s->ActiveDRAMClockChangeLatencyMargin[k];
-
- p->VActiveLatencyHidingUs[k] = s->ActiveClockChangeLatencyHiding;
-
- if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.enable) {
- s->WritebackLatencyHiding = (double)p->WritebackInterfaceBufferSize * 1024.0 / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height * (double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height * (double)h_total / pixel_clock_mhz) * 4.0);
- if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format == dml2_444_64) {
- s->WritebackLatencyHiding = s->WritebackLatencyHiding / 2;
- }
- s->WritebackDRAMClockChangeLatencyMargin = s->WritebackLatencyHiding - p->Watermark->WritebackDRAMClockChangeWatermark;
-
- s->WritebackFCLKChangeLatencyMargin = s->WritebackLatencyHiding - p->Watermark->WritebackFCLKChangeWatermark;
-
- s->ActiveDRAMClockChangeLatencyMargin[k] = math_min2(s->ActiveDRAMClockChangeLatencyMargin[k], s->WritebackDRAMClockChangeLatencyMargin);
- s->ActiveFCLKChangeLatencyMargin[k] = math_min2(s->ActiveFCLKChangeLatencyMargin[k], s->WritebackFCLKChangeLatencyMargin);
- }
- p->MaxActiveDRAMClockChangeLatencySupported[k] = dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k]) ? 0 : (s->ActiveDRAMClockChangeLatencyMargin[k] + p->mmSOCParameters.DRAMClockChangeLatency);
-
- enum dml2_uclk_pstate_change_strategy uclk_pstate_change_strategy = p->display_cfg->plane_descriptors[k].overrides.uclk_pstate_change_strategy;
- double reserved_vblank_time_us = (double)p->display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns / 1000;
-
- p->FCLKChangeSupport[k] = dml2_fclock_change_unsupported;
- if (s->ActiveFCLKChangeLatencyMargin[k] > 0)
- p->FCLKChangeSupport[k] = dml2_fclock_change_vactive;
- else if (reserved_vblank_time_us >= p->mmSOCParameters.FCLKChangeLatency)
- p->FCLKChangeSupport[k] = dml2_fclock_change_vblank;
-
- if (p->FCLKChangeSupport[k] == dml2_fclock_change_unsupported)
- *p->global_fclk_change_supported = false;
-
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_unsupported;
- if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_auto) {
- if (s->ActiveDRAMClockChangeLatencyMargin[k] > 0 && reserved_vblank_time_us >= p->mmSOCParameters.DRAMClockChangeLatency)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vblank_and_vactive;
- else if (s->ActiveDRAMClockChangeLatencyMargin[k] > 0)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vactive;
- else if (reserved_vblank_time_us >= p->mmSOCParameters.DRAMClockChangeLatency)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vblank;
- } else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_vactive && s->ActiveDRAMClockChangeLatencyMargin[k] > 0)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vactive;
- else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_vblank && reserved_vblank_time_us >= p->mmSOCParameters.DRAMClockChangeLatency)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_vblank;
- else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_drr)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_drr;
- else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_mall_svp)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_mall_svp;
- else if (uclk_pstate_change_strategy == dml2_uclk_pstate_change_strategy_force_mall_full_frame)
- p->DRAMClockChangeSupport[k] = dml2_dram_clock_change_mall_full_frame;
-
- if (p->DRAMClockChangeSupport[k] == dml2_dram_clock_change_unsupported)
- *p->global_dram_clock_change_supported = false;
-
- s->dst_y_pstate = (unsigned int)(math_ceil2((p->mmSOCParameters.DRAMClockChangeLatency + p->mmSOCParameters.UrgentLatency) / (h_total / pixel_clock_mhz), 1));
- s->src_y_pstate_l = (unsigned int)(math_ceil2(s->dst_y_pstate * v_ratio, p->SwathHeightY[k]));
- s->src_y_ahead_l = (unsigned int)(math_floor2(p->DETBufferSizeY[k] / p->BytePerPixelDETY[k] / p->SwathWidthY[k], p->SwathHeightY[k]) + s->LBLatencyHidingSourceLinesY[k]);
- s->sub_vp_lines_l = s->src_y_pstate_l + s->src_y_ahead_l + p->meta_row_height_l[k];
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, DETBufferSizeY = %u\n", __func__, k, p->DETBufferSizeY[k]);
- dml2_printf("DML::%s: k=%u, BytePerPixelDETY = %f\n", __func__, k, p->BytePerPixelDETY[k]);
- dml2_printf("DML::%s: k=%u, SwathWidthY = %u\n", __func__, k, p->SwathWidthY[k]);
- dml2_printf("DML::%s: k=%u, SwathHeightY = %u\n", __func__, k, p->SwathHeightY[k]);
- dml2_printf("DML::%s: k=%u, LBLatencyHidingSourceLinesY = %u\n", __func__, k, s->LBLatencyHidingSourceLinesY[k]);
- dml2_printf("DML::%s: k=%u, dst_y_pstate = %u\n", __func__, k, s->dst_y_pstate);
- dml2_printf("DML::%s: k=%u, src_y_pstate_l = %u\n", __func__, k, s->src_y_pstate_l);
- dml2_printf("DML::%s: k=%u, src_y_ahead_l = %u\n", __func__, k, s->src_y_ahead_l);
- dml2_printf("DML::%s: k=%u, meta_row_height_l = %u\n", __func__, p->meta_row_height_l[k]);
- dml2_printf("DML::%s: k=%u, sub_vp_lines_l = %u\n", __func__, k, s->sub_vp_lines_l);
-#endif
- p->SubViewportLinesNeededInMALL[k] = s->sub_vp_lines_l;
-
- if (p->BytePerPixelDETC[k] > 0) {
- s->src_y_pstate_c = (unsigned int)(math_ceil2(s->dst_y_pstate * v_ratio_c, p->SwathHeightC[k]));
- s->src_y_ahead_c = (unsigned int)(math_floor2(p->DETBufferSizeC[k] / p->BytePerPixelDETC[k] / p->SwathWidthC[k], p->SwathHeightC[k]) + s->LBLatencyHidingSourceLinesC[k]);
- s->sub_vp_lines_c = s->src_y_pstate_c + s->src_y_ahead_c + p->meta_row_height_c[k];
-
- if (dml2_core_shared_is_420(p->display_cfg->plane_descriptors[k].pixel_format))
- p->SubViewportLinesNeededInMALL[k] = (unsigned int)(math_max2(s->sub_vp_lines_l, 2 * s->sub_vp_lines_c));
- else
- p->SubViewportLinesNeededInMALL[k] = (unsigned int)(math_max2(s->sub_vp_lines_l, s->sub_vp_lines_c));
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, meta_row_height_c = %u\n", __func__, p->meta_row_height_c[k]);
- dml2_printf("DML::%s: k=%u, src_y_pstate_c = %u\n", __func__, k, s->src_y_pstate_c);
- dml2_printf("DML::%s: k=%u, src_y_ahead_c = %u\n", __func__, k, s->src_y_ahead_c);
- dml2_printf("DML::%s: k=%u, sub_vp_lines_c = %u\n", __func__, k, s->sub_vp_lines_c);
-#endif
- }
- }
-
- bool FoundCriticalSurface = false;
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if ((!dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) && ((!FoundCriticalSurface)
- || ((s->ActiveFCLKChangeLatencyMargin[k] + p->mmSOCParameters.FCLKChangeLatency) < *p->MaxActiveFCLKChangeLatencySupported))) {
- FoundCriticalSurface = true;
- *p->MaxActiveFCLKChangeLatencySupported = s->ActiveFCLKChangeLatencyMargin[k] + p->mmSOCParameters.FCLKChangeLatency;
- }
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: DRAMClockChangeSupport = %u\n", __func__, *p->global_dram_clock_change_supported);
- dml2_printf("DML::%s: FCLKChangeSupport = %u\n", __func__, *p->global_fclk_change_supported);
- dml2_printf("DML::%s: MaxActiveFCLKChangeLatencySupported = %f\n", __func__, *p->MaxActiveFCLKChangeLatencySupported);
- dml2_printf("DML::%s: USRRetrainingSupport = %u\n", __func__, *p->USRRetrainingSupport);
-#endif
-}
-
-static double uclk_khz_to_dram_bw_mbps(unsigned long uclk_khz, const struct dml2_dram_params *dram_config)
-{
- double bw_mbps = 0;
- bw_mbps = ((double)uclk_khz * dram_config->channel_count * dram_config->channel_width_bytes * dram_config->transactions_per_clock) / 1000.0;
-
- return bw_mbps;
-}
-
-static double dram_bw_kbps_to_uclk_mhz(unsigned long long bw_kbps, const struct dml2_dram_params *dram_config)
-{
- double uclk_mhz = 0;
-
- uclk_mhz = (double)bw_kbps / (dram_config->channel_count * dram_config->channel_width_bytes * dram_config->transactions_per_clock) / 1000.0;
-
- return uclk_mhz;
-}
-
-static unsigned int get_qos_param_index(unsigned long uclk_freq_khz, const struct dml2_dcn4_uclk_dpm_dependent_qos_params *per_uclk_dpm_params)
-{
- unsigned int i;
- unsigned int index = 0;
-
- for (i = 0; i < DML_MAX_CLK_TABLE_SIZE; i++) {
- dml2_printf("DML::%s: per_uclk_dpm_params[%d].minimum_uclk_khz = %d\n", __func__, i, per_uclk_dpm_params[i].minimum_uclk_khz);
-
- if (i == 0)
- index = 0;
- else
- index = i - 1;
-
- if (uclk_freq_khz < per_uclk_dpm_params[i].minimum_uclk_khz ||
- per_uclk_dpm_params[i].minimum_uclk_khz == 0) {
- break;
- }
- }
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: uclk_freq_khz = %d\n", __func__, uclk_freq_khz);
- dml2_printf("DML::%s: index = %d\n", __func__, index);
-#endif
- return index;
-}
-
-static unsigned int get_active_min_uclk_dpm_index(unsigned long uclk_freq_khz, const struct dml2_soc_state_table *clk_table)
-{
- unsigned int i;
- bool clk_entry_found = 0;
-
- for (i = 0; i < clk_table->uclk.num_clk_values; i++) {
- dml2_printf("DML::%s: clk_table.uclk.clk_values_khz[%d] = %d\n", __func__, i, clk_table->uclk.clk_values_khz[i]);
-
- if (uclk_freq_khz == clk_table->uclk.clk_values_khz[i]) {
- clk_entry_found = 1;
- break;
- }
- }
-
- dml2_assert(clk_entry_found);
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: uclk_freq_khz = %ld\n", __func__, uclk_freq_khz);
- dml2_printf("DML::%s: index = %d\n", __func__, i);
-#endif
- return i;
-}
-
-static unsigned int get_pipe_flip_bytes(
- double hostvm_inefficiency_factor,
- unsigned int vm_bytes,
- unsigned int dpte_row_bytes,
- unsigned int meta_row_bytes)
-{
- unsigned int flip_bytes = 0;
-
- flip_bytes += (unsigned int)((vm_bytes * hostvm_inefficiency_factor) + 2 * meta_row_bytes);
- flip_bytes += (unsigned int)(2 * dpte_row_bytes * hostvm_inefficiency_factor);
-
- return flip_bytes;
-}
-
-static void calculate_hostvm_inefficiency_factor(
- double *HostVMInefficiencyFactor,
- double *HostVMInefficiencyFactorPrefetch,
-
- bool gpuvm_enable,
- bool hostvm_enable,
- unsigned int remote_iommu_outstanding_translations,
- unsigned int max_outstanding_reqs,
- double urg_bandwidth_avail_active_pixel_and_vm,
- double urg_bandwidth_avail_active_vm_only)
-{
- *HostVMInefficiencyFactor = 1;
- *HostVMInefficiencyFactorPrefetch = 1;
-
- if (gpuvm_enable && hostvm_enable) {
- *HostVMInefficiencyFactor = urg_bandwidth_avail_active_pixel_and_vm / urg_bandwidth_avail_active_vm_only;
- *HostVMInefficiencyFactorPrefetch = *HostVMInefficiencyFactor;
-
- if ((*HostVMInefficiencyFactorPrefetch < 4) && (remote_iommu_outstanding_translations < max_outstanding_reqs))
- *HostVMInefficiencyFactorPrefetch = 4;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: urg_bandwidth_avail_active_pixel_and_vm = %f\n", __func__, urg_bandwidth_avail_active_pixel_and_vm);
- dml2_printf("DML::%s: urg_bandwidth_avail_active_vm_only = %f\n", __func__, urg_bandwidth_avail_active_vm_only);
- dml2_printf("DML::%s: HostVMInefficiencyFactor = %f\n", __func__, *HostVMInefficiencyFactor);
- dml2_printf("DML::%s: HostVMInefficiencyFactorPrefetch = %f\n", __func__, *HostVMInefficiencyFactorPrefetch);
-#endif
- }
-}
-
-static void CalculatePixelDeliveryTimes(
- const struct dml2_display_cfg *display_cfg,
- const struct core_display_cfg_support_info *cfg_support_info,
- unsigned int NumberOfActiveSurfaces,
- double VRatioPrefetchY[],
- double VRatioPrefetchC[],
- unsigned int swath_width_luma_ub[],
- unsigned int swath_width_chroma_ub[],
- double PSCL_THROUGHPUT[],
- double PSCL_THROUGHPUT_CHROMA[],
- double Dppclk[],
- unsigned int BytePerPixelC[],
- unsigned int req_per_swath_ub_l[],
- unsigned int req_per_swath_ub_c[],
-
- // Output
- double DisplayPipeLineDeliveryTimeLuma[],
- double DisplayPipeLineDeliveryTimeChroma[],
- double DisplayPipeLineDeliveryTimeLumaPrefetch[],
- double DisplayPipeLineDeliveryTimeChromaPrefetch[],
- double DisplayPipeRequestDeliveryTimeLuma[],
- double DisplayPipeRequestDeliveryTimeChroma[],
- double DisplayPipeRequestDeliveryTimeLumaPrefetch[],
- double DisplayPipeRequestDeliveryTimeChromaPrefetch[])
-{
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- double pixel_clock_mhz = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u : HRatio = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
- dml2_printf("DML::%s: k=%u : VRatio = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
- dml2_printf("DML::%s: k=%u : HRatioChroma = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio);
- dml2_printf("DML::%s: k=%u : VRatioChroma = %f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio);
- dml2_printf("DML::%s: k=%u : VRatioPrefetchY = %f\n", __func__, k, VRatioPrefetchY[k]);
- dml2_printf("DML::%s: k=%u : VRatioPrefetchC = %f\n", __func__, k, VRatioPrefetchC[k]);
- dml2_printf("DML::%s: k=%u : swath_width_luma_ub = %u\n", __func__, k, swath_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u : swath_width_chroma_ub = %u\n", __func__, k, swath_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u : PSCL_THROUGHPUT = %f\n", __func__, k, PSCL_THROUGHPUT[k]);
- dml2_printf("DML::%s: k=%u : PSCL_THROUGHPUT_CHROMA = %f\n", __func__, k, PSCL_THROUGHPUT_CHROMA[k]);
- dml2_printf("DML::%s: k=%u : DPPPerSurface = %u\n", __func__, k, cfg_support_info->plane_support_info[k].dpps_used);
- dml2_printf("DML::%s: k=%u : pixel_clock_mhz = %f\n", __func__, k, pixel_clock_mhz);
- dml2_printf("DML::%s: k=%u : Dppclk = %f\n", __func__, k, Dppclk[k]);
-#endif
- if (display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio <= 1) {
- DisplayPipeLineDeliveryTimeLuma[k] = swath_width_luma_ub[k] * cfg_support_info->plane_support_info[k].dpps_used / display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio / pixel_clock_mhz;
- } else {
- DisplayPipeLineDeliveryTimeLuma[k] = swath_width_luma_ub[k] / PSCL_THROUGHPUT[k] / Dppclk[k];
- }
-
- if (BytePerPixelC[k] == 0) {
- DisplayPipeLineDeliveryTimeChroma[k] = 0;
- } else {
- if (display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio <= 1) {
- DisplayPipeLineDeliveryTimeChroma[k] = swath_width_chroma_ub[k] * cfg_support_info->plane_support_info[k].dpps_used / display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio / pixel_clock_mhz;
- } else {
- DisplayPipeLineDeliveryTimeChroma[k] = swath_width_chroma_ub[k] / PSCL_THROUGHPUT_CHROMA[k] / Dppclk[k];
- }
- }
-
- if (VRatioPrefetchY[k] <= 1) {
- DisplayPipeLineDeliveryTimeLumaPrefetch[k] = swath_width_luma_ub[k] * cfg_support_info->plane_support_info[k].dpps_used / display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio / pixel_clock_mhz;
- } else {
- DisplayPipeLineDeliveryTimeLumaPrefetch[k] = swath_width_luma_ub[k] / PSCL_THROUGHPUT[k] / Dppclk[k];
- }
-
- if (BytePerPixelC[k] == 0) {
- DisplayPipeLineDeliveryTimeChromaPrefetch[k] = 0;
- } else {
- if (VRatioPrefetchC[k] <= 1) {
- DisplayPipeLineDeliveryTimeChromaPrefetch[k] = swath_width_chroma_ub[k] * cfg_support_info->plane_support_info[k].dpps_used / display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio / pixel_clock_mhz;
- } else {
- DisplayPipeLineDeliveryTimeChromaPrefetch[k] = swath_width_chroma_ub[k] / PSCL_THROUGHPUT_CHROMA[k] / Dppclk[k];
- }
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u : DisplayPipeLineDeliveryTimeLuma = %f\n", __func__, k, DisplayPipeLineDeliveryTimeLuma[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeLineDeliveryTimeLumaPrefetch = %f\n", __func__, k, DisplayPipeLineDeliveryTimeLumaPrefetch[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeLineDeliveryTimeChroma = %f\n", __func__, k, DisplayPipeLineDeliveryTimeChroma[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeLineDeliveryTimeChromaPrefetch = %f\n", __func__, k, DisplayPipeLineDeliveryTimeChromaPrefetch[k]);
-#endif
- }
-
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
-
- DisplayPipeRequestDeliveryTimeLuma[k] = DisplayPipeLineDeliveryTimeLuma[k] / req_per_swath_ub_l[k];
- DisplayPipeRequestDeliveryTimeLumaPrefetch[k] = DisplayPipeLineDeliveryTimeLumaPrefetch[k] / req_per_swath_ub_l[k];
- if (BytePerPixelC[k] == 0) {
- DisplayPipeRequestDeliveryTimeChroma[k] = 0;
- DisplayPipeRequestDeliveryTimeChromaPrefetch[k] = 0;
- } else {
- DisplayPipeRequestDeliveryTimeChroma[k] = DisplayPipeLineDeliveryTimeChroma[k] / req_per_swath_ub_c[k];
- DisplayPipeRequestDeliveryTimeChromaPrefetch[k] = DisplayPipeLineDeliveryTimeChromaPrefetch[k] / req_per_swath_ub_c[k];
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeLuma = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeLuma[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeLumaPrefetch = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeLumaPrefetch[k]);
- dml2_printf("DML::%s: k=%u : req_per_swath_ub_l = %d\n", __func__, k, req_per_swath_ub_l[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeChroma = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeChroma[k]);
- dml2_printf("DML::%s: k=%u : DisplayPipeRequestDeliveryTimeChromaPrefetch = %f\n", __func__, k, DisplayPipeRequestDeliveryTimeChromaPrefetch[k]);
- dml2_printf("DML::%s: k=%u : req_per_swath_ub_c = %d\n", __func__, k, req_per_swath_ub_c[k]);
-#endif
- }
-}
-
-static void CalculateMetaAndPTETimes(struct dml2_core_shared_CalculateMetaAndPTETimes_params *p)
-{
- unsigned int meta_chunk_width;
- unsigned int min_meta_chunk_width;
- unsigned int meta_chunk_per_row_int;
- unsigned int meta_row_remainder;
- unsigned int meta_chunk_threshold;
- unsigned int meta_chunks_per_row_ub;
- unsigned int meta_chunk_width_chroma;
- unsigned int min_meta_chunk_width_chroma;
- unsigned int meta_chunk_per_row_int_chroma;
- unsigned int meta_row_remainder_chroma;
- unsigned int meta_chunk_threshold_chroma;
- unsigned int meta_chunks_per_row_ub_chroma;
- unsigned int dpte_group_width_luma;
- unsigned int dpte_groups_per_row_luma_ub;
- unsigned int dpte_group_width_chroma;
- unsigned int dpte_groups_per_row_chroma_ub;
- double pixel_clock_mhz;
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- p->DST_Y_PER_PTE_ROW_NOM_L[k] = p->dpte_row_height[k] / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- if (p->BytePerPixelC[k] == 0) {
- p->DST_Y_PER_PTE_ROW_NOM_C[k] = 0;
- } else {
- p->DST_Y_PER_PTE_ROW_NOM_C[k] = p->dpte_row_height_chroma[k] / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- }
- p->DST_Y_PER_META_ROW_NOM_L[k] = p->meta_row_height[k] / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- if (p->BytePerPixelC[k] == 0) {
- p->DST_Y_PER_META_ROW_NOM_C[k] = 0;
- } else {
- p->DST_Y_PER_META_ROW_NOM_C[k] = p->meta_row_height_chroma[k] / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- }
- }
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (p->display_cfg->plane_descriptors[k].surface.dcc.enable == true && p->mrq_present) {
- meta_chunk_width = p->MetaChunkSize * 1024 * 256 / p->BytePerPixelY[k] / p->meta_row_height[k];
- min_meta_chunk_width = p->MinMetaChunkSizeBytes * 256 / p->BytePerPixelY[k] / p->meta_row_height[k];
- meta_chunk_per_row_int = p->meta_row_width[k] / meta_chunk_width;
- meta_row_remainder = p->meta_row_width[k] % meta_chunk_width;
- if (!dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle)) {
- meta_chunk_threshold = 2 * min_meta_chunk_width - p->meta_req_width[k];
- } else {
- meta_chunk_threshold = 2 * min_meta_chunk_width - p->meta_req_height[k];
- }
- if (meta_row_remainder <= meta_chunk_threshold) {
- meta_chunks_per_row_ub = meta_chunk_per_row_int + 1;
- } else {
- meta_chunks_per_row_ub = meta_chunk_per_row_int + 2;
- }
- p->TimePerMetaChunkNominal[k] = p->meta_row_height[k] / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio *
- p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total /
- (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) / meta_chunks_per_row_ub;
- p->TimePerMetaChunkVBlank[k] = p->dst_y_per_row_vblank[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total /
- (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) / meta_chunks_per_row_ub;
- p->TimePerMetaChunkFlip[k] = p->dst_y_per_row_flip[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total /
- (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) / meta_chunks_per_row_ub;
- if (p->BytePerPixelC[k] == 0) {
- p->TimePerChromaMetaChunkNominal[k] = 0;
- p->TimePerChromaMetaChunkVBlank[k] = 0;
- p->TimePerChromaMetaChunkFlip[k] = 0;
- } else {
- meta_chunk_width_chroma = p->MetaChunkSize * 1024 * 256 / p->BytePerPixelC[k] / p->meta_row_height_chroma[k];
- min_meta_chunk_width_chroma = p->MinMetaChunkSizeBytes * 256 / p->BytePerPixelC[k] / p->meta_row_height_chroma[k];
- meta_chunk_per_row_int_chroma = (unsigned int)((double)p->meta_row_width_chroma[k] / meta_chunk_width_chroma);
- meta_row_remainder_chroma = p->meta_row_width_chroma[k] % meta_chunk_width_chroma;
- if (!dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle)) {
- meta_chunk_threshold_chroma = 2 * min_meta_chunk_width_chroma - p->meta_req_width_chroma[k];
- } else {
- meta_chunk_threshold_chroma = 2 * min_meta_chunk_width_chroma - p->meta_req_height_chroma[k];
- }
- if (meta_row_remainder_chroma <= meta_chunk_threshold_chroma) {
- meta_chunks_per_row_ub_chroma = meta_chunk_per_row_int_chroma + 1;
- } else {
- meta_chunks_per_row_ub_chroma = meta_chunk_per_row_int_chroma + 2;
- }
- p->TimePerChromaMetaChunkNominal[k] = p->meta_row_height_chroma[k] / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) / meta_chunks_per_row_ub_chroma;
- p->TimePerChromaMetaChunkVBlank[k] = p->dst_y_per_row_vblank[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) / meta_chunks_per_row_ub_chroma;
- p->TimePerChromaMetaChunkFlip[k] = p->dst_y_per_row_flip[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) / meta_chunks_per_row_ub_chroma;
- }
- } else {
- p->TimePerMetaChunkNominal[k] = 0;
- p->TimePerMetaChunkVBlank[k] = 0;
- p->TimePerMetaChunkFlip[k] = 0;
- p->TimePerChromaMetaChunkNominal[k] = 0;
- p->TimePerChromaMetaChunkVBlank[k] = 0;
- p->TimePerChromaMetaChunkFlip[k] = 0;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d, DST_Y_PER_META_ROW_NOM_L = %f\n", __func__, k, p->DST_Y_PER_META_ROW_NOM_L[k]);
- dml2_printf("DML::%s: k=%d, DST_Y_PER_META_ROW_NOM_C = %f\n", __func__, k, p->DST_Y_PER_META_ROW_NOM_C[k]);
- dml2_printf("DML::%s: k=%d, TimePerMetaChunkNominal = %f\n", __func__, k, p->TimePerMetaChunkNominal[k]);
- dml2_printf("DML::%s: k=%d, TimePerMetaChunkVBlank = %f\n", __func__, k, p->TimePerMetaChunkVBlank[k]);
- dml2_printf("DML::%s: k=%d, TimePerMetaChunkFlip = %f\n", __func__, k, p->TimePerMetaChunkFlip[k]);
- dml2_printf("DML::%s: k=%d, TimePerChromaMetaChunkNominal = %f\n", __func__, k, p->TimePerChromaMetaChunkNominal[k]);
- dml2_printf("DML::%s: k=%d, TimePerChromaMetaChunkVBlank = %f\n", __func__, k, p->TimePerChromaMetaChunkVBlank[k]);
- dml2_printf("DML::%s: k=%d, TimePerChromaMetaChunkFlip = %f\n", __func__, k, p->TimePerChromaMetaChunkFlip[k]);
-#endif
- }
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- p->DST_Y_PER_PTE_ROW_NOM_L[k] = p->dpte_row_height[k] / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- if (p->BytePerPixelC[k] == 0) {
- p->DST_Y_PER_PTE_ROW_NOM_C[k] = 0;
- } else {
- p->DST_Y_PER_PTE_ROW_NOM_C[k] = p->dpte_row_height_chroma[k] / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- }
- }
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- pixel_clock_mhz = ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
-
- if (p->display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut)
- p->time_per_tdlut_group[k] = 2 * p->dst_y_per_row_vblank[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / p->tdlut_groups_per_2row_ub[k];
- else
- p->time_per_tdlut_group[k] = 0;
-
- dml2_printf("DML::%s: k=%u, time_per_tdlut_group = %f\n", __func__, k, p->time_per_tdlut_group[k]);
-
- if (p->display_cfg->gpuvm_enable == true) {
- if (!dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle)) {
- dpte_group_width_luma = (unsigned int)((double)p->dpte_group_bytes[k] / (double)p->PTERequestSizeY[k] * p->PixelPTEReqWidthY[k]);
- } else {
- dpte_group_width_luma = (unsigned int)((double)p->dpte_group_bytes[k] / (double)p->PTERequestSizeY[k] * p->PixelPTEReqHeightY[k]);
- }
- if (p->use_one_row_for_frame[k]) {
- dpte_groups_per_row_luma_ub = (unsigned int)(math_ceil2((double)p->dpte_row_width_luma_ub[k] / (double)dpte_group_width_luma / 2.0, 1.0));
- } else {
- dpte_groups_per_row_luma_ub = (unsigned int)(math_ceil2((double)p->dpte_row_width_luma_ub[k] / (double)dpte_group_width_luma, 1.0));
- }
-
- if (dpte_groups_per_row_luma_ub <= 2) {
- dpte_groups_per_row_luma_ub = dpte_groups_per_row_luma_ub + 1;
- }
-
- dml2_printf("DML::%s: k=%u, use_one_row_for_frame = %u\n", __func__, k, p->use_one_row_for_frame[k]);
- dml2_printf("DML::%s: k=%u, dpte_group_bytes = %u\n", __func__, k, p->dpte_group_bytes[k]);
- dml2_printf("DML::%s: k=%u, PTERequestSizeY = %u\n", __func__, k, p->PTERequestSizeY[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEReqWidthY = %u\n", __func__, k, p->PixelPTEReqWidthY[k]);
- dml2_printf("DML::%s: k=%u, PixelPTEReqHeightY = %u\n", __func__, k, p->PixelPTEReqHeightY[k]);
- dml2_printf("DML::%s: k=%u, dpte_row_width_luma_ub = %u\n", __func__, k, p->dpte_row_width_luma_ub[k]);
- dml2_printf("DML::%s: k=%u, dpte_group_width_luma = %u\n", __func__, k, dpte_group_width_luma);
- dml2_printf("DML::%s: k=%u, dpte_groups_per_row_luma_ub = %u\n", __func__, k, dpte_groups_per_row_luma_ub);
-
- p->time_per_pte_group_nom_luma[k] = p->DST_Y_PER_PTE_ROW_NOM_L[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_luma_ub;
- p->time_per_pte_group_vblank_luma[k] = p->dst_y_per_row_vblank[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_luma_ub;
- p->time_per_pte_group_flip_luma[k] = p->dst_y_per_row_flip[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_luma_ub;
- if (p->BytePerPixelC[k] == 0) {
- p->time_per_pte_group_nom_chroma[k] = 0;
- p->time_per_pte_group_vblank_chroma[k] = 0;
- p->time_per_pte_group_flip_chroma[k] = 0;
- } else {
- if (!dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle)) {
- dpte_group_width_chroma = (unsigned int)((double)p->dpte_group_bytes[k] / (double)p->PTERequestSizeC[k] * p->PixelPTEReqWidthC[k]);
- } else {
- dpte_group_width_chroma = (unsigned int)((double)p->dpte_group_bytes[k] / (double)p->PTERequestSizeC[k] * p->PixelPTEReqHeightC[k]);
- }
-
- if (p->use_one_row_for_frame[k]) {
- dpte_groups_per_row_chroma_ub = (unsigned int)(math_ceil2((double)p->dpte_row_width_chroma_ub[k] / (double)dpte_group_width_chroma / 2.0, 1.0));
- } else {
- dpte_groups_per_row_chroma_ub = (unsigned int)(math_ceil2((double)p->dpte_row_width_chroma_ub[k] / (double)dpte_group_width_chroma, 1.0));
- }
- if (dpte_groups_per_row_chroma_ub <= 2) {
- dpte_groups_per_row_chroma_ub = dpte_groups_per_row_chroma_ub + 1;
- }
- dml2_printf("DML::%s: k=%u, dpte_row_width_chroma_ub = %u\n", __func__, k, p->dpte_row_width_chroma_ub[k]);
- dml2_printf("DML::%s: k=%u, dpte_group_width_chroma = %u\n", __func__, k, dpte_group_width_chroma);
- dml2_printf("DML::%s: k=%u, dpte_groups_per_row_chroma_ub = %u\n", __func__, k, dpte_groups_per_row_chroma_ub);
-
- p->time_per_pte_group_nom_chroma[k] = p->DST_Y_PER_PTE_ROW_NOM_C[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_chroma_ub;
- p->time_per_pte_group_vblank_chroma[k] = p->dst_y_per_row_vblank[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_chroma_ub;
- p->time_per_pte_group_flip_chroma[k] = p->dst_y_per_row_flip[k] * p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz / dpte_groups_per_row_chroma_ub;
- }
- } else {
- p->time_per_pte_group_nom_luma[k] = 0;
- p->time_per_pte_group_vblank_luma[k] = 0;
- p->time_per_pte_group_flip_luma[k] = 0;
- p->time_per_pte_group_nom_chroma[k] = 0;
- p->time_per_pte_group_vblank_chroma[k] = 0;
- p->time_per_pte_group_flip_chroma[k] = 0;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, dst_y_per_row_vblank = %f\n", __func__, k, p->dst_y_per_row_vblank[k]);
- dml2_printf("DML::%s: k=%u, dst_y_per_row_flip = %f\n", __func__, k, p->dst_y_per_row_flip[k]);
-
- dml2_printf("DML::%s: k=%u, DST_Y_PER_PTE_ROW_NOM_L = %f\n", __func__, k, p->DST_Y_PER_PTE_ROW_NOM_L[k]);
- dml2_printf("DML::%s: k=%u, DST_Y_PER_PTE_ROW_NOM_C = %f\n", __func__, k, p->DST_Y_PER_PTE_ROW_NOM_C[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_nom_luma = %f\n", __func__, k, p->time_per_pte_group_nom_luma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_vblank_luma = %f\n", __func__, k, p->time_per_pte_group_vblank_luma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_flip_luma = %f\n", __func__, k, p->time_per_pte_group_flip_luma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_nom_chroma = %f\n", __func__, k, p->time_per_pte_group_nom_chroma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_vblank_chroma = %f\n", __func__, k, p->time_per_pte_group_vblank_chroma[k]);
- dml2_printf("DML::%s: k=%u, time_per_pte_group_flip_chroma = %f\n", __func__, k, p->time_per_pte_group_flip_chroma[k]);
-#endif
- }
-} // CalculateMetaAndPTETimes
-
-static void CalculateVMGroupAndRequestTimes(
- const struct dml2_display_cfg *display_cfg,
- unsigned int NumberOfActiveSurfaces,
- unsigned int BytePerPixelC[],
- double dst_y_per_vm_vblank[],
- double dst_y_per_vm_flip[],
- unsigned int dpte_row_width_luma_ub[],
- unsigned int dpte_row_width_chroma_ub[],
- unsigned int vm_group_bytes[],
- unsigned int dpde0_bytes_per_frame_ub_l[],
- unsigned int dpde0_bytes_per_frame_ub_c[],
- unsigned int tdlut_pte_bytes_per_frame[],
- unsigned int meta_pte_bytes_per_frame_ub_l[],
- unsigned int meta_pte_bytes_per_frame_ub_c[],
- bool mrq_present,
-
- // Output
- double TimePerVMGroupVBlank[],
- double TimePerVMGroupFlip[],
- double TimePerVMRequestVBlank[],
- double TimePerVMRequestFlip[])
-{
- unsigned int num_group_per_lower_vm_stage = 1;
- unsigned int num_req_per_lower_vm_stage = 1;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: NumberOfActiveSurfaces = %u\n", __func__, NumberOfActiveSurfaces);
-#endif
- for (unsigned int k = 0; k < NumberOfActiveSurfaces; ++k) {
- double pixel_clock_mhz = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- bool dcc_mrq_enable = display_cfg->plane_descriptors[k].surface.dcc.enable && mrq_present;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, dcc_mrq_enable = %u\n", __func__, k, dcc_mrq_enable);
- dml2_printf("DML::%s: k=%u, vm_group_bytes = %u\n", __func__, k, vm_group_bytes[k]);
- dml2_printf("DML::%s: k=%u, dpde0_bytes_per_frame_ub_l = %u\n", __func__, k, dpde0_bytes_per_frame_ub_l[k]);
- dml2_printf("DML::%s: k=%u, dpde0_bytes_per_frame_ub_c = %u\n", __func__, k, dpde0_bytes_per_frame_ub_c[k]);
- dml2_printf("DML::%s: k=%d, meta_pte_bytes_per_frame_ub_l = %d\n", __func__, k, meta_pte_bytes_per_frame_ub_l[k]);
- dml2_printf("DML::%s: k=%d, meta_pte_bytes_per_frame_ub_c = %d\n", __func__, k, meta_pte_bytes_per_frame_ub_c[k]);
-#endif
-
- if (display_cfg->gpuvm_enable) {
- if (display_cfg->gpuvm_max_page_table_levels >= 2) {
- num_group_per_lower_vm_stage += (unsigned int)math_ceil2((double)(dpde0_bytes_per_frame_ub_l[k]) / (double)(vm_group_bytes[k]), 1);
-
- if (BytePerPixelC[k] > 0)
- num_group_per_lower_vm_stage += (unsigned int)math_ceil2((double)(dpde0_bytes_per_frame_ub_c[k]) / (double)(vm_group_bytes[k]), 1);
- }
-
- if (dcc_mrq_enable) {
- if (BytePerPixelC[k] > 0) {
- num_group_per_lower_vm_stage += (unsigned int)(2.0 /*for each mpde0 group*/ + math_ceil2((double)(meta_pte_bytes_per_frame_ub_l[k]) / (double)(vm_group_bytes[k]), 1) +
- math_ceil2((double)(meta_pte_bytes_per_frame_ub_c[k]) / (double)(vm_group_bytes[k]), 1));
- } else {
- num_group_per_lower_vm_stage += (unsigned int)(1.0 + math_ceil2((double)(meta_pte_bytes_per_frame_ub_l[k]) / (double)(vm_group_bytes[k]), 1));
- }
- }
-
- unsigned int num_group_per_lower_vm_stage_flip = num_group_per_lower_vm_stage;
- unsigned int num_group_per_lower_vm_stage_pref = num_group_per_lower_vm_stage;
-
- if (display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut && display_cfg->gpuvm_enable) {
- num_group_per_lower_vm_stage_pref += (unsigned int)math_ceil2(tdlut_pte_bytes_per_frame[k] / vm_group_bytes[k], 1);
- if (display_cfg->gpuvm_max_page_table_levels >= 2)
- num_group_per_lower_vm_stage_pref += 1; // tdpe0 group
- }
-
- if (display_cfg->gpuvm_max_page_table_levels >= 2) {
- num_req_per_lower_vm_stage += dpde0_bytes_per_frame_ub_l[k] / 64;
- if (BytePerPixelC[k] > 0)
- num_req_per_lower_vm_stage += dpde0_bytes_per_frame_ub_c[k];
- }
-
- if (dcc_mrq_enable) {
- num_req_per_lower_vm_stage += meta_pte_bytes_per_frame_ub_l[k] / 64;
- if (BytePerPixelC[k] > 0)
- num_req_per_lower_vm_stage += meta_pte_bytes_per_frame_ub_c[k] / 64;
- }
-
- unsigned int num_req_per_lower_vm_stage_flip = num_req_per_lower_vm_stage;
- unsigned int num_req_per_lower_vm_stage_pref = num_req_per_lower_vm_stage;
-
- if (display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut && display_cfg->gpuvm_enable) {
- num_req_per_lower_vm_stage_pref += tdlut_pte_bytes_per_frame[k] / 64;
- }
-
- double line_time = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / pixel_clock_mhz;
-
- if (num_group_per_lower_vm_stage_flip <= 2) {
- num_group_per_lower_vm_stage_flip = num_group_per_lower_vm_stage_flip + 1;
- }
-
- if (num_group_per_lower_vm_stage_pref <= 2) {
- num_group_per_lower_vm_stage_pref = num_group_per_lower_vm_stage_pref + 1;
- }
-
- TimePerVMGroupVBlank[k] = dst_y_per_vm_vblank[k] * line_time / num_group_per_lower_vm_stage_pref;
- TimePerVMGroupFlip[k] = dst_y_per_vm_flip[k] * line_time / num_group_per_lower_vm_stage_flip;
- TimePerVMRequestVBlank[k] = dst_y_per_vm_vblank[k] * line_time / num_req_per_lower_vm_stage_pref;
- TimePerVMRequestFlip[k] = dst_y_per_vm_flip[k] * line_time / num_req_per_lower_vm_stage_flip;
-
- dml2_printf("DML::%s: k=%u, dst_y_per_vm_vblank = %f\n", __func__, k, dst_y_per_vm_vblank[k]);
- dml2_printf("DML::%s: k=%u, dst_y_per_vm_flip = %f\n", __func__, k, dst_y_per_vm_flip[k]);
- dml2_printf("DML::%s: k=%u, line_time = %f\n", __func__, k, line_time);
- dml2_printf("DML::%s: k=%u, num_group_per_lower_vm_stage_pref = %f\n", __func__, k, num_group_per_lower_vm_stage_pref);
- dml2_printf("DML::%s: k=%u, num_group_per_lower_vm_stage_flip = %f\n", __func__, k, num_group_per_lower_vm_stage_flip);
- dml2_printf("DML::%s: k=%u, num_req_per_lower_vm_stage_pref = %f\n", __func__, k, num_req_per_lower_vm_stage_pref);
- dml2_printf("DML::%s: k=%u, num_req_per_lower_vm_stage_flip = %f\n", __func__, k, num_req_per_lower_vm_stage_flip);
-
- if (display_cfg->gpuvm_max_page_table_levels > 2) {
- TimePerVMGroupVBlank[k] = TimePerVMGroupVBlank[k] / 2;
- TimePerVMGroupFlip[k] = TimePerVMGroupFlip[k] / 2;
- TimePerVMRequestVBlank[k] = TimePerVMRequestVBlank[k] / 2;
- TimePerVMRequestFlip[k] = TimePerVMRequestFlip[k] / 2;
- }
-
- } else {
- TimePerVMGroupVBlank[k] = 0;
- TimePerVMGroupFlip[k] = 0;
- TimePerVMRequestVBlank[k] = 0;
- TimePerVMRequestFlip[k] = 0;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, TimePerVMGroupVBlank = %f\n", __func__, k, TimePerVMGroupVBlank[k]);
- dml2_printf("DML::%s: k=%u, TimePerVMGroupFlip = %f\n", __func__, k, TimePerVMGroupFlip[k]);
- dml2_printf("DML::%s: k=%u, TimePerVMRequestVBlank = %f\n", __func__, k, TimePerVMRequestVBlank[k]);
- dml2_printf("DML::%s: k=%u, TimePerVMRequestFlip = %f\n", __func__, k, TimePerVMRequestFlip[k]);
-#endif
- }
-}
-
-static void CalculateStutterEfficiency(struct dml2_core_internal_scratch *scratch,
- struct dml2_core_calcs_CalculateStutterEfficiency_params *p)
-{
- struct dml2_core_calcs_CalculateStutterEfficiency_locals *l = &scratch->CalculateStutterEfficiency_locals;
-
- memset(l, 0, sizeof(struct dml2_core_calcs_CalculateStutterEfficiency_locals));
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (!dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) {
- if (p->display_cfg->plane_descriptors[k].surface.dcc.enable == true) {
- if ((dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle) && p->BlockWidth256BytesY[k] > p->SwathHeightY[k]) || (!dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle) && p->BlockHeight256BytesY[k] > p->SwathHeightY[k]) || p->DCCYMaxUncompressedBlock[k] < 256) {
- l->MaximumEffectiveCompressionLuma = 2;
- } else {
- l->MaximumEffectiveCompressionLuma = 4;
- }
- l->TotalCompressedReadBandwidth = l->TotalCompressedReadBandwidth + p->ReadBandwidthSurfaceLuma[k] / math_min2(p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane0, l->MaximumEffectiveCompressionLuma);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceLuma = %f\n", __func__, k, p->ReadBandwidthSurfaceLuma[k]);
- dml2_printf("DML::%s: k=%u, NetDCCRateLuma = %f\n", __func__, k, p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane0);
- dml2_printf("DML::%s: k=%u, MaximumEffectiveCompressionLuma = %f\n", __func__, k, l->MaximumEffectiveCompressionLuma);
-#endif
- l->TotalZeroSizeRequestReadBandwidth = l->TotalZeroSizeRequestReadBandwidth + p->ReadBandwidthSurfaceLuma[k] * p->display_cfg->plane_descriptors[k].surface.dcc.informative.fraction_of_zero_size_request_plane0;
- l->TotalZeroSizeCompressedReadBandwidth = l->TotalZeroSizeCompressedReadBandwidth + p->ReadBandwidthSurfaceLuma[k] * p->display_cfg->plane_descriptors[k].surface.dcc.informative.fraction_of_zero_size_request_plane0 / l->MaximumEffectiveCompressionLuma;
-
- if (p->ReadBandwidthSurfaceChroma[k] > 0) {
- if ((dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle) && p->BlockWidth256BytesC[k] > p->SwathHeightC[k]) || (!dml_is_vertical_rotation(p->display_cfg->plane_descriptors[k].composition.rotation_angle) && p->BlockHeight256BytesC[k] > p->SwathHeightC[k]) || p->DCCCMaxUncompressedBlock[k] < 256) {
- l->MaximumEffectiveCompressionChroma = 2;
- } else {
- l->MaximumEffectiveCompressionChroma = 4;
- }
- l->TotalCompressedReadBandwidth = l->TotalCompressedReadBandwidth + p->ReadBandwidthSurfaceChroma[k] / math_min2(p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane1, l->MaximumEffectiveCompressionChroma);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceChroma = %f\n", __func__, k, p->ReadBandwidthSurfaceChroma[k]);
- dml2_printf("DML::%s: k=%u, NetDCCRateChroma = %f\n", __func__, k, p->display_cfg->plane_descriptors[k].surface.dcc.informative.dcc_rate_plane1);
- dml2_printf("DML::%s: k=%u, MaximumEffectiveCompressionChroma = %f\n", __func__, k, l->MaximumEffectiveCompressionChroma);
-#endif
- l->TotalZeroSizeRequestReadBandwidth = l->TotalZeroSizeRequestReadBandwidth + p->ReadBandwidthSurfaceChroma[k] * p->display_cfg->plane_descriptors[k].surface.dcc.informative.fraction_of_zero_size_request_plane1;
- l->TotalZeroSizeCompressedReadBandwidth = l->TotalZeroSizeCompressedReadBandwidth + p->ReadBandwidthSurfaceChroma[k] * p->display_cfg->plane_descriptors[k].surface.dcc.informative.fraction_of_zero_size_request_plane1 / l->MaximumEffectiveCompressionChroma;
- }
- } else {
- l->TotalCompressedReadBandwidth = l->TotalCompressedReadBandwidth + p->ReadBandwidthSurfaceLuma[k] + p->ReadBandwidthSurfaceChroma[k];
- }
- l->TotalRowReadBandwidth = l->TotalRowReadBandwidth + p->DPPPerSurface[k] * (p->meta_row_bw[k] + p->dpte_row_bw[k]);
- }
- }
-
- l->AverageDCCCompressionRate = p->TotalDataReadBandwidth / l->TotalCompressedReadBandwidth;
- l->AverageDCCZeroSizeFraction = l->TotalZeroSizeRequestReadBandwidth / p->TotalDataReadBandwidth;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: UnboundedRequestEnabled = %u\n", __func__, p->UnboundedRequestEnabled);
- dml2_printf("DML::%s: TotalCompressedReadBandwidth = %f\n", __func__, l->TotalCompressedReadBandwidth);
- dml2_printf("DML::%s: TotalZeroSizeRequestReadBandwidth = %f\n", __func__, l->TotalZeroSizeRequestReadBandwidth);
- dml2_printf("DML::%s: TotalZeroSizeCompressedReadBandwidth = %f\n", __func__, l->TotalZeroSizeCompressedReadBandwidth);
- dml2_printf("DML::%s: MaximumEffectiveCompressionLuma = %f\n", __func__, l->MaximumEffectiveCompressionLuma);
- dml2_printf("DML::%s: MaximumEffectiveCompressionChroma = %f\n", __func__, l->MaximumEffectiveCompressionChroma);
- dml2_printf("DML::%s: AverageDCCCompressionRate = %f\n", __func__, l->AverageDCCCompressionRate);
- dml2_printf("DML::%s: AverageDCCZeroSizeFraction = %f\n", __func__, l->AverageDCCZeroSizeFraction);
-
- dml2_printf("DML::%s: CompbufReservedSpace64B = %u (%f kbytes)\n", __func__, p->CompbufReservedSpace64B, p->CompbufReservedSpace64B * 64 / 1024.0);
- dml2_printf("DML::%s: CompbufReservedSpaceZs = %u\n", __func__, p->CompbufReservedSpaceZs);
- dml2_printf("DML::%s: CompressedBufferSizeInkByte = %u kbytes\n", __func__, p->CompressedBufferSizeInkByte);
- dml2_printf("DML::%s: ROBBufferSizeInKByte = %u kbytes\n", __func__, p->ROBBufferSizeInKByte);
-#endif
- if (l->AverageDCCZeroSizeFraction == 1) {
- l->AverageZeroSizeCompressionRate = l->TotalZeroSizeRequestReadBandwidth / l->TotalZeroSizeCompressedReadBandwidth;
- l->EffectiveCompressedBufferSize = (double)p->MetaFIFOSizeInKEntries * 1024 * 64 * l->AverageZeroSizeCompressionRate + ((double)p->ZeroSizeBufferEntries - p->CompbufReservedSpaceZs) * 64 * l->AverageZeroSizeCompressionRate;
-
-
- } else if (l->AverageDCCZeroSizeFraction > 0) {
- l->AverageZeroSizeCompressionRate = l->TotalZeroSizeRequestReadBandwidth / l->TotalZeroSizeCompressedReadBandwidth;
- l->EffectiveCompressedBufferSize = math_min2((double)p->CompressedBufferSizeInkByte * 1024 * l->AverageDCCCompressionRate,
- (double)p->MetaFIFOSizeInKEntries * 1024 * 64 / (l->AverageDCCZeroSizeFraction / l->AverageZeroSizeCompressionRate + 1 / l->AverageDCCCompressionRate)) +
- (p->rob_alloc_compressed ? math_min2(((double)p->ROBBufferSizeInKByte * 1024 - p->CompbufReservedSpace64B * 64) * l->AverageDCCCompressionRate,
- ((double)p->ZeroSizeBufferEntries - p->CompbufReservedSpaceZs) * 64 / (l->AverageDCCZeroSizeFraction / l->AverageZeroSizeCompressionRate))
- : ((double)p->ROBBufferSizeInKByte * 1024 - p->CompbufReservedSpace64B * 64));
-
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: min 1 = %f\n", __func__, p->CompressedBufferSizeInkByte * 1024 * l->AverageDCCCompressionRate);
- dml2_printf("DML::%s: min 2 = %f\n", __func__, p->MetaFIFOSizeInKEntries * 1024 * 64 / (l->AverageDCCZeroSizeFraction / l->AverageZeroSizeCompressionRate + 1 / l->AverageDCCCompressionRate));
- dml2_printf("DML::%s: min 3 = %d\n", __func__, (p->ROBBufferSizeInKByte * 1024 - p->CompbufReservedSpace64B * 64));
- dml2_printf("DML::%s: min 4 = %f\n", __func__, (p->ZeroSizeBufferEntries - p->CompbufReservedSpaceZs) * 64 / (l->AverageDCCZeroSizeFraction / l->AverageZeroSizeCompressionRate));
-#endif
- } else {
- l->EffectiveCompressedBufferSize = math_min2((double)p->CompressedBufferSizeInkByte * 1024 * l->AverageDCCCompressionRate,
- (double)p->MetaFIFOSizeInKEntries * 1024 * 64 * l->AverageDCCCompressionRate) +
- ((double)p->ROBBufferSizeInKByte * 1024 - p->CompbufReservedSpace64B * 64) * (p->rob_alloc_compressed ? l->AverageDCCCompressionRate : 1.0);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: min 1 = %f\n", __func__, p->CompressedBufferSizeInkByte * 1024 * l->AverageDCCCompressionRate);
- dml2_printf("DML::%s: min 2 = %f\n", __func__, p->MetaFIFOSizeInKEntries * 1024 * 64 * l->AverageDCCCompressionRate);
-#endif
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: MetaFIFOSizeInKEntries = %u\n", __func__, p->MetaFIFOSizeInKEntries);
- dml2_printf("DML::%s: AverageZeroSizeCompressionRate = %f\n", __func__, l->AverageZeroSizeCompressionRate);
- dml2_printf("DML::%s: EffectiveCompressedBufferSize = %f (%f kbytes)\n", __func__, l->EffectiveCompressedBufferSize, l->EffectiveCompressedBufferSize / 1024.0);
-#endif
-
- bool FoundCriticalSurface = false;
- *p->StutterPeriod = 0;
-
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (!dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) {
- l->LinesInDETY = ((double)p->DETBufferSizeY[k] + (p->UnboundedRequestEnabled == true ? l->EffectiveCompressedBufferSize : 0) * p->ReadBandwidthSurfaceLuma[k] / p->TotalDataReadBandwidth) / p->BytePerPixelDETY[k] / p->SwathWidthY[k];
- l->LinesInDETYRoundedDownToSwath = math_floor2(l->LinesInDETY, p->SwathHeightY[k]);
- l->DETBufferingTimeY = l->LinesInDETYRoundedDownToSwath * ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) / p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, DETBufferSizeY = %u (%u kbytes)\n", __func__, k, p->DETBufferSizeY[k], p->DETBufferSizeY[k] / 1024);
- dml2_printf("DML::%s: k=%u, BytePerPixelDETY = %f\n", __func__, k, p->BytePerPixelDETY[k]);
- dml2_printf("DML::%s: k=%u, SwathWidthY = %u\n", __func__, k, p->SwathWidthY[k]);
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceLuma = %f\n", __func__, k, p->ReadBandwidthSurfaceLuma[k]);
- dml2_printf("DML::%s: k=%u, TotalDataReadBandwidth = %f\n", __func__, k, p->TotalDataReadBandwidth);
- dml2_printf("DML::%s: k=%u, LinesInDETY = %f\n", __func__, k, l->LinesInDETY);
- dml2_printf("DML::%s: k=%u, LinesInDETYRoundedDownToSwath = %f\n", __func__, k, l->LinesInDETYRoundedDownToSwath);
- dml2_printf("DML::%s: k=%u, VRatio = %f\n", __func__, k, p->display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
- dml2_printf("DML::%s: k=%u, DETBufferingTimeY = %f\n", __func__, k, l->DETBufferingTimeY);
-#endif
-
- if (!FoundCriticalSurface || l->DETBufferingTimeY < *p->StutterPeriod) {
- bool isInterlaceTiming = p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.interlaced && !p->ProgressiveToInterlaceUnitInOPP;
-
- FoundCriticalSurface = true;
- *p->StutterPeriod = l->DETBufferingTimeY;
- l->FrameTimeCriticalSurface = (isInterlaceTiming ? math_floor2((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.v_total / 2.0, 1.0) : p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.v_total) * (double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- l->VActiveTimeCriticalSurface = (isInterlaceTiming ? math_floor2((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.v_active / 2.0, 1.0) : p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.v_active) * (double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- l->BytePerPixelYCriticalSurface = p->BytePerPixelY[k];
- l->SwathWidthYCriticalSurface = p->SwathWidthY[k];
- l->SwathHeightYCriticalSurface = p->SwathHeightY[k];
- l->BlockWidth256BytesYCriticalSurface = p->BlockWidth256BytesY[k];
- l->DETBufferSizeYCriticalSurface = p->DETBufferSizeY[k];
- l->MinTTUVBlankCriticalSurface = p->MinTTUVBlank[k];
- l->SinglePlaneCriticalSurface = (p->ReadBandwidthSurfaceChroma[k] == 0);
- l->SinglePipeCriticalSurface = (p->DPPPerSurface[k] == 1);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, FoundCriticalSurface = %u\n", __func__, k, FoundCriticalSurface);
- dml2_printf("DML::%s: k=%u, StutterPeriod = %f\n", __func__, k, *p->StutterPeriod);
- dml2_printf("DML::%s: k=%u, MinTTUVBlankCriticalSurface = %f\n", __func__, k, l->MinTTUVBlankCriticalSurface);
- dml2_printf("DML::%s: k=%u, FrameTimeCriticalSurface= %f\n", __func__, k, l->FrameTimeCriticalSurface);
- dml2_printf("DML::%s: k=%u, VActiveTimeCriticalSurface = %f\n", __func__, k, l->VActiveTimeCriticalSurface);
- dml2_printf("DML::%s: k=%u, BytePerPixelYCriticalSurface = %u\n", __func__, k, l->BytePerPixelYCriticalSurface);
- dml2_printf("DML::%s: k=%u, SwathWidthYCriticalSurface = %f\n", __func__, k, l->SwathWidthYCriticalSurface);
- dml2_printf("DML::%s: k=%u, SwathHeightYCriticalSurface = %f\n", __func__, k, l->SwathHeightYCriticalSurface);
- dml2_printf("DML::%s: k=%u, BlockWidth256BytesYCriticalSurface = %u\n", __func__, k, l->BlockWidth256BytesYCriticalSurface);
- dml2_printf("DML::%s: k=%u, SinglePlaneCriticalSurface = %u\n", __func__, k, l->SinglePlaneCriticalSurface);
- dml2_printf("DML::%s: k=%u, SinglePipeCriticalSurface = %u\n", __func__, k, l->SinglePipeCriticalSurface);
-#endif
- }
- }
- }
-
- // for bounded req, the stutter period is calculated only based on DET size, but during burst there can be some return inside ROB/compressed buffer
- // stutter period is calculated only on the det sizing
- // if (cdb + rob >= det) the stutter burst will be absorbed by the cdb + rob which is before decompress
- // else
- // the cdb + rob part will be in compressed rate with urg bw (idea bw)
- // the det part will be return at uncompressed rate with 64B/dcfclk
- //
- // for unbounded req, the stutter period should be calculated as total of CDB+ROB+DET, so the term "PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer"
- // should be == EffectiveCompressedBufferSize which will returned a compressed rate, the rest of stutter period is from the DET will be returned at uncompressed rate with 64B/dcfclk
-
- l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer = math_min2(*p->StutterPeriod * p->TotalDataReadBandwidth, l->EffectiveCompressedBufferSize);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: AverageDCCCompressionRate = %f\n", __func__, l->AverageDCCCompressionRate);
- dml2_printf("DML::%s: StutterPeriod*TotalDataReadBandwidth = %f (%f kbytes)\n", __func__, *p->StutterPeriod * p->TotalDataReadBandwidth, (*p->StutterPeriod * p->TotalDataReadBandwidth) / 1024.0);
- dml2_printf("DML::%s: EffectiveCompressedBufferSize = %f (%f kbytes)\n", __func__, l->EffectiveCompressedBufferSize, l->EffectiveCompressedBufferSize / 1024.0);
- dml2_printf("DML::%s: PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer = %f (%f kbytes)\n", __func__, l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer, l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer / 1024);
- dml2_printf("DML::%s: ReturnBW = %f\n", __func__, p->ReturnBW);
- dml2_printf("DML::%s: TotalDataReadBandwidth = %f\n", __func__, p->TotalDataReadBandwidth);
- dml2_printf("DML::%s: TotalRowReadBandwidth = %f\n", __func__, l->TotalRowReadBandwidth);
- dml2_printf("DML::%s: DCFCLK = %f\n", __func__, p->DCFCLK);
-#endif
-
- l->StutterBurstTime = l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer
- / (p->ReturnBW * (p->hw_debug5 ? 1 : l->AverageDCCCompressionRate)) +
- (*p->StutterPeriod * p->TotalDataReadBandwidth - l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer)
- / math_max2(p->DCFCLK * 64, p->ReturnBW * (p->hw_debug5 ? 1 : l->AverageDCCCompressionRate)) +
- *p->StutterPeriod * l->TotalRowReadBandwidth / p->ReturnBW;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Part 1 = %f\n", __func__, l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer / p->ReturnBW / (p->hw_debug5 ? 1 : l->AverageDCCCompressionRate));
- dml2_printf("DML::%s: Part 2 = %f\n", __func__, (*p->StutterPeriod * p->TotalDataReadBandwidth - l->PartOfUncompressedPixelBurstThatFitsInROBAndCompressedBuffer) / (p->DCFCLK * 64));
- dml2_printf("DML::%s: Part 3 = %f\n", __func__, *p->StutterPeriod * l->TotalRowReadBandwidth / p->ReturnBW);
- dml2_printf("DML::%s: StutterBurstTime = %f\n", __func__, l->StutterBurstTime);
-#endif
-
- l->TotalActiveWriteback = 0;
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].writeback.enable) {
- l->TotalActiveWriteback = l->TotalActiveWriteback + 1;
- }
- }
-
- if (l->TotalActiveWriteback == 0) {
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: SRExitTime = %f\n", __func__, p->SRExitTime);
- dml2_printf("DML::%s: SRExitZ8Time = %f\n", __func__, p->SRExitZ8Time);
- dml2_printf("DML::%s: StutterPeriod = %f\n", __func__, *p->StutterPeriod);
-#endif
- *p->StutterEfficiencyNotIncludingVBlank = math_max2(0., 1 - (p->SRExitTime + l->StutterBurstTime) / *p->StutterPeriod) * 100;
- *p->Z8StutterEfficiencyNotIncludingVBlank = math_max2(0., 1 - (p->SRExitZ8Time + l->StutterBurstTime) / *p->StutterPeriod) * 100;
- *p->NumberOfStutterBurstsPerFrame = (*p->StutterEfficiencyNotIncludingVBlank > 0 ? (unsigned int)(math_ceil2(l->VActiveTimeCriticalSurface / *p->StutterPeriod, 1)) : 0);
- *p->Z8NumberOfStutterBurstsPerFrame = (*p->Z8StutterEfficiencyNotIncludingVBlank > 0 ? (unsigned int)(math_ceil2(l->VActiveTimeCriticalSurface / *p->StutterPeriod, 1)) : 0);
- } else {
- *p->StutterEfficiencyNotIncludingVBlank = 0.;
- *p->Z8StutterEfficiencyNotIncludingVBlank = 0.;
- *p->NumberOfStutterBurstsPerFrame = 0;
- *p->Z8NumberOfStutterBurstsPerFrame = 0;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: VActiveTimeCriticalSurface = %f\n", __func__, l->VActiveTimeCriticalSurface);
- dml2_printf("DML::%s: StutterEfficiencyNotIncludingVBlank = %f\n", __func__, *p->StutterEfficiencyNotIncludingVBlank);
- dml2_printf("DML::%s: Z8StutterEfficiencyNotIncludingVBlank = %f\n", __func__, *p->Z8StutterEfficiencyNotIncludingVBlank);
- dml2_printf("DML::%s: NumberOfStutterBurstsPerFrame = %u\n", __func__, *p->NumberOfStutterBurstsPerFrame);
- dml2_printf("DML::%s: Z8NumberOfStutterBurstsPerFrame = %u\n", __func__, *p->Z8NumberOfStutterBurstsPerFrame);
-#endif
-
- unsigned int TotalNumberOfActiveOTG = 0;
- double SinglePixelClock = 0;
- unsigned int SingleHTotal = 0;
- unsigned int SingleVTotal = 0;
- bool SameTiming = true;
- for (unsigned int k = 0; k < p->NumberOfActiveSurfaces; ++k) {
- if (!dml_is_phantom_pipe(&p->display_cfg->plane_descriptors[k])) {
- if (p->display_cfg->plane_descriptors[k].stream_index == k) {
- if (TotalNumberOfActiveOTG == 0) {
- SinglePixelClock = ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- SingleHTotal = p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total;
- SingleVTotal = p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.v_total;
- } else if (SinglePixelClock != ((double)p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) || SingleHTotal != p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.h_total || SingleVTotal != p->display_cfg->stream_descriptors[p->display_cfg->plane_descriptors[k].stream_index].timing.v_total) {
- SameTiming = false;
- }
- TotalNumberOfActiveOTG = TotalNumberOfActiveOTG + 1;
- }
- }
- }
-
- if (*p->StutterEfficiencyNotIncludingVBlank > 0) {
- if (!((p->SynchronizeTimings || TotalNumberOfActiveOTG == 1) && SameTiming)) {
- *p->StutterEfficiency = *p->StutterEfficiencyNotIncludingVBlank;
- } else {
- *p->StutterEfficiency = (1 - (*p->NumberOfStutterBurstsPerFrame * p->SRExitTime + l->StutterBurstTime * l->VActiveTimeCriticalSurface / *p->StutterPeriod) / l->FrameTimeCriticalSurface) * 100;
- }
- } else {
- *p->StutterEfficiency = 0;
- *p->NumberOfStutterBurstsPerFrame = 0;
- }
-
- double LastZ8StutterPeriod = 0.0;
-
- if (*p->Z8StutterEfficiencyNotIncludingVBlank > 0) {
- LastZ8StutterPeriod = l->VActiveTimeCriticalSurface - (*p->Z8NumberOfStutterBurstsPerFrame - 1) * *p->StutterPeriod;
- if (!((p->SynchronizeTimings || TotalNumberOfActiveOTG == 1) && SameTiming)) {
- *p->Z8StutterEfficiency = *p->Z8StutterEfficiencyNotIncludingVBlank;
- } else {
- *p->Z8StutterEfficiency = (1 - (*p->Z8NumberOfStutterBurstsPerFrame * p->SRExitZ8Time + l->StutterBurstTime * l->VActiveTimeCriticalSurface / *p->StutterPeriod) / l->FrameTimeCriticalSurface) * 100;
- }
- } else {
- *p->Z8StutterEfficiency = 0.;
- *p->Z8NumberOfStutterBurstsPerFrame = 0;
- }
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: LastZ8StutterPeriod = %f\n", __func__, LastZ8StutterPeriod);
- dml2_printf("DML::%s: Z8StutterEnterPlusExitWatermark = %f\n", __func__, p->Z8StutterEnterPlusExitWatermark);
- dml2_printf("DML::%s: StutterBurstTime = %f\n", __func__, l->StutterBurstTime);
- dml2_printf("DML::%s: StutterPeriod = %f\n", __func__, *p->StutterPeriod);
- dml2_printf("DML::%s: StutterEfficiency = %f\n", __func__, *p->StutterEfficiency);
- dml2_printf("DML::%s: Z8StutterEfficiency = %f\n", __func__, *p->Z8StutterEfficiency);
- dml2_printf("DML::%s: StutterEfficiencyNotIncludingVBlank = %f\n", __func__, *p->StutterEfficiencyNotIncludingVBlank);
- dml2_printf("DML::%s: Z8NumberOfStutterBurstsPerFrame = %u\n", __func__, *p->Z8NumberOfStutterBurstsPerFrame);
-#endif
-
-
- unsigned int SwathSizeCriticalSurface;
- unsigned int LastChunkOfSwathSize;
- unsigned int MissingPartOfLastSwathOfDETSize;
-
- SwathSizeCriticalSurface = (unsigned int)(l->BytePerPixelYCriticalSurface * l->SwathHeightYCriticalSurface * math_ceil2(l->SwathWidthYCriticalSurface, l->BlockWidth256BytesYCriticalSurface));
- LastChunkOfSwathSize = SwathSizeCriticalSurface % (p->PixelChunkSizeInKByte * 1024);
- MissingPartOfLastSwathOfDETSize = (unsigned int)(math_ceil2(l->DETBufferSizeYCriticalSurface, SwathSizeCriticalSurface) - l->DETBufferSizeYCriticalSurface);
-
- *p->DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE = !(!p->UnboundedRequestEnabled && (p->NumberOfActiveSurfaces == 1) && l->SinglePlaneCriticalSurface && l->SinglePipeCriticalSurface && (LastChunkOfSwathSize > 0) &&
- (LastChunkOfSwathSize <= 4096) && (MissingPartOfLastSwathOfDETSize > 0) && (MissingPartOfLastSwathOfDETSize <= LastChunkOfSwathSize));
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: SwathSizeCriticalSurface = %u\n", __func__, SwathSizeCriticalSurface);
- dml2_printf("DML::%s: DETBufferSizeYCriticalSurface = %u\n", __func__, l->DETBufferSizeYCriticalSurface);
- dml2_printf("DML::%s: PixelChunkSizeInKByte = %u\n", __func__, p->PixelChunkSizeInKByte);
- dml2_printf("DML::%s: LastChunkOfSwathSize = %u\n", __func__, LastChunkOfSwathSize);
- dml2_printf("DML::%s: MissingPartOfLastSwathOfDETSize = %u\n", __func__, MissingPartOfLastSwathOfDETSize);
- dml2_printf("DML::%s: DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE = %u\n", __func__, *p->DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE);
-#endif
-}
-
-bool dml2_core_shared_mode_programming(struct dml2_core_calcs_mode_programming_ex *in_out_params)
-{
- const struct dml2_display_cfg *display_cfg = in_out_params->in_display_cfg;
- const struct dml2_mcg_min_clock_table *min_clk_table = in_out_params->min_clk_table;
- const struct core_display_cfg_support_info *cfg_support_info = in_out_params->cfg_support_info;
- struct dml2_core_internal_display_mode_lib *mode_lib = in_out_params->mode_lib;
- struct dml2_display_cfg_programming *programming = in_out_params->programming;
-
- struct dml2_core_calcs_mode_programming_locals *s = &mode_lib->scratch.dml_core_mode_programming_locals;
- struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params *CalculateWatermarks_params = &mode_lib->scratch.CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_params;
- struct dml2_core_calcs_CalculateVMRowAndSwath_params *CalculateVMRowAndSwath_params = &mode_lib->scratch.CalculateVMRowAndSwath_params;
- struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params *CalculateSwathAndDETConfiguration_params = &mode_lib->scratch.CalculateSwathAndDETConfiguration_params;
- struct dml2_core_calcs_CalculateStutterEfficiency_params *CalculateStutterEfficiency_params = &mode_lib->scratch.CalculateStutterEfficiency_params;
- struct dml2_core_calcs_CalculatePrefetchSchedule_params *CalculatePrefetchSchedule_params = &mode_lib->scratch.CalculatePrefetchSchedule_params;
- struct dml2_core_calcs_calculate_mcache_setting_params *calculate_mcache_setting_params = &mode_lib->scratch.calculate_mcache_setting_params;
- struct dml2_core_calcs_calculate_tdlut_setting_params *calculate_tdlut_setting_params = &mode_lib->scratch.calculate_tdlut_setting_params;
- struct dml2_core_shared_CalculateMetaAndPTETimes_params *CalculateMetaAndPTETimes_params = &mode_lib->scratch.CalculateMetaAndPTETimes_params;
-
- unsigned int j, k;
-
- dml2_printf("DML::%s: --- START --- \n", __func__);
-
- memset(&mode_lib->mp, 0, sizeof(struct dml2_core_internal_mode_program));
-
- s->num_active_planes = display_cfg->num_planes;
- get_stream_output_bpp(s->OutputBpp, display_cfg);
-
- mode_lib->mp.num_active_pipes = dml_get_num_active_pipes(display_cfg->num_planes, cfg_support_info);
- dml_calc_pipe_plane_mapping(cfg_support_info, mode_lib->mp.pipe_plane);
-
- mode_lib->mp.Dcfclk = programming->min_clocks.dcn4x.active.dcfclk_khz / 1000.0;
- mode_lib->mp.FabricClock = programming->min_clocks.dcn4x.active.fclk_khz / 1000.0;
- mode_lib->mp.dram_bw_mbps = uclk_khz_to_dram_bw_mbps(programming->min_clocks.dcn4x.active.uclk_khz, &mode_lib->soc.clk_table.dram_config);
- mode_lib->mp.uclk_freq_mhz = programming->min_clocks.dcn4x.active.uclk_khz / 1000.0;
- mode_lib->mp.GlobalDPPCLK = programming->min_clocks.dcn4x.dpprefclk_khz / 1000.0;
- s->SOCCLK = (double)programming->min_clocks.dcn4x.socclk_khz / 1000;
- mode_lib->mp.qos_param_index = get_qos_param_index(programming->min_clocks.dcn4x.active.uclk_khz, mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params);
- mode_lib->mp.active_min_uclk_dpm_index = get_active_min_uclk_dpm_index(programming->min_clocks.dcn4x.active.uclk_khz, &mode_lib->soc.clk_table);
-
- for (k = 0; k < s->num_active_planes; ++k) {
- unsigned int stream_index = display_cfg->plane_descriptors[k].stream_index;
- dml2_assert(cfg_support_info->stream_support_info[stream_index].odms_used <= 4);
- dml2_assert(cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 4 ||
- cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 2 ||
- cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 1);
-
- if (cfg_support_info->stream_support_info[stream_index].odms_used > 1)
- dml2_assert(cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 1);
-
- switch (cfg_support_info->stream_support_info[stream_index].odms_used) {
- case (4):
- mode_lib->mp.ODMMode[k] = dml2_odm_mode_combine_4to1;
- break;
- case (3):
- mode_lib->mp.ODMMode[k] = dml2_odm_mode_combine_3to1;
- break;
- case (2):
- mode_lib->mp.ODMMode[k] = dml2_odm_mode_combine_2to1;
- break;
- default:
- if (cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 4)
- mode_lib->mp.ODMMode[k] = dml2_odm_mode_mso_1to4;
- else if (cfg_support_info->stream_support_info[stream_index].num_odm_output_segments == 2)
- mode_lib->mp.ODMMode[k] = dml2_odm_mode_mso_1to2;
- else
- mode_lib->mp.ODMMode[k] = dml2_odm_mode_bypass;
- break;
- }
- }
-
- for (k = 0; k < s->num_active_planes; ++k) {
- mode_lib->mp.NoOfDPP[k] = cfg_support_info->plane_support_info[k].dpps_used;
- mode_lib->mp.Dppclk[k] = programming->plane_programming[k].min_clocks.dcn4x.dppclk_khz / 1000.0;
- dml2_assert(mode_lib->mp.Dppclk[k] > 0);
- }
-
- for (k = 0; k < s->num_active_planes; ++k) {
- unsigned int stream_index = display_cfg->plane_descriptors[k].stream_index;
- mode_lib->mp.DSCCLK[k] = programming->stream_programming[stream_index].min_clocks.dcn4x.dscclk_khz / 1000.0;
- dml2_printf("DML::%s: k=%d stream_index=%d, mode_lib->mp.DSCCLK = %f\n", __func__, k, stream_index, mode_lib->mp.DSCCLK[k]);
- }
-
- mode_lib->mp.Dispclk = programming->min_clocks.dcn4x.dispclk_khz / 1000.0;
- mode_lib->mp.DCFCLKDeepSleep = programming->min_clocks.dcn4x.deepsleep_dcfclk_khz / 1000.0;
-
- dml2_assert(mode_lib->mp.Dcfclk > 0);
- dml2_assert(mode_lib->mp.FabricClock > 0);
- dml2_assert(mode_lib->mp.dram_bw_mbps > 0);
- dml2_assert(mode_lib->mp.uclk_freq_mhz > 0);
- dml2_assert(mode_lib->mp.GlobalDPPCLK > 0);
- dml2_assert(mode_lib->mp.Dispclk > 0);
- dml2_assert(mode_lib->mp.DCFCLKDeepSleep > 0);
- dml2_assert(s->SOCCLK > 0);
-
-#ifdef __DML_VBA_DEBUG__
- // dml2_printf_dml_display_cfg_timing(&display_cfg->timing, s->num_active_planes);
- // dml2_printf_dml_display_cfg_plane(&display_cfg->plane, s->num_active_planes);
- // dml2_printf_dml_display_cfg_surface(&display_cfg->surface, s->num_active_planes);
- // dml2_printf_dml_display_cfg_output(&display_cfg->output, s->num_active_planes);
- // dml2_printf_dml_display_cfg_hw_resource(&display_cfg->hw, s->num_active_planes);
-
- dml2_printf("DML::%s: num_active_planes = %u\n", __func__, s->num_active_planes);
- dml2_printf("DML::%s: num_active_pipes = %u\n", __func__, mode_lib->mp.num_active_pipes);
- dml2_printf("DML::%s: Dcfclk = %f\n", __func__, mode_lib->mp.Dcfclk);
- dml2_printf("DML::%s: FabricClock = %f\n", __func__, mode_lib->mp.FabricClock);
- dml2_printf("DML::%s: dram_bw_mbps = %f\n", __func__, mode_lib->mp.dram_bw_mbps);
- dml2_printf("DML::%s: uclk_freq_mhz = %f\n", __func__, mode_lib->mp.uclk_freq_mhz);
- dml2_printf("DML::%s: Dispclk = %f\n", __func__, mode_lib->mp.Dispclk);
- for (k = 0; k < s->num_active_planes; ++k) {
- dml2_printf("DML::%s: Dppclk[%0d] = %f\n", __func__, k, mode_lib->mp.Dppclk[k]);
- }
- dml2_printf("DML::%s: GlobalDPPCLK = %f\n", __func__, mode_lib->mp.GlobalDPPCLK);
- dml2_printf("DML::%s: DCFCLKDeepSleep = %f\n", __func__, mode_lib->mp.DCFCLKDeepSleep);
- dml2_printf("DML::%s: SOCCLK = %f\n", __func__, s->SOCCLK);
- dml2_printf("DML::%s: min_clk_index = %0d\n", __func__, in_out_params->min_clk_index);
- dml2_printf("DML::%s: min_clk_table min_fclk_khz = %d\n", __func__, min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].min_fclk_khz);
- dml2_printf("DML::%s: min_clk_table uclk_mhz = %f\n", __func__, dram_bw_kbps_to_uclk_mhz(min_clk_table->dram_bw_table.entries[in_out_params->min_clk_index].pre_derate_dram_bw_kbps, &mode_lib->soc.clk_table.dram_config));
- for (k = 0; k < mode_lib->mp.num_active_pipes; ++k) {
- dml2_printf("DML::%s: pipe=%d is in plane=%d\n", __func__, k, mode_lib->mp.pipe_plane[k]);
- dml2_printf("DML::%s: Per-plane DPPPerSurface[%0d] = %d\n", __func__, k, mode_lib->mp.NoOfDPP[k]);
- }
-
- for (k = 0; k < s->num_active_planes; k++)
- dml2_printf("DML::%s: plane_%d: reserved_vblank_time_ns = %u\n", __func__, k, display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns);
-#endif
-
- CalculateMaxDETAndMinCompressedBufferSize(
- mode_lib->ip.config_return_buffer_size_in_kbytes,
- mode_lib->ip.config_return_buffer_segment_size_in_kbytes,
- mode_lib->ip.rob_buffer_size_kbytes,
- mode_lib->ip.max_num_dpp,
- display_cfg->overrides.hw.force_nom_det_size_kbytes.enable,
- display_cfg->overrides.hw.force_nom_det_size_kbytes.value,
- mode_lib->ip.dcn_mrq_present,
-
- /* Output */
- &s->MaxTotalDETInKByte,
- &s->NomDETInKByte,
- &s->MinCompressedBufferSizeInKByte);
-
-
- PixelClockAdjustmentForProgressiveToInterlaceUnit(display_cfg, mode_lib->ip.ptoi_supported, s->PixelClockBackEnd);
-
- for (k = 0; k < s->num_active_planes; ++k) {
- CalculateSinglePipeDPPCLKAndSCLThroughput(
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio,
- mode_lib->ip.max_dchub_pscl_bw_pix_per_clk,
- mode_lib->ip.max_pscl_lb_bw_pix_per_clk,
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000),
- display_cfg->plane_descriptors[k].pixel_format,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_taps,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.h_taps,
-
- /* Output */
- &mode_lib->mp.PSCL_THROUGHPUT[k],
- &mode_lib->mp.PSCL_THROUGHPUT_CHROMA[k],
- &mode_lib->mp.DPPCLKUsingSingleDPP[k]);
- }
-
- for (k = 0; k < s->num_active_planes; ++k) {
- CalculateBytePerPixelAndBlockSizes(
- display_cfg->plane_descriptors[k].pixel_format,
- display_cfg->plane_descriptors[k].surface.tiling,
- display_cfg->plane_descriptors[k].surface.plane0.pitch,
- display_cfg->plane_descriptors[k].surface.plane1.pitch,
-
- // Output
- &mode_lib->mp.BytePerPixelY[k],
- &mode_lib->mp.BytePerPixelC[k],
- &mode_lib->mp.BytePerPixelInDETY[k],
- &mode_lib->mp.BytePerPixelInDETC[k],
- &mode_lib->mp.Read256BlockHeightY[k],
- &mode_lib->mp.Read256BlockHeightC[k],
- &mode_lib->mp.Read256BlockWidthY[k],
- &mode_lib->mp.Read256BlockWidthC[k],
- &mode_lib->mp.MacroTileHeightY[k],
- &mode_lib->mp.MacroTileHeightC[k],
- &mode_lib->mp.MacroTileWidthY[k],
- &mode_lib->mp.MacroTileWidthC[k],
- &mode_lib->mp.surf_linear128_l[k],
- &mode_lib->mp.surf_linear128_c[k]);
- }
-
- CalculateSwathWidth(
- display_cfg,
- false, // ForceSingleDPP
- s->num_active_planes,
- mode_lib->mp.ODMMode,
- mode_lib->mp.BytePerPixelY,
- mode_lib->mp.BytePerPixelC,
- mode_lib->mp.Read256BlockHeightY,
- mode_lib->mp.Read256BlockHeightC,
- mode_lib->mp.Read256BlockWidthY,
- mode_lib->mp.Read256BlockWidthC,
- mode_lib->mp.surf_linear128_l,
- mode_lib->mp.surf_linear128_c,
- mode_lib->mp.NoOfDPP,
-
- /* Output */
- mode_lib->mp.req_per_swath_ub_l,
- mode_lib->mp.req_per_swath_ub_c,
- mode_lib->mp.SwathWidthSingleDPPY,
- mode_lib->mp.SwathWidthSingleDPPC,
- mode_lib->mp.SwathWidthY,
- mode_lib->mp.SwathWidthC,
- s->dummy_integer_array[0], // unsigned int MaximumSwathHeightY[]
- s->dummy_integer_array[1], // unsigned int MaximumSwathHeightC[]
- mode_lib->mp.swath_width_luma_ub,
- mode_lib->mp.swath_width_chroma_ub);
-
- for (k = 0; k < s->num_active_planes; ++k) {
- mode_lib->mp.cursor_bw[k] = display_cfg->plane_descriptors[k].cursor.num_cursors * display_cfg->plane_descriptors[k].cursor.cursor_width * display_cfg->plane_descriptors[k].cursor.cursor_bpp / 8.0 /
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000));
- mode_lib->mp.SurfaceReadBandwidthLuma[k] = mode_lib->mp.SwathWidthSingleDPPY[k] * mode_lib->mp.BytePerPixelY[k] / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- mode_lib->mp.SurfaceReadBandwidthChroma[k] = mode_lib->mp.SwathWidthSingleDPPC[k] * mode_lib->mp.BytePerPixelC[k] / (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- dml2_printf("DML::%s: ReadBandwidthSurfaceLuma[%i] = %fBps\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthLuma[k]);
- dml2_printf("DML::%s: ReadBandwidthSurfaceChroma[%i] = %fBps\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthChroma[k]);
- }
-
- CalculateSwathAndDETConfiguration_params->display_cfg = display_cfg;
- CalculateSwathAndDETConfiguration_params->ConfigReturnBufferSizeInKByte = mode_lib->ip.config_return_buffer_size_in_kbytes;
- CalculateSwathAndDETConfiguration_params->MaxTotalDETInKByte = s->MaxTotalDETInKByte;
- CalculateSwathAndDETConfiguration_params->MinCompressedBufferSizeInKByte = s->MinCompressedBufferSizeInKByte;
- CalculateSwathAndDETConfiguration_params->rob_buffer_size_kbytes = mode_lib->ip.rob_buffer_size_kbytes;
- CalculateSwathAndDETConfiguration_params->pixel_chunk_size_kbytes = mode_lib->ip.pixel_chunk_size_kbytes;
- CalculateSwathAndDETConfiguration_params->rob_buffer_size_kbytes = mode_lib->ip.rob_buffer_size_kbytes;
- CalculateSwathAndDETConfiguration_params->pixel_chunk_size_kbytes = mode_lib->ip.pixel_chunk_size_kbytes;
-
- CalculateSwathAndDETConfiguration_params->ForceSingleDPP = false;
- CalculateSwathAndDETConfiguration_params->NumberOfActiveSurfaces = s->num_active_planes;
- CalculateSwathAndDETConfiguration_params->nomDETInKByte = s->NomDETInKByte;
- CalculateSwathAndDETConfiguration_params->ConfigReturnBufferSegmentSizeInkByte = mode_lib->ip.config_return_buffer_segment_size_in_kbytes;
- CalculateSwathAndDETConfiguration_params->CompressedBufferSegmentSizeInkByte = mode_lib->ip.compressed_buffer_segment_size_in_kbytes;
- CalculateSwathAndDETConfiguration_params->ReadBandwidthLuma = mode_lib->mp.SurfaceReadBandwidthLuma;
- CalculateSwathAndDETConfiguration_params->ReadBandwidthChroma = mode_lib->mp.SurfaceReadBandwidthChroma;
- CalculateSwathAndDETConfiguration_params->MaximumSwathWidthLuma = s->dummy_single_array[0];
- CalculateSwathAndDETConfiguration_params->MaximumSwathWidthChroma = s->dummy_single_array[1];
- CalculateSwathAndDETConfiguration_params->Read256BytesBlockHeightY = mode_lib->mp.Read256BlockHeightY;
- CalculateSwathAndDETConfiguration_params->Read256BytesBlockHeightC = mode_lib->mp.Read256BlockHeightC;
- CalculateSwathAndDETConfiguration_params->Read256BytesBlockWidthY = mode_lib->mp.Read256BlockWidthY;
- CalculateSwathAndDETConfiguration_params->Read256BytesBlockWidthC = mode_lib->mp.Read256BlockWidthC;
- CalculateSwathAndDETConfiguration_params->surf_linear128_l = mode_lib->mp.surf_linear128_l;
- CalculateSwathAndDETConfiguration_params->surf_linear128_c = mode_lib->mp.surf_linear128_c;
- CalculateSwathAndDETConfiguration_params->ODMMode = mode_lib->mp.ODMMode;
- CalculateSwathAndDETConfiguration_params->DPPPerSurface = mode_lib->mp.NoOfDPP;
- CalculateSwathAndDETConfiguration_params->BytePerPixY = mode_lib->mp.BytePerPixelY;
- CalculateSwathAndDETConfiguration_params->BytePerPixC = mode_lib->mp.BytePerPixelC;
- CalculateSwathAndDETConfiguration_params->BytePerPixDETY = mode_lib->mp.BytePerPixelInDETY;
- CalculateSwathAndDETConfiguration_params->BytePerPixDETC = mode_lib->mp.BytePerPixelInDETC;
-
- // output
- CalculateSwathAndDETConfiguration_params->req_per_swath_ub_l = mode_lib->mp.req_per_swath_ub_l;
- CalculateSwathAndDETConfiguration_params->req_per_swath_ub_c = mode_lib->mp.req_per_swath_ub_c;
- CalculateSwathAndDETConfiguration_params->swath_width_luma_ub = s->dummy_long_array[0];
- CalculateSwathAndDETConfiguration_params->swath_width_chroma_ub = s->dummy_long_array[1];
- CalculateSwathAndDETConfiguration_params->SwathWidth = s->dummy_long_array[2];
- CalculateSwathAndDETConfiguration_params->SwathWidthChroma = s->dummy_long_array[3];
- CalculateSwathAndDETConfiguration_params->SwathHeightY = mode_lib->mp.SwathHeightY;
- CalculateSwathAndDETConfiguration_params->SwathHeightC = mode_lib->mp.SwathHeightC;
- CalculateSwathAndDETConfiguration_params->request_size_bytes_luma = mode_lib->mp.request_size_bytes_luma;
- CalculateSwathAndDETConfiguration_params->request_size_bytes_chroma = mode_lib->mp.request_size_bytes_chroma;
- CalculateSwathAndDETConfiguration_params->DETBufferSizeInKByte = mode_lib->mp.DETBufferSizeInKByte;
- CalculateSwathAndDETConfiguration_params->DETBufferSizeY = mode_lib->mp.DETBufferSizeY;
- CalculateSwathAndDETConfiguration_params->DETBufferSizeC = mode_lib->mp.DETBufferSizeC;
- CalculateSwathAndDETConfiguration_params->full_swath_bytes_l = s->full_swath_bytes_l;
- CalculateSwathAndDETConfiguration_params->full_swath_bytes_c = s->full_swath_bytes_c;
- CalculateSwathAndDETConfiguration_params->UnboundedRequestEnabled = &mode_lib->mp.UnboundedRequestEnabled;
- CalculateSwathAndDETConfiguration_params->compbuf_reserved_space_64b = &mode_lib->mp.compbuf_reserved_space_64b;
- CalculateSwathAndDETConfiguration_params->hw_debug5 = &mode_lib->mp.hw_debug5;
- CalculateSwathAndDETConfiguration_params->CompressedBufferSizeInkByte = &mode_lib->mp.CompressedBufferSizeInkByte;
- CalculateSwathAndDETConfiguration_params->ViewportSizeSupportPerSurface = &s->dummy_boolean_array[0][0];
- CalculateSwathAndDETConfiguration_params->ViewportSizeSupport = &s->dummy_boolean[0];
- CalculateSwathAndDETConfiguration_params->funcs = &mode_lib->funcs;
-
- // VBA_DELTA
- // Calculate DET size, swath height here. In VBA, they are calculated in mode check stage
- CalculateSwathAndDETConfiguration(&mode_lib->scratch, CalculateSwathAndDETConfiguration_params);
-
- // DSCCLK
- /*
- s->DSCFormatFactor = 0;
- for (k = 0; k < s->num_active_planes; ++k) {
- if ((display_cfg->plane_descriptors[k].stream_index != k) || !cfg_support_info->stream_support_info[display_cfg->plane_descriptors[k].stream_index].dsc_enable) {
- } else {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format == dml2_420)
- s->DSCFormatFactor = 2;
- else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format == dml2_444)
- s->DSCFormatFactor = 1;
- else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format == dml2_n422 ||
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder == dml2_hdmifrl)
- s->DSCFormatFactor = 2;
- else
- s->DSCFormatFactor = 1;
-
- s->PixelClockBackEndFactor = 3.0;
-
- if (mode_lib->mp.ODMMode[k] == dml2_odm_mode_combine_4to1)
- s->PixelClockBackEndFactor = 12.0;
- else if (mode_lib->mp.ODMMode[k] == dml2_odm_mode_combine_3to1)
- s->PixelClockBackEndFactor = 9.0;
- else if (mode_lib->mp.ODMMode[k] == dml2_odm_mode_combine_2to1)
- s->PixelClockBackEndFactor = 6.0;
-
- }
- #ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, DSCEnabled = %u\n", __func__, k, cfg_support_info->stream_support_info[display_cfg->plane_descriptors[k].stream_index].dsc_enable);
- dml2_printf("DML::%s: k=%u, BlendingAndTiming = %u\n", __func__, k, display_cfg->plane_descriptors[k].stream_index);
- dml2_printf("DML::%s: k=%u, PixelClockBackEndFactor = %f\n", __func__, k, s->PixelClockBackEndFactor);
- dml2_printf("DML::%s: k=%u, PixelClockBackEnd = %f\n", __func__, k, s->PixelClockBackEnd[k]);
- dml2_printf("DML::%s: k=%u, DSCFormatFactor = %u\n", __func__, k, s->DSCFormatFactor);
- dml2_printf("DML::%s: k=%u, DSCCLK = %f\n", __func__, k, mode_lib->mp.DSCCLK[k]);
- #endif
- }
- */
-
- // DSC Delay
- for (k = 0; k < s->num_active_planes; ++k) {
- mode_lib->mp.DSCDelay[k] = DSCDelayRequirement(cfg_support_info->stream_support_info[display_cfg->plane_descriptors[k].stream_index].dsc_enable,
- mode_lib->mp.ODMMode[k],
- mode_lib->ip.maximum_dsc_bits_per_component,
- s->OutputBpp[k],
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total,
- cfg_support_info->stream_support_info[display_cfg->plane_descriptors[k].stream_index].num_dsc_slices,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_encoder,
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000),
- s->PixelClockBackEnd[k]);
- }
-
- for (k = 0; k < s->num_active_planes; ++k)
- for (j = 0; j < s->num_active_planes; ++j) // NumberOfSurfaces
- if (j != k && display_cfg->plane_descriptors[k].stream_index == j && cfg_support_info->stream_support_info[display_cfg->plane_descriptors[j].stream_index].dsc_enable)
- mode_lib->mp.DSCDelay[k] = mode_lib->mp.DSCDelay[j];
-
- // Prefetch
- if (mode_lib->soc.mall_allocated_for_dcn_mbytes == 0) {
- for (k = 0; k < s->num_active_planes; ++k)
- mode_lib->mp.SurfaceSizeInTheMALL[k] = 0;
- } else {
- CalculateSurfaceSizeInMall(
- display_cfg,
- s->num_active_planes,
- mode_lib->soc.mall_allocated_for_dcn_mbytes,
- mode_lib->mp.BytePerPixelY,
- mode_lib->mp.BytePerPixelC,
- mode_lib->mp.Read256BlockWidthY,
- mode_lib->mp.Read256BlockWidthC,
- mode_lib->mp.Read256BlockHeightY,
- mode_lib->mp.Read256BlockHeightC,
- mode_lib->mp.MacroTileWidthY,
- mode_lib->mp.MacroTileWidthC,
- mode_lib->mp.MacroTileHeightY,
- mode_lib->mp.MacroTileHeightC,
-
- /* Output */
- mode_lib->mp.SurfaceSizeInTheMALL,
- &s->dummy_boolean[0]); /* bool *ExceededMALLSize */
- }
-
- for (k = 0; k < s->num_active_planes; ++k) {
- s->SurfaceParameters[k].PixelClock = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- s->SurfaceParameters[k].DPPPerSurface = mode_lib->mp.NoOfDPP[k];
- s->SurfaceParameters[k].RotationAngle = display_cfg->plane_descriptors[k].composition.rotation_angle;
- s->SurfaceParameters[k].ViewportHeight = display_cfg->plane_descriptors[k].composition.viewport.plane0.height;
- s->SurfaceParameters[k].ViewportHeightC = display_cfg->plane_descriptors[k].composition.viewport.plane1.height;
- s->SurfaceParameters[k].BlockWidth256BytesY = mode_lib->mp.Read256BlockWidthY[k];
- s->SurfaceParameters[k].BlockHeight256BytesY = mode_lib->mp.Read256BlockHeightY[k];
- s->SurfaceParameters[k].BlockWidth256BytesC = mode_lib->mp.Read256BlockWidthC[k];
- s->SurfaceParameters[k].BlockHeight256BytesC = mode_lib->mp.Read256BlockHeightC[k];
- s->SurfaceParameters[k].BlockWidthY = mode_lib->mp.MacroTileWidthY[k];
- s->SurfaceParameters[k].BlockHeightY = mode_lib->mp.MacroTileHeightY[k];
- s->SurfaceParameters[k].BlockWidthC = mode_lib->mp.MacroTileWidthC[k];
- s->SurfaceParameters[k].BlockHeightC = mode_lib->mp.MacroTileHeightC[k];
- s->SurfaceParameters[k].InterlaceEnable = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced;
- s->SurfaceParameters[k].HTotal = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total;
- s->SurfaceParameters[k].DCCEnable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- s->SurfaceParameters[k].SourcePixelFormat = display_cfg->plane_descriptors[k].pixel_format;
- s->SurfaceParameters[k].SurfaceTiling = display_cfg->plane_descriptors[k].surface.tiling;
- s->SurfaceParameters[k].BytePerPixelY = mode_lib->mp.BytePerPixelY[k];
- s->SurfaceParameters[k].BytePerPixelC = mode_lib->mp.BytePerPixelC[k];
- s->SurfaceParameters[k].ProgressiveToInterlaceUnitInOPP = mode_lib->ip.ptoi_supported;
- s->SurfaceParameters[k].VRatio = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- s->SurfaceParameters[k].VRatioChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- s->SurfaceParameters[k].VTaps = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps;
- s->SurfaceParameters[k].VTapsChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps;
- s->SurfaceParameters[k].PitchY = display_cfg->plane_descriptors[k].surface.plane0.pitch;
- s->SurfaceParameters[k].PitchC = display_cfg->plane_descriptors[k].surface.plane1.pitch;
- s->SurfaceParameters[k].ViewportStationary = display_cfg->plane_descriptors[k].composition.viewport.stationary;
- s->SurfaceParameters[k].ViewportXStart = display_cfg->plane_descriptors[k].composition.viewport.plane0.x_start;
- s->SurfaceParameters[k].ViewportYStart = display_cfg->plane_descriptors[k].composition.viewport.plane0.y_start;
- s->SurfaceParameters[k].ViewportXStartC = display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start;
- s->SurfaceParameters[k].ViewportYStartC = display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start;
- s->SurfaceParameters[k].FORCE_ONE_ROW_FOR_FRAME = display_cfg->plane_descriptors[k].overrides.hw.force_one_row_for_frame;
- s->SurfaceParameters[k].SwathHeightY = mode_lib->mp.SwathHeightY[k];
- s->SurfaceParameters[k].SwathHeightC = mode_lib->mp.SwathHeightC[k];
- s->SurfaceParameters[k].DCCMetaPitchY = display_cfg->plane_descriptors[k].surface.dcc.plane0.pitch;
- s->SurfaceParameters[k].DCCMetaPitchC = display_cfg->plane_descriptors[k].surface.dcc.plane1.pitch;
- }
-
- CalculateVMRowAndSwath_params->display_cfg = display_cfg;
- CalculateVMRowAndSwath_params->NumberOfActiveSurfaces = s->num_active_planes;
- CalculateVMRowAndSwath_params->myPipe = s->SurfaceParameters;
- CalculateVMRowAndSwath_params->SurfaceSizeInMALL = mode_lib->mp.SurfaceSizeInTheMALL;
- CalculateVMRowAndSwath_params->PTEBufferSizeInRequestsLuma = mode_lib->ip.dpte_buffer_size_in_pte_reqs_luma;
- CalculateVMRowAndSwath_params->PTEBufferSizeInRequestsChroma = mode_lib->ip.dpte_buffer_size_in_pte_reqs_chroma;
- CalculateVMRowAndSwath_params->MALLAllocatedForDCN = mode_lib->soc.mall_allocated_for_dcn_mbytes;
- CalculateVMRowAndSwath_params->SwathWidthY = mode_lib->mp.SwathWidthY;
- CalculateVMRowAndSwath_params->SwathWidthC = mode_lib->mp.SwathWidthC;
- CalculateVMRowAndSwath_params->HostVMMinPageSize = mode_lib->soc.hostvm_min_page_size_kbytes;
- CalculateVMRowAndSwath_params->DCCMetaBufferSizeBytes = mode_lib->ip.dcc_meta_buffer_size_bytes;
- CalculateVMRowAndSwath_params->mrq_present = mode_lib->ip.dcn_mrq_present;
-
- // output
- CalculateVMRowAndSwath_params->PTEBufferSizeNotExceeded = s->dummy_boolean_array[0];
- CalculateVMRowAndSwath_params->dpte_row_width_luma_ub = mode_lib->mp.dpte_row_width_luma_ub;
- CalculateVMRowAndSwath_params->dpte_row_width_chroma_ub = mode_lib->mp.dpte_row_width_chroma_ub;
- CalculateVMRowAndSwath_params->dpte_row_height_luma = mode_lib->mp.dpte_row_height;
- CalculateVMRowAndSwath_params->dpte_row_height_chroma = mode_lib->mp.dpte_row_height_chroma;
- CalculateVMRowAndSwath_params->dpte_row_height_linear_luma = mode_lib->mp.dpte_row_height_linear;
- CalculateVMRowAndSwath_params->dpte_row_height_linear_chroma = mode_lib->mp.dpte_row_height_linear_chroma;
- CalculateVMRowAndSwath_params->vm_group_bytes = mode_lib->mp.vm_group_bytes;
- CalculateVMRowAndSwath_params->dpte_group_bytes = mode_lib->mp.dpte_group_bytes;
- CalculateVMRowAndSwath_params->PixelPTEReqWidthY = mode_lib->mp.PixelPTEReqWidthY;
- CalculateVMRowAndSwath_params->PixelPTEReqHeightY = mode_lib->mp.PixelPTEReqHeightY;
- CalculateVMRowAndSwath_params->PTERequestSizeY = mode_lib->mp.PTERequestSizeY;
- CalculateVMRowAndSwath_params->PixelPTEReqWidthC = mode_lib->mp.PixelPTEReqWidthC;
- CalculateVMRowAndSwath_params->PixelPTEReqHeightC = mode_lib->mp.PixelPTEReqHeightC;
- CalculateVMRowAndSwath_params->PTERequestSizeC = mode_lib->mp.PTERequestSizeC;
- CalculateVMRowAndSwath_params->vmpg_width_y = s->vmpg_width_y;
- CalculateVMRowAndSwath_params->vmpg_height_y = s->vmpg_height_y;
- CalculateVMRowAndSwath_params->vmpg_width_c = s->vmpg_width_c;
- CalculateVMRowAndSwath_params->vmpg_height_c = s->vmpg_height_c;
- CalculateVMRowAndSwath_params->dpde0_bytes_per_frame_ub_l = mode_lib->mp.dpde0_bytes_per_frame_ub_l;
- CalculateVMRowAndSwath_params->dpde0_bytes_per_frame_ub_c = mode_lib->mp.dpde0_bytes_per_frame_ub_c;
- CalculateVMRowAndSwath_params->PrefetchSourceLinesY = mode_lib->mp.PrefetchSourceLinesY;
- CalculateVMRowAndSwath_params->PrefetchSourceLinesC = mode_lib->mp.PrefetchSourceLinesC;
- CalculateVMRowAndSwath_params->VInitPreFillY = mode_lib->mp.VInitPreFillY;
- CalculateVMRowAndSwath_params->VInitPreFillC = mode_lib->mp.VInitPreFillC;
- CalculateVMRowAndSwath_params->MaxNumSwathY = mode_lib->mp.MaxNumSwathY;
- CalculateVMRowAndSwath_params->MaxNumSwathC = mode_lib->mp.MaxNumSwathC;
- CalculateVMRowAndSwath_params->dpte_row_bw = mode_lib->mp.dpte_row_bw;
- CalculateVMRowAndSwath_params->PixelPTEBytesPerRow = mode_lib->mp.PixelPTEBytesPerRow;
- CalculateVMRowAndSwath_params->vm_bytes = mode_lib->mp.vm_bytes;
- CalculateVMRowAndSwath_params->use_one_row_for_frame = mode_lib->mp.use_one_row_for_frame;
- CalculateVMRowAndSwath_params->use_one_row_for_frame_flip = mode_lib->mp.use_one_row_for_frame_flip;
- CalculateVMRowAndSwath_params->is_using_mall_for_ss = mode_lib->mp.is_using_mall_for_ss;
- CalculateVMRowAndSwath_params->PTE_BUFFER_MODE = mode_lib->mp.PTE_BUFFER_MODE;
- CalculateVMRowAndSwath_params->BIGK_FRAGMENT_SIZE = mode_lib->mp.BIGK_FRAGMENT_SIZE;
- CalculateVMRowAndSwath_params->DCCMetaBufferSizeNotExceeded = s->dummy_boolean_array[1];
- CalculateVMRowAndSwath_params->meta_row_bw = mode_lib->mp.meta_row_bw;
- CalculateVMRowAndSwath_params->meta_row_bytes = mode_lib->mp.meta_row_bytes;
- CalculateVMRowAndSwath_params->meta_req_width_luma = mode_lib->mp.meta_req_width;
- CalculateVMRowAndSwath_params->meta_req_height_luma = mode_lib->mp.meta_req_height;
- CalculateVMRowAndSwath_params->meta_row_width_luma = mode_lib->mp.meta_row_width;
- CalculateVMRowAndSwath_params->meta_row_height_luma = mode_lib->mp.meta_row_height;
- CalculateVMRowAndSwath_params->meta_pte_bytes_per_frame_ub_l = mode_lib->mp.meta_pte_bytes_per_frame_ub_l;
- CalculateVMRowAndSwath_params->meta_req_width_chroma = mode_lib->mp.meta_req_width_chroma;
- CalculateVMRowAndSwath_params->meta_row_height_chroma = mode_lib->mp.meta_row_height_chroma;
- CalculateVMRowAndSwath_params->meta_row_width_chroma = mode_lib->mp.meta_row_width_chroma;
- CalculateVMRowAndSwath_params->meta_req_height_chroma = mode_lib->mp.meta_req_height_chroma;
- CalculateVMRowAndSwath_params->meta_pte_bytes_per_frame_ub_c = mode_lib->mp.meta_pte_bytes_per_frame_ub_c;
-
- CalculateVMRowAndSwath(&mode_lib->scratch, CalculateVMRowAndSwath_params);
-
- memset(calculate_mcache_setting_params, 0, sizeof(struct dml2_core_calcs_calculate_mcache_setting_params));
- if (mode_lib->soc.mall_allocated_for_dcn_mbytes == 0 || mode_lib->ip.dcn_mrq_present) {
- for (k = 0; k < s->num_active_planes; k++) {
- mode_lib->mp.mall_prefetch_sdp_overhead_factor[k] = 1.0;
- mode_lib->mp.mall_prefetch_dram_overhead_factor[k] = 1.0;
- mode_lib->mp.dcc_dram_bw_nom_overhead_factor_p0[k] = 1.0;
- mode_lib->mp.dcc_dram_bw_pref_overhead_factor_p0[k] = 1.0;
- mode_lib->mp.dcc_dram_bw_nom_overhead_factor_p1[k] = 1.0;
- mode_lib->mp.dcc_dram_bw_pref_overhead_factor_p1[k] = 1.0;
- }
- } else {
- for (k = 0; k < s->num_active_planes; k++) {
- calculate_mcache_setting_params->dcc_enable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- calculate_mcache_setting_params->num_chans = mode_lib->soc.clk_table.dram_config.channel_count;
- calculate_mcache_setting_params->mem_word_bytes = mode_lib->soc.mem_word_bytes;
- calculate_mcache_setting_params->mcache_size_bytes = mode_lib->soc.mcache_size_bytes;
- calculate_mcache_setting_params->mcache_line_size_bytes = mode_lib->soc.mcache_line_size_bytes;
- calculate_mcache_setting_params->gpuvm_enable = display_cfg->gpuvm_enable;
- calculate_mcache_setting_params->gpuvm_page_size_kbytes = display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes;
-
- calculate_mcache_setting_params->source_format = display_cfg->plane_descriptors[k].pixel_format;
- calculate_mcache_setting_params->surf_vert = dml_is_vertical_rotation(display_cfg->plane_descriptors[k].composition.rotation_angle);
- calculate_mcache_setting_params->vp_stationary = display_cfg->plane_descriptors[k].composition.viewport.stationary;
- calculate_mcache_setting_params->tiling_mode = display_cfg->plane_descriptors[k].surface.tiling;
- calculate_mcache_setting_params->imall_enable = mode_lib->ip.imall_supported && display_cfg->plane_descriptors[k].overrides.legacy_svp_config == dml2_svp_mode_override_imall;
-
- calculate_mcache_setting_params->vp_start_x_l = display_cfg->plane_descriptors[k].composition.viewport.plane0.x_start;
- calculate_mcache_setting_params->vp_start_y_l = display_cfg->plane_descriptors[k].composition.viewport.plane0.y_start;
- calculate_mcache_setting_params->full_vp_width_l = display_cfg->plane_descriptors[k].composition.viewport.plane0.width;
- calculate_mcache_setting_params->full_vp_height_l = display_cfg->plane_descriptors[k].composition.viewport.plane0.height;
- calculate_mcache_setting_params->blk_width_l = mode_lib->mp.MacroTileWidthY[k];
- calculate_mcache_setting_params->blk_height_l = mode_lib->mp.MacroTileHeightY[k];
- calculate_mcache_setting_params->vmpg_width_l = s->vmpg_width_y[k];
- calculate_mcache_setting_params->vmpg_height_l = s->vmpg_height_y[k];
- calculate_mcache_setting_params->full_swath_bytes_l = s->full_swath_bytes_l[k];
- calculate_mcache_setting_params->bytes_per_pixel_l = mode_lib->mp.BytePerPixelY[k];
-
- calculate_mcache_setting_params->vp_start_x_c = display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start;
- calculate_mcache_setting_params->vp_start_y_c = display_cfg->plane_descriptors[k].composition.viewport.plane1.y_start;
- calculate_mcache_setting_params->full_vp_width_c = display_cfg->plane_descriptors[k].composition.viewport.plane1.width;
- calculate_mcache_setting_params->full_vp_height_c = display_cfg->plane_descriptors[k].composition.viewport.plane1.height;
- calculate_mcache_setting_params->blk_width_c = mode_lib->mp.MacroTileWidthC[k];
- calculate_mcache_setting_params->blk_height_c = mode_lib->mp.MacroTileHeightC[k];
- calculate_mcache_setting_params->vmpg_width_c = s->vmpg_width_c[k];
- calculate_mcache_setting_params->vmpg_height_c = s->vmpg_height_c[k];
- calculate_mcache_setting_params->full_swath_bytes_c = s->full_swath_bytes_c[k];
- calculate_mcache_setting_params->bytes_per_pixel_c = mode_lib->mp.BytePerPixelC[k];
-
- // output
- calculate_mcache_setting_params->dcc_dram_bw_nom_overhead_factor_l = &mode_lib->mp.dcc_dram_bw_nom_overhead_factor_p0[k];
- calculate_mcache_setting_params->dcc_dram_bw_pref_overhead_factor_l = &mode_lib->mp.dcc_dram_bw_pref_overhead_factor_p0[k];
- calculate_mcache_setting_params->dcc_dram_bw_nom_overhead_factor_c = &mode_lib->mp.dcc_dram_bw_nom_overhead_factor_p1[k];
- calculate_mcache_setting_params->dcc_dram_bw_pref_overhead_factor_c = &mode_lib->mp.dcc_dram_bw_pref_overhead_factor_p1[k];
-
- calculate_mcache_setting_params->num_mcaches_l = &mode_lib->mp.num_mcaches_l[k];
- calculate_mcache_setting_params->mcache_row_bytes_l = &mode_lib->mp.mcache_row_bytes_l[k];
- calculate_mcache_setting_params->mcache_offsets_l = mode_lib->mp.mcache_offsets_l[k];
- calculate_mcache_setting_params->mcache_shift_granularity_l = &mode_lib->mp.mcache_shift_granularity_l[k];
-
- calculate_mcache_setting_params->num_mcaches_c = &mode_lib->mp.num_mcaches_c[k];
- calculate_mcache_setting_params->mcache_row_bytes_c = &mode_lib->mp.mcache_row_bytes_c[k];
- calculate_mcache_setting_params->mcache_offsets_c = mode_lib->mp.mcache_offsets_c[k];
- calculate_mcache_setting_params->mcache_shift_granularity_c = &mode_lib->mp.mcache_shift_granularity_c[k];
-
- calculate_mcache_setting_params->mall_comb_mcache_l = &mode_lib->mp.mall_comb_mcache_l[k];
- calculate_mcache_setting_params->mall_comb_mcache_c = &mode_lib->mp.mall_comb_mcache_c[k];
- calculate_mcache_setting_params->lc_comb_mcache = &mode_lib->mp.lc_comb_mcache[k];
- calculate_mcache_setting(&mode_lib->scratch, calculate_mcache_setting_params);
- }
-
- calculate_mall_bw_overhead_factor(
- mode_lib->mp.mall_prefetch_sdp_overhead_factor,
- mode_lib->mp.mall_prefetch_dram_overhead_factor,
-
- // input
- display_cfg,
- s->num_active_planes);
- }
-
- // Calculate all the bandwidth availabe
- calculate_bandwidth_available(
- mode_lib->mp.avg_bandwidth_available_min,
- mode_lib->mp.avg_bandwidth_available,
- mode_lib->mp.urg_bandwidth_available_min,
- mode_lib->mp.urg_bandwidth_available,
- mode_lib->mp.urg_bandwidth_available_vm_only,
- mode_lib->mp.urg_bandwidth_available_pixel_and_vm,
-
- &mode_lib->soc,
- display_cfg->hostvm_enable,
- mode_lib->mp.Dcfclk,
- mode_lib->mp.FabricClock,
- mode_lib->mp.dram_bw_mbps);
-
-
- calculate_hostvm_inefficiency_factor(
- &s->HostVMInefficiencyFactor,
- &s->HostVMInefficiencyFactorPrefetch,
-
- display_cfg->gpuvm_enable,
- display_cfg->hostvm_enable,
- mode_lib->ip.remote_iommu_outstanding_translations,
- mode_lib->soc.max_outstanding_reqs,
- mode_lib->mp.urg_bandwidth_available_pixel_and_vm[dml2_core_internal_soc_state_sys_active],
- mode_lib->mp.urg_bandwidth_available_vm_only[dml2_core_internal_soc_state_sys_active]);
-
- s->TotalDCCActiveDPP = 0;
- s->TotalActiveDPP = 0;
- for (k = 0; k < s->num_active_planes; ++k) {
- s->TotalActiveDPP = s->TotalActiveDPP + mode_lib->mp.NoOfDPP[k];
- if (display_cfg->plane_descriptors[k].surface.dcc.enable)
- s->TotalDCCActiveDPP = s->TotalDCCActiveDPP + mode_lib->mp.NoOfDPP[k];
- }
- // Calculate tdlut schedule related terms
- for (k = 0; k <= s->num_active_planes - 1; k++) {
- calculate_tdlut_setting_params->dispclk_mhz = mode_lib->mp.Dispclk;
- calculate_tdlut_setting_params->setup_for_tdlut = display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut;
- calculate_tdlut_setting_params->tdlut_width_mode = display_cfg->plane_descriptors[k].tdlut.tdlut_width_mode;
- calculate_tdlut_setting_params->tdlut_addressing_mode = display_cfg->plane_descriptors[k].tdlut.tdlut_addressing_mode;
- calculate_tdlut_setting_params->cursor_buffer_size = mode_lib->ip.cursor_buffer_size;
- calculate_tdlut_setting_params->gpuvm_enable = display_cfg->gpuvm_enable;
- calculate_tdlut_setting_params->gpuvm_page_size_kbytes = display_cfg->plane_descriptors[k].overrides.gpuvm_min_page_size_kbytes;
-
- // output
- calculate_tdlut_setting_params->tdlut_pte_bytes_per_frame = &s->tdlut_pte_bytes_per_frame[k];
- calculate_tdlut_setting_params->tdlut_bytes_per_frame = &s->tdlut_bytes_per_frame[k];
- calculate_tdlut_setting_params->tdlut_groups_per_2row_ub = &s->tdlut_groups_per_2row_ub[k];
- calculate_tdlut_setting_params->tdlut_opt_time = &s->tdlut_opt_time[k];
- calculate_tdlut_setting_params->tdlut_drain_time = &s->tdlut_drain_time[k];
- calculate_tdlut_setting_params->tdlut_bytes_per_group = &s->tdlut_bytes_per_group[k];
-
- calculate_tdlut_setting(&mode_lib->scratch, calculate_tdlut_setting_params);
- }
-
- if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn3)
- s->ReorderingBytes = (unsigned int)(mode_lib->soc.clk_table.dram_config.channel_count * math_max3(mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_pixel_only_bytes,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_pixel_and_vm_bytes,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_out_of_order_return_per_channel_vm_only_bytes));
-
- CalculateExtraLatency(
- display_cfg,
- mode_lib->ip.rob_buffer_size_kbytes,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.loaded_round_trip_latency_fclk_cycles,
- s->ReorderingBytes,
- mode_lib->mp.Dcfclk,
- mode_lib->mp.FabricClock,
- mode_lib->ip.pixel_chunk_size_kbytes,
- mode_lib->mp.urg_bandwidth_available_min[dml2_core_internal_soc_state_sys_active],
- s->num_active_planes,
- mode_lib->mp.NoOfDPP,
- mode_lib->mp.dpte_group_bytes,
- s->tdlut_bytes_per_group,
- s->HostVMInefficiencyFactor,
- s->HostVMInefficiencyFactorPrefetch,
- mode_lib->soc.hostvm_min_page_size_kbytes,
- mode_lib->soc.qos_parameters.qos_type,
- !(display_cfg->overrides.max_outstanding_when_urgent_expected_disable),
- mode_lib->soc.max_outstanding_reqs,
- mode_lib->mp.request_size_bytes_luma,
- mode_lib->mp.request_size_bytes_chroma,
- mode_lib->ip.meta_chunk_size_kbytes,
- mode_lib->ip.dchub_arb_to_ret_delay,
- mode_lib->mp.TripToMemory,
- mode_lib->ip.hostvm_mode,
-
- // output
- &mode_lib->mp.ExtraLatency,
- &mode_lib->mp.ExtraLatency_sr,
- &mode_lib->mp.ExtraLatencyPrefetch);
-
- mode_lib->mp.TCalc = 24.0 / mode_lib->mp.DCFCLKDeepSleep;
-
- for (k = 0; k < s->num_active_planes; ++k) {
- if (display_cfg->plane_descriptors[k].stream_index == k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- mode_lib->mp.WritebackDelay[k] =
- mode_lib->soc.qos_parameters.writeback.base_latency_us
- + CalculateWriteBackDelay(
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.h_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.v_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total) / mode_lib->mp.Dispclk;
- } else
- mode_lib->mp.WritebackDelay[k] = 0;
-
- for (j = 0; j < s->num_active_planes; ++j) {
- if (display_cfg->plane_descriptors[j].stream_index == k
- && display_cfg->stream_descriptors[display_cfg->plane_descriptors[j].stream_index].writeback.enable == true) {
- mode_lib->mp.WritebackDelay[k] =
- math_max2(
- mode_lib->mp.WritebackDelay[k],
- mode_lib->soc.qos_parameters.writeback.base_latency_us
- + CalculateWriteBackDelay(
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[j].stream_index].writeback.pixel_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[j].stream_index].writeback.scaling_info.h_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[j].stream_index].writeback.scaling_info.v_ratio,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[j].stream_index].writeback.scaling_info.v_taps,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[j].stream_index].writeback.scaling_info.output_width,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[j].stream_index].writeback.scaling_info.output_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[j].stream_index].writeback.scaling_info.input_height,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total) / mode_lib->mp.Dispclk);
- }
- }
- }
- }
-
- for (k = 0; k < s->num_active_planes; ++k)
- for (j = 0; j < s->num_active_planes; ++j)
- if (display_cfg->plane_descriptors[k].stream_index == j)
- mode_lib->mp.WritebackDelay[k] = mode_lib->mp.WritebackDelay[j];
-
- mode_lib->mp.UrgentLatency = CalculateUrgentLatency(
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.base_latency_us,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.base_latency_pixel_vm_us,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.base_latency_vm_us,
- mode_lib->soc.do_urgent_latency_adjustment,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.scaling_factor_fclk_us,
- mode_lib->soc.qos_parameters.qos_params.dcn32x.urgent_latency_us.scaling_factor_mhz,
- mode_lib->mp.FabricClock,
- mode_lib->mp.uclk_freq_mhz,
- mode_lib->soc.qos_parameters.qos_type,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->mp.qos_param_index].urgent_ramp_uclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.df_qos_response_time_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.max_round_trip_to_furthest_cs_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.mall_overhead_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_urgent_ramp_latency_margin,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin);
-
- mode_lib->mp.TripToMemory = CalculateTripToMemory(
- mode_lib->mp.UrgentLatency,
- mode_lib->mp.FabricClock,
- mode_lib->mp.uclk_freq_mhz,
- mode_lib->soc.qos_parameters.qos_type,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->mp.qos_param_index].trip_to_memory_uclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.max_round_trip_to_furthest_cs_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.mall_overhead_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_max_latency_margin,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin);
-
- mode_lib->mp.TripToMemory = math_max2(mode_lib->mp.UrgentLatency, mode_lib->mp.TripToMemory);
-
- mode_lib->mp.MetaTripToMemory = CalculateMetaTripToMemory(
- mode_lib->mp.UrgentLatency,
- mode_lib->mp.FabricClock,
- mode_lib->mp.uclk_freq_mhz,
- mode_lib->soc.qos_parameters.qos_type,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->mp.qos_param_index].meta_trip_to_memory_uclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.meta_trip_adder_fclk_cycles,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_max_latency_margin,
- mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin);
-
- for (k = 0; k < s->num_active_planes; ++k) {
- calculate_cursor_req_attributes(
- display_cfg->plane_descriptors[k].cursor.cursor_width,
- display_cfg->plane_descriptors[k].cursor.cursor_bpp,
-
- // output
- &s->cursor_lines_per_chunk[k],
- &s->cursor_bytes_per_line[k],
- &s->cursor_bytes_per_chunk[k],
- &s->cursor_bytes[k]);
-
- bool cursor_not_enough_urgent_latency_hiding = 0;
- double line_time_us = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
-
- calculate_cursor_urgent_burst_factor(
- mode_lib->ip.cursor_buffer_size,
- display_cfg->plane_descriptors[k].cursor.cursor_width,
- s->cursor_bytes_per_chunk[k],
- s->cursor_lines_per_chunk[k],
- line_time_us,
- mode_lib->mp.UrgentLatency,
-
- // output
- &mode_lib->mp.UrgentBurstFactorCursor[k],
- &cursor_not_enough_urgent_latency_hiding);
- mode_lib->mp.UrgentBurstFactorCursorPre[k] = mode_lib->mp.UrgentBurstFactorCursor[k];
-
- CalculateUrgentBurstFactor(
- &display_cfg->plane_descriptors[k],
- mode_lib->mp.swath_width_luma_ub[k],
- mode_lib->mp.swath_width_chroma_ub[k],
- mode_lib->mp.SwathHeightY[k],
- mode_lib->mp.SwathHeightC[k],
- line_time_us,
- mode_lib->mp.UrgentLatency,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio,
- mode_lib->mp.BytePerPixelInDETY[k],
- mode_lib->mp.BytePerPixelInDETC[k],
- mode_lib->mp.DETBufferSizeY[k],
- mode_lib->mp.DETBufferSizeC[k],
-
- /* output */
- &mode_lib->mp.UrgentBurstFactorLuma[k],
- &mode_lib->mp.UrgentBurstFactorChroma[k],
- &mode_lib->mp.NotEnoughUrgentLatencyHiding[k]);
-
- mode_lib->mp.NotEnoughUrgentLatencyHiding[k] = mode_lib->mp.NotEnoughUrgentLatencyHiding[k] || cursor_not_enough_urgent_latency_hiding;
- }
-
- for (k = 0; k < s->num_active_planes; ++k) {
- s->MaxVStartupLines[k] = CalculateMaxVStartup(
- mode_lib->ip.ptoi_supported,
- mode_lib->ip.vblank_nom_default_us,
- &display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing,
- mode_lib->mp.WritebackDelay[k]);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
- dml2_printf("DML::%s: k=%u WritebackDelay = %f\n", __func__, k, mode_lib->mp.WritebackDelay[k]);
-#endif
- }
-
- s->immediate_flip_required = false;
- for (k = 0; k < s->num_active_planes; ++k) {
- s->immediate_flip_required = s->immediate_flip_required || display_cfg->plane_descriptors[k].immediate_flip;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: immediate_flip_required = %u\n", __func__, s->immediate_flip_required);
-#endif
-
- {
- s->DestinationLineTimesForPrefetchLessThan2 = false;
- s->VRatioPrefetchMoreThanMax = false;
-
- dml2_printf("DML::%s: Start one iteration of prefetch schedule evaluation\n", __func__);
-
- for (k = 0; k < s->num_active_planes; ++k) {
- dml2_printf("DML::%s: k=%d MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
-
- mode_lib->mp.TWait[k] = CalculateTWait(
- display_cfg->plane_descriptors[k].overrides.reserved_vblank_time_ns,
- mode_lib->mp.UrgentLatency,
- mode_lib->mp.TripToMemory);
-
- struct dml2_core_internal_DmlPipe *myPipe = &s->myPipe;
- myPipe->Dppclk = mode_lib->mp.Dppclk[k];
- myPipe->Dispclk = mode_lib->mp.Dispclk;
- myPipe->PixelClock = ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- myPipe->DCFClkDeepSleep = mode_lib->mp.DCFCLKDeepSleep;
- myPipe->DPPPerSurface = mode_lib->mp.NoOfDPP[k];
- myPipe->ScalerEnabled = display_cfg->plane_descriptors[k].composition.scaler_info.enabled;
- myPipe->VRatio = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio;
- myPipe->VRatioChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio;
- myPipe->VTaps = display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_taps;
- myPipe->VTapsChroma = display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_taps;
- myPipe->RotationAngle = display_cfg->plane_descriptors[k].composition.rotation_angle;
- myPipe->mirrored = display_cfg->plane_descriptors[k].composition.mirrored;
- myPipe->BlockWidth256BytesY = mode_lib->mp.Read256BlockWidthY[k];
- myPipe->BlockHeight256BytesY = mode_lib->mp.Read256BlockHeightY[k];
- myPipe->BlockWidth256BytesC = mode_lib->mp.Read256BlockWidthC[k];
- myPipe->BlockHeight256BytesC = mode_lib->mp.Read256BlockHeightC[k];
- myPipe->InterlaceEnable = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced;
- myPipe->NumberOfCursors = display_cfg->plane_descriptors[k].cursor.num_cursors;
- myPipe->VBlank = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active;
- myPipe->HTotal = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total;
- myPipe->HActive = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_active;
- myPipe->DCCEnable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- myPipe->ODMMode = mode_lib->mp.ODMMode[k];
- myPipe->SourcePixelFormat = display_cfg->plane_descriptors[k].pixel_format;
- myPipe->BytePerPixelY = mode_lib->mp.BytePerPixelY[k];
- myPipe->BytePerPixelC = mode_lib->mp.BytePerPixelC[k];
- myPipe->ProgressiveToInterlaceUnitInOPP = mode_lib->ip.ptoi_supported;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Calling CalculatePrefetchSchedule for k=%u\n", __func__, k);
-#endif
- CalculatePrefetchSchedule_params->display_cfg = display_cfg;
- CalculatePrefetchSchedule_params->HostVMInefficiencyFactor = s->HostVMInefficiencyFactorPrefetch;
- CalculatePrefetchSchedule_params->myPipe = myPipe;
- CalculatePrefetchSchedule_params->DSCDelay = mode_lib->mp.DSCDelay[k];
- CalculatePrefetchSchedule_params->DPPCLKDelaySubtotalPlusCNVCFormater = mode_lib->ip.dppclk_delay_subtotal + mode_lib->ip.dppclk_delay_cnvc_formatter;
- CalculatePrefetchSchedule_params->DPPCLKDelaySCL = mode_lib->ip.dppclk_delay_scl;
- CalculatePrefetchSchedule_params->DPPCLKDelaySCLLBOnly = mode_lib->ip.dppclk_delay_scl_lb_only;
- CalculatePrefetchSchedule_params->DPPCLKDelayCNVCCursor = mode_lib->ip.dppclk_delay_cnvc_cursor;
- CalculatePrefetchSchedule_params->DISPCLKDelaySubtotal = mode_lib->ip.dispclk_delay_subtotal;
- CalculatePrefetchSchedule_params->DPP_RECOUT_WIDTH = (unsigned int)(mode_lib->mp.SwathWidthY[k] / display_cfg->plane_descriptors[k].composition.scaler_info.plane0.h_ratio);
- CalculatePrefetchSchedule_params->OutputFormat = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].output.output_format;
- CalculatePrefetchSchedule_params->MaxInterDCNTileRepeaters = mode_lib->ip.max_inter_dcn_tile_repeaters;
- CalculatePrefetchSchedule_params->VStartup = s->MaxVStartupLines[k];
- CalculatePrefetchSchedule_params->MaxVStartup = s->MaxVStartupLines[k];
- CalculatePrefetchSchedule_params->HostVMMinPageSize = mode_lib->soc.hostvm_min_page_size_kbytes;
- CalculatePrefetchSchedule_params->DynamicMetadataEnable = display_cfg->plane_descriptors[k].dynamic_meta_data.enable;
- CalculatePrefetchSchedule_params->DynamicMetadataVMEnabled = mode_lib->ip.dynamic_metadata_vm_enabled;
- CalculatePrefetchSchedule_params->DynamicMetadataLinesBeforeActiveRequired = display_cfg->plane_descriptors[k].dynamic_meta_data.lines_before_active_required;
- CalculatePrefetchSchedule_params->DynamicMetadataTransmittedBytes = display_cfg->plane_descriptors[k].dynamic_meta_data.transmitted_bytes;
- CalculatePrefetchSchedule_params->UrgentLatency = mode_lib->mp.UrgentLatency;
- CalculatePrefetchSchedule_params->ExtraLatencyPrefetch = mode_lib->mp.ExtraLatencyPrefetch;
- CalculatePrefetchSchedule_params->TCalc = mode_lib->mp.TCalc;
- CalculatePrefetchSchedule_params->vm_bytes = mode_lib->mp.vm_bytes[k];
- CalculatePrefetchSchedule_params->PixelPTEBytesPerRow = mode_lib->mp.PixelPTEBytesPerRow[k];
- CalculatePrefetchSchedule_params->PrefetchSourceLinesY = mode_lib->mp.PrefetchSourceLinesY[k];
- CalculatePrefetchSchedule_params->VInitPreFillY = mode_lib->mp.VInitPreFillY[k];
- CalculatePrefetchSchedule_params->MaxNumSwathY = mode_lib->mp.MaxNumSwathY[k];
- CalculatePrefetchSchedule_params->PrefetchSourceLinesC = mode_lib->mp.PrefetchSourceLinesC[k];
- CalculatePrefetchSchedule_params->VInitPreFillC = mode_lib->mp.VInitPreFillC[k];
- CalculatePrefetchSchedule_params->MaxNumSwathC = mode_lib->mp.MaxNumSwathC[k];
- CalculatePrefetchSchedule_params->swath_width_luma_ub = mode_lib->mp.swath_width_luma_ub[k];
- CalculatePrefetchSchedule_params->swath_width_chroma_ub = mode_lib->mp.swath_width_chroma_ub[k];
- CalculatePrefetchSchedule_params->SwathHeightY = mode_lib->mp.SwathHeightY[k];
- CalculatePrefetchSchedule_params->SwathHeightC = mode_lib->mp.SwathHeightC[k];
- CalculatePrefetchSchedule_params->TWait = mode_lib->mp.TWait[k];
- CalculatePrefetchSchedule_params->Ttrip = mode_lib->mp.TripToMemory;
- CalculatePrefetchSchedule_params->setup_for_tdlut = display_cfg->plane_descriptors[k].tdlut.setup_for_tdlut;
- CalculatePrefetchSchedule_params->tdlut_pte_bytes_per_frame = s->tdlut_pte_bytes_per_frame[k];
- CalculatePrefetchSchedule_params->tdlut_bytes_per_frame = s->tdlut_bytes_per_frame[k];
- CalculatePrefetchSchedule_params->tdlut_opt_time = s->tdlut_opt_time[k];
- CalculatePrefetchSchedule_params->tdlut_drain_time = s->tdlut_drain_time[k];
- CalculatePrefetchSchedule_params->num_cursors = (display_cfg->plane_descriptors[k].cursor.cursor_width > 0);
- CalculatePrefetchSchedule_params->cursor_bytes_per_chunk = s->cursor_bytes_per_chunk[k];
- CalculatePrefetchSchedule_params->cursor_bytes_per_line = s->cursor_bytes_per_line[k];
- CalculatePrefetchSchedule_params->dcc_enable = display_cfg->plane_descriptors[k].surface.dcc.enable;
- CalculatePrefetchSchedule_params->mrq_present = mode_lib->ip.dcn_mrq_present;
- CalculatePrefetchSchedule_params->meta_row_bytes = mode_lib->mp.meta_row_bytes[k];
- CalculatePrefetchSchedule_params->mall_prefetch_sdp_overhead_factor = mode_lib->mp.mall_prefetch_sdp_overhead_factor[k];
-
- // output
- CalculatePrefetchSchedule_params->DSTXAfterScaler = &mode_lib->mp.DSTXAfterScaler[k];
- CalculatePrefetchSchedule_params->DSTYAfterScaler = &mode_lib->mp.DSTYAfterScaler[k];
- CalculatePrefetchSchedule_params->dst_y_prefetch = &mode_lib->mp.dst_y_prefetch[k];
- CalculatePrefetchSchedule_params->dst_y_per_vm_vblank = &mode_lib->mp.dst_y_per_vm_vblank[k];
- CalculatePrefetchSchedule_params->dst_y_per_row_vblank = &mode_lib->mp.dst_y_per_row_vblank[k];
- CalculatePrefetchSchedule_params->VRatioPrefetchY = &mode_lib->mp.VRatioPrefetchY[k];
- CalculatePrefetchSchedule_params->VRatioPrefetchC = &mode_lib->mp.VRatioPrefetchC[k];
- CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWLuma = &mode_lib->mp.RequiredPrefetchPixelDataBWLuma[k];
- CalculatePrefetchSchedule_params->RequiredPrefetchPixelDataBWChroma = &mode_lib->mp.RequiredPrefetchPixelDataBWChroma[k];
- CalculatePrefetchSchedule_params->NotEnoughTimeForDynamicMetadata = &mode_lib->mp.NotEnoughTimeForDynamicMetadata[k];
- CalculatePrefetchSchedule_params->Tno_bw = &mode_lib->mp.Tno_bw[k];
- CalculatePrefetchSchedule_params->Tno_bw_flip = &mode_lib->mp.Tno_bw_flip[k];
- CalculatePrefetchSchedule_params->prefetch_vmrow_bw = &mode_lib->mp.prefetch_vmrow_bw[k];
- CalculatePrefetchSchedule_params->Tdmdl_vm = &mode_lib->mp.Tdmdl_vm[k];
- CalculatePrefetchSchedule_params->Tdmdl = &mode_lib->mp.Tdmdl[k];
- CalculatePrefetchSchedule_params->TSetup = &mode_lib->mp.TSetup[k];
- CalculatePrefetchSchedule_params->Tvm_trips = &s->Tvm_trips[k];
- CalculatePrefetchSchedule_params->Tr0_trips = &s->Tr0_trips[k];
- CalculatePrefetchSchedule_params->Tvm_trips_flip = &s->Tvm_trips_flip[k];
- CalculatePrefetchSchedule_params->Tr0_trips_flip = &s->Tr0_trips_flip[k];
- CalculatePrefetchSchedule_params->Tvm_trips_flip_rounded = &s->Tvm_trips_flip_rounded[k];
- CalculatePrefetchSchedule_params->Tr0_trips_flip_rounded = &s->Tr0_trips_flip_rounded[k];
- CalculatePrefetchSchedule_params->VUpdateOffsetPix = &mode_lib->mp.VUpdateOffsetPix[k];
- CalculatePrefetchSchedule_params->VUpdateWidthPix = &mode_lib->mp.VUpdateWidthPix[k];
- CalculatePrefetchSchedule_params->VReadyOffsetPix = &mode_lib->mp.VReadyOffsetPix[k];
- CalculatePrefetchSchedule_params->prefetch_cursor_bw = &mode_lib->mp.prefetch_cursor_bw[k];
-
- mode_lib->mp.NoTimeToPrefetch[k] = CalculatePrefetchSchedule(&mode_lib->scratch, CalculatePrefetchSchedule_params);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%0u NoTimeToPrefetch=%0d\n", __func__, k, mode_lib->mp.NoTimeToPrefetch[k]);
-#endif
- mode_lib->mp.VStartupMin[k] = s->MaxVStartupLines[k];
- } // for k
-
- mode_lib->mp.PrefetchModeSupported = true;
- for (k = 0; k < s->num_active_planes; ++k) {
- if (mode_lib->mp.NoTimeToPrefetch[k] == true ||
- mode_lib->mp.NotEnoughTimeForDynamicMetadata[k] ||
- mode_lib->mp.DSTYAfterScaler[k] > 8) {
- dml2_printf("DML::%s: k=%u, NoTimeToPrefetch = %0d\n", __func__, k, mode_lib->mp.NoTimeToPrefetch[k]);
- dml2_printf("DML::%s: k=%u, NotEnoughTimeForDynamicMetadata=%u\n", __func__, k, mode_lib->mp.NotEnoughTimeForDynamicMetadata[k]);
- dml2_printf("DML::%s: k=%u, DSTYAfterScaler=%u (should be <= 0)\n", __func__, k, mode_lib->mp.DSTYAfterScaler[k]);
- mode_lib->mp.PrefetchModeSupported = false;
- }
- if (mode_lib->mp.dst_y_prefetch[k] < 2)
- s->DestinationLineTimesForPrefetchLessThan2 = true;
-
- if (mode_lib->mp.VRatioPrefetchY[k] > __DML2_CALCS_MAX_VRATIO_PRE_ENHANCE_PREFETCH_ACC__ ||
- mode_lib->mp.VRatioPrefetchC[k] > __DML2_CALCS_MAX_VRATIO_PRE_ENHANCE_PREFETCH_ACC__)
- s->VRatioPrefetchMoreThanMax = true;
-
- if (mode_lib->mp.NotEnoughUrgentLatencyHiding[k]) {
- dml2_printf("DML::%s: k=%u, NotEnoughUrgentLatencyHiding = %u\n", __func__, k, mode_lib->mp.NotEnoughUrgentLatencyHiding[k]);
- mode_lib->mp.PrefetchModeSupported = false;
- }
- }
-
- if (s->VRatioPrefetchMoreThanMax == true || s->DestinationLineTimesForPrefetchLessThan2 == true) {
- dml2_printf("DML::%s: VRatioPrefetchMoreThanMax = %u\n", __func__, s->VRatioPrefetchMoreThanMax);
- dml2_printf("DML::%s: DestinationLineTimesForPrefetchLessThan2 = %u\n", __func__, s->DestinationLineTimesForPrefetchLessThan2);
- mode_lib->mp.PrefetchModeSupported = false;
- }
-
- dml2_printf("DML::%s: Prefetch schedule is %sOK at vstartup = %u\n", __func__,
- mode_lib->mp.PrefetchModeSupported ? "" : "NOT ", CalculatePrefetchSchedule_params->VStartup);
-
- // Prefetch schedule OK, now check prefetch bw
- if (mode_lib->mp.PrefetchModeSupported == true) {
- for (k = 0; k < s->num_active_planes; ++k) {
- double line_time_us = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total /
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- CalculateUrgentBurstFactor(
- &display_cfg->plane_descriptors[k],
- mode_lib->mp.swath_width_luma_ub[k],
- mode_lib->mp.swath_width_chroma_ub[k],
- mode_lib->mp.SwathHeightY[k],
- mode_lib->mp.SwathHeightC[k],
- line_time_us,
- mode_lib->mp.UrgentLatency,
- mode_lib->mp.VRatioPrefetchY[k],
- mode_lib->mp.VRatioPrefetchC[k],
- mode_lib->mp.BytePerPixelInDETY[k],
- mode_lib->mp.BytePerPixelInDETC[k],
- mode_lib->mp.DETBufferSizeY[k],
- mode_lib->mp.DETBufferSizeC[k],
- /* Output */
- &mode_lib->mp.UrgentBurstFactorLumaPre[k],
- &mode_lib->mp.UrgentBurstFactorChromaPre[k],
- &mode_lib->mp.NotEnoughUrgentLatencyHidingPre[k]);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%0u DPPPerSurface=%u\n", __func__, k, mode_lib->mp.NoOfDPP[k]);
- dml2_printf("DML::%s: k=%0u UrgentBurstFactorLuma=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorLuma[k]);
- dml2_printf("DML::%s: k=%0u UrgentBurstFactorChroma=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorChroma[k]);
- dml2_printf("DML::%s: k=%0u UrgentBurstFactorLumaPre=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorLumaPre[k]);
- dml2_printf("DML::%s: k=%0u UrgentBurstFactorChromaPre=%f\n", __func__, k, mode_lib->mp.UrgentBurstFactorChromaPre[k]);
-
- dml2_printf("DML::%s: k=%0u VRatioPrefetchY=%f\n", __func__, k, mode_lib->mp.VRatioPrefetchY[k]);
- dml2_printf("DML::%s: k=%0u VRatioY=%f\n", __func__, k, display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio);
-
- dml2_printf("DML::%s: k=%0u prefetch_vmrow_bw=%f\n", __func__, k, mode_lib->mp.prefetch_vmrow_bw[k]);
- dml2_printf("DML::%s: k=%0u ReadBandwidthSurfaceLuma=%f\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%0u ReadBandwidthSurfaceChroma=%f\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthChroma[k]);
- dml2_printf("DML::%s: k=%0u cursor_bw=%f\n", __func__, k, mode_lib->mp.cursor_bw[k]);
- dml2_printf("DML::%s: k=%0u dpte_row_bw=%f\n", __func__, k, mode_lib->mp.dpte_row_bw[k]);
- dml2_printf("DML::%s: k=%0u meta_row_bw=%f\n", __func__, k, mode_lib->mp.meta_row_bw[k]);
- dml2_printf("DML::%s: k=%0u RequiredPrefetchPixelDataBWLuma=%f\n", __func__, k, mode_lib->mp.RequiredPrefetchPixelDataBWLuma[k]);
- dml2_printf("DML::%s: k=%0u RequiredPrefetchPixelDataBWChroma=%f\n", __func__, k, mode_lib->mp.RequiredPrefetchPixelDataBWChroma[k]);
- dml2_printf("DML::%s: k=%0u prefetch_cursor_bw=%f\n", __func__, k, mode_lib->mp.prefetch_cursor_bw[k]);
-#endif
- }
-
- for (k = 0; k <= s->num_active_planes - 1; k++)
- mode_lib->mp.final_flip_bw[k] = 0;
-
- calculate_peak_bandwidth_required(
- &mode_lib->scratch,
- mode_lib->mp.urg_vactive_bandwidth_required,
- mode_lib->mp.urg_bandwidth_required,
- mode_lib->mp.non_urg_bandwidth_required,
-
- // Input
- display_cfg,
- 0, // inc_flip_bw
- s->num_active_planes,
- mode_lib->mp.NoOfDPP,
- mode_lib->mp.dcc_dram_bw_nom_overhead_factor_p0,
- mode_lib->mp.dcc_dram_bw_nom_overhead_factor_p1,
- mode_lib->mp.dcc_dram_bw_pref_overhead_factor_p0,
- mode_lib->mp.dcc_dram_bw_pref_overhead_factor_p1,
- mode_lib->mp.mall_prefetch_sdp_overhead_factor,
- mode_lib->mp.mall_prefetch_dram_overhead_factor,
- mode_lib->mp.SurfaceReadBandwidthLuma,
- mode_lib->mp.SurfaceReadBandwidthChroma,
- mode_lib->mp.RequiredPrefetchPixelDataBWLuma,
- mode_lib->mp.RequiredPrefetchPixelDataBWChroma,
- mode_lib->mp.cursor_bw,
- mode_lib->mp.dpte_row_bw,
- mode_lib->mp.meta_row_bw,
- mode_lib->mp.prefetch_cursor_bw,
- mode_lib->mp.prefetch_vmrow_bw,
- mode_lib->mp.final_flip_bw,
- mode_lib->mp.UrgentBurstFactorLuma,
- mode_lib->mp.UrgentBurstFactorChroma,
- mode_lib->mp.UrgentBurstFactorCursor,
- mode_lib->mp.UrgentBurstFactorLumaPre,
- mode_lib->mp.UrgentBurstFactorChromaPre,
- mode_lib->mp.UrgentBurstFactorCursorPre);
-
- // Check urg peak bandwidth against available urg bw
- // check at SDP and DRAM, for all soc states (SVP prefetch an Sys Active)
- check_urgent_bandwidth_support(
- &mode_lib->mp.FractionOfUrgentBandwidth, // double* frac_urg_bandwidth
- &mode_lib->mp.FractionOfUrgentBandwidthMALL, // double* frac_urg_bandwidth_mall
- &s->dummy_boolean[1], // vactive bw ok
- &mode_lib->mp.PrefetchModeSupported, // prefetch bw ok
-
- mode_lib->soc.mall_allocated_for_dcn_mbytes,
- mode_lib->mp.non_urg_bandwidth_required,
- mode_lib->mp.urg_vactive_bandwidth_required,
- mode_lib->mp.urg_bandwidth_required,
- mode_lib->mp.urg_bandwidth_available);
-
- for (k = 0; k < s->num_active_planes; ++k) {
- if (mode_lib->mp.NotEnoughUrgentLatencyHidingPre[k]) {
- dml2_printf("DML::%s: k=%u, NotEnoughUrgentLatencyHidingPre = %u\n", __func__, k, mode_lib->mp.NotEnoughUrgentLatencyHidingPre[k]);
- mode_lib->mp.PrefetchModeSupported = false;
- }
- }
- } // prefetch schedule ok
-
- // Prefetch schedule and prefetch bw ok, now check flip bw
- if (mode_lib->mp.PrefetchModeSupported == true) { // prefetch schedule and prefetch bw ok, now check flip bw
-
- mode_lib->mp.BandwidthAvailableForImmediateFlip =
- get_bandwidth_available_for_immediate_flip(dml2_core_internal_soc_state_sys_active,
- mode_lib->mp.urg_bandwidth_required, // no flip
- mode_lib->mp.urg_bandwidth_available);
- mode_lib->mp.TotImmediateFlipBytes = 0;
- for (k = 0; k < s->num_active_planes; ++k) {
- if (display_cfg->plane_descriptors[k].immediate_flip) {
- s->per_pipe_flip_bytes[k] = get_pipe_flip_bytes(
- s->HostVMInefficiencyFactor,
- mode_lib->mp.vm_bytes[k],
- mode_lib->mp.PixelPTEBytesPerRow[k],
- mode_lib->mp.meta_row_bytes[k]);
- } else {
- s->per_pipe_flip_bytes[k] = 0;
- }
- mode_lib->mp.TotImmediateFlipBytes += s->per_pipe_flip_bytes[k] * mode_lib->mp.NoOfDPP[k];
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k = %u\n", __func__, k);
- dml2_printf("DML::%s: DPPPerSurface = %u\n", __func__, mode_lib->mp.NoOfDPP[k]);
- dml2_printf("DML::%s: vm_bytes = %u\n", __func__, mode_lib->mp.vm_bytes[k]);
- dml2_printf("DML::%s: PixelPTEBytesPerRow = %u\n", __func__, mode_lib->mp.PixelPTEBytesPerRow[k]);
- dml2_printf("DML::%s: meta_row_bytes = %u\n", __func__, mode_lib->mp.meta_row_bytes[k]);
- dml2_printf("DML::%s: TotImmediateFlipBytes = %u\n", __func__, mode_lib->mp.TotImmediateFlipBytes);
-#endif
- }
- for (k = 0; k < s->num_active_planes; ++k) {
- CalculateFlipSchedule(
- &mode_lib->scratch,
- display_cfg->plane_descriptors[k].immediate_flip,
- 0, // use_lb_flip_bw
- s->HostVMInefficiencyFactor,
- s->Tvm_trips_flip[k],
- s->Tr0_trips_flip[k],
- s->Tvm_trips_flip_rounded[k],
- s->Tr0_trips_flip_rounded[k],
- display_cfg->gpuvm_enable,
- mode_lib->mp.vm_bytes[k],
- mode_lib->mp.PixelPTEBytesPerRow[k],
- mode_lib->mp.BandwidthAvailableForImmediateFlip,
- mode_lib->mp.TotImmediateFlipBytes,
- display_cfg->plane_descriptors[k].pixel_format,
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000),
- display_cfg->plane_descriptors[k].composition.scaler_info.plane0.v_ratio,
- display_cfg->plane_descriptors[k].composition.scaler_info.plane1.v_ratio,
- mode_lib->mp.Tno_bw[k],
- mode_lib->mp.dpte_row_height[k],
- mode_lib->mp.dpte_row_height_chroma[k],
- mode_lib->mp.use_one_row_for_frame_flip[k],
- mode_lib->ip.max_flip_time_us,
- s->per_pipe_flip_bytes[k],
- mode_lib->mp.meta_row_bytes[k],
- mode_lib->mp.meta_row_height[k],
- mode_lib->mp.meta_row_height_chroma[k],
- mode_lib->ip.dcn_mrq_present && display_cfg->plane_descriptors[k].surface.dcc.enable,
-
- // Output
- &mode_lib->mp.dst_y_per_vm_flip[k],
- &mode_lib->mp.dst_y_per_row_flip[k],
- &mode_lib->mp.final_flip_bw[k],
- &mode_lib->mp.ImmediateFlipSupportedForPipe[k]);
- }
-
- calculate_peak_bandwidth_required(
- &mode_lib->scratch,
- s->dummy_bw,
- mode_lib->mp.urg_bandwidth_required_flip,
- mode_lib->mp.non_urg_bandwidth_required_flip,
-
- // Input
- display_cfg,
- 1, // inc_flip_bw
- s->num_active_planes,
- mode_lib->mp.NoOfDPP,
- mode_lib->mp.dcc_dram_bw_nom_overhead_factor_p0,
- mode_lib->mp.dcc_dram_bw_nom_overhead_factor_p1,
- mode_lib->mp.dcc_dram_bw_pref_overhead_factor_p0,
- mode_lib->mp.dcc_dram_bw_pref_overhead_factor_p1,
- mode_lib->mp.mall_prefetch_sdp_overhead_factor,
- mode_lib->mp.mall_prefetch_dram_overhead_factor,
- mode_lib->mp.SurfaceReadBandwidthLuma,
- mode_lib->mp.SurfaceReadBandwidthChroma,
- mode_lib->mp.RequiredPrefetchPixelDataBWLuma,
- mode_lib->mp.RequiredPrefetchPixelDataBWChroma,
- mode_lib->mp.cursor_bw,
- mode_lib->mp.dpte_row_bw,
- mode_lib->mp.meta_row_bw,
- mode_lib->mp.prefetch_cursor_bw,
- mode_lib->mp.prefetch_vmrow_bw,
- mode_lib->mp.final_flip_bw,
- mode_lib->mp.UrgentBurstFactorLuma,
- mode_lib->mp.UrgentBurstFactorChroma,
- mode_lib->mp.UrgentBurstFactorCursor,
- mode_lib->mp.UrgentBurstFactorLumaPre,
- mode_lib->mp.UrgentBurstFactorChromaPre,
- mode_lib->mp.UrgentBurstFactorCursorPre);
-
- calculate_immediate_flip_bandwidth_support(
- &mode_lib->mp.FractionOfUrgentBandwidthImmediateFlip, // double* frac_urg_bandwidth_flip
- &mode_lib->mp.ImmediateFlipSupported, // bool* flip_bandwidth_support_ok
-
- dml2_core_internal_soc_state_sys_active,
- mode_lib->mp.urg_bandwidth_required_flip,
- mode_lib->mp.non_urg_bandwidth_required_flip,
- mode_lib->mp.urg_bandwidth_available);
-
- for (k = 0; k < s->num_active_planes; ++k) {
- if (display_cfg->plane_descriptors[k].immediate_flip && mode_lib->mp.ImmediateFlipSupportedForPipe[k] == false) {
- mode_lib->mp.ImmediateFlipSupported = false;
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Pipe %0d not supporting iflip!\n", __func__, k);
-#endif
- }
- }
- } else { // flip or prefetch not support
- mode_lib->mp.ImmediateFlipSupported = false;
- }
-
- // consider flip support is okay if the flip bw is ok or (when user does't require a iflip and there is no host vm)
- bool must_support_iflip = display_cfg->hostvm_enable || s->immediate_flip_required;
- mode_lib->mp.PrefetchAndImmediateFlipSupported = (mode_lib->mp.PrefetchModeSupported == true && (!must_support_iflip || mode_lib->mp.ImmediateFlipSupported));
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: PrefetchModeSupported = %u\n", __func__, mode_lib->mp.PrefetchModeSupported);
- for (k = 0; k < s->num_active_planes; ++k)
- dml2_printf("DML::%s: immediate_flip_required[%u] = %u\n", __func__, k, display_cfg->plane_descriptors[k].immediate_flip);
- dml2_printf("DML::%s: HostVMEnable = %u\n", __func__, display_cfg->hostvm_enable);
- dml2_printf("DML::%s: ImmediateFlipSupported = %u\n", __func__, mode_lib->mp.ImmediateFlipSupported);
- dml2_printf("DML::%s: PrefetchAndImmediateFlipSupported = %u\n", __func__, mode_lib->mp.PrefetchAndImmediateFlipSupported);
-#endif
- dml2_printf("DML::%s: Done one iteration: k=%d, MaxVStartupLines=%u\n", __func__, k, s->MaxVStartupLines[k]);
- }
-
- for (k = 0; k < s->num_active_planes; ++k)
- dml2_printf("DML::%s: k=%d MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
-
- if (!mode_lib->mp.PrefetchAndImmediateFlipSupported) {
- dml2_printf("DML::%s: Bad, Prefetch and flip scheduling solution NOT found!\n", __func__);
- } else {
- dml2_printf("DML::%s: Good, Prefetch and flip scheduling solution found\n", __func__);
-
- // DCC Configuration
- for (k = 0; k < s->num_active_planes; ++k) {
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: Calculate DCC configuration for surface k=%u\n", __func__, k);
-#endif
- CalculateDCCConfiguration(
- display_cfg->plane_descriptors[k].surface.dcc.enable,
- display_cfg->overrides.dcc_programming_assumes_scan_direction_unknown,
- display_cfg->plane_descriptors[k].pixel_format,
- display_cfg->plane_descriptors[k].surface.plane0.width,
- display_cfg->plane_descriptors[k].surface.plane1.width,
- display_cfg->plane_descriptors[k].surface.plane0.height,
- display_cfg->plane_descriptors[k].surface.plane1.height,
- s->NomDETInKByte,
- mode_lib->mp.Read256BlockHeightY[k],
- mode_lib->mp.Read256BlockHeightC[k],
- display_cfg->plane_descriptors[k].surface.tiling,
- mode_lib->mp.BytePerPixelY[k],
- mode_lib->mp.BytePerPixelC[k],
- mode_lib->mp.BytePerPixelInDETY[k],
- mode_lib->mp.BytePerPixelInDETC[k],
- display_cfg->plane_descriptors[k].composition.rotation_angle,
-
- /* Output */
- &mode_lib->mp.RequestLuma[k],
- &mode_lib->mp.RequestChroma[k],
- &mode_lib->mp.DCCYMaxUncompressedBlock[k],
- &mode_lib->mp.DCCCMaxUncompressedBlock[k],
- &mode_lib->mp.DCCYMaxCompressedBlock[k],
- &mode_lib->mp.DCCCMaxCompressedBlock[k],
- &mode_lib->mp.DCCYIndependentBlock[k],
- &mode_lib->mp.DCCCIndependentBlock[k]);
- }
-
- //Watermarks and NB P-State/DRAM Clock Change Support
- s->mmSOCParameters.UrgentLatency = mode_lib->mp.UrgentLatency;
- s->mmSOCParameters.ExtraLatency = mode_lib->mp.ExtraLatency;
- s->mmSOCParameters.ExtraLatency_sr = mode_lib->mp.ExtraLatency_sr;
- s->mmSOCParameters.WritebackLatency = mode_lib->soc.qos_parameters.writeback.base_latency_us;
- s->mmSOCParameters.DRAMClockChangeLatency = mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us;
- s->mmSOCParameters.FCLKChangeLatency = mode_lib->soc.power_management_parameters.fclk_change_blackout_us;
- s->mmSOCParameters.SRExitTime = mode_lib->soc.power_management_parameters.stutter_exit_latency_us;
- s->mmSOCParameters.SREnterPlusExitTime = mode_lib->soc.power_management_parameters.stutter_enter_plus_exit_latency_us;
- s->mmSOCParameters.SRExitZ8Time = mode_lib->soc.power_management_parameters.z8_stutter_exit_latency_us;
- s->mmSOCParameters.SREnterPlusExitZ8Time = mode_lib->soc.power_management_parameters.z8_stutter_enter_plus_exit_latency_us;
- s->mmSOCParameters.USRRetrainingLatency = 0; //0; //FIXME_STAGE2
- s->mmSOCParameters.SMNLatency = 0; //mode_lib->soc.smn_latency_us; //FIXME_STAGE2
-
- CalculateWatermarks_params->display_cfg = display_cfg;
- CalculateWatermarks_params->USRRetrainingRequired = false/*FIXME_STAGE2 was: mode_lib->ms.policy.USRRetrainingRequired, no new dml2 replacement*/;
- CalculateWatermarks_params->NumberOfActiveSurfaces = s->num_active_planes;
- CalculateWatermarks_params->MaxLineBufferLines = mode_lib->ip.max_line_buffer_lines;
- CalculateWatermarks_params->LineBufferSize = mode_lib->ip.line_buffer_size_bits;
- CalculateWatermarks_params->WritebackInterfaceBufferSize = mode_lib->ip.writeback_interface_buffer_size_kbytes;
- CalculateWatermarks_params->DCFCLK = mode_lib->mp.Dcfclk;
- CalculateWatermarks_params->SynchronizeTimings = display_cfg->overrides.synchronize_timings;
- CalculateWatermarks_params->SynchronizeDRRDisplaysForUCLKPStateChange = display_cfg->overrides.synchronize_ddr_displays_for_uclk_pstate_change;
- CalculateWatermarks_params->dpte_group_bytes = mode_lib->mp.dpte_group_bytes;
- CalculateWatermarks_params->mmSOCParameters = s->mmSOCParameters;
- CalculateWatermarks_params->WritebackChunkSize = mode_lib->ip.writeback_chunk_size_kbytes;
- CalculateWatermarks_params->SOCCLK = s->SOCCLK;
- CalculateWatermarks_params->DCFClkDeepSleep = mode_lib->mp.DCFCLKDeepSleep;
- CalculateWatermarks_params->DETBufferSizeY = mode_lib->mp.DETBufferSizeY;
- CalculateWatermarks_params->DETBufferSizeC = mode_lib->mp.DETBufferSizeC;
- CalculateWatermarks_params->SwathHeightY = mode_lib->mp.SwathHeightY;
- CalculateWatermarks_params->SwathHeightC = mode_lib->mp.SwathHeightC;
- //CalculateWatermarks_params->LBBitPerPixel = 57; //FIXME_STAGE2
- CalculateWatermarks_params->SwathWidthY = mode_lib->mp.SwathWidthY;
- CalculateWatermarks_params->SwathWidthC = mode_lib->mp.SwathWidthC;
- CalculateWatermarks_params->BytePerPixelDETY = mode_lib->mp.BytePerPixelInDETY;
- CalculateWatermarks_params->BytePerPixelDETC = mode_lib->mp.BytePerPixelInDETC;
- CalculateWatermarks_params->DSTXAfterScaler = mode_lib->mp.DSTXAfterScaler;
- CalculateWatermarks_params->DSTYAfterScaler = mode_lib->mp.DSTYAfterScaler;
- CalculateWatermarks_params->UnboundedRequestEnabled = mode_lib->mp.UnboundedRequestEnabled;
- CalculateWatermarks_params->CompressedBufferSizeInkByte = mode_lib->mp.CompressedBufferSizeInkByte;
- CalculateWatermarks_params->meta_row_height_l = mode_lib->mp.meta_row_height;
- CalculateWatermarks_params->meta_row_height_c = mode_lib->mp.meta_row_height_chroma;
-
- // Output
- CalculateWatermarks_params->Watermark = &mode_lib->mp.Watermark;
- CalculateWatermarks_params->DRAMClockChangeSupport = mode_lib->mp.DRAMClockChangeSupport;
- CalculateWatermarks_params->global_dram_clock_change_supported = &mode_lib->mp.global_dram_clock_change_supported;
- CalculateWatermarks_params->MaxActiveDRAMClockChangeLatencySupported = mode_lib->mp.MaxActiveDRAMClockChangeLatencySupported;
- CalculateWatermarks_params->SubViewportLinesNeededInMALL = mode_lib->mp.SubViewportLinesNeededInMALL;
- CalculateWatermarks_params->FCLKChangeSupport = mode_lib->mp.FCLKChangeSupport;
- CalculateWatermarks_params->global_fclk_change_supported = &mode_lib->mp.global_fclk_change_supported;
- CalculateWatermarks_params->MaxActiveFCLKChangeLatencySupported = &mode_lib->mp.MaxActiveFCLKChangeLatencySupported;
- CalculateWatermarks_params->USRRetrainingSupport = &mode_lib->mp.USRRetrainingSupport;
- CalculateWatermarks_params->VActiveLatencyHidingMargin = 0;
-
- CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport(&mode_lib->scratch, CalculateWatermarks_params);
-
- for (k = 0; k < s->num_active_planes; ++k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- mode_lib->mp.WritebackAllowDRAMClockChangeEndPosition[k] = math_max2(0, mode_lib->mp.VStartupMin[k] * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total /
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) - mode_lib->mp.Watermark.WritebackDRAMClockChangeWatermark);
- mode_lib->mp.WritebackAllowFCLKChangeEndPosition[k] = math_max2(0, mode_lib->mp.VStartupMin[k] * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total /
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000) - mode_lib->mp.Watermark.WritebackFCLKChangeWatermark);
- } else {
- mode_lib->mp.WritebackAllowDRAMClockChangeEndPosition[k] = 0;
- mode_lib->mp.WritebackAllowFCLKChangeEndPosition[k] = 0;
- }
- }
-
- dml2_printf("DML::%s: DEBUG stream_index = %0d\n", __func__, display_cfg->plane_descriptors[0].stream_index);
- dml2_printf("DML::%s: DEBUG PixelClock = %d kHz\n", __func__, (display_cfg->stream_descriptors[display_cfg->plane_descriptors[0].stream_index].timing.pixel_clock_khz));
-
- //Display Pipeline Delivery Time in Prefetch, Groups
- CalculatePixelDeliveryTimes(
- display_cfg,
- cfg_support_info,
- s->num_active_planes,
- mode_lib->mp.VRatioPrefetchY,
- mode_lib->mp.VRatioPrefetchC,
- mode_lib->mp.swath_width_luma_ub,
- mode_lib->mp.swath_width_chroma_ub,
- mode_lib->mp.PSCL_THROUGHPUT,
- mode_lib->mp.PSCL_THROUGHPUT_CHROMA,
- mode_lib->mp.Dppclk,
- mode_lib->mp.BytePerPixelC,
- mode_lib->mp.req_per_swath_ub_l,
- mode_lib->mp.req_per_swath_ub_c,
-
- /* Output */
- mode_lib->mp.DisplayPipeLineDeliveryTimeLuma,
- mode_lib->mp.DisplayPipeLineDeliveryTimeChroma,
- mode_lib->mp.DisplayPipeLineDeliveryTimeLumaPrefetch,
- mode_lib->mp.DisplayPipeLineDeliveryTimeChromaPrefetch,
- mode_lib->mp.DisplayPipeRequestDeliveryTimeLuma,
- mode_lib->mp.DisplayPipeRequestDeliveryTimeChroma,
- mode_lib->mp.DisplayPipeRequestDeliveryTimeLumaPrefetch,
- mode_lib->mp.DisplayPipeRequestDeliveryTimeChromaPrefetch);
-
- CalculateMetaAndPTETimes_params->scratch = &mode_lib->scratch;
- CalculateMetaAndPTETimes_params->display_cfg = display_cfg;
- CalculateMetaAndPTETimes_params->NumberOfActiveSurfaces = s->num_active_planes;
- CalculateMetaAndPTETimes_params->use_one_row_for_frame = mode_lib->mp.use_one_row_for_frame;
- CalculateMetaAndPTETimes_params->dst_y_per_row_vblank = mode_lib->mp.dst_y_per_row_vblank;
- CalculateMetaAndPTETimes_params->dst_y_per_row_flip = mode_lib->mp.dst_y_per_row_flip;
- CalculateMetaAndPTETimes_params->BytePerPixelY = mode_lib->mp.BytePerPixelY;
- CalculateMetaAndPTETimes_params->BytePerPixelC = mode_lib->mp.BytePerPixelC;
- CalculateMetaAndPTETimes_params->dpte_row_height = mode_lib->mp.dpte_row_height;
- CalculateMetaAndPTETimes_params->dpte_row_height_chroma = mode_lib->mp.dpte_row_height_chroma;
- CalculateMetaAndPTETimes_params->dpte_group_bytes = mode_lib->mp.dpte_group_bytes;
- CalculateMetaAndPTETimes_params->PTERequestSizeY = mode_lib->mp.PTERequestSizeY;
- CalculateMetaAndPTETimes_params->PTERequestSizeC = mode_lib->mp.PTERequestSizeC;
- CalculateMetaAndPTETimes_params->PixelPTEReqWidthY = mode_lib->mp.PixelPTEReqWidthY;
- CalculateMetaAndPTETimes_params->PixelPTEReqHeightY = mode_lib->mp.PixelPTEReqHeightY;
- CalculateMetaAndPTETimes_params->PixelPTEReqWidthC = mode_lib->mp.PixelPTEReqWidthC;
- CalculateMetaAndPTETimes_params->PixelPTEReqHeightC = mode_lib->mp.PixelPTEReqHeightC;
- CalculateMetaAndPTETimes_params->dpte_row_width_luma_ub = mode_lib->mp.dpte_row_width_luma_ub;
- CalculateMetaAndPTETimes_params->dpte_row_width_chroma_ub = mode_lib->mp.dpte_row_width_chroma_ub;
- CalculateMetaAndPTETimes_params->tdlut_groups_per_2row_ub = s->tdlut_groups_per_2row_ub;
- CalculateMetaAndPTETimes_params->mrq_present = mode_lib->ip.dcn_mrq_present;
-
- CalculateMetaAndPTETimes_params->MetaChunkSize = mode_lib->ip.meta_chunk_size_kbytes;
- CalculateMetaAndPTETimes_params->MinMetaChunkSizeBytes = mode_lib->ip.min_meta_chunk_size_bytes;
- CalculateMetaAndPTETimes_params->meta_row_width = mode_lib->mp.meta_row_width;
- CalculateMetaAndPTETimes_params->meta_row_width_chroma = mode_lib->mp.meta_row_width_chroma;
- CalculateMetaAndPTETimes_params->meta_row_height = mode_lib->mp.meta_row_height;
- CalculateMetaAndPTETimes_params->meta_row_height_chroma = mode_lib->mp.meta_row_height_chroma;
- CalculateMetaAndPTETimes_params->meta_req_width = mode_lib->mp.meta_req_width;
- CalculateMetaAndPTETimes_params->meta_req_width_chroma = mode_lib->mp.meta_req_width_chroma;
- CalculateMetaAndPTETimes_params->meta_req_height = mode_lib->mp.meta_req_height;
- CalculateMetaAndPTETimes_params->meta_req_height_chroma = mode_lib->mp.meta_req_height_chroma;
-
- CalculateMetaAndPTETimes_params->time_per_tdlut_group = mode_lib->mp.time_per_tdlut_group;
- CalculateMetaAndPTETimes_params->DST_Y_PER_PTE_ROW_NOM_L = mode_lib->mp.DST_Y_PER_PTE_ROW_NOM_L;
- CalculateMetaAndPTETimes_params->DST_Y_PER_PTE_ROW_NOM_C = mode_lib->mp.DST_Y_PER_PTE_ROW_NOM_C;
- CalculateMetaAndPTETimes_params->time_per_pte_group_nom_luma = mode_lib->mp.time_per_pte_group_nom_luma;
- CalculateMetaAndPTETimes_params->time_per_pte_group_vblank_luma = mode_lib->mp.time_per_pte_group_vblank_luma;
- CalculateMetaAndPTETimes_params->time_per_pte_group_flip_luma = mode_lib->mp.time_per_pte_group_flip_luma;
- CalculateMetaAndPTETimes_params->time_per_pte_group_nom_chroma = mode_lib->mp.time_per_pte_group_nom_chroma;
- CalculateMetaAndPTETimes_params->time_per_pte_group_vblank_chroma = mode_lib->mp.time_per_pte_group_vblank_chroma;
- CalculateMetaAndPTETimes_params->time_per_pte_group_flip_chroma = mode_lib->mp.time_per_pte_group_flip_chroma;
- CalculateMetaAndPTETimes_params->DST_Y_PER_META_ROW_NOM_L = mode_lib->mp.DST_Y_PER_META_ROW_NOM_L;
- CalculateMetaAndPTETimes_params->DST_Y_PER_META_ROW_NOM_C = mode_lib->mp.DST_Y_PER_META_ROW_NOM_C;
- CalculateMetaAndPTETimes_params->TimePerMetaChunkNominal = mode_lib->mp.TimePerMetaChunkNominal;
- CalculateMetaAndPTETimes_params->TimePerChromaMetaChunkNominal = mode_lib->mp.TimePerChromaMetaChunkNominal;
- CalculateMetaAndPTETimes_params->TimePerMetaChunkVBlank = mode_lib->mp.TimePerMetaChunkVBlank;
- CalculateMetaAndPTETimes_params->TimePerChromaMetaChunkVBlank = mode_lib->mp.TimePerChromaMetaChunkVBlank;
- CalculateMetaAndPTETimes_params->TimePerMetaChunkFlip = mode_lib->mp.TimePerMetaChunkFlip;
- CalculateMetaAndPTETimes_params->TimePerChromaMetaChunkFlip = mode_lib->mp.TimePerChromaMetaChunkFlip;
-
- CalculateMetaAndPTETimes(CalculateMetaAndPTETimes_params);
-
- CalculateVMGroupAndRequestTimes(
- display_cfg,
- s->num_active_planes,
- mode_lib->mp.BytePerPixelC,
- mode_lib->mp.dst_y_per_vm_vblank,
- mode_lib->mp.dst_y_per_vm_flip,
- mode_lib->mp.dpte_row_width_luma_ub,
- mode_lib->mp.dpte_row_width_chroma_ub,
- mode_lib->mp.vm_group_bytes,
- mode_lib->mp.dpde0_bytes_per_frame_ub_l,
- mode_lib->mp.dpde0_bytes_per_frame_ub_c,
- s->tdlut_pte_bytes_per_frame,
- mode_lib->mp.meta_pte_bytes_per_frame_ub_l,
- mode_lib->mp.meta_pte_bytes_per_frame_ub_c,
- mode_lib->ip.dcn_mrq_present,
-
- /* Output */
- mode_lib->mp.TimePerVMGroupVBlank,
- mode_lib->mp.TimePerVMGroupFlip,
- mode_lib->mp.TimePerVMRequestVBlank,
- mode_lib->mp.TimePerVMRequestFlip);
-
- // VStartup Adjustment
- for (k = 0; k < s->num_active_planes; ++k) {
-
- mode_lib->mp.MinTTUVBlank[k] = mode_lib->mp.TWait[k] + mode_lib->mp.ExtraLatency;
- if (!display_cfg->plane_descriptors[k].dynamic_meta_data.enable)
- mode_lib->mp.MinTTUVBlank[k] = mode_lib->mp.TCalc + mode_lib->mp.MinTTUVBlank[k];
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, MinTTUVBlank = %f (before vstartup margin)\n", __func__, k, mode_lib->mp.MinTTUVBlank[k]);
-#endif
- s->Tvstartup_margin = (s->MaxVStartupLines[k] - mode_lib->mp.VStartupMin[k]) * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000);
- mode_lib->mp.MinTTUVBlank[k] = mode_lib->mp.MinTTUVBlank[k] + s->Tvstartup_margin;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, Tvstartup_margin = %f\n", __func__, k, s->Tvstartup_margin);
- dml2_printf("DML::%s: k=%u, MaxVStartupLines = %u\n", __func__, k, s->MaxVStartupLines[k]);
- dml2_printf("DML::%s: k=%u, MinTTUVBlank = %f\n", __func__, k, mode_lib->mp.MinTTUVBlank[k]);
-#endif
-
- mode_lib->mp.Tdmdl[k] = mode_lib->mp.Tdmdl[k] + s->Tvstartup_margin;
- if (display_cfg->plane_descriptors[k].dynamic_meta_data.enable && mode_lib->ip.dynamic_metadata_vm_enabled) {
- mode_lib->mp.Tdmdl_vm[k] = mode_lib->mp.Tdmdl_vm[k] + s->Tvstartup_margin;
- }
-
- bool isInterlaceTiming = (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.interlaced && !mode_lib->ip.ptoi_supported);
-
- // The actual positioning of the vstartup
- mode_lib->mp.VStartup[k] = (isInterlaceTiming ? (2 * s->MaxVStartupLines[k]) : s->MaxVStartupLines[k]);
-
- s->dlg_vblank_start = ((isInterlaceTiming ? math_floor2((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_front_porch) / 2.0, 1.0) :
- display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total) - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_front_porch);
- s->LSetup = math_floor2(4.0 * mode_lib->mp.TSetup[k] / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)), 1.0) / 4.0;
- s->blank_lines_remaining = (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active) - mode_lib->mp.VStartup[k];
-
- if (s->blank_lines_remaining < 0) {
- dml2_printf("ERROR: Vstartup is larger than vblank!?\n");
- s->blank_lines_remaining = 0;
- DML2_ASSERT(0);
- }
- mode_lib->mp.MIN_DST_Y_NEXT_START[k] = s->dlg_vblank_start + s->blank_lines_remaining + s->LSetup;
-
- // debug only
- if (((mode_lib->mp.VUpdateOffsetPix[k] + mode_lib->mp.VUpdateWidthPix[k] + (double) mode_lib->mp.VReadyOffsetPix[k]) / display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total) <=
- (isInterlaceTiming ?
- math_floor2((display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_front_porch - mode_lib->mp.VStartup[k]) / 2.0, 1.0) :
- (int)(display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active - display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_front_porch - mode_lib->mp.VStartup[k]))) {
- mode_lib->mp.VREADY_AT_OR_AFTER_VSYNC[k] = true;
- } else {
- mode_lib->mp.VREADY_AT_OR_AFTER_VSYNC[k] = false;
- }
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, VStartup = %u (max)\n", __func__, k, mode_lib->mp.VStartup[k]);
- dml2_printf("DML::%s: k=%u, VStartupMin = %u (max)\n", __func__, k, mode_lib->mp.VStartupMin[k]);
- dml2_printf("DML::%s: k=%u, VUpdateOffsetPix = %u\n", __func__, k, mode_lib->mp.VUpdateOffsetPix[k]);
- dml2_printf("DML::%s: k=%u, VUpdateWidthPix = %u\n", __func__, k, mode_lib->mp.VUpdateWidthPix[k]);
- dml2_printf("DML::%s: k=%u, VReadyOffsetPix = %u\n", __func__, k, mode_lib->mp.VReadyOffsetPix[k]);
- dml2_printf("DML::%s: k=%u, HTotal = %u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total);
- dml2_printf("DML::%s: k=%u, VTotal = %u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_total);
- dml2_printf("DML::%s: k=%u, VActive = %u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_active);
- dml2_printf("DML::%s: k=%u, VFrontPorch = %u\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.v_front_porch);
- dml2_printf("DML::%s: k=%u, TSetup = %f\n", __func__, k, mode_lib->mp.TSetup[k]);
- dml2_printf("DML::%s: k=%u, MIN_DST_Y_NEXT_START = %f\n", __func__, k, mode_lib->mp.MIN_DST_Y_NEXT_START[k]);
- dml2_printf("DML::%s: k=%u, VREADY_AT_OR_AFTER_VSYNC = %u\n", __func__, k, mode_lib->mp.VREADY_AT_OR_AFTER_VSYNC[k]);
-#endif
- }
-
- //Maximum Bandwidth Used
- s->TotalWRBandwidth = 0;
- s->WRBandwidth = 0;
- for (k = 0; k < s->num_active_planes; ++k) {
- if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true && display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.pixel_format == dml2_444_32) {
- s->WRBandwidth = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width /
- (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * 4;
- } else if (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.enable == true) {
- s->WRBandwidth = display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_height * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.output_width /
- (display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.h_total * display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].writeback.scaling_info.input_height / ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.pixel_clock_khz / 1000)) * 8;
- }
- s->TotalWRBandwidth = s->TotalWRBandwidth + s->WRBandwidth;
- }
-
- mode_lib->mp.TotalDataReadBandwidth = 0;
- for (k = 0; k < s->num_active_planes; ++k) {
- mode_lib->mp.TotalDataReadBandwidth = mode_lib->mp.TotalDataReadBandwidth + mode_lib->mp.SurfaceReadBandwidthLuma[k] + mode_lib->mp.SurfaceReadBandwidthChroma[k];
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%u, TotalDataReadBandwidth = %f\n", __func__, k, mode_lib->mp.TotalDataReadBandwidth);
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceLuma = %f\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthLuma[k]);
- dml2_printf("DML::%s: k=%u, ReadBandwidthSurfaceChroma = %f\n", __func__, k, mode_lib->mp.SurfaceReadBandwidthChroma[k]);
-#endif
- }
-
- CalculateStutterEfficiency_params->display_cfg = display_cfg;
- CalculateStutterEfficiency_params->CompressedBufferSizeInkByte = mode_lib->mp.CompressedBufferSizeInkByte;
- CalculateStutterEfficiency_params->UnboundedRequestEnabled = mode_lib->mp.UnboundedRequestEnabled;
- CalculateStutterEfficiency_params->MetaFIFOSizeInKEntries = mode_lib->ip.meta_fifo_size_in_kentries;
- CalculateStutterEfficiency_params->ZeroSizeBufferEntries = mode_lib->ip.zero_size_buffer_entries;
- CalculateStutterEfficiency_params->PixelChunkSizeInKByte = mode_lib->ip.pixel_chunk_size_kbytes;
- CalculateStutterEfficiency_params->NumberOfActiveSurfaces = s->num_active_planes;
- CalculateStutterEfficiency_params->ROBBufferSizeInKByte = mode_lib->ip.rob_buffer_size_kbytes;
- CalculateStutterEfficiency_params->TotalDataReadBandwidth = mode_lib->mp.TotalDataReadBandwidth;
- CalculateStutterEfficiency_params->DCFCLK = mode_lib->mp.Dcfclk;
- CalculateStutterEfficiency_params->ReturnBW = mode_lib->mp.urg_bandwidth_available_min[dml2_core_internal_soc_state_sys_active];
- CalculateStutterEfficiency_params->CompbufReservedSpace64B = mode_lib->mp.compbuf_reserved_space_64b;
- CalculateStutterEfficiency_params->CompbufReservedSpaceZs = mode_lib->ip.compbuf_reserved_space_zs;
- CalculateStutterEfficiency_params->SRExitTime = mode_lib->soc.power_management_parameters.stutter_exit_latency_us;
- CalculateStutterEfficiency_params->SRExitZ8Time = mode_lib->soc.power_management_parameters.z8_stutter_exit_latency_us;
- CalculateStutterEfficiency_params->SynchronizeTimings = display_cfg->overrides.synchronize_timings;
- CalculateStutterEfficiency_params->StutterEnterPlusExitWatermark = mode_lib->mp.Watermark.StutterEnterPlusExitWatermark;
- CalculateStutterEfficiency_params->Z8StutterEnterPlusExitWatermark = mode_lib->mp.Watermark.Z8StutterEnterPlusExitWatermark;
- CalculateStutterEfficiency_params->ProgressiveToInterlaceUnitInOPP = mode_lib->ip.ptoi_supported;
- CalculateStutterEfficiency_params->MinTTUVBlank = mode_lib->mp.MinTTUVBlank;
- CalculateStutterEfficiency_params->DPPPerSurface = mode_lib->mp.NoOfDPP;
- CalculateStutterEfficiency_params->DETBufferSizeY = mode_lib->mp.DETBufferSizeY;
- CalculateStutterEfficiency_params->BytePerPixelY = mode_lib->mp.BytePerPixelY;
- CalculateStutterEfficiency_params->BytePerPixelDETY = mode_lib->mp.BytePerPixelInDETY;
- CalculateStutterEfficiency_params->SwathWidthY = mode_lib->mp.SwathWidthY;
- CalculateStutterEfficiency_params->SwathHeightY = mode_lib->mp.SwathHeightY;
- CalculateStutterEfficiency_params->SwathHeightC = mode_lib->mp.SwathHeightC;
- CalculateStutterEfficiency_params->BlockHeight256BytesY = mode_lib->mp.Read256BlockHeightY;
- CalculateStutterEfficiency_params->BlockWidth256BytesY = mode_lib->mp.Read256BlockWidthY;
- CalculateStutterEfficiency_params->BlockHeight256BytesC = mode_lib->mp.Read256BlockHeightC;
- CalculateStutterEfficiency_params->BlockWidth256BytesC = mode_lib->mp.Read256BlockWidthC;
- CalculateStutterEfficiency_params->DCCYMaxUncompressedBlock = mode_lib->mp.DCCYMaxUncompressedBlock;
- CalculateStutterEfficiency_params->DCCCMaxUncompressedBlock = mode_lib->mp.DCCCMaxUncompressedBlock;
- CalculateStutterEfficiency_params->ReadBandwidthSurfaceLuma = mode_lib->mp.SurfaceReadBandwidthLuma;
- CalculateStutterEfficiency_params->ReadBandwidthSurfaceChroma = mode_lib->mp.SurfaceReadBandwidthChroma;
- CalculateStutterEfficiency_params->dpte_row_bw = mode_lib->mp.dpte_row_bw;
- CalculateStutterEfficiency_params->meta_row_bw = mode_lib->mp.meta_row_bw;
- CalculateStutterEfficiency_params->rob_alloc_compressed = mode_lib->ip.dcn_mrq_present;
-
- // output
- CalculateStutterEfficiency_params->StutterEfficiencyNotIncludingVBlank = &mode_lib->mp.StutterEfficiencyNotIncludingVBlank;
- CalculateStutterEfficiency_params->StutterEfficiency = &mode_lib->mp.StutterEfficiency;
- CalculateStutterEfficiency_params->NumberOfStutterBurstsPerFrame = &mode_lib->mp.NumberOfStutterBurstsPerFrame;
- CalculateStutterEfficiency_params->Z8StutterEfficiencyNotIncludingVBlank = &mode_lib->mp.Z8StutterEfficiencyNotIncludingVBlank;
- CalculateStutterEfficiency_params->Z8StutterEfficiency = &mode_lib->mp.Z8StutterEfficiency;
- CalculateStutterEfficiency_params->Z8NumberOfStutterBurstsPerFrame = &mode_lib->mp.Z8NumberOfStutterBurstsPerFrame;
- CalculateStutterEfficiency_params->StutterPeriod = &mode_lib->mp.StutterPeriod;
- CalculateStutterEfficiency_params->DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE = &mode_lib->mp.DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE;
-
- // Stutter Efficiency
- CalculateStutterEfficiency(&mode_lib->scratch, CalculateStutterEfficiency_params);
-
-#ifdef __DML_VBA_ALLOW_DELTA__
- // Calculate z8 stutter eff assuming 0 reserved space
- CalculateStutterEfficiency_params->CompbufReservedSpace64B = 0;
- CalculateStutterEfficiency_params->CompbufReservedSpaceZs = 0;
-
- CalculateStutterEfficiency_params->Z8StutterEfficiencyNotIncludingVBlank = &mode_lib->mp.Z8StutterEfficiencyNotIncludingVBlankBestCase;
- CalculateStutterEfficiency_params->Z8StutterEfficiency = &mode_lib->mp.Z8StutterEfficiencyBestCase;
- CalculateStutterEfficiency_params->Z8NumberOfStutterBurstsPerFrame = &mode_lib->mp.Z8NumberOfStutterBurstsPerFrameBestCase;
- CalculateStutterEfficiency_params->StutterPeriod = &mode_lib->mp.StutterPeriodBestCase;
-
- // Stutter Efficiency
- CalculateStutterEfficiency(&mode_lib->scratch, CalculateStutterEfficiency_params);
-#else
- mode_lib->mp.Z8StutterEfficiencyNotIncludingVBlankBestCase = mode_lib->mp.Z8StutterEfficiencyNotIncludingVBlank;
- mode_lib->mp.Z8StutterEfficiencyBestCase = mode_lib->mp.Z8StutterEfficiency;
- mode_lib->mp.Z8NumberOfStutterBurstsPerFrameBestCase = mode_lib->mp.Z8NumberOfStutterBurstsPerFrame;
- mode_lib->mp.StutterPeriodBestCase = mode_lib->mp.StutterPeriod;
-#endif
- } // PrefetchAndImmediateFlipSupported
-
- const long min_return_uclk_cycles = 83;
- const long min_return_fclk_cycles = 75;
- double max_fclk_mhz = min_clk_table->max_clocks_khz.fclk / 1000.0;
- double max_uclk_mhz = mode_lib->soc.clk_table.uclk.clk_values_khz[mode_lib->soc.clk_table.uclk.num_clk_values - 1] / 1000.0;
- double hard_minimum_dcfclk_mhz = (double)min_clk_table->dram_bw_table.entries[0].min_dcfclk_khz / 1000.0;
- double min_return_latency_in_DCFCLK_cycles = (min_return_uclk_cycles / max_uclk_mhz + min_return_fclk_cycles / max_fclk_mhz) * hard_minimum_dcfclk_mhz;
- mode_lib->mp.min_return_latency_in_dcfclk = (unsigned int)min_return_latency_in_DCFCLK_cycles;
- mode_lib->mp.dcfclk_deep_sleep_hysteresis = (unsigned int)math_max2(32, (double)mode_lib->ip.pixel_chunk_size_kbytes * 1024 * 3 / 4 / 64 - min_return_latency_in_DCFCLK_cycles);
- mode_lib->mp.dcfclk_deep_sleep_hysteresis = 255;
- DML2_ASSERT(mode_lib->mp.dcfclk_deep_sleep_hysteresis < 256);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: max_fclk_mhz = %f\n", __func__, max_fclk_mhz);
- dml2_printf("DML::%s: max_uclk_mhz = %f\n", __func__, max_uclk_mhz);
- dml2_printf("DML::%s: hard_minimum_dcfclk_mhz = %f\n", __func__, hard_minimum_dcfclk_mhz);
- dml2_printf("DML::%s: min_return_uclk_cycles = %d\n", __func__, min_return_uclk_cycles);
- dml2_printf("DML::%s: min_return_fclk_cycles = %d\n", __func__, min_return_fclk_cycles);
- dml2_printf("DML::%s: min_return_latency_in_DCFCLK_cycles = %f\n", __func__, min_return_latency_in_DCFCLK_cycles);
- dml2_printf("DML::%s: dcfclk_deep_sleep_hysteresis = %d \n", __func__, mode_lib->mp.dcfclk_deep_sleep_hysteresis);
- dml2_printf("DML::%s: --- END --- \n", __func__);
-#endif
-
- return (in_out_params->mode_lib->mp.PrefetchAndImmediateFlipSupported);
-}
-
-static bool dml_is_dual_plane(enum dml2_source_format_class source_format)
-{
- bool ret_val = 0;
-
- if ((source_format == dml2_420_12) || (source_format == dml2_420_8) || (source_format == dml2_420_10) || (source_format == dml2_rgbe_alpha))
- ret_val = 1;
-
- return ret_val;
-}
-
-static unsigned int dml_get_plane_idx(const struct dml2_core_internal_display_mode_lib *mode_lib, unsigned int pipe_idx)
-{
- unsigned int plane_idx = mode_lib->mp.pipe_plane[pipe_idx];
- return plane_idx;
-}
-
-static void rq_dlg_get_wm_regs(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_dchub_watermark_regs *wm_regs)
-{
- double refclk_freq_in_mhz = (display_cfg->overrides.hw.dlg_ref_clk_mhz > 0) ? (double)display_cfg->overrides.hw.dlg_ref_clk_mhz : mode_lib->soc.dchub_refclk_mhz;
-
- wm_regs->fclk_pstate = (int unsigned)(mode_lib->mp.Watermark.FCLKChangeWatermark * refclk_freq_in_mhz);
- wm_regs->sr_enter = (int unsigned)(mode_lib->mp.Watermark.StutterEnterPlusExitWatermark * refclk_freq_in_mhz);
- wm_regs->sr_exit = (int unsigned)(mode_lib->mp.Watermark.StutterExitWatermark * refclk_freq_in_mhz);
- wm_regs->temp_read_or_ppt = 0;
- wm_regs->uclk_pstate = (int unsigned)(mode_lib->mp.Watermark.DRAMClockChangeWatermark * refclk_freq_in_mhz);
- wm_regs->urgent = (int unsigned)(mode_lib->mp.Watermark.UrgentWatermark * refclk_freq_in_mhz);
-}
-
-static unsigned int log_and_substract_if_non_zero(unsigned int a, unsigned int subtrahend)
-{
- if (a == 0)
- return 0;
-
- return (math_log2_approx(a) - subtrahend);
-}
-
-void dml2_core_shared_cursor_dlg_reg(struct dml2_cursor_dlg_regs *cursor_dlg_regs, const struct dml2_get_cursor_dlg_reg *p)
-{
- int dst_x_offset = (int)((p->cursor_x_position + (p->cursor_stereo_en == 0 ? 0 : math_max2(p->cursor_primary_offset, p->cursor_secondary_offset)) -
- (p->cursor_hotspot_x * (p->cursor_2x_magnify == 0 ? 1 : 2))) * p->dlg_refclk_mhz / p->pixel_rate_mhz / p->hratio);
- cursor_dlg_regs->dst_x_offset = (unsigned int)((dst_x_offset > 0) ? dst_x_offset : 0);
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG::%s: cursor_x_position=%d\n", __func__, p->cursor_x_position);
- dml2_printf("DML_DLG::%s: dlg_refclk_mhz=%f\n", __func__, p->dlg_refclk_mhz);
- dml2_printf("DML_DLG::%s: pixel_rate_mhz=%f\n", __func__, p->pixel_rate_mhz);
- dml2_printf("DML_DLG::%s: dst_x_offset=%d\n", __func__, dst_x_offset);
- dml2_printf("DML_DLG::%s: dst_x_offset=%d (reg)\n", __func__, cursor_dlg_regs->dst_x_offset);
-#endif
-
- cursor_dlg_regs->chunk_hdl_adjust = 3;
- cursor_dlg_regs->dst_y_offset = 0;
-
- cursor_dlg_regs->qos_level_fixed = 8;
- cursor_dlg_regs->qos_ramp_disable = 0;
-}
-
-static void rq_dlg_get_rq_reg(struct dml2_display_rq_regs *rq_regs,
- const struct dml2_display_cfg *display_cfg,
- const struct dml2_core_internal_display_mode_lib *mode_lib,
- unsigned int pipe_idx)
-{
- dml2_printf("DML_DLG::%s: Calculation for pipe[%d] start\n", __func__, pipe_idx);
-
- unsigned int plane_idx = dml_get_plane_idx(mode_lib, pipe_idx);
- enum dml2_source_format_class source_format = display_cfg->plane_descriptors[plane_idx].pixel_format;
- enum dml2_swizzle_mode sw_mode = display_cfg->plane_descriptors[plane_idx].surface.tiling;
- bool dual_plane = dml_is_dual_plane((enum dml2_source_format_class)(source_format));
-
- unsigned int pixel_chunk_bytes = 0;
- unsigned int min_pixel_chunk_bytes = 0;
- unsigned int dpte_group_bytes = 0;
- unsigned int mpte_group_bytes = 0;
-
- unsigned int p1_pixel_chunk_bytes = 0;
- unsigned int p1_min_pixel_chunk_bytes = 0;
- unsigned int p1_dpte_group_bytes = 0;
- unsigned int p1_mpte_group_bytes = 0;
-
- pixel_chunk_bytes = (unsigned int)(mode_lib->ip.pixel_chunk_size_kbytes * 1024);
- min_pixel_chunk_bytes = (unsigned int)(mode_lib->ip.min_pixel_chunk_size_bytes);
-
- if (pixel_chunk_bytes == 64 * 1024)
- min_pixel_chunk_bytes = 0;
-
- dpte_group_bytes = (unsigned int)(mode_lib->mp.dpte_group_bytes[mode_lib->mp.pipe_plane[pipe_idx]]);
- mpte_group_bytes = (unsigned int)(mode_lib->mp.vm_group_bytes[mode_lib->mp.pipe_plane[pipe_idx]]);
-
- p1_pixel_chunk_bytes = pixel_chunk_bytes;
- p1_min_pixel_chunk_bytes = min_pixel_chunk_bytes;
- p1_dpte_group_bytes = dpte_group_bytes;
- p1_mpte_group_bytes = mpte_group_bytes;
-
- if (source_format == dml2_rgbe_alpha)
- p1_pixel_chunk_bytes = (unsigned int)(mode_lib->ip.alpha_pixel_chunk_size_kbytes * 1024);
-
- rq_regs->unbounded_request_enabled = mode_lib->mp.UnboundedRequestEnabled;
- rq_regs->rq_regs_l.chunk_size = log_and_substract_if_non_zero(pixel_chunk_bytes, 10);
- rq_regs->rq_regs_c.chunk_size = log_and_substract_if_non_zero(p1_pixel_chunk_bytes, 10);
-
- if (min_pixel_chunk_bytes == 0)
- rq_regs->rq_regs_l.min_chunk_size = 0;
- else
- rq_regs->rq_regs_l.min_chunk_size = log_and_substract_if_non_zero(min_pixel_chunk_bytes, 8 - 1);
-
- if (p1_min_pixel_chunk_bytes == 0)
- rq_regs->rq_regs_c.min_chunk_size = 0;
- else
- rq_regs->rq_regs_c.min_chunk_size = log_and_substract_if_non_zero(p1_min_pixel_chunk_bytes, 8 - 1);
-
- rq_regs->rq_regs_l.dpte_group_size = log_and_substract_if_non_zero(dpte_group_bytes, 6);
- rq_regs->rq_regs_l.mpte_group_size = log_and_substract_if_non_zero(mpte_group_bytes, 6);
- rq_regs->rq_regs_c.dpte_group_size = log_and_substract_if_non_zero(p1_dpte_group_bytes, 6);
- rq_regs->rq_regs_c.mpte_group_size = log_and_substract_if_non_zero(p1_mpte_group_bytes, 6);
-
- unsigned int detile_buf_size_in_bytes = (unsigned int)(mode_lib->mp.DETBufferSizeInKByte[mode_lib->mp.pipe_plane[pipe_idx]] * 1024);
- unsigned int detile_buf_plane1_addr = 0;
-
- if (sw_mode == dml2_sw_linear && display_cfg->gpuvm_enable) {
- unsigned int p0_pte_row_height_linear = (unsigned int)(mode_lib->mp.dpte_row_height_linear[mode_lib->mp.pipe_plane[pipe_idx]]);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: p0_pte_row_height_linear = %u\n", __func__, p0_pte_row_height_linear);
-#endif
- DML2_ASSERT(p0_pte_row_height_linear >= 8);
-
- rq_regs->rq_regs_l.pte_row_height_linear = math_log2_approx(p0_pte_row_height_linear) - 3;
- if (dual_plane) {
- unsigned int p1_pte_row_height_linear = (unsigned int)(mode_lib->mp.dpte_row_height_linear_chroma[mode_lib->mp.pipe_plane[pipe_idx]]);
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: p1_pte_row_height_linear = %u\n", __func__, p1_pte_row_height_linear);
-#endif
- if (sw_mode == dml2_sw_linear) {
- DML2_ASSERT(p1_pte_row_height_linear >= 8);
- }
-
- rq_regs->rq_regs_c.pte_row_height_linear = math_log2_approx(p1_pte_row_height_linear) - 3;
- }
- } else {
- rq_regs->rq_regs_l.pte_row_height_linear = 0;
- rq_regs->rq_regs_c.pte_row_height_linear = 0;
- }
-
- rq_regs->rq_regs_l.swath_height = log_and_substract_if_non_zero(mode_lib->mp.SwathHeightY[mode_lib->mp.pipe_plane[pipe_idx]], 0);
- rq_regs->rq_regs_c.swath_height = log_and_substract_if_non_zero(mode_lib->mp.SwathHeightC[mode_lib->mp.pipe_plane[pipe_idx]], 0);
-
- // FIXME_DCN4, programming guide has dGPU condition
- if (pixel_chunk_bytes >= 32 * 1024 || (dual_plane && p1_pixel_chunk_bytes >= 32 * 1024)) { //32kb
- rq_regs->drq_expansion_mode = 0;
- } else {
- rq_regs->drq_expansion_mode = 2;
- }
- rq_regs->prq_expansion_mode = 1;
- rq_regs->crq_expansion_mode = 1;
- rq_regs->mrq_expansion_mode = 1;
-
- double stored_swath_l_bytes = mode_lib->mp.DETBufferSizeY[mode_lib->mp.pipe_plane[pipe_idx]];
- double stored_swath_c_bytes = mode_lib->mp.DETBufferSizeC[mode_lib->mp.pipe_plane[pipe_idx]];
- bool is_phantom_pipe = dml_get_is_phantom_pipe(display_cfg, mode_lib, pipe_idx);
-
- // Note: detile_buf_plane1_addr is in unit of 1KB
- if (dual_plane) {
- if (is_phantom_pipe) {
- detile_buf_plane1_addr = (unsigned int)((1024.0 * 1024.0) / 2.0 / 1024.0); // half to chroma
- } else {
- if (stored_swath_l_bytes / stored_swath_c_bytes <= 1.5) {
- detile_buf_plane1_addr = (unsigned int)(detile_buf_size_in_bytes / 2.0 / 1024.0); // half to chroma
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: detile_buf_plane1_addr = %d (1/2 to chroma)\n", __func__, detile_buf_plane1_addr);
-#endif
- } else {
- detile_buf_plane1_addr = (unsigned int)(dml_round_to_multiple((unsigned int)((2.0 * detile_buf_size_in_bytes) / 3.0), 1024, 0) / 1024.0); // 2/3 to luma
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: detile_buf_plane1_addr = %d (1/3 chroma)\n", __func__, detile_buf_plane1_addr);
-#endif
- }
- }
- }
- rq_regs->plane1_base_address = detile_buf_plane1_addr;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML_DLG: %s: is_phantom_pipe = %d\n", __func__, is_phantom_pipe);
- dml2_printf("DML_DLG: %s: stored_swath_l_bytes = %f\n", __func__, stored_swath_l_bytes);
- dml2_printf("DML_DLG: %s: stored_swath_c_bytes = %f\n", __func__, stored_swath_c_bytes);
- dml2_printf("DML_DLG: %s: detile_buf_size_in_bytes = %d\n", __func__, detile_buf_size_in_bytes);
- dml2_printf("DML_DLG: %s: detile_buf_plane1_addr = %d\n", __func__, detile_buf_plane1_addr);
- dml2_printf("DML_DLG: %s: plane1_base_address = %d\n", __func__, rq_regs->plane1_base_address);
-#endif
- //dml2_printf_rq_regs_st(rq_regs);
- dml2_printf("DML_DLG::%s: Calculation for pipe[%d] done\n", __func__, pipe_idx);
-}
-
-static void rq_dlg_get_dlg_reg(
- struct dml2_core_internal_scratch *s,
- struct dml2_display_dlg_regs *disp_dlg_regs,
- struct dml2_display_ttu_regs *disp_ttu_regs,
- const struct dml2_display_cfg *display_cfg,
- const struct dml2_core_internal_display_mode_lib *mode_lib,
- const unsigned int pipe_idx)
-{
- struct dml2_core_shared_rq_dlg_get_dlg_reg_locals *l = &s->rq_dlg_get_dlg_reg_locals;
-
- memset(l, 0, sizeof(struct dml2_core_shared_rq_dlg_get_dlg_reg_locals));
-
- dml2_printf("DML_DLG::%s: Calculation for pipe_idx=%d\n", __func__, pipe_idx);
-
- l->plane_idx = dml_get_plane_idx(mode_lib, pipe_idx);
- dml2_assert(l->plane_idx < DML2_MAX_PLANES);
-
- l->source_format = dml2_444_8;
- l->dual_plane = dml_is_dual_plane(l->source_format);
- l->odm_mode = dml2_odm_mode_bypass;
-
- l->htotal = 0;
- l->hactive = 0;
- l->hblank_end = 0;
- l->vblank_end = 0;
- l->interlaced = false;
- l->pclk_freq_in_mhz = 0.0;
- l->refclk_freq_in_mhz = (display_cfg->overrides.hw.dlg_ref_clk_mhz > 0) ? (double)display_cfg->overrides.hw.dlg_ref_clk_mhz : mode_lib->soc.dchub_refclk_mhz;
- l->ref_freq_to_pix_freq = 0.0;
-
- if (l->plane_idx < DML2_MAX_PLANES) {
-
- l->timing = &display_cfg->stream_descriptors[display_cfg->plane_descriptors[l->plane_idx].stream_index].timing;
- l->source_format = display_cfg->plane_descriptors[l->plane_idx].pixel_format;
- l->odm_mode = mode_lib->mp.ODMMode[l->plane_idx];
-
- l->htotal = l->timing->h_total;
- l->hactive = l->timing->h_active;
- l->hblank_end = l->timing->h_blank_end;
- l->vblank_end = l->timing->v_blank_end;
- l->interlaced = l->timing->interlaced;
- l->pclk_freq_in_mhz = (double)l->timing->pixel_clock_khz / 1000;
- l->ref_freq_to_pix_freq = l->refclk_freq_in_mhz / l->pclk_freq_in_mhz;
-
- dml2_printf("DML_DLG::%s: plane_idx = %d\n", __func__, l->plane_idx);
- dml2_printf("DML_DLG: %s: htotal = %d\n", __func__, l->htotal);
- dml2_printf("DML_DLG: %s: refclk_freq_in_mhz = %3.2f\n", __func__, l->refclk_freq_in_mhz);
- dml2_printf("DML_DLG: %s: dlg_ref_clk_mhz = %3.2f\n", __func__, display_cfg->overrides.hw.dlg_ref_clk_mhz);
- dml2_printf("DML_DLG: %s: soc.refclk_mhz = %3.2f\n", __func__, mode_lib->soc.dchub_refclk_mhz);
- dml2_printf("DML_DLG: %s: pclk_freq_in_mhz = %3.2f\n", __func__, l->pclk_freq_in_mhz);
- dml2_printf("DML_DLG: %s: ref_freq_to_pix_freq = %3.2f\n", __func__, l->ref_freq_to_pix_freq);
- dml2_printf("DML_DLG: %s: interlaced = %d\n", __func__, l->interlaced);
-
- DML2_ASSERT(l->refclk_freq_in_mhz != 0);
- DML2_ASSERT(l->pclk_freq_in_mhz != 0);
- DML2_ASSERT(l->ref_freq_to_pix_freq < 4.0);
-
- // Need to figure out which side of odm combine we're in
- // Assume the pipe instance under the same plane is in order
-
- if (l->odm_mode == dml2_odm_mode_bypass) {
- disp_dlg_regs->refcyc_h_blank_end = (unsigned int)((double)l->hblank_end * l->ref_freq_to_pix_freq);
- } else if (l->odm_mode == dml2_odm_mode_combine_2to1 || l->odm_mode == dml2_odm_mode_combine_3to1 || l->odm_mode == dml2_odm_mode_combine_4to1) {
- // find out how many pipe are in this plane
- l->num_active_pipes = mode_lib->mp.num_active_pipes;
- l->first_pipe_idx_in_plane = DML2_MAX_PLANES;
- l->pipe_idx_in_combine = 0; // pipe index within the plane
- l->odm_combine_factor = 2;
-
- if (l->odm_mode == dml2_odm_mode_combine_3to1)
- l->odm_combine_factor = 3;
- else if (l->odm_mode == dml2_odm_mode_combine_4to1)
- l->odm_combine_factor = 4;
-
- for (unsigned int i = 0; i < l->num_active_pipes; i++) {
- if (dml_get_plane_idx(mode_lib, i) == l->plane_idx) {
- if (i < l->first_pipe_idx_in_plane) {
- l->first_pipe_idx_in_plane = i;
- }
- }
- }
- l->pipe_idx_in_combine = pipe_idx - l->first_pipe_idx_in_plane; // DML assumes the pipes in the same plane will have continuous indexing (i.e. plane 0 use pipe 0, 1, and plane 1 uses pipe 2, 3, etc.)
-
- disp_dlg_regs->refcyc_h_blank_end = (unsigned int)(((double)l->hblank_end + (double)l->pipe_idx_in_combine * (double)l->hactive / (double)l->odm_combine_factor) * l->ref_freq_to_pix_freq);
- dml2_printf("DML_DLG: %s: pipe_idx = %d\n", __func__, pipe_idx);
- dml2_printf("DML_DLG: %s: first_pipe_idx_in_plane = %d\n", __func__, l->first_pipe_idx_in_plane);
- dml2_printf("DML_DLG: %s: pipe_idx_in_combine = %d\n", __func__, l->pipe_idx_in_combine);
- dml2_printf("DML_DLG: %s: odm_combine_factor = %d\n", __func__, l->odm_combine_factor);
- }
- dml2_printf("DML_DLG: %s: refcyc_h_blank_end = %d\n", __func__, disp_dlg_regs->refcyc_h_blank_end);
-
- DML2_ASSERT(disp_dlg_regs->refcyc_h_blank_end < (unsigned int)math_pow(2, 13));
-
- disp_dlg_regs->ref_freq_to_pix_freq = (unsigned int)(l->ref_freq_to_pix_freq * math_pow(2, 19));
- disp_dlg_regs->refcyc_per_htotal = (unsigned int)(l->ref_freq_to_pix_freq * (double)l->htotal * math_pow(2, 8));
- disp_dlg_regs->dlg_vblank_end = l->interlaced ? (l->vblank_end / 2) : l->vblank_end; // 15 bits
-
- l->min_ttu_vblank = mode_lib->mp.MinTTUVBlank[mode_lib->mp.pipe_plane[pipe_idx]];
- l->min_dst_y_next_start = (unsigned int)(mode_lib->mp.MIN_DST_Y_NEXT_START[mode_lib->mp.pipe_plane[pipe_idx]]);
-
- dml2_printf("DML_DLG: %s: min_ttu_vblank (us) = %3.2f\n", __func__, l->min_ttu_vblank);
- dml2_printf("DML_DLG: %s: min_dst_y_next_start = %d\n", __func__, l->min_dst_y_next_start);
- dml2_printf("DML_DLG: %s: ref_freq_to_pix_freq = %3.2f\n", __func__, l->ref_freq_to_pix_freq);
-
- l->vready_after_vcount0 = (unsigned int)(mode_lib->mp.VREADY_AT_OR_AFTER_VSYNC[mode_lib->mp.pipe_plane[pipe_idx]]);
- disp_dlg_regs->vready_after_vcount0 = l->vready_after_vcount0;
-
- dml2_printf("DML_DLG: %s: vready_after_vcount0 = %d\n", __func__, disp_dlg_regs->vready_after_vcount0);
-
- l->dst_x_after_scaler = (unsigned int)(mode_lib->mp.DSTXAfterScaler[mode_lib->mp.pipe_plane[pipe_idx]]);
- l->dst_y_after_scaler = (unsigned int)(mode_lib->mp.DSTYAfterScaler[mode_lib->mp.pipe_plane[pipe_idx]]);
-
- dml2_printf("DML_DLG: %s: dst_x_after_scaler = %d\n", __func__, l->dst_x_after_scaler);
- dml2_printf("DML_DLG: %s: dst_y_after_scaler = %d\n", __func__, l->dst_y_after_scaler);
-
- l->dst_y_prefetch = mode_lib->mp.dst_y_prefetch[mode_lib->mp.pipe_plane[pipe_idx]];
- l->dst_y_per_vm_vblank = mode_lib->mp.dst_y_per_vm_vblank[mode_lib->mp.pipe_plane[pipe_idx]];
- l->dst_y_per_row_vblank = mode_lib->mp.dst_y_per_row_vblank[mode_lib->mp.pipe_plane[pipe_idx]];
- l->dst_y_per_vm_flip = mode_lib->mp.dst_y_per_vm_flip[mode_lib->mp.pipe_plane[pipe_idx]];
- l->dst_y_per_row_flip = mode_lib->mp.dst_y_per_row_flip[mode_lib->mp.pipe_plane[pipe_idx]];
-
- l->max_dst_y_per_vm_vblank = 32.0; //U5.2
- l->max_dst_y_per_row_vblank = 16.0; //U4.2
-
- // magic!
- if (l->htotal <= 75) {
- l->max_dst_y_per_vm_vblank = 100.0;
- l->max_dst_y_per_row_vblank = 100.0;
- }
-
- dml2_printf("DML_DLG: %s: dst_y_prefetch (after rnd) = %3.2f\n", __func__, l->dst_y_prefetch);
- dml2_printf("DML_DLG: %s: dst_y_per_vm_flip = %3.2f\n", __func__, l->dst_y_per_vm_flip);
- dml2_printf("DML_DLG: %s: dst_y_per_row_flip = %3.2f\n", __func__, l->dst_y_per_row_flip);
- dml2_printf("DML_DLG: %s: dst_y_per_vm_vblank = %3.2f\n", __func__, l->dst_y_per_vm_vblank);
- dml2_printf("DML_DLG: %s: dst_y_per_row_vblank = %3.2f\n", __func__, l->dst_y_per_row_vblank);
-
- DML2_ASSERT(l->dst_y_per_vm_vblank < l->max_dst_y_per_vm_vblank);
- DML2_ASSERT(l->dst_y_per_row_vblank < l->max_dst_y_per_row_vblank);
- if (l->dst_y_prefetch > 0 && l->dst_y_per_vm_vblank > 0 && l->dst_y_per_row_vblank > 0) {
- DML2_ASSERT(l->dst_y_prefetch > (l->dst_y_per_vm_vblank + l->dst_y_per_row_vblank));
- }
-
- l->vratio_pre_l = mode_lib->mp.VRatioPrefetchY[mode_lib->mp.pipe_plane[pipe_idx]];
- l->vratio_pre_c = mode_lib->mp.VRatioPrefetchC[mode_lib->mp.pipe_plane[pipe_idx]];
-
- dml2_printf("DML_DLG: %s: vratio_pre_l = %3.2f\n", __func__, l->vratio_pre_l);
- dml2_printf("DML_DLG: %s: vratio_pre_c = %3.2f\n", __func__, l->vratio_pre_c);
-
- // Active
- l->refcyc_per_line_delivery_pre_l = mode_lib->mp.DisplayPipeLineDeliveryTimeLumaPrefetch[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_line_delivery_l = mode_lib->mp.DisplayPipeLineDeliveryTimeLuma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
-
- dml2_printf("DML_DLG: %s: refcyc_per_line_delivery_pre_l = %3.2f\n", __func__, l->refcyc_per_line_delivery_pre_l);
- dml2_printf("DML_DLG: %s: refcyc_per_line_delivery_l = %3.2f\n", __func__, l->refcyc_per_line_delivery_l);
-
- l->refcyc_per_line_delivery_pre_c = 0.0;
- l->refcyc_per_line_delivery_c = 0.0;
-
- if (l->dual_plane) {
- l->refcyc_per_line_delivery_pre_c = mode_lib->mp.DisplayPipeLineDeliveryTimeChromaPrefetch[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_line_delivery_c = mode_lib->mp.DisplayPipeLineDeliveryTimeChroma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
-
- dml2_printf("DML_DLG: %s: refcyc_per_line_delivery_pre_c = %3.2f\n", __func__, l->refcyc_per_line_delivery_pre_c);
- dml2_printf("DML_DLG: %s: refcyc_per_line_delivery_c = %3.2f\n", __func__, l->refcyc_per_line_delivery_c);
- }
-
- disp_dlg_regs->refcyc_per_vm_dmdata = (unsigned int)(mode_lib->mp.Tdmdl_vm[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz);
- disp_dlg_regs->dmdata_dl_delta = (unsigned int)(mode_lib->mp.Tdmdl[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz);
-
- l->refcyc_per_req_delivery_pre_l = mode_lib->mp.DisplayPipeRequestDeliveryTimeLumaPrefetch[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_req_delivery_l = mode_lib->mp.DisplayPipeRequestDeliveryTimeLuma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
-
- dml2_printf("DML_DLG: %s: refcyc_per_req_delivery_pre_l = %3.2f\n", __func__, l->refcyc_per_req_delivery_pre_l);
- dml2_printf("DML_DLG: %s: refcyc_per_req_delivery_l = %3.2f\n", __func__, l->refcyc_per_req_delivery_l);
-
- l->refcyc_per_req_delivery_pre_c = 0.0;
- l->refcyc_per_req_delivery_c = 0.0;
- if (l->dual_plane) {
- l->refcyc_per_req_delivery_pre_c = mode_lib->mp.DisplayPipeRequestDeliveryTimeChromaPrefetch[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_req_delivery_c = mode_lib->mp.DisplayPipeRequestDeliveryTimeChroma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
-
- dml2_printf("DML_DLG: %s: refcyc_per_req_delivery_pre_c = %3.2f\n", __func__, l->refcyc_per_req_delivery_pre_c);
- dml2_printf("DML_DLG: %s: refcyc_per_req_delivery_c = %3.2f\n", __func__, l->refcyc_per_req_delivery_c);
- }
-
- // TTU - Cursor
- DML2_ASSERT(display_cfg->plane_descriptors[l->plane_idx].cursor.num_cursors <= 1);
-
- // Assign to register structures
- disp_dlg_regs->min_dst_y_next_start = (unsigned int)((double)l->min_dst_y_next_start * math_pow(2, 2));
- DML2_ASSERT(disp_dlg_regs->min_dst_y_next_start < (unsigned int)math_pow(2, 18));
-
- disp_dlg_regs->dst_y_after_scaler = l->dst_y_after_scaler; // in terms of line
- disp_dlg_regs->refcyc_x_after_scaler = (unsigned int)((double)l->dst_x_after_scaler * l->ref_freq_to_pix_freq); // in terms of refclk
- disp_dlg_regs->dst_y_prefetch = (unsigned int)(l->dst_y_prefetch * math_pow(2, 2));
- disp_dlg_regs->dst_y_per_vm_vblank = (unsigned int)(l->dst_y_per_vm_vblank * math_pow(2, 2));
- disp_dlg_regs->dst_y_per_row_vblank = (unsigned int)(l->dst_y_per_row_vblank * math_pow(2, 2));
- disp_dlg_regs->dst_y_per_vm_flip = (unsigned int)(l->dst_y_per_vm_flip * math_pow(2, 2));
- disp_dlg_regs->dst_y_per_row_flip = (unsigned int)(l->dst_y_per_row_flip * math_pow(2, 2));
-
- disp_dlg_regs->vratio_prefetch = (unsigned int)(l->vratio_pre_l * math_pow(2, 19));
- disp_dlg_regs->vratio_prefetch_c = (unsigned int)(l->vratio_pre_c * math_pow(2, 19));
-
- dml2_printf("DML_DLG: %s: disp_dlg_regs->dst_y_per_vm_vblank = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_vm_vblank);
- dml2_printf("DML_DLG: %s: disp_dlg_regs->dst_y_per_row_vblank = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_row_vblank);
- dml2_printf("DML_DLG: %s: disp_dlg_regs->dst_y_per_vm_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_vm_flip);
- dml2_printf("DML_DLG: %s: disp_dlg_regs->dst_y_per_row_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_row_flip);
-
- disp_dlg_regs->refcyc_per_vm_group_vblank = (unsigned int)(mode_lib->mp.TimePerVMGroupVBlank[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz);
- disp_dlg_regs->refcyc_per_vm_group_flip = (unsigned int)(mode_lib->mp.TimePerVMGroupFlip[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz);
- disp_dlg_regs->refcyc_per_vm_req_vblank = (unsigned int)(mode_lib->mp.TimePerVMRequestVBlank[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz * math_pow(2, 10));
- disp_dlg_regs->refcyc_per_vm_req_flip = (unsigned int)(mode_lib->mp.TimePerVMRequestFlip[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz * math_pow(2, 10));
-
- l->dst_y_per_pte_row_nom_l = mode_lib->mp.DST_Y_PER_PTE_ROW_NOM_L[mode_lib->mp.pipe_plane[pipe_idx]];
- l->dst_y_per_pte_row_nom_c = mode_lib->mp.DST_Y_PER_PTE_ROW_NOM_C[mode_lib->mp.pipe_plane[pipe_idx]];
- l->refcyc_per_pte_group_nom_l = mode_lib->mp.time_per_pte_group_nom_luma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_pte_group_nom_c = mode_lib->mp.time_per_pte_group_nom_chroma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_pte_group_vblank_l = mode_lib->mp.time_per_pte_group_vblank_luma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_pte_group_vblank_c = mode_lib->mp.time_per_pte_group_vblank_chroma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_pte_group_flip_l = mode_lib->mp.time_per_pte_group_flip_luma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_pte_group_flip_c = mode_lib->mp.time_per_pte_group_flip_chroma[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_tdlut_group = mode_lib->mp.time_per_tdlut_group[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
-
- disp_dlg_regs->dst_y_per_pte_row_nom_l = (unsigned int)(l->dst_y_per_pte_row_nom_l * math_pow(2, 2));
- disp_dlg_regs->dst_y_per_pte_row_nom_c = (unsigned int)(l->dst_y_per_pte_row_nom_c * math_pow(2, 2));
-
- disp_dlg_regs->refcyc_per_pte_group_nom_l = (unsigned int)(l->refcyc_per_pte_group_nom_l);
- disp_dlg_regs->refcyc_per_pte_group_nom_c = (unsigned int)(l->refcyc_per_pte_group_nom_c);
- disp_dlg_regs->refcyc_per_pte_group_vblank_l = (unsigned int)(l->refcyc_per_pte_group_vblank_l);
- disp_dlg_regs->refcyc_per_pte_group_vblank_c = (unsigned int)(l->refcyc_per_pte_group_vblank_c);
- disp_dlg_regs->refcyc_per_pte_group_flip_l = (unsigned int)(l->refcyc_per_pte_group_flip_l);
- disp_dlg_regs->refcyc_per_pte_group_flip_c = (unsigned int)(l->refcyc_per_pte_group_flip_c);
- disp_dlg_regs->refcyc_per_line_delivery_pre_l = (unsigned int)math_floor2(l->refcyc_per_line_delivery_pre_l, 1);
- disp_dlg_regs->refcyc_per_line_delivery_l = (unsigned int)math_floor2(l->refcyc_per_line_delivery_l, 1);
- disp_dlg_regs->refcyc_per_line_delivery_pre_c = (unsigned int)math_floor2(l->refcyc_per_line_delivery_pre_c, 1);
- disp_dlg_regs->refcyc_per_line_delivery_c = (unsigned int)math_floor2(l->refcyc_per_line_delivery_c, 1);
-
- l->dst_y_per_meta_row_nom_l = mode_lib->mp.DST_Y_PER_META_ROW_NOM_L[mode_lib->mp.pipe_plane[pipe_idx]];
- l->dst_y_per_meta_row_nom_c = mode_lib->mp.DST_Y_PER_META_ROW_NOM_C[mode_lib->mp.pipe_plane[pipe_idx]];
- l->refcyc_per_meta_chunk_nom_l = mode_lib->mp.TimePerMetaChunkNominal[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_meta_chunk_nom_c = mode_lib->mp.TimePerChromaMetaChunkNominal[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_meta_chunk_vblank_l = mode_lib->mp.TimePerMetaChunkVBlank[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_meta_chunk_vblank_c = mode_lib->mp.TimePerChromaMetaChunkVBlank[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_meta_chunk_flip_l = mode_lib->mp.TimePerMetaChunkFlip[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
- l->refcyc_per_meta_chunk_flip_c = mode_lib->mp.TimePerChromaMetaChunkFlip[mode_lib->mp.pipe_plane[pipe_idx]] * l->refclk_freq_in_mhz;
-
- disp_dlg_regs->dst_y_per_meta_row_nom_l = (unsigned int)(l->dst_y_per_meta_row_nom_l * math_pow(2, 2));
- disp_dlg_regs->dst_y_per_meta_row_nom_c = (unsigned int)(l->dst_y_per_meta_row_nom_c * math_pow(2, 2));
- disp_dlg_regs->refcyc_per_meta_chunk_nom_l = (unsigned int)(l->refcyc_per_meta_chunk_nom_l);
- disp_dlg_regs->refcyc_per_meta_chunk_nom_c = (unsigned int)(l->refcyc_per_meta_chunk_nom_c);
- disp_dlg_regs->refcyc_per_meta_chunk_vblank_l = (unsigned int)(l->refcyc_per_meta_chunk_vblank_l);
- disp_dlg_regs->refcyc_per_meta_chunk_vblank_c = (unsigned int)(l->refcyc_per_meta_chunk_vblank_c);
- disp_dlg_regs->refcyc_per_meta_chunk_flip_l = (unsigned int)(l->refcyc_per_meta_chunk_flip_l);
- disp_dlg_regs->refcyc_per_meta_chunk_flip_c = (unsigned int)(l->refcyc_per_meta_chunk_flip_c);
-
- disp_dlg_regs->refcyc_per_tdlut_group = (unsigned int)(l->refcyc_per_tdlut_group);
- disp_dlg_regs->dst_y_delta_drq_limit = 0x7fff; // off
-
- disp_ttu_regs->refcyc_per_req_delivery_pre_l = (unsigned int)(l->refcyc_per_req_delivery_pre_l * math_pow(2, 10));
- disp_ttu_regs->refcyc_per_req_delivery_l = (unsigned int)(l->refcyc_per_req_delivery_l * math_pow(2, 10));
- disp_ttu_regs->refcyc_per_req_delivery_pre_c = (unsigned int)(l->refcyc_per_req_delivery_pre_c * math_pow(2, 10));
- disp_ttu_regs->refcyc_per_req_delivery_c = (unsigned int)(l->refcyc_per_req_delivery_c * math_pow(2, 10));
- disp_ttu_regs->qos_level_low_wm = 0;
-
- disp_ttu_regs->qos_level_high_wm = (unsigned int)(4.0 * (double)l->htotal * l->ref_freq_to_pix_freq);
-
- disp_ttu_regs->qos_level_flip = 14;
- disp_ttu_regs->qos_level_fixed_l = 8;
- disp_ttu_regs->qos_level_fixed_c = 8;
- disp_ttu_regs->qos_ramp_disable_l = 0;
- disp_ttu_regs->qos_ramp_disable_c = 0;
- disp_ttu_regs->min_ttu_vblank = (unsigned int)(l->min_ttu_vblank * l->refclk_freq_in_mhz);
-
- // CHECK for HW registers' range, DML2_ASSERT or clamp
- DML2_ASSERT(l->refcyc_per_req_delivery_pre_l < math_pow(2, 13));
- DML2_ASSERT(l->refcyc_per_req_delivery_l < math_pow(2, 13));
- DML2_ASSERT(l->refcyc_per_req_delivery_pre_c < math_pow(2, 13));
- DML2_ASSERT(l->refcyc_per_req_delivery_c < math_pow(2, 13));
- if (disp_dlg_regs->refcyc_per_vm_group_vblank >= (unsigned int)math_pow(2, 23))
- disp_dlg_regs->refcyc_per_vm_group_vblank = (unsigned int)(math_pow(2, 23) - 1);
-
- if (disp_dlg_regs->refcyc_per_vm_group_flip >= (unsigned int)math_pow(2, 23))
- disp_dlg_regs->refcyc_per_vm_group_flip = (unsigned int)(math_pow(2, 23) - 1);
-
- if (disp_dlg_regs->refcyc_per_vm_req_vblank >= (unsigned int)math_pow(2, 23))
- disp_dlg_regs->refcyc_per_vm_req_vblank = (unsigned int)(math_pow(2, 23) - 1);
-
- if (disp_dlg_regs->refcyc_per_vm_req_flip >= (unsigned int)math_pow(2, 23))
- disp_dlg_regs->refcyc_per_vm_req_flip = (unsigned int)(math_pow(2, 23) - 1);
-
-
- DML2_ASSERT(disp_dlg_regs->dst_y_after_scaler < (unsigned int)8);
- DML2_ASSERT(disp_dlg_regs->refcyc_x_after_scaler < (unsigned int)math_pow(2, 13));
-
- if (disp_dlg_regs->dst_y_per_pte_row_nom_l >= (unsigned int)math_pow(2, 17)) {
- dml2_printf("DML_DLG: %s: Warning DST_Y_PER_PTE_ROW_NOM_L %u > register max U15.2 %u, clamp to max\n", __func__, disp_dlg_regs->dst_y_per_pte_row_nom_l, (unsigned int)math_pow(2, 17) - 1);
- l->dst_y_per_pte_row_nom_l = (unsigned int)math_pow(2, 17) - 1;
- }
- if (l->dual_plane) {
- if (disp_dlg_regs->dst_y_per_pte_row_nom_c >= (unsigned int)math_pow(2, 17)) {
- dml2_printf("DML_DLG: %s: Warning DST_Y_PER_PTE_ROW_NOM_C %u > register max U15.2 %u, clamp to max\n", __func__, disp_dlg_regs->dst_y_per_pte_row_nom_c, (unsigned int)math_pow(2, 17) - 1);
- l->dst_y_per_pte_row_nom_c = (unsigned int)math_pow(2, 17) - 1;
- }
- }
-
- if (disp_dlg_regs->refcyc_per_pte_group_nom_l >= (unsigned int)math_pow(2, 23))
- disp_dlg_regs->refcyc_per_pte_group_nom_l = (unsigned int)(math_pow(2, 23) - 1);
- if (l->dual_plane) {
- if (disp_dlg_regs->refcyc_per_pte_group_nom_c >= (unsigned int)math_pow(2, 23))
- disp_dlg_regs->refcyc_per_pte_group_nom_c = (unsigned int)(math_pow(2, 23) - 1);
- }
- DML2_ASSERT(disp_dlg_regs->refcyc_per_pte_group_vblank_l < (unsigned int)math_pow(2, 13));
- if (l->dual_plane) {
- DML2_ASSERT(disp_dlg_regs->refcyc_per_pte_group_vblank_c < (unsigned int)math_pow(2, 13));
- }
-
- DML2_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_pre_l < (unsigned int)math_pow(2, 13));
- DML2_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_l < (unsigned int)math_pow(2, 13));
- DML2_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_pre_c < (unsigned int)math_pow(2, 13));
- DML2_ASSERT(disp_dlg_regs->refcyc_per_line_delivery_c < (unsigned int)math_pow(2, 13));
- DML2_ASSERT(disp_ttu_regs->qos_level_low_wm < (unsigned int)math_pow(2, 14));
- DML2_ASSERT(disp_ttu_regs->qos_level_high_wm < (unsigned int)math_pow(2, 14));
- DML2_ASSERT(disp_ttu_regs->min_ttu_vblank < (unsigned int)math_pow(2, 24));
-
- dml2_printf("DML_DLG::%s: Calculation for pipe[%d] done\n", __func__, pipe_idx);
-
- }
-}
-
-static void rq_dlg_get_arb_params(const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_display_arb_regs *arb_param)
-{
- arb_param->max_req_outstanding = mode_lib->soc.max_outstanding_reqs;
- arb_param->min_req_outstanding = mode_lib->soc.max_outstanding_reqs; // turn off the sat level feature if this set to max
- arb_param->sdpif_request_rate_limit = (3 * mode_lib->ip.words_per_channel * mode_lib->soc.clk_table.dram_config.channel_count) / 4;
- arb_param->sdpif_request_rate_limit = arb_param->sdpif_request_rate_limit < 96 ? 96 : arb_param->sdpif_request_rate_limit;
- arb_param->sat_level_us = 60;
- arb_param->hvm_max_qos_commit_threshold = 0xf;
- arb_param->hvm_min_req_outstand_commit_threshold = 0xa;
- arb_param->compbuf_reserved_space_kbytes = mode_lib->mp.compbuf_reserved_space_64b * 64 / 1024;
- arb_param->allow_sdpif_rate_limit_when_cstate_req = mode_lib->mp.hw_debug5;
- arb_param->dcfclk_deep_sleep_hysteresis = mode_lib->mp.dcfclk_deep_sleep_hysteresis;
-
-#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: max_req_outstanding = %d\n", __func__, arb_param->max_req_outstanding);
- dml2_printf("DML::%s: sdpif_request_rate_limit = %d\n", __func__, arb_param->sdpif_request_rate_limit);
- dml2_printf("DML::%s: compbuf_reserved_space_kbytes = %d\n", __func__, arb_param->compbuf_reserved_space_kbytes);
- dml2_printf("DML::%s: allow_sdpif_rate_limit_when_cstate_req = %d\n", __func__, arb_param->allow_sdpif_rate_limit_when_cstate_req);
- dml2_printf("DML::%s: dcfclk_deep_sleep_hysteresis = %d\n", __func__, arb_param->dcfclk_deep_sleep_hysteresis);
-#endif
-
-}
-
-void dml2_core_shared_get_watermarks(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_dchub_watermark_regs *out)
-{
- rq_dlg_get_wm_regs(display_cfg, mode_lib, out);
-}
-
-void dml2_core_shared_get_arb_params(const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_display_arb_regs *out)
-{
- rq_dlg_get_arb_params(mode_lib, out);
-}
-
-void dml2_core_shared_get_pipe_regs(const struct dml2_display_cfg *display_cfg,
- struct dml2_core_internal_display_mode_lib *mode_lib,
- struct dml2_dchub_per_pipe_register_set *out, int pipe_index)
-{
- rq_dlg_get_rq_reg(&out->rq_regs, display_cfg, mode_lib, pipe_index);
- rq_dlg_get_dlg_reg(&mode_lib->scratch, &out->dlg_regs, &out->ttu_regs, display_cfg, mode_lib, pipe_index);
- out->det_size = mode_lib->mp.DETBufferSizeInKByte[mode_lib->mp.pipe_plane[pipe_index]] / mode_lib->ip.config_return_buffer_segment_size_in_kbytes;
-}
-
-void dml2_core_shared_get_stream_programming(const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_per_stream_programming *out, int pipe_index)
-{
- // out->min_clocks.dcn4x.dscclk_khz = (unsigned int)(dml_get_dscclk_calculated(mode_lib, pipe_index) * 1000); // FIXME_STAGE2
- // out->min_clocks.dcn4x.dtbclk_khz = (unsigned int)(dml_get_dscclk_calculated(mode_lib, pipe_index) * 1000);
- // out->min_clocks.dcn4x.phyclk_khz = (unsigned int)(dml_get_dscclk_calculated(mode_lib, pipe_index) * 1000);
-
- out->global_sync.dcn4x.vready_offset_pixels = mode_lib->mp.VReadyOffsetPix[mode_lib->mp.pipe_plane[pipe_index]];
- out->global_sync.dcn4x.vstartup_lines = mode_lib->mp.VStartup[mode_lib->mp.pipe_plane[pipe_index]];
- out->global_sync.dcn4x.vupdate_offset_pixels = mode_lib->mp.VUpdateOffsetPix[mode_lib->mp.pipe_plane[pipe_index]];
- out->global_sync.dcn4x.vupdate_vupdate_width_pixels = mode_lib->mp.VUpdateWidthPix[mode_lib->mp.pipe_plane[pipe_index]];
-}
-
-void dml2_core_shared_get_mcache_allocation(const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_mcache_surface_allocation *out, int plane_idx)
-{
- unsigned int n;
-
- out->num_mcaches_plane0 = mode_lib->ms.num_mcaches_l[plane_idx];
- out->num_mcaches_plane1 = mode_lib->ms.num_mcaches_c[plane_idx];
- out->shift_granularity.p0 = mode_lib->ms.mcache_shift_granularity_l[plane_idx];
- out->shift_granularity.p1 = mode_lib->ms.mcache_shift_granularity_c[plane_idx];
-
- for (n = 0; n < out->num_mcaches_plane0; n++)
- out->mcache_x_offsets_plane0[n] = mode_lib->ms.mcache_offsets_l[plane_idx][n];
-
- for (n = 0; n < out->num_mcaches_plane1; n++)
- out->mcache_x_offsets_plane1[n] = mode_lib->ms.mcache_offsets_l[plane_idx][n];
-
- out->last_slice_sharing.mall_comb_mcache_p0 = mode_lib->ms.mall_comb_mcache_l[plane_idx];
- out->last_slice_sharing.mall_comb_mcache_p1 = mode_lib->ms.mall_comb_mcache_c[plane_idx];
- out->last_slice_sharing.plane0_plane1 = mode_lib->ms.lc_comb_mcache[plane_idx];
- out->informative.meta_row_bytes_plane0 = mode_lib->ms.mcache_row_bytes_l[plane_idx];
- out->informative.meta_row_bytes_plane1 = mode_lib->ms.mcache_row_bytes_c[plane_idx];
-
- out->valid = true;
-}
-
-void dml2_core_shared_get_mall_allocation(struct dml2_core_internal_display_mode_lib *mode_lib, unsigned int *out, int pipe_index)
-{
- *out = mode_lib->mp.SurfaceSizeInTheMALL[mode_lib->mp.pipe_plane[pipe_index]];
-}
-
-void dml2_core_shared_get_plane_support_info(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, struct core_plane_support_info *out, int plane_idx)
-{
- out->mall_svp_size_requirement_ways = 0;
-
- out->nominal_vblank_pstate_latency_hiding_us =
- (int)(display_cfg->stream_descriptors[display_cfg->plane_descriptors[plane_idx].stream_index].timing.h_total /
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[plane_idx].stream_index].timing.pixel_clock_khz / 1000) * mode_lib->ms.TWait[plane_idx]);
-
- out->dram_change_latency_hiding_margin_in_active = (int)mode_lib->ms.VActiveLatencyHidingMargin[plane_idx];
-
- out->active_latency_hiding_us = (int)mode_lib->ms.VActiveLatencyHidingUs[plane_idx];
-}
-
-void dml2_core_shared_get_stream_support_info(const struct dml2_display_cfg *display_cfg, const struct dml2_core_internal_display_mode_lib *mode_lib, struct core_stream_support_info *out, int plane_index)
-{
- double phantom_processing_delay_pix;
- unsigned int phantom_processing_delay_lines;
- unsigned int phantom_v_active_lines;
- unsigned int phantom_v_startup_lines;
- unsigned int phantom_v_blank_lines;
- unsigned int main_v_blank_lines;
- unsigned int rem;
-
- phantom_processing_delay_pix = (double)((mode_lib->ip.subvp_fw_processing_delay_us + mode_lib->ip.subvp_pstate_allow_width_us) *
- ((double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[plane_index].stream_index].timing.pixel_clock_khz / 1000));
- phantom_processing_delay_lines = (unsigned int)(phantom_processing_delay_pix / (double)display_cfg->stream_descriptors[display_cfg->plane_descriptors[plane_index].stream_index].timing.h_total);
- dml2_core_shared_div_rem(phantom_processing_delay_pix, display_cfg->stream_descriptors[display_cfg->plane_descriptors[plane_index].stream_index].timing.h_total, &rem);
- if (rem)
- phantom_processing_delay_lines++;
-
- phantom_v_startup_lines = mode_lib->ms.MaxVStartupLines[plane_index];
- phantom_v_active_lines = phantom_processing_delay_lines + mode_lib->ms.SubViewportLinesNeededInMALL[plane_index] + mode_lib->ip.subvp_swath_height_margin_lines;
-
- // phantom_vblank = max(vbp(vstartup) + vactive + vfp(always 1) + vsync(can be 1), main_vblank)
- phantom_v_blank_lines = phantom_v_startup_lines + 1 + 1;
- main_v_blank_lines = display_cfg->stream_descriptors[display_cfg->plane_descriptors[plane_index].stream_index].timing.v_total - display_cfg->stream_descriptors[display_cfg->plane_descriptors[plane_index].stream_index].timing.v_active;
- if (phantom_v_blank_lines > main_v_blank_lines)
- phantom_v_blank_lines = main_v_blank_lines;
-
- out->phantom_v_active = phantom_v_active_lines;
- // phantom_vtotal = vactive + vblank
- out->phantom_v_total = phantom_v_active_lines + phantom_v_blank_lines;
-
- out->phantom_min_v_active = mode_lib->ms.SubViewportLinesNeededInMALL[plane_index];
- out->phantom_v_startup = mode_lib->ms.MaxVStartupLines[plane_index];
-
- out->vblank_reserved_time_us = display_cfg->plane_descriptors[plane_index].overrides.reserved_vblank_time_ns / 1000;
-#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: subvp_fw_processing_delay_us = %d\n", __func__, mode_lib->ip.subvp_fw_processing_delay_us);
- dml2_printf("DML::%s: subvp_pstate_allow_width_us = %d\n", __func__, mode_lib->ip.subvp_pstate_allow_width_us);
- dml2_printf("DML::%s: subvp_swath_height_margin_lines = %d\n", __func__, mode_lib->ip.subvp_swath_height_margin_lines);
- dml2_printf("DML::%s: vblank_reserved_time_us = %f\n", __func__, out->vblank_reserved_time_us);
-#endif
-}
-
-void dml2_core_shared_get_informative(const struct dml2_core_internal_display_mode_lib *mode_lib, struct dml2_display_cfg_programming *out)
-{
- unsigned int k, n;
-
- out->informative.mode_support_info.ModeIsSupported = mode_lib->ms.support.ModeSupport;
- out->informative.mode_support_info.ImmediateFlipSupport = mode_lib->ms.support.ImmediateFlipSupport;
- out->informative.mode_support_info.WritebackLatencySupport = mode_lib->ms.support.WritebackLatencySupport;
- out->informative.mode_support_info.ScaleRatioAndTapsSupport = mode_lib->ms.support.ScaleRatioAndTapsSupport;
- out->informative.mode_support_info.SourceFormatPixelAndScanSupport = mode_lib->ms.support.SourceFormatPixelAndScanSupport;
- out->informative.mode_support_info.P2IWith420 = mode_lib->ms.support.P2IWith420;
- out->informative.mode_support_info.DSCOnlyIfNecessaryWithBPP = mode_lib->ms.support.DSCOnlyIfNecessaryWithBPP;
- out->informative.mode_support_info.DSC422NativeNotSupported = mode_lib->ms.support.DSC422NativeNotSupported;
- out->informative.mode_support_info.LinkRateDoesNotMatchDPVersion = mode_lib->ms.support.LinkRateDoesNotMatchDPVersion;
- out->informative.mode_support_info.LinkRateForMultistreamNotIndicated = mode_lib->ms.support.LinkRateForMultistreamNotIndicated;
- out->informative.mode_support_info.BPPForMultistreamNotIndicated = mode_lib->ms.support.BPPForMultistreamNotIndicated;
- out->informative.mode_support_info.MultistreamWithHDMIOreDP = mode_lib->ms.support.MultistreamWithHDMIOreDP;
- out->informative.mode_support_info.MSOOrODMSplitWithNonDPLink = mode_lib->ms.support.MSOOrODMSplitWithNonDPLink;
- out->informative.mode_support_info.NotEnoughLanesForMSO = mode_lib->ms.support.NotEnoughLanesForMSO;
- out->informative.mode_support_info.NumberOfOTGSupport = mode_lib->ms.support.NumberOfOTGSupport;
- out->informative.mode_support_info.NumberOfHDMIFRLSupport = mode_lib->ms.support.NumberOfHDMIFRLSupport;
- out->informative.mode_support_info.NumberOfDP2p0Support = mode_lib->ms.support.NumberOfDP2p0Support;
- out->informative.mode_support_info.WritebackScaleRatioAndTapsSupport = mode_lib->ms.support.WritebackScaleRatioAndTapsSupport;
- out->informative.mode_support_info.CursorSupport = mode_lib->ms.support.CursorSupport;
- out->informative.mode_support_info.PitchSupport = mode_lib->ms.support.PitchSupport;
- out->informative.mode_support_info.ViewportExceedsSurface = mode_lib->ms.support.ViewportExceedsSurface;
- out->informative.mode_support_info.ImmediateFlipRequiredButTheRequirementForEachSurfaceIsNotSpecified = false;
- out->informative.mode_support_info.ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe = mode_lib->ms.support.ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe;
- out->informative.mode_support_info.InvalidCombinationOfMALLUseForPStateAndStaticScreen = mode_lib->ms.support.InvalidCombinationOfMALLUseForPStateAndStaticScreen;
- out->informative.mode_support_info.InvalidCombinationOfMALLUseForPState = mode_lib->ms.support.InvalidCombinationOfMALLUseForPState;
- out->informative.mode_support_info.ExceededMALLSize = mode_lib->ms.support.ExceededMALLSize;
- out->informative.mode_support_info.EnoughWritebackUnits = mode_lib->ms.support.EnoughWritebackUnits;
-
- out->informative.mode_support_info.ExceededMultistreamSlots = mode_lib->ms.support.ExceededMultistreamSlots;
- out->informative.mode_support_info.NotEnoughDSCUnits = mode_lib->ms.support.NotEnoughDSCUnits;
- out->informative.mode_support_info.NotEnoughDSCSlices = mode_lib->ms.support.NotEnoughDSCSlices;
- out->informative.mode_support_info.PixelsPerLinePerDSCUnitSupport = mode_lib->ms.support.PixelsPerLinePerDSCUnitSupport;
- out->informative.mode_support_info.DSCCLKRequiredMoreThanSupported = mode_lib->ms.support.DSCCLKRequiredMoreThanSupported;
- out->informative.mode_support_info.DTBCLKRequiredMoreThanSupported = mode_lib->ms.support.DTBCLKRequiredMoreThanSupported;
- out->informative.mode_support_info.LinkCapacitySupport = mode_lib->ms.support.LinkCapacitySupport;
-
- out->informative.mode_support_info.ROBSupport = mode_lib->ms.support.ROBSupport;
- out->informative.mode_support_info.OutstandingRequestsSupport = mode_lib->ms.support.OutstandingRequestsSupport;
- out->informative.mode_support_info.OutstandingRequestsUrgencyAvoidance = mode_lib->ms.support.OutstandingRequestsUrgencyAvoidance;
- out->informative.mode_support_info.PTEBufferSizeNotExceeded = mode_lib->ms.support.PTEBufferSizeNotExceeded;
- out->informative.mode_support_info.DCCMetaBufferSizeNotExceeded = mode_lib->ms.support.DCCMetaBufferSizeNotExceeded;
-
- out->informative.mode_support_info.TotalVerticalActiveBandwidthSupport = mode_lib->ms.support.AvgBandwidthSupport;
- out->informative.mode_support_info.VActiveBandwidthSupport = mode_lib->ms.support.UrgVactiveBandwidthSupport;
- out->informative.mode_support_info.USRRetrainingSupport = mode_lib->ms.support.USRRetrainingSupport;
-
- out->informative.mode_support_info.PrefetchSupported = mode_lib->ms.support.PrefetchSupported;
- out->informative.mode_support_info.DynamicMetadataSupported = mode_lib->ms.support.DynamicMetadataSupported;
- out->informative.mode_support_info.VRatioInPrefetchSupported = mode_lib->ms.support.VRatioInPrefetchSupported;
- out->informative.mode_support_info.DISPCLK_DPPCLK_Support = mode_lib->ms.support.DISPCLK_DPPCLK_Support;
- out->informative.mode_support_info.TotalAvailablePipesSupport = mode_lib->ms.support.TotalAvailablePipesSupport;
- out->informative.mode_support_info.ViewportSizeSupport = mode_lib->ms.support.ViewportSizeSupport;
-
- for (k = 0; k < out->display_config.num_planes; k++) {
-
- out->informative.mode_support_info.FCLKChangeSupport[k] = mode_lib->ms.support.FCLKChangeSupport[k];
- out->informative.mode_support_info.MPCCombineEnable[k] = mode_lib->ms.support.MPCCombineEnable[k];
- out->informative.mode_support_info.ODMMode[k] = mode_lib->ms.support.ODMMode[k];
- out->informative.mode_support_info.DPPPerSurface[k] = mode_lib->ms.support.DPPPerSurface[k];
- out->informative.mode_support_info.DSCEnabled[k] = mode_lib->ms.support.DSCEnabled[k];
- out->informative.mode_support_info.FECEnabled[k] = mode_lib->ms.support.FECEnabled[k];
- out->informative.mode_support_info.NumberOfDSCSlices[k] = mode_lib->ms.support.NumberOfDSCSlices[k];
- out->informative.mode_support_info.OutputBpp[k] = mode_lib->ms.support.OutputBpp[k];
-
- if (mode_lib->ms.support.OutputType[k] == dml2_core_internal_output_type_unknown)
- out->informative.mode_support_info.OutputType[k] = dml2_output_type_unknown;
- else if (mode_lib->ms.support.OutputType[k] == dml2_core_internal_output_type_dp)
- out->informative.mode_support_info.OutputType[k] = dml2_output_type_dp;
- else if (mode_lib->ms.support.OutputType[k] == dml2_core_internal_output_type_edp)
- out->informative.mode_support_info.OutputType[k] = dml2_output_type_edp;
- else if (mode_lib->ms.support.OutputType[k] == dml2_core_internal_output_type_dp2p0)
- out->informative.mode_support_info.OutputType[k] = dml2_output_type_dp2p0;
- else if (mode_lib->ms.support.OutputType[k] == dml2_core_internal_output_type_hdmi)
- out->informative.mode_support_info.OutputType[k] = dml2_output_type_hdmi;
- else if (mode_lib->ms.support.OutputType[k] == dml2_core_internal_output_type_hdmifrl)
- out->informative.mode_support_info.OutputType[k] = dml2_output_type_hdmifrl;
-
- if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_unknown)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_unknown;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_dp_rate_hbr)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_dp_rate_hbr;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_dp_rate_hbr2)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_dp_rate_hbr2;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_dp_rate_hbr3)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_dp_rate_hbr3;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_dp_rate_uhbr10)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_dp_rate_uhbr10;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_dp_rate_uhbr13p5)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_dp_rate_uhbr13p5;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_dp_rate_uhbr20)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_dp_rate_uhbr20;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_hdmi_rate_3x3)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_3x3;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_hdmi_rate_6x3)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_6x3;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_hdmi_rate_6x4)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_6x4;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_hdmi_rate_8x4)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_8x4;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_hdmi_rate_10x4)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_10x4;
- else if (mode_lib->ms.support.OutputRate[k] == dml2_core_internal_output_rate_hdmi_rate_12x4)
- out->informative.mode_support_info.OutputRate[k] = dml2_output_rate_hdmi_rate_12x4;
-
- out->informative.mode_support_info.AlignedYPitch[k] = mode_lib->ms.support.AlignedYPitch[k];
- out->informative.mode_support_info.AlignedCPitch[k] = mode_lib->ms.support.AlignedCPitch[k];
- }
-
- out->informative.watermarks.urgent_us = mode_lib->mp.Watermark.UrgentWatermark;
- out->informative.watermarks.writeback_urgent_us = mode_lib->mp.Watermark.WritebackUrgentWatermark;
- out->informative.watermarks.writeback_pstate_us = mode_lib->mp.Watermark.WritebackDRAMClockChangeWatermark;
- out->informative.watermarks.writeback_fclk_pstate_us = mode_lib->mp.Watermark.WritebackFCLKChangeWatermark;
-
- out->informative.watermarks.cstate_exit_us = mode_lib->mp.Watermark.StutterExitWatermark;
- out->informative.watermarks.cstate_enter_plus_exit_us = mode_lib->mp.Watermark.StutterEnterPlusExitWatermark;
- out->informative.watermarks.z8_cstate_exit_us = mode_lib->mp.Watermark.Z8StutterExitWatermark;
- out->informative.watermarks.z8_cstate_enter_plus_exit_us = mode_lib->mp.Watermark.Z8StutterEnterPlusExitWatermark;
- out->informative.watermarks.pstate_change_us = mode_lib->mp.Watermark.DRAMClockChangeWatermark;
- out->informative.watermarks.fclk_pstate_change_us = mode_lib->mp.Watermark.FCLKChangeWatermark;
- out->informative.watermarks.usr_retraining_us = mode_lib->mp.Watermark.USRRetrainingWatermark;
-
- out->informative.mall.total_surface_size_in_mall_bytes = 0;
- for (k = 0; k < out->display_config.num_planes; ++k)
- out->informative.mall.total_surface_size_in_mall_bytes += mode_lib->mp.SurfaceSizeInTheMALL[k];
-
- out->informative.qos.min_return_latency_in_dcfclk = mode_lib->mp.min_return_latency_in_dcfclk;
- out->informative.qos.urgent_latency_us = mode_lib->mp.UrgentLatency;
-
- out->informative.qos.max_urgent_latency_us = mode_lib->ms.support.max_urgent_latency_us;
- out->informative.qos.avg_non_urgent_latency_us = mode_lib->ms.support.avg_non_urgent_latency_us;
- out->informative.qos.avg_urgent_latency_us = mode_lib->ms.support.avg_urgent_latency_us;
-
- out->informative.qos.wm_memory_trip_us = mode_lib->mp.UrgentLatency;
- out->informative.qos.meta_trip_memory_us = mode_lib->mp.MetaTripToMemory;
- out->informative.qos.fraction_of_urgent_bandwidth = mode_lib->mp.FractionOfUrgentBandwidth;
- out->informative.qos.fraction_of_urgent_bandwidth_immediate_flip = mode_lib->mp.FractionOfUrgentBandwidthImmediateFlip;
- out->informative.qos.fraction_of_urgent_bandwidth_mall = mode_lib->mp.FractionOfUrgentBandwidthMALL;
-
- out->informative.qos.avg_bw_required.sys_active.sdp_bw_mbps =
- mode_lib->ms.support.avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- out->informative.qos.avg_bw_required.sys_active.dram_bw_mbps =
- mode_lib->ms.support.avg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
- out->informative.qos.avg_bw_required.svp_prefetch.sdp_bw_mbps =
- mode_lib->ms.support.avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- out->informative.qos.avg_bw_required.svp_prefetch.dram_bw_mbps =
- mode_lib->ms.support.avg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
-
- out->informative.qos.avg_bw_available.sys_active.sdp_bw_mbps =
- mode_lib->mp.avg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- out->informative.qos.avg_bw_available.sys_active.dram_bw_mbps =
- mode_lib->mp.avg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
- out->informative.qos.avg_bw_available.svp_prefetch.sdp_bw_mbps =
- mode_lib->mp.avg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- out->informative.qos.avg_bw_available.svp_prefetch.dram_bw_mbps =
- mode_lib->mp.avg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
-
- out->informative.qos.urg_bw_available.sys_active.sdp_bw_mbps =
- mode_lib->mp.urg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- out->informative.qos.urg_bw_available.sys_active.dram_bw_mbps =
- mode_lib->mp.urg_bandwidth_available[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
- out->informative.qos.urg_bw_available.sys_active.dram_vm_only_bw_mbps =
- mode_lib->mp.urg_bandwidth_available_vm_only[dml2_core_internal_soc_state_sys_active];
-
- out->informative.qos.urg_bw_available.svp_prefetch.sdp_bw_mbps =
- mode_lib->mp.urg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- out->informative.qos.urg_bw_available.svp_prefetch.dram_bw_mbps =
- mode_lib->mp.urg_bandwidth_available[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
- out->informative.qos.urg_bw_available.svp_prefetch.dram_vm_only_bw_mbps =
- mode_lib->mp.urg_bandwidth_available_vm_only[dml2_core_internal_soc_state_svp_prefetch];
-
- out->informative.qos.urg_bw_required.sys_active.sdp_bw_mbps = mode_lib->mp.urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- out->informative.qos.urg_bw_required.sys_active.dram_bw_mbps = mode_lib->mp.urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
- out->informative.qos.urg_bw_required.svp_prefetch.sdp_bw_mbps = mode_lib->mp.urg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- out->informative.qos.urg_bw_required.svp_prefetch.dram_bw_mbps = mode_lib->mp.urg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
-
- out->informative.qos.non_urg_bw_required.sys_active.sdp_bw_mbps = mode_lib->mp.non_urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- out->informative.qos.non_urg_bw_required.sys_active.dram_bw_mbps = mode_lib->mp.non_urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
- out->informative.qos.non_urg_bw_required.svp_prefetch.sdp_bw_mbps = mode_lib->mp.non_urg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- out->informative.qos.non_urg_bw_required.svp_prefetch.dram_bw_mbps = mode_lib->mp.non_urg_bandwidth_required[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
-
- out->informative.qos.urg_bw_required_with_flip.sys_active.sdp_bw_mbps = mode_lib->mp.urg_bandwidth_required_flip[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- out->informative.qos.urg_bw_required_with_flip.sys_active.dram_bw_mbps = mode_lib->mp.urg_bandwidth_required_flip[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
- out->informative.qos.urg_bw_required_with_flip.svp_prefetch.sdp_bw_mbps = mode_lib->mp.urg_bandwidth_required_flip[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- out->informative.qos.urg_bw_required_with_flip.svp_prefetch.dram_bw_mbps = mode_lib->mp.urg_bandwidth_required_flip[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
-
- out->informative.qos.non_urg_bw_required_with_flip.sys_active.sdp_bw_mbps = mode_lib->mp.non_urg_bandwidth_required_flip[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp];
- out->informative.qos.non_urg_bw_required_with_flip.sys_active.dram_bw_mbps = mode_lib->mp.non_urg_bandwidth_required_flip[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_dram];
- out->informative.qos.non_urg_bw_required_with_flip.svp_prefetch.sdp_bw_mbps = mode_lib->mp.non_urg_bandwidth_required_flip[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_sdp];
- out->informative.qos.non_urg_bw_required_with_flip.svp_prefetch.dram_bw_mbps = mode_lib->mp.non_urg_bandwidth_required_flip[dml2_core_internal_soc_state_svp_prefetch][dml2_core_internal_bw_dram];
-
- out->informative.crb.comp_buffer_size_kbytes = mode_lib->mp.CompressedBufferSizeInkByte;
- out->informative.crb.UnboundedRequestEnabled = mode_lib->mp.UnboundedRequestEnabled;
-
- out->informative.crb.compbuf_reserved_space_64b = mode_lib->mp.compbuf_reserved_space_64b;
- out->informative.misc.hw_debug5 = mode_lib->mp.hw_debug5;
- out->informative.misc.dcfclk_deep_sleep_hysteresis = mode_lib->mp.dcfclk_deep_sleep_hysteresis;
-
- out->informative.power_management.stutter_efficiency = mode_lib->mp.StutterEfficiencyNotIncludingVBlank;
- out->informative.power_management.stutter_efficiency_with_vblank = mode_lib->mp.StutterEfficiency;
- out->informative.power_management.stutter_num_bursts = mode_lib->mp.NumberOfStutterBurstsPerFrame;
-
- out->informative.power_management.z8.stutter_efficiency = mode_lib->mp.Z8StutterEfficiency;
- out->informative.power_management.z8.stutter_efficiency_with_vblank = mode_lib->mp.StutterEfficiency;
- out->informative.power_management.z8.stutter_num_bursts = mode_lib->mp.Z8NumberOfStutterBurstsPerFrame;
- out->informative.power_management.z8.stutter_period = mode_lib->mp.StutterPeriod;
-
- out->informative.power_management.z8.bestcase.stutter_efficiency = mode_lib->mp.Z8StutterEfficiencyBestCase;
- out->informative.power_management.z8.bestcase.stutter_num_bursts = mode_lib->mp.Z8NumberOfStutterBurstsPerFrameBestCase;
- out->informative.power_management.z8.bestcase.stutter_period = mode_lib->mp.StutterPeriodBestCase;
-
- out->informative.misc.cstate_max_cap_mode = mode_lib->mp.DCHUBBUB_ARB_CSTATE_MAX_CAP_MODE;
-
- out->min_clocks.dcn4x.dpprefclk_khz = (int unsigned)(mode_lib->mp.GlobalDPPCLK * 1000.0);
-
- out->informative.qos.max_active_fclk_change_latency_supported = mode_lib->mp.MaxActiveFCLKChangeLatencySupported;
-
- for (k = 0; k < out->display_config.num_planes; k++) {
-
- if ((out->display_config.plane_descriptors->overrides.reserved_vblank_time_ns >= 1000.0 * mode_lib->soc.power_management_parameters.dram_clk_change_blackout_us)
- && (out->display_config.plane_descriptors->overrides.reserved_vblank_time_ns >= 1000.0 * mode_lib->soc.power_management_parameters.fclk_change_blackout_us)
- && (out->display_config.plane_descriptors->overrides.reserved_vblank_time_ns >= 1000.0 * mode_lib->soc.power_management_parameters.stutter_enter_plus_exit_latency_us))
- out->informative.misc.PrefetchMode[k] = 0;
- else if ((out->display_config.plane_descriptors->overrides.reserved_vblank_time_ns >= 1000.0 * mode_lib->soc.power_management_parameters.fclk_change_blackout_us)
- && (out->display_config.plane_descriptors->overrides.reserved_vblank_time_ns >= 1000.0 * mode_lib->soc.power_management_parameters.stutter_enter_plus_exit_latency_us))
- out->informative.misc.PrefetchMode[k] = 1;
- else if (out->display_config.plane_descriptors->overrides.reserved_vblank_time_ns >= 1000.0 * mode_lib->soc.power_management_parameters.stutter_enter_plus_exit_latency_us)
- out->informative.misc.PrefetchMode[k] = 2;
- else
- out->informative.misc.PrefetchMode[k] = 3;
-
- out->informative.misc.min_ttu_vblank_us[k] = mode_lib->mp.MinTTUVBlank[k];
- out->informative.mall.subviewport_lines_needed_in_mall[k] = mode_lib->mp.SubViewportLinesNeededInMALL[k];
- out->informative.crb.det_size_in_kbytes[k] = mode_lib->mp.DETBufferSizeInKByte[k];
- out->informative.crb.DETBufferSizeY[k] = mode_lib->mp.DETBufferSizeY[k];
- out->informative.misc.ImmediateFlipSupportedForPipe[k] = mode_lib->mp.ImmediateFlipSupportedForPipe[k];
- out->informative.misc.UsesMALLForStaticScreen[k] = mode_lib->mp.is_using_mall_for_ss[k];
- out->informative.plane_info[k].dpte_row_height_plane0 = mode_lib->mp.dpte_row_height[k];
- out->informative.plane_info[k].dpte_row_height_plane1 = mode_lib->mp.dpte_row_height_chroma[k];
- out->informative.plane_info[k].meta_row_height_plane0 = mode_lib->mp.meta_row_height[k];
- out->informative.plane_info[k].meta_row_height_plane1 = mode_lib->mp.meta_row_height_chroma[k];
- out->informative.dcc_control[k].max_uncompressed_block_plane0 = mode_lib->mp.DCCYMaxUncompressedBlock[k];
- out->informative.dcc_control[k].max_compressed_block_plane0 = mode_lib->mp.DCCYMaxCompressedBlock[k];
- out->informative.dcc_control[k].independent_block_plane0 = mode_lib->mp.DCCYIndependentBlock[k];
- out->informative.dcc_control[k].max_uncompressed_block_plane1 = mode_lib->mp.DCCCMaxUncompressedBlock[k];
- out->informative.dcc_control[k].max_compressed_block_plane1 = mode_lib->mp.DCCCMaxCompressedBlock[k];
- out->informative.dcc_control[k].independent_block_plane1 = mode_lib->mp.DCCCIndependentBlock[k];
- out->informative.misc.dst_x_after_scaler[k] = mode_lib->mp.DSTXAfterScaler[k];
- out->informative.misc.dst_y_after_scaler[k] = mode_lib->mp.DSTYAfterScaler[k];
- out->informative.misc.prefetch_source_lines_plane0[k] = mode_lib->mp.PrefetchSourceLinesY[k];
- out->informative.misc.prefetch_source_lines_plane1[k] = mode_lib->mp.PrefetchSourceLinesC[k];
- out->informative.misc.vready_at_or_after_vsync[k] = mode_lib->mp.VREADY_AT_OR_AFTER_VSYNC[k];
- out->informative.misc.min_dst_y_next_start[k] = mode_lib->mp.MIN_DST_Y_NEXT_START[k];
- out->informative.plane_info[k].swath_width_plane0 = mode_lib->mp.SwathWidthY[k];
- out->informative.plane_info[k].swath_height_plane0 = mode_lib->mp.SwathHeightY[k];
- out->informative.plane_info[k].swath_height_plane1 = mode_lib->mp.SwathHeightC[k];
- out->informative.misc.CursorDstXOffset[k] = mode_lib->mp.CursorDstXOffset[k];
- out->informative.misc.CursorDstYOffset[k] = mode_lib->mp.CursorDstYOffset[k];
- out->informative.misc.CursorChunkHDLAdjust[k] = mode_lib->mp.CursorChunkHDLAdjust[k];
- out->informative.misc.dpte_group_bytes[k] = mode_lib->mp.dpte_group_bytes[k];
- out->informative.misc.vm_group_bytes[k] = mode_lib->mp.vm_group_bytes[k];
- out->informative.misc.DisplayPipeRequestDeliveryTimeLuma[k] = mode_lib->mp.DisplayPipeRequestDeliveryTimeLuma[k];
- out->informative.misc.DisplayPipeRequestDeliveryTimeChroma[k] = mode_lib->mp.DisplayPipeRequestDeliveryTimeChroma[k];
- out->informative.misc.DisplayPipeRequestDeliveryTimeLumaPrefetch[k] = mode_lib->mp.DisplayPipeRequestDeliveryTimeLumaPrefetch[k];
- out->informative.misc.DisplayPipeRequestDeliveryTimeChromaPrefetch[k] = mode_lib->mp.DisplayPipeRequestDeliveryTimeChromaPrefetch[k];
- out->informative.misc.TimePerVMGroupVBlank[k] = mode_lib->mp.TimePerVMGroupVBlank[k];
- out->informative.misc.TimePerVMGroupFlip[k] = mode_lib->mp.TimePerVMGroupFlip[k];
- out->informative.misc.TimePerVMRequestVBlank[k] = mode_lib->mp.TimePerVMRequestVBlank[k];
- out->informative.misc.TimePerVMRequestFlip[k] = mode_lib->mp.TimePerVMRequestFlip[k];
- out->informative.misc.Tdmdl_vm[k] = mode_lib->mp.Tdmdl_vm[k];
- out->informative.misc.Tdmdl[k] = mode_lib->mp.Tdmdl[k];
- out->informative.misc.VStartup[k] = mode_lib->mp.VStartup[k];
- out->informative.misc.VUpdateOffsetPix[k] = mode_lib->mp.VUpdateOffsetPix[k];
- out->informative.misc.VUpdateWidthPix[k] = mode_lib->mp.VUpdateWidthPix[k];
- out->informative.misc.VReadyOffsetPix[k] = mode_lib->mp.VReadyOffsetPix[k];
-
- out->informative.misc.DST_Y_PER_PTE_ROW_NOM_L[k] = mode_lib->mp.DST_Y_PER_PTE_ROW_NOM_L[k];
- out->informative.misc.DST_Y_PER_PTE_ROW_NOM_C[k] = mode_lib->mp.DST_Y_PER_PTE_ROW_NOM_C[k];
- out->informative.misc.time_per_pte_group_nom_luma[k] = mode_lib->mp.time_per_pte_group_nom_luma[k];
- out->informative.misc.time_per_pte_group_nom_chroma[k] = mode_lib->mp.time_per_pte_group_nom_chroma[k];
- out->informative.misc.time_per_pte_group_vblank_luma[k] = mode_lib->mp.time_per_pte_group_vblank_luma[k];
- out->informative.misc.time_per_pte_group_vblank_chroma[k] = mode_lib->mp.time_per_pte_group_vblank_chroma[k];
- out->informative.misc.time_per_pte_group_flip_luma[k] = mode_lib->mp.time_per_pte_group_flip_luma[k];
- out->informative.misc.time_per_pte_group_flip_chroma[k] = mode_lib->mp.time_per_pte_group_flip_chroma[k];
- out->informative.misc.VRatioPrefetchY[k] = mode_lib->mp.VRatioPrefetchY[k];
- out->informative.misc.VRatioPrefetchC[k] = mode_lib->mp.VRatioPrefetchC[k];
- out->informative.misc.DestinationLinesForPrefetch[k] = mode_lib->mp.dst_y_prefetch[k];
- out->informative.misc.DestinationLinesToRequestVMInVBlank[k] = mode_lib->mp.dst_y_per_vm_vblank[k];
- out->informative.misc.DestinationLinesToRequestRowInVBlank[k] = mode_lib->mp.dst_y_per_row_vblank[k];
- out->informative.misc.DestinationLinesToRequestVMInImmediateFlip[k] = mode_lib->mp.dst_y_per_vm_flip[k];
- out->informative.misc.DestinationLinesToRequestRowInImmediateFlip[k] = mode_lib->mp.dst_y_per_row_flip[k];
- out->informative.misc.DisplayPipeLineDeliveryTimeLuma[k] = mode_lib->mp.DisplayPipeLineDeliveryTimeLuma[k];
- out->informative.misc.DisplayPipeLineDeliveryTimeChroma[k] = mode_lib->mp.DisplayPipeLineDeliveryTimeChroma[k];
- out->informative.misc.DisplayPipeLineDeliveryTimeLumaPrefetch[k] = mode_lib->mp.DisplayPipeLineDeliveryTimeLumaPrefetch[k];
- out->informative.misc.DisplayPipeLineDeliveryTimeChromaPrefetch[k] = mode_lib->mp.DisplayPipeLineDeliveryTimeChromaPrefetch[k];
-
- out->informative.misc.WritebackAllowDRAMClockChangeEndPosition[k] = mode_lib->mp.WritebackAllowDRAMClockChangeEndPosition[k];
- out->informative.misc.WritebackAllowFCLKChangeEndPosition[k] = mode_lib->mp.WritebackAllowFCLKChangeEndPosition[k];
- out->informative.misc.DSCCLK_calculated[k] = mode_lib->mp.DSCCLK[k];
- out->informative.misc.BIGK_FRAGMENT_SIZE[k] = mode_lib->mp.BIGK_FRAGMENT_SIZE[k];
- out->informative.misc.PTE_BUFFER_MODE[k] = mode_lib->mp.PTE_BUFFER_MODE[k];
- out->informative.misc.DSCDelay[k] = mode_lib->mp.DSCDelay[k];
- out->informative.misc.MaxActiveDRAMClockChangeLatencySupported[k] = mode_lib->mp.MaxActiveDRAMClockChangeLatencySupported[k];
- }
-
- // For this DV informative layer, all pipes in the same planes will just use the same id
- // will have the optimization and helper layer later on
- // only work when we can have high "mcache" that fit everything without thrashing the cache
- for (k = 0; k < out->display_config.num_planes; k++) {
- out->informative.non_optimized_mcache_allocation[k].num_mcaches_plane0 = mode_lib->ms.num_mcaches_l[k];
- out->informative.non_optimized_mcache_allocation[k].informative.meta_row_bytes_plane0 = mode_lib->ms.mcache_row_bytes_l[k];
-
- for (n = 0; n < out->informative.non_optimized_mcache_allocation[k].num_mcaches_plane0; n++) {
- out->informative.non_optimized_mcache_allocation[k].mcache_x_offsets_plane0[n] = mode_lib->ms.mcache_offsets_l[k][n];
- out->informative.non_optimized_mcache_allocation[k].global_mcache_ids_plane0[n] = k;
- }
-
- out->informative.non_optimized_mcache_allocation[k].num_mcaches_plane1 = mode_lib->ms.num_mcaches_c[k];
- out->informative.non_optimized_mcache_allocation[k].informative.meta_row_bytes_plane1 = mode_lib->ms.mcache_row_bytes_c[k];
-
- for (n = 0; n < out->informative.non_optimized_mcache_allocation[k].num_mcaches_plane1; n++) {
- out->informative.non_optimized_mcache_allocation[k].mcache_x_offsets_plane1[n] = mode_lib->ms.mcache_offsets_c[k][n];
- out->informative.non_optimized_mcache_allocation[k].global_mcache_ids_plane1[n] = k;
- }
- }
-
- out->informative.qos.max_non_urgent_latency_us = mode_lib->soc.qos_parameters.qos_params.dcn4x.per_uclk_dpm_params[mode_lib->mp.qos_param_index].maximum_latency_when_non_urgent_uclk_cycles
- / mode_lib->mp.uclk_freq_mhz * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.umc_max_latency_margin / 100.0)
- + mode_lib->soc.qos_parameters.qos_params.dcn4x.mall_overhead_fclk_cycles / mode_lib->mp.FabricClock
- + mode_lib->soc.qos_parameters.qos_params.dcn4x.max_round_trip_to_furthest_cs_fclk_cycles / mode_lib->mp.FabricClock
- * (1 + mode_lib->soc.qos_parameters.qos_params.dcn4x.fabric_max_transport_latency_margin / 100.0);
-
- if (mode_lib->soc.qos_parameters.qos_type == dml2_qos_param_type_dcn4x) {
- if (((mode_lib->ip.rob_buffer_size_kbytes - mode_lib->ip.pixel_chunk_size_kbytes) * 1024
- / mode_lib->mp.non_urg_bandwidth_required[dml2_core_internal_soc_state_sys_active][dml2_core_internal_bw_sdp]) >= out->informative.qos.max_non_urgent_latency_us) {
- out->informative.misc.ROBUrgencyAvoidance = true;
- } else {
- out->informative.misc.ROBUrgencyAvoidance = false;
- }
- } else {
- out->informative.misc.ROBUrgencyAvoidance = true;
- }
-}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_shared_types.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_shared_types.h
index cbdfbd5a0bde..28687565ac22 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_shared_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_shared_types.h
@@ -102,6 +102,7 @@ struct dml2_core_internal_DmlPipe {
double DCFClkDeepSleep;
unsigned int DPPPerSurface;
bool ScalerEnabled;
+ bool UPSPEnabled;
enum dml2_rotation_angle RotationAngle;
bool mirrored;
unsigned int ViewportHeight;
@@ -186,7 +187,9 @@ enum dml2_core_internal_output_type_rate {
dml2_core_internal_output_rate_hdmi_rate_6x4 = 9,
dml2_core_internal_output_rate_hdmi_rate_8x4 = 10,
dml2_core_internal_output_rate_hdmi_rate_10x4 = 11,
- dml2_core_internal_output_rate_hdmi_rate_12x4 = 12
+ dml2_core_internal_output_rate_hdmi_rate_12x4 = 12,
+ dml2_core_internal_output_rate_hdmi_rate_16x4 = 13,
+ dml2_core_internal_output_rate_hdmi_rate_20x4 = 14
};
struct dml2_core_internal_watermarks {
@@ -201,7 +204,7 @@ struct dml2_core_internal_watermarks {
double Z8StutterExitWatermark;
double Z8StutterEnterPlusExitWatermark;
double USRRetrainingWatermark;
- double g6_temp_read_watermark_us;
+ double temp_read_or_ppt_watermark_us;
};
struct dml2_core_internal_mode_support_info {
@@ -252,20 +255,22 @@ struct dml2_core_internal_mode_support_info {
bool PTEBufferSizeNotExceeded;
bool DCCMetaBufferSizeNotExceeded;
- enum dml2_dram_clock_change_support DRAMClockChangeSupport[DML2_MAX_PLANES];
- enum dml2_fclock_change_support FCLKChangeSupport[DML2_MAX_PLANES];
+ enum dml2_pstate_change_support DRAMClockChangeSupport[DML2_MAX_PLANES];
+ enum dml2_pstate_change_support FCLKChangeSupport[DML2_MAX_PLANES];
bool global_dram_clock_change_supported;
bool global_fclk_change_supported;
bool USRRetrainingSupport;
bool AvgBandwidthSupport;
bool UrgVactiveBandwidthSupport;
bool EnoughUrgentLatencyHidingSupport;
+ bool PrefetchScheduleSupported;
bool PrefetchSupported;
bool PrefetchBandwidthSupported;
bool DynamicMetadataSupported;
bool VRatioInPrefetchSupported;
bool DISPCLK_DPPCLK_Support;
bool TotalAvailablePipesSupport;
+ bool ODMSupport;
bool ModeSupport;
bool ViewportSizeSupport;
@@ -314,18 +319,21 @@ struct dml2_core_internal_mode_support_info {
double non_urg_bandwidth_required[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max]; // same as urg_bandwidth, except not scaled by urg burst factor
double non_urg_bandwidth_required_flip[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max];
-
bool avg_bandwidth_support_ok[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max];
-
double max_urgent_latency_us;
+ double max_non_urgent_latency_us;
double avg_non_urgent_latency_us;
double avg_urgent_latency_us;
+ double df_response_time_us;
bool incorrect_imall_usage;
bool g6_temp_read_support;
+ bool temp_read_or_ppt_support;
struct dml2_core_internal_watermarks watermarks;
+ bool dcfclk_support;
+ bool qos_bandwidth_support;
};
struct dml2_core_internal_mode_support {
@@ -347,9 +355,11 @@ struct dml2_core_internal_mode_support {
double SOCCLK; /// <brief Basically just the clock freq at the min (or given) state
double DCFCLK; /// <brief Basically just the clock freq at the min (or given) state and max combine setting
double GlobalDPPCLK; /// <brief the Max DPPCLK freq out of all pipes
+ double GlobalDTBCLK; /// <brief the Max DTBCLK freq out of all pipes
double uclk_freq_mhz;
double dram_bw_mbps;
double max_dram_bw_mbps;
+ double min_available_urgent_bandwidth_MBps; /// <brief Minimum guaranteed available urgent return bandwidth in MBps
double MaxFabricClock; /// <brief Basically just the clock freq at the min (or given) state
double MaxDCFCLK; /// <brief Basically just the clock freq at the min (or given) state and max combine setting
@@ -378,8 +388,8 @@ struct dml2_core_internal_mode_support {
unsigned int DETBufferSizeC[DML2_MAX_PLANES];
unsigned int SwathHeightY[DML2_MAX_PLANES];
unsigned int SwathHeightC[DML2_MAX_PLANES];
- unsigned int SwathWidthY[DML2_MAX_PLANES];
- unsigned int SwathWidthC[DML2_MAX_PLANES];
+ unsigned int SwathWidthY[DML2_MAX_PLANES]; // per-pipe
+ unsigned int SwathWidthC[DML2_MAX_PLANES]; // per-pipe
// ----------------------------------
// Intermediates/Informational
@@ -391,9 +401,13 @@ struct dml2_core_internal_mode_support {
double TWait[DML2_MAX_PLANES];
bool UnboundedRequestEnabled;
+ unsigned int compbuf_reserved_space_64b;
+ bool hw_debug5;
unsigned int CompressedBufferSizeInkByte;
double VRatioPreY[DML2_MAX_PLANES];
double VRatioPreC[DML2_MAX_PLANES];
+ unsigned int req_per_swath_ub_l[DML2_MAX_PLANES];
+ unsigned int req_per_swath_ub_c[DML2_MAX_PLANES];
unsigned int swath_width_luma_ub[DML2_MAX_PLANES];
unsigned int swath_width_chroma_ub[DML2_MAX_PLANES];
unsigned int RequiredSlots[DML2_MAX_PLANES];
@@ -414,8 +428,8 @@ struct dml2_core_internal_mode_support {
double dst_y_prefetch[DML2_MAX_PLANES];
double LinesForVM[DML2_MAX_PLANES];
double LinesForDPTERow[DML2_MAX_PLANES];
- double SwathWidthYSingleDPP[DML2_MAX_PLANES];
- double SwathWidthCSingleDPP[DML2_MAX_PLANES];
+ unsigned int SwathWidthYSingleDPP[DML2_MAX_PLANES];
+ unsigned int SwathWidthCSingleDPP[DML2_MAX_PLANES];
unsigned int BytePerPixelY[DML2_MAX_PLANES];
unsigned int BytePerPixelC[DML2_MAX_PLANES];
double BytePerPixelInDETY[DML2_MAX_PLANES];
@@ -466,21 +480,74 @@ struct dml2_core_internal_mode_support {
double mall_prefetch_sdp_overhead_factor[DML2_MAX_PLANES]; // overhead to the imall or phantom pipe
double mall_prefetch_dram_overhead_factor[DML2_MAX_PLANES];
+ bool is_using_mall_for_ss[DML2_MAX_PLANES];
+ unsigned int meta_row_width_chroma[DML2_MAX_PLANES];
+ unsigned int PixelPTEReqHeightC[DML2_MAX_PLANES];
+ bool PTE_BUFFER_MODE[DML2_MAX_PLANES];
+ unsigned int meta_req_height_chroma[DML2_MAX_PLANES];
+ unsigned int meta_pte_bytes_per_frame_ub_c[DML2_MAX_PLANES];
+ unsigned int dpde0_bytes_per_frame_ub_c[DML2_MAX_PLANES];
+ unsigned int dpte_row_width_luma_ub[DML2_MAX_PLANES];
+ unsigned int meta_req_width[DML2_MAX_PLANES];
+ unsigned int meta_row_width[DML2_MAX_PLANES];
+ unsigned int PixelPTEReqWidthY[DML2_MAX_PLANES];
+ unsigned int dpte_row_height_linear[DML2_MAX_PLANES];
+ unsigned int PTERequestSizeY[DML2_MAX_PLANES];
+ unsigned int dpte_row_width_chroma_ub[DML2_MAX_PLANES];
+ unsigned int PixelPTEReqWidthC[DML2_MAX_PLANES];
+ unsigned int meta_pte_bytes_per_frame_ub_l[DML2_MAX_PLANES];
+ unsigned int dpte_row_height_linear_chroma[DML2_MAX_PLANES];
+ unsigned int PTERequestSizeC[DML2_MAX_PLANES];
+ unsigned int meta_req_height[DML2_MAX_PLANES];
+ unsigned int dpde0_bytes_per_frame_ub_l[DML2_MAX_PLANES];
+ unsigned int meta_req_width_chroma[DML2_MAX_PLANES];
+ unsigned int PixelPTEReqHeightY[DML2_MAX_PLANES];
+ unsigned int BIGK_FRAGMENT_SIZE[DML2_MAX_PLANES];
+ unsigned int vm_group_bytes[DML2_MAX_PLANES];
+ unsigned int VReadyOffsetPix[DML2_MAX_PLANES];
+ unsigned int VUpdateOffsetPix[DML2_MAX_PLANES];
+ unsigned int VUpdateWidthPix[DML2_MAX_PLANES];
+ double TSetup[DML2_MAX_PLANES];
+ double Tdmdl_vm_raw[DML2_MAX_PLANES];
+ double Tdmdl_raw[DML2_MAX_PLANES];
+ unsigned int VStartupMin[DML2_MAX_PLANES]; /// <brief Minimum vstartup to meet the prefetch schedule (i.e. the prefetch solution can be found at this vstartup time); not the actual global sync vstartup pos.
+ double MaxActiveDRAMClockChangeLatencySupported[DML2_MAX_PLANES];
+ double MaxActiveFCLKChangeLatencySupported;
+
// Backend
bool RequiresDSC[DML2_MAX_PLANES];
bool RequiresFEC[DML2_MAX_PLANES];
double OutputBpp[DML2_MAX_PLANES];
+ double DesiredOutputBpp[DML2_MAX_PLANES];
+ double PixelClockBackEnd[DML2_MAX_PLANES];
unsigned int DSCDelay[DML2_MAX_PLANES];
enum dml2_core_internal_output_type OutputType[DML2_MAX_PLANES];
enum dml2_core_internal_output_type_rate OutputRate[DML2_MAX_PLANES];
+ bool TotalAvailablePipesSupportNoDSC;
+ bool TotalAvailablePipesSupportDSC;
+ unsigned int NumberOfDPPNoDSC;
+ unsigned int NumberOfDPPDSC;
+ enum dml2_odm_mode ODMModeNoDSC;
+ enum dml2_odm_mode ODMModeDSC;
+ double RequiredDISPCLKPerSurfaceNoDSC;
+ double RequiredDISPCLKPerSurfaceDSC;
+ unsigned int EstimatedNumberOfDSCSlices[DML2_MAX_PLANES];
// Bandwidth Related Info
double BandwidthAvailableForImmediateFlip;
- double SurfaceReadBandwidthLuma[DML2_MAX_PLANES]; // no dcc overhead, for the plane
- double SurfaceReadBandwidthChroma[DML2_MAX_PLANES];
- double WriteBandwidth[DML2_MAX_PLANES];
+ double vactive_sw_bw_l[DML2_MAX_PLANES]; // no dcc overhead, for the plane
+ double vactive_sw_bw_c[DML2_MAX_PLANES];
+ double WriteBandwidth[DML2_MAX_PLANES][DML2_MAX_WRITEBACK];
double RequiredPrefetchPixelDataBWLuma[DML2_MAX_PLANES];
double RequiredPrefetchPixelDataBWChroma[DML2_MAX_PLANES];
+ /* Max bandwidth calculated from prefetch schedule should be considered in addition to the pixel data bw to avoid ms/mp mismatches.
+ * 1. oto bw should also be considered when calculating peak urgent bw to avoid situations oto/equ mismatches between ms and mp
+ *
+ * 2. equ bandwidth needs to be considered for calculating peak urgent bw when equ schedule is used in mode support.
+ * Some slight difference in variables may cause the pixel data bandwidth to be higher
+ * even though overall equ prefetch bandwidths can be lower going from ms to mp
+ */
+ double RequiredPrefetchBWMax[DML2_MAX_PLANES];
double cursor_bw[DML2_MAX_PLANES];
double prefetch_cursor_bw[DML2_MAX_PLANES];
double prefetch_vmrow_bw[DML2_MAX_PLANES];
@@ -519,11 +586,13 @@ struct dml2_core_internal_mode_support {
unsigned int num_mcaches_l[DML2_MAX_PLANES];
unsigned int mcache_row_bytes_l[DML2_MAX_PLANES];
+ unsigned int mcache_row_bytes_per_channel_l[DML2_MAX_PLANES];
unsigned int mcache_offsets_l[DML2_MAX_PLANES][DML2_MAX_MCACHES + 1];
unsigned int mcache_shift_granularity_l[DML2_MAX_PLANES];
unsigned int num_mcaches_c[DML2_MAX_PLANES];
unsigned int mcache_row_bytes_c[DML2_MAX_PLANES];
+ unsigned int mcache_row_bytes_per_channel_c[DML2_MAX_PLANES];
unsigned int mcache_offsets_c[DML2_MAX_PLANES][DML2_MAX_MCACHES + 1];
unsigned int mcache_shift_granularity_c[DML2_MAX_PLANES];
@@ -531,7 +600,44 @@ struct dml2_core_internal_mode_support {
bool mall_comb_mcache_c[DML2_MAX_PLANES];
bool lc_comb_mcache[DML2_MAX_PLANES];
+ unsigned int vmpg_width_y[DML2_MAX_PLANES];
+ unsigned int vmpg_height_y[DML2_MAX_PLANES];
+ unsigned int vmpg_width_c[DML2_MAX_PLANES];
+ unsigned int vmpg_height_c[DML2_MAX_PLANES];
+ unsigned int meta_row_height_luma[DML2_MAX_PLANES];
+ unsigned int meta_row_height_chroma[DML2_MAX_PLANES];
+ unsigned int meta_row_bytes_per_row_ub_l[DML2_MAX_PLANES];
+ unsigned int meta_row_bytes_per_row_ub_c[DML2_MAX_PLANES];
+ unsigned int dpte_row_bytes_per_row_l[DML2_MAX_PLANES];
+ unsigned int dpte_row_bytes_per_row_c[DML2_MAX_PLANES];
+
+ unsigned int pstate_bytes_required_l[DML2_MAX_PLANES];
+ unsigned int pstate_bytes_required_c[DML2_MAX_PLANES];
+ unsigned int cursor_bytes_per_chunk[DML2_MAX_PLANES];
+ unsigned int cursor_bytes_per_line[DML2_MAX_PLANES];
+
+ unsigned int MaximumVStartup[DML2_MAX_PLANES];
+
+ double HostVMInefficiencyFactor;
+ double HostVMInefficiencyFactorPrefetch;
+
+ unsigned int tdlut_pte_bytes_per_frame[DML2_MAX_PLANES];
+ unsigned int tdlut_bytes_per_frame[DML2_MAX_PLANES];
+ unsigned int tdlut_groups_per_2row_ub[DML2_MAX_PLANES];
+ double tdlut_opt_time[DML2_MAX_PLANES];
+ double tdlut_drain_time[DML2_MAX_PLANES];
+ unsigned int tdlut_bytes_per_group[DML2_MAX_PLANES];
+
+ double Tvm_trips_flip[DML2_MAX_PLANES];
+ double Tr0_trips_flip[DML2_MAX_PLANES];
+ double Tvm_trips_flip_rounded[DML2_MAX_PLANES];
+ double Tr0_trips_flip_rounded[DML2_MAX_PLANES];
+
+ unsigned int DSTYAfterScaler[DML2_MAX_PLANES];
+ unsigned int DSTXAfterScaler[DML2_MAX_PLANES];
+
+ enum dml2_pstate_method pstate_switch_modes[DML2_MAX_PLANES];
};
/// @brief A mega structure that houses various info for model programming step.
@@ -539,8 +645,9 @@ struct dml2_core_internal_mode_program {
unsigned int qos_param_index; // to access the uclk dependent dpm table
unsigned int active_min_uclk_dpm_index; // to access the min_clk table
double FabricClock; /// <brief Basically just the clock freq at the min (or given) state
- double DCFCLK; /// <brief Basically just the clock freq at the min (or given) state and max combine setting
+ //double DCFCLK; /// <brief Basically just the clock freq at the min (or given) state and max combine setting
double dram_bw_mbps;
+ double min_available_urgent_bandwidth_MBps; /// <brief Minimum guaranteed available urgent return bandwidth in MBps
double uclk_freq_mhz;
unsigned int NoOfDPP[DML2_MAX_PLANES];
enum dml2_odm_mode ODMMode[DML2_MAX_PLANES];
@@ -562,14 +669,14 @@ struct dml2_core_internal_mode_program {
double BytePerPixelInDETC[DML2_MAX_PLANES];
unsigned int BytePerPixelY[DML2_MAX_PLANES];
unsigned int BytePerPixelC[DML2_MAX_PLANES];
- unsigned int SwathWidthY[DML2_MAX_PLANES];
- unsigned int SwathWidthC[DML2_MAX_PLANES];
+ unsigned int SwathWidthY[DML2_MAX_PLANES]; // per-pipe
+ unsigned int SwathWidthC[DML2_MAX_PLANES]; // per-pipe
unsigned int req_per_swath_ub_l[DML2_MAX_PLANES];
unsigned int req_per_swath_ub_c[DML2_MAX_PLANES];
unsigned int SwathWidthSingleDPPY[DML2_MAX_PLANES];
unsigned int SwathWidthSingleDPPC[DML2_MAX_PLANES];
- double SurfaceReadBandwidthLuma[DML2_MAX_PLANES];
- double SurfaceReadBandwidthChroma[DML2_MAX_PLANES];
+ double vactive_sw_bw_l[DML2_MAX_PLANES];
+ double vactive_sw_bw_c[DML2_MAX_PLANES];
double excess_vactive_fill_bw_l[DML2_MAX_PLANES];
double excess_vactive_fill_bw_c[DML2_MAX_PLANES];
@@ -592,6 +699,8 @@ struct dml2_core_internal_mode_program {
unsigned int MacroTileHeightC[DML2_MAX_PLANES];
unsigned int MacroTileWidthY[DML2_MAX_PLANES];
unsigned int MacroTileWidthC[DML2_MAX_PLANES];
+ double MaximumSwathWidthLuma[DML2_MAX_PLANES];
+ double MaximumSwathWidthChroma[DML2_MAX_PLANES];
bool surf_linear128_l[DML2_MAX_PLANES];
bool surf_linear128_c[DML2_MAX_PLANES];
@@ -624,6 +733,14 @@ struct dml2_core_internal_mode_program {
double UrgentBurstFactorChroma[DML2_MAX_PLANES];
double UrgentBurstFactorChromaPre[DML2_MAX_PLANES];
+ double MaximumSwathWidthInLineBufferLuma;
+ double MaximumSwathWidthInLineBufferChroma;
+
+ unsigned int vmpg_width_y[DML2_MAX_PLANES];
+ unsigned int vmpg_height_y[DML2_MAX_PLANES];
+ unsigned int vmpg_width_c[DML2_MAX_PLANES];
+ unsigned int vmpg_height_c[DML2_MAX_PLANES];
+
double meta_row_bw[DML2_MAX_PLANES];
unsigned int meta_row_bytes[DML2_MAX_PLANES];
unsigned int meta_req_width[DML2_MAX_PLANES];
@@ -645,7 +762,9 @@ struct dml2_core_internal_mode_program {
unsigned int PTERequestSizeC[DML2_MAX_PLANES];
double TWait[DML2_MAX_PLANES];
+ double Tdmdl_vm_raw[DML2_MAX_PLANES];
double Tdmdl_vm[DML2_MAX_PLANES];
+ double Tdmdl_raw[DML2_MAX_PLANES];
double Tdmdl[DML2_MAX_PLANES];
double TSetup[DML2_MAX_PLANES];
unsigned int dpde0_bytes_per_frame_ub_l[DML2_MAX_PLANES];
@@ -677,6 +796,38 @@ struct dml2_core_internal_mode_program {
double TCalc;
unsigned int TotImmediateFlipBytes;
+ unsigned int MaxTotalDETInKByte;
+ unsigned int NomDETInKByte;
+ unsigned int MinCompressedBufferSizeInKByte;
+ double PixelClockBackEnd[DML2_MAX_PLANES];
+ double OutputBpp[DML2_MAX_PLANES];
+ bool dsc_enable[DML2_MAX_PLANES];
+ unsigned int num_dsc_slices[DML2_MAX_PLANES];
+ unsigned int meta_row_bytes_per_row_ub_l[DML2_MAX_PLANES];
+ unsigned int meta_row_bytes_per_row_ub_c[DML2_MAX_PLANES];
+ unsigned int dpte_row_bytes_per_row_l[DML2_MAX_PLANES];
+ unsigned int dpte_row_bytes_per_row_c[DML2_MAX_PLANES];
+ unsigned int cursor_bytes_per_chunk[DML2_MAX_PLANES];
+ unsigned int cursor_bytes_per_line[DML2_MAX_PLANES];
+ unsigned int MaxVStartupLines[DML2_MAX_PLANES]; /// <brief more like vblank for the plane's OTG
+ double HostVMInefficiencyFactor;
+ double HostVMInefficiencyFactorPrefetch;
+ unsigned int tdlut_pte_bytes_per_frame[DML2_MAX_PLANES];
+ unsigned int tdlut_bytes_per_frame[DML2_MAX_PLANES];
+ unsigned int tdlut_groups_per_2row_ub[DML2_MAX_PLANES];
+ double tdlut_opt_time[DML2_MAX_PLANES];
+ double tdlut_drain_time[DML2_MAX_PLANES];
+ unsigned int tdlut_bytes_per_group[DML2_MAX_PLANES];
+ double Tvm_trips_flip[DML2_MAX_PLANES];
+ double Tr0_trips_flip[DML2_MAX_PLANES];
+ double Tvm_trips_flip_rounded[DML2_MAX_PLANES];
+ double Tr0_trips_flip_rounded[DML2_MAX_PLANES];
+ bool immediate_flip_required; // any pipes need immediate flip
+ double SOCCLK; /// <brief Basically just the clock freq at the min (or given) state
+ double TotalWRBandwidth;
+ double max_urgent_latency_us;
+ double df_response_time_us;
+
// -------------------
// Output
// -------------------
@@ -687,9 +838,12 @@ struct dml2_core_internal_mode_program {
// Support
bool UrgVactiveBandwidthSupport;
+ bool PrefetchScheduleSupported;
+ bool UrgentBandwidthSupport;
bool PrefetchModeSupported; // <brief Is the prefetch mode (bandwidth and latency) supported
bool ImmediateFlipSupported;
bool ImmediateFlipSupportedForPipe[DML2_MAX_PLANES];
+ bool dcfclk_support;
// Clock
double Dcfclk;
@@ -781,7 +935,7 @@ struct dml2_core_internal_mode_program {
// RQ registers
bool PTE_BUFFER_MODE[DML2_MAX_PLANES];
unsigned int BIGK_FRAGMENT_SIZE[DML2_MAX_PLANES];
-
+ double VActiveLatencyHidingUs[DML2_MAX_PLANES];
unsigned int SubViewportLinesNeededInMALL[DML2_MAX_PLANES];
bool is_using_mall_for_ss[DML2_MAX_PLANES];
@@ -797,8 +951,9 @@ struct dml2_core_internal_mode_program {
double MaxActiveFCLKChangeLatencySupported;
bool USRRetrainingSupport;
bool g6_temp_read_support;
- enum dml2_fclock_change_support FCLKChangeSupport[DML2_MAX_PLANES];
- enum dml2_dram_clock_change_support DRAMClockChangeSupport[DML2_MAX_PLANES];
+ bool temp_read_or_ppt_support;
+ enum dml2_pstate_change_support FCLKChangeSupport[DML2_MAX_PLANES];
+ enum dml2_pstate_change_support DRAMClockChangeSupport[DML2_MAX_PLANES];
bool global_dram_clock_change_supported;
bool global_fclk_change_supported;
double MaxActiveDRAMClockChangeLatencySupported[DML2_MAX_PLANES];
@@ -835,17 +990,21 @@ struct dml2_core_internal_mode_program {
unsigned int num_mcaches_l[DML2_MAX_PLANES];
unsigned int mcache_row_bytes_l[DML2_MAX_PLANES];
+ unsigned int mcache_row_bytes_per_channel_l[DML2_MAX_PLANES];
unsigned int mcache_offsets_l[DML2_MAX_PLANES][DML2_MAX_MCACHES + 1];
unsigned int mcache_shift_granularity_l[DML2_MAX_PLANES];
unsigned int num_mcaches_c[DML2_MAX_PLANES];
unsigned int mcache_row_bytes_c[DML2_MAX_PLANES];
+ unsigned int mcache_row_bytes_per_channel_c[DML2_MAX_PLANES];
unsigned int mcache_offsets_c[DML2_MAX_PLANES][DML2_MAX_MCACHES + 1];
unsigned int mcache_shift_granularity_c[DML2_MAX_PLANES];
bool mall_comb_mcache_l[DML2_MAX_PLANES];
bool mall_comb_mcache_c[DML2_MAX_PLANES];
bool lc_comb_mcache[DML2_MAX_PLANES];
+
+ double impacted_prefetch_margin_us[DML2_MAX_PLANES];
};
struct dml2_core_internal_SOCParametersList {
@@ -862,6 +1021,7 @@ struct dml2_core_internal_SOCParametersList {
double USRRetrainingLatency;
double SMNLatency;
double g6_temp_read_blackout_us;
+ double temp_read_or_ppt_blackout_us;
double max_urgent_latency_us;
double df_response_time_us;
enum dml2_qos_param_type qos_type;
@@ -951,6 +1111,7 @@ struct dml2_core_calcs_mode_support_locals {
unsigned int tdlut_groups_per_2row_ub[DML2_MAX_PLANES];
double tdlut_opt_time[DML2_MAX_PLANES];
double tdlut_drain_time[DML2_MAX_PLANES];
+ unsigned int tdlut_bytes_to_deliver[DML2_MAX_PLANES];
unsigned int tdlut_bytes_per_group[DML2_MAX_PLANES];
unsigned int cursor_bytes_per_chunk[DML2_MAX_PLANES];
@@ -961,6 +1122,18 @@ struct dml2_core_calcs_mode_support_locals {
unsigned int pstate_bytes_required_l[DML2_MAX_PLANES];
unsigned int pstate_bytes_required_c[DML2_MAX_PLANES];
+
+ double prefetch_sw_bytes[DML2_MAX_PLANES];
+ double Tpre_rounded[DML2_MAX_PLANES];
+ double Tpre_oto[DML2_MAX_PLANES];
+ bool recalc_prefetch_schedule;
+ bool recalc_prefetch_done;
+ double impacted_dst_y_pre[DML2_MAX_PLANES];
+ double line_times[DML2_MAX_PLANES];
+ enum dml2_source_format_class pixel_format[DML2_MAX_PLANES];
+ unsigned int lb_source_lines_l[DML2_MAX_PLANES];
+ unsigned int lb_source_lines_c[DML2_MAX_PLANES];
+ double prefetch_swath_time_us[DML2_MAX_PLANES];
};
struct dml2_core_calcs_mode_programming_locals {
@@ -975,10 +1148,10 @@ struct dml2_core_calcs_mode_programming_locals {
double dummy_bw[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max];
double surface_dummy_bw[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max][DML2_MAX_PLANES];
double surface_dummy_bw0[dml2_core_internal_soc_state_max][dml2_core_internal_bw_max][DML2_MAX_PLANES];
- unsigned int dummy_integer_array[2][DML2_MAX_PLANES];
+ unsigned int dummy_integer_array[4][DML2_MAX_PLANES];
enum dml2_output_encoder_class dummy_output_encoder_array[DML2_MAX_PLANES];
double dummy_single_array[2][DML2_MAX_PLANES];
- unsigned int dummy_long_array[4][DML2_MAX_PLANES];
+ unsigned int dummy_long_array[8][DML2_MAX_PLANES];
bool dummy_boolean_array[2][DML2_MAX_PLANES];
bool dummy_boolean[2];
double dummy_single[2];
@@ -1002,7 +1175,6 @@ struct dml2_core_calcs_mode_programming_locals {
double dlg_vblank_start;
double LSetup;
double blank_lines_remaining;
- double TotalWRBandwidth;
double WRBandwidth;
struct dml2_core_internal_DmlPipe myPipe;
double PixelClockBackEndFactor;
@@ -1024,6 +1196,7 @@ struct dml2_core_calcs_mode_programming_locals {
unsigned int tdlut_groups_per_2row_ub[DML2_MAX_PLANES];
double tdlut_opt_time[DML2_MAX_PLANES];
double tdlut_drain_time[DML2_MAX_PLANES];
+ unsigned int tdlut_bytes_to_deliver[DML2_MAX_PLANES];
unsigned int tdlut_bytes_per_group[DML2_MAX_PLANES];
unsigned int cursor_bytes_per_chunk[DML2_MAX_PLANES];
@@ -1041,6 +1214,18 @@ struct dml2_core_calcs_mode_programming_locals {
unsigned int pstate_bytes_required_l[DML2_MAX_PLANES];
unsigned int pstate_bytes_required_c[DML2_MAX_PLANES];
+
+ double prefetch_sw_bytes[DML2_MAX_PLANES];
+ double Tpre_rounded[DML2_MAX_PLANES];
+ double Tpre_oto[DML2_MAX_PLANES];
+ bool recalc_prefetch_schedule;
+ double impacted_dst_y_pre[DML2_MAX_PLANES];
+ double line_times[DML2_MAX_PLANES];
+ enum dml2_source_format_class pixel_format[DML2_MAX_PLANES];
+ unsigned int lb_source_lines_l[DML2_MAX_PLANES];
+ unsigned int lb_source_lines_c[DML2_MAX_PLANES];
+ unsigned int num_dsc_slices[DML2_MAX_PLANES];
+ bool dsc_enable[DML2_MAX_PLANES];
};
struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_locals {
@@ -1048,6 +1233,7 @@ struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_local
double ActiveFCLKChangeLatencyMargin[DML2_MAX_PLANES];
double USRRetrainingLatencyMargin[DML2_MAX_PLANES];
double g6_temp_read_latency_margin[DML2_MAX_PLANES];
+ double temp_read_or_ppt_latency_margin[DML2_MAX_PLANES];
double EffectiveLBLatencyHidingY;
double EffectiveLBLatencyHidingC;
@@ -1113,6 +1299,7 @@ struct dml2_core_calcs_CalculateVMRowAndSwath_params {
unsigned int HostVMMinPageSize;
unsigned int DCCMetaBufferSizeBytes;
bool mrq_present;
+ enum dml2_pstate_method pstate_switch_modes[DML2_MAX_PLANES];
// Output
bool *PTEBufferSizeNotExceeded;
@@ -1185,17 +1372,14 @@ struct dml2_core_calcs_CalculatePrefetchSchedule_locals {
double LineTime;
double dst_y_prefetch_equ;
double prefetch_bw_oto;
+ double per_pipe_vactive_sw_bw;
double Tvm_oto;
double Tr0_oto;
- double Tvm_no_trip_oto;
- double Tr0_no_trip_oto;
double Tvm_oto_lines;
double Tr0_oto_lines;
double dst_y_prefetch_oto;
double TimeForFetchingVM;
double TimeForFetchingRowInVBlank;
- double dst_y_per_vm_no_trip_vblank;
- double dst_y_per_row_no_trip_vblank;
double LinesToRequestPrefetchPixelData;
unsigned int HostVMDynamicLevelsTrips;
double trip_to_mem;
@@ -1203,15 +1387,12 @@ struct dml2_core_calcs_CalculatePrefetchSchedule_locals {
double Tr0_trips_rounded;
double max_Tsw;
double Lsw_oto;
- double Lsw_equ;
- double Tpre_rounded;
double prefetch_bw_equ;
double Tvm_equ;
double Tr0_equ;
double Tdmbf;
double Tdmec;
double Tdmsks;
- double prefetch_sw_bytes;
double total_row_bytes;
double prefetch_bw_pr;
double bytes_pp;
@@ -1225,6 +1406,7 @@ struct dml2_core_calcs_CalculatePrefetchSchedule_locals {
double prefetch_bw2;
double prefetch_bw3;
double prefetch_bw4;
+ double dst_y_prefetch_equ_impacted;
double TWait_p;
unsigned int cursor_prefetch_bytes;
@@ -1354,6 +1536,7 @@ struct dml2_core_shared_get_urgent_bandwidth_required_locals {
double vm_row_bw;
double flip_and_active_bw;
double flip_and_prefetch_bw;
+ double flip_and_prefetch_bw_max;
double active_and_excess_bw;
};
@@ -1382,6 +1565,7 @@ struct dml2_core_shared_CalculateFlipSchedule_locals {
struct dml2_core_shared_rq_dlg_get_dlg_reg_locals {
unsigned int plane_idx;
+ unsigned int stream_idx;
enum dml2_source_format_class source_format;
const struct dml2_timing_cfg *timing;
bool dual_plane;
@@ -1537,7 +1721,7 @@ struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_param
unsigned int *DSTYAfterScaler;
bool UnboundedRequestEnabled;
unsigned int CompressedBufferSizeInkByte;
- bool max_oustanding_when_urgent_expected;
+ bool max_outstanding_when_urgent_expected;
unsigned int max_outstanding_requests;
unsigned int max_request_size_bytes;
unsigned int *meta_row_height_l;
@@ -1545,17 +1729,18 @@ struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_param
// Output
struct dml2_core_internal_watermarks *Watermark;
- enum dml2_dram_clock_change_support *DRAMClockChangeSupport;
+ enum dml2_pstate_change_support *DRAMClockChangeSupport;
bool *global_dram_clock_change_supported;
double *MaxActiveDRAMClockChangeLatencySupported;
unsigned int *SubViewportLinesNeededInMALL;
- enum dml2_fclock_change_support *FCLKChangeSupport;
+ enum dml2_pstate_change_support *FCLKChangeSupport;
bool *global_fclk_change_supported;
double *MaxActiveFCLKChangeLatencySupported;
bool *USRRetrainingSupport;
double *VActiveLatencyHidingMargin;
double *VActiveLatencyHidingUs;
bool *g6_temp_read_support;
+ bool *temp_read_or_ppt_support;
};
@@ -1588,6 +1773,9 @@ struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params {
double *BytePerPixDETC;
unsigned int *DPPPerSurface;
bool mrq_present;
+ unsigned int dummy[2][DML2_MAX_PLANES];
+ unsigned int swath_width_luma_ub_single_dpp[DML2_MAX_PLANES];
+ unsigned int swath_width_chroma_ub_single_dpp[DML2_MAX_PLANES];
// output
unsigned int *req_per_swath_ub_l;
@@ -1605,6 +1793,8 @@ struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params {
unsigned int *DETBufferSizeC;
unsigned int *full_swath_bytes_l;
unsigned int *full_swath_bytes_c;
+ unsigned int *full_swath_bytes_single_dpp_l;
+ unsigned int *full_swath_bytes_single_dpp_c;
bool *UnboundedRequestEnabled;
unsigned int *compbuf_reserved_space_64b;
unsigned int *CompressedBufferSizeInkByte;
@@ -1727,8 +1917,8 @@ struct dml2_core_calcs_CalculatePrefetchSchedule_params {
double PrefetchSourceLinesC;
unsigned int VInitPreFillC;
unsigned int MaxNumSwathC;
- unsigned int swath_width_luma_ub;
- unsigned int swath_width_chroma_ub;
+ unsigned int swath_width_luma_ub; // per-pipe
+ unsigned int swath_width_chroma_ub; // per-pipe
unsigned int SwathHeightY;
unsigned int SwathHeightC;
double TWait;
@@ -1750,6 +1940,10 @@ struct dml2_core_calcs_CalculatePrefetchSchedule_params {
unsigned int meta_row_bytes;
double mall_prefetch_sdp_overhead_factor;
+ double impacted_dst_y_pre;
+ double vactive_sw_bw_l; // per surface bw
+ double vactive_sw_bw_c; // per surface bw
+
// output
unsigned int *DSTXAfterScaler;
unsigned int *DSTYAfterScaler;
@@ -1760,6 +1954,7 @@ struct dml2_core_calcs_CalculatePrefetchSchedule_params {
double *VRatioPrefetchC;
double *RequiredPrefetchPixelDataBWLuma;
double *RequiredPrefetchPixelDataBWChroma;
+ double *RequiredPrefetchBWMax;
bool *NotEnoughTimeForDynamicMetadata;
double *Tno_bw;
double *Tno_bw_flip;
@@ -1767,6 +1962,8 @@ struct dml2_core_calcs_CalculatePrefetchSchedule_params {
double *Tdmdl_vm;
double *Tdmdl;
double *TSetup;
+ double *Tpre_rounded;
+ double *Tpre_oto;
double *Tvm_trips;
double *Tr0_trips;
double *Tvm_trips_flip;
@@ -1777,6 +1974,48 @@ struct dml2_core_calcs_CalculatePrefetchSchedule_params {
unsigned int *VUpdateWidthPix;
unsigned int *VReadyOffsetPix;
double *prefetch_cursor_bw;
+ double *prefetch_sw_bytes;
+ double *prefetch_swath_time_us;
+};
+
+struct dml2_core_calcs_CheckGlobalPrefetchAdmissibility_params {
+ unsigned int num_active_planes;
+ enum dml2_source_format_class *pixel_format;
+ unsigned int rob_buffer_size_kbytes;
+ unsigned int compressed_buffer_size_kbytes;
+ unsigned int chunk_bytes_l; // same for all planes
+ unsigned int chunk_bytes_c;
+ unsigned int *detile_buffer_size_bytes_l;
+ unsigned int *detile_buffer_size_bytes_c;
+ unsigned int *full_swath_bytes_l;
+ unsigned int *full_swath_bytes_c;
+ unsigned int *lb_source_lines_l;
+ unsigned int *lb_source_lines_c;
+ unsigned int *swath_height_l;
+ unsigned int *swath_height_c;
+ double *prefetch_sw_bytes;
+ double *Tpre_rounded;
+ double *Tpre_oto;
+ double estimated_dcfclk_mhz;
+ double estimated_urg_bandwidth_required_mbps;
+ double *line_time;
+ double *dst_y_prefetch;
+
+ // output
+ bool *recalc_prefetch_schedule;
+ double *impacted_dst_y_pre;
+};
+
+struct dml2_core_calcs_CheckGlobalPrefetchAdmissibility_locals {
+ unsigned int max_Trpd_dcfclk_cycles;
+ unsigned int burst_bytes_to_fill_det;
+ double time_to_fill_det_us;
+ unsigned int accumulated_return_path_dcfclk_cycles[DML2_MAX_PLANES];
+ bool prefetch_global_check_passed;
+ unsigned int src_swath_bytes_l[DML2_MAX_PLANES];
+ unsigned int src_swath_bytes_c[DML2_MAX_PLANES];
+ unsigned int src_detile_buf_size_bytes_l[DML2_MAX_PLANES];
+ unsigned int src_detile_buf_size_bytes_c[DML2_MAX_PLANES];
};
struct dml2_core_calcs_calculate_mcache_row_bytes_params {
@@ -1807,6 +2046,7 @@ struct dml2_core_calcs_calculate_mcache_row_bytes_params {
// output
unsigned int *num_mcaches;
unsigned int *mcache_row_bytes;
+ unsigned int *mcache_row_bytes_per_channel;
unsigned int *meta_row_width_ub;
double *dcc_dram_bw_nom_overhead_factor;
double *dcc_dram_bw_pref_overhead_factor;
@@ -1886,6 +2126,7 @@ struct dml2_core_calcs_calculate_mcache_setting_params {
// output
unsigned int *num_mcaches_l;
unsigned int *mcache_row_bytes_l;
+ unsigned int *mcache_row_bytes_per_channel_l;
unsigned int *mcache_offsets_l;
unsigned int *mcache_shift_granularity_l;
double *dcc_dram_bw_nom_overhead_factor_l;
@@ -1893,6 +2134,7 @@ struct dml2_core_calcs_calculate_mcache_setting_params {
unsigned int *num_mcaches_c;
unsigned int *mcache_row_bytes_c;
+ unsigned int *mcache_row_bytes_per_channel_c;
unsigned int *mcache_offsets_c;
unsigned int *mcache_shift_granularity_c;
double *dcc_dram_bw_nom_overhead_factor_c;
@@ -1921,6 +2163,7 @@ struct dml2_core_calcs_calculate_tdlut_setting_params {
unsigned int *tdlut_groups_per_2row_ub;
double *tdlut_opt_time;
double *tdlut_drain_time;
+ unsigned int *tdlut_bytes_to_deliver;
unsigned int *tdlut_bytes_per_group;
};
@@ -1948,6 +2191,7 @@ struct dml2_core_calcs_calculate_peak_bandwidth_required_params {
double *surface_read_bandwidth_c;
double *prefetch_bandwidth_l;
double *prefetch_bandwidth_c;
+ double *prefetch_bandwidth_max;
double *excess_vactive_fill_bw_l;
double *excess_vactive_fill_bw_c;
double *cursor_bw;
@@ -2004,6 +2248,7 @@ struct dml2_core_internal_scratch {
struct dml2_core_calcs_CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_locals CalculateWatermarksMALLUseAndDRAMSpeedChangeSupport_locals;
struct dml2_core_calcs_CalculateVMRowAndSwath_locals CalculateVMRowAndSwath_locals;
struct dml2_core_calcs_CalculatePrefetchSchedule_locals CalculatePrefetchSchedule_locals;
+ struct dml2_core_calcs_CheckGlobalPrefetchAdmissibility_locals CheckGlobalPrefetchAdmissibility_locals;
struct dml2_core_shared_CalculateSwathAndDETConfiguration_locals CalculateSwathAndDETConfiguration_locals;
struct dml2_core_shared_TruncToValidBPP_locals TruncToValidBPP_locals;
struct dml2_core_shared_CalculateDETBufferSize_locals CalculateDETBufferSize_locals;
@@ -2019,6 +2264,7 @@ struct dml2_core_internal_scratch {
struct dml2_core_calcs_CalculateSwathAndDETConfiguration_params CalculateSwathAndDETConfiguration_params;
struct dml2_core_calcs_CalculateStutterEfficiency_params CalculateStutterEfficiency_params;
struct dml2_core_calcs_CalculatePrefetchSchedule_params CalculatePrefetchSchedule_params;
+ struct dml2_core_calcs_CheckGlobalPrefetchAdmissibility_params CheckGlobalPrefetchAdmissibility_params;
struct dml2_core_calcs_calculate_mcache_setting_params calculate_mcache_setting_params;
struct dml2_core_calcs_calculate_tdlut_setting_params calculate_tdlut_setting_params;
struct dml2_core_shared_calculate_vm_and_row_bytes_params calculate_vm_and_row_bytes_params;
@@ -2038,7 +2284,6 @@ struct dml2_core_internal_display_mode_lib {
// Used to hold input; intermediate and output of the calculations
struct dml2_core_internal_mode_support ms; // struct for mode support
struct dml2_core_internal_mode_program mp; // struct for mode programming
-
// Available overridable calculators for core_shared.
// if null, core_shared will use default calculators.
struct dml2_core_shared_calculation_funcs funcs;
@@ -2051,7 +2296,6 @@ struct dml2_core_calcs_mode_support_ex {
const struct dml2_display_cfg *in_display_cfg;
const struct dml2_mcg_min_clock_table *min_clk_table;
int min_clk_index;
-
//unsigned int in_state_index;
struct dml2_core_internal_mode_support_info *out_evaluation_info;
};
@@ -2064,9 +2308,7 @@ struct dml2_core_calcs_mode_programming_ex {
const struct dml2_mcg_min_clock_table *min_clk_table;
const struct core_display_cfg_support_info *cfg_support_info;
int min_clk_index;
-
struct dml2_display_cfg_programming *programming;
-
};
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.c
index 714b5c39b7e6..5f301befed16 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.c
@@ -63,8 +63,152 @@ bool dml2_core_utils_is_420(enum dml2_source_format_class source_format)
case dml2_mono_16:
val = 0;
break;
+ case dml2_422_planar_8:
+ val = 0;
+ break;
+ case dml2_422_planar_10:
+ val = 0;
+ break;
+ case dml2_422_planar_12:
+ val = 0;
+ break;
+ case dml2_422_packed_8:
+ val = 0;
+ break;
+ case dml2_422_packed_10:
+ val = 0;
+ break;
+ case dml2_422_packed_12:
+ val = 0;
+ break;
+ default:
+ DML_ASSERT(0);
+ break;
+ }
+ return val;
+}
+
+bool dml2_core_utils_is_422_planar(enum dml2_source_format_class source_format)
+{
+ bool val = false;
+
+ switch (source_format) {
+ case dml2_444_8:
+ val = 0;
+ break;
+ case dml2_444_16:
+ val = 0;
+ break;
+ case dml2_444_32:
+ val = 0;
+ break;
+ case dml2_444_64:
+ val = 0;
+ break;
+ case dml2_420_8:
+ val = 0;
+ break;
+ case dml2_420_10:
+ val = 0;
+ break;
+ case dml2_420_12:
+ val = 0;
+ break;
+ case dml2_rgbe_alpha:
+ val = 0;
+ break;
+ case dml2_rgbe:
+ val = 0;
+ break;
+ case dml2_mono_8:
+ val = 0;
+ break;
+ case dml2_mono_16:
+ val = 0;
+ break;
+ case dml2_422_planar_8:
+ val = 1;
+ break;
+ case dml2_422_planar_10:
+ val = 1;
+ break;
+ case dml2_422_planar_12:
+ val = 1;
+ break;
+ case dml2_422_packed_8:
+ val = 0;
+ break;
+ case dml2_422_packed_10:
+ val = 0;
+ break;
+ case dml2_422_packed_12:
+ val = 0;
+ break;
default:
- DML2_ASSERT(0);
+ DML_ASSERT(0);
+ break;
+ }
+ return val;
+}
+
+bool dml2_core_utils_is_422_packed(enum dml2_source_format_class source_format)
+{
+ bool val = false;
+
+ switch (source_format) {
+ case dml2_444_8:
+ val = 0;
+ break;
+ case dml2_444_16:
+ val = 0;
+ break;
+ case dml2_444_32:
+ val = 0;
+ break;
+ case dml2_444_64:
+ val = 0;
+ break;
+ case dml2_420_8:
+ val = 0;
+ break;
+ case dml2_420_10:
+ val = 0;
+ break;
+ case dml2_420_12:
+ val = 0;
+ break;
+ case dml2_rgbe_alpha:
+ val = 0;
+ break;
+ case dml2_rgbe:
+ val = 0;
+ break;
+ case dml2_mono_8:
+ val = 0;
+ break;
+ case dml2_mono_16:
+ val = 0;
+ break;
+ case dml2_422_planar_8:
+ val = 0;
+ break;
+ case dml2_422_planar_10:
+ val = 0;
+ break;
+ case dml2_422_planar_12:
+ val = 0;
+ break;
+ case dml2_422_packed_8:
+ val = 1;
+ break;
+ case dml2_422_packed_10:
+ val = 1;
+ break;
+ case dml2_422_packed_12:
+ val = 1;
+ break;
+ default:
+ DML_ASSERT(0);
break;
}
return val;
@@ -72,104 +216,104 @@ bool dml2_core_utils_is_420(enum dml2_source_format_class source_format)
void dml2_core_utils_print_mode_support_info(const struct dml2_core_internal_mode_support_info *support, bool fail_only)
{
- dml2_printf("DML: ===================================== \n");
- dml2_printf("DML: DML_MODE_SUPPORT_INFO_ST\n");
+ DML_LOG_VERBOSE("DML: ===================================== \n");
+ DML_LOG_VERBOSE("DML: DML_MODE_SUPPORT_INFO_ST\n");
if (!fail_only || support->ScaleRatioAndTapsSupport == 0)
- dml2_printf("DML: support: ScaleRatioAndTapsSupport = %d\n", support->ScaleRatioAndTapsSupport);
+ DML_LOG_VERBOSE("DML: support: ScaleRatioAndTapsSupport = %d\n", support->ScaleRatioAndTapsSupport);
if (!fail_only || support->SourceFormatPixelAndScanSupport == 0)
- dml2_printf("DML: support: SourceFormatPixelAndScanSupport = %d\n", support->SourceFormatPixelAndScanSupport);
+ DML_LOG_VERBOSE("DML: support: SourceFormatPixelAndScanSupport = %d\n", support->SourceFormatPixelAndScanSupport);
if (!fail_only || support->ViewportSizeSupport == 0)
- dml2_printf("DML: support: ViewportSizeSupport = %d\n", support->ViewportSizeSupport);
+ DML_LOG_VERBOSE("DML: support: ViewportSizeSupport = %d\n", support->ViewportSizeSupport);
if (!fail_only || support->LinkRateDoesNotMatchDPVersion == 1)
- dml2_printf("DML: support: LinkRateDoesNotMatchDPVersion = %d\n", support->LinkRateDoesNotMatchDPVersion);
+ DML_LOG_VERBOSE("DML: support: LinkRateDoesNotMatchDPVersion = %d\n", support->LinkRateDoesNotMatchDPVersion);
if (!fail_only || support->LinkRateForMultistreamNotIndicated == 1)
- dml2_printf("DML: support: LinkRateForMultistreamNotIndicated = %d\n", support->LinkRateForMultistreamNotIndicated);
+ DML_LOG_VERBOSE("DML: support: LinkRateForMultistreamNotIndicated = %d\n", support->LinkRateForMultistreamNotIndicated);
if (!fail_only || support->BPPForMultistreamNotIndicated == 1)
- dml2_printf("DML: support: BPPForMultistreamNotIndicated = %d\n", support->BPPForMultistreamNotIndicated);
+ DML_LOG_VERBOSE("DML: support: BPPForMultistreamNotIndicated = %d\n", support->BPPForMultistreamNotIndicated);
if (!fail_only || support->MultistreamWithHDMIOreDP == 1)
- dml2_printf("DML: support: MultistreamWithHDMIOreDP = %d\n", support->MultistreamWithHDMIOreDP);
+ DML_LOG_VERBOSE("DML: support: MultistreamWithHDMIOreDP = %d\n", support->MultistreamWithHDMIOreDP);
if (!fail_only || support->ExceededMultistreamSlots == 1)
- dml2_printf("DML: support: ExceededMultistreamSlots = %d\n", support->ExceededMultistreamSlots);
+ DML_LOG_VERBOSE("DML: support: ExceededMultistreamSlots = %d\n", support->ExceededMultistreamSlots);
if (!fail_only || support->MSOOrODMSplitWithNonDPLink == 1)
- dml2_printf("DML: support: MSOOrODMSplitWithNonDPLink = %d\n", support->MSOOrODMSplitWithNonDPLink);
+ DML_LOG_VERBOSE("DML: support: MSOOrODMSplitWithNonDPLink = %d\n", support->MSOOrODMSplitWithNonDPLink);
if (!fail_only || support->NotEnoughLanesForMSO == 1)
- dml2_printf("DML: support: NotEnoughLanesForMSO = %d\n", support->NotEnoughLanesForMSO);
+ DML_LOG_VERBOSE("DML: support: NotEnoughLanesForMSO = %d\n", support->NotEnoughLanesForMSO);
if (!fail_only || support->P2IWith420 == 1)
- dml2_printf("DML: support: P2IWith420 = %d\n", support->P2IWith420);
+ DML_LOG_VERBOSE("DML: support: P2IWith420 = %d\n", support->P2IWith420);
if (!fail_only || support->DSC422NativeNotSupported == 1)
- dml2_printf("DML: support: DSC422NativeNotSupported = %d\n", support->DSC422NativeNotSupported);
+ DML_LOG_VERBOSE("DML: support: DSC422NativeNotSupported = %d\n", support->DSC422NativeNotSupported);
if (!fail_only || support->DSCSlicesODMModeSupported == 0)
- dml2_printf("DML: support: DSCSlicesODMModeSupported = %d\n", support->DSCSlicesODMModeSupported);
+ DML_LOG_VERBOSE("DML: support: DSCSlicesODMModeSupported = %d\n", support->DSCSlicesODMModeSupported);
if (!fail_only || support->NotEnoughDSCUnits == 1)
- dml2_printf("DML: support: NotEnoughDSCUnits = %d\n", support->NotEnoughDSCUnits);
+ DML_LOG_VERBOSE("DML: support: NotEnoughDSCUnits = %d\n", support->NotEnoughDSCUnits);
if (!fail_only || support->NotEnoughDSCSlices == 1)
- dml2_printf("DML: support: NotEnoughDSCSlices = %d\n", support->NotEnoughDSCSlices);
+ DML_LOG_VERBOSE("DML: support: NotEnoughDSCSlices = %d\n", support->NotEnoughDSCSlices);
if (!fail_only || support->ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe == 1)
- dml2_printf("DML: support: ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe = %d\n", support->ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe);
+ DML_LOG_VERBOSE("DML: support: ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe = %d\n", support->ImmediateFlipOrHostVMAndPStateWithMALLFullFrameOrPhantomPipe);
if (!fail_only || support->InvalidCombinationOfMALLUseForPStateAndStaticScreen == 1)
- dml2_printf("DML: support: InvalidCombinationOfMALLUseForPStateAndStaticScreen = %d\n", support->InvalidCombinationOfMALLUseForPStateAndStaticScreen);
+ DML_LOG_VERBOSE("DML: support: InvalidCombinationOfMALLUseForPStateAndStaticScreen = %d\n", support->InvalidCombinationOfMALLUseForPStateAndStaticScreen);
if (!fail_only || support->DSCCLKRequiredMoreThanSupported == 1)
- dml2_printf("DML: support: DSCCLKRequiredMoreThanSupported = %d\n", support->DSCCLKRequiredMoreThanSupported);
+ DML_LOG_VERBOSE("DML: support: DSCCLKRequiredMoreThanSupported = %d\n", support->DSCCLKRequiredMoreThanSupported);
if (!fail_only || support->PixelsPerLinePerDSCUnitSupport == 0)
- dml2_printf("DML: support: PixelsPerLinePerDSCUnitSupport = %d\n", support->PixelsPerLinePerDSCUnitSupport);
+ DML_LOG_VERBOSE("DML: support: PixelsPerLinePerDSCUnitSupport = %d\n", support->PixelsPerLinePerDSCUnitSupport);
if (!fail_only || support->DTBCLKRequiredMoreThanSupported == 1)
- dml2_printf("DML: support: DTBCLKRequiredMoreThanSupported = %d\n", support->DTBCLKRequiredMoreThanSupported);
+ DML_LOG_VERBOSE("DML: support: DTBCLKRequiredMoreThanSupported = %d\n", support->DTBCLKRequiredMoreThanSupported);
if (!fail_only || support->InvalidCombinationOfMALLUseForPState == 1)
- dml2_printf("DML: support: InvalidCombinationOfMALLUseForPState = %d\n", support->InvalidCombinationOfMALLUseForPState);
+ DML_LOG_VERBOSE("DML: support: InvalidCombinationOfMALLUseForPState = %d\n", support->InvalidCombinationOfMALLUseForPState);
if (!fail_only || support->ROBSupport == 0)
- dml2_printf("DML: support: ROBSupport = %d\n", support->ROBSupport);
+ DML_LOG_VERBOSE("DML: support: ROBSupport = %d\n", support->ROBSupport);
if (!fail_only || support->OutstandingRequestsSupport == 0)
- dml2_printf("DML: support: OutstandingRequestsSupport = %d\n", support->OutstandingRequestsSupport);
+ DML_LOG_VERBOSE("DML: support: OutstandingRequestsSupport = %d\n", support->OutstandingRequestsSupport);
if (!fail_only || support->OutstandingRequestsUrgencyAvoidance == 0)
- dml2_printf("DML: support: OutstandingRequestsUrgencyAvoidance = %d\n", support->OutstandingRequestsUrgencyAvoidance);
+ DML_LOG_VERBOSE("DML: support: OutstandingRequestsUrgencyAvoidance = %d\n", support->OutstandingRequestsUrgencyAvoidance);
if (!fail_only || support->DISPCLK_DPPCLK_Support == 0)
- dml2_printf("DML: support: DISPCLK_DPPCLK_Support = %d\n", support->DISPCLK_DPPCLK_Support);
+ DML_LOG_VERBOSE("DML: support: DISPCLK_DPPCLK_Support = %d\n", support->DISPCLK_DPPCLK_Support);
if (!fail_only || support->TotalAvailablePipesSupport == 0)
- dml2_printf("DML: support: TotalAvailablePipesSupport = %d\n", support->TotalAvailablePipesSupport);
+ DML_LOG_VERBOSE("DML: support: TotalAvailablePipesSupport = %d\n", support->TotalAvailablePipesSupport);
if (!fail_only || support->NumberOfOTGSupport == 0)
- dml2_printf("DML: support: NumberOfOTGSupport = %d\n", support->NumberOfOTGSupport);
+ DML_LOG_VERBOSE("DML: support: NumberOfOTGSupport = %d\n", support->NumberOfOTGSupport);
if (!fail_only || support->NumberOfHDMIFRLSupport == 0)
- dml2_printf("DML: support: NumberOfHDMIFRLSupport = %d\n", support->NumberOfHDMIFRLSupport);
+ DML_LOG_VERBOSE("DML: support: NumberOfHDMIFRLSupport = %d\n", support->NumberOfHDMIFRLSupport);
if (!fail_only || support->NumberOfDP2p0Support == 0)
- dml2_printf("DML: support: NumberOfDP2p0Support = %d\n", support->NumberOfDP2p0Support);
+ DML_LOG_VERBOSE("DML: support: NumberOfDP2p0Support = %d\n", support->NumberOfDP2p0Support);
if (!fail_only || support->EnoughWritebackUnits == 0)
- dml2_printf("DML: support: EnoughWritebackUnits = %d\n", support->EnoughWritebackUnits);
+ DML_LOG_VERBOSE("DML: support: EnoughWritebackUnits = %d\n", support->EnoughWritebackUnits);
if (!fail_only || support->WritebackScaleRatioAndTapsSupport == 0)
- dml2_printf("DML: support: WritebackScaleRatioAndTapsSupport = %d\n", support->WritebackScaleRatioAndTapsSupport);
+ DML_LOG_VERBOSE("DML: support: WritebackScaleRatioAndTapsSupport = %d\n", support->WritebackScaleRatioAndTapsSupport);
if (!fail_only || support->WritebackLatencySupport == 0)
- dml2_printf("DML: support: WritebackLatencySupport = %d\n", support->WritebackLatencySupport);
+ DML_LOG_VERBOSE("DML: support: WritebackLatencySupport = %d\n", support->WritebackLatencySupport);
if (!fail_only || support->CursorSupport == 0)
- dml2_printf("DML: support: CursorSupport = %d\n", support->CursorSupport);
+ DML_LOG_VERBOSE("DML: support: CursorSupport = %d\n", support->CursorSupport);
if (!fail_only || support->PitchSupport == 0)
- dml2_printf("DML: support: PitchSupport = %d\n", support->PitchSupport);
+ DML_LOG_VERBOSE("DML: support: PitchSupport = %d\n", support->PitchSupport);
if (!fail_only || support->ViewportExceedsSurface == 1)
- dml2_printf("DML: support: ViewportExceedsSurface = %d\n", support->ViewportExceedsSurface);
+ DML_LOG_VERBOSE("DML: support: ViewportExceedsSurface = %d\n", support->ViewportExceedsSurface);
if (!fail_only || support->PrefetchSupported == 0)
- dml2_printf("DML: support: PrefetchSupported = %d\n", support->PrefetchSupported);
+ DML_LOG_VERBOSE("DML: support: PrefetchSupported = %d\n", support->PrefetchSupported);
if (!fail_only || support->EnoughUrgentLatencyHidingSupport == 0)
- dml2_printf("DML: support: EnoughUrgentLatencyHidingSupport = %d\n", support->EnoughUrgentLatencyHidingSupport);
+ DML_LOG_VERBOSE("DML: support: EnoughUrgentLatencyHidingSupport = %d\n", support->EnoughUrgentLatencyHidingSupport);
if (!fail_only || support->AvgBandwidthSupport == 0)
- dml2_printf("DML: support: AvgBandwidthSupport = %d\n", support->AvgBandwidthSupport);
+ DML_LOG_VERBOSE("DML: support: AvgBandwidthSupport = %d\n", support->AvgBandwidthSupport);
if (!fail_only || support->DynamicMetadataSupported == 0)
- dml2_printf("DML: support: DynamicMetadataSupported = %d\n", support->DynamicMetadataSupported);
+ DML_LOG_VERBOSE("DML: support: DynamicMetadataSupported = %d\n", support->DynamicMetadataSupported);
if (!fail_only || support->VRatioInPrefetchSupported == 0)
- dml2_printf("DML: support: VRatioInPrefetchSupported = %d\n", support->VRatioInPrefetchSupported);
- if (!fail_only || support->PTEBufferSizeNotExceeded == 1)
- dml2_printf("DML: support: PTEBufferSizeNotExceeded = %d\n", support->PTEBufferSizeNotExceeded);
- if (!fail_only || support->DCCMetaBufferSizeNotExceeded == 1)
- dml2_printf("DML: support: DCCMetaBufferSizeNotExceeded = %d\n", support->DCCMetaBufferSizeNotExceeded);
+ DML_LOG_VERBOSE("DML: support: VRatioInPrefetchSupported = %d\n", support->VRatioInPrefetchSupported);
+ if (!fail_only || support->PTEBufferSizeNotExceeded == 0)
+ DML_LOG_VERBOSE("DML: support: PTEBufferSizeNotExceeded = %d\n", support->PTEBufferSizeNotExceeded);
+ if (!fail_only || support->DCCMetaBufferSizeNotExceeded == 0)
+ DML_LOG_VERBOSE("DML: support: DCCMetaBufferSizeNotExceeded = %d\n", support->DCCMetaBufferSizeNotExceeded);
if (!fail_only || support->ExceededMALLSize == 1)
- dml2_printf("DML: support: ExceededMALLSize = %d\n", support->ExceededMALLSize);
+ DML_LOG_VERBOSE("DML: support: ExceededMALLSize = %d\n", support->ExceededMALLSize);
if (!fail_only || support->g6_temp_read_support == 0)
- dml2_printf("DML: support: g6_temp_read_support = %d\n", support->g6_temp_read_support);
+ DML_LOG_VERBOSE("DML: support: g6_temp_read_support = %d\n", support->g6_temp_read_support);
if (!fail_only || support->ImmediateFlipSupport == 0)
- dml2_printf("DML: support: ImmediateFlipSupport = %d\n", support->ImmediateFlipSupport);
+ DML_LOG_VERBOSE("DML: support: ImmediateFlipSupport = %d\n", support->ImmediateFlipSupport);
if (!fail_only || support->LinkCapacitySupport == 0)
- dml2_printf("DML: support: LinkCapacitySupport = %d\n", support->LinkCapacitySupport);
+ DML_LOG_VERBOSE("DML: support: LinkCapacitySupport = %d\n", support->LinkCapacitySupport);
if (!fail_only || support->ModeSupport == 0)
- dml2_printf("DML: support: ModeSupport = %d\n", support->ModeSupport);
- dml2_printf("DML: ===================================== \n");
+ DML_LOG_VERBOSE("DML: support: ModeSupport = %d\n", support->ModeSupport);
+ DML_LOG_VERBOSE("DML: ===================================== \n");
}
const char *dml2_core_utils_internal_soc_state_type_str(enum dml2_core_internal_soc_state_type dml2_core_internal_soc_state_type)
@@ -214,9 +358,9 @@ void dml2_core_utils_get_stream_output_bpp(double *out_bpp, const struct dml2_di
out_bpp[k] = 0;
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: k=%d bpc=%f\n", __func__, k, bpc);
- dml2_printf("DML::%s: k=%d dsc.enable=%d\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable);
- dml2_printf("DML::%s: k=%d out_bpp=%f\n", __func__, k, out_bpp[k]);
+ DML_LOG_VERBOSE("DML::%s: k=%d bpc=%f\n", __func__, k, bpc);
+ DML_LOG_VERBOSE("DML::%s: k=%d dsc.enable=%d\n", __func__, k, display_cfg->stream_descriptors[display_cfg->plane_descriptors[k].stream_index].timing.dsc.enable);
+ DML_LOG_VERBOSE("DML::%s: k=%d out_bpp=%f\n", __func__, k, out_bpp[k]);
#endif
}
}
@@ -247,7 +391,7 @@ unsigned int dml2_core_util_get_num_active_pipes(int unsigned num_planes, const
}
#ifdef __DML_VBA_DEBUG__
- dml2_printf("DML::%s: num_active_pipes = %d\n", __func__, num_active_pipes);
+ DML_LOG_VERBOSE("DML::%s: num_active_pipes = %d\n", __func__, num_active_pipes);
#endif
return num_active_pipes;
}
@@ -280,39 +424,49 @@ bool dml2_core_utils_is_phantom_pipe(const struct dml2_plane_parameters *plane_c
return is_phantom;
}
-unsigned int dml2_core_utils_get_tile_block_size_bytes(enum dml2_swizzle_mode sw_mode)
-{
- switch (sw_mode) {
- case (dml2_sw_linear):
- return 256; break;
- case (dml2_sw_256b_2d):
- return 256; break;
- case (dml2_sw_4kb_2d):
- return 4096; break;
- case (dml2_sw_64kb_2d):
- return 65536; break;
- case (dml2_sw_256kb_2d):
- return 262144; break;
- case (dml2_gfx11_sw_linear):
- return 256; break;
- case (dml2_gfx11_sw_64kb_d):
- return 65536; break;
- case (dml2_gfx11_sw_64kb_d_t):
- return 65536; break;
- case (dml2_gfx11_sw_64kb_d_x):
- return 65536; break;
- case (dml2_gfx11_sw_64kb_r_x):
- return 65536; break;
- case (dml2_gfx11_sw_256kb_d_x):
- return 262144; break;
- case (dml2_gfx11_sw_256kb_r_x):
- return 262144; break;
- default:
- DML2_ASSERT(0);
+unsigned int dml2_core_utils_get_tile_block_size_bytes(enum dml2_swizzle_mode sw_mode, unsigned int byte_per_pixel)
+{
+
+ if (sw_mode == dml2_sw_linear)
+ return 256;
+ else if (sw_mode == dml2_sw_256b_2d)
+ return 256;
+ else if (sw_mode == dml2_sw_4kb_2d)
+ return 4096;
+ else if (sw_mode == dml2_sw_64kb_2d)
+ return 65536;
+ else if (sw_mode == dml2_sw_256kb_2d)
+ return 262144;
+ else if (sw_mode == dml2_gfx11_sw_linear)
+ return 256;
+ else if (sw_mode == dml2_gfx11_sw_64kb_d)
+ return 65536;
+ else if (sw_mode == dml2_gfx11_sw_64kb_d_t)
+ return 65536;
+ else if (sw_mode == dml2_gfx11_sw_64kb_d_x)
+ return 65536;
+ else if (sw_mode == dml2_gfx11_sw_64kb_r_x)
+ return 65536;
+ else if (sw_mode == dml2_gfx11_sw_256kb_d_x)
+ return 262144;
+ else if (sw_mode == dml2_gfx11_sw_256kb_r_x)
+ return 262144;
+ else {
+ DML_ASSERT(0);
return 256;
};
}
+bool dml2_core_utils_get_segment_horizontal_contiguous(enum dml2_swizzle_mode sw_mode, unsigned int byte_per_pixel)
+{
+ return (byte_per_pixel != 2);
+}
+
+bool dml2_core_utils_is_linear(enum dml2_swizzle_mode sw_mode)
+{
+ return sw_mode == dml2_sw_linear;
+};
+
bool dml2_core_utils_is_vertical_rotation(enum dml2_rotation_angle Scan)
{
@@ -325,7 +479,6 @@ bool dml2_core_utils_is_vertical_rotation(enum dml2_rotation_angle Scan)
return is_vert;
}
-
int unsigned dml2_core_utils_get_gfx_version(enum dml2_swizzle_mode sw_mode)
{
int unsigned version = 0;
@@ -334,19 +487,19 @@ int unsigned dml2_core_utils_get_gfx_version(enum dml2_swizzle_mode sw_mode)
sw_mode == dml2_sw_256b_2d ||
sw_mode == dml2_sw_4kb_2d ||
sw_mode == dml2_sw_64kb_2d ||
- sw_mode == dml2_sw_256kb_2d) {
+ sw_mode == dml2_sw_256kb_2d)
version = 12;
- } else if (sw_mode == dml2_gfx11_sw_linear ||
+ else if (sw_mode == dml2_gfx11_sw_linear ||
sw_mode == dml2_gfx11_sw_64kb_d ||
sw_mode == dml2_gfx11_sw_64kb_d_t ||
sw_mode == dml2_gfx11_sw_64kb_d_x ||
sw_mode == dml2_gfx11_sw_64kb_r_x ||
sw_mode == dml2_gfx11_sw_256kb_d_x ||
- sw_mode == dml2_gfx11_sw_256kb_r_x) {
+ sw_mode == dml2_gfx11_sw_256kb_r_x)
version = 11;
- } else {
- dml2_printf("ERROR: Invalid sw_mode setting! val=%u\n", sw_mode);
- DML2_ASSERT(0);
+ else {
+ DML_LOG_VERBOSE("ERROR: Invalid sw_mode setting! val=%u\n", sw_mode);
+ DML_ASSERT(0);
}
return version;
@@ -358,7 +511,7 @@ unsigned int dml2_core_utils_get_qos_param_index(unsigned long uclk_freq_khz, co
unsigned int index = 0;
for (i = 0; i < DML_MAX_CLK_TABLE_SIZE; i++) {
- dml2_printf("DML::%s: per_uclk_dpm_params[%d].minimum_uclk_khz = %d\n", __func__, i, per_uclk_dpm_params[i].minimum_uclk_khz);
+ DML_LOG_VERBOSE("DML::%s: per_uclk_dpm_params[%d].minimum_uclk_khz = %ld\n", __func__, i, per_uclk_dpm_params[i].minimum_uclk_khz);
if (i == 0)
index = 0;
@@ -371,8 +524,8 @@ unsigned int dml2_core_utils_get_qos_param_index(unsigned long uclk_freq_khz, co
}
}
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: uclk_freq_khz = %d\n", __func__, uclk_freq_khz);
- dml2_printf("DML::%s: index = %d\n", __func__, index);
+ DML_LOG_VERBOSE("DML::%s: uclk_freq_khz = %ld\n", __func__, uclk_freq_khz);
+ DML_LOG_VERBOSE("DML::%s: index = %d\n", __func__, index);
#endif
return index;
}
@@ -380,31 +533,32 @@ unsigned int dml2_core_utils_get_qos_param_index(unsigned long uclk_freq_khz, co
unsigned int dml2_core_utils_get_active_min_uclk_dpm_index(unsigned long uclk_freq_khz, const struct dml2_soc_state_table *clk_table)
{
unsigned int i;
- bool clk_entry_found = 0;
+ bool clk_entry_found = false;
for (i = 0; i < clk_table->uclk.num_clk_values; i++) {
- dml2_printf("DML::%s: clk_table.uclk.clk_values_khz[%d] = %d\n", __func__, i, clk_table->uclk.clk_values_khz[i]);
+ DML_LOG_VERBOSE("DML::%s: clk_table.uclk.clk_values_khz[%d] = %ld\n", __func__, i, clk_table->uclk.clk_values_khz[i]);
if (uclk_freq_khz == clk_table->uclk.clk_values_khz[i]) {
- clk_entry_found = 1;
+ clk_entry_found = true;
break;
}
}
- dml2_assert(clk_entry_found);
+ if (!clk_entry_found)
+ DML_ASSERT(clk_entry_found);
#if defined(__DML_VBA_DEBUG__)
- dml2_printf("DML::%s: uclk_freq_khz = %ld\n", __func__, uclk_freq_khz);
- dml2_printf("DML::%s: index = %d\n", __func__, i);
+ DML_LOG_VERBOSE("DML::%s: uclk_freq_khz = %ld\n", __func__, uclk_freq_khz);
+ DML_LOG_VERBOSE("DML::%s: index = %d\n", __func__, i);
#endif
return i;
}
bool dml2_core_utils_is_dual_plane(enum dml2_source_format_class source_format)
{
- bool ret_val = 0;
+ bool ret_val = false;
- if ((source_format == dml2_420_12) || (source_format == dml2_420_8) || (source_format == dml2_420_10) || (source_format == dml2_rgbe_alpha))
- ret_val = 1;
+ if (dml2_core_utils_is_420(source_format) || dml2_core_utils_is_422_planar(source_format) || (source_format == dml2_rgbe_alpha))
+ ret_val = true;
return ret_val;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.h
index a5cc6a07167a..95f0d017add4 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_core/dml2_core_utils.h
@@ -11,6 +11,8 @@
double dml2_core_utils_div_rem(double dividend, unsigned int divisor, unsigned int *remainder);
const char *dml2_core_utils_internal_bw_type_str(enum dml2_core_internal_bw_type bw_type);
bool dml2_core_utils_is_420(enum dml2_source_format_class source_format);
+bool dml2_core_utils_is_422_planar(enum dml2_source_format_class source_format);
+bool dml2_core_utils_is_422_packed(enum dml2_source_format_class source_format);
void dml2_core_utils_print_mode_support_info(const struct dml2_core_internal_mode_support_info *support, bool fail_only);
const char *dml2_core_utils_internal_soc_state_type_str(enum dml2_core_internal_soc_state_type dml2_core_internal_soc_state_type);
void dml2_core_utils_get_stream_output_bpp(double *out_bpp, const struct dml2_display_cfg *display_cfg);
@@ -18,8 +20,10 @@ unsigned int dml2_core_utils_round_to_multiple(unsigned int num, unsigned int mu
unsigned int dml2_core_util_get_num_active_pipes(int unsigned num_planes, const struct core_display_cfg_support_info *cfg_support_info);
void dml2_core_utils_pipe_plane_mapping(const struct core_display_cfg_support_info *cfg_support_info, unsigned int *pipe_plane);
bool dml2_core_utils_is_phantom_pipe(const struct dml2_plane_parameters *plane_cfg);
-unsigned int dml2_core_utils_get_tile_block_size_bytes(enum dml2_swizzle_mode sw_mode);
+unsigned int dml2_core_utils_get_tile_block_size_bytes(enum dml2_swizzle_mode sw_mode, unsigned int byte_per_pixel);
+bool dml2_core_utils_get_segment_horizontal_contiguous(enum dml2_swizzle_mode sw_mode, unsigned int byte_per_pixel);
bool dml2_core_utils_is_vertical_rotation(enum dml2_rotation_angle Scan);
+bool dml2_core_utils_is_linear(enum dml2_swizzle_mode sw_mode);
int unsigned dml2_core_utils_get_gfx_version(enum dml2_swizzle_mode sw_mode);
unsigned int dml2_core_utils_get_qos_param_index(unsigned long uclk_freq_khz, const struct dml2_dcn4_uclk_dpm_dependent_qos_params *per_uclk_dpm_params);
unsigned int dml2_core_utils_get_active_min_uclk_dpm_index(unsigned long uclk_freq_khz, const struct dml2_soc_state_table *clk_table);
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.c
index 8869ea089312..22969a533a7b 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.c
@@ -96,6 +96,7 @@ static void calculate_svp_prefetch_minimums(struct dml2_dpmm_map_mode_to_soc_dpm
double min_uclk_latency;
const struct dml2_core_mode_support_result *mode_support_result = &in_out->display_cfg->mode_support_result;
+ /* assumes DF throttling is enabled */
min_uclk_avg = dram_bw_kbps_to_uclk_khz(mode_support_result->global.svp_prefetch.average_bw_dram_kbps, &in_out->soc_bb->clk_table.dram_config);
min_uclk_avg = (double)min_uclk_avg / ((double)in_out->soc_bb->qos_parameters.derate_table.dcn_mall_prefetch_average.dram_derate_percent_pixel / 100);
@@ -125,6 +126,37 @@ static void calculate_svp_prefetch_minimums(struct dml2_dpmm_map_mode_to_soc_dpm
in_out->programming->min_clocks.dcn4x.svp_prefetch.uclk_khz = dml_round_up(min_uclk_bw > min_uclk_latency ? min_uclk_bw : min_uclk_latency);
in_out->programming->min_clocks.dcn4x.svp_prefetch.fclk_khz = dml_round_up(min_fclk_bw > min_fclk_latency ? min_fclk_bw : min_fclk_latency);
in_out->programming->min_clocks.dcn4x.svp_prefetch.dcfclk_khz = dml_round_up(min_dcfclk_bw > min_dcfclk_latency ? min_dcfclk_bw : min_dcfclk_latency);
+
+ /* assumes DF throttling is disabled */
+ min_uclk_avg = dram_bw_kbps_to_uclk_khz(mode_support_result->global.svp_prefetch.average_bw_dram_kbps, &in_out->soc_bb->clk_table.dram_config);
+ min_uclk_avg = (double)min_uclk_avg / ((double)in_out->soc_bb->qos_parameters.derate_table.system_active_average.dram_derate_percent_pixel / 100);
+
+ min_uclk_urgent = dram_bw_kbps_to_uclk_khz(mode_support_result->global.svp_prefetch.urgent_bw_dram_kbps, &in_out->soc_bb->clk_table.dram_config);
+ min_uclk_urgent = (double)min_uclk_urgent / ((double)in_out->soc_bb->qos_parameters.derate_table.system_active_urgent.dram_derate_percent_pixel / 100);
+
+ min_uclk_bw = min_uclk_urgent > min_uclk_avg ? min_uclk_urgent : min_uclk_avg;
+
+ min_fclk_avg = (double)mode_support_result->global.svp_prefetch.average_bw_sdp_kbps / in_out->soc_bb->fabric_datapath_to_dcn_data_return_bytes;
+ min_fclk_avg = (double)min_fclk_avg / ((double)in_out->soc_bb->qos_parameters.derate_table.system_active_average.fclk_derate_percent / 100);
+
+ min_fclk_urgent = (double)mode_support_result->global.svp_prefetch.urgent_bw_sdp_kbps / in_out->soc_bb->fabric_datapath_to_dcn_data_return_bytes;
+ min_fclk_urgent = (double)min_fclk_urgent / ((double)in_out->soc_bb->qos_parameters.derate_table.system_active_urgent.fclk_derate_percent / 100);
+
+ min_fclk_bw = min_fclk_urgent > min_fclk_avg ? min_fclk_urgent : min_fclk_avg;
+
+ min_dcfclk_avg = (double)mode_support_result->global.svp_prefetch.average_bw_sdp_kbps / in_out->soc_bb->return_bus_width_bytes;
+ min_dcfclk_avg = (double)min_dcfclk_avg / ((double)in_out->soc_bb->qos_parameters.derate_table.system_active_average.dcfclk_derate_percent / 100);
+
+ min_dcfclk_urgent = (double)mode_support_result->global.svp_prefetch.urgent_bw_sdp_kbps / in_out->soc_bb->return_bus_width_bytes;
+ min_dcfclk_urgent = (double)min_dcfclk_urgent / ((double)in_out->soc_bb->qos_parameters.derate_table.system_active_urgent.dcfclk_derate_percent / 100);
+
+ min_dcfclk_bw = min_dcfclk_urgent > min_dcfclk_avg ? min_dcfclk_urgent : min_dcfclk_avg;
+
+ get_minimum_clocks_for_latency(in_out, &min_uclk_latency, &min_fclk_latency, &min_dcfclk_latency);
+
+ in_out->programming->min_clocks.dcn4x.svp_prefetch_no_throttle.uclk_khz = dml_round_up(min_uclk_bw > min_uclk_latency ? min_uclk_bw : min_uclk_latency);
+ in_out->programming->min_clocks.dcn4x.svp_prefetch_no_throttle.fclk_khz = dml_round_up(min_fclk_bw > min_fclk_latency ? min_fclk_bw : min_fclk_latency);
+ in_out->programming->min_clocks.dcn4x.svp_prefetch_no_throttle.dcfclk_khz = dml_round_up(min_dcfclk_bw > min_dcfclk_latency ? min_dcfclk_bw : min_dcfclk_latency);
}
static void calculate_idle_minimums(struct dml2_dpmm_map_mode_to_soc_dpm_params_in_out *in_out)
@@ -272,6 +304,17 @@ static bool map_soc_min_clocks_to_dpm_fine_grained(struct dml2_display_cfg_progr
if (result)
result = round_up_to_next_dpm(&display_cfg->min_clocks.dcn4x.idle.uclk_khz, &state_table->uclk);
+ /* these clocks are optional, so they can fail to map, in which case map all to 0 */
+ if (result) {
+ if (!round_up_to_next_dpm(&display_cfg->min_clocks.dcn4x.svp_prefetch_no_throttle.dcfclk_khz, &state_table->dcfclk) ||
+ !round_up_to_next_dpm(&display_cfg->min_clocks.dcn4x.svp_prefetch_no_throttle.fclk_khz, &state_table->fclk) ||
+ !round_up_to_next_dpm(&display_cfg->min_clocks.dcn4x.svp_prefetch_no_throttle.uclk_khz, &state_table->uclk)) {
+ display_cfg->min_clocks.dcn4x.svp_prefetch_no_throttle.dcfclk_khz = 0;
+ display_cfg->min_clocks.dcn4x.svp_prefetch_no_throttle.fclk_khz = 0;
+ display_cfg->min_clocks.dcn4x.svp_prefetch_no_throttle.uclk_khz = 0;
+ }
+ }
+
return result;
}
@@ -346,9 +389,6 @@ static bool map_min_clocks_to_dpm(const struct dml2_core_mode_support_result *mo
if (result)
result = round_up_to_next_dpm(&display_cfg->min_clocks.dcn4x.dispclk_khz, &state_table->dispclk);
- if (result)
- result = round_up_to_next_dpm(&display_cfg->min_clocks.dcn4x.deepsleep_dcfclk_khz, &state_table->dcfclk);
-
for (i = 0; i < DML2_MAX_DCN_PIPES; i++) {
if (result)
result = round_up_to_next_dpm(&display_cfg->plane_programming[i].min_clocks.dcn4x.dppclk_khz, &state_table->dppclk);
@@ -374,11 +414,11 @@ static bool map_min_clocks_to_dpm(const struct dml2_core_mode_support_result *mo
static bool are_timings_trivially_synchronizable(struct dml2_display_cfg *display_config, int mask)
{
- unsigned char i;
+ unsigned int i;
bool identical = true;
bool contains_drr = false;
- unsigned char remap_array[DML2_MAX_PLANES];
- unsigned char remap_array_size = 0;
+ unsigned int remap_array[DML2_MAX_PLANES];
+ unsigned int remap_array_size = 0;
// Create a remap array to enable simple iteration through only masked stream indicies
for (i = 0; i < display_config->num_streams; i++) {
@@ -413,10 +453,10 @@ static bool are_timings_trivially_synchronizable(struct dml2_display_cfg *displa
static int find_smallest_idle_time_in_vblank_us(struct dml2_dpmm_map_mode_to_soc_dpm_params_in_out *in_out, int mask)
{
- unsigned char i;
+ unsigned int i;
int min_idle_us = 0;
- unsigned char remap_array[DML2_MAX_PLANES];
- unsigned char remap_array_size = 0;
+ unsigned int remap_array[DML2_MAX_PLANES];
+ unsigned int remap_array_size = 0;
const struct dml2_core_mode_support_result *mode_support_result = &in_out->display_cfg->mode_support_result;
// Create a remap array to enable simple iteration through only masked stream indicies
@@ -711,7 +751,9 @@ bool dpmm_dcn4_map_watermarks(struct dml2_dpmm_map_watermarks_params_in_out *in_
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].fclk_pstate = (int unsigned)(mode_lib->mp.Watermark.FCLKChangeWatermark * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].sr_enter = (int unsigned)(mode_lib->mp.Watermark.StutterEnterPlusExitWatermark * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].sr_exit = (int unsigned)(mode_lib->mp.Watermark.StutterExitWatermark * refclk_freq_in_mhz);
- dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].temp_read_or_ppt = (int unsigned)(mode_lib->mp.Watermark.g6_temp_read_watermark_us * refclk_freq_in_mhz);
+ dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].sr_enter_z8 = (int unsigned)(mode_lib->mp.Watermark.Z8StutterEnterPlusExitWatermark * refclk_freq_in_mhz);
+ dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].sr_exit_z8 = (int unsigned)(mode_lib->mp.Watermark.Z8StutterExitWatermark * refclk_freq_in_mhz);
+ dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].temp_read_or_ppt = (int unsigned)(mode_lib->mp.Watermark.temp_read_or_ppt_watermark_us * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].uclk_pstate = (int unsigned)(mode_lib->mp.Watermark.DRAMClockChangeWatermark * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].urgent = (int unsigned)(mode_lib->mp.Watermark.UrgentWatermark * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_A].usr = (int unsigned)(mode_lib->mp.Watermark.USRRetrainingWatermark * refclk_freq_in_mhz);
@@ -725,7 +767,9 @@ bool dpmm_dcn4_map_watermarks(struct dml2_dpmm_map_watermarks_params_in_out *in_
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].fclk_pstate = (int unsigned)(mode_lib->mp.Watermark.FCLKChangeWatermark * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].sr_enter = (int unsigned)(mode_lib->mp.Watermark.StutterEnterPlusExitWatermark * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].sr_exit = (int unsigned)(mode_lib->mp.Watermark.StutterExitWatermark * refclk_freq_in_mhz);
- dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].temp_read_or_ppt = (int unsigned)(mode_lib->mp.Watermark.g6_temp_read_watermark_us * refclk_freq_in_mhz);
+ dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].sr_enter_z8 = (int unsigned)(mode_lib->mp.Watermark.Z8StutterEnterPlusExitWatermark * refclk_freq_in_mhz);
+ dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].sr_exit_z8 = (int unsigned)(mode_lib->mp.Watermark.Z8StutterExitWatermark * refclk_freq_in_mhz);
+ dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].temp_read_or_ppt = (int unsigned)(mode_lib->mp.Watermark.temp_read_or_ppt_watermark_us * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].uclk_pstate = (int unsigned)(mode_lib->mp.Watermark.DRAMClockChangeWatermark * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].urgent = (int unsigned)(mode_lib->mp.Watermark.UrgentWatermark * refclk_freq_in_mhz);
dchubbub_regs->wm_regs[DML2_DCHUB_WATERMARK_SET_B].usr = (int unsigned)(mode_lib->mp.Watermark.USRRetrainingWatermark * refclk_freq_in_mhz);
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.h
index b165c58dfd11..e7b58f2efda4 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_dpmm/dml2_dpmm_dcn4.h
@@ -11,6 +11,4 @@ bool dpmm_dcn3_map_mode_to_soc_dpm(struct dml2_dpmm_map_mode_to_soc_dpm_params_i
bool dpmm_dcn4_map_mode_to_soc_dpm(struct dml2_dpmm_map_mode_to_soc_dpm_params_in_out *in_out);
bool dpmm_dcn4_map_watermarks(struct dml2_dpmm_map_watermarks_params_in_out *in_out);
-bool dpmm_dcn4_unit_test(void);
-
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_dcn4.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_dcn4.c
index f4b1a7d02d42..a265f254152c 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_dcn4.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_dcn4.c
@@ -182,6 +182,10 @@ static bool build_min_clock_table(const struct dml2_soc_bb *soc_bb, struct dml2_
min_table->max_clocks_khz.dtbclk = soc_bb->clk_table.dtbclk.clk_values_khz[soc_bb->clk_table.dtbclk.num_clk_values - 1];
min_table->max_clocks_khz.phyclk = soc_bb->clk_table.phyclk.clk_values_khz[soc_bb->clk_table.phyclk.num_clk_values - 1];
+ min_table->max_ss_clocks_khz.dispclk = (unsigned int)((double)min_table->max_clocks_khz.dispclk / (1.0 + soc_bb->dcn_downspread_percent / 100.0));
+ min_table->max_ss_clocks_khz.dppclk = (unsigned int)((double)min_table->max_clocks_khz.dppclk / (1.0 + soc_bb->dcn_downspread_percent / 100.0));
+ min_table->max_ss_clocks_khz.dtbclk = (unsigned int)((double)min_table->max_clocks_khz.dtbclk / (1.0 + soc_bb->dcn_downspread_percent / 100.0));
+
min_table->max_clocks_khz.dcfclk = soc_bb->clk_table.dcfclk.clk_values_khz[soc_bb->clk_table.dcfclk.num_clk_values - 1];
min_table->max_clocks_khz.fclk = soc_bb->clk_table.fclk.clk_values_khz[soc_bb->clk_table.fclk.num_clk_values - 1];
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_factory.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_factory.c
index c60b8fe90819..cd3fbc0591d8 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_factory.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_mcg/dml2_mcg_factory.c
@@ -15,7 +15,7 @@ bool dml2_mcg_create(enum dml2_project_id project_id, struct dml2_mcg_instance *
{
bool result = false;
- if (!out)
+ if (out == 0)
return false;
memset(out, 0, sizeof(struct dml2_mcg_instance));
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn3.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn3.c
index a31db5742675..e763c8e45da8 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn3.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn3.c
@@ -195,11 +195,11 @@ static int count_planes_with_stream_index(const struct dml2_display_cfg *display
static bool are_timings_trivially_synchronizable(struct display_configuation_with_meta *display_config, int mask)
{
- unsigned char i;
+ unsigned int i;
bool identical = true;
bool contains_drr = false;
- unsigned char remap_array[DML2_MAX_PLANES];
- unsigned char remap_array_size = 0;
+ unsigned int remap_array[DML2_MAX_PLANES];
+ unsigned int remap_array_size = 0;
// Create a remap array to enable simple iteration through only masked stream indicies
for (i = 0; i < display_config->display_config.num_streams; i++) {
@@ -347,8 +347,12 @@ static int find_highest_odm_load_stream_index(
int odm_load, highest_odm_load = -1, highest_odm_load_index = -1;
for (i = 0; i < display_config->num_streams; i++) {
- odm_load = display_config->stream_descriptors[i].timing.pixel_clock_khz
+ if (mode_support_result->cfg_support_info.stream_support_info[i].odms_used > 0)
+ odm_load = display_config->stream_descriptors[i].timing.pixel_clock_khz
/ mode_support_result->cfg_support_info.stream_support_info[i].odms_used;
+ else
+ odm_load = 0;
+
if (odm_load > highest_odm_load) {
highest_odm_load_index = i;
highest_odm_load = odm_load;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c
index 92269f0e50ed..d88b3e0082dd 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.c
@@ -13,32 +13,32 @@ static const double MIN_BLANK_STUTTER_FACTOR = 3.0;
static const struct dml2_pmo_pstate_strategy base_strategy_list_1_display[] = {
// VActive Preferred
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vactive, dml2_pstate_method_na, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
// Then SVP
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_fw_svp, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_fw_svp, dml2_pstate_method_na, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
// Then VBlank
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vblank, dml2_pstate_method_na, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = false,
},
// Then DRR
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_fw_drr, dml2_pstate_method_na, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
// Finally VBlank, but allow base clocks for latency to increase
/*
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vblank, dml2_pstate_method_na, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
*/
@@ -49,56 +49,56 @@ static const int base_strategy_list_1_display_size = sizeof(base_strategy_list_1
static const struct dml2_pmo_pstate_strategy base_strategy_list_2_display[] = {
// VActive only is preferred
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vactive, dml2_pstate_method_vactive, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
// Then VActive + VBlank
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vactive, dml2_pstate_method_vblank, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = false,
},
// Then VBlank only
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = false,
},
// Then SVP + VBlank
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_fw_svp, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_fw_svp, dml2_pstate_method_vblank, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = false,
},
// Then SVP + DRR
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_fw_svp, dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_fw_svp, dml2_pstate_method_fw_drr, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
// Then SVP + SVP
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_fw_svp, dml2_pmo_pstate_strategy_fw_svp, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_fw_svp, dml2_pstate_method_fw_svp, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
// Then DRR + VActive
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vactive, dml2_pstate_method_fw_drr, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
// Then DRR + DRR
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_fw_drr, dml2_pstate_method_fw_drr, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
// Finally VBlank, but allow base clocks for latency to increase
/*
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_na, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_na, dml2_pstate_method_na },
.allow_state_increase = true,
},
*/
@@ -109,32 +109,32 @@ static const int base_strategy_list_2_display_size = sizeof(base_strategy_list_2
static const struct dml2_pmo_pstate_strategy base_strategy_list_3_display[] = {
// All VActive
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vactive, dml2_pstate_method_vactive, dml2_pstate_method_vactive, dml2_pstate_method_na },
.allow_state_increase = true,
},
// VActive + 1 VBlank
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vactive, dml2_pstate_method_vactive, dml2_pstate_method_vblank, dml2_pstate_method_na },
.allow_state_increase = false,
},
// All VBlank
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_na },
.allow_state_increase = false,
},
// All DRR
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_fw_drr, dml2_pstate_method_fw_drr, dml2_pstate_method_fw_drr, dml2_pstate_method_na },
.allow_state_increase = true,
},
// All VBlank, with state increase allowed
/*
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_na },
+ .per_stream_pstate_method = { dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_na },
.allow_state_increase = true,
},
*/
@@ -145,32 +145,32 @@ static const int base_strategy_list_3_display_size = sizeof(base_strategy_list_3
static const struct dml2_pmo_pstate_strategy base_strategy_list_4_display[] = {
// All VActive
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vactive },
+ .per_stream_pstate_method = { dml2_pstate_method_vactive, dml2_pstate_method_vactive, dml2_pstate_method_vactive, dml2_pstate_method_vactive },
.allow_state_increase = true,
},
// VActive + 1 VBlank
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vactive, dml2_pmo_pstate_strategy_vblank },
+ .per_stream_pstate_method = { dml2_pstate_method_vactive, dml2_pstate_method_vactive, dml2_pstate_method_vactive, dml2_pstate_method_vblank },
.allow_state_increase = false,
},
// All Vblank
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank },
+ .per_stream_pstate_method = { dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_vblank },
.allow_state_increase = false,
},
// All DRR
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_fw_drr, dml2_pmo_pstate_strategy_fw_drr },
+ .per_stream_pstate_method = { dml2_pstate_method_fw_drr, dml2_pstate_method_fw_drr, dml2_pstate_method_fw_drr, dml2_pstate_method_fw_drr },
.allow_state_increase = true,
},
// All VBlank, with state increase allowed
/*
{
- .per_stream_pstate_method = { dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank, dml2_pmo_pstate_strategy_vblank },
+ .per_stream_pstate_method = { dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_vblank, dml2_pstate_method_vblank },
.allow_state_increase = true,
},
*/
@@ -355,29 +355,30 @@ bool pmo_dcn4_fams2_optimize_dcc_mcache(struct dml2_pmo_optimize_dcc_mcache_in_o
return result;
}
-static enum dml2_pmo_pstate_method convert_strategy_to_drr_variant(const enum dml2_pmo_pstate_method base_strategy)
+static enum dml2_pstate_method convert_strategy_to_drr_variant(const enum dml2_pstate_method base_strategy)
{
- enum dml2_pmo_pstate_method variant_strategy = 0;
+ enum dml2_pstate_method variant_strategy = 0;
switch (base_strategy) {
- case dml2_pmo_pstate_strategy_vactive:
- variant_strategy = dml2_pmo_pstate_strategy_fw_vactive_drr;
+ case dml2_pstate_method_vactive:
+ variant_strategy = dml2_pstate_method_fw_vactive_drr;
break;
- case dml2_pmo_pstate_strategy_vblank:
- variant_strategy = dml2_pmo_pstate_strategy_fw_vblank_drr;
+ case dml2_pstate_method_vblank:
+ variant_strategy = dml2_pstate_method_fw_vblank_drr;
break;
- case dml2_pmo_pstate_strategy_fw_svp:
- variant_strategy = dml2_pmo_pstate_strategy_fw_svp_drr;
+ case dml2_pstate_method_fw_svp:
+ variant_strategy = dml2_pstate_method_fw_svp_drr;
break;
- case dml2_pmo_pstate_strategy_fw_vactive_drr:
- case dml2_pmo_pstate_strategy_fw_vblank_drr:
- case dml2_pmo_pstate_strategy_fw_svp_drr:
- case dml2_pmo_pstate_strategy_fw_drr:
- case dml2_pmo_pstate_strategy_reserved_hw:
- case dml2_pmo_pstate_strategy_reserved_fw:
- case dml2_pmo_pstate_strategy_reserved_fw_drr_clamped:
- case dml2_pmo_pstate_strategy_reserved_fw_drr_var:
- case dml2_pmo_pstate_strategy_na:
+ case dml2_pstate_method_fw_vactive_drr:
+ case dml2_pstate_method_fw_vblank_drr:
+ case dml2_pstate_method_fw_svp_drr:
+ case dml2_pstate_method_fw_drr:
+ case dml2_pstate_method_reserved_hw:
+ case dml2_pstate_method_reserved_fw:
+ case dml2_pstate_method_reserved_fw_drr_clamped:
+ case dml2_pstate_method_reserved_fw_drr_var:
+ case dml2_pstate_method_count:
+ case dml2_pstate_method_na:
default:
/* no variant for this mode */
variant_strategy = base_strategy;
@@ -419,23 +420,22 @@ static unsigned int get_num_expanded_strategies(
static void insert_strategy_into_expanded_list(
const struct dml2_pmo_pstate_strategy *per_stream_pstate_strategy,
- int stream_count,
- struct dml2_pmo_init_data *init_data)
+ const int stream_count,
+ struct dml2_pmo_pstate_strategy *expanded_strategy_list,
+ unsigned int *num_expanded_strategies)
{
- struct dml2_pmo_pstate_strategy *expanded_strategy_list = NULL;
-
- expanded_strategy_list = get_expanded_strategy_list(init_data, stream_count);
+ if (expanded_strategy_list && num_expanded_strategies) {
+ memcpy(&expanded_strategy_list[*num_expanded_strategies], per_stream_pstate_strategy, sizeof(struct dml2_pmo_pstate_strategy));
- if (expanded_strategy_list) {
- memcpy(&expanded_strategy_list[init_data->pmo_dcn4.num_expanded_strategies_per_list[stream_count - 1]], per_stream_pstate_strategy, sizeof(struct dml2_pmo_pstate_strategy));
-
- init_data->pmo_dcn4.num_expanded_strategies_per_list[stream_count - 1]++;
+ (*num_expanded_strategies)++;
}
}
-static void expand_base_strategy(struct dml2_pmo_instance *pmo,
+static void expand_base_strategy(
const struct dml2_pmo_pstate_strategy *base_strategy,
- unsigned int stream_count)
+ const unsigned int stream_count,
+ struct dml2_pmo_pstate_strategy *expanded_strategy_list,
+ unsigned int *num_expanded_strategies)
{
bool skip_to_next_stream;
bool expanded_strategy_added;
@@ -473,7 +473,7 @@ static void expand_base_strategy(struct dml2_pmo_instance *pmo,
if (i >= stream_count - 1) {
/* insert into strategy list */
- insert_strategy_into_expanded_list(&cur_strategy_list, stream_count, &pmo->init_data);
+ insert_strategy_into_expanded_list(&cur_strategy_list, stream_count, expanded_strategy_list, num_expanded_strategies);
expanded_strategy_added = true;
} else {
/* skip to next stream */
@@ -512,9 +512,9 @@ static void expand_base_strategy(struct dml2_pmo_instance *pmo,
static bool is_variant_method_valid(const struct dml2_pmo_pstate_strategy *base_strategy,
const struct dml2_pmo_pstate_strategy *variant_strategy,
- unsigned int num_streams_per_base_method[PMO_DCN4_MAX_DISPLAYS],
- unsigned int num_streams_per_variant_method[PMO_DCN4_MAX_DISPLAYS],
- unsigned int stream_count)
+ const unsigned int num_streams_per_base_method[PMO_DCN4_MAX_DISPLAYS],
+ const unsigned int num_streams_per_variant_method[PMO_DCN4_MAX_DISPLAYS],
+ const unsigned int stream_count)
{
bool valid = true;
unsigned int i;
@@ -522,7 +522,7 @@ static bool is_variant_method_valid(const struct dml2_pmo_pstate_strategy *base_
/* check all restrictions are met */
for (i = 0; i < stream_count; i++) {
/* vblank + vblank_drr variants are invalid */
- if (base_strategy->per_stream_pstate_method[i] == dml2_pmo_pstate_strategy_vblank &&
+ if (base_strategy->per_stream_pstate_method[i] == dml2_pstate_method_vblank &&
((num_streams_per_base_method[i] > 0 && num_streams_per_variant_method[i] > 0) ||
num_streams_per_variant_method[i] > 1)) {
valid = false;
@@ -533,9 +533,12 @@ static bool is_variant_method_valid(const struct dml2_pmo_pstate_strategy *base_
return valid;
}
-static void expand_variant_strategy(struct dml2_pmo_instance *pmo,
+static void expand_variant_strategy(
const struct dml2_pmo_pstate_strategy *base_strategy,
- unsigned int stream_count)
+ const unsigned int stream_count,
+ const bool should_permute,
+ struct dml2_pmo_pstate_strategy *expanded_strategy_list,
+ unsigned int *num_expanded_strategies)
{
bool variant_found;
unsigned int i, j;
@@ -544,7 +547,7 @@ static void expand_variant_strategy(struct dml2_pmo_instance *pmo,
unsigned int num_streams_per_method[PMO_DCN4_MAX_DISPLAYS] = { 0 };
unsigned int num_streams_per_base_method[PMO_DCN4_MAX_DISPLAYS] = { 0 };
unsigned int num_streams_per_variant_method[PMO_DCN4_MAX_DISPLAYS] = { 0 };
- enum dml2_pmo_pstate_method per_stream_variant_method[DML2_MAX_PLANES];
+ enum dml2_pstate_method per_stream_variant_method[DML2_MAX_PLANES];
struct dml2_pmo_pstate_strategy variant_strategy = { 0 };
/* determine number of displays per method */
@@ -585,7 +588,13 @@ static void expand_variant_strategy(struct dml2_pmo_instance *pmo,
}
if (variant_found && is_variant_method_valid(base_strategy, &variant_strategy, num_streams_per_base_method, num_streams_per_variant_method, stream_count)) {
- expand_base_strategy(pmo, &variant_strategy, stream_count);
+ if (should_permute) {
+ /* permutations are permitted, proceed to expand */
+ expand_base_strategy(&variant_strategy, stream_count, expanded_strategy_list, num_expanded_strategies);
+ } else {
+ /* no permutations allowed, so add to list now */
+ insert_strategy_into_expanded_list(&variant_strategy, stream_count, expanded_strategy_list, num_expanded_strategies);
+ }
}
/* rollback to earliest method with bases remaining */
@@ -612,18 +621,19 @@ static void expand_variant_strategy(struct dml2_pmo_instance *pmo,
}
}
-static void expand_base_strategies(
- struct dml2_pmo_instance *pmo,
- const struct dml2_pmo_pstate_strategy *base_strategies_list,
- const unsigned int num_base_strategies,
- unsigned int stream_count)
+void pmo_dcn4_fams2_expand_base_pstate_strategies(
+ const struct dml2_pmo_pstate_strategy *base_strategies_list,
+ const unsigned int num_base_strategies,
+ const unsigned int stream_count,
+ struct dml2_pmo_pstate_strategy *expanded_strategy_list,
+ unsigned int *num_expanded_strategies)
{
unsigned int i;
/* expand every explicit base strategy (except all DRR) */
for (i = 0; i < num_base_strategies; i++) {
- expand_base_strategy(pmo, &base_strategies_list[i], stream_count);
- expand_variant_strategy(pmo, &base_strategies_list[i], stream_count);
+ expand_base_strategy(&base_strategies_list[i], stream_count, expanded_strategy_list, num_expanded_strategies);
+ expand_variant_strategy(&base_strategies_list[i], stream_count, true, expanded_strategy_list, num_expanded_strategies);
}
}
@@ -649,28 +659,48 @@ bool pmo_dcn4_fams2_initialize(struct dml2_pmo_initialize_in_out *in_out)
for (i = 1; i <= PMO_DCN4_MAX_DISPLAYS; i++) {
switch (i) {
case 1:
- DML2_ASSERT(base_strategy_list_1_display_size <= PMO_DCN4_MAX_BASE_STRATEGIES);
+ DML_ASSERT(base_strategy_list_1_display_size <= PMO_DCN4_MAX_BASE_STRATEGIES);
/* populate list */
- expand_base_strategies(pmo, base_strategy_list_1_display, base_strategy_list_1_display_size, 1);
+ pmo_dcn4_fams2_expand_base_pstate_strategies(
+ base_strategy_list_1_display,
+ base_strategy_list_1_display_size,
+ i,
+ pmo->init_data.pmo_dcn4.expanded_strategy_list_1_display,
+ &pmo->init_data.pmo_dcn4.num_expanded_strategies_per_list[i - 1]);
break;
case 2:
- DML2_ASSERT(base_strategy_list_2_display_size <= PMO_DCN4_MAX_BASE_STRATEGIES);
+ DML_ASSERT(base_strategy_list_2_display_size <= PMO_DCN4_MAX_BASE_STRATEGIES);
/* populate list */
- expand_base_strategies(pmo, base_strategy_list_2_display, base_strategy_list_2_display_size, 2);
+ pmo_dcn4_fams2_expand_base_pstate_strategies(
+ base_strategy_list_2_display,
+ base_strategy_list_2_display_size,
+ i,
+ pmo->init_data.pmo_dcn4.expanded_strategy_list_2_display,
+ &pmo->init_data.pmo_dcn4.num_expanded_strategies_per_list[i - 1]);
break;
case 3:
- DML2_ASSERT(base_strategy_list_3_display_size <= PMO_DCN4_MAX_BASE_STRATEGIES);
+ DML_ASSERT(base_strategy_list_3_display_size <= PMO_DCN4_MAX_BASE_STRATEGIES);
/* populate list */
- expand_base_strategies(pmo, base_strategy_list_3_display, base_strategy_list_3_display_size, 3);
+ pmo_dcn4_fams2_expand_base_pstate_strategies(
+ base_strategy_list_3_display,
+ base_strategy_list_3_display_size,
+ i,
+ pmo->init_data.pmo_dcn4.expanded_strategy_list_3_display,
+ &pmo->init_data.pmo_dcn4.num_expanded_strategies_per_list[i - 1]);
break;
case 4:
- DML2_ASSERT(base_strategy_list_4_display_size <= PMO_DCN4_MAX_BASE_STRATEGIES);
+ DML_ASSERT(base_strategy_list_4_display_size <= PMO_DCN4_MAX_BASE_STRATEGIES);
/* populate list */
- expand_base_strategies(pmo, base_strategy_list_4_display, base_strategy_list_4_display_size, 4);
+ pmo_dcn4_fams2_expand_base_pstate_strategies(
+ base_strategy_list_4_display,
+ base_strategy_list_4_display_size,
+ i,
+ pmo->init_data.pmo_dcn4.expanded_strategy_list_4_display,
+ &pmo->init_data.pmo_dcn4.num_expanded_strategies_per_list[i - 1]);
break;
}
}
@@ -783,8 +813,12 @@ static int find_highest_odm_load_stream_index(
int odm_load, highest_odm_load = -1, highest_odm_load_index = -1;
for (i = 0; i < display_config->num_streams; i++) {
- odm_load = display_config->stream_descriptors[i].timing.pixel_clock_khz
+ if (mode_support_result->cfg_support_info.stream_support_info[i].odms_used > 0)
+ odm_load = display_config->stream_descriptors[i].timing.pixel_clock_khz
/ mode_support_result->cfg_support_info.stream_support_info[i].odms_used;
+ else
+ odm_load = 0;
+
if (odm_load > highest_odm_load) {
highest_odm_load_index = i;
highest_odm_load = odm_load;
@@ -941,11 +975,8 @@ static void build_synchronized_timing_groups(
/* find synchronizable timing groups */
for (j = i + 1; j < display_config->display_config.num_streams; j++) {
if (memcmp(master_timing,
- &display_config->display_config.stream_descriptors[j].timing,
- sizeof(struct dml2_timing_cfg)) == 0 &&
- display_config->display_config.stream_descriptors[i].output.output_encoder == display_config->display_config.stream_descriptors[j].output.output_encoder &&
- (display_config->display_config.stream_descriptors[i].output.output_encoder != dml2_hdmi || //hdmi requires formats match
- display_config->display_config.stream_descriptors[i].output.output_format == display_config->display_config.stream_descriptors[j].output.output_format)) {
+ &display_config->display_config.stream_descriptors[j].timing,
+ sizeof(struct dml2_timing_cfg)) == 0) {
set_bit_in_bitfield(&pmo->scratch.pmo_dcn4.synchronized_timing_group_masks[timing_group_idx], j);
set_bit_in_bitfield(&stream_mapped_mask, j);
}
@@ -959,7 +990,7 @@ static bool all_timings_support_vactive(const struct dml2_pmo_instance *pmo,
const struct display_configuation_with_meta *display_config,
unsigned int mask)
{
- unsigned char i;
+ unsigned int i;
bool valid = true;
// Create a remap array to enable simple iteration through only masked stream indicies
@@ -1008,7 +1039,7 @@ static bool all_timings_support_drr(const struct dml2_pmo_instance *pmo,
const struct display_configuation_with_meta *display_config,
unsigned int mask)
{
- unsigned char i;
+ unsigned int i;
for (i = 0; i < DML2_MAX_PLANES; i++) {
const struct dml2_stream_parameters *stream_descriptor;
const struct dml2_fams2_meta *stream_fams2_meta;
@@ -1050,13 +1081,22 @@ static bool all_timings_support_svp(const struct dml2_pmo_instance *pmo,
const struct dml2_plane_parameters *plane_descriptor;
const struct dml2_fams2_meta *stream_fams2_meta;
unsigned int microschedule_vlines;
- unsigned char i;
+ unsigned int i;
+ unsigned int mcaches_per_plane;
+ unsigned int total_mcaches_required = 0;
unsigned int num_planes_per_stream[DML2_MAX_PLANES] = { 0 };
/* confirm timing it is not a centered timing */
for (i = 0; i < display_config->display_config.num_planes; i++) {
plane_descriptor = &display_config->display_config.plane_descriptors[i];
+ mcaches_per_plane = 0;
+
+ if (plane_descriptor->surface.dcc.enable) {
+ mcaches_per_plane += display_config->stage2.mcache_allocations[i].num_mcaches_plane0 +
+ display_config->stage2.mcache_allocations[i].num_mcaches_plane1 -
+ (display_config->stage2.mcache_allocations[i].last_slice_sharing.plane0_plane1 ? 1 : 0);
+ }
if (is_bit_set_in_bitfield(mask, (unsigned char)plane_descriptor->stream_index)) {
num_planes_per_stream[plane_descriptor->stream_index]++;
@@ -1067,7 +1107,18 @@ static bool all_timings_support_svp(const struct dml2_pmo_instance *pmo,
plane_descriptor->composition.rotation_angle != dml2_rotation_0) {
return false;
}
+
+ /* phantom requires same number of mcaches as main */
+ if (plane_descriptor->surface.dcc.enable) {
+ mcaches_per_plane *= 2;
+ }
}
+ total_mcaches_required += mcaches_per_plane;
+ }
+
+ if (total_mcaches_required > pmo->soc_bb->num_dcc_mcaches) {
+ /* too many mcaches required */
+ return false;
}
for (i = 0; i < DML2_MAX_PLANES; i++) {
@@ -1106,24 +1157,73 @@ static void insert_into_candidate_list(const struct dml2_pmo_pstate_strategy *ps
scratch->pmo_dcn4.num_pstate_candidates++;
}
-static bool all_planes_match_method(const struct display_configuation_with_meta *display_cfg, int plane_mask, enum dml2_pmo_pstate_method method)
+static enum dml2_pstate_method uclk_pstate_strategy_override_to_pstate_method(const enum dml2_uclk_pstate_change_strategy override_strategy)
{
- unsigned char i;
- enum dml2_uclk_pstate_change_strategy matching_strategy = (enum dml2_uclk_pstate_change_strategy) dml2_pmo_pstate_strategy_na;
+ enum dml2_pstate_method method = dml2_pstate_method_na;
+
+ switch (override_strategy) {
+ case dml2_uclk_pstate_change_strategy_force_vactive:
+ method = dml2_pstate_method_vactive;
+ break;
+ case dml2_uclk_pstate_change_strategy_force_vblank:
+ method = dml2_pstate_method_vblank;
+ break;
+ case dml2_uclk_pstate_change_strategy_force_drr:
+ method = dml2_pstate_method_fw_drr;
+ break;
+ case dml2_uclk_pstate_change_strategy_force_mall_svp:
+ method = dml2_pstate_method_fw_svp;
+ break;
+ case dml2_uclk_pstate_change_strategy_force_mall_full_frame:
+ case dml2_uclk_pstate_change_strategy_auto:
+ default:
+ method = dml2_pstate_method_na;
+ }
+
+ return method;
+}
- if (method == dml2_pmo_pstate_strategy_vactive || method == dml2_pmo_pstate_strategy_fw_vactive_drr)
- matching_strategy = dml2_uclk_pstate_change_strategy_force_vactive;
- else if (method == dml2_pmo_pstate_strategy_vblank || method == dml2_pmo_pstate_strategy_fw_vblank_drr)
- matching_strategy = dml2_uclk_pstate_change_strategy_force_vblank;
- else if (method == dml2_pmo_pstate_strategy_fw_svp)
- matching_strategy = dml2_uclk_pstate_change_strategy_force_mall_svp;
- else if (method == dml2_pmo_pstate_strategy_fw_drr)
- matching_strategy = dml2_uclk_pstate_change_strategy_force_drr;
+static enum dml2_uclk_pstate_change_strategy pstate_method_to_uclk_pstate_strategy_override(const enum dml2_pstate_method method)
+{
+ enum dml2_uclk_pstate_change_strategy override_strategy = dml2_uclk_pstate_change_strategy_auto;
+
+ switch (method) {
+ case dml2_pstate_method_vactive:
+ case dml2_pstate_method_fw_vactive_drr:
+ override_strategy = dml2_uclk_pstate_change_strategy_force_vactive;
+ break;
+ case dml2_pstate_method_vblank:
+ case dml2_pstate_method_fw_vblank_drr:
+ override_strategy = dml2_uclk_pstate_change_strategy_force_vblank;
+ break;
+ case dml2_pstate_method_fw_svp:
+ case dml2_pstate_method_fw_svp_drr:
+ override_strategy = dml2_uclk_pstate_change_strategy_force_mall_svp;
+ break;
+ case dml2_pstate_method_fw_drr:
+ override_strategy = dml2_uclk_pstate_change_strategy_force_drr;
+ break;
+ case dml2_pstate_method_reserved_hw:
+ case dml2_pstate_method_reserved_fw:
+ case dml2_pstate_method_reserved_fw_drr_clamped:
+ case dml2_pstate_method_reserved_fw_drr_var:
+ case dml2_pstate_method_count:
+ case dml2_pstate_method_na:
+ default:
+ override_strategy = dml2_uclk_pstate_change_strategy_auto;
+ }
+
+ return override_strategy;
+}
+
+static bool all_planes_match_method(const struct display_configuation_with_meta *display_cfg, int plane_mask, enum dml2_pstate_method method)
+{
+ unsigned int i;
for (i = 0; i < DML2_MAX_PLANES; i++) {
if (is_bit_set_in_bitfield(plane_mask, i)) {
if (display_cfg->display_config.plane_descriptors[i].overrides.uclk_pstate_change_strategy != dml2_uclk_pstate_change_strategy_auto &&
- display_cfg->display_config.plane_descriptors[i].overrides.uclk_pstate_change_strategy != matching_strategy)
+ display_cfg->display_config.plane_descriptors[i].overrides.uclk_pstate_change_strategy != pstate_method_to_uclk_pstate_strategy_override(method))
return false;
}
}
@@ -1149,32 +1249,33 @@ static void build_method_scheduling_params(
static struct dml2_fams2_per_method_common_meta *get_per_method_common_meta(
struct dml2_pmo_instance *pmo,
- enum dml2_pmo_pstate_method stream_pstate_method,
+ enum dml2_pstate_method stream_pstate_method,
int stream_idx)
{
struct dml2_fams2_per_method_common_meta *stream_method_fams2_meta = NULL;
switch (stream_pstate_method) {
- case dml2_pmo_pstate_strategy_vactive:
- case dml2_pmo_pstate_strategy_fw_vactive_drr:
+ case dml2_pstate_method_vactive:
+ case dml2_pstate_method_fw_vactive_drr:
stream_method_fams2_meta = &pmo->scratch.pmo_dcn4.stream_fams2_meta[stream_idx].method_vactive.common;
break;
- case dml2_pmo_pstate_strategy_vblank:
- case dml2_pmo_pstate_strategy_fw_vblank_drr:
+ case dml2_pstate_method_vblank:
+ case dml2_pstate_method_fw_vblank_drr:
stream_method_fams2_meta = &pmo->scratch.pmo_dcn4.stream_fams2_meta[stream_idx].method_vblank.common;
break;
- case dml2_pmo_pstate_strategy_fw_svp:
- case dml2_pmo_pstate_strategy_fw_svp_drr:
+ case dml2_pstate_method_fw_svp:
+ case dml2_pstate_method_fw_svp_drr:
stream_method_fams2_meta = &pmo->scratch.pmo_dcn4.stream_fams2_meta[stream_idx].method_subvp.common;
break;
- case dml2_pmo_pstate_strategy_fw_drr:
+ case dml2_pstate_method_fw_drr:
stream_method_fams2_meta = &pmo->scratch.pmo_dcn4.stream_fams2_meta[stream_idx].method_drr.common;
break;
- case dml2_pmo_pstate_strategy_reserved_hw:
- case dml2_pmo_pstate_strategy_reserved_fw:
- case dml2_pmo_pstate_strategy_reserved_fw_drr_clamped:
- case dml2_pmo_pstate_strategy_reserved_fw_drr_var:
- case dml2_pmo_pstate_strategy_na:
+ case dml2_pstate_method_reserved_hw:
+ case dml2_pstate_method_reserved_fw:
+ case dml2_pstate_method_reserved_fw_drr_clamped:
+ case dml2_pstate_method_reserved_fw_drr_var:
+ case dml2_pstate_method_count:
+ case dml2_pstate_method_na:
default:
stream_method_fams2_meta = NULL;
}
@@ -1215,7 +1316,7 @@ static bool is_timing_group_schedulable(
if (is_bit_set_in_bitfield(pmo->scratch.pmo_dcn4.synchronized_timing_group_masks[timing_group_idx], i)) {
stream_method_fams2_meta = get_per_method_common_meta(pmo, pstate_strategy->per_stream_pstate_method[i], i);
if (!stream_method_fams2_meta)
- return false;
+ continue;
if (group_fams2_meta->allow_start_otg_vline < stream_method_fams2_meta->allow_start_otg_vline) {
/* set group allow start to larger otg vline */
@@ -1295,7 +1396,7 @@ static bool is_config_schedulable(
if (j_disallow_us < jp1_disallow_us) {
/* swap as A < B */
swap(s->pmo_dcn4.sorted_group_gtl_disallow_index[j],
- s->pmo_dcn4.sorted_group_gtl_disallow_index[j+1]);
+ s->pmo_dcn4.sorted_group_gtl_disallow_index[j + 1]);
swapped = true;
}
}
@@ -1354,7 +1455,7 @@ static bool is_config_schedulable(
if (j_period_us < jp1_period_us) {
/* swap as A < B */
swap(s->pmo_dcn4.sorted_group_gtl_period_index[j],
- s->pmo_dcn4.sorted_group_gtl_period_index[j+1]);
+ s->pmo_dcn4.sorted_group_gtl_period_index[j + 1]);
swapped = true;
}
}
@@ -1413,7 +1514,7 @@ static bool is_config_schedulable(
static bool stream_matches_drr_policy(struct dml2_pmo_instance *pmo,
const struct display_configuation_with_meta *display_cfg,
- const enum dml2_pmo_pstate_method stream_pstate_method,
+ const enum dml2_pstate_method stream_pstate_method,
unsigned int stream_index)
{
const struct dml2_stream_parameters *stream_descriptor = &display_cfg->display_config.stream_descriptors[stream_index];
@@ -1468,7 +1569,7 @@ static bool validate_pstate_support_strategy_cofunctionality(struct dml2_pmo_ins
{
struct dml2_pmo_scratch *s = &pmo->scratch;
- unsigned char stream_index = 0;
+ unsigned int stream_index = 0;
unsigned int svp_count = 0;
unsigned int svp_stream_mask = 0;
@@ -1494,19 +1595,19 @@ static bool validate_pstate_support_strategy_cofunctionality(struct dml2_pmo_ins
strategy_matches_drr_requirements &=
stream_matches_drr_policy(pmo, display_cfg, pstate_strategy->per_stream_pstate_method[stream_index], stream_index);
- if (pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_svp ||
- pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_svp_drr) {
+ if (pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_svp ||
+ pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_svp_drr) {
svp_count++;
set_bit_in_bitfield(&svp_stream_mask, stream_index);
- } else if (pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_drr) {
+ } else if (pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_drr) {
drr_count++;
set_bit_in_bitfield(&drr_stream_mask, stream_index);
- } else if (pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_vactive ||
- pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_vactive_drr) {
+ } else if (pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pstate_method_vactive ||
+ pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_vactive_drr) {
vactive_count++;
set_bit_in_bitfield(&vactive_stream_mask, stream_index);
- } else if (pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_vblank ||
- pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_vblank_drr) {
+ } else if (pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pstate_method_vblank ||
+ pstate_strategy->per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_vblank_drr) {
vblank_count++;
set_bit_in_bitfield(&vblank_stream_mask, stream_index);
}
@@ -1532,7 +1633,7 @@ static bool validate_pstate_support_strategy_cofunctionality(struct dml2_pmo_ins
static int get_vactive_pstate_margin(const struct display_configuation_with_meta *display_cfg, int plane_mask)
{
- unsigned char i;
+ unsigned int i;
int min_vactive_margin_us = 0xFFFFFFF;
for (i = 0; i < DML2_MAX_PLANES; i++) {
@@ -1625,7 +1726,7 @@ static void build_fams2_meta_per_stream(struct dml2_pmo_instance *pmo,
/* for single stream, guarantee at least an instant of allow */
stream_fams2_meta->method_vactive.max_vactive_det_fill_delay_otg_vlines = (unsigned int)math_floor(
math_max2(0.0,
- timing->v_active - stream_fams2_meta->min_allow_width_otg_vlines - stream_fams2_meta->dram_clk_change_blackout_otg_vlines));
+ timing->v_active - math_max2(1.0, stream_fams2_meta->min_allow_width_otg_vlines) - stream_fams2_meta->dram_clk_change_blackout_otg_vlines));
} else {
/* for multi stream, bound to a max fill time defined by IP caps */
stream_fams2_meta->method_vactive.max_vactive_det_fill_delay_otg_vlines =
@@ -1738,8 +1839,10 @@ bool pmo_dcn4_fams2_init_for_pstate_support(struct dml2_pmo_init_for_pstate_supp
struct display_configuation_with_meta *display_config;
const struct dml2_plane_parameters *plane_descriptor;
const struct dml2_pmo_pstate_strategy *strategy_list = NULL;
+ struct dml2_pmo_pstate_strategy override_base_strategy = { 0 };
unsigned int strategy_list_size = 0;
- unsigned char plane_index, stream_index, i;
+ unsigned int plane_index, stream_index, i;
+ bool build_override_strategy = true;
state->performed = true;
in_out->base_display_config->stage3.min_clk_index_for_latency = in_out->base_display_config->stage1.min_clk_index_for_latency;
@@ -1763,7 +1866,11 @@ bool pmo_dcn4_fams2_init_for_pstate_support(struct dml2_pmo_init_for_pstate_supp
set_bit_in_bitfield(&s->pmo_dcn4.stream_plane_mask[plane_descriptor->stream_index], plane_index);
- state->pstate_switch_modes[plane_index] = dml2_uclk_pstate_support_method_vactive;
+ state->pstate_switch_modes[plane_index] = dml2_pstate_method_vactive;
+
+ build_override_strategy &= plane_descriptor->overrides.uclk_pstate_change_strategy != dml2_uclk_pstate_change_strategy_auto;
+ override_base_strategy.per_stream_pstate_method[plane_descriptor->stream_index] =
+ uclk_pstate_strategy_override_to_pstate_method(plane_descriptor->overrides.uclk_pstate_change_strategy);
}
// Figure out which streams can do vactive, and also build up implicit SVP and FAMS2 meta
@@ -1781,13 +1888,30 @@ bool pmo_dcn4_fams2_init_for_pstate_support(struct dml2_pmo_init_for_pstate_supp
/* get synchronized timing groups */
build_synchronized_timing_groups(pmo, display_config);
- strategy_list = get_expanded_strategy_list(&pmo->init_data, display_config->display_config.num_streams);
- if (!strategy_list)
- return false;
-
- strategy_list_size = get_num_expanded_strategies(&pmo->init_data, display_config->display_config.num_streams);
+ if (build_override_strategy) {
+ /* build expanded override strategy list (no permutations) */
+ override_base_strategy.allow_state_increase = true;
+ s->pmo_dcn4.num_expanded_override_strategies = 0;
+ insert_strategy_into_expanded_list(&override_base_strategy,
+ display_config->display_config.num_streams,
+ s->pmo_dcn4.expanded_override_strategy_list,
+ &s->pmo_dcn4.num_expanded_override_strategies);
+ expand_variant_strategy(&override_base_strategy,
+ display_config->display_config.num_streams,
+ false,
+ s->pmo_dcn4.expanded_override_strategy_list,
+ &s->pmo_dcn4.num_expanded_override_strategies);
+
+ /* use override strategy list */
+ strategy_list = s->pmo_dcn4.expanded_override_strategy_list;
+ strategy_list_size = s->pmo_dcn4.num_expanded_override_strategies;
+ } else {
+ /* use predefined strategy list */
+ strategy_list = get_expanded_strategy_list(&pmo->init_data, display_config->display_config.num_streams);
+ strategy_list_size = get_num_expanded_strategies(&pmo->init_data, display_config->display_config.num_streams);
+ }
- if (strategy_list_size == 0)
+ if (!strategy_list || strategy_list_size == 0)
return false;
s->pmo_dcn4.num_pstate_candidates = 0;
@@ -1799,7 +1923,7 @@ bool pmo_dcn4_fams2_init_for_pstate_support(struct dml2_pmo_init_for_pstate_supp
}
if (s->pmo_dcn4.num_pstate_candidates > 0) {
- s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.num_pstate_candidates - 1].allow_state_increase = true;
+ s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.num_pstate_candidates-1].allow_state_increase = true;
s->pmo_dcn4.cur_pstate_candidate = -1;
return true;
} else {
@@ -1832,7 +1956,7 @@ static void reset_display_configuration(struct display_configuation_with_meta *d
// Reset strategy to auto
plane->overrides.uclk_pstate_change_strategy = dml2_uclk_pstate_change_strategy_auto;
- display_config->stage3.pstate_switch_modes[plane_index] = dml2_uclk_pstate_support_method_not_supported;
+ display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_na;
}
}
@@ -1840,7 +1964,7 @@ static void setup_planes_for_drr_by_mask(struct display_configuation_with_meta *
struct dml2_pmo_instance *pmo,
int plane_mask)
{
- unsigned char plane_index;
+ unsigned int plane_index;
struct dml2_plane_parameters *plane;
for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
@@ -1849,7 +1973,7 @@ static void setup_planes_for_drr_by_mask(struct display_configuation_with_meta *
plane->overrides.uclk_pstate_change_strategy = dml2_uclk_pstate_change_strategy_force_drr;
- display_config->stage3.pstate_switch_modes[plane_index] = dml2_uclk_pstate_support_method_fw_drr;
+ display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_fw_drr;
}
}
@@ -1861,13 +1985,13 @@ static void setup_planes_for_svp_by_mask(struct display_configuation_with_meta *
{
struct dml2_pmo_scratch *scratch = &pmo->scratch;
- unsigned char plane_index;
+ unsigned int plane_index;
int stream_index = -1;
for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
if (is_bit_set_in_bitfield(plane_mask, plane_index)) {
stream_index = (char)display_config->display_config.plane_descriptors[plane_index].stream_index;
- display_config->stage3.pstate_switch_modes[plane_index] = dml2_uclk_pstate_support_method_fw_subvp_phantom;
+ display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_fw_svp;
}
}
@@ -1884,13 +2008,13 @@ static void setup_planes_for_svp_drr_by_mask(struct display_configuation_with_me
{
struct dml2_pmo_scratch *scratch = &pmo->scratch;
- unsigned char plane_index;
+ unsigned int plane_index;
int stream_index = -1;
for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
if (is_bit_set_in_bitfield(plane_mask, plane_index)) {
stream_index = (char)display_config->display_config.plane_descriptors[plane_index].stream_index;
- display_config->stage3.pstate_switch_modes[plane_index] = dml2_uclk_pstate_support_method_fw_subvp_phantom_drr;
+ display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_fw_svp_drr;
}
}
@@ -1905,7 +2029,7 @@ static void setup_planes_for_vblank_by_mask(struct display_configuation_with_met
struct dml2_pmo_instance *pmo,
int plane_mask)
{
- unsigned char plane_index;
+ unsigned int plane_index;
struct dml2_plane_parameters *plane;
for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
@@ -1915,7 +2039,7 @@ static void setup_planes_for_vblank_by_mask(struct display_configuation_with_met
plane->overrides.reserved_vblank_time_ns = (long)math_max2(pmo->soc_bb->power_management_parameters.dram_clk_change_blackout_us * 1000.0,
plane->overrides.reserved_vblank_time_ns);
- display_config->stage3.pstate_switch_modes[plane_index] = dml2_uclk_pstate_support_method_vblank;
+ display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_vblank;
}
}
@@ -1925,7 +2049,7 @@ static void setup_planes_for_vblank_drr_by_mask(struct display_configuation_with
struct dml2_pmo_instance *pmo,
int plane_mask)
{
- unsigned char plane_index;
+ unsigned int plane_index;
struct dml2_plane_parameters *plane;
for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
@@ -1933,7 +2057,7 @@ static void setup_planes_for_vblank_drr_by_mask(struct display_configuation_with
plane = &display_config->display_config.plane_descriptors[plane_index];
plane->overrides.reserved_vblank_time_ns = (long)(pmo->soc_bb->power_management_parameters.dram_clk_change_blackout_us * 1000);
- display_config->stage3.pstate_switch_modes[plane_index] = dml2_uclk_pstate_support_method_fw_vblank_drr;
+ display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_fw_vblank_drr;
}
}
}
@@ -1942,14 +2066,14 @@ static void setup_planes_for_vactive_by_mask(struct display_configuation_with_me
struct dml2_pmo_instance *pmo,
int plane_mask)
{
- unsigned char plane_index;
+ unsigned int plane_index;
unsigned int stream_index;
for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
if (is_bit_set_in_bitfield(plane_mask, plane_index)) {
stream_index = display_config->display_config.plane_descriptors[plane_index].stream_index;
- display_config->stage3.pstate_switch_modes[plane_index] = dml2_uclk_pstate_support_method_vactive;
+ display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_vactive;
if (!pmo->options->disable_vactive_det_fill_bw_pad) {
display_config->display_config.plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us =
@@ -1963,14 +2087,14 @@ static void setup_planes_for_vactive_drr_by_mask(struct display_configuation_wit
struct dml2_pmo_instance *pmo,
int plane_mask)
{
- unsigned char plane_index;
+ unsigned int plane_index;
unsigned int stream_index;
for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
if (is_bit_set_in_bitfield(plane_mask, plane_index)) {
stream_index = display_config->display_config.plane_descriptors[plane_index].stream_index;
- display_config->stage3.pstate_switch_modes[plane_index] = dml2_uclk_pstate_support_method_fw_vactive_drr;
+ display_config->stage3.pstate_switch_modes[plane_index] = dml2_pstate_method_fw_vactive_drr;
if (!pmo->options->disable_vactive_det_fill_bw_pad) {
display_config->display_config.plane_descriptors[plane_index].overrides.max_vactive_det_fill_delay_us =
@@ -1992,26 +2116,26 @@ static bool setup_display_config(struct display_configuation_with_meta *display_
for (stream_index = 0; stream_index < display_config->display_config.num_streams; stream_index++) {
- if (pmo->scratch.pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_na) {
+ if (pmo->scratch.pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pstate_method_na) {
success = false;
break;
- } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_vactive) {
+ } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pstate_method_vactive) {
setup_planes_for_vactive_by_mask(display_config, pmo, scratch->pmo_dcn4.stream_plane_mask[stream_index]);
- } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_vblank) {
+ } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pstate_method_vblank) {
setup_planes_for_vblank_by_mask(display_config, pmo, scratch->pmo_dcn4.stream_plane_mask[stream_index]);
- } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_svp) {
+ } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_svp) {
fams2_required = true;
setup_planes_for_svp_by_mask(display_config, pmo, scratch->pmo_dcn4.stream_plane_mask[stream_index]);
- } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_vactive_drr) {
+ } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_vactive_drr) {
fams2_required = true;
setup_planes_for_vactive_drr_by_mask(display_config, pmo, scratch->pmo_dcn4.stream_plane_mask[stream_index]);
- } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_vblank_drr) {
+ } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_vblank_drr) {
fams2_required = true;
setup_planes_for_vblank_drr_by_mask(display_config, pmo, scratch->pmo_dcn4.stream_plane_mask[stream_index]);
- } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_svp_drr) {
+ } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_svp_drr) {
fams2_required = true;
setup_planes_for_svp_drr_by_mask(display_config, pmo, scratch->pmo_dcn4.stream_plane_mask[stream_index]);
- } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_drr) {
+ } else if (scratch->pmo_dcn4.pstate_strategy_candidates[strategy_index].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_drr) {
fams2_required = true;
setup_planes_for_drr_by_mask(display_config, pmo, scratch->pmo_dcn4.stream_plane_mask[stream_index]);
}
@@ -2031,7 +2155,7 @@ static bool setup_display_config(struct display_configuation_with_meta *display_
static int get_minimum_reserved_time_us_for_planes(struct display_configuation_with_meta *display_config, int plane_mask)
{
int min_time_us = 0xFFFFFF;
- unsigned char plane_index = 0;
+ unsigned int plane_index = 0;
for (plane_index = 0; plane_index < display_config->display_config.num_planes; plane_index++) {
if (is_bit_set_in_bitfield(plane_mask, plane_index)) {
@@ -2066,34 +2190,34 @@ bool pmo_dcn4_fams2_test_for_pstate_support(struct dml2_pmo_test_for_pstate_supp
for (stream_index = 0; stream_index < in_out->base_display_config->display_config.num_streams; stream_index++) {
struct dml2_fams2_meta *stream_fams2_meta = &s->pmo_dcn4.stream_fams2_meta[stream_index];
- if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_vactive ||
- s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_vactive_drr) {
+ if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pstate_method_vactive ||
+ s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_vactive_drr) {
if (get_vactive_pstate_margin(in_out->base_display_config, s->pmo_dcn4.stream_plane_mask[stream_index]) < (MIN_VACTIVE_MARGIN_PCT * in_out->instance->soc_bb->power_management_parameters.dram_clk_change_blackout_us) ||
get_vactive_det_fill_latency_delay_us(in_out->base_display_config, s->pmo_dcn4.stream_plane_mask[stream_index]) > stream_fams2_meta->method_vactive.max_vactive_det_fill_delay_us) {
p_state_supported = false;
break;
}
- } else if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_vblank ||
- s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_vblank_drr) {
+ } else if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pstate_method_vblank ||
+ s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_vblank_drr) {
if (get_minimum_reserved_time_us_for_planes(in_out->base_display_config, s->pmo_dcn4.stream_plane_mask[stream_index]) <
REQUIRED_RESERVED_TIME ||
get_vactive_pstate_margin(in_out->base_display_config, s->pmo_dcn4.stream_plane_mask[stream_index]) < MIN_VACTIVE_MARGIN_VBLANK) {
p_state_supported = false;
break;
}
- } else if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_svp ||
- s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_svp_drr) {
+ } else if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_svp ||
+ s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_svp_drr) {
if (in_out->base_display_config->stage3.stream_svp_meta[stream_index].valid == false) {
p_state_supported = false;
break;
}
- } else if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_fw_drr) {
- if (!all_planes_match_method(in_out->base_display_config, s->pmo_dcn4.stream_plane_mask[stream_index], dml2_pmo_pstate_strategy_fw_drr) ||
+ } else if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pstate_method_fw_drr) {
+ if (!all_planes_match_method(in_out->base_display_config, s->pmo_dcn4.stream_plane_mask[stream_index], dml2_pstate_method_fw_drr) ||
get_vactive_pstate_margin(in_out->base_display_config, s->pmo_dcn4.stream_plane_mask[stream_index]) < MIN_VACTIVE_MARGIN_DRR) {
p_state_supported = false;
break;
}
- } else if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pmo_pstate_strategy_na) {
+ } else if (s->pmo_dcn4.pstate_strategy_candidates[s->pmo_dcn4.cur_pstate_candidate].per_stream_pstate_method[stream_index] == dml2_pstate_method_na) {
p_state_supported = false;
break;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.h
index 0c25bd3e9ac0..6baab7ad6ecc 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_dcn4_fams2.h
@@ -23,4 +23,11 @@ bool pmo_dcn4_fams2_init_for_stutter(struct dml2_pmo_init_for_stutter_in_out *in
bool pmo_dcn4_fams2_test_for_stutter(struct dml2_pmo_test_for_stutter_in_out *in_out);
bool pmo_dcn4_fams2_optimize_for_stutter(struct dml2_pmo_optimize_for_stutter_in_out *in_out);
+void pmo_dcn4_fams2_expand_base_pstate_strategies(
+ const struct dml2_pmo_pstate_strategy *base_strategies_list,
+ const unsigned int num_base_strategies,
+ const unsigned int stream_count,
+ struct dml2_pmo_pstate_strategy *expanded_strategy_list,
+ unsigned int *num_expanded_strategies);
+
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_factory.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_factory.c
index add51d41a515..7ed0242a4b33 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_factory.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_pmo/dml2_pmo_factory.c
@@ -72,7 +72,6 @@ bool dml2_pmo_create(enum dml2_project_id project_id, struct dml2_pmo_instance *
out->init_for_stutter = pmo_dcn4_fams2_init_for_stutter;
out->test_for_stutter = pmo_dcn4_fams2_test_for_stutter;
out->optimize_for_stutter = pmo_dcn4_fams2_optimize_for_stutter;
-
result = true;
break;
case dml2_project_invalid:
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_interfaces.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_interfaces.c
new file mode 100644
index 000000000000..5a33e2f357f4
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_interfaces.c
@@ -0,0 +1,49 @@
+// SPDX-License-Identifier: MIT
+//
+// Copyright 2024 Advanced Micro Devices, Inc.
+
+#include "dml_top.h"
+#include "dml2_internal_shared_types.h"
+#include "dml2_top_soc15.h"
+
+unsigned int dml2_get_instance_size_bytes(void)
+{
+ return sizeof(struct dml2_instance);
+}
+
+bool dml2_initialize_instance(struct dml2_initialize_instance_in_out *in_out)
+{
+ switch (in_out->options.project_id) {
+ case dml2_project_dcn4x_stage1:
+ case dml2_project_dcn4x_stage2:
+ case dml2_project_dcn4x_stage2_auto_drr_svp:
+ return dml2_top_soc15_initialize_instance(in_out);
+ case dml2_project_invalid:
+ default:
+ return false;
+ }
+}
+
+bool dml2_check_mode_supported(struct dml2_check_mode_supported_in_out *in_out)
+{
+ if (!in_out->dml2_instance->funcs.check_mode_supported)
+ return false;
+
+ return in_out->dml2_instance->funcs.check_mode_supported(in_out);
+}
+
+bool dml2_build_mode_programming(struct dml2_build_mode_programming_in_out *in_out)
+{
+ if (!in_out->dml2_instance->funcs.build_mode_programming)
+ return false;
+
+ return in_out->dml2_instance->funcs.build_mode_programming(in_out);
+}
+
+bool dml2_build_mcache_programming(struct dml2_build_mcache_programming_in_out *in_out)
+{
+ if (!in_out->dml2_instance->funcs.build_mcache_programming)
+ return false;
+
+ return in_out->dml2_instance->funcs.build_mcache_programming(in_out);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_legacy.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_legacy.c
new file mode 100644
index 000000000000..5e14d85821e2
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_legacy.c
@@ -0,0 +1,10 @@
+// SPDX-License-Identifier: MIT
+//
+// Copyright 2024 Advanced Micro Devices, Inc.
+
+#include "dml2_top_legacy.h"
+#include "dml2_top_soc15.h"
+#include "dml2_core_factory.h"
+#include "dml2_pmo_factory.h"
+#include "display_mode_core_structs.h"
+
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_legacy.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_legacy.h
new file mode 100644
index 000000000000..14d0ae03dce6
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_legacy.h
@@ -0,0 +1,9 @@
+// SPDX-License-Identifier: MIT
+//
+// Copyright 2024 Advanced Micro Devices, Inc.
+
+#ifndef __DML2_TOP_LEGACY_H__
+#define __DML2_TOP_LEGACY_H__
+#include "dml2_internal_shared_types.h"
+bool dml2_top_legacy_initialize_instance(struct dml2_initialize_instance_in_out *in_out);
+#endif /* __DML2_TOP_LEGACY_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_optimization.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_optimization.c
deleted file mode 100644
index d0e026d981b5..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_optimization.c
+++ /dev/null
@@ -1,307 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-#include "dml2_top_optimization.h"
-#include "dml2_internal_shared_types.h"
-#include "dml_top_mcache.h"
-
-static void copy_display_configuration_with_meta(struct display_configuation_with_meta *dst, const struct display_configuation_with_meta *src)
-{
- memcpy(dst, src, sizeof(struct display_configuation_with_meta));
-}
-
-bool dml2_top_optimization_init_function_min_clk_for_latency(const struct optimization_init_function_params *params)
-{
- struct dml2_optimization_stage1_state *state = &params->display_config->stage1;
-
- state->performed = true;
-
- return true;
-}
-
-bool dml2_top_optimization_test_function_min_clk_for_latency(const struct optimization_test_function_params *params)
-{
- struct dml2_optimization_stage1_state *state = &params->display_config->stage1;
-
- return state->min_clk_index_for_latency == 0;
-}
-
-bool dml2_top_optimization_optimize_function_min_clk_for_latency(const struct optimization_optimize_function_params *params)
-{
- bool result = false;
-
- if (params->display_config->stage1.min_clk_index_for_latency > 0) {
- copy_display_configuration_with_meta(params->optimized_display_config, params->display_config);
- params->optimized_display_config->stage1.min_clk_index_for_latency--;
- result = true;
- }
-
- return result;
-}
-
-bool dml2_top_optimization_test_function_mcache(const struct optimization_test_function_params *params)
-{
- struct dml2_optimization_test_function_locals *l = params->locals;
- bool mcache_success = false;
- bool result = false;
-
- memset(l, 0, sizeof(struct dml2_optimization_test_function_locals));
-
- l->test_mcache.calc_mcache_count_params.dml2_instance = params->dml;
- l->test_mcache.calc_mcache_count_params.display_config = &params->display_config->display_config;
- l->test_mcache.calc_mcache_count_params.mcache_allocations = params->display_config->stage2.mcache_allocations;
-
- result = dml2_top_mcache_calc_mcache_count_and_offsets(&l->test_mcache.calc_mcache_count_params); // use core to get the basic mcache_allocations
-
- if (result) {
- l->test_mcache.assign_global_mcache_ids_params.allocations = params->display_config->stage2.mcache_allocations;
- l->test_mcache.assign_global_mcache_ids_params.num_allocations = params->display_config->display_config.num_planes;
-
- dml2_top_mcache_assign_global_mcache_ids(&l->test_mcache.assign_global_mcache_ids_params);
-
- l->test_mcache.validate_admissibility_params.dml2_instance = params->dml;
- l->test_mcache.validate_admissibility_params.display_cfg = &params->display_config->display_config;
- l->test_mcache.validate_admissibility_params.mcache_allocations = params->display_config->stage2.mcache_allocations;
- l->test_mcache.validate_admissibility_params.cfg_support_info = &params->display_config->mode_support_result.cfg_support_info;
-
- mcache_success = dml2_top_mcache_validate_admissability(&l->test_mcache.validate_admissibility_params); // also find the shift to make mcache allocation works
-
- memcpy(params->display_config->stage2.per_plane_mcache_support, l->test_mcache.validate_admissibility_params.per_plane_status, sizeof(bool) * DML2_MAX_PLANES);
- }
-
- return mcache_success;
-}
-
-bool dml2_top_optimization_optimize_function_mcache(const struct optimization_optimize_function_params *params)
-{
- struct dml2_optimization_optimize_function_locals *l = params->locals;
- bool optimize_success = false;
-
- if (params->last_candidate_supported == false)
- return false;
-
- copy_display_configuration_with_meta(params->optimized_display_config, params->display_config);
-
- l->optimize_mcache.optimize_mcache_params.instance = &params->dml->pmo_instance;
- l->optimize_mcache.optimize_mcache_params.dcc_mcache_supported = params->display_config->stage2.per_plane_mcache_support;
- l->optimize_mcache.optimize_mcache_params.display_config = &params->display_config->display_config;
- l->optimize_mcache.optimize_mcache_params.optimized_display_cfg = &params->optimized_display_config->display_config;
- l->optimize_mcache.optimize_mcache_params.cfg_support_info = &params->optimized_display_config->mode_support_result.cfg_support_info;
-
- optimize_success = params->dml->pmo_instance.optimize_dcc_mcache(&l->optimize_mcache.optimize_mcache_params);
-
- return optimize_success;
-}
-
-bool dml2_top_optimization_init_function_vmin(const struct optimization_init_function_params *params)
-{
- struct dml2_optimization_init_function_locals *l = params->locals;
-
- l->vmin.init_params.instance = &params->dml->pmo_instance;
- l->vmin.init_params.base_display_config = params->display_config;
- return params->dml->pmo_instance.init_for_vmin(&l->vmin.init_params);
-}
-
-bool dml2_top_optimization_test_function_vmin(const struct optimization_test_function_params *params)
-{
- struct dml2_optimization_test_function_locals *l = params->locals;
-
- l->test_vmin.pmo_test_vmin_params.instance = &params->dml->pmo_instance;
- l->test_vmin.pmo_test_vmin_params.display_config = params->display_config;
- l->test_vmin.pmo_test_vmin_params.vmin_limits = &params->dml->soc_bbox.vmin_limit;
- return params->dml->pmo_instance.test_for_vmin(&l->test_vmin.pmo_test_vmin_params);
-}
-
-bool dml2_top_optimization_optimize_function_vmin(const struct optimization_optimize_function_params *params)
-{
- struct dml2_optimization_optimize_function_locals *l = params->locals;
-
- if (params->last_candidate_supported == false)
- return false;
-
- l->optimize_vmin.pmo_optimize_vmin_params.instance = &params->dml->pmo_instance;
- l->optimize_vmin.pmo_optimize_vmin_params.base_display_config = params->display_config;
- l->optimize_vmin.pmo_optimize_vmin_params.optimized_display_config = params->optimized_display_config;
- return params->dml->pmo_instance.optimize_for_vmin(&l->optimize_vmin.pmo_optimize_vmin_params);
-}
-
-bool dml2_top_optimization_perform_optimization_phase(struct dml2_optimization_phase_locals *l, const struct optimization_phase_params *params)
-{
- bool test_passed = false;
- bool optimize_succeeded = true;
- bool candidate_validation_passed = true;
- struct optimization_init_function_params init_params = { 0 };
- struct optimization_test_function_params test_params = { 0 };
- struct optimization_optimize_function_params optimize_params = { 0 };
-
- if (!params->dml ||
- !params->optimize_function ||
- !params->test_function ||
- !params->display_config ||
- !params->optimized_display_config)
- return false;
-
- copy_display_configuration_with_meta(&l->cur_candidate_display_cfg, params->display_config);
-
- init_params.locals = &l->init_function_locals;
- init_params.dml = params->dml;
- init_params.display_config = &l->cur_candidate_display_cfg;
-
- if (params->init_function && !params->init_function(&init_params))
- return false;
-
- test_params.locals = &l->test_function_locals;
- test_params.dml = params->dml;
- test_params.display_config = &l->cur_candidate_display_cfg;
-
- test_passed = params->test_function(&test_params);
-
- while (!test_passed && optimize_succeeded) {
- memset(&optimize_params, 0, sizeof(struct optimization_optimize_function_params));
-
- optimize_params.locals = &l->optimize_function_locals;
- optimize_params.dml = params->dml;
- optimize_params.display_config = &l->cur_candidate_display_cfg;
- optimize_params.optimized_display_config = &l->next_candidate_display_cfg;
- optimize_params.last_candidate_supported = candidate_validation_passed;
-
- optimize_succeeded = params->optimize_function(&optimize_params);
-
- if (optimize_succeeded) {
- l->mode_support_params.instance = &params->dml->core_instance;
- l->mode_support_params.display_cfg = &l->next_candidate_display_cfg;
- l->mode_support_params.min_clk_table = &params->dml->min_clk_table;
-
- if (l->next_candidate_display_cfg.stage3.performed)
- l->mode_support_params.min_clk_index = l->next_candidate_display_cfg.stage3.min_clk_index_for_latency;
- else
- l->mode_support_params.min_clk_index = l->next_candidate_display_cfg.stage1.min_clk_index_for_latency;
-
- candidate_validation_passed = params->dml->core_instance.mode_support(&l->mode_support_params);
-
- l->next_candidate_display_cfg.mode_support_result = l->mode_support_params.mode_support_result;
- }
-
- if (optimize_succeeded && candidate_validation_passed) {
- memset(&test_params, 0, sizeof(struct optimization_test_function_params));
- test_params.locals = &l->test_function_locals;
- test_params.dml = params->dml;
- test_params.display_config = &l->next_candidate_display_cfg;
- test_passed = params->test_function(&test_params);
-
- copy_display_configuration_with_meta(&l->cur_candidate_display_cfg, &l->next_candidate_display_cfg);
-
- // If optimization is not all or nothing, then store partial progress in output
- if (!params->all_or_nothing)
- copy_display_configuration_with_meta(params->optimized_display_config, &l->next_candidate_display_cfg);
- }
- }
-
- if (test_passed)
- copy_display_configuration_with_meta(params->optimized_display_config, &l->cur_candidate_display_cfg);
-
- return test_passed;
-}
-
-bool dml2_top_optimization_perform_optimization_phase_1(struct dml2_optimization_phase_locals *l, const struct optimization_phase_params *params)
-{
- int highest_state, lowest_state, cur_state;
- bool supported = false;
-
- if (!params->dml ||
- !params->optimize_function ||
- !params->test_function ||
- !params->display_config ||
- !params->optimized_display_config)
- return false;
-
- copy_display_configuration_with_meta(&l->cur_candidate_display_cfg, params->display_config);
- highest_state = l->cur_candidate_display_cfg.stage1.min_clk_index_for_latency;
- lowest_state = 0;
-
- while (highest_state > lowest_state) {
- cur_state = (highest_state + lowest_state) / 2;
-
- l->mode_support_params.instance = &params->dml->core_instance;
- l->mode_support_params.display_cfg = &l->cur_candidate_display_cfg;
- l->mode_support_params.min_clk_table = &params->dml->min_clk_table;
- l->mode_support_params.min_clk_index = cur_state;
-
- supported = params->dml->core_instance.mode_support(&l->mode_support_params);
-
- if (supported) {
- l->cur_candidate_display_cfg.mode_support_result = l->mode_support_params.mode_support_result;
- highest_state = cur_state;
- } else {
- lowest_state = cur_state + 1;
- }
- }
- l->cur_candidate_display_cfg.stage1.min_clk_index_for_latency = lowest_state;
-
- copy_display_configuration_with_meta(params->optimized_display_config, &l->cur_candidate_display_cfg);
-
- return true;
-}
-
-bool dml2_top_optimization_init_function_uclk_pstate(const struct optimization_init_function_params *params)
-{
- struct dml2_optimization_init_function_locals *l = params->locals;
-
- l->uclk_pstate.init_params.instance = &params->dml->pmo_instance;
- l->uclk_pstate.init_params.base_display_config = params->display_config;
-
- return params->dml->pmo_instance.init_for_uclk_pstate(&l->uclk_pstate.init_params);
-}
-
-bool dml2_top_optimization_test_function_uclk_pstate(const struct optimization_test_function_params *params)
-{
- struct dml2_optimization_test_function_locals *l = params->locals;
-
- l->uclk_pstate.test_params.instance = &params->dml->pmo_instance;
- l->uclk_pstate.test_params.base_display_config = params->display_config;
-
- return params->dml->pmo_instance.test_for_uclk_pstate(&l->uclk_pstate.test_params);
-}
-
-bool dml2_top_optimization_optimize_function_uclk_pstate(const struct optimization_optimize_function_params *params)
-{
- struct dml2_optimization_optimize_function_locals *l = params->locals;
-
- l->uclk_pstate.optimize_params.instance = &params->dml->pmo_instance;
- l->uclk_pstate.optimize_params.base_display_config = params->display_config;
- l->uclk_pstate.optimize_params.optimized_display_config = params->optimized_display_config;
- l->uclk_pstate.optimize_params.last_candidate_failed = !params->last_candidate_supported;
-
- return params->dml->pmo_instance.optimize_for_uclk_pstate(&l->uclk_pstate.optimize_params);
-}
-
-bool dml2_top_optimization_init_function_stutter(const struct optimization_init_function_params *params)
-{
- struct dml2_optimization_init_function_locals *l = params->locals;
-
- l->uclk_pstate.init_params.instance = &params->dml->pmo_instance;
- l->uclk_pstate.init_params.base_display_config = params->display_config;
-
- return params->dml->pmo_instance.init_for_stutter(&l->stutter.stutter_params);
-}
-
-bool dml2_top_optimization_test_function_stutter(const struct optimization_test_function_params *params)
-{
- struct dml2_optimization_test_function_locals *l = params->locals;
-
- l->stutter.stutter_params.instance = &params->dml->pmo_instance;
- l->stutter.stutter_params.base_display_config = params->display_config;
- return params->dml->pmo_instance.test_for_stutter(&l->stutter.stutter_params);
-}
-
-bool dml2_top_optimization_optimize_function_stutter(const struct optimization_optimize_function_params *params)
-{
- struct dml2_optimization_optimize_function_locals *l = params->locals;
-
- l->stutter.stutter_params.instance = &params->dml->pmo_instance;
- l->stutter.stutter_params.base_display_config = params->display_config;
- l->stutter.stutter_params.optimized_display_config = params->optimized_display_config;
- l->stutter.stutter_params.last_candidate_failed = !params->last_candidate_supported;
- return params->dml->pmo_instance.optimize_for_stutter(&l->stutter.stutter_params);
-}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_optimization.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_optimization.h
deleted file mode 100644
index 9f22ab33eab1..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_optimization.h
+++ /dev/null
@@ -1,33 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-#ifndef __DML2_TOP_OPTIMIZATION_H__
-#define __DML2_TOP_OPTIMIZATION_H__
-
-#include "dml2_external_lib_deps.h"
-#include "dml2_internal_shared_types.h"
-
-bool dml2_top_optimization_perform_optimization_phase(struct dml2_optimization_phase_locals *l, const struct optimization_phase_params *params);
-bool dml2_top_optimization_perform_optimization_phase_1(struct dml2_optimization_phase_locals *l, const struct optimization_phase_params *params);
-
-bool dml2_top_optimization_init_function_min_clk_for_latency(const struct optimization_init_function_params *params);
-bool dml2_top_optimization_test_function_min_clk_for_latency(const struct optimization_test_function_params *params);
-bool dml2_top_optimization_optimize_function_min_clk_for_latency(const struct optimization_optimize_function_params *params);
-
-bool dml2_top_optimization_test_function_mcache(const struct optimization_test_function_params *params);
-bool dml2_top_optimization_optimize_function_mcache(const struct optimization_optimize_function_params *params);
-
-bool dml2_top_optimization_init_function_uclk_pstate(const struct optimization_init_function_params *params);
-bool dml2_top_optimization_test_function_uclk_pstate(const struct optimization_test_function_params *params);
-bool dml2_top_optimization_optimize_function_uclk_pstate(const struct optimization_optimize_function_params *params);
-
-bool dml2_top_optimization_init_function_vmin(const struct optimization_init_function_params *params);
-bool dml2_top_optimization_test_function_vmin(const struct optimization_test_function_params *params);
-bool dml2_top_optimization_optimize_function_vmin(const struct optimization_optimize_function_params *params);
-
-bool dml2_top_optimization_init_function_stutter(const struct optimization_init_function_params *params);
-bool dml2_top_optimization_test_function_stutter(const struct optimization_test_function_params *params);
-bool dml2_top_optimization_optimize_function_stutter(const struct optimization_optimize_function_params *params);
-
-#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_soc15.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_soc15.c
new file mode 100644
index 000000000000..4a7c4c62111e
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_soc15.c
@@ -0,0 +1,1170 @@
+// SPDX-License-Identifier: MIT
+//
+// Copyright 2024 Advanced Micro Devices, Inc.
+
+#include "dml2_top_soc15.h"
+#include "dml2_mcg_factory.h"
+#include "dml2_dpmm_factory.h"
+#include "dml2_core_factory.h"
+#include "dml2_pmo_factory.h"
+#include "lib_float_math.h"
+#include "dml2_debug.h"
+static void setup_unoptimized_display_config_with_meta(const struct dml2_instance *dml, struct display_configuation_with_meta *out, const struct dml2_display_cfg *display_config)
+{
+ memcpy(&out->display_config, display_config, sizeof(struct dml2_display_cfg));
+ out->stage1.min_clk_index_for_latency = dml->min_clk_table.dram_bw_table.num_entries - 1; //dml->min_clk_table.clean_me_up.soc_bb.num_states - 1;
+}
+
+static void setup_speculative_display_config_with_meta(const struct dml2_instance *dml, struct display_configuation_with_meta *out, const struct dml2_display_cfg *display_config)
+{
+ memcpy(&out->display_config, display_config, sizeof(struct dml2_display_cfg));
+ out->stage1.min_clk_index_for_latency = 0;
+}
+
+static void copy_display_configuration_with_meta(struct display_configuation_with_meta *dst, const struct display_configuation_with_meta *src)
+{
+ memcpy(dst, src, sizeof(struct display_configuation_with_meta));
+}
+
+static bool dml2_top_optimization_init_function_min_clk_for_latency(const struct optimization_init_function_params *params)
+{
+ struct dml2_optimization_stage1_state *state = &params->display_config->stage1;
+
+ state->performed = true;
+
+ return true;
+}
+
+static bool dml2_top_optimization_test_function_min_clk_for_latency(const struct optimization_test_function_params *params)
+{
+ struct dml2_optimization_stage1_state *state = &params->display_config->stage1;
+
+ return state->min_clk_index_for_latency == 0;
+}
+
+static bool dml2_top_optimization_optimize_function_min_clk_for_latency(const struct optimization_optimize_function_params *params)
+{
+ bool result = false;
+
+ if (params->display_config->stage1.min_clk_index_for_latency > 0) {
+ copy_display_configuration_with_meta(params->optimized_display_config, params->display_config);
+ params->optimized_display_config->stage1.min_clk_index_for_latency--;
+ result = true;
+ }
+
+ return result;
+}
+
+static bool dml2_top_optimization_test_function_mcache(const struct optimization_test_function_params *params)
+{
+ struct dml2_optimization_test_function_locals *l = params->locals;
+ bool mcache_success = false;
+ bool result = false;
+
+ memset(l, 0, sizeof(struct dml2_optimization_test_function_locals));
+
+ l->test_mcache.calc_mcache_count_params.dml2_instance = params->dml;
+ l->test_mcache.calc_mcache_count_params.display_config = &params->display_config->display_config;
+ l->test_mcache.calc_mcache_count_params.mcache_allocations = params->display_config->stage2.mcache_allocations;
+
+ result = dml2_top_mcache_calc_mcache_count_and_offsets(&l->test_mcache.calc_mcache_count_params); // use core to get the basic mcache_allocations
+
+ if (result) {
+ l->test_mcache.assign_global_mcache_ids_params.allocations = params->display_config->stage2.mcache_allocations;
+ l->test_mcache.assign_global_mcache_ids_params.num_allocations = params->display_config->display_config.num_planes;
+
+ dml2_top_mcache_assign_global_mcache_ids(&l->test_mcache.assign_global_mcache_ids_params);
+
+ l->test_mcache.validate_admissibility_params.dml2_instance = params->dml;
+ l->test_mcache.validate_admissibility_params.display_cfg = &params->display_config->display_config;
+ l->test_mcache.validate_admissibility_params.mcache_allocations = params->display_config->stage2.mcache_allocations;
+ l->test_mcache.validate_admissibility_params.cfg_support_info = &params->display_config->mode_support_result.cfg_support_info;
+
+ mcache_success = dml2_top_mcache_validate_admissability(&l->test_mcache.validate_admissibility_params); // also find the shift to make mcache allocation works
+
+ memcpy(params->display_config->stage2.per_plane_mcache_support, l->test_mcache.validate_admissibility_params.per_plane_status, sizeof(bool) * DML2_MAX_PLANES);
+ }
+
+ return mcache_success;
+}
+
+static bool dml2_top_optimization_optimize_function_mcache(const struct optimization_optimize_function_params *params)
+{
+ struct dml2_optimization_optimize_function_locals *l = params->locals;
+ bool optimize_success = false;
+
+ if (params->last_candidate_supported == false)
+ return false;
+
+ copy_display_configuration_with_meta(params->optimized_display_config, params->display_config);
+
+ l->optimize_mcache.optimize_mcache_params.instance = &params->dml->pmo_instance;
+ l->optimize_mcache.optimize_mcache_params.dcc_mcache_supported = params->display_config->stage2.per_plane_mcache_support;
+ l->optimize_mcache.optimize_mcache_params.display_config = &params->display_config->display_config;
+ l->optimize_mcache.optimize_mcache_params.optimized_display_cfg = &params->optimized_display_config->display_config;
+ l->optimize_mcache.optimize_mcache_params.cfg_support_info = &params->optimized_display_config->mode_support_result.cfg_support_info;
+
+ optimize_success = params->dml->pmo_instance.optimize_dcc_mcache(&l->optimize_mcache.optimize_mcache_params);
+
+ return optimize_success;
+}
+
+static bool dml2_top_optimization_init_function_vmin(const struct optimization_init_function_params *params)
+{
+ struct dml2_optimization_init_function_locals *l = params->locals;
+
+ l->vmin.init_params.instance = &params->dml->pmo_instance;
+ l->vmin.init_params.base_display_config = params->display_config;
+ return params->dml->pmo_instance.init_for_vmin(&l->vmin.init_params);
+}
+
+static bool dml2_top_optimization_test_function_vmin(const struct optimization_test_function_params *params)
+{
+ struct dml2_optimization_test_function_locals *l = params->locals;
+
+ l->test_vmin.pmo_test_vmin_params.instance = &params->dml->pmo_instance;
+ l->test_vmin.pmo_test_vmin_params.display_config = params->display_config;
+ l->test_vmin.pmo_test_vmin_params.vmin_limits = &params->dml->soc_bbox.vmin_limit;
+ return params->dml->pmo_instance.test_for_vmin(&l->test_vmin.pmo_test_vmin_params);
+}
+
+static bool dml2_top_optimization_optimize_function_vmin(const struct optimization_optimize_function_params *params)
+{
+ struct dml2_optimization_optimize_function_locals *l = params->locals;
+
+ if (params->last_candidate_supported == false)
+ return false;
+
+ l->optimize_vmin.pmo_optimize_vmin_params.instance = &params->dml->pmo_instance;
+ l->optimize_vmin.pmo_optimize_vmin_params.base_display_config = params->display_config;
+ l->optimize_vmin.pmo_optimize_vmin_params.optimized_display_config = params->optimized_display_config;
+ return params->dml->pmo_instance.optimize_for_vmin(&l->optimize_vmin.pmo_optimize_vmin_params);
+}
+
+static bool dml2_top_optimization_init_function_uclk_pstate(const struct optimization_init_function_params *params)
+{
+ struct dml2_optimization_init_function_locals *l = params->locals;
+
+ l->uclk_pstate.init_params.instance = &params->dml->pmo_instance;
+ l->uclk_pstate.init_params.base_display_config = params->display_config;
+
+ return params->dml->pmo_instance.init_for_uclk_pstate(&l->uclk_pstate.init_params);
+}
+
+static bool dml2_top_optimization_test_function_uclk_pstate(const struct optimization_test_function_params *params)
+{
+ struct dml2_optimization_test_function_locals *l = params->locals;
+
+ l->uclk_pstate.test_params.instance = &params->dml->pmo_instance;
+ l->uclk_pstate.test_params.base_display_config = params->display_config;
+
+ return params->dml->pmo_instance.test_for_uclk_pstate(&l->uclk_pstate.test_params);
+}
+
+static bool dml2_top_optimization_optimize_function_uclk_pstate(const struct optimization_optimize_function_params *params)
+{
+ struct dml2_optimization_optimize_function_locals *l = params->locals;
+
+ l->uclk_pstate.optimize_params.instance = &params->dml->pmo_instance;
+ l->uclk_pstate.optimize_params.base_display_config = params->display_config;
+ l->uclk_pstate.optimize_params.optimized_display_config = params->optimized_display_config;
+ l->uclk_pstate.optimize_params.last_candidate_failed = !params->last_candidate_supported;
+
+ return params->dml->pmo_instance.optimize_for_uclk_pstate(&l->uclk_pstate.optimize_params);
+}
+
+static bool dml2_top_optimization_init_function_stutter(const struct optimization_init_function_params *params)
+{
+ struct dml2_optimization_init_function_locals *l = params->locals;
+
+ l->uclk_pstate.init_params.instance = &params->dml->pmo_instance;
+ l->uclk_pstate.init_params.base_display_config = params->display_config;
+
+ return params->dml->pmo_instance.init_for_stutter(&l->stutter.stutter_params);
+}
+
+static bool dml2_top_optimization_test_function_stutter(const struct optimization_test_function_params *params)
+{
+ struct dml2_optimization_test_function_locals *l = params->locals;
+
+ l->stutter.stutter_params.instance = &params->dml->pmo_instance;
+ l->stutter.stutter_params.base_display_config = params->display_config;
+ return params->dml->pmo_instance.test_for_stutter(&l->stutter.stutter_params);
+}
+
+static bool dml2_top_optimization_optimize_function_stutter(const struct optimization_optimize_function_params *params)
+{
+ struct dml2_optimization_optimize_function_locals *l = params->locals;
+
+ l->stutter.stutter_params.instance = &params->dml->pmo_instance;
+ l->stutter.stutter_params.base_display_config = params->display_config;
+ l->stutter.stutter_params.optimized_display_config = params->optimized_display_config;
+ l->stutter.stutter_params.last_candidate_failed = !params->last_candidate_supported;
+ return params->dml->pmo_instance.optimize_for_stutter(&l->stutter.stutter_params);
+}
+
+static bool dml2_top_optimization_perform_optimization_phase(struct dml2_optimization_phase_locals *l, const struct optimization_phase_params *params)
+{
+ bool test_passed = false;
+ bool optimize_succeeded = true;
+ bool candidate_validation_passed = true;
+ struct optimization_init_function_params init_params = { 0 };
+ struct optimization_test_function_params test_params = { 0 };
+ struct optimization_optimize_function_params optimize_params = { 0 };
+
+ if (!params->dml ||
+ !params->optimize_function ||
+ !params->test_function ||
+ !params->display_config ||
+ !params->optimized_display_config)
+ return false;
+
+ copy_display_configuration_with_meta(&l->cur_candidate_display_cfg, params->display_config);
+
+ init_params.locals = &l->init_function_locals;
+ init_params.dml = params->dml;
+ init_params.display_config = &l->cur_candidate_display_cfg;
+
+ if (params->init_function && !params->init_function(&init_params))
+ return false;
+
+ test_params.locals = &l->test_function_locals;
+ test_params.dml = params->dml;
+ test_params.display_config = &l->cur_candidate_display_cfg;
+
+ test_passed = params->test_function(&test_params);
+
+ while (!test_passed && optimize_succeeded) {
+ memset(&optimize_params, 0, sizeof(struct optimization_optimize_function_params));
+
+ optimize_params.locals = &l->optimize_function_locals;
+ optimize_params.dml = params->dml;
+ optimize_params.display_config = &l->cur_candidate_display_cfg;
+ optimize_params.optimized_display_config = &l->next_candidate_display_cfg;
+ optimize_params.last_candidate_supported = candidate_validation_passed;
+
+ optimize_succeeded = params->optimize_function(&optimize_params);
+
+ if (optimize_succeeded) {
+ l->mode_support_params.instance = &params->dml->core_instance;
+ l->mode_support_params.display_cfg = &l->next_candidate_display_cfg;
+ l->mode_support_params.min_clk_table = &params->dml->min_clk_table;
+
+ if (l->next_candidate_display_cfg.stage3.performed)
+ l->mode_support_params.min_clk_index = l->next_candidate_display_cfg.stage3.min_clk_index_for_latency;
+ else
+ l->mode_support_params.min_clk_index = l->next_candidate_display_cfg.stage1.min_clk_index_for_latency;
+ candidate_validation_passed = params->dml->core_instance.mode_support(&l->mode_support_params);
+ l->next_candidate_display_cfg.mode_support_result = l->mode_support_params.mode_support_result;
+ }
+
+ if (optimize_succeeded && candidate_validation_passed) {
+ memset(&test_params, 0, sizeof(struct optimization_test_function_params));
+ test_params.locals = &l->test_function_locals;
+ test_params.dml = params->dml;
+ test_params.display_config = &l->next_candidate_display_cfg;
+ test_passed = params->test_function(&test_params);
+
+ copy_display_configuration_with_meta(&l->cur_candidate_display_cfg, &l->next_candidate_display_cfg);
+
+ // If optimization is not all or nothing, then store partial progress in output
+ if (!params->all_or_nothing)
+ copy_display_configuration_with_meta(params->optimized_display_config, &l->next_candidate_display_cfg);
+ }
+ }
+
+ if (test_passed)
+ copy_display_configuration_with_meta(params->optimized_display_config, &l->cur_candidate_display_cfg);
+
+ return test_passed;
+}
+
+static bool dml2_top_optimization_perform_optimization_phase_1(struct dml2_optimization_phase_locals *l, const struct optimization_phase_params *params)
+{
+ int highest_state, lowest_state, cur_state;
+ bool supported = false;
+
+ if (!params->dml ||
+ !params->optimize_function ||
+ !params->test_function ||
+ !params->display_config ||
+ !params->optimized_display_config)
+ return false;
+
+ copy_display_configuration_with_meta(&l->cur_candidate_display_cfg, params->display_config);
+ highest_state = l->cur_candidate_display_cfg.stage1.min_clk_index_for_latency;
+ lowest_state = 0;
+
+ while (highest_state > lowest_state) {
+ cur_state = (highest_state + lowest_state) / 2;
+
+ l->mode_support_params.instance = &params->dml->core_instance;
+ l->mode_support_params.display_cfg = &l->cur_candidate_display_cfg;
+ l->mode_support_params.min_clk_table = &params->dml->min_clk_table;
+ l->mode_support_params.min_clk_index = cur_state;
+ supported = params->dml->core_instance.mode_support(&l->mode_support_params);
+
+ if (supported) {
+ l->cur_candidate_display_cfg.mode_support_result = l->mode_support_params.mode_support_result;
+ highest_state = cur_state;
+ } else {
+ lowest_state = cur_state + 1;
+ }
+ }
+ l->cur_candidate_display_cfg.stage1.min_clk_index_for_latency = lowest_state;
+
+ copy_display_configuration_with_meta(params->optimized_display_config, &l->cur_candidate_display_cfg);
+
+ return true;
+}
+
+/*
+* Takes an input set of mcache boundaries and finds the appropriate setting of cache programming.
+* Returns true if a valid set of programming can be made, and false otherwise. "Valid" means
+* that the horizontal viewport does not span more than 2 cache slices.
+*
+* It optionally also can apply a constant shift to all the cache boundaries.
+*/
+static const uint32_t MCACHE_ID_UNASSIGNED = 0xF;
+static const uint32_t SPLIT_LOCATION_UNDEFINED = 0xFFFF;
+
+static bool calculate_first_second_splitting(const int *mcache_boundaries, int num_boundaries, int shift,
+ int pipe_h_vp_start, int pipe_h_vp_end, int *first_offset, int *second_offset)
+{
+ const int MAX_VP = 0xFFFFFF;
+ int left_cache_id;
+ int right_cache_id;
+ int range_start;
+ int range_end;
+ bool success = false;
+
+ if (num_boundaries <= 1) {
+ if (first_offset && second_offset) {
+ *first_offset = 0;
+ *second_offset = -1;
+ }
+ success = true;
+ return success;
+ } else {
+ range_start = 0;
+ for (left_cache_id = 0; left_cache_id < num_boundaries; left_cache_id++) {
+ range_end = mcache_boundaries[left_cache_id] - shift - 1;
+
+ if (range_start <= pipe_h_vp_start && pipe_h_vp_start <= range_end)
+ break;
+
+ range_start = range_end + 1;
+ }
+
+ range_end = MAX_VP;
+ for (right_cache_id = num_boundaries - 1; right_cache_id >= -1; right_cache_id--) {
+ if (right_cache_id >= 0)
+ range_start = mcache_boundaries[right_cache_id] - shift;
+ else
+ range_start = 0;
+
+ if (range_start <= pipe_h_vp_end && pipe_h_vp_end <= range_end) {
+ break;
+ }
+ range_end = range_start - 1;
+ }
+ right_cache_id = (right_cache_id + 1) % num_boundaries;
+
+ if (right_cache_id == left_cache_id) {
+ if (first_offset && second_offset) {
+ *first_offset = left_cache_id;
+ *second_offset = -1;
+ }
+ success = true;
+ } else if (right_cache_id == (left_cache_id + 1) % num_boundaries) {
+ if (first_offset && second_offset) {
+ *first_offset = left_cache_id;
+ *second_offset = right_cache_id;
+ }
+ success = true;
+ }
+ }
+
+ return success;
+}
+
+/*
+* For a given set of pipe start/end x positions, checks to see it can support the input mcache splitting.
+* It also attempts to "optimize" by finding a shift if the default 0 shift does not work.
+*/
+static bool find_shift_for_valid_cache_id_assignment(int *mcache_boundaries, unsigned int num_boundaries,
+ int *pipe_vp_startx, int *pipe_vp_endx, unsigned int pipe_count, int shift_granularity, int *shift)
+{
+ int max_shift = 0xFFFF;
+ unsigned int pipe_index;
+ unsigned int i, slice_width;
+ bool success = false;
+
+ for (i = 0; i < num_boundaries; i++) {
+ if (i == 0)
+ slice_width = mcache_boundaries[i];
+ else
+ slice_width = mcache_boundaries[i] - mcache_boundaries[i - 1];
+
+ if (max_shift > (int)slice_width) {
+ max_shift = slice_width;
+ }
+ }
+
+ for (*shift = 0; *shift <= max_shift; *shift += shift_granularity) {
+ success = true;
+ for (pipe_index = 0; pipe_index < pipe_count; pipe_index++) {
+ if (!calculate_first_second_splitting(mcache_boundaries, num_boundaries, *shift,
+ pipe_vp_startx[pipe_index], pipe_vp_endx[pipe_index], 0, 0)) {
+ success = false;
+ break;
+ }
+ }
+ if (success)
+ break;
+ }
+
+ return success;
+}
+
+/*
+* Counts the number of elements inside input array within the given span length.
+* Formally, what is the size of the largest subset of the array where the largest and smallest element
+* differ no more than the span.
+*/
+static unsigned int count_elements_in_span(int *array, unsigned int array_size, unsigned int span)
+{
+ unsigned int i;
+ unsigned int span_start_value;
+ unsigned int span_start_index;
+ unsigned int greatest_element_count;
+
+ if (array_size == 0)
+ return 1;
+
+ if (span == 0)
+ return array_size > 0 ? 1 : 0;
+
+ span_start_value = 0;
+ span_start_index = 0;
+ greatest_element_count = 0;
+
+ while (span_start_index < array_size) {
+ for (i = span_start_index; i < array_size; i++) {
+ if (array[i] - span_start_value <= span) {
+ if (i - span_start_index + 1 > greatest_element_count) {
+ greatest_element_count = i - span_start_index + 1;
+ }
+ } else
+ break;
+ }
+
+ span_start_index++;
+
+ if (span_start_index < array_size) {
+ span_start_value = array[span_start_index - 1] + 1;
+ }
+ }
+
+ return greatest_element_count;
+}
+
+static bool calculate_h_split_for_scaling_transform(int full_vp_width, int h_active, int num_pipes,
+ enum dml2_scaling_transform scaling_transform, int *pipe_vp_x_start, int *pipe_vp_x_end)
+{
+ int i, slice_width;
+ const char MAX_SCL_VP_OVERLAP = 3;
+ bool success = false;
+
+ switch (scaling_transform) {
+ case dml2_scaling_transform_centered:
+ case dml2_scaling_transform_aspect_ratio:
+ case dml2_scaling_transform_fullscreen:
+ slice_width = full_vp_width / num_pipes;
+ for (i = 0; i < num_pipes; i++) {
+ pipe_vp_x_start[i] = i * slice_width;
+ pipe_vp_x_end[i] = (i + 1) * slice_width - 1;
+
+ if (pipe_vp_x_start[i] < MAX_SCL_VP_OVERLAP)
+ pipe_vp_x_start[i] = 0;
+ else
+ pipe_vp_x_start[i] -= MAX_SCL_VP_OVERLAP;
+
+ if (pipe_vp_x_end[i] > full_vp_width - MAX_SCL_VP_OVERLAP - 1)
+ pipe_vp_x_end[i] = full_vp_width - 1;
+ else
+ pipe_vp_x_end[i] += MAX_SCL_VP_OVERLAP;
+ }
+ break;
+ case dml2_scaling_transform_explicit:
+ default:
+ success = false;
+ break;
+ }
+
+ return success;
+}
+
+bool dml2_top_mcache_validate_admissability(struct top_mcache_validate_admissability_in_out *params)
+{
+ struct dml2_instance *dml = (struct dml2_instance *)params->dml2_instance;
+ struct dml2_top_mcache_validate_admissability_locals *l = &dml->scratch.mcache_validate_admissability_locals;
+
+ const int MAX_PIXEL_OVERLAP = 6;
+ int max_per_pipe_vp_p0 = 0;
+ int max_per_pipe_vp_p1 = 0;
+ int temp, p0shift, p1shift;
+ unsigned int plane_index = 0;
+ unsigned int i;
+ unsigned int odm_combine_factor;
+ unsigned int mpc_combine_factor;
+ unsigned int num_dpps;
+ unsigned int num_boundaries;
+ enum dml2_scaling_transform scaling_transform;
+ const struct dml2_plane_parameters *plane;
+ const struct dml2_stream_parameters *stream;
+
+ bool p0pass = false;
+ bool p1pass = false;
+ bool all_pass = true;
+
+ for (plane_index = 0; plane_index < params->display_cfg->num_planes; plane_index++) {
+ if (!params->display_cfg->plane_descriptors[plane_index].surface.dcc.enable)
+ continue;
+
+ plane = &params->display_cfg->plane_descriptors[plane_index];
+ stream = &params->display_cfg->stream_descriptors[plane->stream_index];
+
+ num_dpps = odm_combine_factor = params->cfg_support_info->stream_support_info[plane->stream_index].odms_used;
+
+ if (odm_combine_factor == 1)
+ num_dpps = mpc_combine_factor = (unsigned int)params->cfg_support_info->plane_support_info[plane_index].dpps_used;
+ else
+ mpc_combine_factor = 1;
+
+ if (odm_combine_factor > 1) {
+ max_per_pipe_vp_p0 = plane->surface.plane0.width;
+ temp = (unsigned int)math_ceil(plane->composition.scaler_info.plane0.h_ratio * stream->timing.h_active / odm_combine_factor);
+
+ if (temp < max_per_pipe_vp_p0)
+ max_per_pipe_vp_p0 = temp;
+
+ max_per_pipe_vp_p1 = plane->surface.plane1.width;
+ temp = (unsigned int)math_ceil(plane->composition.scaler_info.plane1.h_ratio * stream->timing.h_active / odm_combine_factor);
+
+ if (temp < max_per_pipe_vp_p1)
+ max_per_pipe_vp_p1 = temp;
+ } else {
+ max_per_pipe_vp_p0 = plane->surface.plane0.width / mpc_combine_factor;
+ max_per_pipe_vp_p1 = plane->surface.plane1.width / mpc_combine_factor;
+ }
+
+ max_per_pipe_vp_p0 += 2 * MAX_PIXEL_OVERLAP;
+ max_per_pipe_vp_p1 += MAX_PIXEL_OVERLAP;
+
+ p0shift = 0;
+ p1shift = 0;
+
+ // The last element in the unshifted boundary array will always be the first pixel outside the
+ // plane, which means theres no mcache associated with it, so -1
+ num_boundaries = params->mcache_allocations[plane_index].num_mcaches_plane0 == 0 ? 0 : params->mcache_allocations[plane_index].num_mcaches_plane0 - 1;
+ if ((count_elements_in_span(params->mcache_allocations[plane_index].mcache_x_offsets_plane0,
+ num_boundaries, max_per_pipe_vp_p0) <= 1) && (num_boundaries <= num_dpps)) {
+ p0pass = true;
+ }
+ num_boundaries = params->mcache_allocations[plane_index].num_mcaches_plane1 == 0 ? 0 : params->mcache_allocations[plane_index].num_mcaches_plane1 - 1;
+ if ((count_elements_in_span(params->mcache_allocations[plane_index].mcache_x_offsets_plane1,
+ num_boundaries, max_per_pipe_vp_p1) <= 1) && (num_boundaries <= num_dpps)) {
+ p1pass = true;
+ }
+
+ if (!p0pass || !p1pass) {
+ if (odm_combine_factor > 1) {
+ num_dpps = odm_combine_factor;
+ scaling_transform = plane->composition.scaling_transform;
+ } else {
+ num_dpps = mpc_combine_factor;
+ scaling_transform = dml2_scaling_transform_fullscreen;
+ }
+
+ if (!p0pass) {
+ if (plane->composition.viewport.stationary) {
+ calculate_h_split_for_scaling_transform(plane->surface.plane0.width,
+ stream->timing.h_active, num_dpps, scaling_transform,
+ &l->plane0.pipe_vp_startx[plane_index], &l->plane0.pipe_vp_endx[plane_index]);
+ p0pass = find_shift_for_valid_cache_id_assignment(params->mcache_allocations[plane_index].mcache_x_offsets_plane0,
+ params->mcache_allocations[plane_index].num_mcaches_plane0,
+ &l->plane0.pipe_vp_startx[plane_index], &l->plane0.pipe_vp_endx[plane_index], num_dpps,
+ params->mcache_allocations[plane_index].shift_granularity.p0, &p0shift);
+ }
+ }
+ if (!p1pass) {
+ if (plane->composition.viewport.stationary) {
+ calculate_h_split_for_scaling_transform(plane->surface.plane1.width,
+ stream->timing.h_active, num_dpps, scaling_transform,
+ &l->plane0.pipe_vp_startx[plane_index], &l->plane0.pipe_vp_endx[plane_index]);
+ p1pass = find_shift_for_valid_cache_id_assignment(params->mcache_allocations[plane_index].mcache_x_offsets_plane1,
+ params->mcache_allocations[plane_index].num_mcaches_plane1,
+ &l->plane1.pipe_vp_startx[plane_index], &l->plane1.pipe_vp_endx[plane_index], num_dpps,
+ params->mcache_allocations[plane_index].shift_granularity.p1, &p1shift);
+ }
+ }
+ }
+
+ if (p0pass && p1pass) {
+ for (i = 0; i < params->mcache_allocations[plane_index].num_mcaches_plane0; i++) {
+ params->mcache_allocations[plane_index].mcache_x_offsets_plane0[i] -= p0shift;
+ }
+ for (i = 0; i < params->mcache_allocations[plane_index].num_mcaches_plane1; i++) {
+ params->mcache_allocations[plane_index].mcache_x_offsets_plane1[i] -= p1shift;
+ }
+ }
+
+ params->per_plane_status[plane_index] = p0pass && p1pass;
+ all_pass &= p0pass && p1pass;
+ }
+
+ return all_pass;
+}
+
+static void reset_mcache_allocations(struct dml2_hubp_pipe_mcache_regs *per_plane_pipe_mcache_regs)
+{
+ // Initialize all entries to special valid MCache ID and special valid split coordinate
+ per_plane_pipe_mcache_regs->main.p0.mcache_id_first = MCACHE_ID_UNASSIGNED;
+ per_plane_pipe_mcache_regs->main.p0.mcache_id_second = MCACHE_ID_UNASSIGNED;
+ per_plane_pipe_mcache_regs->main.p0.split_location = SPLIT_LOCATION_UNDEFINED;
+
+ per_plane_pipe_mcache_regs->mall.p0.mcache_id_first = MCACHE_ID_UNASSIGNED;
+ per_plane_pipe_mcache_regs->mall.p0.mcache_id_second = MCACHE_ID_UNASSIGNED;
+ per_plane_pipe_mcache_regs->mall.p0.split_location = SPLIT_LOCATION_UNDEFINED;
+
+ per_plane_pipe_mcache_regs->main.p1.mcache_id_first = MCACHE_ID_UNASSIGNED;
+ per_plane_pipe_mcache_regs->main.p1.mcache_id_second = MCACHE_ID_UNASSIGNED;
+ per_plane_pipe_mcache_regs->main.p1.split_location = SPLIT_LOCATION_UNDEFINED;
+
+ per_plane_pipe_mcache_regs->mall.p1.mcache_id_first = MCACHE_ID_UNASSIGNED;
+ per_plane_pipe_mcache_regs->mall.p1.mcache_id_second = MCACHE_ID_UNASSIGNED;
+ per_plane_pipe_mcache_regs->mall.p1.split_location = SPLIT_LOCATION_UNDEFINED;
+}
+
+void dml2_top_mcache_assign_global_mcache_ids(struct top_mcache_assign_global_mcache_ids_in_out *params)
+{
+ int i;
+ unsigned int j;
+ int next_unused_cache_id = 0;
+
+ for (i = 0; i < params->num_allocations; i++) {
+ if (!params->allocations[i].valid)
+ continue;
+
+ for (j = 0; j < params->allocations[i].num_mcaches_plane0; j++) {
+ params->allocations[i].global_mcache_ids_plane0[j] = next_unused_cache_id++;
+ }
+ for (j = 0; j < params->allocations[i].num_mcaches_plane1; j++) {
+ params->allocations[i].global_mcache_ids_plane1[j] = next_unused_cache_id++;
+ }
+
+ // The "psuedo-last" slice is always wrapped around
+ params->allocations[i].global_mcache_ids_plane0[params->allocations[i].num_mcaches_plane0] =
+ params->allocations[i].global_mcache_ids_plane0[0];
+ params->allocations[i].global_mcache_ids_plane1[params->allocations[i].num_mcaches_plane1] =
+ params->allocations[i].global_mcache_ids_plane1[0];
+
+ // If we need dedicated caches for mall requesting, then we assign them here.
+ if (params->allocations[i].requires_dedicated_mall_mcache) {
+ for (j = 0; j < params->allocations[i].num_mcaches_plane0; j++) {
+ params->allocations[i].global_mcache_ids_mall_plane0[j] = next_unused_cache_id++;
+ }
+ for (j = 0; j < params->allocations[i].num_mcaches_plane1; j++) {
+ params->allocations[i].global_mcache_ids_mall_plane1[j] = next_unused_cache_id++;
+ }
+
+ // The "psuedo-last" slice is always wrapped around
+ params->allocations[i].global_mcache_ids_mall_plane0[params->allocations[i].num_mcaches_plane0] =
+ params->allocations[i].global_mcache_ids_mall_plane0[0];
+ params->allocations[i].global_mcache_ids_mall_plane1[params->allocations[i].num_mcaches_plane1] =
+ params->allocations[i].global_mcache_ids_mall_plane1[0];
+ }
+
+ // If P0 and P1 are sharing caches, then it means the largest mcache IDs for p0 and p1 can be the same
+ // since mcache IDs are always ascending, then it means the largest mcacheID of p1 should be the
+ // largest mcacheID of P0
+ if (params->allocations[i].num_mcaches_plane0 > 0 && params->allocations[i].num_mcaches_plane1 > 0 &&
+ params->allocations[i].last_slice_sharing.plane0_plane1) {
+ params->allocations[i].global_mcache_ids_plane1[params->allocations[i].num_mcaches_plane1 - 1] =
+ params->allocations[i].global_mcache_ids_plane0[params->allocations[i].num_mcaches_plane0 - 1];
+ }
+
+ // If we need dedicated caches handle last slice sharing
+ if (params->allocations[i].requires_dedicated_mall_mcache) {
+ if (params->allocations[i].num_mcaches_plane0 > 0 && params->allocations[i].num_mcaches_plane1 > 0 &&
+ params->allocations[i].last_slice_sharing.plane0_plane1) {
+ params->allocations[i].global_mcache_ids_mall_plane1[params->allocations[i].num_mcaches_plane1 - 1] =
+ params->allocations[i].global_mcache_ids_mall_plane0[params->allocations[i].num_mcaches_plane0 - 1];
+ }
+ // If mall_comb_mcache_l is set then it means that largest mcache ID for MALL p0 can be same as regular read p0
+ if (params->allocations[i].num_mcaches_plane0 > 0 && params->allocations[i].last_slice_sharing.mall_comb_mcache_p0) {
+ params->allocations[i].global_mcache_ids_mall_plane0[params->allocations[i].num_mcaches_plane0 - 1] =
+ params->allocations[i].global_mcache_ids_plane0[params->allocations[i].num_mcaches_plane0 - 1];
+ }
+ // If mall_comb_mcache_c is set then it means that largest mcache ID for MALL p1 can be same as regular
+ // read p1 (which can be same as regular read p0 if plane0_plane1 is also set)
+ if (params->allocations[i].num_mcaches_plane1 > 0 && params->allocations[i].last_slice_sharing.mall_comb_mcache_p1) {
+ params->allocations[i].global_mcache_ids_mall_plane1[params->allocations[i].num_mcaches_plane1 - 1] =
+ params->allocations[i].global_mcache_ids_plane1[params->allocations[i].num_mcaches_plane1 - 1];
+ }
+ }
+
+ // If you don't need dedicated mall mcaches, the mall mcache assignments are identical to the normal requesting
+ if (!params->allocations[i].requires_dedicated_mall_mcache) {
+ memcpy(params->allocations[i].global_mcache_ids_mall_plane0, params->allocations[i].global_mcache_ids_plane0,
+ sizeof(params->allocations[i].global_mcache_ids_mall_plane0));
+ memcpy(params->allocations[i].global_mcache_ids_mall_plane1, params->allocations[i].global_mcache_ids_plane1,
+ sizeof(params->allocations[i].global_mcache_ids_mall_plane1));
+ }
+ }
+}
+
+bool dml2_top_mcache_calc_mcache_count_and_offsets(struct top_mcache_calc_mcache_count_and_offsets_in_out *params)
+{
+ struct dml2_instance *dml = (struct dml2_instance *)params->dml2_instance;
+ struct dml2_top_mcache_verify_mcache_size_locals *l = &dml->scratch.mcache_verify_mcache_size_locals;
+
+ unsigned int total_mcaches_required;
+ unsigned int i;
+ bool result = false;
+
+ if (dml->soc_bbox.num_dcc_mcaches == 0) {
+ return true;
+ }
+
+ total_mcaches_required = 0;
+ l->calc_mcache_params.instance = &dml->core_instance;
+ for (i = 0; i < params->display_config->num_planes; i++) {
+ if (!params->display_config->plane_descriptors[i].surface.dcc.enable) {
+ memset(&params->mcache_allocations[i], 0, sizeof(struct dml2_mcache_surface_allocation));
+ continue;
+ }
+
+ l->calc_mcache_params.plane_descriptor = &params->display_config->plane_descriptors[i];
+ l->calc_mcache_params.mcache_allocation = &params->mcache_allocations[i];
+ l->calc_mcache_params.plane_index = i;
+
+ if (!dml->core_instance.calculate_mcache_allocation(&l->calc_mcache_params)) {
+ result = false;
+ break;
+ }
+
+ if (params->mcache_allocations[i].valid) {
+ total_mcaches_required += params->mcache_allocations[i].num_mcaches_plane0 + params->mcache_allocations[i].num_mcaches_plane1;
+ if (params->mcache_allocations[i].last_slice_sharing.plane0_plane1)
+ total_mcaches_required--;
+ }
+ }
+ DML_LOG_VERBOSE("DML_CORE_DCN3::%s: plane_%d, total_mcaches_required=%d\n", __func__, i, total_mcaches_required);
+
+ if (total_mcaches_required > dml->soc_bbox.num_dcc_mcaches) {
+ result = false;
+ } else {
+ result = true;
+ }
+
+ return result;
+}
+
+static bool dml2_top_soc15_check_mode_supported(struct dml2_check_mode_supported_in_out *in_out)
+{
+ struct dml2_instance *dml = (struct dml2_instance *)in_out->dml2_instance;
+ struct dml2_check_mode_supported_locals *l = &dml->scratch.check_mode_supported_locals;
+ struct dml2_display_cfg_programming *dpmm_programming = &dml->dpmm_instance.dpmm_scratch.programming;
+
+ bool result = false;
+ bool mcache_success = false;
+ memset(dpmm_programming, 0, sizeof(struct dml2_display_cfg_programming));
+
+ setup_unoptimized_display_config_with_meta(dml, &l->base_display_config_with_meta, in_out->display_config);
+
+ l->mode_support_params.instance = &dml->core_instance;
+ l->mode_support_params.display_cfg = &l->base_display_config_with_meta;
+ l->mode_support_params.min_clk_table = &dml->min_clk_table;
+ l->mode_support_params.min_clk_index = l->base_display_config_with_meta.stage1.min_clk_index_for_latency;
+ result = dml->core_instance.mode_support(&l->mode_support_params);
+ l->base_display_config_with_meta.mode_support_result = l->mode_support_params.mode_support_result;
+
+ if (result) {
+ struct optimization_phase_params mcache_phase = {
+ .dml = dml,
+ .display_config = &l->base_display_config_with_meta,
+ .test_function = dml2_top_optimization_test_function_mcache,
+ .optimize_function = dml2_top_optimization_optimize_function_mcache,
+ .optimized_display_config = &l->optimized_display_config_with_meta,
+ .all_or_nothing = false,
+ };
+ mcache_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &mcache_phase);
+ }
+
+ /*
+ * Call DPMM to map all requirements to minimum clock state
+ */
+ if (result) {
+ l->dppm_map_mode_params.min_clk_table = &dml->min_clk_table;
+ l->dppm_map_mode_params.display_cfg = &l->base_display_config_with_meta;
+ l->dppm_map_mode_params.programming = dpmm_programming;
+ l->dppm_map_mode_params.soc_bb = &dml->soc_bbox;
+ l->dppm_map_mode_params.ip = &dml->core_instance.clean_me_up.mode_lib.ip;
+ result = dml->dpmm_instance.map_mode_to_soc_dpm(&l->dppm_map_mode_params);
+ }
+
+ in_out->is_supported = mcache_success;
+ result = result && in_out->is_supported;
+
+ return result;
+}
+
+static bool dml2_top_soc15_build_mode_programming(struct dml2_build_mode_programming_in_out *in_out)
+{
+ struct dml2_instance *dml = (struct dml2_instance *)in_out->dml2_instance;
+ struct dml2_build_mode_programming_locals *l = &dml->scratch.build_mode_programming_locals;
+
+ bool result = false;
+ bool mcache_success = false;
+ bool uclk_pstate_success = false;
+ bool vmin_success = false;
+ bool stutter_success = false;
+
+ memset(l, 0, sizeof(struct dml2_build_mode_programming_locals));
+ memset(in_out->programming, 0, sizeof(struct dml2_display_cfg_programming));
+
+ memcpy(&in_out->programming->display_config, in_out->display_config, sizeof(struct dml2_display_cfg));
+
+ setup_speculative_display_config_with_meta(dml, &l->base_display_config_with_meta, in_out->display_config);
+
+ l->mode_support_params.instance = &dml->core_instance;
+ l->mode_support_params.display_cfg = &l->base_display_config_with_meta;
+ l->mode_support_params.min_clk_table = &dml->min_clk_table;
+ l->mode_support_params.min_clk_index = l->base_display_config_with_meta.stage1.min_clk_index_for_latency;
+ result = dml->core_instance.mode_support(&l->mode_support_params);
+
+ l->base_display_config_with_meta.mode_support_result = l->mode_support_params.mode_support_result;
+
+ if (!result) {
+ setup_unoptimized_display_config_with_meta(dml, &l->base_display_config_with_meta, in_out->display_config);
+
+ l->mode_support_params.instance = &dml->core_instance;
+ l->mode_support_params.display_cfg = &l->base_display_config_with_meta;
+ l->mode_support_params.min_clk_table = &dml->min_clk_table;
+ l->mode_support_params.min_clk_index = l->base_display_config_with_meta.stage1.min_clk_index_for_latency;
+ result = dml->core_instance.mode_support(&l->mode_support_params);
+ l->base_display_config_with_meta.mode_support_result = l->mode_support_params.mode_support_result;
+
+ if (!result) {
+ l->informative_params.instance = &dml->core_instance;
+ l->informative_params.programming = in_out->programming;
+ l->informative_params.mode_is_supported = false;
+ dml->core_instance.populate_informative(&l->informative_params);
+
+ return false;
+ }
+
+ /*
+ * Phase 1: Determine minimum clocks to satisfy latency requirements for this mode
+ */
+ memset(&l->min_clock_for_latency_phase, 0, sizeof(struct optimization_phase_params));
+ l->min_clock_for_latency_phase.dml = dml;
+ l->min_clock_for_latency_phase.display_config = &l->base_display_config_with_meta;
+ l->min_clock_for_latency_phase.init_function = dml2_top_optimization_init_function_min_clk_for_latency;
+ l->min_clock_for_latency_phase.test_function = dml2_top_optimization_test_function_min_clk_for_latency;
+ l->min_clock_for_latency_phase.optimize_function = dml2_top_optimization_optimize_function_min_clk_for_latency;
+ l->min_clock_for_latency_phase.optimized_display_config = &l->optimized_display_config_with_meta;
+ l->min_clock_for_latency_phase.all_or_nothing = false;
+
+ dml2_top_optimization_perform_optimization_phase_1(&l->optimization_phase_locals, &l->min_clock_for_latency_phase);
+
+ memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
+ }
+
+ /*
+ * Phase 2: Satisfy DCC mcache requirements
+ */
+ memset(&l->mcache_phase, 0, sizeof(struct optimization_phase_params));
+ l->mcache_phase.dml = dml;
+ l->mcache_phase.display_config = &l->base_display_config_with_meta;
+ l->mcache_phase.test_function = dml2_top_optimization_test_function_mcache;
+ l->mcache_phase.optimize_function = dml2_top_optimization_optimize_function_mcache;
+ l->mcache_phase.optimized_display_config = &l->optimized_display_config_with_meta;
+ l->mcache_phase.all_or_nothing = true;
+
+ mcache_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &l->mcache_phase);
+
+ if (!mcache_success) {
+ l->informative_params.instance = &dml->core_instance;
+ l->informative_params.programming = in_out->programming;
+ l->informative_params.mode_is_supported = false;
+
+ dml->core_instance.populate_informative(&l->informative_params);
+
+ in_out->programming->informative.failed_mcache_validation = true;
+ return false;
+ }
+
+ memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
+
+ /*
+ * Phase 3: Optimize for Pstate
+ */
+ memset(&l->uclk_pstate_phase, 0, sizeof(struct optimization_phase_params));
+ l->uclk_pstate_phase.dml = dml;
+ l->uclk_pstate_phase.display_config = &l->base_display_config_with_meta;
+ l->uclk_pstate_phase.init_function = dml2_top_optimization_init_function_uclk_pstate;
+ l->uclk_pstate_phase.test_function = dml2_top_optimization_test_function_uclk_pstate;
+ l->uclk_pstate_phase.optimize_function = dml2_top_optimization_optimize_function_uclk_pstate;
+ l->uclk_pstate_phase.optimized_display_config = &l->optimized_display_config_with_meta;
+ l->uclk_pstate_phase.all_or_nothing = true;
+
+ uclk_pstate_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &l->uclk_pstate_phase);
+
+ if (uclk_pstate_success) {
+ memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
+ l->base_display_config_with_meta.stage3.success = true;
+ }
+
+ /*
+ * Phase 4: Optimize for Vmin
+ */
+ memset(&l->vmin_phase, 0, sizeof(struct optimization_phase_params));
+ l->vmin_phase.dml = dml;
+ l->vmin_phase.display_config = &l->base_display_config_with_meta;
+ l->vmin_phase.init_function = dml2_top_optimization_init_function_vmin;
+ l->vmin_phase.test_function = dml2_top_optimization_test_function_vmin;
+ l->vmin_phase.optimize_function = dml2_top_optimization_optimize_function_vmin;
+ l->vmin_phase.optimized_display_config = &l->optimized_display_config_with_meta;
+ l->vmin_phase.all_or_nothing = false;
+
+ vmin_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &l->vmin_phase);
+
+ if (l->optimized_display_config_with_meta.stage4.performed) {
+ /*
+ * when performed is true, optimization has applied to
+ * optimized_display_config_with_meta and it has passed mode
+ * support. However it may or may not pass the test function to
+ * reach actual Vmin. As long as voltage is optimized even if it
+ * doesn't reach Vmin level, there is still power benefit so in
+ * this case we will still copy this optimization into base
+ * display config.
+ */
+ memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
+ l->base_display_config_with_meta.stage4.success = vmin_success;
+ }
+
+ /*
+ * Phase 5: Optimize for Stutter
+ */
+ memset(&l->stutter_phase, 0, sizeof(struct optimization_phase_params));
+ l->stutter_phase.dml = dml;
+ l->stutter_phase.display_config = &l->base_display_config_with_meta;
+ l->stutter_phase.init_function = dml2_top_optimization_init_function_stutter;
+ l->stutter_phase.test_function = dml2_top_optimization_test_function_stutter;
+ l->stutter_phase.optimize_function = dml2_top_optimization_optimize_function_stutter;
+ l->stutter_phase.optimized_display_config = &l->optimized_display_config_with_meta;
+ l->stutter_phase.all_or_nothing = true;
+
+ stutter_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &l->stutter_phase);
+
+ if (stutter_success) {
+ memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
+ l->base_display_config_with_meta.stage5.success = true;
+ }
+
+ /*
+ * Call DPMM to map all requirements to minimum clock state
+ */
+ if (result) {
+ l->dppm_map_mode_params.min_clk_table = &dml->min_clk_table;
+ l->dppm_map_mode_params.display_cfg = &l->base_display_config_with_meta;
+ l->dppm_map_mode_params.programming = in_out->programming;
+ l->dppm_map_mode_params.soc_bb = &dml->soc_bbox;
+ l->dppm_map_mode_params.ip = &dml->core_instance.clean_me_up.mode_lib.ip;
+ result = dml->dpmm_instance.map_mode_to_soc_dpm(&l->dppm_map_mode_params);
+ if (!result)
+ in_out->programming->informative.failed_dpmm = true;
+ }
+
+ if (result) {
+ l->mode_programming_params.instance = &dml->core_instance;
+ l->mode_programming_params.display_cfg = &l->base_display_config_with_meta;
+ l->mode_programming_params.cfg_support_info = &l->base_display_config_with_meta.mode_support_result.cfg_support_info;
+ l->mode_programming_params.programming = in_out->programming;
+ result = dml->core_instance.mode_programming(&l->mode_programming_params);
+ if (!result)
+ in_out->programming->informative.failed_mode_programming = true;
+ }
+
+ if (result) {
+ l->dppm_map_watermarks_params.core = &dml->core_instance;
+ l->dppm_map_watermarks_params.display_cfg = &l->base_display_config_with_meta;
+ l->dppm_map_watermarks_params.programming = in_out->programming;
+ result = dml->dpmm_instance.map_watermarks(&l->dppm_map_watermarks_params);
+ }
+
+ l->informative_params.instance = &dml->core_instance;
+ l->informative_params.programming = in_out->programming;
+ l->informative_params.mode_is_supported = result;
+
+ dml->core_instance.populate_informative(&l->informative_params);
+
+ return result;
+}
+
+bool dml2_top_soc15_build_mcache_programming(struct dml2_build_mcache_programming_in_out *params)
+{
+ bool success = true;
+ int config_index, pipe_index;
+ int first_offset, second_offset;
+ int free_per_plane_reg_index = 0;
+
+ memset(params->per_plane_pipe_mcache_regs, 0, DML2_MAX_PLANES * DML2_MAX_DCN_PIPES * sizeof(struct dml2_hubp_pipe_mcache_regs *));
+
+ for (config_index = 0; config_index < params->num_configurations; config_index++) {
+ for (pipe_index = 0; pipe_index < params->mcache_configurations[config_index].num_pipes; pipe_index++) {
+ // Allocate storage for the mcache regs
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index] = &params->mcache_regs_set[free_per_plane_reg_index++];
+
+ reset_mcache_allocations(params->per_plane_pipe_mcache_regs[config_index][pipe_index]);
+
+ if (params->mcache_configurations[config_index].plane_descriptor->surface.dcc.enable) {
+ // P0 always enabled
+ if (!calculate_first_second_splitting(params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane0,
+ params->mcache_configurations[config_index].mcache_allocation->num_mcaches_plane0,
+ 0,
+ params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane0.viewport_x_start,
+ params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane0.viewport_x_start +
+ params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane0.viewport_width - 1,
+ &first_offset, &second_offset)) {
+ success = false;
+ break;
+ }
+
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p0.mcache_id_first =
+ params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_plane0[first_offset];
+
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p0.mcache_id_first =
+ params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_mall_plane0[first_offset];
+
+ if (second_offset >= 0) {
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p0.mcache_id_second =
+ params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_plane0[second_offset];
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p0.split_location =
+ params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane0[first_offset] - 1;
+
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p0.mcache_id_second =
+ params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_mall_plane0[second_offset];
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p0.split_location =
+ params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane0[first_offset] - 1;
+ }
+
+ // Populate P1 if enabled
+ if (params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane1_enabled) {
+ if (!calculate_first_second_splitting(params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane1,
+ params->mcache_configurations[config_index].mcache_allocation->num_mcaches_plane1,
+ 0,
+ params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane1.viewport_x_start,
+ params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane1.viewport_x_start +
+ params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane1.viewport_width - 1,
+ &first_offset, &second_offset)) {
+ success = false;
+ break;
+ }
+
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p1.mcache_id_first =
+ params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_plane1[first_offset];
+
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p1.mcache_id_first =
+ params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_mall_plane1[first_offset];
+
+ if (second_offset >= 0) {
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p1.mcache_id_second =
+ params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_plane1[second_offset];
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p1.split_location =
+ params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane1[first_offset] - 1;
+
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p1.mcache_id_second =
+ params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_mall_plane1[second_offset];
+ params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p1.split_location =
+ params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane1[first_offset] - 1;
+ }
+ }
+ }
+ }
+ }
+
+ return success;
+}
+
+static const struct dml2_top_funcs soc15_funcs = {
+ .check_mode_supported = dml2_top_soc15_check_mode_supported,
+ .build_mode_programming = dml2_top_soc15_build_mode_programming,
+ .build_mcache_programming = dml2_top_soc15_build_mcache_programming,
+};
+
+bool dml2_top_soc15_initialize_instance(struct dml2_initialize_instance_in_out *in_out)
+{
+ struct dml2_instance *dml = (struct dml2_instance *)in_out->dml2_instance;
+ struct dml2_initialize_instance_locals *l = &dml->scratch.initialize_instance_locals;
+ struct dml2_core_initialize_in_out core_init_params = { 0 };
+ struct dml2_mcg_build_min_clock_table_params_in_out mcg_build_min_clk_params = { 0 };
+ struct dml2_pmo_initialize_in_out pmo_init_params = { 0 };
+ bool result = false;
+
+ memset(l, 0, sizeof(struct dml2_initialize_instance_locals));
+ memset(dml, 0, sizeof(struct dml2_instance));
+
+ memcpy(&dml->ip_caps, &in_out->ip_caps, sizeof(struct dml2_ip_capabilities));
+ memcpy(&dml->soc_bbox, &in_out->soc_bb, sizeof(struct dml2_soc_bb));
+
+ dml->project_id = in_out->options.project_id;
+ dml->pmo_options = in_out->options.pmo_options;
+
+ // Initialize All Components
+ result = dml2_mcg_create(in_out->options.project_id, &dml->mcg_instance);
+
+ if (result)
+ result = dml2_dpmm_create(in_out->options.project_id, &dml->dpmm_instance);
+
+ if (result)
+ result = dml2_core_create(in_out->options.project_id, &dml->core_instance);
+
+ if (result) {
+ mcg_build_min_clk_params.soc_bb = &in_out->soc_bb;
+ mcg_build_min_clk_params.min_clk_table = &dml->min_clk_table;
+ result = dml->mcg_instance.build_min_clock_table(&mcg_build_min_clk_params);
+ }
+
+ if (result) {
+ core_init_params.project_id = in_out->options.project_id;
+ core_init_params.instance = &dml->core_instance;
+ core_init_params.minimum_clock_table = &dml->min_clk_table;
+ core_init_params.explicit_ip_bb = in_out->overrides.explicit_ip_bb;
+ core_init_params.explicit_ip_bb_size = in_out->overrides.explicit_ip_bb_size;
+ core_init_params.ip_caps = &in_out->ip_caps;
+ core_init_params.soc_bb = &in_out->soc_bb;
+ result = dml->core_instance.initialize(&core_init_params);
+
+ if (core_init_params.explicit_ip_bb && core_init_params.explicit_ip_bb_size > 0) {
+ memcpy(&dml->ip_caps, &in_out->ip_caps, sizeof(struct dml2_ip_capabilities));
+ }
+ }
+
+ if (result)
+ result = dml2_pmo_create(in_out->options.project_id, &dml->pmo_instance);
+
+ if (result) {
+ pmo_init_params.instance = &dml->pmo_instance;
+ pmo_init_params.soc_bb = &dml->soc_bbox;
+ pmo_init_params.ip_caps = &dml->ip_caps;
+ pmo_init_params.mcg_clock_table_size = dml->min_clk_table.dram_bw_table.num_entries;
+ pmo_init_params.options = &dml->pmo_options;
+ dml->pmo_instance.initialize(&pmo_init_params);
+ }
+ dml->funcs = soc15_funcs;
+ return result;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top_mcache.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_soc15.h
index 7b1f6f7143d0..53bd8602f9ef 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top_mcache.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml2_top_soc15.h
@@ -2,22 +2,13 @@
//
// Copyright 2024 Advanced Micro Devices, Inc.
-#ifndef __DML_TOP_MCACHE_H__
-#define __DML_TOP_MCACHE_H__
-
-#include "dml2_external_lib_deps.h"
-#include "dml_top_display_cfg_types.h"
-#include "dml_top_types.h"
+#ifndef __DML2_TOP_SOC15_H__
+#define __DML2_TOP_SOC15_H__
#include "dml2_internal_shared_types.h"
+bool dml2_top_soc15_initialize_instance(struct dml2_initialize_instance_in_out *in_out);
bool dml2_top_mcache_calc_mcache_count_and_offsets(struct top_mcache_calc_mcache_count_and_offsets_in_out *params);
-
void dml2_top_mcache_assign_global_mcache_ids(struct top_mcache_assign_global_mcache_ids_in_out *params);
-
bool dml2_top_mcache_validate_admissability(struct top_mcache_validate_admissability_in_out *params);
-
-bool dml2_top_mcache_build_mcache_programming(struct dml2_build_mcache_programming_in_out *params);
-
-bool dml2_top_mcache_unit_test(void);
-
-#endif
+bool dml2_top_soc15_build_mcache_programming(struct dml2_build_mcache_programming_in_out *params);
+#endif /* __DML2_TOP_SOC15_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top.c
deleted file mode 100644
index f9f8869cd8b8..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top.c
+++ /dev/null
@@ -1,354 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-#include "dml2_internal_shared_types.h"
-#include "dml_top.h"
-#include "dml2_mcg_factory.h"
-#include "dml2_core_factory.h"
-#include "dml2_dpmm_factory.h"
-#include "dml2_pmo_factory.h"
-#include "dml_top_mcache.h"
-#include "dml2_top_optimization.h"
-#include "dml2_external_lib_deps.h"
-
-unsigned int dml2_get_instance_size_bytes(void)
-{
- return sizeof(struct dml2_instance);
-}
-
-bool dml2_initialize_instance(struct dml2_initialize_instance_in_out *in_out)
-{
- struct dml2_instance *dml = (struct dml2_instance *)in_out->dml2_instance;
- struct dml2_initialize_instance_locals *l = &dml->scratch.initialize_instance_locals;
- struct dml2_core_initialize_in_out core_init_params = { 0 };
- struct dml2_mcg_build_min_clock_table_params_in_out mcg_build_min_clk_params = { 0 };
- struct dml2_pmo_initialize_in_out pmo_init_params = { 0 };
- bool result = false;
-
- memset(l, 0, sizeof(struct dml2_initialize_instance_locals));
- memset(dml, 0, sizeof(struct dml2_instance));
-
- memcpy(&dml->ip_caps, &in_out->ip_caps, sizeof(struct dml2_ip_capabilities));
- memcpy(&dml->soc_bbox, &in_out->soc_bb, sizeof(struct dml2_soc_bb));
-
- dml->project_id = in_out->options.project_id;
- dml->pmo_options = in_out->options.pmo_options;
-
- // Initialize All Components
- result = dml2_mcg_create(in_out->options.project_id, &dml->mcg_instance);
-
- if (result)
- result = dml2_dpmm_create(in_out->options.project_id, &dml->dpmm_instance);
-
- if (result)
- result = dml2_core_create(in_out->options.project_id, &dml->core_instance);
-
- if (result) {
- mcg_build_min_clk_params.soc_bb = &in_out->soc_bb;
- mcg_build_min_clk_params.min_clk_table = &dml->min_clk_table;
- result = dml->mcg_instance.build_min_clock_table(&mcg_build_min_clk_params);
- }
-
- if (result) {
- core_init_params.project_id = in_out->options.project_id;
- core_init_params.instance = &dml->core_instance;
- core_init_params.minimum_clock_table = &dml->min_clk_table;
- core_init_params.explicit_ip_bb = in_out->overrides.explicit_ip_bb;
- core_init_params.explicit_ip_bb_size = in_out->overrides.explicit_ip_bb_size;
- core_init_params.ip_caps = &in_out->ip_caps;
- core_init_params.soc_bb = &in_out->soc_bb;
- result = dml->core_instance.initialize(&core_init_params);
-
- if (core_init_params.explicit_ip_bb && core_init_params.explicit_ip_bb_size > 0) {
- memcpy(&dml->ip_caps, &in_out->ip_caps, sizeof(struct dml2_ip_capabilities));
- }
- }
-
- if (result)
- result = dml2_pmo_create(in_out->options.project_id, &dml->pmo_instance);
-
- if (result) {
- pmo_init_params.instance = &dml->pmo_instance;
- pmo_init_params.soc_bb = &dml->soc_bbox;
- pmo_init_params.ip_caps = &dml->ip_caps;
- pmo_init_params.mcg_clock_table_size = dml->min_clk_table.dram_bw_table.num_entries;
- pmo_init_params.options = &dml->pmo_options;
- dml->pmo_instance.initialize(&pmo_init_params);
- }
-
- return result;
-}
-
-static void setup_unoptimized_display_config_with_meta(const struct dml2_instance *dml, struct display_configuation_with_meta *out, const struct dml2_display_cfg *display_config)
-{
- memcpy(&out->display_config, display_config, sizeof(struct dml2_display_cfg));
- out->stage1.min_clk_index_for_latency = dml->min_clk_table.dram_bw_table.num_entries - 1; //dml->min_clk_table.clean_me_up.soc_bb.num_states - 1;
-}
-
-static void setup_speculative_display_config_with_meta(const struct dml2_instance *dml, struct display_configuation_with_meta *out, const struct dml2_display_cfg *display_config)
-{
- memcpy(&out->display_config, display_config, sizeof(struct dml2_display_cfg));
- out->stage1.min_clk_index_for_latency = 0;
-}
-
-bool dml2_check_mode_supported(struct dml2_check_mode_supported_in_out *in_out)
-{
- struct dml2_instance *dml = (struct dml2_instance *)in_out->dml2_instance;
- struct dml2_check_mode_supported_locals *l = &dml->scratch.check_mode_supported_locals;
- struct dml2_display_cfg_programming *dpmm_programming = &dml->dpmm_instance.dpmm_scratch.programming;
-
- bool result = false;
- bool mcache_success = false;
-
- memset(dpmm_programming, 0, sizeof(struct dml2_display_cfg_programming));
-
- setup_unoptimized_display_config_with_meta(dml, &l->base_display_config_with_meta, in_out->display_config);
-
- l->mode_support_params.instance = &dml->core_instance;
- l->mode_support_params.display_cfg = &l->base_display_config_with_meta;
- l->mode_support_params.min_clk_table = &dml->min_clk_table;
- l->mode_support_params.min_clk_index = l->base_display_config_with_meta.stage1.min_clk_index_for_latency;
-
- result = dml->core_instance.mode_support(&l->mode_support_params);
- l->base_display_config_with_meta.mode_support_result = l->mode_support_params.mode_support_result;
-
- if (result) {
- struct optimization_phase_params mcache_phase = {
- .dml = dml,
- .display_config = &l->base_display_config_with_meta,
- .test_function = dml2_top_optimization_test_function_mcache,
- .optimize_function = dml2_top_optimization_optimize_function_mcache,
- .optimized_display_config = &l->optimized_display_config_with_meta,
- .all_or_nothing = false,
- };
- mcache_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &mcache_phase);
- }
-
- /*
- * Call DPMM to map all requirements to minimum clock state
- */
- if (result) {
- l->dppm_map_mode_params.min_clk_table = &dml->min_clk_table;
- l->dppm_map_mode_params.display_cfg = &l->base_display_config_with_meta;
- l->dppm_map_mode_params.programming = dpmm_programming;
- l->dppm_map_mode_params.soc_bb = &dml->soc_bbox;
- l->dppm_map_mode_params.ip = &dml->core_instance.clean_me_up.mode_lib.ip;
- result = dml->dpmm_instance.map_mode_to_soc_dpm(&l->dppm_map_mode_params);
- }
-
- in_out->is_supported = mcache_success;
- result = result && in_out->is_supported;
-
- return result;
-}
-
-bool dml2_build_mode_programming(struct dml2_build_mode_programming_in_out *in_out)
-{
- struct dml2_instance *dml = (struct dml2_instance *)in_out->dml2_instance;
- struct dml2_build_mode_programming_locals *l = &dml->scratch.build_mode_programming_locals;
-
- bool result = false;
- bool mcache_success = false;
- bool uclk_pstate_success = false;
- bool vmin_success = false;
- bool stutter_success = false;
- unsigned int i;
-
- memset(l, 0, sizeof(struct dml2_build_mode_programming_locals));
- memset(in_out->programming, 0, sizeof(struct dml2_display_cfg_programming));
-
- memcpy(&in_out->programming->display_config, in_out->display_config, sizeof(struct dml2_display_cfg));
-
- setup_speculative_display_config_with_meta(dml, &l->base_display_config_with_meta, in_out->display_config);
-
- l->mode_support_params.instance = &dml->core_instance;
- l->mode_support_params.display_cfg = &l->base_display_config_with_meta;
- l->mode_support_params.min_clk_table = &dml->min_clk_table;
- l->mode_support_params.min_clk_index = l->base_display_config_with_meta.stage1.min_clk_index_for_latency;
-
- result = dml->core_instance.mode_support(&l->mode_support_params);
- l->base_display_config_with_meta.mode_support_result = l->mode_support_params.mode_support_result;
-
- if (!result) {
- setup_unoptimized_display_config_with_meta(dml, &l->base_display_config_with_meta, in_out->display_config);
-
- l->mode_support_params.instance = &dml->core_instance;
- l->mode_support_params.display_cfg = &l->base_display_config_with_meta;
- l->mode_support_params.min_clk_table = &dml->min_clk_table;
- l->mode_support_params.min_clk_index = l->base_display_config_with_meta.stage1.min_clk_index_for_latency;
-
- result = dml->core_instance.mode_support(&l->mode_support_params);
- l->base_display_config_with_meta.mode_support_result = l->mode_support_params.mode_support_result;
-
- if (!result) {
- l->informative_params.instance = &dml->core_instance;
- l->informative_params.programming = in_out->programming;
- l->informative_params.mode_is_supported = false;
- dml->core_instance.populate_informative(&l->informative_params);
-
- return false;
- }
-
- /*
- * Phase 1: Determine minimum clocks to satisfy latency requirements for this mode
- */
- memset(&l->min_clock_for_latency_phase, 0, sizeof(struct optimization_phase_params));
- l->min_clock_for_latency_phase.dml = dml;
- l->min_clock_for_latency_phase.display_config = &l->base_display_config_with_meta;
- l->min_clock_for_latency_phase.init_function = dml2_top_optimization_init_function_min_clk_for_latency;
- l->min_clock_for_latency_phase.test_function = dml2_top_optimization_test_function_min_clk_for_latency;
- l->min_clock_for_latency_phase.optimize_function = dml2_top_optimization_optimize_function_min_clk_for_latency;
- l->min_clock_for_latency_phase.optimized_display_config = &l->optimized_display_config_with_meta;
- l->min_clock_for_latency_phase.all_or_nothing = false;
-
- dml2_top_optimization_perform_optimization_phase_1(&l->optimization_phase_locals, &l->min_clock_for_latency_phase);
-
- memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
- }
-
- /*
- * Phase 2: Satisfy DCC mcache requirements
- */
- memset(&l->mcache_phase, 0, sizeof(struct optimization_phase_params));
- l->mcache_phase.dml = dml;
- l->mcache_phase.display_config = &l->base_display_config_with_meta;
- l->mcache_phase.test_function = dml2_top_optimization_test_function_mcache;
- l->mcache_phase.optimize_function = dml2_top_optimization_optimize_function_mcache;
- l->mcache_phase.optimized_display_config = &l->optimized_display_config_with_meta;
- l->mcache_phase.all_or_nothing = true;
-
- mcache_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &l->mcache_phase);
-
- if (!mcache_success) {
- l->informative_params.instance = &dml->core_instance;
- l->informative_params.programming = in_out->programming;
- l->informative_params.mode_is_supported = false;
-
- dml->core_instance.populate_informative(&l->informative_params);
-
- in_out->programming->informative.failed_mcache_validation = true;
- return false;
- }
-
- memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
-
- /*
- * Phase 3: Optimize for Pstate
- */
- memset(&l->uclk_pstate_phase, 0, sizeof(struct optimization_phase_params));
- l->uclk_pstate_phase.dml = dml;
- l->uclk_pstate_phase.display_config = &l->base_display_config_with_meta;
- l->uclk_pstate_phase.init_function = dml2_top_optimization_init_function_uclk_pstate;
- l->uclk_pstate_phase.test_function = dml2_top_optimization_test_function_uclk_pstate;
- l->uclk_pstate_phase.optimize_function = dml2_top_optimization_optimize_function_uclk_pstate;
- l->uclk_pstate_phase.optimized_display_config = &l->optimized_display_config_with_meta;
- l->uclk_pstate_phase.all_or_nothing = true;
-
- uclk_pstate_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &l->uclk_pstate_phase);
-
- if (uclk_pstate_success) {
- memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
- l->base_display_config_with_meta.stage3.success = true;
- }
-
- /*
- * Phase 4: Optimize for Vmin
- */
- memset(&l->vmin_phase, 0, sizeof(struct optimization_phase_params));
- l->vmin_phase.dml = dml;
- l->vmin_phase.display_config = &l->base_display_config_with_meta;
- l->vmin_phase.init_function = dml2_top_optimization_init_function_vmin;
- l->vmin_phase.test_function = dml2_top_optimization_test_function_vmin;
- l->vmin_phase.optimize_function = dml2_top_optimization_optimize_function_vmin;
- l->vmin_phase.optimized_display_config = &l->optimized_display_config_with_meta;
- l->vmin_phase.all_or_nothing = false;
-
- vmin_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &l->vmin_phase);
-
- if (l->optimized_display_config_with_meta.stage4.performed) {
- /*
- * when performed is true, optimization has applied to
- * optimized_display_config_with_meta and it has passed mode
- * support. However it may or may not pass the test function to
- * reach actual Vmin. As long as voltage is optimized even if it
- * doesn't reach Vmin level, there is still power benefit so in
- * this case we will still copy this optimization into base
- * display config.
- */
- memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
- l->base_display_config_with_meta.stage4.success = vmin_success;
- }
-
- /*
- * Phase 5: Optimize for Stutter
- */
- memset(&l->stutter_phase, 0, sizeof(struct optimization_phase_params));
- l->stutter_phase.dml = dml;
- l->stutter_phase.display_config = &l->base_display_config_with_meta;
- l->stutter_phase.init_function = dml2_top_optimization_init_function_stutter;
- l->stutter_phase.test_function = dml2_top_optimization_test_function_stutter;
- l->stutter_phase.optimize_function = dml2_top_optimization_optimize_function_stutter;
- l->stutter_phase.optimized_display_config = &l->optimized_display_config_with_meta;
- l->stutter_phase.all_or_nothing = true;
-
- stutter_success = dml2_top_optimization_perform_optimization_phase(&l->optimization_phase_locals, &l->stutter_phase);
-
- if (stutter_success) {
- memcpy(&l->base_display_config_with_meta, &l->optimized_display_config_with_meta, sizeof(struct display_configuation_with_meta));
- l->base_display_config_with_meta.stage5.success = true;
- }
-
- /*
- * Populate mcache programming
- */
- for (i = 0; i < in_out->display_config->num_planes; i++) {
- in_out->programming->plane_programming[i].mcache_allocation = l->base_display_config_with_meta.stage2.mcache_allocations[i];
- }
-
- /*
- * Call DPMM to map all requirements to minimum clock state
- */
- if (result) {
- l->dppm_map_mode_params.min_clk_table = &dml->min_clk_table;
- l->dppm_map_mode_params.display_cfg = &l->base_display_config_with_meta;
- l->dppm_map_mode_params.programming = in_out->programming;
- l->dppm_map_mode_params.soc_bb = &dml->soc_bbox;
- l->dppm_map_mode_params.ip = &dml->core_instance.clean_me_up.mode_lib.ip;
- result = dml->dpmm_instance.map_mode_to_soc_dpm(&l->dppm_map_mode_params);
- if (!result)
- in_out->programming->informative.failed_dpmm = true;
- }
-
- if (result) {
- l->mode_programming_params.instance = &dml->core_instance;
- l->mode_programming_params.display_cfg = &l->base_display_config_with_meta;
- l->mode_programming_params.cfg_support_info = &l->base_display_config_with_meta.mode_support_result.cfg_support_info;
- l->mode_programming_params.programming = in_out->programming;
-
- result = dml->core_instance.mode_programming(&l->mode_programming_params);
- if (!result)
- in_out->programming->informative.failed_mode_programming = true;
- }
-
- if (result) {
- l->dppm_map_watermarks_params.core = &dml->core_instance;
- l->dppm_map_watermarks_params.display_cfg = &l->base_display_config_with_meta;
- l->dppm_map_watermarks_params.programming = in_out->programming;
- result = dml->dpmm_instance.map_watermarks(&l->dppm_map_watermarks_params);
- }
-
- l->informative_params.instance = &dml->core_instance;
- l->informative_params.programming = in_out->programming;
- l->informative_params.mode_is_supported = result;
-
- dml->core_instance.populate_informative(&l->informative_params);
-
- return result;
-}
-
-bool dml2_build_mcache_programming(struct dml2_build_mcache_programming_in_out *in_out)
-{
- return dml2_top_mcache_build_mcache_programming(in_out);
-}
-
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top_mcache.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top_mcache.c
deleted file mode 100644
index a342ebfbe4e7..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/dml2_top/dml_top_mcache.c
+++ /dev/null
@@ -1,549 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-#include "dml2_debug.h"
-
-#include "dml_top_mcache.h"
-#include "lib_float_math.h"
-
-#include "dml2_internal_shared_types.h"
-
-/*
-* Takes an input set of mcache boundaries and finds the appropriate setting of cache programming.
-* Returns true if a valid set of programming can be made, and false otherwise. "Valid" means
-* that the horizontal viewport does not span more than 2 cache slices.
-*
-* It optionally also can apply a constant shift to all the cache boundaries.
-*/
-static const uint32_t MCACHE_ID_UNASSIGNED = 0xF;
-static const uint32_t SPLIT_LOCATION_UNDEFINED = 0xFFFF;
-
-static bool calculate_first_second_splitting(const int *mcache_boundaries, int num_boundaries, int shift,
- int pipe_h_vp_start, int pipe_h_vp_end, int *first_offset, int *second_offset)
-{
- const int MAX_VP = 0xFFFFFF;
- int left_cache_id;
- int right_cache_id;
- int range_start;
- int range_end;
- bool success = false;
-
- if (num_boundaries <= 1) {
- if (first_offset && second_offset) {
- *first_offset = 0;
- *second_offset = -1;
- }
- success = true;
- return success;
- } else {
- range_start = 0;
- for (left_cache_id = 0; left_cache_id < num_boundaries; left_cache_id++) {
- range_end = mcache_boundaries[left_cache_id] - shift - 1;
-
- if (range_start <= pipe_h_vp_start && pipe_h_vp_start <= range_end)
- break;
-
- range_start = range_end + 1;
- }
-
- range_end = MAX_VP;
- for (right_cache_id = num_boundaries - 1; right_cache_id >= -1; right_cache_id--) {
- if (right_cache_id >= 0)
- range_start = mcache_boundaries[right_cache_id] - shift;
- else
- range_start = 0;
-
- if (range_start <= pipe_h_vp_end && pipe_h_vp_end <= range_end) {
- break;
- }
- range_end = range_start - 1;
- }
- right_cache_id = (right_cache_id + 1) % num_boundaries;
-
- if (right_cache_id == left_cache_id) {
- if (first_offset && second_offset) {
- *first_offset = left_cache_id;
- *second_offset = -1;
- }
- success = true;
- } else if (right_cache_id == (left_cache_id + 1) % num_boundaries) {
- if (first_offset && second_offset) {
- *first_offset = left_cache_id;
- *second_offset = right_cache_id;
- }
- success = true;
- }
- }
-
- return success;
-}
-
-/*
-* For a given set of pipe start/end x positions, checks to see it can support the input mcache splitting.
-* It also attempts to "optimize" by finding a shift if the default 0 shift does not work.
-*/
-static bool find_shift_for_valid_cache_id_assignment(int *mcache_boundaries, unsigned int num_boundaries,
- int *pipe_vp_startx, int *pipe_vp_endx, unsigned int pipe_count, int shift_granularity, int *shift)
-{
- int max_shift = 0xFFFF;
- unsigned int pipe_index;
- unsigned int i, slice_width;
- bool success = false;
-
- for (i = 0; i < num_boundaries; i++) {
- if (i == 0)
- slice_width = mcache_boundaries[i];
- else
- slice_width = mcache_boundaries[i] - mcache_boundaries[i - 1];
-
- if (max_shift > (int)slice_width) {
- max_shift = slice_width;
- }
- }
-
- for (*shift = 0; *shift <= max_shift; *shift += shift_granularity) {
- success = true;
- for (pipe_index = 0; pipe_index < pipe_count; pipe_index++) {
- if (!calculate_first_second_splitting(mcache_boundaries, num_boundaries, *shift,
- pipe_vp_startx[pipe_index], pipe_vp_endx[pipe_index], 0, 0)) {
- success = false;
- break;
- }
- }
- if (success)
- break;
- }
-
- return success;
-}
-
-/*
-* Counts the number of elements inside input array within the given span length.
-* Formally, what is the size of the largest subset of the array where the largest and smallest element
-* differ no more than the span.
-*/
-static unsigned int count_elements_in_span(int *array, unsigned int array_size, unsigned int span)
-{
- unsigned int i;
- unsigned int span_start_value;
- unsigned int span_start_index;
- unsigned int greatest_element_count;
-
- if (array_size == 0)
- return 1;
-
- if (span == 0)
- return array_size > 0 ? 1 : 0;
-
- span_start_value = 0;
- span_start_index = 0;
- greatest_element_count = 0;
-
- while (span_start_index < array_size) {
- for (i = span_start_index; i < array_size; i++) {
- if (array[i] - span_start_value <= span) {
- if (i - span_start_index + 1 > greatest_element_count) {
- greatest_element_count = i - span_start_index + 1;
- }
- } else
- break;
- }
-
- span_start_index++;
-
- if (span_start_index < array_size) {
- span_start_value = array[span_start_index - 1] + 1;
- }
- }
-
- return greatest_element_count;
-}
-
-static bool calculate_h_split_for_scaling_transform(int full_vp_width, int h_active, int num_pipes,
- enum dml2_scaling_transform scaling_transform, int *pipe_vp_x_start, int *pipe_vp_x_end)
-{
- int i, slice_width;
- const char MAX_SCL_VP_OVERLAP = 3;
- bool success = false;
-
- switch (scaling_transform) {
- case dml2_scaling_transform_centered:
- case dml2_scaling_transform_aspect_ratio:
- case dml2_scaling_transform_fullscreen:
- slice_width = full_vp_width / num_pipes;
- for (i = 0; i < num_pipes; i++) {
- pipe_vp_x_start[i] = i * slice_width;
- pipe_vp_x_end[i] = (i + 1) * slice_width - 1;
-
- if (pipe_vp_x_start[i] < MAX_SCL_VP_OVERLAP)
- pipe_vp_x_start[i] = 0;
- else
- pipe_vp_x_start[i] -= MAX_SCL_VP_OVERLAP;
-
- if (pipe_vp_x_end[i] > full_vp_width - MAX_SCL_VP_OVERLAP - 1)
- pipe_vp_x_end[i] = full_vp_width - 1;
- else
- pipe_vp_x_end[i] += MAX_SCL_VP_OVERLAP;
- }
- break;
- case dml2_scaling_transform_explicit:
- default:
- success = false;
- break;
- }
-
- return success;
-}
-
-bool dml2_top_mcache_validate_admissability(struct top_mcache_validate_admissability_in_out *params)
-{
- struct dml2_instance *dml = (struct dml2_instance *)params->dml2_instance;
- struct dml2_top_mcache_validate_admissability_locals *l = &dml->scratch.mcache_validate_admissability_locals;
-
- const int MAX_PIXEL_OVERLAP = 6;
- int max_per_pipe_vp_p0 = 0;
- int max_per_pipe_vp_p1 = 0;
- int temp, p0shift, p1shift;
- unsigned int plane_index = 0;
- unsigned int i;
- unsigned int odm_combine_factor;
- unsigned int mpc_combine_factor;
- unsigned int num_dpps;
- unsigned int num_boundaries;
- enum dml2_scaling_transform scaling_transform;
- const struct dml2_plane_parameters *plane;
- const struct dml2_stream_parameters *stream;
-
- bool p0pass = false;
- bool p1pass = false;
- bool all_pass = true;
-
- for (plane_index = 0; plane_index < params->display_cfg->num_planes; plane_index++) {
- if (!params->display_cfg->plane_descriptors[plane_index].surface.dcc.enable)
- continue;
-
- plane = &params->display_cfg->plane_descriptors[plane_index];
- stream = &params->display_cfg->stream_descriptors[plane->stream_index];
-
- num_dpps = odm_combine_factor = params->cfg_support_info->stream_support_info[plane->stream_index].odms_used;
-
- if (odm_combine_factor == 1)
- num_dpps = mpc_combine_factor = (unsigned int)params->cfg_support_info->plane_support_info[plane_index].dpps_used;
- else
- mpc_combine_factor = 1;
-
- if (odm_combine_factor > 1) {
- max_per_pipe_vp_p0 = plane->surface.plane0.width;
- temp = (unsigned int)math_ceil(plane->composition.scaler_info.plane0.h_ratio * stream->timing.h_active / odm_combine_factor);
-
- if (temp < max_per_pipe_vp_p0)
- max_per_pipe_vp_p0 = temp;
-
- max_per_pipe_vp_p1 = plane->surface.plane1.width;
- temp = (unsigned int)math_ceil(plane->composition.scaler_info.plane1.h_ratio * stream->timing.h_active / odm_combine_factor);
-
- if (temp < max_per_pipe_vp_p1)
- max_per_pipe_vp_p1 = temp;
- } else {
- max_per_pipe_vp_p0 = plane->surface.plane0.width / mpc_combine_factor;
- max_per_pipe_vp_p1 = plane->surface.plane1.width / mpc_combine_factor;
- }
-
- max_per_pipe_vp_p0 += 2 * MAX_PIXEL_OVERLAP;
- max_per_pipe_vp_p1 += MAX_PIXEL_OVERLAP;
-
- p0shift = 0;
- p1shift = 0;
-
- // The last element in the unshifted boundary array will always be the first pixel outside the
- // plane, which means theres no mcache associated with it, so -1
- num_boundaries = params->mcache_allocations[plane_index].num_mcaches_plane0 == 0 ? 0 : params->mcache_allocations[plane_index].num_mcaches_plane0 - 1;
- if ((count_elements_in_span(params->mcache_allocations[plane_index].mcache_x_offsets_plane0,
- num_boundaries, max_per_pipe_vp_p0) <= 1) && (num_boundaries <= num_dpps)) {
- p0pass = true;
- }
- num_boundaries = params->mcache_allocations[plane_index].num_mcaches_plane1 == 0 ? 0 : params->mcache_allocations[plane_index].num_mcaches_plane1 - 1;
- if ((count_elements_in_span(params->mcache_allocations[plane_index].mcache_x_offsets_plane1,
- num_boundaries, max_per_pipe_vp_p1) <= 1) && (num_boundaries <= num_dpps)) {
- p1pass = true;
- }
-
- if (!p0pass || !p1pass) {
- if (odm_combine_factor > 1) {
- num_dpps = odm_combine_factor;
- scaling_transform = plane->composition.scaling_transform;
- } else {
- num_dpps = mpc_combine_factor;
- scaling_transform = dml2_scaling_transform_fullscreen;
- }
-
- if (!p0pass) {
- if (plane->composition.viewport.stationary) {
- calculate_h_split_for_scaling_transform(plane->surface.plane0.width,
- stream->timing.h_active, num_dpps, scaling_transform,
- &l->plane0.pipe_vp_startx[plane_index], &l->plane0.pipe_vp_endx[plane_index]);
- p0pass = find_shift_for_valid_cache_id_assignment(params->mcache_allocations[plane_index].mcache_x_offsets_plane0,
- params->mcache_allocations[plane_index].num_mcaches_plane0,
- &l->plane0.pipe_vp_startx[plane_index], &l->plane0.pipe_vp_endx[plane_index], num_dpps,
- params->mcache_allocations[plane_index].shift_granularity.p0, &p0shift);
- }
- }
- if (!p1pass) {
- if (plane->composition.viewport.stationary) {
- calculate_h_split_for_scaling_transform(plane->surface.plane1.width,
- stream->timing.h_active, num_dpps, scaling_transform,
- &l->plane0.pipe_vp_startx[plane_index], &l->plane0.pipe_vp_endx[plane_index]);
- p1pass = find_shift_for_valid_cache_id_assignment(params->mcache_allocations[plane_index].mcache_x_offsets_plane1,
- params->mcache_allocations[plane_index].num_mcaches_plane1,
- &l->plane1.pipe_vp_startx[plane_index], &l->plane1.pipe_vp_endx[plane_index], num_dpps,
- params->mcache_allocations[plane_index].shift_granularity.p1, &p1shift);
- }
- }
- }
-
- if (p0pass && p1pass) {
- for (i = 0; i < params->mcache_allocations[plane_index].num_mcaches_plane0; i++) {
- params->mcache_allocations[plane_index].mcache_x_offsets_plane0[i] -= p0shift;
- }
- for (i = 0; i < params->mcache_allocations[plane_index].num_mcaches_plane1; i++) {
- params->mcache_allocations[plane_index].mcache_x_offsets_plane1[i] -= p1shift;
- }
- }
-
- params->per_plane_status[plane_index] = p0pass && p1pass;
- all_pass &= p0pass && p1pass;
- }
-
- return all_pass;
-}
-
-static void reset_mcache_allocations(struct dml2_hubp_pipe_mcache_regs *per_plane_pipe_mcache_regs)
-{
- // Initialize all entries to special valid MCache ID and special valid split coordinate
- per_plane_pipe_mcache_regs->main.p0.mcache_id_first = MCACHE_ID_UNASSIGNED;
- per_plane_pipe_mcache_regs->main.p0.mcache_id_second = MCACHE_ID_UNASSIGNED;
- per_plane_pipe_mcache_regs->main.p0.split_location = SPLIT_LOCATION_UNDEFINED;
-
- per_plane_pipe_mcache_regs->mall.p0.mcache_id_first = MCACHE_ID_UNASSIGNED;
- per_plane_pipe_mcache_regs->mall.p0.mcache_id_second = MCACHE_ID_UNASSIGNED;
- per_plane_pipe_mcache_regs->mall.p0.split_location = SPLIT_LOCATION_UNDEFINED;
-
- per_plane_pipe_mcache_regs->main.p1.mcache_id_first = MCACHE_ID_UNASSIGNED;
- per_plane_pipe_mcache_regs->main.p1.mcache_id_second = MCACHE_ID_UNASSIGNED;
- per_plane_pipe_mcache_regs->main.p1.split_location = SPLIT_LOCATION_UNDEFINED;
-
- per_plane_pipe_mcache_regs->mall.p1.mcache_id_first = MCACHE_ID_UNASSIGNED;
- per_plane_pipe_mcache_regs->mall.p1.mcache_id_second = MCACHE_ID_UNASSIGNED;
- per_plane_pipe_mcache_regs->mall.p1.split_location = SPLIT_LOCATION_UNDEFINED;
-}
-
-bool dml2_top_mcache_build_mcache_programming(struct dml2_build_mcache_programming_in_out *params)
-{
- bool success = true;
- int config_index, pipe_index;
- int first_offset, second_offset;
- int free_per_plane_reg_index = 0;
-
- memset(params->per_plane_pipe_mcache_regs, 0, DML2_MAX_PLANES * DML2_MAX_DCN_PIPES * sizeof(struct dml2_hubp_pipe_mcache_regs *));
-
- for (config_index = 0; config_index < params->num_configurations; config_index++) {
- for (pipe_index = 0; pipe_index < params->mcache_configurations[config_index].num_pipes; pipe_index++) {
- // Allocate storage for the mcache regs
- params->per_plane_pipe_mcache_regs[config_index][pipe_index] = &params->mcache_regs_set[free_per_plane_reg_index++];
-
- reset_mcache_allocations(params->per_plane_pipe_mcache_regs[config_index][pipe_index]);
-
- if (params->mcache_configurations[config_index].plane_descriptor->surface.dcc.enable) {
- // P0 always enabled
- if (!calculate_first_second_splitting(params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane0,
- params->mcache_configurations[config_index].mcache_allocation->num_mcaches_plane0,
- 0,
- params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane0.viewport_x_start,
- params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane0.viewport_x_start +
- params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane0.viewport_width - 1,
- &first_offset, &second_offset)) {
- success = false;
- break;
- }
-
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p0.mcache_id_first =
- params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_plane0[first_offset];
-
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p0.mcache_id_first =
- params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_mall_plane0[first_offset];
-
- if (second_offset >= 0) {
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p0.mcache_id_second =
- params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_plane0[second_offset];
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p0.split_location =
- params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane0[first_offset] - 1;
-
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p0.mcache_id_second =
- params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_mall_plane0[second_offset];
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p0.split_location =
- params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane0[first_offset] - 1;
- }
-
- // Populate P1 if enabled
- if (params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane1_enabled) {
- if (!calculate_first_second_splitting(params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane1,
- params->mcache_configurations[config_index].mcache_allocation->num_mcaches_plane1,
- 0,
- params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane1.viewport_x_start,
- params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane1.viewport_x_start +
- params->mcache_configurations[config_index].pipe_configurations[pipe_index].plane1.viewport_width - 1,
- &first_offset, &second_offset)) {
- success = false;
- break;
- }
-
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p1.mcache_id_first =
- params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_plane1[first_offset];
-
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p1.mcache_id_first =
- params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_mall_plane1[first_offset];
-
- if (second_offset >= 0) {
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p1.mcache_id_second =
- params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_plane1[second_offset];
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->main.p1.split_location =
- params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane1[first_offset] - 1;
-
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p1.mcache_id_second =
- params->mcache_configurations[config_index].mcache_allocation->global_mcache_ids_mall_plane1[second_offset];
- params->per_plane_pipe_mcache_regs[config_index][pipe_index]->mall.p1.split_location =
- params->mcache_configurations[config_index].mcache_allocation->mcache_x_offsets_plane1[first_offset] - 1;
- }
- }
- }
- }
- }
-
- return success;
-}
-
-void dml2_top_mcache_assign_global_mcache_ids(struct top_mcache_assign_global_mcache_ids_in_out *params)
-{
- int i;
- unsigned int j;
- int next_unused_cache_id = 0;
-
- for (i = 0; i < params->num_allocations; i++) {
- if (!params->allocations[i].valid)
- continue;
-
- for (j = 0; j < params->allocations[i].num_mcaches_plane0; j++) {
- params->allocations[i].global_mcache_ids_plane0[j] = next_unused_cache_id++;
- }
- for (j = 0; j < params->allocations[i].num_mcaches_plane1; j++) {
- params->allocations[i].global_mcache_ids_plane1[j] = next_unused_cache_id++;
- }
-
- // The "psuedo-last" slice is always wrapped around
- params->allocations[i].global_mcache_ids_plane0[params->allocations[i].num_mcaches_plane0] =
- params->allocations[i].global_mcache_ids_plane0[0];
- params->allocations[i].global_mcache_ids_plane1[params->allocations[i].num_mcaches_plane1] =
- params->allocations[i].global_mcache_ids_plane1[0];
-
- // If we need dedicated caches for mall requesting, then we assign them here.
- if (params->allocations[i].requires_dedicated_mall_mcache) {
- for (j = 0; j < params->allocations[i].num_mcaches_plane0; j++) {
- params->allocations[i].global_mcache_ids_mall_plane0[j] = next_unused_cache_id++;
- }
- for (j = 0; j < params->allocations[i].num_mcaches_plane1; j++) {
- params->allocations[i].global_mcache_ids_mall_plane1[j] = next_unused_cache_id++;
- }
-
- // The "psuedo-last" slice is always wrapped around
- params->allocations[i].global_mcache_ids_mall_plane0[params->allocations[i].num_mcaches_plane0] =
- params->allocations[i].global_mcache_ids_mall_plane0[0];
- params->allocations[i].global_mcache_ids_mall_plane1[params->allocations[i].num_mcaches_plane1] =
- params->allocations[i].global_mcache_ids_mall_plane1[0];
- }
-
- // If P0 and P1 are sharing caches, then it means the largest mcache IDs for p0 and p1 can be the same
- // since mcache IDs are always ascending, then it means the largest mcacheID of p1 should be the
- // largest mcacheID of P0
- if (params->allocations[i].num_mcaches_plane0 > 0 && params->allocations[i].num_mcaches_plane1 > 0 &&
- params->allocations[i].last_slice_sharing.plane0_plane1) {
- params->allocations[i].global_mcache_ids_plane1[params->allocations[i].num_mcaches_plane1 - 1] =
- params->allocations[i].global_mcache_ids_plane0[params->allocations[i].num_mcaches_plane0 - 1];
- }
-
- // If we need dedicated caches handle last slice sharing
- if (params->allocations[i].requires_dedicated_mall_mcache) {
- if (params->allocations[i].num_mcaches_plane0 > 0 && params->allocations[i].num_mcaches_plane1 > 0 &&
- params->allocations[i].last_slice_sharing.plane0_plane1) {
- params->allocations[i].global_mcache_ids_mall_plane1[params->allocations[i].num_mcaches_plane1 - 1] =
- params->allocations[i].global_mcache_ids_mall_plane0[params->allocations[i].num_mcaches_plane0 - 1];
- }
- // If mall_comb_mcache_l is set then it means that largest mcache ID for MALL p0 can be same as regular read p0
- if (params->allocations[i].num_mcaches_plane0 > 0 && params->allocations[i].last_slice_sharing.mall_comb_mcache_p0) {
- params->allocations[i].global_mcache_ids_mall_plane0[params->allocations[i].num_mcaches_plane0 - 1] =
- params->allocations[i].global_mcache_ids_plane0[params->allocations[i].num_mcaches_plane0 - 1];
- }
- // If mall_comb_mcache_c is set then it means that largest mcache ID for MALL p1 can be same as regular
- // read p1 (which can be same as regular read p0 if plane0_plane1 is also set)
- if (params->allocations[i].num_mcaches_plane1 > 0 && params->allocations[i].last_slice_sharing.mall_comb_mcache_p1) {
- params->allocations[i].global_mcache_ids_mall_plane1[params->allocations[i].num_mcaches_plane1 - 1] =
- params->allocations[i].global_mcache_ids_plane1[params->allocations[i].num_mcaches_plane1 - 1];
- }
- }
-
- // If you don't need dedicated mall mcaches, the mall mcache assignments are identical to the normal requesting
- if (!params->allocations[i].requires_dedicated_mall_mcache) {
- memcpy(params->allocations[i].global_mcache_ids_mall_plane0, params->allocations[i].global_mcache_ids_plane0,
- sizeof(params->allocations[i].global_mcache_ids_mall_plane0));
- memcpy(params->allocations[i].global_mcache_ids_mall_plane1, params->allocations[i].global_mcache_ids_plane1,
- sizeof(params->allocations[i].global_mcache_ids_mall_plane1));
- }
- }
-}
-
-bool dml2_top_mcache_calc_mcache_count_and_offsets(struct top_mcache_calc_mcache_count_and_offsets_in_out *params)
-{
- struct dml2_instance *dml = (struct dml2_instance *)params->dml2_instance;
- struct dml2_top_mcache_verify_mcache_size_locals *l = &dml->scratch.mcache_verify_mcache_size_locals;
-
- unsigned int total_mcaches_required;
- unsigned int i;
- bool result = false;
-
- if (dml->soc_bbox.num_dcc_mcaches == 0) {
- return true;
- }
-
- total_mcaches_required = 0;
- l->calc_mcache_params.instance = &dml->core_instance;
- for (i = 0; i < params->display_config->num_planes; i++) {
- if (!params->display_config->plane_descriptors[i].surface.dcc.enable) {
- memset(&params->mcache_allocations[i], 0, sizeof(struct dml2_mcache_surface_allocation));
- continue;
- }
-
- l->calc_mcache_params.plane_descriptor = &params->display_config->plane_descriptors[i];
- l->calc_mcache_params.mcache_allocation = &params->mcache_allocations[i];
- l->calc_mcache_params.plane_index = i;
-
- if (!dml->core_instance.calculate_mcache_allocation(&l->calc_mcache_params)) {
- result = false;
- break;
- }
-
- if (params->mcache_allocations[i].valid) {
- total_mcaches_required += params->mcache_allocations[i].num_mcaches_plane0 + params->mcache_allocations[i].num_mcaches_plane1;
- if (params->mcache_allocations[i].last_slice_sharing.plane0_plane1)
- total_mcaches_required--;
- }
- }
- dml2_printf("DML_CORE_DCN3::%s: plane_%d, total_mcaches_required=%d\n", __func__, i, total_mcaches_required);
-
- if (total_mcaches_required > dml->soc_bbox.num_dcc_mcaches) {
- result = false;
- } else {
- result = true;
- }
-
- return result;
-}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_debug.c b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_debug.c
deleted file mode 100644
index e9b8e10695ae..000000000000
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_debug.c
+++ /dev/null
@@ -1,31 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-#include "dml2_debug.h"
-
-int dml2_printf(const char *format, ...)
-{
-#ifdef _DEBUG
-#ifdef _DEBUG_PRINTS
- int result;
- va_list args;
- va_start(args, format);
-
- result = vprintf(format, args);
-
- va_end(args);
-
- return result;
-#else
- return 0;
-#endif
-#else
- return 0;
-#endif
-}
-
-void dml2_assert(int condition)
-{
- //ASSERT(condition);
-}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_debug.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_debug.h
index d51a1b6c62f2..611c80f4f1bf 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_debug.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_debug.h
@@ -5,13 +5,185 @@
#ifndef __DML2_DEBUG_H__
#define __DML2_DEBUG_H__
-#ifdef _DEBUG
-#define DML2_ASSERT(condition) dml2_assert(condition)
+#include "os_types.h"
+#define DML_ASSERT(condition) ASSERT(condition)
+#define DML_LOG_LEVEL_DEFAULT DML_LOG_LEVEL_WARN
+#define DML_LOG_INTERNAL(fmt, ...) dm_output_to_console(fmt, ## __VA_ARGS__)
+
+/* private helper macros */
+#define _BOOL_FORMAT(field) "%s", field ? "true" : "false"
+#define _UINT_FORMAT(field) "%u", field
+#define _INT_FORMAT(field) "%d", field
+#define _DOUBLE_FORMAT(field) "%lf", field
+#define _ELEMENT_FUNC "function"
+#define _ELEMENT_COMP_IF "component_interface"
+#define _ELEMENT_TOP_IF "top_interface"
+#define _LOG_ENTRY(element) do { \
+ DML_LOG_INTERNAL("<"element" name=\""); \
+ DML_LOG_INTERNAL(__func__); \
+ DML_LOG_INTERNAL("\">\n"); \
+} while (0)
+#define _LOG_EXIT(element) DML_LOG_INTERNAL("</"element">\n")
+#define _LOG_SCALAR(field, format) do { \
+ DML_LOG_INTERNAL(#field" = "format(field)); \
+ DML_LOG_INTERNAL("\n"); \
+} while (0)
+#define _LOG_ARRAY(field, size, format) do { \
+ DML_LOG_INTERNAL(#field " = ["); \
+ for (int _i = 0; _i < (int) size; _i++) { \
+ DML_LOG_INTERNAL(format(field[_i])); \
+ if (_i + 1 == (int) size) \
+ DML_LOG_INTERNAL("]\n"); \
+ else \
+ DML_LOG_INTERNAL(", "); \
+}} while (0)
+#define _LOG_2D_ARRAY(field, size0, size1, format) do { \
+ DML_LOG_INTERNAL(#field" = ["); \
+ for (int _i = 0; _i < (int) size0; _i++) { \
+ DML_LOG_INTERNAL("\n\t["); \
+ for (int _j = 0; _j < (int) size1; _j++) { \
+ DML_LOG_INTERNAL(format(field[_i][_j])); \
+ if (_j + 1 == (int) size1) \
+ DML_LOG_INTERNAL("]"); \
+ else \
+ DML_LOG_INTERNAL(", "); \
+ } \
+ if (_i + 1 == (int) size0) \
+ DML_LOG_INTERNAL("]\n"); \
+ else \
+ DML_LOG_INTERNAL(", "); \
+ } \
+} while (0)
+#define _LOG_3D_ARRAY(field, size0, size1, size2, format) do { \
+ DML_LOG_INTERNAL(#field" = ["); \
+ for (int _i = 0; _i < (int) size0; _i++) { \
+ DML_LOG_INTERNAL("\n\t["); \
+ for (int _j = 0; _j < (int) size1; _j++) { \
+ DML_LOG_INTERNAL("["); \
+ for (int _k = 0; _k < (int) size2; _k++) { \
+ DML_LOG_INTERNAL(format(field[_i][_j][_k])); \
+ if (_k + 1 == (int) size2) \
+ DML_LOG_INTERNAL("]"); \
+ else \
+ DML_LOG_INTERNAL(", "); \
+ } \
+ if (_j + 1 == (int) size1) \
+ DML_LOG_INTERNAL("]"); \
+ else \
+ DML_LOG_INTERNAL(", "); \
+ } \
+ if (_i + 1 == (int) size0) \
+ DML_LOG_INTERNAL("]\n"); \
+ else \
+ DML_LOG_INTERNAL(", "); \
+ } \
+} while (0)
+
+/* fatal errors for unrecoverable DML states until a full reset */
+#define DML_LOG_LEVEL_FATAL 0
+/* unexpected but recoverable failures inside DML */
+#define DML_LOG_LEVEL_ERROR 1
+/* unexpected inputs or events to DML */
+#define DML_LOG_LEVEL_WARN 2
+/* high level tracing of DML interfaces */
+#define DML_LOG_LEVEL_INFO 3
+/* tracing of DML internal executions */
+#define DML_LOG_LEVEL_DEBUG 4
+/* detailed tracing of DML calculation procedure */
+#define DML_LOG_LEVEL_VERBOSE 5
+
+#ifndef DML_LOG_LEVEL
+#define DML_LOG_LEVEL DML_LOG_LEVEL_DEFAULT
+#endif /* #ifndef DML_LOG_LEVEL */
+
+/* public macros for DML_LOG_LEVEL_FATAL and up */
+#define DML_LOG_FATAL(fmt, ...) DML_LOG_INTERNAL("[DML FATAL] " fmt, ## __VA_ARGS__)
+
+/* public macros for DML_LOG_LEVEL_ERROR and up */
+#if DML_LOG_LEVEL >= DML_LOG_LEVEL_ERROR
+#define DML_LOG_ERROR(fmt, ...) DML_LOG_INTERNAL("[DML ERROR] "fmt, ## __VA_ARGS__)
+#define DML_ASSERT_MSG(condition, fmt, ...) \
+ do { \
+ if (!(condition)) { \
+ DML_LOG_ERROR("ASSERT hit in %s line %d\n", __func__, __LINE__); \
+ DML_LOG_ERROR(fmt, ## __VA_ARGS__); \
+ DML_ASSERT(condition); \
+ } \
+ } while (0)
#else
-#define DML2_ASSERT(condition)
+#define DML_LOG_ERROR(fmt, ...) ((void)0)
+#define DML_ASSERT_MSG(condition, fmt, ...) ((void)0)
#endif
-int dml2_printf(const char *format, ...);
-void dml2_assert(int condition);
+/* public macros for DML_LOG_LEVEL_WARN and up */
+#if DML_LOG_LEVEL >= DML_LOG_LEVEL_WARN
+#define DML_LOG_WARN(fmt, ...) DML_LOG_INTERNAL("[DML WARN] "fmt, ## __VA_ARGS__)
+#else
+#define DML_LOG_WARN(fmt, ...) ((void)0)
+#endif
+/* public macros for DML_LOG_LEVEL_INFO and up */
+#if DML_LOG_LEVEL >= DML_LOG_LEVEL_INFO
+#define DML_LOG_INFO(fmt, ...) DML_LOG_INTERNAL("[DML INFO] "fmt, ## __VA_ARGS__)
+#define DML_LOG_TOP_IF_ENTER() _LOG_ENTRY(_ELEMENT_TOP_IF)
+#define DML_LOG_TOP_IF_EXIT() _LOG_EXIT(_ELEMENT_TOP_IF)
+#else
+#define DML_LOG_INFO(fmt, ...) ((void)0)
+#define DML_LOG_TOP_IF_ENTER() ((void)0)
+#define DML_LOG_TOP_IF_EXIT() ((void)0)
+#endif
+
+/* public macros for DML_LOG_LEVEL_DEBUG and up */
+#if DML_LOG_LEVEL >= DML_LOG_LEVEL_DEBUG
+#define DML_LOG_DEBUG(fmt, ...) DML_LOG_INTERNAL(fmt, ## __VA_ARGS__)
+#define DML_LOG_COMP_IF_ENTER() _LOG_ENTRY(_ELEMENT_COMP_IF)
+#define DML_LOG_COMP_IF_EXIT() _LOG_EXIT(_ELEMENT_COMP_IF)
+#define DML_LOG_FUNC_ENTER() _LOG_ENTRY(_ELEMENT_FUNC)
+#define DML_LOG_FUNC_EXIT() _LOG_EXIT(_ELEMENT_FUNC)
+#define DML_LOG_DEBUG_BOOL(field) _LOG_SCALAR(field, _BOOL_FORMAT)
+#define DML_LOG_DEBUG_UINT(field) _LOG_SCALAR(field, _UINT_FORMAT)
+#define DML_LOG_DEBUG_INT(field) _LOG_SCALAR(field, _INT_FORMAT)
+#define DML_LOG_DEBUG_DOUBLE(field) _LOG_SCALAR(field, _DOUBLE_FORMAT)
+#define DML_LOG_DEBUG_ARRAY_BOOL(field, size) _LOG_ARRAY(field, size, _BOOL_FORMAT)
+#define DML_LOG_DEBUG_ARRAY_UINT(field, size) _LOG_ARRAY(field, size, _UINT_FORMAT)
+#define DML_LOG_DEBUG_ARRAY_INT(field, size) _LOG_ARRAY(field, size, _INT_FORMAT)
+#define DML_LOG_DEBUG_ARRAY_DOUBLE(field, size) _LOG_ARRAY(field, size, _DOUBLE_FORMAT)
+#define DML_LOG_DEBUG_2D_ARRAY_BOOL(field, size0, size1) _LOG_2D_ARRAY(field, size0, size1, _BOOL_FORMAT)
+#define DML_LOG_DEBUG_2D_ARRAY_UINT(field, size0, size1) _LOG_2D_ARRAY(field, size0, size1, _UINT_FORMAT)
+#define DML_LOG_DEBUG_2D_ARRAY_INT(field, size0, size1) _LOG_2D_ARRAY(field, size0, size1, _INT_FORMAT)
+#define DML_LOG_DEBUG_2D_ARRAY_DOUBLE(field, size0, size1) _LOG_2D_ARRAY(field, size0, size1, _DOUBLE_FORMAT)
+#define DML_LOG_DEBUG_3D_ARRAY_BOOL(field, size0, size1, size2) _LOG_3D_ARRAY(field, size0, size1, size2, _BOOL_FORMAT)
+#define DML_LOG_DEBUG_3D_ARRAY_UINT(field, size0, size1, size2) _LOG_3D_ARRAY(field, size0, size1, size2, _UINT_FORMAT)
+#define DML_LOG_DEBUG_3D_ARRAY_INT(field, size0, size1, size2) _LOG_3D_ARRAY(field, size0, size1, size2, _INT_FORMAT)
+#define DML_LOG_DEBUG_3D_ARRAY_DOUBLE(field, size0, size1, size2) _LOG_3D_ARRAY(field, size0, size1, size2, _DOUBLE_FORMAT)
+#else
+#define DML_LOG_DEBUG(fmt, ...) ((void)0)
+#define DML_LOG_COMP_IF_ENTER() ((void)0)
+#define DML_LOG_COMP_IF_EXIT() ((void)0)
+#define DML_LOG_FUNC_ENTER() ((void)0)
+#define DML_LOG_FUNC_EXIT() ((void)0)
+#define DML_LOG_DEBUG_BOOL(field) ((void)0)
+#define DML_LOG_DEBUG_UINT(field) ((void)0)
+#define DML_LOG_DEBUG_INT(field) ((void)0)
+#define DML_LOG_DEBUG_DOUBLE(field) ((void)0)
+#define DML_LOG_DEBUG_ARRAY_BOOL(field, size) ((void)0)
+#define DML_LOG_DEBUG_ARRAY_UINT(field, size) ((void)0)
+#define DML_LOG_DEBUG_ARRAY_INT(field, size) ((void)0)
+#define DML_LOG_DEBUG_ARRAY_DOUBLE(field, size) ((void)0)
+#define DML_LOG_DEBUG_2D_ARRAY_BOOL(field, size0, size1) ((void)0)
+#define DML_LOG_DEBUG_2D_ARRAY_UINT(field, size0, size1) ((void)0)
+#define DML_LOG_DEBUG_2D_ARRAY_INT(field, size0, size1) ((void)0)
+#define DML_LOG_DEBUG_2D_ARRAY_DOUBLE(field, size0, size1) ((void)0)
+#define DML_LOG_DEBUG_3D_ARRAY_BOOL(field, size0, size1, size2) ((void)0)
+#define DML_LOG_DEBUG_3D_ARRAY_UINT(field, size0, size1, size2) ((void)0)
+#define DML_LOG_DEBUG_3D_ARRAY_INT(field, size0, size1, size2) ((void)0)
+#define DML_LOG_DEBUG_3D_ARRAY_DOUBLE(field, size0, size1, size2) ((void)0)
#endif
+
+/* public macros for DML_LOG_LEVEL_VERBOSE */
+#if DML_LOG_LEVEL >= DML_LOG_LEVEL_VERBOSE
+#define DML_LOG_VERBOSE(fmt, ...) DML_LOG_INTERNAL(fmt, ## __VA_ARGS__)
+#else
+#define DML_LOG_VERBOSE(fmt, ...) ((void)0)
+#endif /* #if DML_LOG_LEVEL >= DML_LOG_LEVEL_VERBOSE */
+#endif /* __DML2_DEBUG_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_internal_shared_types.h b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_internal_shared_types.h
index aeac9f159fa5..d52aa82283b3 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_internal_shared_types.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml21/src/inc/dml2_internal_shared_types.h
@@ -8,7 +8,6 @@
#include "dml2_external_lib_deps.h"
#include "dml_top_types.h"
#include "dml2_core_shared_types.h"
-
/*
* DML2 MCG Types and Interfaces
*/
@@ -39,6 +38,12 @@ struct dml2_mcg_min_clock_table {
} max_clocks_khz;
struct {
+ unsigned int dispclk;
+ unsigned int dppclk;
+ unsigned int dtbclk;
+ } max_ss_clocks_khz;
+
+ struct {
unsigned int dprefclk;
unsigned int xtalclk;
unsigned int pcierefclk;
@@ -63,10 +68,8 @@ struct dml2_mcg_build_min_clock_table_params_in_out {
*/
struct dml2_mcg_min_clock_table *min_clk_table;
};
-
struct dml2_mcg_instance {
bool (*build_min_clock_table)(struct dml2_mcg_build_min_clock_table_params_in_out *in_out);
- bool (*unit_test)(void);
};
/*
@@ -81,7 +84,6 @@ struct dml2_dpmm_map_mode_to_soc_dpm_params_in_out {
struct dml2_soc_bb *soc_bb;
struct dml2_mcg_min_clock_table *min_clk_table;
const struct display_configuation_with_meta *display_cfg;
-
struct {
bool perform_pseudo_map;
struct dml2_core_internal_soc_bb *soc_bb;
@@ -113,7 +115,6 @@ struct dml2_dpmm_scratch {
struct dml2_dpmm_instance {
bool (*map_mode_to_soc_dpm)(struct dml2_dpmm_map_mode_to_soc_dpm_params_in_out *in_out);
bool (*map_watermarks)(struct dml2_dpmm_map_watermarks_params_in_out *in_out);
- bool (*unit_test)(void);
struct dml2_dpmm_scratch dpmm_scratch;
};
@@ -201,6 +202,8 @@ struct dml2_core_mode_support_result {
} active;
unsigned int dispclk_khz;
+ unsigned int dpprefclk_khz;
+ unsigned int dtbrefclk_khz;
unsigned int dcfclk_deepsleep_khz;
unsigned int socclk_khz;
@@ -309,7 +312,7 @@ struct dml2_optimization_stage3_state {
// The pstate support mode for each plane
// The number of valid elements == display_cfg.num_planes
// The indexing of pstate_switch_modes matches plane_descriptors[]
- enum dml2_uclk_pstate_support_method pstate_switch_modes[DML2_MAX_PLANES];
+ enum dml2_pstate_method pstate_switch_modes[DML2_MAX_PLANES];
// Meta-data for implicit SVP generation, indexed by stream index
struct dml2_implicit_svp_meta stream_svp_meta[DML2_MAX_PLANES];
@@ -356,6 +359,10 @@ struct display_configuation_with_meta {
struct dml2_optimization_stage5_state stage5;
};
+struct dml2_pmo_pstate_strategy {
+ enum dml2_pstate_method per_stream_pstate_method[DML2_MAX_PLANES];
+ bool allow_state_increase;
+};
struct dml2_core_mode_support_in_out {
/*
* Inputs
@@ -365,7 +372,6 @@ struct dml2_core_mode_support_in_out {
struct dml2_mcg_min_clock_table *min_clk_table;
int min_clk_index;
-
/*
* Outputs
*/
@@ -395,7 +401,6 @@ struct dml2_core_mode_programming_in_out {
struct dml2_core_instance *instance;
const struct display_configuation_with_meta *display_cfg;
const struct core_display_cfg_support_info *cfg_support_info;
-
/*
* Outputs (also Input the clk freq are also from programming struct)
*/
@@ -443,12 +448,17 @@ struct dml2_core_internal_state_intermediates {
};
struct dml2_core_mode_support_locals {
- struct dml2_core_calcs_mode_support_ex mode_support_ex_params;
+ union {
+ struct dml2_core_calcs_mode_support_ex mode_support_ex_params;
+ };
struct dml2_display_cfg svp_expanded_display_cfg;
+ struct dml2_calculate_mcache_allocation_in_out calc_mcache_allocation_params;
};
struct dml2_core_mode_programming_locals {
- struct dml2_core_calcs_mode_programming_ex mode_programming_ex_params;
+ union {
+ struct dml2_core_calcs_mode_programming_ex mode_programming_ex_params;
+ };
struct dml2_display_cfg svp_expanded_display_cfg;
};
@@ -473,7 +483,6 @@ struct dml2_core_instance {
bool (*mode_programming)(struct dml2_core_mode_programming_in_out *in_out);
bool (*populate_informative)(struct dml2_core_populate_informative_in_out *in_out);
bool (*calculate_mcache_allocation)(struct dml2_calculate_mcache_allocation_in_out *in_out);
- bool (*unit_test)(void);
struct {
struct dml2_core_internal_display_mode_lib mode_lib;
@@ -600,34 +609,11 @@ struct dml2_pmo_optimize_for_stutter_in_out {
struct display_configuation_with_meta *optimized_display_config;
};
-enum dml2_pmo_pstate_method {
- dml2_pmo_pstate_strategy_na = 0,
- /* hw exclusive modes */
- dml2_pmo_pstate_strategy_vactive = 1,
- dml2_pmo_pstate_strategy_vblank = 2,
- dml2_pmo_pstate_strategy_reserved_hw = 5,
- /* fw assisted exclusive modes */
- dml2_pmo_pstate_strategy_fw_svp = 6,
- dml2_pmo_pstate_strategy_reserved_fw = 10,
- /* fw assisted modes requiring drr modulation */
- dml2_pmo_pstate_strategy_fw_vactive_drr = 11,
- dml2_pmo_pstate_strategy_fw_vblank_drr = 12,
- dml2_pmo_pstate_strategy_fw_svp_drr = 13,
- dml2_pmo_pstate_strategy_reserved_fw_drr_clamped = 20,
- dml2_pmo_pstate_strategy_fw_drr = 21,
- dml2_pmo_pstate_strategy_reserved_fw_drr_var = 22,
-};
-
-struct dml2_pmo_pstate_strategy {
- enum dml2_pmo_pstate_method per_stream_pstate_method[DML2_MAX_PLANES];
- bool allow_state_increase;
-};
-
-#define PMO_NO_DRR_STRATEGY_MASK (((1 << (dml2_pmo_pstate_strategy_reserved_fw - dml2_pmo_pstate_strategy_na + 1)) - 1) << dml2_pmo_pstate_strategy_na)
-#define PMO_DRR_STRATEGY_MASK (((1 << (dml2_pmo_pstate_strategy_reserved_fw_drr_var - dml2_pmo_pstate_strategy_fw_vactive_drr + 1)) - 1) << dml2_pmo_pstate_strategy_fw_vactive_drr)
-#define PMO_DRR_CLAMPED_STRATEGY_MASK (((1 << (dml2_pmo_pstate_strategy_reserved_fw_drr_clamped - dml2_pmo_pstate_strategy_fw_vactive_drr + 1)) - 1) << dml2_pmo_pstate_strategy_fw_vactive_drr)
-#define PMO_DRR_VAR_STRATEGY_MASK (((1 << (dml2_pmo_pstate_strategy_reserved_fw_drr_var - dml2_pmo_pstate_strategy_fw_drr + 1)) - 1) << dml2_pmo_pstate_strategy_fw_drr)
-#define PMO_FW_STRATEGY_MASK (((1 << (dml2_pmo_pstate_strategy_reserved_fw_drr_var - dml2_pmo_pstate_strategy_fw_svp + 1)) - 1) << dml2_pmo_pstate_strategy_fw_svp)
+#define PMO_NO_DRR_STRATEGY_MASK (((1 << (dml2_pstate_method_reserved_fw - dml2_pstate_method_na + 1)) - 1) << dml2_pstate_method_na)
+#define PMO_DRR_STRATEGY_MASK (((1 << (dml2_pstate_method_reserved_fw_drr_var - dml2_pstate_method_fw_vactive_drr + 1)) - 1) << dml2_pstate_method_fw_vactive_drr)
+#define PMO_DRR_CLAMPED_STRATEGY_MASK (((1 << (dml2_pstate_method_reserved_fw_drr_clamped - dml2_pstate_method_fw_vactive_drr + 1)) - 1) << dml2_pstate_method_fw_vactive_drr)
+#define PMO_DRR_VAR_STRATEGY_MASK (((1 << (dml2_pstate_method_reserved_fw_drr_var - dml2_pstate_method_fw_drr + 1)) - 1) << dml2_pstate_method_fw_drr)
+#define PMO_FW_STRATEGY_MASK (((1 << (dml2_pstate_method_reserved_fw_drr_var - dml2_pstate_method_fw_svp + 1)) - 1) << dml2_pstate_method_fw_svp)
#define PMO_DCN4_MAX_DISPLAYS 4
#define PMO_DCN4_MAX_NUM_VARIANTS 2
@@ -645,6 +631,8 @@ struct dml2_pmo_scratch {
int stream_mask;
} pmo_dcn3;
struct {
+ struct dml2_pmo_pstate_strategy expanded_override_strategy_list[2 * 2 * 2 * 2];
+ unsigned int num_expanded_override_strategies;
struct dml2_pmo_pstate_strategy pstate_strategy_candidates[DML2_PMO_PSTATE_CANDIDATE_LIST_SIZE];
int num_pstate_candidates;
int cur_pstate_candidate;
@@ -706,7 +694,6 @@ struct dml2_pmo_instance {
int mpc_combine_limit;
int odm_combine_limit;
int mcg_clock_table_size;
-
union {
struct {
struct {
@@ -743,8 +730,6 @@ struct dml2_pmo_instance {
bool (*test_for_stutter)(struct dml2_pmo_test_for_stutter_in_out *in_out);
bool (*optimize_for_stutter)(struct dml2_pmo_optimize_for_stutter_in_out *in_out);
- bool (*unit_test)(void);
-
struct dml2_pmo_init_data init_data;
struct dml2_pmo_scratch scratch;
};
@@ -963,7 +948,12 @@ struct dml2_top_mcache_validate_admissability_locals {
struct dml2_top_display_cfg_support_info {
const struct dml2_display_cfg *display_config;
struct core_display_cfg_support_info core_info;
- enum dml2_pstate_support_method per_plane_pstate_method[DML2_MAX_PLANES];
+};
+
+struct dml2_top_funcs {
+ bool (*check_mode_supported)(struct dml2_check_mode_supported_in_out *in_out);
+ bool (*build_mode_programming)(struct dml2_build_mode_programming_in_out *in_out);
+ bool (*build_mcache_programming)(struct dml2_build_mcache_programming_in_out *in_out);
};
struct dml2_instance {
@@ -978,8 +968,8 @@ struct dml2_instance {
struct dml2_ip_capabilities ip_caps;
struct dml2_mcg_min_clock_table min_clk_table;
-
struct dml2_pmo_options pmo_options;
+ struct dml2_top_funcs funcs;
struct {
struct dml2_initialize_instance_locals initialize_instance_locals;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.c
index 1ed21c1b86a5..5f1b49a50049 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_dc_resource_mgmt.c
@@ -532,26 +532,6 @@ static void calculate_odm_slices(const struct dc_stream_state *stream, unsigned
odm_slice_end_x[odm_factor - 1] = stream->src.width - 1;
}
-static bool is_plane_in_odm_slice(const struct dc_plane_state *plane, unsigned int slice_index, unsigned int *odm_slice_end_x, unsigned int num_slices)
-{
- unsigned int slice_start_x, slice_end_x;
-
- if (slice_index == 0)
- slice_start_x = 0;
- else
- slice_start_x = odm_slice_end_x[slice_index - 1] + 1;
-
- slice_end_x = odm_slice_end_x[slice_index];
-
- if (plane->clip_rect.x + plane->clip_rect.width < slice_start_x)
- return false;
-
- if (plane->clip_rect.x > slice_end_x)
- return false;
-
- return true;
-}
-
static void add_odm_slice_to_odm_tree(struct dml2_context *ctx,
struct dc_state *state,
struct dc_pipe_mapping_scratch *scratch,
@@ -791,12 +771,6 @@ static void map_pipes_for_plane(struct dml2_context *ctx, struct dc_state *state
sort_pipes_for_splitting(&scratch->pipe_pool);
for (odm_slice_index = 0; odm_slice_index < scratch->odm_info.odm_factor; odm_slice_index++) {
- // We build the tree for one ODM slice at a time.
- // Each ODM slice shares a common OPP
- if (!is_plane_in_odm_slice(plane, odm_slice_index, scratch->odm_info.odm_slice_end_x, scratch->odm_info.odm_factor)) {
- continue;
- }
-
// Now we have a list of all pipes to be used for this plane/stream, now setup the tree.
scratch->odm_info.next_higher_pipe_for_odm_slice[odm_slice_index] = add_plane_to_blend_tree(ctx, state,
plane,
@@ -1108,22 +1082,22 @@ bool dml2_map_dc_pipes(struct dml2_context *ctx, struct dc_state *state, const s
if (stream_disp_cfg_index >= disp_cfg_index_max)
continue;
- if (ODMMode[stream_disp_cfg_index] == dml_odm_mode_bypass) {
- scratch.odm_info.odm_factor = 1;
- } else if (ODMMode[stream_disp_cfg_index] == dml_odm_mode_combine_2to1) {
- scratch.odm_info.odm_factor = 2;
- } else if (ODMMode[stream_disp_cfg_index] == dml_odm_mode_combine_4to1) {
- scratch.odm_info.odm_factor = 4;
- } else {
- ASSERT(false);
- scratch.odm_info.odm_factor = 1;
- }
-
+ if (ctx->architecture == dml2_architecture_20) {
+ if (ODMMode[stream_disp_cfg_index] == dml_odm_mode_bypass) {
+ scratch.odm_info.odm_factor = 1;
+ } else if (ODMMode[stream_disp_cfg_index] == dml_odm_mode_combine_2to1) {
+ scratch.odm_info.odm_factor = 2;
+ } else if (ODMMode[stream_disp_cfg_index] == dml_odm_mode_combine_4to1) {
+ scratch.odm_info.odm_factor = 4;
+ } else {
+ ASSERT(false);
+ scratch.odm_info.odm_factor = 1;
+ }
+ } else if (ctx->architecture == dml2_architecture_21) {
/* After DML2.1 update, ODM interpretation needs to change and is no longer same as for DML2.0.
* This is not an issue with new resource management logic. This block ensure backcompat
* with legacy pipe management with updated DML.
* */
- if (ctx->architecture == dml2_architecture_21) {
if (ODMMode[stream_disp_cfg_index] == 1) {
scratch.odm_info.odm_factor = 1;
} else if (ODMMode[stream_disp_cfg_index] == 2) {
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_mall_phantom.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_mall_phantom.c
index 3d29169dd6bb..a56e75cdf712 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_mall_phantom.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_mall_phantom.c
@@ -813,7 +813,7 @@ static bool remove_all_phantom_planes_for_stream(struct dml2_context *ctx, struc
{
int i, old_plane_count;
struct dc_stream_status *stream_status = NULL;
- struct dc_plane_state *del_planes[MAX_SURFACE_NUM] = { 0 };
+ struct dc_plane_state *del_planes[MAX_SURFACES] = { 0 };
for (i = 0; i < context->stream_count; i++)
if (context->streams[i] == stream) {
@@ -868,7 +868,7 @@ bool dml2_svp_remove_all_phantom_pipes(struct dml2_context *ctx, struct dc_state
/* Conditions for setting up phantom pipes for SubVP:
* 1. Not force disable SubVP
- * 2. Full update (i.e. !fast_validate)
+ * 2. Full update (i.e. DC_VALIDATE_MODE_AND_PROGRAMMING)
* 3. Enough pipes are available to support SubVP (TODO: Which pipes will use VACTIVE / VBLANK / SUBVP?)
* 4. Display configuration passes validation
* 5. (Config doesn't support MCLK in VACTIVE/VBLANK || dc->debug.force_subvp_mclk_switch)
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_policy.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_policy.c
index c4c52173ef22..ef693f608d59 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_policy.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_policy.c
@@ -301,6 +301,7 @@ void build_unoptimized_policy_settings(enum dml_project_id project, struct dml_m
policy->AssumeModeSupportAtMaxPwrStateEvenDRAMClockChangeNotSupported = true; // TOREVIEW: What does this mean?
policy->AssumeModeSupportAtMaxPwrStateEvenFClockChangeNotSupported = true; // TOREVIEW: What does this mean?
if (project == dml_project_dcn35 ||
+ project == dml_project_dcn36 ||
project == dml_project_dcn351) {
policy->DCCProgrammingAssumesScanDirectionUnknownFinal = false;
policy->EnhancedPrefetchScheduleAccelerationFinal = 0;
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c
index bde4250853b1..3b866e876bf4 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_translation_helper.c
@@ -107,6 +107,7 @@ void dml2_init_ip_params(struct dml2_context *dml2, const struct dc *in_dc, stru
case dml_project_dcn35:
case dml_project_dcn351:
+ case dml_project_dcn36:
out->rob_buffer_size_kbytes = 64;
out->config_return_buffer_size_in_kbytes = 1792;
out->compressed_buffer_segment_size_in_kbytes = 64;
@@ -292,6 +293,7 @@ void dml2_init_socbb_params(struct dml2_context *dml2, const struct dc *in_dc, s
case dml_project_dcn35:
case dml_project_dcn351:
+ case dml_project_dcn36:
out->num_chans = 4;
out->round_trip_ping_latency_dcfclk_cycles = 106;
out->smn_latency_us = 2;
@@ -506,6 +508,7 @@ void dml2_init_soc_states(struct dml2_context *dml2, const struct dc *in_dc,
p->dcfclk_stas_mhz[3] = 1324;
p->dcfclk_stas_mhz[4] = p->in_states->state_array[1].dcfclk_mhz;
} else if (dml2->v20.dml_core_ctx.project != dml_project_dcn35 &&
+ dml2->v20.dml_core_ctx.project != dml_project_dcn36 &&
dml2->v20.dml_core_ctx.project != dml_project_dcn351) {
p->dcfclk_stas_mhz[0] = 300;
p->dcfclk_stas_mhz[1] = 615;
@@ -553,13 +556,54 @@ void dml2_init_soc_states(struct dml2_context *dml2, const struct dc *in_dc,
}
}
- dml2_policy_build_synthetic_soc_states(s, p);
- if (dml2->v20.dml_core_ctx.project == dml_project_dcn35) {
- // Override last out_state with data from last in_state
- // This will ensure that out_state contains max fclk
- memcpy(&p->out_states->state_array[p->out_states->num_states - 1],
- &p->in_states->state_array[p->in_states->num_states - 1],
- sizeof(struct soc_state_bounding_box_st));
+ if (dml2->v20.dml_core_ctx.project == dml_project_dcn35 ||
+ dml2->v20.dml_core_ctx.project == dml_project_dcn36 ||
+ dml2->v20.dml_core_ctx.project == dml_project_dcn351) {
+ int max_dcfclk_mhz = 0, max_dispclk_mhz = 0, max_dppclk_mhz = 0, max_phyclk_mhz = 0,
+ max_dtbclk_mhz = 0, max_fclk_mhz = 0, max_uclk_mhz = 0, max_socclk_mhz = 0;
+
+ for (i = 0; i < p->in_states->num_states; i++) {
+ if (p->in_states->state_array[i].dcfclk_mhz > max_dcfclk_mhz)
+ max_dcfclk_mhz = (int)p->in_states->state_array[i].dcfclk_mhz;
+ if (p->in_states->state_array[i].fabricclk_mhz > max_fclk_mhz)
+ max_fclk_mhz = (int)p->in_states->state_array[i].fabricclk_mhz;
+ if (p->in_states->state_array[i].socclk_mhz > max_socclk_mhz)
+ max_socclk_mhz = (int)p->in_states->state_array[i].socclk_mhz;
+ if (p->in_states->state_array[i].dram_speed_mts > max_uclk_mhz)
+ max_uclk_mhz = (int)p->in_states->state_array[i].dram_speed_mts;
+ if (p->in_states->state_array[i].dispclk_mhz > max_dispclk_mhz)
+ max_dispclk_mhz = (int)p->in_states->state_array[i].dispclk_mhz;
+ if (p->in_states->state_array[i].dppclk_mhz > max_dppclk_mhz)
+ max_dppclk_mhz = (int)p->in_states->state_array[i].dppclk_mhz;
+ if (p->in_states->state_array[i].phyclk_mhz > max_phyclk_mhz)
+ max_phyclk_mhz = (int)p->in_states->state_array[i].phyclk_mhz;
+ if (p->in_states->state_array[i].dtbclk_mhz > max_dtbclk_mhz)
+ max_dtbclk_mhz = (int)p->in_states->state_array[i].dtbclk_mhz;
+ }
+
+ for (i = 0; i < p->in_states->num_states; i++) {
+ /* Independent states - including base (unlisted) parameters from state 0. */
+ p->out_states->state_array[i] = p->in_states->state_array[0];
+
+ p->out_states->state_array[i].dispclk_mhz = max_dispclk_mhz;
+ p->out_states->state_array[i].dppclk_mhz = max_dppclk_mhz;
+ p->out_states->state_array[i].dtbclk_mhz = max_dtbclk_mhz;
+ p->out_states->state_array[i].phyclk_mhz = max_phyclk_mhz;
+
+ p->out_states->state_array[i].dscclk_mhz = max_dispclk_mhz / 3.0;
+ p->out_states->state_array[i].phyclk_mhz = max_phyclk_mhz;
+ p->out_states->state_array[i].dtbclk_mhz = max_dtbclk_mhz;
+
+ /* Dependent states. */
+ p->out_states->state_array[i].dram_speed_mts = p->in_states->state_array[i].dram_speed_mts;
+ p->out_states->state_array[i].fabricclk_mhz = p->in_states->state_array[i].fabricclk_mhz;
+ p->out_states->state_array[i].socclk_mhz = p->in_states->state_array[i].socclk_mhz;
+ p->out_states->state_array[i].dcfclk_mhz = p->in_states->state_array[i].dcfclk_mhz;
+ }
+
+ p->out_states->num_states = p->in_states->num_states;
+ } else {
+ dml2_policy_build_synthetic_soc_states(s, p);
}
}
@@ -746,7 +790,7 @@ static void populate_dml_output_cfg_from_stream_state(struct dml_output_cfg_st *
case SIGNAL_TYPE_DISPLAY_PORT_MST:
case SIGNAL_TYPE_DISPLAY_PORT:
out->OutputEncoder[location] = dml_dp;
- if (dml2->v20.scratch.hpo_stream_to_link_encoder_mapping[location] != -1)
+ if (location < MAX_HPO_DP2_ENCODERS && dml2->v20.scratch.hpo_stream_to_link_encoder_mapping[location] != -1)
out->OutputEncoder[dml2->v20.scratch.hpo_stream_to_link_encoder_mapping[location]] = dml_dp2p0;
break;
case SIGNAL_TYPE_EDP:
@@ -852,7 +896,7 @@ static void populate_dummy_dml_surface_cfg(struct dml_surface_cfg_st *out, unsig
out->SurfaceWidthC[location] = in->timing.h_addressable;
out->SurfaceHeightC[location] = in->timing.v_addressable;
out->PitchY[location] = ((out->SurfaceWidthY[location] + 127) / 128) * 128;
- out->PitchC[location] = 0;
+ out->PitchC[location] = 1;
out->DCCEnable[location] = false;
out->DCCMetaPitchY[location] = 0;
out->DCCMetaPitchC[location] = 0;
@@ -909,6 +953,7 @@ static void populate_dml_surface_cfg_from_plane_state(enum dml_project_id dml2_p
out->SourcePixelFormat[location] = dml_420_10;
break;
case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616:
+ case SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616:
case SURFACE_PIXEL_FORMAT_GRPH_ARGB16161616F:
case SURFACE_PIXEL_FORMAT_GRPH_ABGR16161616F:
out->SourcePixelFormat[location] = dml_444_64;
@@ -929,7 +974,9 @@ static void populate_dml_surface_cfg_from_plane_state(enum dml_project_id dml2_p
}
}
-static void get_scaler_data_for_plane(const struct dc_plane_state *in, struct dc_state *context, struct scaler_data *out)
+static struct scaler_data *get_scaler_data_for_plane(
+ const struct dc_plane_state *in,
+ struct dc_state *context)
{
int i;
struct pipe_ctx *temp_pipe = &context->res_ctx.temp_pipe;
@@ -950,7 +997,7 @@ static void get_scaler_data_for_plane(const struct dc_plane_state *in, struct dc
}
ASSERT(i < MAX_PIPES);
- memcpy(out, &temp_pipe->plane_res.scl_data, sizeof(*out));
+ return &temp_pipe->plane_res.scl_data;
}
static void populate_dummy_dml_plane_cfg(struct dml_plane_cfg_st *out, unsigned int location,
@@ -1013,11 +1060,7 @@ static void populate_dml_plane_cfg_from_plane_state(struct dml_plane_cfg_st *out
const struct dc_plane_state *in, struct dc_state *context,
const struct soc_bounding_box_st *soc)
{
- struct scaler_data *scaler_data = kzalloc(sizeof(*scaler_data), GFP_KERNEL);
- if (!scaler_data)
- return;
-
- get_scaler_data_for_plane(in, context, scaler_data);
+ struct scaler_data *scaler_data = get_scaler_data_for_plane(in, context);
out->CursorBPP[location] = dml_cur_32bit;
out->CursorWidth[location] = 256;
@@ -1082,8 +1125,6 @@ static void populate_dml_plane_cfg_from_plane_state(struct dml_plane_cfg_st *out
out->DynamicMetadataTransmittedBytes[location] = 0;
out->NumberOfCursors[location] = 1;
-
- kfree(scaler_data);
}
static unsigned int map_stream_to_dml_display_cfg(const struct dml2_context *dml2,
@@ -1148,22 +1189,6 @@ static unsigned int map_plane_to_dml_display_cfg(const struct dml2_context *dml2
return location;
}
-static void apply_legacy_svp_drr_settings(struct dml2_context *dml2, const struct dc_state *state, struct dml_display_cfg_st *dml_dispcfg)
-{
- int i;
-
- if (state->bw_ctx.bw.dcn.clk.fw_based_mclk_switching) {
- ASSERT(state->stream_count == 1);
- dml_dispcfg->timing.DRRDisplay[0] = true;
- } else if (state->bw_ctx.bw.dcn.legacy_svp_drr_stream_index_valid) {
-
- for (i = 0; i < dml_dispcfg->num_timings; i++) {
- if (dml2->v20.scratch.dml_to_dc_pipe_mapping.disp_cfg_to_stream_id[i] == state->streams[state->bw_ctx.bw.dcn.legacy_svp_drr_stream_index]->stream_id)
- dml_dispcfg->timing.DRRDisplay[i] = true;
- }
- }
-}
-
static void dml2_populate_pipe_to_plane_index_mapping(struct dml2_context *dml2, struct dc_state *state)
{
unsigned int i;
@@ -1303,7 +1328,7 @@ void map_dc_state_into_dml_display_cfg(struct dml2_context *dml2, struct dc_stat
if (disp_cfg_stream_location < 0)
disp_cfg_stream_location = dml_dispcfg->num_timings++;
- ASSERT(disp_cfg_stream_location >= 0 && disp_cfg_stream_location <= __DML2_WRAPPER_MAX_STREAMS_PLANES__);
+ ASSERT(disp_cfg_stream_location >= 0 && disp_cfg_stream_location < __DML2_WRAPPER_MAX_STREAMS_PLANES__);
populate_dml_timing_cfg_from_stream_state(&dml_dispcfg->timing, disp_cfg_stream_location, context->streams[i]);
populate_dml_output_cfg_from_stream_state(&dml_dispcfg->output, disp_cfg_stream_location, context->streams[i], current_pipe_context, dml2);
@@ -1343,7 +1368,7 @@ void map_dc_state_into_dml_display_cfg(struct dml2_context *dml2, struct dc_stat
if (disp_cfg_plane_location < 0)
disp_cfg_plane_location = dml_dispcfg->num_surfaces++;
- ASSERT(disp_cfg_plane_location >= 0 && disp_cfg_plane_location <= __DML2_WRAPPER_MAX_STREAMS_PLANES__);
+ ASSERT(disp_cfg_plane_location >= 0 && disp_cfg_plane_location < __DML2_WRAPPER_MAX_STREAMS_PLANES__);
populate_dml_surface_cfg_from_plane_state(dml2->v20.dml_core_ctx.project, &dml_dispcfg->surface, disp_cfg_plane_location, context->stream_status[i].plane_states[j]);
populate_dml_plane_cfg_from_plane_state(
@@ -1396,9 +1421,6 @@ void map_dc_state_into_dml_display_cfg(struct dml2_context *dml2, struct dc_stat
}
}
}
-
- if (!dml2->config.use_native_pstate_optimization)
- apply_legacy_svp_drr_settings(dml2, context, dml_dispcfg);
}
void dml2_update_pipe_ctx_dchub_regs(struct _vcs_dpi_dml_display_rq_regs_st *rq_regs,
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c
index 9190c1328d5b..0318260370ed 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.c
@@ -33,7 +33,6 @@
#include "dml2_dc_resource_mgmt.h"
#include "dml21_wrapper.h"
-
static void initialize_dml2_ip_params(struct dml2_context *dml2, const struct dc *in_dc, struct ip_params_st *out)
{
if (dml2->config.use_native_soc_bb_construction)
@@ -72,6 +71,7 @@ static void map_hw_resources(struct dml2_context *dml2,
in_out_display_cfg->hw.NumberOfDSCSlices[i] = mode_support_info->NumberOfDSCSlices[i];
in_out_display_cfg->hw.DLGRefClkFreqMHz = 24;
if (dml2->v20.dml_core_ctx.project != dml_project_dcn35 &&
+ dml2->v20.dml_core_ctx.project != dml_project_dcn36 &&
dml2->v20.dml_core_ctx.project != dml_project_dcn351) {
/*dGPU default as 50Mhz*/
in_out_display_cfg->hw.DLGRefClkFreqMHz = 50;
@@ -93,12 +93,17 @@ static void map_hw_resources(struct dml2_context *dml2,
static unsigned int pack_and_call_dml_mode_support_ex(struct dml2_context *dml2,
const struct dml_display_cfg_st *display_cfg,
- struct dml_mode_support_info_st *evaluation_info)
+ struct dml_mode_support_info_st *evaluation_info,
+ enum dc_validate_mode validate_mode)
{
struct dml2_wrapper_scratch *s = &dml2->v20.scratch;
s->mode_support_params.mode_lib = &dml2->v20.dml_core_ctx;
s->mode_support_params.in_display_cfg = display_cfg;
+ if (validate_mode == DC_VALIDATE_MODE_ONLY)
+ s->mode_support_params.in_start_state_idx = dml2->v20.dml_core_ctx.states.num_states - 1;
+ else
+ s->mode_support_params.in_start_state_idx = 0;
s->mode_support_params.out_evaluation_info = evaluation_info;
memset(evaluation_info, 0, sizeof(struct dml_mode_support_info_st));
@@ -110,10 +115,8 @@ static unsigned int pack_and_call_dml_mode_support_ex(struct dml2_context *dml2,
static bool optimize_configuration(struct dml2_context *dml2, struct dml2_wrapper_optimize_configuration_params *p)
{
int unused_dpps = p->ip_params->max_num_dpp;
- int i, j;
- int odms_needed, refresh_rate_hz, dpps_needed, subvp_height, pstate_width_fw_delay_lines, surface_count;
- int subvp_timing_to_add, new_timing_index, subvp_surface_to_add, new_surface_index;
- float frame_time_sec, max_frame_time_sec;
+ int i;
+ int odms_needed;
int largest_blend_and_timing = 0;
bool optimization_done = false;
@@ -128,79 +131,6 @@ static bool optimize_configuration(struct dml2_context *dml2, struct dml2_wrappe
if (p->new_display_config != p->cur_display_config)
*p->new_display_config = *p->cur_display_config;
- // Optimize P-State Support
- if (dml2->config.use_native_pstate_optimization) {
- if (p->cur_mode_support_info->DRAMClockChangeSupport[0] == dml_dram_clock_change_unsupported) {
- // Find a display with < 120Hz refresh rate with maximal refresh rate that's not already subvp
- subvp_timing_to_add = -1;
- subvp_surface_to_add = -1;
- max_frame_time_sec = 0;
- surface_count = 0;
- for (i = 0; i < (int) p->cur_display_config->num_timings; i++) {
- refresh_rate_hz = (int)div_u64((unsigned long long) p->cur_display_config->timing.PixelClock[i] * 1000 * 1000,
- (p->cur_display_config->timing.HTotal[i] * p->cur_display_config->timing.VTotal[i]));
- if (refresh_rate_hz < 120) {
- // Check its upstream surfaces to see if this one could be converted to subvp.
- dpps_needed = 0;
- for (j = 0; j < (int) p->cur_display_config->num_surfaces; j++) {
- if (p->cur_display_config->plane.BlendingAndTiming[j] == i &&
- p->cur_display_config->plane.UseMALLForPStateChange[j] == dml_use_mall_pstate_change_disable) {
- dpps_needed += p->cur_mode_support_info->DPPPerSurface[j];
- subvp_surface_to_add = j;
- surface_count++;
- }
- }
-
- if (surface_count == 1 && dpps_needed > 0 && dpps_needed <= unused_dpps) {
- frame_time_sec = (float)1 / refresh_rate_hz;
- if (frame_time_sec > max_frame_time_sec) {
- max_frame_time_sec = frame_time_sec;
- subvp_timing_to_add = i;
- }
- }
- }
- }
- if (subvp_timing_to_add >= 0) {
- new_timing_index = p->new_display_config->num_timings++;
- new_surface_index = p->new_display_config->num_surfaces++;
- // Add a phantom pipe reflecting the main pipe's timing
- dml2_util_copy_dml_timing(&p->new_display_config->timing, new_timing_index, subvp_timing_to_add);
-
- pstate_width_fw_delay_lines = (int)(((double)(p->config->svp_pstate.subvp_fw_processing_delay_us +
- p->config->svp_pstate.subvp_pstate_allow_width_us) / 1000000) *
- (p->new_display_config->timing.PixelClock[subvp_timing_to_add] * 1000 * 1000) /
- (double)p->new_display_config->timing.HTotal[subvp_timing_to_add]);
-
- subvp_height = p->cur_mode_support_info->SubViewportLinesNeededInMALL[subvp_timing_to_add] + pstate_width_fw_delay_lines;
-
- p->new_display_config->timing.VActive[new_timing_index] = subvp_height;
- p->new_display_config->timing.VTotal[new_timing_index] = subvp_height +
- p->new_display_config->timing.VTotal[subvp_timing_to_add] - p->new_display_config->timing.VActive[subvp_timing_to_add];
-
- p->new_display_config->output.OutputDisabled[new_timing_index] = true;
-
- p->new_display_config->plane.UseMALLForPStateChange[subvp_surface_to_add] = dml_use_mall_pstate_change_sub_viewport;
-
- dml2_util_copy_dml_plane(&p->new_display_config->plane, new_surface_index, subvp_surface_to_add);
- dml2_util_copy_dml_surface(&p->new_display_config->surface, new_surface_index, subvp_surface_to_add);
-
- p->new_display_config->plane.ViewportHeight[new_surface_index] = subvp_height;
- p->new_display_config->plane.ViewportHeightChroma[new_surface_index] = subvp_height;
- p->new_display_config->plane.ViewportStationary[new_surface_index] = false;
-
- p->new_display_config->plane.UseMALLForStaticScreen[new_surface_index] = dml_use_mall_static_screen_disable;
- p->new_display_config->plane.UseMALLForPStateChange[new_surface_index] = dml_use_mall_pstate_change_phantom_pipe;
-
- p->new_display_config->plane.NumberOfCursors[new_surface_index] = 0;
-
- p->new_policy->ImmediateFlipRequirement[new_surface_index] = dml_immediate_flip_not_required;
-
- p->new_display_config->plane.BlendingAndTiming[new_surface_index] = new_timing_index;
-
- optimization_done = true;
- }
- }
- }
// Optimize Clocks
if (!optimization_done) {
@@ -224,7 +154,8 @@ static bool optimize_configuration(struct dml2_context *dml2, struct dml2_wrappe
return optimization_done;
}
-static int calculate_lowest_supported_state_for_temp_read(struct dml2_context *dml2, struct dc_state *display_state)
+static int calculate_lowest_supported_state_for_temp_read(struct dml2_context *dml2, struct dc_state *display_state,
+ enum dc_validate_mode validate_mode)
{
struct dml2_calculate_lowest_supported_state_for_temp_read_scratch *s = &dml2->v20.scratch.dml2_calculate_lowest_supported_state_for_temp_read_scratch;
struct dml2_wrapper_scratch *s_global = &dml2->v20.scratch;
@@ -266,7 +197,8 @@ static int calculate_lowest_supported_state_for_temp_read(struct dml2_context *d
dml2->v20.dml_core_ctx.states.state_array[j].dram_clock_change_latency_us = s_global->dummy_pstate_table[i].dummy_pstate_latency_us;
}
- dml_result = pack_and_call_dml_mode_support_ex(dml2, &s->cur_display_config, &s->evaluation_info);
+ dml_result = pack_and_call_dml_mode_support_ex(dml2, &s->cur_display_config, &s->evaluation_info,
+ validate_mode);
if (dml_result && s->evaluation_info.DRAMClockChangeSupport[0] == dml_dram_clock_change_vactive) {
map_hw_resources(dml2, &s->cur_display_config, &s->evaluation_info);
@@ -331,7 +263,8 @@ static bool does_configuration_meet_sw_policies(struct dml2_context *ctx, const
}
static bool dml_mode_support_wrapper(struct dml2_context *dml2,
- struct dc_state *display_state)
+ struct dc_state *display_state,
+ enum dc_validate_mode validate_mode)
{
struct dml2_wrapper_scratch *s = &dml2->v20.scratch;
unsigned int result = 0, i;
@@ -367,7 +300,8 @@ static bool dml_mode_support_wrapper(struct dml2_context *dml2,
result = pack_and_call_dml_mode_support_ex(dml2,
&s->cur_display_config,
- &s->mode_support_info);
+ &s->mode_support_info,
+ validate_mode);
if (result)
result = does_configuration_meet_sw_policies(dml2, &s->cur_display_config, &s->mode_support_info);
@@ -388,7 +322,8 @@ static bool dml_mode_support_wrapper(struct dml2_context *dml2,
dml2->v20.dml_core_ctx.policy = s->new_policy;
optimized_result = pack_and_call_dml_mode_support_ex(dml2,
&s->new_display_config,
- &s->mode_support_info);
+ &s->mode_support_info,
+ validate_mode);
if (optimized_result)
optimized_result = does_configuration_meet_sw_policies(dml2, &s->new_display_config, &s->mode_support_info);
@@ -407,7 +342,8 @@ static bool dml_mode_support_wrapper(struct dml2_context *dml2,
if (!optimized_result) {
result = pack_and_call_dml_mode_support_ex(dml2,
&s->cur_display_config,
- &s->mode_support_info);
+ &s->mode_support_info,
+ validate_mode);
}
}
@@ -417,150 +353,50 @@ static bool dml_mode_support_wrapper(struct dml2_context *dml2,
return result;
}
-static int find_drr_eligible_stream(struct dc_state *display_state)
-{
- int i;
-
- for (i = 0; i < display_state->stream_count; i++) {
- if (dc_state_get_stream_subvp_type(display_state, display_state->streams[i]) == SUBVP_NONE
- && display_state->streams[i]->ignore_msa_timing_param) {
- // Use ignore_msa_timing_param flag to identify as DRR
- return i;
- }
- }
-
- return -1;
-}
-
-static bool optimize_pstate_with_svp_and_drr(struct dml2_context *dml2, struct dc_state *display_state)
+static bool call_dml_mode_support_and_programming(struct dc_state *context, enum dc_validate_mode validate_mode)
{
- struct dml2_wrapper_scratch *s = &dml2->v20.scratch;
- bool pstate_optimization_done = false;
- bool pstate_optimization_success = false;
- bool result = false;
- int drr_display_index = 0, non_svp_streams = 0;
- bool force_svp = dml2->config.svp_pstate.force_enable_subvp;
-
- display_state->bw_ctx.bw.dcn.clk.fw_based_mclk_switching = false;
- display_state->bw_ctx.bw.dcn.legacy_svp_drr_stream_index_valid = false;
-
- result = dml_mode_support_wrapper(dml2, display_state);
-
- if (!result) {
- pstate_optimization_done = true;
- } else if (s->mode_support_info.DRAMClockChangeSupport[0] != dml_dram_clock_change_unsupported && !force_svp) {
- pstate_optimization_success = true;
- pstate_optimization_done = true;
- }
-
- if (display_state->stream_count == 1 && dml2->config.callbacks.can_support_mclk_switch_using_fw_based_vblank_stretch(dml2->config.callbacks.dc, display_state)) {
- display_state->bw_ctx.bw.dcn.clk.fw_based_mclk_switching = true;
-
- result = dml_mode_support_wrapper(dml2, display_state);
- } else {
- non_svp_streams = display_state->stream_count;
-
- while (!pstate_optimization_done) {
- result = dml_mode_programming(&dml2->v20.dml_core_ctx, s->mode_support_params.out_lowest_state_idx, &s->cur_display_config, true);
-
- // Always try adding SVP first
- if (result)
- result = dml2_svp_add_phantom_pipe_to_dc_state(dml2, display_state, &s->mode_support_info);
- else
- pstate_optimization_done = true;
-
-
- if (result) {
- result = dml_mode_support_wrapper(dml2, display_state);
- } else {
- pstate_optimization_done = true;
- }
-
- if (result) {
- non_svp_streams--;
-
- if (s->mode_support_info.DRAMClockChangeSupport[0] != dml_dram_clock_change_unsupported) {
- if (dml2_svp_validate_static_schedulability(dml2, display_state, s->mode_support_info.DRAMClockChangeSupport[0])) {
- pstate_optimization_success = true;
- pstate_optimization_done = true;
- } else {
- pstate_optimization_success = false;
- pstate_optimization_done = false;
- }
- } else {
- drr_display_index = find_drr_eligible_stream(display_state);
-
- // If there is only 1 remaining non SubVP pipe that is DRR, check static
- // schedulability for SubVP + DRR.
- if (non_svp_streams == 1 && drr_display_index >= 0) {
- if (dml2_svp_drr_schedulable(dml2, display_state, &display_state->streams[drr_display_index]->timing)) {
- display_state->bw_ctx.bw.dcn.legacy_svp_drr_stream_index_valid = true;
- display_state->bw_ctx.bw.dcn.legacy_svp_drr_stream_index = drr_display_index;
- result = dml_mode_support_wrapper(dml2, display_state);
- }
-
- if (result && s->mode_support_info.DRAMClockChangeSupport[0] != dml_dram_clock_change_unsupported) {
- pstate_optimization_success = true;
- pstate_optimization_done = true;
- } else {
- pstate_optimization_success = false;
- pstate_optimization_done = false;
- }
- }
-
- if (pstate_optimization_success) {
- pstate_optimization_done = true;
- } else {
- pstate_optimization_done = false;
- }
- }
- }
- }
- }
+ unsigned int result = 0;
+ unsigned int min_state = 0;
+ int min_state_for_g6_temp_read = 0;
- if (!pstate_optimization_success) {
- dml2_svp_remove_all_phantom_pipes(dml2, display_state);
- display_state->bw_ctx.bw.dcn.clk.fw_based_mclk_switching = false;
- display_state->bw_ctx.bw.dcn.legacy_svp_drr_stream_index_valid = false;
- result = dml_mode_support_wrapper(dml2, display_state);
- }
- return result;
-}
+ if (!context)
+ return false;
-static bool call_dml_mode_support_and_programming(struct dc_state *context)
-{
- unsigned int result = 0;
- unsigned int min_state;
- int min_state_for_g6_temp_read = 0;
struct dml2_context *dml2 = context->bw_ctx.dml2;
struct dml2_wrapper_scratch *s = &dml2->v20.scratch;
- min_state_for_g6_temp_read = calculate_lowest_supported_state_for_temp_read(dml2, context);
+ if (!context->streams[0]->sink->link->dc->caps.is_apu) {
+ min_state_for_g6_temp_read = calculate_lowest_supported_state_for_temp_read(dml2, context,
+ validate_mode);
- ASSERT(min_state_for_g6_temp_read >= 0);
-
- if (!dml2->config.use_native_pstate_optimization) {
- result = optimize_pstate_with_svp_and_drr(dml2, context);
- } else {
- result = dml_mode_support_wrapper(dml2, context);
+ ASSERT(min_state_for_g6_temp_read >= 0);
}
+ result = dml_mode_support_wrapper(dml2, context, validate_mode);
+
/* Upon trying to sett certain frequencies in FRL, min_state_for_g6_temp_read is reported as -1. This leads to an invalid value of min_state causing crashes later on.
* Use the default logic for min_state only when min_state_for_g6_temp_read is a valid value. In other cases, use the value calculated by the DML directly.
*/
- if (min_state_for_g6_temp_read >= 0)
- min_state = min_state_for_g6_temp_read > s->mode_support_params.out_lowest_state_idx ? min_state_for_g6_temp_read : s->mode_support_params.out_lowest_state_idx;
- else
- min_state = s->mode_support_params.out_lowest_state_idx;
-
- if (result)
- result = dml_mode_programming(&dml2->v20.dml_core_ctx, min_state, &s->cur_display_config, true);
+ if (!context->streams[0]->sink->link->dc->caps.is_apu) {
+ if (min_state_for_g6_temp_read >= 0)
+ min_state = min_state_for_g6_temp_read > s->mode_support_params.out_lowest_state_idx ? min_state_for_g6_temp_read : s->mode_support_params.out_lowest_state_idx;
+ else
+ min_state = s->mode_support_params.out_lowest_state_idx;
+ }
+ if (result) {
+ if (!context->streams[0]->sink->link->dc->caps.is_apu) {
+ result = dml_mode_programming(&dml2->v20.dml_core_ctx, min_state, &s->cur_display_config, true);
+ } else {
+ result = dml_mode_programming(&dml2->v20.dml_core_ctx, s->mode_support_params.out_lowest_state_idx, &s->cur_display_config, true);
+ }
+ }
return result;
}
-static bool dml2_validate_and_build_resource(const struct dc *in_dc, struct dc_state *context)
+static bool dml2_validate_and_build_resource(const struct dc *in_dc, struct dc_state *context,
+ enum dc_validate_mode validate_mode)
{
struct dml2_context *dml2 = context->bw_ctx.dml2;
struct dml2_wrapper_scratch *s = &dml2->v20.scratch;
@@ -596,7 +432,7 @@ static bool dml2_validate_and_build_resource(const struct dc *in_dc, struct dc_s
copy_dummy_pstate_table(s->dummy_pstate_table, in_dc->clk_mgr->bw_params->dummy_pstate_table, 4);
- result = call_dml_mode_support_and_programming(context);
+ result = call_dml_mode_support_and_programming(context, validate_mode);
/* Call map dc pipes to map the pipes based on the DML output. For correctly determining if recalculation
* is required or not, the resource context needs to correctly reflect the number of active pipes. We would
* only know the correct number if active pipes after dml2_map_dc_pipes is called.
@@ -613,7 +449,7 @@ static bool dml2_validate_and_build_resource(const struct dc *in_dc, struct dc_s
need_recalculation = dml2_verify_det_buffer_configuration(dml2, context, &dml2->det_helper_scratch);
if (need_recalculation) {
/* Engage the DML again if recalculation is required. */
- call_dml_mode_support_and_programming(context);
+ call_dml_mode_support_and_programming(context, validate_mode);
if (!dml2->config.skip_hw_state_mapping) {
dml2_map_dc_pipes(dml2, context, &s->cur_display_config, &s->dml_to_dc_pipe_mapping, in_dc->current_state);
}
@@ -648,7 +484,10 @@ static bool dml2_validate_and_build_resource(const struct dc *in_dc, struct dc_s
dml2_copy_clocks_to_dc_state(&out_clks, context);
dml2_extract_watermark_set(&context->bw_ctx.bw.dcn.watermarks.a, &dml2->v20.dml_core_ctx);
dml2_extract_watermark_set(&context->bw_ctx.bw.dcn.watermarks.b, &dml2->v20.dml_core_ctx);
- memcpy(&context->bw_ctx.bw.dcn.watermarks.c, &dml2->v20.g6_temp_read_watermark_set, sizeof(context->bw_ctx.bw.dcn.watermarks.c));
+ if (context->streams[0]->sink->link->dc->caps.is_apu)
+ dml2_extract_watermark_set(&context->bw_ctx.bw.dcn.watermarks.c, &dml2->v20.dml_core_ctx);
+ else
+ memcpy(&context->bw_ctx.bw.dcn.watermarks.c, &dml2->v20.g6_temp_read_watermark_set, sizeof(context->bw_ctx.bw.dcn.watermarks.c));
dml2_extract_watermark_set(&context->bw_ctx.bw.dcn.watermarks.d, &dml2->v20.dml_core_ctx);
dml2_extract_writeback_wm(context, &dml2->v20.dml_core_ctx);
//copy for deciding zstate use
@@ -666,7 +505,7 @@ static bool dml2_validate_and_build_resource(const struct dc *in_dc, struct dc_s
return result;
}
-static bool dml2_validate_only(struct dc_state *context)
+static bool dml2_validate_only(struct dc_state *context, enum dc_validate_mode validate_mode)
{
struct dml2_context *dml2;
unsigned int result = 0;
@@ -685,10 +524,13 @@ static bool dml2_validate_only(struct dc_state *context)
build_unoptimized_policy_settings(dml2->v20.dml_core_ctx.project, &dml2->v20.dml_core_ctx.policy);
map_dc_state_into_dml_display_cfg(dml2, context, &dml2->v20.scratch.cur_display_config);
+ if (!dml2->config.skip_hw_state_mapping)
+ dml2_apply_det_buffer_allocation_policy(dml2, &dml2->v20.scratch.cur_display_config);
result = pack_and_call_dml_mode_support_ex(dml2,
&dml2->v20.scratch.cur_display_config,
- &dml2->v20.scratch.mode_support_info);
+ &dml2->v20.scratch.mode_support_info,
+ validate_mode);
if (result)
result = does_configuration_meet_sw_policies(dml2, &dml2->v20.scratch.cur_display_config, &dml2->v20.scratch.mode_support_info);
@@ -703,7 +545,8 @@ static void dml2_apply_debug_options(const struct dc *dc, struct dml2_context *d
}
}
-bool dml2_validate(const struct dc *in_dc, struct dc_state *context, struct dml2_context *dml2, bool fast_validate)
+bool dml2_validate(const struct dc *in_dc, struct dc_state *context, struct dml2_context *dml2,
+ enum dc_validate_mode validate_mode)
{
bool out = false;
@@ -713,30 +556,34 @@ bool dml2_validate(const struct dc *in_dc, struct dc_state *context, struct dml2
/* DML2.1 validation path */
if (dml2->architecture == dml2_architecture_21) {
- out = dml21_validate(in_dc, context, dml2, fast_validate);
+ out = dml21_validate(in_dc, context, dml2, validate_mode);
return out;
}
- /* Use dml_validate_only for fast_validate path */
- if (fast_validate)
- out = dml2_validate_only(context);
+ DC_FP_START();
+
+ /* Use dml_validate_only for DC_VALIDATE_MODE_ONLY and DC_VALIDATE_MODE_AND_STATE_INDEX path */
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING)
+ out = dml2_validate_only(context, validate_mode);
else
- out = dml2_validate_and_build_resource(in_dc, context);
+ out = dml2_validate_and_build_resource(in_dc, context, validate_mode);
+
+ DC_FP_END();
+
return out;
}
static inline struct dml2_context *dml2_allocate_memory(void)
{
- return (struct dml2_context *) kzalloc(sizeof(struct dml2_context), GFP_KERNEL);
+ return (struct dml2_context *) vzalloc(sizeof(struct dml2_context));
}
static void dml2_init(const struct dc *in_dc, const struct dml2_configuration_options *config, struct dml2_context **dml2)
{
- // TODO : Temporarily add DCN_VERSION_3_2 for N-1 validation. Remove DCN_VERSION_3_2 after N-1 validation phase is complete.
- if ((in_dc->debug.using_dml21) && (in_dc->ctx->dce_version == DCN_VERSION_4_01 || in_dc->ctx->dce_version == DCN_VERSION_3_2)) {
- dml21_reinit(in_dc, dml2, config);
+ if ((in_dc->debug.using_dml21) && (in_dc->ctx->dce_version >= DCN_VERSION_4_01)) {
+ dml21_reinit(in_dc, *dml2, config);
return;
- }
+ }
// Store config options
(*dml2)->config = *config;
@@ -748,6 +595,9 @@ static void dml2_init(const struct dc *in_dc, const struct dml2_configuration_op
case DCN_VERSION_3_51:
(*dml2)->v20.dml_core_ctx.project = dml_project_dcn351;
break;
+ case DCN_VERSION_3_6:
+ (*dml2)->v20.dml_core_ctx.project = dml_project_dcn36;
+ break;
case DCN_VERSION_3_2:
(*dml2)->v20.dml_core_ctx.project = dml_project_dcn32;
break;
@@ -762,19 +612,22 @@ static void dml2_init(const struct dc *in_dc, const struct dml2_configuration_op
break;
}
+ DC_FP_START();
+
initialize_dml2_ip_params(*dml2, in_dc, &(*dml2)->v20.dml_core_ctx.ip);
initialize_dml2_soc_bbox(*dml2, in_dc, &(*dml2)->v20.dml_core_ctx.soc);
initialize_dml2_soc_states(*dml2, in_dc, &(*dml2)->v20.dml_core_ctx.soc, &(*dml2)->v20.dml_core_ctx.states);
+
+ DC_FP_END();
}
bool dml2_create(const struct dc *in_dc, const struct dml2_configuration_options *config, struct dml2_context **dml2)
{
// TODO : Temporarily add DCN_VERSION_3_2 for N-1 validation. Remove DCN_VERSION_3_2 after N-1 validation phase is complete.
- if ((in_dc->debug.using_dml21) && (in_dc->ctx->dce_version == DCN_VERSION_4_01 || in_dc->ctx->dce_version == DCN_VERSION_3_2)) {
+ if ((in_dc->debug.using_dml21) && (in_dc->ctx->dce_version >= DCN_VERSION_4_01))
return dml21_create(in_dc, dml2, config);
- }
// Allocate Mode Lib Ctx
*dml2 = dml2_allocate_memory();
@@ -794,7 +647,7 @@ void dml2_destroy(struct dml2_context *dml2)
if (dml2->architecture == dml2_architecture_21)
dml21_destroy(dml2);
- kfree(dml2);
+ vfree(dml2);
}
void dml2_extract_dram_and_fclk_change_support(struct dml2_context *dml2,
@@ -842,9 +695,8 @@ void dml2_reinit(const struct dc *in_dc,
const struct dml2_configuration_options *config,
struct dml2_context **dml2)
{
- // TODO : Temporarily add DCN_VERSION_3_2 for N-1 validation. Remove DCN_VERSION_3_2 after N-1 validation phase is complete.
- if ((in_dc->debug.using_dml21) && (in_dc->ctx->dce_version == DCN_VERSION_4_01 || in_dc->ctx->dce_version == DCN_VERSION_3_2)) {
- dml21_reinit(in_dc, dml2, config);
+ if ((in_dc->debug.using_dml21) && (in_dc->ctx->dce_version >= DCN_VERSION_4_01)) {
+ dml21_reinit(in_dc, *dml2, config);
return;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h
index 0f944fcfd5a5..c384e141cebc 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_wrapper.h
@@ -40,6 +40,7 @@ struct dc_sink;
struct dc_stream_state;
struct resource_context;
struct display_stream_compressor;
+struct dc_mcache_params;
// Configuration of the MALL on the SoC
struct dml2_soc_mall_info {
@@ -107,6 +108,7 @@ struct dml2_dc_callbacks {
unsigned int (*get_max_flickerless_instant_vtotal_increase)(
struct dc_stream_state *stream,
bool is_gaming);
+ bool (*allocate_mcache)(struct dc_state *context, const struct dc_mcache_params *mcache_params);
};
struct dml2_dc_svp_callbacks {
@@ -159,6 +161,7 @@ struct dml2_clks_table_entry {
unsigned int dtbclk_mhz;
unsigned int dispclk_mhz;
unsigned int dppclk_mhz;
+ unsigned int dram_speed_mts; /*which is based on wck_ratio*/
};
struct dml2_clks_num_entries {
@@ -237,7 +240,7 @@ struct dml2_configuration_options {
bool use_clock_dc_limits;
bool gpuvm_enable;
bool force_tdlut_enable;
- struct dml2_soc_bb *bb_from_dmub;
+ void *bb_from_dmub;
};
/*
@@ -269,7 +272,7 @@ void dml2_reinit(const struct dc *in_dc,
* dml2_validate - Determines if a display configuration is supported or not.
* @in_dc: dc.
* @context: dc_state to be validated.
- * @fast_validate: Fast validate will not populate context.res_ctx.
+ * @validate_mode: DC_VALIDATE_MODE_ONLY and DC_VALIDATE_MODE_AND_STATE_INDEX will not populate context.res_ctx.
*
* DML1.0 compatible interface for validation.
*
@@ -292,7 +295,7 @@ void dml2_reinit(const struct dc *in_dc,
bool dml2_validate(const struct dc *in_dc,
struct dc_state *context,
struct dml2_context *dml2,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
/*
* dml2_extract_dram_and_fclk_change_support - Extracts the FCLK and UCLK change support info.
diff --git a/drivers/gpu/drm/amd/display/dc/dml2/dml_display_rq_dlg_calc.c b/drivers/gpu/drm/amd/display/dc/dml2/dml_display_rq_dlg_calc.c
index 377ef6d01ae5..00d22e542469 100644
--- a/drivers/gpu/drm/amd/display/dc/dml2/dml_display_rq_dlg_calc.c
+++ b/drivers/gpu/drm/amd/display/dc/dml2/dml_display_rq_dlg_calc.c
@@ -427,18 +427,6 @@ void dml_rq_dlg_get_dlg_reg(dml_display_dlg_regs_st *disp_dlg_regs,
dml_print("DML_DLG: %s: disp_dlg_regs->dst_y_per_vm_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_vm_flip);
dml_print("DML_DLG: %s: disp_dlg_regs->dst_y_per_row_flip = 0x%x\n", __func__, disp_dlg_regs->dst_y_per_row_flip);
- // hack for FPGA
- /* NOTE: We dont have getenv defined in driver and it does not make any sense in the driver */
- /*char* fpga_env = getenv("FPGA_FPDIV");
- if(fpga_env !=NULL)
- {
- if(disp_dlg_regs->vratio_prefetch >= (dml_uint_t)dml_pow(2, 22))
- {
- disp_dlg_regs->vratio_prefetch = (dml_uint_t)dml_pow(2, 22)-1;
- dml_print("FPGA msg: vratio_prefetch exceed the max value, the register field is [21:0]\n");
- }
- }*/
-
disp_dlg_regs->refcyc_per_vm_group_vblank = (dml_uint_t)(dml_get_refcyc_per_vm_group_vblank_in_us(mode_lib, pipe_idx) * refclk_freq_in_mhz);
disp_dlg_regs->refcyc_per_vm_group_flip = (dml_uint_t)(dml_get_refcyc_per_vm_group_flip_in_us(mode_lib, pipe_idx) * refclk_freq_in_mhz);
disp_dlg_regs->refcyc_per_vm_req_vblank = (dml_uint_t)(dml_get_refcyc_per_vm_req_vblank_in_us(mode_lib, pipe_idx) * refclk_freq_in_mhz * dml_pow(2, 10));
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c
index e1da48b05d00..75fb77bca83b 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn10/dcn10_dpp.c
@@ -194,6 +194,9 @@ void dpp_reset(struct dpp *dpp_base)
dpp->filter_h = NULL;
dpp->filter_v = NULL;
+ memset(&dpp_base->pos, 0, sizeof(dpp_base->pos));
+ memset(&dpp_base->att, 0, sizeof(dpp_base->att));
+
memset(&dpp->scl_data, 0, sizeof(dpp->scl_data));
memset(&dpp->pwl_data, 0, sizeof(dpp->pwl_data));
}
@@ -480,10 +483,11 @@ void dpp1_set_cursor_position(
if (src_y_offset + cursor_height <= 0)
cur_en = 0; /* not visible beyond top edge*/
- REG_UPDATE(CURSOR0_CONTROL,
- CUR0_ENABLE, cur_en);
+ if (dpp_base->pos.cur0_ctl.bits.cur0_enable != cur_en) {
+ REG_UPDATE(CURSOR0_CONTROL, CUR0_ENABLE, cur_en);
- dpp_base->pos.cur0_ctl.bits.cur0_enable = cur_en;
+ dpp_base->pos.cur0_ctl.bits.cur0_enable = cur_en;
+ }
}
void dpp1_cnv_set_optional_cursor_attributes(
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h
index f09cba8e29cc..85f359b5da67 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn20/dcn20_dpp.h
@@ -26,7 +26,6 @@
#define __DCN20_DPP_H__
#include "dcn10/dcn10_dpp.h"
-#include "spl/dc_spl_types.h"
#define TO_DCN20_DPP(dpp)\
container_of(dpp, struct dcn20_dpp, base)
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c
index 40acebd13e46..2d70586cef40 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn30/dcn30_dpp.c
@@ -425,11 +425,6 @@ bool dpp3_get_optimal_number_of_taps(
int min_taps_y, min_taps_c;
enum lb_memory_config lb_config;
- if (scl_data->viewport.width > scl_data->h_active &&
- dpp->ctx->dc->debug.max_downscale_src_width != 0 &&
- scl_data->viewport.width > dpp->ctx->dc->debug.max_downscale_src_width)
- return false;
-
/*
* Set default taps if none are provided
* From programming guide: taps = min{ ceil(2*H_RATIO,1), 8} for downscaling
@@ -467,6 +462,12 @@ bool dpp3_get_optimal_number_of_taps(
else
scl_data->taps.h_taps_c = in_taps->h_taps_c;
+ // Avoid null data in the scl data with this early return, proceed non-adaptive calcualtion first
+ if (scl_data->viewport.width > scl_data->h_active &&
+ dpp->ctx->dc->debug.max_downscale_src_width != 0 &&
+ scl_data->viewport.width > dpp->ctx->dc->debug.max_downscale_src_width)
+ return false;
+
/*Ensure we can support the requested number of vtaps*/
min_taps_y = dc_fixpt_ceil(scl_data->ratios.vert);
min_taps_c = dc_fixpt_ceil(scl_data->ratios.vert_c);
@@ -789,8 +790,7 @@ static bool dpp3_program_blnd_lut(struct dpp *dpp_base,
if (params == NULL) {
REG_SET(CM_BLNDGAM_CONTROL, 0, CM_BLNDGAM_MODE, 0);
- if (dpp_base->ctx->dc->debug.enable_mem_low_power.bits.cm)
- dpp3_power_on_blnd_lut(dpp_base, false);
+ dpp3_power_on_blnd_lut(dpp_base, false);
return false;
}
@@ -1203,8 +1203,7 @@ static bool dpp3_program_shaper(struct dpp *dpp_base,
if (params == NULL) {
REG_SET(CM_SHAPER_CONTROL, 0, CM_SHAPER_LUT_MODE, 0);
- if (dpp_base->ctx->dc->debug.enable_mem_low_power.bits.cm)
- dpp3_power_on_shaper(dpp_base, false);
+ dpp3_power_on_shaper(dpp_base, false);
return false;
}
@@ -1398,8 +1397,7 @@ static bool dpp3_program_3dlut(struct dpp *dpp_base,
if (params == NULL) {
dpp3_set_3dlut_mode(dpp_base, LUT_BYPASS, false, false);
- if (dpp_base->ctx->dc->debug.enable_mem_low_power.bits.cm)
- dpp3_power_on_hdr3dlut(dpp_base, false);
+ dpp3_power_on_hdr3dlut(dpp_base, false);
return false;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn32/dcn32_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn32/dcn32_dpp.h
index 992df172378c..f33dddbfcc31 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn32/dcn32_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn32/dcn32_dpp.h
@@ -27,7 +27,6 @@
#include "dcn20/dcn20_dpp.h"
#include "dcn30/dcn30_dpp.h"
-#include "spl/dc_spl_types.h"
bool dpp32_construct(struct dcn3_dpp *dpp3,
struct dc_context *ctx,
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c
index 62b7012cda43..f7a373a3d70a 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn35/dcn35_dpp.c
@@ -138,7 +138,7 @@ bool dpp35_construct(
dpp->base.funcs = &dcn35_dpp_funcs;
// w/a for cursor memory stuck in LS by programming DISPCLK_R_GATE_DISABLE, limit w/a to some ASIC revs
- if (dpp->base.ctx->asic_id.hw_internal_rev <= 0x10)
+ if (dpp->base.ctx->asic_id.hw_internal_rev < 0x40)
dpp->dispclk_r_gate_disable = true;
return ret;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.c
index 97bf26fa3573..36187f890d5d 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.c
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.c
@@ -231,7 +231,7 @@ static struct dpp_funcs dcn401_dpp_funcs = {
.dpp_program_regamma_pwl = NULL,
.dpp_set_pre_degam = dpp3_set_pre_degam,
.dpp_program_input_lut = NULL,
- .dpp_full_bypass = dpp401_full_bypass,
+ .dpp_full_bypass = NULL,
.dpp_setup = dpp401_dpp_setup,
.dpp_program_degamma_pwl = NULL,
.dpp_program_cm_dealpha = dpp3_program_cm_dealpha,
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.h b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.h
index 4bc85aaf17da..5a6a861402b3 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp.h
@@ -567,80 +567,83 @@
type ISHARP_NLDELTA_SCLIP_PIVOT_N; \
type ISHARP_NLDELTA_SCLIP_SLOPE_N
+#define DPP_REG_VARIABLE_LIST_DCN401 \
+ DPP_DCN3_REG_VARIABLE_LIST_COMMON; \
+ uint32_t CURSOR0_FP_SCALE_BIAS_G_Y; \
+ uint32_t CURSOR0_FP_SCALE_BIAS_RB_CRCB; \
+ uint32_t CUR0_MATRIX_MODE; \
+ uint32_t CUR0_MATRIX_C11_C12_A; \
+ uint32_t CUR0_MATRIX_C13_C14_A; \
+ uint32_t CUR0_MATRIX_C21_C22_A; \
+ uint32_t CUR0_MATRIX_C23_C24_A; \
+ uint32_t CUR0_MATRIX_C31_C32_A; \
+ uint32_t CUR0_MATRIX_C33_C34_A; \
+ uint32_t CUR0_MATRIX_C11_C12_B; \
+ uint32_t CUR0_MATRIX_C13_C14_B; \
+ uint32_t CUR0_MATRIX_C21_C22_B; \
+ uint32_t CUR0_MATRIX_C23_C24_B; \
+ uint32_t CUR0_MATRIX_C31_C32_B; \
+ uint32_t CUR0_MATRIX_C33_C34_B; \
+ uint32_t DSCL_SC_MODE; \
+ uint32_t DSCL_EASF_H_MODE; \
+ uint32_t DSCL_EASF_H_BF_CNTL; \
+ uint32_t DSCL_EASF_H_RINGEST_EVENTAP_REDUCE; \
+ uint32_t DSCL_EASF_H_RINGEST_EVENTAP_GAIN; \
+ uint32_t DSCL_EASF_H_BF_FINAL_MAX_MIN; \
+ uint32_t DSCL_EASF_H_BF1_PWL_SEG0; \
+ uint32_t DSCL_EASF_H_BF1_PWL_SEG1; \
+ uint32_t DSCL_EASF_H_BF1_PWL_SEG2; \
+ uint32_t DSCL_EASF_H_BF1_PWL_SEG3; \
+ uint32_t DSCL_EASF_H_BF1_PWL_SEG4; \
+ uint32_t DSCL_EASF_H_BF1_PWL_SEG5; \
+ uint32_t DSCL_EASF_H_BF1_PWL_SEG6; \
+ uint32_t DSCL_EASF_H_BF1_PWL_SEG7; \
+ uint32_t DSCL_EASF_H_BF3_PWL_SEG0; \
+ uint32_t DSCL_EASF_H_BF3_PWL_SEG1; \
+ uint32_t DSCL_EASF_H_BF3_PWL_SEG2; \
+ uint32_t DSCL_EASF_H_BF3_PWL_SEG3; \
+ uint32_t DSCL_EASF_H_BF3_PWL_SEG4; \
+ uint32_t DSCL_EASF_H_BF3_PWL_SEG5; \
+ uint32_t DSCL_EASF_V_MODE; \
+ uint32_t DSCL_EASF_V_BF_CNTL; \
+ uint32_t DSCL_EASF_V_RINGEST_3TAP_CNTL1; \
+ uint32_t DSCL_EASF_V_RINGEST_3TAP_CNTL2; \
+ uint32_t DSCL_EASF_V_RINGEST_3TAP_CNTL3; \
+ uint32_t DSCL_EASF_V_RINGEST_EVENTAP_REDUCE; \
+ uint32_t DSCL_EASF_V_RINGEST_EVENTAP_GAIN; \
+ uint32_t DSCL_EASF_V_BF_FINAL_MAX_MIN; \
+ uint32_t DSCL_EASF_V_BF1_PWL_SEG0; \
+ uint32_t DSCL_EASF_V_BF1_PWL_SEG1; \
+ uint32_t DSCL_EASF_V_BF1_PWL_SEG2; \
+ uint32_t DSCL_EASF_V_BF1_PWL_SEG3; \
+ uint32_t DSCL_EASF_V_BF1_PWL_SEG4; \
+ uint32_t DSCL_EASF_V_BF1_PWL_SEG5; \
+ uint32_t DSCL_EASF_V_BF1_PWL_SEG6; \
+ uint32_t DSCL_EASF_V_BF1_PWL_SEG7; \
+ uint32_t DSCL_EASF_V_BF3_PWL_SEG0; \
+ uint32_t DSCL_EASF_V_BF3_PWL_SEG1; \
+ uint32_t DSCL_EASF_V_BF3_PWL_SEG2; \
+ uint32_t DSCL_EASF_V_BF3_PWL_SEG3; \
+ uint32_t DSCL_EASF_V_BF3_PWL_SEG4; \
+ uint32_t DSCL_EASF_V_BF3_PWL_SEG5; \
+ uint32_t DSCL_SC_MATRIX_C0C1; \
+ uint32_t DSCL_SC_MATRIX_C2C3; \
+ uint32_t ISHARP_MODE; \
+ uint32_t ISHARP_NOISEDET_THRESHOLD; \
+ uint32_t ISHARP_NOISE_GAIN_PWL; \
+ uint32_t ISHARP_LBA_PWL_SEG0; \
+ uint32_t ISHARP_LBA_PWL_SEG1; \
+ uint32_t ISHARP_LBA_PWL_SEG2; \
+ uint32_t ISHARP_LBA_PWL_SEG3; \
+ uint32_t ISHARP_LBA_PWL_SEG4; \
+ uint32_t ISHARP_LBA_PWL_SEG5; \
+ uint32_t ISHARP_DELTA_CTRL; \
+ uint32_t ISHARP_DELTA_DATA; \
+ uint32_t ISHARP_DELTA_INDEX; \
+ uint32_t ISHARP_NLDELTA_SOFT_CLIP
+
struct dcn401_dpp_registers {
- DPP_DCN3_REG_VARIABLE_LIST_COMMON;
- uint32_t CURSOR0_FP_SCALE_BIAS_G_Y;
- uint32_t CURSOR0_FP_SCALE_BIAS_RB_CRCB;
- uint32_t CUR0_MATRIX_MODE;
- uint32_t CUR0_MATRIX_C11_C12_A;
- uint32_t CUR0_MATRIX_C13_C14_A;
- uint32_t CUR0_MATRIX_C21_C22_A;
- uint32_t CUR0_MATRIX_C23_C24_A;
- uint32_t CUR0_MATRIX_C31_C32_A;
- uint32_t CUR0_MATRIX_C33_C34_A;
- uint32_t CUR0_MATRIX_C11_C12_B;
- uint32_t CUR0_MATRIX_C13_C14_B;
- uint32_t CUR0_MATRIX_C21_C22_B;
- uint32_t CUR0_MATRIX_C23_C24_B;
- uint32_t CUR0_MATRIX_C31_C32_B;
- uint32_t CUR0_MATRIX_C33_C34_B;
- uint32_t DSCL_SC_MODE;
- uint32_t DSCL_EASF_H_MODE;
- uint32_t DSCL_EASF_H_BF_CNTL;
- uint32_t DSCL_EASF_H_RINGEST_EVENTAP_REDUCE;
- uint32_t DSCL_EASF_H_RINGEST_EVENTAP_GAIN;
- uint32_t DSCL_EASF_H_BF_FINAL_MAX_MIN;
- uint32_t DSCL_EASF_H_BF1_PWL_SEG0;
- uint32_t DSCL_EASF_H_BF1_PWL_SEG1;
- uint32_t DSCL_EASF_H_BF1_PWL_SEG2;
- uint32_t DSCL_EASF_H_BF1_PWL_SEG3;
- uint32_t DSCL_EASF_H_BF1_PWL_SEG4;
- uint32_t DSCL_EASF_H_BF1_PWL_SEG5;
- uint32_t DSCL_EASF_H_BF1_PWL_SEG6;
- uint32_t DSCL_EASF_H_BF1_PWL_SEG7;
- uint32_t DSCL_EASF_H_BF3_PWL_SEG0;
- uint32_t DSCL_EASF_H_BF3_PWL_SEG1;
- uint32_t DSCL_EASF_H_BF3_PWL_SEG2;
- uint32_t DSCL_EASF_H_BF3_PWL_SEG3;
- uint32_t DSCL_EASF_H_BF3_PWL_SEG4;
- uint32_t DSCL_EASF_H_BF3_PWL_SEG5;
- uint32_t DSCL_EASF_V_MODE;
- uint32_t DSCL_EASF_V_BF_CNTL;
- uint32_t DSCL_EASF_V_RINGEST_3TAP_CNTL1;
- uint32_t DSCL_EASF_V_RINGEST_3TAP_CNTL2;
- uint32_t DSCL_EASF_V_RINGEST_3TAP_CNTL3;
- uint32_t DSCL_EASF_V_RINGEST_EVENTAP_REDUCE;
- uint32_t DSCL_EASF_V_RINGEST_EVENTAP_GAIN;
- uint32_t DSCL_EASF_V_BF_FINAL_MAX_MIN;
- uint32_t DSCL_EASF_V_BF1_PWL_SEG0;
- uint32_t DSCL_EASF_V_BF1_PWL_SEG1;
- uint32_t DSCL_EASF_V_BF1_PWL_SEG2;
- uint32_t DSCL_EASF_V_BF1_PWL_SEG3;
- uint32_t DSCL_EASF_V_BF1_PWL_SEG4;
- uint32_t DSCL_EASF_V_BF1_PWL_SEG5;
- uint32_t DSCL_EASF_V_BF1_PWL_SEG6;
- uint32_t DSCL_EASF_V_BF1_PWL_SEG7;
- uint32_t DSCL_EASF_V_BF3_PWL_SEG0;
- uint32_t DSCL_EASF_V_BF3_PWL_SEG1;
- uint32_t DSCL_EASF_V_BF3_PWL_SEG2;
- uint32_t DSCL_EASF_V_BF3_PWL_SEG3;
- uint32_t DSCL_EASF_V_BF3_PWL_SEG4;
- uint32_t DSCL_EASF_V_BF3_PWL_SEG5;
- uint32_t DSCL_SC_MATRIX_C0C1;
- uint32_t DSCL_SC_MATRIX_C2C3;
- uint32_t ISHARP_MODE;
- uint32_t ISHARP_NOISEDET_THRESHOLD;
- uint32_t ISHARP_NOISE_GAIN_PWL;
- uint32_t ISHARP_LBA_PWL_SEG0;
- uint32_t ISHARP_LBA_PWL_SEG1;
- uint32_t ISHARP_LBA_PWL_SEG2;
- uint32_t ISHARP_LBA_PWL_SEG3;
- uint32_t ISHARP_LBA_PWL_SEG4;
- uint32_t ISHARP_LBA_PWL_SEG5;
- uint32_t ISHARP_DELTA_CTRL;
- uint32_t ISHARP_DELTA_DATA;
- uint32_t ISHARP_DELTA_INDEX;
- uint32_t ISHARP_NLDELTA_SOFT_CLIP;
+ DPP_REG_VARIABLE_LIST_DCN401;
};
struct dcn401_dpp_shift {
@@ -681,8 +684,6 @@ void dpp401_dscl_set_scaler_manual_scale(
struct dpp *dpp_base,
const struct scaler_data *scl_data);
-void dpp401_full_bypass(struct dpp *dpp_base);
-
void dpp401_dpp_setup(
struct dpp *dpp_base,
enum surface_pixel_format format,
diff --git a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c
index 3b6ca7974e18..7aab77b58869 100644
--- a/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c
+++ b/drivers/gpu/drm/amd/display/dc/dpp/dcn401/dcn401_dpp_cm.c
@@ -88,30 +88,6 @@ enum dscl_mode_sel {
DSCL_MODE_DSCL_BYPASS = 6
};
-void dpp401_full_bypass(struct dpp *dpp_base)
-{
- struct dcn401_dpp *dpp = TO_DCN401_DPP(dpp_base);
-
- /* Input pixel format: ARGB8888 */
- REG_SET(CNVC_SURFACE_PIXEL_FORMAT, 0,
- CNVC_SURFACE_PIXEL_FORMAT, 0x8);
-
- /* Zero expansion */
- REG_SET_3(FORMAT_CONTROL, 0,
- CNVC_BYPASS, 0,
- FORMAT_CONTROL__ALPHA_EN, 0,
- FORMAT_EXPANSION_MODE, 0);
-
- /* COLOR_KEYER_CONTROL.COLOR_KEYER_EN = 0 this should be default */
- if (dpp->tf_mask->CM_BYPASS_EN)
- REG_SET(CM_CONTROL, 0, CM_BYPASS_EN, 1);
- else
- REG_SET(CM_CONTROL, 0, CM_BYPASS, 1);
-
- /* Setting degamma bypass for now */
- REG_SET(CM_DGAM_CONTROL, 0, CM_DGAM_LUT_MODE, 0);
-}
-
void dpp401_set_cursor_attributes(
struct dpp *dpp_base,
struct dc_cursor_attributes *cursor_attributes)
@@ -120,10 +96,11 @@ void dpp401_set_cursor_attributes(
enum dc_cursor_color_format color_format = cursor_attributes->color_format;
int cur_rom_en = 0;
- // DCN4 should always do Cursor degamma for Cursor Color modes
if (color_format == CURSOR_MODE_COLOR_PRE_MULTIPLIED_ALPHA ||
color_format == CURSOR_MODE_COLOR_UN_PRE_MULTIPLIED_ALPHA) {
- cur_rom_en = 1;
+ if (cursor_attributes->attribute_flags.bits.ENABLE_CURSOR_DEGAMMA) {
+ cur_rom_en = 1;
+ }
}
REG_UPDATE_3(CURSOR0_CONTROL,
@@ -154,9 +131,11 @@ void dpp401_set_cursor_position(
struct dcn401_dpp *dpp = TO_DCN401_DPP(dpp_base);
uint32_t cur_en = pos->enable ? 1 : 0;
- REG_UPDATE(CURSOR0_CONTROL, CUR0_ENABLE, cur_en);
+ if (dpp_base->pos.cur0_ctl.bits.cur0_enable != cur_en) {
+ REG_UPDATE(CURSOR0_CONTROL, CUR0_ENABLE, cur_en);
- dpp_base->pos.cur0_ctl.bits.cur0_enable = cur_en;
+ dpp_base->pos.cur0_ctl.bits.cur0_enable = cur_en;
+ }
}
void dpp401_set_optional_cursor_attributes(
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c
index d9aaebfa3a0a..1f53a9f0c0ac 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c
+++ b/drivers/gpu/drm/amd/display/dc/dsc/dc_dsc.c
@@ -30,6 +30,12 @@
#include "rc_calc.h"
#include "fixed31_32.h"
+#include "clk_mgr.h"
+#include "resource.h"
+
+#define DC_LOGGER \
+ dsc->ctx->logger
+
/* This module's internal functions */
/* default DSC policy target bitrate limit is 16bpp */
@@ -146,6 +152,11 @@ uint32_t dc_bandwidth_in_kbps_from_timing(
}
/* Forward Declerations */
+static unsigned int get_min_dsc_slice_count_for_odm(
+ const struct display_stream_compressor *dsc,
+ const struct dsc_enc_caps *dsc_enc_caps,
+ const struct dc_crtc_timing *timing);
+
static bool decide_dsc_bandwidth_range(
const uint32_t min_bpp_x16,
const uint32_t max_bpp_x16,
@@ -180,6 +191,7 @@ static bool setup_dsc_config(
const struct dc_crtc_timing *timing,
const struct dc_dsc_config_options *options,
const enum dc_link_encoding_format link_encoding,
+ int min_slice_count,
struct dc_dsc_config *dsc_cfg);
static bool dsc_buff_block_size_from_dpcd(int dpcd_buff_block_size, int *buff_block_size)
@@ -439,7 +451,6 @@ bool dc_dsc_parse_dsc_dpcd(const struct dc *dc,
return true;
}
-
/* If DSC is possbile, get DSC bandwidth range based on [min_bpp, max_bpp] target bitrate range and
* timing's pixel clock and uncompressed bandwidth.
* If DSC is not possible, leave '*range' untouched.
@@ -455,6 +466,7 @@ bool dc_dsc_compute_bandwidth_range(
struct dc_dsc_bw_range *range)
{
bool is_dsc_possible = false;
+ unsigned int min_dsc_slice_count;
struct dsc_enc_caps dsc_enc_caps;
struct dsc_enc_caps dsc_common_caps;
struct dc_dsc_config config = {0};
@@ -466,12 +478,14 @@ bool dc_dsc_compute_bandwidth_range(
get_dsc_enc_caps(dsc, &dsc_enc_caps, timing->pix_clk_100hz);
+ min_dsc_slice_count = get_min_dsc_slice_count_for_odm(dsc, &dsc_enc_caps, timing);
+
is_dsc_possible = intersect_dsc_caps(dsc_sink_caps, &dsc_enc_caps,
timing->pixel_encoding, &dsc_common_caps);
if (is_dsc_possible)
is_dsc_possible = setup_dsc_config(dsc_sink_caps, &dsc_enc_caps, 0, timing,
- &options, link_encoding, &config);
+ &options, link_encoding, min_dsc_slice_count, &config);
if (is_dsc_possible)
is_dsc_possible = decide_dsc_bandwidth_range(min_bpp_x16, max_bpp_x16,
@@ -480,20 +494,195 @@ bool dc_dsc_compute_bandwidth_range(
return is_dsc_possible;
}
+void dc_dsc_dump_encoder_caps(const struct display_stream_compressor *dsc,
+ const struct dc_crtc_timing *timing)
+{
+ struct dsc_enc_caps dsc_enc_caps;
+
+ get_dsc_enc_caps(dsc, &dsc_enc_caps, timing->pix_clk_100hz);
+
+ DC_LOG_DSC("dsc encoder caps:");
+ DC_LOG_DSC("\tdsc_version 0x%x", dsc_enc_caps.dsc_version);
+ DC_LOG_DSC("\tslice_caps 0x%x", dsc_enc_caps.slice_caps.raw);
+ DC_LOG_DSC("\tlb_bit_depth %d", dsc_enc_caps.lb_bit_depth);
+ DC_LOG_DSC("\tis_block_pred_supported %d", dsc_enc_caps.is_block_pred_supported);
+ DC_LOG_DSC("\tcolor_formats 0x%x", dsc_enc_caps.color_formats.raw);
+ DC_LOG_DSC("\tcolor_depth 0x%x", dsc_enc_caps.color_depth.raw);
+ DC_LOG_DSC("\tmax_total_throughput_mps %d", dsc_enc_caps.max_total_throughput_mps);
+ DC_LOG_DSC("\tmax_slice_width %d", dsc_enc_caps.max_slice_width);
+ DC_LOG_DSC("\tbpp_increment_div %d", dsc_enc_caps.bpp_increment_div);
+}
+
+void dc_dsc_dump_decoder_caps(const struct display_stream_compressor *dsc,
+ const struct dsc_dec_dpcd_caps *dsc_sink_caps)
+{
+ DC_LOG_DSC("dsc decoder caps:");
+ DC_LOG_DSC("\tis_dsc_supported %d", dsc_sink_caps->is_dsc_supported);
+ DC_LOG_DSC("\tdsc_version 0x%x", dsc_sink_caps->dsc_version);
+ DC_LOG_DSC("\trc_buffer_size %d", dsc_sink_caps->rc_buffer_size);
+ DC_LOG_DSC("\tslice_caps1 0x%x", dsc_sink_caps->slice_caps1.raw);
+ DC_LOG_DSC("\tslice_caps2 0x%x", dsc_sink_caps->slice_caps2.raw);
+ DC_LOG_DSC("\tlb_bit_depth %d", dsc_sink_caps->lb_bit_depth);
+ DC_LOG_DSC("\tis_block_pred_supported %d", dsc_sink_caps->is_block_pred_supported);
+ DC_LOG_DSC("\tedp_max_bits_per_pixel %d", dsc_sink_caps->edp_max_bits_per_pixel);
+ DC_LOG_DSC("\tcolor_formats 0x%x", dsc_sink_caps->color_formats.raw);
+ DC_LOG_DSC("\tthroughput_mode_0_mps %d", dsc_sink_caps->throughput_mode_0_mps);
+ DC_LOG_DSC("\tthroughput_mode_1_mps %d", dsc_sink_caps->throughput_mode_1_mps);
+ DC_LOG_DSC("\tmax_slice_width %d", dsc_sink_caps->max_slice_width);
+ DC_LOG_DSC("\tbpp_increment_div %d", dsc_sink_caps->bpp_increment_div);
+ DC_LOG_DSC("\tbranch_overall_throughput_0_mps %d", dsc_sink_caps->branch_overall_throughput_0_mps);
+ DC_LOG_DSC("\tbranch_overall_throughput_1_mps %d", dsc_sink_caps->branch_overall_throughput_1_mps);
+ DC_LOG_DSC("\tbranch_max_line_width %d", dsc_sink_caps->branch_max_line_width);
+ DC_LOG_DSC("\tis_dp %d", dsc_sink_caps->is_dp);
+}
+
+
+static void build_dsc_enc_combined_slice_caps(
+ const struct dsc_enc_caps *single_dsc_enc_caps,
+ struct dsc_enc_caps *dsc_enc_caps,
+ unsigned int max_odm_combine_factor)
+{
+ /* 1-16 slice configurations, single DSC */
+ dsc_enc_caps->slice_caps.raw |= single_dsc_enc_caps->slice_caps.raw;
+
+ /* 2x DSC's */
+ if (max_odm_combine_factor >= 2) {
+ /* 1 + 1 */
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_2 |= single_dsc_enc_caps->slice_caps.bits.NUM_SLICES_1;
+
+ /* 2 + 2 */
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_4 |= single_dsc_enc_caps->slice_caps.bits.NUM_SLICES_2;
+
+ /* 4 + 4 */
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_8 |= single_dsc_enc_caps->slice_caps.bits.NUM_SLICES_4;
+
+ /* 8 + 8 */
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_16 |= single_dsc_enc_caps->slice_caps.bits.NUM_SLICES_8;
+ }
+
+ /* 3x DSC's */
+ if (max_odm_combine_factor >= 3) {
+ /* 4 + 4 + 4 */
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_12 |= single_dsc_enc_caps->slice_caps.bits.NUM_SLICES_4;
+ }
+
+ /* 4x DSC's */
+ if (max_odm_combine_factor >= 4) {
+ /* 1 + 1 + 1 + 1 */
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_4 |= single_dsc_enc_caps->slice_caps.bits.NUM_SLICES_1;
+
+ /* 2 + 2 + 2 + 2 */
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_8 |= single_dsc_enc_caps->slice_caps.bits.NUM_SLICES_2;
+
+ /* 3 + 3 + 3 + 3 */
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_12 |= single_dsc_enc_caps->slice_caps.bits.NUM_SLICES_3;
+
+ /* 4 + 4 + 4 + 4 */
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_16 |= single_dsc_enc_caps->slice_caps.bits.NUM_SLICES_4;
+ }
+}
+
+static void build_dsc_enc_caps(
+ const struct display_stream_compressor *dsc,
+ struct dsc_enc_caps *dsc_enc_caps)
+{
+ unsigned int max_dscclk_khz;
+ unsigned int num_dsc;
+ unsigned int max_odm_combine_factor;
+ struct dsc_enc_caps single_dsc_enc_caps;
+
+ struct dc *dc;
+
+ if (!dsc || !dsc->ctx || !dsc->ctx->dc || !dsc->funcs->dsc_get_single_enc_caps)
+ return;
+
+ dc = dsc->ctx->dc;
+
+ if (!dc->clk_mgr || !dc->clk_mgr->funcs->get_max_clock_khz || !dc->res_pool || dc->debug.disable_dsc)
+ return;
+
+ /* get max DSCCLK from clk_mgr */
+ max_dscclk_khz = dc->clk_mgr->funcs->get_max_clock_khz(dc->clk_mgr, CLK_TYPE_DSCCLK);
+
+ dsc->funcs->dsc_get_single_enc_caps(&single_dsc_enc_caps, max_dscclk_khz);
+
+ /* global capabilities */
+ dsc_enc_caps->dsc_version = single_dsc_enc_caps.dsc_version;
+ dsc_enc_caps->lb_bit_depth = single_dsc_enc_caps.lb_bit_depth;
+ dsc_enc_caps->is_block_pred_supported = single_dsc_enc_caps.is_block_pred_supported;
+ dsc_enc_caps->max_slice_width = single_dsc_enc_caps.max_slice_width;
+ dsc_enc_caps->bpp_increment_div = single_dsc_enc_caps.bpp_increment_div;
+ dsc_enc_caps->color_formats.raw = single_dsc_enc_caps.color_formats.raw;
+ dsc_enc_caps->color_depth.raw = single_dsc_enc_caps.color_depth.raw;
+
+ /* expand per DSC capabilities to global */
+ max_odm_combine_factor = dc->caps.max_odm_combine_factor;
+ num_dsc = dc->res_pool->res_cap->num_dsc;
+ max_odm_combine_factor = min(max_odm_combine_factor, num_dsc);
+ dsc_enc_caps->max_total_throughput_mps =
+ single_dsc_enc_caps.max_total_throughput_mps *
+ max_odm_combine_factor;
+
+ /* check slice counts possible for with ODM combine */
+ build_dsc_enc_combined_slice_caps(&single_dsc_enc_caps, dsc_enc_caps, max_odm_combine_factor);
+}
+
+static inline uint32_t dsc_div_by_10_round_up(uint32_t value)
+{
+ return (value + 9) / 10;
+}
+
+static unsigned int get_min_dsc_slice_count_for_odm(
+ const struct display_stream_compressor *dsc,
+ const struct dsc_enc_caps *dsc_enc_caps,
+ const struct dc_crtc_timing *timing)
+{
+ unsigned int max_dispclk_khz;
+
+ /* get max pixel rate and combine caps */
+ max_dispclk_khz = dsc_enc_caps->max_total_throughput_mps * 1000;
+ if (dsc && dsc->ctx->dc) {
+ if (dsc->ctx->dc->clk_mgr &&
+ dsc->ctx->dc->clk_mgr->funcs->get_max_clock_khz) {
+ /* dispclk is available */
+ max_dispclk_khz = dsc->ctx->dc->clk_mgr->funcs->get_max_clock_khz(dsc->ctx->dc->clk_mgr, CLK_TYPE_DISPCLK);
+ }
+ }
+
+ /* validate parameters */
+ if (max_dispclk_khz == 0 || dsc_enc_caps->max_slice_width == 0)
+ return 1;
+
+ /* consider minimum odm slices required due to
+ * 1) display pipe throughput (dispclk)
+ * 2) max image width per slice
+ */
+ return dc_fixpt_ceil(dc_fixpt_max(
+ dc_fixpt_div_int(dc_fixpt_from_int(dsc_div_by_10_round_up(timing->pix_clk_100hz)),
+ max_dispclk_khz), // throughput
+ dc_fixpt_div_int(dc_fixpt_from_int(timing->h_addressable + timing->h_border_left + timing->h_border_right),
+ dsc_enc_caps->max_slice_width))); // slice width
+}
+
static void get_dsc_enc_caps(
const struct display_stream_compressor *dsc,
struct dsc_enc_caps *dsc_enc_caps,
int pixel_clock_100Hz)
{
- // This is a static HW query, so we can use any DSC
-
memset(dsc_enc_caps, 0, sizeof(struct dsc_enc_caps));
- if (dsc) {
- if (!dsc->ctx->dc->debug.disable_dsc)
- dsc->funcs->dsc_get_enc_caps(dsc_enc_caps, pixel_clock_100Hz);
- if (dsc->ctx->dc->debug.native422_support)
- dsc_enc_caps->color_formats.bits.YCBCR_NATIVE_422 = 1;
+
+ if (!dsc || !dsc->ctx || !dsc->ctx->dc || dsc->ctx->dc->debug.disable_dsc)
+ return;
+
+ /* check if reported cap global or only for a single DCN DSC enc */
+ if (dsc->funcs->dsc_get_enc_caps) {
+ dsc->funcs->dsc_get_enc_caps(dsc_enc_caps, pixel_clock_100Hz);
+ } else {
+ build_dsc_enc_caps(dsc, dsc_enc_caps);
}
+
+ if (dsc->ctx->dc->debug.native422_support)
+ dsc_enc_caps->color_formats.bits.YCBCR_NATIVE_422 = 1;
}
/* Returns 'false' if no intersection was found for at least one capability.
@@ -576,11 +765,6 @@ static bool intersect_dsc_caps(
return true;
}
-static inline uint32_t dsc_div_by_10_round_up(uint32_t value)
-{
- return (value + 9) / 10;
-}
-
static uint32_t compute_bpp_x16_from_target_bandwidth(
const uint32_t bandwidth_in_kbps,
const struct dc_crtc_timing *timing,
@@ -865,11 +1049,11 @@ static bool setup_dsc_config(
const struct dc_crtc_timing *timing,
const struct dc_dsc_config_options *options,
const enum dc_link_encoding_format link_encoding,
+ int min_slices_h,
struct dc_dsc_config *dsc_cfg)
{
struct dsc_enc_caps dsc_common_caps;
int max_slices_h = 0;
- int min_slices_h = 0;
int num_slices_h = 0;
int pic_width;
int slice_width;
@@ -973,12 +1157,9 @@ static bool setup_dsc_config(
if (!is_dsc_possible)
goto done;
- min_slices_h = pic_width / dsc_common_caps.max_slice_width;
- if (pic_width % dsc_common_caps.max_slice_width)
- min_slices_h++;
-
min_slices_h = fit_num_slices_up(dsc_common_caps.slice_caps, min_slices_h);
+ /* increase minimum slice count to meet sink throughput limitations */
while (min_slices_h <= max_slices_h) {
int pix_clk_per_slice_khz = dsc_div_by_10_round_up(timing->pix_clk_100hz) / min_slices_h;
if (pix_clk_per_slice_khz <= sink_per_slice_throughput_mps * 1000)
@@ -987,14 +1168,12 @@ static bool setup_dsc_config(
min_slices_h = inc_num_slices(dsc_common_caps.slice_caps, min_slices_h);
}
- is_dsc_possible = (min_slices_h <= max_slices_h);
-
- if (pic_width % min_slices_h != 0)
- min_slices_h = 0; // DSC TODO: Maybe try increasing the number of slices first?
-
- if (min_slices_h == 0 && max_slices_h == 0)
- is_dsc_possible = false;
+ /* increase minimum slice count to meet divisibility requirements */
+ while (pic_width % min_slices_h != 0 && min_slices_h <= max_slices_h) {
+ min_slices_h = inc_num_slices(dsc_common_caps.slice_caps, min_slices_h);
+ }
+ is_dsc_possible = (min_slices_h <= max_slices_h) && max_slices_h != 0;
if (!is_dsc_possible)
goto done;
@@ -1117,12 +1296,19 @@ bool dc_dsc_compute_config(
{
bool is_dsc_possible = false;
struct dsc_enc_caps dsc_enc_caps;
-
+ unsigned int min_dsc_slice_count;
get_dsc_enc_caps(dsc, &dsc_enc_caps, timing->pix_clk_100hz);
+
+ min_dsc_slice_count = get_min_dsc_slice_count_for_odm(dsc, &dsc_enc_caps, timing);
+
is_dsc_possible = setup_dsc_config(dsc_sink_caps,
&dsc_enc_caps,
target_bandwidth_kbps,
- timing, options, link_encoding, dsc_cfg);
+ timing,
+ options,
+ link_encoding,
+ min_dsc_slice_count,
+ dsc_cfg);
return is_dsc_possible;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.c b/drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.c
index 75128fd34306..bd1b9aef6d5c 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.c
+++ b/drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.c
@@ -57,13 +57,6 @@ static const struct dsc_funcs dcn20_dsc_funcs = {
#define DC_LOGGER \
dsc->ctx->logger
-enum dsc_bits_per_comp {
- DSC_BPC_8 = 8,
- DSC_BPC_10 = 10,
- DSC_BPC_12 = 12,
- DSC_BPC_UNKNOWN
-};
-
/* API functions (external or via structure->function_pointer) */
void dsc2_construct(struct dcn20_dsc *dsc,
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.h b/drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.h
index 1fb90b52b814..a9c04fc95bd1 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.h
+++ b/drivers/gpu/drm/amd/display/dc/dsc/dcn20/dcn20_dsc.h
@@ -457,6 +457,12 @@
type DSCRM_DSC_DOUBLE_BUFFER_REG_UPDATE_PENDING; \
type DSCRM_DSC_FORWARD_EN_STATUS
+enum dsc_bits_per_comp {
+ DSC_BPC_8 = 8,
+ DSC_BPC_10 = 10,
+ DSC_BPC_12 = 12,
+ DSC_BPC_UNKNOWN
+};
struct dcn20_dsc_registers {
uint32_t DSC_TOP_CONTROL;
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.c b/drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.c
index 61678b0a5a1e..7bd92ae8b13e 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.c
+++ b/drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.c
@@ -9,26 +9,14 @@
#include "dsc/dscc_types.h"
#include "dsc/rc_calc.h"
-#define MAX_THROUGHPUT_PER_DSC_100HZ 20000000
-#define MAX_DSC_UNIT_COMBINE 4
-
static void dsc_write_to_registers(struct display_stream_compressor *dsc, const struct dsc_reg_values *reg_vals);
/* Object I/F functions */
//static void dsc401_get_enc_caps(struct dsc_enc_caps *dsc_enc_caps, int pixel_clock_100Hz);
-static void dsc401_read_state(struct display_stream_compressor *dsc, struct dcn_dsc_state *s);
-static bool dsc401_validate_stream(struct display_stream_compressor *dsc, const struct dsc_config *dsc_cfg);
-static void dsc401_set_config(struct display_stream_compressor *dsc, const struct dsc_config *dsc_cfg,
- struct dsc_optc_config *dsc_optc_cfg);
//static bool dsc401_get_packed_pps(struct display_stream_compressor *dsc, const struct dsc_config *dsc_cfg, uint8_t *dsc_packed_pps);
-static void dsc401_enable(struct display_stream_compressor *dsc, int opp_pipe);
-static void dsc401_disable(struct display_stream_compressor *dsc);
-static void dsc401_disconnect(struct display_stream_compressor *dsc);
-static void dsc401_wait_disconnect_pending_clear(struct display_stream_compressor *dsc);
-static void dsc401_get_enc_caps(struct dsc_enc_caps *dsc_enc_caps, int pixel_clock_100Hz);
+static void dsc401_get_single_enc_caps(struct dsc_enc_caps *dsc_enc_caps, unsigned int max_dscclk_khz);
static const struct dsc_funcs dcn401_dsc_funcs = {
- .dsc_get_enc_caps = dsc401_get_enc_caps,
.dsc_read_state = dsc401_read_state,
.dsc_validate_stream = dsc401_validate_stream,
.dsc_set_config = dsc401_set_config,
@@ -37,6 +25,7 @@ static const struct dsc_funcs dcn401_dsc_funcs = {
.dsc_disable = dsc401_disable,
.dsc_disconnect = dsc401_disconnect,
.dsc_wait_disconnect_pending_clear = dsc401_wait_disconnect_pending_clear,
+ .dsc_get_single_enc_caps = dsc401_get_single_enc_caps,
};
/* Macro definitios for REG_SET macros*/
@@ -52,12 +41,6 @@ static const struct dsc_funcs dcn401_dsc_funcs = {
#define DC_LOGGER \
dsc->ctx->logger
-enum dsc_bits_per_comp {
- DSC_BPC_8 = 8,
- DSC_BPC_10 = 10,
- DSC_BPC_12 = 12,
- DSC_BPC_UNKNOWN
-};
/* API functions (external or via structure->function_pointer) */
@@ -79,22 +62,14 @@ void dsc401_construct(struct dcn401_dsc *dsc,
dsc->max_image_width = 5184;
}
-static void dsc401_get_enc_caps(struct dsc_enc_caps *dsc_enc_caps, int pixel_clock_100Hz)
+static void dsc401_get_single_enc_caps(struct dsc_enc_caps *dsc_enc_caps, unsigned int max_dscclk_khz)
{
- int min_dsc_unit_required = (pixel_clock_100Hz + MAX_THROUGHPUT_PER_DSC_100HZ - 1) / MAX_THROUGHPUT_PER_DSC_100HZ;
-
dsc_enc_caps->dsc_version = 0x21; /* v1.2 - DP spec defined it in reverse order and we kept it */
- /* 1 slice is only supported with 1 DSC unit */
- dsc_enc_caps->slice_caps.bits.NUM_SLICES_1 = min_dsc_unit_required == 1 ? 1 : 0;
- /* 2 slice is only supported with 1 or 2 DSC units */
- dsc_enc_caps->slice_caps.bits.NUM_SLICES_2 = (min_dsc_unit_required == 1 || min_dsc_unit_required == 2) ? 1 : 0;
- /* 3 slice is only supported with 1 DSC unit */
- dsc_enc_caps->slice_caps.bits.NUM_SLICES_3 = min_dsc_unit_required == 1 ? 1 : 0;
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_1 = 1;
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_2 = 1;
+ dsc_enc_caps->slice_caps.bits.NUM_SLICES_3 = 1;
dsc_enc_caps->slice_caps.bits.NUM_SLICES_4 = 1;
- dsc_enc_caps->slice_caps.bits.NUM_SLICES_8 = 1;
- dsc_enc_caps->slice_caps.bits.NUM_SLICES_12 = 1;
- dsc_enc_caps->slice_caps.bits.NUM_SLICES_16 = 1;
dsc_enc_caps->lb_bit_depth = 13;
dsc_enc_caps->is_block_pred_supported = true;
@@ -108,7 +83,7 @@ static void dsc401_get_enc_caps(struct dsc_enc_caps *dsc_enc_caps, int pixel_clo
dsc_enc_caps->color_depth.bits.COLOR_DEPTH_8_BPC = 1;
dsc_enc_caps->color_depth.bits.COLOR_DEPTH_10_BPC = 1;
dsc_enc_caps->color_depth.bits.COLOR_DEPTH_12_BPC = 1;
- dsc_enc_caps->max_total_throughput_mps = MAX_THROUGHPUT_PER_DSC_100HZ * MAX_DSC_UNIT_COMBINE;
+ dsc_enc_caps->max_total_throughput_mps = max_dscclk_khz * 3 / 1000;
dsc_enc_caps->max_slice_width = 5184; /* (including 64 overlap pixels for eDP MSO mode) */
dsc_enc_caps->bpp_increment_div = 16; /* 1/16th of a bit */
@@ -117,7 +92,7 @@ static void dsc401_get_enc_caps(struct dsc_enc_caps *dsc_enc_caps, int pixel_clo
/* this function read dsc related register fields to be logged later in dcn10_log_hw_state
* into a dcn_dsc_state struct.
*/
-static void dsc401_read_state(struct display_stream_compressor *dsc, struct dcn_dsc_state *s)
+void dsc401_read_state(struct display_stream_compressor *dsc, struct dcn_dsc_state *s)
{
struct dcn401_dsc *dsc401 = TO_DCN401_DSC(dsc);
@@ -134,7 +109,7 @@ static void dsc401_read_state(struct display_stream_compressor *dsc, struct dcn_
}
-static bool dsc401_validate_stream(struct display_stream_compressor *dsc, const struct dsc_config *dsc_cfg)
+bool dsc401_validate_stream(struct display_stream_compressor *dsc, const struct dsc_config *dsc_cfg)
{
struct dsc_optc_config dsc_optc_cfg;
struct dcn401_dsc *dsc401 = TO_DCN401_DSC(dsc);
@@ -145,7 +120,7 @@ static bool dsc401_validate_stream(struct display_stream_compressor *dsc, const
return dsc_prepare_config(dsc_cfg, &dsc401->reg_vals, &dsc_optc_cfg);
}
-static void dsc401_set_config(struct display_stream_compressor *dsc, const struct dsc_config *dsc_cfg,
+void dsc401_set_config(struct display_stream_compressor *dsc, const struct dsc_config *dsc_cfg,
struct dsc_optc_config *dsc_optc_cfg)
{
bool is_config_ok;
@@ -160,7 +135,7 @@ static void dsc401_set_config(struct display_stream_compressor *dsc, const struc
dsc_write_to_registers(dsc, &dsc401->reg_vals);
}
-static void dsc401_enable(struct display_stream_compressor *dsc, int opp_pipe)
+void dsc401_enable(struct display_stream_compressor *dsc, int opp_pipe)
{
struct dcn401_dsc *dsc401 = TO_DCN401_DSC(dsc);
int dsc_clock_en;
@@ -185,7 +160,7 @@ static void dsc401_enable(struct display_stream_compressor *dsc, int opp_pipe)
}
-static void dsc401_disable(struct display_stream_compressor *dsc)
+void dsc401_disable(struct display_stream_compressor *dsc)
{
struct dcn401_dsc *dsc401 = TO_DCN401_DSC(dsc);
int dsc_clock_en;
@@ -204,14 +179,14 @@ static void dsc401_disable(struct display_stream_compressor *dsc)
DSC_CLOCK_EN, 0);
}
-static void dsc401_wait_disconnect_pending_clear(struct display_stream_compressor *dsc)
+void dsc401_wait_disconnect_pending_clear(struct display_stream_compressor *dsc)
{
struct dcn401_dsc *dsc401 = TO_DCN401_DSC(dsc);
REG_WAIT(DSCRM_DSC_FORWARD_CONFIG, DSCRM_DSC_FORWARD_EN_STATUS, 0, 2, 50000);
}
-static void dsc401_disconnect(struct display_stream_compressor *dsc)
+void dsc401_disconnect(struct display_stream_compressor *dsc)
{
struct dcn401_dsc *dsc401 = TO_DCN401_DSC(dsc);
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.h b/drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.h
index 3c9fa8988974..7acd57eb4f42 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.h
+++ b/drivers/gpu/drm/amd/display/dc/dsc/dcn401/dcn401_dsc.h
@@ -334,5 +334,13 @@ void dsc401_construct(struct dcn401_dsc *dsc,
void dsc401_set_fgcg(struct dcn401_dsc *dsc401, bool enable);
+void dsc401_read_state(struct display_stream_compressor *dsc, struct dcn_dsc_state *s);
+bool dsc401_validate_stream(struct display_stream_compressor *dsc, const struct dsc_config *dsc_cfg);
+void dsc401_set_config(struct display_stream_compressor *dsc, const struct dsc_config *dsc_cfg,
+ struct dsc_optc_config *dsc_optc_cfg);
+void dsc401_enable(struct display_stream_compressor *dsc, int opp_pipe);
+void dsc401_disable(struct display_stream_compressor *dsc);
+void dsc401_disconnect(struct display_stream_compressor *dsc);
+void dsc401_wait_disconnect_pending_clear(struct display_stream_compressor *dsc);
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dsc/dsc.h b/drivers/gpu/drm/amd/display/dc/dsc/dsc.h
index 1ebce5426a58..b0bd1f9425b5 100644
--- a/drivers/gpu/drm/amd/display/dc/dsc/dsc.h
+++ b/drivers/gpu/drm/amd/display/dc/dsc/dsc.h
@@ -108,6 +108,7 @@ struct dsc_funcs {
void (*dsc_disable)(struct display_stream_compressor *dsc);
void (*dsc_disconnect)(struct display_stream_compressor *dsc);
void (*dsc_wait_disconnect_pending_clear)(struct display_stream_compressor *dsc);
+ void (*dsc_get_single_enc_caps)(struct dsc_enc_caps *dsc_enc_caps, unsigned int max_dscclk_khz);
};
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.c b/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.c
index fae98cf52020..bc058f682438 100644
--- a/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.c
+++ b/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.c
@@ -270,16 +270,3 @@ void dcn30_dwbc_construct(struct dcn30_dwbc *dwbc30,
dwbc30->dwbc_shift = dwbc_shift;
dwbc30->dwbc_mask = dwbc_mask;
}
-
-void dwb3_set_host_read_rate_control(struct dwbc *dwbc, bool host_read_delay)
-{
- struct dcn30_dwbc *dwbc30 = TO_DCN30_DWBC(dwbc);
-
- /*
- * Set maximum delay of host read access to DWBSCL LUT or OGAM LUT if there are no
- * idle cycles in HW pipeline (in number of clock cycles times 4)
- */
- REG_UPDATE(DWB_HOST_READ_CONTROL, DWB_HOST_READ_RATE_CONTROL, host_read_delay);
-
- DC_LOG_DWB("%s dwb3_rate_control at inst = %d", __func__, dwbc->inst);
-}
diff --git a/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.h b/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.h
index 0f3f7c5fbaec..7f053f49ec6a 100644
--- a/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.h
+++ b/drivers/gpu/drm/amd/display/dc/dwb/dcn30/dcn30_dwb.h
@@ -914,7 +914,6 @@ bool dwb3_ogam_set_input_transfer_func(
struct dwbc *dwbc,
const struct dc_transfer_func *in_transfer_func_dwb_ogam);
-void dwb3_set_host_read_rate_control(struct dwbc *dwbc, bool host_read_delay);
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c
index b099989d9364..942d9f0b6df2 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/gpio_service.c
@@ -411,6 +411,20 @@ enum dc_irq_source dal_irq_get_rx_source(
}
}
+enum dc_irq_source dal_irq_get_read_request(
+ const struct gpio *irq)
+{
+ enum gpio_id id = dal_gpio_get_id(irq);
+
+ switch (id) {
+ case GPIO_ID_HPD:
+ return (enum dc_irq_source)(DC_IRQ_SOURCE_DCI2C_RR_DDC1 +
+ dal_gpio_get_enum(irq));
+ default:
+ return DC_IRQ_SOURCE_INVALID;
+ }
+}
+
enum gpio_result dal_irq_setup_hpd_filter(
struct gpio *irq,
struct gpio_hpd_config *config)
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
index 9a0952f9004f..8bc67ca42197 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_factory.c
@@ -112,6 +112,7 @@ bool dal_hw_factory_init(
case DCN_VERSION_3_21:
case DCN_VERSION_3_5:
case DCN_VERSION_3_51:
+ case DCN_VERSION_3_6:
dal_hw_factory_dcn32_init(factory);
return true;
case DCN_VERSION_4_01:
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_hpd.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_hpd.c
index 3f13a744d07d..01ec451004f7 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/hw_hpd.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_hpd.c
@@ -62,7 +62,7 @@ static void dal_hw_hpd_destroy(
*ptr = NULL;
}
-static enum gpio_result get_value(
+static enum gpio_result dal_hw_hpd_get_value(
const struct hw_gpio_pin *ptr,
uint32_t *value)
{
@@ -85,7 +85,7 @@ static enum gpio_result get_value(
return dal_hw_gpio_get_value(ptr, value);
}
-static enum gpio_result set_config(
+static enum gpio_result dal_hw_hpd_set_config(
struct hw_gpio_pin *ptr,
const struct gpio_config_data *config_data)
{
@@ -104,9 +104,9 @@ static enum gpio_result set_config(
static const struct hw_gpio_pin_funcs funcs = {
.destroy = dal_hw_hpd_destroy,
.open = dal_hw_gpio_open,
- .get_value = get_value,
+ .get_value = dal_hw_hpd_get_value,
.set_value = dal_hw_gpio_set_value,
- .set_config = set_config,
+ .set_config = dal_hw_hpd_set_config,
.change_mode = dal_hw_gpio_change_mode,
.close = dal_hw_gpio_close,
};
diff --git a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
index 9832247ee739..cb79a2832287 100644
--- a/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
+++ b/drivers/gpu/drm/amd/display/dc/gpio/hw_translate.c
@@ -113,6 +113,7 @@ bool dal_hw_translate_init(
case DCN_VERSION_3_21:
case DCN_VERSION_3_5:
case DCN_VERSION_3_51:
+ case DCN_VERSION_3_6:
dal_hw_translate_dcn32_init(translate);
return true;
case DCN_VERSION_4_01:
diff --git a/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.c b/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.c
index 03b4ac2f1991..0d2ae21abbdd 100644
--- a/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.c
@@ -262,7 +262,7 @@ void dcn31_hpo_dp_link_enc_set_link_test_pattern(
}
}
-static void fill_stream_allocation_row_info(
+void dcn31_fill_stream_allocation_row_info(
const struct link_mst_stream_allocation *stream_allocation,
uint32_t *src,
uint32_t *slots)
@@ -296,7 +296,7 @@ void dcn31_hpo_dp_link_enc_update_stream_allocation_table(
/* we should clean-up table each time */
if (table->stream_count >= 1) {
- fill_stream_allocation_row_info(
+ dcn31_fill_stream_allocation_row_info(
&table->stream_allocations[0],
&src,
&slots);
@@ -310,7 +310,7 @@ void dcn31_hpo_dp_link_enc_update_stream_allocation_table(
SAT_SLOT_COUNT, slots);
if (table->stream_count >= 2) {
- fill_stream_allocation_row_info(
+ dcn31_fill_stream_allocation_row_info(
&table->stream_allocations[1],
&src,
&slots);
@@ -324,7 +324,7 @@ void dcn31_hpo_dp_link_enc_update_stream_allocation_table(
SAT_SLOT_COUNT, slots);
if (table->stream_count >= 3) {
- fill_stream_allocation_row_info(
+ dcn31_fill_stream_allocation_row_info(
&table->stream_allocations[2],
&src,
&slots);
@@ -338,7 +338,7 @@ void dcn31_hpo_dp_link_enc_update_stream_allocation_table(
SAT_SLOT_COUNT, slots);
if (table->stream_count >= 4) {
- fill_stream_allocation_row_info(
+ dcn31_fill_stream_allocation_row_info(
&table->stream_allocations[3],
&src,
&slots);
diff --git a/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.h b/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.h
index 51f5781325e8..40859660e4dc 100644
--- a/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_link_encoder.h
@@ -226,4 +226,10 @@ void dcn31_hpo_dp_link_enc_set_ffe(
const struct dc_link_settings *link_settings,
uint8_t ffe_preset);
+
+void dcn31_fill_stream_allocation_row_info(
+ const struct link_mst_stream_allocation *stream_allocation,
+ uint32_t *src,
+ uint32_t *slots);
+
#endif // __DAL_DCN31_HPO_LINK_ENCODER_H__
diff --git a/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_stream_encoder.c b/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_stream_encoder.c
index 678db949cfe3..759b453385c4 100644
--- a/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_stream_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/hpo/dcn31/dcn31_hpo_dp_stream_encoder.c
@@ -323,7 +323,7 @@ static void dcn31_hpo_dp_stream_enc_set_stream_attribute(
break;
case COLOR_SPACE_2020_RGB_LIMITEDRANGE:
case COLOR_SPACE_2020_RGB_FULLRANGE:
- case COLOR_SPACE_2020_YCBCR:
+ case COLOR_SPACE_2020_YCBCR_LIMITED:
case COLOR_SPACE_XR_RGB:
case COLOR_SPACE_MSREF_SCRGB:
case COLOR_SPACE_ADOBERGB:
@@ -336,6 +336,7 @@ static void dcn31_hpo_dp_stream_enc_set_stream_attribute(
case COLOR_SPACE_CUSTOMPOINTS:
case COLOR_SPACE_UNKNOWN:
case COLOR_SPACE_YCBCR709_BLACK:
+ default:
/* do nothing */
break;
}
diff --git a/drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.c b/drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.c
index 8af01f579690..de3ec4fcade2 100644
--- a/drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.c
+++ b/drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.c
@@ -41,7 +41,7 @@
#define CTX \
enc3->base.ctx
-static bool dcn32_hpo_dp_link_enc_is_in_alt_mode(
+bool dcn32_hpo_dp_link_enc_is_in_alt_mode(
struct hpo_dp_link_encoder *enc)
{
struct dcn31_hpo_dp_link_encoder *enc3 = DCN3_1_HPO_DP_LINK_ENC_FROM_HPO_LINK_ENC(enc);
diff --git a/drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.h b/drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.h
index 176b1537d2a1..bea4e1a8ff90 100644
--- a/drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/hpo/dcn32/dcn32_hpo_dp_link_encoder.h
@@ -54,6 +54,7 @@
SE_SF(DP_DPHY_SYM320_DP_DPHY_SYM32_VC_RATE_CNTL0, STREAM_VC_RATE_Y, mask_sh),\
SE_SF(DP_DPHY_SYM320_DP_DPHY_SYM32_SAT_UPDATE, SAT_UPDATE, mask_sh)
+bool dcn32_hpo_dp_link_enc_is_in_alt_mode(struct hpo_dp_link_encoder *enc);
void hpo_dp_link_encoder32_construct(struct dcn31_hpo_dp_link_encoder *enc31,
struct dc_context *ctx,
uint32_t inst,
@@ -61,4 +62,7 @@ void hpo_dp_link_encoder32_construct(struct dcn31_hpo_dp_link_encoder *enc31,
const struct dcn31_hpo_dp_link_encoder_shift *hpo_le_shift,
const struct dcn31_hpo_dp_link_encoder_mask *hpo_le_mask);
+bool dcn32_hpo_dp_link_enc_is_in_alt_mode(
+ struct hpo_dp_link_encoder *enc);
+
#endif // __DAL_DCN32_HPO_DP_LINK_ENCODER_H__
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.c b/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.c
index d738a36f2132..7847c1c4927b 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.c
@@ -679,24 +679,6 @@ void hubbub1_update_dchub(
dh_data->dchub_info_valid = false;
}
-void hubbub1_toggle_watermark_change_req(struct hubbub *hubbub)
-{
- struct dcn10_hubbub *hubbub1 = TO_DCN10_HUBBUB(hubbub);
-
- uint32_t watermark_change_req;
-
- REG_GET(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL,
- DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, &watermark_change_req);
-
- if (watermark_change_req)
- watermark_change_req = 0;
- else
- watermark_change_req = 1;
-
- REG_UPDATE(DCHUBBUB_ARB_WATERMARK_CHANGE_CNTL,
- DCHUBBUB_ARB_WATERMARK_CHANGE_REQUEST, watermark_change_req);
-}
-
void hubbub1_soft_reset(struct hubbub *hubbub, bool reset)
{
struct dcn10_hubbub *hubbub1 = TO_DCN10_HUBBUB(hubbub);
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.h b/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.h
index 9fbd45c7dfef..fa5c4c18ed59 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn10/dcn10_hubbub.h
@@ -489,9 +489,6 @@ void hubbub1_allow_self_refresh_control(struct hubbub *hubbub, bool allow);
bool hubbub1_is_allow_self_refresh_enabled(struct hubbub *hubub);
-void hubbub1_toggle_watermark_change_req(
- struct hubbub *hubbub);
-
void hubbub1_wm_read_state(struct hubbub *hubbub,
struct dcn_hubbub_wm *wm);
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn21/dcn21_hubbub.c b/drivers/gpu/drm/amd/display/dc/hubbub/dcn21/dcn21_hubbub.c
index 2546224b326a..e4496ad203b2 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn21/dcn21_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn21/dcn21_hubbub.c
@@ -132,9 +132,9 @@ int hubbub21_init_dchub(struct hubbub *hubbub,
// Init VMID 0 based on PA config
dcn20_vmid_setup(&hubbub1->vmid[0], &phys_config);
}
-
- dcn21_dchvm_init(hubbub);
-
+ if (!hubbub1->base.ctx->dc->config.skip_riommu_prefetch_wa) {
+ dcn21_dchvm_init(hubbub);
+ }
return hubbub1->num_vmid;
}
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn30/dcn30_hubbub.c b/drivers/gpu/drm/amd/display/dc/hubbub/dcn30/dcn30_hubbub.c
index fe741100c0f8..d347bb06577a 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn30/dcn30_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn30/dcn30_hubbub.c
@@ -129,7 +129,8 @@ bool hubbub3_program_watermarks(
REG_UPDATE(DCHUBBUB_ARB_DF_REQ_OUTSTAND,
DCHUBBUB_ARB_MIN_REQ_OUTSTAND, 0x1FF);
- hubbub1_allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
+ if (safe_to_lower || hubbub->ctx->dc->debug.disable_stutter)
+ hubbub1_allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
return wm_pending;
}
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c b/drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c
index 7fb5523f9722..b98505b240a7 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn31/dcn31_hubbub.c
@@ -750,7 +750,8 @@ static bool hubbub31_program_watermarks(
REG_UPDATE(DCHUBBUB_ARB_DF_REQ_OUTSTAND,
DCHUBBUB_ARB_MIN_REQ_OUTSTAND, 0x1FF);*/
- hubbub1_allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
+ if (safe_to_lower || hubbub->ctx->dc->debug.disable_stutter)
+ hubbub1_allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
return wm_pending;
}
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn32/dcn32_hubbub.c b/drivers/gpu/drm/amd/display/dc/hubbub/dcn32/dcn32_hubbub.c
index 5264dc26cce1..32a6be543105 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn32/dcn32_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn32/dcn32_hubbub.c
@@ -786,7 +786,8 @@ static bool hubbub32_program_watermarks(
REG_UPDATE(DCHUBBUB_ARB_DF_REQ_OUTSTAND,
DCHUBBUB_ARB_MIN_REQ_OUTSTAND, 0x1FF);*/
- hubbub1_allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
+ if (safe_to_lower || hubbub->ctx->dc->debug.disable_stutter)
+ hubbub1_allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
hubbub32_force_usr_retraining_allow(hubbub, hubbub->ctx->dc->debug.force_usr_allow);
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c b/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c
index 5eb3da8d5206..6d41953011f5 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.c
@@ -46,7 +46,7 @@
#define DCN35_CRB_SEGMENT_SIZE_KB 64
-static void dcn35_init_crb(struct hubbub *hubbub)
+void dcn35_init_crb(struct hubbub *hubbub)
{
struct dcn20_hubbub *hubbub2 = TO_DCN20_HUBBUB(hubbub);
@@ -71,7 +71,7 @@ static void dcn35_init_crb(struct hubbub *hubbub)
REG_UPDATE(DCHUBBUB_DEBUG_CTRL_0, DET_DEPTH, 0x5FF);
}
-static void dcn35_program_compbuf_size(struct hubbub *hubbub, unsigned int compbuf_size_kb, bool safe_to_increase)
+void dcn35_program_compbuf_size(struct hubbub *hubbub, unsigned int compbuf_size_kb, bool safe_to_increase)
{
struct dcn20_hubbub *hubbub2 = TO_DCN20_HUBBUB(hubbub);
unsigned int compbuf_size_segments = (compbuf_size_kb + DCN35_CRB_SEGMENT_SIZE_KB - 1) / DCN35_CRB_SEGMENT_SIZE_KB;
@@ -255,7 +255,7 @@ static bool hubbub35_program_stutter_z8_watermarks(
return wm_pending;
}
-static void hubbub35_get_dchub_ref_freq(struct hubbub *hubbub,
+void hubbub35_get_dchub_ref_freq(struct hubbub *hubbub,
unsigned int dccg_ref_freq_inKhz,
unsigned int *dchub_ref_freq_inKhz)
{
@@ -295,7 +295,7 @@ static void hubbub35_get_dchub_ref_freq(struct hubbub *hubbub,
}
-static bool hubbub35_program_watermarks(
+bool hubbub35_program_watermarks(
struct hubbub *hubbub,
union dcn_watermark_set *watermarks,
unsigned int refclk_mhz,
@@ -326,7 +326,8 @@ static bool hubbub35_program_watermarks(
DCHUBBUB_ARB_MIN_REQ_OUTSTAND_COMMIT_THRESHOLD, 0xA);/*hw delta*/
REG_UPDATE(DCHUBBUB_ARB_HOSTVM_CNTL, DCHUBBUB_ARB_MAX_QOS_COMMIT_THRESHOLD, 0xF);
- hubbub1_allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
+ if (safe_to_lower || hubbub->ctx->dc->debug.disable_stutter)
+ hubbub1_allow_self_refresh_control(hubbub, !hubbub->ctx->dc->debug.disable_stutter);
hubbub32_force_usr_retraining_allow(hubbub, hubbub->ctx->dc->debug.force_usr_allow);
@@ -334,7 +335,7 @@ static bool hubbub35_program_watermarks(
}
/* Copy values from WM set A to all other sets */
-static void hubbub35_init_watermarks(struct hubbub *hubbub)
+void hubbub35_init_watermarks(struct hubbub *hubbub)
{
struct dcn20_hubbub *hubbub2 = TO_DCN20_HUBBUB(hubbub);
uint32_t reg;
@@ -396,7 +397,7 @@ static void hubbub35_init_watermarks(struct hubbub *hubbub)
}
-static void hubbub35_wm_read_state(struct hubbub *hubbub,
+void hubbub35_wm_read_state(struct hubbub *hubbub,
struct dcn_hubbub_wm *wm)
{
struct dcn20_hubbub *hubbub2 = TO_DCN20_HUBBUB(hubbub);
@@ -513,7 +514,7 @@ static void hubbub35_set_fgcg(struct dcn20_hubbub *hubbub2, bool enable)
REG_UPDATE(DCHUBBUB_CLOCK_CNTL, DCHUBBUB_FGCG_REP_DIS, !enable);
}
-static void hubbub35_init(struct hubbub *hubbub)
+void hubbub35_init(struct hubbub *hubbub)
{
struct dcn20_hubbub *hubbub2 = TO_DCN20_HUBBUB(hubbub);
/*Enable clock gaters*/
diff --git a/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.h b/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.h
index 54cf00ffceb8..23fecf88556c 100644
--- a/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.h
+++ b/drivers/gpu/drm/amd/display/dc/hubbub/dcn35/dcn35_hubbub.h
@@ -152,4 +152,20 @@ void hubbub35_construct(struct dcn20_hubbub *hubbub2,
int det_size_kb,
int pixel_chunk_size_kb,
int config_return_buffer_size_kb);
+
+void hubbub35_wm_read_state(struct hubbub *hubbub,
+ struct dcn_hubbub_wm *wm);
+void hubbub35_get_dchub_ref_freq(struct hubbub *hubbub,
+ unsigned int dccg_ref_freq_inKhz,
+ unsigned int *dchub_ref_freq_inKhz);
+bool hubbub35_program_watermarks(
+ struct hubbub *hubbub,
+ union dcn_watermark_set *watermarks,
+ unsigned int refclk_mhz,
+ bool safe_to_lower);
+void hubbub35_init_watermarks(struct hubbub *hubbub);
+void dcn35_program_compbuf_size(struct hubbub *hubbub,
+ unsigned int compbuf_size_kb, bool safe_to_increase);
+void dcn35_init_crb(struct hubbub *hubbub);
+void hubbub35_init(struct hubbub *hubbub);
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.c b/drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.c
index 22ac2b7e49ae..9b026600b90e 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.c
@@ -140,7 +140,7 @@ void hubp1_vready_workaround(struct hubp *hubp,
void hubp1_program_tiling(
struct hubp *hubp,
- const union dc_tiling_info *info,
+ const struct dc_tiling_info *info,
const enum surface_pixel_format pixel_format)
{
struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
@@ -518,6 +518,20 @@ bool hubp1_program_surface_flip_and_addr(
return true;
}
+void hubp1_clear_tiling(struct hubp *hubp)
+{
+ struct dcn10_hubp *hubp1 = TO_DCN10_HUBP(hubp);
+
+ REG_UPDATE(DCHUBP_REQ_SIZE_CONFIG, SWATH_HEIGHT, 0);
+ REG_UPDATE(DCSURF_TILING_CONFIG, SW_MODE, DC_SW_LINEAR);
+
+ REG_UPDATE_4(DCSURF_SURFACE_CONTROL,
+ PRIMARY_SURFACE_DCC_EN, 0,
+ PRIMARY_SURFACE_DCC_IND_64B_BLK, 0,
+ SECONDARY_SURFACE_DCC_EN, 0,
+ SECONDARY_SURFACE_DCC_IND_64B_BLK, 0);
+}
+
void hubp1_dcc_control(struct hubp *hubp, bool enable,
enum hubp_ind_block_size independent_64b_blks)
{
@@ -532,10 +546,16 @@ void hubp1_dcc_control(struct hubp *hubp, bool enable,
SECONDARY_SURFACE_DCC_IND_64B_BLK, dcc_ind_64b_blk);
}
+void hubp_reset(struct hubp *hubp)
+{
+ memset(&hubp->pos, 0, sizeof(hubp->pos));
+ memset(&hubp->att, 0, sizeof(hubp->att));
+}
+
void hubp1_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -1337,8 +1357,9 @@ static void hubp1_wait_pipe_read_start(struct hubp *hubp)
void hubp1_init(struct hubp *hubp)
{
- //do nothing
+ hubp_reset(hubp);
}
+
static const struct hubp_funcs dcn10_hubp_funcs = {
.hubp_program_surface_flip_and_addr =
hubp1_program_surface_flip_and_addr,
@@ -1351,6 +1372,7 @@ static const struct hubp_funcs dcn10_hubp_funcs = {
.hubp_set_vm_context0_settings = hubp1_set_vm_context0_settings,
.set_blank = hubp1_set_blank,
.dcc_control = hubp1_dcc_control,
+ .hubp_reset = hubp_reset,
.mem_program_viewport = min_set_viewport,
.set_hubp_blank_en = hubp1_set_hubp_blank_en,
.set_cursor_attributes = hubp1_cursor_set_attributes,
@@ -1363,6 +1385,7 @@ static const struct hubp_funcs dcn10_hubp_funcs = {
.hubp_disable_control = hubp1_disable_control,
.hubp_get_underflow_status = hubp1_get_underflow_status,
.hubp_init = hubp1_init,
+ .hubp_clear_tiling = hubp1_clear_tiling,
.dmdata_set_attributes = NULL,
.dmdata_load = NULL,
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.h b/drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.h
index 69119b2fdce2..f8f991785d4f 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn10/dcn10_hubp.h
@@ -666,10 +666,29 @@ struct dcn_mi_mask {
DCN_HUBP_REG_FIELD_LIST(uint32_t);
};
+struct dcn_fl_regs_st {
+ uint32_t lut_enable;
+ uint32_t lut_done;
+ uint32_t lut_addr_mode;
+ uint32_t lut_width;
+ uint32_t lut_tmz;
+ uint32_t lut_crossbar_sel_r;
+ uint32_t lut_crossbar_sel_g;
+ uint32_t lut_crossbar_sel_b;
+ uint32_t lut_addr_hi;
+ uint32_t lut_addr_lo;
+ uint32_t refcyc_3dlut_group;
+ uint32_t lut_fl_bias;
+ uint32_t lut_fl_scale;
+ uint32_t lut_fl_mode;
+ uint32_t lut_fl_format;
+};
+
struct dcn_hubp_state {
struct _vcs_dpi_display_dlg_regs_st dlg_attr;
struct _vcs_dpi_display_ttu_regs_st ttu_attr;
struct _vcs_dpi_display_rq_regs_st rq_regs;
+ struct dcn_fl_regs_st fl_regs;
uint32_t pixel_format;
uint32_t inuse_addr_hi;
uint32_t inuse_addr_lo;
@@ -706,7 +725,7 @@ struct dcn10_hubp {
void hubp1_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -739,13 +758,15 @@ void hubp1_program_rotation(
void hubp1_program_tiling(
struct hubp *hubp,
- const union dc_tiling_info *info,
+ const struct dc_tiling_info *info,
const enum surface_pixel_format pixel_format);
void hubp1_dcc_control(struct hubp *hubp,
bool enable,
enum hubp_ind_block_size independent_64b_blks);
+void hubp_reset(struct hubp *hubp);
+
bool hubp1_program_surface_flip_and_addr(
struct hubp *hubp,
const struct dc_plane_address *address,
@@ -794,4 +815,6 @@ void hubp1_soft_reset(struct hubp *hubp, bool reset);
void hubp1_set_flip_int(struct hubp *hubp);
+void hubp1_clear_tiling(struct hubp *hubp);
+
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.c b/drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.c
index 0637e4c552d8..91259b896e03 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.c
@@ -310,7 +310,7 @@ void hubp2_setup_interdependent(
*/
static void hubp2_program_tiling(
struct dcn20_hubp *hubp2,
- const union dc_tiling_info *info,
+ const struct dc_tiling_info *info,
const enum surface_pixel_format pixel_format)
{
REG_UPDATE_3(DCSURF_ADDR_CONFIG,
@@ -406,6 +406,20 @@ void hubp2_program_rotation(
H_MIRROR_EN, mirror);
}
+void hubp2_clear_tiling(struct hubp *hubp)
+{
+ struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
+
+ REG_UPDATE(DCHUBP_REQ_SIZE_CONFIG, SWATH_HEIGHT, 0);
+ REG_UPDATE(DCSURF_TILING_CONFIG, SW_MODE, DC_SW_LINEAR);
+
+ REG_UPDATE_4(DCSURF_SURFACE_CONTROL,
+ PRIMARY_SURFACE_DCC_EN, 0,
+ PRIMARY_SURFACE_DCC_IND_64B_BLK, 0,
+ SECONDARY_SURFACE_DCC_EN, 0,
+ SECONDARY_SURFACE_DCC_IND_64B_BLK, 0);
+}
+
void hubp2_dcc_control(struct hubp *hubp, bool enable,
enum hubp_ind_block_size independent_64b_blks)
{
@@ -536,7 +550,7 @@ void hubp2_program_pixel_format(
void hubp2_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -1044,11 +1058,13 @@ void hubp2_cursor_set_position(
if (src_y_offset + cursor_height <= 0)
cur_en = 0; /* not visible beyond top edge*/
- if (cur_en && REG_READ(CURSOR_SURFACE_ADDRESS) == 0)
- hubp->funcs->set_cursor_attributes(hubp, &hubp->curs_attr);
+ if (hubp->pos.cur_ctl.bits.cur_enable != cur_en) {
+ if (cur_en && REG_READ(CURSOR_SURFACE_ADDRESS) == 0)
+ hubp->funcs->set_cursor_attributes(hubp, &hubp->curs_attr);
- REG_UPDATE(CURSOR_CONTROL,
+ REG_UPDATE(CURSOR_CONTROL,
CURSOR_ENABLE, cur_en);
+ }
REG_SET_2(CURSOR_POSITION, 0,
CURSOR_X_POSITION, pos->x,
@@ -1660,6 +1676,7 @@ static struct hubp_funcs dcn20_hubp_funcs = {
.set_blank = hubp2_set_blank,
.set_blank_regs = hubp2_set_blank_regs,
.dcc_control = hubp2_dcc_control,
+ .hubp_reset = hubp_reset,
.mem_program_viewport = min_set_viewport,
.set_cursor_attributes = hubp2_cursor_set_attributes,
.set_cursor_position = hubp2_cursor_set_position,
@@ -1676,6 +1693,7 @@ static struct hubp_funcs dcn20_hubp_funcs = {
.hubp_in_blank = hubp1_in_blank,
.hubp_soft_reset = hubp1_soft_reset,
.hubp_set_flip_int = hubp1_set_flip_int,
+ .hubp_clear_tiling = hubp2_clear_tiling,
};
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.h b/drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.h
index 18e194507e36..62369be070ea 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn20/dcn20_hubp.h
@@ -280,9 +280,8 @@
type MCACHEID_MALL_PREF_1H_P0;\
type MCACHEID_MALL_PREF_2H_P0;\
type MCACHEID_MALL_PREF_1H_P1;\
- type MCACHEID_MALL_PREF_2H_P1
-
-
+ type MCACHEID_MALL_PREF_2H_P1;\
+ type HUBP_FGCG_REP_DIS
struct dcn_hubp2_registers {
DCN401_HUBP_REG_COMMON_VARIABLE_LIST;
@@ -382,7 +381,7 @@ void hubp2_program_pixel_format(
void hubp2_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -409,6 +408,8 @@ void hubp2_read_state_common(struct hubp *hubp);
void hubp2_read_state(struct hubp *hubp);
+void hubp2_clear_tiling(struct hubp *hubp);
+
#endif /* __DC_MEM_INPUT_DCN20_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn201/dcn201_hubp.c b/drivers/gpu/drm/amd/display/dc/hubp/dcn201/dcn201_hubp.c
index cd2bfcc51276..ec88ee424a7f 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn201/dcn201_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn201/dcn201_hubp.c
@@ -42,7 +42,7 @@
static void hubp201_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -121,6 +121,7 @@ static struct hubp_funcs dcn201_hubp_funcs = {
.set_cursor_position = hubp1_cursor_set_position,
.set_blank = hubp1_set_blank,
.dcc_control = hubp1_dcc_control,
+ .hubp_reset = hubp_reset,
.mem_program_viewport = min_set_viewport,
.hubp_clk_cntl = hubp1_clk_cntl,
.hubp_vtg_sel = hubp1_vtg_sel,
@@ -131,6 +132,7 @@ static struct hubp_funcs dcn201_hubp_funcs = {
.hubp_clear_underflow = hubp1_clear_underflow,
.hubp_set_flip_control_surface_gsl = hubp2_set_flip_control_surface_gsl,
.hubp_init = hubp1_init,
+ .hubp_clear_tiling = hubp1_clear_tiling,
};
bool dcn201_hubp_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn21/dcn21_hubp.c b/drivers/gpu/drm/amd/display/dc/hubp/dcn21/dcn21_hubp.c
index e13d69a22c1c..e2740482e1cf 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn21/dcn21_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn21/dcn21_hubp.c
@@ -811,6 +811,8 @@ static void hubp21_init(struct hubp *hubp)
struct dcn21_hubp *hubp21 = TO_DCN21_HUBP(hubp);
//hubp[i].HUBPREQ_DEBUG.HUBPREQ_DEBUG[26] = 1;
REG_WRITE(HUBPREQ_DEBUG, 1 << 26);
+
+ hubp_reset(hubp);
}
static struct hubp_funcs dcn21_hubp_funcs = {
.hubp_enable_tripleBuffer = hubp2_enable_triplebuffer,
@@ -823,6 +825,7 @@ static struct hubp_funcs dcn21_hubp_funcs = {
.hubp_set_vm_system_aperture_settings = hubp21_set_vm_system_aperture_settings,
.set_blank = hubp1_set_blank,
.dcc_control = hubp1_dcc_control,
+ .hubp_reset = hubp_reset,
.mem_program_viewport = hubp21_set_viewport,
.set_cursor_attributes = hubp2_cursor_set_attributes,
.set_cursor_position = hubp1_cursor_set_position,
@@ -837,6 +840,7 @@ static struct hubp_funcs dcn21_hubp_funcs = {
.hubp_init = hubp21_init,
.validate_dml_output = hubp21_validate_dml_output,
.hubp_set_flip_int = hubp1_set_flip_int,
+ .hubp_clear_tiling = hubp1_clear_tiling,
};
bool hubp21_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.c b/drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.c
index 60a64d290352..0da70b50e86d 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.c
@@ -318,7 +318,7 @@ bool hubp3_program_surface_flip_and_addr(
void hubp3_program_tiling(
struct dcn20_hubp *hubp2,
- const union dc_tiling_info *info,
+ const struct dc_tiling_info *info,
const enum surface_pixel_format pixel_format)
{
REG_UPDATE_4(DCSURF_ADDR_CONFIG,
@@ -334,6 +334,22 @@ void hubp3_program_tiling(
}
+void hubp3_clear_tiling(struct hubp *hubp)
+{
+ struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
+
+ REG_UPDATE(DCHUBP_REQ_SIZE_CONFIG, SWATH_HEIGHT, 0);
+ REG_UPDATE(DCSURF_TILING_CONFIG, SW_MODE, DC_SW_LINEAR);
+
+ REG_UPDATE_6(DCSURF_SURFACE_CONTROL,
+ PRIMARY_SURFACE_DCC_EN, 0,
+ PRIMARY_SURFACE_DCC_IND_BLK, 0,
+ PRIMARY_SURFACE_DCC_IND_BLK_C, 0,
+ SECONDARY_SURFACE_DCC_EN, 0,
+ SECONDARY_SURFACE_DCC_IND_BLK, 0,
+ SECONDARY_SURFACE_DCC_IND_BLK_C, 0);
+}
+
void hubp3_dcc_control(struct hubp *hubp, bool enable,
enum hubp_ind_block_size blk_size)
{
@@ -395,7 +411,7 @@ void hubp3_dmdata_set_attributes(
void hubp3_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -483,6 +499,10 @@ void hubp3_init(struct hubp *hubp)
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
//hubp[i].HUBPREQ_DEBUG.HUBPREQ_DEBUG[26] = 1;
REG_WRITE(HUBPREQ_DEBUG, 1 << 26);
+
+ REG_UPDATE(DCHUBP_CNTL, HUBP_TTU_DISABLE, 0);
+
+ hubp_reset(hubp);
}
static struct hubp_funcs dcn30_hubp_funcs = {
@@ -497,6 +517,7 @@ static struct hubp_funcs dcn30_hubp_funcs = {
.set_blank = hubp2_set_blank,
.set_blank_regs = hubp2_set_blank_regs,
.dcc_control = hubp3_dcc_control,
+ .hubp_reset = hubp_reset,
.mem_program_viewport = min_set_viewport,
.set_cursor_attributes = hubp2_cursor_set_attributes,
.set_cursor_position = hubp2_cursor_set_position,
@@ -512,6 +533,7 @@ static struct hubp_funcs dcn30_hubp_funcs = {
.hubp_in_blank = hubp1_in_blank,
.hubp_soft_reset = hubp1_soft_reset,
.hubp_set_flip_int = hubp1_set_flip_int,
+ .hubp_clear_tiling = hubp3_clear_tiling,
};
bool hubp3_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.h b/drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.h
index b010531a7fe8..b7d7adf0b58c 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn30/dcn30_hubp.h
@@ -264,7 +264,7 @@ bool hubp3_program_surface_flip_and_addr(
void hubp3_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -280,7 +280,7 @@ void hubp3_setup(
void hubp3_program_tiling(
struct dcn20_hubp *hubp2,
- const union dc_tiling_info *info,
+ const struct dc_tiling_info *info,
const enum surface_pixel_format pixel_format);
void hubp3_dcc_control(struct hubp *hubp, bool enable,
@@ -297,6 +297,8 @@ void hubp3_read_state(struct hubp *hubp);
void hubp3_init(struct hubp *hubp);
+void hubp3_clear_tiling(struct hubp *hubp);
+
#endif /* __DC_HUBP_DCN30_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn31/dcn31_hubp.c b/drivers/gpu/drm/amd/display/dc/hubp/dcn31/dcn31_hubp.c
index 8394e8c06919..7fd582a8a4ba 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn31/dcn31_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn31/dcn31_hubp.c
@@ -44,7 +44,7 @@ void hubp31_set_unbounded_requesting(struct hubp *hubp, bool enable)
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
REG_UPDATE(DCHUBP_CNTL, HUBP_UNBOUNDED_REQ_MODE, enable);
- REG_UPDATE(CURSOR_CONTROL, CURSOR_REQ_MODE, enable);
+ REG_UPDATE(CURSOR_CONTROL, CURSOR_REQ_MODE, 1);
}
void hubp31_soft_reset(struct hubp *hubp, bool reset)
@@ -79,6 +79,7 @@ static struct hubp_funcs dcn31_hubp_funcs = {
.hubp_set_vm_system_aperture_settings = hubp3_set_vm_system_aperture_settings,
.set_blank = hubp2_set_blank,
.dcc_control = hubp3_dcc_control,
+ .hubp_reset = hubp_reset,
.mem_program_viewport = min_set_viewport,
.set_cursor_attributes = hubp2_cursor_set_attributes,
.set_cursor_position = hubp2_cursor_set_position,
@@ -96,6 +97,7 @@ static struct hubp_funcs dcn31_hubp_funcs = {
.hubp_set_flip_int = hubp1_set_flip_int,
.hubp_in_blank = hubp1_in_blank,
.program_extended_blank = hubp31_program_extended_blank,
+ .hubp_clear_tiling = hubp3_clear_tiling,
};
bool hubp31_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn32/dcn32_hubp.c b/drivers/gpu/drm/amd/display/dc/hubp/dcn32/dcn32_hubp.c
index ca5b4b28a664..f3a21c623f44 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn32/dcn32_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn32/dcn32_hubp.c
@@ -168,6 +168,8 @@ void hubp32_init(struct hubp *hubp)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
REG_WRITE(HUBPREQ_DEBUG_DB, 1 << 8);
+
+ REG_UPDATE(DCHUBP_CNTL, HUBP_TTU_DISABLE, 0);
}
static struct hubp_funcs dcn32_hubp_funcs = {
.hubp_enable_tripleBuffer = hubp2_enable_triplebuffer,
@@ -181,6 +183,7 @@ static struct hubp_funcs dcn32_hubp_funcs = {
.set_blank = hubp2_set_blank,
.set_blank_regs = hubp2_set_blank_regs,
.dcc_control = hubp3_dcc_control,
+ .hubp_reset = hubp_reset,
.mem_program_viewport = min_set_viewport,
.set_cursor_attributes = hubp32_cursor_set_attributes,
.set_cursor_position = hubp2_cursor_set_position,
@@ -201,7 +204,8 @@ static struct hubp_funcs dcn32_hubp_funcs = {
.hubp_update_force_cursor_pstate_disallow = hubp32_update_force_cursor_pstate_disallow,
.phantom_hubp_post_enable = hubp32_phantom_hubp_post_enable,
.hubp_update_mall_sel = hubp32_update_mall_sel,
- .hubp_prepare_subvp_buffering = hubp32_prepare_subvp_buffering
+ .hubp_prepare_subvp_buffering = hubp32_prepare_subvp_buffering,
+ .hubp_clear_tiling = hubp3_clear_tiling,
};
bool hubp32_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.c b/drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.c
index d1f05b82b3dd..6d060ba12da8 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.c
@@ -45,7 +45,7 @@ void hubp35_set_fgcg(struct hubp *hubp, bool enable)
REG_UPDATE(HUBP_CLK_CNTL, HUBP_FGCG_REP_DIS, !enable);
}
-static void hubp35_init(struct hubp *hubp)
+void hubp35_init(struct hubp *hubp)
{
hubp3_init(hubp);
@@ -172,7 +172,7 @@ void hubp35_program_pixel_format(
void hubp35_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -199,6 +199,7 @@ static struct hubp_funcs dcn35_hubp_funcs = {
.hubp_set_vm_system_aperture_settings = hubp3_set_vm_system_aperture_settings,
.set_blank = hubp2_set_blank,
.dcc_control = hubp3_dcc_control,
+ .hubp_reset = hubp_reset,
.mem_program_viewport = min_set_viewport,
.set_cursor_attributes = hubp2_cursor_set_attributes,
.set_cursor_position = hubp2_cursor_set_position,
@@ -216,6 +217,7 @@ static struct hubp_funcs dcn35_hubp_funcs = {
.hubp_set_flip_int = hubp1_set_flip_int,
.hubp_in_blank = hubp1_in_blank,
.program_extended_blank = hubp31_program_extended_blank_value,
+ .hubp_clear_tiling = hubp3_clear_tiling,
};
bool hubp35_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.h b/drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.h
index 586b43aa5834..934836717f32 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn35/dcn35_hubp.h
@@ -65,11 +65,12 @@ void hubp35_program_pixel_format(
void hubp35_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
bool horizontal_mirror,
unsigned int compat_level);
+void hubp35_init(struct hubp *hubp);
#endif /* __DC_HUBP_DCN35_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.c b/drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.c
index b1ebf5053b4f..705b98b1b6cc 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.c
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.c
@@ -40,7 +40,7 @@
#define FN(reg_name, field_name) \
hubp2->hubp_shift->field_name, hubp2->hubp_mask->field_name
-static void hubp401_program_3dlut_fl_addr(struct hubp *hubp,
+void hubp401_program_3dlut_fl_addr(struct hubp *hubp,
const struct dc_plane_address address)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
@@ -49,14 +49,14 @@ static void hubp401_program_3dlut_fl_addr(struct hubp *hubp,
REG_WRITE(HUBP_3DLUT_ADDRESS_LOW, address.lut3d.addr.low_part);
}
-static void hubp401_program_3dlut_fl_dlg_param(struct hubp *hubp, int refcyc_per_3dlut_group)
+void hubp401_program_3dlut_fl_dlg_param(struct hubp *hubp, int refcyc_per_3dlut_group)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
REG_UPDATE(HUBP_3DLUT_DLG_PARAM, REFCYC_PER_3DLUT_GROUP, refcyc_per_3dlut_group);
}
-static void hubp401_enable_3dlut_fl(struct hubp *hubp, bool enable)
+void hubp401_enable_3dlut_fl(struct hubp *hubp, bool enable)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
@@ -72,28 +72,28 @@ int hubp401_get_3dlut_fl_done(struct hubp *hubp)
return ret;
}
-static void hubp401_program_3dlut_fl_addressing_mode(struct hubp *hubp, enum hubp_3dlut_fl_addressing_mode addr_mode)
+void hubp401_program_3dlut_fl_addressing_mode(struct hubp *hubp, enum hubp_3dlut_fl_addressing_mode addr_mode)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
REG_UPDATE(HUBP_3DLUT_CONTROL, HUBP_3DLUT_ADDRESSING_MODE, addr_mode);
}
-static void hubp401_program_3dlut_fl_width(struct hubp *hubp, enum hubp_3dlut_fl_width width)
+void hubp401_program_3dlut_fl_width(struct hubp *hubp, enum hubp_3dlut_fl_width width)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
REG_UPDATE(HUBP_3DLUT_CONTROL, HUBP_3DLUT_WIDTH, width);
}
-static void hubp401_program_3dlut_fl_tmz_protected(struct hubp *hubp, bool protection_enabled)
+void hubp401_program_3dlut_fl_tmz_protected(struct hubp *hubp, uint8_t protection_bits)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
- REG_UPDATE(HUBP_3DLUT_CONTROL, HUBP_3DLUT_TMZ, protection_enabled ? 1 : 0);
+ REG_UPDATE(HUBP_3DLUT_CONTROL, HUBP_3DLUT_TMZ, protection_bits);
}
-static void hubp401_program_3dlut_fl_crossbar(struct hubp *hubp,
+void hubp401_program_3dlut_fl_crossbar(struct hubp *hubp,
enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_y_g,
enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_cb_b,
enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_cr_r)
@@ -106,21 +106,21 @@ static void hubp401_program_3dlut_fl_crossbar(struct hubp *hubp,
HUBP_3DLUT_CROSSBAR_SELECT_CR_R, bit_slice_cr_r);
}
-static void hubp401_update_3dlut_fl_bias_scale(struct hubp *hubp, uint16_t bias, uint16_t scale)
+void hubp401_update_3dlut_fl_bias_scale(struct hubp *hubp, uint16_t bias, uint16_t scale)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
REG_UPDATE_2(_3DLUT_FL_BIAS_SCALE, HUBP0_3DLUT_FL_BIAS, bias, HUBP0_3DLUT_FL_SCALE, scale);
}
-static void hubp401_program_3dlut_fl_mode(struct hubp *hubp, enum hubp_3dlut_fl_mode mode)
+void hubp401_program_3dlut_fl_mode(struct hubp *hubp, enum hubp_3dlut_fl_mode mode)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
REG_UPDATE(_3DLUT_FL_CONFIG, HUBP0_3DLUT_FL_MODE, mode);
}
-static void hubp401_program_3dlut_fl_format(struct hubp *hubp, enum hubp_3dlut_fl_format format)
+void hubp401_program_3dlut_fl_format(struct hubp *hubp, enum hubp_3dlut_fl_format format)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
@@ -141,34 +141,48 @@ void hubp401_update_mall_sel(struct hubp *hubp, uint32_t mall_sel, bool c_cursor
void hubp401_init(struct hubp *hubp)
{
- //For now nothing to do, HUBPREQ_DEBUG_DB register is removed on DCN4x.
+ hubp_reset(hubp);
}
void hubp401_vready_at_or_After_vsync(struct hubp *hubp,
- struct _vcs_dpi_display_pipe_dest_params_st *pipe_dest)
+ union dml2_global_sync_programming *pipe_global_sync,
+ struct dc_crtc_timing *timing)
{
- uint32_t value = 0;
+ unsigned int vstartup_lines = pipe_global_sync->dcn4x.vstartup_lines;
+ unsigned int vupdate_offset_pixels = pipe_global_sync->dcn4x.vupdate_offset_pixels;
+ unsigned int vupdate_width_pixels = pipe_global_sync->dcn4x.vupdate_vupdate_width_pixels;
+ unsigned int vready_offset_pixels = pipe_global_sync->dcn4x.vready_offset_pixels;
+ unsigned int htotal = timing->h_total;
+ unsigned int vblank_start = 0;
+ unsigned int vblank_end = 0;
+ unsigned int pixel_width = 0;
+ uint32_t reg_value = 0;
+ bool is_vready_at_or_after_vsync = false;
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
+
/*
* if (VSTARTUP_START - (VREADY_OFFSET+VUPDATE_WIDTH+VUPDATE_OFFSET)/htotal) <= OTG_V_BLANK_END
* Set HUBP_VREADY_AT_OR_AFTER_VSYNC = 1
* else
* Set HUBP_VREADY_AT_OR_AFTER_VSYNC = 0
*/
- if (pipe_dest->htotal != 0) {
- if ((pipe_dest->vstartup_start - (pipe_dest->vready_offset+pipe_dest->vupdate_width
- + pipe_dest->vupdate_offset) / pipe_dest->htotal) <= pipe_dest->vblank_end) {
- value = 1;
- } else
- value = 0;
+ if (htotal != 0) {
+ vblank_start = timing->v_total - timing->v_front_porch;
+ vblank_end = vblank_start - timing->v_addressable - timing->v_border_top - timing->v_border_bottom;
+ pixel_width = vready_offset_pixels + vupdate_width_pixels + vupdate_offset_pixels;
+
+ is_vready_at_or_after_vsync = (vstartup_lines - pixel_width / htotal) <= vblank_end;
+
+ if (is_vready_at_or_after_vsync)
+ reg_value = 1;
}
- REG_UPDATE(DCHUBP_CNTL, HUBP_VREADY_AT_OR_AFTER_VSYNC, value);
+ REG_UPDATE(DCHUBP_CNTL, HUBP_VREADY_AT_OR_AFTER_VSYNC, reg_value);
}
void hubp401_program_requestor(
struct hubp *hubp,
- struct _vcs_dpi_display_rq_regs_st *rq_regs)
+ struct dml2_display_rq_regs *rq_regs)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
@@ -196,8 +210,8 @@ void hubp401_program_requestor(
void hubp401_program_deadline(
struct hubp *hubp,
- struct _vcs_dpi_display_dlg_regs_st *dlg_attr,
- struct _vcs_dpi_display_ttu_regs_st *ttu_attr)
+ struct dml2_display_dlg_regs *dlg_attr,
+ struct dml2_display_ttu_regs *ttu_attr)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
@@ -294,66 +308,64 @@ void hubp401_program_deadline(
void hubp401_setup(
struct hubp *hubp,
- struct _vcs_dpi_display_dlg_regs_st *dlg_attr,
- struct _vcs_dpi_display_ttu_regs_st *ttu_attr,
- struct _vcs_dpi_display_rq_regs_st *rq_regs,
- struct _vcs_dpi_display_pipe_dest_params_st *pipe_dest)
+ struct dml2_dchub_per_pipe_register_set *pipe_regs,
+ union dml2_global_sync_programming *pipe_global_sync,
+ struct dc_crtc_timing *timing)
{
/* otg is locked when this func is called. Register are double buffered.
* disable the requestors is not needed
*/
- hubp401_vready_at_or_After_vsync(hubp, pipe_dest);
- hubp401_program_requestor(hubp, rq_regs);
- hubp401_program_deadline(hubp, dlg_attr, ttu_attr);
+ hubp401_vready_at_or_After_vsync(hubp, pipe_global_sync, timing);
+ hubp401_program_requestor(hubp, &pipe_regs->rq_regs);
+ hubp401_program_deadline(hubp, &pipe_regs->dlg_regs, &pipe_regs->ttu_regs);
}
void hubp401_setup_interdependent(
struct hubp *hubp,
- struct _vcs_dpi_display_dlg_regs_st *dlg_attr,
- struct _vcs_dpi_display_ttu_regs_st *ttu_attr)
+ struct dml2_dchub_per_pipe_register_set *pipe_regs)
{
struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
REG_SET_2(PREFETCH_SETTINGS, 0,
- DST_Y_PREFETCH, dlg_attr->dst_y_prefetch,
- VRATIO_PREFETCH, dlg_attr->vratio_prefetch);
+ DST_Y_PREFETCH, pipe_regs->dlg_regs.dst_y_prefetch,
+ VRATIO_PREFETCH, pipe_regs->dlg_regs.vratio_prefetch);
REG_SET(PREFETCH_SETTINGS_C, 0,
- VRATIO_PREFETCH_C, dlg_attr->vratio_prefetch_c);
+ VRATIO_PREFETCH_C, pipe_regs->dlg_regs.vratio_prefetch_c);
REG_SET_2(VBLANK_PARAMETERS_0, 0,
- DST_Y_PER_VM_VBLANK, dlg_attr->dst_y_per_vm_vblank,
- DST_Y_PER_ROW_VBLANK, dlg_attr->dst_y_per_row_vblank);
+ DST_Y_PER_VM_VBLANK, pipe_regs->dlg_regs.dst_y_per_vm_vblank,
+ DST_Y_PER_ROW_VBLANK, pipe_regs->dlg_regs.dst_y_per_row_vblank);
REG_SET_2(FLIP_PARAMETERS_0, 0,
- DST_Y_PER_VM_FLIP, dlg_attr->dst_y_per_vm_flip,
- DST_Y_PER_ROW_FLIP, dlg_attr->dst_y_per_row_flip);
+ DST_Y_PER_VM_FLIP, pipe_regs->dlg_regs.dst_y_per_vm_flip,
+ DST_Y_PER_ROW_FLIP, pipe_regs->dlg_regs.dst_y_per_row_flip);
REG_SET(VBLANK_PARAMETERS_3, 0,
- REFCYC_PER_META_CHUNK_VBLANK_L, dlg_attr->refcyc_per_meta_chunk_vblank_l);
+ REFCYC_PER_META_CHUNK_VBLANK_L, pipe_regs->dlg_regs.refcyc_per_meta_chunk_vblank_l);
REG_SET(VBLANK_PARAMETERS_4, 0,
- REFCYC_PER_META_CHUNK_VBLANK_C, dlg_attr->refcyc_per_meta_chunk_vblank_c);
+ REFCYC_PER_META_CHUNK_VBLANK_C, pipe_regs->dlg_regs.refcyc_per_meta_chunk_vblank_c);
REG_SET(FLIP_PARAMETERS_2, 0,
- REFCYC_PER_META_CHUNK_FLIP_L, dlg_attr->refcyc_per_meta_chunk_flip_l);
+ REFCYC_PER_META_CHUNK_FLIP_L, pipe_regs->dlg_regs.refcyc_per_meta_chunk_flip_l);
REG_SET_2(PER_LINE_DELIVERY_PRE, 0,
- REFCYC_PER_LINE_DELIVERY_PRE_L, dlg_attr->refcyc_per_line_delivery_pre_l,
- REFCYC_PER_LINE_DELIVERY_PRE_C, dlg_attr->refcyc_per_line_delivery_pre_c);
+ REFCYC_PER_LINE_DELIVERY_PRE_L, pipe_regs->dlg_regs.refcyc_per_line_delivery_pre_l,
+ REFCYC_PER_LINE_DELIVERY_PRE_C, pipe_regs->dlg_regs.refcyc_per_line_delivery_pre_c);
REG_SET(DCN_SURF0_TTU_CNTL1, 0,
REFCYC_PER_REQ_DELIVERY_PRE,
- ttu_attr->refcyc_per_req_delivery_pre_l);
+ pipe_regs->ttu_regs.refcyc_per_req_delivery_pre_l);
REG_SET(DCN_SURF1_TTU_CNTL1, 0,
REFCYC_PER_REQ_DELIVERY_PRE,
- ttu_attr->refcyc_per_req_delivery_pre_c);
+ pipe_regs->ttu_regs.refcyc_per_req_delivery_pre_c);
REG_SET(DCN_CUR0_TTU_CNTL1, 0,
- REFCYC_PER_REQ_DELIVERY_PRE, ttu_attr->refcyc_per_req_delivery_pre_cur0);
+ REFCYC_PER_REQ_DELIVERY_PRE, pipe_regs->ttu_regs.refcyc_per_req_delivery_pre_cur0);
REG_SET_2(DCN_GLOBAL_TTU_CNTL, 0,
- MIN_TTU_VBLANK, ttu_attr->min_ttu_vblank,
- QoS_LEVEL_FLIP, ttu_attr->qos_level_flip);
+ MIN_TTU_VBLANK, pipe_regs->ttu_regs.min_ttu_vblank,
+ QoS_LEVEL_FLIP, pipe_regs->ttu_regs.qos_level_flip);
}
@@ -508,6 +520,18 @@ bool hubp401_program_surface_flip_and_addr(
return true;
}
+void hubp401_clear_tiling(struct hubp *hubp)
+{
+ struct dcn20_hubp *hubp2 = TO_DCN20_HUBP(hubp);
+
+ REG_UPDATE(DCHUBP_REQ_SIZE_CONFIG, SWATH_HEIGHT, 0);
+ REG_UPDATE(DCSURF_TILING_CONFIG, SW_MODE, DC_SW_LINEAR);
+
+ REG_UPDATE_2(DCSURF_SURFACE_CONTROL,
+ PRIMARY_SURFACE_DCC_EN, 0,
+ SECONDARY_SURFACE_DCC_EN, 0);
+}
+
void hubp401_dcc_control(struct hubp *hubp,
struct dc_plane_dcc_param *dcc)
{
@@ -520,7 +544,7 @@ void hubp401_dcc_control(struct hubp *hubp,
void hubp401_program_tiling(
struct dcn20_hubp *hubp2,
- const union dc_tiling_info *info,
+ const struct dc_tiling_info *info,
const enum surface_pixel_format pixel_format)
{
/* DCSURF_ADDR_CONFIG still shows up in reg spec, but does not need to be programmed for DCN4x
@@ -568,7 +592,7 @@ void hubp401_program_size(
void hubp401_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -718,11 +742,13 @@ void hubp401_cursor_set_position(
dc_fixpt_from_int(dst_x_offset),
param->h_scale_ratio));
- if (cur_en && REG_READ(CURSOR_SURFACE_ADDRESS) == 0)
- hubp->funcs->set_cursor_attributes(hubp, &hubp->curs_attr);
+ if (hubp->pos.cur_ctl.bits.cur_enable != cur_en) {
+ if (cur_en && REG_READ(CURSOR_SURFACE_ADDRESS) == 0)
+ hubp->funcs->set_cursor_attributes(hubp, &hubp->curs_attr);
- REG_UPDATE(CURSOR_CONTROL,
- CURSOR_ENABLE, cur_en);
+ REG_UPDATE(CURSOR_CONTROL,
+ CURSOR_ENABLE, cur_en);
+ }
REG_SET_2(CURSOR_POSITION, 0,
CURSOR_X_POSITION, x_pos,
@@ -969,11 +995,12 @@ static struct hubp_funcs dcn401_hubp_funcs = {
.hubp_program_surface_flip_and_addr = hubp401_program_surface_flip_and_addr,
.hubp_program_surface_config = hubp401_program_surface_config,
.hubp_is_flip_pending = hubp2_is_flip_pending,
- .hubp_setup = hubp401_setup,
- .hubp_setup_interdependent = hubp401_setup_interdependent,
+ .hubp_setup2 = hubp401_setup,
+ .hubp_setup_interdependent2 = hubp401_setup_interdependent,
.hubp_set_vm_system_aperture_settings = hubp3_set_vm_system_aperture_settings,
.set_blank = hubp2_set_blank,
.set_blank_regs = hubp2_set_blank_regs,
+ .hubp_reset = hubp_reset,
.mem_program_viewport = hubp401_set_viewport,
.set_cursor_attributes = hubp32_cursor_set_attributes,
.set_cursor_position = hubp401_cursor_set_position,
@@ -1004,7 +1031,8 @@ static struct hubp_funcs dcn401_hubp_funcs = {
.hubp_program_3dlut_fl_width = hubp401_program_3dlut_fl_width,
.hubp_program_3dlut_fl_tmz_protected = hubp401_program_3dlut_fl_tmz_protected,
.hubp_program_3dlut_fl_crossbar = hubp401_program_3dlut_fl_crossbar,
- .hubp_get_3dlut_fl_done = hubp401_get_3dlut_fl_done
+ .hubp_get_3dlut_fl_done = hubp401_get_3dlut_fl_done,
+ .hubp_clear_tiling = hubp401_clear_tiling,
};
bool hubp401_construct(
diff --git a/drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.h b/drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.h
index e52fdb5b0cd0..608e6153fa68 100644
--- a/drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/hubp/dcn401/dcn401_hubp.h
@@ -256,29 +256,15 @@
void hubp401_update_mall_sel(struct hubp *hubp, uint32_t mall_sel, bool c_cursor);
-void hubp401_vready_at_or_After_vsync(struct hubp *hubp,
- struct _vcs_dpi_display_pipe_dest_params_st *pipe_dest);
-
-void hubp401_program_requestor(
- struct hubp *hubp,
- struct _vcs_dpi_display_rq_regs_st *rq_regs);
-
-void hubp401_program_deadline(
- struct hubp *hubp,
- struct _vcs_dpi_display_dlg_regs_st *dlg_attr,
- struct _vcs_dpi_display_ttu_regs_st *ttu_attr);
-
void hubp401_setup(
struct hubp *hubp,
- struct _vcs_dpi_display_dlg_regs_st *dlg_attr,
- struct _vcs_dpi_display_ttu_regs_st *ttu_attr,
- struct _vcs_dpi_display_rq_regs_st *rq_regs,
- struct _vcs_dpi_display_pipe_dest_params_st *pipe_dest);
+ struct dml2_dchub_per_pipe_register_set *pipe_regs,
+ union dml2_global_sync_programming *pipe_global_sync,
+ struct dc_crtc_timing *timing);
void hubp401_setup_interdependent(
struct hubp *hubp,
- struct _vcs_dpi_display_dlg_regs_st *dlg_attr,
- struct _vcs_dpi_display_ttu_regs_st *ttu_attr);
+ struct dml2_dchub_per_pipe_register_set *pipe_regs);
bool hubp401_program_surface_flip_and_addr(
struct hubp *hubp,
@@ -290,7 +276,7 @@ void hubp401_dcc_control(struct hubp *hubp,
void hubp401_program_tiling(
struct dcn20_hubp *hubp2,
- const union dc_tiling_info *info,
+ const struct dc_tiling_info *info,
const enum surface_pixel_format pixel_format);
void hubp401_program_size(
@@ -302,7 +288,7 @@ void hubp401_program_size(
void hubp401_program_surface_config(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -340,4 +326,42 @@ int hubp401_get_3dlut_fl_done(struct hubp *hubp);
void hubp401_set_unbounded_requesting(struct hubp *hubp, bool enable);
+void hubp401_update_3dlut_fl_bias_scale(struct hubp *hubp, uint16_t bias, uint16_t scale);
+
+void hubp401_program_3dlut_fl_crossbar(struct hubp *hubp,
+ enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_y_g,
+ enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_cb_b,
+ enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_cr_r);
+
+void hubp401_program_3dlut_fl_tmz_protected(struct hubp *hubp, uint8_t protection_bits);
+
+void hubp401_program_3dlut_fl_width(struct hubp *hubp, enum hubp_3dlut_fl_width width);
+
+void hubp401_program_3dlut_fl_addressing_mode(struct hubp *hubp, enum hubp_3dlut_fl_addressing_mode addr_mode);
+
+void hubp401_enable_3dlut_fl(struct hubp *hubp, bool enable);
+
+void hubp401_program_3dlut_fl_dlg_param(struct hubp *hubp, int refcyc_per_3dlut_group);
+
+void hubp401_program_3dlut_fl_addr(struct hubp *hubp, const struct dc_plane_address address);
+
+void hubp401_program_3dlut_fl_format(struct hubp *hubp, enum hubp_3dlut_fl_format format);
+
+void hubp401_program_3dlut_fl_mode(struct hubp *hubp, enum hubp_3dlut_fl_mode mode);
+
+void hubp401_clear_tiling(struct hubp *hubp);
+
+void hubp401_vready_at_or_After_vsync(struct hubp *hubp,
+ union dml2_global_sync_programming *pipe_global_sync,
+ struct dc_crtc_timing *timing);
+
+void hubp401_program_requestor(
+ struct hubp *hubp,
+ struct dml2_display_rq_regs *rq_regs);
+
+void hubp401_program_deadline(
+ struct hubp *hubp,
+ struct dml2_display_dlg_regs *dlg_attr,
+ struct dml2_display_ttu_regs *ttu_attr);
+
#endif /* __DC_HUBP_DCN401_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/Makefile b/drivers/gpu/drm/amd/display/dc/hwss/Makefile
index 40ecebea1ba0..bee617ca0838 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/hwss/Makefile
@@ -27,6 +27,24 @@
# DCE
###############################################################################
+ifdef CONFIG_DRM_AMD_DC_SI
+HWSS_DCE60 = dce60_hwseq.o
+
+AMD_DAL_HWSS_DCE60 = $(addprefix $(AMDDALPATH)/dc/hwss/dce60/,$(HWSS_DCE60))
+
+AMD_DISPLAY_FILES += $(AMD_DAL_HWSS_DCE60)
+endif
+
+###############################################################################
+
+HWSS_DCE80 = dce80_hwseq.o
+
+AMD_DAL_HWSS_DCE80 = $(addprefix $(AMDDALPATH)/dc/hwss/dce80/,$(HWSS_DCE80))
+
+AMD_DISPLAY_FILES += $(AMD_DAL_HWSS_DCE80)
+
+###############################################################################
+
HWSS_DCE = dce_hwseq.o
AMD_DAL_HWSS_DCE = $(addprefix $(AMDDALPATH)/dc/hwss/dce/,$(HWSS_DCE))
@@ -65,14 +83,6 @@ AMD_DAL_HWSS_DCE120 = $(addprefix $(AMDDALPATH)/dc/hwss/dce120/,$(HWSS_DCE120))
AMD_DISPLAY_FILES += $(AMD_DAL_HWSS_DCE120)
-###############################################################################
-
-HWSS_DCE80 = dce80_hwseq.o
-
-AMD_DAL_HWSS_DCE80 = $(addprefix $(AMDDALPATH)/dc/hwss/dce80/,$(HWSS_DCE80))
-
-AMD_DISPLAY_FILES += $(AMD_DAL_HWSS_DCE80)
-
ifdef CONFIG_DRM_AMD_DC_FP
###############################################################################
# DCN
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce/dce_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dce/dce_hwseq.h
index 84c8f8707c5d..f66a38f43a09 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dce/dce_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce/dce_hwseq.h
@@ -644,10 +644,18 @@ struct dce_hwseq_registers {
uint32_t DPP_TOP0_DPP_CRC_CTRL;
uint32_t DPP_TOP0_DPP_CRC_VAL_R_G;
uint32_t DPP_TOP0_DPP_CRC_VAL_B_A;
+ uint32_t DPP_TOP0_DPP_CRC_VAL_R;
+ uint32_t DPP_TOP0_DPP_CRC_VAL_G;
+ uint32_t DPP_TOP0_DPP_CRC_VAL_B;
+ uint32_t DPP_TOP0_DPP_CRC_VAL_A;
uint32_t MPC_CRC_CTRL;
uint32_t MPC_CRC_RESULT_GB;
uint32_t MPC_CRC_RESULT_C;
uint32_t MPC_CRC_RESULT_AR;
+ uint32_t MPC_CRC_RESULT_R;
+ uint32_t MPC_CRC_RESULT_G;
+ uint32_t MPC_CRC_RESULT_B;
+ uint32_t MPC_CRC_RESULT_A;
uint32_t D1VGA_CONTROL;
uint32_t D2VGA_CONTROL;
uint32_t D3VGA_CONTROL;
@@ -1236,6 +1244,7 @@ struct dce_hwseq_registers {
type DOMAIN24_PGFSM_PWR_STATUS; \
type DOMAIN25_PGFSM_PWR_STATUS; \
type DOMAIN_DESIRED_PWR_STATE;
+
struct dce_hwseq_shift {
HWSEQ_REG_FIELD_LIST(uint8_t)
HWSEQ_DCN_REG_FIELD_LIST(uint8_t)
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.c
index f1f14796a3da..0d7e28260db1 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.c
@@ -138,5 +138,35 @@ void dce100_hw_sequencer_construct(struct dc *dc)
dc->hwseq->funcs.enable_display_power_gating = dce100_enable_display_power_gating;
dc->hwss.prepare_bandwidth = dce100_prepare_bandwidth;
dc->hwss.optimize_bandwidth = dce100_optimize_bandwidth;
+ dc->hwss.clear_surface_dcc_and_tiling = dce100_reset_surface_dcc_and_tiling;
+}
+
+/**
+ * dce100_reset_surface_dcc_and_tiling - Set DCC and tiling in DCE to their disable mode.
+ *
+ * @pipe_ctx: Pointer to the pipe context structure.
+ * @plane_state: Surface state
+ * @clear_tiling: If true set tiling to Linear, otherwise does not change tiling
+ *
+ * This function is responsible for call the HUBP block to disable DCC and set
+ * tiling to the linear mode.
+ */
+void dce100_reset_surface_dcc_and_tiling(struct pipe_ctx *pipe_ctx,
+ struct dc_plane_state *plane_state,
+ bool clear_tiling)
+{
+ struct mem_input *mi = pipe_ctx->plane_res.mi;
+
+ if (!mi)
+ return;
+
+ /* if framebuffer is tiled, disable tiling */
+ if (clear_tiling && mi->funcs->mem_input_clear_tiling)
+ mi->funcs->mem_input_clear_tiling(mi);
+
+ /* force page flip to see the new content of the framebuffer */
+ mi->funcs->mem_input_program_surface_flip_and_addr(mi,
+ &plane_state->address,
+ true);
}
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.h
index 34518da20009..fadfa794f96b 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce100/dce100_hwseq.h
@@ -46,5 +46,9 @@ bool dce100_enable_display_power_gating(struct dc *dc, uint8_t controller_id,
struct dc_bios *dcb,
enum pipe_gating_control power_gating);
+void dce100_reset_surface_dcc_and_tiling(struct pipe_ctx *pipe_ctx,
+ struct dc_plane_state *plane_state,
+ bool clear_tiling);
+
#endif /* __DC_HWSS_DCE100_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c
index 81f4c386c287..4ea13d0bf815 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.c
@@ -33,6 +33,7 @@
#include "dce110_hwseq.h"
#include "dce110/dce110_timing_generator.h"
#include "dce/dce_hwseq.h"
+#include "dce100/dce100_hwseq.h"
#include "gpio_service_interface.h"
#include "dce110/dce110_compressor.h"
@@ -951,8 +952,8 @@ void dce110_edp_backlight_control(
struct dc_context *ctx = link->ctx;
struct bp_transmitter_control cntl = { 0 };
uint8_t pwrseq_instance = 0;
- unsigned int pre_T11_delay = OLED_PRE_T11_DELAY;
- unsigned int post_T7_delay = OLED_POST_T7_DELAY;
+ unsigned int pre_T11_delay = (link->dpcd_sink_ext_caps.bits.oled ? OLED_PRE_T11_DELAY : 0);
+ unsigned int post_T7_delay = (link->dpcd_sink_ext_caps.bits.oled ? OLED_POST_T7_DELAY : 0);
if (dal_graphics_object_id_get_connector_id(link->link_enc->connector)
!= CONNECTOR_ID_EDP) {
@@ -1065,9 +1066,11 @@ void dce110_edp_backlight_control(
DC_LOG_DC("edp_receiver_ready_T9 skipped\n");
}
- if (!enable && link->dpcd_sink_ext_caps.bits.oled) {
+ if (!enable) {
+ /*follow oem panel config's requirement*/
pre_T11_delay += link->panel_config.pps.extra_pre_t11_ms;
- msleep(pre_T11_delay);
+ if (pre_T11_delay)
+ msleep(pre_T11_delay);
}
}
@@ -1152,9 +1155,12 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx)
struct timing_generator *tg = pipe_ctx->stream_res.tg;
struct dtbclk_dto_params dto_params = {0};
int dp_hpo_inst;
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(pipe_ctx->stream->link);
+ struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
struct stream_encoder *stream_enc = pipe_ctx->stream_res.stream_enc;
+ if (!dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+
if (dc_is_hdmi_tmds_signal(pipe_ctx->stream->signal)) {
pipe_ctx->stream_res.stream_enc->funcs->stop_hdmi_info_packets(
pipe_ctx->stream_res.stream_enc);
@@ -1180,8 +1186,10 @@ void dce110_disable_stream(struct pipe_ctx *pipe_ctx)
if (dccg) {
dccg->funcs->disable_symclk32_se(dccg, dp_hpo_inst);
dccg->funcs->set_dpstreamclk(dccg, REFCLK, tg->inst, dp_hpo_inst);
- if (dccg && dccg->funcs->set_dtbclk_dto)
- dccg->funcs->set_dtbclk_dto(dccg, &dto_params);
+ if (!(dc->ctx->dce_version >= DCN_VERSION_3_5)) {
+ if (dccg && dccg->funcs->set_dtbclk_dto)
+ dccg->funcs->set_dtbclk_dto(dccg, &dto_params);
+ }
}
} else if (dccg && dccg->funcs->disable_symclk_se) {
dccg->funcs->disable_symclk_se(dccg, stream_enc->stream_enc_inst,
@@ -1215,8 +1223,11 @@ void dce110_blank_stream(struct pipe_ctx *pipe_ctx)
struct dc_link *link = stream->link;
struct dce_hwseq *hws = link->dc->hwseq;
+ if (hws && hws->wa_state.skip_blank_stream)
+ return;
+
if (link->local_sink && link->local_sink->sink_signal == SIGNAL_TYPE_EDP) {
- if (!link->skip_implict_edp_power_control)
+ if (!link->skip_implict_edp_power_control && hws)
hws->funcs.edp_backlight_control(link, false);
link->dc->hwss.set_abm_immediate_disable(pipe_ctx);
}
@@ -1370,7 +1381,7 @@ static void populate_audio_dp_link_info(
}
}
-static void build_audio_output(
+void build_audio_output(
struct dc_state *state,
const struct pipe_ctx *pipe_ctx,
struct audio_output *audio_output)
@@ -1654,9 +1665,7 @@ enum dc_status dce110_apply_single_controller_ctx_to_hw(
params.vertical_total_min = stream->adjust.v_total_min;
params.vertical_total_max = stream->adjust.v_total_max;
- if (pipe_ctx->stream_res.tg->funcs->set_drr)
- pipe_ctx->stream_res.tg->funcs->set_drr(
- pipe_ctx->stream_res.tg, &params);
+ set_drr_and_clear_adjust_pending(pipe_ctx, stream, &params);
// DRR should set trigger event to monitor surface update event
if (stream->adjust.v_total_min != 0 && stream->adjust.v_total_max != 0)
@@ -1677,6 +1686,19 @@ enum dc_status dce110_apply_single_controller_ctx_to_hw(
if (dc_is_dp_signal(pipe_ctx->stream->signal))
dc->link_srv->dp_trace_source_sequence(link, DPCD_SOURCE_SEQ_AFTER_CONNECT_DIG_FE_OTG);
+ /* Temporary workaround to perform DSC programming ahead of stream enablement
+ * for smartmux/SPRS
+ * TODO: Remove SmartMux/SPRS checks once movement of DSC programming is generalized
+ */
+ if (pipe_ctx->stream->timing.flags.DSC) {
+ if ((pipe_ctx->stream->signal == SIGNAL_TYPE_EDP &&
+ ((link->dc->config.smart_mux_version && link->dc->is_switch_in_progress_dest)
+ || link->is_dds || link->skip_implict_edp_power_control)) &&
+ (dc_is_dp_signal(pipe_ctx->stream->signal) ||
+ dc_is_virtual_signal(pipe_ctx->stream->signal)))
+ dc->link_srv->set_dsc_enable(pipe_ctx, true);
+ }
+
if (!stream->dpms_off)
dc->link_srv->set_dpms_on(context, pipe_ctx);
@@ -1834,10 +1856,10 @@ static void clean_up_dsc_blocks(struct dc *dc)
struct pg_cntl *pg_cntl = dc->res_pool->pg_cntl;
int i;
- if (dc->ctx->dce_version != DCN_VERSION_3_5 &&
- dc->ctx->dce_version != DCN_VERSION_3_51)
+ if (!dc->caps.is_apu ||
+ dc->ctx->dce_version < DCN_VERSION_3_15)
return;
-
+ /*VBIOS supports dsc starts from dcn315*/
for (i = 0; i < dc->res_pool->res_cap->num_dsc; i++) {
struct dcn_dsc_state s = {0};
@@ -1918,6 +1940,13 @@ void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context)
can_apply_edp_fast_boot = dc_validate_boot_timing(dc,
edp_stream->sink, &edp_stream->timing);
+
+ // For Mux-platform, the default value is false.
+ // Disable fast boot during mux switching.
+ // The flag would be clean after switching done.
+ if (dc->is_switch_in_progress_dest && edp_link->is_dds)
+ can_apply_edp_fast_boot = false;
+
edp_stream->apply_edp_fast_boot_optimization = can_apply_edp_fast_boot;
if (can_apply_edp_fast_boot) {
DC_LOG_EVENT_LINK_TRAINING("eDP fast boot Enable\n");
@@ -1961,6 +1990,10 @@ void dce110_enable_accelerated_mode(struct dc *dc, struct dc_state *context)
if (edp_with_sink_num)
edp_link_with_sink = edp_links_with_sink[0];
+ // During a mux switch, powering down the HW blocks and then enabling
+ // the link via a DPCD SET_POWER write causes a brief flash
+ keep_edp_vdd_on |= dc->is_switch_in_progress_dest;
+
if (!can_apply_edp_fast_boot && !can_apply_seamless_boot) {
if (edp_link_with_sink && !keep_edp_vdd_on) {
/*turn off backlight before DP_blank and encoder powered down*/
@@ -2104,8 +2137,7 @@ static void set_drr(struct pipe_ctx **pipe_ctx,
struct timing_generator *tg = pipe_ctx[i]->stream_res.tg;
if ((tg != NULL) && tg->funcs) {
- if (tg->funcs->set_drr)
- tg->funcs->set_drr(tg, &params);
+ set_drr_and_clear_adjust_pending(pipe_ctx[i], pipe_ctx[i]->stream, &params);
if (adjust.v_total_max != 0 && adjust.v_total_min != 0)
if (tg->funcs->set_static_screen_control)
tg->funcs->set_static_screen_control(
@@ -2761,12 +2793,12 @@ static void dce110_enable_per_frame_crtc_position_reset(
}
-static void init_pipes(struct dc *dc, struct dc_state *context)
+static void dce110_init_pipes(struct dc *dc, struct dc_state *context)
{
// Do nothing
}
-static void init_hw(struct dc *dc)
+static void dce110_init_hw(struct dc *dc)
{
int i;
struct dc_bios *bp;
@@ -3325,12 +3357,13 @@ void dce110_disable_link_output(struct dc_link *link,
static const struct hw_sequencer_funcs dce110_funcs = {
.program_gamut_remap = program_gamut_remap,
.program_output_csc = program_output_csc,
- .init_hw = init_hw,
+ .init_hw = dce110_init_hw,
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = dce110_apply_ctx_for_surface,
.post_unlock_program_front_end = dce110_post_unlock_program_front_end,
.update_plane_addr = update_plane_addr,
.update_pending_status = dce110_update_pending_status,
+ .clear_surface_dcc_and_tiling = dce100_reset_surface_dcc_and_tiling,
.enable_accelerated_mode = dce110_enable_accelerated_mode,
.enable_timing_synchronization = dce110_enable_timing_synchronization,
.enable_per_frame_crtc_position_reset = dce110_enable_per_frame_crtc_position_reset,
@@ -3368,7 +3401,7 @@ static const struct hw_sequencer_funcs dce110_funcs = {
};
static const struct hwseq_private_funcs dce110_private_funcs = {
- .init_pipes = init_pipes,
+ .init_pipes = dce110_init_pipes,
.set_input_transfer_func = dce110_set_input_transfer_func,
.set_output_transfer_func = dce110_set_output_transfer_func,
.power_down = dce110_power_down,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.h
index 06789ac3a224..7cd8c1576988 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce110/dce110_hwseq.h
@@ -110,5 +110,9 @@ void dce110_enable_dp_link_output(
enum signal_type signal,
enum clock_source_id clock_source,
const struct dc_link_settings *link_settings);
+void build_audio_output(
+ struct dc_state *state,
+ const struct pipe_ctx *pipe_ctx,
+ struct audio_output *audio_output);
#endif /* __DC_HWSS_DCE110_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce120/dce120_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dce120/dce120_hwseq.c
index 22ee304ef9cf..2a62f63d0357 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dce120/dce120_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce120/dce120_hwseq.c
@@ -29,6 +29,7 @@
#include "dce120_hwseq.h"
#include "dce/dce_hwseq.h"
+#include "dce100/dce100_hwseq.h"
#include "dce110/dce110_hwseq.h"
#include "dce/dce_12_0_offset.h"
@@ -264,5 +265,6 @@ void dce120_hw_sequencer_construct(struct dc *dc)
dce110_hw_sequencer_construct(dc);
dc->hwseq->funcs.enable_display_power_gating = dce120_enable_display_power_gating;
dc->hwss.update_dchub = dce120_update_dchub;
+ dc->hwss.clear_surface_dcc_and_tiling = dce100_reset_surface_dcc_and_tiling;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dce60/dce60_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/hwss/dce60/dce60_hwseq.c
index 1fdeef47e4dc..a08e9f9eec17 100644
--- a/drivers/gpu/drm/amd/display/dc/dce60/dce60_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce60/dce60_hwseq.c
@@ -26,7 +26,7 @@
#include "dm_services.h"
#include "dc.h"
#include "core_types.h"
-#include "dce60_hw_sequencer.h"
+#include "dce60_hwseq.h"
#include "dce/dce_hwseq.h"
#include "dce110/dce110_hwseq.h"
@@ -428,5 +428,6 @@ void dce60_hw_sequencer_construct(struct dc *dc)
dc->hwss.pipe_control_lock = dce60_pipe_control_lock;
dc->hwss.prepare_bandwidth = dce100_prepare_bandwidth;
dc->hwss.optimize_bandwidth = dce100_optimize_bandwidth;
+ dc->hwss.clear_surface_dcc_and_tiling = dce100_reset_surface_dcc_and_tiling;
}
diff --git a/drivers/gpu/drm/amd/display/dc/dce60/dce60_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/hwss/dce60/dce60_hwseq.h
index f3b2d8b60d5b..f3b2d8b60d5b 100644
--- a/drivers/gpu/drm/amd/display/dc/dce60/dce60_hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce60/dce60_hwseq.h
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dce80/dce80_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dce80/dce80_hwseq.c
index 0a054e880801..76fd45550c5e 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dce80/dce80_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dce80/dce80_hwseq.c
@@ -50,5 +50,6 @@ void dce80_hw_sequencer_construct(struct dc *dc)
dc->hwss.pipe_control_lock = dce_pipe_control_lock;
dc->hwss.prepare_bandwidth = dce100_prepare_bandwidth;
dc->hwss.optimize_bandwidth = dce100_optimize_bandwidth;
+ dc->hwss.clear_surface_dcc_and_tiling = dce100_reset_surface_dcc_and_tiling;
}
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c
index 681bb92c6069..39910f73ecd0 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.c
@@ -94,6 +94,128 @@ static void print_microsec(struct dc_context *dc_ctx,
us_x10 % frac);
}
+/*
+ * Delay until we passed busy-until-point to which we can
+ * do necessary locking/programming on consecutive full updates
+ */
+void dcn10_wait_for_pipe_update_if_needed(struct dc *dc, struct pipe_ctx *pipe_ctx, bool is_surface_update_only)
+{
+ struct crtc_position position;
+ struct dc_stream_state *stream = pipe_ctx->stream;
+ unsigned int vpos, frame_count;
+ uint32_t vupdate_start, vupdate_end, vblank_start;
+ unsigned int lines_to_vupdate, us_to_vupdate;
+ unsigned int us_per_line, us_vupdate;
+
+ if (!pipe_ctx->stream ||
+ !pipe_ctx->stream_res.tg ||
+ !pipe_ctx->stream_res.stream_enc)
+ return;
+
+ if (pipe_ctx->prev_odm_pipe &&
+ pipe_ctx->stream)
+ return;
+
+ if (!pipe_ctx->wait_is_required)
+ return;
+
+ struct timing_generator *tg = pipe_ctx->stream_res.tg;
+
+ if (tg->funcs->is_tg_enabled && !tg->funcs->is_tg_enabled(tg))
+ return;
+
+ dc->hwss.calc_vupdate_position(dc, pipe_ctx, &vupdate_start,
+ &vupdate_end);
+
+ dc->hwss.get_position(&pipe_ctx, 1, &position);
+ vpos = position.vertical_count;
+
+ frame_count = tg->funcs->get_frame_count(tg);
+
+ if (frame_count - pipe_ctx->wait_frame_count > 2)
+ return;
+
+ vblank_start = pipe_ctx->pipe_dlg_param.vblank_start;
+
+ if (vpos >= vupdate_start && vupdate_start >= vblank_start)
+ lines_to_vupdate = stream->timing.v_total - vpos + vupdate_start;
+ else
+ lines_to_vupdate = vupdate_start - vpos;
+
+ us_per_line =
+ stream->timing.h_total * 10000u / stream->timing.pix_clk_100hz;
+ us_to_vupdate = lines_to_vupdate * us_per_line;
+
+ if (vupdate_end < vupdate_start)
+ vupdate_end += stream->timing.v_total;
+
+ if (lines_to_vupdate > stream->timing.v_total - vupdate_end + vupdate_start)
+ us_to_vupdate = 0;
+
+ us_vupdate = (vupdate_end - vupdate_start + 1) * us_per_line;
+
+ if (is_surface_update_only && us_to_vupdate + us_vupdate > 200) {
+ //surface updates come in at high irql
+ pipe_ctx->wait_is_required = true;
+ return;
+ }
+
+ fsleep(us_to_vupdate + us_vupdate);
+
+ //clear
+ pipe_ctx->next_vupdate = 0;
+ pipe_ctx->wait_frame_count = 0;
+ pipe_ctx->wait_is_required = false;
+}
+
+/*
+ * On pipe unlock and programming, indicate pipe will be busy
+ * until some frame and line (vupdate), this is required for consecutive
+ * full updates, need to wait for updates
+ * to latch to try and program the next update
+ */
+void dcn10_set_wait_for_update_needed_for_pipe(struct dc *dc, struct pipe_ctx *pipe_ctx)
+{
+ uint32_t vupdate_start, vupdate_end;
+ struct crtc_position position;
+ unsigned int vpos, cur_frame;
+
+ if (!pipe_ctx->stream ||
+ !pipe_ctx->stream_res.tg ||
+ !pipe_ctx->stream_res.stream_enc)
+ return;
+
+ dc->hwss.get_position(&pipe_ctx, 1, &position);
+ vpos = position.vertical_count;
+
+ dc->hwss.calc_vupdate_position(dc, pipe_ctx, &vupdate_start,
+ &vupdate_end);
+
+ struct timing_generator *tg = pipe_ctx->stream_res.tg;
+
+ struct optc *optc1 = DCN10TG_FROM_TG(tg);
+
+ ASSERT(optc1->max_frame_count != 0);
+
+ if (tg->funcs->is_tg_enabled && !tg->funcs->is_tg_enabled(tg))
+ return;
+
+ pipe_ctx->next_vupdate = vupdate_start;
+
+ cur_frame = tg->funcs->get_frame_count(tg);
+
+ if (vpos < vupdate_start) {
+ pipe_ctx->wait_frame_count = cur_frame;
+ } else {
+ if (cur_frame + 1 > optc1->max_frame_count)
+ pipe_ctx->wait_frame_count = cur_frame + 1 - optc1->max_frame_count;
+ else
+ pipe_ctx->wait_frame_count = cur_frame + 1;
+ }
+
+ pipe_ctx->wait_is_required = true;
+}
+
void dcn10_lock_all_pipes(struct dc *dc,
struct dc_state *context,
bool lock)
@@ -205,6 +327,35 @@ static void dcn10_log_hubp_states(struct dc *dc, void *log_ctx)
}
}
+ DTN_INFO("\n=======HUBP FL======\n");
+ DTN_INFO(
+ "HUBP FL: Enabled Done adr_mode width tmz xbar_sel_R xbar_sel_G xbar_sel_B adr_hi adr_low REFCYC Bias Scale Mode Format\n");
+ for (i = 0; i < pool->pipe_count; i++) {
+ struct dcn_hubp_state *s = &(TO_DCN10_HUBP(pool->hubps[i])->state);
+ struct dcn_fl_regs_st *fl_regs = &s->fl_regs;
+
+ if (!s->blank_en) {
+ DTN_INFO("[%2d]: %5xh %6xh %5d %6d %8xh %2xh %6xh %6d %8d %8d %7d %8xh %5x %5x %5x",
+ pool->hubps[i]->inst,
+ fl_regs->lut_enable,
+ fl_regs->lut_done,
+ fl_regs->lut_addr_mode,
+ fl_regs->lut_width,
+ fl_regs->lut_tmz,
+ fl_regs->lut_crossbar_sel_r,
+ fl_regs->lut_crossbar_sel_g,
+ fl_regs->lut_crossbar_sel_b,
+ fl_regs->lut_addr_hi,
+ fl_regs->lut_addr_lo,
+ fl_regs->refcyc_3dlut_group,
+ fl_regs->lut_fl_bias,
+ fl_regs->lut_fl_scale,
+ fl_regs->lut_fl_mode,
+ fl_regs->lut_fl_format);
+ DTN_INFO("\n");
+ }
+ }
+
DTN_INFO("\n=========RQ========\n");
DTN_INFO("HUBP: drq_exp_m prq_exp_m mrq_exp_m crq_exp_m plane1_ba L:chunk_s min_chu_s meta_ch_s"
" min_m_c_s dpte_gr_s mpte_gr_s swath_hei pte_row_h C:chunk_s min_chu_s meta_ch_s"
@@ -389,6 +540,36 @@ static void dcn10_log_color_state(struct dc *dc,
dc->caps.color.mpc.num_3dluts,
dc->caps.color.mpc.ogam_ram,
dc->caps.color.mpc.ocsc);
+ DTN_INFO("===== MPC RMCM 3DLUT =====\n");
+ DTN_INFO("MPCC: SIZE MODE MODE_CUR RD_SEL 30BIT_EN WR_EN_MASK RAM_SEL OUT_NORM_FACTOR FL_SEL OUT_OFFSET OUT_SCALE FL_DONE SOFT_UNDERFLOW HARD_UNDERFLOW MEM_PWR_ST FORCE DIS MODE\n");
+ for (i = 0; i < pool->mpcc_count; i++) {
+ struct mpcc_state s = {0};
+
+ pool->mpc->funcs->read_mpcc_state(pool->mpc, i, &s);
+ if (s.opp_id != 0xf)
+ DTN_INFO("[%2d]: %4xh %4xh %6xh %4x %4x %4x %4x %4x %4xh %4xh %6xh %4x %4x %4x %4x %4x %4x %4x\n",
+ i, s.rmcm_regs.rmcm_3dlut_size, s.rmcm_regs.rmcm_3dlut_mode, s.rmcm_regs.rmcm_3dlut_mode_cur,
+ s.rmcm_regs.rmcm_3dlut_read_sel, s.rmcm_regs.rmcm_3dlut_30bit_en, s.rmcm_regs.rmcm_3dlut_wr_en_mask,
+ s.rmcm_regs.rmcm_3dlut_ram_sel, s.rmcm_regs.rmcm_3dlut_out_norm_factor, s.rmcm_regs.rmcm_3dlut_fl_sel,
+ s.rmcm_regs.rmcm_3dlut_out_offset_r, s.rmcm_regs.rmcm_3dlut_out_scale_r, s.rmcm_regs.rmcm_3dlut_fl_done,
+ s.rmcm_regs.rmcm_3dlut_fl_soft_underflow, s.rmcm_regs.rmcm_3dlut_fl_hard_underflow, s.rmcm_regs.rmcm_3dlut_mem_pwr_state,
+ s.rmcm_regs.rmcm_3dlut_mem_pwr_force, s.rmcm_regs.rmcm_3dlut_mem_pwr_dis, s.rmcm_regs.rmcm_3dlut_mem_pwr_mode);
+ }
+ DTN_INFO("\n");
+ DTN_INFO("===== MPC RMCM Shaper =====\n");
+ DTN_INFO("MPCC: CNTL LUT_MODE MODE_CUR WR_EN_MASK WR_SEL OFFSET SCALE START_B START_SEG_B END_B END_BASE_B MEM_PWR_ST FORCE DIS MODE\n");
+ for (i = 0; i < pool->mpcc_count; i++) {
+ struct mpcc_state s = {0};
+
+ pool->mpc->funcs->read_mpcc_state(pool->mpc, i, &s);
+ if (s.opp_id != 0xf)
+ DTN_INFO("[%2d]: %4xh %4xh %6xh %4x %4x %4x %4x %4x %4xh %4xh %6xh %4x %4x %4x %4x\n",
+ i, s.rmcm_regs.rmcm_cntl, s.rmcm_regs.rmcm_shaper_lut_mode, s.rmcm_regs.rmcm_shaper_mode_cur,
+ s.rmcm_regs.rmcm_shaper_lut_write_en_mask, s.rmcm_regs.rmcm_shaper_lut_write_sel, s.rmcm_regs.rmcm_shaper_offset_b,
+ s.rmcm_regs.rmcm_shaper_scale_b, s.rmcm_regs.rmcm_shaper_rama_exp_region_start_b, s.rmcm_regs.rmcm_shaper_rama_exp_region_start_seg_b,
+ s.rmcm_regs.rmcm_shaper_rama_exp_region_end_b, s.rmcm_regs.rmcm_shaper_rama_exp_region_end_base_b, s.rmcm_regs.rmcm_shaper_mem_pwr_state,
+ s.rmcm_regs.rmcm_shaper_mem_pwr_force, s.rmcm_regs.rmcm_shaper_mem_pwr_dis, s.rmcm_regs.rmcm_shaper_mem_pwr_mode);
+ }
}
void dcn10_log_hw_state(struct dc *dc,
@@ -415,7 +596,8 @@ void dcn10_log_hw_state(struct dc *dc,
struct timing_generator *tg = pool->timing_generators[i];
struct dcn_otg_state s = {0};
/* Read shared OTG state registers for all DCNx */
- optc1_read_otg_state(DCN10TG_FROM_TG(tg), &s);
+ if (tg->funcs->read_otg_state)
+ tg->funcs->read_otg_state(tg, &s);
/*
* For DCN2 and greater, a register on the OPP is used to
@@ -1112,9 +1294,7 @@ static void dcn10_reset_back_end_for_pipe(
pipe_ctx->stream_res.tg->funcs->disable_crtc(pipe_ctx->stream_res.tg);
pipe_ctx->stream_res.tg->funcs->enable_optc_clock(pipe_ctx->stream_res.tg, false);
- if (pipe_ctx->stream_res.tg->funcs->set_drr)
- pipe_ctx->stream_res.tg->funcs->set_drr(
- pipe_ctx->stream_res.tg, NULL);
+ set_drr_and_clear_adjust_pending(pipe_ctx, pipe_ctx->stream, NULL);
if (dc_is_hdmi_tmds_signal(pipe_ctx->stream->signal))
pipe_ctx->stream->link->phy_state.symclk_ref_cnts.otg = 0;
}
@@ -1286,6 +1466,7 @@ void dcn10_plane_atomic_power_down(struct dc *dc,
if (hws->funcs.hubp_pg_control)
hws->funcs.hubp_pg_control(hws, hubp->inst, false);
+ hubp->funcs->hubp_reset(hubp);
dpp->funcs->dpp_reset(dpp);
REG_SET(DC_IP_REQUEST_CNTL, 0,
@@ -1447,6 +1628,7 @@ void dcn10_init_pipes(struct dc *dc, struct dc_state *context)
/* Disable on the current state so the new one isn't cleared. */
pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
+ hubp->funcs->hubp_reset(hubp);
dpp->funcs->dpp_reset(dpp);
pipe_ctx->stream_res.tg = tg;
@@ -1990,20 +2172,11 @@ static void delay_cursor_until_vupdate(struct dc *dc, struct pipe_ctx *pipe_ctx)
dc->hwss.get_position(&pipe_ctx, 1, &position);
vpos = position.vertical_count;
- /* Avoid wraparound calculation issues */
- vupdate_start += stream->timing.v_total;
- vupdate_end += stream->timing.v_total;
- vpos += stream->timing.v_total;
-
if (vpos <= vupdate_start) {
/* VPOS is in VACTIVE or back porch. */
lines_to_vupdate = vupdate_start - vpos;
- } else if (vpos > vupdate_end) {
- /* VPOS is in the front porch. */
- return;
} else {
- /* VPOS is in VUPDATE. */
- lines_to_vupdate = 0;
+ lines_to_vupdate = stream->timing.v_total - vpos + vupdate_start;
}
/* Calculate time until VUPDATE in microseconds. */
@@ -2011,13 +2184,18 @@ static void delay_cursor_until_vupdate(struct dc *dc, struct pipe_ctx *pipe_ctx)
stream->timing.h_total * 10000u / stream->timing.pix_clk_100hz;
us_to_vupdate = lines_to_vupdate * us_per_line;
+ /* Stall out until the cursor update completes. */
+ if (vupdate_end < vupdate_start)
+ vupdate_end += stream->timing.v_total;
+
+ /* Position is in the range of vupdate start and end*/
+ if (lines_to_vupdate > stream->timing.v_total - vupdate_end + vupdate_start)
+ us_to_vupdate = 0;
+
/* 70 us is a conservative estimate of cursor update time*/
if (us_to_vupdate > 70)
return;
- /* Stall out until the cursor update completes. */
- if (vupdate_end < vupdate_start)
- vupdate_end += stream->timing.v_total;
us_vupdate = (vupdate_end - vupdate_start + 1) * us_per_line;
udelay(us_to_vupdate + us_vupdate);
}
@@ -2667,7 +2845,6 @@ void dcn10_update_visual_confirm_color(struct dc *dc,
struct mpc *mpc = dc->res_pool->mpc;
if (mpc->funcs->set_bg_color) {
- memcpy(&pipe_ctx->plane_state->visual_confirm_color, &(pipe_ctx->visual_confirm_color), sizeof(struct tg_color));
mpc->funcs->set_bg_color(mpc, &(pipe_ctx->visual_confirm_color), mpcc_id);
}
}
@@ -3219,8 +3396,7 @@ void dcn10_set_drr(struct pipe_ctx **pipe_ctx,
struct timing_generator *tg = pipe_ctx[i]->stream_res.tg;
if ((tg != NULL) && tg->funcs) {
- if (tg->funcs->set_drr)
- tg->funcs->set_drr(tg, &params);
+ set_drr_and_clear_adjust_pending(pipe_ctx[i], pipe_ctx[i]->stream, &params);
if (adjust.v_total_max != 0 && adjust.v_total_min != 0)
if (tg->funcs->set_static_screen_control)
tg->funcs->set_static_screen_control(
@@ -3429,52 +3605,6 @@ void dcn10_update_dchub(struct dce_hwseq *hws, struct dchub_init_data *dh_data)
hubbub->funcs->update_dchub(hubbub, dh_data);
}
-static bool dcn10_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx)
-{
- struct pipe_ctx *test_pipe, *split_pipe;
- const struct scaler_data *scl_data = &pipe_ctx->plane_res.scl_data;
- struct rect r1 = scl_data->recout, r2, r2_half;
- int r1_r = r1.x + r1.width, r1_b = r1.y + r1.height, r2_r, r2_b;
- int cur_layer = pipe_ctx->plane_state->layer_index;
-
- /**
- * Disable the cursor if there's another pipe above this with a
- * plane that contains this pipe's viewport to prevent double cursor
- * and incorrect scaling artifacts.
- */
- for (test_pipe = pipe_ctx->top_pipe; test_pipe;
- test_pipe = test_pipe->top_pipe) {
- // Skip invisible layer and pipe-split plane on same layer
- if (!test_pipe->plane_state ||
- !test_pipe->plane_state->visible ||
- test_pipe->plane_state->layer_index == cur_layer)
- continue;
-
- r2 = test_pipe->plane_res.scl_data.recout;
- r2_r = r2.x + r2.width;
- r2_b = r2.y + r2.height;
-
- /**
- * There is another half plane on same layer because of
- * pipe-split, merge together per same height.
- */
- for (split_pipe = pipe_ctx->top_pipe; split_pipe;
- split_pipe = split_pipe->top_pipe)
- if (split_pipe->plane_state->layer_index == test_pipe->plane_state->layer_index) {
- r2_half = split_pipe->plane_res.scl_data.recout;
- r2.x = (r2_half.x < r2.x) ? r2_half.x : r2.x;
- r2.width = r2.width + r2_half.width;
- r2_r = r2.x + r2.width;
- break;
- }
-
- if (r1.x >= r2.x && r1.y >= r2.y && r1_r <= r2_r && r1_b <= r2_b)
- return true;
- }
-
- return false;
-}
-
void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx)
{
struct dc_cursor_position pos_cpy = pipe_ctx->stream->cursor_position;
@@ -3574,7 +3704,7 @@ void dcn10_set_cursor_position(struct pipe_ctx *pipe_ctx)
== PLN_ADDR_TYPE_VIDEO_PROGRESSIVE)
pos_cpy.enable = false;
- if (pos_cpy.enable && dcn10_can_pipe_disable_cursor(pipe_ctx))
+ if (pos_cpy.enable && resource_can_pipe_disable_cursor(pipe_ctx))
pos_cpy.enable = false;
@@ -3967,3 +4097,32 @@ void dcn10_get_dcc_en_bits(struct dc *dc, int *dcc_en_bits)
dcc_en_bits[i] = s->dcc_en ? 1 : 0;
}
}
+
+/**
+ * dcn10_reset_surface_dcc_and_tiling - Set DCC and tiling in DCN to their disable mode.
+ *
+ * @pipe_ctx: Pointer to the pipe context structure.
+ * @plane_state: Surface state
+ * @clear_tiling: If true set tiling to Linear, otherwise does not change tiling
+ *
+ * This function is responsible for call the HUBP block to disable DCC and set
+ * tiling to the linear mode.
+ */
+void dcn10_reset_surface_dcc_and_tiling(struct pipe_ctx *pipe_ctx,
+ struct dc_plane_state *plane_state,
+ bool clear_tiling)
+{
+ struct hubp *hubp = pipe_ctx->plane_res.hubp;
+
+ if (!hubp)
+ return;
+
+ /* if framebuffer is tiled, disable tiling */
+ if (clear_tiling && hubp->funcs->hubp_clear_tiling)
+ hubp->funcs->hubp_clear_tiling(hubp);
+
+ /* force page flip to see the new content of the framebuffer */
+ hubp->funcs->hubp_program_surface_flip_and_addr(hubp,
+ &plane_state->address,
+ true);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.h
index bc5dd68a2408..57d30ea225f2 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_hwseq.h
@@ -50,6 +50,13 @@ void dcn10_optimize_bandwidth(
void dcn10_prepare_bandwidth(
struct dc *dc,
struct dc_state *context);
+void dcn10_wait_for_pipe_update_if_needed(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ bool is_surface_update_only);
+void dcn10_set_wait_for_update_needed_for_pipe(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx);
void dcn10_pipe_control_lock(
struct dc *dc,
struct pipe_ctx *pipe,
@@ -207,4 +214,8 @@ void dcn10_update_visual_confirm_color(
struct pipe_ctx *pipe_ctx,
int mpcc_id);
+void dcn10_reset_surface_dcc_and_tiling(struct pipe_ctx *pipe_ctx,
+ struct dc_plane_state *plane_state,
+ bool clear_tiling);
+
#endif /* __DC_HWSS_DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_init.c
index 5e51e1761707..079c226c1097 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn10/dcn10_init.c
@@ -40,6 +40,7 @@ static const struct hw_sequencer_funcs dcn10_funcs = {
.update_plane_addr = dcn10_update_plane_addr,
.update_dchub = dcn10_update_dchub,
.update_pending_status = dcn10_update_pending_status,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.program_output_csc = dcn10_program_output_csc,
.enable_accelerated_mode = dce110_enable_accelerated_mode,
.enable_timing_synchronization = dcn10_enable_timing_synchronization,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c
index b029ec1b26d3..3207addbd4eb 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.c
@@ -76,6 +76,7 @@ void dcn20_log_color_state(struct dc *dc,
{
struct dc_context *dc_ctx = dc->ctx;
struct resource_pool *pool = dc->res_pool;
+ bool is_gamut_remap_available = false;
int i;
DTN_INFO("DPP: DGAM mode SHAPER mode 3DLUT mode 3DLUT bit depth"
@@ -89,15 +90,15 @@ void dcn20_log_color_state(struct dc *dc,
struct dcn_dpp_state s = {0};
dpp->funcs->dpp_read_state(dpp, &s);
- dpp->funcs->dpp_get_gamut_remap(dpp, &s.gamut_remap);
+ if (dpp->funcs->dpp_get_gamut_remap) {
+ dpp->funcs->dpp_get_gamut_remap(dpp, &s.gamut_remap);
+ is_gamut_remap_available = true;
+ }
if (!s.is_enabled)
continue;
- DTN_INFO("[%2d]: %8s %11s %10s %15s %10s %9s %12s "
- "%010lld %010lld %010lld %010lld "
- "%010lld %010lld %010lld %010lld "
- "%010lld %010lld %010lld %010lld",
+ DTN_INFO("[%2d]: %8s %11s %10s %15s %10s %9s",
dpp->inst,
(s.dgam_lut_mode == 0) ? "Bypass" :
((s.dgam_lut_mode == 1) ? "sRGB" :
@@ -114,10 +115,17 @@ void dcn20_log_color_state(struct dc *dc,
(s.lut3d_bit_depth <= 0) ? "12-bit" : "10-bit",
(s.lut3d_size == 0) ? "17x17x17" : "9x9x9",
(s.rgam_lut_mode == 1) ? "RAM A" :
- ((s.rgam_lut_mode == 1) ? "RAM B" : "Bypass"),
+ ((s.rgam_lut_mode == 1) ? "RAM B" : "Bypass"));
+
+ if (is_gamut_remap_available) {
+ DTN_INFO(" %12s "
+ "%010lld %010lld %010lld %010lld "
+ "%010lld %010lld %010lld %010lld "
+ "%010lld %010lld %010lld %010lld",
+
(s.gamut_remap.gamut_adjust_type == 0) ? "Bypass" :
- ((s.gamut_remap.gamut_adjust_type == 1) ? "HW" :
- "SW"),
+ ((s.gamut_remap.gamut_adjust_type == 1) ? "HW" :
+ "SW"),
s.gamut_remap.temperature_matrix[0].value,
s.gamut_remap.temperature_matrix[1].value,
s.gamut_remap.temperature_matrix[2].value,
@@ -130,6 +138,8 @@ void dcn20_log_color_state(struct dc *dc,
s.gamut_remap.temperature_matrix[9].value,
s.gamut_remap.temperature_matrix[10].value,
s.gamut_remap.temperature_matrix[11].value);
+ }
+
DTN_INFO("\n");
}
DTN_INFO("\n");
@@ -273,14 +283,13 @@ void dcn20_setup_gsl_group_as_lock(
}
/* at this point we want to program whether it's to enable or disable */
- if (pipe_ctx->stream_res.tg->funcs->set_gsl != NULL &&
- pipe_ctx->stream_res.tg->funcs->set_gsl_source_select != NULL) {
+ if (pipe_ctx->stream_res.tg->funcs->set_gsl != NULL) {
pipe_ctx->stream_res.tg->funcs->set_gsl(
pipe_ctx->stream_res.tg,
&gsl);
-
- pipe_ctx->stream_res.tg->funcs->set_gsl_source_select(
- pipe_ctx->stream_res.tg, group_idx, enable ? 4 : 0);
+ if (pipe_ctx->stream_res.tg->funcs->set_gsl_source_select != NULL)
+ pipe_ctx->stream_res.tg->funcs->set_gsl_source_select(
+ pipe_ctx->stream_res.tg, group_idx, enable ? 4 : 0);
} else
BREAK_TO_DEBUGGER();
}
@@ -946,15 +955,13 @@ enum dc_status dcn20_enable_stream_timing(
return DC_ERROR_UNEXPECTED;
}
- hws->funcs.wait_for_blank_complete(pipe_ctx->stream_res.opp);
+ fsleep(stream->timing.v_total * (stream->timing.h_total * 10000u / stream->timing.pix_clk_100hz));
params.vertical_total_min = stream->adjust.v_total_min;
params.vertical_total_max = stream->adjust.v_total_max;
params.vertical_total_mid = stream->adjust.v_total_mid;
params.vertical_total_mid_frame_num = stream->adjust.v_total_mid_frame_num;
- if (pipe_ctx->stream_res.tg->funcs->set_drr)
- pipe_ctx->stream_res.tg->funcs->set_drr(
- pipe_ctx->stream_res.tg, &params);
+ set_drr_and_clear_adjust_pending(pipe_ctx, stream, &params);
// DRR should set trigger event to monitor surface update event
if (stream->adjust.v_total_min != 0 && stream->adjust.v_total_max != 0)
@@ -1266,14 +1273,18 @@ static void dcn20_power_on_plane_resources(
struct dce_hwseq *hws,
struct pipe_ctx *pipe_ctx)
{
+ uint32_t org_ip_request_cntl = 0;
+
DC_LOGGER_INIT(hws->ctx->logger);
if (hws->funcs.dpp_root_clock_control)
hws->funcs.dpp_root_clock_control(hws, pipe_ctx->plane_res.dpp->inst, true);
if (REG(DC_IP_REQUEST_CNTL)) {
- REG_SET(DC_IP_REQUEST_CNTL, 0,
- IP_REQUEST_EN, 1);
+ REG_GET(DC_IP_REQUEST_CNTL, IP_REQUEST_EN, &org_ip_request_cntl);
+ if (org_ip_request_cntl == 0)
+ REG_SET(DC_IP_REQUEST_CNTL, 0,
+ IP_REQUEST_EN, 1);
if (hws->funcs.dpp_pg_control)
hws->funcs.dpp_pg_control(hws, pipe_ctx->plane_res.dpp->inst, true);
@@ -1281,14 +1292,16 @@ static void dcn20_power_on_plane_resources(
if (hws->funcs.hubp_pg_control)
hws->funcs.hubp_pg_control(hws, pipe_ctx->plane_res.hubp->inst, true);
- REG_SET(DC_IP_REQUEST_CNTL, 0,
- IP_REQUEST_EN, 0);
+ if (org_ip_request_cntl == 0)
+ REG_SET(DC_IP_REQUEST_CNTL, 0,
+ IP_REQUEST_EN, 0);
+
DC_LOG_DEBUG(
"Un-gated front end for pipe %d\n", pipe_ctx->plane_res.hubp->inst);
}
}
-static void dcn20_enable_plane(struct dc *dc, struct pipe_ctx *pipe_ctx,
+void dcn20_enable_plane(struct dc *dc, struct pipe_ctx *pipe_ctx,
struct dc_state *context)
{
//if (dc->debug.sanity_checks) {
@@ -1467,7 +1480,7 @@ void dcn20_pipe_control_lock(
}
}
-static void dcn20_detect_pipe_changes(struct dc_state *old_state,
+void dcn20_detect_pipe_changes(struct dc_state *old_state,
struct dc_state *new_state,
struct pipe_ctx *old_pipe,
struct pipe_ctx *new_pipe)
@@ -1655,7 +1668,7 @@ static void dcn20_detect_pipe_changes(struct dc_state *old_state,
}
}
-static void dcn20_update_dchubp_dpp(
+void dcn20_update_dchubp_dpp(
struct dc *dc,
struct pipe_ctx *pipe_ctx,
struct dc_state *context)
@@ -1678,25 +1691,41 @@ static void dcn20_update_dchubp_dpp(
* VTG is within DCHUBBUB which is commond block share by each pipe HUBP.
* VTG is 1:1 mapping with OTG. Each pipe HUBP will select which VTG
*/
+
if (pipe_ctx->update_flags.bits.hubp_rq_dlg_ttu) {
hubp->funcs->hubp_vtg_sel(hubp, pipe_ctx->stream_res.tg->inst);
- hubp->funcs->hubp_setup(
- hubp,
- &pipe_ctx->dlg_regs,
- &pipe_ctx->ttu_regs,
- &pipe_ctx->rq_regs,
- &pipe_ctx->pipe_dlg_param);
+ if (hubp->funcs->hubp_setup2) {
+ hubp->funcs->hubp_setup2(
+ hubp,
+ &pipe_ctx->hubp_regs,
+ &pipe_ctx->global_sync,
+ &pipe_ctx->stream->timing);
+ } else {
+ hubp->funcs->hubp_setup(
+ hubp,
+ &pipe_ctx->dlg_regs,
+ &pipe_ctx->ttu_regs,
+ &pipe_ctx->rq_regs,
+ &pipe_ctx->pipe_dlg_param);
+ }
}
if (pipe_ctx->update_flags.bits.unbounded_req && hubp->funcs->set_unbounded_requesting)
hubp->funcs->set_unbounded_requesting(hubp, pipe_ctx->unbounded_req);
- if (pipe_ctx->update_flags.bits.hubp_interdependent)
- hubp->funcs->hubp_setup_interdependent(
- hubp,
- &pipe_ctx->dlg_regs,
- &pipe_ctx->ttu_regs);
+ if (pipe_ctx->update_flags.bits.hubp_interdependent) {
+ if (hubp->funcs->hubp_setup_interdependent2) {
+ hubp->funcs->hubp_setup_interdependent2(
+ hubp,
+ &pipe_ctx->hubp_regs);
+ } else {
+ hubp->funcs->hubp_setup_interdependent(
+ hubp,
+ &pipe_ctx->dlg_regs,
+ &pipe_ctx->ttu_regs);
+ }
+ }
if (pipe_ctx->update_flags.bits.enable ||
pipe_ctx->update_flags.bits.plane_changed ||
@@ -1756,10 +1785,9 @@ static void dcn20_update_dchubp_dpp(
&pipe_ctx->plane_res.scl_data.viewport_c);
viewport_changed = true;
}
- if (hubp->funcs->hubp_program_mcache_id_and_split_coordinate)
- hubp->funcs->hubp_program_mcache_id_and_split_coordinate(
- hubp,
- &pipe_ctx->mcache_regs);
+
+ if (hubp->funcs->hubp_program_mcache_id_and_split_coordinate)
+ hubp->funcs->hubp_program_mcache_id_and_split_coordinate(hubp, &pipe_ctx->mcache_regs);
/* Any updates are handled in dc interface, just need to apply existing for plane enable */
if ((pipe_ctx->update_flags.bits.enable || pipe_ctx->update_flags.bits.opp_changed ||
@@ -1838,7 +1866,7 @@ static void dcn20_update_dchubp_dpp(
hubp->funcs->phantom_hubp_post_enable(hubp);
}
-static int calculate_vready_offset_for_group(struct pipe_ctx *pipe)
+static int dcn20_calculate_vready_offset_for_group(struct pipe_ctx *pipe)
{
struct pipe_ctx *other_pipe;
int vready_offset = pipe->pipe_dlg_param.vready_offset;
@@ -1864,6 +1892,30 @@ static int calculate_vready_offset_for_group(struct pipe_ctx *pipe)
return vready_offset;
}
+static void dcn20_program_tg(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_state *context,
+ struct dce_hwseq *hws)
+{
+ pipe_ctx->stream_res.tg->funcs->program_global_sync(
+ pipe_ctx->stream_res.tg,
+ dcn20_calculate_vready_offset_for_group(pipe_ctx),
+ pipe_ctx->pipe_dlg_param.vstartup_start,
+ pipe_ctx->pipe_dlg_param.vupdate_offset,
+ pipe_ctx->pipe_dlg_param.vupdate_width,
+ pipe_ctx->pipe_dlg_param.pstate_keepout);
+
+ if (dc_state_get_pipe_subvp_type(context, pipe_ctx) != SUBVP_PHANTOM)
+ pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VACTIVE);
+
+ pipe_ctx->stream_res.tg->funcs->set_vtg_params(
+ pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing, true);
+
+ if (hws->funcs.setup_vupdate_interrupt)
+ hws->funcs.setup_vupdate_interrupt(dc, pipe_ctx);
+}
+
static void dcn20_program_pipe(
struct dc *dc,
struct pipe_ctx *pipe_ctx,
@@ -1874,33 +1926,17 @@ static void dcn20_program_pipe(
/* Only need to unblank on top pipe */
if (resource_is_pipe_type(pipe_ctx, OTG_MASTER)) {
if (pipe_ctx->update_flags.bits.enable ||
- pipe_ctx->update_flags.bits.odm ||
- pipe_ctx->stream->update_flags.bits.abm_level)
+ pipe_ctx->update_flags.bits.odm ||
+ pipe_ctx->stream->update_flags.bits.abm_level)
hws->funcs.blank_pixel_data(dc, pipe_ctx,
- !pipe_ctx->plane_state ||
- !pipe_ctx->plane_state->visible);
+ !pipe_ctx->plane_state ||
+ !pipe_ctx->plane_state->visible);
}
/* Only update TG on top pipe */
if (pipe_ctx->update_flags.bits.global_sync && !pipe_ctx->top_pipe
- && !pipe_ctx->prev_odm_pipe) {
- pipe_ctx->stream_res.tg->funcs->program_global_sync(
- pipe_ctx->stream_res.tg,
- calculate_vready_offset_for_group(pipe_ctx),
- pipe_ctx->pipe_dlg_param.vstartup_start,
- pipe_ctx->pipe_dlg_param.vupdate_offset,
- pipe_ctx->pipe_dlg_param.vupdate_width,
- pipe_ctx->pipe_dlg_param.pstate_keepout);
-
- if (dc_state_get_pipe_subvp_type(context, pipe_ctx) != SUBVP_PHANTOM)
- pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VACTIVE);
-
- pipe_ctx->stream_res.tg->funcs->set_vtg_params(
- pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing, true);
-
- if (hws->funcs.setup_vupdate_interrupt)
- hws->funcs.setup_vupdate_interrupt(dc, pipe_ctx);
- }
+ && !pipe_ctx->prev_odm_pipe)
+ dcn20_program_tg(dc, pipe_ctx, context, hws);
if (pipe_ctx->update_flags.bits.odm)
hws->funcs.update_odm(dc, context, pipe_ctx);
@@ -1931,22 +1967,14 @@ static void dcn20_program_pipe(
dcn20_update_dchubp_dpp(dc, pipe_ctx, context);
if (pipe_ctx->plane_state && (pipe_ctx->update_flags.bits.enable ||
- pipe_ctx->plane_state->update_flags.bits.hdr_mult))
+ pipe_ctx->plane_state->update_flags.bits.hdr_mult))
hws->funcs.set_hdr_multiplier(pipe_ctx);
- if (hws->funcs.populate_mcm_luts) {
- if (pipe_ctx->plane_state) {
- hws->funcs.populate_mcm_luts(dc, pipe_ctx, pipe_ctx->plane_state->mcm_luts,
- pipe_ctx->plane_state->lut_bank_a);
- pipe_ctx->plane_state->lut_bank_a = !pipe_ctx->plane_state->lut_bank_a;
- }
- }
-
if (pipe_ctx->plane_state &&
- (pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change ||
- pipe_ctx->plane_state->update_flags.bits.gamma_change ||
- pipe_ctx->plane_state->update_flags.bits.lut_3d ||
- pipe_ctx->update_flags.bits.enable))
+ (pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change ||
+ pipe_ctx->plane_state->update_flags.bits.gamma_change ||
+ pipe_ctx->plane_state->update_flags.bits.lut_3d ||
+ pipe_ctx->update_flags.bits.enable))
hws->funcs.set_input_transfer_func(dc, pipe_ctx, pipe_ctx->plane_state);
/* dcn10_translate_regamma_to_hw_format takes 750us to finish
@@ -1954,10 +1982,10 @@ static void dcn20_program_pipe(
* updating on slave planes
*/
if (pipe_ctx->update_flags.bits.enable ||
- pipe_ctx->update_flags.bits.plane_changed ||
- pipe_ctx->stream->update_flags.bits.out_tf ||
- (pipe_ctx->plane_state &&
- pipe_ctx->plane_state->update_flags.bits.output_tf_change))
+ pipe_ctx->update_flags.bits.plane_changed ||
+ pipe_ctx->stream->update_flags.bits.out_tf ||
+ (pipe_ctx->plane_state &&
+ pipe_ctx->plane_state->update_flags.bits.output_tf_change))
hws->funcs.set_output_transfer_func(dc, pipe_ctx, pipe_ctx->stream);
/* If the pipe has been enabled or has a different opp, we
@@ -1966,7 +1994,7 @@ static void dcn20_program_pipe(
* causes a different pipe to be chosen to odm combine with.
*/
if (pipe_ctx->update_flags.bits.enable
- || pipe_ctx->update_flags.bits.opp_changed) {
+ || pipe_ctx->update_flags.bits.opp_changed) {
pipe_ctx->stream_res.opp->funcs->opp_set_dyn_expansion(
pipe_ctx->stream_res.opp,
@@ -1996,14 +2024,14 @@ static void dcn20_program_pipe(
memset(&params, 0, sizeof(params));
odm_opp->funcs->opp_program_bit_depth_reduction(odm_opp, &params);
dc->hwss.set_disp_pattern_generator(dc,
- pipe_ctx,
- pipe_ctx->stream_res.test_pattern_params.test_pattern,
- pipe_ctx->stream_res.test_pattern_params.color_space,
- pipe_ctx->stream_res.test_pattern_params.color_depth,
- NULL,
- pipe_ctx->stream_res.test_pattern_params.width,
- pipe_ctx->stream_res.test_pattern_params.height,
- pipe_ctx->stream_res.test_pattern_params.offset);
+ pipe_ctx,
+ pipe_ctx->stream_res.test_pattern_params.test_pattern,
+ pipe_ctx->stream_res.test_pattern_params.color_space,
+ pipe_ctx->stream_res.test_pattern_params.color_depth,
+ NULL,
+ pipe_ctx->stream_res.test_pattern_params.width,
+ pipe_ctx->stream_res.test_pattern_params.height,
+ pipe_ctx->stream_res.test_pattern_params.offset);
}
}
@@ -2012,11 +2040,12 @@ void dcn20_program_front_end_for_ctx(
struct dc_state *context)
{
int i;
- struct dce_hwseq *hws = dc->hwseq;
- DC_LOGGER_INIT(dc->ctx->logger);
unsigned int prev_hubp_count = 0;
unsigned int hubp_count = 0;
- struct pipe_ctx *pipe;
+ struct dce_hwseq *hws = dc->hwseq;
+ struct pipe_ctx *pipe = NULL;
+
+ DC_LOGGER_INIT(dc->ctx->logger);
if (resource_is_pipe_topology_changed(dc->current_state, context))
resource_log_pipe_topology_update(dc, context);
@@ -2025,11 +2054,11 @@ void dcn20_program_front_end_for_ctx(
for (i = 0; i < dc->res_pool->pipe_count; i++) {
pipe = &context->res_ctx.pipe_ctx[i];
- if (!pipe->top_pipe && !pipe->prev_odm_pipe && pipe->plane_state) {
+ if (pipe->plane_state) {
ASSERT(!pipe->plane_state->triplebuffer_flips);
/*turn off triple buffer for full update*/
dc->hwss.program_triplebuffer(
- dc, pipe, pipe->plane_state->triplebuffer_flips);
+ dc, pipe, pipe->plane_state->triplebuffer_flips);
}
}
}
@@ -2044,30 +2073,31 @@ void dcn20_program_front_end_for_ctx(
if (prev_hubp_count == 0 && hubp_count > 0) {
if (dc->res_pool->hubbub->funcs->force_pstate_change_control)
dc->res_pool->hubbub->funcs->force_pstate_change_control(
- dc->res_pool->hubbub, true, false);
+ dc->res_pool->hubbub, true, false);
udelay(500);
}
/* Set pipe update flags and lock pipes */
for (i = 0; i < dc->res_pool->pipe_count; i++)
dcn20_detect_pipe_changes(dc->current_state, context, &dc->current_state->res_ctx.pipe_ctx[i],
- &context->res_ctx.pipe_ctx[i]);
+ &context->res_ctx.pipe_ctx[i]);
/* When disabling phantom pipes, turn on phantom OTG first (so we can get double
* buffer updates properly)
*/
for (i = 0; i < dc->res_pool->pipe_count; i++) {
struct dc_stream_state *stream = dc->current_state->res_ctx.pipe_ctx[i].stream;
+
pipe = &dc->current_state->res_ctx.pipe_ctx[i];
if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable && stream &&
- dc_state_get_pipe_subvp_type(dc->current_state, pipe) == SUBVP_PHANTOM) {
+ dc_state_get_pipe_subvp_type(dc->current_state, pipe) == SUBVP_PHANTOM) {
struct timing_generator *tg = dc->current_state->res_ctx.pipe_ctx[i].stream_res.tg;
if (tg->funcs->enable_crtc) {
- if (dc->hwseq->funcs.blank_pixel_data) {
+ if (dc->hwseq->funcs.blank_pixel_data)
dc->hwseq->funcs.blank_pixel_data(dc, pipe, true);
- }
+
tg->funcs->enable_crtc(tg);
}
}
@@ -2075,15 +2105,15 @@ void dcn20_program_front_end_for_ctx(
/* OTG blank before disabling all front ends */
for (i = 0; i < dc->res_pool->pipe_count; i++)
if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable
- && !context->res_ctx.pipe_ctx[i].top_pipe
- && !context->res_ctx.pipe_ctx[i].prev_odm_pipe
- && context->res_ctx.pipe_ctx[i].stream)
+ && !context->res_ctx.pipe_ctx[i].top_pipe
+ && !context->res_ctx.pipe_ctx[i].prev_odm_pipe
+ && context->res_ctx.pipe_ctx[i].stream)
hws->funcs.blank_pixel_data(dc, &context->res_ctx.pipe_ctx[i], true);
/* Disconnect mpcc */
for (i = 0; i < dc->res_pool->pipe_count; i++)
if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable
- || context->res_ctx.pipe_ctx[i].update_flags.bits.opp_changed) {
+ || context->res_ctx.pipe_ctx[i].update_flags.bits.opp_changed) {
struct hubbub *hubbub = dc->res_pool->hubbub;
/* Phantom pipe DET should be 0, but if a pipe in use is being transitioned to phantom
@@ -2093,13 +2123,18 @@ void dcn20_program_front_end_for_ctx(
* DET allocation.
*/
if ((context->res_ctx.pipe_ctx[i].update_flags.bits.disable ||
- (context->res_ctx.pipe_ctx[i].plane_state && dc_state_get_pipe_subvp_type(context, &context->res_ctx.pipe_ctx[i]) == SUBVP_PHANTOM))) {
+ (context->res_ctx.pipe_ctx[i].plane_state &&
+ dc_state_get_pipe_subvp_type(context, &context->res_ctx.pipe_ctx[i])
+ == SUBVP_PHANTOM))) {
if (hubbub->funcs->program_det_size)
- hubbub->funcs->program_det_size(hubbub, dc->current_state->res_ctx.pipe_ctx[i].plane_res.hubp->inst, 0);
+ hubbub->funcs->program_det_size(hubbub,
+ dc->current_state->res_ctx.pipe_ctx[i].plane_res.hubp->inst, 0);
if (dc->res_pool->hubbub->funcs->program_det_segments)
- dc->res_pool->hubbub->funcs->program_det_segments(hubbub, dc->current_state->res_ctx.pipe_ctx[i].plane_res.hubp->inst, 0);
+ dc->res_pool->hubbub->funcs->program_det_segments(
+ hubbub, dc->current_state->res_ctx.pipe_ctx[i].plane_res.hubp->inst, 0);
}
- hws->funcs.plane_atomic_disconnect(dc, dc->current_state, &dc->current_state->res_ctx.pipe_ctx[i]);
+ hws->funcs.plane_atomic_disconnect(dc, dc->current_state,
+ &dc->current_state->res_ctx.pipe_ctx[i]);
DC_LOG_DC("Reset mpcc for pipe %d\n", dc->current_state->res_ctx.pipe_ctx[i].pipe_idx);
}
@@ -2107,9 +2142,9 @@ void dcn20_program_front_end_for_ctx(
for (i = 0; i < dc->res_pool->pipe_count; i++) {
pipe = &context->res_ctx.pipe_ctx[i];
if (resource_is_pipe_type(pipe, OTG_MASTER) &&
- !resource_is_pipe_type(pipe, DPP_PIPE) &&
- pipe->update_flags.bits.odm &&
- hws->funcs.update_odm)
+ !resource_is_pipe_type(pipe, DPP_PIPE) &&
+ pipe->update_flags.bits.odm &&
+ hws->funcs.update_odm)
hws->funcs.update_odm(dc, context, pipe);
}
@@ -2127,25 +2162,28 @@ void dcn20_program_front_end_for_ctx(
else {
/* Don't program phantom pipes in the regular front end programming sequence.
* There is an MPO transition case where a pipe being used by a video plane is
- * transitioned directly to be a phantom pipe when closing the MPO video. However
- * the phantom pipe will program a new HUBP_VTG_SEL (update takes place right away),
- * but the MPO still exists until the double buffered update of the main pipe so we
- * will get a frame of underflow if the phantom pipe is programmed here.
+ * transitioned directly to be a phantom pipe when closing the MPO video.
+ * However the phantom pipe will program a new HUBP_VTG_SEL (update takes place
+ * right away) but the MPO still exists until the double buffered update of the
+ * main pipe so we will get a frame of underflow if the phantom pipe is
+ * programmed here.
*/
- if (pipe->stream && dc_state_get_pipe_subvp_type(context, pipe) != SUBVP_PHANTOM)
+ if (pipe->stream &&
+ dc_state_get_pipe_subvp_type(context, pipe) != SUBVP_PHANTOM)
dcn20_program_pipe(dc, pipe, context);
}
pipe = pipe->bottom_pipe;
}
}
+
/* Program secondary blending tree and writeback pipes */
pipe = &context->res_ctx.pipe_ctx[i];
if (!pipe->top_pipe && !pipe->prev_odm_pipe
- && pipe->stream && pipe->stream->num_wb_info > 0
- && (pipe->update_flags.raw || (pipe->plane_state && pipe->plane_state->update_flags.raw)
- || pipe->stream->update_flags.raw)
- && hws->funcs.program_all_writeback_pipes_in_tree)
+ && pipe->stream && pipe->stream->num_wb_info > 0
+ && (pipe->update_flags.raw || (pipe->plane_state && pipe->plane_state->update_flags.raw)
+ || pipe->stream->update_flags.raw)
+ && hws->funcs.program_all_writeback_pipes_in_tree)
hws->funcs.program_all_writeback_pipes_in_tree(dc, pipe->stream, context);
/* Avoid underflow by check of pipe line read when adding 2nd plane. */
@@ -2164,7 +2202,7 @@ void dcn20_program_front_end_for_ctx(
* buffered pending status clear and reset opp head pipe's none double buffered
* registers to their initial state.
*/
-static void post_unlock_reset_opp(struct dc *dc,
+void dcn20_post_unlock_reset_opp(struct dc *dc,
struct pipe_ctx *opp_head)
{
struct display_stream_compressor *dsc = opp_head->stream_res.dsc;
@@ -2201,16 +2239,17 @@ void dcn20_post_unlock_program_front_end(
struct dc *dc,
struct dc_state *context)
{
- int i;
- const unsigned int TIMEOUT_FOR_PIPE_ENABLE_US = 100000;
+ // Timeout for pipe enable
+ unsigned int timeout_us = 100000;
unsigned int polling_interval_us = 1;
struct dce_hwseq *hwseq = dc->hwseq;
+ int i;
for (i = 0; i < dc->res_pool->pipe_count; i++)
if (resource_is_pipe_type(&dc->current_state->res_ctx.pipe_ctx[i], OPP_HEAD) &&
- !resource_is_pipe_type(&context->res_ctx.pipe_ctx[i], OPP_HEAD))
- post_unlock_reset_opp(dc,
- &dc->current_state->res_ctx.pipe_ctx[i]);
+ !resource_is_pipe_type(&context->res_ctx.pipe_ctx[i], OPP_HEAD))
+ dcn20_post_unlock_reset_opp(dc,
+ &dc->current_state->res_ctx.pipe_ctx[i]);
for (i = 0; i < dc->res_pool->pipe_count; i++)
if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable)
@@ -2226,11 +2265,12 @@ void dcn20_post_unlock_program_front_end(
struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
// Don't check flip pending on phantom pipes
if (pipe->plane_state && !pipe->top_pipe && pipe->update_flags.bits.enable &&
- dc_state_get_pipe_subvp_type(context, pipe) != SUBVP_PHANTOM) {
+ dc_state_get_pipe_subvp_type(context, pipe) != SUBVP_PHANTOM) {
struct hubp *hubp = pipe->plane_res.hubp;
int j = 0;
- for (j = 0; j < TIMEOUT_FOR_PIPE_ENABLE_US / polling_interval_us
- && hubp->funcs->hubp_is_flip_pending(hubp); j++)
+
+ for (j = 0; j < timeout_us / polling_interval_us
+ && hubp->funcs->hubp_is_flip_pending(hubp); j++)
udelay(polling_interval_us);
}
}
@@ -2244,15 +2284,14 @@ void dcn20_post_unlock_program_front_end(
* before we've transitioned to 2:1 or 4:1
*/
if (resource_is_pipe_type(old_pipe, OTG_MASTER) && resource_is_pipe_type(pipe, OTG_MASTER) &&
- resource_get_odm_slice_count(old_pipe) < resource_get_odm_slice_count(pipe) &&
- dc_state_get_pipe_subvp_type(context, pipe) != SUBVP_PHANTOM) {
+ resource_get_odm_slice_count(old_pipe) < resource_get_odm_slice_count(pipe) &&
+ dc_state_get_pipe_subvp_type(context, pipe) != SUBVP_PHANTOM) {
int j = 0;
struct timing_generator *tg = pipe->stream_res.tg;
-
if (tg->funcs->get_optc_double_buffer_pending) {
- for (j = 0; j < TIMEOUT_FOR_PIPE_ENABLE_US / polling_interval_us
- && tg->funcs->get_optc_double_buffer_pending(tg); j++)
+ for (j = 0; j < timeout_us / polling_interval_us
+ && tg->funcs->get_optc_double_buffer_pending(tg); j++)
udelay(polling_interval_us);
}
}
@@ -2260,7 +2299,7 @@ void dcn20_post_unlock_program_front_end(
if (dc->res_pool->hubbub->funcs->force_pstate_change_control)
dc->res_pool->hubbub->funcs->force_pstate_change_control(
- dc->res_pool->hubbub, false, false);
+ dc->res_pool->hubbub, false, false);
for (i = 0; i < dc->res_pool->pipe_count; i++) {
struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
@@ -2291,11 +2330,11 @@ void dcn20_post_unlock_program_front_end(
return;
/* P-State support transitions:
- * Natural -> FPO: P-State disabled in prepare, force disallow anytime is safe
- * FPO -> Natural: Unforce anytime after FW disable is safe (P-State will assert naturally)
- * Unsupported -> FPO: P-State enabled in optimize, force disallow anytime is safe
- * FPO -> Unsupported: P-State disabled in prepare, unforce disallow anytime is safe
- * FPO <-> SubVP: Force disallow is maintained on the FPO / SubVP pipes
+ * Natural -> FPO: P-State disabled in prepare, force disallow anytime is safe
+ * FPO -> Natural: Unforce anytime after FW disable is safe (P-State will assert naturally)
+ * Unsupported -> FPO: P-State enabled in optimize, force disallow anytime is safe
+ * FPO -> Unsupported: P-State disabled in prepare, unforce disallow anytime is safe
+ * FPO <-> SubVP: Force disallow is maintained on the FPO / SubVP pipes
*/
if (hwseq->funcs.update_force_pstate)
dc->hwseq->funcs.update_force_pstate(dc, context);
@@ -2310,12 +2349,11 @@ void dcn20_post_unlock_program_front_end(
if (hwseq->wa.DEGVIDCN21)
dc->res_pool->hubbub->funcs->apply_DEDCN21_147_wa(dc->res_pool->hubbub);
-
/* WA for stutter underflow during MPO transitions when adding 2nd plane */
if (hwseq->wa.disallow_self_refresh_during_multi_plane_transition) {
if (dc->current_state->stream_status[0].plane_count == 1 &&
- context->stream_status[0].plane_count > 1) {
+ context->stream_status[0].plane_count > 1) {
struct timing_generator *tg = dc->res_pool->timing_generators[0];
@@ -2445,7 +2483,7 @@ bool dcn20_update_bandwidth(
struct dce_hwseq *hws = dc->hwseq;
/* recalculate DML parameters */
- if (!dc->res_pool->funcs->validate_bandwidth(dc, context, false))
+ if (dc->res_pool->funcs->validate_bandwidth(dc, context, DC_VALIDATE_MODE_AND_PROGRAMMING) != DC_OK)
return false;
/* apply updated bandwidth parameters */
@@ -2463,7 +2501,7 @@ bool dcn20_update_bandwidth(
pipe_ctx->stream_res.tg->funcs->program_global_sync(
pipe_ctx->stream_res.tg,
- calculate_vready_offset_for_group(pipe_ctx),
+ dcn20_calculate_vready_offset_for_group(pipe_ctx),
pipe_ctx->pipe_dlg_param.vstartup_start,
pipe_ctx->pipe_dlg_param.vupdate_offset,
pipe_ctx->pipe_dlg_param.vupdate_width,
@@ -2739,7 +2777,8 @@ void dcn20_unblank_stream(struct pipe_ctx *pipe_ctx,
} else if (dc_is_dp_signal(pipe_ctx->stream->signal)) {
if (is_two_pixels_per_container || params.opp_cnt > 1)
params.timing.pix_clk_100hz /= 2;
- pipe_ctx->stream_res.stream_enc->funcs->dp_set_odm_combine(
+ if (pipe_ctx->stream_res.stream_enc->funcs->dp_set_odm_combine)
+ pipe_ctx->stream_res.stream_enc->funcs->dp_set_odm_combine(
pipe_ctx->stream_res.stream_enc, params.opp_cnt > 1);
pipe_ctx->stream_res.stream_enc->funcs->dp_unblank(link, pipe_ctx->stream_res.stream_enc, &params);
}
@@ -2768,6 +2807,8 @@ void dcn20_reset_back_end_for_pipe(
{
struct dc_link *link = pipe_ctx->stream->link;
const struct link_hwss *link_hwss = get_link_hwss(link, &pipe_ctx->link_res);
+ struct dccg *dccg = dc->res_pool->dccg;
+ struct dtbclk_dto_params dto_params = {0};
DC_LOGGER_INIT(dc->ctx->logger);
if (pipe_ctx->stream_res.stream_enc == NULL) {
@@ -2816,9 +2857,7 @@ void dcn20_reset_back_end_for_pipe(
pipe_ctx->stream_res.tg->funcs->set_odm_bypass(
pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing);
- if (pipe_ctx->stream_res.tg->funcs->set_drr)
- pipe_ctx->stream_res.tg->funcs->set_drr(
- pipe_ctx->stream_res.tg, NULL);
+ set_drr_and_clear_adjust_pending(pipe_ctx, pipe_ctx->stream, NULL);
/* TODO - convert symclk_ref_cnts for otg to a bit map to solve
* the case where the same symclk is shared across multiple otg
* instances
@@ -2830,6 +2869,13 @@ void dcn20_reset_back_end_for_pipe(
&pipe_ctx->link_res, pipe_ctx->stream->signal);
link->phy_state.symclk_state = SYMCLK_OFF_TX_OFF;
}
+ if (dc->link_srv->dp_is_128b_132b_signal(pipe_ctx) && dccg
+ && dc->ctx->dce_version >= DCN_VERSION_3_5) {
+ dto_params.otg_inst = pipe_ctx->stream_res.tg->inst;
+ dto_params.timing = &pipe_ctx->stream->timing;
+ if (dccg && dccg->funcs->set_dtbclk_dto)
+ dccg->funcs->set_dtbclk_dto(dccg, &dto_params);
+ }
}
/*
@@ -2980,9 +3026,12 @@ void dcn20_enable_stream(struct pipe_ctx *pipe_ctx)
enum phyd32clk_clock_source phyd32clk;
int dp_hpo_inst;
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(pipe_ctx->stream->link);
+ struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
struct stream_encoder *stream_enc = pipe_ctx->stream_res.stream_enc;
+ if (!dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+
if (dc->link_srv->dp_is_128b_132b_signal(pipe_ctx)) {
dto_params.otg_inst = tg->inst;
dto_params.pixclk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10;
@@ -2994,7 +3043,11 @@ void dcn20_enable_stream(struct pipe_ctx *pipe_ctx)
dccg->funcs->set_dpstreamclk(dccg, DTBCLK0, tg->inst, dp_hpo_inst);
phyd32clk = get_phyd32clk_src(link);
- dccg->funcs->enable_symclk32_se(dccg, dp_hpo_inst, phyd32clk);
+ if (link->cur_link_settings.link_rate == LINK_RATE_UNKNOWN) {
+ dccg->funcs->disable_symclk32_se(dccg, dp_hpo_inst);
+ } else {
+ dccg->funcs->enable_symclk32_se(dccg, dp_hpo_inst, phyd32clk);
+ }
} else {
if (dccg->funcs->enable_symclk_se)
dccg->funcs->enable_symclk_se(dccg, stream_enc->stream_enc_inst,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.h
index 5c874f7b0683..9d1ad3b29ca5 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_hwseq.h
@@ -154,6 +154,21 @@ void dcn20_setup_gsl_group_as_lock(
const struct dc *dc,
struct pipe_ctx *pipe_ctx,
bool enable);
-
+void dcn20_detect_pipe_changes(
+ struct dc_state *old_state,
+ struct dc_state *new_state,
+ struct pipe_ctx *old_pipe,
+ struct pipe_ctx *new_pipe);
+void dcn20_enable_plane(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_state *context);
+void dcn20_update_dchubp_dpp(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_state *context);
+void dcn20_post_unlock_reset_opp(
+ struct dc *dc,
+ struct pipe_ctx *opp_head);
#endif /* __DC_HWSS_DCN20_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_init.c
index 32707b344f0b..ad253c586ea1 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn20/dcn20_init.c
@@ -36,6 +36,7 @@ static const struct hw_sequencer_funcs dcn20_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn201/dcn201_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn201/dcn201_init.c
index 78351408e864..dec57fb4c05c 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn201/dcn201_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn201/dcn201_init.c
@@ -36,6 +36,7 @@ static const struct hw_sequencer_funcs dcn201_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn10_post_unlock_program_front_end,
.update_plane_addr = dcn201_update_plane_addr,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_init.c
index e044e9e0a3a1..c7701a8b574a 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn21/dcn21_init.c
@@ -37,6 +37,7 @@ static const struct hw_sequencer_funcs dcn21_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c
index e89ebfda4873..37a239219dfe 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_hwseq.c
@@ -74,6 +74,7 @@ void dcn30_log_color_state(struct dc *dc,
{
struct dc_context *dc_ctx = dc->ctx;
struct resource_pool *pool = dc->res_pool;
+ bool is_gamut_remap_available = false;
int i;
DTN_INFO("DPP: DGAM ROM DGAM ROM type DGAM LUT SHAPER mode"
@@ -88,16 +89,16 @@ void dcn30_log_color_state(struct dc *dc,
struct dcn_dpp_state s = {0};
dpp->funcs->dpp_read_state(dpp, &s);
- dpp->funcs->dpp_get_gamut_remap(dpp, &s.gamut_remap);
+
+ if (dpp->funcs->dpp_get_gamut_remap) {
+ dpp->funcs->dpp_get_gamut_remap(dpp, &s.gamut_remap);
+ is_gamut_remap_available = true;
+ }
if (!s.is_enabled)
continue;
- DTN_INFO("[%2d]: %7x %13s %8s %11s %10s %15s %10s %9s"
- " %12s "
- "%010lld %010lld %010lld %010lld "
- "%010lld %010lld %010lld %010lld "
- "%010lld %010lld %010lld %010lld",
+ DTN_INFO("[%2d]: %7x %13s %8s %11s %10s %15s %10s %9s",
dpp->inst,
s.pre_dgam_mode,
(s.pre_dgam_select == 0) ? "sRGB" :
@@ -121,7 +122,14 @@ void dcn30_log_color_state(struct dc *dc,
(s.lut3d_size == 0) ? "17x17x17" : "9x9x9",
(s.rgam_lut_mode == 0) ? "Bypass" :
((s.rgam_lut_mode == 1) ? "RAM A" :
- "RAM B"),
+ "RAM B"));
+
+ if (is_gamut_remap_available) {
+ DTN_INFO(" %12s "
+ "%010lld %010lld %010lld %010lld "
+ "%010lld %010lld %010lld %010lld "
+ "%010lld %010lld %010lld %010lld",
+
(s.gamut_remap.gamut_adjust_type == 0) ? "Bypass" :
((s.gamut_remap.gamut_adjust_type == 1) ? "HW" :
"SW"),
@@ -137,6 +145,8 @@ void dcn30_log_color_state(struct dc *dc,
s.gamut_remap.temperature_matrix[9].value,
s.gamut_remap.temperature_matrix[10].value,
s.gamut_remap.temperature_matrix[11].value);
+ }
+
DTN_INFO("\n");
}
DTN_INFO("\n");
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c
index 0e8d32e3dbae..2ac5d54d1626 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn30/dcn30_init.c
@@ -37,6 +37,7 @@ static const struct hw_sequencer_funcs dcn30_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
@@ -86,7 +87,6 @@ static const struct hw_sequencer_funcs dcn30_funcs = {
.enable_writeback = dcn30_enable_writeback,
.disable_writeback = dcn30_disable_writeback,
.update_writeback = dcn30_update_writeback,
- .mmhubbub_warmup = dcn30_mmhubbub_warmup,
.dmdata_status_done = dcn20_dmdata_status_done,
.program_dmdata_engine = dcn30_program_dmdata_engine,
.set_dmdata_attributes = dcn20_set_dmdata_attributes,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.c
index 780ce4c064aa..8d7ceb7b32b8 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn301/dcn301_init.c
@@ -39,6 +39,7 @@ static const struct hw_sequencer_funcs dcn301_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
@@ -86,7 +87,6 @@ static const struct hw_sequencer_funcs dcn301_funcs = {
.enable_writeback = dcn30_enable_writeback,
.disable_writeback = dcn30_disable_writeback,
.update_writeback = dcn30_update_writeback,
- .mmhubbub_warmup = dcn30_mmhubbub_warmup,
.dmdata_status_done = dcn20_dmdata_status_done,
.program_dmdata_engine = dcn30_program_dmdata_engine,
.set_dmdata_attributes = dcn20_set_dmdata_attributes,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c
index 03ba01f4ace1..8ba934b83957 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_hwseq.c
@@ -396,6 +396,11 @@ void dcn31_update_info_frame(struct pipe_ctx *pipe_ctx)
pipe_ctx->stream_res.stream_enc,
&pipe_ctx->stream_res.encoder_info_frame);
else if (pipe_ctx->stream->ctx->dc->link_srv->dp_is_128b_132b_signal(pipe_ctx)) {
+ if (pipe_ctx->stream_res.hpo_dp_stream_enc->funcs->update_dp_info_packets_sdp_line_num)
+ pipe_ctx->stream_res.hpo_dp_stream_enc->funcs->update_dp_info_packets_sdp_line_num(
+ pipe_ctx->stream_res.hpo_dp_stream_enc,
+ &pipe_ctx->stream_res.encoder_info_frame);
+
pipe_ctx->stream_res.hpo_dp_stream_enc->funcs->update_dp_info_packets(
pipe_ctx->stream_res.hpo_dp_stream_enc,
&pipe_ctx->stream_res.encoder_info_frame);
@@ -521,9 +526,15 @@ static void dcn31_reset_back_end_for_pipe(
link = pipe_ctx->stream->link;
+ if (dc->hwseq)
+ dc->hwseq->wa_state.skip_blank_stream = false;
+
if ((!pipe_ctx->stream->dpms_off || link->link_status.link_active) &&
- (link->connector_signal == SIGNAL_TYPE_EDP))
+ (link->connector_signal == SIGNAL_TYPE_EDP)) {
dc->hwss.blank_stream(pipe_ctx);
+ if (dc->hwseq)
+ dc->hwseq->wa_state.skip_blank_stream = true;
+ }
pipe_ctx->stream_res.tg->funcs->set_dsc_config(
pipe_ctx->stream_res.tg,
@@ -538,9 +549,7 @@ static void dcn31_reset_back_end_for_pipe(
if (dc_is_hdmi_tmds_signal(pipe_ctx->stream->signal))
pipe_ctx->stream->link->phy_state.symclk_ref_cnts.otg = 0;
- if (pipe_ctx->stream_res.tg->funcs->set_drr)
- pipe_ctx->stream_res.tg->funcs->set_drr(
- pipe_ctx->stream_res.tg, NULL);
+ set_drr_and_clear_adjust_pending(pipe_ctx, pipe_ctx->stream, NULL);
/* DPMS may already disable or */
/* dpms_off status is incorrect due to fastboot
@@ -553,6 +562,19 @@ static void dcn31_reset_back_end_for_pipe(
else if (pipe_ctx->stream_res.audio)
dc->hwss.disable_audio_stream(pipe_ctx);
+ /* Temporary workaround to perform DSC programming ahead of pipe reset
+ * for smartmux/SPRS
+ * TODO: Remove SmartMux/SPRS checks once movement of DSC programming is generalized
+ */
+ if (pipe_ctx->stream->timing.flags.DSC) {
+ if ((pipe_ctx->stream->signal == SIGNAL_TYPE_EDP &&
+ ((link->dc->config.smart_mux_version && link->dc->is_switch_in_progress_dest)
+ || link->is_dds || link->skip_implict_edp_power_control)) &&
+ (dc_is_dp_signal(pipe_ctx->stream->signal) ||
+ dc_is_virtual_signal(pipe_ctx->stream->signal)))
+ dc->link_srv->set_dsc_enable(pipe_ctx, false);
+ }
+
/* free acquired resources */
if (pipe_ctx->stream_res.audio) {
/*disable az_endpoint*/
@@ -567,7 +589,8 @@ static void dcn31_reset_back_end_for_pipe(
pipe_ctx->stream_res.audio = NULL;
}
}
-
+ if (dc->hwseq)
+ dc->hwseq->wa_state.skip_blank_stream = false;
pipe_ctx->stream = NULL;
DC_LOG_DEBUG("Reset back end for pipe %d, tg:%d\n",
pipe_ctx->pipe_idx, pipe_ctx->stream_res.tg->inst);
@@ -616,7 +639,8 @@ void dcn31_reset_hw_ctx_wrap(
}
/* New dc_state in the process of being applied to hardware. */
- link_enc_cfg_set_transient_mode(dc, dc->current_state, context);
+ if (!dc->config.unify_link_enc_assignment)
+ link_enc_cfg_set_transient_mode(dc, dc->current_state, context);
}
void dcn31_setup_hpo_hw_control(const struct dce_hwseq *hws, bool enable)
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c
index 5f8f45b48720..556f4fe57eda 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn31/dcn31_init.c
@@ -40,6 +40,7 @@ static const struct hw_sequencer_funcs dcn31_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
@@ -89,7 +90,6 @@ static const struct hw_sequencer_funcs dcn31_funcs = {
.enable_writeback = dcn30_enable_writeback,
.disable_writeback = dcn30_disable_writeback,
.update_writeback = dcn30_update_writeback,
- .mmhubbub_warmup = dcn30_mmhubbub_warmup,
.dmdata_status_done = dcn20_dmdata_status_done,
.program_dmdata_engine = dcn30_program_dmdata_engine,
.set_dmdata_attributes = dcn20_set_dmdata_attributes,
@@ -98,7 +98,7 @@ static const struct hw_sequencer_funcs dcn31_funcs = {
.set_flip_control_gsl = dcn20_set_flip_control_gsl,
.get_vupdate_offset_from_vsync = dcn10_get_vupdate_offset_from_vsync,
.calc_vupdate_position = dcn10_calc_vupdate_position,
- .set_backlight_level = dcn31_set_backlight_level,
+ .set_backlight_level = dcn21_set_backlight_level,
.set_abm_immediate_disable = dcn21_set_abm_immediate_disable,
.set_pipe = dcn21_set_pipe,
.enable_lvds_link_output = dce110_enable_lvds_link_output,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c
index 9b88eb72086d..e68f21fd5f0f 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_hwseq.c
@@ -84,6 +84,20 @@ static void update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable)
struct dsc_config dsc_cfg;
struct dsc_optc_config dsc_optc_cfg = {0};
enum optc_dsc_mode optc_dsc_mode;
+ struct dcn_dsc_state dsc_state = {0};
+
+ if (!dsc) {
+ DC_LOG_DSC("DSC is NULL for tg instance %d:", pipe_ctx->stream_res.tg->inst);
+ return;
+ }
+
+ if (dsc->funcs->dsc_read_state) {
+ dsc->funcs->dsc_read_state(dsc, &dsc_state);
+ if (!dsc_state.dsc_fw_en) {
+ DC_LOG_DSC("DSC has been disabled for tg instance %d:", pipe_ctx->stream_res.tg->inst);
+ return;
+ }
+ }
/* Enable DSC hw block */
dsc_cfg.pic_width = (stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right) / opp_cnt;
@@ -162,6 +176,8 @@ void dcn314_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx
int opp_inst[MAX_PIPES] = {0};
int odm_slice_width = resource_get_odm_slice_dst_width(pipe_ctx, false);
int last_odm_slice_width = resource_get_odm_slice_dst_width(pipe_ctx, true);
+ struct mpc *mpc = dc->res_pool->mpc;
+ int i;
opp_cnt = get_odm_config(pipe_ctx, opp_inst);
@@ -174,6 +190,16 @@ void dcn314_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx
pipe_ctx->stream_res.tg->funcs->set_odm_bypass(
pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing);
+ if (mpc->funcs->set_out_rate_control) {
+ for (i = 0; i < opp_cnt; ++i) {
+ mpc->funcs->set_out_rate_control(
+ mpc, opp_inst[i],
+ false,
+ 0,
+ NULL);
+ }
+ }
+
for (odm_pipe = pipe_ctx->next_odm_pipe; odm_pipe; odm_pipe = odm_pipe->next_odm_pipe) {
odm_pipe->stream_res.opp->funcs->opp_pipe_clock_control(
odm_pipe->stream_res.opp,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c
index 6bdfbf22ce87..f5112742edf9 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn314/dcn314_init.c
@@ -42,6 +42,7 @@ static const struct hw_sequencer_funcs dcn314_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
@@ -91,7 +92,6 @@ static const struct hw_sequencer_funcs dcn314_funcs = {
.enable_writeback = dcn30_enable_writeback,
.disable_writeback = dcn30_disable_writeback,
.update_writeback = dcn30_update_writeback,
- .mmhubbub_warmup = dcn30_mmhubbub_warmup,
.dmdata_status_done = dcn20_dmdata_status_done,
.program_dmdata_engine = dcn30_program_dmdata_engine,
.set_dmdata_attributes = dcn20_set_dmdata_attributes,
@@ -100,7 +100,7 @@ static const struct hw_sequencer_funcs dcn314_funcs = {
.set_flip_control_gsl = dcn20_set_flip_control_gsl,
.get_vupdate_offset_from_vsync = dcn10_get_vupdate_offset_from_vsync,
.calc_vupdate_position = dcn10_calc_vupdate_position,
- .set_backlight_level = dcn31_set_backlight_level,
+ .set_backlight_level = dcn21_set_backlight_level,
.set_abm_immediate_disable = dcn21_set_abm_immediate_disable,
.set_pipe = dcn21_set_pipe,
.enable_lvds_link_output = dce110_enable_lvds_link_output,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c
index d7f8b2dcaa6b..416b1dca3dac 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_hwseq.c
@@ -316,10 +316,12 @@ bool dcn32_apply_idle_power_optimizations(struct dc *dc, bool enable)
cmd.cab.cab_alloc_ways = (uint8_t)ways;
dc_wake_and_execute_dmub_cmd(dc->ctx, &cmd, DM_DMUB_WAIT_TYPE_NO_WAIT);
+ DC_LOG_MALL("enable scanout from MALL");
return true;
}
+ DC_LOG_MALL("surface cannot fit in CAB, disabling scanout from MALL\n");
return false;
}
@@ -985,6 +987,7 @@ void dcn32_init_hw(struct dc *dc)
dc->caps.dmub_caps.subvp_psr = dc->ctx->dmub_srv->dmub->feature_caps.subvp_psr_support;
dc->caps.dmub_caps.gecc_enable = dc->ctx->dmub_srv->dmub->feature_caps.gecc_enable;
dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver;
+ dc->caps.dmub_caps.aux_backlight_support = dc->ctx->dmub_srv->dmub->feature_caps.abm_aux_backlight_support;
/* for DCN401 testing only */
dc->caps.dmub_caps.fams_ver = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver;
@@ -1049,7 +1052,8 @@ void dcn32_update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable)
}
/* Enable DSC hw block */
- dsc_cfg.pic_width = (stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right) / opp_cnt;
+ dsc_cfg.pic_width = (stream->timing.h_addressable + pipe_ctx->hblank_borrow +
+ stream->timing.h_border_left + stream->timing.h_border_right) / opp_cnt;
dsc_cfg.pic_height = stream->timing.v_addressable + stream->timing.v_border_top + stream->timing.v_border_bottom;
dsc_cfg.pixel_encoding = stream->timing.pixel_encoding;
dsc_cfg.color_depth = stream->timing.display_color_depth;
@@ -1059,15 +1063,17 @@ void dcn32_update_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable)
dsc_cfg.dc_dsc_cfg.num_slices_h /= opp_cnt;
if (should_use_dto_dscclk)
- dccg->funcs->set_dto_dscclk(dccg, dsc->inst);
+ dccg->funcs->set_dto_dscclk(dccg, dsc->inst, dsc_cfg.dc_dsc_cfg.num_slices_h);
dsc->funcs->dsc_set_config(dsc, &dsc_cfg, &dsc_optc_cfg);
dsc->funcs->dsc_enable(dsc, pipe_ctx->stream_res.opp->inst);
for (odm_pipe = pipe_ctx->next_odm_pipe; odm_pipe; odm_pipe = odm_pipe->next_odm_pipe) {
struct display_stream_compressor *odm_dsc = odm_pipe->stream_res.dsc;
ASSERT(odm_dsc);
+ if (!odm_dsc)
+ continue;
if (should_use_dto_dscclk)
- dccg->funcs->set_dto_dscclk(dccg, odm_dsc->inst);
+ dccg->funcs->set_dto_dscclk(dccg, odm_dsc->inst, dsc_cfg.dc_dsc_cfg.num_slices_h);
odm_dsc->funcs->dsc_set_config(odm_dsc, &dsc_cfg, &dsc_optc_cfg);
odm_dsc->funcs->dsc_enable(odm_dsc, odm_pipe->stream_res.opp->inst);
}
@@ -1177,6 +1183,7 @@ unsigned int dcn32_calculate_dccg_k1_k2_values(struct pipe_ctx *pipe_ctx, unsign
struct dc_stream_state *stream = pipe_ctx->stream;
unsigned int odm_combine_factor = 0;
bool two_pix_per_container = false;
+ struct dce_hwseq *hws = stream->ctx->dc->hwseq;
two_pix_per_container = pipe_ctx->stream_res.tg->funcs->is_two_pixels_per_container(&stream->timing);
odm_combine_factor = get_odm_config(pipe_ctx, NULL);
@@ -1197,7 +1204,8 @@ unsigned int dcn32_calculate_dccg_k1_k2_values(struct pipe_ctx *pipe_ctx, unsign
} else {
*k1_div = PIXEL_RATE_DIV_BY_1;
*k2_div = PIXEL_RATE_DIV_BY_4;
- if ((odm_combine_factor == 2) || dcn32_is_dp_dig_pixel_rate_div_policy(pipe_ctx))
+ if ((odm_combine_factor == 2) || (hws->funcs.is_dp_dig_pixel_rate_div_policy &&
+ hws->funcs.is_dp_dig_pixel_rate_div_policy(pipe_ctx)))
*k2_div = PIXEL_RATE_DIV_BY_2;
}
}
@@ -1320,7 +1328,8 @@ void dcn32_unblank_stream(struct pipe_ctx *pipe_ctx,
params.timing.pix_clk_100hz /= 2;
params.pix_per_cycle = 2;
}
- pipe_ctx->stream_res.stream_enc->funcs->dp_set_odm_combine(
+ if (pipe_ctx->stream_res.stream_enc->funcs->dp_set_odm_combine)
+ pipe_ctx->stream_res.stream_enc->funcs->dp_set_odm_combine(
pipe_ctx->stream_res.stream_enc, params.pix_per_cycle > 1);
pipe_ctx->stream_res.stream_enc->funcs->dp_unblank(link, pipe_ctx->stream_res.stream_enc, &params);
}
@@ -1397,12 +1406,12 @@ void dcn32_disable_link_output(struct dc_link *link,
link_hwss->disable_link_output(link, link_res, signal);
link->phy_state.symclk_state = SYMCLK_OFF_TX_OFF;
-
- if (signal == SIGNAL_TYPE_EDP &&
- link->dc->hwss.edp_power_control &&
- !link->skip_implict_edp_power_control)
- link->dc->hwss.edp_power_control(link, false);
- else if (dmcu != NULL && dmcu->funcs->unlock_phy)
+ /*
+ * Add the logic to extract BOTH power up and power down sequences
+ * from enable/disable link output and only call edp panel control
+ * in enable_link_dp and disable_link_dp once.
+ */
+ if (dmcu != NULL && dmcu->funcs->unlock_phy)
dmcu->funcs->unlock_phy(dmcu);
dc->link_srv->dp_trace_source_sequence(link, DPCD_SOURCE_SEQ_AFTER_DISABLE_LINK_PHY);
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c
index 5ecee7e320da..b971356d30b1 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn32/dcn32_init.c
@@ -39,6 +39,7 @@ static const struct hw_sequencer_funcs dcn32_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
@@ -87,7 +88,6 @@ static const struct hw_sequencer_funcs dcn32_funcs = {
.enable_writeback = dcn30_enable_writeback,
.disable_writeback = dcn30_disable_writeback,
.update_writeback = dcn30_update_writeback,
- .mmhubbub_warmup = dcn30_mmhubbub_warmup,
.dmdata_status_done = dcn20_dmdata_status_done,
.program_dmdata_engine = dcn30_program_dmdata_engine,
.set_dmdata_attributes = dcn20_set_dmdata_attributes,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c
index e599cdc465bf..a267f574b619 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c
@@ -236,15 +236,11 @@ void dcn35_init_hw(struct dc *dc)
}
hws->funcs.init_pipes(dc, dc->current_state);
- if (dc->res_pool->hubbub->funcs->allow_self_refresh_control)
+ if (dc->res_pool->hubbub->funcs->allow_self_refresh_control &&
+ !dc->res_pool->hubbub->ctx->dc->debug.disable_stutter)
dc->res_pool->hubbub->funcs->allow_self_refresh_control(dc->res_pool->hubbub,
!dc->res_pool->hubbub->ctx->dc->debug.disable_stutter);
}
- if (res_pool->dccg->funcs->dccg_root_gate_disable_control) {
- for (i = 0; i < res_pool->pipe_count; i++)
- res_pool->dccg->funcs->dccg_root_gate_disable_control(res_pool->dccg, i, 0);
- }
-
for (i = 0; i < res_pool->audio_count; i++) {
struct audio *audio = res_pool->audios[i];
@@ -426,6 +422,8 @@ void dcn35_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx *
int opp_inst[MAX_PIPES] = {0};
int odm_slice_width = resource_get_odm_slice_dst_width(pipe_ctx, false);
int last_odm_slice_width = resource_get_odm_slice_dst_width(pipe_ctx, true);
+ struct mpc *mpc = dc->res_pool->mpc;
+ int i;
opp_cnt = get_odm_config(pipe_ctx, opp_inst);
@@ -438,6 +436,16 @@ void dcn35_update_odm(struct dc *dc, struct dc_state *context, struct pipe_ctx *
pipe_ctx->stream_res.tg->funcs->set_odm_bypass(
pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing);
+ if (mpc->funcs->set_out_rate_control) {
+ for (i = 0; i < opp_cnt; ++i) {
+ mpc->funcs->set_out_rate_control(
+ mpc, opp_inst[i],
+ false,
+ 0,
+ NULL);
+ }
+ }
+
for (odm_pipe = pipe_ctx->next_odm_pipe; odm_pipe; odm_pipe = odm_pipe->next_odm_pipe) {
odm_pipe->stream_res.opp->funcs->opp_pipe_clock_control(
odm_pipe->stream_res.opp,
@@ -788,6 +796,7 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context)
/* Disable on the current state so the new one isn't cleared. */
pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
+ hubp->funcs->hubp_reset(hubp);
dpp->funcs->dpp_reset(dpp);
pipe_ctx->stream_res.tg = tg;
@@ -887,12 +896,18 @@ void dcn35_init_pipes(struct dc *dc, struct dc_state *context)
void dcn35_enable_plane(struct dc *dc, struct pipe_ctx *pipe_ctx,
struct dc_state *context)
{
+ struct dpp *dpp = pipe_ctx->plane_res.dpp;
+ struct dccg *dccg = dc->res_pool->dccg;
+
+
/* enable DCFCLK current DCHUB */
pipe_ctx->plane_res.hubp->funcs->hubp_clk_cntl(pipe_ctx->plane_res.hubp, true);
/* initialize HUBP on power up */
pipe_ctx->plane_res.hubp->funcs->hubp_init(pipe_ctx->plane_res.hubp);
-
+ /*make sure DPPCLK is on*/
+ dccg->funcs->dccg_root_gate_disable_control(dccg, dpp->inst, true);
+ dpp->funcs->dpp_dppclk_control(dpp, false, true);
/* make sure OPP_PIPE_CLOCK_EN = 1 */
pipe_ctx->stream_res.opp->funcs->opp_pipe_clock_control(
pipe_ctx->stream_res.opp,
@@ -909,6 +924,7 @@ void dcn35_enable_plane(struct dc *dc, struct pipe_ctx *pipe_ctx,
// Program system aperture settings
pipe_ctx->plane_res.hubp->funcs->hubp_set_vm_system_aperture_settings(pipe_ctx->plane_res.hubp, &apt);
}
+ //DC_LOG_DEBUG("%s: dpp_inst(%d) =\n", __func__, dpp->inst);
if (!pipe_ctx->top_pipe
&& pipe_ctx->plane_state
@@ -924,6 +940,8 @@ void dcn35_plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx)
{
struct hubp *hubp = pipe_ctx->plane_res.hubp;
struct dpp *dpp = pipe_ctx->plane_res.dpp;
+ struct dccg *dccg = dc->res_pool->dccg;
+
dc->hwss.wait_for_mpcc_disconnect(dc, dc->res_pool, pipe_ctx);
@@ -941,9 +959,11 @@ void dcn35_plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx)
hubp->funcs->hubp_clk_cntl(hubp, false);
dpp->funcs->dpp_dppclk_control(dpp, false, false);
-/*to do, need to support both case*/
+ dccg->funcs->dccg_root_gate_disable_control(dccg, dpp->inst, false);
+
hubp->power_gated = true;
+ hubp->funcs->hubp_reset(hubp);
dpp->funcs->dpp_reset(dpp);
pipe_ctx->stream = NULL;
@@ -952,6 +972,8 @@ void dcn35_plane_atomic_disable(struct dc *dc, struct pipe_ctx *pipe_ctx)
pipe_ctx->top_pipe = NULL;
pipe_ctx->bottom_pipe = NULL;
pipe_ctx->plane_state = NULL;
+ //DC_LOG_DEBUG("%s: dpp_inst(%d)=\n", __func__, dpp->inst);
+
}
void dcn35_disable_plane(struct dc *dc, struct dc_state *state, struct pipe_ctx *pipe_ctx)
@@ -1020,8 +1042,22 @@ void dcn35_calc_blocks_to_gate(struct dc *dc, struct dc_state *context,
if (pipe_ctx->plane_res.dpp || pipe_ctx->stream_res.opp)
update_state->pg_pipe_res_update[PG_MPCC][pipe_ctx->plane_res.mpcc_inst] = false;
- if (pipe_ctx->stream_res.dsc)
+ if (pipe_ctx->stream_res.dsc) {
update_state->pg_pipe_res_update[PG_DSC][pipe_ctx->stream_res.dsc->inst] = false;
+ if (dc->caps.sequential_ono) {
+ update_state->pg_pipe_res_update[PG_HUBP][pipe_ctx->stream_res.dsc->inst] = false;
+ update_state->pg_pipe_res_update[PG_DPP][pipe_ctx->stream_res.dsc->inst] = false;
+
+ /* All HUBP/DPP instances must be powered if the DSC inst != HUBP inst */
+ if (!pipe_ctx->top_pipe && pipe_ctx->plane_res.hubp &&
+ pipe_ctx->plane_res.hubp->inst != pipe_ctx->stream_res.dsc->inst) {
+ for (j = 0; j < dc->res_pool->pipe_count; ++j) {
+ update_state->pg_pipe_res_update[PG_HUBP][j] = false;
+ update_state->pg_pipe_res_update[PG_DPP][j] = false;
+ }
+ }
+ }
+ }
if (pipe_ctx->stream_res.opp)
update_state->pg_pipe_res_update[PG_OPP][pipe_ctx->stream_res.opp->inst] = false;
@@ -1166,6 +1202,25 @@ void dcn35_calc_blocks_to_ungate(struct dc *dc, struct dc_state *context,
update_state->pg_pipe_res_update[PG_HDMISTREAM][0] = true;
if (dc->caps.sequential_ono) {
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ struct pipe_ctx *new_pipe = &context->res_ctx.pipe_ctx[i];
+
+ if (new_pipe->stream_res.dsc && !new_pipe->top_pipe &&
+ update_state->pg_pipe_res_update[PG_DSC][new_pipe->stream_res.dsc->inst]) {
+ update_state->pg_pipe_res_update[PG_HUBP][new_pipe->stream_res.dsc->inst] = true;
+ update_state->pg_pipe_res_update[PG_DPP][new_pipe->stream_res.dsc->inst] = true;
+
+ /* All HUBP/DPP instances must be powered if the DSC inst != HUBP inst */
+ if (new_pipe->plane_res.hubp &&
+ new_pipe->plane_res.hubp->inst != new_pipe->stream_res.dsc->inst) {
+ for (j = 0; j < dc->res_pool->pipe_count; ++j) {
+ update_state->pg_pipe_res_update[PG_HUBP][j] = true;
+ update_state->pg_pipe_res_update[PG_DPP][j] = true;
+ }
+ }
+ }
+ }
+
for (i = dc->res_pool->pipe_count - 1; i >= 0; i--) {
if (update_state->pg_pipe_res_update[PG_HUBP][i] &&
update_state->pg_pipe_res_update[PG_DPP][i]) {
@@ -1453,8 +1508,7 @@ void dcn35_set_drr(struct pipe_ctx **pipe_ctx,
num_frames = 2 * (frame_rate % 60);
}
}
- if (tg->funcs->set_drr)
- tg->funcs->set_drr(tg, &params);
+ set_drr_and_clear_adjust_pending(pipe_ctx[i], pipe_ctx[i]->stream, &params);
if (adjust.v_total_max != 0 && adjust.v_total_min != 0)
if (tg->funcs->set_static_screen_control)
tg->funcs->set_static_screen_control(
@@ -1524,7 +1578,7 @@ static bool should_avoid_empty_tu(struct pipe_ctx *pipe_ctx)
struct dc_link_settings *link_settings = &pipe_ctx->link_config.dp_link_settings;
const struct dc *dc = pipe_ctx->stream->link->dc;
- if (pipe_ctx->stream->link->ep_type != DISPLAY_ENDPOINT_USB4_DPIA)
+ if (pipe_ctx->link_config.dp_tunnel_settings.should_enable_dp_tunneling == false)
return false;
// Not necessary for MST configurations
@@ -1579,3 +1633,37 @@ bool dcn35_is_dp_dig_pixel_rate_div_policy(struct pipe_ctx *pipe_ctx)
return false;
}
+
+/*
+ * Set powerup to true for every pipe to match pre-OS configuration.
+ */
+static void dcn35_calc_blocks_to_ungate_for_hw_release(struct dc *dc, struct pg_block_update *update_state)
+{
+ int i = 0, j = 0;
+
+ memset(update_state, 0, sizeof(struct pg_block_update));
+
+ for (i = 0; i < dc->res_pool->pipe_count; i++)
+ for (j = 0; j < PG_HW_PIPE_RESOURCES_NUM_ELEMENT; j++)
+ update_state->pg_pipe_res_update[j][i] = true;
+
+ update_state->pg_res_update[PG_HPO] = true;
+ update_state->pg_res_update[PG_DWB] = true;
+}
+
+/*
+ * The purpose is to power up all gatings to restore optimization to pre-OS env.
+ * Re-use hwss func and existing PG&RCG flags to decide powerup sequence.
+ */
+void dcn35_hardware_release(struct dc *dc)
+{
+ struct pg_block_update pg_update_state;
+
+ dcn35_calc_blocks_to_ungate_for_hw_release(dc, &pg_update_state);
+
+ if (dc->hwss.root_clock_control)
+ dc->hwss.root_clock_control(dc, &pg_update_state, true);
+ /*power up required HW block*/
+ if (dc->hwss.hw_block_power_up)
+ dc->hwss.hw_block_power_up(dc, &pg_update_state);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h
index e27b3609020f..0b1d6f608edd 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.h
@@ -99,4 +99,6 @@ void dcn35_set_long_vblank(struct pipe_ctx **pipe_ctx,
bool dcn35_is_dp_dig_pixel_rate_div_policy(struct pipe_ctx *pipe_ctx);
+void dcn35_hardware_release(struct dc *dc);
+
#endif /* __DC_HWSS_DCN35_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c
index fd67779c27a9..a3ccf805bd16 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_init.c
@@ -44,6 +44,7 @@ static const struct hw_sequencer_funcs dcn35_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
@@ -92,7 +93,6 @@ static const struct hw_sequencer_funcs dcn35_funcs = {
.enable_writeback = dcn30_enable_writeback,
.disable_writeback = dcn30_disable_writeback,
.update_writeback = dcn30_update_writeback,
- .mmhubbub_warmup = dcn30_mmhubbub_warmup,
.dmdata_status_done = dcn20_dmdata_status_done,
.program_dmdata_engine = dcn30_program_dmdata_engine,
.set_dmdata_attributes = dcn20_set_dmdata_attributes,
@@ -123,6 +123,11 @@ static const struct hw_sequencer_funcs dcn35_funcs = {
.root_clock_control = dcn35_root_clock_control,
.set_long_vtotal = dcn35_set_long_vblank,
.calculate_pix_rate_divider = dcn32_calculate_pix_rate_divider,
+ .hardware_release = dcn35_hardware_release,
+ .detect_pipe_changes = dcn20_detect_pipe_changes,
+ .enable_plane = dcn20_enable_plane,
+ .update_dchubp_dpp = dcn20_update_dchubp_dpp,
+ .post_unlock_reset_opp = dcn20_post_unlock_reset_opp,
};
static const struct hwseq_private_funcs dcn35_private_funcs = {
@@ -163,6 +168,8 @@ static const struct hwseq_private_funcs dcn35_private_funcs = {
.dsc_pg_control = dcn35_dsc_pg_control,
.dsc_pg_status = dcn32_dsc_pg_status,
.enable_plane = dcn35_enable_plane,
+ .wait_for_pipe_update_if_needed = dcn10_wait_for_pipe_update_if_needed,
+ .set_wait_for_update_needed_for_pipe = dcn10_set_wait_for_update_needed_for_pipe,
};
void dcn35_hw_sequencer_construct(struct dc *dc)
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c
index 3c275a1eff58..58f2be2a326b 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn351/dcn351_init.c
@@ -43,6 +43,7 @@ static const struct hw_sequencer_funcs dcn351_funcs = {
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
@@ -91,7 +92,6 @@ static const struct hw_sequencer_funcs dcn351_funcs = {
.enable_writeback = dcn30_enable_writeback,
.disable_writeback = dcn30_disable_writeback,
.update_writeback = dcn30_update_writeback,
- .mmhubbub_warmup = dcn30_mmhubbub_warmup,
.dmdata_status_done = dcn20_dmdata_status_done,
.program_dmdata_engine = dcn30_program_dmdata_engine,
.set_dmdata_attributes = dcn20_set_dmdata_attributes,
@@ -158,10 +158,12 @@ static const struct hwseq_private_funcs dcn351_private_funcs = {
.set_mcm_luts = dcn32_set_mcm_luts,
.setup_hpo_hw_control = dcn35_setup_hpo_hw_control,
.calculate_dccg_k1_k2_values = dcn32_calculate_dccg_k1_k2_values,
- .is_dp_dig_pixel_rate_div_policy = dcn32_is_dp_dig_pixel_rate_div_policy,
+ .is_dp_dig_pixel_rate_div_policy = dcn35_is_dp_dig_pixel_rate_div_policy,
.dsc_pg_control = dcn35_dsc_pg_control,
.dsc_pg_status = dcn32_dsc_pg_status,
.enable_plane = dcn35_enable_plane,
+ .wait_for_pipe_update_if_needed = dcn10_wait_for_pipe_update_if_needed,
+ .set_wait_for_update_needed_for_pipe = dcn10_set_wait_for_update_needed_for_pipe,
};
void dcn351_hw_sequencer_construct(struct dc *dc)
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c
index 5de11e2837c0..cc9f40d97af2 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.c
@@ -2,7 +2,10 @@
//
// Copyright 2024 Advanced Micro Devices, Inc.
+
+#include "os_types.h"
#include "dm_services.h"
+#include "basics/dc_common.h"
#include "dm_helpers.h"
#include "core_types.h"
#include "resource.h"
@@ -48,7 +51,7 @@
#define FN(reg_name, field_name) \
hws->shifts->field_name, hws->masks->field_name
-static void dcn401_initialize_min_clocks(struct dc *dc)
+void dcn401_initialize_min_clocks(struct dc *dc)
{
struct dc_clocks *clocks = &dc->current_state->bw_ctx.bw.dcn.clk;
@@ -126,91 +129,6 @@ void dcn401_program_gamut_remap(struct pipe_ctx *pipe_ctx)
mpc->funcs->set_gamut_remap(mpc, mpcc_id, &mpc_adjust);
}
-struct ips_ono_region_state dcn401_read_ono_state(struct dc *dc, uint8_t region)
-{
- struct dce_hwseq *hws = dc->hwseq;
- struct ips_ono_region_state state = {0, 0};
-
- switch (region) {
- case 0:
- /* dccg, dio, dcio */
- REG_GET_2(DOMAIN22_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 1:
- /* dchubbub, dchvm, dchubbubmem */
- REG_GET_2(DOMAIN23_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 2:
- /* mpc, opp, optc, dwb */
- REG_GET_2(DOMAIN24_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 3:
- /* hpo */
- REG_GET_2(DOMAIN25_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 4:
- /* dchubp0, dpp0 */
- REG_GET_2(DOMAIN0_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 5:
- /* dsc0 */
- REG_GET_2(DOMAIN16_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 6:
- /* dchubp1, dpp1 */
- REG_GET_2(DOMAIN1_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 7:
- /* dsc1 */
- REG_GET_2(DOMAIN17_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 8:
- /* dchubp2, dpp2 */
- REG_GET_2(DOMAIN2_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 9:
- /* dsc2 */
- REG_GET_2(DOMAIN18_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 10:
- /* dchubp3, dpp3 */
- REG_GET_2(DOMAIN3_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- case 11:
- /* dsc3 */
- REG_GET_2(DOMAIN19_PG_STATUS,
- DOMAIN_DESIRED_PWR_STATE, &state.desire_pwr_state,
- DOMAIN_PGFSM_PWR_STATUS, &state.current_pwr_state);
- break;
- default:
- break;
- }
-
- return state;
-}
-
void dcn401_init_hw(struct dc *dc)
{
struct abm **abms = dc->res_pool->multiple_abms;
@@ -227,13 +145,8 @@ void dcn401_init_hw(struct dc *dc)
dc->clk_mgr->funcs->init_clocks(dc->clk_mgr);
// mark dcmode limits present if any clock has distinct AC and DC values from SMU
- dc->caps.dcmode_power_limits_present =
- (dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dcfclk_levels && dc->clk_mgr->bw_params->dc_mode_limit.dcfclk_mhz) ||
- (dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dispclk_levels && dc->clk_mgr->bw_params->dc_mode_limit.dispclk_mhz) ||
- (dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_dtbclk_levels && dc->clk_mgr->bw_params->dc_mode_limit.dtbclk_mhz) ||
- (dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_fclk_levels && dc->clk_mgr->bw_params->dc_mode_limit.fclk_mhz) ||
- (dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_memclk_levels && dc->clk_mgr->bw_params->dc_mode_limit.memclk_mhz) ||
- (dc->clk_mgr->bw_params->clk_table.num_entries_per_clk.num_socclk_levels && dc->clk_mgr->bw_params->dc_mode_limit.socclk_mhz);
+ dc->caps.dcmode_power_limits_present = dc->clk_mgr->funcs->is_dc_mode_present &&
+ dc->clk_mgr->funcs->is_dc_mode_present(dc->clk_mgr);
}
// Initialize the dccg
@@ -435,7 +348,8 @@ void dcn401_init_hw(struct dc *dc)
dc->caps.dmub_caps.psr = dc->ctx->dmub_srv->dmub->feature_caps.psr;
dc->caps.dmub_caps.mclk_sw = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver > 0;
dc->caps.dmub_caps.fams_ver = dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver;
- dc->debug.fams2_config.bits.enable &= dc->ctx->dmub_srv->dmub->feature_caps.fw_assisted_mclk_switch_ver == 2;
+ dc->debug.fams2_config.bits.enable &=
+ dc->caps.dmub_caps.fams_ver == dc->debug.fams_version.ver; // sw & fw fams versions must match for support
if ((!dc->debug.fams2_config.bits.enable && dc->res_pool->funcs->update_bw_bounding_box)
|| res_pool->ref_clocks.dchub_ref_clock_inKhz / 1000 != current_dchub_ref_freq) {
/* update bounding box if FAMS2 disabled, or if dchub clk has changed */
@@ -490,17 +404,16 @@ void dcn401_populate_mcm_luts(struct dc *dc,
struct mpc *mpc = dc->res_pool->mpc;
union mcm_lut_params m_lut_params;
enum dc_cm2_transfer_func_source lut3d_src = mcm_luts.lut3d_data.lut3d_src;
- enum hubp_3dlut_fl_format format;
+ enum hubp_3dlut_fl_format format = 0;
enum hubp_3dlut_fl_mode mode;
- enum hubp_3dlut_fl_width width;
+ enum hubp_3dlut_fl_width width = 0;
enum hubp_3dlut_fl_addressing_mode addr_mode;
- enum hubp_3dlut_fl_crossbar_bit_slice crossbar_bit_slice_y_g;
- enum hubp_3dlut_fl_crossbar_bit_slice crossbar_bit_slice_cb_b;
- enum hubp_3dlut_fl_crossbar_bit_slice crossbar_bit_slice_cr_r;
+ enum hubp_3dlut_fl_crossbar_bit_slice crossbar_bit_slice_y_g = 0;
+ enum hubp_3dlut_fl_crossbar_bit_slice crossbar_bit_slice_cb_b = 0;
+ enum hubp_3dlut_fl_crossbar_bit_slice crossbar_bit_slice_cr_r = 0;
enum MCM_LUT_XABLE shaper_xable = MCM_LUT_DISABLE;
enum MCM_LUT_XABLE lut3d_xable = MCM_LUT_DISABLE;
enum MCM_LUT_XABLE lut1d_xable = MCM_LUT_DISABLE;
- bool is_17x17x17 = true;
bool rval;
dcn401_get_mcm_lut_xable_from_pipe_ctx(dc, pipe_ctx, &shaper_xable, &lut3d_xable, &lut1d_xable);
@@ -525,7 +438,7 @@ void dcn401_populate_mcm_luts(struct dc *dc,
}
/* Shaper */
- if (mcm_luts.shaper) {
+ if (mcm_luts.shaper && mcm_luts.lut3d_data.mpc_3dlut_enable) {
memset(&m_lut_params, 0, sizeof(m_lut_params));
if (mcm_luts.shaper->type == TF_TYPE_HWPWL)
m_lut_params.pwl = &mcm_luts.shaper->pwl;
@@ -537,11 +450,11 @@ void dcn401_populate_mcm_luts(struct dc *dc,
m_lut_params.pwl = rval ? &dpp_base->regamma_params : NULL;
}
if (m_lut_params.pwl) {
- if (mpc->funcs->populate_lut)
- mpc->funcs->populate_lut(mpc, MCM_LUT_SHAPER, m_lut_params, lut_bank_a, mpcc_id);
+ if (mpc->funcs->mcm.populate_lut)
+ mpc->funcs->mcm.populate_lut(mpc, m_lut_params, lut_bank_a, mpcc_id);
+ if (mpc->funcs->program_lut_mode)
+ mpc->funcs->program_lut_mode(mpc, MCM_LUT_SHAPER, MCM_LUT_ENABLE, lut_bank_a, mpcc_id);
}
- if (mpc->funcs->program_lut_mode)
- mpc->funcs->program_lut_mode(mpc, MCM_LUT_SHAPER, shaper_xable, lut_bank_a, mpcc_id);
}
/* 3DLUT */
@@ -550,6 +463,7 @@ void dcn401_populate_mcm_luts(struct dc *dc,
memset(&m_lut_params, 0, sizeof(m_lut_params));
if (hubp->funcs->hubp_enable_3dlut_fl)
hubp->funcs->hubp_enable_3dlut_fl(hubp, false);
+
if (mcm_luts.lut3d_data.lut3d_func && mcm_luts.lut3d_data.lut3d_func->state.bits.initialized) {
m_lut_params.lut3d = &mcm_luts.lut3d_data.lut3d_func->lut_3d;
if (mpc->funcs->populate_lut)
@@ -559,16 +473,35 @@ void dcn401_populate_mcm_luts(struct dc *dc,
mpcc_id);
}
break;
- case DC_CM2_TRANSFER_FUNC_SOURCE_VIDMEM:
+ case DC_CM2_TRANSFER_FUNC_SOURCE_VIDMEM:
+ switch (mcm_luts.lut3d_data.gpu_mem_params.size) {
+ case DC_CM2_GPU_MEM_SIZE_171717:
+ width = hubp_3dlut_fl_width_17;
+ break;
+ case DC_CM2_GPU_MEM_SIZE_TRANSFORMED:
+ width = hubp_3dlut_fl_width_transformed;
+ break;
+ default:
+ //TODO: handle default case
+ break;
+ }
+
+ //check for support
+ if (mpc->funcs->mcm.is_config_supported &&
+ !mpc->funcs->mcm.is_config_supported(width))
+ break;
if (mpc->funcs->program_lut_read_write_control)
mpc->funcs->program_lut_read_write_control(mpc, MCM_LUT_3DLUT, lut_bank_a, mpcc_id);
if (mpc->funcs->program_lut_mode)
mpc->funcs->program_lut_mode(mpc, MCM_LUT_3DLUT, lut3d_xable, lut_bank_a, mpcc_id);
- if (mpc->funcs->program_3dlut_size)
- mpc->funcs->program_3dlut_size(mpc, is_17x17x17, mpcc_id);
+
if (hubp->funcs->hubp_program_3dlut_fl_addr)
hubp->funcs->hubp_program_3dlut_fl_addr(hubp, mcm_luts.lut3d_data.gpu_mem_params.addr);
+
+ if (mpc->funcs->mcm.program_bit_depth)
+ mpc->funcs->mcm.program_bit_depth(mpc, mcm_luts.lut3d_data.gpu_mem_params.bit_depth, mpcc_id);
+
switch (mcm_luts.lut3d_data.gpu_mem_params.layout) {
case DC_CM2_GPU_MEM_LAYOUT_3D_SWIZZLE_LINEAR_RGB:
mode = hubp_3dlut_fl_mode_native_1;
@@ -595,7 +528,6 @@ void dcn401_populate_mcm_luts(struct dc *dc,
switch (mcm_luts.lut3d_data.gpu_mem_params.format_params.format) {
case DC_CM2_GPU_MEM_FORMAT_16161616_UNORM_12MSB:
- default:
format = hubp_3dlut_fl_format_unorm_12msb_bitslice;
break;
case DC_CM2_GPU_MEM_FORMAT_16161616_UNORM_12LSB:
@@ -607,11 +539,19 @@ void dcn401_populate_mcm_luts(struct dc *dc,
}
if (hubp->funcs->hubp_program_3dlut_fl_format)
hubp->funcs->hubp_program_3dlut_fl_format(hubp, format);
- if (hubp->funcs->hubp_update_3dlut_fl_bias_scale)
+ if (hubp->funcs->hubp_update_3dlut_fl_bias_scale &&
+ mpc->funcs->mcm.program_bias_scale) {
+ mpc->funcs->mcm.program_bias_scale(mpc,
+ mcm_luts.lut3d_data.gpu_mem_params.format_params.float_params.bias,
+ mcm_luts.lut3d_data.gpu_mem_params.format_params.float_params.scale,
+ mpcc_id);
hubp->funcs->hubp_update_3dlut_fl_bias_scale(hubp,
- mcm_luts.lut3d_data.gpu_mem_params.format_params.float_params.bias,
- mcm_luts.lut3d_data.gpu_mem_params.format_params.float_params.scale);
+ mcm_luts.lut3d_data.gpu_mem_params.format_params.float_params.bias,
+ mcm_luts.lut3d_data.gpu_mem_params.format_params.float_params.scale);
+ }
+ //navi 4x has a bug and r and blue are swapped and need to be worked around here in
+ //TODO: need to make a method for get_xbar per asic OR do the workaround in program_crossbar for 4x
switch (mcm_luts.lut3d_data.gpu_mem_params.component_order) {
case DC_CM2_GPU_MEM_PIXEL_COMPONENT_ORDER_RGBA:
default:
@@ -623,21 +563,16 @@ void dcn401_populate_mcm_luts(struct dc *dc,
if (hubp->funcs->hubp_program_3dlut_fl_crossbar)
hubp->funcs->hubp_program_3dlut_fl_crossbar(hubp,
+ crossbar_bit_slice_cr_r,
crossbar_bit_slice_y_g,
- crossbar_bit_slice_cb_b,
- crossbar_bit_slice_cr_r);
+ crossbar_bit_slice_cb_b);
+
+ if (mpc->funcs->mcm.program_lut_read_write_control)
+ mpc->funcs->mcm.program_lut_read_write_control(mpc, MCM_LUT_3DLUT, lut_bank_a, true, mpcc_id);
+
+ if (mpc->funcs->mcm.program_3dlut_size)
+ mpc->funcs->mcm.program_3dlut_size(mpc, width, mpcc_id);
- switch (mcm_luts.lut3d_data.gpu_mem_params.size) {
- case DC_CM2_GPU_MEM_SIZE_171717:
- default:
- width = hubp_3dlut_fl_width_17;
- break;
- case DC_CM2_GPU_MEM_SIZE_TRANSFORMED:
- width = hubp_3dlut_fl_width_transformed;
- break;
- }
- if (hubp->funcs->hubp_program_3dlut_fl_width)
- hubp->funcs->hubp_program_3dlut_fl_width(hubp, width);
if (mpc->funcs->update_3dlut_fast_load_select)
mpc->funcs->update_3dlut_fast_load_select(mpc, mpcc_id, hubp->inst);
@@ -820,7 +755,8 @@ enum dc_status dcn401_enable_stream_timing(
int opp_cnt = 1;
int opp_inst[MAX_PIPES] = {0};
struct pipe_ctx *opp_heads[MAX_PIPES] = {0};
- bool manual_mode;
+ struct dc_crtc_timing patched_crtc_timing = stream->timing;
+ bool manual_mode = false;
unsigned int tmds_div = PIXEL_RATE_DIV_NA;
unsigned int unused_div = PIXEL_RATE_DIV_NA;
int odm_slice_width;
@@ -874,16 +810,20 @@ enum dc_status dcn401_enable_stream_timing(
if (dc->hwseq->funcs.PLAT_58856_wa && (!dc_is_dp_signal(stream->signal)))
dc->hwseq->funcs.PLAT_58856_wa(context, pipe_ctx);
+ /* if we are borrowing from hblank, h_addressable needs to be adjusted */
+ if (dc->debug.enable_hblank_borrow)
+ patched_crtc_timing.h_addressable = patched_crtc_timing.h_addressable + pipe_ctx->hblank_borrow;
+
pipe_ctx->stream_res.tg->funcs->program_timing(
- pipe_ctx->stream_res.tg,
- &stream->timing,
- pipe_ctx->pipe_dlg_param.vready_offset,
- pipe_ctx->pipe_dlg_param.vstartup_start,
- pipe_ctx->pipe_dlg_param.vupdate_offset,
- pipe_ctx->pipe_dlg_param.vupdate_width,
- pipe_ctx->pipe_dlg_param.pstate_keepout,
- pipe_ctx->stream->signal,
- true);
+ pipe_ctx->stream_res.tg,
+ &patched_crtc_timing,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vready_offset_pixels,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vstartup_lines,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vupdate_offset_pixels,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vupdate_vupdate_width_pixels,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.pstate_keepout_start_lines,
+ pipe_ctx->stream->signal,
+ true);
for (i = 0; i < opp_cnt; i++) {
opp_heads[i]->stream_res.opp->funcs->opp_pipe_clock_control(
@@ -908,10 +848,7 @@ enum dc_status dcn401_enable_stream_timing(
}
hws->funcs.wait_for_blank_complete(pipe_ctx->stream_res.opp);
-
- if (pipe_ctx->stream_res.tg->funcs->set_drr)
- pipe_ctx->stream_res.tg->funcs->set_drr(
- pipe_ctx->stream_res.tg, &params);
+ set_drr_and_clear_adjust_pending(pipe_ctx, stream, &params);
/* Event triggers and num frames initialized for DRR, but can be
* later updated for PSR use. Note DRR trigger events are generated
@@ -1005,17 +942,23 @@ void dcn401_enable_stream(struct pipe_ctx *pipe_ctx)
int dp_hpo_inst = 0;
unsigned int tmds_div = PIXEL_RATE_DIV_NA;
unsigned int unused_div = PIXEL_RATE_DIV_NA;
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(pipe_ctx->stream->link);
+ struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
struct stream_encoder *stream_enc = pipe_ctx->stream_res.stream_enc;
+ if (!dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+
dcn401_enable_stream_calc(pipe_ctx, &dp_hpo_inst, &phyd32clk,
&tmds_div, &early_control);
if (dc_is_dp_signal(pipe_ctx->stream->signal) || dc_is_virtual_signal(pipe_ctx->stream->signal)) {
if (dc->link_srv->dp_is_128b_132b_signal(pipe_ctx)) {
dccg->funcs->set_dpstreamclk(dccg, DPREFCLK, tg->inst, dp_hpo_inst);
-
- dccg->funcs->enable_symclk32_se(dccg, dp_hpo_inst, phyd32clk);
+ if (link->cur_link_settings.link_rate == LINK_RATE_UNKNOWN) {
+ dccg->funcs->disable_symclk32_se(dccg, dp_hpo_inst);
+ } else {
+ dccg->funcs->enable_symclk32_se(dccg, dp_hpo_inst, phyd32clk);
+ }
} else {
dccg->funcs->enable_symclk_se(dccg, stream_enc->stream_enc_inst,
link_enc->transmitter - TRANSMITTER_UNIPHY_A);
@@ -1050,52 +993,6 @@ void dcn401_setup_hpo_hw_control(const struct dce_hwseq *hws, bool enable)
REG_UPDATE(HPO_TOP_HW_CONTROL, HPO_IO_EN, enable);
}
-static bool dcn401_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx)
-{
- struct pipe_ctx *test_pipe, *split_pipe;
- const struct scaler_data *scl_data = &pipe_ctx->plane_res.scl_data;
- struct rect r1 = scl_data->recout, r2, r2_half;
- int r1_r = r1.x + r1.width, r1_b = r1.y + r1.height, r2_r, r2_b;
- int cur_layer = pipe_ctx->plane_state->layer_index;
-
- /**
- * Disable the cursor if there's another pipe above this with a
- * plane that contains this pipe's viewport to prevent double cursor
- * and incorrect scaling artifacts.
- */
- for (test_pipe = pipe_ctx->top_pipe; test_pipe;
- test_pipe = test_pipe->top_pipe) {
- // Skip invisible layer and pipe-split plane on same layer
- if (!test_pipe->plane_state ||
- !test_pipe->plane_state->visible ||
- test_pipe->plane_state->layer_index == cur_layer)
- continue;
-
- r2 = test_pipe->plane_res.scl_data.recout;
- r2_r = r2.x + r2.width;
- r2_b = r2.y + r2.height;
-
- /**
- * There is another half plane on same layer because of
- * pipe-split, merge together per same height.
- */
- for (split_pipe = pipe_ctx->top_pipe; split_pipe;
- split_pipe = split_pipe->top_pipe)
- if (split_pipe->plane_state->layer_index == test_pipe->plane_state->layer_index) {
- r2_half = split_pipe->plane_res.scl_data.recout;
- r2.x = (r2_half.x < r2.x) ? r2_half.x : r2.x;
- r2.width = r2.width + r2_half.width;
- r2_r = r2.x + r2.width;
- break;
- }
-
- if (r1.x >= r2.x && r1.y >= r2.y && r1_r <= r2_r && r1_b <= r2_b)
- return true;
- }
-
- return false;
-}
-
void adjust_hotspot_between_slices_for_2x_magnify(uint32_t cursor_width, struct dc_cursor_position *pos_cpy)
{
if (cursor_width <= 128) {
@@ -1286,7 +1183,7 @@ void dcn401_set_cursor_position(struct pipe_ctx *pipe_ctx)
pos_cpy.x = (uint32_t)x_pos;
pos_cpy.y = (uint32_t)y_pos;
- if (pos_cpy.enable && dcn401_can_pipe_disable_cursor(pipe_ctx))
+ if (pos_cpy.enable && resource_can_pipe_disable_cursor(pipe_ctx))
pos_cpy.enable = false;
x_pos = pos_cpy.x - param.recout.x;
@@ -1941,9 +1838,8 @@ void dcn401_reset_back_end_for_pipe(
pipe_ctx->stream_res.tg->funcs->set_odm_bypass(
pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing);
- if (pipe_ctx->stream_res.tg->funcs->set_drr)
- pipe_ctx->stream_res.tg->funcs->set_drr(
- pipe_ctx->stream_res.tg, NULL);
+ set_drr_and_clear_adjust_pending(pipe_ctx, pipe_ctx->stream, NULL);
+
/* TODO - convert symclk_ref_cnts for otg to a bit map to solve
* the case where the same symclk is shared across multiple otg
* instances
@@ -2007,3 +1903,758 @@ void dcn401_reset_hw_ctx_wrap(
}
}
}
+
+static unsigned int dcn401_calculate_vready_offset_for_group(struct pipe_ctx *pipe)
+{
+ struct pipe_ctx *other_pipe;
+ unsigned int vready_offset = pipe->global_sync.dcn4x.vready_offset_pixels;
+
+ /* Always use the largest vready_offset of all connected pipes */
+ for (other_pipe = pipe->bottom_pipe; other_pipe != NULL; other_pipe = other_pipe->bottom_pipe) {
+ if (other_pipe->global_sync.dcn4x.vready_offset_pixels > vready_offset)
+ vready_offset = other_pipe->global_sync.dcn4x.vready_offset_pixels;
+ }
+ for (other_pipe = pipe->top_pipe; other_pipe != NULL; other_pipe = other_pipe->top_pipe) {
+ if (other_pipe->global_sync.dcn4x.vready_offset_pixels > vready_offset)
+ vready_offset = other_pipe->global_sync.dcn4x.vready_offset_pixels;
+ }
+ for (other_pipe = pipe->next_odm_pipe; other_pipe != NULL; other_pipe = other_pipe->next_odm_pipe) {
+ if (other_pipe->global_sync.dcn4x.vready_offset_pixels > vready_offset)
+ vready_offset = other_pipe->global_sync.dcn4x.vready_offset_pixels;
+ }
+ for (other_pipe = pipe->prev_odm_pipe; other_pipe != NULL; other_pipe = other_pipe->prev_odm_pipe) {
+ if (other_pipe->global_sync.dcn4x.vready_offset_pixels > vready_offset)
+ vready_offset = other_pipe->global_sync.dcn4x.vready_offset_pixels;
+ }
+
+ return vready_offset;
+}
+
+static void dcn401_program_tg(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_state *context,
+ struct dce_hwseq *hws)
+{
+ pipe_ctx->stream_res.tg->funcs->program_global_sync(
+ pipe_ctx->stream_res.tg,
+ dcn401_calculate_vready_offset_for_group(pipe_ctx),
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vstartup_lines,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vupdate_offset_pixels,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vupdate_vupdate_width_pixels,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.pstate_keepout_start_lines);
+
+ if (dc_state_get_pipe_subvp_type(context, pipe_ctx) != SUBVP_PHANTOM)
+ pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VACTIVE);
+
+ pipe_ctx->stream_res.tg->funcs->set_vtg_params(
+ pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing, true);
+
+ if (hws->funcs.setup_vupdate_interrupt)
+ hws->funcs.setup_vupdate_interrupt(dc, pipe_ctx);
+}
+
+void dcn401_program_pipe(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_state *context)
+{
+ struct dce_hwseq *hws = dc->hwseq;
+
+ /* Only need to unblank on top pipe */
+ if (resource_is_pipe_type(pipe_ctx, OTG_MASTER)) {
+ if (pipe_ctx->update_flags.bits.enable ||
+ pipe_ctx->update_flags.bits.odm ||
+ pipe_ctx->stream->update_flags.bits.abm_level)
+ hws->funcs.blank_pixel_data(dc, pipe_ctx,
+ !pipe_ctx->plane_state ||
+ !pipe_ctx->plane_state->visible);
+ }
+
+ /* Only update TG on top pipe */
+ if (pipe_ctx->update_flags.bits.global_sync && !pipe_ctx->top_pipe
+ && !pipe_ctx->prev_odm_pipe)
+ dcn401_program_tg(dc, pipe_ctx, context, hws);
+
+ if (pipe_ctx->update_flags.bits.odm)
+ hws->funcs.update_odm(dc, context, pipe_ctx);
+
+ if (pipe_ctx->update_flags.bits.enable) {
+ if (hws->funcs.enable_plane)
+ hws->funcs.enable_plane(dc, pipe_ctx, context);
+ else
+ dc->hwss.enable_plane(dc, pipe_ctx, context);
+
+ if (dc->res_pool->hubbub->funcs->force_wm_propagate_to_pipes)
+ dc->res_pool->hubbub->funcs->force_wm_propagate_to_pipes(dc->res_pool->hubbub);
+ }
+
+ if (pipe_ctx->update_flags.bits.det_size) {
+ if (dc->res_pool->hubbub->funcs->program_det_size)
+ dc->res_pool->hubbub->funcs->program_det_size(
+ dc->res_pool->hubbub, pipe_ctx->plane_res.hubp->inst, pipe_ctx->det_buffer_size_kb);
+ if (dc->res_pool->hubbub->funcs->program_det_segments)
+ dc->res_pool->hubbub->funcs->program_det_segments(
+ dc->res_pool->hubbub, pipe_ctx->plane_res.hubp->inst, pipe_ctx->hubp_regs.det_size);
+ }
+
+ if (pipe_ctx->plane_state && (pipe_ctx->update_flags.raw ||
+ pipe_ctx->plane_state->update_flags.raw ||
+ pipe_ctx->stream->update_flags.raw))
+ dc->hwss.update_dchubp_dpp(dc, pipe_ctx, context);
+
+ if (pipe_ctx->plane_state && (pipe_ctx->update_flags.bits.enable ||
+ pipe_ctx->plane_state->update_flags.bits.hdr_mult))
+ hws->funcs.set_hdr_multiplier(pipe_ctx);
+
+ if (pipe_ctx->plane_state &&
+ (pipe_ctx->plane_state->update_flags.bits.in_transfer_func_change ||
+ pipe_ctx->plane_state->update_flags.bits.gamma_change ||
+ pipe_ctx->plane_state->update_flags.bits.lut_3d ||
+ pipe_ctx->update_flags.bits.enable))
+ hws->funcs.set_input_transfer_func(dc, pipe_ctx, pipe_ctx->plane_state);
+
+ /* dcn10_translate_regamma_to_hw_format takes 750us to finish
+ * only do gamma programming for powering on, internal memcmp to avoid
+ * updating on slave planes
+ */
+ if (pipe_ctx->update_flags.bits.enable ||
+ pipe_ctx->update_flags.bits.plane_changed ||
+ pipe_ctx->stream->update_flags.bits.out_tf ||
+ (pipe_ctx->plane_state &&
+ pipe_ctx->plane_state->update_flags.bits.output_tf_change))
+ hws->funcs.set_output_transfer_func(dc, pipe_ctx, pipe_ctx->stream);
+
+ /* If the pipe has been enabled or has a different opp, we
+ * should reprogram the fmt. This deals with cases where
+ * interation between mpc and odm combine on different streams
+ * causes a different pipe to be chosen to odm combine with.
+ */
+ if (pipe_ctx->update_flags.bits.enable
+ || pipe_ctx->update_flags.bits.opp_changed) {
+
+ pipe_ctx->stream_res.opp->funcs->opp_set_dyn_expansion(
+ pipe_ctx->stream_res.opp,
+ COLOR_SPACE_YCBCR601,
+ pipe_ctx->stream->timing.display_color_depth,
+ pipe_ctx->stream->signal);
+
+ pipe_ctx->stream_res.opp->funcs->opp_program_fmt(
+ pipe_ctx->stream_res.opp,
+ &pipe_ctx->stream->bit_depth_params,
+ &pipe_ctx->stream->clamping);
+ }
+
+ /* Set ABM pipe after other pipe configurations done */
+ if ((pipe_ctx->plane_state && pipe_ctx->plane_state->visible)) {
+ if (pipe_ctx->stream_res.abm) {
+ dc->hwss.set_pipe(pipe_ctx);
+ pipe_ctx->stream_res.abm->funcs->set_abm_level(pipe_ctx->stream_res.abm,
+ pipe_ctx->stream->abm_level);
+ }
+ }
+
+ if (pipe_ctx->update_flags.bits.test_pattern_changed) {
+ struct output_pixel_processor *odm_opp = pipe_ctx->stream_res.opp;
+ struct bit_depth_reduction_params params;
+
+ memset(&params, 0, sizeof(params));
+ odm_opp->funcs->opp_program_bit_depth_reduction(odm_opp, &params);
+ dc->hwss.set_disp_pattern_generator(dc,
+ pipe_ctx,
+ pipe_ctx->stream_res.test_pattern_params.test_pattern,
+ pipe_ctx->stream_res.test_pattern_params.color_space,
+ pipe_ctx->stream_res.test_pattern_params.color_depth,
+ NULL,
+ pipe_ctx->stream_res.test_pattern_params.width,
+ pipe_ctx->stream_res.test_pattern_params.height,
+ pipe_ctx->stream_res.test_pattern_params.offset);
+ }
+}
+
+void dcn401_program_front_end_for_ctx(
+ struct dc *dc,
+ struct dc_state *context)
+{
+ int i;
+ unsigned int prev_hubp_count = 0;
+ unsigned int hubp_count = 0;
+ struct dce_hwseq *hws = dc->hwseq;
+ struct pipe_ctx *pipe = NULL;
+
+ DC_LOGGER_INIT(dc->ctx->logger);
+
+ if (resource_is_pipe_topology_changed(dc->current_state, context))
+ resource_log_pipe_topology_update(dc, context);
+
+ if (dc->hwss.program_triplebuffer != NULL && dc->debug.enable_tri_buf) {
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ pipe = &context->res_ctx.pipe_ctx[i];
+
+ if (pipe->plane_state) {
+ if (pipe->plane_state->triplebuffer_flips)
+ BREAK_TO_DEBUGGER();
+
+ /*turn off triple buffer for full update*/
+ dc->hwss.program_triplebuffer(
+ dc, pipe, pipe->plane_state->triplebuffer_flips);
+ }
+ }
+ }
+
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ if (dc->current_state->res_ctx.pipe_ctx[i].plane_state)
+ prev_hubp_count++;
+ if (context->res_ctx.pipe_ctx[i].plane_state)
+ hubp_count++;
+ }
+
+ if (prev_hubp_count == 0 && hubp_count > 0) {
+ if (dc->res_pool->hubbub->funcs->force_pstate_change_control)
+ dc->res_pool->hubbub->funcs->force_pstate_change_control(
+ dc->res_pool->hubbub, true, false);
+ udelay(500);
+ }
+
+ /* Set pipe update flags and lock pipes */
+ for (i = 0; i < dc->res_pool->pipe_count; i++)
+ dc->hwss.detect_pipe_changes(dc->current_state, context, &dc->current_state->res_ctx.pipe_ctx[i],
+ &context->res_ctx.pipe_ctx[i]);
+
+ /* When disabling phantom pipes, turn on phantom OTG first (so we can get double
+ * buffer updates properly)
+ */
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ struct dc_stream_state *stream = dc->current_state->res_ctx.pipe_ctx[i].stream;
+
+ pipe = &dc->current_state->res_ctx.pipe_ctx[i];
+
+ if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable && stream &&
+ dc_state_get_pipe_subvp_type(dc->current_state, pipe) == SUBVP_PHANTOM) {
+ struct timing_generator *tg = dc->current_state->res_ctx.pipe_ctx[i].stream_res.tg;
+
+ if (tg->funcs->enable_crtc) {
+ if (dc->hwseq->funcs.blank_pixel_data)
+ dc->hwseq->funcs.blank_pixel_data(dc, pipe, true);
+
+ tg->funcs->enable_crtc(tg);
+ }
+ }
+ }
+ /* OTG blank before disabling all front ends */
+ for (i = 0; i < dc->res_pool->pipe_count; i++)
+ if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable
+ && !context->res_ctx.pipe_ctx[i].top_pipe
+ && !context->res_ctx.pipe_ctx[i].prev_odm_pipe
+ && context->res_ctx.pipe_ctx[i].stream)
+ hws->funcs.blank_pixel_data(dc, &context->res_ctx.pipe_ctx[i], true);
+
+
+ /* Disconnect mpcc */
+ for (i = 0; i < dc->res_pool->pipe_count; i++)
+ if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable
+ || context->res_ctx.pipe_ctx[i].update_flags.bits.opp_changed) {
+ struct hubbub *hubbub = dc->res_pool->hubbub;
+
+ /* Phantom pipe DET should be 0, but if a pipe in use is being transitioned to phantom
+ * then we want to do the programming here (effectively it's being disabled). If we do
+ * the programming later the DET won't be updated until the OTG for the phantom pipe is
+ * turned on (i.e. in an MCLK switch) which can come in too late and cause issues with
+ * DET allocation.
+ */
+ if ((context->res_ctx.pipe_ctx[i].update_flags.bits.disable ||
+ (context->res_ctx.pipe_ctx[i].plane_state &&
+ dc_state_get_pipe_subvp_type(context, &context->res_ctx.pipe_ctx[i]) ==
+ SUBVP_PHANTOM))) {
+ if (hubbub->funcs->program_det_size)
+ hubbub->funcs->program_det_size(hubbub,
+ dc->current_state->res_ctx.pipe_ctx[i].plane_res.hubp->inst, 0);
+ if (dc->res_pool->hubbub->funcs->program_det_segments)
+ dc->res_pool->hubbub->funcs->program_det_segments(
+ hubbub, dc->current_state->res_ctx.pipe_ctx[i].plane_res.hubp->inst, 0);
+ }
+ hws->funcs.plane_atomic_disconnect(dc, dc->current_state,
+ &dc->current_state->res_ctx.pipe_ctx[i]);
+ DC_LOG_DC("Reset mpcc for pipe %d\n", dc->current_state->res_ctx.pipe_ctx[i].pipe_idx);
+ }
+
+ /* update ODM for blanked OTG master pipes */
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ pipe = &context->res_ctx.pipe_ctx[i];
+ if (resource_is_pipe_type(pipe, OTG_MASTER) &&
+ !resource_is_pipe_type(pipe, DPP_PIPE) &&
+ pipe->update_flags.bits.odm &&
+ hws->funcs.update_odm)
+ hws->funcs.update_odm(dc, context, pipe);
+ }
+
+ /*
+ * Program all updated pipes, order matters for mpcc setup. Start with
+ * top pipe and program all pipes that follow in order
+ */
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ pipe = &context->res_ctx.pipe_ctx[i];
+
+ if (pipe->plane_state && !pipe->top_pipe) {
+ while (pipe) {
+ if (hws->funcs.program_pipe)
+ hws->funcs.program_pipe(dc, pipe, context);
+ else {
+ /* Don't program phantom pipes in the regular front end programming sequence.
+ * There is an MPO transition case where a pipe being used by a video plane is
+ * transitioned directly to be a phantom pipe when closing the MPO video.
+ * However the phantom pipe will program a new HUBP_VTG_SEL (update takes place
+ * right away) but the MPO still exists until the double buffered update of the
+ * main pipe so we will get a frame of underflow if the phantom pipe is
+ * programmed here.
+ */
+ if (pipe->stream &&
+ dc_state_get_pipe_subvp_type(context, pipe) != SUBVP_PHANTOM)
+ dcn401_program_pipe(dc, pipe, context);
+ }
+
+ pipe = pipe->bottom_pipe;
+ }
+ }
+
+ /* Program secondary blending tree and writeback pipes */
+ pipe = &context->res_ctx.pipe_ctx[i];
+ if (!pipe->top_pipe && !pipe->prev_odm_pipe
+ && pipe->stream && pipe->stream->num_wb_info > 0
+ && (pipe->update_flags.raw || (pipe->plane_state && pipe->plane_state->update_flags.raw)
+ || pipe->stream->update_flags.raw)
+ && hws->funcs.program_all_writeback_pipes_in_tree)
+ hws->funcs.program_all_writeback_pipes_in_tree(dc, pipe->stream, context);
+
+ /* Avoid underflow by check of pipe line read when adding 2nd plane. */
+ if (hws->wa.wait_hubpret_read_start_during_mpo_transition &&
+ !pipe->top_pipe &&
+ pipe->stream &&
+ pipe->plane_res.hubp->funcs->hubp_wait_pipe_read_start &&
+ dc->current_state->stream_status[0].plane_count == 1 &&
+ context->stream_status[0].plane_count > 1) {
+ pipe->plane_res.hubp->funcs->hubp_wait_pipe_read_start(pipe->plane_res.hubp);
+ }
+ }
+}
+
+void dcn401_post_unlock_program_front_end(
+ struct dc *dc,
+ struct dc_state *context)
+{
+ // Timeout for pipe enable
+ unsigned int timeout_us = 100000;
+ unsigned int polling_interval_us = 1;
+ struct dce_hwseq *hwseq = dc->hwseq;
+ int i;
+
+ DC_LOGGER_INIT(dc->ctx->logger);
+
+ for (i = 0; i < dc->res_pool->pipe_count; i++)
+ if (resource_is_pipe_type(&dc->current_state->res_ctx.pipe_ctx[i], OPP_HEAD) &&
+ !resource_is_pipe_type(&context->res_ctx.pipe_ctx[i], OPP_HEAD))
+ dc->hwss.post_unlock_reset_opp(dc,
+ &dc->current_state->res_ctx.pipe_ctx[i]);
+
+ for (i = 0; i < dc->res_pool->pipe_count; i++)
+ if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable)
+ dc->hwss.disable_plane(dc, dc->current_state, &dc->current_state->res_ctx.pipe_ctx[i]);
+
+ /*
+ * If we are enabling a pipe, we need to wait for pending clear as this is a critical
+ * part of the enable operation otherwise, DM may request an immediate flip which
+ * will cause HW to perform an "immediate enable" (as opposed to "vsync enable") which
+ * is unsupported on DCN.
+ */
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
+ // Don't check flip pending on phantom pipes
+ if (pipe->plane_state && !pipe->top_pipe && pipe->update_flags.bits.enable &&
+ dc_state_get_pipe_subvp_type(context, pipe) != SUBVP_PHANTOM) {
+ struct hubp *hubp = pipe->plane_res.hubp;
+ int j = 0;
+
+ for (j = 0; j < timeout_us / polling_interval_us
+ && hubp->funcs->hubp_is_flip_pending(hubp); j++)
+ udelay(polling_interval_us);
+ }
+ }
+
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
+ struct pipe_ctx *old_pipe = &dc->current_state->res_ctx.pipe_ctx[i];
+
+ /* When going from a smaller ODM slice count to larger, we must ensure double
+ * buffer update completes before we return to ensure we don't reduce DISPCLK
+ * before we've transitioned to 2:1 or 4:1
+ */
+ if (resource_is_pipe_type(old_pipe, OTG_MASTER) && resource_is_pipe_type(pipe, OTG_MASTER) &&
+ resource_get_odm_slice_count(old_pipe) < resource_get_odm_slice_count(pipe) &&
+ dc_state_get_pipe_subvp_type(context, pipe) != SUBVP_PHANTOM) {
+ int j = 0;
+ struct timing_generator *tg = pipe->stream_res.tg;
+
+ if (tg->funcs->get_optc_double_buffer_pending) {
+ for (j = 0; j < timeout_us / polling_interval_us
+ && tg->funcs->get_optc_double_buffer_pending(tg); j++)
+ udelay(polling_interval_us);
+ }
+ }
+ }
+
+ if (dc->res_pool->hubbub->funcs->force_pstate_change_control)
+ dc->res_pool->hubbub->funcs->force_pstate_change_control(
+ dc->res_pool->hubbub, false, false);
+
+
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ struct pipe_ctx *pipe = &context->res_ctx.pipe_ctx[i];
+
+ if (pipe->plane_state && !pipe->top_pipe) {
+ /* Program phantom pipe here to prevent a frame of underflow in the MPO transition
+ * case (if a pipe being used for a video plane transitions to a phantom pipe, it
+ * can underflow due to HUBP_VTG_SEL programming if done in the regular front end
+ * programming sequence).
+ */
+ while (pipe) {
+ if (pipe->stream && dc_state_get_pipe_subvp_type(context, pipe) == SUBVP_PHANTOM) {
+ /* When turning on the phantom pipe we want to run through the
+ * entire enable sequence, so apply all the "enable" flags.
+ */
+ if (dc->hwss.apply_update_flags_for_phantom)
+ dc->hwss.apply_update_flags_for_phantom(pipe);
+ if (dc->hwss.update_phantom_vp_position)
+ dc->hwss.update_phantom_vp_position(dc, context, pipe);
+ dcn401_program_pipe(dc, pipe, context);
+ }
+ pipe = pipe->bottom_pipe;
+ }
+ }
+ }
+
+ if (!hwseq)
+ return;
+
+ /* P-State support transitions:
+ * Natural -> FPO: P-State disabled in prepare, force disallow anytime is safe
+ * FPO -> Natural: Unforce anytime after FW disable is safe (P-State will assert naturally)
+ * Unsupported -> FPO: P-State enabled in optimize, force disallow anytime is safe
+ * FPO -> Unsupported: P-State disabled in prepare, unforce disallow anytime is safe
+ * FPO <-> SubVP: Force disallow is maintained on the FPO / SubVP pipes
+ */
+ if (hwseq->funcs.update_force_pstate)
+ dc->hwseq->funcs.update_force_pstate(dc, context);
+
+ /* Only program the MALL registers after all the main and phantom pipes
+ * are done programming.
+ */
+ if (hwseq->funcs.program_mall_pipe_config)
+ hwseq->funcs.program_mall_pipe_config(dc, context);
+
+ /* WA to apply WM setting*/
+ if (hwseq->wa.DEGVIDCN21)
+ dc->res_pool->hubbub->funcs->apply_DEDCN21_147_wa(dc->res_pool->hubbub);
+
+
+ /* WA for stutter underflow during MPO transitions when adding 2nd plane */
+ if (hwseq->wa.disallow_self_refresh_during_multi_plane_transition) {
+
+ if (dc->current_state->stream_status[0].plane_count == 1 &&
+ context->stream_status[0].plane_count > 1) {
+
+ struct timing_generator *tg = dc->res_pool->timing_generators[0];
+
+ dc->res_pool->hubbub->funcs->allow_self_refresh_control(dc->res_pool->hubbub, false);
+
+ hwseq->wa_state.disallow_self_refresh_during_multi_plane_transition_applied = true;
+ hwseq->wa_state.disallow_self_refresh_during_multi_plane_transition_applied_on_frame =
+ tg->funcs->get_frame_count(tg);
+ }
+ }
+}
+
+bool dcn401_update_bandwidth(
+ struct dc *dc,
+ struct dc_state *context)
+{
+ int i;
+ struct dce_hwseq *hws = dc->hwseq;
+
+ /* recalculate DML parameters */
+ if (dc->res_pool->funcs->validate_bandwidth(dc, context, DC_VALIDATE_MODE_AND_PROGRAMMING) != DC_OK)
+ return false;
+
+ /* apply updated bandwidth parameters */
+ dc->hwss.prepare_bandwidth(dc, context);
+
+ /* update hubp configs for all pipes */
+ for (i = 0; i < dc->res_pool->pipe_count; i++) {
+ struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+
+ if (pipe_ctx->plane_state == NULL)
+ continue;
+
+ if (pipe_ctx->top_pipe == NULL) {
+ bool blank = !is_pipe_tree_visible(pipe_ctx);
+
+ pipe_ctx->stream_res.tg->funcs->program_global_sync(
+ pipe_ctx->stream_res.tg,
+ dcn401_calculate_vready_offset_for_group(pipe_ctx),
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vstartup_lines,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vupdate_offset_pixels,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.vupdate_vupdate_width_pixels,
+ (unsigned int)pipe_ctx->global_sync.dcn4x.pstate_keepout_start_lines);
+
+ pipe_ctx->stream_res.tg->funcs->set_vtg_params(
+ pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing, false);
+
+ if (pipe_ctx->prev_odm_pipe == NULL)
+ hws->funcs.blank_pixel_data(dc, pipe_ctx, blank);
+
+ if (hws->funcs.setup_vupdate_interrupt)
+ hws->funcs.setup_vupdate_interrupt(dc, pipe_ctx);
+ }
+
+ if (pipe_ctx->plane_res.hubp->funcs->hubp_setup2)
+ pipe_ctx->plane_res.hubp->funcs->hubp_setup2(
+ pipe_ctx->plane_res.hubp,
+ &pipe_ctx->hubp_regs,
+ &pipe_ctx->global_sync,
+ &pipe_ctx->stream->timing);
+ }
+
+ return true;
+}
+
+void dcn401_detect_pipe_changes(struct dc_state *old_state,
+ struct dc_state *new_state,
+ struct pipe_ctx *old_pipe,
+ struct pipe_ctx *new_pipe)
+{
+ bool old_is_phantom = dc_state_get_pipe_subvp_type(old_state, old_pipe) == SUBVP_PHANTOM;
+ bool new_is_phantom = dc_state_get_pipe_subvp_type(new_state, new_pipe) == SUBVP_PHANTOM;
+
+ unsigned int old_pipe_vready_offset_pixels = old_pipe->global_sync.dcn4x.vready_offset_pixels;
+ unsigned int new_pipe_vready_offset_pixels = new_pipe->global_sync.dcn4x.vready_offset_pixels;
+ unsigned int old_pipe_vstartup_lines = old_pipe->global_sync.dcn4x.vstartup_lines;
+ unsigned int new_pipe_vstartup_lines = new_pipe->global_sync.dcn4x.vstartup_lines;
+ unsigned int old_pipe_vupdate_offset_pixels = old_pipe->global_sync.dcn4x.vupdate_offset_pixels;
+ unsigned int new_pipe_vupdate_offset_pixels = new_pipe->global_sync.dcn4x.vupdate_offset_pixels;
+ unsigned int old_pipe_vupdate_width_pixels = old_pipe->global_sync.dcn4x.vupdate_vupdate_width_pixels;
+ unsigned int new_pipe_vupdate_width_pixels = new_pipe->global_sync.dcn4x.vupdate_vupdate_width_pixels;
+
+ new_pipe->update_flags.raw = 0;
+
+ /* If non-phantom pipe is being transitioned to a phantom pipe,
+ * set disable and return immediately. This is because the pipe
+ * that was previously in use must be fully disabled before we
+ * can "enable" it as a phantom pipe (since the OTG will certainly
+ * be different). The post_unlock sequence will set the correct
+ * update flags to enable the phantom pipe.
+ */
+ if (old_pipe->plane_state && !old_is_phantom &&
+ new_pipe->plane_state && new_is_phantom) {
+ new_pipe->update_flags.bits.disable = 1;
+ return;
+ }
+
+ if (resource_is_pipe_type(new_pipe, OTG_MASTER) &&
+ resource_is_odm_topology_changed(new_pipe, old_pipe))
+ /* Detect odm changes */
+ new_pipe->update_flags.bits.odm = 1;
+
+ /* Exit on unchanged, unused pipe */
+ if (!old_pipe->plane_state && !new_pipe->plane_state)
+ return;
+ /* Detect pipe enable/disable */
+ if (!old_pipe->plane_state && new_pipe->plane_state) {
+ new_pipe->update_flags.bits.enable = 1;
+ new_pipe->update_flags.bits.mpcc = 1;
+ new_pipe->update_flags.bits.dppclk = 1;
+ new_pipe->update_flags.bits.hubp_interdependent = 1;
+ new_pipe->update_flags.bits.hubp_rq_dlg_ttu = 1;
+ new_pipe->update_flags.bits.unbounded_req = 1;
+ new_pipe->update_flags.bits.gamut_remap = 1;
+ new_pipe->update_flags.bits.scaler = 1;
+ new_pipe->update_flags.bits.viewport = 1;
+ new_pipe->update_flags.bits.det_size = 1;
+ if (new_pipe->stream->test_pattern.type != DP_TEST_PATTERN_VIDEO_MODE &&
+ new_pipe->stream_res.test_pattern_params.width != 0 &&
+ new_pipe->stream_res.test_pattern_params.height != 0)
+ new_pipe->update_flags.bits.test_pattern_changed = 1;
+ if (!new_pipe->top_pipe && !new_pipe->prev_odm_pipe) {
+ new_pipe->update_flags.bits.odm = 1;
+ new_pipe->update_flags.bits.global_sync = 1;
+ }
+ return;
+ }
+
+ /* For SubVP we need to unconditionally enable because any phantom pipes are
+ * always removed then newly added for every full updates whenever SubVP is in use.
+ * The remove-add sequence of the phantom pipe always results in the pipe
+ * being blanked in enable_stream_timing (DPG).
+ */
+ if (new_pipe->stream && dc_state_get_pipe_subvp_type(new_state, new_pipe) == SUBVP_PHANTOM)
+ new_pipe->update_flags.bits.enable = 1;
+
+ /* Phantom pipes are effectively disabled, if the pipe was previously phantom
+ * we have to enable
+ */
+ if (old_pipe->plane_state && old_is_phantom &&
+ new_pipe->plane_state && !new_is_phantom)
+ new_pipe->update_flags.bits.enable = 1;
+
+ if (old_pipe->plane_state && !new_pipe->plane_state) {
+ new_pipe->update_flags.bits.disable = 1;
+ return;
+ }
+
+ /* Detect plane change */
+ if (old_pipe->plane_state != new_pipe->plane_state)
+ new_pipe->update_flags.bits.plane_changed = true;
+
+ /* Detect top pipe only changes */
+ if (resource_is_pipe_type(new_pipe, OTG_MASTER)) {
+ /* Detect global sync changes */
+ if ((old_pipe_vready_offset_pixels != new_pipe_vready_offset_pixels)
+ || (old_pipe_vstartup_lines != new_pipe_vstartup_lines)
+ || (old_pipe_vupdate_offset_pixels != new_pipe_vupdate_offset_pixels)
+ || (old_pipe_vupdate_width_pixels != new_pipe_vupdate_width_pixels))
+ new_pipe->update_flags.bits.global_sync = 1;
+ }
+
+ if (old_pipe->det_buffer_size_kb != new_pipe->det_buffer_size_kb)
+ new_pipe->update_flags.bits.det_size = 1;
+
+ /*
+ * Detect opp / tg change, only set on change, not on enable
+ * Assume mpcc inst = pipe index, if not this code needs to be updated
+ * since mpcc is what is affected by these. In fact all of our sequence
+ * makes this assumption at the moment with how hubp reset is matched to
+ * same index mpcc reset.
+ */
+ if (old_pipe->stream_res.opp != new_pipe->stream_res.opp)
+ new_pipe->update_flags.bits.opp_changed = 1;
+ if (old_pipe->stream_res.tg != new_pipe->stream_res.tg)
+ new_pipe->update_flags.bits.tg_changed = 1;
+
+ /*
+ * Detect mpcc blending changes, only dpp inst and opp matter here,
+ * mpccs getting removed/inserted update connected ones during their own
+ * programming
+ */
+ if (old_pipe->plane_res.dpp != new_pipe->plane_res.dpp
+ || old_pipe->stream_res.opp != new_pipe->stream_res.opp)
+ new_pipe->update_flags.bits.mpcc = 1;
+
+ /* Detect dppclk change */
+ if (old_pipe->plane_res.bw.dppclk_khz != new_pipe->plane_res.bw.dppclk_khz)
+ new_pipe->update_flags.bits.dppclk = 1;
+
+ /* Check for scl update */
+ if (memcmp(&old_pipe->plane_res.scl_data, &new_pipe->plane_res.scl_data, sizeof(struct scaler_data)))
+ new_pipe->update_flags.bits.scaler = 1;
+ /* Check for vp update */
+ if (memcmp(&old_pipe->plane_res.scl_data.viewport, &new_pipe->plane_res.scl_data.viewport, sizeof(struct rect))
+ || memcmp(&old_pipe->plane_res.scl_data.viewport_c,
+ &new_pipe->plane_res.scl_data.viewport_c, sizeof(struct rect)))
+ new_pipe->update_flags.bits.viewport = 1;
+
+ /* Detect dlg/ttu/rq updates */
+ {
+ struct dml2_display_dlg_regs old_dlg_regs = old_pipe->hubp_regs.dlg_regs;
+ struct dml2_display_ttu_regs old_ttu_regs = old_pipe->hubp_regs.ttu_regs;
+ struct dml2_display_rq_regs old_rq_regs = old_pipe->hubp_regs.rq_regs;
+ struct dml2_display_dlg_regs *new_dlg_regs = &new_pipe->hubp_regs.dlg_regs;
+ struct dml2_display_ttu_regs *new_ttu_regs = &new_pipe->hubp_regs.ttu_regs;
+ struct dml2_display_rq_regs *new_rq_regs = &new_pipe->hubp_regs.rq_regs;
+
+ /* Detect pipe interdependent updates */
+ if ((old_dlg_regs.dst_y_prefetch != new_dlg_regs->dst_y_prefetch)
+ || (old_dlg_regs.vratio_prefetch != new_dlg_regs->vratio_prefetch)
+ || (old_dlg_regs.vratio_prefetch_c != new_dlg_regs->vratio_prefetch_c)
+ || (old_dlg_regs.dst_y_per_vm_vblank != new_dlg_regs->dst_y_per_vm_vblank)
+ || (old_dlg_regs.dst_y_per_row_vblank != new_dlg_regs->dst_y_per_row_vblank)
+ || (old_dlg_regs.dst_y_per_vm_flip != new_dlg_regs->dst_y_per_vm_flip)
+ || (old_dlg_regs.dst_y_per_row_flip != new_dlg_regs->dst_y_per_row_flip)
+ || (old_dlg_regs.refcyc_per_meta_chunk_vblank_l != new_dlg_regs->refcyc_per_meta_chunk_vblank_l)
+ || (old_dlg_regs.refcyc_per_meta_chunk_vblank_c != new_dlg_regs->refcyc_per_meta_chunk_vblank_c)
+ || (old_dlg_regs.refcyc_per_meta_chunk_flip_l != new_dlg_regs->refcyc_per_meta_chunk_flip_l)
+ || (old_dlg_regs.refcyc_per_line_delivery_pre_l != new_dlg_regs->refcyc_per_line_delivery_pre_l)
+ || (old_dlg_regs.refcyc_per_line_delivery_pre_c != new_dlg_regs->refcyc_per_line_delivery_pre_c)
+ || (old_ttu_regs.refcyc_per_req_delivery_pre_l != new_ttu_regs->refcyc_per_req_delivery_pre_l)
+ || (old_ttu_regs.refcyc_per_req_delivery_pre_c != new_ttu_regs->refcyc_per_req_delivery_pre_c)
+ || (old_ttu_regs.refcyc_per_req_delivery_pre_cur0 !=
+ new_ttu_regs->refcyc_per_req_delivery_pre_cur0)
+ || (old_ttu_regs.min_ttu_vblank != new_ttu_regs->min_ttu_vblank)
+ || (old_ttu_regs.qos_level_flip != new_ttu_regs->qos_level_flip)) {
+ old_dlg_regs.dst_y_prefetch = new_dlg_regs->dst_y_prefetch;
+ old_dlg_regs.vratio_prefetch = new_dlg_regs->vratio_prefetch;
+ old_dlg_regs.vratio_prefetch_c = new_dlg_regs->vratio_prefetch_c;
+ old_dlg_regs.dst_y_per_vm_vblank = new_dlg_regs->dst_y_per_vm_vblank;
+ old_dlg_regs.dst_y_per_row_vblank = new_dlg_regs->dst_y_per_row_vblank;
+ old_dlg_regs.dst_y_per_vm_flip = new_dlg_regs->dst_y_per_vm_flip;
+ old_dlg_regs.dst_y_per_row_flip = new_dlg_regs->dst_y_per_row_flip;
+ old_dlg_regs.refcyc_per_meta_chunk_vblank_l = new_dlg_regs->refcyc_per_meta_chunk_vblank_l;
+ old_dlg_regs.refcyc_per_meta_chunk_vblank_c = new_dlg_regs->refcyc_per_meta_chunk_vblank_c;
+ old_dlg_regs.refcyc_per_meta_chunk_flip_l = new_dlg_regs->refcyc_per_meta_chunk_flip_l;
+ old_dlg_regs.refcyc_per_line_delivery_pre_l = new_dlg_regs->refcyc_per_line_delivery_pre_l;
+ old_dlg_regs.refcyc_per_line_delivery_pre_c = new_dlg_regs->refcyc_per_line_delivery_pre_c;
+ old_ttu_regs.refcyc_per_req_delivery_pre_l = new_ttu_regs->refcyc_per_req_delivery_pre_l;
+ old_ttu_regs.refcyc_per_req_delivery_pre_c = new_ttu_regs->refcyc_per_req_delivery_pre_c;
+ old_ttu_regs.refcyc_per_req_delivery_pre_cur0 = new_ttu_regs->refcyc_per_req_delivery_pre_cur0;
+ old_ttu_regs.min_ttu_vblank = new_ttu_regs->min_ttu_vblank;
+ old_ttu_regs.qos_level_flip = new_ttu_regs->qos_level_flip;
+ new_pipe->update_flags.bits.hubp_interdependent = 1;
+ }
+ /* Detect any other updates to ttu/rq/dlg */
+ if (memcmp(&old_dlg_regs, new_dlg_regs, sizeof(old_dlg_regs)) ||
+ memcmp(&old_ttu_regs, new_ttu_regs, sizeof(old_ttu_regs)) ||
+ memcmp(&old_rq_regs, new_rq_regs, sizeof(old_rq_regs)))
+ new_pipe->update_flags.bits.hubp_rq_dlg_ttu = 1;
+ }
+
+ if (old_pipe->unbounded_req != new_pipe->unbounded_req)
+ new_pipe->update_flags.bits.unbounded_req = 1;
+
+ if (memcmp(&old_pipe->stream_res.test_pattern_params,
+ &new_pipe->stream_res.test_pattern_params, sizeof(struct test_pattern_params))) {
+ new_pipe->update_flags.bits.test_pattern_changed = 1;
+ }
+}
+
+void dcn401_plane_atomic_power_down(struct dc *dc,
+ struct dpp *dpp,
+ struct hubp *hubp)
+{
+ struct dce_hwseq *hws = dc->hwseq;
+ uint32_t org_ip_request_cntl = 0;
+
+ DC_LOGGER_INIT(dc->ctx->logger);
+
+ if (REG(DC_IP_REQUEST_CNTL)) {
+ REG_GET(DC_IP_REQUEST_CNTL, IP_REQUEST_EN, &org_ip_request_cntl);
+ if (org_ip_request_cntl == 0)
+ REG_SET(DC_IP_REQUEST_CNTL, 0,
+ IP_REQUEST_EN, 1);
+ }
+
+ if (hws->funcs.dpp_pg_control)
+ hws->funcs.dpp_pg_control(hws, dpp->inst, false);
+
+ if (hws->funcs.hubp_pg_control)
+ hws->funcs.hubp_pg_control(hws, hubp->inst, false);
+
+ hubp->funcs->hubp_reset(hubp);
+ dpp->funcs->dpp_reset(dpp);
+
+ if (org_ip_request_cntl == 0 && REG(DC_IP_REQUEST_CNTL))
+ REG_SET(DC_IP_REQUEST_CNTL, 0,
+ IP_REQUEST_EN, 0);
+
+ DC_LOG_DEBUG(
+ "Power gated front end %d\n", hubp->inst);
+
+ if (hws->funcs.dpp_root_clock_control)
+ hws->funcs.dpp_root_clock_control(hws, dpp->inst, false);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.h b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.h
index 28a513dfc005..2621b7725267 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_hwseq.h
@@ -63,8 +63,6 @@ void dcn401_set_cursor_position(struct pipe_ctx *pipe_ctx);
bool dcn401_apply_idle_power_optimizations(struct dc *dc, bool enable);
-struct ips_ono_region_state dcn401_read_ono_state(struct dc *dc,
- uint8_t region);
void dcn401_wait_for_dcc_meta_propagation(const struct dc *dc,
const struct pipe_ctx *top_pipe_to_program);
@@ -95,6 +93,21 @@ void dcn401_reset_back_end_for_pipe(
void dcn401_reset_hw_ctx_wrap(
struct dc *dc,
struct dc_state *context);
+void dcn401_program_pipe(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_state *context);
void dcn401_perform_3dlut_wa_unlock(struct pipe_ctx *pipe_ctx);
-
+void dcn401_program_front_end_for_ctx(struct dc *dc, struct dc_state *context);
+void dcn401_post_unlock_program_front_end(struct dc *dc, struct dc_state *context);
+bool dcn401_update_bandwidth(struct dc *dc, struct dc_state *context);
+void dcn401_detect_pipe_changes(
+ struct dc_state *old_state,
+ struct dc_state *new_state,
+ struct pipe_ctx *old_pipe,
+ struct pipe_ctx *new_pipe);
+void dcn401_plane_atomic_power_down(struct dc *dc,
+ struct dpp *dpp,
+ struct hubp *hubp);
+void dcn401_initialize_min_clocks(struct dc *dc);
#endif /* __DC_HWSS_DCN401_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c
index 23e4f208152e..fe7aceb2f510 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn401/dcn401_init.c
@@ -17,9 +17,10 @@ static const struct hw_sequencer_funcs dcn401_funcs = {
.init_hw = dcn401_init_hw,
.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
.apply_ctx_for_surface = NULL,
- .program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+ .program_front_end_for_ctx = dcn401_program_front_end_for_ctx,
+ .clear_surface_dcc_and_tiling = dcn10_reset_surface_dcc_and_tiling,
.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
- .post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
+ .post_unlock_program_front_end = dcn401_post_unlock_program_front_end,
.update_plane_addr = dcn20_update_plane_addr,
.update_dchub = dcn10_update_dchub,
.update_pending_status = dcn10_update_pending_status,
@@ -42,7 +43,7 @@ static const struct hw_sequencer_funcs dcn401_funcs = {
.cursor_lock = dcn10_cursor_lock,
.prepare_bandwidth = dcn401_prepare_bandwidth,
.optimize_bandwidth = dcn401_optimize_bandwidth,
- .update_bandwidth = dcn20_update_bandwidth,
+ .update_bandwidth = dcn401_update_bandwidth,
.set_drr = dcn10_set_drr,
.get_position = dcn10_get_position,
.set_static_screen_control = dcn31_set_static_screen_control,
@@ -66,7 +67,6 @@ static const struct hw_sequencer_funcs dcn401_funcs = {
.enable_writeback = dcn30_enable_writeback,
.disable_writeback = dcn30_disable_writeback,
.update_writeback = dcn30_update_writeback,
- .mmhubbub_warmup = dcn30_mmhubbub_warmup,
.dmdata_status_done = dcn20_dmdata_status_done,
.program_dmdata_engine = dcn30_program_dmdata_engine,
.set_dmdata_attributes = dcn20_set_dmdata_attributes,
@@ -100,6 +100,10 @@ static const struct hw_sequencer_funcs dcn401_funcs = {
.fams2_global_control_lock_fast = dcn401_fams2_global_control_lock_fast,
.program_outstanding_updates = dcn401_program_outstanding_updates,
.wait_for_all_pending_updates = dcn30_wait_for_all_pending_updates,
+ .detect_pipe_changes = dcn401_detect_pipe_changes,
+ .enable_plane = dcn20_enable_plane,
+ .update_dchubp_dpp = dcn20_update_dchubp_dpp,
+ .post_unlock_reset_opp = dcn20_post_unlock_reset_opp,
};
static const struct hwseq_private_funcs dcn401_private_funcs = {
@@ -120,7 +124,7 @@ static const struct hwseq_private_funcs dcn401_private_funcs = {
.disable_vga = dcn20_disable_vga,
.bios_golden_init = dcn10_bios_golden_init,
.plane_atomic_disable = dcn20_plane_atomic_disable,
- .plane_atomic_power_down = dcn10_plane_atomic_power_down,
+ .plane_atomic_power_down = dcn401_plane_atomic_power_down,
.enable_power_gating_plane = dcn32_enable_power_gating_plane,
.hubp_pg_control = dcn32_hubp_pg_control,
.program_all_writeback_pipes_in_tree = dcn30_program_all_writeback_pipes_in_tree,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h
index 66fdc5805d0a..9df8030e37f7 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer.h
@@ -46,6 +46,7 @@ struct dce_hwseq;
struct link_resource;
struct dc_dmub_cmd;
struct pg_block_update;
+struct drr_params;
struct subvp_pipe_control_lock_fast_params {
struct dc *dc;
@@ -194,7 +195,8 @@ enum block_sequence_func {
DMUB_SUBVP_SAVE_SURF_ADDR,
HUBP_WAIT_FOR_DCC_META_PROP,
DMUB_FAMS2_GLOBAL_CONTROL_LOCK_FAST,
-
+ /* This must be the last value in this enum, add new ones above */
+ HWSS_BLOCK_SEQUENCE_FUNC_COUNT
};
struct block_sequence {
@@ -202,6 +204,8 @@ struct block_sequence {
enum block_sequence_func func;
};
+#define MAX_HWSS_BLOCK_SEQUENCE_SIZE (HWSS_BLOCK_SEQUENCE_FUNC_COUNT * MAX_PIPES)
+
struct hw_sequencer_funcs {
void (*hardware_release)(struct dc *dc);
/* Embedded Display Related */
@@ -241,6 +245,7 @@ struct hw_sequencer_funcs {
struct pipe_ctx *pipe_ctx, bool enableTripleBuffer);
void (*update_pending_status)(struct pipe_ctx *pipe_ctx);
void (*update_dsc_pg)(struct dc *dc, struct dc_state *context, bool safe_to_disable);
+ void (*clear_surface_dcc_and_tiling)(struct pipe_ctx *pipe_ctx, struct dc_plane_state *plane_state, bool clear_tiling);
/* Pipe Lock Related */
void (*pipe_control_lock)(struct dc *dc,
@@ -331,10 +336,6 @@ struct hw_sequencer_funcs {
void (*disable_writeback)(struct dc *dc,
unsigned int dwb_pipe_inst);
- bool (*mmhubbub_warmup)(struct dc *dc,
- unsigned int num_dwb,
- struct dc_writeback_info *wb_info);
-
/* Clock Related */
enum dc_status (*set_clock)(struct dc *dc,
enum dc_clock_type clock_type,
@@ -462,6 +463,18 @@ struct hw_sequencer_funcs {
struct dc_state *context);
void (*setup_hpo_hw_control)(const struct dce_hwseq *hws, bool enable);
void (*wait_for_all_pending_updates)(const struct pipe_ctx *pipe_ctx);
+ void (*detect_pipe_changes)(struct dc_state *old_state,
+ struct dc_state *new_state,
+ struct pipe_ctx *old_pipe,
+ struct pipe_ctx *new_pipe);
+ void (*enable_plane)(struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_state *context);
+ void (*update_dchubp_dpp)(struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct dc_state *context);
+ void (*post_unlock_reset_opp)(struct dc *dc,
+ struct pipe_ctx *opp_head);
};
void color_space_to_black_color(
@@ -489,11 +502,15 @@ void get_hdr_visual_confirm_color(
void get_mpctree_visual_confirm_color(
struct pipe_ctx *pipe_ctx,
struct tg_color *color);
-
+void get_smartmux_visual_confirm_color(
+ struct dc *dc,
+ struct tg_color *color);
+void get_vabc_visual_confirm_color(
+ struct pipe_ctx *pipe_ctx,
+ struct tg_color *color);
void get_subvp_visual_confirm_color(
struct pipe_ctx *pipe_ctx,
struct tg_color *color);
-
void get_fams2_visual_confirm_color(
struct dc *dc,
struct dc_state *context,
@@ -508,19 +525,29 @@ void get_cursor_visual_confirm_color(
struct pipe_ctx *pipe_ctx,
struct tg_color *color);
+void get_dcc_visual_confirm_color(
+ struct dc *dc,
+ struct pipe_ctx *pipe_ctx,
+ struct tg_color *color);
+
void set_p_state_switch_method(
struct dc *dc,
struct dc_state *context,
struct pipe_ctx *pipe_ctx);
+void set_drr_and_clear_adjust_pending(
+ struct pipe_ctx *pipe_ctx,
+ struct dc_stream_state *stream,
+ struct drr_params *params);
+
void hwss_execute_sequence(struct dc *dc,
- struct block_sequence block_sequence[],
+ struct block_sequence block_sequence[MAX_HWSS_BLOCK_SEQUENCE_SIZE],
int num_steps);
void hwss_build_fast_sequence(struct dc *dc,
struct dc_dmub_cmd *dc_dmub_cmd,
unsigned int dmub_cmd_count,
- struct block_sequence block_sequence[],
+ struct block_sequence block_sequence[MAX_HWSS_BLOCK_SEQUENCE_SIZE],
unsigned int *num_steps,
struct pipe_ctx *pipe_ctx,
struct dc_stream_status *stream_status,
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h
index 22a5d4a03c98..1e2d247fbbac 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h
+++ b/drivers/gpu/drm/amd/display/dc/hwss/hw_sequencer_private.h
@@ -49,6 +49,7 @@ struct hwseq_wa_state {
bool DEGVIDCN10_253_applied;
bool disallow_self_refresh_during_multi_plane_transition_applied;
unsigned int disallow_self_refresh_during_multi_plane_transition_applied_on_frame;
+ bool skip_blank_stream;
};
struct pipe_ctx;
@@ -183,6 +184,8 @@ struct hwseq_private_funcs {
struct dc_cm2_func_luts mcm_luts,
bool lut_bank_a);
void (*perform_3dlut_wa_unlock)(struct pipe_ctx *pipe_ctx);
+ void (*wait_for_pipe_update_if_needed)(struct dc *dc, struct pipe_ctx *pipe_ctx, bool is_surface_update_only);
+ void (*set_wait_for_update_needed_for_pipe)(struct dc *dc, struct pipe_ctx *pipe_ctx);
};
struct dce_hwseq {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_status.h b/drivers/gpu/drm/amd/display/dc/inc/core_status.h
index b5afd8c3103d..82085d9c3f40 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/core_status.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/core_status.h
@@ -26,6 +26,8 @@
#ifndef _CORE_STATUS_H_
#define _CORE_STATUS_H_
+#include "dc_hw_types.h"
+
enum dc_status {
DC_OK = 1,
@@ -56,6 +58,8 @@ enum dc_status {
DC_NO_LINK_ENC_RESOURCE = 26,
DC_FAIL_DP_PAYLOAD_ALLOCATION = 27,
DC_FAIL_DP_LINK_BANDWIDTH = 28,
+ DC_FAIL_HW_CURSOR_SUPPORT = 29,
+ DC_FAIL_DP_TUNNEL_BW_VALIDATE = 30,
DC_ERROR_UNEXPECTED = -1
};
diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
index 8597e866bfe6..f0d7185153b2 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
@@ -39,15 +39,12 @@
#include "panel_cntl.h"
#include "dmub/inc/dmub_cmd.h"
#include "pg_cntl.h"
-#include "spl/dc_spl.h"
+#include "sspl/dc_spl.h"
#define MAX_CLOCK_SOURCES 7
#define MAX_SVP_PHANTOM_STREAMS 2
#define MAX_SVP_PHANTOM_PLANES 2
-void enable_surface_flip_reporting(struct dc_plane_state *plane_state,
- uint32_t controller_id);
-
#include "grph_object_id.h"
#include "link_encoder.h"
#include "stream_encoder.h"
@@ -68,6 +65,9 @@ struct resource_pool;
struct dc_state;
struct resource_context;
struct clk_bw_params;
+struct dc_mcache_params;
+
+#define MAX_RMCM_INST 2
struct resource_funcs {
enum engine_id (*get_preferred_eng_id_dpia)(unsigned int dpia_index);
@@ -81,11 +81,10 @@ struct resource_funcs {
/* Create a minimal link encoder object with no dc_link object
* associated with it. */
struct link_encoder *(*link_enc_create_minimal)(struct dc_context *ctx, enum engine_id eng_id);
-
- bool (*validate_bandwidth)(
+ enum dc_status (*validate_bandwidth)(
struct dc *dc,
struct dc_state *context,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void (*calculate_wm_and_dlg)(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
@@ -110,7 +109,7 @@ struct resource_funcs {
struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
/*
* Algorithm for assigning available link encoders to links.
@@ -219,6 +218,18 @@ struct resource_funcs {
* Get indicator of power from a context that went through full validation
*/
int (*get_power_profile)(const struct dc_state *context);
+ unsigned int (*get_det_buffer_size)(const struct dc_state *context);
+ unsigned int (*get_vstartup_for_pipe)(struct pipe_ctx *pipe_ctx);
+ unsigned int (*get_max_hw_cursor_size)(const struct dc *dc,
+ struct dc_state *state,
+ const struct dc_stream_state *stream);
+ bool (*program_mcache_pipe_config)(struct dc_state *context,
+ const struct dc_mcache_params *mcache_params);
+ enum dc_status (*update_dc_state_for_encoder_switch)(struct dc_link *link,
+ struct dc_link_settings *link_setting,
+ uint8_t pipe_count,
+ struct pipe_ctx *pipes,
+ struct audio_output *audio_output);
};
struct audio_support{
@@ -277,6 +288,7 @@ struct resource_pool {
struct hpo_dp_link_encoder *hpo_dp_link_enc[MAX_HPO_DP2_LINK_ENCODERS];
struct dc_3dlut *mpc_lut[MAX_PIPES];
struct dc_transfer_func *mpc_shaper[MAX_PIPES];
+ struct dc_rmcm_3dlut rmcm_3dlut[MAX_RMCM_INST];
struct {
unsigned int xtalin_clock_inKhz;
@@ -377,12 +389,15 @@ struct plane_resource {
/* all mappable hardware resources used to enable a link */
struct link_resource {
+ struct link_encoder *dio_link_enc;
struct hpo_dp_link_encoder *hpo_dp_link_enc;
};
struct link_config {
struct dc_link_settings dp_link_settings;
+ struct dc_tunnel_settings dp_tunnel_settings;
};
+
union pipe_update_flags {
struct {
uint32_t enable : 1;
@@ -467,6 +482,7 @@ struct pipe_ctx {
unsigned int surface_size_in_mall_bytes;
struct dml2_dchub_per_pipe_register_set hubp_regs;
struct dml2_hubp_pipe_mcache_regs mcache_regs;
+ union dml2_global_sync_programming global_sync;
struct dwbc *dwbc;
struct mcif_wb *mcif_wb;
@@ -477,6 +493,12 @@ struct pipe_ctx {
/* subvp_index: only valid if the pipe is a SUBVP_MAIN*/
uint8_t subvp_index;
struct pixel_rate_divider pixel_rate_divider;
+ /* pixels borrowed from hblank to hactive */
+ uint8_t hblank_borrow;
+ /* next vupdate */
+ uint32_t next_vupdate;
+ uint32_t wait_frame_count;
+ bool wait_is_required;
};
/* Data used for dynamic link encoder assignment.
@@ -498,11 +520,13 @@ struct resource_context {
uint8_t dp_clock_source_ref_count;
bool is_dsc_acquired[MAX_PIPES];
struct link_enc_cfg_context link_enc_cfg_ctx;
+ unsigned int dio_link_enc_to_link_idx[MAX_DIG_LINK_ENCODERS];
+ int dio_link_enc_ref_cnts[MAX_DIG_LINK_ENCODERS];
bool is_hpo_dp_stream_enc_acquired[MAX_HPO_DP2_ENCODERS];
unsigned int hpo_dp_link_enc_to_link_idx[MAX_HPO_DP2_LINK_ENCODERS];
int hpo_dp_link_enc_ref_cnts[MAX_HPO_DP2_LINK_ENCODERS];
bool is_mpc_3dlut_acquired[MAX_PIPES];
- /* solely used for build scalar data in dml2 */
+ /* used to build scalar data in dml2 and for edp backlight programming */
struct pipe_ctx temp_pipe;
};
@@ -539,7 +563,11 @@ struct dcn_bw_output {
bool legacy_svp_drr_stream_index_valid;
struct dml2_mcache_surface_allocation mcache_allocations[DML2_MAX_PLANES];
struct dmub_cmd_fams2_global_config fams2_global_config;
- struct dmub_fams2_stream_static_state fams2_stream_params[DML2_MAX_PLANES];
+ union dmub_cmd_fams2_config fams2_stream_base_params[DML2_MAX_PLANES];
+ union {
+ union dmub_cmd_fams2_config fams2_stream_sub_params[DML2_MAX_PLANES];
+ union dmub_fams2_stream_static_sub_state_v2 fams2_stream_sub_params_v2[DML2_MAX_PLANES];
+ };
struct dml2_display_arb_regs arb_regs;
};
@@ -624,7 +652,7 @@ struct dc_state {
*/
struct bw_context bw_ctx;
- struct block_sequence block_sequence[50];
+ struct block_sequence block_sequence[MAX_HWSS_BLOCK_SEQUENCE_SIZE];
unsigned int block_sequence_steps;
struct dc_dmub_cmd dc_dmub_cmd[10];
unsigned int dmub_cmd_count;
diff --git a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h
index 55529c5f471c..134091d5842d 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/dcn_calcs.h
@@ -622,11 +622,7 @@ extern const struct dcn_ip_params dcn10_ip_defaults;
bool dcn_validate_bandwidth(
struct dc *dc,
struct dc_state *context,
- bool fast_validate);
-
-unsigned int dcn_find_dcfclk_suits_all(
- const struct dc *dc,
- struct dc_clocks *clocks);
+ enum dc_validate_mode validate_mode);
void dcn_get_soc_clks(
struct dc *dc,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h
index 2d06067ff36d..2c9a4a12bd8a 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h
@@ -100,6 +100,17 @@ struct dcn301_clk_internal {
#define MAX_NUM_DPM_LVL 8
#define WM_SET_COUNT 4
+enum clk_type {
+ CLK_TYPE_DCFCLK,
+ CLK_TYPE_FCLK,
+ CLK_TYPE_MCLK,
+ CLK_TYPE_SOCCLK,
+ CLK_TYPE_DTBCLK,
+ CLK_TYPE_DISPCLK,
+ CLK_TYPE_DPPCLK,
+ CLK_TYPE_DSCCLK,
+ CLK_TYPE_COUNT
+};
struct clk_limit_table_entry {
unsigned int voltage; /* milivolts withh 2 fractional bits */
@@ -306,6 +317,9 @@ struct clk_mgr_funcs {
*/
void (*set_hard_min_memclk)(struct clk_mgr *clk_mgr, bool current_mode);
+ int (*get_hard_min_memclk)(struct clk_mgr *clk_mgr);
+ int (*get_hard_min_fclk)(struct clk_mgr *clk_mgr);
+
/* Send message to PMFW to set hard max memclk frequency to highest DPM */
void (*set_hard_max_memclk)(struct clk_mgr *clk_mgr);
@@ -321,6 +335,11 @@ struct clk_mgr_funcs {
int (*get_dispclk_from_dentist)(struct clk_mgr *clk_mgr_base);
+ bool (*is_dc_mode_present)(struct clk_mgr *clk_mgr);
+
+ uint32_t (*set_smartmux_switch)(struct clk_mgr *clk_mgr, uint32_t pins_to_set);
+
+ unsigned int (*get_max_clock_khz)(struct clk_mgr *clk_mgr_base, enum clk_type clk_type);
};
struct clk_mgr {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h
index c2dd061892f4..bac8febad69a 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr_internal.h
@@ -166,6 +166,42 @@ enum dentist_divider_range {
CLK_SR_DCN32(CLK1_CLK4_CURRENT_CNT), \
CLK_SR_DCN32(CLK4_CLK0_CURRENT_CNT)
+#define CLK_REG_LIST_DCN35() \
+ CLK_SR_DCN35(CLK1_CLK_PLL_REQ), \
+ CLK_SR_DCN35(CLK1_CLK0_DFS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK1_DFS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK2_DFS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK3_DFS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK4_DFS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK5_DFS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK0_CURRENT_CNT), \
+ CLK_SR_DCN35(CLK1_CLK1_CURRENT_CNT), \
+ CLK_SR_DCN35(CLK1_CLK2_CURRENT_CNT), \
+ CLK_SR_DCN35(CLK1_CLK3_CURRENT_CNT), \
+ CLK_SR_DCN35(CLK1_CLK4_CURRENT_CNT), \
+ CLK_SR_DCN35(CLK1_CLK5_CURRENT_CNT), \
+ CLK_SR_DCN35(CLK1_CLK0_BYPASS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK1_BYPASS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK2_BYPASS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK3_BYPASS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK4_BYPASS_CNTL),\
+ CLK_SR_DCN35(CLK1_CLK5_BYPASS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK0_DS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK1_DS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK2_DS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK3_DS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK4_DS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK5_DS_CNTL), \
+ CLK_SR_DCN35(CLK1_CLK0_ALLOW_DS), \
+ CLK_SR_DCN35(CLK1_CLK1_ALLOW_DS), \
+ CLK_SR_DCN35(CLK1_CLK2_ALLOW_DS), \
+ CLK_SR_DCN35(CLK1_CLK3_ALLOW_DS), \
+ CLK_SR_DCN35(CLK1_CLK4_ALLOW_DS), \
+ CLK_SR_DCN35(CLK1_CLK5_ALLOW_DS), \
+ CLK_SR_DCN35(CLK5_spll_field_8), \
+ CLK_SR_DCN35(CLK6_spll_field_8), \
+ SR(DENTIST_DISPCLK_CNTL), \
+
#define CLK_COMMON_MASK_SH_LIST_DCN32(mask_sh) \
CLK_COMMON_MASK_SH_LIST_DCN20_BASE(mask_sh),\
CLK_SF(CLK1_CLK_PLL_REQ, FbMult_int, mask_sh),\
@@ -186,6 +222,7 @@ enum dentist_divider_range {
CLK_SF(CLK0_CLK_PLL_REQ, FbMult_frac, mask_sh)
#define CLK_REG_LIST_DCN401() \
+ SR(DENTIST_DISPCLK_CNTL), \
CLK_SR_DCN401(CLK0_CLK_PLL_REQ, CLK01, 0), \
CLK_SR_DCN401(CLK0_CLK0_DFS_CNTL, CLK01, 0), \
CLK_SR_DCN401(CLK0_CLK1_DFS_CNTL, CLK01, 0), \
@@ -236,6 +273,7 @@ struct clk_mgr_registers {
uint32_t CLK1_CLK2_DFS_CNTL;
uint32_t CLK1_CLK3_DFS_CNTL;
uint32_t CLK1_CLK4_DFS_CNTL;
+ uint32_t CLK1_CLK5_DFS_CNTL;
uint32_t CLK2_CLK2_DFS_CNTL;
uint32_t CLK1_CLK0_CURRENT_CNT;
@@ -243,11 +281,34 @@ struct clk_mgr_registers {
uint32_t CLK1_CLK2_CURRENT_CNT;
uint32_t CLK1_CLK3_CURRENT_CNT;
uint32_t CLK1_CLK4_CURRENT_CNT;
+ uint32_t CLK1_CLK5_CURRENT_CNT;
uint32_t CLK0_CLK0_DFS_CNTL;
uint32_t CLK0_CLK1_DFS_CNTL;
uint32_t CLK0_CLK3_DFS_CNTL;
uint32_t CLK0_CLK4_DFS_CNTL;
+ uint32_t CLK1_CLK0_BYPASS_CNTL;
+ uint32_t CLK1_CLK1_BYPASS_CNTL;
+ uint32_t CLK1_CLK2_BYPASS_CNTL;
+ uint32_t CLK1_CLK3_BYPASS_CNTL;
+ uint32_t CLK1_CLK4_BYPASS_CNTL;
+ uint32_t CLK1_CLK5_BYPASS_CNTL;
+
+ uint32_t CLK1_CLK0_DS_CNTL;
+ uint32_t CLK1_CLK1_DS_CNTL;
+ uint32_t CLK1_CLK2_DS_CNTL;
+ uint32_t CLK1_CLK3_DS_CNTL;
+ uint32_t CLK1_CLK4_DS_CNTL;
+ uint32_t CLK1_CLK5_DS_CNTL;
+
+ uint32_t CLK1_CLK0_ALLOW_DS;
+ uint32_t CLK1_CLK1_ALLOW_DS;
+ uint32_t CLK1_CLK2_ALLOW_DS;
+ uint32_t CLK1_CLK3_ALLOW_DS;
+ uint32_t CLK1_CLK4_ALLOW_DS;
+ uint32_t CLK1_CLK5_ALLOW_DS;
+ uint32_t CLK5_spll_field_8;
+ uint32_t CLK6_spll_field_8;
};
struct clk_mgr_shift {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h
index e94e9ba60f55..61c4d2a7db1c 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h
@@ -211,7 +211,7 @@ struct dccg_funcs {
struct dccg *dccg,
enum streamclk_source src,
uint32_t otg_inst);
- void (*set_dto_dscclk)(struct dccg *dccg, uint32_t dsc_inst);
+ void (*set_dto_dscclk)(struct dccg *dccg, uint32_t dsc_inst, uint32_t num_slices_h);
void (*set_ref_dscclk)(struct dccg *dccg, uint32_t dsc_inst);
void (*dccg_root_gate_disable_control)(struct dccg *dccg, uint32_t pipe_idx, uint32_t disable_clock_gating);
};
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
index 0150f2581ee4..0c5675d1c593 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dpp.h
@@ -119,10 +119,14 @@ static const struct dpp_input_csc_matrix __maybe_unused dpp_input_csc_matrix[] =
{ 0x39a6, 0x2568, 0, 0xe0d6,
0xeedd, 0x2568, 0xf925, 0x9a8,
0, 0x2568, 0x43ee, 0xdbb2 } },
- { COLOR_SPACE_2020_YCBCR,
+ { COLOR_SPACE_2020_YCBCR_FULL,
{ 0x2F30, 0x2000, 0, 0xE869,
0xEDB7, 0x2000, 0xFABC, 0xBC6,
0, 0x2000, 0x3C34, 0xE1E6 } },
+ { COLOR_SPACE_2020_YCBCR_LIMITED,
+ { 0x35B9, 0x2543, 0, 0xE2B2,
+ 0xEB2F, 0x2543, 0xFA01, 0x0B1F,
+ 0, 0x2543, 0x4489, 0xDB42 } },
{ COLOR_SPACE_2020_RGB_LIMITEDRANGE,
{ 0x35E0, 0x255F, 0, 0xE2B3,
0xEB20, 0x255F, 0xF9FD, 0xB1E,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
index 16580d624278..cee29e89ec5c 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/hubp.h
@@ -42,6 +42,7 @@
#include "cursor_reg_cache.h"
#include "dml2/dml21/inc/dml_top_dchub_registers.h"
+#include "dml2/dml21/inc/dml_top_types.h"
#define OPP_ID_INVALID 0xf
#define MAX_TTU 0xffffff
@@ -144,14 +145,26 @@ struct hubp_funcs {
struct _vcs_dpi_display_rq_regs_st *rq_regs,
struct _vcs_dpi_display_pipe_dest_params_st *pipe_dest);
+ void (*hubp_setup2)(
+ struct hubp *hubp,
+ struct dml2_dchub_per_pipe_register_set *pipe_regs,
+ union dml2_global_sync_programming *pipe_global_sync,
+ struct dc_crtc_timing *timing);
+
void (*hubp_setup_interdependent)(
struct hubp *hubp,
struct _vcs_dpi_display_dlg_regs_st *dlg_regs,
struct _vcs_dpi_display_ttu_regs_st *ttu_regs);
+ void (*hubp_setup_interdependent2)(
+ struct hubp *hubp,
+ struct dml2_dchub_per_pipe_register_set *pipe_regs);
+
void (*dcc_control)(struct hubp *hubp, bool enable,
enum hubp_ind_block_size blk_size);
+ void (*hubp_reset)(struct hubp *hubp);
+
void (*mem_program_viewport)(
struct hubp *hubp,
const struct rect *viewport,
@@ -165,7 +178,7 @@ struct hubp_funcs {
void (*hubp_program_pte_vm)(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
enum dc_rotation_angle rotation);
void (*hubp_set_vm_system_aperture_settings)(
@@ -179,7 +192,7 @@ struct hubp_funcs {
void (*hubp_program_surface_config)(
struct hubp *hubp,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -269,12 +282,13 @@ struct hubp_funcs {
void (*hubp_enable_3dlut_fl)(struct hubp *hubp, bool enable);
void (*hubp_program_3dlut_fl_addressing_mode)(struct hubp *hubp, enum hubp_3dlut_fl_addressing_mode addr_mode);
void (*hubp_program_3dlut_fl_width)(struct hubp *hubp, enum hubp_3dlut_fl_width width);
- void (*hubp_program_3dlut_fl_tmz_protected)(struct hubp *hubp, bool protection_enabled);
+ void (*hubp_program_3dlut_fl_tmz_protected)(struct hubp *hubp, uint8_t protection_bits);
void (*hubp_program_3dlut_fl_crossbar)(struct hubp *hubp,
enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_y_g,
enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_cb_b,
enum hubp_3dlut_fl_crossbar_bit_slice bit_slice_cr_r);
int (*hubp_get_3dlut_fl_done)(struct hubp *hubp);
+ void (*hubp_clear_tiling)(struct hubp *hubp);
};
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
index af9183f5d69b..08c16ba52a51 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/link_encoder.h
@@ -168,6 +168,14 @@ struct link_encoder_funcs {
struct link_encoder *enc,
enum encoder_type_select sel,
uint32_t hpo_inst);
+ void (*enable_dpia_output)(struct link_encoder *enc,
+ const struct dc_link_settings *link_settings,
+ uint8_t dpia_id,
+ uint8_t digmode,
+ uint8_t fec_rdy);
+ void (*disable_dpia_output)(struct link_encoder *link_enc,
+ uint8_t dpia_id,
+ uint8_t digmode);
};
/*
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
index a8b44f398ce6..42fbc70f7056 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mem_input.h
@@ -150,7 +150,7 @@ struct mem_input_funcs {
void (*mem_input_program_pte_vm)(
struct mem_input *mem_input,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
enum dc_rotation_angle rotation);
void (*mem_input_set_vm_system_aperture_settings)(
@@ -164,7 +164,7 @@ struct mem_input_funcs {
void (*mem_input_program_surface_config)(
struct mem_input *mem_input,
enum surface_pixel_format format,
- union dc_tiling_info *tiling_info,
+ struct dc_tiling_info *tiling_info,
struct plane_size *plane_size,
enum dc_rotation_angle rotation,
struct dc_plane_dcc_param *dcc,
@@ -187,6 +187,8 @@ struct mem_input_funcs {
const struct dc_cursor_position *pos,
const struct dc_cursor_mi_param *param);
+ void (*mem_input_clear_tiling)(
+ struct mem_input *mem_input);
};
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h b/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
index 3a89cc0cffc1..7641439f6ca0 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h
@@ -190,6 +190,42 @@ struct mpc_grph_gamut_adjustment {
enum mpcc_gamut_remap_id mpcc_gamut_remap_block_id;
};
+struct mpc_rmcm_regs {
+ uint32_t rmcm_3dlut_mem_pwr_state;
+ uint32_t rmcm_3dlut_mem_pwr_force;
+ uint32_t rmcm_3dlut_mem_pwr_dis;
+ uint32_t rmcm_3dlut_mem_pwr_mode;
+ uint32_t rmcm_3dlut_size;
+ uint32_t rmcm_3dlut_mode;
+ uint32_t rmcm_3dlut_mode_cur;
+ uint32_t rmcm_3dlut_read_sel;
+ uint32_t rmcm_3dlut_30bit_en;
+ uint32_t rmcm_3dlut_wr_en_mask;
+ uint32_t rmcm_3dlut_ram_sel;
+ uint32_t rmcm_3dlut_out_norm_factor;
+ uint32_t rmcm_3dlut_fl_sel;
+ uint32_t rmcm_3dlut_out_offset_r;
+ uint32_t rmcm_3dlut_out_scale_r;
+ uint32_t rmcm_3dlut_fl_done;
+ uint32_t rmcm_3dlut_fl_soft_underflow;
+ uint32_t rmcm_3dlut_fl_hard_underflow;
+ uint32_t rmcm_cntl;
+ uint32_t rmcm_shaper_mem_pwr_state;
+ uint32_t rmcm_shaper_mem_pwr_force;
+ uint32_t rmcm_shaper_mem_pwr_dis;
+ uint32_t rmcm_shaper_mem_pwr_mode;
+ uint32_t rmcm_shaper_lut_mode;
+ uint32_t rmcm_shaper_mode_cur;
+ uint32_t rmcm_shaper_lut_write_en_mask;
+ uint32_t rmcm_shaper_lut_write_sel;
+ uint32_t rmcm_shaper_offset_b;
+ uint32_t rmcm_shaper_scale_b;
+ uint32_t rmcm_shaper_rama_exp_region_start_b;
+ uint32_t rmcm_shaper_rama_exp_region_start_seg_b;
+ uint32_t rmcm_shaper_rama_exp_region_end_b;
+ uint32_t rmcm_shaper_rama_exp_region_end_base_b;
+};
+
struct mpcc_sm_cfg {
bool enable;
/* 0-single plane,2-row subsampling,4-column subsampling,6-checkboard subsampling */
@@ -301,6 +337,7 @@ struct mpcc_state {
uint32_t rgam_mode;
uint32_t rgam_lut;
struct mpc_grph_gamut_adjustment gamut_remap;
+ struct mpc_rmcm_regs rmcm_regs;
};
/**
@@ -967,23 +1004,6 @@ struct mpc_funcs {
*/
void (*update_3dlut_fast_load_select)(struct mpc *mpc, int mpcc_id, int hubp_idx);
- /**
- * @get_3dlut_fast_load_status:
- *
- * Get 3D LUT fast load status and reference them with done, soft_underflow and hard_underflow pointers.
- *
- * Parameters:
- * - [in/out] mpc - MPC context.
- * - [in] mpcc_id
- * - [in/out] done
- * - [in/out] soft_underflow
- * - [in/out] hard_underflow
- *
- * Return:
- *
- * void
- */
- void (*get_3dlut_fast_load_status)(struct mpc *mpc, int mpcc_id, uint32_t *done, uint32_t *soft_underflow, uint32_t *hard_underflow);
/**
* @populate_lut:
@@ -1054,6 +1074,45 @@ struct mpc_funcs {
* void
*/
void (*program_3dlut_size)(struct mpc *mpc, bool is_17x17x17, int mpcc_id);
+
+ /**
+ * @mcm:
+ *
+ * MPC MCM new HW sequential programming functions
+ */
+ struct {
+ void (*program_3dlut_size)(struct mpc *mpc, uint32_t width, int mpcc_id);
+ void (*program_bias_scale)(struct mpc *mpc, uint16_t bias, uint16_t scale, int mpcc_id);
+ void (*program_bit_depth)(struct mpc *mpc, uint16_t bit_depth, int mpcc_id);
+ bool (*is_config_supported)(uint32_t width);
+ void (*program_lut_read_write_control)(struct mpc *mpc, const enum MCM_LUT_ID id,
+ bool lut_bank_a, bool enabled, int mpcc_id);
+
+ void (*populate_lut)(struct mpc *mpc, const union mcm_lut_params params,
+ bool lut_bank_a, int mpcc_id);
+ } mcm;
+
+ /**
+ * @rmcm:
+ *
+ * MPC RMCM new HW sequential programming functions
+ */
+ struct {
+ void (*enable_3dlut_fl)(struct mpc *mpc, bool enable, int mpcc_id);
+ void (*update_3dlut_fast_load_select)(struct mpc *mpc, int mpcc_id, int hubp_idx);
+ void (*program_lut_read_write_control)(struct mpc *mpc, const enum MCM_LUT_ID id,
+ bool lut_bank_a, bool enabled, int mpcc_id);
+ void (*program_lut_mode)(struct mpc *mpc, const enum MCM_LUT_XABLE xable,
+ bool lut_bank_a, int mpcc_id);
+ void (*program_3dlut_size)(struct mpc *mpc, uint32_t width, int mpcc_id);
+ void (*program_bias_scale)(struct mpc *mpc, uint16_t bias, uint16_t scale, int mpcc_id);
+ void (*program_bit_depth)(struct mpc *mpc, uint16_t bit_depth, int mpcc_id);
+ bool (*is_config_supported)(uint32_t width);
+
+ void (*power_on_shaper_3dlut)(struct mpc *mpc, uint32_t mpcc_id, bool power_on);
+ void (*populate_lut)(struct mpc *mpc, const union mcm_lut_params params,
+ bool lut_bank_a, int mpcc_id);
+ } rmcm;
};
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/optc.h b/drivers/gpu/drm/amd/display/dc/inc/hw/optc.h
index 03cbcbb36f1c..0d5a8358a778 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/optc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/optc.h
@@ -68,37 +68,10 @@ struct optc {
int pstate_keepout;
struct dc_crtc_timing orginal_patched_timing;
enum signal_type signal;
+ uint32_t max_frame_count;
};
-struct dcn_otg_state {
- uint32_t v_blank_start;
- uint32_t v_blank_end;
- uint32_t v_sync_a_pol;
- uint32_t v_total;
- uint32_t v_total_max;
- uint32_t v_total_min;
- uint32_t v_total_min_sel;
- uint32_t v_total_max_sel;
- uint32_t v_sync_a_start;
- uint32_t v_sync_a_end;
- uint32_t h_blank_start;
- uint32_t h_blank_end;
- uint32_t h_sync_a_start;
- uint32_t h_sync_a_end;
- uint32_t h_sync_a_pol;
- uint32_t h_total;
- uint32_t underflow_occurred_status;
- uint32_t otg_enabled;
- uint32_t blank_enabled;
- uint32_t vertical_interrupt1_en;
- uint32_t vertical_interrupt1_line;
- uint32_t vertical_interrupt2_en;
- uint32_t vertical_interrupt2_line;
- uint32_t otg_master_update_lock;
- uint32_t otg_double_buffer_control;
-};
-
-void optc1_read_otg_state(struct optc *optc1, struct dcn_otg_state *s);
+void optc1_read_otg_state(struct timing_generator *optc, struct dcn_otg_state *s);
bool optc1_get_hw_timing(struct timing_generator *tg, struct dc_crtc_timing *hw_crtc_timing);
@@ -210,7 +183,7 @@ void optc1_enable_crtc_reset(struct timing_generator *optc,
bool optc1_configure_crc(struct timing_generator *optc, const struct crc_params *params);
-bool optc1_get_crc(struct timing_generator *optc,
+bool optc1_get_crc(struct timing_generator *optc, uint8_t idx,
uint32_t *r_cr,
uint32_t *g_y,
uint32_t *b_cb);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/pg_cntl.h b/drivers/gpu/drm/amd/display/dc/inc/hw/pg_cntl.h
index 00ea3864dd4d..44f86cc2d1d6 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/pg_cntl.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/pg_cntl.h
@@ -46,6 +46,8 @@ struct pg_cntl_funcs {
void (*opp_pg_control)(struct pg_cntl *pg_cntl, unsigned int opp_inst, bool power_on);
void (*optc_pg_control)(struct pg_cntl *pg_cntl, unsigned int optc_inst, bool power_on);
void (*dwb_pg_control)(struct pg_cntl *pg_cntl, bool power_on);
+ void (*mem_pg_control)(struct pg_cntl *pg_cntl, bool power_on);
+ void (*dio_pg_control)(struct pg_cntl *pg_cntl, bool power_on);
void (*init_pg_status)(struct pg_cntl *pg_cntl);
};
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
index fe7f3137f228..27f950ae45ee 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/stream_encoder.h
@@ -117,6 +117,7 @@ struct stream_encoder {
uint32_t stream_enc_inst;
struct vpg *vpg;
struct afmt *afmt;
+ struct apg *apg;
};
struct enc_state {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
index b74e18cc1e66..267ace4eef8a 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
@@ -141,6 +141,38 @@ struct crc_params {
bool continuous_mode;
bool enable;
+
+ uint8_t crc_eng_inst;
+ bool reset;
+};
+
+struct dcn_otg_state {
+ uint32_t v_blank_start;
+ uint32_t v_blank_end;
+ uint32_t v_sync_a_pol;
+ uint32_t v_total;
+ uint32_t v_total_max;
+ uint32_t v_total_min;
+ uint32_t v_total_min_sel;
+ uint32_t v_total_max_sel;
+ uint32_t v_sync_a_start;
+ uint32_t v_sync_a_end;
+ uint32_t h_blank_start;
+ uint32_t h_blank_end;
+ uint32_t h_sync_a_start;
+ uint32_t h_sync_a_end;
+ uint32_t h_sync_a_pol;
+ uint32_t h_total;
+ uint32_t underflow_occurred_status;
+ uint32_t otg_enabled;
+ uint32_t blank_enabled;
+ uint32_t vertical_interrupt1_en;
+ uint32_t vertical_interrupt1_line;
+ uint32_t vertical_interrupt2_en;
+ uint32_t vertical_interrupt2_line;
+ uint32_t vertical_interrupt2_dest;
+ uint32_t otg_master_update_lock;
+ uint32_t otg_double_buffer_control;
};
/**
@@ -291,7 +323,7 @@ struct timing_generator_funcs {
* @get_crc: Get CRCs for the given timing generator. Return false if
* CRCs are not enabled (via configure_crc).
*/
- bool (*get_crc)(struct timing_generator *tg,
+ bool (*get_crc)(struct timing_generator *tg, uint8_t idx,
uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb);
void (*program_manual_trigger)(struct timing_generator *optc);
@@ -347,6 +379,7 @@ struct timing_generator_funcs {
bool (*get_pipe_update_pending)(struct timing_generator *tg);
void (*set_vupdate_keepout)(struct timing_generator *tg, bool enable);
bool (*wait_update_lock_status)(struct timing_generator *tg, bool locked);
+ void (*read_otg_state)(struct timing_generator *tg, struct dcn_otg_state *s);
};
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h b/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
index 45262cba675e..5a1d9b708a9d 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/transform.h
@@ -29,7 +29,7 @@
#include "hw_shared.h"
#include "dc_hw_types.h"
#include "fixed31_32.h"
-#include "spl/dc_spl_types.h"
+#include "sspl/dc_spl_types.h"
#define CSC_TEMPERATURE_MATRIX_SIZE 12
diff --git a/drivers/gpu/drm/amd/display/dc/inc/link.h b/drivers/gpu/drm/amd/display/dc/inc/link.h
index f04292086c08..f2503402c10e 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/link.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/link.h
@@ -144,9 +144,13 @@ struct link_service {
uint32_t (*dp_link_bandwidth_kbps)(
const struct dc_link *link,
const struct dc_link_settings *link_settings);
- bool (*validate_dpia_bandwidth)(
- const struct dc_stream_state *stream,
- const unsigned int num_streams);
+ enum dc_status (*validate_dp_tunnel_bandwidth)(
+ const struct dc *dc,
+ const struct dc_state *new_ctx);
+
+ uint32_t (*dp_required_hblank_size_bytes)(
+ const struct dc_link *link,
+ struct dp_audio_bandwidth_params *audio_params);
/*************************** DPMS *************************************/
@@ -203,6 +207,9 @@ struct link_service {
bool (*dp_decide_link_settings)(
struct dc_stream_state *stream,
struct dc_link_settings *link_setting);
+ void (*dp_decide_tunnel_settings)(
+ struct dc_stream_state *stream,
+ struct dc_tunnel_settings *dp_tunnel_setting);
enum dp_link_encoding (*mst_decide_link_encoding_format)(
const struct dc_link *link);
bool (*edp_decide_link_settings)(struct dc_link *link,
@@ -214,10 +221,8 @@ struct link_service {
/*************************** DP DPIA/PHY ******************************/
- int (*dpia_handle_usb4_bandwidth_allocation_for_link)(
+ void (*dpia_handle_usb4_bandwidth_allocation_for_link)(
struct dc_link *link, int peak_bw);
- void (*dpia_handle_bw_alloc_response)(
- struct dc_link *link, uint8_t bw, uint8_t result);
void (*dp_set_drive_settings)(
struct dc_link *link,
const struct link_resource *link_res,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/link_enc_cfg.h b/drivers/gpu/drm/amd/display/dc/inc/link_enc_cfg.h
index dc650be3837e..f1afb31ac70b 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/link_enc_cfg.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/link_enc_cfg.h
@@ -96,11 +96,6 @@ struct link_encoder *link_enc_cfg_get_link_enc_used_by_link(
/* Return next available DIG link encoder. NULL if none available. */
struct link_encoder *link_enc_cfg_get_next_avail_link_enc(struct dc *dc);
-/* Return DIG link encoder used by stream. NULL if unused. */
-struct link_encoder *link_enc_cfg_get_link_enc_used_by_stream(
- struct dc *dc,
- const struct dc_stream_state *stream);
-
/* Return DIG link encoder. NULL if unused. */
struct link_encoder *link_enc_cfg_get_link_enc(const struct dc_link *link);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h
index a402df225a76..26cb1459b743 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/reg_helper.h
@@ -508,6 +508,10 @@ uint32_t generic_indirect_reg_update_ex(const struct dc_context *ctx,
initial_val, \
n, __VA_ARGS__)
+#define IX_REG_SET_SYNC(index, init_value, f1, v1) \
+ IX_REG_SET_N_SYNC(index, 1, init_value, \
+ FN(reg, f1), v1)
+
#define IX_REG_SET_2_SYNC(index, init_value, f1, v1, f2, v2) \
IX_REG_SET_N_SYNC(index, 2, init_value, \
FN(reg, f1), v1,\
diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h
index cd1157d225ab..a890f581f4e8 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/resource.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h
@@ -29,10 +29,10 @@
#include "core_status.h"
#include "dal_asic_id.h"
#include "dm_pp_smu.h"
-#include "spl/dc_spl.h"
#define MEMORY_TYPE_MULTIPLIER_CZ 4
#define MEMORY_TYPE_HBM 2
+#define MAX_MCACHES 8
#define IS_PIPE_SYNCD_VALID(pipe) ((((pipe)->pipe_idx_syncd) & 0x80)?1:0)
@@ -66,6 +66,13 @@ struct resource_straps {
uint32_t audio_stream_number;
};
+struct dc_mcache_allocations {
+ int global_mcache_ids_plane0[MAX_MCACHES + 1];
+ int global_mcache_ids_plane1[MAX_MCACHES + 1];
+ int global_mcache_ids_mall_plane0[MAX_MCACHES + 1];
+ int global_mcache_ids_mall_plane1[MAX_MCACHES + 1];
+};
+
struct resource_create_funcs {
void (*read_dce_straps)(
struct dc_context *ctx, struct resource_straps *straps);
@@ -152,6 +159,8 @@ bool resource_attach_surfaces_to_context(
struct dc_state *context,
const struct resource_pool *pool);
+bool resource_can_pipe_disable_cursor(struct pipe_ctx *pipe_ctx);
+
#define FREE_PIPE_INDEX_NOT_FOUND -1
/*
@@ -627,8 +636,6 @@ enum dc_status update_dp_encoder_resources_for_test_harness(const struct dc *dc,
struct dc_state *context,
struct pipe_ctx *pipe_ctx);
-bool check_subvp_sw_cursor_fallback_req(const struct dc *dc, struct dc_stream_state *stream);
-
/* Get hw programming parameters container from pipe context
* @pipe_ctx: pipe context
* @dscl_prog_data: struct to hold programmable hw reg values
@@ -646,4 +653,9 @@ void resource_init_common_dml2_callbacks(struct dc *dc, struct dml2_configuratio
int resource_calculate_det_for_stream(struct dc_state *state, struct pipe_ctx *otg_master);
bool resource_is_hpo_acquired(struct dc_state *context);
+
+struct link_encoder *get_temp_dio_link_enc(
+ const struct resource_context *res_ctx,
+ const struct resource_pool *const pool,
+ const struct dc_link *link);
#endif /* DRIVERS_GPU_DRM_AMD_DC_DEV_DC_INC_RESOURCE_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/irq/Makefile b/drivers/gpu/drm/amd/display/dc/irq/Makefile
index 8ac36bdd4e1e..b5e14d792378 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/irq/Makefile
@@ -182,6 +182,15 @@ AMD_DAL_IRQ_DCN351= $(addprefix $(AMDDALPATH)/dc/irq/dcn351/,$(IRQ_DCN351))
AMD_DISPLAY_FILES += $(AMD_DAL_IRQ_DCN351)
###############################################################################
+# DCN 36
+###############################################################################
+IRQ_DCN36 = irq_service_dcn36.o
+
+AMD_DAL_IRQ_DCN36= $(addprefix $(AMDDALPATH)/dc/irq/dcn36/,$(IRQ_DCN36))
+
+AMD_DISPLAY_FILES += $(AMD_DAL_IRQ_DCN36)
+
+###############################################################################
# DCN 401
###############################################################################
IRQ_DCN401 = irq_service_dcn401.o
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c b/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c
index 953f4a4dacad..33ce470e4c88 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dce120/irq_service_dce120.c
@@ -37,36 +37,9 @@
#include "ivsrcid/ivsrcid_vislands30.h"
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dce60/irq_service_dce60.c b/drivers/gpu/drm/amd/display/dc/irq/dce60/irq_service_dce60.c
index 2c72074310c7..d777b85e70da 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dce60/irq_service_dce60.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dce60/irq_service_dce60.c
@@ -46,36 +46,9 @@
#include "dc_types.h"
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- DC_HPD1_INT_STATUS,
- DC_HPD1_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- DC_HPD1_INT_CONTROL,
- DC_HPD1_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd1_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
@@ -391,5 +364,3 @@ struct irq_service *dal_irq_service_dce60_create(
dce60_irq_construct(irq_service, init_data);
return irq_service;
}
-
-
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dce80/irq_service_dce80.c b/drivers/gpu/drm/amd/display/dc/irq/dce80/irq_service_dce80.c
index 49317934ef4f..3a9163acb49b 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dce80/irq_service_dce80.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dce80/irq_service_dce80.c
@@ -37,36 +37,9 @@
#include "dc_types.h"
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- DC_HPD1_INT_STATUS,
- DC_HPD1_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- DC_HPD1_INT_CONTROL,
- DC_HPD1_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd1_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
@@ -303,5 +276,3 @@ struct irq_service *dal_irq_service_dce80_create(
dce80_irq_construct(irq_service, init_data);
return irq_service;
}
-
-
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c b/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c
index 9ca28565a9d1..4ce9edd16344 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn10/irq_service_dcn10.c
@@ -129,36 +129,9 @@ static enum dc_irq_source to_dal_irq_source_dcn10(struct irq_service *irq_servic
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn20/irq_service_dcn20.c b/drivers/gpu/drm/amd/display/dc/irq/dcn20/irq_service_dcn20.c
index 916f0c974637..5847af0e66cb 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn20/irq_service_dcn20.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn20/irq_service_dcn20.c
@@ -130,36 +130,9 @@ static enum dc_irq_source to_dal_irq_source_dcn20(
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn201/irq_service_dcn201.c b/drivers/gpu/drm/amd/display/dc/irq/dcn201/irq_service_dcn201.c
index 4fb9cd6708d5..6417011d2246 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn201/irq_service_dcn201.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn201/irq_service_dcn201.c
@@ -30,8 +30,8 @@
#include "../dce110/irq_service_dce110.h"
#include "irq_service_dcn201.h"
-#include "dcn/dcn_2_0_3_offset.h"
-#include "dcn/dcn_2_0_3_sh_mask.h"
+#include "dcn/dcn_2_0_1_offset.h"
+#include "dcn/dcn_2_0_1_sh_mask.h"
#include "cyan_skillfish_ip_offset.h"
#include "soc15_hw_ip.h"
@@ -80,36 +80,9 @@ static enum dc_irq_source to_dal_irq_source_dcn201(
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn21/irq_service_dcn21.c b/drivers/gpu/drm/amd/display/dc/irq/dcn21/irq_service_dcn21.c
index 42cdfe6c3538..71d2f065140b 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn21/irq_service_dcn21.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn21/irq_service_dcn21.c
@@ -132,36 +132,9 @@ static enum dc_irq_source to_dal_irq_source_dcn21(struct irq_service *irq_servic
return DC_IRQ_SOURCE_INVALID;
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn30/irq_service_dcn30.c b/drivers/gpu/drm/amd/display/dc/irq/dcn30/irq_service_dcn30.c
index a443a8abb1ea..2a4080bdcf6b 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn30/irq_service_dcn30.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn30/irq_service_dcn30.c
@@ -139,36 +139,9 @@ static enum dc_irq_source to_dal_irq_source_dcn30(
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
@@ -447,4 +420,3 @@ struct irq_service *dal_irq_service_dcn30_create(
dcn30_irq_construct(irq_service, init_data);
return irq_service;
}
-
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn302/irq_service_dcn302.c b/drivers/gpu/drm/amd/display/dc/irq/dcn302/irq_service_dcn302.c
index 8ffc7e2c681a..624f1ac309f8 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn302/irq_service_dcn302.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn302/irq_service_dcn302.c
@@ -126,26 +126,9 @@ static enum dc_irq_source to_dal_irq_source_dcn302(struct irq_service *irq_servi
}
}
-static bool hpd_ack(struct irq_service *irq_service, const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status = get_reg_field_value(value, HPD0_DC_HPD_INT_STATUS, DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(value, current_status ? 0 : 1, HPD0_DC_HPD_INT_CONTROL, DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn303/irq_service_dcn303.c b/drivers/gpu/drm/amd/display/dc/irq/dcn303/irq_service_dcn303.c
index 262bb8b74b15..137caffae916 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn303/irq_service_dcn303.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn303/irq_service_dcn303.c
@@ -77,26 +77,9 @@ static enum dc_irq_source to_dal_irq_source_dcn303(struct irq_service *irq_servi
}
}
-static bool hpd_ack(struct irq_service *irq_service, const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status = get_reg_field_value(value, HPD0_DC_HPD_INT_STATUS, DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(value, current_status ? 0 : 1, HPD0_DC_HPD_INT_CONTROL, DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn31/irq_service_dcn31.c b/drivers/gpu/drm/amd/display/dc/irq/dcn31/irq_service_dcn31.c
index 53e78ae7eecf..921cb167d920 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn31/irq_service_dcn31.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn31/irq_service_dcn31.c
@@ -128,36 +128,9 @@ static enum dc_irq_source to_dal_irq_source_dcn31(struct irq_service *irq_servic
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn314/irq_service_dcn314.c b/drivers/gpu/drm/amd/display/dc/irq/dcn314/irq_service_dcn314.c
index e0563e880432..0118fd6e5db0 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn314/irq_service_dcn314.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn314/irq_service_dcn314.c
@@ -130,36 +130,9 @@ static enum dc_irq_source to_dal_irq_source_dcn314(struct irq_service *irq_servi
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn315/irq_service_dcn315.c b/drivers/gpu/drm/amd/display/dc/irq/dcn315/irq_service_dcn315.c
index 2ef22299101a..adebfc888618 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn315/irq_service_dcn315.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn315/irq_service_dcn315.c
@@ -135,36 +135,9 @@ static enum dc_irq_source to_dal_irq_source_dcn315(
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn32/irq_service_dcn32.c b/drivers/gpu/drm/amd/display/dc/irq/dcn32/irq_service_dcn32.c
index f0ac0aeeac51..e9e315c75d76 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn32/irq_service_dcn32.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn32/irq_service_dcn32.c
@@ -129,36 +129,9 @@ static enum dc_irq_source to_dal_irq_source_dcn32(
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
@@ -191,6 +164,16 @@ static struct irq_source_info_funcs vline0_irq_info_funcs = {
.ack = NULL
};
+static struct irq_source_info_funcs vline1_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
+static struct irq_source_info_funcs vline2_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
#undef BASE_INNER
#define BASE_INNER(seg) DCN_BASE__INST0_SEG ## seg
@@ -259,6 +242,13 @@ static struct irq_source_info_funcs vline0_irq_info_funcs = {
.funcs = &pflip_irq_info_funcs\
}
+#define vblank_int_entry(reg_num)\
+ [DC_IRQ_SOURCE_VBLANK1 + reg_num] = {\
+ IRQ_REG_ENTRY(OTG, reg_num,\
+ OTG_GLOBAL_SYNC_STATUS, VSTARTUP_INT_EN,\
+ OTG_GLOBAL_SYNC_STATUS, VSTARTUP_EVENT_CLEAR),\
+ .funcs = &vblank_irq_info_funcs\
+ }
/* vupdate_no_lock_int_entry maps to DC_IRQ_SOURCE_VUPDATEx, to match semantic
* of DCE's DC_IRQ_SOURCE_VUPDATEx.
*/
@@ -270,14 +260,6 @@ static struct irq_source_info_funcs vline0_irq_info_funcs = {
.funcs = &vupdate_no_lock_irq_info_funcs\
}
-#define vblank_int_entry(reg_num)\
- [DC_IRQ_SOURCE_VBLANK1 + reg_num] = {\
- IRQ_REG_ENTRY(OTG, reg_num,\
- OTG_GLOBAL_SYNC_STATUS, VSTARTUP_INT_EN,\
- OTG_GLOBAL_SYNC_STATUS, VSTARTUP_EVENT_CLEAR),\
- .funcs = &vblank_irq_info_funcs\
-}
-
#define vline0_int_entry(reg_num)\
[DC_IRQ_SOURCE_DC1_VLINE0 + reg_num] = {\
IRQ_REG_ENTRY(OTG, reg_num,\
@@ -285,6 +267,20 @@ static struct irq_source_info_funcs vline0_irq_info_funcs = {
OTG_VERTICAL_INTERRUPT0_CONTROL, OTG_VERTICAL_INTERRUPT0_CLEAR),\
.funcs = &vline0_irq_info_funcs\
}
+#define vline1_int_entry(reg_num)\
+ [DC_IRQ_SOURCE_DC1_VLINE1 + reg_num] = {\
+ IRQ_REG_ENTRY(OTG, reg_num,\
+ OTG_VERTICAL_INTERRUPT1_CONTROL, OTG_VERTICAL_INTERRUPT1_INT_ENABLE,\
+ OTG_VERTICAL_INTERRUPT1_CONTROL, OTG_VERTICAL_INTERRUPT1_CLEAR),\
+ .funcs = &vline1_irq_info_funcs\
+ }
+#define vline2_int_entry(reg_num)\
+ [DC_IRQ_SOURCE_DC1_VLINE2 + reg_num] = {\
+ IRQ_REG_ENTRY(OTG, reg_num,\
+ OTG_VERTICAL_INTERRUPT2_CONTROL, OTG_VERTICAL_INTERRUPT2_INT_ENABLE,\
+ OTG_VERTICAL_INTERRUPT2_CONTROL, OTG_VERTICAL_INTERRUPT2_CLEAR),\
+ .funcs = &vline2_irq_info_funcs\
+ }
#define dmub_outbox_int_entry()\
[DC_IRQ_SOURCE_DMCUB_OUTBOX] = {\
IRQ_REG_ENTRY_DMUB(\
@@ -387,21 +383,29 @@ irq_source_info_dcn32[DAL_IRQ_SOURCES_NUMBER] = {
dc_underflow_int_entry(6),
[DC_IRQ_SOURCE_DMCU_SCP] = dummy_irq_entry(),
[DC_IRQ_SOURCE_VBIOS_SW] = dummy_irq_entry(),
- vupdate_no_lock_int_entry(0),
- vupdate_no_lock_int_entry(1),
- vupdate_no_lock_int_entry(2),
- vupdate_no_lock_int_entry(3),
vblank_int_entry(0),
vblank_int_entry(1),
vblank_int_entry(2),
vblank_int_entry(3),
+ [DC_IRQ_SOURCE_DC5_VLINE1] = dummy_irq_entry(),
+ [DC_IRQ_SOURCE_DC6_VLINE1] = dummy_irq_entry(),
+ dmub_outbox_int_entry(),
+ vupdate_no_lock_int_entry(0),
+ vupdate_no_lock_int_entry(1),
+ vupdate_no_lock_int_entry(2),
+ vupdate_no_lock_int_entry(3),
vline0_int_entry(0),
vline0_int_entry(1),
vline0_int_entry(2),
vline0_int_entry(3),
- [DC_IRQ_SOURCE_DC5_VLINE1] = dummy_irq_entry(),
- [DC_IRQ_SOURCE_DC6_VLINE1] = dummy_irq_entry(),
- dmub_outbox_int_entry(),
+ vline1_int_entry(0),
+ vline1_int_entry(1),
+ vline1_int_entry(2),
+ vline1_int_entry(3),
+ vline2_int_entry(0),
+ vline2_int_entry(1),
+ vline2_int_entry(2),
+ vline2_int_entry(3)
};
static const struct irq_service_funcs irq_service_funcs_dcn32 = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn35/irq_service_dcn35.c b/drivers/gpu/drm/amd/display/dc/irq/dcn35/irq_service_dcn35.c
index ea8c271171bc..79e5e8c137ca 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn35/irq_service_dcn35.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn35/irq_service_dcn35.c
@@ -127,36 +127,9 @@ static enum dc_irq_source to_dal_irq_source_dcn35(
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn351/irq_service_dcn351.c b/drivers/gpu/drm/amd/display/dc/irq/dcn351/irq_service_dcn351.c
index 7ec8e0de2f01..163b8ee9ebf7 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn351/irq_service_dcn351.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn351/irq_service_dcn351.c
@@ -106,36 +106,9 @@ static enum dc_irq_source to_dal_irq_source_dcn351(
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn36/irq_service_dcn36.c b/drivers/gpu/drm/amd/display/dc/irq/dcn36/irq_service_dcn36.c
new file mode 100644
index 000000000000..f716ab0fd30e
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn36/irq_service_dcn36.c
@@ -0,0 +1,381 @@
+/* SPDX-License-Identifier: MIT */
+/* Copyright 2025 Advanced Micro Devices, Inc. */
+
+#include "dm_services.h"
+#include "include/logger_interface.h"
+#include "../dce110/irq_service_dce110.h"
+
+#include "dcn/dcn_3_6_0_offset.h"
+#include "dcn/dcn_3_6_0_sh_mask.h"
+
+#include "irq_service_dcn36.h"
+
+#include "ivsrcid/dcn/irqsrcs_dcn_1_0.h"
+
+static enum dc_irq_source to_dal_irq_source_dcn36(
+ struct irq_service *irq_service,
+ uint32_t src_id,
+ uint32_t ext_id)
+{
+ switch (src_id) {
+ case DCN_1_0__SRCID__DC_D1_OTG_VSTARTUP:
+ return DC_IRQ_SOURCE_VBLANK1;
+ case DCN_1_0__SRCID__DC_D2_OTG_VSTARTUP:
+ return DC_IRQ_SOURCE_VBLANK2;
+ case DCN_1_0__SRCID__DC_D3_OTG_VSTARTUP:
+ return DC_IRQ_SOURCE_VBLANK3;
+ case DCN_1_0__SRCID__DC_D4_OTG_VSTARTUP:
+ return DC_IRQ_SOURCE_VBLANK4;
+ case DCN_1_0__SRCID__DC_D5_OTG_VSTARTUP:
+ return DC_IRQ_SOURCE_VBLANK5;
+ case DCN_1_0__SRCID__DC_D6_OTG_VSTARTUP:
+ return DC_IRQ_SOURCE_VBLANK6;
+ case DCN_1_0__SRCID__OTG1_VERTICAL_INTERRUPT0_CONTROL:
+ return DC_IRQ_SOURCE_DC1_VLINE0;
+ case DCN_1_0__SRCID__OTG2_VERTICAL_INTERRUPT0_CONTROL:
+ return DC_IRQ_SOURCE_DC2_VLINE0;
+ case DCN_1_0__SRCID__OTG3_VERTICAL_INTERRUPT0_CONTROL:
+ return DC_IRQ_SOURCE_DC3_VLINE0;
+ case DCN_1_0__SRCID__OTG4_VERTICAL_INTERRUPT0_CONTROL:
+ return DC_IRQ_SOURCE_DC4_VLINE0;
+ case DCN_1_0__SRCID__OTG5_VERTICAL_INTERRUPT0_CONTROL:
+ return DC_IRQ_SOURCE_DC5_VLINE0;
+ case DCN_1_0__SRCID__OTG6_VERTICAL_INTERRUPT0_CONTROL:
+ return DC_IRQ_SOURCE_DC6_VLINE0;
+ case DCN_1_0__SRCID__HUBP0_FLIP_INTERRUPT:
+ return DC_IRQ_SOURCE_PFLIP1;
+ case DCN_1_0__SRCID__HUBP1_FLIP_INTERRUPT:
+ return DC_IRQ_SOURCE_PFLIP2;
+ case DCN_1_0__SRCID__HUBP2_FLIP_INTERRUPT:
+ return DC_IRQ_SOURCE_PFLIP3;
+ case DCN_1_0__SRCID__HUBP3_FLIP_INTERRUPT:
+ return DC_IRQ_SOURCE_PFLIP4;
+ case DCN_1_0__SRCID__HUBP4_FLIP_INTERRUPT:
+ return DC_IRQ_SOURCE_PFLIP5;
+ case DCN_1_0__SRCID__HUBP5_FLIP_INTERRUPT:
+ return DC_IRQ_SOURCE_PFLIP6;
+ case DCN_1_0__SRCID__OTG0_IHC_V_UPDATE_NO_LOCK_INTERRUPT:
+ return DC_IRQ_SOURCE_VUPDATE1;
+ case DCN_1_0__SRCID__OTG1_IHC_V_UPDATE_NO_LOCK_INTERRUPT:
+ return DC_IRQ_SOURCE_VUPDATE2;
+ case DCN_1_0__SRCID__OTG2_IHC_V_UPDATE_NO_LOCK_INTERRUPT:
+ return DC_IRQ_SOURCE_VUPDATE3;
+ case DCN_1_0__SRCID__OTG3_IHC_V_UPDATE_NO_LOCK_INTERRUPT:
+ return DC_IRQ_SOURCE_VUPDATE4;
+ case DCN_1_0__SRCID__OTG4_IHC_V_UPDATE_NO_LOCK_INTERRUPT:
+ return DC_IRQ_SOURCE_VUPDATE5;
+ case DCN_1_0__SRCID__OTG5_IHC_V_UPDATE_NO_LOCK_INTERRUPT:
+ return DC_IRQ_SOURCE_VUPDATE6;
+ case DCN_1_0__SRCID__DMCUB_OUTBOX_LOW_PRIORITY_READY_INT:
+ return DC_IRQ_SOURCE_DMCUB_OUTBOX;
+ case DCN_1_0__SRCID__DC_HPD1_INT:
+ /* generic src_id for all HPD and HPDRX interrupts */
+ switch (ext_id) {
+ case DCN_1_0__CTXID__DC_HPD1_INT:
+ return DC_IRQ_SOURCE_HPD1;
+ case DCN_1_0__CTXID__DC_HPD2_INT:
+ return DC_IRQ_SOURCE_HPD2;
+ case DCN_1_0__CTXID__DC_HPD3_INT:
+ return DC_IRQ_SOURCE_HPD3;
+ case DCN_1_0__CTXID__DC_HPD4_INT:
+ return DC_IRQ_SOURCE_HPD4;
+ case DCN_1_0__CTXID__DC_HPD5_INT:
+ return DC_IRQ_SOURCE_HPD5;
+ case DCN_1_0__CTXID__DC_HPD6_INT:
+ return DC_IRQ_SOURCE_HPD6;
+ case DCN_1_0__CTXID__DC_HPD1_RX_INT:
+ return DC_IRQ_SOURCE_HPD1RX;
+ case DCN_1_0__CTXID__DC_HPD2_RX_INT:
+ return DC_IRQ_SOURCE_HPD2RX;
+ case DCN_1_0__CTXID__DC_HPD3_RX_INT:
+ return DC_IRQ_SOURCE_HPD3RX;
+ case DCN_1_0__CTXID__DC_HPD4_RX_INT:
+ return DC_IRQ_SOURCE_HPD4RX;
+ case DCN_1_0__CTXID__DC_HPD5_RX_INT:
+ return DC_IRQ_SOURCE_HPD5RX;
+ case DCN_1_0__CTXID__DC_HPD6_RX_INT:
+ return DC_IRQ_SOURCE_HPD6RX;
+ default:
+ return DC_IRQ_SOURCE_INVALID;
+ }
+ break;
+
+ default:
+ return DC_IRQ_SOURCE_INVALID;
+ }
+}
+
+static struct irq_source_info_funcs hpd_irq_info_funcs = {
+ .set = NULL,
+ .ack = hpd0_ack
+};
+
+static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
+static struct irq_source_info_funcs pflip_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
+static struct irq_source_info_funcs vupdate_no_lock_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
+static struct irq_source_info_funcs vblank_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
+static struct irq_source_info_funcs outbox_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
+static struct irq_source_info_funcs vline0_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
+#undef BASE_INNER
+#define BASE_INNER(seg) ctx->dcn_reg_offsets[seg]
+
+/* compile time expand base address. */
+#define BASE(seg) \
+ BASE_INNER(seg)
+
+#define SRI(reg_name, block, id)\
+ BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define SRI_DMUB(reg_name)\
+ BASE(reg ## reg_name ## _BASE_IDX) + \
+ reg ## reg_name
+
+#define IRQ_REG_ENTRY(base, block, reg_num, reg1, mask1, reg2, mask2)\
+ REG_STRUCT[base + reg_num].enable_reg = SRI(reg1, block, reg_num),\
+ REG_STRUCT[base + reg_num].enable_mask = \
+ block ## reg_num ## _ ## reg1 ## __ ## mask1 ## _MASK,\
+ REG_STRUCT[base + reg_num].enable_value[0] = \
+ block ## reg_num ## _ ## reg1 ## __ ## mask1 ## _MASK,\
+ REG_STRUCT[base + reg_num].enable_value[1] = \
+ ~block ## reg_num ## _ ## reg1 ## __ ## mask1 ## _MASK, \
+ REG_STRUCT[base + reg_num].ack_reg = SRI(reg2, block, reg_num),\
+ REG_STRUCT[base + reg_num].ack_mask = \
+ block ## reg_num ## _ ## reg2 ## __ ## mask2 ## _MASK,\
+ REG_STRUCT[base + reg_num].ack_value = \
+ block ## reg_num ## _ ## reg2 ## __ ## mask2 ## _MASK \
+
+#define IRQ_REG_ENTRY_DMUB(base, reg1, mask1, reg2, mask2)\
+ REG_STRUCT[base].enable_reg = SRI_DMUB(reg1),\
+ REG_STRUCT[base].enable_mask = \
+ reg1 ## __ ## mask1 ## _MASK,\
+ REG_STRUCT[base].enable_value[0] = \
+ reg1 ## __ ## mask1 ## _MASK,\
+ REG_STRUCT[base].enable_value[1] = \
+ ~reg1 ## __ ## mask1 ## _MASK, \
+ REG_STRUCT[base].ack_reg = SRI_DMUB(reg2),\
+ REG_STRUCT[base].ack_mask = \
+ reg2 ## __ ## mask2 ## _MASK,\
+ REG_STRUCT[base].ack_value = \
+ reg2 ## __ ## mask2 ## _MASK \
+
+#define hpd_int_entry(reg_num)\
+ IRQ_REG_ENTRY(DC_IRQ_SOURCE_HPD1, HPD, reg_num,\
+ DC_HPD_INT_CONTROL, DC_HPD_INT_EN,\
+ DC_HPD_INT_CONTROL, DC_HPD_INT_ACK),\
+ REG_STRUCT[DC_IRQ_SOURCE_HPD1 + reg_num].funcs = &hpd_irq_info_funcs;\
+ REG_STRUCT[DC_IRQ_SOURCE_HPD1 + reg_num].status_reg = SRI(DC_HPD_INT_STATUS, HPD, reg_num);\
+
+#define hpd_rx_int_entry(reg_num)\
+ IRQ_REG_ENTRY(DC_IRQ_SOURCE_HPD1RX, HPD, reg_num,\
+ DC_HPD_INT_CONTROL, DC_HPD_RX_INT_EN,\
+ DC_HPD_INT_CONTROL, DC_HPD_RX_INT_ACK),\
+ REG_STRUCT[DC_IRQ_SOURCE_HPD1RX + reg_num].status_reg = SRI(DC_HPD_INT_STATUS, HPD, reg_num);\
+ REG_STRUCT[DC_IRQ_SOURCE_HPD1RX + reg_num].funcs = &hpd_rx_irq_info_funcs;\
+
+#define pflip_int_entry(reg_num)\
+ IRQ_REG_ENTRY(DC_IRQ_SOURCE_PFLIP1, HUBPREQ, reg_num,\
+ DCSURF_SURFACE_FLIP_INTERRUPT, SURFACE_FLIP_INT_MASK,\
+ DCSURF_SURFACE_FLIP_INTERRUPT, SURFACE_FLIP_CLEAR),\
+ REG_STRUCT[DC_IRQ_SOURCE_PFLIP1 + reg_num].funcs = &pflip_irq_info_funcs\
+
+/* vupdate_no_lock_int_entry maps to DC_IRQ_SOURCE_VUPDATEx, to match semantic
+ * of DCE's DC_IRQ_SOURCE_VUPDATEx.
+ */
+#define vupdate_no_lock_int_entry(reg_num)\
+ IRQ_REG_ENTRY(DC_IRQ_SOURCE_VUPDATE1, OTG, reg_num,\
+ OTG_GLOBAL_SYNC_STATUS, VUPDATE_NO_LOCK_INT_EN,\
+ OTG_GLOBAL_SYNC_STATUS, VUPDATE_NO_LOCK_EVENT_CLEAR),\
+ REG_STRUCT[DC_IRQ_SOURCE_VUPDATE1 + reg_num].funcs = &vupdate_no_lock_irq_info_funcs\
+
+#define vblank_int_entry(reg_num)\
+ IRQ_REG_ENTRY(DC_IRQ_SOURCE_VBLANK1, OTG, reg_num,\
+ OTG_GLOBAL_SYNC_STATUS, VSTARTUP_INT_EN,\
+ OTG_GLOBAL_SYNC_STATUS, VSTARTUP_EVENT_CLEAR),\
+ REG_STRUCT[DC_IRQ_SOURCE_VBLANK1 + reg_num].funcs = &vblank_irq_info_funcs\
+
+#define vline0_int_entry(reg_num)\
+ IRQ_REG_ENTRY(DC_IRQ_SOURCE_DC1_VLINE0, OTG, reg_num,\
+ OTG_VERTICAL_INTERRUPT0_CONTROL, OTG_VERTICAL_INTERRUPT0_INT_ENABLE,\
+ OTG_VERTICAL_INTERRUPT0_CONTROL, OTG_VERTICAL_INTERRUPT0_CLEAR),\
+ REG_STRUCT[DC_IRQ_SOURCE_DC1_VLINE0 + reg_num].funcs = &vline0_irq_info_funcs\
+
+#define dmub_outbox_int_entry()\
+ IRQ_REG_ENTRY_DMUB(DC_IRQ_SOURCE_DMCUB_OUTBOX, \
+ DMCUB_INTERRUPT_ENABLE, DMCUB_OUTBOX1_READY_INT_EN,\
+ DMCUB_INTERRUPT_ACK, DMCUB_OUTBOX1_READY_INT_ACK),\
+ REG_STRUCT[DC_IRQ_SOURCE_DMCUB_OUTBOX].funcs = &outbox_irq_info_funcs
+
+#define dummy_irq_entry(irqno) \
+ REG_STRUCT[irqno].funcs = &dummy_irq_info_funcs\
+
+#define i2c_int_entry(reg_num) \
+ dummy_irq_entry(DC_IRQ_SOURCE_I2C_DDC ## reg_num)
+
+#define dp_sink_int_entry(reg_num) \
+ dummy_irq_entry(DC_IRQ_SOURCE_DPSINK ## reg_num)
+
+#define gpio_pad_int_entry(reg_num) \
+ dummy_irq_entry(DC_IRQ_SOURCE_GPIOPAD ## reg_num)
+
+#define dc_underflow_int_entry(reg_num) \
+ dummy_irq_entry(DC_IRQ_SOURCE_DC ## reg_num ## UNDERFLOW)
+
+static struct irq_source_info_funcs dummy_irq_info_funcs = {
+ .set = dal_irq_service_dummy_set,
+ .ack = dal_irq_service_dummy_ack
+};
+
+#define dcn36_irq_init_part_1() {\
+ dummy_irq_entry(DC_IRQ_SOURCE_INVALID); \
+ hpd_int_entry(0); \
+ hpd_int_entry(1); \
+ hpd_int_entry(2); \
+ hpd_int_entry(3); \
+ hpd_int_entry(4); \
+ hpd_rx_int_entry(0); \
+ hpd_rx_int_entry(1); \
+ hpd_rx_int_entry(2); \
+ hpd_rx_int_entry(3); \
+ hpd_rx_int_entry(4); \
+ i2c_int_entry(1); \
+ i2c_int_entry(2); \
+ i2c_int_entry(3); \
+ i2c_int_entry(4); \
+ i2c_int_entry(5); \
+ i2c_int_entry(6); \
+ dp_sink_int_entry(1); \
+ dp_sink_int_entry(2); \
+ dp_sink_int_entry(3); \
+ dp_sink_int_entry(4); \
+ dp_sink_int_entry(5); \
+ dp_sink_int_entry(6); \
+ dummy_irq_entry(DC_IRQ_SOURCE_TIMER); \
+ pflip_int_entry(0); \
+ pflip_int_entry(1); \
+ pflip_int_entry(2); \
+ pflip_int_entry(3); \
+ dummy_irq_entry(DC_IRQ_SOURCE_PFLIP5); \
+ dummy_irq_entry(DC_IRQ_SOURCE_PFLIP6); \
+ dummy_irq_entry(DC_IRQ_SOURCE_PFLIP_UNDERLAY0); \
+ gpio_pad_int_entry(0); \
+ gpio_pad_int_entry(1); \
+ gpio_pad_int_entry(2); \
+ gpio_pad_int_entry(3); \
+ gpio_pad_int_entry(4); \
+ gpio_pad_int_entry(5); \
+ gpio_pad_int_entry(6); \
+ gpio_pad_int_entry(7); \
+ gpio_pad_int_entry(8); \
+ gpio_pad_int_entry(9); \
+ gpio_pad_int_entry(10); \
+ gpio_pad_int_entry(11); \
+ gpio_pad_int_entry(12); \
+ gpio_pad_int_entry(13); \
+ gpio_pad_int_entry(14); \
+ gpio_pad_int_entry(15); \
+ gpio_pad_int_entry(16); \
+ gpio_pad_int_entry(17); \
+ gpio_pad_int_entry(18); \
+ gpio_pad_int_entry(19); \
+ gpio_pad_int_entry(20); \
+ gpio_pad_int_entry(21); \
+ gpio_pad_int_entry(22); \
+ gpio_pad_int_entry(23); \
+ gpio_pad_int_entry(24); \
+ gpio_pad_int_entry(25); \
+ gpio_pad_int_entry(26); \
+ gpio_pad_int_entry(27); \
+ gpio_pad_int_entry(28); \
+ gpio_pad_int_entry(29); \
+ gpio_pad_int_entry(30); \
+ dc_underflow_int_entry(1); \
+ dc_underflow_int_entry(2); \
+ dc_underflow_int_entry(3); \
+ dc_underflow_int_entry(4); \
+ dc_underflow_int_entry(5); \
+ dc_underflow_int_entry(6); \
+ dummy_irq_entry(DC_IRQ_SOURCE_DMCU_SCP); \
+ dummy_irq_entry(DC_IRQ_SOURCE_VBIOS_SW); \
+}
+
+#define dcn36_irq_init_part_2() {\
+ vupdate_no_lock_int_entry(0); \
+ vupdate_no_lock_int_entry(1); \
+ vupdate_no_lock_int_entry(2); \
+ vupdate_no_lock_int_entry(3); \
+ vblank_int_entry(0); \
+ vblank_int_entry(1); \
+ vblank_int_entry(2); \
+ vblank_int_entry(3); \
+ vline0_int_entry(0); \
+ vline0_int_entry(1); \
+ vline0_int_entry(2); \
+ vline0_int_entry(3); \
+ dummy_irq_entry(DC_IRQ_SOURCE_DC5_VLINE1); \
+ dummy_irq_entry(DC_IRQ_SOURCE_DC6_VLINE1); \
+ dmub_outbox_int_entry(); \
+}
+
+#define dcn36_irq_init() {\
+ dcn36_irq_init_part_1(); \
+ dcn36_irq_init_part_2(); \
+}
+
+static struct irq_source_info irq_source_info_dcn36[DAL_IRQ_SOURCES_NUMBER] = {0};
+
+static struct irq_service_funcs irq_service_funcs_dcn36 = {
+ .to_dal_irq_source = to_dal_irq_source_dcn36
+};
+
+static void dcn36_irq_construct(
+ struct irq_service *irq_service,
+ struct irq_service_init_data *init_data)
+{
+ struct dc_context *ctx = init_data->ctx;
+
+#define REG_STRUCT irq_source_info_dcn36
+ dcn36_irq_init();
+
+ dal_irq_service_construct(irq_service, init_data);
+
+ irq_service->info = irq_source_info_dcn36;
+ irq_service->funcs = &irq_service_funcs_dcn36;
+}
+
+struct irq_service *dal_irq_service_dcn36_create(
+ struct irq_service_init_data *init_data)
+{
+ struct irq_service *irq_service = kzalloc(sizeof(*irq_service),
+ GFP_KERNEL);
+
+ if (!irq_service)
+ return NULL;
+
+ dcn36_irq_construct(irq_service, init_data);
+ return irq_service;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn36/irq_service_dcn36.h b/drivers/gpu/drm/amd/display/dc/irq/dcn36/irq_service_dcn36.h
new file mode 100644
index 000000000000..21ff95f6562d
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn36/irq_service_dcn36.h
@@ -0,0 +1,12 @@
+/* SPDX-License-Identifier: MIT */
+/* Copyright 2025 Advanced Micro Devices, Inc. */
+
+#ifndef __DAL_IRQ_SERVICE_DCN36_H__
+#define __DAL_IRQ_SERVICE_DCN36_H__
+
+#include "../irq_service.h"
+
+struct irq_service *dal_irq_service_dcn36_create(
+ struct irq_service_init_data *init_data);
+
+#endif /* __DAL_IRQ_SERVICE_DCN36_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/irq/dcn401/irq_service_dcn401.c b/drivers/gpu/drm/amd/display/dc/irq/dcn401/irq_service_dcn401.c
index b43c9524b0de..fd9bb1950c20 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/dcn401/irq_service_dcn401.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/dcn401/irq_service_dcn401.c
@@ -109,36 +109,9 @@ static enum dc_irq_source to_dal_irq_source_dcn401(
}
}
-static bool hpd_ack(
- struct irq_service *irq_service,
- const struct irq_source_info *info)
-{
- uint32_t addr = info->status_reg;
- uint32_t value = dm_read_reg(irq_service->ctx, addr);
- uint32_t current_status =
- get_reg_field_value(
- value,
- HPD0_DC_HPD_INT_STATUS,
- DC_HPD_SENSE_DELAYED);
-
- dal_irq_service_ack_generic(irq_service, info);
-
- value = dm_read_reg(irq_service->ctx, info->enable_reg);
-
- set_reg_field_value(
- value,
- current_status ? 0 : 1,
- HPD0_DC_HPD_INT_CONTROL,
- DC_HPD_INT_POLARITY);
-
- dm_write_reg(irq_service->ctx, info->enable_reg, value);
-
- return true;
-}
-
static struct irq_source_info_funcs hpd_irq_info_funcs = {
.set = NULL,
- .ack = hpd_ack
+ .ack = hpd0_ack
};
static struct irq_source_info_funcs hpd_rx_irq_info_funcs = {
@@ -171,6 +144,16 @@ static struct irq_source_info_funcs vline0_irq_info_funcs = {
.ack = NULL
};
+static struct irq_source_info_funcs vline1_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
+static struct irq_source_info_funcs vline2_irq_info_funcs = {
+ .set = NULL,
+ .ack = NULL
+};
+
#undef BASE_INNER
#define BASE_INNER(seg) DCN_BASE__INST0_SEG ## seg
@@ -239,6 +222,13 @@ static struct irq_source_info_funcs vline0_irq_info_funcs = {
.funcs = &pflip_irq_info_funcs\
}
+#define vblank_int_entry(reg_num)\
+ [DC_IRQ_SOURCE_VBLANK1 + reg_num] = {\
+ IRQ_REG_ENTRY(OTG, reg_num,\
+ OTG_GLOBAL_SYNC_STATUS, VSTARTUP_INT_EN,\
+ OTG_GLOBAL_SYNC_STATUS, VSTARTUP_EVENT_CLEAR),\
+ .funcs = &vblank_irq_info_funcs\
+ }
/* vupdate_no_lock_int_entry maps to DC_IRQ_SOURCE_VUPDATEx, to match semantic
* of DCE's DC_IRQ_SOURCE_VUPDATEx.
*/
@@ -250,13 +240,6 @@ static struct irq_source_info_funcs vline0_irq_info_funcs = {
.funcs = &vupdate_no_lock_irq_info_funcs\
}
-#define vblank_int_entry(reg_num)\
- [DC_IRQ_SOURCE_VBLANK1 + reg_num] = {\
- IRQ_REG_ENTRY(OTG, reg_num,\
- OTG_GLOBAL_SYNC_STATUS, VSTARTUP_INT_EN,\
- OTG_GLOBAL_SYNC_STATUS, VSTARTUP_EVENT_CLEAR),\
- .funcs = &vblank_irq_info_funcs\
- }
#define vline0_int_entry(reg_num)\
[DC_IRQ_SOURCE_DC1_VLINE0 + reg_num] = {\
IRQ_REG_ENTRY(OTG, reg_num,\
@@ -264,6 +247,20 @@ static struct irq_source_info_funcs vline0_irq_info_funcs = {
OTG_VERTICAL_INTERRUPT0_CONTROL, OTG_VERTICAL_INTERRUPT0_CLEAR),\
.funcs = &vline0_irq_info_funcs\
}
+#define vline1_int_entry(reg_num)\
+ [DC_IRQ_SOURCE_DC1_VLINE1 + reg_num] = {\
+ IRQ_REG_ENTRY(OTG, reg_num,\
+ OTG_VERTICAL_INTERRUPT1_CONTROL, OTG_VERTICAL_INTERRUPT1_INT_ENABLE,\
+ OTG_VERTICAL_INTERRUPT1_CONTROL, OTG_VERTICAL_INTERRUPT1_CLEAR),\
+ .funcs = &vline1_irq_info_funcs\
+ }
+#define vline2_int_entry(reg_num)\
+ [DC_IRQ_SOURCE_DC1_VLINE2 + reg_num] = {\
+ IRQ_REG_ENTRY(OTG, reg_num,\
+ OTG_VERTICAL_INTERRUPT2_CONTROL, OTG_VERTICAL_INTERRUPT2_INT_ENABLE,\
+ OTG_VERTICAL_INTERRUPT2_CONTROL, OTG_VERTICAL_INTERRUPT2_CLEAR),\
+ .funcs = &vline2_irq_info_funcs\
+ }
#define dmub_outbox_int_entry()\
[DC_IRQ_SOURCE_DMCUB_OUTBOX] = {\
IRQ_REG_ENTRY_DMUB(\
@@ -364,21 +361,29 @@ irq_source_info_dcn401[DAL_IRQ_SOURCES_NUMBER] = {
dc_underflow_int_entry(6),
[DC_IRQ_SOURCE_DMCU_SCP] = dummy_irq_entry(),
[DC_IRQ_SOURCE_VBIOS_SW] = dummy_irq_entry(),
- vupdate_no_lock_int_entry(0),
- vupdate_no_lock_int_entry(1),
- vupdate_no_lock_int_entry(2),
- vupdate_no_lock_int_entry(3),
vblank_int_entry(0),
vblank_int_entry(1),
vblank_int_entry(2),
vblank_int_entry(3),
+ [DC_IRQ_SOURCE_DC5_VLINE1] = dummy_irq_entry(),
+ [DC_IRQ_SOURCE_DC6_VLINE1] = dummy_irq_entry(),
+ dmub_outbox_int_entry(),
+ vupdate_no_lock_int_entry(0),
+ vupdate_no_lock_int_entry(1),
+ vupdate_no_lock_int_entry(2),
+ vupdate_no_lock_int_entry(3),
vline0_int_entry(0),
vline0_int_entry(1),
vline0_int_entry(2),
vline0_int_entry(3),
- [DC_IRQ_SOURCE_DC5_VLINE1] = dummy_irq_entry(),
- [DC_IRQ_SOURCE_DC6_VLINE1] = dummy_irq_entry(),
- dmub_outbox_int_entry(),
+ vline1_int_entry(0),
+ vline1_int_entry(1),
+ vline1_int_entry(2),
+ vline1_int_entry(3),
+ vline2_int_entry(0),
+ vline2_int_entry(1),
+ vline2_int_entry(2),
+ vline2_int_entry(3),
};
static const struct irq_service_funcs irq_service_funcs_dcn401 = {
diff --git a/drivers/gpu/drm/amd/display/dc/irq/irq_service.c b/drivers/gpu/drm/amd/display/dc/irq/irq_service.c
index eca3d7ee7e4e..b595a11c5eaf 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/irq_service.c
+++ b/drivers/gpu/drm/amd/display/dc/irq/irq_service.c
@@ -41,6 +41,16 @@
#include "reg_helper.h"
#include "irq_service.h"
+//HPD0_DC_HPD_INT_STATUS
+#define HPD0_DC_HPD_INT_STATUS__DC_HPD_SENSE_DELAYED_MASK 0x00000010L
+#define HPD0_DC_HPD_INT_CONTROL__DC_HPD_INT_POLARITY_MASK 0x00000100L
+#define HPD0_DC_HPD_INT_STATUS__DC_HPD_SENSE_DELAYED__SHIFT 0x4
+#define HPD0_DC_HPD_INT_CONTROL__DC_HPD_INT_POLARITY__SHIFT 0x8
+//HPD1_DC_HPD_INT_STATUS
+#define DC_HPD1_INT_STATUS__DC_HPD1_SENSE_DELAYED_MASK 0x10
+#define DC_HPD1_INT_STATUS__DC_HPD1_SENSE_DELAYED__SHIFT 0x4
+#define DC_HPD1_INT_CONTROL__DC_HPD1_INT_POLARITY_MASK 0x100
+#define DC_HPD1_INT_CONTROL__DC_HPD1_INT_POLARITY__SHIFT 0x8
#define CTX \
@@ -177,3 +187,57 @@ enum dc_irq_source dal_irq_service_to_irq_source(
src_id,
ext_id);
}
+
+bool hpd0_ack(
+ struct irq_service *irq_service,
+ const struct irq_source_info *info)
+{
+ uint32_t addr = info->status_reg;
+ uint32_t value = dm_read_reg(irq_service->ctx, addr);
+ uint32_t current_status =
+ get_reg_field_value(
+ value,
+ HPD0_DC_HPD_INT_STATUS,
+ DC_HPD_SENSE_DELAYED);
+
+ dal_irq_service_ack_generic(irq_service, info);
+
+ value = dm_read_reg(irq_service->ctx, info->enable_reg);
+
+ set_reg_field_value(
+ value,
+ current_status ? 0 : 1,
+ HPD0_DC_HPD_INT_CONTROL,
+ DC_HPD_INT_POLARITY);
+
+ dm_write_reg(irq_service->ctx, info->enable_reg, value);
+
+ return true;
+}
+
+bool hpd1_ack(
+ struct irq_service *irq_service,
+ const struct irq_source_info *info)
+{
+ uint32_t addr = info->status_reg;
+ uint32_t value = dm_read_reg(irq_service->ctx, addr);
+ uint32_t current_status =
+ get_reg_field_value(
+ value,
+ DC_HPD1_INT_STATUS,
+ DC_HPD1_SENSE_DELAYED);
+
+ dal_irq_service_ack_generic(irq_service, info);
+
+ value = dm_read_reg(irq_service->ctx, info->enable_reg);
+
+ set_reg_field_value(
+ value,
+ current_status ? 0 : 1,
+ DC_HPD1_INT_CONTROL,
+ DC_HPD1_INT_POLARITY);
+
+ dm_write_reg(irq_service->ctx, info->enable_reg, value);
+
+ return true;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/irq/irq_service.h b/drivers/gpu/drm/amd/display/dc/irq/irq_service.h
index b178f85944cd..bbcef3d2fe33 100644
--- a/drivers/gpu/drm/amd/display/dc/irq/irq_service.h
+++ b/drivers/gpu/drm/amd/display/dc/irq/irq_service.h
@@ -82,4 +82,12 @@ void dal_irq_service_set_generic(
const struct irq_source_info *info,
bool enable);
+bool hpd0_ack(
+ struct irq_service *irq_service,
+ const struct irq_source_info *info);
+
+bool hpd1_ack(
+ struct irq_service *irq_service,
+ const struct irq_source_info *info);
+
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/irq_types.h b/drivers/gpu/drm/amd/display/dc/irq_types.h
index e962c426beda..a2f7b933bebf 100644
--- a/drivers/gpu/drm/amd/display/dc/irq_types.h
+++ b/drivers/gpu/drm/amd/display/dc/irq_types.h
@@ -161,6 +161,20 @@ enum dc_irq_source {
DC_IRQ_SOURCE_DPCX_TX_PHYE,
DC_IRQ_SOURCE_DPCX_TX_PHYF,
+ DC_IRQ_SOURCE_DC1_VLINE2,
+ DC_IRQ_SOURCE_DC2_VLINE2,
+ DC_IRQ_SOURCE_DC3_VLINE2,
+ DC_IRQ_SOURCE_DC4_VLINE2,
+ DC_IRQ_SOURCE_DC5_VLINE2,
+ DC_IRQ_SOURCE_DC6_VLINE2,
+
+ DC_IRQ_SOURCE_DCI2C_RR_DDC1,
+ DC_IRQ_SOURCE_DCI2C_RR_DDC2,
+ DC_IRQ_SOURCE_DCI2C_RR_DDC3,
+ DC_IRQ_SOURCE_DCI2C_RR_DDC4,
+ DC_IRQ_SOURCE_DCI2C_RR_DDC5,
+ DC_IRQ_SOURCE_DCI2C_RR_DDC6,
+
DAL_IRQ_SOURCES_NUMBER
};
@@ -170,6 +184,9 @@ enum irq_type
IRQ_TYPE_VUPDATE = DC_IRQ_SOURCE_VUPDATE1,
IRQ_TYPE_VBLANK = DC_IRQ_SOURCE_VBLANK1,
IRQ_TYPE_VLINE0 = DC_IRQ_SOURCE_DC1_VLINE0,
+ IRQ_TYPE_VLINE1 = DC_IRQ_SOURCE_DC1_VLINE1,
+ IRQ_TYPE_VLINE2 = DC_IRQ_SOURCE_DC1_VLINE2,
+ IRQ_TYPE_DCUNDERFLOW = DC_IRQ_SOURCE_DC1UNDERFLOW,
};
#define DAL_VALID_IRQ_SRC_NUM(src) \
diff --git a/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_cts.c b/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_cts.c
index ff8fe1a94965..2956c2b3ad1a 100644
--- a/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_cts.c
+++ b/drivers/gpu/drm/amd/display/dc/link/accessories/link_dp_cts.c
@@ -34,6 +34,7 @@
#include "dm_helpers.h"
#include "dc_dmub_srv.h"
#include "dce/dmub_hw_lock_mgr.h"
+#include "clk_mgr.h"
#define DC_LOGGER \
link->ctx->logger
@@ -67,10 +68,17 @@ static void dp_retrain_link_dp_test(struct dc_link *link,
{
struct pipe_ctx *pipes[MAX_PIPES];
struct dc_state *state = link->dc->current_state;
+ struct dc_stream_update stream_update = { 0 };
+ bool dpms_off = false;
+ bool needs_divider_update = false;
bool was_hpo_acquired = resource_is_hpo_acquired(link->dc->current_state);
bool is_hpo_acquired;
uint8_t count;
int i;
+ struct audio_output audio_output[MAX_PIPES];
+
+ needs_divider_update = (link->dc->link_srv->dp_get_encoding_format(link_setting) !=
+ link->dc->link_srv->dp_get_encoding_format((const struct dc_link_settings *) &link->cur_link_settings));
udelay(100);
@@ -83,16 +91,59 @@ static void dp_retrain_link_dp_test(struct dc_link *link,
link->dc,
state,
pipes[i]);
+
+ // Disable OTG and re-enable after updating clocks
+ pipes[i]->stream_res.tg->funcs->disable_crtc(pipes[i]->stream_res.tg);
}
- if (link->dc->hwss.setup_hpo_hw_control) {
- is_hpo_acquired = resource_is_hpo_acquired(state);
- if (was_hpo_acquired != is_hpo_acquired)
- link->dc->hwss.setup_hpo_hw_control(link->dc->hwseq, is_hpo_acquired);
+ if (needs_divider_update && link->dc->res_pool->funcs->update_dc_state_for_encoder_switch) {
+ link->dc->res_pool->funcs->update_dc_state_for_encoder_switch(link,
+ link_setting, count,
+ *pipes, &audio_output[0]);
+ for (i = 0; i < count; i++) {
+ pipes[i]->clock_source->funcs->program_pix_clk(
+ pipes[i]->clock_source,
+ &pipes[i]->stream_res.pix_clk_params,
+ link->dc->link_srv->dp_get_encoding_format(&pipes[i]->link_config.dp_link_settings),
+ &pipes[i]->pll_settings);
+
+ if (pipes[i]->stream_res.audio != NULL) {
+ const struct link_hwss *link_hwss = get_link_hwss(
+ link, &pipes[i]->link_res);
+
+ link_hwss->setup_audio_output(pipes[i], &audio_output[i],
+ pipes[i]->stream_res.audio->inst);
+
+ pipes[i]->stream_res.audio->funcs->az_configure(
+ pipes[i]->stream_res.audio,
+ pipes[i]->stream->signal,
+ &audio_output[i].crtc_info,
+ &pipes[i]->stream->audio_info,
+ &audio_output[i].dp_link_info);
+
+ if (link->dc->config.disable_hbr_audio_dp2 &&
+ pipes[i]->stream_res.audio->funcs->az_disable_hbr_audio &&
+ link->dc->link_srv->dp_is_128b_132b_signal(pipes[i]))
+ pipes[i]->stream_res.audio->funcs->az_disable_hbr_audio(pipes[i]->stream_res.audio);
+ }
+ }
}
- for (i = count-1; i >= 0; i--)
- link_set_dpms_on(state, pipes[i]);
+ // Toggle on HPO I/O if necessary
+ is_hpo_acquired = resource_is_hpo_acquired(state);
+ if (was_hpo_acquired != is_hpo_acquired && link->dc->hwss.setup_hpo_hw_control)
+ link->dc->hwss.setup_hpo_hw_control(link->dc->hwseq, is_hpo_acquired);
+
+ for (i = 0; i < count; i++)
+ pipes[i]->stream_res.tg->funcs->enable_crtc(pipes[i]->stream_res.tg);
+
+ // Set DPMS on with stream update
+ for (i = 0; i < state->stream_count; i++)
+ if (state->streams[i] && state->streams[i]->link && state->streams[i]->link == link) {
+ stream_update.stream = state->streams[i];
+ stream_update.dpms_off = &dpms_off;
+ dc_update_planes_and_stream(state->clk_mgr->ctx->dc, NULL, 0, state->streams[i], &stream_update);
+ }
}
static void dp_test_send_link_training(struct dc_link *link)
@@ -251,7 +302,7 @@ static void dp_test_send_phy_test_pattern(struct dc_link *link)
link_training_settings.lttpr_mode = dp_decide_lttpr_mode(link, &link->cur_link_settings);
- if ((link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
+ if (((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
link_training_settings.lttpr_mode == LTTPR_MODE_TRANSPARENT)
dp_fixed_vs_pe_read_lane_adjust(
link,
@@ -646,7 +697,7 @@ bool dp_set_test_pattern(
if (IS_DP_PHY_PATTERN(test_pattern)) {
/* Set DPCD Lane Settings before running test pattern */
if (p_link_settings != NULL) {
- if ((link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
+ if (((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
p_link_settings->lttpr_mode == LTTPR_MODE_TRANSPARENT) {
dp_fixed_vs_pe_set_retimer_lane_settings(
link,
diff --git a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio.c b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio.c
index 3e47a6735912..b68bcc9fca0a 100644
--- a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio.c
+++ b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio.c
@@ -48,9 +48,16 @@ void set_dio_throttled_vcp_size(struct pipe_ctx *pipe_ctx,
void setup_dio_stream_encoder(struct pipe_ctx *pipe_ctx)
{
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(pipe_ctx->stream->link);
+ struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
struct stream_encoder *stream_enc = pipe_ctx->stream_res.stream_enc;
+ if (!pipe_ctx->stream->ctx->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(pipe_ctx->stream->link);
+ if (!link_enc) {
+ ASSERT(link_enc);
+ return;
+ }
+
link_enc->funcs->connect_dig_be_to_fe(link_enc,
pipe_ctx->stream_res.stream_enc->id, true);
if (dc_is_dp_signal(pipe_ctx->stream->signal))
@@ -71,9 +78,16 @@ void setup_dio_stream_encoder(struct pipe_ctx *pipe_ctx)
void reset_dio_stream_encoder(struct pipe_ctx *pipe_ctx)
{
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(pipe_ctx->stream->link);
+ struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
struct stream_encoder *stream_enc = pipe_ctx->stream_res.stream_enc;
+ if (!pipe_ctx->stream->ctx->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(pipe_ctx->stream->link);
+ if (!link_enc) {
+ ASSERT(link_enc);
+ return;
+ }
+
if (!stream_enc)
return;
@@ -142,7 +156,14 @@ void enable_dio_dp_link_output(struct dc_link *link,
enum clock_source_id clock_source,
const struct dc_link_settings *link_settings)
{
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link);
+ struct link_encoder *link_enc = link_res->dio_link_enc;
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+ if (!link_enc) {
+ ASSERT(link_enc);
+ return;
+ }
if (dc_is_dp_sst_signal(signal))
link_enc->funcs->enable_dp_output(
@@ -162,7 +183,14 @@ void disable_dio_link_output(struct dc_link *link,
const struct link_resource *link_res,
enum signal_type signal)
{
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link);
+ struct link_encoder *link_enc = link_res->dio_link_enc;
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+ if (!link_enc) {
+ ASSERT(link_enc);
+ return;
+ }
link_enc->funcs->disable_output(link_enc, signal);
link->dc->link_srv->dp_trace_source_sequence(link,
@@ -173,7 +201,14 @@ void set_dio_dp_link_test_pattern(struct dc_link *link,
const struct link_resource *link_res,
struct encoder_set_dp_phy_pattern_param *tp_params)
{
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link);
+ struct link_encoder *link_enc = link_res->dio_link_enc;
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+ if (!link_enc) {
+ ASSERT(link_enc);
+ return;
+ }
link_enc->funcs->dp_set_phy_pattern(link_enc, tp_params);
link->dc->link_srv->dp_trace_source_sequence(link, DPCD_SOURCE_SEQ_AFTER_SET_SOURCE_PATTERN);
@@ -184,7 +219,14 @@ void set_dio_dp_lane_settings(struct dc_link *link,
const struct dc_link_settings *link_settings,
const struct dc_lane_settings lane_settings[LANE_COUNT_DP_MAX])
{
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link);
+ struct link_encoder *link_enc = link_res->dio_link_enc;
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+ if (!link_enc) {
+ ASSERT(link_enc);
+ return;
+ }
link_enc->funcs->dp_set_lane_settings(link_enc, link_settings, lane_settings);
}
@@ -193,9 +235,15 @@ void update_dio_stream_allocation_table(struct dc_link *link,
const struct link_resource *link_res,
const struct link_mst_stream_allocation_table *table)
{
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link);
+ struct link_encoder *link_enc = link_res->dio_link_enc;
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+ if (!link_enc) {
+ ASSERT(link_enc);
+ return;
+ }
- ASSERT(link_enc);
link_enc->funcs->update_mst_stream_allocation_table(link_enc, table);
}
@@ -280,7 +328,10 @@ static const struct link_hwss dio_link_hwss = {
bool can_use_dio_link_hwss(const struct dc_link *link,
const struct link_resource *link_res)
{
- return link->link_enc != NULL;
+ if (!link->dc->config.unify_link_enc_assignment)
+ return link->link_enc != NULL;
+ else
+ return link_res->dio_link_enc != NULL;
}
/**
diff --git a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio_fixed_vs_pe_retimer.c b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio_fixed_vs_pe_retimer.c
index 348ea4cb832d..e1dff4e3f446 100644
--- a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio_fixed_vs_pe_retimer.c
+++ b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dio_fixed_vs_pe_retimer.c
@@ -127,7 +127,10 @@ static void set_dio_fixed_vs_pe_retimer_dp_link_test_pattern(struct dc_link *lin
const struct link_resource *link_res,
struct encoder_set_dp_phy_pattern_param *tp_params)
{
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link);
+ struct link_encoder *link_enc = link_res->dio_link_enc;
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
if (!set_dio_fixed_vs_pe_retimer_dp_link_test_pattern_override(
link, link_res, tp_params, get_dio_link_hwss())) {
@@ -187,7 +190,7 @@ static const struct link_hwss dio_fixed_vs_pe_retimer_link_hwss = {
bool requires_fixed_vs_pe_retimer_dio_link_hwss(const struct dc_link *link)
{
- return (link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN);
+ return ((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN);
}
const struct link_hwss *get_dio_fixed_vs_pe_retimer_link_hwss(void)
diff --git a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.c b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.c
index 6499807af72a..81bf3c5e1fdf 100644
--- a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.c
+++ b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.c
@@ -35,12 +35,15 @@ static void update_dpia_stream_allocation_table(struct dc_link *link,
const struct link_resource *link_res,
const struct link_mst_stream_allocation_table *table)
{
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link);
+ struct link_encoder *link_enc = link_res->dio_link_enc;
static enum dc_status status;
uint8_t mst_alloc_slots = 0, prev_mst_slots_in_use = 0xFF;
int i;
DC_LOGGER_INIT(link->ctx->logger);
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+
for (i = 0; i < table->stream_count; i++)
mst_alloc_slots += table->stream_allocations[i].slot_count;
@@ -61,7 +64,10 @@ static void set_dio_dpia_link_test_pattern(struct dc_link *link,
if (tp_params->dp_phy_pattern != DP_TEST_PATTERN_VIDEO_MODE)
return;
- struct link_encoder *link_enc = link_enc_cfg_get_link_enc(link);
+ struct link_encoder *link_enc = link_res->dio_link_enc;
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
if (!link_enc)
return;
@@ -77,17 +83,78 @@ static void set_dio_dpia_lane_settings(struct dc_link *link,
{
}
+static void enable_dpia_link_output(struct dc_link *link,
+ const struct link_resource *link_res,
+ enum signal_type signal,
+ enum clock_source_id clock_source,
+ const struct dc_link_settings *link_settings)
+{
+ struct link_encoder *link_enc = link_res->dio_link_enc;
+ DC_LOGGER_INIT(link->ctx->logger);
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+
+ if (link_enc != NULL) {
+ if (link->dc->config.enable_dpia_pre_training || link->dc->config.unify_link_enc_assignment) {
+ uint8_t fec_rdy = link->dc->link_srv->dp_should_enable_fec(link);
+ uint8_t digmode = dc_is_dp_sst_signal(signal) ? DIG_SST_MODE : DIG_MST_MODE;
+
+ if (link_enc->funcs->enable_dpia_output)
+ link_enc->funcs->enable_dpia_output(
+ link_enc,
+ link_settings,
+ link->ddc_hw_inst,
+ digmode,
+ fec_rdy);
+ else
+ DC_LOG_ERROR("%s: link encoder does not support enable_dpia_output\n", __func__);
+ } else
+ enable_dio_dp_link_output(link, link_res, signal, clock_source, link_settings);
+
+ }
+
+ link->dc->link_srv->dp_trace_source_sequence(link,
+ DPCD_SOURCE_SEQ_AFTER_ENABLE_LINK_PHY);
+}
+
+static void disable_dpia_link_output(struct dc_link *link,
+ const struct link_resource *link_res,
+ enum signal_type signal)
+{
+ struct link_encoder *link_enc = link_res->dio_link_enc;
+ DC_LOGGER_INIT(link->ctx->logger);
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+
+ if (link_enc != NULL) {
+ if (link->dc->config.enable_dpia_pre_training || link->dc->config.unify_link_enc_assignment) {
+ uint8_t digmode = dc_is_dp_sst_signal(signal) ? DIG_SST_MODE : DIG_MST_MODE;
+
+ if (link_enc->funcs->disable_dpia_output)
+ link_enc->funcs->disable_dpia_output(link_enc, link->ddc_hw_inst, digmode);
+ else
+ DC_LOG_ERROR("%s: link encoder does not support disable_dpia_output\n", __func__);
+ } else
+ link_enc->funcs->disable_output(link_enc, signal);
+ }
+
+ link->dc->link_srv->dp_trace_source_sequence(link,
+ DPCD_SOURCE_SEQ_AFTER_DISABLE_LINK_PHY);
+}
+
static const struct link_hwss dpia_link_hwss = {
.setup_stream_encoder = setup_dio_stream_encoder,
.reset_stream_encoder = reset_dio_stream_encoder,
.setup_stream_attribute = setup_dio_stream_attribute,
- .disable_link_output = disable_dio_link_output,
+ .disable_link_output = disable_dpia_link_output,
.setup_audio_output = setup_dio_audio_output,
.enable_audio_packet = enable_dio_audio_packet,
.disable_audio_packet = disable_dio_audio_packet,
.ext = {
.set_throttled_vcp_size = set_dio_throttled_vcp_size,
- .enable_dp_link_output = enable_dio_dp_link_output,
+ .enable_dp_link_output = enable_dpia_link_output,
.set_dp_link_test_pattern = set_dio_dpia_link_test_pattern,
.set_dp_lane_settings = set_dio_dpia_lane_settings,
.update_stream_allocation_table = update_dpia_stream_allocation_table,
@@ -97,8 +164,10 @@ static const struct link_hwss dpia_link_hwss = {
bool can_use_dpia_link_hwss(const struct dc_link *link,
const struct link_resource *link_res)
{
- return link->is_dig_mapping_flexible &&
- link->dc->res_pool->funcs->link_encs_assign;
+ if (!link->dc->config.unify_link_enc_assignment)
+ return link->is_dig_mapping_flexible && link->dc->res_pool->funcs->link_encs_assign;
+ else
+ return link->is_dig_mapping_flexible && link_res->dio_link_enc != NULL;
}
const struct link_hwss *get_dpia_link_hwss(void)
diff --git a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.h b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.h
index ad16ec5d9bb7..259e0f4775e1 100644
--- a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.h
+++ b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_dpia.h
@@ -27,6 +27,9 @@
#include "link_hwss.h"
+#define DIG_SST_MODE 0
+#define DIG_MST_MODE 5
+
const struct link_hwss *get_dpia_link_hwss(void);
bool can_use_dpia_link_hwss(const struct dc_link *link,
const struct link_resource *link_res);
diff --git a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_hpo_fixed_vs_pe_retimer_dp.c b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_hpo_fixed_vs_pe_retimer_dp.c
index 116ff37126e7..55c5148de800 100644
--- a/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_hpo_fixed_vs_pe_retimer_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/link/hwss/link_hwss_hpo_fixed_vs_pe_retimer_dp.c
@@ -74,7 +74,7 @@ static void dp_hpo_fixed_vs_pe_retimer_set_tx_ffe(struct dc_link *link,
static void dp_hpo_fixed_vs_pe_retimer_program_override_test_pattern(struct dc_link *link,
struct encoder_set_dp_phy_pattern_param *tp_params)
{
- uint8_t clk_src = 0x4C;
+ uint8_t clk_src = 0xC4;
uint8_t pattern = 0x4F; /* SQ128 */
const uint8_t vendor_lttpr_write_data_pg0[4] = {0x1, 0x11, 0x0, 0x0};
diff --git a/drivers/gpu/drm/amd/display/dc/link/link_detection.c b/drivers/gpu/drm/amd/display/dc/link/link_detection.c
index e026c728042a..827b630daf49 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_detection.c
+++ b/drivers/gpu/drm/amd/display/dc/link/link_detection.c
@@ -593,8 +593,9 @@ static bool detect_dp(struct dc_link *link,
if (sink_caps->transaction_type == DDC_TRANSACTION_TYPE_I2C_OVER_AUX) {
sink_caps->signal = SIGNAL_TYPE_DISPLAY_PORT;
- if (!detect_dp_sink_caps(link))
+ if (!detect_dp_sink_caps(link)) {
return false;
+ }
if (is_dp_branch_device(link))
/* DP SST branch */
@@ -611,6 +612,7 @@ static bool detect_dp(struct dc_link *link,
link->dpcd_caps.dongle_type = sink_caps->dongle_type;
link->dpcd_caps.is_dongle_type_one = sink_caps->is_dongle_type_one;
link->dpcd_caps.dpcd_rev.raw = 0;
+ link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.raw = 0;
}
return true;
@@ -816,7 +818,10 @@ static bool should_verify_link_capability_destructively(struct dc_link *link,
{
bool destrictive = false;
struct dc_link_settings max_link_cap;
- bool is_link_enc_unavailable = link->link_enc &&
+ bool is_link_enc_unavailable = false;
+
+ if (!link->dc->config.unify_link_enc_assignment)
+ is_link_enc_unavailable = link->link_enc &&
link->dc->res_pool->funcs->link_encs_assign &&
!link_enc_cfg_is_link_enc_avail(
link->ctx->dc,
@@ -829,7 +834,8 @@ static bool should_verify_link_capability_destructively(struct dc_link *link,
if (link->dc->debug.skip_detection_link_training ||
dc_is_embedded_signal(link->local_sink->sink_signal) ||
- link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) {
+ (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA &&
+ !link->dc->config.enable_dpia_pre_training)) {
destrictive = false;
} else if (link_dp_get_encoding_format(&max_link_cap) ==
DP_8b_10b_ENCODING) {
@@ -1003,21 +1009,11 @@ static bool detect_link_and_local_sink(struct dc_link *link,
link->reported_link_cap.link_rate > LINK_RATE_HIGH3)
link->reported_link_cap.link_rate = LINK_RATE_HIGH3;
- /*
- * If this is DP over USB4 link then we need to:
- * - Enable BW ALLOC support on DPtx if applicable
- */
- if (dc->config.usb4_bw_alloc_support) {
- if (link_dp_dpia_set_dptx_usb4_bw_alloc_support(link)) {
- /* update with non reduced link cap if bw allocation mode is supported */
- if (link->dpia_bw_alloc_config.nrd_max_link_rate &&
- link->dpia_bw_alloc_config.nrd_max_lane_count) {
- link->reported_link_cap.link_rate =
- link->dpia_bw_alloc_config.nrd_max_link_rate;
- link->reported_link_cap.lane_count =
- link->dpia_bw_alloc_config.nrd_max_lane_count;
- }
- }
+ if (link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dp_tunneling
+ && link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dpia_bw_alloc
+ && link->dpcd_caps.usb4_dp_tun_info.driver_bw_cap.bits.driver_bw_alloc_support) {
+ if (link_dpia_enable_usb4_dp_bw_alloc_mode(link) == false)
+ link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dpia_bw_alloc = false;
}
break;
}
diff --git a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
index 41cab9ad6885..8c8682f743d6 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
+++ b/drivers/gpu/drm/amd/display/dc/link/link_dpms.c
@@ -140,7 +140,8 @@ void link_blank_dp_stream(struct dc_link *link, bool hw_init)
}
}
- if ((!link->wa_flags.dp_keep_receiver_powered) || hw_init)
+ if (((!dc->is_switch_in_progress_dest) && ((!link->wa_flags.dp_keep_receiver_powered) || hw_init)) &&
+ (link->type != dc_connection_none))
dpcd_write_rx_power_ctrl(link, false);
}
}
@@ -148,6 +149,7 @@ void link_blank_dp_stream(struct dc_link *link, bool hw_init)
void link_set_all_streams_dpms_off_for_link(struct dc_link *link)
{
struct pipe_ctx *pipes[MAX_PIPES];
+ struct dc_stream_state *streams[MAX_PIPES];
struct dc_state *state = link->dc->current_state;
uint8_t count;
int i;
@@ -160,10 +162,18 @@ void link_set_all_streams_dpms_off_for_link(struct dc_link *link)
link_get_master_pipes_with_dpms_on(link, state, &count, pipes);
+ /* The subsequent call to dc_commit_updates_for_stream for a full update
+ * will release the current state and swap to a new state. Releasing the
+ * current state results in the stream pointers in the pipe_ctx structs
+ * to be zero'd. Hence, cache all streams prior to dc_commit_updates_for_stream.
+ */
+ for (i = 0; i < count; i++)
+ streams[i] = pipes[i]->stream;
+
for (i = 0; i < count; i++) {
- stream_update.stream = pipes[i]->stream;
+ stream_update.stream = streams[i];
dc_commit_updates_for_stream(link->ctx->dc, NULL, 0,
- pipes[i]->stream, &stream_update,
+ streams[i], &stream_update,
state);
}
@@ -652,15 +662,15 @@ static void write_i2c_redriver_setting(
static void update_psp_stream_config(struct pipe_ctx *pipe_ctx, bool dpms_off)
{
struct cp_psp *cp_psp = &pipe_ctx->stream->ctx->cp_psp;
- struct link_encoder *link_enc = NULL;
+ struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
struct cp_psp_stream_config config = {0};
enum dp_panel_mode panel_mode =
dp_get_panel_mode(pipe_ctx->stream->link);
if (cp_psp == NULL || cp_psp->funcs.update_stream_config == NULL)
return;
-
- link_enc = link_enc_cfg_get_link_enc(pipe_ctx->stream->link);
+ if (!pipe_ctx->stream->ctx->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(pipe_ctx->stream->link);
ASSERT(link_enc);
if (link_enc == NULL)
return;
@@ -772,6 +782,20 @@ static bool dp_set_dsc_on_rx(struct pipe_ctx *pipe_ctx, bool enable)
return result;
}
+static bool dp_set_hblank_reduction_on_rx(struct pipe_ctx *pipe_ctx)
+{
+ struct dc *dc = pipe_ctx->stream->ctx->dc;
+ struct dc_stream_state *stream = pipe_ctx->stream;
+ bool result = false;
+
+ if (dc_is_virtual_signal(stream->signal))
+ result = true;
+ else
+ result = dm_helpers_dp_write_hblank_reduction(dc->ctx, stream);
+ return result;
+}
+
+
/* The stream with these settings can be sent (unblanked) only after DSC was enabled on RX first,
* i.e. after dp_enable_dsc_on_rx() had been called
*/
@@ -808,7 +832,8 @@ void link_set_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable)
enum optc_dsc_mode optc_dsc_mode;
/* Enable DSC hw block */
- dsc_cfg.pic_width = (stream->timing.h_addressable + stream->timing.h_border_left + stream->timing.h_border_right) / opp_cnt;
+ dsc_cfg.pic_width = (stream->timing.h_addressable + pipe_ctx->hblank_borrow +
+ stream->timing.h_border_left + stream->timing.h_border_right) / opp_cnt;
dsc_cfg.pic_height = stream->timing.v_addressable + stream->timing.v_border_top + stream->timing.v_border_bottom;
dsc_cfg.pixel_encoding = stream->timing.pixel_encoding;
dsc_cfg.color_depth = stream->timing.display_color_depth;
@@ -818,14 +843,14 @@ void link_set_dsc_on_stream(struct pipe_ctx *pipe_ctx, bool enable)
dsc_cfg.dc_dsc_cfg.num_slices_h /= opp_cnt;
if (should_use_dto_dscclk)
- dccg->funcs->set_dto_dscclk(dccg, dsc->inst);
+ dccg->funcs->set_dto_dscclk(dccg, dsc->inst, dsc_cfg.dc_dsc_cfg.num_slices_h);
dsc->funcs->dsc_set_config(dsc, &dsc_cfg, &dsc_optc_cfg);
dsc->funcs->dsc_enable(dsc, pipe_ctx->stream_res.opp->inst);
for (odm_pipe = pipe_ctx->next_odm_pipe; odm_pipe; odm_pipe = odm_pipe->next_odm_pipe) {
struct display_stream_compressor *odm_dsc = odm_pipe->stream_res.dsc;
if (should_use_dto_dscclk)
- dccg->funcs->set_dto_dscclk(dccg, odm_dsc->inst);
+ dccg->funcs->set_dto_dscclk(dccg, odm_dsc->inst, dsc_cfg.dc_dsc_cfg.num_slices_h);
odm_dsc->funcs->dsc_set_config(odm_dsc, &dsc_cfg, &dsc_optc_cfg);
odm_dsc->funcs->dsc_enable(odm_dsc, odm_pipe->stream_res.opp->inst);
}
@@ -1909,7 +1934,7 @@ static void disable_link_dp(struct dc_link *link,
if (link_dp_get_encoding_format(&link_settings) ==
DP_8b_10b_ENCODING) {
- dp_set_fec_enable(link, false);
+ dp_set_fec_enable(link, link_res, false);
dp_set_fec_ready(link, link_res, false);
}
}
@@ -1952,11 +1977,15 @@ static void enable_link_hdmi(struct pipe_ctx *pipe_ctx)
stream->phy_pix_clk = stream->timing.pix_clk_100hz / 10;
if (stream->phy_pix_clk > 340000)
is_over_340mhz = true;
+ if (dc_is_tmds_signal(stream->signal) && stream->phy_pix_clk > 6000000UL) {
+ ASSERT(false);
+ return;
+ }
if (dc_is_hdmi_signal(pipe_ctx->stream->signal)) {
unsigned short masked_chip_caps = pipe_ctx->stream->link->chip_caps &
- EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK;
- if (masked_chip_caps == EXT_DISPLAY_PATH_CAPS__HDMI20_TISN65DP159RSBT) {
+ AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK;
+ if (masked_chip_caps == AMD_EXT_DISPLAY_PATH_CAPS__HDMI20_TISN65DP159RSBT) {
/* DP159, Retimer settings */
eng_id = pipe_ctx->stream_res.stream_enc->id;
@@ -1967,7 +1996,7 @@ static void enable_link_hdmi(struct pipe_ctx *pipe_ctx)
write_i2c_default_retimer_setting(pipe_ctx,
is_vga_mode, is_over_340mhz);
}
- } else if (masked_chip_caps == EXT_DISPLAY_PATH_CAPS__HDMI20_PI3EQX1204) {
+ } else if (masked_chip_caps == AMD_EXT_DISPLAY_PATH_CAPS__HDMI20_PI3EQX1204) {
/* PI3EQX1204, Redriver settings */
write_i2c_redriver_setting(pipe_ctx, is_over_340mhz);
}
@@ -2023,7 +2052,7 @@ static enum dc_status enable_link_dp(struct dc_state *state,
int lt_attempts = LINK_TRAINING_ATTEMPTS;
// Increase retry count if attempting DP1.x on FIXED_VS link
- if ((link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
+ if (((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
link_dp_get_encoding_format(link_settings) == DP_8b_10b_ENCODING)
lt_attempts = 10;
@@ -2038,7 +2067,8 @@ static enum dc_status enable_link_dp(struct dc_state *state,
/* Train with fallback when enabling DPIA link. Conventional links are
* trained with fallback during sink detection.
*/
- if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA)
+ if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA &&
+ !link->dc->config.enable_dpia_pre_training)
do_fallback = true;
/*
@@ -2102,7 +2132,7 @@ static enum dc_status enable_link_dp(struct dc_state *state,
fec_enable = true;
if (link_dp_get_encoding_format(link_settings) == DP_8b_10b_ENCODING)
- dp_set_fec_enable(link, fec_enable);
+ dp_set_fec_enable(link, &pipe_ctx->link_res, fec_enable);
// during mode set we do DP_SET_POWER off then on, aux writes are lost
if (link->dpcd_sink_ext_caps.bits.oled == 1 ||
@@ -2267,26 +2297,10 @@ static bool allocate_usb4_bandwidth_for_stream(struct dc_stream_state *stream, i
link->dpia_bw_alloc_config.remote_sink_req_bw[sink_index] = bw;
}
- /* get dp overhead for dp tunneling */
- link->dpia_bw_alloc_config.dp_overhead = link_dp_dpia_get_dp_overhead_in_dp_tunneling(link);
+ link->dpia_bw_alloc_config.dp_overhead = link_dpia_get_dp_overhead(link);
req_bw += link->dpia_bw_alloc_config.dp_overhead;
- if (link_dp_dpia_allocate_usb4_bandwidth_for_stream(link, req_bw)) {
- if (req_bw <= link->dpia_bw_alloc_config.allocated_bw) {
- DC_LOG_DEBUG("%s, Success in allocate bw for link(%d), allocated_bw(%d), dp_overhead(%d)\n",
- __func__, link->link_index, link->dpia_bw_alloc_config.allocated_bw,
- link->dpia_bw_alloc_config.dp_overhead);
- } else {
- // Cannot get the required bandwidth.
- DC_LOG_ERROR("%s, Failed to allocate bw for link(%d), allocated_bw(%d), dp_overhead(%d)\n",
- __func__, link->link_index, link->dpia_bw_alloc_config.allocated_bw,
- link->dpia_bw_alloc_config.dp_overhead);
- return false;
- }
- } else {
- DC_LOG_DEBUG("%s, usb4 request bw timeout\n", __func__);
- return false;
- }
+ link_dp_dpia_allocate_usb4_bandwidth_for_stream(link, req_bw);
if (stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST) {
int i = 0;
@@ -2360,7 +2374,7 @@ void link_set_dpms_off(struct pipe_ctx *pipe_ctx)
update_psp_stream_config(pipe_ctx, true);
dc->hwss.blank_stream(pipe_ctx);
- if (pipe_ctx->stream->link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA)
+ if (pipe_ctx->link_config.dp_tunnel_settings.should_use_dp_bw_allocation)
deallocate_usb4_bandwidth(pipe_ctx->stream);
if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
@@ -2374,13 +2388,13 @@ void link_set_dpms_off(struct pipe_ctx *pipe_ctx)
enum engine_id eng_id = pipe_ctx->stream_res.stream_enc->id;
unsigned short masked_chip_caps = link->chip_caps &
- EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK;
+ AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK;
//Need to inform that sink is going to use legacy HDMI mode.
write_scdc_data(
link->ddc,
165000,//vbios only handles 165Mhz.
false);
- if (masked_chip_caps == EXT_DISPLAY_PATH_CAPS__HDMI20_TISN65DP159RSBT) {
+ if (masked_chip_caps == AMD_EXT_DISPLAY_PATH_CAPS__HDMI20_TISN65DP159RSBT) {
/* DP159, Retimer settings */
if (get_ext_hdmi_settings(pipe_ctx, eng_id, &settings))
write_i2c_retimer_setting(pipe_ctx,
@@ -2388,7 +2402,7 @@ void link_set_dpms_off(struct pipe_ctx *pipe_ctx)
else
write_i2c_default_retimer_setting(pipe_ctx,
false, false);
- } else if (masked_chip_caps == EXT_DISPLAY_PATH_CAPS__HDMI20_PI3EQX1204) {
+ } else if (masked_chip_caps == AMD_EXT_DISPLAY_PATH_CAPS__HDMI20_PI3EQX1204) {
/* PI3EQX1204, Redriver settings */
write_i2c_redriver_setting(pipe_ctx, false);
}
@@ -2428,7 +2442,7 @@ void link_set_dpms_off(struct pipe_ctx *pipe_ctx)
if (link->connector_signal == SIGNAL_TYPE_EDP && dc->debug.psp_disabled_wa) {
/* reset internal save state to default since eDP is off */
enum dp_panel_mode panel_mode = dp_get_panel_mode(pipe_ctx->stream->link);
- /* since current psp not loaded, we need to reset it to default*/
+ /* since current psp not loaded, we need to reset it to default */
link->panel_mode = panel_mode;
}
}
@@ -2441,7 +2455,7 @@ void link_set_dpms_on(
struct dc_stream_state *stream = pipe_ctx->stream;
struct dc_link *link = stream->sink->link;
enum dc_status status;
- struct link_encoder *link_enc;
+ struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
enum otg_out_mux_dest otg_out_dest = OUT_MUX_DIO;
struct vpg *vpg = pipe_ctx->stream_res.stream_enc->vpg;
const struct link_hwss *link_hwss = get_link_hwss(link, &pipe_ctx->link_res);
@@ -2466,7 +2480,8 @@ void link_set_dpms_on(
}
}
- link_enc = link_enc_cfg_get_link_enc(link);
+ if (!dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
ASSERT(link_enc);
if (!dc_is_virtual_signal(pipe_ctx->stream->signal)
@@ -2522,12 +2537,29 @@ void link_set_dpms_on(
!pipe_ctx->next_odm_pipe) {
pipe_ctx->stream->dpms_off = false;
update_psp_stream_config(pipe_ctx, false);
+
+ if (link->is_dds) {
+ uint32_t post_oui_delay = 30; // 30ms
+
+ dpcd_set_source_specific_data(link);
+ msleep(post_oui_delay);
+ }
+
return;
}
if (pipe_ctx->stream->dpms_off)
return;
+ /* For Dp tunneling link, a pending HPD means that we have a race condition between processing
+ * current link and processing the pending HPD. If we enable the link now, we may end up with a
+ * link that is not actually connected to a sink. So we skip enabling the link in this case.
+ */
+ if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA && link->is_hpd_pending) {
+ DC_LOG_DEBUG("%s, Link%d HPD is pending, not enable it.\n", __func__, link->link_index);
+ return;
+ }
+
/* Have to setup DSC before DIG FE and BE are connected (which happens before the
* link training). This is to make sure the bandwidth sent to DIG BE won't be
* bigger than what the link and/or DIG BE can handle. VBID[6]/CompressedStream_flag
@@ -2593,7 +2625,10 @@ void link_set_dpms_on(
}
}
- if (pipe_ctx->stream->link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA)
+ if (dc_is_dp_signal(pipe_ctx->stream->signal))
+ dp_set_hblank_reduction_on_rx(pipe_ctx);
+
+ if (pipe_ctx->link_config.dp_tunnel_settings.should_use_dp_bw_allocation)
allocate_usb4_bandwidth(pipe_ctx->stream);
if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
@@ -2602,6 +2637,15 @@ void link_set_dpms_on(
dp_is_128b_132b_signal(pipe_ctx))
update_sst_payload(pipe_ctx, true);
+ /* Corruption was observed on systems with display mux when stream gets
+ * enabled after the mux switch. Having a small delay between link
+ * training and stream unblank resolves the corruption issue.
+ * This is workaround.
+ */
+ if (pipe_ctx->stream->signal == SIGNAL_TYPE_EDP &&
+ link->is_display_mux_present)
+ msleep(20);
+
dc->hwss.unblank_stream(pipe_ctx,
&pipe_ctx->stream->link->cur_link_settings);
diff --git a/drivers/gpu/drm/amd/display/dc/link/link_factory.c b/drivers/gpu/drm/amd/display/dc/link/link_factory.c
index 5e1b5ab9fbc6..de1143dbbd25 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_factory.c
+++ b/drivers/gpu/drm/amd/display/dc/link/link_factory.c
@@ -100,7 +100,8 @@ static void construct_link_service_validation(struct link_service *link_srv)
{
link_srv->validate_mode_timing = link_validate_mode_timing;
link_srv->dp_link_bandwidth_kbps = dp_link_bandwidth_kbps;
- link_srv->validate_dpia_bandwidth = link_validate_dpia_bandwidth;
+ link_srv->validate_dp_tunnel_bandwidth = link_validate_dp_tunnel_bandwidth;
+ link_srv->dp_required_hblank_size_bytes = dp_required_hblank_size_bytes;
}
/* link dpms owns the programming sequence of stream's dpms state associated
@@ -155,6 +156,7 @@ static void construct_link_service_dp_capability(struct link_service *link_srv)
link_srv->dp_get_encoding_format = link_dp_get_encoding_format;
link_srv->dp_should_enable_fec = dp_should_enable_fec;
link_srv->dp_decide_link_settings = link_decide_link_settings;
+ link_srv->dp_decide_tunnel_settings = link_decide_dp_tunnel_settings;
link_srv->mst_decide_link_encoding_format =
mst_decide_link_encoding_format;
link_srv->edp_decide_link_settings = edp_decide_link_settings;
@@ -174,7 +176,6 @@ static void construct_link_service_dp_phy_or_dpia(struct link_service *link_srv)
{
link_srv->dpia_handle_usb4_bandwidth_allocation_for_link =
dpia_handle_usb4_bandwidth_allocation_for_link;
- link_srv->dpia_handle_bw_alloc_response = dpia_handle_bw_alloc_response;
link_srv->dp_set_drive_settings = dp_set_drive_settings;
link_srv->dpcd_write_rx_power_ctrl = dpcd_write_rx_power_ctrl;
}
@@ -464,6 +465,7 @@ static bool construct_phy(struct dc_link *link,
link->irq_source_hpd = DC_IRQ_SOURCE_INVALID;
link->irq_source_hpd_rx = DC_IRQ_SOURCE_INVALID;
+ link->irq_source_read_request = DC_IRQ_SOURCE_INVALID;
link->link_status.dpcd_caps = &link->dpcd_caps;
link->dc = init_params->dc;
@@ -514,6 +516,9 @@ static bool construct_phy(struct dc_link *link,
case CONNECTOR_ID_HDMI_TYPE_A:
link->connector_signal = SIGNAL_TYPE_HDMI_TYPE_A;
+ if (link->hpd_gpio)
+ link->irq_source_read_request =
+ dal_irq_get_read_request(link->hpd_gpio);
break;
case CONNECTOR_ID_SINGLE_LINK_DVID:
case CONNECTOR_ID_SINGLE_LINK_DVII:
@@ -534,10 +539,16 @@ static bool construct_phy(struct dc_link *link,
break;
case CONNECTOR_ID_EDP:
+ // If smartmux is supported, only create the link on the primary eDP.
+ // Dual eDP is not supported with smartmux.
+ if (!(!link->dc->config.smart_mux_version || dc_ctx->dc_edp_id_count == 0))
+ goto create_fail;
+
link->connector_signal = SIGNAL_TYPE_EDP;
if (link->hpd_gpio) {
- if (!link->dc->config.allow_edp_hotplug_detection)
+ if (!link->dc->config.allow_edp_hotplug_detection
+ && !is_smartmux_suported(link))
link->irq_source_hpd = DC_IRQ_SOURCE_INVALID;
switch (link->dc->config.allow_edp_hotplug_detection) {
@@ -653,7 +664,7 @@ static bool construct_phy(struct dc_link *link,
}
/* Look for device tag that matches connector signal,
- * CRT for rgb, LCD for other supported signal tyes
+ * CRT for rgb, LCD for other supported signal types
*/
if (!bp_funcs->is_device_id_supported(dc_ctx->dc_bios,
link->device_tag.dev_id))
@@ -698,7 +709,7 @@ static bool construct_phy(struct dc_link *link,
link->chip_caps);
}
- if (link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) {
+ if ((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) {
link->bios_forced_drive_settings.VOLTAGE_SWING =
(bios->integrated_info->ext_disp_conn_info.fixdpvoltageswing & 0x3);
link->bios_forced_drive_settings.PRE_EMPHASIS =
diff --git a/drivers/gpu/drm/amd/display/dc/link/link_validation.c b/drivers/gpu/drm/amd/display/dc/link/link_validation.c
index 60f15a9ba7a5..aecaf37eee35 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_validation.c
+++ b/drivers/gpu/drm/amd/display/dc/link/link_validation.c
@@ -86,6 +86,10 @@ static bool dp_active_dongle_validate_timing(
if (!dongle_caps->is_dp_hdmi_ycbcr420_pass_through)
return false;
break;
+ case PIXEL_ENCODING_UNDEFINED:
+ /* These color depths are currently not supported */
+ ASSERT(false);
+ break;
default:
/* Invalid Pixel Encoding*/
return false;
@@ -104,6 +108,10 @@ static bool dp_active_dongle_validate_timing(
if (dongle_caps->dp_hdmi_max_bpc < 12)
return false;
break;
+ case COLOR_DEPTH_UNDEFINED:
+ /* These color depths are currently not supported */
+ ASSERT(false);
+ break;
case COLOR_DEPTH_141414:
case COLOR_DEPTH_161616:
default:
@@ -255,6 +263,14 @@ uint32_t dp_link_bandwidth_kbps(
return link_rate_per_lane_kbps * link_settings->lane_count / 10000 * total_data_bw_efficiency_x10000;
}
+static uint32_t dp_get_timing_bandwidth_kbps(
+ const struct dc_crtc_timing *timing,
+ const struct dc_link *link)
+{
+ return dc_bandwidth_in_kbps_from_timing(timing,
+ dc_link_get_highest_encoding_format(link));
+}
+
static bool dp_validate_mode_timing(
struct dc_link *link,
const struct dc_crtc_timing *timing)
@@ -351,61 +367,258 @@ enum dc_status link_validate_mode_timing(
return DC_OK;
}
+static const struct dc_tunnel_settings *get_dp_tunnel_settings(const struct dc_state *context,
+ const struct dc_stream_state *stream)
+{
+ int i;
+ const struct dc_tunnel_settings *dp_tunnel_settings = NULL;
+
+ for (i = 0; i < MAX_PIPES; i++) {
+ if (context->res_ctx.pipe_ctx[i].stream && (context->res_ctx.pipe_ctx[i].stream == stream)) {
+ dp_tunnel_settings = &context->res_ctx.pipe_ctx[i].link_config.dp_tunnel_settings;
+ break;
+ }
+ }
+
+ return dp_tunnel_settings;
+}
+
/*
- * This function calculates the bandwidth required for the stream timing
- * and aggregates the stream bandwidth for the respective dpia link
- *
- * @stream: pointer to the dc_stream_state struct instance
- * @num_streams: number of streams to be validated
+ * Calculates the DP tunneling bandwidth required for the stream timing
+ * and aggregates the stream bandwidth for the respective DP tunneling link
*
- * return: true if validation is succeeded
+ * return: dc_status
*/
-bool link_validate_dpia_bandwidth(const struct dc_stream_state *stream, const unsigned int num_streams)
+enum dc_status link_validate_dp_tunnel_bandwidth(const struct dc *dc, const struct dc_state *new_ctx)
{
- int bw_needed[MAX_DPIA_NUM] = {0};
- struct dc_link *dpia_link[MAX_DPIA_NUM] = {0};
- int num_dpias = 0;
-
- for (unsigned int i = 0; i < num_streams; ++i) {
- if (stream[i].signal == SIGNAL_TYPE_DISPLAY_PORT) {
- /* new dpia sst stream, check whether it exceeds max dpia */
- if (num_dpias >= MAX_DPIA_NUM)
- return false;
+ struct dc_validation_dpia_set dpia_link_sets[MAX_DPIA_NUM] = { 0 };
+ uint8_t link_count = 0;
+ enum dc_status result = DC_OK;
+
+ // Iterate through streams in the new context
+ for (uint8_t i = 0; (i < MAX_PIPES && i < new_ctx->stream_count); i++) {
+ const struct dc_stream_state *stream = new_ctx->streams[i];
+ const struct dc_link *link;
+ const struct dc_tunnel_settings *dp_tunnel_settings;
+ uint32_t timing_bw;
+
+ if (stream == NULL)
+ continue;
+
+ link = stream->link;
+
+ if (!(link && (stream->signal == SIGNAL_TYPE_DISPLAY_PORT
+ || stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
+ && link->hpd_status))
+ continue;
+
+ dp_tunnel_settings = get_dp_tunnel_settings(new_ctx, stream);
+
+ if ((dp_tunnel_settings == NULL) || (dp_tunnel_settings->should_use_dp_bw_allocation == false))
+ continue;
+
+ timing_bw = dp_get_timing_bandwidth_kbps(&stream->timing, link);
- dpia_link[num_dpias] = stream[i].link;
- bw_needed[num_dpias] = dc_bandwidth_in_kbps_from_timing(&stream[i].timing,
- dc_link_get_highest_encoding_format(dpia_link[num_dpias]));
- num_dpias++;
- } else if (stream[i].signal == SIGNAL_TYPE_DISPLAY_PORT_MST) {
- uint8_t j = 0;
- /* check whether its a known dpia link */
- for (; j < num_dpias; ++j) {
- if (dpia_link[j] == stream[i].link)
- break;
+ // Find an existing entry for this 'link' in 'dpia_link_sets'
+ for (uint8_t j = 0; j < MAX_DPIA_NUM; j++) {
+ bool is_new_slot = false;
+
+ if (dpia_link_sets[j].link == NULL) {
+ is_new_slot = true;
+ link_count++;
+ dpia_link_sets[j].required_bw = 0;
+ dpia_link_sets[j].link = link;
}
- if (j == num_dpias) {
- /* new dpia mst stream, check whether it exceeds max dpia */
- if (num_dpias >= MAX_DPIA_NUM)
- return false;
- else {
- dpia_link[j] = stream[i].link;
- num_dpias++;
- }
+ if (is_new_slot || (dpia_link_sets[j].link == link)) {
+ dpia_link_sets[j].tunnel_settings = dp_tunnel_settings;
+ dpia_link_sets[j].required_bw += timing_bw;
+ break;
}
+ }
+ }
+
+ if (link_count && link_dpia_validate_dp_tunnel_bandwidth(dpia_link_sets, link_count) == false)
+ result = DC_FAIL_DP_TUNNEL_BW_VALIDATE;
+
+ return result;
+}
+
+struct dp_audio_layout_config {
+ uint8_t layouts_per_sample_denom;
+ uint8_t symbols_per_layout;
+ uint8_t max_layouts_per_audio_sdp;
+};
+
+static void get_audio_layout_config(
+ uint32_t channel_count,
+ enum dp_link_encoding encoding,
+ struct dp_audio_layout_config *output)
+{
+ memset(output, 0, sizeof(struct dp_audio_layout_config));
- bw_needed[j] += dc_bandwidth_in_kbps_from_timing(&stream[i].timing,
- dc_link_get_highest_encoding_format(dpia_link[j]));
+ /* Assuming L-PCM audio. Current implementation uses max 1 layout per SDP,
+ * with each layout being the same size (8ch layout).
+ */
+ if (encoding == DP_8b_10b_ENCODING) {
+ if (channel_count == 2) {
+ output->layouts_per_sample_denom = 4;
+ output->symbols_per_layout = 40;
+ output->max_layouts_per_audio_sdp = 1;
+ } else if (channel_count == 8 || channel_count == 6) {
+ output->layouts_per_sample_denom = 1;
+ output->symbols_per_layout = 40;
+ output->max_layouts_per_audio_sdp = 1;
}
+ } else if (encoding == DP_128b_132b_ENCODING) {
+ if (channel_count == 2) {
+ output->layouts_per_sample_denom = 4;
+ output->symbols_per_layout = 10;
+ output->max_layouts_per_audio_sdp = 1;
+ } else if (channel_count == 8 || channel_count == 6) {
+ output->layouts_per_sample_denom = 1;
+ output->symbols_per_layout = 10;
+ output->max_layouts_per_audio_sdp = 1;
+ }
+ }
+}
+
+static uint32_t get_av_stream_map_lane_count(
+ enum dp_link_encoding encoding,
+ enum dc_lane_count lane_count,
+ bool is_mst)
+{
+ uint32_t av_stream_map_lane_count = 0;
+
+ if (encoding == DP_8b_10b_ENCODING) {
+ if (!is_mst)
+ av_stream_map_lane_count = lane_count;
+ else
+ av_stream_map_lane_count = 4;
+ } else if (encoding == DP_128b_132b_ENCODING) {
+ av_stream_map_lane_count = 4;
+ }
+
+ ASSERT(av_stream_map_lane_count != 0);
+
+ return av_stream_map_lane_count;
+}
+
+static uint32_t get_audio_sdp_overhead(
+ enum dp_link_encoding encoding,
+ enum dc_lane_count lane_count,
+ bool is_mst)
+{
+ uint32_t audio_sdp_overhead = 0;
+
+ if (encoding == DP_8b_10b_ENCODING) {
+ if (is_mst)
+ audio_sdp_overhead = 16; /* 4 * 2 + 8 */
+ else
+ audio_sdp_overhead = lane_count * 2 + 8;
+ } else if (encoding == DP_128b_132b_ENCODING) {
+ audio_sdp_overhead = 10; /* 4 x 2.5 */
}
- /* Include dp overheads */
- for (uint8_t i = 0; i < num_dpias; ++i) {
- int dp_overhead = 0;
+ ASSERT(audio_sdp_overhead != 0);
+
+ return audio_sdp_overhead;
+}
+
+/* Current calculation only applicable for 8b/10b MST and 128b/132b SST/MST.
+ */
+static uint32_t calculate_overhead_hblank_bw_in_symbols(
+ uint32_t max_slice_h)
+{
+ uint32_t overhead_hblank_bw = 0; /* in stream symbols */
+
+ overhead_hblank_bw += max_slice_h * 4; /* EOC overhead */
+ overhead_hblank_bw += 12; /* Main link overhead (VBID, BS/BE) */
+
+ return overhead_hblank_bw;
+}
+
+uint32_t dp_required_hblank_size_bytes(
+ const struct dc_link *link,
+ struct dp_audio_bandwidth_params *audio_params)
+{
+ /* Main logic from dce_audio is duplicated here, with the main
+ * difference being:
+ * - Pre-determined lane count of 4
+ * - Assumed 16 dsc slices for worst case
+ * - Assumed SDP split disabled for worst case
+ * TODO: Unify logic from dce_audio to prevent duplicated logic.
+ */
- dp_overhead = link_dp_dpia_get_dp_overhead_in_dp_tunneling(dpia_link[i]);
- bw_needed[i] += dp_overhead;
+ const struct dc_crtc_timing *timing = audio_params->crtc_timing;
+ const uint32_t channel_count = audio_params->channel_count;
+ const uint32_t sample_rate_hz = audio_params->sample_rate_hz;
+ const enum dp_link_encoding link_encoding = audio_params->link_encoding;
+
+ // 8b/10b MST and 128b/132b are always 4 logical lanes.
+ const uint32_t lane_count = 4;
+ const bool is_mst = (link->connector_signal == SIGNAL_TYPE_DISPLAY_PORT);
+ // Maximum slice count is with ODM 4:1, 4 slices per DSC
+ const uint32_t max_slices_h = 16;
+
+ const uint32_t av_stream_map_lane_count = get_av_stream_map_lane_count(
+ link_encoding, lane_count, is_mst);
+ const uint32_t audio_sdp_overhead = get_audio_sdp_overhead(
+ link_encoding, lane_count, is_mst);
+ struct dp_audio_layout_config layout_config;
+
+ if (link_encoding == DP_8b_10b_ENCODING && link->connector_signal == SIGNAL_TYPE_DISPLAY_PORT)
+ return 0;
+
+ get_audio_layout_config(
+ channel_count, link_encoding, &layout_config);
+
+ /* DP spec recommends between 1.05 to 1.1 safety margin to prevent sample under-run */
+ struct fixed31_32 audio_sdp_margin = dc_fixpt_from_fraction(110, 100);
+ struct fixed31_32 horizontal_line_freq_khz = dc_fixpt_from_fraction(
+ timing->pix_clk_100hz, (long long)timing->h_total * 10);
+ struct fixed31_32 samples_per_line;
+ struct fixed31_32 layouts_per_line;
+ struct fixed31_32 symbols_per_sdp_max_layout;
+ struct fixed31_32 remainder;
+ uint32_t num_sdp_with_max_layouts;
+ uint32_t required_symbols_per_hblank;
+ uint32_t required_bytes_per_hblank = 0;
+
+ samples_per_line = dc_fixpt_from_fraction(sample_rate_hz, 1000);
+ samples_per_line = dc_fixpt_div(samples_per_line, horizontal_line_freq_khz);
+ layouts_per_line = dc_fixpt_div_int(samples_per_line, layout_config.layouts_per_sample_denom);
+ // HBlank expansion usage assumes SDP split disabled to allow for worst case.
+ layouts_per_line = dc_fixpt_from_int(dc_fixpt_ceil(layouts_per_line));
+
+ num_sdp_with_max_layouts = dc_fixpt_floor(
+ dc_fixpt_div_int(layouts_per_line, layout_config.max_layouts_per_audio_sdp));
+ symbols_per_sdp_max_layout = dc_fixpt_from_int(
+ layout_config.max_layouts_per_audio_sdp * layout_config.symbols_per_layout);
+ symbols_per_sdp_max_layout = dc_fixpt_add_int(symbols_per_sdp_max_layout, audio_sdp_overhead);
+ symbols_per_sdp_max_layout = dc_fixpt_mul(symbols_per_sdp_max_layout, audio_sdp_margin);
+ required_symbols_per_hblank = num_sdp_with_max_layouts;
+ required_symbols_per_hblank *= ((dc_fixpt_ceil(symbols_per_sdp_max_layout) + av_stream_map_lane_count) /
+ av_stream_map_lane_count) * av_stream_map_lane_count;
+
+ if (num_sdp_with_max_layouts != dc_fixpt_ceil(
+ dc_fixpt_div_int(layouts_per_line, layout_config.max_layouts_per_audio_sdp))) {
+ remainder = dc_fixpt_sub_int(layouts_per_line,
+ num_sdp_with_max_layouts * layout_config.max_layouts_per_audio_sdp);
+ remainder = dc_fixpt_mul_int(remainder, layout_config.symbols_per_layout);
+ remainder = dc_fixpt_add_int(remainder, audio_sdp_overhead);
+ remainder = dc_fixpt_mul(remainder, audio_sdp_margin);
+ required_symbols_per_hblank += ((dc_fixpt_ceil(remainder) + av_stream_map_lane_count) /
+ av_stream_map_lane_count) * av_stream_map_lane_count;
}
- return dpia_validate_usb4_bw(dpia_link, bw_needed, num_dpias);
+ required_symbols_per_hblank += calculate_overhead_hblank_bw_in_symbols(max_slices_h);
+
+ if (link_encoding == DP_8b_10b_ENCODING)
+ required_bytes_per_hblank = required_symbols_per_hblank; // 8 bits per 8b/10b symbol
+ else if (link_encoding == DP_128b_132b_ENCODING)
+ required_bytes_per_hblank = required_symbols_per_hblank * 4; // 32 bits per 128b/132b symbol
+
+ return required_bytes_per_hblank;
}
+
diff --git a/drivers/gpu/drm/amd/display/dc/link/link_validation.h b/drivers/gpu/drm/amd/display/dc/link/link_validation.h
index 595fb05946e9..9553c81053fe 100644
--- a/drivers/gpu/drm/amd/display/dc/link/link_validation.h
+++ b/drivers/gpu/drm/amd/display/dc/link/link_validation.h
@@ -30,11 +30,16 @@ enum dc_status link_validate_mode_timing(
const struct dc_stream_state *stream,
struct dc_link *link,
const struct dc_crtc_timing *timing);
-bool link_validate_dpia_bandwidth(
- const struct dc_stream_state *stream,
- const unsigned int num_streams);
+enum dc_status link_validate_dp_tunnel_bandwidth(
+ const struct dc *dc,
+ const struct dc_state *new_ctx);
uint32_t dp_link_bandwidth_kbps(
const struct dc_link *link,
const struct dc_link_settings *link_settings);
+
+uint32_t dp_required_hblank_size_bytes(
+ const struct dc_link *link,
+ struct dp_audio_bandwidth_params *audio_params);
+
#endif /* __LINK_VALIDATION_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_ddc.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_ddc.c
index d6d5bbf2108c..267180e7bc48 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_ddc.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_ddc.c
@@ -505,7 +505,7 @@ bool try_to_configure_aux_timeout(struct ddc_service *ddc,
bool result = false;
struct ddc *ddc_pin = ddc->ddc_pin;
- if ((ddc->link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
+ if (((ddc->link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
!ddc->link->dc->debug.disable_fixed_vs_aux_timeout_wa &&
ddc->ctx->dce_version == DCN_VERSION_3_1) {
/* Fixed VS workaround for AUX timeout */
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c
index 9dabaf682171..651926e547b9 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.c
@@ -158,6 +158,14 @@ uint8_t dp_parse_lttpr_repeater_count(uint8_t lttpr_repeater_count)
return 0; // invalid value
}
+uint32_t dp_get_closest_lttpr_offset(uint8_t lttpr_count)
+{
+ /* Calculate offset for LTTPR closest to DPTX which is highest in the chain
+ * Offset is 0 for single LTTPR cases as base LTTPR DPCD addresses target LTTPR 1
+ */
+ return DP_REPEATER_CONFIGURATION_AND_STATUS_SIZE * (lttpr_count - 1);
+}
+
uint32_t link_bw_kbps_from_raw_frl_link_rate_data(uint8_t bw)
{
switch (bw) {
@@ -250,21 +258,23 @@ static uint32_t intersect_frl_link_bw_support(
{
uint32_t supported_bw_in_kbps = max_supported_frl_bw_in_kbps;
- // HDMI_ENCODED_LINK_BW bits are only valid if HDMI Link Configuration bit is 1 (FRL mode)
- if (hdmi_encoded_link_bw.bits.FRL_MODE) {
- if (hdmi_encoded_link_bw.bits.BW_48Gbps)
- supported_bw_in_kbps = 48000000;
- else if (hdmi_encoded_link_bw.bits.BW_40Gbps)
- supported_bw_in_kbps = 40000000;
- else if (hdmi_encoded_link_bw.bits.BW_32Gbps)
- supported_bw_in_kbps = 32000000;
- else if (hdmi_encoded_link_bw.bits.BW_24Gbps)
- supported_bw_in_kbps = 24000000;
- else if (hdmi_encoded_link_bw.bits.BW_18Gbps)
- supported_bw_in_kbps = 18000000;
- else if (hdmi_encoded_link_bw.bits.BW_9Gbps)
- supported_bw_in_kbps = 9000000;
- }
+ /* Skip checking FRL_MODE bit, as certain PCON will clear
+ * it despite supporting the link BW indicated in the other bits.
+ */
+ if (hdmi_encoded_link_bw.bits.BW_48Gbps)
+ supported_bw_in_kbps = 48000000;
+ else if (hdmi_encoded_link_bw.bits.BW_40Gbps)
+ supported_bw_in_kbps = 40000000;
+ else if (hdmi_encoded_link_bw.bits.BW_32Gbps)
+ supported_bw_in_kbps = 32000000;
+ else if (hdmi_encoded_link_bw.bits.BW_24Gbps)
+ supported_bw_in_kbps = 24000000;
+ else if (hdmi_encoded_link_bw.bits.BW_18Gbps)
+ supported_bw_in_kbps = 18000000;
+ else if (hdmi_encoded_link_bw.bits.BW_9Gbps)
+ supported_bw_in_kbps = 9000000;
+ else if (hdmi_encoded_link_bw.bits.FRL_LINK_TRAINING_FINISHED)
+ supported_bw_in_kbps = 0; /* This case should only get hit in regulated autonomous mode. */
return supported_bw_in_kbps;
}
@@ -330,9 +340,12 @@ bool dp_is_fec_supported(const struct dc_link *link)
/* TODO - use asic cap instead of link_enc->features
* we no longer know which link enc to use for this link before commit
*/
- struct link_encoder *link_enc = NULL;
+ struct resource_context *res_ctx = &link->dc->current_state->res_ctx;
+ struct resource_pool *res_pool = link->dc->res_pool;
+ struct link_encoder *link_enc = get_temp_dio_link_enc(res_ctx, res_pool, link);
- link_enc = link_enc_cfg_get_link_enc(link);
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
ASSERT(link_enc);
return (dc_is_dp_signal(link->connector_signal) && link_enc &&
@@ -372,9 +385,15 @@ bool dp_is_128b_132b_signal(struct pipe_ctx *pipe_ctx)
bool dp_is_lttpr_present(struct dc_link *link)
{
/* Some sink devices report invalid LTTPR revision, so don't validate against that cap */
- return (dp_parse_lttpr_repeater_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt) != 0 &&
+ uint32_t lttpr_count = dp_parse_lttpr_repeater_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt);
+ bool is_lttpr_present = (lttpr_count > 0 &&
link->dpcd_caps.lttpr_caps.max_lane_count > 0 &&
link->dpcd_caps.lttpr_caps.max_lane_count <= 4);
+
+ if (lttpr_count > 0 && !is_lttpr_present)
+ DC_LOG_ERROR("LTTPR count is nonzero but invalid lane count reported. Assuming no LTTPR present.\n");
+
+ return is_lttpr_present;
}
/* in DP compliance test, DPR-120 may have
@@ -945,6 +964,9 @@ bool link_decide_link_settings(struct dc_stream_state *stream,
* TODO: add MST specific link training routine
*/
decide_mst_link_settings(link, link_setting);
+ } else if (stream->signal == SIGNAL_TYPE_VIRTUAL) {
+ link_setting->lane_count = LANE_COUNT_FOUR;
+ link_setting->link_rate = LINK_RATE_HIGH3;
} else if (link->connector_signal == SIGNAL_TYPE_EDP) {
/* enable edp link optimization for DSC eDP case */
if (stream->timing.flags.DSC) {
@@ -967,9 +989,6 @@ bool link_decide_link_settings(struct dc_stream_state *stream,
} else {
edp_decide_link_settings(link, link_setting, req_bw);
}
- } else if (stream->signal == SIGNAL_TYPE_VIRTUAL) {
- link_setting->lane_count = LANE_COUNT_FOUR;
- link_setting->link_rate = LINK_RATE_HIGH3;
} else {
decide_dp_link_settings(link, link_setting, req_bw);
}
@@ -1072,6 +1091,48 @@ static enum dc_status wake_up_aux_channel(struct dc_link *link)
return DC_OK;
}
+static void read_and_intersect_post_frl_lt_status(
+ struct dc_link *link)
+{
+ union autonomous_mode_and_frl_link_status autonomous_mode_caps = {0};
+ union hdmi_tx_link_status hdmi_tx_link_status = {0};
+ union hdmi_encoded_link_bw hdmi_encoded_link_bw = {0};
+
+ /* Check if dongle supports regulated autonomous mode. */
+ core_link_read_dpcd(link, DP_REGULATED_AUTONOMOUS_MODE_SUPPORTED_AND_HDMI_LINK_TRAINING_STATUS,
+ &autonomous_mode_caps.raw, sizeof(autonomous_mode_caps));
+
+ link->dpcd_caps.dongle_caps.dp_hdmi_regulated_autonomous_mode_support =
+ autonomous_mode_caps.bits.REGULATED_AUTONOMOUS_MODE_SUPPORTED;
+
+ if (link->dpcd_caps.dongle_caps.dp_hdmi_regulated_autonomous_mode_support) {
+ DC_LOG_DC("%s: PCON supports regulated autonomous mode.\n", __func__);
+
+ core_link_read_dpcd(link, DP_PCON_HDMI_TX_LINK_STATUS,
+ &hdmi_tx_link_status.raw, sizeof(hdmi_tx_link_status));
+ }
+
+ // Intersect reported max link bw support with the supported link rate post FRL link training
+ if (core_link_read_dpcd(link, DP_PCON_HDMI_POST_FRL_STATUS,
+ &hdmi_encoded_link_bw.raw, sizeof(hdmi_encoded_link_bw)) == DC_OK) {
+
+ if (link->dpcd_caps.dongle_caps.dp_hdmi_regulated_autonomous_mode_support &&
+ (!hdmi_tx_link_status.bits.HDMI_TX_READY_STATUS ||
+ !hdmi_encoded_link_bw.bits.FRL_LINK_TRAINING_FINISHED)) {
+ DC_LOG_WARNING("%s: PCON TX link training has not finished.\n", __func__);
+
+ /* Link training not finished, ignore values from this DPCD reg. */
+ return;
+ }
+
+ link->dpcd_caps.dongle_caps.dp_hdmi_frl_max_link_bw_in_kbps = intersect_frl_link_bw_support(
+ link->dpcd_caps.dongle_caps.dp_hdmi_frl_max_link_bw_in_kbps,
+ hdmi_encoded_link_bw);
+ DC_LOG_DC("%s: pcon frl link bw = %u\n", __func__,
+ link->dpcd_caps.dongle_caps.dp_hdmi_frl_max_link_bw_in_kbps);
+ }
+}
+
static void get_active_converter_info(
uint8_t data, struct dc_link *link)
{
@@ -1160,21 +1221,12 @@ static void get_active_converter_info(
hdmi_color_caps.bits.MAX_BITS_PER_COLOR_COMPONENT);
if (link->dc->caps.dp_hdmi21_pcon_support) {
- union hdmi_encoded_link_bw hdmi_encoded_link_bw;
link->dpcd_caps.dongle_caps.dp_hdmi_frl_max_link_bw_in_kbps =
link_bw_kbps_from_raw_frl_link_rate_data(
hdmi_color_caps.bits.MAX_ENCODED_LINK_BW_SUPPORT);
- // Intersect reported max link bw support with the supported link rate post FRL link training
- if (core_link_read_dpcd(link, DP_PCON_HDMI_POST_FRL_STATUS,
- &hdmi_encoded_link_bw.raw, sizeof(hdmi_encoded_link_bw)) == DC_OK) {
- link->dpcd_caps.dongle_caps.dp_hdmi_frl_max_link_bw_in_kbps = intersect_frl_link_bw_support(
- link->dpcd_caps.dongle_caps.dp_hdmi_frl_max_link_bw_in_kbps,
- hdmi_encoded_link_bw);
- DC_LOG_DC("%s: pcon frl link bw = %u\n", __func__,
- link->dpcd_caps.dongle_caps.dp_hdmi_frl_max_link_bw_in_kbps);
- }
+ read_and_intersect_post_frl_lt_status(link);
if (link->dpcd_caps.dongle_caps.dp_hdmi_frl_max_link_bw_in_kbps > 0)
link->dpcd_caps.dongle_caps.extendedCapValid = true;
@@ -1336,6 +1388,21 @@ void dpcd_set_source_specific_data(struct dc_link *link)
struct dpcd_amd_signature amd_signature = {0};
struct dpcd_amd_device_id amd_device_id = {0};
+ if (link->is_dds) {
+ uint8_t dpcd_dp_edp_backlight_mode = 0;
+
+ /*
+ * Write 0 to bits 0:1 for dp_edp_backlight_mode_set register
+ * if platform is DDS
+ */
+ core_link_read_dpcd(link, DP_EDP_BACKLIGHT_MODE_SET_REGISTER,
+ &dpcd_dp_edp_backlight_mode, sizeof(uint8_t));
+ dpcd_dp_edp_backlight_mode &= ~0x3;
+
+ core_link_write_dpcd(link, DP_EDP_BACKLIGHT_MODE_SET_REGISTER,
+ &dpcd_dp_edp_backlight_mode, sizeof(uint8_t));
+ }
+
amd_device_id.device_id_byte1 =
(uint8_t)(link->ctx->asic_id.chip_id);
amd_device_id.device_id_byte2 =
@@ -1491,6 +1558,10 @@ static bool dpcd_read_sink_ext_caps(struct dc_link *link)
return false;
link->dpcd_sink_ext_caps.raw = dpcd_data;
+ if (link->is_dds && !link->dpcd_sink_ext_caps.bits.oled) {
+ link->dpcd_sink_ext_caps.raw = 0;
+ return false;
+ }
if (core_link_read_dpcd(link, DP_EDP_GENERAL_CAP_2, &edp_general_cap2, 1) != DC_OK)
return false;
@@ -1502,9 +1573,11 @@ static bool dpcd_read_sink_ext_caps(struct dc_link *link)
enum dc_status dp_retrieve_lttpr_cap(struct dc_link *link)
{
- uint8_t lttpr_dpcd_data[8] = {0};
+ uint8_t lttpr_dpcd_data[10] = {0};
enum dc_status status;
bool is_lttpr_present;
+ uint32_t lttpr_count;
+ uint32_t closest_lttpr_offset;
/* Logic to determine LTTPR support*/
bool vbios_lttpr_interop = link->dc->caps.vbios_lttpr_aware;
@@ -1552,26 +1625,54 @@ enum dc_status dp_retrieve_lttpr_cap(struct dc_link *link)
lttpr_dpcd_data[DP_PHY_REPEATER_128B132B_RATES -
DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV];
+ link->dpcd_caps.lttpr_caps.alpm.raw =
+ lttpr_dpcd_data[DP_LTTPR_ALPM_CAPABILITIES -
+ DP_LT_TUNABLE_PHY_REPEATER_FIELD_DATA_STRUCTURE_REV];
+
+ lttpr_count = dp_parse_lttpr_repeater_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt);
+
/* If this chip cap is set, at least one retimer must exist in the chain
* Override count to 1 if we receive a known bad count (0 or an invalid value) */
- if ((link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
- (dp_parse_lttpr_repeater_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt) == 0)) {
+ if (((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
+ lttpr_count == 0) {
/* If you see this message consistently, either the host platform has FIXED_VS flag
* incorrectly configured or the sink device is returning an invalid count.
*/
DC_LOG_ERROR("lttpr_caps phy_repeater_cnt is 0x%x, forcing it to 0x80.",
link->dpcd_caps.lttpr_caps.phy_repeater_cnt);
link->dpcd_caps.lttpr_caps.phy_repeater_cnt = 0x80;
+ lttpr_count = 1;
DC_LOG_DC("lttpr_caps forced phy_repeater_cnt = %d\n", link->dpcd_caps.lttpr_caps.phy_repeater_cnt);
}
- /* Attempt to train in LTTPR transparent mode if repeater count exceeds 8. */
is_lttpr_present = dp_is_lttpr_present(link);
- if (is_lttpr_present)
+ DC_LOG_DC("is_lttpr_present = %d\n", is_lttpr_present);
+
+ if (is_lttpr_present) {
CONN_DATA_DETECT(link, lttpr_dpcd_data, sizeof(lttpr_dpcd_data), "LTTPR Caps: ");
- DC_LOG_DC("is_lttpr_present = %d\n", is_lttpr_present);
+ // Identify closest LTTPR to determine if workarounds required for known embedded LTTPR
+ closest_lttpr_offset = dp_get_closest_lttpr_offset(lttpr_count);
+
+ core_link_read_dpcd(link, (DP_LTTPR_IEEE_OUI + closest_lttpr_offset),
+ link->dpcd_caps.lttpr_caps.lttpr_ieee_oui, sizeof(link->dpcd_caps.lttpr_caps.lttpr_ieee_oui));
+ core_link_read_dpcd(link, (DP_LTTPR_DEVICE_ID + closest_lttpr_offset),
+ link->dpcd_caps.lttpr_caps.lttpr_device_id, sizeof(link->dpcd_caps.lttpr_caps.lttpr_device_id));
+
+ if (lttpr_count > 1) {
+ CONN_DATA_DETECT(link, link->dpcd_caps.lttpr_caps.lttpr_ieee_oui, sizeof(link->dpcd_caps.lttpr_caps.lttpr_ieee_oui),
+ "Closest LTTPR To Host's IEEE OUI: ");
+ CONN_DATA_DETECT(link, link->dpcd_caps.lttpr_caps.lttpr_device_id, sizeof(link->dpcd_caps.lttpr_caps.lttpr_device_id),
+ "Closest LTTPR To Host's LTTPR Device ID: ");
+ } else {
+ CONN_DATA_DETECT(link, link->dpcd_caps.lttpr_caps.lttpr_ieee_oui, sizeof(link->dpcd_caps.lttpr_caps.lttpr_ieee_oui),
+ "LTTPR IEEE OUI: ");
+ CONN_DATA_DETECT(link, link->dpcd_caps.lttpr_caps.lttpr_device_id, sizeof(link->dpcd_caps.lttpr_caps.lttpr_device_id),
+ "LTTPR Device ID: ");
+ }
+ }
+
return status;
}
@@ -1632,13 +1733,6 @@ static bool retrieve_link_cap(struct dc_link *link)
sizeof(link->dpcd_caps.lttpr_caps.phy_repeater_cnt));
}
- /* Read DP tunneling information. */
- if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA) {
- status = dpcd_get_tunneling_device_data(link);
- if (status != DC_OK)
- dm_error("%s: Read tunneling device data failed.\n", __func__);
- }
-
dpcd_set_source_specific_data(link);
/* Sink may need to configure internals based on vendor, so allow some
* time before proceeding with possibly vendor specific transactions
@@ -1711,7 +1805,7 @@ static bool retrieve_link_cap(struct dc_link *link)
link->dpcd_caps.dprx_feature.raw = dpcd_dprx_data;
if (status != DC_OK)
- dm_error("%s: Read DPRX caps data failed.\n", __func__);
+ dm_error("%s: Read DPRX feature list failed.\n", __func__);
/* AdaptiveSyncCapability */
dpcd_dprx_data = 0;
@@ -1726,15 +1820,13 @@ static bool retrieve_link_cap(struct dc_link *link)
link->dpcd_caps.adaptive_sync_caps.dp_adap_sync_caps.raw = dpcd_dprx_data;
if (status != DC_OK)
- dm_error("%s: Read DPRX caps data failed. Addr:%#x\n",
+ dm_error("%s: Read DPRX feature list_1 failed. Addr:%#x\n",
__func__, DP_DPRX_FEATURE_ENUMERATION_LIST_CONT_1);
}
-
else {
link->dpcd_caps.dprx_feature.raw = 0;
}
-
/* Error condition checking...
* It is impossible for Sink to report Max Lane Count = 0.
* It is possible for Sink to report Max Link Rate = 0, if it is
@@ -1788,6 +1880,11 @@ static bool retrieve_link_cap(struct dc_link *link)
link->test_pattern_enabled = false;
link->compliance_test_state.raw = 0;
+ link->dpcd_caps.receive_port0_cap.raw[0] =
+ dpcd_data[DP_RECEIVE_PORT_0_CAP_0 - DP_DPCD_REV];
+ link->dpcd_caps.receive_port0_cap.raw[1] =
+ dpcd_data[DP_RECEIVE_PORT_0_BUFFER_SIZE - DP_DPCD_REV];
+
/* read sink count */
core_link_read_dpcd(link,
DP_SINK_COUNT,
@@ -1918,6 +2015,7 @@ static bool retrieve_link_cap(struct dc_link *link)
if (link->dpcd_caps.channel_coding_cap.bits.DP_128b_132b_SUPPORTED) {
DC_LOG_DP2("128b/132b encoding is supported at link %d", link->link_index);
+ /* Read 128b/132b suppoerted link rates */
core_link_read_dpcd(link,
DP_128B132B_SUPPORTED_LINK_RATES,
&link->dpcd_caps.dp_128b_132b_supported_link_rates.raw,
@@ -1965,6 +2063,11 @@ static bool retrieve_link_cap(struct dc_link *link)
link->dpcd_caps.max_uncompressed_pixel_rate_cap.raw,
sizeof(link->dpcd_caps.max_uncompressed_pixel_rate_cap.raw));
+ /* Read DP tunneling information. */
+ status = dpcd_get_tunneling_device_data(link);
+ if (status != DC_OK)
+ DC_LOG_DP2("%s: Read DP tunneling device data failed.\n", __func__);
+
retrieve_cable_id(link);
dpcd_write_cable_id_to_dprx(link);
@@ -2081,18 +2184,32 @@ void detect_edp_sink_caps(struct dc_link *link)
core_link_read_dpcd(link, DP_SINK_EMISSION_RATE,
(uint8_t *)&link->dpcd_caps.edp_oled_emission_rate,
sizeof(link->dpcd_caps.edp_oled_emission_rate));
+
+ /*
+ * Read Multi-SST (Single Stream Transport) capability
+ * for eDP version 1.4 or higher.
+ */
+ if (link->dpcd_caps.dpcd_rev.raw >= DP_EDP_14)
+ core_link_read_dpcd(
+ link,
+ DP_EDP_MSO_LINK_CAPABILITIES,
+ (uint8_t *)&link->dpcd_caps.mso_cap_sst_links_supported,
+ sizeof(link->dpcd_caps.mso_cap_sst_links_supported));
}
bool dp_get_max_link_enc_cap(const struct dc_link *link, struct dc_link_settings *max_link_enc_cap)
{
- struct link_encoder *link_enc = NULL;
+ struct resource_context *res_ctx = &link->dc->current_state->res_ctx;
+ struct resource_pool *res_pool = link->dc->res_pool;
+ struct link_encoder *link_enc = get_temp_dio_link_enc(res_ctx, res_pool, link);
if (!max_link_enc_cap) {
DC_LOG_ERROR("%s: Could not return max link encoder caps", __func__);
return false;
}
- link_enc = link_enc_cfg_get_link_enc(link);
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
ASSERT(link_enc);
if (link_enc && link_enc->funcs->get_max_link_cap) {
@@ -2120,10 +2237,13 @@ struct dc_link_settings dp_get_max_link_cap(struct dc_link *link)
struct dc_link_settings max_link_cap = {0};
enum dc_link_rate lttpr_max_link_rate;
enum dc_link_rate cable_max_link_rate;
- struct link_encoder *link_enc = NULL;
+ struct resource_context *res_ctx = &link->dc->current_state->res_ctx;
+ struct resource_pool *res_pool = link->dc->res_pool;
+ struct link_encoder *link_enc = get_temp_dio_link_enc(res_ctx, res_pool, link);
bool is_uhbr13_5_supported = true;
- link_enc = link_enc_cfg_get_link_enc(link);
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
ASSERT(link_enc);
/* get max link encoder capability */
@@ -2308,6 +2428,14 @@ bool dp_verify_link_cap_with_retries(
} else {
link->verified_link_cap = last_verified_link_cap;
}
+
+ /* For Dp tunneling link, a pending HPD means that we have a race condition between processing
+ * current link and processing the pending HPD. Since the training is failed, we should just brak
+ * the loop so that we have chance to process the pending HPD.
+ */
+ if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA && link->is_hpd_pending)
+ break;
+
fsleep(10 * 1000);
}
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.h
index 8f0ce97f2362..940b147cc5d4 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_capability.h
@@ -48,6 +48,9 @@ enum dc_status dp_retrieve_lttpr_cap(struct dc_link *link);
/* Convert PHY repeater count read from DPCD uint8_t. */
uint8_t dp_parse_lttpr_repeater_count(uint8_t lttpr_repeater_count);
+/* Calculate embedded LTTPR address offset for vendor-specific behaviour */
+uint32_t dp_get_closest_lttpr_offset(uint8_t lttpr_count);
+
bool dp_is_sink_present(struct dc_link *link);
bool dp_is_lttpr_present(struct dc_link *link);
@@ -67,6 +70,7 @@ bool dp_is_128b_132b_signal(struct pipe_ctx *pipe_ctx);
/* Initialize output parameter lt_settings. */
void dp_decide_training_settings(
struct dc_link *link,
+ const struct link_resource *link_res,
const struct dc_link_settings *link_setting,
struct link_training_settings *lt_settings);
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.c
index 0d123e647652..9b2f1a7da1d1 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.c
@@ -62,6 +62,39 @@ enum dc_status dpcd_get_tunneling_device_data(struct dc_link *link)
if (status != DC_OK)
goto err;
+ link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.raw =
+ dpcd_dp_tun_data[DP_TUNNELING_CAPABILITIES_SUPPORT - DP_TUNNELING_CAPABILITIES_SUPPORT];
+
+ if (link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dp_tunneling == false)
+ goto err;
+
+ link->dpcd_caps.usb4_dp_tun_info.dpia_info.raw =
+ dpcd_dp_tun_data[DP_IN_ADAPTER_INFO - DP_TUNNELING_CAPABILITIES_SUPPORT];
+ link->dpcd_caps.usb4_dp_tun_info.usb4_driver_id =
+ dpcd_dp_tun_data[DP_USB4_DRIVER_ID - DP_TUNNELING_CAPABILITIES_SUPPORT];
+
+ if (link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dpia_bw_alloc) {
+ status = core_link_read_dpcd(link, USB4_DRIVER_BW_CAPABILITY,
+ dpcd_dp_tun_data, 2);
+
+ if (status != DC_OK)
+ goto err;
+
+ link->dpcd_caps.usb4_dp_tun_info.driver_bw_cap.raw =
+ dpcd_dp_tun_data[USB4_DRIVER_BW_CAPABILITY - USB4_DRIVER_BW_CAPABILITY];
+ link->dpcd_caps.usb4_dp_tun_info.dpia_tunnel_info.raw =
+ dpcd_dp_tun_data[DP_IN_ADAPTER_TUNNEL_INFO - USB4_DRIVER_BW_CAPABILITY];
+ }
+
+ DC_LOG_DEBUG("%s: Link[%d] DP tunneling support (RouterId=%d AdapterId=%d) "
+ "DPIA_BW_Alloc_support=%d "
+ "CM_BW_Alloc_support=%d ",
+ __func__, link->link_index,
+ link->dpcd_caps.usb4_dp_tun_info.usb4_driver_id,
+ link->dpcd_caps.usb4_dp_tun_info.dpia_info.bits.dpia_num,
+ link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dpia_bw_alloc,
+ link->dpcd_caps.usb4_dp_tun_info.driver_bw_cap.bits.driver_bw_alloc_support);
+
status = core_link_read_dpcd(
link,
DP_USB4_ROUTER_TOPOLOGY_ID,
@@ -71,13 +104,6 @@ enum dc_status dpcd_get_tunneling_device_data(struct dc_link *link)
if (status != DC_OK)
goto err;
- link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.raw =
- dpcd_dp_tun_data[DP_TUNNELING_CAPABILITIES_SUPPORT - DP_TUNNELING_CAPABILITIES_SUPPORT];
- link->dpcd_caps.usb4_dp_tun_info.dpia_info.raw =
- dpcd_dp_tun_data[DP_IN_ADAPTER_INFO - DP_TUNNELING_CAPABILITIES_SUPPORT];
- link->dpcd_caps.usb4_dp_tun_info.usb4_driver_id =
- dpcd_dp_tun_data[DP_USB4_DRIVER_ID - DP_TUNNELING_CAPABILITIES_SUPPORT];
-
for (i = 0; i < DPCD_USB4_TOPOLOGY_ID_LEN; i++)
link->dpcd_caps.usb4_dp_tun_info.usb4_topology_id[i] = dpcd_topology_data[i];
@@ -92,6 +118,7 @@ bool dpia_query_hpd_status(struct dc_link *link)
/* prepare QUERY_HPD command */
cmd.query_hpd.header.type = DMUB_CMD__QUERY_HPD_STATE;
+ cmd.query_hpd.header.payload_bytes = sizeof(cmd.query_hpd.data);
cmd.query_hpd.data.instance = link->link_id.enum_id - ENUM_ID_1;
cmd.query_hpd.data.ch_type = AUX_CHANNEL_DPIA;
@@ -119,3 +146,26 @@ bool dpia_query_hpd_status(struct dc_link *link)
return link->hpd_status;
}
+void link_decide_dp_tunnel_settings(struct dc_stream_state *stream,
+ struct dc_tunnel_settings *dp_tunnel_setting)
+{
+ struct dc_link *link = stream->link;
+
+ memset(dp_tunnel_setting, 0, sizeof(*dp_tunnel_setting));
+
+ if ((stream->signal == SIGNAL_TYPE_DISPLAY_PORT) || (stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)) {
+ dp_tunnel_setting->should_enable_dp_tunneling =
+ link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dp_tunneling;
+
+ if (link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dpia_bw_alloc
+ && link->dpcd_caps.usb4_dp_tun_info.driver_bw_cap.bits.driver_bw_alloc_support) {
+ dp_tunnel_setting->should_use_dp_bw_allocation = true;
+ dp_tunnel_setting->cm_id = link->dpcd_caps.usb4_dp_tun_info.usb4_driver_id & 0x0F;
+ dp_tunnel_setting->group_id = link->dpcd_caps.usb4_dp_tun_info.dpia_tunnel_info.bits.group_id;
+ dp_tunnel_setting->estimated_bw = link->dpia_bw_alloc_config.estimated_bw;
+ dp_tunnel_setting->allocated_bw = link->dpia_bw_alloc_config.allocated_bw;
+ dp_tunnel_setting->bw_granularity = link->dpia_bw_alloc_config.bw_granularity;
+ }
+ }
+}
+
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.h
index 363f45a1a964..a61edfc9ca7a 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia.h
@@ -38,4 +38,10 @@ enum dc_status dpcd_get_tunneling_device_data(struct dc_link *link);
* Returns true if HPD high.
*/
bool dpia_query_hpd_status(struct dc_link *link);
+
+/* Decide the DP tunneling settings based on the DPCD capabilities
+ */
+void link_decide_dp_tunnel_settings(struct dc_stream_state *stream,
+ struct dc_tunnel_settings *dp_tunnel_setting);
+
#endif /* __DC_LINK_DPIA_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.c
index 0f1c411523a2..819bf2d8ba53 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.c
@@ -24,7 +24,7 @@
*
*/
/*********************************************************************/
-// USB4 DPIA BANDWIDTH ALLOCATION LOGIC
+// USB4 DPIA BANDWIDTH ALLOCATION LOGIC
/*********************************************************************/
#include "link_dp_dpia_bw.h"
#include "link_dpcd.h"
@@ -35,8 +35,10 @@
#define Kbps_TO_Gbps (1000 * 1000)
+#define MST_TIME_SLOT_COUNT 64
+
// ------------------------------------------------------------------
-// PRIVATE FUNCTIONS
+// PRIVATE FUNCTIONS
// ------------------------------------------------------------------
/*
* Always Check the following:
@@ -44,11 +46,12 @@
* - Is HPD HIGH?
* - Is BW Allocation Support Mode enabled on DP-Tx?
*/
-static bool get_bw_alloc_proceed_flag(struct dc_link *tmp)
+static bool link_dp_is_bw_alloc_available(struct dc_link *link)
{
- return (tmp && DISPLAY_ENDPOINT_USB4_DPIA == tmp->ep_type
- && tmp->hpd_status
- && tmp->dpia_bw_alloc_config.bw_alloc_enabled);
+ return (link && link->hpd_status
+ && link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dp_tunneling
+ && link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dpia_bw_alloc
+ && link->dpcd_caps.usb4_dp_tun_info.driver_bw_cap.bits.driver_bw_alloc_support);
}
static void reset_bw_alloc_struct(struct dc_link *link)
@@ -60,7 +63,6 @@ static void reset_bw_alloc_struct(struct dc_link *link)
link->dpia_bw_alloc_config.estimated_bw = 0;
link->dpia_bw_alloc_config.bw_granularity = 0;
link->dpia_bw_alloc_config.dp_overhead = 0;
- link->dpia_bw_alloc_config.response_ready = false;
link->dpia_bw_alloc_config.nrd_max_lane_count = 0;
link->dpia_bw_alloc_config.nrd_max_link_rate = 0;
for (int i = 0; i < MAX_SINKS_PER_LINK; i++)
@@ -142,7 +144,7 @@ static int get_non_reduced_max_lane_count(struct dc_link *link)
* granuality, Driver_ID, CM_Group, & populate the BW allocation structs
* for host router and dpia
*/
-static void init_usb4_bw_struct(struct dc_link *link)
+static void retrieve_usb4_dp_bw_allocation_info(struct dc_link *link)
{
reset_bw_alloc_struct(link);
@@ -160,78 +162,6 @@ static void init_usb4_bw_struct(struct dc_link *link)
link->dpia_bw_alloc_config.nrd_max_lane_count);
}
-static uint8_t get_lowest_dpia_index(struct dc_link *link)
-{
- const struct dc *dc_struct = link->dc;
- uint8_t idx = 0xFF;
- int i;
-
- for (i = 0; i < MAX_LINKS; ++i) {
-
- if (!dc_struct->links[i] ||
- dc_struct->links[i]->ep_type != DISPLAY_ENDPOINT_USB4_DPIA)
- continue;
-
- if (idx > dc_struct->links[i]->link_index) {
- idx = dc_struct->links[i]->link_index;
- break;
- }
- }
-
- return idx;
-}
-
-/*
- * Get the maximum dp tunnel banwidth of host router
- *
- * @dc: pointer to the dc struct instance
- * @hr_index: host router index
- *
- * return: host router maximum dp tunnel bandwidth
- */
-static int get_host_router_total_dp_tunnel_bw(const struct dc *dc, uint8_t hr_index)
-{
- uint8_t lowest_dpia_index = get_lowest_dpia_index(dc->links[0]);
- uint8_t hr_index_temp = 0;
- struct dc_link *link_dpia_primary, *link_dpia_secondary;
- int total_bw = 0;
-
- for (uint8_t i = 0; i < MAX_LINKS - 1; ++i) {
-
- if (!dc->links[i] || dc->links[i]->ep_type != DISPLAY_ENDPOINT_USB4_DPIA)
- continue;
-
- hr_index_temp = (dc->links[i]->link_index - lowest_dpia_index) / 2;
-
- if (hr_index_temp == hr_index) {
- link_dpia_primary = dc->links[i];
- link_dpia_secondary = dc->links[i + 1];
-
- /**
- * If BW allocation enabled on both DPIAs, then
- * HR BW = Estimated(dpia_primary) + Allocated(dpia_secondary)
- * otherwise HR BW = Estimated(bw alloc enabled dpia)
- */
- if ((link_dpia_primary->hpd_status &&
- link_dpia_primary->dpia_bw_alloc_config.bw_alloc_enabled) &&
- (link_dpia_secondary->hpd_status &&
- link_dpia_secondary->dpia_bw_alloc_config.bw_alloc_enabled)) {
- total_bw += link_dpia_primary->dpia_bw_alloc_config.estimated_bw +
- link_dpia_secondary->dpia_bw_alloc_config.allocated_bw;
- } else if (link_dpia_primary->hpd_status &&
- link_dpia_primary->dpia_bw_alloc_config.bw_alloc_enabled) {
- total_bw = link_dpia_primary->dpia_bw_alloc_config.estimated_bw;
- } else if (link_dpia_secondary->hpd_status &&
- link_dpia_secondary->dpia_bw_alloc_config.bw_alloc_enabled) {
- total_bw += link_dpia_secondary->dpia_bw_alloc_config.estimated_bw;
- }
- break;
- }
- }
-
- return total_bw;
-}
-
/*
* Cleanup function for when the dpia is unplugged to reset struct
* and perform any required clean up
@@ -243,323 +173,269 @@ static int get_host_router_total_dp_tunnel_bw(const struct dc *dc, uint8_t hr_in
static void dpia_bw_alloc_unplug(struct dc_link *link)
{
if (link) {
- DC_LOG_DEBUG("%s: resetting bw alloc config for link(%d)\n",
+ DC_LOG_DEBUG("%s: resetting BW alloc config for link(%d)\n",
__func__, link->link_index);
reset_bw_alloc_struct(link);
}
}
-static void set_usb4_req_bw_req(struct dc_link *link, int req_bw)
+static void link_dpia_send_bw_alloc_request(struct dc_link *link, int req_bw)
{
- uint8_t requested_bw;
- uint32_t temp;
+ uint8_t request_reg_val;
+ uint32_t temp, request_bw;
- /* Error check whether request bw greater than allocated */
- if (req_bw > link->dpia_bw_alloc_config.estimated_bw) {
- DC_LOG_ERROR("%s: Request bw greater than estimated bw for link(%d)\n",
- __func__, link->link_index);
- req_bw = link->dpia_bw_alloc_config.estimated_bw;
+ if (link->dpia_bw_alloc_config.bw_granularity == 0) {
+ DC_LOG_ERROR("%s: Link[%d]: bw_granularity is zero!", __func__, link->link_index);
+ return;
}
temp = req_bw * link->dpia_bw_alloc_config.bw_granularity;
- requested_bw = temp / Kbps_TO_Gbps;
+ request_reg_val = temp / Kbps_TO_Gbps;
/* Always make sure to add more to account for floating points */
if (temp % Kbps_TO_Gbps)
- ++requested_bw;
+ ++request_reg_val;
- /* Error check whether requested and allocated are equal */
- req_bw = requested_bw * (Kbps_TO_Gbps / link->dpia_bw_alloc_config.bw_granularity);
- if (req_bw && (req_bw == link->dpia_bw_alloc_config.allocated_bw)) {
- DC_LOG_ERROR("%s: Request bw equals to allocated bw for link(%d)\n",
- __func__, link->link_index);
+ request_bw = request_reg_val * (Kbps_TO_Gbps / link->dpia_bw_alloc_config.bw_granularity);
+
+ if (request_bw > link->dpia_bw_alloc_config.estimated_bw) {
+ DC_LOG_ERROR("%s: Link[%d]: Request BW (%d --> %d) > Estimated BW (%d)... Set to Estimated BW!",
+ __func__, link->link_index,
+ req_bw, request_bw, link->dpia_bw_alloc_config.estimated_bw);
+ req_bw = link->dpia_bw_alloc_config.estimated_bw;
+
+ temp = req_bw * link->dpia_bw_alloc_config.bw_granularity;
+ request_reg_val = temp / Kbps_TO_Gbps;
+ if (temp % Kbps_TO_Gbps)
+ ++request_reg_val;
}
- link->dpia_bw_alloc_config.response_ready = false; // Reset flag
- core_link_write_dpcd(
- link,
- REQUESTED_BW,
- &requested_bw,
- sizeof(uint8_t));
-}
+ link->dpia_bw_alloc_config.allocated_bw = request_bw;
+ DC_LOG_DC("%s: Link[%d]: Request BW: %d", __func__, link->link_index, request_bw);
-/*
- * Return the response_ready flag from dc_link struct
- *
- * @link: pointer to the dc_link struct instance
- *
- * return: response_ready flag from dc_link struct
- */
-static bool get_cm_response_ready_flag(struct dc_link *link)
-{
- return link->dpia_bw_alloc_config.response_ready;
+ core_link_write_dpcd(link, REQUESTED_BW,
+ &request_reg_val,
+ sizeof(uint8_t));
}
// ------------------------------------------------------------------
-// PUBLIC FUNCTIONS
+// PUBLIC FUNCTIONS
// ------------------------------------------------------------------
-bool link_dp_dpia_set_dptx_usb4_bw_alloc_support(struct dc_link *link)
+bool link_dpia_enable_usb4_dp_bw_alloc_mode(struct dc_link *link)
{
bool ret = false;
- uint8_t response = 0,
- bw_support_dpia = 0,
- bw_support_cm = 0;
+ uint8_t val;
- if (!(link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA && link->hpd_status))
- goto out;
+ if (link->hpd_status) {
+ val = DPTX_BW_ALLOC_MODE_ENABLE | DPTX_BW_ALLOC_UNMASK_IRQ;
- if (core_link_read_dpcd(
- link,
- DP_TUNNELING_CAPABILITIES,
- &response,
- sizeof(uint8_t)) == DC_OK)
- bw_support_dpia = (response >> 7) & 1;
-
- if (core_link_read_dpcd(
- link,
- USB4_DRIVER_BW_CAPABILITY,
- &response,
- sizeof(uint8_t)) == DC_OK)
- bw_support_cm = (response >> 7) & 1;
-
- /* Send request acknowledgment to Turn ON DPTX support */
- if (bw_support_cm && bw_support_dpia) {
-
- response = 0x80;
- if (core_link_write_dpcd(
- link,
- DPTX_BW_ALLOCATION_MODE_CONTROL,
- &response,
- sizeof(uint8_t)) != DC_OK) {
- DC_LOG_DEBUG("%s: FAILURE Enabling DPtx BW Allocation Mode Support for link(%d)\n",
- __func__, link->link_index);
- } else {
- // SUCCESS Enabled DPtx BW Allocation Mode Support
- DC_LOG_DEBUG("%s: SUCCESS Enabling DPtx BW Allocation Mode Support for link(%d)\n",
- __func__, link->link_index);
+ if (core_link_write_dpcd(link, DPTX_BW_ALLOCATION_MODE_CONTROL, &val, sizeof(uint8_t)) == DC_OK) {
+ DC_LOG_DEBUG("%s: link[%d] DPTX BW allocation mode enabled", __func__, link->link_index);
+
+ retrieve_usb4_dp_bw_allocation_info(link);
+
+ if (link->dpia_bw_alloc_config.nrd_max_link_rate && link->dpia_bw_alloc_config.nrd_max_lane_count) {
+ link->reported_link_cap.link_rate = link->dpia_bw_alloc_config.nrd_max_link_rate;
+ link->reported_link_cap.lane_count = link->dpia_bw_alloc_config.nrd_max_lane_count;
+ }
- ret = true;
- init_usb4_bw_struct(link);
link->dpia_bw_alloc_config.bw_alloc_enabled = true;
+ ret = true;
- /*
- * During DP tunnel creation, CM preallocates BW and reduces estimated BW of other
- * DPIA. CM release preallocation only when allocation is complete. Do zero alloc
- * to make the CM to release preallocation and update estimated BW correctly for
- * all DPIAs per host router
- */
- link_dp_dpia_allocate_usb4_bandwidth_for_stream(link, 0);
- }
+ if (link->dc->debug.dpia_debug.bits.enable_usb4_bw_zero_alloc_patch) {
+ /*
+ * During DP tunnel creation, the CM preallocates BW
+ * and reduces the estimated BW of other DPIAs.
+ * The CM releases the preallocation only when the allocation is complete.
+ * Perform a zero allocation to make the CM release the preallocation
+ * and correctly update the estimated BW for all DPIAs per host router.
+ */
+ link_dp_dpia_allocate_usb4_bandwidth_for_stream(link, 0);
+ }
+ } else
+ DC_LOG_DEBUG("%s: link[%d] failed to enable DPTX BW allocation mode", __func__, link->link_index);
}
-out:
return ret;
}
-void dpia_handle_bw_alloc_response(struct dc_link *link, uint8_t bw, uint8_t result)
+/*
+ * Handle DP BW allocation status register
+ *
+ * @link: pointer to the dc_link struct instance
+ * @status: content of DP tunneling status DPCD register
+ *
+ * return: none
+ */
+void link_dp_dpia_handle_bw_alloc_status(struct dc_link *link, uint8_t status)
{
- int bw_needed = 0;
- int estimated = 0;
-
- if (!get_bw_alloc_proceed_flag((link)))
- return;
-
- switch (result) {
-
- case DPIA_BW_REQ_FAILED:
-
- /*
- * Ideally, we shouldn't run into this case as we always validate available
- * bandwidth and request within that limit
- */
- estimated = bw * (Kbps_TO_Gbps / link->dpia_bw_alloc_config.bw_granularity);
-
- DC_LOG_ERROR("%s: BW REQ FAILURE for DP-TX Request for link(%d)\n",
- __func__, link->link_index);
- DC_LOG_ERROR("%s: current estimated_bw(%d), new estimated_bw(%d)\n",
- __func__, link->dpia_bw_alloc_config.estimated_bw, estimated);
-
- /* Update the new Estimated BW value updated by CM */
- link->dpia_bw_alloc_config.estimated_bw = estimated;
-
- /* Allocate the previously requested bandwidth */
- set_usb4_req_bw_req(link, link->dpia_bw_alloc_config.estimated_bw);
-
- /*
- * If FAIL then it is either:
- * 1. Due to DP-Tx trying to allocate more than available i.e. it failed locally
- * => get estimated and allocate that
- * 2. Due to the fact that DP-Tx tried to allocated ESTIMATED BW and failed then
- * CM will have to update 0xE0023 with new ESTIMATED BW value.
- */
- break;
-
- case DPIA_BW_REQ_SUCCESS:
-
- bw_needed = bw * (Kbps_TO_Gbps / link->dpia_bw_alloc_config.bw_granularity);
-
- DC_LOG_DEBUG("%s: BW REQ SUCCESS for DP-TX Request for link(%d)\n",
- __func__, link->link_index);
- DC_LOG_DEBUG("%s: current allocated_bw(%d), new allocated_bw(%d)\n",
- __func__, link->dpia_bw_alloc_config.allocated_bw, bw_needed);
-
- link->dpia_bw_alloc_config.allocated_bw = bw_needed;
-
- link->dpia_bw_alloc_config.response_ready = true;
- break;
-
- case DPIA_EST_BW_CHANGED:
-
- estimated = bw * (Kbps_TO_Gbps / link->dpia_bw_alloc_config.bw_granularity);
-
- DC_LOG_DEBUG("%s: ESTIMATED BW CHANGED for link(%d)\n",
- __func__, link->link_index);
- DC_LOG_DEBUG("%s: current estimated_bw(%d), new estimated_bw(%d)\n",
- __func__, link->dpia_bw_alloc_config.estimated_bw, estimated);
-
- link->dpia_bw_alloc_config.estimated_bw = estimated;
- break;
-
- case DPIA_BW_ALLOC_CAPS_CHANGED:
+ link->dpia_bw_alloc_config.estimated_bw = get_estimated_bw(link);
- DC_LOG_ERROR("%s: BW ALLOC CAPABILITY CHANGED to Disabled for link(%d)\n",
- __func__, link->link_index);
- link->dpia_bw_alloc_config.bw_alloc_enabled = false;
- break;
+ if (status & DP_TUNNELING_BW_REQUEST_SUCCEEDED) {
+ DC_LOG_DEBUG("%s: BW Allocation request succeeded on link(%d)",
+ __func__, link->link_index);
+ } else if (status & DP_TUNNELING_BW_REQUEST_FAILED) {
+ DC_LOG_DEBUG("%s: BW Allocation request failed on link(%d) allocated/estimated BW=%d",
+ __func__, link->link_index, link->dpia_bw_alloc_config.estimated_bw);
+
+ link_dpia_send_bw_alloc_request(link, link->dpia_bw_alloc_config.estimated_bw);
+ } else if (status & DP_TUNNELING_ESTIMATED_BW_CHANGED) {
+ DC_LOG_DEBUG("%s: Estimated BW changed on link(%d) new estimated BW=%d",
+ __func__, link->link_index, link->dpia_bw_alloc_config.estimated_bw);
}
-}
-int dpia_handle_usb4_bandwidth_allocation_for_link(struct dc_link *link, int peak_bw)
-{
- int ret = 0;
- uint8_t timeout = 10;
-
- if (!(link && DISPLAY_ENDPOINT_USB4_DPIA == link->ep_type
- && link->dpia_bw_alloc_config.bw_alloc_enabled))
- goto out;
-
- //1. Hot Plug
- if (link->hpd_status && peak_bw > 0) {
- // If DP over USB4 then we need to check BW allocation
- link->dpia_bw_alloc_config.link_max_bw = peak_bw;
- set_usb4_req_bw_req(link, link->dpia_bw_alloc_config.link_max_bw);
-
- do {
- if (timeout > 0)
- timeout--;
- else
- break;
- msleep(10);
- } while (!get_cm_response_ready_flag(link));
+ core_link_write_dpcd(
+ link, DP_TUNNELING_STATUS,
+ &status, sizeof(status));
+}
- if (!timeout)
- ret = 0;// ERROR TIMEOUT waiting for response for allocating bw
- else if (link->dpia_bw_alloc_config.allocated_bw > 0)
- ret = link->dpia_bw_alloc_config.allocated_bw;
+/*
+ * Handle the DP Bandwidth allocation for DPIA
+ *
+ */
+void dpia_handle_usb4_bandwidth_allocation_for_link(struct dc_link *link, int peak_bw)
+{
+ if (link && link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dp_tunneling
+ && link->dpia_bw_alloc_config.bw_alloc_enabled) {
+ //1. Hot Plug
+ if (link->hpd_status && peak_bw > 0) {
+ // If DP over USB4 then we need to check BW allocation
+ link->dpia_bw_alloc_config.link_max_bw = peak_bw;
+
+ link_dpia_send_bw_alloc_request(link, peak_bw);
+ }
+ //2. Cold Unplug
+ else if (!link->hpd_status)
+ dpia_bw_alloc_unplug(link);
}
- //2. Cold Unplug
- else if (!link->hpd_status)
- dpia_bw_alloc_unplug(link);
-
-out:
- return ret;
}
-bool link_dp_dpia_allocate_usb4_bandwidth_for_stream(struct dc_link *link, int req_bw)
+
+void link_dp_dpia_allocate_usb4_bandwidth_for_stream(struct dc_link *link, int req_bw)
{
- bool ret = false;
- uint8_t timeout = 10;
+ link->dpia_bw_alloc_config.estimated_bw = get_estimated_bw(link);
- DC_LOG_DEBUG("%s: ENTER: link(%d), hpd_status(%d), current allocated_bw(%d), req_bw(%d)\n",
+ DC_LOG_DEBUG("%s: ENTER: link[%d] hpd(%d) Allocated_BW: %d Estimated_BW: %d Req_BW: %d",
__func__, link->link_index, link->hpd_status,
- link->dpia_bw_alloc_config.allocated_bw, req_bw);
+ link->dpia_bw_alloc_config.allocated_bw,
+ link->dpia_bw_alloc_config.estimated_bw,
+ req_bw);
+
+ if (link_dp_is_bw_alloc_available(link))
+ link_dpia_send_bw_alloc_request(link, req_bw);
+ else
+ DC_LOG_DEBUG("%s: BW Allocation mode not available", __func__);
+}
- if (!get_bw_alloc_proceed_flag(link))
- goto out;
+uint32_t link_dpia_get_dp_overhead(const struct dc_link *link)
+{
+ uint32_t link_dp_overhead = 0;
- set_usb4_req_bw_req(link, req_bw);
- do {
- if (timeout > 0)
- timeout--;
- else
- break;
- msleep(10);
- } while (!get_cm_response_ready_flag(link));
+ if ((link->type == dc_connection_mst_branch) &&
+ !link->dpcd_caps.channel_coding_cap.bits.DP_128b_132b_SUPPORTED) {
+ /* For 8b/10b encoding: MTP is 64 time slots long, slot 0 is used for MTPH
+ * MST overhead is 1/64 of link bandwidth (excluding any overhead)
+ */
+ const struct dc_link_settings *link_cap = dc_link_get_link_cap(link);
- if (timeout)
- ret = true;
+ if (link_cap) {
+ uint32_t link_bw_in_kbps = (uint32_t)link_cap->link_rate *
+ (uint32_t)link_cap->lane_count *
+ LINK_RATE_REF_FREQ_IN_KHZ * 8;
+ link_dp_overhead = (link_bw_in_kbps / MST_TIME_SLOT_COUNT)
+ + ((link_bw_in_kbps % MST_TIME_SLOT_COUNT) ? 1 : 0);
+ }
+ }
-out:
- DC_LOG_DEBUG("%s: EXIT: timeout(%d), ret(%d)\n", __func__, timeout, ret);
- return ret;
+ return link_dp_overhead;
}
-bool dpia_validate_usb4_bw(struct dc_link **link, int *bw_needed_per_dpia, const unsigned int num_dpias)
+/*
+ * Aggregates the DPIA bandwidth usage for the respective USB4 Router.
+ * And then validate if the required bandwidth is within the router's capacity.
+ *
+ * @dc_validation_dpia_set: pointer to the dc_validation_dpia_set
+ * @count: number of DPIA validation sets
+ *
+ * return: true if validation is succeeded
+ */
+bool link_dpia_validate_dp_tunnel_bandwidth(const struct dc_validation_dpia_set *dpia_link_sets, uint8_t count)
{
- bool ret = true;
- int bw_needed_per_hr[MAX_HR_NUM] = { 0, 0 }, host_router_total_dp_bw = 0;
- uint8_t lowest_dpia_index, i, hr_index;
+ uint32_t granularity_Gbps;
+ const struct dc_link *link;
+ uint32_t link_bw_granularity;
+ uint32_t link_required_bw;
+ struct usb4_router_validation_set router_sets[MAX_HOST_ROUTERS_NUM] = { 0 };
+ uint8_t i;
+ bool is_success = true;
+ uint8_t router_count = 0;
+
+ if ((dpia_link_sets == NULL) || (count == 0))
+ return is_success;
+
+ // Iterate through each DP tunneling link (DPIA).
+ // Aggregate its bandwidth requirements onto the respective USB4 router.
+ for (i = 0; i < count; i++) {
+ link = dpia_link_sets[i].link;
+ link_required_bw = dpia_link_sets[i].required_bw;
+ const struct dc_tunnel_settings *dp_tunnel_settings = dpia_link_sets[i].tunnel_settings;
+
+ if ((link == NULL) || (dp_tunnel_settings == NULL) || dp_tunnel_settings->bw_granularity == 0)
+ break;
- if (!num_dpias || num_dpias > MAX_DPIA_NUM)
- return ret;
+ if (link->type == dc_connection_mst_branch)
+ link_required_bw += link_dpia_get_dp_overhead(link);
- lowest_dpia_index = get_lowest_dpia_index(link[0]);
+ granularity_Gbps = (Kbps_TO_Gbps / dp_tunnel_settings->bw_granularity);
+ link_bw_granularity = (link_required_bw / granularity_Gbps) * granularity_Gbps +
+ ((link_required_bw % granularity_Gbps) ? granularity_Gbps : 0);
- /* get total Host Router BW with granularity for the given modes */
- for (i = 0; i < num_dpias; ++i) {
- int granularity_Gbps = 0;
- int bw_granularity = 0;
+ // Find or add the USB4 router associated with the current DPIA link
+ for (uint8_t j = 0; j < MAX_HOST_ROUTERS_NUM; j++) {
+ if (router_sets[j].is_valid == false) {
+ router_sets[j].is_valid = true;
+ router_sets[j].cm_id = dp_tunnel_settings->cm_id;
+ router_count++;
+ }
- if (!link[i]->dpia_bw_alloc_config.bw_alloc_enabled)
- continue;
+ if (router_sets[j].cm_id == dp_tunnel_settings->cm_id) {
+ uint32_t remaining_bw =
+ dp_tunnel_settings->estimated_bw - dp_tunnel_settings->allocated_bw;
- if (link[i]->link_index < lowest_dpia_index)
- continue;
+ router_sets[j].allocated_bw += dp_tunnel_settings->allocated_bw;
- granularity_Gbps = (Kbps_TO_Gbps / link[i]->dpia_bw_alloc_config.bw_granularity);
- bw_granularity = (bw_needed_per_dpia[i] / granularity_Gbps) * granularity_Gbps +
- ((bw_needed_per_dpia[i] % granularity_Gbps) ? granularity_Gbps : 0);
+ if (remaining_bw > router_sets[j].remaining_bw)
+ router_sets[j].remaining_bw = remaining_bw;
- hr_index = (link[i]->link_index - lowest_dpia_index) / 2;
- bw_needed_per_hr[hr_index] += bw_granularity;
- }
+ // Get the max estimated BW within the same CM_ID
+ if (dp_tunnel_settings->estimated_bw > router_sets[j].estimated_bw)
+ router_sets[j].estimated_bw = dp_tunnel_settings->estimated_bw;
- /* validate against each Host Router max BW */
- for (hr_index = 0; hr_index < MAX_HR_NUM; ++hr_index) {
- if (bw_needed_per_hr[hr_index]) {
- host_router_total_dp_bw = get_host_router_total_dp_tunnel_bw(link[0]->dc, hr_index);
- if (bw_needed_per_hr[hr_index] > host_router_total_dp_bw) {
- ret = false;
+ router_sets[j].required_bw += link_bw_granularity;
+ router_sets[j].dpia_count++;
break;
}
}
}
- return ret;
-}
+ // Validate bandwidth for each unique router found.
+ for (i = 0; i < router_count; i++) {
+ uint32_t total_bw = 0;
-int link_dp_dpia_get_dp_overhead_in_dp_tunneling(struct dc_link *link)
-{
- int dp_overhead = 0, link_mst_overhead = 0;
+ if (router_sets[i].is_valid == false)
+ break;
- if (!get_bw_alloc_proceed_flag((link)))
- return dp_overhead;
+ // Determine the total available bandwidth for the current router based on aggregated data
+ if ((router_sets[i].dpia_count == 1) || (router_sets[i].allocated_bw == 0))
+ total_bw = router_sets[i].estimated_bw;
+ else
+ total_bw = router_sets[i].allocated_bw + router_sets[i].remaining_bw;
- /* if its mst link, add MTPH overhead */
- if ((link->type == dc_connection_mst_branch) &&
- !link->dpcd_caps.channel_coding_cap.bits.DP_128b_132b_SUPPORTED) {
- /* For 8b/10b encoding: MTP is 64 time slots long, slot 0 is used for MTPH
- * MST overhead is 1/64 of link bandwidth (excluding any overhead)
- */
- const struct dc_link_settings *link_cap =
- dc_link_get_link_cap(link);
- uint32_t link_bw_in_kbps = (uint32_t)link_cap->link_rate *
- (uint32_t)link_cap->lane_count *
- LINK_RATE_REF_FREQ_IN_KHZ * 8;
- link_mst_overhead = (link_bw_in_kbps / 64) + ((link_bw_in_kbps % 64) ? 1 : 0);
+ if (router_sets[i].required_bw > total_bw) {
+ is_success = false;
+ break;
+ }
}
- /* add all the overheads */
- dp_overhead = link_mst_overhead;
-
- return dp_overhead;
+ return is_success;
}
+
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.h
index 3b6d8494f9d5..41efcb3e44e2 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_dpia_bw.h
@@ -28,10 +28,6 @@
#include "link.h"
-/* Number of Host Routers per motherboard is 2 */
-#define MAX_HR_NUM 2
-/* Number of DPIA per host router is 2 */
-#define MAX_DPIA_NUM (MAX_HR_NUM * 2)
/*
* Host Router BW type
@@ -42,14 +38,24 @@ enum bw_type {
HOST_ROUTER_BW_INVALID,
};
+struct usb4_router_validation_set {
+ bool is_valid;
+ uint8_t cm_id;
+ uint8_t dpia_count;
+ uint32_t required_bw;
+ uint32_t allocated_bw;
+ uint32_t estimated_bw;
+ uint32_t remaining_bw;
+};
+
/*
- * Enable BW Allocation Mode Support from the DP-Tx side
+ * Enable USB4 DP BW allocation mode
*
* @link: pointer to the dc_link struct instance
*
* return: SUCCESS or FAILURE
*/
-bool link_dp_dpia_set_dptx_usb4_bw_alloc_support(struct dc_link *link);
+bool link_dpia_enable_usb4_dp_bw_alloc_mode(struct dc_link *link);
/*
* Allocates only what the stream needs for bw, so if:
@@ -59,9 +65,8 @@ bool link_dp_dpia_set_dptx_usb4_bw_alloc_support(struct dc_link *link);
* @link: pointer to the dc_link struct instance
* @req_bw: Bw requested by the stream
*
- * return: true if allocated successfully
*/
-bool link_dp_dpia_allocate_usb4_bandwidth_for_stream(struct dc_link *link, int req_bw);
+void link_dp_dpia_allocate_usb4_bandwidth_for_stream(struct dc_link *link, int req_bw);
/*
* Handle the USB4 BW Allocation related functionality here:
@@ -71,41 +76,37 @@ bool link_dp_dpia_allocate_usb4_bandwidth_for_stream(struct dc_link *link, int r
* @link: pointer to the dc_link struct instance
* @peak_bw: Peak bw used by the link/sink
*
- * return: allocated bw else return 0
*/
-int dpia_handle_usb4_bandwidth_allocation_for_link(struct dc_link *link, int peak_bw);
+void dpia_handle_usb4_bandwidth_allocation_for_link(struct dc_link *link, int peak_bw);
/*
- * Handle function for when the status of the Request above is complete.
- * We will find out the result of allocating on CM and update structs.
+ * Obtain all the DP overheads in dp tunneling for the dpia link
*
* @link: pointer to the dc_link struct instance
- * @bw: Allocated or Estimated BW depending on the result
- * @result: Response type
*
- * return: none
+ * return: DP overheads in DP tunneling
*/
-void dpia_handle_bw_alloc_response(struct dc_link *link, uint8_t bw, uint8_t result);
+uint32_t link_dpia_get_dp_overhead(const struct dc_link *link);
/*
- * Handle the validation of total BW here and confirm that the bw used by each
- * DPIA doesn't exceed available BW for each host router (HR)
+ * Handle DP BW allocation status register
*
- * @link[]: array of link pointer to all possible DPIA links
- * @bw_needed[]: bw needed for each DPIA link based on timing
- * @num_dpias: Number of DPIAs for the above 2 arrays. Should always be <= MAX_DPIA_NUM
+ * @link: pointer to the dc_link struct instance
+ * @status: content of DP tunneling status register
*
- * return: TRUE if bw used by DPIAs doesn't exceed available BW else return FALSE
+ * return: none
*/
-bool dpia_validate_usb4_bw(struct dc_link **link, int *bw_needed, const unsigned int num_dpias);
+void link_dp_dpia_handle_bw_alloc_status(struct dc_link *link, uint8_t status);
/*
- * Obtain all the DP overheads in dp tunneling for the dpia link
+ * Aggregates the DPIA bandwidth usage for the respective USB4 Router.
*
- * @link: pointer to the dc_link struct instance
+ * @dc_validation_dpia_set: pointer to the dc_validation_dpia_set
+ * @count: number of DPIA validation sets
*
- * return: DP overheads in DP tunneling
+ * return: true if validation is succeeded
*/
-int link_dp_dpia_get_dp_overhead_in_dp_tunneling(struct dc_link *link);
+bool link_dpia_validate_dp_tunnel_bandwidth(const struct dc_validation_dpia_set *dpia_link_sets, uint8_t count);
#endif /* DC_INC_LINK_DP_DPIA_BW_H_ */
+
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c
index 48abeaa88678..693477413347 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_irq_handler.c
@@ -37,6 +37,7 @@
#include "link/accessories/link_dp_trace.h"
#include "link/link_dpms.h"
#include "dm_helpers.h"
+#include "link_dp_dpia_bw.h"
#define DC_LOGGER \
link->ctx->logger
@@ -226,6 +227,12 @@ static void handle_hpd_irq_replay_sink(struct dc_link *link)
replay_configuration.bits.STATE_TRANSITION_ERROR_STATUS) {
bool allow_active;
+ link->replay_settings.config.replay_error_status.raw |= replay_error_status.raw;
+
+ /* Increment desync error counter if a desync error is detected */
+ if (replay_configuration.bits.DESYNC_ERROR_STATUS)
+ link->replay_settings.replay_desync_error_fail_count++;
+
if (link->replay_settings.config.force_disable_desync_error_check)
return;
@@ -281,6 +288,30 @@ void dp_handle_link_loss(struct dc_link *link)
}
}
+static void dp_handle_tunneling_irq(struct dc_link *link)
+{
+ enum dc_status retval;
+ uint8_t tunneling_status = 0;
+
+ retval = core_link_read_dpcd(
+ link, DP_TUNNELING_STATUS,
+ &tunneling_status,
+ sizeof(tunneling_status));
+
+ if (retval == DC_OK) {
+ DC_LOG_HW_HPD_IRQ("%s: Got DP tunneling status on link %d status=0x%x",
+ __func__, link->link_index, tunneling_status);
+
+ if (tunneling_status & DP_TUNNELING_BW_ALLOC_BITS_MASK)
+ link_dp_dpia_handle_bw_alloc_status(link, tunneling_status);
+ }
+
+ tunneling_status = DP_TUNNELING_IRQ;
+ core_link_write_dpcd(
+ link, DP_LINK_SERVICE_IRQ_VECTOR_ESI0,
+ &tunneling_status, 1);
+}
+
static void read_dpcd204h_on_irq_hpd(struct dc_link *link, union hpd_irq_data *irq_data)
{
enum dc_status retval;
@@ -314,13 +345,19 @@ enum dc_status dp_read_hpd_rx_irq_data(
*
* For DP 1.4 we need to read those from 2002h range.
*/
- if (link->dpcd_caps.dpcd_rev.raw < DPCD_REV_14)
+ if (link->dpcd_caps.dpcd_rev.raw < DPCD_REV_14) {
retval = core_link_read_dpcd(
link,
DP_SINK_COUNT,
irq_data->raw,
- sizeof(union hpd_irq_data));
- else {
+ DP_SINK_STATUS - DP_SINK_COUNT + 1);
+
+ if (link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dp_tunneling) {
+ retval = core_link_read_dpcd(
+ link, DP_LINK_SERVICE_IRQ_VECTOR_ESI0,
+ &irq_data->bytes.link_service_irq_esi0.raw, 1);
+ }
+ } else {
/* Read 14 bytes in a single read and then copy only the required fields.
* This is more efficient than doing it in two separate AUX reads. */
@@ -341,6 +378,7 @@ enum dc_status dp_read_hpd_rx_irq_data(
irq_data->bytes.lane23_status.raw = tmp[DP_LANE2_3_STATUS_ESI - DP_SINK_COUNT_ESI];
irq_data->bytes.lane_status_updated.raw = tmp[DP_LANE_ALIGN_STATUS_UPDATED_ESI - DP_SINK_COUNT_ESI];
irq_data->bytes.sink_status.raw = tmp[DP_SINK_STATUS_ESI - DP_SINK_COUNT_ESI];
+ irq_data->bytes.link_service_irq_esi0.raw = tmp[DP_LINK_SERVICE_IRQ_VECTOR_ESI0 - DP_SINK_COUNT_ESI];
/*
* This display doesn't have correct values in DPCD200Eh.
@@ -408,7 +446,8 @@ bool dp_handle_hpd_rx_irq(struct dc_link *link,
if (hpd_irq_dpcd_data.bytes.device_service_irq.bits.AUTOMATED_TEST) {
// Workaround for DP 1.4a LL Compliance CTS as USB4 has to share encoders unlike DP and USBC
- if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA)
+ if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA &&
+ !link->dc->config.enable_dpia_pre_training)
link->skip_fallback_on_link_loss = true;
device_service_clear.bits.AUTOMATED_TEST = 1;
@@ -482,6 +521,11 @@ bool dp_handle_hpd_rx_irq(struct dc_link *link,
dp_trace_link_loss_increment(link);
}
+ if (link->dpcd_caps.usb4_dp_tun_info.dp_tun_cap.bits.dp_tunneling) {
+ if (hpd_irq_dpcd_data.bytes.link_service_irq_esi0.bits.DP_LINK_TUNNELING_IRQ)
+ dp_handle_tunneling_irq(link);
+ }
+
if (link->type == dc_connection_sst_branch &&
hpd_irq_dpcd_data.bytes.sink_cnt.bits.SINK_COUNT
!= link->dpcd_sink_count)
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.c
index bafa52a0165a..49521ac4b0e8 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.c
@@ -75,7 +75,8 @@ void dp_disable_link_phy(struct dc_link *link,
struct dc *dc = link->ctx->dc;
if (!link->wa_flags.dp_keep_receiver_powered &&
- !link->skip_implict_edp_power_control)
+ !link->skip_implict_edp_power_control &&
+ link->type != dc_connection_none)
dpcd_write_rx_power_ctrl(link, false);
dc->hwss.disable_link_output(link, link_res, signal);
@@ -104,7 +105,7 @@ void dp_set_hw_lane_settings(
// Don't return here if using FIXED_VS link HWSS and encoding is 128b/132b
if ((link_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) &&
!is_immediate_downstream(link, offset) &&
- (!(link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) ||
+ (!((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) ||
link_dp_get_encoding_format(&link_settings->link_settings) == DP_8b_10b_ENCODING))
return;
@@ -141,11 +142,12 @@ enum dc_status dp_set_fec_ready(struct dc_link *link, const struct link_resource
* if the sink supports it and leave it enabled on link.
* If FEC is not supported, disable it.
*/
- struct link_encoder *link_enc = NULL;
+ struct link_encoder *link_enc = link_res->dio_link_enc;
enum dc_status status = DC_OK;
uint8_t fec_config = 0;
- link_enc = link_enc_cfg_get_link_enc(link);
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
ASSERT(link_enc);
if (link_enc->funcs->fec_set_ready == NULL)
return DC_NOT_SUPPORTED;
@@ -163,8 +165,9 @@ enum dc_status dp_set_fec_ready(struct dc_link *link, const struct link_resource
} else {
if (link->fec_state == dc_link_fec_ready) {
fec_config = 0;
- core_link_write_dpcd(link, DP_FEC_CONFIGURATION,
- &fec_config, sizeof(fec_config));
+ if (link->type != dc_connection_none)
+ core_link_write_dpcd(link, DP_FEC_CONFIGURATION,
+ &fec_config, sizeof(fec_config));
link_enc->funcs->fec_set_ready(link_enc, false);
link->fec_state = dc_link_fec_not_ready;
@@ -174,13 +177,14 @@ enum dc_status dp_set_fec_ready(struct dc_link *link, const struct link_resource
return status;
}
-void dp_set_fec_enable(struct dc_link *link, bool enable)
+void dp_set_fec_enable(struct dc_link *link, const struct link_resource *link_res, bool enable)
{
- struct link_encoder *link_enc = NULL;
+ struct link_encoder *link_enc = link_res->dio_link_enc;
- link_enc = link_enc_cfg_get_link_enc(link);
- ASSERT(link_enc);
- if (link_enc->funcs->fec_set_enable == NULL)
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+
+ if (link_enc == NULL || link_enc->funcs == NULL || link_enc->funcs->fec_set_enable == NULL)
return;
if (enable && dp_should_enable_fec(link)) {
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.h
index 1eb0619d6710..ab1c1f8f1f8b 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_phy.h
@@ -52,7 +52,8 @@ void dp_set_drive_settings(
enum dc_status dp_set_fec_ready(struct dc_link *link,
const struct link_resource *link_res, bool ready);
-void dp_set_fec_enable(struct dc_link *link, bool enable);
+void dp_set_fec_enable(struct dc_link *link,
+ const struct link_resource *link_res, bool enable);
void dpcd_write_rx_power_ctrl(struct dc_link *link, bool on);
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c
index 754c895e1bfb..2dc1a660e504 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.c
@@ -736,10 +736,12 @@ void override_training_settings(
lt_settings->pre_emphasis = overrides->pre_emphasis;
if (overrides->post_cursor2 != NULL)
lt_settings->post_cursor2 = overrides->post_cursor2;
+ if (link->wa_flags.force_dp_ffe_preset && !dp_is_lttpr_present(link))
+ lt_settings->ffe_preset = &link->forced_dp_ffe_preset;
if (overrides->ffe_preset != NULL)
lt_settings->ffe_preset = overrides->ffe_preset;
/* Override HW lane settings with BIOS forced values if present */
- if ((link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
+ if ((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) &&
lt_settings->lttpr_mode == LTTPR_MODE_TRANSPARENT) {
lt_settings->voltage_swing = &link->bios_forced_drive_settings.VOLTAGE_SWING;
lt_settings->pre_emphasis = &link->bios_forced_drive_settings.PRE_EMPHASIS;
@@ -783,7 +785,6 @@ void override_training_settings(
lt_settings->lttpr_mode = LTTPR_MODE_NON_LTTPR;
dp_get_lttpr_mode_override(link, &lt_settings->lttpr_mode);
-
}
enum dc_dp_training_pattern decide_cr_training_pattern(
@@ -799,19 +800,23 @@ enum dc_dp_training_pattern decide_cr_training_pattern(
}
enum dc_dp_training_pattern decide_eq_training_pattern(struct dc_link *link,
+ const struct link_resource *link_res,
const struct dc_link_settings *link_settings)
{
- struct link_encoder *link_enc;
+ struct link_encoder *link_enc = link_res->dio_link_enc;
struct encoder_feature_support *enc_caps;
struct dpcd_caps *rx_caps = &link->dpcd_caps;
enum dc_dp_training_pattern pattern = DP_TRAINING_PATTERN_SEQUENCE_2;
- link_enc = link_enc_cfg_get_link_enc(link);
- ASSERT(link_enc);
- enc_caps = &link_enc->features;
-
switch (link_dp_get_encoding_format(link_settings)) {
case DP_8b_10b_ENCODING:
+ if (!link->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
+
+ if (!link_enc)
+ break;
+
+ enc_caps = &link_enc->features;
if (enc_caps->flags.bits.IS_TPS4_CAPABLE &&
rx_caps->max_down_spread.bits.TPS4_SUPPORTED)
pattern = DP_TRAINING_PATTERN_SEQUENCE_4;
@@ -884,13 +889,14 @@ void dp_decide_lane_settings(
void dp_decide_training_settings(
struct dc_link *link,
+ const struct link_resource *link_res,
const struct dc_link_settings *link_settings,
struct link_training_settings *lt_settings)
{
if (link_dp_get_encoding_format(link_settings) == DP_8b_10b_ENCODING)
- decide_8b_10b_training_settings(link, link_settings, lt_settings);
+ decide_8b_10b_training_settings(link, link_res, link_settings, lt_settings);
else if (link_dp_get_encoding_format(link_settings) == DP_128b_132b_ENCODING)
- decide_128b_132b_training_settings(link, link_settings, lt_settings);
+ decide_128b_132b_training_settings(link, link_res, link_settings, lt_settings);
}
@@ -1554,6 +1560,7 @@ enum link_training_result dp_perform_link_training(
/* decide training settings */
dp_decide_training_settings(
link,
+ link_res,
link_settings,
&lt_settings);
@@ -1567,14 +1574,15 @@ enum link_training_result dp_perform_link_training(
/* configure link prior to entering training mode */
dpcd_configure_lttpr_mode(link, &lt_settings);
- dp_set_fec_ready(link, link_res, lt_settings.should_set_fec_ready);
+ if (link_dp_get_encoding_format(link_settings) == DP_8b_10b_ENCODING)
+ dp_set_fec_ready(link, link_res, lt_settings.should_set_fec_ready);
dpcd_configure_channel_coding(link, &lt_settings);
/* enter training mode:
* Per DP specs starting from here, DPTX device shall not issue
* Non-LT AUX transactions inside training mode.
*/
- if ((link->chip_caps & EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) && encoding == DP_8b_10b_ENCODING)
+ if (((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) == AMD_EXT_DISPLAY_PATH_CAPS__DP_FIXED_VS_EN) && encoding == DP_8b_10b_ENCODING)
status = dp_perform_fixed_vs_pe_training_sequence(link, link_res, &lt_settings);
else if (encoding == DP_8b_10b_ENCODING)
status = dp_perform_8b_10b_link_training(link, link_res, &lt_settings);
@@ -1780,13 +1788,10 @@ bool perform_link_training_with_retries(
is_link_bw_min = ((cur_link_settings.link_rate <= LINK_RATE_LOW) &&
(cur_link_settings.lane_count <= LANE_COUNT_ONE));
- if (is_link_bw_low) {
+ if (is_link_bw_low)
DC_LOG_WARNING(
"%s: Link(%d) bandwidth too low after fallback req_bw(%d) > link_bw(%d)\n",
__func__, link->link_index, req_bw, link_bw);
-
- return false;
- }
}
msleep(delay_between_attempts);
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.h
index 0b18aa35c33c..574b083e0936 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training.h
@@ -104,6 +104,7 @@ void start_clock_recovery_pattern_early(struct dc_link *link,
void dp_decide_training_settings(
struct dc_link *link,
+ const struct link_resource *link_res,
const struct dc_link_settings *link_settings,
struct link_training_settings *lt_settings);
@@ -117,6 +118,7 @@ enum dc_dp_training_pattern decide_cr_training_pattern(
const struct dc_link_settings *link_settings);
enum dc_dp_training_pattern decide_eq_training_pattern(struct dc_link *link,
+ const struct link_resource *link_res,
const struct dc_link_settings *link_settings);
enum lttpr_mode dp_decide_lttpr_mode(struct dc_link *link,
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.c
index db87cfe37b5c..11565f187ac7 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.c
@@ -204,6 +204,7 @@ enum link_training_result dp_perform_128b_132b_link_training(
struct link_training_settings legacy_settings;
decide_8b_10b_training_settings(link,
+ link_res,
&lt_settings->link_settings,
&legacy_settings);
return dp_perform_8b_10b_link_training(link, link_res, &legacy_settings);
@@ -227,6 +228,7 @@ enum link_training_result dp_perform_128b_132b_link_training(
}
void decide_128b_132b_training_settings(struct dc_link *link,
+ const struct link_resource *link_res,
const struct dc_link_settings *link_settings,
struct link_training_settings *lt_settings)
{
@@ -238,7 +240,7 @@ void decide_128b_132b_training_settings(struct dc_link *link,
LINK_SPREAD_05_DOWNSPREAD_30KHZ;
lt_settings->pattern_for_cr = decide_cr_training_pattern(link_settings);
- lt_settings->pattern_for_eq = decide_eq_training_pattern(link, link_settings);
+ lt_settings->pattern_for_eq = decide_eq_training_pattern(link, link_res, link_settings);
lt_settings->eq_pattern_time = 2500;
lt_settings->eq_wait_time_limit = 400000;
lt_settings->eq_loop_count_limit = 20;
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.h
index 2147f24efc8b..901a42edafa1 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_128b_132b.h
@@ -34,6 +34,7 @@ enum link_training_result dp_perform_128b_132b_link_training(
struct link_training_settings *lt_settings);
void decide_128b_132b_training_settings(struct dc_link *link,
+ const struct link_resource *link_res,
const struct dc_link_settings *link_settings,
struct link_training_settings *lt_settings);
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.c
index 3bdce32a85e3..66d0fb1b9b9d 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.c
@@ -35,22 +35,41 @@
#define DC_LOGGER \
link->ctx->logger
+static void get_default_8b_10b_lttpr_aux_rd_interval(
+ union training_aux_rd_interval *training_rd_interval)
+{
+ /* LTTPR are required to program DPCD 0000Eh to 0x4 (16ms) upon AUX
+ * read reply to this register. Since old sinks with DPCD rev 1.1
+ * and earlier may not support this register, assume the mandatory
+ * value is programmed by the LTTPR to avoid AUX timeout issues.
+ */
+ training_rd_interval->raw = 0x4;
+}
+
static int32_t get_cr_training_aux_rd_interval(struct dc_link *link,
- const struct dc_link_settings *link_settings)
+ const struct dc_link_settings *link_settings,
+ enum lttpr_mode lttpr_mode)
{
union training_aux_rd_interval training_rd_interval;
uint32_t wait_in_micro_secs = 100;
memset(&training_rd_interval, 0, sizeof(training_rd_interval));
- if (link_dp_get_encoding_format(link_settings) == DP_8b_10b_ENCODING &&
- link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_12) {
- core_link_read_dpcd(
- link,
- DP_TRAINING_AUX_RD_INTERVAL,
- (uint8_t *)&training_rd_interval,
- sizeof(training_rd_interval));
- if (training_rd_interval.bits.TRAINIG_AUX_RD_INTERVAL)
- wait_in_micro_secs = training_rd_interval.bits.TRAINIG_AUX_RD_INTERVAL * 4000;
+ if (link_dp_get_encoding_format(link_settings) == DP_8b_10b_ENCODING) {
+ if (link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_12)
+ core_link_read_dpcd(
+ link,
+ DP_TRAINING_AUX_RD_INTERVAL,
+ (uint8_t *)&training_rd_interval,
+ sizeof(training_rd_interval));
+ else if (dp_is_lttpr_present(link))
+ get_default_8b_10b_lttpr_aux_rd_interval(&training_rd_interval);
+
+ if (training_rd_interval.raw != 0) {
+ if (lttpr_mode != LTTPR_MODE_NON_TRANSPARENT)
+ wait_in_micro_secs = 400;
+ if (training_rd_interval.bits.TRAINIG_AUX_RD_INTERVAL)
+ wait_in_micro_secs = training_rd_interval.bits.TRAINIG_AUX_RD_INTERVAL * 4000;
+ }
}
return wait_in_micro_secs;
}
@@ -68,13 +87,15 @@ static uint32_t get_eq_training_aux_rd_interval(
DP_128B132B_TRAINING_AUX_RD_INTERVAL,
(uint8_t *)&training_rd_interval,
sizeof(training_rd_interval));
- } else if (link_dp_get_encoding_format(link_settings) == DP_8b_10b_ENCODING &&
- link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_12) {
- core_link_read_dpcd(
- link,
- DP_TRAINING_AUX_RD_INTERVAL,
- (uint8_t *)&training_rd_interval,
- sizeof(training_rd_interval));
+ } else if (link_dp_get_encoding_format(link_settings) == DP_8b_10b_ENCODING) {
+ if (link->dpcd_caps.dpcd_rev.raw >= DPCD_REV_12)
+ core_link_read_dpcd(
+ link,
+ DP_TRAINING_AUX_RD_INTERVAL,
+ (uint8_t *)&training_rd_interval,
+ sizeof(training_rd_interval));
+ else if (dp_is_lttpr_present(link))
+ get_default_8b_10b_lttpr_aux_rd_interval(&training_rd_interval);
}
switch (training_rd_interval.bits.TRAINIG_AUX_RD_INTERVAL) {
@@ -90,7 +111,8 @@ static uint32_t get_eq_training_aux_rd_interval(
}
void decide_8b_10b_training_settings(
- struct dc_link *link,
+ struct dc_link *link,
+ const struct link_resource *link_res,
const struct dc_link_settings *link_setting,
struct link_training_settings *lt_settings)
{
@@ -110,16 +132,24 @@ void decide_8b_10b_training_settings(
*/
lt_settings->link_settings.link_spread = link->dp_ss_off ?
LINK_SPREAD_DISABLED : LINK_SPREAD_05_DOWNSPREAD_30KHZ;
- lt_settings->cr_pattern_time = get_cr_training_aux_rd_interval(link, link_setting);
lt_settings->eq_pattern_time = get_eq_training_aux_rd_interval(link, link_setting);
lt_settings->pattern_for_cr = decide_cr_training_pattern(link_setting);
- lt_settings->pattern_for_eq = decide_eq_training_pattern(link, link_setting);
+ lt_settings->pattern_for_eq = decide_eq_training_pattern(link, link_res, link_setting);
lt_settings->enhanced_framing = 1;
lt_settings->should_set_fec_ready = true;
lt_settings->disallow_per_lane_settings = true;
lt_settings->always_match_dpcd_with_hw_lane_settings = true;
lt_settings->lttpr_mode = dp_decide_8b_10b_lttpr_mode(link);
+ lt_settings->cr_pattern_time = get_cr_training_aux_rd_interval(link, link_setting, lt_settings->lttpr_mode);
dp_hw_to_dpcd_lane_settings(lt_settings, lt_settings->hw_lane_settings, lt_settings->dpcd_lane_settings);
+
+ /* Some embedded LTTPRs rely on receiving TPS2 before LT to interop reliably with sensitive VGA dongles
+ * This allows these LTTPRs to minimize freq/phase and skew variation during lock and deskew sequences
+ */
+ if ((link->chip_caps & AMD_EXT_DISPLAY_PATH_CAPS__EXT_CHIP_MASK) ==
+ AMD_EXT_DISPLAY_PATH_CAPS__DP_EARLY_8B10B_TPS2) {
+ lt_settings->lttpr_early_tps2 = true;
+ }
}
enum lttpr_mode dp_decide_8b_10b_lttpr_mode(struct dc_link *link)
@@ -151,6 +181,42 @@ enum lttpr_mode dp_decide_8b_10b_lttpr_mode(struct dc_link *link)
return LTTPR_MODE_NON_LTTPR;
}
+static void set_link_settings_and_perform_early_tps2_retimer_pre_lt_sequence(struct dc_link *link,
+ const struct link_resource *link_res,
+ struct link_training_settings *lt_settings,
+ uint32_t lttpr_count)
+{
+ /* Vendor-specific LTTPR early TPS2 sequence:
+ * 1. Output TPS2
+ * 2. Wait 400us
+ * 3. Set link settings as usual
+ * 4. Write TPS1 to DP_TRAINING_PATTERN_SET_PHY_REPEATERx targeting LTTPR closest to host
+ * 5. Wait 1ms
+ * 6. Begin link training as usual
+ * */
+
+ uint32_t closest_lttpr_address_offset = dp_get_closest_lttpr_offset(lttpr_count);
+
+ union dpcd_training_pattern dpcd_pattern = {0};
+
+ dpcd_pattern.v1_4.TRAINING_PATTERN_SET = 1;
+ dpcd_pattern.v1_4.SCRAMBLING_DISABLE = 1;
+
+ DC_LOG_HW_LINK_TRAINING("%s\n GPU sends TPS2. Wait 400us.\n", __func__);
+
+ dp_set_hw_training_pattern(link, link_res, DP_TRAINING_PATTERN_SEQUENCE_2, DPRX);
+
+ dp_set_hw_lane_settings(link, link_res, lt_settings, DPRX);
+
+ udelay(400);
+
+ dpcd_set_link_settings(link, lt_settings);
+
+ core_link_write_dpcd(link, DP_TRAINING_PATTERN_SET_PHY_REPEATER1 + closest_lttpr_address_offset, &dpcd_pattern.raw, 1);
+
+ udelay(1000);
+ }
+
enum link_training_result perform_8b_10b_clock_recovery_sequence(
struct dc_link *link,
const struct link_resource *link_res,
@@ -361,7 +427,7 @@ enum link_training_result dp_perform_8b_10b_link_training(
{
enum link_training_result status = LINK_TRAINING_SUCCESS;
- uint8_t repeater_cnt;
+ uint8_t repeater_cnt = dp_parse_lttpr_repeater_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt);
uint8_t repeater_id;
uint8_t lane = 0;
@@ -369,14 +435,16 @@ enum link_training_result dp_perform_8b_10b_link_training(
start_clock_recovery_pattern_early(link, link_res, lt_settings, DPRX);
/* 1. set link rate, lane count and spread. */
- dpcd_set_link_settings(link, lt_settings);
+ if (lt_settings->lttpr_early_tps2)
+ set_link_settings_and_perform_early_tps2_retimer_pre_lt_sequence(link, link_res, lt_settings, repeater_cnt);
+ else
+ dpcd_set_link_settings(link, lt_settings);
if (lt_settings->lttpr_mode == LTTPR_MODE_NON_TRANSPARENT) {
/* 2. perform link training (set link training done
* to false is done as well)
*/
- repeater_cnt = dp_parse_lttpr_repeater_count(link->dpcd_caps.lttpr_caps.phy_repeater_cnt);
for (repeater_id = repeater_cnt; (repeater_id > 0 && status == LINK_TRAINING_SUCCESS);
repeater_id--) {
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.h
index d26de15ce954..ea0de701d83f 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_8b_10b.h
@@ -54,7 +54,8 @@ enum link_training_result perform_8b_10b_channel_equalization_sequence(
enum lttpr_mode dp_decide_8b_10b_lttpr_mode(struct dc_link *link);
void decide_8b_10b_training_settings(
- struct dc_link *link,
+ struct dc_link *link,
+ const struct link_resource *link_res,
const struct dc_link_settings *link_setting,
struct link_training_settings *lt_settings);
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_auxless.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_auxless.c
index 4c6b886a9da8..f99d26290bc0 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_auxless.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_auxless.c
@@ -39,6 +39,7 @@ bool dp_perform_link_training_skip_aux(
dp_decide_training_settings(
link,
+ link_res,
link_setting,
&lt_settings);
override_training_settings(
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_dpia.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_dpia.c
index 39e4b7dc9588..603537ffd128 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_dpia.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_dpia.c
@@ -110,6 +110,7 @@ static enum link_training_result dpia_configure_link(
dp_decide_training_settings(
link,
+ link_res,
link_setting,
lt_settings);
@@ -129,11 +130,14 @@ static enum link_training_result dpia_configure_link(
if (status != DC_OK && link->is_hpd_pending)
return LINK_TRAINING_ABORT;
- if (link->preferred_training_settings.fec_enable != NULL)
- fec_enable = *link->preferred_training_settings.fec_enable;
- else
- fec_enable = true;
- status = dp_set_fec_ready(link, link_res, fec_enable);
+ if (link_dp_get_encoding_format(link_setting) == DP_8b_10b_ENCODING) {
+ if (link->preferred_training_settings.fec_enable != NULL)
+ fec_enable = *link->preferred_training_settings.fec_enable;
+ else
+ fec_enable = true;
+ status = dp_set_fec_ready(link, link_res, fec_enable);
+ }
+
if (status != DC_OK && link->is_hpd_pending)
return LINK_TRAINING_ABORT;
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c
index ccf8096dde29..ce174ce5579c 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_dp_training_fixed_vs_pe_retimer.c
@@ -270,7 +270,8 @@ enum link_training_result dp_perform_fixed_vs_pe_training_sequence(
rate = get_dpcd_link_rate(&lt_settings->link_settings);
- if (!link->dpcd_caps.lttpr_caps.main_link_channel_coding.bits.DP_128b_132b_SUPPORTED) {
+ // Only perform toggle if FIXED_VS LTTPR reports no IEEE OUI
+ if (memcmp("\x0,\x0,\x0", &link->dpcd_caps.lttpr_caps.lttpr_ieee_oui[0], 3) == 0) {
/* Vendor specific: Toggle link rate */
toggle_rate = (rate == 0x6) ? 0xA : 0x6;
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c
index e0e3bb865359..e7927b8f5ba3 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.c
@@ -161,6 +161,9 @@ bool edp_set_backlight_level_nits(struct dc_link *link,
link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT))
return false;
+ if (link->is_dds && !link->dpcd_caps.panel_luminance_control)
+ return true;
+
// use internal backlight control if dmub capabilities are not present
if (link->backlight_control_type == BACKLIGHT_CONTROL_VESA_AUX &&
!link->dc->caps.dmub_caps.aux_backlight_support) {
@@ -173,6 +176,15 @@ bool edp_set_backlight_level_nits(struct dc_link *link,
target_luminance = (struct target_luminance_value *)&backlight_millinits;
+ //make sure we disable AMD ABC first.
+ core_link_read_dpcd(link, DP_SOURCE_BACKLIGHT_CONTROL,
+ &backlight_enable, sizeof(uint8_t));
+ if (backlight_enable) {
+ backlight_enable = 0;
+ core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_CONTROL,
+ &backlight_enable, 1);
+ }
+
core_link_read_dpcd(link, DP_EDP_BACKLIGHT_MODE_SET_REGISTER,
&backlight_enable, sizeof(uint8_t));
@@ -193,10 +205,22 @@ bool edp_set_backlight_level_nits(struct dc_link *link,
*(uint16_t *)&dpcd_backlight_set.backlight_transition_time_ms = (uint16_t)transition_time_in_ms;
uint8_t backlight_control = isHDR ? 1 : 0;
+ uint8_t backlight_enable = 0;
+
// OLEDs have no PWM, they can only use AUX
if (link->dpcd_sink_ext_caps.bits.oled == 1)
backlight_control = 1;
+ //make sure we disable VESA ABC first.
+ core_link_read_dpcd(link, DP_EDP_BACKLIGHT_MODE_SET_REGISTER,
+ &backlight_enable, sizeof(uint8_t));
+
+ if (backlight_enable & DP_EDP_PANEL_LUMINANCE_CONTROL_ENABLE) {
+ backlight_enable &= ~DP_EDP_PANEL_LUMINANCE_CONTROL_ENABLE;
+ core_link_write_dpcd(link, DP_EDP_BACKLIGHT_MODE_SET_REGISTER,
+ &backlight_enable, sizeof(backlight_enable));
+ }
+
if (core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_LEVEL,
(uint8_t *)(&dpcd_backlight_set),
sizeof(dpcd_backlight_set)) != DC_OK)
@@ -222,6 +246,8 @@ bool edp_get_backlight_level_nits(struct dc_link *link,
link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT))
return false;
+ if (link->is_dds)
+ return false;
if (!core_link_read_dpcd(link, DP_SOURCE_BACKLIGHT_CURRENT_PEAK,
dpcd_backlight_get.raw,
sizeof(union dpcd_source_backlight_get)))
@@ -248,6 +274,8 @@ bool edp_backlight_enable_aux(struct dc_link *link, bool enable)
link->connector_signal != SIGNAL_TYPE_DISPLAY_PORT))
return false;
+ if (link->is_dds)
+ return true;
if (core_link_write_dpcd(link, DP_SOURCE_BACKLIGHT_ENABLE,
&backlight_enable, 1) != DC_OK)
return false;
@@ -524,7 +552,7 @@ bool edp_set_backlight_level(const struct dc_link *link,
struct dc *dc = link->ctx->dc;
uint32_t backlight_pwm_u16_16 = backlight_level_params->backlight_pwm_u16_16;
uint32_t frame_ramp = backlight_level_params->frame_ramp;
- DC_LOGGER_INIT(link->ctx->logger);
+
DC_LOG_BACKLIGHT("New Backlight level: %d (0x%X)\n",
backlight_pwm_u16_16, backlight_pwm_u16_16);
@@ -675,6 +703,18 @@ bool edp_setup_psr(struct dc_link *link,
if (!link)
return false;
+ //Clear PSR cfg
+ memset(&psr_configuration, 0, sizeof(psr_configuration));
+ dm_helpers_dp_write_dpcd(
+ link->ctx,
+ link,
+ DP_PSR_EN_CFG,
+ &psr_configuration.raw,
+ sizeof(psr_configuration.raw));
+
+ if (link->psr_settings.psr_version == DC_PSR_VERSION_UNSUPPORTED)
+ return false;
+
dc = link->ctx->dc;
dmcu = dc->res_pool->dmcu;
psr = dc->res_pool->psr;
@@ -685,9 +725,6 @@ bool edp_setup_psr(struct dc_link *link,
if (!dc_get_edp_link_panel_inst(dc, link, &panel_inst))
return false;
-
- memset(&psr_configuration, 0, sizeof(psr_configuration));
-
psr_configuration.bits.ENABLE = 1;
psr_configuration.bits.CRC_VERIFICATION = 1;
psr_configuration.bits.FRAME_CAPTURE_INDICATION =
@@ -907,7 +944,7 @@ bool edp_set_replay_allow_active(struct dc_link *link, const bool *allow_active,
// TODO: Handle mux change case if force_static is set
// If force_static is set, just change the replay_allow_active state directly
if (replay != NULL && link->replay_settings.replay_feature_enabled)
- replay->funcs->replay_enable(replay, *allow_active, wait, panel_inst);
+ replay->funcs->replay_enable(replay, *allow_active, wait, panel_inst, link);
link->replay_settings.replay_allow_active = *allow_active;
}
@@ -950,6 +987,16 @@ bool edp_setup_replay(struct dc_link *link, const struct dc_stream_state *stream
if (!link)
return false;
+ //Clear Replay config
+ dm_helpers_dp_write_dpcd(link->ctx, link,
+ DP_SINK_PR_ENABLE_AND_CONFIGURATION,
+ (uint8_t *)&(replay_config.raw), sizeof(uint8_t));
+
+ if (!(link->replay_settings.config.replay_supported))
+ return false;
+
+ link->replay_settings.config.replay_error_status.raw = 0;
+
dc = link->ctx->dc;
replay = dc->res_pool->replay;
@@ -1003,6 +1050,9 @@ bool edp_setup_replay(struct dc_link *link, const struct dc_stream_state *stream
&alpm_config.raw,
sizeof(alpm_config.raw));
}
+
+ link->replay_settings.config.replay_video_conferencing_optimization_enabled = false;
+
return true;
}
@@ -1111,11 +1161,11 @@ static struct abm *get_abm_from_stream_res(const struct dc_link *link)
struct abm *abm = NULL;
for (i = 0; i < MAX_PIPES; i++) {
- struct pipe_ctx pipe_ctx = dc->current_state->res_ctx.pipe_ctx[i];
- struct dc_stream_state *stream = pipe_ctx.stream;
+ struct pipe_ctx *pipe_ctx = &dc->current_state->res_ctx.pipe_ctx[i];
+ struct dc_stream_state *stream = pipe_ctx->stream;
if (stream && stream->link == link) {
- abm = pipe_ctx.stream_res.abm;
+ abm = pipe_ctx->stream_res.abm;
break;
}
}
@@ -1151,6 +1201,16 @@ int edp_get_target_backlight_pwm(const struct dc_link *link)
return (int) abm->funcs->get_target_backlight(abm);
}
+bool is_smartmux_suported(struct dc_link *link)
+{
+ if (link->dc->caps.is_apu)
+ return false;
+ if (!link->dc->config.smart_mux_version)
+ return false;
+
+ return true;
+}
+
static void edp_set_assr_enable(const struct dc *pDC, struct dc_link *link,
struct link_resource *link_res, bool enable)
{
diff --git a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h
index bcfa6ac5d4e7..4a475d5b9dde 100644
--- a/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h
+++ b/drivers/gpu/drm/amd/display/dc/link/protocols/link_edp_panel_control.h
@@ -30,6 +30,7 @@
enum dp_panel_mode dp_get_panel_mode(struct dc_link *link);
void dp_set_panel_mode(struct dc_link *link, enum dp_panel_mode panel_mode);
bool set_default_brightness_aux(struct dc_link *link);
+bool is_smartmux_suported(struct dc_link *link);
void edp_panel_backlight_power_on(struct dc_link *link, bool wait_for_hpd);
int edp_get_backlight_level(const struct dc_link *link);
bool edp_get_backlight_level_nits(struct dc_link *link,
diff --git a/drivers/gpu/drm/amd/display/dc/mmhubbub/Makefile b/drivers/gpu/drm/amd/display/dc/mmhubbub/Makefile
index eab196c57c6c..2d4b7a85847d 100644
--- a/drivers/gpu/drm/amd/display/dc/mmhubbub/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/mmhubbub/Makefile
@@ -50,5 +50,5 @@ MMHUBBUB_DCN35 = dcn35_mmhubbub.o
AMD_DAL_MMHUBBUB_DCN35 = $(addprefix $(AMDDALPATH)/dc/mmhubbub/dcn35/,$(MMHUBBUB_DCN35))
AMD_DISPLAY_FILES += $(AMD_DAL_MMHUBBUB_DCN35)
-
endif
+
diff --git a/drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.c b/drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.c
index f2f55565e98a..b23c64004dd5 100644
--- a/drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.c
+++ b/drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.c
@@ -142,22 +142,6 @@ struct mpcc *mpc1_get_mpcc_for_dpp(struct mpc_tree *tree, int dpp_id)
return NULL;
}
-bool mpc1_is_mpcc_idle(struct mpc *mpc, int mpcc_id)
-{
- struct dcn10_mpc *mpc10 = TO_DCN10_MPC(mpc);
- unsigned int top_sel;
- unsigned int opp_id;
- unsigned int idle;
-
- REG_GET(MPCC_TOP_SEL[mpcc_id], MPCC_TOP_SEL, &top_sel);
- REG_GET(MPCC_OPP_ID[mpcc_id], MPCC_OPP_ID, &opp_id);
- REG_GET(MPCC_STATUS[mpcc_id], MPCC_IDLE, &idle);
- if (top_sel == 0xf && opp_id == 0xf && idle)
- return true;
- else
- return false;
-}
-
void mpc1_assert_mpcc_idle_before_connect(struct mpc *mpc, int mpcc_id)
{
struct dcn10_mpc *mpc10 = TO_DCN10_MPC(mpc);
diff --git a/drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.h b/drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.h
index dbfffc6383dc..874e36e39e1b 100644
--- a/drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/mpc/dcn10/dcn10_mpc.h
@@ -173,10 +173,6 @@ void mpc1_update_stereo_mix(
struct mpcc_sm_cfg *sm_cfg,
int mpcc_id);
-bool mpc1_is_mpcc_idle(
- struct mpc *mpc,
- int mpcc_id);
-
void mpc1_assert_mpcc_idle_before_connect(
struct mpc *mpc,
int mpcc_id);
diff --git a/drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.c b/drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.c
index fe26fde12eeb..85298b8a1b5e 100644
--- a/drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.c
+++ b/drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.c
@@ -110,6 +110,23 @@ void mpc3_disable_dwb_mux(
MPC_DWB0_MUX, 0xf);
}
+void mpc3_set_out_rate_control(
+ struct mpc *mpc,
+ int opp_id,
+ bool enable,
+ bool rate_2x_mode,
+ struct mpc_dwb_flow_control *flow_control)
+{
+ struct dcn30_mpc *mpc30 = TO_DCN30_MPC(mpc);
+
+ /* Always disable mpc out rate and flow control.
+ * MPC flow rate control is not needed for DCN30 and above.
+ */
+ REG_UPDATE_2(MUX[opp_id],
+ MPC_OUT_RATE_CONTROL_DISABLE, 1,
+ MPC_OUT_RATE_CONTROL, 0);
+}
+
enum dc_lut_mode mpc3_get_ogam_current(struct mpc *mpc, int mpcc_id)
{
/*Contrary to DCN2 and DCN1 wherein a single status register field holds this info;
@@ -1519,6 +1536,7 @@ static const struct mpc_funcs dcn30_mpc_funcs = {
.set_dwb_mux = mpc3_set_dwb_mux,
.disable_dwb_mux = mpc3_disable_dwb_mux,
.is_dwb_idle = mpc3_is_dwb_idle,
+ .set_out_rate_control = mpc3_set_out_rate_control,
.set_gamut_remap = mpc3_set_gamut_remap,
.program_shaper = mpc3_program_shaper,
.acquire_rmu = mpcc3_acquire_rmu,
diff --git a/drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.h b/drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.h
index ce93003dae01..103f29900a2c 100644
--- a/drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/mpc/dcn30/dcn30_mpc.h
@@ -1085,6 +1085,13 @@ bool mpc3_is_dwb_idle(
struct mpc *mpc,
int dwb_id);
+void mpc3_set_out_rate_control(
+ struct mpc *mpc,
+ int opp_id,
+ bool enable,
+ bool rate_2x_mode,
+ struct mpc_dwb_flow_control *flow_control);
+
void mpc3_power_on_ogam_lut(
struct mpc *mpc, int mpcc_id,
bool power_on);
diff --git a/drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.c b/drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.c
index a0e9e9f0441a..6f0e017a8ae2 100644
--- a/drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.c
+++ b/drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.c
@@ -30,7 +30,6 @@
#include "basics/conversion.h"
#include "dcn10/dcn10_cm_common.h"
#include "dc.h"
-#include "dcn401/dcn401_mpc.h"
#define REG(reg)\
mpc30->mpc_regs->reg
@@ -370,275 +369,279 @@ void mpc32_program_shaper_luta_settings(
MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_BASE_B, params->corner_points[1].red.custom_float_y);
curve = params->arr_curve_points;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_0_1[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_2_3[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_4_5[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_6_7[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_8_9[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_10_11[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_12_13[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_14_15[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_16_17[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_18_19[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_20_21[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_22_23[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_24_25[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_26_27[mpcc_id], 0,
+ if (curve) {
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_0_1[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_28_29[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_30_31[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_32_33[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-}
-
-
-void mpc32_program_shaper_lutb_settings(
- struct mpc *mpc,
- const struct pwl_params *params,
- uint32_t mpcc_id)
-{
- const struct gamma_curve *curve;
- struct dcn30_mpc *mpc30 = TO_DCN30_MPC(mpc);
-
- REG_SET_2(MPCC_MCM_SHAPER_RAMB_START_CNTL_B[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_B, params->corner_points[0].blue.custom_float_x,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_SEGMENT_B, 0);
- REG_SET_2(MPCC_MCM_SHAPER_RAMB_START_CNTL_G[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_B, params->corner_points[0].green.custom_float_x,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_SEGMENT_B, 0);
- REG_SET_2(MPCC_MCM_SHAPER_RAMB_START_CNTL_R[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_B, params->corner_points[0].red.custom_float_x,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_SEGMENT_B, 0);
-
- REG_SET_2(MPCC_MCM_SHAPER_RAMB_END_CNTL_B[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_B, params->corner_points[1].blue.custom_float_x,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_BASE_B, params->corner_points[1].blue.custom_float_y);
- REG_SET_2(MPCC_MCM_SHAPER_RAMB_END_CNTL_G[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_B, params->corner_points[1].green.custom_float_x,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_BASE_B, params->corner_points[1].green.custom_float_y);
- REG_SET_2(MPCC_MCM_SHAPER_RAMB_END_CNTL_R[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_B, params->corner_points[1].red.custom_float_x,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_BASE_B, params->corner_points[1].red.custom_float_y);
-
- curve = params->arr_curve_points;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_0_1[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_2_3[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_2_3[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
-
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_4_5[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_4_5[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_6_7[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_6_7[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_8_9[mpcc_id], 0,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
- MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_8_9[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_10_11[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_10_11[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_12_13[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_12_13[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_14_15[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_14_15[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_16_17[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_16_17[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_18_19[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_18_19[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_20_21[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_20_21[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_22_23[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_22_23[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_24_25[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_24_25[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_26_27[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_26_27[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_28_29[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_28_29[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_30_31[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_30_31[mpcc_id], 0,
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMA_REGION_32_33[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+ }
+}
+
+
+void mpc32_program_shaper_lutb_settings(
+ struct mpc *mpc,
+ const struct pwl_params *params,
+ uint32_t mpcc_id)
+{
+ const struct gamma_curve *curve;
+ struct dcn30_mpc *mpc30 = TO_DCN30_MPC(mpc);
+
+ REG_SET_2(MPCC_MCM_SHAPER_RAMB_START_CNTL_B[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_B, params->corner_points[0].blue.custom_float_x,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_SEGMENT_B, 0);
+ REG_SET_2(MPCC_MCM_SHAPER_RAMB_START_CNTL_G[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_B, params->corner_points[0].green.custom_float_x,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_SEGMENT_B, 0);
+ REG_SET_2(MPCC_MCM_SHAPER_RAMB_START_CNTL_R[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_B, params->corner_points[0].red.custom_float_x,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_START_SEGMENT_B, 0);
- curve += 2;
- REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_32_33[mpcc_id], 0,
+ REG_SET_2(MPCC_MCM_SHAPER_RAMB_END_CNTL_B[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_B, params->corner_points[1].blue.custom_float_x,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_BASE_B, params->corner_points[1].blue.custom_float_y);
+ REG_SET_2(MPCC_MCM_SHAPER_RAMB_END_CNTL_G[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_B, params->corner_points[1].green.custom_float_x,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_BASE_B, params->corner_points[1].green.custom_float_y);
+ REG_SET_2(MPCC_MCM_SHAPER_RAMB_END_CNTL_R[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_B, params->corner_points[1].red.custom_float_x,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION_END_BASE_B, params->corner_points[1].red.custom_float_y);
+
+ curve = params->arr_curve_points;
+ if (curve) {
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_0_1[mpcc_id], 0,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_2_3[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_4_5[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_6_7[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_8_9[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_10_11[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_12_13[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_14_15[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_16_17[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_18_19[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_20_21[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_22_23[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_24_25[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_26_27[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_28_29[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_30_31[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+
+ curve += 2;
+ REG_SET_4(MPCC_MCM_SHAPER_RAMB_REGION_32_33[mpcc_id], 0,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_LUT_OFFSET, curve[0].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION0_NUM_SEGMENTS, curve[0].segments_num,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_LUT_OFFSET, curve[1].offset,
+ MPCC_MCM_SHAPER_RAMA_EXP_REGION1_NUM_SEGMENTS, curve[1].segments_num);
+ }
}
@@ -875,7 +878,7 @@ void mpc32_set3dlut_ram10(
}
-static void mpc32_set_3dlut_mode(
+void mpc32_set_3dlut_mode(
struct mpc *mpc,
enum dc_lut_mode mode,
bool is_color_channel_12bits,
@@ -1018,8 +1021,6 @@ static const struct mpc_funcs dcn32_mpc_funcs = {
.power_on_mpc_mem_pwr = mpc3_power_on_ogam_lut,
.get_mpc_out_mux = mpc1_get_mpc_out_mux,
.set_bg_color = mpc1_set_bg_color,
- .set_movable_cm_location = mpc401_set_movable_cm_location,
- .populate_lut = mpc401_populate_lut,
};
diff --git a/drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.h b/drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.h
index 9622518826c9..8c9b20bcca85 100644
--- a/drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/mpc/dcn32/dcn32_mpc.h
@@ -391,4 +391,12 @@ void mpc32_select_3dlut_ram(
enum dc_lut_mode mode,
bool is_color_channel_12bits,
uint32_t mpcc_id);
+
+void mpc32_set_3dlut_mode(
+ struct mpc *mpc,
+ enum dc_lut_mode mode,
+ bool is_color_channel_12bits,
+ bool is_lut_size17x17x17,
+ uint32_t mpcc_id);
+
#endif //__DC_MPCC_DCN32_H__
diff --git a/drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.c b/drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.c
index 37ab5a4eefc7..f3fb3fe13757 100644
--- a/drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.c
+++ b/drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.c
@@ -40,23 +40,13 @@
#define FN(reg_name, field_name) \
mpc401->mpc_shift->field_name, mpc401->mpc_mask->field_name
-static void mpc401_update_3dlut_fast_load_select(struct mpc *mpc, int mpcc_id, int hubp_idx)
+void mpc401_update_3dlut_fast_load_select(struct mpc *mpc, int mpcc_id, int hubp_idx)
{
struct dcn401_mpc *mpc401 = TO_DCN401_MPC(mpc);
REG_SET(MPCC_MCM_3DLUT_FAST_LOAD_SELECT[mpcc_id], 0, MPCC_MCM_3DLUT_FL_SEL, hubp_idx);
}
-static void mpc401_get_3dlut_fast_load_status(struct mpc *mpc, int mpcc_id, uint32_t *done, uint32_t *soft_underflow, uint32_t *hard_underflow)
-{
- struct dcn401_mpc *mpc401 = TO_DCN401_MPC(mpc);
-
- REG_GET_3(MPCC_MCM_3DLUT_FAST_LOAD_STATUS[mpcc_id],
- MPCC_MCM_3DLUT_FL_DONE, done,
- MPCC_MCM_3DLUT_FL_SOFT_UNDERFLOW, soft_underflow,
- MPCC_MCM_3DLUT_FL_HARD_UNDERFLOW, hard_underflow);
-}
-
void mpc401_set_movable_cm_location(struct mpc *mpc, enum mpcc_movable_cm_location location, int mpcc_id)
{
struct dcn401_mpc *mpc401 = TO_DCN401_MPC(mpc);
@@ -304,7 +294,7 @@ void mpc401_program_3dlut_size(struct mpc *mpc, bool is_17x17x17, int mpcc_id)
REG_UPDATE(MPCC_MCM_3DLUT_MODE[mpcc_id], MPCC_MCM_3DLUT_SIZE, is_17x17x17 ? 0 : 1);
}
-static void program_gamut_remap(
+void mpc_program_gamut_remap(
struct mpc *mpc,
unsigned int mpcc_id,
const uint16_t *regval,
@@ -436,7 +426,7 @@ void mpc401_set_gamut_remap(
if (adjust->gamut_adjust_type != GRAPHICS_GAMUT_ADJUST_TYPE_SW) {
/* Bypass / Disable if type is bypass or hw */
- program_gamut_remap(mpc, mpcc_id, NULL,
+ mpc_program_gamut_remap(mpc, mpcc_id, NULL,
adjust->mpcc_gamut_remap_block_id, MPCC_GAMUT_REMAP_MODE_SELECT_0);
} else {
struct fixed31_32 arr_matrix[12];
@@ -470,12 +460,12 @@ void mpc401_set_gamut_remap(
else
mode_select = MPCC_GAMUT_REMAP_MODE_SELECT_2;
- program_gamut_remap(mpc, mpcc_id, arr_reg_val,
+ mpc_program_gamut_remap(mpc, mpcc_id, arr_reg_val,
adjust->mpcc_gamut_remap_block_id, mode_select);
}
}
-static void read_gamut_remap(struct mpc *mpc,
+void mpc_read_gamut_remap(struct mpc *mpc,
int mpcc_id,
uint16_t *regval,
enum mpcc_gamut_remap_id gamut_remap_block_id,
@@ -571,9 +561,9 @@ void mpc401_get_gamut_remap(struct mpc *mpc,
struct mpc_grph_gamut_adjustment *adjust)
{
uint16_t arr_reg_val[12] = {0};
- uint32_t mode_select;
+ uint32_t mode_select = MPCC_GAMUT_REMAP_MODE_SELECT_0;
- read_gamut_remap(mpc, mpcc_id, arr_reg_val, adjust->mpcc_gamut_remap_block_id, &mode_select);
+ mpc_read_gamut_remap(mpc, mpcc_id, arr_reg_val, adjust->mpcc_gamut_remap_block_id, &mode_select);
if (mode_select == MPCC_GAMUT_REMAP_MODE_SELECT_0) {
adjust->gamut_adjust_type = GRAPHICS_GAMUT_ADJUST_TYPE_BYPASS;
@@ -618,7 +608,6 @@ static const struct mpc_funcs dcn401_mpc_funcs = {
.set_bg_color = mpc1_set_bg_color,
.set_movable_cm_location = mpc401_set_movable_cm_location,
.update_3dlut_fast_load_select = mpc401_update_3dlut_fast_load_select,
- .get_3dlut_fast_load_status = mpc401_get_3dlut_fast_load_status,
.populate_lut = mpc401_populate_lut,
.program_lut_read_write_control = mpc401_program_lut_read_write_control,
.program_lut_mode = mpc401_program_lut_mode,
diff --git a/drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.h b/drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.h
index af44054c2477..eb0c68d0b0c7 100644
--- a/drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.h
+++ b/drivers/gpu/drm/amd/display/dc/mpc/dcn401/dcn401_mpc.h
@@ -63,7 +63,7 @@
uint32_t MPC_MCM_SECOND_GAMUT_REMAP_C31_C32_B[MAX_MPCC]; \
uint32_t MPC_MCM_SECOND_GAMUT_REMAP_C33_C34_B[MAX_MPCC]; \
uint32_t MPCC_MCM_3DLUT_FAST_LOAD_SELECT[MAX_MPCC]; \
- uint32_t MPCC_MCM_3DLUT_FAST_LOAD_STATUS[MAX_MPCC]
+ uint32_t MPCC_MCM_3DLUT_FAST_LOAD_STATUS[MAX_MPCC];
#define MPC_COMMON_MASK_SH_LIST_DCN4_01(mask_sh) \
MPC_COMMON_MASK_SH_LIST_DCN32(mask_sh), \
@@ -183,7 +183,7 @@ struct dcn401_mpc_mask {
};
struct dcn401_mpc_registers {
- MPC_REG_VARIABLE_LIST_DCN4_01;
+ MPC_REG_VARIABLE_LIST_DCN4_01
};
struct dcn401_mpc {
@@ -236,4 +236,27 @@ void mpc401_get_gamut_remap(
int mpcc_id,
struct mpc_grph_gamut_adjustment *adjust);
+void mpc401_update_3dlut_fast_load_select(
+ struct mpc *mpc,
+ int mpcc_id,
+ int hubp_idx);
+
+void mpc_program_gamut_remap(
+ struct mpc *mpc,
+ unsigned int mpcc_id,
+ const uint16_t *regval,
+ enum mpcc_gamut_remap_id gamut_remap_block_id,
+ enum mpcc_gamut_remap_mode_select mode_select);
+
+void mpc_read_gamut_remap(struct mpc *mpc,
+ int mpcc_id,
+ uint16_t *regval,
+ enum mpcc_gamut_remap_id gamut_remap_block_id,
+ uint32_t *mode_select);
+
+void mpc401_update_3dlut_fast_load_select(
+ struct mpc *mpc,
+ int mpcc_id,
+ int hubp_idx);
+
#endif
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.c
index 097d06023e64..6f7b0f816f2a 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.c
@@ -302,7 +302,6 @@ void optc1_program_timing(
/* Enable stereo - only when we need to pack 3D frame. Other types
* of stereo handled in explicit call
*/
-
if (optc->funcs->is_two_pixels_per_container(&patched_crtc_timing) || optc1->opp_count == 2)
h_div = H_TIMING_DIV_BY2;
@@ -1313,7 +1312,7 @@ bool optc1_get_hw_timing(struct timing_generator *tg,
if (tg == NULL || hw_crtc_timing == NULL)
return false;
- optc1_read_otg_state(DCN10TG_FROM_TG(tg), &s);
+ optc1_read_otg_state(tg, &s);
hw_crtc_timing->h_total = s.h_total + 1;
hw_crtc_timing->h_addressable = s.h_total - ((s.h_total - s.h_blank_start) + s.h_blank_end);
@@ -1329,9 +1328,11 @@ bool optc1_get_hw_timing(struct timing_generator *tg,
}
-void optc1_read_otg_state(struct optc *optc1,
+void optc1_read_otg_state(struct timing_generator *optc,
struct dcn_otg_state *s)
{
+ struct optc *optc1 = DCN10TG_FROM_TG(optc);
+
REG_GET(OTG_CONTROL,
OTG_MASTER_EN, &s->otg_enabled);
@@ -1471,37 +1472,71 @@ bool optc1_configure_crc(struct timing_generator *optc,
if (!optc1_is_tg_enabled(optc))
return false;
- REG_WRITE(OTG_CRC_CNTL, 0);
+ if (!params->enable || params->reset)
+ REG_WRITE(OTG_CRC_CNTL, 0);
if (!params->enable)
return true;
/* Program frame boundaries */
- /* Window A x axis start and end. */
- REG_UPDATE_2(OTG_CRC0_WINDOWA_X_CONTROL,
- OTG_CRC0_WINDOWA_X_START, params->windowa_x_start,
- OTG_CRC0_WINDOWA_X_END, params->windowa_x_end);
-
- /* Window A y axis start and end. */
- REG_UPDATE_2(OTG_CRC0_WINDOWA_Y_CONTROL,
- OTG_CRC0_WINDOWA_Y_START, params->windowa_y_start,
- OTG_CRC0_WINDOWA_Y_END, params->windowa_y_end);
-
- /* Window B x axis start and end. */
- REG_UPDATE_2(OTG_CRC0_WINDOWB_X_CONTROL,
- OTG_CRC0_WINDOWB_X_START, params->windowb_x_start,
- OTG_CRC0_WINDOWB_X_END, params->windowb_x_end);
-
- /* Window B y axis start and end. */
- REG_UPDATE_2(OTG_CRC0_WINDOWB_Y_CONTROL,
- OTG_CRC0_WINDOWB_Y_START, params->windowb_y_start,
- OTG_CRC0_WINDOWB_Y_END, params->windowb_y_end);
-
- /* Set crc mode and selection, and enable. Only using CRC0*/
- REG_UPDATE_3(OTG_CRC_CNTL,
- OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
- OTG_CRC0_SELECT, params->selection,
- OTG_CRC_EN, 1);
+ switch (params->crc_eng_inst) {
+ case 0:
+ /* Window A x axis start and end. */
+ REG_UPDATE_2(OTG_CRC0_WINDOWA_X_CONTROL,
+ OTG_CRC0_WINDOWA_X_START, params->windowa_x_start,
+ OTG_CRC0_WINDOWA_X_END, params->windowa_x_end);
+
+ /* Window A y axis start and end. */
+ REG_UPDATE_2(OTG_CRC0_WINDOWA_Y_CONTROL,
+ OTG_CRC0_WINDOWA_Y_START, params->windowa_y_start,
+ OTG_CRC0_WINDOWA_Y_END, params->windowa_y_end);
+
+ /* Window B x axis start and end. */
+ REG_UPDATE_2(OTG_CRC0_WINDOWB_X_CONTROL,
+ OTG_CRC0_WINDOWB_X_START, params->windowb_x_start,
+ OTG_CRC0_WINDOWB_X_END, params->windowb_x_end);
+
+ /* Window B y axis start and end. */
+ REG_UPDATE_2(OTG_CRC0_WINDOWB_Y_CONTROL,
+ OTG_CRC0_WINDOWB_Y_START, params->windowb_y_start,
+ OTG_CRC0_WINDOWB_Y_END, params->windowb_y_end);
+
+ /* Set crc mode and selection, and enable.*/
+ REG_UPDATE_3(OTG_CRC_CNTL,
+ OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
+ OTG_CRC0_SELECT, params->selection,
+ OTG_CRC_EN, 1);
+ break;
+ case 1:
+ /* Window A x axis start and end. */
+ REG_UPDATE_2(OTG_CRC1_WINDOWA_X_CONTROL,
+ OTG_CRC1_WINDOWA_X_START, params->windowa_x_start,
+ OTG_CRC1_WINDOWA_X_END, params->windowa_x_end);
+
+ /* Window A y axis start and end. */
+ REG_UPDATE_2(OTG_CRC1_WINDOWA_Y_CONTROL,
+ OTG_CRC1_WINDOWA_Y_START, params->windowa_y_start,
+ OTG_CRC1_WINDOWA_Y_END, params->windowa_y_end);
+
+ /* Window B x axis start and end. */
+ REG_UPDATE_2(OTG_CRC1_WINDOWB_X_CONTROL,
+ OTG_CRC1_WINDOWB_X_START, params->windowb_x_start,
+ OTG_CRC1_WINDOWB_X_END, params->windowb_x_end);
+
+ /* Window B y axis start and end. */
+ REG_UPDATE_2(OTG_CRC1_WINDOWB_Y_CONTROL,
+ OTG_CRC1_WINDOWB_Y_START, params->windowb_y_start,
+ OTG_CRC1_WINDOWB_Y_END, params->windowb_y_end);
+
+ /* Set crc mode and selection, and enable.*/
+ REG_UPDATE_3(OTG_CRC_CNTL,
+ OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
+ OTG_CRC1_SELECT, params->selection,
+ OTG_CRC_EN, 1);
+ break;
+ default:
+ return false;
+ }
return true;
}
@@ -1510,6 +1545,7 @@ bool optc1_configure_crc(struct timing_generator *optc,
* optc1_get_crc - Capture CRC result per component
*
* @optc: timing_generator instance.
+ * @idx: index of crc engine to get CRC from
* @r_cr: 16-bit primary CRC signature for red data.
* @g_y: 16-bit primary CRC signature for green data.
* @b_cb: 16-bit primary CRC signature for blue data.
@@ -1521,7 +1557,7 @@ bool optc1_configure_crc(struct timing_generator *optc,
* If CRC is disabled, return false; otherwise, return true, and the CRC
* results in the parameters.
*/
-bool optc1_get_crc(struct timing_generator *optc,
+bool optc1_get_crc(struct timing_generator *optc, uint8_t idx,
uint32_t *r_cr, uint32_t *g_y, uint32_t *b_cb)
{
uint32_t field = 0;
@@ -1533,14 +1569,30 @@ bool optc1_get_crc(struct timing_generator *optc,
if (!field)
return false;
- /* OTG_CRC0_DATA_RG has the CRC16 results for the red and green component */
- REG_GET_2(OTG_CRC0_DATA_RG,
- CRC0_R_CR, r_cr,
- CRC0_G_Y, g_y);
+ switch (idx) {
+ case 0:
+ /* OTG_CRC0_DATA_RG has the CRC16 results for the red and green component */
+ REG_GET_2(OTG_CRC0_DATA_RG,
+ CRC0_R_CR, r_cr,
+ CRC0_G_Y, g_y);
- /* OTG_CRC0_DATA_B has the CRC16 results for the blue component */
- REG_GET(OTG_CRC0_DATA_B,
- CRC0_B_CB, b_cb);
+ /* OTG_CRC0_DATA_B has the CRC16 results for the blue component */
+ REG_GET(OTG_CRC0_DATA_B,
+ CRC0_B_CB, b_cb);
+ break;
+ case 1:
+ /* OTG_CRC1_DATA_RG has the CRC16 results for the red and green component */
+ REG_GET_2(OTG_CRC1_DATA_RG,
+ CRC1_R_CR, r_cr,
+ CRC1_G_Y, g_y);
+
+ /* OTG_CRC1_DATA_B has the CRC16 results for the blue component */
+ REG_GET(OTG_CRC1_DATA_B,
+ CRC1_B_CB, b_cb);
+ break;
+ default:
+ return false;
+ }
return true;
}
@@ -1613,6 +1665,7 @@ static const struct timing_generator_funcs dcn10_tg_funcs = {
.setup_manual_trigger = optc1_setup_manual_trigger,
.get_hw_timing = optc1_get_hw_timing,
.is_two_pixels_per_container = optc1_is_two_pixels_per_container,
+ .read_otg_state = optc1_read_otg_state,
};
void dcn10_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h
index 40757f20d73f..8b2a8455eb56 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn10/dcn10_optc.h
@@ -86,6 +86,12 @@
SRI(OTG_CRC0_WINDOWA_Y_CONTROL, OTG, inst),\
SRI(OTG_CRC0_WINDOWB_X_CONTROL, OTG, inst),\
SRI(OTG_CRC0_WINDOWB_Y_CONTROL, OTG, inst),\
+ SRI(OTG_CRC1_DATA_RG, OTG, inst),\
+ SRI(OTG_CRC1_DATA_B, OTG, inst),\
+ SRI(OTG_CRC1_WINDOWA_X_CONTROL, OTG, inst),\
+ SRI(OTG_CRC1_WINDOWA_Y_CONTROL, OTG, inst),\
+ SRI(OTG_CRC1_WINDOWB_X_CONTROL, OTG, inst),\
+ SRI(OTG_CRC1_WINDOWB_Y_CONTROL, OTG, inst),\
SR(GSL_SOURCE_SELECT),\
SRI(OTG_GLOBAL_CONTROL2, OTG, inst),\
SRI(OTG_TRIGA_MANUAL_TRIG, OTG, inst)
@@ -98,111 +104,115 @@
SRI(OTG_MANUAL_FLOW_CONTROL, OTG, inst)
+#define OPTC_REG_VARIABLE_LIST_DCN \
+ uint32_t OTG_GLOBAL_CONTROL1; \
+ uint32_t OTG_GLOBAL_CONTROL2; \
+ uint32_t OTG_VERT_SYNC_CONTROL; \
+ uint32_t OTG_MASTER_UPDATE_MODE; \
+ uint32_t OTG_GSL_CONTROL; \
+ uint32_t OTG_VSTARTUP_PARAM; \
+ uint32_t OTG_VUPDATE_PARAM; \
+ uint32_t OTG_VREADY_PARAM; \
+ uint32_t OTG_BLANK_CONTROL; \
+ uint32_t OTG_MASTER_UPDATE_LOCK; \
+ uint32_t OTG_GLOBAL_CONTROL0; \
+ uint32_t OTG_DOUBLE_BUFFER_CONTROL; \
+ uint32_t OTG_H_TOTAL; \
+ uint32_t OTG_H_BLANK_START_END; \
+ uint32_t OTG_H_SYNC_A; \
+ uint32_t OTG_H_SYNC_A_CNTL; \
+ uint32_t OTG_H_TIMING_CNTL; \
+ uint32_t OTG_V_TOTAL; \
+ uint32_t OTG_V_BLANK_START_END; \
+ uint32_t OTG_V_SYNC_A; \
+ uint32_t OTG_V_SYNC_A_CNTL; \
+ uint32_t OTG_INTERLACE_CONTROL; \
+ uint32_t OTG_CONTROL; \
+ uint32_t OTG_STEREO_CONTROL; \
+ uint32_t OTG_3D_STRUCTURE_CONTROL; \
+ uint32_t OTG_STEREO_STATUS; \
+ uint32_t OTG_V_TOTAL_MAX; \
+ uint32_t OTG_V_TOTAL_MID; \
+ uint32_t OTG_V_TOTAL_MIN; \
+ uint32_t OTG_V_TOTAL_CONTROL; \
+ uint32_t OTG_V_COUNT_STOP_CONTROL; \
+ uint32_t OTG_V_COUNT_STOP_CONTROL2; \
+ uint32_t OTG_TRIGA_CNTL; \
+ uint32_t OTG_TRIGA_MANUAL_TRIG; \
+ uint32_t OTG_MANUAL_FLOW_CONTROL; \
+ uint32_t OTG_FORCE_COUNT_NOW_CNTL; \
+ uint32_t OTG_STATIC_SCREEN_CONTROL; \
+ uint32_t OTG_STATUS_FRAME_COUNT; \
+ uint32_t OTG_STATUS; \
+ uint32_t OTG_STATUS_POSITION; \
+ uint32_t OTG_NOM_VERT_POSITION; \
+ uint32_t OTG_BLACK_COLOR; \
+ uint32_t OTG_TEST_PATTERN_PARAMETERS; \
+ uint32_t OTG_TEST_PATTERN_CONTROL; \
+ uint32_t OTG_TEST_PATTERN_COLOR; \
+ uint32_t OTG_CLOCK_CONTROL; \
+ uint32_t OTG_VERTICAL_INTERRUPT0_CONTROL; \
+ uint32_t OTG_VERTICAL_INTERRUPT0_POSITION; \
+ uint32_t OTG_VERTICAL_INTERRUPT1_CONTROL; \
+ uint32_t OTG_VERTICAL_INTERRUPT1_POSITION; \
+ uint32_t OTG_VERTICAL_INTERRUPT2_CONTROL; \
+ uint32_t OTG_VERTICAL_INTERRUPT2_POSITION; \
+ uint32_t OPTC_INPUT_CLOCK_CONTROL; \
+ uint32_t OPTC_DATA_SOURCE_SELECT; \
+ uint32_t OPTC_MEMORY_CONFIG; \
+ uint32_t OPTC_INPUT_GLOBAL_CONTROL; \
+ uint32_t CONTROL; \
+ uint32_t OTG_GSL_WINDOW_X; \
+ uint32_t OTG_GSL_WINDOW_Y; \
+ uint32_t OTG_VUPDATE_KEEPOUT; \
+ uint32_t OTG_CRC_CNTL; \
+ uint32_t OTG_CRC_CNTL2; \
+ uint32_t OTG_CRC0_DATA_RG; \
+ uint32_t OTG_CRC0_DATA_B; \
+ uint32_t OTG_CRC1_DATA_B; \
+ uint32_t OTG_CRC2_DATA_B; \
+ uint32_t OTG_CRC3_DATA_B; \
+ uint32_t OTG_CRC1_DATA_RG; \
+ uint32_t OTG_CRC2_DATA_RG; \
+ uint32_t OTG_CRC3_DATA_RG; \
+ uint32_t OTG_CRC0_WINDOWA_X_CONTROL; \
+ uint32_t OTG_CRC0_WINDOWA_Y_CONTROL; \
+ uint32_t OTG_CRC0_WINDOWB_X_CONTROL; \
+ uint32_t OTG_CRC0_WINDOWB_Y_CONTROL; \
+ uint32_t OTG_CRC1_WINDOWA_X_CONTROL; \
+ uint32_t OTG_CRC1_WINDOWA_Y_CONTROL; \
+ uint32_t OTG_CRC1_WINDOWB_X_CONTROL; \
+ uint32_t OTG_CRC1_WINDOWB_Y_CONTROL; \
+ uint32_t GSL_SOURCE_SELECT; \
+ uint32_t DWB_SOURCE_SELECT; \
+ uint32_t OTG_DSC_START_POSITION; \
+ uint32_t OPTC_DATA_FORMAT_CONTROL; \
+ uint32_t OPTC_BYTES_PER_PIXEL; \
+ uint32_t OPTC_WIDTH_CONTROL; \
+ uint32_t OTG_DRR_CONTROL; \
+ uint32_t OTG_BLANK_DATA_COLOR; \
+ uint32_t OTG_BLANK_DATA_COLOR_EXT; \
+ uint32_t OTG_DRR_TRIGGER_WINDOW; \
+ uint32_t OTG_M_CONST_DTO0; \
+ uint32_t OTG_M_CONST_DTO1; \
+ uint32_t OTG_DRR_V_TOTAL_CHANGE; \
+ uint32_t OTG_GLOBAL_CONTROL4; \
+ uint32_t OTG_CRC0_WINDOWA_X_CONTROL_READBACK; \
+ uint32_t OTG_CRC0_WINDOWA_Y_CONTROL_READBACK; \
+ uint32_t OTG_CRC0_WINDOWB_X_CONTROL_READBACK; \
+ uint32_t OTG_CRC0_WINDOWB_Y_CONTROL_READBACK; \
+ uint32_t OTG_CRC1_WINDOWA_X_CONTROL_READBACK; \
+ uint32_t OTG_CRC1_WINDOWA_Y_CONTROL_READBACK; \
+ uint32_t OTG_CRC1_WINDOWB_X_CONTROL_READBACK; \
+ uint32_t OTG_CRC1_WINDOWB_Y_CONTROL_READBACK; \
+ uint32_t OPTC_CLOCK_CONTROL; \
+ uint32_t OPTC_WIDTH_CONTROL2; \
+ uint32_t OTG_PSTATE_REGISTER; \
+ uint32_t OTG_PIPE_UPDATE_STATUS; \
+ uint32_t INTERRUPT_DEST
+
struct dcn_optc_registers {
- uint32_t OTG_GLOBAL_CONTROL1;
- uint32_t OTG_GLOBAL_CONTROL2;
- uint32_t OTG_VERT_SYNC_CONTROL;
- uint32_t OTG_MASTER_UPDATE_MODE;
- uint32_t OTG_GSL_CONTROL;
- uint32_t OTG_VSTARTUP_PARAM;
- uint32_t OTG_VUPDATE_PARAM;
- uint32_t OTG_VREADY_PARAM;
- uint32_t OTG_BLANK_CONTROL;
- uint32_t OTG_MASTER_UPDATE_LOCK;
- uint32_t OTG_GLOBAL_CONTROL0;
- uint32_t OTG_DOUBLE_BUFFER_CONTROL;
- uint32_t OTG_H_TOTAL;
- uint32_t OTG_H_BLANK_START_END;
- uint32_t OTG_H_SYNC_A;
- uint32_t OTG_H_SYNC_A_CNTL;
- uint32_t OTG_H_TIMING_CNTL;
- uint32_t OTG_V_TOTAL;
- uint32_t OTG_V_BLANK_START_END;
- uint32_t OTG_V_SYNC_A;
- uint32_t OTG_V_SYNC_A_CNTL;
- uint32_t OTG_INTERLACE_CONTROL;
- uint32_t OTG_CONTROL;
- uint32_t OTG_STEREO_CONTROL;
- uint32_t OTG_3D_STRUCTURE_CONTROL;
- uint32_t OTG_STEREO_STATUS;
- uint32_t OTG_V_TOTAL_MAX;
- uint32_t OTG_V_TOTAL_MID;
- uint32_t OTG_V_TOTAL_MIN;
- uint32_t OTG_V_TOTAL_CONTROL;
- uint32_t OTG_V_COUNT_STOP_CONTROL;
- uint32_t OTG_V_COUNT_STOP_CONTROL2;
- uint32_t OTG_TRIGA_CNTL;
- uint32_t OTG_TRIGA_MANUAL_TRIG;
- uint32_t OTG_MANUAL_FLOW_CONTROL;
- uint32_t OTG_FORCE_COUNT_NOW_CNTL;
- uint32_t OTG_STATIC_SCREEN_CONTROL;
- uint32_t OTG_STATUS_FRAME_COUNT;
- uint32_t OTG_STATUS;
- uint32_t OTG_STATUS_POSITION;
- uint32_t OTG_NOM_VERT_POSITION;
- uint32_t OTG_BLACK_COLOR;
- uint32_t OTG_TEST_PATTERN_PARAMETERS;
- uint32_t OTG_TEST_PATTERN_CONTROL;
- uint32_t OTG_TEST_PATTERN_COLOR;
- uint32_t OTG_CLOCK_CONTROL;
- uint32_t OTG_VERTICAL_INTERRUPT0_CONTROL;
- uint32_t OTG_VERTICAL_INTERRUPT0_POSITION;
- uint32_t OTG_VERTICAL_INTERRUPT1_CONTROL;
- uint32_t OTG_VERTICAL_INTERRUPT1_POSITION;
- uint32_t OTG_VERTICAL_INTERRUPT2_CONTROL;
- uint32_t OTG_VERTICAL_INTERRUPT2_POSITION;
- uint32_t OPTC_INPUT_CLOCK_CONTROL;
- uint32_t OPTC_DATA_SOURCE_SELECT;
- uint32_t OPTC_MEMORY_CONFIG;
- uint32_t OPTC_INPUT_GLOBAL_CONTROL;
- uint32_t CONTROL;
- uint32_t OTG_GSL_WINDOW_X;
- uint32_t OTG_GSL_WINDOW_Y;
- uint32_t OTG_VUPDATE_KEEPOUT;
- uint32_t OTG_CRC_CNTL;
- uint32_t OTG_CRC_CNTL2;
- uint32_t OTG_CRC0_DATA_RG;
- uint32_t OTG_CRC0_DATA_B;
- uint32_t OTG_CRC1_DATA_B;
- uint32_t OTG_CRC2_DATA_B;
- uint32_t OTG_CRC3_DATA_B;
- uint32_t OTG_CRC1_DATA_RG;
- uint32_t OTG_CRC2_DATA_RG;
- uint32_t OTG_CRC3_DATA_RG;
- uint32_t OTG_CRC0_WINDOWA_X_CONTROL;
- uint32_t OTG_CRC0_WINDOWA_Y_CONTROL;
- uint32_t OTG_CRC0_WINDOWB_X_CONTROL;
- uint32_t OTG_CRC0_WINDOWB_Y_CONTROL;
- uint32_t OTG_CRC1_WINDOWA_X_CONTROL;
- uint32_t OTG_CRC1_WINDOWA_Y_CONTROL;
- uint32_t OTG_CRC1_WINDOWB_X_CONTROL;
- uint32_t OTG_CRC1_WINDOWB_Y_CONTROL;
- uint32_t GSL_SOURCE_SELECT;
- uint32_t DWB_SOURCE_SELECT;
- uint32_t OTG_DSC_START_POSITION;
- uint32_t OPTC_DATA_FORMAT_CONTROL;
- uint32_t OPTC_BYTES_PER_PIXEL;
- uint32_t OPTC_WIDTH_CONTROL;
- uint32_t OTG_DRR_CONTROL;
- uint32_t OTG_BLANK_DATA_COLOR;
- uint32_t OTG_BLANK_DATA_COLOR_EXT;
- uint32_t OTG_DRR_TRIGGER_WINDOW;
- uint32_t OTG_M_CONST_DTO0;
- uint32_t OTG_M_CONST_DTO1;
- uint32_t OTG_DRR_V_TOTAL_CHANGE;
- uint32_t OTG_GLOBAL_CONTROL4;
- uint32_t OTG_CRC0_WINDOWA_X_CONTROL_READBACK;
- uint32_t OTG_CRC0_WINDOWA_Y_CONTROL_READBACK;
- uint32_t OTG_CRC0_WINDOWB_X_CONTROL_READBACK;
- uint32_t OTG_CRC0_WINDOWB_Y_CONTROL_READBACK;
- uint32_t OTG_CRC1_WINDOWA_X_CONTROL_READBACK;
- uint32_t OTG_CRC1_WINDOWA_Y_CONTROL_READBACK;
- uint32_t OTG_CRC1_WINDOWB_X_CONTROL_READBACK;
- uint32_t OTG_CRC1_WINDOWB_Y_CONTROL_READBACK;
- uint32_t OPTC_CLOCK_CONTROL;
- uint32_t OPTC_WIDTH_CONTROL2;
- uint32_t OTG_PSTATE_REGISTER;
- uint32_t OTG_PIPE_UPDATE_STATUS;
+ OPTC_REG_VARIABLE_LIST_DCN;
};
#define TG_COMMON_MASK_SH_LIST_DCN(mask_sh)\
@@ -315,6 +325,7 @@ struct dcn_optc_registers {
SF(OTG0_OTG_GSL_CONTROL, OTG_GSL_CHECK_ALL_FIELDS, mask_sh),\
SF(OTG0_OTG_CRC_CNTL, OTG_CRC_CONT_EN, mask_sh),\
SF(OTG0_OTG_CRC_CNTL, OTG_CRC0_SELECT, mask_sh),\
+ SF(OTG0_OTG_CRC_CNTL, OTG_CRC1_SELECT, mask_sh),\
SF(OTG0_OTG_CRC_CNTL, OTG_CRC_EN, mask_sh),\
SF(OTG0_OTG_CRC0_DATA_RG, CRC0_R_CR, mask_sh),\
SF(OTG0_OTG_CRC0_DATA_RG, CRC0_G_Y, mask_sh),\
@@ -327,6 +338,17 @@ struct dcn_optc_registers {
SF(OTG0_OTG_CRC0_WINDOWB_X_CONTROL, OTG_CRC0_WINDOWB_X_END, mask_sh),\
SF(OTG0_OTG_CRC0_WINDOWB_Y_CONTROL, OTG_CRC0_WINDOWB_Y_START, mask_sh),\
SF(OTG0_OTG_CRC0_WINDOWB_Y_CONTROL, OTG_CRC0_WINDOWB_Y_END, mask_sh),\
+ SF(OTG0_OTG_CRC1_DATA_RG, CRC1_R_CR, mask_sh),\
+ SF(OTG0_OTG_CRC1_DATA_RG, CRC1_G_Y, mask_sh),\
+ SF(OTG0_OTG_CRC1_DATA_B, CRC1_B_CB, mask_sh),\
+ SF(OTG0_OTG_CRC1_WINDOWA_X_CONTROL, OTG_CRC1_WINDOWA_X_START, mask_sh),\
+ SF(OTG0_OTG_CRC1_WINDOWA_X_CONTROL, OTG_CRC1_WINDOWA_X_END, mask_sh),\
+ SF(OTG0_OTG_CRC1_WINDOWA_Y_CONTROL, OTG_CRC1_WINDOWA_Y_START, mask_sh),\
+ SF(OTG0_OTG_CRC1_WINDOWA_Y_CONTROL, OTG_CRC1_WINDOWA_Y_END, mask_sh),\
+ SF(OTG0_OTG_CRC1_WINDOWB_X_CONTROL, OTG_CRC1_WINDOWB_X_START, mask_sh),\
+ SF(OTG0_OTG_CRC1_WINDOWB_X_CONTROL, OTG_CRC1_WINDOWB_X_END, mask_sh),\
+ SF(OTG0_OTG_CRC1_WINDOWB_Y_CONTROL, OTG_CRC1_WINDOWB_Y_START, mask_sh),\
+ SF(OTG0_OTG_CRC1_WINDOWB_Y_CONTROL, OTG_CRC1_WINDOWB_Y_END, mask_sh),\
SF(GSL_SOURCE_SELECT, GSL0_READY_SOURCE_SEL, mask_sh),\
SF(GSL_SOURCE_SELECT, GSL1_READY_SOURCE_SEL, mask_sh),\
SF(GSL_SOURCE_SELECT, GSL2_READY_SOURCE_SEL, mask_sh),\
@@ -482,6 +504,7 @@ struct dcn_optc_registers {
type OTG_MASTER_UPDATE_LOCK_VUPDATE_KEEPOUT_EN;\
type OTG_CRC_CONT_EN;\
type OTG_CRC0_SELECT;\
+ type OTG_CRC1_SELECT;\
type OTG_CRC_EN;\
type CRC0_R_CR;\
type CRC0_G_Y;\
@@ -572,6 +595,7 @@ struct dcn_optc_registers {
type OTG_DC_REG_UPDATE_PENDING;\
type OTG_CURSOR_UPDATE_PENDING;\
type OTG_VUPDATE_KEEPOUT_STATUS;\
+ type OTG0_IHC_OTG_VERTICAL_INTERRUPT2_DEST;
#define TG_REG_FIELD_LIST_DCN3_2(type) \
type OTG_H_TIMING_DIV_MODE_MANUAL;
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.c
index b4694985a40a..e7a90a437fff 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn20/dcn20_optc.c
@@ -502,7 +502,7 @@ void optc2_get_last_used_drr_vtotal(struct timing_generator *optc, uint32_t *ref
REG_GET(OTG_DRR_CONTROL, OTG_V_TOTAL_LAST_USED_BY_DRR, refresh_rate);
}
-static struct timing_generator_funcs dcn20_tg_funcs = {
+static const struct timing_generator_funcs dcn20_tg_funcs = {
.validate_timing = optc1_validate_timing,
.program_timing = optc1_program_timing,
.setup_vertical_interrupt0 = optc1_setup_vertical_interrupt0,
@@ -562,6 +562,7 @@ static struct timing_generator_funcs dcn20_tg_funcs = {
.get_hw_timing = optc1_get_hw_timing,
.align_vblanks = optc2_align_vblanks,
.is_two_pixels_per_container = optc1_is_two_pixels_per_container,
+ .read_otg_state = optc1_read_otg_state,
};
void dcn20_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn201/dcn201_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn201/dcn201_optc.c
index 49c2efdfa403..772a8bfb949c 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn201/dcn201_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn201/dcn201_optc.c
@@ -129,7 +129,7 @@ static void optc201_get_optc_source(struct timing_generator *optc,
*num_of_src_opp = 1;
}
-static struct timing_generator_funcs dcn201_tg_funcs = {
+static const struct timing_generator_funcs dcn201_tg_funcs = {
.validate_timing = optc201_validate_timing,
.program_timing = optc1_program_timing,
.setup_vertical_interrupt0 = optc1_setup_vertical_interrupt0,
@@ -180,6 +180,7 @@ static struct timing_generator_funcs dcn201_tg_funcs = {
.setup_manual_trigger = optc2_setup_manual_trigger,
.get_hw_timing = optc1_get_hw_timing,
.is_two_pixels_per_container = optc1_is_two_pixels_per_container,
+ .read_otg_state = optc1_read_otg_state,
};
void dcn201_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.c
index 4c95c0958612..ee4665aa49e9 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn30/dcn30_optc.c
@@ -357,7 +357,7 @@ void optc3_tg_init(struct timing_generator *optc)
optc1_clear_optc_underflow(optc);
}
-static struct timing_generator_funcs dcn30_tg_funcs = {
+static const struct timing_generator_funcs dcn30_tg_funcs = {
.validate_timing = optc1_validate_timing,
.program_timing = optc1_program_timing,
.setup_vertical_interrupt0 = optc1_setup_vertical_interrupt0,
@@ -420,6 +420,7 @@ static struct timing_generator_funcs dcn30_tg_funcs = {
.get_optc_double_buffer_pending = optc3_get_optc_double_buffer_pending,
.get_otg_double_buffer_pending = optc3_get_otg_update_pending,
.get_pipe_update_pending = optc3_get_pipe_update_pending,
+ .read_otg_state = optc1_read_otg_state,
};
void dcn30_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn301/dcn301_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn301/dcn301_optc.c
index d7a45ef2d01b..38f85bc2681a 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn301/dcn301_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn301/dcn301_optc.c
@@ -109,7 +109,7 @@ void optc301_setup_manual_trigger(struct timing_generator *optc)
OTG_TRIGA_CLEAR, 1);
}
-static struct timing_generator_funcs dcn30_tg_funcs = {
+static const struct timing_generator_funcs dcn30_tg_funcs = {
.validate_timing = optc1_validate_timing,
.program_timing = optc1_program_timing,
.setup_vertical_interrupt0 = optc1_setup_vertical_interrupt0,
@@ -172,6 +172,7 @@ static struct timing_generator_funcs dcn30_tg_funcs = {
.get_optc_double_buffer_pending = optc3_get_optc_double_buffer_pending,
.get_otg_double_buffer_pending = optc3_get_otg_update_pending,
.get_pipe_update_pending = optc3_get_pipe_update_pending,
+ .read_otg_state = optc1_read_otg_state,
};
void dcn301_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.c
index 4b6446ed4ce4..4f1830ba619f 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.c
@@ -245,7 +245,77 @@ void optc3_init_odm(struct timing_generator *optc)
optc1->opp_count = 1;
}
-static struct timing_generator_funcs dcn31_tg_funcs = {
+void optc31_read_otg_state(struct timing_generator *optc,
+ struct dcn_otg_state *s)
+{
+ struct optc *optc1 = DCN10TG_FROM_TG(optc);
+
+ REG_GET(OTG_CONTROL,
+ OTG_MASTER_EN, &s->otg_enabled);
+
+ REG_GET_2(OTG_V_BLANK_START_END,
+ OTG_V_BLANK_START, &s->v_blank_start,
+ OTG_V_BLANK_END, &s->v_blank_end);
+
+ REG_GET(OTG_V_SYNC_A_CNTL,
+ OTG_V_SYNC_A_POL, &s->v_sync_a_pol);
+
+ REG_GET(OTG_V_TOTAL,
+ OTG_V_TOTAL, &s->v_total);
+
+ REG_GET(OTG_V_TOTAL_MAX,
+ OTG_V_TOTAL_MAX, &s->v_total_max);
+
+ REG_GET(OTG_V_TOTAL_MIN,
+ OTG_V_TOTAL_MIN, &s->v_total_min);
+
+ REG_GET(OTG_V_TOTAL_CONTROL,
+ OTG_V_TOTAL_MAX_SEL, &s->v_total_max_sel);
+
+ REG_GET(OTG_V_TOTAL_CONTROL,
+ OTG_V_TOTAL_MIN_SEL, &s->v_total_min_sel);
+
+ REG_GET_2(OTG_V_SYNC_A,
+ OTG_V_SYNC_A_START, &s->v_sync_a_start,
+ OTG_V_SYNC_A_END, &s->v_sync_a_end);
+
+ REG_GET_2(OTG_H_BLANK_START_END,
+ OTG_H_BLANK_START, &s->h_blank_start,
+ OTG_H_BLANK_END, &s->h_blank_end);
+
+ REG_GET_2(OTG_H_SYNC_A,
+ OTG_H_SYNC_A_START, &s->h_sync_a_start,
+ OTG_H_SYNC_A_END, &s->h_sync_a_end);
+
+ REG_GET(OTG_H_SYNC_A_CNTL,
+ OTG_H_SYNC_A_POL, &s->h_sync_a_pol);
+
+ REG_GET(OTG_H_TOTAL,
+ OTG_H_TOTAL, &s->h_total);
+
+ REG_GET(OPTC_INPUT_GLOBAL_CONTROL,
+ OPTC_UNDERFLOW_OCCURRED_STATUS, &s->underflow_occurred_status);
+
+ REG_GET(OTG_VERTICAL_INTERRUPT1_CONTROL,
+ OTG_VERTICAL_INTERRUPT1_INT_ENABLE, &s->vertical_interrupt1_en);
+
+ REG_GET(OTG_VERTICAL_INTERRUPT1_POSITION,
+ OTG_VERTICAL_INTERRUPT1_LINE_START, &s->vertical_interrupt1_line);
+
+ REG_GET(OTG_VERTICAL_INTERRUPT2_CONTROL,
+ OTG_VERTICAL_INTERRUPT2_INT_ENABLE, &s->vertical_interrupt2_en);
+
+ REG_GET(OTG_VERTICAL_INTERRUPT2_POSITION,
+ OTG_VERTICAL_INTERRUPT2_LINE_START, &s->vertical_interrupt2_line);
+
+ REG_GET(INTERRUPT_DEST,
+ OTG0_IHC_OTG_VERTICAL_INTERRUPT2_DEST, &s->vertical_interrupt2_dest);
+
+ s->otg_master_update_lock = REG_READ(OTG_MASTER_UPDATE_LOCK);
+ s->otg_double_buffer_control = REG_READ(OTG_DOUBLE_BUFFER_CONTROL);
+}
+
+static const struct timing_generator_funcs dcn31_tg_funcs = {
.validate_timing = optc1_validate_timing,
.program_timing = optc1_program_timing,
.setup_vertical_interrupt0 = optc1_setup_vertical_interrupt0,
@@ -306,6 +376,7 @@ static struct timing_generator_funcs dcn31_tg_funcs = {
.get_hw_timing = optc1_get_hw_timing,
.init_odm = optc3_init_odm,
.is_two_pixels_per_container = optc1_is_two_pixels_per_container,
+ .read_otg_state = optc31_read_otg_state,
};
void dcn31_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.h
index fbbe86d00c2e..0f72c274f40b 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn31/dcn31_optc.h
@@ -100,7 +100,8 @@
SRI(OTG_CRC_CNTL2, OTG, inst),\
SR(DWB_SOURCE_SELECT),\
SRI(OTG_DRR_CONTROL, OTG, inst),\
- SRI(OTG_PIPE_UPDATE_STATUS, OTG, inst)
+ SRI(OTG_PIPE_UPDATE_STATUS, OTG, inst),\
+ SRI(INTERRUPT_DEST, OTG, inst)
#define OPTC_COMMON_MASK_SH_LIST_DCN3_1(mask_sh)\
SF(OTG0_OTG_VSTARTUP_PARAM, VSTARTUP_START, mask_sh),\
@@ -260,6 +261,7 @@
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh),\
+ SF(OTG0_INTERRUPT_DEST, OTG0_IHC_OTG_VERTICAL_INTERRUPT2_DEST, mask_sh)
void dcn31_timing_generator_init(struct optc *optc1);
@@ -269,4 +271,7 @@ void optc31_set_drr(struct timing_generator *optc, const struct drr_params *para
void optc3_init_odm(struct timing_generator *optc);
+void optc31_read_otg_state(struct timing_generator *optc,
+ struct dcn_otg_state *s);
+
#endif /* __DC_OPTC_DCN31_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.c
index 633d62addd4d..4a2caca37255 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.c
@@ -192,7 +192,7 @@ static void optc314_set_h_timing_div_manual_mode(struct timing_generator *optc,
}
-static struct timing_generator_funcs dcn314_tg_funcs = {
+static const struct timing_generator_funcs dcn314_tg_funcs = {
.validate_timing = optc1_validate_timing,
.program_timing = optc1_program_timing,
.setup_vertical_interrupt0 = optc1_setup_vertical_interrupt0,
@@ -255,6 +255,7 @@ static struct timing_generator_funcs dcn314_tg_funcs = {
.set_odm_combine = optc314_set_odm_combine,
.set_h_timing_div_manual_mode = optc314_set_h_timing_div_manual_mode,
.is_two_pixels_per_container = optc1_is_two_pixels_per_container,
+ .read_otg_state = optc31_read_otg_state,
};
void dcn314_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.h
index 0ff72b97b465..6bfdee3fcf5f 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn314/dcn314_optc.h
@@ -99,7 +99,8 @@
SRI(OPTC_WIDTH_CONTROL, ODM, inst),\
SRI(OPTC_MEMORY_CONFIG, ODM, inst),\
SRI(OTG_DRR_CONTROL, OTG, inst),\
- SRI(OTG_PIPE_UPDATE_STATUS, OTG, inst)
+ SRI(OTG_PIPE_UPDATE_STATUS, OTG, inst),\
+ SRI(INTERRUPT_DEST, OTG, inst)
#define OPTC_COMMON_MASK_SH_LIST_DCN3_14(mask_sh)\
SF(OTG0_OTG_VSTARTUP_PARAM, VSTARTUP_START, mask_sh),\
@@ -254,6 +255,7 @@
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh),\
+ SF(OTG0_INTERRUPT_DEST, OTG0_IHC_OTG_VERTICAL_INTERRUPT2_DEST, mask_sh)
void dcn314_timing_generator_init(struct optc *optc1);
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c
index c217f653b3c8..b2b226bcd871 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.c
@@ -297,7 +297,7 @@ static void optc32_set_drr(
optc32_setup_manual_trigger(optc);
}
-static struct timing_generator_funcs dcn32_tg_funcs = {
+static const struct timing_generator_funcs dcn32_tg_funcs = {
.validate_timing = optc1_validate_timing,
.program_timing = optc1_program_timing,
.setup_vertical_interrupt0 = optc1_setup_vertical_interrupt0,
@@ -364,6 +364,7 @@ static struct timing_generator_funcs dcn32_tg_funcs = {
.get_optc_double_buffer_pending = optc3_get_optc_double_buffer_pending,
.get_otg_double_buffer_pending = optc3_get_otg_update_pending,
.get_pipe_update_pending = optc3_get_pipe_update_pending,
+ .read_otg_state = optc31_read_otg_state,
};
void dcn32_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h
index 0b0964a9da74..d159e3ed3bb3 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn32/dcn32_optc.h
@@ -181,7 +181,8 @@
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\
- SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh)
+ SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh),\
+ SF(OTG0_INTERRUPT_DEST, OTG0_IHC_OTG_VERTICAL_INTERRUPT2_DEST, mask_sh)
void dcn32_timing_generator_init(struct optc *optc1);
void optc32_set_h_timing_div_manual_mode(struct timing_generator *optc, bool manual_mode);
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.c
index dfa9364fe5a6..72bff94cb57d 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.c
@@ -183,34 +183,87 @@ static bool optc35_configure_crc(struct timing_generator *optc,
{
struct optc *optc1 = DCN10TG_FROM_TG(optc);
+ /* Cannot configure crc on a CRTC that is disabled */
if (!optc1_is_tg_enabled(optc))
return false;
- REG_WRITE(OTG_CRC_CNTL, 0);
+
+ if (!params->enable || params->reset)
+ REG_WRITE(OTG_CRC_CNTL, 0);
+
if (!params->enable)
return true;
- REG_UPDATE_2(OTG_CRC0_WINDOWA_X_CONTROL,
- OTG_CRC0_WINDOWA_X_START, params->windowa_x_start,
- OTG_CRC0_WINDOWA_X_END, params->windowa_x_end);
- REG_UPDATE_2(OTG_CRC0_WINDOWA_Y_CONTROL,
- OTG_CRC0_WINDOWA_Y_START, params->windowa_y_start,
- OTG_CRC0_WINDOWA_Y_END, params->windowa_y_end);
- REG_UPDATE_2(OTG_CRC0_WINDOWB_X_CONTROL,
- OTG_CRC0_WINDOWB_X_START, params->windowb_x_start,
- OTG_CRC0_WINDOWB_X_END, params->windowb_x_end);
- REG_UPDATE_2(OTG_CRC0_WINDOWB_Y_CONTROL,
- OTG_CRC0_WINDOWB_Y_START, params->windowb_y_start,
- OTG_CRC0_WINDOWB_Y_END, params->windowb_y_end);
- if (optc1->base.ctx->dc->debug.otg_crc_db && optc1->tg_mask->OTG_CRC_WINDOW_DB_EN != 0) {
- REG_UPDATE_4(OTG_CRC_CNTL,
- OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
- OTG_CRC0_SELECT, params->selection,
- OTG_CRC_EN, 1,
- OTG_CRC_WINDOW_DB_EN, 1);
- } else
- REG_UPDATE_3(OTG_CRC_CNTL,
- OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
- OTG_CRC0_SELECT, params->selection,
- OTG_CRC_EN, 1);
+
+ /* Program frame boundaries */
+ switch (params->crc_eng_inst) {
+ case 0:
+ /* Window A x axis start and end. */
+ REG_UPDATE_2(OTG_CRC0_WINDOWA_X_CONTROL,
+ OTG_CRC0_WINDOWA_X_START, params->windowa_x_start,
+ OTG_CRC0_WINDOWA_X_END, params->windowa_x_end);
+
+ /* Window A y axis start and end. */
+ REG_UPDATE_2(OTG_CRC0_WINDOWA_Y_CONTROL,
+ OTG_CRC0_WINDOWA_Y_START, params->windowa_y_start,
+ OTG_CRC0_WINDOWA_Y_END, params->windowa_y_end);
+
+ /* Window B x axis start and end. */
+ REG_UPDATE_2(OTG_CRC0_WINDOWB_X_CONTROL,
+ OTG_CRC0_WINDOWB_X_START, params->windowb_x_start,
+ OTG_CRC0_WINDOWB_X_END, params->windowb_x_end);
+
+ /* Window B y axis start and end. */
+ REG_UPDATE_2(OTG_CRC0_WINDOWB_Y_CONTROL,
+ OTG_CRC0_WINDOWB_Y_START, params->windowb_y_start,
+ OTG_CRC0_WINDOWB_Y_END, params->windowb_y_end);
+
+ if (optc1->base.ctx->dc->debug.otg_crc_db && optc1->tg_mask->OTG_CRC_WINDOW_DB_EN != 0)
+ REG_UPDATE_4(OTG_CRC_CNTL,
+ OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
+ OTG_CRC0_SELECT, params->selection,
+ OTG_CRC_EN, 1,
+ OTG_CRC_WINDOW_DB_EN, 1);
+ else
+ REG_UPDATE_3(OTG_CRC_CNTL,
+ OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
+ OTG_CRC0_SELECT, params->selection,
+ OTG_CRC_EN, 1);
+ break;
+ case 1:
+ /* Window A x axis start and end. */
+ REG_UPDATE_2(OTG_CRC1_WINDOWA_X_CONTROL,
+ OTG_CRC1_WINDOWA_X_START, params->windowa_x_start,
+ OTG_CRC1_WINDOWA_X_END, params->windowa_x_end);
+
+ /* Window A y axis start and end. */
+ REG_UPDATE_2(OTG_CRC1_WINDOWA_Y_CONTROL,
+ OTG_CRC1_WINDOWA_Y_START, params->windowa_y_start,
+ OTG_CRC1_WINDOWA_Y_END, params->windowa_y_end);
+
+ /* Window B x axis start and end. */
+ REG_UPDATE_2(OTG_CRC1_WINDOWB_X_CONTROL,
+ OTG_CRC1_WINDOWB_X_START, params->windowb_x_start,
+ OTG_CRC1_WINDOWB_X_END, params->windowb_x_end);
+
+ /* Window B y axis start and end. */
+ REG_UPDATE_2(OTG_CRC1_WINDOWB_Y_CONTROL,
+ OTG_CRC1_WINDOWB_Y_START, params->windowb_y_start,
+ OTG_CRC1_WINDOWB_Y_END, params->windowb_y_end);
+
+ if (optc1->base.ctx->dc->debug.otg_crc_db && optc1->tg_mask->OTG_CRC_WINDOW_DB_EN != 0)
+ REG_UPDATE_4(OTG_CRC_CNTL,
+ OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
+ OTG_CRC1_SELECT, params->selection,
+ OTG_CRC_EN, 1,
+ OTG_CRC_WINDOW_DB_EN, 1);
+ else
+ REG_UPDATE_3(OTG_CRC_CNTL,
+ OTG_CRC_CONT_EN, params->continuous_mode ? 1 : 0,
+ OTG_CRC1_SELECT, params->selection,
+ OTG_CRC_EN, 1);
+ break;
+ default:
+ return false;
+ }
return true;
}
@@ -375,7 +428,7 @@ static void optc35_set_long_vtotal(
}
}
-static struct timing_generator_funcs dcn35_tg_funcs = {
+static const struct timing_generator_funcs dcn35_tg_funcs = {
.validate_timing = optc1_validate_timing,
.program_timing = optc1_program_timing,
.setup_vertical_interrupt0 = optc1_setup_vertical_interrupt0,
@@ -439,6 +492,7 @@ static struct timing_generator_funcs dcn35_tg_funcs = {
.init_odm = optc3_init_odm,
.set_long_vtotal = optc35_set_long_vtotal,
.is_two_pixels_per_container = optc1_is_two_pixels_per_container,
+ .read_otg_state = optc31_read_otg_state,
};
void dcn35_timing_generator_init(struct optc *optc1)
@@ -453,6 +507,7 @@ void dcn35_timing_generator_init(struct optc *optc1)
optc1->min_v_blank_interlace = 5;
optc1->min_h_sync_width = 4;
optc1->min_v_sync_width = 1;
+ optc1->max_frame_count = 0xFFFFFF;
dcn35_timing_generator_set_fgcg(
optc1, CTX->dc->debug.enable_fine_grain_clock_gating.bits.optc);
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h
index be749ab41dce..733a2f149d9a 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn35/dcn35_optc.h
@@ -71,7 +71,8 @@
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\
- SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh)
+ SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh),\
+ SF(OTG0_INTERRUPT_DEST, OTG0_IHC_OTG_VERTICAL_INTERRUPT2_DEST, mask_sh)
void dcn35_timing_generator_init(struct optc *optc1);
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.c b/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.c
index 783ca9acc762..ff79c38287df 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.c
@@ -101,7 +101,7 @@ static uint32_t decide_odm_mem_bit_map(int *opp_id, int opp_cnt, int h_active)
return memory_bit_map;
}
-static void optc401_set_odm_combine(struct timing_generator *optc, int *opp_id,
+void optc401_set_odm_combine(struct timing_generator *optc, int *opp_id,
int opp_cnt, int segment_width, int last_segment_width)
{
struct optc *optc1 = DCN10TG_FROM_TG(optc);
@@ -162,7 +162,7 @@ static void optc401_set_odm_combine(struct timing_generator *optc, int *opp_id,
optc1->opp_count = opp_cnt;
}
-static void optc401_set_h_timing_div_manual_mode(struct timing_generator *optc, bool manual_mode)
+void optc401_set_h_timing_div_manual_mode(struct timing_generator *optc, bool manual_mode)
{
struct optc *optc1 = DCN10TG_FROM_TG(optc);
@@ -177,7 +177,7 @@ static void optc401_set_h_timing_div_manual_mode(struct timing_generator *optc,
*
* Return: Always returns true
*/
-static bool optc401_enable_crtc(struct timing_generator *optc)
+bool optc401_enable_crtc(struct timing_generator *optc)
{
struct optc *optc1 = DCN10TG_FROM_TG(optc);
@@ -203,7 +203,7 @@ static bool optc401_enable_crtc(struct timing_generator *optc)
}
/* disable_crtc */
-static bool optc401_disable_crtc(struct timing_generator *optc)
+bool optc401_disable_crtc(struct timing_generator *optc)
{
struct optc *optc1 = DCN10TG_FROM_TG(optc);
@@ -234,7 +234,7 @@ static bool optc401_disable_crtc(struct timing_generator *optc)
return true;
}
-static void optc401_phantom_crtc_post_enable(struct timing_generator *optc)
+void optc401_phantom_crtc_post_enable(struct timing_generator *optc)
{
struct optc *optc1 = DCN10TG_FROM_TG(optc);
@@ -245,7 +245,7 @@ static void optc401_phantom_crtc_post_enable(struct timing_generator *optc)
REG_WAIT(OTG_CLOCK_CONTROL, OTG_BUSY, 0, 1, 100000);
}
-static void optc401_disable_phantom_otg(struct timing_generator *optc)
+void optc401_disable_phantom_otg(struct timing_generator *optc)
{
struct optc *optc1 = DCN10TG_FROM_TG(optc);
@@ -259,7 +259,7 @@ static void optc401_disable_phantom_otg(struct timing_generator *optc)
REG_UPDATE(OTG_CONTROL, OTG_MASTER_EN, 0);
}
-static void optc401_set_odm_bypass(struct timing_generator *optc,
+void optc401_set_odm_bypass(struct timing_generator *optc,
const struct dc_crtc_timing *dc_crtc_timing)
{
struct optc *optc1 = DCN10TG_FROM_TG(optc);
@@ -315,7 +315,7 @@ void optc401_set_drr(
struct drr_params amended_params = { 0 };
bool program_manual_trigger = false;
- if (dc->caps.dmub_caps.fams_ver >= 2 && dc->debug.fams2_config.bits.enable) {
+ if (dc->caps.dmub_caps.fams_ver == dc->debug.fams_version.ver && dc->debug.fams2_config.bits.enable) {
if (params != NULL &&
params->vertical_total_max > 0 &&
params->vertical_total_min > 0) {
@@ -365,7 +365,7 @@ void optc401_set_drr(
}
}
-static void optc401_set_out_mux(struct timing_generator *optc, enum otg_out_mux_dest dest)
+void optc401_set_out_mux(struct timing_generator *optc, enum otg_out_mux_dest dest)
{
struct optc *optc1 = DCN10TG_FROM_TG(optc);
@@ -380,7 +380,7 @@ void optc401_set_vtotal_min_max(struct timing_generator *optc, int vtotal_min, i
{
struct dc *dc = optc->ctx->dc;
- if (dc->caps.dmub_caps.fams_ver >= 2 && dc->debug.fams2_config.bits.enable) {
+ if (dc->caps.dmub_caps.fams_ver == dc->debug.fams_version.ver && dc->debug.fams2_config.bits.enable) {
/* FAMS2 */
dc_dmub_srv_fams2_drr_update(dc, optc->inst,
vtotal_min,
@@ -396,7 +396,7 @@ void optc401_set_vtotal_min_max(struct timing_generator *optc, int vtotal_min, i
}
}
-static void optc401_program_global_sync(
+void optc401_program_global_sync(
struct timing_generator *optc,
int vready_offset,
int vstartup_start,
@@ -430,7 +430,7 @@ static void optc401_program_global_sync(
REG_UPDATE(OTG_PSTATE_REGISTER, OTG_PSTATE_KEEPOUT_START, pstate_keepout);
}
-static void optc401_set_vupdate_keepout(struct timing_generator *tg, bool enable)
+void optc401_set_vupdate_keepout(struct timing_generator *tg, bool enable)
{
struct optc *optc1 = DCN10TG_FROM_TG(tg);
@@ -442,7 +442,7 @@ static void optc401_set_vupdate_keepout(struct timing_generator *tg, bool enable
return;
}
-static bool optc401_wait_update_lock_status(struct timing_generator *tg, bool locked)
+bool optc401_wait_update_lock_status(struct timing_generator *tg, bool locked)
{
struct optc *optc1 = DCN10TG_FROM_TG(tg);
uint32_t lock_status = 0;
@@ -459,7 +459,7 @@ static bool optc401_wait_update_lock_status(struct timing_generator *tg, bool lo
return true;
}
-static struct timing_generator_funcs dcn401_tg_funcs = {
+static const struct timing_generator_funcs dcn401_tg_funcs = {
.validate_timing = optc1_validate_timing,
.program_timing = optc1_program_timing,
.setup_vertical_interrupt0 = optc1_setup_vertical_interrupt0,
@@ -527,6 +527,7 @@ static struct timing_generator_funcs dcn401_tg_funcs = {
.get_pipe_update_pending = optc3_get_pipe_update_pending,
.set_vupdate_keepout = optc401_set_vupdate_keepout,
.wait_update_lock_status = optc401_wait_update_lock_status,
+ .read_otg_state = optc31_read_otg_state,
};
void dcn401_timing_generator_init(struct optc *optc1)
diff --git a/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.h b/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.h
index 1be89571986f..fa62737b5b1b 100644
--- a/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/optc/dcn401/dcn401_optc.h
@@ -163,7 +163,8 @@
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_FLIP_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_DC_REG_UPDATE_PENDING, mask_sh),\
SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_CURSOR_UPDATE_PENDING, mask_sh),\
- SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh)
+ SF(OTG0_OTG_PIPE_UPDATE_STATUS, OTG_VUPDATE_KEEPOUT_STATUS, mask_sh),\
+ SF(OTG0_INTERRUPT_DEST, OTG0_IHC_OTG_VERTICAL_INTERRUPT2_DEST, mask_sh)
void dcn401_timing_generator_init(struct optc *optc1);
@@ -172,5 +173,24 @@ void optc401_set_drr(
const struct drr_params *params);
void optc401_set_vtotal_min_max(struct timing_generator *optc, int vtotal_min, int vtotal_max);
void optc401_setup_manual_trigger(struct timing_generator *optc);
+void optc401_program_global_sync(
+ struct timing_generator *optc,
+ int vready_offset,
+ int vstartup_start,
+ int vupdate_offset,
+ int vupdate_width,
+ int pstate_keepout);
+bool optc401_enable_crtc(struct timing_generator *optc);
+bool optc401_disable_crtc(struct timing_generator *optc);
+void optc401_phantom_crtc_post_enable(struct timing_generator *optc);
+void optc401_disable_phantom_otg(struct timing_generator *optc);
+void optc401_set_odm_bypass(struct timing_generator *optc,
+ const struct dc_crtc_timing *dc_crtc_timing);
+void optc401_set_odm_combine(struct timing_generator *optc, int *opp_id,
+ int opp_cnt, int segment_width, int last_segment_width);
+void optc401_set_h_timing_div_manual_mode(struct timing_generator *optc, bool manual_mode);
+void optc401_set_out_mux(struct timing_generator *optc, enum otg_out_mux_dest dest);
+bool optc401_wait_update_lock_status(struct timing_generator *tg, bool locked);
+void optc401_set_vupdate_keepout(struct timing_generator *tg, bool enable);
#endif /* __DC_OPTC_DCN401_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/os_types.h b/drivers/gpu/drm/amd/display/dc/os_types.h
index f2ba76c1e0c0..782316348941 100644
--- a/drivers/gpu/drm/amd/display/dc/os_types.h
+++ b/drivers/gpu/drm/amd/display/dc/os_types.h
@@ -31,6 +31,7 @@
#include <linux/kgdb.h>
#include <linux/delay.h>
#include <linux/mm.h>
+#include <linux/vmalloc.h>
#include <asm/byteorder.h>
diff --git a/drivers/gpu/drm/amd/display/dc/resource/Makefile b/drivers/gpu/drm/amd/display/dc/resource/Makefile
index 09320344d8e9..5b42da8b79c2 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/resource/Makefile
@@ -27,6 +27,24 @@
# DCE
###############################################################################
+ifdef CONFIG_DRM_AMD_DC_SI
+RESOURCE_DCE60 = dce60_resource.o
+
+AMD_DAL_RESOURCE_DCE60 = $(addprefix $(AMDDALPATH)/dc/resource/dce60/,$(RESOURCE_DCE60))
+
+AMD_DISPLAY_FILES += $(AMD_DAL_RESOURCE_DCE60)
+endif
+
+###############################################################################
+
+RESOURCE_DCE80 = dce80_resource.o
+
+AMD_DAL_RESOURCE_DCE80 = $(addprefix $(AMDDALPATH)/dc/resource/dce80/,$(RESOURCE_DCE80))
+
+AMD_DISPLAY_FILES += $(AMD_DAL_RESOURCE_DCE80)
+
+###############################################################################
+
RESOURCE_DCE100 = dce100_resource.o
AMD_DAL_RESOURCE_DCE100 = $(addprefix $(AMDDALPATH)/dc/resource/dce100/,$(RESOURCE_DCE100))
@@ -57,14 +75,6 @@ AMD_DAL_RESOURCE_DCE120 = $(addprefix $(AMDDALPATH)/dc/resource/dce120/,$(RESOUR
AMD_DISPLAY_FILES += $(AMD_DAL_RESOURCE_DCE120)
-###############################################################################
-
-RESOURCE_DCE80 = dce80_resource.o
-
-AMD_DAL_RESOURCE_DCE80 = $(addprefix $(AMDDALPATH)/dc/resource/dce80/,$(RESOURCE_DCE80))
-
-AMD_DISPLAY_FILES += $(AMD_DAL_RESOURCE_DCE80)
-
ifdef CONFIG_DRM_AMD_DC_FP
###############################################################################
# DCN
@@ -198,6 +208,14 @@ AMD_DISPLAY_FILES += $(AMD_DAL_RESOURCE_DCN351)
###############################################################################
+RESOURCE_DCN36 = dcn36_resource.o
+
+AMD_DAL_RESOURCE_DCN36 = $(addprefix $(AMDDALPATH)/dc/resource/dcn36/,$(RESOURCE_DCN36))
+
+AMD_DISPLAY_FILES += $(AMD_DAL_RESOURCE_DCN36)
+
+###############################################################################
+
RESOURCE_DCN401 = dcn401_resource.o
AMD_DAL_RESOURCE_DCN401 = $(addprefix $(AMDDALPATH)/dc/resource/dcn401/,$(RESOURCE_DCN401))
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce100/dce100_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce100/dce100_resource.c
index e698543ec937..3a51be63f020 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dce100/dce100_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dce100/dce100_resource.c
@@ -836,10 +836,10 @@ static enum dc_status build_mapped_resource(
return DC_OK;
}
-static bool dce100_validate_bandwidth(
+static enum dc_status dce100_validate_bandwidth(
struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i;
bool at_least_one_pipe = false;
@@ -858,7 +858,7 @@ static bool dce100_validate_bandwidth(
context->bw_ctx.bw.dce.yclk_khz = 0;
}
- return true;
+ return DC_OK;
}
static bool dce100_validate_surface_sets(
@@ -1069,7 +1069,7 @@ static bool dce100_resource_construct(
pool->base.timing_generator_count = pool->base.res_cap->num_timing_generator;
dc->caps.max_downscale_ratio = 200;
dc->caps.i2c_speed_in_khz = 40;
- dc->caps.i2c_speed_in_khz = 40;
+ dc->caps.i2c_speed_in_khz_hdcp = 40;
dc->caps.max_cursor_size = 128;
dc->caps.min_horizontal_blanking_period = 80;
dc->caps.dual_link_dvi = true;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce110/dce110_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce110/dce110_resource.c
index 035c6cfdaee5..cccde5a6f3cd 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dce110/dce110_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dce110/dce110_resource.c
@@ -960,10 +960,10 @@ static enum dc_status build_mapped_resource(
return DC_OK;
}
-static bool dce110_validate_bandwidth(
+static enum dc_status dce110_validate_bandwidth(
struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool result = false;
@@ -1031,7 +1031,7 @@ static bool dce110_validate_bandwidth(
context->bw_ctx.bw.dce.yclk_khz,
context->bw_ctx.bw.dce.blackout_recovery_time_us);
}
- return result;
+ return result ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
}
static enum dc_status dce110_validate_plane(const struct dc_plane_state *plane_state,
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c
index 480a50967385..164ba796f64c 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.c
@@ -883,10 +883,10 @@ static enum dc_status build_mapped_resource(
return DC_OK;
}
-bool dce112_validate_bandwidth(
+enum dc_status dce112_validate_bandwidth(
struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool result = false;
@@ -952,7 +952,7 @@ bool dce112_validate_bandwidth(
context->bw_ctx.bw.dce.yclk_khz,
context->bw_ctx.bw.dce.blackout_recovery_time_us);
}
- return result;
+ return result ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
}
enum dc_status resource_map_phy_clock_resources(
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.h
index 1f57ebc6f9b4..3efc4c55d2d2 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dce112/dce112_resource.h
@@ -42,10 +42,10 @@ enum dc_status dce112_validate_with_context(
struct dc_state *context,
struct dc_state *old_context);
-bool dce112_validate_bandwidth(
+enum dc_status dce112_validate_bandwidth(
struct dc *dc,
struct dc_state *context,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
enum dc_status dce112_add_stream_to_ctx(
struct dc *dc,
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c
index c63c59623433..eb1e158d3436 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dce120/dce120_resource.c
@@ -67,6 +67,7 @@
#include "reg_helper.h"
#include "dce100/dce100_resource.h"
+#include "link.h"
#ifndef mmDP0_DP_DPHY_INTERNAL_CTRL
#define mmDP0_DP_DPHY_INTERNAL_CTRL 0x210f
@@ -659,6 +660,12 @@ static void dce120_resource_destruct(struct dce110_resource_pool *pool)
if (pool->base.dmcu != NULL)
dce_dmcu_destroy(&pool->base.dmcu);
+
+ if (pool->base.oem_device != NULL) {
+ struct dc *dc = pool->base.oem_device->ctx->dc;
+
+ dc->link_srv->destroy_ddc_service(&pool->base.oem_device);
+ }
}
static void read_dce_straps(
@@ -1054,6 +1061,7 @@ static bool dce120_resource_construct(
struct dc *dc,
struct dce110_resource_pool *pool)
{
+ struct ddc_service_init_data ddc_init_data = {0};
unsigned int i;
int j;
struct dc_context *ctx = dc->ctx;
@@ -1257,6 +1265,15 @@ static bool dce120_resource_construct(
bw_calcs_data_update_from_pplib(dc);
+ if (dc->ctx->dc_bios->fw_info.oem_i2c_present) {
+ ddc_init_data.ctx = dc->ctx;
+ ddc_init_data.link = NULL;
+ ddc_init_data.id.id = dc->ctx->dc_bios->fw_info.oem_i2c_obj_id;
+ ddc_init_data.id.enum_id = 0;
+ ddc_init_data.id.type = OBJECT_TYPE_GENERIC;
+ pool->base.oem_device = dc->link_srv->create_ddc_service(&ddc_init_data);
+ }
+
return true;
irqs_create_fail:
diff --git a/drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce60/dce60_resource.c
index 889f314cac65..53b60044653f 100644
--- a/drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dce60/dce60_resource.c
@@ -48,7 +48,7 @@
#include "dce/dce_clock_source.h"
#include "dce/dce_audio.h"
#include "dce/dce_hwseq.h"
-#include "dce60/dce60_hw_sequencer.h"
+#include "dce60/dce60_hwseq.h"
#include "dce100/dce100_resource.h"
#include "dce/dce_panel_cntl.h"
@@ -373,7 +373,7 @@ static const struct resource_caps res_cap = {
.num_timing_generator = 6,
.num_audio = 6,
.num_stream_encoder = 6,
- .num_pll = 2,
+ .num_pll = 3,
.num_ddc = 6,
};
@@ -389,7 +389,7 @@ static const struct resource_caps res_cap_64 = {
.num_timing_generator = 2,
.num_audio = 2,
.num_stream_encoder = 2,
- .num_pll = 2,
+ .num_pll = 3,
.num_ddc = 2,
};
@@ -863,10 +863,10 @@ static void dce60_resource_destruct(struct dce110_resource_pool *pool)
}
}
-static bool dce60_validate_bandwidth(
+static enum dc_status dce60_validate_bandwidth(
struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i;
bool at_least_one_pipe = false;
@@ -885,7 +885,7 @@ static bool dce60_validate_bandwidth(
context->bw_ctx.bw.dce.yclk_khz = 0;
}
- return true;
+ return DC_OK;
}
static bool dce60_validate_surface_sets(
@@ -973,21 +973,24 @@ static bool dce60_construct(
if (bp->fw_info_valid && bp->fw_info.external_clock_source_frequency_for_dp != 0) {
pool->base.dp_clock_source =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_EXTERNAL, NULL, true);
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_EXTERNAL, NULL, true);
+ /* DCE 6.0 and 6.4: PLL0 can only be used with DP. Don't initialize it here. */
pool->base.clock_sources[0] =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL0, &clk_src_regs[0], false);
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL1, &clk_src_regs[1], false);
pool->base.clock_sources[1] =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL1, &clk_src_regs[1], false);
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL2, &clk_src_regs[2], false);
pool->base.clk_src_count = 2;
} else {
pool->base.dp_clock_source =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL0, &clk_src_regs[0], true);
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL0, &clk_src_regs[0], true);
pool->base.clock_sources[0] =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL1, &clk_src_regs[1], false);
- pool->base.clk_src_count = 1;
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL1, &clk_src_regs[1], false);
+ pool->base.clock_sources[1] =
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL2, &clk_src_regs[2], false);
+ pool->base.clk_src_count = 2;
}
if (pool->base.dp_clock_source == NULL) {
@@ -1365,21 +1368,24 @@ static bool dce64_construct(
if (bp->fw_info_valid && bp->fw_info.external_clock_source_frequency_for_dp != 0) {
pool->base.dp_clock_source =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_EXTERNAL, NULL, true);
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_EXTERNAL, NULL, true);
+ /* DCE 6.0 and 6.4: PLL0 can only be used with DP. Don't initialize it here. */
pool->base.clock_sources[0] =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL1, &clk_src_regs[0], false);
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL1, &clk_src_regs[1], false);
pool->base.clock_sources[1] =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL2, &clk_src_regs[1], false);
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL2, &clk_src_regs[2], false);
pool->base.clk_src_count = 2;
} else {
pool->base.dp_clock_source =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL1, &clk_src_regs[0], true);
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL0, &clk_src_regs[0], true);
pool->base.clock_sources[0] =
- dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL2, &clk_src_regs[1], false);
- pool->base.clk_src_count = 1;
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL1, &clk_src_regs[1], false);
+ pool->base.clock_sources[1] =
+ dce60_clock_source_create(ctx, bp, CLOCK_SOURCE_ID_PLL2, &clk_src_regs[2], false);
+ pool->base.clk_src_count = 2;
}
if (pool->base.dp_clock_source == NULL) {
diff --git a/drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dce60/dce60_resource.h
index 5d653a76b0b0..5d653a76b0b0 100644
--- a/drivers/gpu/drm/amd/display/dc/dce60/dce60_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dce60/dce60_resource.h
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c
index 3d5113f010bb..3e8b0ac11d90 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dce80/dce80_resource.c
@@ -869,10 +869,10 @@ static void dce80_resource_destruct(struct dce110_resource_pool *pool)
}
}
-static bool dce80_validate_bandwidth(
+static enum dc_status dce80_validate_bandwidth(
struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i;
bool at_least_one_pipe = false;
@@ -891,7 +891,7 @@ static bool dce80_validate_bandwidth(
context->bw_ctx.bw.dce.yclk_khz = 0;
}
- return true;
+ return DC_OK;
}
static bool dce80_validate_surface_sets(
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c
index 770a380cc03d..652c05c35494 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.c
@@ -23,6 +23,7 @@
*
*/
+#include "core_status.h"
#include "dm_services.h"
#include "dc.h"
@@ -1125,18 +1126,18 @@ static void dcn10_destroy_resource_pool(struct resource_pool **pool)
*pool = NULL;
}
-static bool dcn10_validate_bandwidth(
+static enum dc_status dcn10_validate_bandwidth(
struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool voltage_supported;
DC_FP_START();
- voltage_supported = dcn_validate_bandwidth(dc, context, fast_validate);
+ voltage_supported = dcn_validate_bandwidth(dc, context, validate_mode);
DC_FP_END();
- return voltage_supported;
+ return voltage_supported ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
}
static enum dc_status dcn10_validate_plane(const struct dc_plane_state *plane_state, struct dc_caps *caps)
@@ -1245,6 +1246,10 @@ struct stream_encoder *dcn10_find_first_free_match_stream_enc_for_link(
if (link->ep_type == DISPLAY_ENDPOINT_PHY && pool->stream_enc[i]->id ==
link->link_enc->preferred_engine)
return pool->stream_enc[i];
+
+ if (link->ep_type == DISPLAY_ENDPOINT_USB4_DPIA && pool->stream_enc[i]->id ==
+ link->dpia_preferred_eng_id)
+ return pool->stream_enc[i];
}
}
@@ -1258,6 +1263,11 @@ struct stream_encoder *dcn10_find_first_free_match_stream_enc_for_link(
return NULL;
}
+unsigned int dcn10_get_vstartup_for_pipe(struct pipe_ctx *pipe_ctx)
+{
+ return pipe_ctx->pipe_dlg_param.vstartup_start;
+}
+
static const struct dc_cap_funcs cap_funcs = {
.get_dcc_compression_cap = dcn10_get_dcc_compression_cap
};
@@ -1272,7 +1282,8 @@ static const struct resource_funcs dcn10_res_pool_funcs = {
.validate_global = dcn10_validate_global,
.add_stream_to_ctx = dcn10_add_stream_to_ctx,
.patch_unknown_plane_state = dcn10_patch_unknown_plane_state,
- .find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link
+ .find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe
};
static uint32_t read_pipe_fuses(struct dc_context *ctx)
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.h
index bf8e33cd8147..7bc1be53e800 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn10/dcn10_resource.h
@@ -51,6 +51,7 @@ struct stream_encoder *dcn10_find_first_free_match_stream_enc_for_link(
const struct resource_pool *pool,
struct dc_stream_state *stream);
+unsigned int dcn10_get_vstartup_for_pipe(struct pipe_ctx *pipe_ctx);
#endif /* __DC_RESOURCE_DCN10_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c
index 189d0c85872e..f9cbdad3ef37 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.c
@@ -1220,7 +1220,7 @@ static void get_pixel_clock_parameters(
struct pipe_ctx *odm_pipe;
int opp_cnt = 1;
struct dc_link *link = stream->link;
- struct link_encoder *link_enc = NULL;
+ struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
struct dc *dc = pipe_ctx->stream->ctx->dc;
struct dce_hwseq *hws = dc->hwseq;
@@ -1229,7 +1229,8 @@ static void get_pixel_clock_parameters(
pixel_clk_params->requested_pix_clk_100hz = stream->timing.pix_clk_100hz;
- link_enc = link_enc_cfg_get_link_enc(link);
+ if (!dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
if (link_enc)
pixel_clk_params->encoder_object_id = link_enc->id;
@@ -1509,41 +1510,12 @@ bool dcn20_split_stream_for_odm(
next_odm_pipe->prev_odm_pipe = prev_odm_pipe;
if (prev_odm_pipe->plane_state) {
- struct scaler_data *sd = &prev_odm_pipe->plane_res.scl_data;
- int new_width;
-
- /* HACTIVE halved for odm combine */
- sd->h_active /= 2;
- /* Calculate new vp and recout for left pipe */
- /* Need at least 16 pixels width per side */
- if (sd->recout.x + 16 >= sd->h_active)
- return false;
- new_width = sd->h_active - sd->recout.x;
- sd->viewport.width -= dc_fixpt_floor(dc_fixpt_mul_int(
- sd->ratios.horz, sd->recout.width - new_width));
- sd->viewport_c.width -= dc_fixpt_floor(dc_fixpt_mul_int(
- sd->ratios.horz_c, sd->recout.width - new_width));
- sd->recout.width = new_width;
-
- /* Calculate new vp and recout for right pipe */
- sd = &next_odm_pipe->plane_res.scl_data;
- /* HACTIVE halved for odm combine */
- sd->h_active /= 2;
- /* Need at least 16 pixels width per side */
- if (new_width <= 16)
- return false;
- new_width = sd->recout.width + sd->recout.x - sd->h_active;
- sd->viewport.width -= dc_fixpt_floor(dc_fixpt_mul_int(
- sd->ratios.horz, sd->recout.width - new_width));
- sd->viewport_c.width -= dc_fixpt_floor(dc_fixpt_mul_int(
- sd->ratios.horz_c, sd->recout.width - new_width));
- sd->recout.width = new_width;
- sd->viewport.x += dc_fixpt_floor(dc_fixpt_mul_int(
- sd->ratios.horz, sd->h_active - sd->recout.x));
- sd->viewport_c.x += dc_fixpt_floor(dc_fixpt_mul_int(
- sd->ratios.horz_c, sd->h_active - sd->recout.x));
- sd->recout.x = 0;
+ if (!resource_build_scaling_params(prev_odm_pipe) ||
+ !resource_build_scaling_params(next_odm_pipe)) {
+ return false;
+ }
}
+
if (!next_odm_pipe->top_pipe)
next_odm_pipe->stream_res.opp = pool->opps[next_odm_pipe->pipe_idx];
else
@@ -2035,7 +2007,7 @@ bool dcn20_fast_validate_bw(
int *pipe_cnt_out,
int *pipe_split_from,
int *vlevel_out,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool out = false;
int split[MAX_PIPES] = { 0 };
@@ -2049,7 +2021,7 @@ bool dcn20_fast_validate_bw(
dcn20_merge_pipes_for_validate(dc, context);
DC_FP_START();
- pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate);
+ pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, validate_mode);
DC_FP_END();
*pipe_cnt_out = pipe_cnt;
@@ -2132,6 +2104,7 @@ bool dcn20_fast_validate_bw(
ASSERT(0);
}
}
+
/* Actual dsc count per stream dsc validation*/
if (!dcn20_validate_dsc(dc, context)) {
context->bw_ctx.dml.vba.ValidationStatus[context->bw_ctx.dml.vba.soc.num_states] =
@@ -2151,22 +2124,22 @@ validate_out:
return out;
}
-bool dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context,
- bool fast_validate)
+enum dc_status dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context,
+ enum dc_validate_mode validate_mode)
{
bool voltage_supported;
display_e2e_pipe_params_st *pipes;
pipes = kcalloc(dc->res_pool->pipe_count, sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
if (!pipes)
- return false;
+ return DC_FAIL_BANDWIDTH_VALIDATE;
DC_FP_START();
- voltage_supported = dcn20_validate_bandwidth_fp(dc, context, fast_validate, pipes);
+ voltage_supported = dcn20_validate_bandwidth_fp(dc, context, validate_mode, pipes);
DC_FP_END();
kfree(pipes);
- return voltage_supported;
+ return voltage_supported ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
}
struct pipe_ctx *dcn20_acquire_free_pipe_for_layer(
@@ -2257,7 +2230,8 @@ static const struct resource_funcs dcn20_res_pool_funcs = {
.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
.set_mcif_arb_params = dcn20_set_mcif_arb_params,
.populate_dml_pipes = dcn20_populate_dml_pipes_from_context,
- .find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link
+ .find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe
};
bool dcn20_dwbc_create(struct dc_context *ctx, struct resource_pool *pool)
@@ -2762,6 +2736,8 @@ static bool dcn20_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 2;
+
dc->cap_funcs = cap_funcs;
if (dc->ctx->dc_bios->fw_info.oem_i2c_present) {
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.h
index 4cee3fa11a7f..e997d35a8b86 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn20/dcn20_resource.h
@@ -119,7 +119,7 @@ void dcn20_set_mcif_arb_params(
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
int pipe_cnt);
-bool dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context, bool fast_validate);
+enum dc_status dcn20_validate_bandwidth(struct dc *dc, struct dc_state *context, enum dc_validate_mode validate_mode);
void dcn20_merge_pipes_for_validate(
struct dc *dc,
struct dc_state *context);
@@ -158,7 +158,7 @@ bool dcn20_fast_validate_bw(
int *pipe_cnt_out,
int *pipe_split_from,
int *vlevel_out,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
enum dc_status dcn20_build_mapped_resource(const struct dc *dc, struct dc_state *context, struct dc_stream_state *stream);
enum dc_status dcn20_add_stream_to_ctx(struct dc *dc, struct dc_state *new_ctx, struct dc_stream_state *dc_stream);
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c
index d3d67d366523..e4a1338d21e0 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn201/dcn201_resource.c
@@ -59,8 +59,8 @@
#include "cyan_skillfish_ip_offset.h"
-#include "dcn/dcn_2_0_3_offset.h"
-#include "dcn/dcn_2_0_3_sh_mask.h"
+#include "dcn/dcn_2_0_1_offset.h"
+#include "dcn/dcn_2_0_1_sh_mask.h"
#include "dpcs/dpcs_2_0_3_offset.h"
#include "dpcs/dpcs_2_0_3_sh_mask.h"
@@ -1079,7 +1079,8 @@ static struct resource_funcs dcn201_res_pool_funcs = {
.populate_dml_writeback_from_context = dcn201_populate_dml_writeback_from_context,
.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
.set_mcif_arb_params = dcn20_set_mcif_arb_params,
- .find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link
+ .find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe
};
static bool dcn201_resource_construct(
@@ -1284,6 +1285,8 @@ static bool dcn201_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 2;
+
dc->cap_funcs = cap_funcs;
return true;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c
index 021ba8ac5c8c..918742a42ded 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.c
@@ -769,7 +769,7 @@ bool dcn21_fast_validate_bw(struct dc *dc,
int *pipe_cnt_out,
int *pipe_split_from,
int *vlevel_out,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool out = false;
int split[MAX_PIPES] = { 0 };
@@ -783,7 +783,7 @@ bool dcn21_fast_validate_bw(struct dc *dc,
dcn20_merge_pipes_for_validate(dc, context);
DC_FP_START();
- pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate);
+ pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, validate_mode);
DC_FP_END();
*pipe_cnt_out = pipe_cnt;
@@ -923,22 +923,22 @@ validate_out:
* with DC_FP_START()/DC_FP_END(). Use the same approach as for
* dcn20_validate_bandwidth in dcn20_resource.c.
*/
-static bool dcn21_validate_bandwidth(struct dc *dc, struct dc_state *context,
- bool fast_validate)
+static enum dc_status dcn21_validate_bandwidth(struct dc *dc, struct dc_state *context,
+ enum dc_validate_mode validate_mode)
{
bool voltage_supported;
display_e2e_pipe_params_st *pipes;
pipes = kcalloc(dc->res_pool->pipe_count, sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
if (!pipes)
- return false;
+ return DC_FAIL_BANDWIDTH_VALIDATE;
DC_FP_START();
- voltage_supported = dcn21_validate_bandwidth_fp(dc, context, fast_validate, pipes);
+ voltage_supported = dcn21_validate_bandwidth_fp(dc, context, validate_mode, pipes);
DC_FP_END();
kfree(pipes);
- return voltage_supported;
+ return voltage_supported ? DC_OK : DC_NOT_SUPPORTED;
}
static void dcn21_destroy_resource_pool(struct resource_pool **pool)
@@ -1378,6 +1378,7 @@ static const struct resource_funcs dcn21_res_pool_funcs = {
.find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link,
.update_bw_bounding_box = dcn21_update_bw_bounding_box,
.get_panel_config_defaults = dcn21_get_panel_config_defaults,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe
};
static bool dcn21_resource_construct(
@@ -1412,9 +1413,9 @@ static bool dcn21_resource_construct(
dc->caps.min_horizontal_blanking_period = 80;
dc->caps.dmdata_alloc_size = 2048;
- dc->caps.max_slave_planes = 1;
- dc->caps.max_slave_yuv_planes = 1;
- dc->caps.max_slave_rgb_planes = 1;
+ dc->caps.max_slave_planes = 3;
+ dc->caps.max_slave_yuv_planes = 3;
+ dc->caps.max_slave_rgb_planes = 3;
dc->caps.post_blend_color_processing = true;
dc->caps.force_dp_tps4_for_cp2520 = true;
dc->caps.extended_aux_timeout_support = true;
@@ -1683,6 +1684,8 @@ static bool dcn21_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 2;
+
dc->cap_funcs = cap_funcs;
return true;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.h
index f7ecc002c2f7..a017fd9854d1 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn21/dcn21_resource.h
@@ -51,6 +51,6 @@ bool dcn21_fast_validate_bw(
int *pipe_cnt_out,
int *pipe_split_from,
int *vlevel_out,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
#endif /* _DCN21_RESOURCE_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c
index cd31e4f16c14..895349d9ca07 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.c
@@ -1319,13 +1319,13 @@ static struct clock_source *dcn30_clock_source_create(
int dcn30_populate_dml_pipes_from_context(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i, pipe_cnt;
struct resource_context *res_ctx = &context->res_ctx;
DC_FP_START();
- dcn20_populate_dml_pipes_from_context(dc, context, pipes, fast_validate);
+ dcn20_populate_dml_pipes_from_context(dc, context, pipes, validate_mode);
DC_FP_END();
for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) {
@@ -1627,7 +1627,7 @@ noinline bool dcn30_internal_validate_bw(
display_e2e_pipe_params_st *pipes,
int *pipe_cnt_out,
int *vlevel_out,
- bool fast_validate,
+ enum dc_validate_mode validate_mode,
bool allow_self_refresh_only)
{
bool out = false;
@@ -1646,7 +1646,7 @@ noinline bool dcn30_internal_validate_bw(
context->bw_ctx.dml.vba.VoltageLevel = 0;
context->bw_ctx.dml.vba.DRAMClockChangeSupport[0][0] = dm_dram_clock_change_vactive;
dc->res_pool->funcs->update_soc_for_wm_a(dc, context);
- pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate);
+ pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, validate_mode);
if (!pipe_cnt) {
out = true;
@@ -1655,7 +1655,7 @@ noinline bool dcn30_internal_validate_bw(
dml_log_pipe_params(&context->bw_ctx.dml, pipes, pipe_cnt);
- if (!fast_validate || !allow_self_refresh_only) {
+ if (validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING || !allow_self_refresh_only) {
/*
* DML favors voltage over p-state, but we're more interested in
* supporting p-state over voltage. We can't support p-state in
@@ -1669,7 +1669,7 @@ noinline bool dcn30_internal_validate_bw(
vlevel = dcn20_validate_apply_pipe_split_flags(dc, context, vlevel, split, merge);
}
if (allow_self_refresh_only &&
- (fast_validate || vlevel == context->bw_ctx.dml.soc.num_states ||
+ (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING || vlevel == context->bw_ctx.dml.soc.num_states ||
vba->DRAMClockChangeSupport[vlevel][vba->maxMpcComb] == dm_dram_clock_change_unsupported)) {
/*
* If mode is unsupported or there's still no p-state support
@@ -1678,7 +1678,7 @@ noinline bool dcn30_internal_validate_bw(
* We don't actually support prefetch mode 2, so require that we
* at least support prefetch mode 1.
*/
- context->bw_ctx.dml.validate_max_state = fast_validate;
+ context->bw_ctx.dml.validate_max_state = (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING);
context->bw_ctx.dml.soc.allow_dram_self_refresh_or_dram_clock_change_in_vblank =
dm_allow_self_refresh;
@@ -1865,7 +1865,7 @@ noinline bool dcn30_internal_validate_bw(
}
if (repopulate_pipes)
- pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, fast_validate);
+ pipe_cnt = dc->res_pool->funcs->populate_dml_pipes(dc, context, pipes, validate_mode);
context->bw_ctx.dml.vba.VoltageLevel = vlevel;
*vlevel_out = vlevel;
*pipe_cnt_out = pipe_cnt;
@@ -1891,8 +1891,6 @@ static int get_refresh_rate(struct dc_state *context)
/* check if refresh rate at least 120hz */
timing = &context->streams[0]->timing;
- if (timing == NULL)
- return 0;
h_v_total = timing->h_total * timing->v_total;
if (h_v_total == 0)
@@ -2037,9 +2035,9 @@ void dcn30_calculate_wm_and_dlg(
DC_FP_END();
}
-bool dcn30_validate_bandwidth(struct dc *dc,
+enum dc_status dcn30_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool out = false;
@@ -2047,7 +2045,8 @@ bool dcn30_validate_bandwidth(struct dc *dc,
int vlevel = 0;
int pipe_cnt = 0;
- display_e2e_pipe_params_st *pipes = kzalloc(dc->res_pool->pipe_count * sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
+ display_e2e_pipe_params_st *pipes = kcalloc(dc->res_pool->pipe_count,
+ sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
DC_LOGGER_INIT(dc->ctx->logger);
BW_VAL_TRACE_COUNT();
@@ -2056,7 +2055,7 @@ bool dcn30_validate_bandwidth(struct dc *dc,
goto validate_fail;
DC_FP_START();
- out = dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, fast_validate, true);
+ out = dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, validate_mode, true);
DC_FP_END();
if (pipe_cnt == 0)
@@ -2067,7 +2066,7 @@ bool dcn30_validate_bandwidth(struct dc *dc,
BW_VAL_TRACE_END_VOLTAGE_LEVEL();
- if (fast_validate) {
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING) {
BW_VAL_TRACE_SKIP(fast);
goto validate_out;
}
@@ -2093,7 +2092,7 @@ validate_out:
BW_VAL_TRACE_FINISH();
- return out;
+ return out ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
}
void dcn30_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params)
@@ -2250,6 +2249,7 @@ static const struct resource_funcs dcn30_res_pool_funcs = {
.update_bw_bounding_box = dcn30_update_bw_bounding_box,
.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
.get_panel_config_defaults = dcn30_get_panel_config_defaults,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe
};
#define CTX ctx
@@ -2353,6 +2353,7 @@ static bool dcn30_resource_construct(
dc->caps.dp_hdmi21_pcon_support = true;
dc->caps.max_v_total = (1 << 15) - 1;
+ dc->caps.vtotal_limited_by_fp2 = true;
/* read VBIOS LTTPR caps */
{
@@ -2585,6 +2586,8 @@ static bool dcn30_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
if (dc->ctx->dc_bios->fw_info.oem_i2c_present) {
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.h
index 8e6b8b7368fd..2c967fe55712 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn30/dcn30_resource.h
@@ -56,15 +56,15 @@ unsigned int dcn30_calc_max_scaled_time(
enum mmhubbub_wbif_mode mode,
unsigned int urgent_watermark);
-bool dcn30_validate_bandwidth(struct dc *dc, struct dc_state *context,
- bool fast_validate);
+enum dc_status dcn30_validate_bandwidth(struct dc *dc, struct dc_state *context,
+ enum dc_validate_mode validate_mode);
bool dcn30_internal_validate_bw(
struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
int *pipe_cnt_out,
int *vlevel_out,
- bool fast_validate,
+ enum dc_validate_mode validate_mode,
bool allow_self_refresh_only);
void dcn30_calculate_wm_and_dlg(
struct dc *dc, struct dc_state *context,
@@ -78,7 +78,7 @@ void dcn30_populate_dml_writeback_from_context(
int dcn30_populate_dml_pipes_from_context(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
bool dcn30_acquire_post_bldn_3dlut(
struct resource_context *res_ctx,
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn301/dcn301_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn301/dcn301_resource.c
index a9816affd312..82a205a7c25c 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn301/dcn301_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn301/dcn301_resource.c
@@ -671,9 +671,9 @@ static const struct dc_plane_cap plane_cap = {
/* 6:1 downscaling ratio: 1000/6 = 166.666 */
.max_downscale_factor = {
- .argb8888 = 167,
- .nv12 = 167,
- .fp16 = 167
+ .argb8888 = 358,
+ .nv12 = 358,
+ .fp16 = 358
},
64,
64
@@ -693,7 +693,7 @@ static const struct dc_debug_options debug_defaults_drv = {
.disable_dcc = DCC_ENABLE,
.vsr_support = true,
.performance_trace = false,
- .max_downscale_src_width = 7680,/*upto 8K*/
+ .max_downscale_src_width = 4096,/*upto true 4k*/
.scl_reset_length10 = true,
.sanity_checks = false,
.underflow_assert_delay_us = 0xFFFFFFFF,
@@ -1400,7 +1400,8 @@ static struct resource_funcs dcn301_res_pool_funcs = {
.acquire_post_bldn_3dlut = dcn30_acquire_post_bldn_3dlut,
.release_post_bldn_3dlut = dcn30_release_post_bldn_3dlut,
.update_bw_bounding_box = dcn301_update_bw_bounding_box,
- .patch_unknown_plane_state = dcn20_patch_unknown_plane_state
+ .patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe
};
static bool dcn301_resource_construct(
@@ -1705,6 +1706,8 @@ static bool dcn301_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
return true;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c
index 02af8b8f4d27..3345068a878c 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn302/dcn302_resource.c
@@ -1151,6 +1151,7 @@ static struct resource_funcs dcn302_res_pool_funcs = {
.update_bw_bounding_box = dcn302_update_bw_bounding_box,
.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
.get_panel_config_defaults = dcn302_get_panel_config_defaults,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe
};
static struct dc_cap_funcs cap_funcs = {
@@ -1233,6 +1234,7 @@ static bool dcn302_resource_construct(
dc->caps.extended_aux_timeout_support = true;
dc->caps.dmcub_support = true;
dc->caps.max_v_total = (1 << 15) - 1;
+ dc->caps.vtotal_limited_by_fp2 = true;
/* Color pipeline capabilities */
dc->caps.color.dpp.dcn_arch = 1;
@@ -1479,6 +1481,8 @@ static bool dcn302_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
if (dc->ctx->dc_bios->fw_info.oem_i2c_present) {
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c
index 7002a8dd358a..3479e1eab4cd 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn303/dcn303_resource.c
@@ -1096,6 +1096,7 @@ static struct resource_funcs dcn303_res_pool_funcs = {
.update_bw_bounding_box = dcn303_update_bw_bounding_box,
.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
.get_panel_config_defaults = dcn303_get_panel_config_defaults,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe
};
static struct dc_cap_funcs cap_funcs = {
@@ -1178,6 +1179,7 @@ static bool dcn303_resource_construct(
dc->caps.extended_aux_timeout_support = true;
dc->caps.dmcub_support = true;
dc->caps.max_v_total = (1 << 15) - 1;
+ dc->caps.vtotal_limited_by_fp2 = true;
/* Color pipeline capabilities */
dc->caps.color.dpp.dcn_arch = 1;
@@ -1412,6 +1414,8 @@ static bool dcn303_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
if (dc->ctx->dc_bios->fw_info.oem_i2c_present) {
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c
index c16cf1c8f7f9..3ed7f50554e2 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.c
@@ -890,7 +890,7 @@ static const struct dc_debug_options debug_defaults_drv = {
.disable_z10 = true,
.enable_legacy_fast_update = true,
.enable_z9_disable_interface = true, /* Allow support for the PMFW interface for disable Z9*/
- .dml_hostvm_override = DML_HOSTVM_NO_OVERRIDE,
+ .dml_hostvm_override = DML_HOSTVM_OVERRIDE_FALSE,
.using_dml2 = false,
};
@@ -1616,14 +1616,14 @@ static bool is_dual_plane(enum surface_pixel_format format)
int dcn31x_populate_dml_pipes_from_context(struct dc *dc,
struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
uint32_t pipe_cnt;
int i;
dc_assert_fp_enabled();
- pipe_cnt = dcn20_populate_dml_pipes_from_context(dc, context, pipes, fast_validate);
+ pipe_cnt = dcn20_populate_dml_pipes_from_context(dc, context, pipes, validate_mode);
for (i = 0; i < pipe_cnt; i++) {
pipes[i].pipe.src.gpuvm = 1;
@@ -1641,7 +1641,7 @@ int dcn31x_populate_dml_pipes_from_context(struct dc *dc,
int dcn31_populate_dml_pipes_from_context(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i, pipe_cnt;
struct resource_context *res_ctx = &context->res_ctx;
@@ -1649,7 +1649,7 @@ int dcn31_populate_dml_pipes_from_context(
bool upscaled = false;
DC_FP_START();
- dcn31x_populate_dml_pipes_from_context(dc, context, pipes, fast_validate);
+ dcn31x_populate_dml_pipes_from_context(dc, context, pipes, validate_mode);
DC_FP_END();
for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) {
@@ -1720,6 +1720,12 @@ int dcn31_populate_dml_pipes_from_context(
return pipe_cnt;
}
+unsigned int dcn31_get_det_buffer_size(
+ const struct dc_state *context)
+{
+ return context->bw_ctx.dml.ip.det_buffer_size_kbytes;
+}
+
void dcn31_calculate_wm_and_dlg(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
@@ -1752,9 +1758,9 @@ dcn31_set_mcif_arb_params(struct dc *dc,
DC_FP_END();
}
-bool dcn31_validate_bandwidth(struct dc *dc,
+enum dc_status dcn31_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool out = false;
@@ -1762,7 +1768,8 @@ bool dcn31_validate_bandwidth(struct dc *dc,
int vlevel = 0;
int pipe_cnt = 0;
- display_e2e_pipe_params_st *pipes = kzalloc(dc->res_pool->pipe_count * sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
+ display_e2e_pipe_params_st *pipes = kcalloc(dc->res_pool->pipe_count,
+ sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
DC_LOGGER_INIT(dc->ctx->logger);
BW_VAL_TRACE_COUNT();
@@ -1771,19 +1778,19 @@ bool dcn31_validate_bandwidth(struct dc *dc,
goto validate_fail;
DC_FP_START();
- out = dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, fast_validate, true);
+ out = dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, validate_mode, true);
DC_FP_END();
- // Disable fast_validate to set min dcfclk in calculate_wm_and_dlg
+ // Disable DC_VALIDATE_MODE_ONLY and DC_VALIDATE_MODE_AND_STATE_INDEX to set min dcfclk in calculate_wm_and_dlg
if (pipe_cnt == 0)
- fast_validate = false;
+ validate_mode = DC_VALIDATE_MODE_AND_PROGRAMMING;
if (!out)
goto validate_fail;
BW_VAL_TRACE_END_VOLTAGE_LEVEL();
- if (fast_validate) {
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING) {
BW_VAL_TRACE_SKIP(fast);
goto validate_out;
}
@@ -1806,7 +1813,7 @@ validate_out:
BW_VAL_TRACE_FINISH();
- return out;
+ return out ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
}
static void dcn31_get_panel_config_defaults(struct dc_panel_config *panel_config)
@@ -1842,6 +1849,10 @@ static struct resource_funcs dcn31_res_pool_funcs = {
.update_bw_bounding_box = dcn31_update_bw_bounding_box,
.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
.get_panel_config_defaults = dcn31_get_panel_config_defaults,
+ .get_det_buffer_size = dcn31_get_det_buffer_size,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe,
+ .update_dc_state_for_encoder_switch = dcn31_update_dc_state_for_encoder_switch,
+ .build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params
};
static struct clock_source *dcn30_clock_source_create(
@@ -1945,6 +1956,9 @@ static bool dcn31_resource_construct(
dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
dc->caps.color.mpc.ocsc = 1;
+ dc->caps.num_of_host_routers = 2;
+ dc->caps.num_of_dpias_per_host_router = 2;
+
/* Use pipe context based otg sync logic */
dc->config.use_pipe_ctx_sync_logic = true;
dc->config.disable_hbr_audio_dp2 = true;
@@ -2190,6 +2204,8 @@ static bool dcn31_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
dc->dcn_ip->max_num_dpp = dcn3_1_ip.max_num_dpp;
@@ -2219,3 +2235,35 @@ struct resource_pool *dcn31_create_resource_pool(
kfree(pool);
return NULL;
}
+
+enum dc_status dcn31_update_dc_state_for_encoder_switch(struct dc_link *link,
+ struct dc_link_settings *link_setting,
+ uint8_t pipe_count,
+ struct pipe_ctx *pipes,
+ struct audio_output *audio_output)
+{
+ struct dc_state *state = link->dc->current_state;
+ int i;
+
+#if defined(CONFIG_DRM_AMD_DC_FP)
+ for (i = 0; i < state->stream_count; i++)
+ if (state->streams[i] && state->streams[i]->link && state->streams[i]->link == link)
+ link->dc->hwss.calculate_pix_rate_divider((struct dc *)link->dc, state, state->streams[i]);
+
+ for (i = 0; i < pipe_count; i++) {
+ link->dc->res_pool->funcs->build_pipe_pix_clk_params(&pipes[i]);
+
+ // Setup audio
+ if (pipes[i].stream_res.audio != NULL)
+ build_audio_output(state, &pipes[i], &audio_output[i]);
+ }
+#else
+ /* This DCN requires rate divider updates and audio reprogramming to allow DP1<-->DP2 link rate switching,
+ * but the above will not compile on architectures without an FPU.
+ */
+ DC_LOG_WARNING("%s: DP1<-->DP2 link retraining will not work on this DCN on non-FPU platforms", __func__);
+ ASSERT(0);
+#endif
+
+ return DC_OK;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.h
index 901436591ed4..c32c85ef0ba4 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn31/dcn31_resource.h
@@ -37,9 +37,9 @@ struct dcn31_resource_pool {
struct resource_pool base;
};
-bool dcn31_validate_bandwidth(struct dc *dc,
+enum dc_status dcn31_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void dcn31_calculate_wm_and_dlg(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
@@ -48,7 +48,7 @@ void dcn31_calculate_wm_and_dlg(
int dcn31_populate_dml_pipes_from_context(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void
dcn31_populate_dml_writeback_from_context(struct dc *dc,
struct resource_context *res_ctx,
@@ -63,6 +63,15 @@ struct resource_pool *dcn31_create_resource_pool(
const struct dc_init_data *init_data,
struct dc *dc);
+unsigned int dcn31_get_det_buffer_size(
+ const struct dc_state *context);
+
+enum dc_status dcn31_update_dc_state_for_encoder_switch(struct dc_link *link,
+ struct dc_link_settings *link_setting,
+ uint8_t pipe_count,
+ struct pipe_ctx *pipes,
+ struct audio_output *audio_output);
+
/*temp: B0 specific before switch to dcn313 headers*/
#ifndef regPHYPLLF_PIXCLK_RESYNC_CNTL
#define regPHYPLLF_PIXCLK_RESYNC_CNTL 0x007e
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c
index c0f48c78e968..663c49cce4aa 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.c
@@ -926,6 +926,7 @@ static const struct dc_debug_options debug_defaults_drv = {
.seamless_boot_odm_combine = true,
.enable_legacy_fast_update = true,
.using_dml2 = false,
+ .disable_dsc_power_gate = true,
};
static const struct dc_panel_config panel_config_defaults = {
@@ -1667,12 +1668,12 @@ static struct clock_source *dcn31_clock_source_create(
static int dcn314_populate_dml_pipes_from_context(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int pipe_cnt;
DC_FP_START();
- pipe_cnt = dcn314_populate_dml_pipes_from_context_fpu(dc, context, pipes, fast_validate);
+ pipe_cnt = dcn314_populate_dml_pipes_from_context_fpu(dc, context, pipes, validate_mode);
DC_FP_END();
return pipe_cnt;
@@ -1694,9 +1695,9 @@ static void dcn314_get_panel_config_defaults(struct dc_panel_config *panel_confi
*panel_config = panel_config_defaults;
}
-bool dcn314_validate_bandwidth(struct dc *dc,
+enum dc_status dcn314_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool out = false;
@@ -1704,7 +1705,8 @@ bool dcn314_validate_bandwidth(struct dc *dc,
int vlevel = 0;
int pipe_cnt = 0;
- display_e2e_pipe_params_st *pipes = kzalloc(dc->res_pool->pipe_count * sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
+ display_e2e_pipe_params_st *pipes = kcalloc(dc->res_pool->pipe_count,
+ sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
DC_LOGGER_INIT(dc->ctx->logger);
BW_VAL_TRACE_COUNT();
@@ -1714,19 +1716,19 @@ bool dcn314_validate_bandwidth(struct dc *dc,
DC_FP_START();
// do not support self refresh only
- out = dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, fast_validate, false);
+ out = dcn30_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, validate_mode, false);
DC_FP_END();
- // Disable fast_validate to set min dcfclk in calculate_wm_and_dlg
+ // Disable DC_VALIDATE_MODE_ONLY and DC_VALIDATE_MODE_AND_STATE_INDEX to set min dcfclk in calculate_wm_and_dlg
if (pipe_cnt == 0)
- fast_validate = false;
+ validate_mode = DC_VALIDATE_MODE_AND_PROGRAMMING;
if (!out)
goto validate_fail;
BW_VAL_TRACE_END_VOLTAGE_LEVEL();
- if (fast_validate) {
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING) {
BW_VAL_TRACE_SKIP(fast);
goto validate_out;
}
@@ -1749,7 +1751,7 @@ validate_out:
BW_VAL_TRACE_FINISH();
- return out;
+ return out ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
}
static struct resource_funcs dcn314_res_pool_funcs = {
@@ -1777,6 +1779,10 @@ static struct resource_funcs dcn314_res_pool_funcs = {
.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
.get_panel_config_defaults = dcn314_get_panel_config_defaults,
.get_preferred_eng_id_dpia = dcn314_get_preferred_eng_id_dpia,
+ .get_det_buffer_size = dcn31_get_det_buffer_size,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe,
+ .update_dc_state_for_encoder_switch = dcn31_update_dc_state_for_encoder_switch,
+ .build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params
};
static struct clock_source *dcn30_clock_source_create(
@@ -1882,6 +1888,9 @@ static bool dcn314_resource_construct(
dc->caps.max_disp_clock_khz_at_vmin = 650000;
+ dc->caps.num_of_host_routers = 2;
+ dc->caps.num_of_dpias_per_host_router = 2;
+
/* Use pipe context based otg sync logic */
dc->config.use_pipe_ctx_sync_logic = true;
@@ -2111,6 +2120,8 @@ static bool dcn314_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
dc->dcn_ip->max_num_dpp = dcn3_14_ip.max_num_dpp;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.h
index 49ffe71018df..ac9bb7f097d5 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn314/dcn314_resource.h
@@ -39,9 +39,9 @@ struct dcn314_resource_pool {
struct resource_pool base;
};
-bool dcn314_validate_bandwidth(struct dc *dc,
+enum dc_status dcn314_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
struct resource_pool *dcn314_create_resource_pool(
const struct dc_init_data *init_data,
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c
index 6c3295259a81..82cc78c291d8 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn315/dcn315_resource.c
@@ -1664,7 +1664,7 @@ static bool allow_pixel_rate_crb(struct dc *dc, struct dc_state *context)
static int dcn315_populate_dml_pipes_from_context(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i, pipe_cnt, crb_idx, crb_pipes;
struct resource_context *res_ctx = &context->res_ctx;
@@ -1674,7 +1674,7 @@ static int dcn315_populate_dml_pipes_from_context(
bool pixel_rate_crb = allow_pixel_rate_crb(dc, context);
DC_FP_START();
- dcn31x_populate_dml_pipes_from_context(dc, context, pipes, fast_validate);
+ dcn31x_populate_dml_pipes_from_context(dc, context, pipes, validate_mode);
DC_FP_END();
for (i = 0, pipe_cnt = 0, crb_pipes = 0; i < dc->res_pool->pipe_count; i++) {
@@ -1698,7 +1698,7 @@ static int dcn315_populate_dml_pipes_from_context(
pipes[pipe_cnt].dout.dsc_input_bpc = 0;
DC_FP_START();
dcn31_zero_pipe_dcc_fraction(pipes, pipe_cnt);
- if (pixel_rate_crb && !pipe->top_pipe && !pipe->prev_odm_pipe) {
+ if (pixel_rate_crb) {
int bpp = source_format_to_bpp(pipes[pipe_cnt].pipe.src.source_format);
/* Ceil to crb segment size */
int approx_det_segs_required_for_pstate = dcn_get_approx_det_segs_required_for_pstate(
@@ -1755,28 +1755,26 @@ static int dcn315_populate_dml_pipes_from_context(
continue;
}
- if (!pipe->top_pipe && !pipe->prev_odm_pipe) {
- bool split_required = pipe->stream->timing.pix_clk_100hz >= dcn_get_max_non_odm_pix_rate_100hz(&dc->dml.soc)
- || (pipe->plane_state && pipe->plane_state->src_rect.width > 5120);
-
- if (remaining_det_segs > MIN_RESERVED_DET_SEGS && crb_pipes != 0)
- pipes[pipe_cnt].pipe.src.det_size_override += (remaining_det_segs - MIN_RESERVED_DET_SEGS) / crb_pipes +
- (crb_idx < (remaining_det_segs - MIN_RESERVED_DET_SEGS) % crb_pipes ? 1 : 0);
- if (pipes[pipe_cnt].pipe.src.det_size_override > 2 * DCN3_15_MAX_DET_SEGS) {
- /* Clamp to 2 pipe split max det segments */
- remaining_det_segs += pipes[pipe_cnt].pipe.src.det_size_override - 2 * (DCN3_15_MAX_DET_SEGS);
- pipes[pipe_cnt].pipe.src.det_size_override = 2 * DCN3_15_MAX_DET_SEGS;
- }
- if (pipes[pipe_cnt].pipe.src.det_size_override > DCN3_15_MAX_DET_SEGS || split_required) {
- /* If we are splitting we must have an even number of segments */
- remaining_det_segs += pipes[pipe_cnt].pipe.src.det_size_override % 2;
- pipes[pipe_cnt].pipe.src.det_size_override -= pipes[pipe_cnt].pipe.src.det_size_override % 2;
- }
- /* Convert segments into size for DML use */
- pipes[pipe_cnt].pipe.src.det_size_override *= DCN3_15_CRB_SEGMENT_SIZE_KB;
-
- crb_idx++;
+ bool split_required = pipe->stream->timing.pix_clk_100hz >= dcn_get_max_non_odm_pix_rate_100hz(&dc->dml.soc)
+ || (pipe->plane_state && pipe->plane_state->src_rect.width > 5120);
+
+ if (remaining_det_segs > MIN_RESERVED_DET_SEGS && crb_pipes != 0)
+ pipes[pipe_cnt].pipe.src.det_size_override += (remaining_det_segs - MIN_RESERVED_DET_SEGS) / crb_pipes +
+ (crb_idx < (remaining_det_segs - MIN_RESERVED_DET_SEGS) % crb_pipes ? 1 : 0);
+ if (pipes[pipe_cnt].pipe.src.det_size_override > 2 * DCN3_15_MAX_DET_SEGS) {
+ /* Clamp to 2 pipe split max det segments */
+ remaining_det_segs += pipes[pipe_cnt].pipe.src.det_size_override - 2 * (DCN3_15_MAX_DET_SEGS);
+ pipes[pipe_cnt].pipe.src.det_size_override = 2 * DCN3_15_MAX_DET_SEGS;
+ }
+ if (pipes[pipe_cnt].pipe.src.det_size_override > DCN3_15_MAX_DET_SEGS || split_required) {
+ /* If we are splitting we must have an even number of segments */
+ remaining_det_segs += pipes[pipe_cnt].pipe.src.det_size_override % 2;
+ pipes[pipe_cnt].pipe.src.det_size_override -= pipes[pipe_cnt].pipe.src.det_size_override % 2;
}
+ /* Convert segments into size for DML use */
+ pipes[pipe_cnt].pipe.src.det_size_override *= DCN3_15_CRB_SEGMENT_SIZE_KB;
+
+ crb_idx++;
pipe_cnt++;
}
}
@@ -1845,6 +1843,10 @@ static struct resource_funcs dcn315_res_pool_funcs = {
.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
.get_panel_config_defaults = dcn315_get_panel_config_defaults,
.get_power_profile = dcn315_get_power_profile,
+ .get_det_buffer_size = dcn31_get_det_buffer_size,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe,
+ .update_dc_state_for_encoder_switch = dcn31_update_dc_state_for_encoder_switch,
+ .build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params
};
static bool dcn315_resource_construct(
@@ -2140,6 +2142,8 @@ static bool dcn315_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
dc->dcn_ip->max_num_dpp = dcn3_15_ip.max_num_dpp;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c
index 6edaaadcb173..636110e48d01 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn316/dcn316_resource.c
@@ -1610,7 +1610,7 @@ static bool is_dual_plane(enum surface_pixel_format format)
static int dcn316_populate_dml_pipes_from_context(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i, pipe_cnt;
struct resource_context *res_ctx = &context->res_ctx;
@@ -1618,7 +1618,7 @@ static int dcn316_populate_dml_pipes_from_context(
const int max_usable_det = context->bw_ctx.dml.ip.config_return_buffer_size_in_kbytes - DCN3_16_MIN_COMPBUF_SIZE_KB;
DC_FP_START();
- dcn31x_populate_dml_pipes_from_context(dc, context, pipes, fast_validate);
+ dcn31x_populate_dml_pipes_from_context(dc, context, pipes, validate_mode);
DC_FP_END();
for (i = 0, pipe_cnt = 0; i < dc->res_pool->pipe_count; i++) {
@@ -1719,6 +1719,10 @@ static struct resource_funcs dcn316_res_pool_funcs = {
.update_bw_bounding_box = dcn316_update_bw_bounding_box,
.patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
.get_panel_config_defaults = dcn316_get_panel_config_defaults,
+ .get_det_buffer_size = dcn31_get_det_buffer_size,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe,
+ .update_dc_state_for_encoder_switch = dcn31_update_dc_state_for_encoder_switch,
+ .build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params
};
static bool dcn316_resource_construct(
@@ -2006,6 +2010,8 @@ static bool dcn316_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
dc->dcn_ip->max_num_dpp = dcn3_16_ip.max_num_dpp;
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c
index 01d1a11d5545..9917b366f00c 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.c
@@ -24,6 +24,7 @@
*
*/
+#include "dc_types.h"
#include "dm_services.h"
#include "dc.h"
@@ -1741,7 +1742,7 @@ void dcn32_add_phantom_pipes(struct dc *dc, struct dc_state *context,
}
}
-static bool dml1_validate(struct dc *dc, struct dc_state *context, bool fast_validate)
+static bool dml1_validate(struct dc *dc, struct dc_state *context, enum dc_validate_mode validate_mode)
{
bool out = false;
@@ -1749,7 +1750,8 @@ static bool dml1_validate(struct dc *dc, struct dc_state *context, bool fast_val
int vlevel = 0;
int pipe_cnt = 0;
- display_e2e_pipe_params_st *pipes = kzalloc(dc->res_pool->pipe_count * sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
+ display_e2e_pipe_params_st *pipes = kcalloc(dc->res_pool->pipe_count,
+ sizeof(display_e2e_pipe_params_st), GFP_KERNEL);
/* To handle Freesync properly, setting FreeSync DML parameters
* to its default state for the first stage of validation
@@ -1765,7 +1767,7 @@ static bool dml1_validate(struct dc *dc, struct dc_state *context, bool fast_val
goto validate_fail;
DC_FP_START();
- out = dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, fast_validate);
+ out = dcn32_internal_validate_bw(dc, context, pipes, &pipe_cnt, &vlevel, validate_mode);
DC_FP_END();
if (pipe_cnt == 0)
@@ -1776,7 +1778,7 @@ static bool dml1_validate(struct dc *dc, struct dc_state *context, bool fast_val
BW_VAL_TRACE_END_VOLTAGE_LEVEL();
- if (fast_validate) {
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING) {
BW_VAL_TRACE_SKIP(fast);
goto validate_out;
}
@@ -1805,25 +1807,62 @@ validate_out:
return out;
}
-bool dcn32_validate_bandwidth(struct dc *dc,
+enum dc_status dcn32_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
- bool out = false;
+ unsigned int i;
+ enum dc_status status;
+ const struct dc_stream_state *stream;
+
+ /* reset cursor limitations on subvp */
+ for (i = 0; i < context->stream_count; i++) {
+ stream = context->streams[i];
+
+ if (dc_state_can_clear_stream_cursor_subvp_limit(stream, context)) {
+ dc_state_set_stream_cursor_subvp_limit(stream, context, false);
+ }
+ }
if (dc->debug.using_dml2)
- out = dml2_validate(dc, context,
+ status = dml2_validate(dc, context,
context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2,
- fast_validate);
+ validate_mode) ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
else
- out = dml1_validate(dc, context, fast_validate);
- return out;
+ status = dml1_validate(dc, context, validate_mode) ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
+
+ if (validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING && status == DC_OK && dc_state_is_subvp_in_use(context)) {
+ /* check new stream configuration still supports cursor if subvp used */
+ for (i = 0; i < context->stream_count; i++) {
+ stream = context->streams[i];
+
+ if (dc_state_get_stream_subvp_type(context, stream) != SUBVP_PHANTOM &&
+ stream->cursor_position.enable &&
+ !dc_stream_check_cursor_attributes(stream, context, &stream->cursor_attributes)) {
+ /* hw cursor cannot be supported with subvp active, so disable subvp for now */
+ dc_state_set_stream_cursor_subvp_limit(stream, context, true);
+ status = DC_FAIL_HW_CURSOR_SUPPORT;
+ }
+ };
+ }
+
+ if (validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING && status == DC_FAIL_HW_CURSOR_SUPPORT) {
+ /* attempt to validate again with subvp disabled due to cursor */
+ if (dc->debug.using_dml2)
+ status = dml2_validate(dc, context,
+ context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2,
+ validate_mode) ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
+ else
+ status = dml1_validate(dc, context, validate_mode) ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
+ }
+
+ return status;
}
int dcn32_populate_dml_pipes_from_context(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
int i, pipe_cnt;
struct resource_context *res_ctx = &context->res_ctx;
@@ -1839,7 +1878,7 @@ int dcn32_populate_dml_pipes_from_context(
int num_subvp_none = 0;
int odm_slice_count;
- dcn20_populate_dml_pipes_from_context(dc, context, pipes, fast_validate);
+ dcn20_populate_dml_pipes_from_context(dc, context, pipes, validate_mode);
/* For single display subvp, look for subvp main so if we have phantom
* pipe, we can set odm policy to match main pipe
@@ -1921,7 +1960,7 @@ int dcn32_populate_dml_pipes_from_context(
/* Only populate DML input with subvp info for full updates.
* This is just a workaround -- needs a proper fix.
*/
- if (!fast_validate) {
+ if (validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING) {
switch (dc_state_get_pipe_subvp_type(context, pipe)) {
case SUBVP_MAIN:
pipes[pipe_cnt].pipe.src.use_mall_for_pstate_change = dm_use_mall_pstate_change_sub_viewport;
@@ -2022,25 +2061,31 @@ void dcn32_calculate_wm_and_dlg(struct dc *dc, struct dc_state *context,
static void dcn32_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params)
{
- struct dml2_configuration_options *dml2_opt = &dc->dml2_tmp;
-
- memcpy(dml2_opt, &dc->dml2_options, sizeof(dc->dml2_options));
-
DC_FP_START();
dcn32_update_bw_bounding_box_fpu(dc, bw_params);
- dml2_opt->use_clock_dc_limits = false;
if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2)
- dml2_reinit(dc, dml2_opt, &dc->current_state->bw_ctx.dml2);
+ dml2_reinit(dc, &dc->dml2_options, &dc->current_state->bw_ctx.dml2);
- dml2_opt->use_clock_dc_limits = true;
if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2_dc_power_source)
- dml2_reinit(dc, dml2_opt, &dc->current_state->bw_ctx.dml2_dc_power_source);
+ dml2_reinit(dc, &dc->dml2_dc_power_options, &dc->current_state->bw_ctx.dml2_dc_power_source);
DC_FP_END();
}
+unsigned int dcn32_get_max_hw_cursor_size(const struct dc *dc,
+ struct dc_state *state,
+ const struct dc_stream_state *stream)
+{
+ bool limit_cur_to_buf;
+
+ limit_cur_to_buf = dc_state_get_stream_subvp_cursor_limit(stream, state) &&
+ !stream->hw_cursor_req;
+
+ return limit_cur_to_buf ? dc->caps.max_buffered_cursor_size : dc->caps.max_cursor_size;
+}
+
static struct resource_funcs dcn32_res_pool_funcs = {
.destroy = dcn32_destroy_resource_pool,
.link_enc_create = dcn32_link_encoder_create,
@@ -2066,6 +2111,8 @@ static struct resource_funcs dcn32_res_pool_funcs = {
.add_phantom_pipes = dcn32_add_phantom_pipes,
.build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params,
.calculate_mall_ways_from_bytes = dcn32_calculate_mall_ways_from_bytes,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe,
+ .get_max_hw_cursor_size = dcn32_get_max_hw_cursor_size,
};
static uint32_t read_pipe_fuses(struct dc_context *ctx)
@@ -2112,8 +2159,6 @@ static bool dcn32_resource_construct(
#define REG_STRUCT dccg_regs
dccg_regs_init();
- DC_FP_START();
-
ctx->dc_bios->regs = &bios_regs;
pool->base.res_cap = &res_cap_dcn32;
@@ -2151,6 +2196,7 @@ static bool dcn32_resource_construct(
dc->caps.i2c_speed_in_khz_hdcp = 100; /*1.4 w/a applied by default*/
/* TODO: Bring max_cursor_size back to 256 after subvp cursor corruption is fixed*/
dc->caps.max_cursor_size = 64;
+ dc->caps.max_buffered_cursor_size = 64; // sqrt(16 * 1024 / 4)
dc->caps.min_horizontal_blanking_period = 80;
dc->caps.dmdata_alloc_size = 2048;
dc->caps.mall_size_per_mem_channel = 4;
@@ -2189,6 +2235,7 @@ static bool dcn32_resource_construct(
dc->caps.dmcub_support = true;
dc->caps.seamless_odm = true;
dc->caps.max_v_total = (1 << 15) - 1;
+ dc->caps.vtotal_limited_by_fp2 = true;
/* Color pipeline capabilities */
dc->caps.color.dpp.dcn_arch = 1;
@@ -2204,7 +2251,7 @@ static bool dcn32_resource_construct(
dc->caps.color.dpp.gamma_corr = 1;
dc->caps.color.dpp.dgam_rom_for_yuv = 0;
- dc->caps.color.dpp.hw_3d_lut = 1;
+ dc->caps.color.dpp.hw_3d_lut = 0;
dc->caps.color.dpp.ogam_ram = 0; // no OGAM in DPP since DCN1
// no OGAM ROM on DCN2 and later ASICs
dc->caps.color.dpp.ogam_rom_caps.srgb = 0;
@@ -2223,6 +2270,7 @@ static bool dcn32_resource_construct(
dc->caps.color.mpc.ogam_rom_caps.pq = 0;
dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
dc->caps.color.mpc.ocsc = 1;
+ dc->caps.color.mpc.preblend = true;
/* Use pipe context based otg sync logic */
dc->config.use_pipe_ctx_sync_logic = true;
@@ -2452,6 +2500,8 @@ static bool dcn32_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
if (dc->ctx->dc_bios->fw_info.oem_i2c_present) {
@@ -2466,7 +2516,6 @@ static bool dcn32_resource_construct(
}
dc->dml2_options.dcn_pipe_count = pool->base.pipe_count;
- dc->dml2_options.use_native_pstate_optimization = false;
dc->dml2_options.use_native_soc_bb_construction = true;
dc->dml2_options.minimize_dispclk_using_odm = true;
@@ -2498,14 +2547,14 @@ static bool dcn32_resource_construct(
if (ASICREV_IS_GC_11_0_3(dc->ctx->asic_id.hw_internal_rev) && (dc->config.sdpif_request_limit_words_per_umc == 0))
dc->config.sdpif_request_limit_words_per_umc = 16;
- DC_FP_END();
+ /* init DC limited DML2 options */
+ memcpy(&dc->dml2_dc_power_options, &dc->dml2_options, sizeof(struct dml2_configuration_options));
+ dc->dml2_dc_power_options.use_clock_dc_limits = true;
return true;
create_fail:
- DC_FP_END();
-
dcn32_resource_destruct(pool);
return false;
@@ -2803,6 +2852,7 @@ struct pipe_ctx *dcn32_acquire_free_pipe_as_secondary_opp_head(
free_pipe->plane_res.xfm = pool->transforms[free_pipe_idx];
free_pipe->plane_res.dpp = pool->dpps[free_pipe_idx];
free_pipe->plane_res.mpcc_inst = pool->dpps[free_pipe_idx]->inst;
+ free_pipe->hblank_borrow = otg_master->hblank_borrow;
if (free_pipe->stream->timing.flags.DSC == 1) {
dcn20_acquire_dsc(free_pipe->stream->ctx->dc,
&new_ctx->res_ctx,
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h
index 86c6e5e8c42e..82f966cf4ed2 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn32/dcn32_resource.h
@@ -98,14 +98,14 @@ void dcn32_add_phantom_pipes(struct dc *dc,
unsigned int pipe_cnt,
unsigned int index);
-bool dcn32_validate_bandwidth(struct dc *dc,
+enum dc_status dcn32_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
int dcn32_populate_dml_pipes_from_context(
struct dc *dc, struct dc_state *context,
display_e2e_pipe_params_st *pipes,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void dcn32_calculate_wm_and_dlg(
struct dc *dc, struct dc_state *context,
@@ -188,6 +188,10 @@ void dcn32_override_min_req_dcfclk(struct dc *dc, struct dc_state *context);
unsigned int dcn32_calculate_mall_ways_from_bytes(const struct dc *dc, unsigned int total_size_in_mall_bytes);
+unsigned int dcn32_get_max_hw_cursor_size(const struct dc *dc,
+ struct dc_state *state,
+ const struct dc_stream_state *stream);
+
/* definitions for run time init of reg offsets */
/* CLK SRC */
@@ -1055,7 +1059,8 @@ unsigned int dcn32_calculate_mall_ways_from_bytes(const struct dc *dc, unsigned
SRI_ARR(OPTC_WIDTH_CONTROL, ODM, inst), \
SRI_ARR(OPTC_MEMORY_CONFIG, ODM, inst), \
SRI_ARR(OTG_DRR_CONTROL, OTG, inst), \
- SRI_ARR(OTG_PIPE_UPDATE_STATUS, OTG, inst)
+ SRI_ARR(OTG_PIPE_UPDATE_STATUS, OTG, inst), \
+ SRI_ARR(INTERRUPT_DEST, OTG, inst)
/* HUBP */
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c
index 5cb74fd9cb7d..061c0907d802 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn321/dcn321_resource.c
@@ -1580,21 +1580,15 @@ static struct dc_cap_funcs cap_funcs = {
static void dcn321_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params)
{
- struct dml2_configuration_options *dml2_opt = &dc->dml2_tmp;
-
- memcpy(dml2_opt, &dc->dml2_options, sizeof(dc->dml2_options));
-
DC_FP_START();
dcn321_update_bw_bounding_box_fpu(dc, bw_params);
- dml2_opt->use_clock_dc_limits = false;
if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2)
- dml2_reinit(dc, dml2_opt, &dc->current_state->bw_ctx.dml2);
+ dml2_reinit(dc, &dc->dml2_options, &dc->current_state->bw_ctx.dml2);
- dml2_opt->use_clock_dc_limits = true;
if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2_dc_power_source)
- dml2_reinit(dc, dml2_opt, &dc->current_state->bw_ctx.dml2_dc_power_source);
+ dml2_reinit(dc, &dc->dml2_dc_power_options, &dc->current_state->bw_ctx.dml2_dc_power_source);
DC_FP_END();
}
@@ -1624,6 +1618,8 @@ static struct resource_funcs dcn321_res_pool_funcs = {
.add_phantom_pipes = dcn32_add_phantom_pipes,
.build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params,
.calculate_mall_ways_from_bytes = dcn32_calculate_mall_ways_from_bytes,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe,
+ .get_max_hw_cursor_size = dcn32_get_max_hw_cursor_size,
};
static uint32_t read_pipe_fuses(struct dc_context *ctx)
@@ -1708,6 +1704,7 @@ static bool dcn321_resource_construct(
dc->caps.i2c_speed_in_khz_hdcp = 100; /*1.4 w/a applied by default*/
/* TODO: Bring max cursor size back to 256 after subvp cursor corruption is fixed*/
dc->caps.max_cursor_size = 64;
+ dc->caps.max_buffered_cursor_size = 64; // sqrt(16 * 1024 / 4)
dc->caps.min_horizontal_blanking_period = 80;
dc->caps.dmdata_alloc_size = 2048;
dc->caps.mall_size_per_mem_channel = 4;
@@ -1742,6 +1739,7 @@ static bool dcn321_resource_construct(
dc->caps.extended_aux_timeout_support = true;
dc->caps.dmcub_support = true;
dc->caps.max_v_total = (1 << 15) - 1;
+ dc->caps.vtotal_limited_by_fp2 = true;
/* Color pipeline capabilities */
dc->caps.color.dpp.dcn_arch = 1;
@@ -1757,8 +1755,8 @@ static bool dcn321_resource_construct(
dc->caps.color.dpp.gamma_corr = 1;
dc->caps.color.dpp.dgam_rom_for_yuv = 0;
- dc->caps.color.dpp.hw_3d_lut = 1;
- dc->caps.color.dpp.ogam_ram = 1;
+ dc->caps.color.dpp.hw_3d_lut = 0;
+ dc->caps.color.dpp.ogam_ram = 0;
// no OGAM ROM on DCN2 and later ASICs
dc->caps.color.dpp.ogam_rom_caps.srgb = 0;
dc->caps.color.dpp.ogam_rom_caps.bt2020 = 0;
@@ -1776,6 +1774,7 @@ static bool dcn321_resource_construct(
dc->caps.color.mpc.ogam_rom_caps.pq = 0;
dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
dc->caps.color.mpc.ocsc = 1;
+ dc->caps.color.mpc.preblend = true;
/* Use pipe context based otg sync logic */
dc->config.use_pipe_ctx_sync_logic = true;
@@ -2000,6 +1999,8 @@ static bool dcn321_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
if (dc->ctx->dc_bios->fw_info.oem_i2c_present) {
@@ -2014,7 +2015,6 @@ static bool dcn321_resource_construct(
}
dc->dml2_options.dcn_pipe_count = pool->base.pipe_count;
- dc->dml2_options.use_native_pstate_optimization = false;
dc->dml2_options.use_native_soc_bb_construction = true;
dc->dml2_options.minimize_dispclk_using_odm = true;
@@ -2042,6 +2042,10 @@ static bool dcn321_resource_construct(
dc->dml2_options.max_segments_per_hubp = 18;
dc->dml2_options.det_segment_size = DCN3_2_DET_SEG_SIZE;
+ /* init DC limited DML2 options */
+ memcpy(&dc->dml2_dc_power_options, &dc->dml2_options, sizeof(struct dml2_configuration_options));
+ dc->dml2_dc_power_options.use_clock_dc_limits = true;
+
return true;
create_fail:
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c
index 6cc2960b6104..8475c6eec547 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.c
@@ -1732,24 +1732,31 @@ static void dcn35_get_panel_config_defaults(struct dc_panel_config *panel_config
}
-static bool dcn35_validate_bandwidth(struct dc *dc,
+static enum dc_status dcn35_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool out = false;
out = dml2_validate(dc, context,
context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2,
- fast_validate);
+ validate_mode);
- if (fast_validate)
- return out;
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING)
+ return out ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
DC_FP_START();
dcn35_decide_zstate_support(dc, context);
DC_FP_END();
- return out;
+ return out ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
+}
+
+enum dc_status dcn35_patch_unknown_plane_state(struct dc_plane_state *plane_state)
+{
+ plane_state->tiling_info.gfxversion = DcGfxVersion9;
+ dcn20_patch_unknown_plane_state(plane_state);
+ return DC_OK;
}
@@ -1775,9 +1782,13 @@ static struct resource_funcs dcn35_res_pool_funcs = {
.acquire_post_bldn_3dlut = dcn30_acquire_post_bldn_3dlut,
.release_post_bldn_3dlut = dcn30_release_post_bldn_3dlut,
.update_bw_bounding_box = dcn35_update_bw_bounding_box_fpu,
- .patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
+ .patch_unknown_plane_state = dcn35_patch_unknown_plane_state,
.get_panel_config_defaults = dcn35_get_panel_config_defaults,
.get_preferred_eng_id_dpia = dcn35_get_preferred_eng_id_dpia,
+ .get_det_buffer_size = dcn31_get_det_buffer_size,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe,
+ .update_dc_state_for_encoder_switch = dcn31_update_dc_state_for_encoder_switch,
+ .build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params
};
static bool dcn35_resource_construct(
@@ -1830,9 +1841,9 @@ static bool dcn35_resource_construct(
dc->caps.max_cursor_size = 256;
dc->caps.min_horizontal_blanking_period = 80;
dc->caps.dmdata_alloc_size = 2048;
- dc->caps.max_slave_planes = 2;
- dc->caps.max_slave_yuv_planes = 2;
- dc->caps.max_slave_rgb_planes = 2;
+ dc->caps.max_slave_planes = 3;
+ dc->caps.max_slave_yuv_planes = 3;
+ dc->caps.max_slave_rgb_planes = 3;
dc->caps.post_blend_color_processing = true;
dc->caps.force_dp_tps4_for_cp2520 = true;
if (dc->config.forceHBR2CP2520)
@@ -1849,6 +1860,7 @@ static bool dcn35_resource_construct(
dc->caps.zstate_support = true;
dc->caps.ips_support = true;
dc->caps.max_v_total = (1 << 15) - 1;
+ dc->caps.vtotal_limited_by_fp2 = true;
/* Color pipeline capabilities */
dc->caps.color.dpp.dcn_arch = 1;
@@ -1864,7 +1876,7 @@ static bool dcn35_resource_construct(
dc->caps.color.dpp.gamma_corr = 1;
dc->caps.color.dpp.dgam_rom_for_yuv = 0;
- dc->caps.color.dpp.hw_3d_lut = 1;
+ dc->caps.color.dpp.hw_3d_lut = 0;
dc->caps.color.dpp.ogam_ram = 0; // no OGAM in DPP since DCN1
// no OGAM ROM on DCN301
dc->caps.color.dpp.ogam_rom_caps.srgb = 0;
@@ -1883,6 +1895,13 @@ static bool dcn35_resource_construct(
dc->caps.color.mpc.ogam_rom_caps.pq = 0;
dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
dc->caps.color.mpc.ocsc = 1;
+ dc->caps.color.mpc.preblend = true;
+
+ dc->caps.num_of_host_routers = 2;
+ dc->caps.num_of_dpias_per_host_router = 2;
+
+ dc->caps.num_of_host_routers = 2;
+ dc->caps.num_of_dpias_per_host_router = 2;
/* max_disp_clock_khz_at_vmin is slightly lower than the STA value in order
* to provide some margin.
@@ -1893,12 +1912,13 @@ static bool dcn35_resource_construct(
dc->caps.max_disp_clock_khz_at_vmin = 650000;
/* Sequential ONO is based on ASIC. */
- if (dc->ctx->asic_id.hw_internal_rev > 0x10)
+ if (dc->ctx->asic_id.hw_internal_rev >= 0x40)
dc->caps.sequential_ono = true;
/* Use pipe context based otg sync logic */
dc->config.use_pipe_ctx_sync_logic = true;
+
dc->config.disable_hbr_audio_dp2 = true;
/* read VBIOS LTTPR caps */
{
@@ -2137,12 +2157,13 @@ static bool dcn35_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
dc->dcn_ip->max_num_dpp = pool->base.pipe_count;
dc->dml2_options.dcn_pipe_count = pool->base.pipe_count;
- dc->dml2_options.use_native_pstate_optimization = true;
dc->dml2_options.use_native_soc_bb_construction = true;
dc->dml2_options.minimize_dispclk_using_odm = false;
if (dc->config.EnableMinDispClkODM)
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.h
index f97bb4cb3761..9c56ae76e0c7 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn35/dcn35_resource.h
@@ -35,6 +35,7 @@
extern struct _vcs_dpi_ip_params_st dcn3_5_ip;
extern struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc;
+enum dc_status dcn35_patch_unknown_plane_state(struct dc_plane_state *plane_state);
struct dcn35_resource_pool {
struct resource_pool base;
@@ -304,7 +305,8 @@ struct resource_pool *dcn35_create_resource_pool(
SRI_ARR(OPTC_WIDTH_CONTROL, ODM, inst),\
SRI_ARR(OPTC_MEMORY_CONFIG, ODM, inst),\
SRI_ARR(OTG_DRR_CONTROL, OTG, inst),\
- SRI2_ARR(OPTC_CLOCK_CONTROL, OPTC, inst)
+ SRI2_ARR(OPTC_CLOCK_CONTROL, OPTC, inst),\
+ SRI_ARR(INTERRUPT_DEST, OTG, inst)
/* DPP */
#define DPP_REG_LIST_DCN35_RI(id)\
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c
index d87e2641cda1..0971c0f74186 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn351/dcn351_resource.c
@@ -1712,24 +1712,24 @@ static void dcn35_get_panel_config_defaults(struct dc_panel_config *panel_config
}
-static bool dcn351_validate_bandwidth(struct dc *dc,
+static enum dc_status dcn351_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
bool out = false;
out = dml2_validate(dc, context,
context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2,
- fast_validate);
+ validate_mode);
- if (fast_validate)
- return out;
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING)
+ return out ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
DC_FP_START();
dcn35_decide_zstate_support(dc, context);
DC_FP_END();
- return out;
+ return out ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
}
static struct resource_funcs dcn351_res_pool_funcs = {
@@ -1754,9 +1754,13 @@ static struct resource_funcs dcn351_res_pool_funcs = {
.acquire_post_bldn_3dlut = dcn30_acquire_post_bldn_3dlut,
.release_post_bldn_3dlut = dcn30_release_post_bldn_3dlut,
.update_bw_bounding_box = dcn351_update_bw_bounding_box_fpu,
- .patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
+ .patch_unknown_plane_state = dcn35_patch_unknown_plane_state,
.get_panel_config_defaults = dcn35_get_panel_config_defaults,
.get_preferred_eng_id_dpia = dcn351_get_preferred_eng_id_dpia,
+ .get_det_buffer_size = dcn31_get_det_buffer_size,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe,
+ .update_dc_state_for_encoder_switch = dcn31_update_dc_state_for_encoder_switch,
+ .build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params
};
static bool dcn351_resource_construct(
@@ -1809,9 +1813,9 @@ static bool dcn351_resource_construct(
dc->caps.max_cursor_size = 256;
dc->caps.min_horizontal_blanking_period = 80;
dc->caps.dmdata_alloc_size = 2048;
- dc->caps.max_slave_planes = 2;
- dc->caps.max_slave_yuv_planes = 2;
- dc->caps.max_slave_rgb_planes = 2;
+ dc->caps.max_slave_planes = 3;
+ dc->caps.max_slave_yuv_planes = 3;
+ dc->caps.max_slave_rgb_planes = 3;
dc->caps.post_blend_color_processing = true;
dc->caps.force_dp_tps4_for_cp2520 = true;
if (dc->config.forceHBR2CP2520)
@@ -1828,6 +1832,7 @@ static bool dcn351_resource_construct(
dc->caps.zstate_support = true;
dc->caps.ips_support = true;
dc->caps.max_v_total = (1 << 15) - 1;
+ dc->caps.vtotal_limited_by_fp2 = true;
/* Color pipeline capabilities */
dc->caps.color.dpp.dcn_arch = 1;
@@ -1843,7 +1848,7 @@ static bool dcn351_resource_construct(
dc->caps.color.dpp.gamma_corr = 1;
dc->caps.color.dpp.dgam_rom_for_yuv = 0;
- dc->caps.color.dpp.hw_3d_lut = 1;
+ dc->caps.color.dpp.hw_3d_lut = 0;
dc->caps.color.dpp.ogam_ram = 0; // no OGAM in DPP since DCN1
// no OGAM ROM on DCN301
dc->caps.color.dpp.ogam_rom_caps.srgb = 0;
@@ -1862,6 +1867,13 @@ static bool dcn351_resource_construct(
dc->caps.color.mpc.ogam_rom_caps.pq = 0;
dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
dc->caps.color.mpc.ocsc = 1;
+ dc->caps.color.mpc.preblend = true;
+
+ dc->caps.num_of_host_routers = 2;
+ dc->caps.num_of_dpias_per_host_router = 2;
+
+ dc->caps.num_of_host_routers = 2;
+ dc->caps.num_of_dpias_per_host_router = 2;
/* max_disp_clock_khz_at_vmin is slightly lower than the STA value in order
* to provide some margin.
@@ -1874,6 +1886,7 @@ static bool dcn351_resource_construct(
/* Use pipe context based otg sync logic */
dc->config.use_pipe_ctx_sync_logic = true;
+
/* Use psp mailbox to enable assr */
dc->config.use_assr_psp_message = true;
@@ -2115,13 +2128,14 @@ static bool dcn351_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
dc->dcn_ip->max_num_dpp = pool->base.pipe_count;
dc->dml2_options.dcn_pipe_count = pool->base.pipe_count;
- dc->dml2_options.use_native_pstate_optimization = true;
dc->dml2_options.use_native_soc_bb_construction = true;
dc->dml2_options.minimize_dispclk_using_odm = false;
if (dc->config.EnableMinDispClkODM)
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn36/dcn36_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn36/dcn36_resource.c
new file mode 100644
index 000000000000..8bae7fcedc22
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn36/dcn36_resource.c
@@ -0,0 +1,2181 @@
+/* SPDX-License-Identifier: MIT */
+/* Copyright 2025 Advanced Micro Devices, Inc. */
+
+#include "dm_services.h"
+#include "dc.h"
+
+#include "dcn31/dcn31_init.h"
+#include "dcn35/dcn35_init.h"
+#include "dcn36/dcn36_resource.h"
+
+#include "resource.h"
+#include "include/irq_service_interface.h"
+#include "dcn36_resource.h"
+#include "dml2/dml2_wrapper.h"
+
+#include "dcn20/dcn20_resource.h"
+#include "dcn30/dcn30_resource.h"
+#include "dcn31/dcn31_resource.h"
+#include "dcn32/dcn32_resource.h"
+#include "dcn35/dcn35_resource.h"
+
+#include "dcn10/dcn10_ipp.h"
+#include "dcn30/dcn30_hubbub.h"
+#include "dcn31/dcn31_hubbub.h"
+#include "dcn35/dcn35_hubbub.h"
+#include "dcn32/dcn32_mpc.h"
+#include "dcn35/dcn35_hubp.h"
+#include "irq/dcn36/irq_service_dcn36.h"
+#include "dcn35/dcn35_dpp.h"
+#include "dcn35/dcn35_optc.h"
+#include "dcn20/dcn20_hwseq.h"
+#include "dcn30/dcn30_hwseq.h"
+#include "dce110/dce110_hwseq.h"
+#include "dcn35/dcn35_opp.h"
+#include "dcn35/dcn35_dsc.h"
+#include "dcn30/dcn30_vpg.h"
+#include "dcn30/dcn30_afmt.h"
+#include "dcn31/dcn31_dio_link_encoder.h"
+#include "dcn35/dcn35_dio_stream_encoder.h"
+#include "dcn31/dcn31_hpo_dp_stream_encoder.h"
+#include "dcn31/dcn31_hpo_dp_link_encoder.h"
+#include "dcn32/dcn32_hpo_dp_link_encoder.h"
+#include "link.h"
+#include "dcn31/dcn31_apg.h"
+#include "dcn32/dcn32_dio_link_encoder.h"
+#include "dcn31/dcn31_vpg.h"
+#include "dcn31/dcn31_afmt.h"
+#include "dce/dce_clock_source.h"
+#include "dce/dce_audio.h"
+#include "dce/dce_hwseq.h"
+#include "clk_mgr.h"
+#include "virtual/virtual_stream_encoder.h"
+#include "dce110/dce110_resource.h"
+#include "dml/display_mode_vba.h"
+#include "dcn35/dcn35_dccg.h"
+#include "dcn35/dcn35_pg_cntl.h"
+#include "dcn10/dcn10_resource.h"
+#include "dcn31/dcn31_panel_cntl.h"
+#include "dcn35/dcn35_hwseq.h"
+#include "dcn35/dcn35_dio_link_encoder.h"
+#include "dml/dcn31/dcn31_fpu.h" /*todo*/
+#include "dml/dcn35/dcn35_fpu.h"
+#include "dcn35/dcn35_dwb.h"
+#include "dcn35/dcn35_mmhubbub.h"
+
+#include "dcn/dcn_3_6_0_offset.h"
+#include "dcn/dcn_3_6_0_sh_mask.h"
+
+#define regBIF_BX2_BIOS_SCRATCH_2 0x2ffc004e
+#define regBIF_BX2_BIOS_SCRATCH_2_BASE_IDX 5
+
+#define regBIF_BX2_BIOS_SCRATCH_3 0x2ffc004f
+#define regBIF_BX2_BIOS_SCRATCH_3_BASE_IDX 5
+
+#define regBIF_BX2_BIOS_SCRATCH_6 0x2ffc0052
+#define regBIF_BX2_BIOS_SCRATCH_6_BASE_IDX 5
+
+#define DSCC0_DSCC_CONFIG0__ICH_RESET_AT_END_OF_LINE__SHIFT 0x0
+#define DSCC0_DSCC_CONFIG0__ICH_RESET_AT_END_OF_LINE_MASK 0x0000000FL
+
+#include "reg_helper.h"
+#include "dce/dmub_abm.h"
+#include "dce/dmub_psr.h"
+#include "dce/dmub_replay.h"
+#include "dce/dce_aux.h"
+#include "dce/dce_i2c.h"
+#include "dml/dcn31/display_mode_vba_31.h" /*temp*/
+#include "vm_helper.h"
+#include "dcn20/dcn20_vmid.h"
+
+#include "dc_state_priv.h"
+
+#include "link_enc_cfg.h"
+#define DC_LOGGER_INIT(logger)
+
+enum dcn36_clk_src_array_id {
+ DCN36_CLK_SRC_PLL0,
+ DCN36_CLK_SRC_PLL1,
+ DCN36_CLK_SRC_PLL2,
+ DCN36_CLK_SRC_PLL3,
+ DCN36_CLK_SRC_PLL4,
+ DCN36_CLK_SRC_TOTAL
+};
+
+/* begin *********************
+ * macros to expend register list macro defined in HW object header file
+ */
+
+/* DCN */
+/* TODO awful hack. fixup dcn20_dwb.h */
+#undef BASE_INNER
+#define BASE_INNER(seg) ctx->dcn_reg_offsets[seg]
+
+#define BASE(seg) BASE_INNER(seg)
+
+#define SR(reg_name)\
+ REG_STRUCT.reg_name = BASE(reg ## reg_name ## _BASE_IDX) + \
+ reg ## reg_name
+
+#define SR_ARR(reg_name, id) \
+ REG_STRUCT[id].reg_name = BASE(reg##reg_name##_BASE_IDX) + reg##reg_name
+
+#define SR_ARR_INIT(reg_name, id, value) \
+ REG_STRUCT[id].reg_name = value
+
+#define SRI(reg_name, block, id)\
+ REG_STRUCT.reg_name = BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define SRI_ARR(reg_name, block, id)\
+ REG_STRUCT[id].reg_name = BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define SR_ARR_I2C(reg_name, id) \
+ REG_STRUCT[id-1].reg_name = BASE(reg##reg_name##_BASE_IDX) + reg##reg_name
+
+#define SRI_ARR_I2C(reg_name, block, id)\
+ REG_STRUCT[id-1].reg_name = BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define SRI_ARR_ALPHABET(reg_name, block, index, id)\
+ REG_STRUCT[index].reg_name = BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define SRI2(reg_name, block, id)\
+ .reg_name = BASE(reg ## reg_name ## _BASE_IDX) + \
+ reg ## reg_name
+
+#define SRI2_ARR(reg_name, block, id)\
+ REG_STRUCT[id].reg_name = BASE(reg ## reg_name ## _BASE_IDX) + \
+ reg ## reg_name
+
+#define SRIR(var_name, reg_name, block, id)\
+ .var_name = BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define SRII(reg_name, block, id)\
+ REG_STRUCT.reg_name[id] = BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define SRII_ARR_2(reg_name, block, id, inst)\
+ REG_STRUCT[inst].reg_name[id] = BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define SRII_MPC_RMU(reg_name, block, id)\
+ .RMU##_##reg_name[id] = BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define SRII_DWB(reg_name, temp_name, block, id)\
+ REG_STRUCT.reg_name[id] = BASE(reg ## block ## id ## _ ## temp_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## temp_name
+
+#define SF_DWB2(reg_name, block, id, field_name, post_fix) \
+ .field_name = reg_name ## __ ## field_name ## post_fix
+
+#define DCCG_SRII(reg_name, block, id)\
+ REG_STRUCT.block ## _ ## reg_name[id] = BASE(reg ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
+ reg ## block ## id ## _ ## reg_name
+
+#define VUPDATE_SRII(reg_name, block, id)\
+ REG_STRUCT.reg_name[id] = BASE(reg ## reg_name ## _ ## block ## id ## _BASE_IDX) + \
+ reg ## reg_name ## _ ## block ## id
+
+/* NBIO */
+#define NBIO_BASE_INNER(seg) ctx->nbio_reg_offsets[seg]
+
+#define NBIO_BASE(seg) \
+ NBIO_BASE_INNER(seg)
+
+#define NBIO_SR(reg_name)\
+ REG_STRUCT.reg_name = NBIO_BASE(regBIF_BX2_ ## reg_name ## _BASE_IDX) + \
+ regBIF_BX2_ ## reg_name
+
+#define NBIO_SR_ARR(reg_name, id)\
+ REG_STRUCT[id].reg_name = NBIO_BASE(regBIF_BX2_ ## reg_name ## _BASE_IDX) + \
+ regBIF_BX2_ ## reg_name
+
+#define bios_regs_init() \
+ ( \
+ NBIO_SR(BIOS_SCRATCH_3),\
+ NBIO_SR(BIOS_SCRATCH_6)\
+ )
+
+static struct bios_registers bios_regs;
+
+#define clk_src_regs_init(index, pllid)\
+ CS_COMMON_REG_LIST_DCN3_0_RI(index, pllid)
+
+static struct dce110_clk_src_regs clk_src_regs[5];
+
+static const struct dce110_clk_src_shift cs_shift = {
+ CS_COMMON_MASK_SH_LIST_DCN3_1_4(__SHIFT)
+};
+
+static const struct dce110_clk_src_mask cs_mask = {
+ CS_COMMON_MASK_SH_LIST_DCN3_1_4(_MASK)
+};
+
+#define abm_regs_init(id)\
+ ABM_DCN32_REG_LIST_RI(id)
+
+static struct dce_abm_registers abm_regs[4];
+
+static const struct dce_abm_shift abm_shift = {
+ ABM_MASK_SH_LIST_DCN35(__SHIFT)
+};
+
+static const struct dce_abm_mask abm_mask = {
+ ABM_MASK_SH_LIST_DCN35(_MASK)
+};
+
+#define audio_regs_init(id)\
+ AUD_COMMON_REG_LIST_RI(id)
+
+static struct dce_audio_registers audio_regs[7];
+
+
+#define DCE120_AUD_COMMON_MASK_SH_LIST(mask_sh)\
+ SF(AZF0ENDPOINT0_AZALIA_F0_CODEC_ENDPOINT_INDEX, AZALIA_ENDPOINT_REG_INDEX, mask_sh),\
+ SF(AZF0ENDPOINT0_AZALIA_F0_CODEC_ENDPOINT_DATA, AZALIA_ENDPOINT_REG_DATA, mask_sh),\
+ AUD_COMMON_MASK_SH_LIST_BASE(mask_sh)
+
+static const struct dce_audio_shift audio_shift = {
+ DCE120_AUD_COMMON_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce_audio_mask audio_mask = {
+ DCE120_AUD_COMMON_MASK_SH_LIST(_MASK)
+};
+
+#define vpg_regs_init(id)\
+ VPG_DCN31_REG_LIST_RI(id)
+
+static struct dcn31_vpg_registers vpg_regs[10];
+
+static const struct dcn31_vpg_shift vpg_shift = {
+ DCN31_VPG_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dcn31_vpg_mask vpg_mask = {
+ DCN31_VPG_MASK_SH_LIST(_MASK)
+};
+
+#define afmt_regs_init(id)\
+ AFMT_DCN31_REG_LIST_RI(id)
+
+static struct dcn31_afmt_registers afmt_regs[6];
+
+static const struct dcn31_afmt_shift afmt_shift = {
+ DCN31_AFMT_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dcn31_afmt_mask afmt_mask = {
+ DCN31_AFMT_MASK_SH_LIST(_MASK)
+};
+
+#define apg_regs_init(id)\
+ APG_DCN31_REG_LIST_RI(id)
+
+static struct dcn31_apg_registers apg_regs[4];
+
+static const struct dcn31_apg_shift apg_shift = {
+ DCN31_APG_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dcn31_apg_mask apg_mask = {
+ DCN31_APG_MASK_SH_LIST(_MASK)
+};
+
+#define stream_enc_regs_init(id)\
+ SE_DCN35_REG_LIST_RI(id)
+
+static struct dcn10_stream_enc_registers stream_enc_regs[5];
+
+static const struct dcn10_stream_encoder_shift se_shift = {
+ SE_COMMON_MASK_SH_LIST_DCN35(__SHIFT)
+};
+
+static const struct dcn10_stream_encoder_mask se_mask = {
+ SE_COMMON_MASK_SH_LIST_DCN35(_MASK)
+};
+
+#define aux_regs_init(id)\
+ DCN2_AUX_REG_LIST_RI(id)
+
+static struct dcn10_link_enc_aux_registers link_enc_aux_regs[5];
+
+#define hpd_regs_init(id)\
+ HPD_REG_LIST_RI(id)
+
+static struct dcn10_link_enc_hpd_registers link_enc_hpd_regs[5];
+
+
+static const struct dce110_aux_registers_shift aux_shift = {
+ DCN_AUX_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce110_aux_registers_mask aux_mask = {
+ DCN_AUX_MASK_SH_LIST(_MASK)
+};
+
+#define link_regs_init(id, phyid)\
+ ( \
+ LE_DCN35_REG_LIST_RI(id), \
+ UNIPHY_DCN2_REG_LIST_RI(id, phyid)\
+ )
+
+static struct dcn10_link_enc_registers link_enc_regs[5];
+
+static const struct dcn10_link_enc_shift le_shift = {
+ LINK_ENCODER_MASK_SH_LIST_DCN35(__SHIFT), \
+ //DPCS_DCN31_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dcn10_link_enc_mask le_mask = {
+ LINK_ENCODER_MASK_SH_LIST_DCN35(_MASK), \
+ //DPCS_DCN31_MASK_SH_LIST(_MASK)
+};
+
+#define hpo_dp_stream_encoder_reg_init(id)\
+ DCN3_1_HPO_DP_STREAM_ENC_REG_LIST_RI(id)
+
+static struct dcn31_hpo_dp_stream_encoder_registers hpo_dp_stream_enc_regs[4];
+
+static const struct dcn31_hpo_dp_stream_encoder_shift hpo_dp_se_shift = {
+ DCN3_1_HPO_DP_STREAM_ENC_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dcn31_hpo_dp_stream_encoder_mask hpo_dp_se_mask = {
+ DCN3_1_HPO_DP_STREAM_ENC_MASK_SH_LIST(_MASK)
+};
+
+#define hpo_dp_link_encoder_reg_init(id)\
+ DCN3_1_HPO_DP_LINK_ENC_REG_LIST_RI(id)
+
+static struct dcn31_hpo_dp_link_encoder_registers hpo_dp_link_enc_regs[2];
+
+static const struct dcn31_hpo_dp_link_encoder_shift hpo_dp_le_shift = {
+ DCN3_1_HPO_DP_LINK_ENC_COMMON_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dcn31_hpo_dp_link_encoder_mask hpo_dp_le_mask = {
+ DCN3_1_HPO_DP_LINK_ENC_COMMON_MASK_SH_LIST(_MASK)
+};
+
+#define dpp_regs_init(id)\
+ DPP_REG_LIST_DCN35_RI(id)
+
+static struct dcn3_dpp_registers dpp_regs[4];
+
+static const struct dcn35_dpp_shift tf_shift = {
+ DPP_REG_LIST_SH_MASK_DCN35(__SHIFT)
+};
+
+static const struct dcn35_dpp_mask tf_mask = {
+ DPP_REG_LIST_SH_MASK_DCN35(_MASK)
+};
+
+#define opp_regs_init(id)\
+ OPP_REG_LIST_DCN35_RI(id)
+
+static struct dcn35_opp_registers opp_regs[4];
+
+static const struct dcn35_opp_shift opp_shift = {
+ OPP_MASK_SH_LIST_DCN35(__SHIFT)
+};
+
+static const struct dcn35_opp_mask opp_mask = {
+ OPP_MASK_SH_LIST_DCN35(_MASK)
+};
+
+#define aux_engine_regs_init(id)\
+ ( \
+ AUX_COMMON_REG_LIST0_RI(id), \
+ SR_ARR_INIT(AUXN_IMPCAL, id, 0), \
+ SR_ARR_INIT(AUXP_IMPCAL, id, 0), \
+ SR_ARR_INIT(AUX_RESET_MASK, id, DP_AUX0_AUX_CONTROL__AUX_RESET_MASK) \
+ )
+
+static struct dce110_aux_registers aux_engine_regs[5];
+
+#define dwbc_regs_dcn3_init(id)\
+ DWBC_COMMON_REG_LIST_DCN30_RI(id)
+
+static struct dcn30_dwbc_registers dwbc35_regs[1];
+
+static const struct dcn35_dwbc_shift dwbc35_shift = {
+ DWBC_COMMON_MASK_SH_LIST_DCN35(__SHIFT)
+};
+
+static const struct dcn35_dwbc_mask dwbc35_mask = {
+ DWBC_COMMON_MASK_SH_LIST_DCN35(_MASK)
+};
+
+#define mcif_wb_regs_dcn3_init(id)\
+ MCIF_WB_COMMON_REG_LIST_DCN3_5_RI(id)
+
+static struct dcn35_mmhubbub_registers mcif_wb35_regs[1];
+
+static const struct dcn35_mmhubbub_shift mcif_wb35_shift = {
+ MCIF_WB_COMMON_MASK_SH_LIST_DCN3_5(__SHIFT)
+};
+
+static const struct dcn35_mmhubbub_mask mcif_wb35_mask = {
+ MCIF_WB_COMMON_MASK_SH_LIST_DCN3_5(_MASK)
+};
+
+#define dsc_regsDCN35_init(id)\
+ DSC_REG_LIST_DCN20_RI(id)
+
+static struct dcn20_dsc_registers dsc_regs[4];
+
+static const struct dcn35_dsc_shift dsc_shift = {
+ DSC_REG_LIST_SH_MASK_DCN35(__SHIFT)
+};
+
+static const struct dcn35_dsc_mask dsc_mask = {
+ DSC_REG_LIST_SH_MASK_DCN35(_MASK)
+};
+
+static struct dcn30_mpc_registers mpc_regs;
+
+#define dcn_mpc_regs_init() \
+ MPC_REG_LIST_DCN3_2_RI(0),\
+ MPC_REG_LIST_DCN3_2_RI(1),\
+ MPC_REG_LIST_DCN3_2_RI(2),\
+ MPC_REG_LIST_DCN3_2_RI(3),\
+ MPC_OUT_MUX_REG_LIST_DCN3_0_RI(0),\
+ MPC_OUT_MUX_REG_LIST_DCN3_0_RI(1),\
+ MPC_OUT_MUX_REG_LIST_DCN3_0_RI(2),\
+ MPC_OUT_MUX_REG_LIST_DCN3_0_RI(3),\
+ MPC_DWB_MUX_REG_LIST_DCN3_0_RI(0)
+
+static const struct dcn30_mpc_shift mpc_shift = {
+ MPC_COMMON_MASK_SH_LIST_DCN32(__SHIFT)
+};
+
+static const struct dcn30_mpc_mask mpc_mask = {
+ MPC_COMMON_MASK_SH_LIST_DCN32(_MASK)
+};
+
+#define optc_regs_init(id)\
+ OPTC_COMMON_REG_LIST_DCN3_5_RI(id)
+
+static struct dcn_optc_registers optc_regs[4];
+
+static const struct dcn_optc_shift optc_shift = {
+ OPTC_COMMON_MASK_SH_LIST_DCN3_5(__SHIFT)
+};
+
+static const struct dcn_optc_mask optc_mask = {
+ OPTC_COMMON_MASK_SH_LIST_DCN3_5(_MASK)
+};
+
+#define hubp_regs_init(id)\
+ HUBP_REG_LIST_DCN30_RI(id)
+
+static struct dcn_hubp2_registers hubp_regs[4];
+
+
+static const struct dcn35_hubp2_shift hubp_shift = {
+ HUBP_MASK_SH_LIST_DCN35(__SHIFT)
+};
+
+static const struct dcn35_hubp2_mask hubp_mask = {
+ HUBP_MASK_SH_LIST_DCN35(_MASK)
+};
+
+static struct dcn_hubbub_registers hubbub_reg;
+
+#define hubbub_reg_init()\
+ HUBBUB_REG_LIST_DCN35(0)
+
+static const struct dcn_hubbub_shift hubbub_shift = {
+ HUBBUB_MASK_SH_LIST_DCN35(__SHIFT)
+};
+
+static const struct dcn_hubbub_mask hubbub_mask = {
+ HUBBUB_MASK_SH_LIST_DCN35(_MASK)
+};
+
+static struct dccg_registers dccg_regs;
+
+#define dccg_regs_init()\
+ DCCG_REG_LIST_DCN35()
+
+static const struct dccg_shift dccg_shift = {
+ DCCG_MASK_SH_LIST_DCN35(__SHIFT)
+};
+
+static const struct dccg_mask dccg_mask = {
+ DCCG_MASK_SH_LIST_DCN35(_MASK)
+};
+
+static struct pg_cntl_registers pg_cntl_regs;
+
+#define pg_cntl_dcn35_regs_init() \
+ PG_CNTL_REG_LIST_DCN35()
+
+static const struct pg_cntl_shift pg_cntl_shift = {
+ PG_CNTL_MASK_SH_LIST_DCN35(__SHIFT)
+};
+
+static const struct pg_cntl_mask pg_cntl_mask = {
+ PG_CNTL_MASK_SH_LIST_DCN35(_MASK)
+};
+
+#define SRII2(reg_name_pre, reg_name_post, id)\
+ .reg_name_pre ## _ ## reg_name_post[id] = BASE(reg ## reg_name_pre \
+ ## id ## _ ## reg_name_post ## _BASE_IDX) + \
+ reg ## reg_name_pre ## id ## _ ## reg_name_post
+
+static struct dce_hwseq_registers hwseq_reg;
+
+#define hwseq_reg_init()\
+ HWSEQ_DCN36_REG_LIST()
+
+#define HWSEQ_DCN36_MASK_SH_LIST(mask_sh)\
+ HWSEQ_DCN_MASK_SH_LIST(mask_sh), \
+ HWS_SF(, DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_REFDIV, mask_sh), \
+ HWS_SF(, DCHUBBUB_ARB_HOSTVM_CNTL, DISABLE_HOSTVM_FORCE_ALLOW_PSTATE, mask_sh), \
+ HWS_SF(, DOMAIN0_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN0_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN1_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN1_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN2_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN2_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN3_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN3_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN16_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN16_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN17_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN17_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN18_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN18_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN19_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN19_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN22_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN22_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN23_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN23_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN24_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN24_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN25_PG_CONFIG, DOMAIN_POWER_FORCEON, mask_sh), \
+ HWS_SF(, DOMAIN25_PG_CONFIG, DOMAIN_POWER_GATE, mask_sh), \
+ HWS_SF(, DOMAIN0_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN1_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN2_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN3_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN16_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN17_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN18_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN19_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN22_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN23_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN24_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DOMAIN25_PG_STATUS, DOMAIN_PGFSM_PWR_STATUS, mask_sh), \
+ HWS_SF(, DC_IP_REQUEST_CNTL, IP_REQUEST_EN, mask_sh), \
+ HWS_SF(, AZALIA_AUDIO_DTO, AZALIA_AUDIO_DTO_MODULE, mask_sh), \
+ HWS_SF(, HPO_TOP_CLOCK_CONTROL, HPO_HDMISTREAMCLK_G_GATE_DIS, mask_sh), \
+ HWS_SF(, ODM_MEM_PWR_CTRL3, ODM_MEM_UNASSIGNED_PWR_MODE, mask_sh), \
+ HWS_SF(, ODM_MEM_PWR_CTRL3, ODM_MEM_VBLANK_PWR_MODE, mask_sh), \
+ HWS_SF(, DIO_MEM_PWR_CTRL, I2C_LIGHT_SLEEP_FORCE, mask_sh), \
+ HWS_SF(, HPO_TOP_HW_CONTROL, HPO_IO_EN, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, DISPCLK_R_DMU_GATE_DIS, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, DISPCLK_G_RBBMIF_GATE_DIS, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, RBBMIF_FGCG_REP_DIS, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, DPREFCLK_ALLOW_DS_CLKSTOP, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, DISPCLK_ALLOW_DS_CLKSTOP, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, DPPCLK_ALLOW_DS_CLKSTOP, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, DTBCLK_ALLOW_DS_CLKSTOP, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, DCFCLK_ALLOW_DS_CLKSTOP, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, DPIACLK_ALLOW_DS_CLKSTOP, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, LONO_FGCG_REP_DIS, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, LONO_DISPCLK_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, LONO_SOCCLK_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DMU_CLK_CNTL, LONO_DMCUBCLK_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKA_FE_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKB_FE_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKC_FE_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKD_FE_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKE_FE_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, HDMICHARCLK0_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKA_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKB_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKC_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKD_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, SYMCLKE_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, PHYASYMCLK_ROOT_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, PHYBSYMCLK_ROOT_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, PHYCSYMCLK_ROOT_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, PHYDSYMCLK_ROOT_GATE_DISABLE, mask_sh), \
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL2, PHYESYMCLK_ROOT_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL5, DTBCLK_P0_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL5, DTBCLK_P1_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL5, DTBCLK_P2_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL5, DTBCLK_P3_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK0_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK1_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK2_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL5, DPSTREAMCLK3_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL4, DPIASYMCLK0_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL4, DPIASYMCLK1_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL4, DPIASYMCLK2_GATE_DISABLE, mask_sh),\
+ HWS_SF(, DCCG_GATE_DISABLE_CNTL4, DPIASYMCLK3_GATE_DISABLE, mask_sh)
+
+static const struct dce_hwseq_shift hwseq_shift = {
+ HWSEQ_DCN36_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dce_hwseq_mask hwseq_mask = {
+ HWSEQ_DCN36_MASK_SH_LIST(_MASK)
+};
+
+#define vmid_regs_init(id)\
+ DCN20_VMID_REG_LIST_RI(id)
+
+static struct dcn_vmid_registers vmid_regs[16];
+
+static const struct dcn20_vmid_shift vmid_shifts = {
+ DCN20_VMID_MASK_SH_LIST(__SHIFT)
+};
+
+static const struct dcn20_vmid_mask vmid_masks = {
+ DCN20_VMID_MASK_SH_LIST(_MASK)
+};
+
+static const struct resource_caps res_cap_dcn36 = {
+ .num_timing_generator = 4,
+ .num_opp = 4,
+ .num_video_plane = 4,
+ .num_audio = 5,
+ .num_stream_encoder = 5,
+ .num_dig_link_enc = 5,
+ .num_hpo_dp_stream_encoder = 4,
+ .num_hpo_dp_link_encoder = 2,
+ .num_pll = 4,/*1 c10 edp, 3xc20 combo PHY*/
+ .num_dwb = 1,
+ .num_ddc = 5,
+ .num_vmid = 16,
+ .num_mpc_3dlut = 2,
+ .num_dsc = 4,
+};
+
+static const struct dc_plane_cap plane_cap = {
+ .type = DC_PLANE_TYPE_DCN_UNIVERSAL,
+ .per_pixel_alpha = true,
+
+ .pixel_format_support = {
+ .argb8888 = true,
+ .nv12 = true,
+ .fp16 = true,
+ .p010 = true,
+ .ayuv = false,
+ },
+
+ .max_upscale_factor = {
+ .argb8888 = 16000,
+ .nv12 = 16000,
+ .fp16 = 16000
+ },
+
+ // 6:1 downscaling ratio: 1000/6 = 166.666
+ .max_downscale_factor = {
+ .argb8888 = 250,
+ .nv12 = 167,
+ .fp16 = 167
+ },
+ 64,
+ 64
+};
+
+static const struct dc_debug_options debug_defaults_drv = {
+ .disable_dmcu = true,
+ .force_abm_enable = false,
+ .clock_trace = true,
+ .disable_pplib_clock_request = false,
+ .pipe_split_policy = MPC_SPLIT_AVOID,
+ .force_single_disp_pipe_split = false,
+ .disable_dcc = DCC_ENABLE,
+ .disable_dpp_power_gate = true,
+ .disable_hubp_power_gate = true,
+ .disable_optc_power_gate = true, /*should the same as above two*/
+ .disable_hpo_power_gate = true, /*dmubfw force domain25 on*/
+ .disable_clock_gate = false,
+ .disable_dsc_power_gate = true,
+ .vsr_support = true,
+ .performance_trace = false,
+ .max_downscale_src_width = 4096,/*upto true 4k*/
+ .disable_pplib_wm_range = false,
+ .scl_reset_length10 = true,
+ .sanity_checks = false,
+ .underflow_assert_delay_us = 0xFFFFFFFF,
+ .dwb_fi_phase = -1, // -1 = disable,
+ .dmub_command_table = true,
+ .pstate_enabled = true,
+ .use_max_lb = true,
+ .enable_mem_low_power = {
+ .bits = {
+ .vga = false,
+ .i2c = true,
+ .dmcu = false, // This is previously known to cause hang on S3 cycles if enabled
+ .dscl = true,
+ .cm = true,
+ .mpc = true,
+ .optc = true,
+ .vpg = true,
+ .afmt = true,
+ }
+ },
+ .root_clock_optimization = {
+ .bits = {
+ .dpp = true,
+ .dsc = true,/*dscclk and dsc pg*/
+ .hdmistream = true,
+ .hdmichar = true,
+ .dpstream = true,
+ .symclk32_se = true,
+ .symclk32_le = true,
+ .symclk_fe = true,
+ .physymclk = false,
+ .dpiasymclk = true,
+ }
+ },
+ .seamless_boot_odm_combine = DML_FAIL_SOURCE_PIXEL_FORMAT,
+ .enable_z9_disable_interface = true, /* Allow support for the PMFW interface for disable Z9*/
+ .minimum_z8_residency_time = 1, /* Always allow when other conditions are met */
+ .using_dml2 = true,
+ .support_eDP1_5 = true,
+ .enable_hpo_pg_support = false,
+ .enable_legacy_fast_update = true,
+ .enable_single_display_2to1_odm_policy = true,
+ .disable_idle_power_optimizations = false,
+ .dmcub_emulation = false,
+ .disable_boot_optimizations = false,
+ .disable_unbounded_requesting = false,
+ .disable_mem_low_power = false,
+ //must match enable_single_display_2to1_odm_policy to support dynamic ODM transitions
+ .enable_double_buffered_dsc_pg_support = true,
+ .enable_dp_dig_pixel_rate_div_policy = 1,
+ .disable_z10 = false,
+ .ignore_pg = true,
+ .psp_disabled_wa = true,
+ .ips2_eval_delay_us = 2000,
+ .ips2_entry_delay_us = 800,
+ .disable_dmub_reallow_idle = false,
+ .static_screen_wait_frames = 2,
+ .disable_timeout = true,
+ .min_disp_clk_khz = 50000,
+};
+
+static const struct dc_panel_config panel_config_defaults = {
+ .psr = {
+ .disable_psr = false,
+ .disallow_psrsu = false,
+ .disallow_replay = false,
+ },
+ .ilr = {
+ .optimize_edp_link_rate = true,
+ },
+};
+
+static void dcn35_dpp_destroy(struct dpp **dpp)
+{
+ kfree(TO_DCN20_DPP(*dpp));
+ *dpp = NULL;
+}
+
+static struct dpp *dcn35_dpp_create(struct dc_context *ctx, uint32_t inst)
+{
+ struct dcn3_dpp *dpp = kzalloc(sizeof(struct dcn3_dpp), GFP_KERNEL);
+ bool success = (dpp != NULL);
+
+ if (!success)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT dpp_regs
+ dpp_regs_init(0),
+ dpp_regs_init(1),
+ dpp_regs_init(2),
+ dpp_regs_init(3);
+
+ success = dpp35_construct(dpp, ctx, inst, &dpp_regs[inst], &tf_shift,
+ &tf_mask);
+ if (success) {
+ dpp35_set_fgcg(
+ dpp,
+ ctx->dc->debug.enable_fine_grain_clock_gating.bits.dpp);
+ return &dpp->base;
+ }
+
+ BREAK_TO_DEBUGGER();
+ kfree(dpp);
+ return NULL;
+}
+
+static struct output_pixel_processor *dcn35_opp_create(
+ struct dc_context *ctx, uint32_t inst)
+{
+ struct dcn20_opp *opp =
+ kzalloc(sizeof(struct dcn20_opp), GFP_KERNEL);
+
+ if (!opp) {
+ BREAK_TO_DEBUGGER();
+ return NULL;
+ }
+
+#undef REG_STRUCT
+#define REG_STRUCT opp_regs
+ opp_regs_init(0),
+ opp_regs_init(1),
+ opp_regs_init(2),
+ opp_regs_init(3);
+
+ dcn35_opp_construct(opp, ctx, inst,
+ &opp_regs[inst], &opp_shift, &opp_mask);
+
+ dcn35_opp_set_fgcg(opp, ctx->dc->debug.enable_fine_grain_clock_gating.bits.opp);
+
+ return &opp->base;
+}
+
+static struct dce_aux *dcn31_aux_engine_create(
+ struct dc_context *ctx,
+ uint32_t inst)
+{
+ struct aux_engine_dce110 *aux_engine =
+ kzalloc(sizeof(struct aux_engine_dce110), GFP_KERNEL);
+
+ if (!aux_engine)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT aux_engine_regs
+ aux_engine_regs_init(0),
+ aux_engine_regs_init(1),
+ aux_engine_regs_init(2),
+ aux_engine_regs_init(3),
+ aux_engine_regs_init(4);
+
+ dce110_aux_engine_construct(aux_engine, ctx, inst,
+ SW_AUX_TIMEOUT_PERIOD_MULTIPLIER * AUX_TIMEOUT_PERIOD,
+ &aux_engine_regs[inst],
+ &aux_mask,
+ &aux_shift,
+ ctx->dc->caps.extended_aux_timeout_support);
+
+ return &aux_engine->base;
+}
+
+#define i2c_inst_regs_init(id)\
+ I2C_HW_ENGINE_COMMON_REG_LIST_DCN30_RI(id)
+
+static struct dce_i2c_registers i2c_hw_regs[5];
+
+static const struct dce_i2c_shift i2c_shifts = {
+ I2C_COMMON_MASK_SH_LIST_DCN35(__SHIFT)
+};
+
+static const struct dce_i2c_mask i2c_masks = {
+ I2C_COMMON_MASK_SH_LIST_DCN35(_MASK)
+};
+
+/* ========================================================== */
+
+/*
+ * DPIA index | Preferred Encoder | Host Router
+ * 0 | C | 0
+ * 1 | First Available | 0
+ * 2 | D | 1
+ * 3 | First Available | 1
+ */
+/* ========================================================== */
+static const enum engine_id dpia_to_preferred_enc_id_table[] = {
+ ENGINE_ID_DIGC,
+ ENGINE_ID_DIGC,
+ ENGINE_ID_DIGD,
+ ENGINE_ID_DIGD
+};
+
+static enum engine_id dcn36_get_preferred_eng_id_dpia(unsigned int dpia_index)
+{
+ return dpia_to_preferred_enc_id_table[dpia_index];
+}
+
+static struct dce_i2c_hw *dcn31_i2c_hw_create(
+ struct dc_context *ctx,
+ uint32_t inst)
+{
+ struct dce_i2c_hw *dce_i2c_hw =
+ kzalloc(sizeof(struct dce_i2c_hw), GFP_KERNEL);
+
+ if (!dce_i2c_hw)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT i2c_hw_regs
+ i2c_inst_regs_init(1),
+ i2c_inst_regs_init(2),
+ i2c_inst_regs_init(3),
+ i2c_inst_regs_init(4),
+ i2c_inst_regs_init(5);
+
+ dcn2_i2c_hw_construct(dce_i2c_hw, ctx, inst,
+ &i2c_hw_regs[inst], &i2c_shifts, &i2c_masks);
+
+ return dce_i2c_hw;
+}
+static struct mpc *dcn35_mpc_create(
+ struct dc_context *ctx,
+ int num_mpcc,
+ int num_rmu)
+{
+ struct dcn30_mpc *mpc30 = kzalloc(sizeof(struct dcn30_mpc), GFP_KERNEL);
+
+ if (!mpc30)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT mpc_regs
+ dcn_mpc_regs_init();
+
+ dcn32_mpc_construct(mpc30, ctx,
+ &mpc_regs,
+ &mpc_shift,
+ &mpc_mask,
+ num_mpcc,
+ num_rmu);
+
+ return &mpc30->base;
+}
+
+static struct hubbub *dcn35_hubbub_create(struct dc_context *ctx)
+{
+ int i;
+
+ struct dcn20_hubbub *hubbub3 = kzalloc(sizeof(struct dcn20_hubbub),
+ GFP_KERNEL);
+
+ if (!hubbub3)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT hubbub_reg
+ hubbub_reg_init();
+
+#undef REG_STRUCT
+#define REG_STRUCT vmid_regs
+ vmid_regs_init(0),
+ vmid_regs_init(1),
+ vmid_regs_init(2),
+ vmid_regs_init(3),
+ vmid_regs_init(4),
+ vmid_regs_init(5),
+ vmid_regs_init(6),
+ vmid_regs_init(7),
+ vmid_regs_init(8),
+ vmid_regs_init(9),
+ vmid_regs_init(10),
+ vmid_regs_init(11),
+ vmid_regs_init(12),
+ vmid_regs_init(13),
+ vmid_regs_init(14),
+ vmid_regs_init(15);
+
+ hubbub35_construct(hubbub3, ctx,
+ &hubbub_reg,
+ &hubbub_shift,
+ &hubbub_mask,
+ 384,/*ctx->dc->dml.ip.det_buffer_size_kbytes,*/
+ 8, /*ctx->dc->dml.ip.pixel_chunk_size_kbytes,*/
+ 1792 /*ctx->dc->dml.ip.config_return_buffer_size_in_kbytes*/);
+
+
+ for (i = 0; i < res_cap_dcn36.num_vmid; i++) {
+ struct dcn20_vmid *vmid = &hubbub3->vmid[i];
+
+ vmid->ctx = ctx;
+
+ vmid->regs = &vmid_regs[i];
+ vmid->shifts = &vmid_shifts;
+ vmid->masks = &vmid_masks;
+ }
+
+ return &hubbub3->base;
+}
+
+static struct timing_generator *dcn35_timing_generator_create(
+ struct dc_context *ctx,
+ uint32_t instance)
+{
+ struct optc *tgn10 =
+ kzalloc(sizeof(struct optc), GFP_KERNEL);
+
+ if (!tgn10)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT optc_regs
+ optc_regs_init(0),
+ optc_regs_init(1),
+ optc_regs_init(2),
+ optc_regs_init(3);
+
+ tgn10->base.inst = instance;
+ tgn10->base.ctx = ctx;
+
+ tgn10->tg_regs = &optc_regs[instance];
+ tgn10->tg_shift = &optc_shift;
+ tgn10->tg_mask = &optc_mask;
+
+ dcn35_timing_generator_init(tgn10);
+
+ return &tgn10->base;
+}
+
+static const struct encoder_feature_support link_enc_feature = {
+ .max_hdmi_deep_color = COLOR_DEPTH_121212,
+ .max_hdmi_pixel_clock = 600000,
+ .hdmi_ycbcr420_supported = true,
+ .dp_ycbcr420_supported = true,
+ .fec_supported = true,
+ .flags.bits.IS_HBR2_CAPABLE = true,
+ .flags.bits.IS_HBR3_CAPABLE = true,
+ .flags.bits.IS_TPS3_CAPABLE = true,
+ .flags.bits.IS_TPS4_CAPABLE = true
+};
+
+static struct link_encoder *dcn35_link_encoder_create(
+ struct dc_context *ctx,
+ const struct encoder_init_data *enc_init_data)
+{
+ struct dcn20_link_encoder *enc20 =
+ kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL);
+
+ if (!enc20 || enc_init_data->hpd_source >= ARRAY_SIZE(link_enc_hpd_regs))
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT link_enc_aux_regs
+ aux_regs_init(0),
+ aux_regs_init(1),
+ aux_regs_init(2),
+ aux_regs_init(3),
+ aux_regs_init(4);
+
+#undef REG_STRUCT
+#define REG_STRUCT link_enc_hpd_regs
+ hpd_regs_init(0),
+ hpd_regs_init(1),
+ hpd_regs_init(2),
+ hpd_regs_init(3),
+ hpd_regs_init(4);
+
+#undef REG_STRUCT
+#define REG_STRUCT link_enc_regs
+ link_regs_init(0, A),
+ link_regs_init(1, B),
+ link_regs_init(2, C),
+ link_regs_init(3, D),
+ link_regs_init(4, E);
+
+ dcn35_link_encoder_construct(enc20,
+ enc_init_data,
+ &link_enc_feature,
+ &link_enc_regs[enc_init_data->transmitter],
+ &link_enc_aux_regs[enc_init_data->channel - 1],
+ &link_enc_hpd_regs[enc_init_data->hpd_source],
+ &le_shift,
+ &le_mask);
+
+ return &enc20->enc10.base;
+}
+
+/* Create a minimal link encoder object not associated with a particular
+ * physical connector.
+ * resource_funcs.link_enc_create_minimal
+ */
+static struct link_encoder *dcn31_link_enc_create_minimal(
+ struct dc_context *ctx, enum engine_id eng_id)
+{
+ struct dcn20_link_encoder *enc20;
+
+ if ((eng_id - ENGINE_ID_DIGA) > ctx->dc->res_pool->res_cap->num_dig_link_enc)
+ return NULL;
+
+ enc20 = kzalloc(sizeof(struct dcn20_link_encoder), GFP_KERNEL);
+ if (!enc20)
+ return NULL;
+
+ dcn31_link_encoder_construct_minimal(
+ enc20,
+ ctx,
+ &link_enc_feature,
+ &link_enc_regs[eng_id - ENGINE_ID_DIGA],
+ eng_id);
+
+ return &enc20->enc10.base;
+}
+
+static struct panel_cntl *dcn31_panel_cntl_create(const struct panel_cntl_init_data *init_data)
+{
+ struct dcn31_panel_cntl *panel_cntl =
+ kzalloc(sizeof(struct dcn31_panel_cntl), GFP_KERNEL);
+
+ if (!panel_cntl)
+ return NULL;
+
+ dcn31_panel_cntl_construct(panel_cntl, init_data);
+
+ return &panel_cntl->base;
+}
+
+static void read_dce_straps(
+ struct dc_context *ctx,
+ struct resource_straps *straps)
+{
+ generic_reg_get(ctx, regDC_PINSTRAPS + BASE(regDC_PINSTRAPS_BASE_IDX),
+ FN(DC_PINSTRAPS, DC_PINSTRAPS_AUDIO), &straps->dc_pinstraps_audio);
+
+}
+
+static struct audio *dcn31_create_audio(
+ struct dc_context *ctx, unsigned int inst)
+{
+
+#undef REG_STRUCT
+#define REG_STRUCT audio_regs
+ audio_regs_init(0),
+ audio_regs_init(1),
+ audio_regs_init(2),
+ audio_regs_init(3),
+ audio_regs_init(4);
+ audio_regs_init(5);
+ audio_regs_init(6);
+
+ return dce_audio_create(ctx, inst,
+ &audio_regs[inst], &audio_shift, &audio_mask);
+}
+
+static struct vpg *dcn31_vpg_create(
+ struct dc_context *ctx,
+ uint32_t inst)
+{
+ struct dcn31_vpg *vpg31 = kzalloc(sizeof(struct dcn31_vpg), GFP_KERNEL);
+
+ if (!vpg31)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT vpg_regs
+ vpg_regs_init(0),
+ vpg_regs_init(1),
+ vpg_regs_init(2),
+ vpg_regs_init(3),
+ vpg_regs_init(4),
+ vpg_regs_init(5),
+ vpg_regs_init(6),
+ vpg_regs_init(7),
+ vpg_regs_init(8),
+ vpg_regs_init(9);
+
+ vpg31_construct(vpg31, ctx, inst,
+ &vpg_regs[inst],
+ &vpg_shift,
+ &vpg_mask);
+
+ return &vpg31->base;
+}
+
+static struct afmt *dcn31_afmt_create(
+ struct dc_context *ctx,
+ uint32_t inst)
+{
+ struct dcn31_afmt *afmt31 = kzalloc(sizeof(struct dcn31_afmt), GFP_KERNEL);
+
+ if (!afmt31)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT afmt_regs
+ afmt_regs_init(0),
+ afmt_regs_init(1),
+ afmt_regs_init(2),
+ afmt_regs_init(3),
+ afmt_regs_init(4),
+ afmt_regs_init(5);
+
+ afmt31_construct(afmt31, ctx, inst,
+ &afmt_regs[inst],
+ &afmt_shift,
+ &afmt_mask);
+
+ // Light sleep by default, no need to power down here
+
+ return &afmt31->base;
+}
+
+static struct apg *dcn31_apg_create(
+ struct dc_context *ctx,
+ uint32_t inst)
+{
+ struct dcn31_apg *apg31 = kzalloc(sizeof(struct dcn31_apg), GFP_KERNEL);
+
+ if (!apg31)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT apg_regs
+ apg_regs_init(0),
+ apg_regs_init(1),
+ apg_regs_init(2),
+ apg_regs_init(3);
+
+ apg31_construct(apg31, ctx, inst,
+ &apg_regs[inst],
+ &apg_shift,
+ &apg_mask);
+
+ return &apg31->base;
+}
+
+static struct stream_encoder *dcn35_stream_encoder_create(
+ enum engine_id eng_id,
+ struct dc_context *ctx)
+{
+ struct dcn10_stream_encoder *enc1;
+ struct vpg *vpg;
+ struct afmt *afmt;
+ int vpg_inst;
+ int afmt_inst;
+
+ /* Mapping of VPG, AFMT, DME register blocks to DIO block instance */
+ if (eng_id <= ENGINE_ID_DIGF) {
+ vpg_inst = eng_id;
+ afmt_inst = eng_id;
+ } else
+ return NULL;
+
+ enc1 = kzalloc(sizeof(struct dcn10_stream_encoder), GFP_KERNEL);
+ vpg = dcn31_vpg_create(ctx, vpg_inst);
+ afmt = dcn31_afmt_create(ctx, afmt_inst);
+
+ if (!enc1 || !vpg || !afmt) {
+ kfree(enc1);
+ kfree(vpg);
+ kfree(afmt);
+ return NULL;
+ }
+
+#undef REG_STRUCT
+#define REG_STRUCT stream_enc_regs
+ stream_enc_regs_init(0),
+ stream_enc_regs_init(1),
+ stream_enc_regs_init(2),
+ stream_enc_regs_init(3),
+ stream_enc_regs_init(4);
+
+ dcn35_dio_stream_encoder_construct(enc1, ctx, ctx->dc_bios,
+ eng_id, vpg, afmt,
+ &stream_enc_regs[eng_id],
+ &se_shift, &se_mask);
+
+ return &enc1->base;
+}
+
+static struct hpo_dp_stream_encoder *dcn31_hpo_dp_stream_encoder_create(
+ enum engine_id eng_id,
+ struct dc_context *ctx)
+{
+ struct dcn31_hpo_dp_stream_encoder *hpo_dp_enc31;
+ struct vpg *vpg;
+ struct apg *apg;
+ uint32_t hpo_dp_inst;
+ uint32_t vpg_inst;
+ uint32_t apg_inst;
+
+ ASSERT((eng_id >= ENGINE_ID_HPO_DP_0) && (eng_id <= ENGINE_ID_HPO_DP_3));
+ hpo_dp_inst = eng_id - ENGINE_ID_HPO_DP_0;
+
+ /* Mapping of VPG register blocks to HPO DP block instance:
+ * VPG[6] -> HPO_DP[0]
+ * VPG[7] -> HPO_DP[1]
+ * VPG[8] -> HPO_DP[2]
+ * VPG[9] -> HPO_DP[3]
+ */
+ vpg_inst = hpo_dp_inst + 6;
+
+ /* Mapping of APG register blocks to HPO DP block instance:
+ * APG[0] -> HPO_DP[0]
+ * APG[1] -> HPO_DP[1]
+ * APG[2] -> HPO_DP[2]
+ * APG[3] -> HPO_DP[3]
+ */
+ apg_inst = hpo_dp_inst;
+
+ /* allocate HPO stream encoder and create VPG sub-block */
+ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_stream_encoder), GFP_KERNEL);
+ vpg = dcn31_vpg_create(ctx, vpg_inst);
+ apg = dcn31_apg_create(ctx, apg_inst);
+
+ if (!hpo_dp_enc31 || !vpg || !apg) {
+ kfree(hpo_dp_enc31);
+ kfree(vpg);
+ kfree(apg);
+ return NULL;
+ }
+
+#undef REG_STRUCT
+#define REG_STRUCT hpo_dp_stream_enc_regs
+ hpo_dp_stream_encoder_reg_init(0),
+ hpo_dp_stream_encoder_reg_init(1),
+ hpo_dp_stream_encoder_reg_init(2),
+ hpo_dp_stream_encoder_reg_init(3);
+
+ dcn31_hpo_dp_stream_encoder_construct(hpo_dp_enc31, ctx, ctx->dc_bios,
+ hpo_dp_inst, eng_id, vpg, apg,
+ &hpo_dp_stream_enc_regs[hpo_dp_inst],
+ &hpo_dp_se_shift, &hpo_dp_se_mask);
+
+ return &hpo_dp_enc31->base;
+}
+
+static struct hpo_dp_link_encoder *dcn31_hpo_dp_link_encoder_create(
+ uint8_t inst,
+ struct dc_context *ctx)
+{
+ struct dcn31_hpo_dp_link_encoder *hpo_dp_enc31;
+
+ /* allocate HPO link encoder */
+ hpo_dp_enc31 = kzalloc(sizeof(struct dcn31_hpo_dp_link_encoder), GFP_KERNEL);
+ if (!hpo_dp_enc31)
+ return NULL; /* out of memory */
+
+#undef REG_STRUCT
+#define REG_STRUCT hpo_dp_link_enc_regs
+ hpo_dp_link_encoder_reg_init(0),
+ hpo_dp_link_encoder_reg_init(1);
+
+ hpo_dp_link_encoder31_construct(hpo_dp_enc31, ctx, inst,
+ &hpo_dp_link_enc_regs[inst],
+ &hpo_dp_le_shift, &hpo_dp_le_mask);
+
+ return &hpo_dp_enc31->base;
+}
+
+static struct dce_hwseq *dcn36_hwseq_create(
+ struct dc_context *ctx)
+{
+ struct dce_hwseq *hws = kzalloc(sizeof(struct dce_hwseq), GFP_KERNEL);
+
+#undef REG_STRUCT
+#define REG_STRUCT hwseq_reg
+ hwseq_reg_init();
+
+ if (hws) {
+ hws->ctx = ctx;
+ hws->regs = &hwseq_reg;
+ hws->shifts = &hwseq_shift;
+ hws->masks = &hwseq_mask;
+ }
+ return hws;
+}
+static const struct resource_create_funcs res_create_funcs = {
+ .read_dce_straps = read_dce_straps,
+ .create_audio = dcn31_create_audio,
+ .create_stream_encoder = dcn35_stream_encoder_create,
+ .create_hpo_dp_stream_encoder = dcn31_hpo_dp_stream_encoder_create,
+ .create_hpo_dp_link_encoder = dcn31_hpo_dp_link_encoder_create,
+ .create_hwseq = dcn36_hwseq_create,
+};
+
+static void dcn36_resource_destruct(struct dcn36_resource_pool *pool)
+{
+ unsigned int i;
+
+ for (i = 0; i < pool->base.stream_enc_count; i++) {
+ if (pool->base.stream_enc[i] != NULL) {
+ if (pool->base.stream_enc[i]->vpg != NULL) {
+ kfree(DCN30_VPG_FROM_VPG(pool->base.stream_enc[i]->vpg));
+ pool->base.stream_enc[i]->vpg = NULL;
+ }
+ if (pool->base.stream_enc[i]->afmt != NULL) {
+ kfree(DCN30_AFMT_FROM_AFMT(pool->base.stream_enc[i]->afmt));
+ pool->base.stream_enc[i]->afmt = NULL;
+ }
+ kfree(DCN10STRENC_FROM_STRENC(pool->base.stream_enc[i]));
+ pool->base.stream_enc[i] = NULL;
+ }
+ }
+
+ for (i = 0; i < pool->base.hpo_dp_stream_enc_count; i++) {
+ if (pool->base.hpo_dp_stream_enc[i] != NULL) {
+ if (pool->base.hpo_dp_stream_enc[i]->vpg != NULL) {
+ kfree(DCN30_VPG_FROM_VPG(pool->base.hpo_dp_stream_enc[i]->vpg));
+ pool->base.hpo_dp_stream_enc[i]->vpg = NULL;
+ }
+ if (pool->base.hpo_dp_stream_enc[i]->apg != NULL) {
+ kfree(DCN31_APG_FROM_APG(pool->base.hpo_dp_stream_enc[i]->apg));
+ pool->base.hpo_dp_stream_enc[i]->apg = NULL;
+ }
+ kfree(DCN3_1_HPO_DP_STREAM_ENC_FROM_HPO_STREAM_ENC(pool->base.hpo_dp_stream_enc[i]));
+ pool->base.hpo_dp_stream_enc[i] = NULL;
+ }
+ }
+
+ for (i = 0; i < pool->base.hpo_dp_link_enc_count; i++) {
+ if (pool->base.hpo_dp_link_enc[i] != NULL) {
+ kfree(DCN3_1_HPO_DP_LINK_ENC_FROM_HPO_LINK_ENC(pool->base.hpo_dp_link_enc[i]));
+ pool->base.hpo_dp_link_enc[i] = NULL;
+ }
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_dsc; i++) {
+ if (pool->base.dscs[i] != NULL)
+ dcn20_dsc_destroy(&pool->base.dscs[i]);
+ }
+
+ if (pool->base.mpc != NULL) {
+ kfree(TO_DCN20_MPC(pool->base.mpc));
+ pool->base.mpc = NULL;
+ }
+ if (pool->base.hubbub != NULL) {
+ kfree(pool->base.hubbub);
+ pool->base.hubbub = NULL;
+ }
+ for (i = 0; i < pool->base.pipe_count; i++) {
+ if (pool->base.dpps[i] != NULL)
+ dcn35_dpp_destroy(&pool->base.dpps[i]);
+
+ if (pool->base.ipps[i] != NULL)
+ pool->base.ipps[i]->funcs->ipp_destroy(&pool->base.ipps[i]);
+
+ if (pool->base.hubps[i] != NULL) {
+ kfree(TO_DCN20_HUBP(pool->base.hubps[i]));
+ pool->base.hubps[i] = NULL;
+ }
+
+ if (pool->base.irqs != NULL) {
+ dal_irq_service_destroy(&pool->base.irqs);
+ }
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_ddc; i++) {
+ if (pool->base.engines[i] != NULL)
+ dce110_engine_destroy(&pool->base.engines[i]);
+ if (pool->base.hw_i2cs[i] != NULL) {
+ kfree(pool->base.hw_i2cs[i]);
+ pool->base.hw_i2cs[i] = NULL;
+ }
+ if (pool->base.sw_i2cs[i] != NULL) {
+ kfree(pool->base.sw_i2cs[i]);
+ pool->base.sw_i2cs[i] = NULL;
+ }
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_opp; i++) {
+ if (pool->base.opps[i] != NULL)
+ pool->base.opps[i]->funcs->opp_destroy(&pool->base.opps[i]);
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_timing_generator; i++) {
+ if (pool->base.timing_generators[i] != NULL) {
+ kfree(DCN10TG_FROM_TG(pool->base.timing_generators[i]));
+ pool->base.timing_generators[i] = NULL;
+ }
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_dwb; i++) {
+ if (pool->base.dwbc[i] != NULL) {
+ kfree(TO_DCN30_DWBC(pool->base.dwbc[i]));
+ pool->base.dwbc[i] = NULL;
+ }
+ if (pool->base.mcif_wb[i] != NULL) {
+ kfree(TO_DCN30_MMHUBBUB(pool->base.mcif_wb[i]));
+ pool->base.mcif_wb[i] = NULL;
+ }
+ }
+
+ for (i = 0; i < pool->base.audio_count; i++) {
+ if (pool->base.audios[i])
+ dce_aud_destroy(&pool->base.audios[i]);
+ }
+
+ for (i = 0; i < pool->base.clk_src_count; i++) {
+ if (pool->base.clock_sources[i] != NULL) {
+ dcn20_clock_source_destroy(&pool->base.clock_sources[i]);
+ pool->base.clock_sources[i] = NULL;
+ }
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_mpc_3dlut; i++) {
+ if (pool->base.mpc_lut[i] != NULL) {
+ dc_3dlut_func_release(pool->base.mpc_lut[i]);
+ pool->base.mpc_lut[i] = NULL;
+ }
+ if (pool->base.mpc_shaper[i] != NULL) {
+ dc_transfer_func_release(pool->base.mpc_shaper[i]);
+ pool->base.mpc_shaper[i] = NULL;
+ }
+ }
+
+ if (pool->base.dp_clock_source != NULL) {
+ dcn20_clock_source_destroy(&pool->base.dp_clock_source);
+ pool->base.dp_clock_source = NULL;
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_timing_generator; i++) {
+ if (pool->base.multiple_abms[i] != NULL)
+ dce_abm_destroy(&pool->base.multiple_abms[i]);
+ }
+
+ if (pool->base.psr != NULL)
+ dmub_psr_destroy(&pool->base.psr);
+
+ if (pool->base.replay != NULL)
+ dmub_replay_destroy(&pool->base.replay);
+
+ if (pool->base.pg_cntl != NULL)
+ dcn_pg_cntl_destroy(&pool->base.pg_cntl);
+
+ if (pool->base.dccg != NULL)
+ dcn_dccg_destroy(&pool->base.dccg);
+}
+
+static struct hubp *dcn35_hubp_create(
+ struct dc_context *ctx,
+ uint32_t inst)
+{
+ struct dcn20_hubp *hubp2 =
+ kzalloc(sizeof(struct dcn20_hubp), GFP_KERNEL);
+
+ if (!hubp2)
+ return NULL;
+
+#undef REG_STRUCT
+#define REG_STRUCT hubp_regs
+ hubp_regs_init(0),
+ hubp_regs_init(1),
+ hubp_regs_init(2),
+ hubp_regs_init(3);
+
+ if (hubp35_construct(hubp2, ctx, inst,
+ &hubp_regs[inst], &hubp_shift, &hubp_mask))
+ return &hubp2->base;
+
+ BREAK_TO_DEBUGGER();
+ kfree(hubp2);
+ return NULL;
+}
+
+static void dcn35_dwbc_init(struct dcn30_dwbc *dwbc30, struct dc_context *ctx)
+{
+ dcn35_dwbc_set_fgcg(
+ dwbc30, ctx->dc->debug.enable_fine_grain_clock_gating.bits.dwb);
+}
+
+static bool dcn35_dwbc_create(struct dc_context *ctx, struct resource_pool *pool)
+{
+ int i;
+ uint32_t pipe_count = pool->res_cap->num_dwb;
+
+ for (i = 0; i < pipe_count; i++) {
+ struct dcn30_dwbc *dwbc30 = kzalloc(sizeof(struct dcn30_dwbc),
+ GFP_KERNEL);
+
+ if (!dwbc30) {
+ dm_error("DC: failed to create dwbc30!\n");
+ return false;
+ }
+
+#undef REG_STRUCT
+#define REG_STRUCT dwbc35_regs
+ dwbc_regs_dcn3_init(0);
+
+ dcn35_dwbc_construct(dwbc30, ctx,
+ &dwbc35_regs[i],
+ &dwbc35_shift,
+ &dwbc35_mask,
+ i);
+
+ pool->dwbc[i] = &dwbc30->base;
+
+ dcn35_dwbc_init(dwbc30, ctx);
+ }
+ return true;
+}
+
+static void dcn35_mmhubbub_init(struct dcn30_mmhubbub *mcif_wb30,
+ struct dc_context *ctx)
+{
+ dcn35_mmhubbub_set_fgcg(
+ mcif_wb30,
+ ctx->dc->debug.enable_fine_grain_clock_gating.bits.mmhubbub);
+}
+
+static bool dcn35_mmhubbub_create(struct dc_context *ctx, struct resource_pool *pool)
+{
+ int i;
+ uint32_t pipe_count = pool->res_cap->num_dwb;
+
+ for (i = 0; i < pipe_count; i++) {
+ struct dcn30_mmhubbub *mcif_wb30 = kzalloc(sizeof(struct dcn30_mmhubbub),
+ GFP_KERNEL);
+
+ if (!mcif_wb30) {
+ dm_error("DC: failed to create mcif_wb30!\n");
+ return false;
+ }
+
+#undef REG_STRUCT
+#define REG_STRUCT mcif_wb35_regs
+ mcif_wb_regs_dcn3_init(0);
+
+ dcn35_mmhubbub_construct(mcif_wb30, ctx,
+ &mcif_wb35_regs[i],
+ &mcif_wb35_shift,
+ &mcif_wb35_mask,
+ i);
+
+ dcn35_mmhubbub_init(mcif_wb30, ctx);
+
+ pool->mcif_wb[i] = &mcif_wb30->base;
+ }
+ return true;
+}
+
+static struct display_stream_compressor *dcn35_dsc_create(
+ struct dc_context *ctx, uint32_t inst)
+{
+ struct dcn20_dsc *dsc =
+ kzalloc(sizeof(struct dcn20_dsc), GFP_KERNEL);
+
+ if (!dsc) {
+ BREAK_TO_DEBUGGER();
+ return NULL;
+ }
+
+#undef REG_STRUCT
+#define REG_STRUCT dsc_regs
+ dsc_regsDCN35_init(0),
+ dsc_regsDCN35_init(1),
+ dsc_regsDCN35_init(2),
+ dsc_regsDCN35_init(3);
+
+ dsc35_construct(dsc, ctx, inst, &dsc_regs[inst], &dsc_shift, &dsc_mask);
+ dsc35_set_fgcg(dsc,
+ ctx->dc->debug.enable_fine_grain_clock_gating.bits.dsc);
+ return &dsc->base;
+}
+
+static void dcn36_destroy_resource_pool(struct resource_pool **pool)
+{
+ struct dcn36_resource_pool *dcn36_pool = TO_DCN36_RES_POOL(*pool);
+
+ dcn36_resource_destruct(dcn36_pool);
+ kfree(dcn36_pool);
+ *pool = NULL;
+}
+
+static struct clock_source *dcn35_clock_source_create(
+ struct dc_context *ctx,
+ struct dc_bios *bios,
+ enum clock_source_id id,
+ const struct dce110_clk_src_regs *regs,
+ bool dp_clk_src)
+{
+ struct dce110_clk_src *clk_src =
+ kzalloc(sizeof(struct dce110_clk_src), GFP_KERNEL);
+
+ if (!clk_src)
+ return NULL;
+
+ if (dcn31_clk_src_construct(clk_src, ctx, bios, id,
+ regs, &cs_shift, &cs_mask)) {
+ clk_src->base.dp_clk_src = dp_clk_src;
+ return &clk_src->base;
+ }
+
+ kfree(clk_src);
+ BREAK_TO_DEBUGGER();
+ return NULL;
+}
+
+static struct dc_cap_funcs cap_funcs = {
+ .get_dcc_compression_cap = dcn20_get_dcc_compression_cap
+};
+
+static void dcn35_get_panel_config_defaults(struct dc_panel_config *panel_config)
+{
+ *panel_config = panel_config_defaults;
+}
+
+
+static enum dc_status dcn35_validate_bandwidth(struct dc *dc,
+ struct dc_state *context,
+ enum dc_validate_mode validate_mode)
+{
+ bool out = false;
+
+ out = dml2_validate(dc, context,
+ context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2,
+ validate_mode);
+
+ if (validate_mode != DC_VALIDATE_MODE_AND_PROGRAMMING)
+ return out ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
+
+ DC_FP_START();
+ dcn35_decide_zstate_support(dc, context);
+ DC_FP_END();
+
+ return out ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
+}
+
+
+static struct resource_funcs dcn36_res_pool_funcs = {
+ .destroy = dcn36_destroy_resource_pool,
+ .link_enc_create = dcn35_link_encoder_create,
+ .link_enc_create_minimal = dcn31_link_enc_create_minimal,
+ .link_encs_assign = link_enc_cfg_link_encs_assign,
+ .link_enc_unassign = link_enc_cfg_link_enc_unassign,
+ .panel_cntl_create = dcn31_panel_cntl_create,
+ .validate_bandwidth = dcn35_validate_bandwidth,
+ .calculate_wm_and_dlg = NULL,
+ .update_soc_for_wm_a = dcn31_update_soc_for_wm_a,
+ .populate_dml_pipes = dcn35_populate_dml_pipes_from_context_fpu,
+ .acquire_free_pipe_as_secondary_dpp_pipe = dcn20_acquire_free_pipe_for_layer,
+ .release_pipe = dcn20_release_pipe,
+ .add_stream_to_ctx = dcn30_add_stream_to_ctx,
+ .add_dsc_to_stream_resource = dcn20_add_dsc_to_stream_resource,
+ .remove_stream_from_ctx = dcn20_remove_stream_from_ctx,
+ .populate_dml_writeback_from_context = dcn30_populate_dml_writeback_from_context,
+ .set_mcif_arb_params = dcn30_set_mcif_arb_params,
+ .find_first_free_match_stream_enc_for_link = dcn10_find_first_free_match_stream_enc_for_link,
+ .acquire_post_bldn_3dlut = dcn30_acquire_post_bldn_3dlut,
+ .release_post_bldn_3dlut = dcn30_release_post_bldn_3dlut,
+ .update_bw_bounding_box = dcn35_update_bw_bounding_box_fpu,
+ .patch_unknown_plane_state = dcn20_patch_unknown_plane_state,
+ .get_panel_config_defaults = dcn35_get_panel_config_defaults,
+ .get_preferred_eng_id_dpia = dcn36_get_preferred_eng_id_dpia,
+ .get_vstartup_for_pipe = dcn10_get_vstartup_for_pipe,
+ .update_dc_state_for_encoder_switch = dcn31_update_dc_state_for_encoder_switch,
+ .build_pipe_pix_clk_params = dcn20_build_pipe_pix_clk_params
+};
+
+static bool dcn36_resource_construct(
+ uint8_t num_virtual_links,
+ struct dc *dc,
+ struct dcn36_resource_pool *pool)
+{
+ int i;
+ struct dc_context *ctx = dc->ctx;
+ struct irq_service_init_data init_data;
+
+#undef REG_STRUCT
+#define REG_STRUCT bios_regs
+ bios_regs_init();
+
+#undef REG_STRUCT
+#define REG_STRUCT clk_src_regs
+ clk_src_regs_init(0, A),
+ clk_src_regs_init(1, B),
+ clk_src_regs_init(2, C),
+ clk_src_regs_init(3, D),
+ clk_src_regs_init(4, E);
+
+#undef REG_STRUCT
+#define REG_STRUCT abm_regs
+ abm_regs_init(0),
+ abm_regs_init(1),
+ abm_regs_init(2),
+ abm_regs_init(3);
+
+#undef REG_STRUCT
+#define REG_STRUCT dccg_regs
+ dccg_regs_init();
+
+ ctx->dc_bios->regs = &bios_regs;
+
+ pool->base.res_cap = &res_cap_dcn36;
+
+ pool->base.funcs = &dcn36_res_pool_funcs;
+
+ /*************************************************
+ * Resource + asic cap harcoding *
+ *************************************************/
+ pool->base.underlay_pipe_index = NO_UNDERLAY_PIPE;
+ pool->base.pipe_count = pool->base.res_cap->num_timing_generator;
+ pool->base.mpcc_count = pool->base.res_cap->num_timing_generator;
+ dc->caps.max_downscale_ratio = 600;
+ dc->caps.i2c_speed_in_khz = 100;
+ dc->caps.i2c_speed_in_khz_hdcp = 100;
+ dc->caps.max_cursor_size = 256;
+ dc->caps.min_horizontal_blanking_period = 80;
+ dc->caps.dmdata_alloc_size = 2048;
+ dc->caps.max_slave_planes = 3;
+ dc->caps.max_slave_yuv_planes = 3;
+ dc->caps.max_slave_rgb_planes = 3;
+ dc->caps.post_blend_color_processing = true;
+ dc->caps.force_dp_tps4_for_cp2520 = true;
+ if (dc->config.forceHBR2CP2520)
+ dc->caps.force_dp_tps4_for_cp2520 = false;
+ dc->caps.dp_hpo = true;
+ dc->caps.dp_hdmi21_pcon_support = true;
+
+ dc->caps.edp_dsc_support = true;
+ dc->caps.extended_aux_timeout_support = true;
+ dc->caps.dmcub_support = true;
+ dc->caps.is_apu = true;
+ dc->caps.seamless_odm = true;
+
+ dc->caps.zstate_support = true;
+ dc->caps.ips_support = true;
+ dc->caps.max_v_total = (1 << 15) - 1;
+ dc->caps.vtotal_limited_by_fp2 = true;
+
+ /* Color pipeline capabilities */
+ dc->caps.color.dpp.dcn_arch = 1;
+ dc->caps.color.dpp.input_lut_shared = 0;
+ dc->caps.color.dpp.icsc = 1;
+ dc->caps.color.dpp.dgam_ram = 0; // must use gamma_corr
+ dc->caps.color.dpp.dgam_rom_caps.srgb = 1;
+ dc->caps.color.dpp.dgam_rom_caps.bt2020 = 1;
+ dc->caps.color.dpp.dgam_rom_caps.gamma2_2 = 1;
+ dc->caps.color.dpp.dgam_rom_caps.pq = 1;
+ dc->caps.color.dpp.dgam_rom_caps.hlg = 1;
+ dc->caps.color.dpp.post_csc = 1;
+ dc->caps.color.dpp.gamma_corr = 1;
+ dc->caps.color.dpp.dgam_rom_for_yuv = 0;
+
+ dc->caps.color.dpp.hw_3d_lut = 0;
+ dc->caps.color.dpp.ogam_ram = 0; // no OGAM in DPP since DCN1
+ // no OGAM ROM on DCN301
+ dc->caps.color.dpp.ogam_rom_caps.srgb = 0;
+ dc->caps.color.dpp.ogam_rom_caps.bt2020 = 0;
+ dc->caps.color.dpp.ogam_rom_caps.gamma2_2 = 0;
+ dc->caps.color.dpp.ogam_rom_caps.pq = 0;
+ dc->caps.color.dpp.ogam_rom_caps.hlg = 0;
+ dc->caps.color.dpp.ocsc = 0;
+
+ dc->caps.color.mpc.gamut_remap = 1;
+ dc->caps.color.mpc.num_3dluts = pool->base.res_cap->num_mpc_3dlut; //2
+ dc->caps.color.mpc.ogam_ram = 1;
+ dc->caps.color.mpc.ogam_rom_caps.srgb = 0;
+ dc->caps.color.mpc.ogam_rom_caps.bt2020 = 0;
+ dc->caps.color.mpc.ogam_rom_caps.gamma2_2 = 0;
+ dc->caps.color.mpc.ogam_rom_caps.pq = 0;
+ dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
+ dc->caps.color.mpc.ocsc = 1;
+ dc->caps.color.mpc.preblend = true;
+
+ dc->caps.num_of_host_routers = 2;
+ dc->caps.num_of_dpias_per_host_router = 2;
+
+ dc->caps.num_of_host_routers = 2;
+ dc->caps.num_of_dpias_per_host_router = 2;
+
+ /* max_disp_clock_khz_at_vmin is slightly lower than the STA value in order
+ * to provide some margin.
+ * It's expected for furture ASIC to have equal or higher value, in order to
+ * have determinstic power improvement from generate to genration.
+ * (i.e., we should not expect new ASIC generation with lower vmin rate)
+ */
+ dc->caps.max_disp_clock_khz_at_vmin = 650000;
+
+ /* Sequential ONO is based on ASIC. */
+ if (dc->ctx->asic_id.hw_internal_rev >= 0x40)
+ dc->caps.sequential_ono = true;
+
+ /* Use pipe context based otg sync logic */
+ dc->config.use_pipe_ctx_sync_logic = true;
+
+ dc->config.disable_hbr_audio_dp2 = true;
+ /* read VBIOS LTTPR caps */
+ {
+ if (ctx->dc_bios->funcs->get_lttpr_caps) {
+ enum bp_result bp_query_result;
+ uint8_t is_vbios_lttpr_enable = 0;
+
+ bp_query_result = ctx->dc_bios->funcs->get_lttpr_caps(ctx->dc_bios, &is_vbios_lttpr_enable);
+ dc->caps.vbios_lttpr_enable = (bp_query_result == BP_RESULT_OK) && !!is_vbios_lttpr_enable;
+ }
+
+ /* interop bit is implicit */
+ {
+ dc->caps.vbios_lttpr_aware = true;
+ }
+ }
+
+ if (dc->ctx->dce_environment == DCE_ENV_PRODUCTION_DRV)
+ dc->debug = debug_defaults_drv;
+
+ /*HW default is to have all the FGCG enabled, SW no need to program them*/
+ dc->debug.enable_fine_grain_clock_gating.u32All = 0xFFFF;
+ // Init the vm_helper
+ if (dc->vm_helper)
+ vm_helper_init(dc->vm_helper, 16);
+
+ /*************************************************
+ * Create resources *
+ *************************************************/
+
+ /* Clock Sources for Pixel Clock*/
+ pool->base.clock_sources[DCN36_CLK_SRC_PLL0] =
+ dcn35_clock_source_create(ctx, ctx->dc_bios,
+ CLOCK_SOURCE_COMBO_PHY_PLL0,
+ &clk_src_regs[0], false);
+ pool->base.clock_sources[DCN36_CLK_SRC_PLL1] =
+ dcn35_clock_source_create(ctx, ctx->dc_bios,
+ CLOCK_SOURCE_COMBO_PHY_PLL1,
+ &clk_src_regs[1], false);
+ pool->base.clock_sources[DCN36_CLK_SRC_PLL2] =
+ dcn35_clock_source_create(ctx, ctx->dc_bios,
+ CLOCK_SOURCE_COMBO_PHY_PLL2,
+ &clk_src_regs[2], false);
+ pool->base.clock_sources[DCN36_CLK_SRC_PLL3] =
+ dcn35_clock_source_create(ctx, ctx->dc_bios,
+ CLOCK_SOURCE_COMBO_PHY_PLL3,
+ &clk_src_regs[3], false);
+ pool->base.clock_sources[DCN36_CLK_SRC_PLL4] =
+ dcn35_clock_source_create(ctx, ctx->dc_bios,
+ CLOCK_SOURCE_COMBO_PHY_PLL4,
+ &clk_src_regs[4], false);
+
+ pool->base.clk_src_count = DCN36_CLK_SRC_TOTAL;
+
+ /* todo: not reuse phy_pll registers */
+ pool->base.dp_clock_source =
+ dcn35_clock_source_create(ctx, ctx->dc_bios,
+ CLOCK_SOURCE_ID_DP_DTO,
+ &clk_src_regs[0], true);
+
+ for (i = 0; i < pool->base.clk_src_count; i++) {
+ if (pool->base.clock_sources[i] == NULL) {
+ dm_error("DC: failed to create clock sources!\n");
+ BREAK_TO_DEBUGGER();
+ goto create_fail;
+ }
+ }
+ /*temp till dml2 fully work without dml1*/
+ dml_init_instance(&dc->dml, &dcn3_5_soc, &dcn3_5_ip, DML_PROJECT_DCN31);
+
+ /* TODO: DCCG */
+ pool->base.dccg = dccg35_create(ctx, &dccg_regs, &dccg_shift, &dccg_mask);
+ if (pool->base.dccg == NULL) {
+ dm_error("DC: failed to create dccg!\n");
+ BREAK_TO_DEBUGGER();
+ goto create_fail;
+ }
+
+#undef REG_STRUCT
+#define REG_STRUCT pg_cntl_regs
+ pg_cntl_dcn35_regs_init();
+
+ pool->base.pg_cntl = pg_cntl35_create(ctx, &pg_cntl_regs, &pg_cntl_shift, &pg_cntl_mask);
+ if (pool->base.pg_cntl == NULL) {
+ dm_error("DC: failed to create power gate control!\n");
+ BREAK_TO_DEBUGGER();
+ goto create_fail;
+ }
+
+ /* TODO: IRQ */
+ init_data.ctx = dc->ctx;
+ pool->base.irqs = dal_irq_service_dcn36_create(&init_data);
+ if (!pool->base.irqs)
+ goto create_fail;
+
+ /* HUBBUB */
+ pool->base.hubbub = dcn35_hubbub_create(ctx);
+ if (pool->base.hubbub == NULL) {
+ BREAK_TO_DEBUGGER();
+ dm_error("DC: failed to create hubbub!\n");
+ goto create_fail;
+ }
+
+ /* HUBPs, DPPs, OPPs and TGs */
+ for (i = 0; i < pool->base.pipe_count; i++) {
+ pool->base.hubps[i] = dcn35_hubp_create(ctx, i);
+ if (pool->base.hubps[i] == NULL) {
+ BREAK_TO_DEBUGGER();
+ dm_error(
+ "DC: failed to create hubps!\n");
+ goto create_fail;
+ }
+
+ pool->base.dpps[i] = dcn35_dpp_create(ctx, i);
+ if (pool->base.dpps[i] == NULL) {
+ BREAK_TO_DEBUGGER();
+ dm_error(
+ "DC: failed to create dpps!\n");
+ goto create_fail;
+ }
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_opp; i++) {
+ pool->base.opps[i] = dcn35_opp_create(ctx, i);
+ if (pool->base.opps[i] == NULL) {
+ BREAK_TO_DEBUGGER();
+ dm_error(
+ "DC: failed to create output pixel processor!\n");
+ goto create_fail;
+ }
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_timing_generator; i++) {
+ pool->base.timing_generators[i] = dcn35_timing_generator_create(
+ ctx, i);
+ if (pool->base.timing_generators[i] == NULL) {
+ BREAK_TO_DEBUGGER();
+ dm_error("DC: failed to create tg!\n");
+ goto create_fail;
+ }
+ }
+ pool->base.timing_generator_count = i;
+
+ /* PSR */
+ pool->base.psr = dmub_psr_create(ctx);
+ if (pool->base.psr == NULL) {
+ dm_error("DC: failed to create psr obj!\n");
+ BREAK_TO_DEBUGGER();
+ goto create_fail;
+ }
+
+ /* Replay */
+ pool->base.replay = dmub_replay_create(ctx);
+ if (pool->base.replay == NULL) {
+ dm_error("DC: failed to create replay obj!\n");
+ BREAK_TO_DEBUGGER();
+ goto create_fail;
+ }
+
+ /* ABM */
+ for (i = 0; i < pool->base.res_cap->num_timing_generator; i++) {
+ pool->base.multiple_abms[i] = dmub_abm_create(ctx,
+ &abm_regs[i],
+ &abm_shift,
+ &abm_mask);
+ if (pool->base.multiple_abms[i] == NULL) {
+ dm_error("DC: failed to create abm for pipe %d!\n", i);
+ BREAK_TO_DEBUGGER();
+ goto create_fail;
+ }
+ }
+
+ /* MPC and DSC */
+ pool->base.mpc = dcn35_mpc_create(ctx, pool->base.mpcc_count, pool->base.res_cap->num_mpc_3dlut);
+ if (pool->base.mpc == NULL) {
+ BREAK_TO_DEBUGGER();
+ dm_error("DC: failed to create mpc!\n");
+ goto create_fail;
+ }
+
+ for (i = 0; i < pool->base.res_cap->num_dsc; i++) {
+ pool->base.dscs[i] = dcn35_dsc_create(ctx, i);
+ if (pool->base.dscs[i] == NULL) {
+ BREAK_TO_DEBUGGER();
+ dm_error("DC: failed to create display stream compressor %d!\n", i);
+ goto create_fail;
+ }
+ }
+
+ /* DWB and MMHUBBUB */
+ if (!dcn35_dwbc_create(ctx, &pool->base)) {
+ BREAK_TO_DEBUGGER();
+ dm_error("DC: failed to create dwbc!\n");
+ goto create_fail;
+ }
+
+ if (!dcn35_mmhubbub_create(ctx, &pool->base)) {
+ BREAK_TO_DEBUGGER();
+ dm_error("DC: failed to create mcif_wb!\n");
+ goto create_fail;
+ }
+
+ /* AUX and I2C */
+ for (i = 0; i < pool->base.res_cap->num_ddc; i++) {
+ pool->base.engines[i] = dcn31_aux_engine_create(ctx, i);
+ if (pool->base.engines[i] == NULL) {
+ BREAK_TO_DEBUGGER();
+ dm_error(
+ "DC:failed to create aux engine!!\n");
+ goto create_fail;
+ }
+ pool->base.hw_i2cs[i] = dcn31_i2c_hw_create(ctx, i);
+ if (pool->base.hw_i2cs[i] == NULL) {
+ BREAK_TO_DEBUGGER();
+ dm_error(
+ "DC:failed to create hw i2c!!\n");
+ goto create_fail;
+ }
+ pool->base.sw_i2cs[i] = NULL;
+ }
+
+ /* DCN3.5 has 6 DPIA */
+ pool->base.usb4_dpia_count = 4;
+ if (dc->debug.dpia_debug.bits.disable_dpia)
+ pool->base.usb4_dpia_count = 0;
+
+ /* Audio, Stream Encoders including HPO and virtual, MPC 3D LUTs */
+ if (!resource_construct(num_virtual_links, dc, &pool->base,
+ &res_create_funcs))
+ goto create_fail;
+
+ /* HW Sequencer and Plane caps */
+ dcn35_hw_sequencer_construct(dc);
+
+ dc->caps.max_planes = pool->base.pipe_count;
+
+ for (i = 0; i < dc->caps.max_planes; ++i)
+ dc->caps.planes[i] = plane_cap;
+
+ dc->caps.max_odm_combine_factor = 4;
+
+ dc->cap_funcs = cap_funcs;
+
+ dc->dcn_ip->max_num_dpp = pool->base.pipe_count;
+
+ dc->dml2_options.dcn_pipe_count = pool->base.pipe_count;
+ dc->dml2_options.use_native_soc_bb_construction = true;
+ dc->dml2_options.minimize_dispclk_using_odm = false;
+ if (dc->config.EnableMinDispClkODM)
+ dc->dml2_options.minimize_dispclk_using_odm = true;
+ dc->dml2_options.enable_windowed_mpo_odm = dc->config.enable_windowed_mpo_odm;
+
+ resource_init_common_dml2_callbacks(dc, &dc->dml2_options);
+ dc->dml2_options.callbacks.can_support_mclk_switch_using_fw_based_vblank_stretch = &dcn30_can_support_mclk_switch_using_fw_based_vblank_stretch;
+
+ dc->dml2_options.max_segments_per_hubp = 24;
+ dc->dml2_options.det_segment_size = DCN3_2_DET_SEG_SIZE;/*todo*/
+ dc->dml2_options.override_det_buffer_size_kbytes = true;
+
+ if (dc->config.sdpif_request_limit_words_per_umc == 0)
+ dc->config.sdpif_request_limit_words_per_umc = 16;/*todo*/
+
+ return true;
+
+create_fail:
+
+ dcn36_resource_destruct(pool);
+
+ return false;
+}
+
+struct resource_pool *dcn36_create_resource_pool(
+ const struct dc_init_data *init_data,
+ struct dc *dc)
+{
+ struct dcn36_resource_pool *pool =
+ kzalloc(sizeof(struct dcn36_resource_pool), GFP_KERNEL);
+
+ if (!pool)
+ return NULL;
+
+ if (dcn36_resource_construct(init_data->num_virtual_links, dc, pool))
+ return &pool->base;
+
+ BREAK_TO_DEBUGGER();
+ kfree(pool);
+ return NULL;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn36/dcn36_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn36/dcn36_resource.h
new file mode 100644
index 000000000000..5490c9975e23
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn36/dcn36_resource.h
@@ -0,0 +1,73 @@
+/* SPDX-License-Identifier: MIT */
+/* Copyright 2025 Advanced Micro Devices, Inc. */
+
+#ifndef _DCN36_RESOURCE_H_
+#define _DCN36_RESOURCE_H_
+
+#include "core_types.h"
+
+extern struct _vcs_dpi_ip_params_st dcn3_6_ip;
+extern struct _vcs_dpi_soc_bounding_box_st dcn3_6_soc;
+
+#define TO_DCN36_RES_POOL(pool)\
+ container_of(pool, struct dcn36_resource_pool, base)
+
+struct dcn36_resource_pool {
+ struct resource_pool base;
+};
+
+struct resource_pool *dcn36_create_resource_pool(
+ const struct dc_init_data *init_data,
+ struct dc *dc);
+
+#define HWSEQ_DCN36_REG_LIST()\
+ SR(DCHUBBUB_GLOBAL_TIMER_CNTL), \
+ SR(DCHUBBUB_ARB_HOSTVM_CNTL), \
+ SR(DIO_MEM_PWR_CTRL), \
+ SR(ODM_MEM_PWR_CTRL3), \
+ SR(MMHUBBUB_MEM_PWR_CNTL), \
+ SR(DCCG_GATE_DISABLE_CNTL), \
+ SR(DCCG_GATE_DISABLE_CNTL2), \
+ SR(DCCG_GATE_DISABLE_CNTL4), \
+ SR(DCCG_GATE_DISABLE_CNTL5), \
+ SR(DCFCLK_CNTL),\
+ SR(DC_MEM_GLOBAL_PWR_REQ_CNTL), \
+ SRII(PIXEL_RATE_CNTL, OTG, 0), \
+ SRII(PIXEL_RATE_CNTL, OTG, 1),\
+ SRII(PIXEL_RATE_CNTL, OTG, 2),\
+ SRII(PIXEL_RATE_CNTL, OTG, 3),\
+ SRII(PHYPLL_PIXEL_RATE_CNTL, OTG, 0),\
+ SRII(PHYPLL_PIXEL_RATE_CNTL, OTG, 1),\
+ SRII(PHYPLL_PIXEL_RATE_CNTL, OTG, 2),\
+ SRII(PHYPLL_PIXEL_RATE_CNTL, OTG, 3),\
+ SR(MICROSECOND_TIME_BASE_DIV), \
+ SR(MILLISECOND_TIME_BASE_DIV), \
+ SR(DISPCLK_FREQ_CHANGE_CNTL), \
+ SR(RBBMIF_TIMEOUT_DIS), \
+ SR(RBBMIF_TIMEOUT_DIS_2), \
+ SR(DCHUBBUB_CRC_CTRL), \
+ SR(DPP_TOP0_DPP_CRC_CTRL), \
+ SR(MPC_CRC_CTRL), \
+ SR(DOMAIN0_PG_CONFIG), \
+ SR(DOMAIN1_PG_CONFIG), \
+ SR(DOMAIN2_PG_CONFIG), \
+ SR(DOMAIN3_PG_CONFIG), \
+ SR(DOMAIN16_PG_CONFIG), \
+ SR(DOMAIN17_PG_CONFIG), \
+ SR(DOMAIN18_PG_CONFIG), \
+ SR(DOMAIN19_PG_CONFIG), \
+ SR(DOMAIN0_PG_STATUS), \
+ SR(DOMAIN1_PG_STATUS), \
+ SR(DOMAIN2_PG_STATUS), \
+ SR(DOMAIN3_PG_STATUS), \
+ SR(DOMAIN16_PG_STATUS), \
+ SR(DOMAIN17_PG_STATUS), \
+ SR(DOMAIN18_PG_STATUS), \
+ SR(DOMAIN19_PG_STATUS), \
+ SR(DC_IP_REQUEST_CNTL), \
+ SR(AZALIA_AUDIO_DTO), \
+ SR(AZALIA_CONTROLLER_CLOCK_GATING), \
+ SR(HPO_TOP_HW_CONTROL),\
+ SR(DMU_CLK_CNTL)
+
+#endif /* _DCN36_RESOURCE_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c
index db93bac247c0..b3988e38d0a6 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.c
@@ -70,15 +70,11 @@
#include "dml/dcn30/display_mode_vba_30.h"
#include "vm_helper.h"
#include "dcn20/dcn20_vmid.h"
-#include "dml/dcn401/dcn401_fpu.h"
#include "dc_state_priv.h"
#include "dml2/dml2_wrapper.h"
-#include "spl/dc_spl_scl_easf_filters.h"
-#include "spl/dc_spl_isharp_filters.h"
-
#define DC_LOGGER_INIT(logger)
enum dcn401_clk_src_array_id {
@@ -726,6 +722,10 @@ static const struct dc_debug_options debug_defaults_drv = {
.disable_unbounded_requesting = false,
.enable_legacy_fast_update = false,
.dcc_meta_propagation_delay_us = 10,
+ .fams_version = {
+ .minor = 1,
+ .major = 2,
+ }, //v2.1
.fams2_config = {
.bits = {
.enable = true,
@@ -733,7 +733,7 @@ static const struct dc_debug_options debug_defaults_drv = {
.enable_stall_recovery = true,
}
},
- .force_cositing = CHROMA_COSITING_TOPLEFT + 1,
+ .force_cositing = CHROMA_COSITING_NONE + 1,
};
static struct dce_aux *dcn401_aux_engine_create(
@@ -1293,6 +1293,29 @@ static struct hpo_dp_link_encoder *dcn401_hpo_dp_link_encoder_create(
return &hpo_dp_enc31->base;
}
+static unsigned int dcn401_calc_num_avail_chans_for_mall(struct dc *dc, unsigned int num_chans)
+{
+ unsigned int num_available_chans = 1;
+
+ /* channels for MALL must be a power of 2 */
+ while (num_chans > 1) {
+ num_available_chans = (num_available_chans << 1);
+ num_chans = (num_chans >> 1);
+ }
+
+ /* cannot be odd */
+ num_available_chans &= ~1;
+
+ /* clamp to max available channels for MALL per ASIC */
+ if (ASICREV_IS_GC_12_0_0_A0(dc->ctx->asic_id.hw_internal_rev)) {
+ num_available_chans = num_available_chans > 16 ? 16 : num_available_chans;
+ } else if (ASICREV_IS_GC_12_0_1_A0(dc->ctx->asic_id.hw_internal_rev)) {
+ num_available_chans = num_available_chans > 8 ? 8 : num_available_chans;
+ }
+
+ return num_available_chans;
+}
+
static struct dce_hwseq *dcn401_hwseq_create(
struct dc_context *ctx)
{
@@ -1584,41 +1607,78 @@ static struct dc_cap_funcs cap_funcs = {
static void dcn401_update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_params)
{
- struct dml2_configuration_options *dml2_opt = &dc->dml2_tmp;
-
- memcpy(dml2_opt, &dc->dml2_options, sizeof(dc->dml2_options));
+ /* re-calculate the available MALL size if required */
+ if (bw_params->num_channels > 0) {
+ dc->caps.max_cab_allocation_bytes = dcn401_calc_num_avail_chans_for_mall(
+ dc, bw_params->num_channels) *
+ dc->caps.mall_size_per_mem_channel * 1024 * 1024;
+ dc->caps.mall_size_total = dc->caps.max_cab_allocation_bytes;
+ }
DC_FP_START();
- dcn401_update_bw_bounding_box_fpu(dc, bw_params);
-
- dml2_opt->use_clock_dc_limits = false;
if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2)
- dml2_reinit(dc, dml2_opt, &dc->current_state->bw_ctx.dml2);
+ dml2_reinit(dc, &dc->dml2_options, &dc->current_state->bw_ctx.dml2);
- dml2_opt->use_clock_dc_limits = true;
if (dc->debug.using_dml2 && dc->current_state && dc->current_state->bw_ctx.dml2_dc_power_source)
- dml2_reinit(dc, dml2_opt, &dc->current_state->bw_ctx.dml2_dc_power_source);
+ dml2_reinit(dc, &dc->dml2_dc_power_options, &dc->current_state->bw_ctx.dml2_dc_power_source);
DC_FP_END();
}
enum dc_status dcn401_patch_unknown_plane_state(struct dc_plane_state *plane_state)
{
+ plane_state->tiling_info.gfxversion = DcGfxAddr3;
plane_state->tiling_info.gfx_addr3.swizzle = DC_ADDR3_SW_64KB_2D;
return DC_OK;
}
-bool dcn401_validate_bandwidth(struct dc *dc,
+enum dc_status dcn401_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate)
+ enum dc_validate_mode validate_mode)
{
- bool out = false;
+ unsigned int i;
+ enum dc_status status = DC_OK;
+ const struct dc_stream_state *stream;
+
+ /* reset cursor limitations on subvp */
+ for (i = 0; i < context->stream_count; i++) {
+ stream = context->streams[i];
+
+ if (dc_state_can_clear_stream_cursor_subvp_limit(stream, context)) {
+ dc_state_set_stream_cursor_subvp_limit(stream, context, false);
+ }
+ }
+
if (dc->debug.using_dml2)
- out = dml2_validate(dc, context,
+ status = dml2_validate(dc, context,
context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2,
- fast_validate);
- return out;
+ validate_mode) ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
+
+ if (validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING && status == DC_OK && dc_state_is_subvp_in_use(context)) {
+ /* check new stream configuration still supports cursor if subvp used */
+ for (i = 0; i < context->stream_count; i++) {
+ stream = context->streams[i];
+
+ if (dc_state_get_stream_subvp_type(context, stream) != SUBVP_PHANTOM &&
+ stream->cursor_position.enable &&
+ !dc_stream_check_cursor_attributes(stream, context, &stream->cursor_attributes)) {
+ /* hw cursor cannot be supported with subvp active, so disable subvp for now */
+ dc_state_set_stream_cursor_subvp_limit(stream, context, true);
+ status = DC_FAIL_HW_CURSOR_SUPPORT;
+ }
+ };
+ }
+
+ if (validate_mode == DC_VALIDATE_MODE_AND_PROGRAMMING && status == DC_FAIL_HW_CURSOR_SUPPORT) {
+ /* attempt to validate again with subvp disabled due to cursor */
+ if (dc->debug.using_dml2)
+ status = dml2_validate(dc, context,
+ context->power_source == DC_POWER_SOURCE_DC ? context->bw_ctx.dml2_dc_power_source : context->bw_ctx.dml2,
+ validate_mode) ? DC_OK : DC_FAIL_BANDWIDTH_VALIDATE;
+ }
+
+ return status;
}
void dcn401_prepare_mcache_programming(struct dc *dc,
@@ -1633,12 +1693,13 @@ static void dcn401_build_pipe_pix_clk_params(struct pipe_ctx *pipe_ctx)
{
const struct dc_stream_state *stream = pipe_ctx->stream;
struct dc_link *link = stream->link;
- struct link_encoder *link_enc = NULL;
+ struct link_encoder *link_enc = pipe_ctx->link_res.dio_link_enc;
struct pixel_clk_params *pixel_clk_params = &pipe_ctx->stream_res.pix_clk_params;
pixel_clk_params->requested_pix_clk_100hz = stream->timing.pix_clk_100hz;
- link_enc = link_enc_cfg_get_link_enc(link);
+ if (!pipe_ctx->stream->ctx->dc->config.unify_link_enc_assignment)
+ link_enc = link_enc_cfg_get_link_enc(link);
if (link_enc)
pixel_clk_params->encoder_object_id = link_enc->id;
@@ -1704,27 +1765,9 @@ static int dcn401_get_power_profile(const struct dc_state *context)
return dpm_level;
}
-static unsigned int dcn401_calc_num_avail_chans_for_mall(struct dc *dc, unsigned int num_chans)
+static unsigned int dcn401_get_vstartup_for_pipe(struct pipe_ctx *pipe_ctx)
{
- unsigned int num_available_chans = 1;
-
- /* channels for MALL must be a power of 2 */
- while (num_chans > 1) {
- num_available_chans = (num_available_chans << 1);
- num_chans = (num_chans >> 1);
- }
-
- /* cannot be odd */
- num_available_chans &= ~1;
-
- /* clamp to max available channels for MALL per ASIC */
- if (ASICREV_IS_GC_12_0_0_A0(dc->ctx->asic_id.hw_internal_rev)) {
- num_available_chans = num_available_chans > 16 ? 16 : num_available_chans;
- } else if (ASICREV_IS_GC_12_0_1_A0(dc->ctx->asic_id.hw_internal_rev)) {
- num_available_chans = num_available_chans > 8 ? 8 : num_available_chans;
- }
-
- return num_available_chans;
+ return pipe_ctx->global_sync.dcn4x.vstartup_lines;
}
static struct resource_funcs dcn401_res_pool_funcs = {
@@ -1754,6 +1797,8 @@ static struct resource_funcs dcn401_res_pool_funcs = {
.build_pipe_pix_clk_params = dcn401_build_pipe_pix_clk_params,
.calculate_mall_ways_from_bytes = dcn32_calculate_mall_ways_from_bytes,
.get_power_profile = dcn401_get_power_profile,
+ .get_vstartup_for_pipe = dcn401_get_vstartup_for_pipe,
+ .get_max_hw_cursor_size = dcn32_get_max_hw_cursor_size
};
static uint32_t read_pipe_fuses(struct dc_context *ctx)
@@ -1829,8 +1874,9 @@ static bool dcn401_resource_construct(
dc->caps.max_downscale_ratio = 600;
dc->caps.i2c_speed_in_khz = 95;
dc->caps.i2c_speed_in_khz_hdcp = 95; /*1.4 w/a applied by default*/
- /* TODO: Bring max cursor size back to 256 after subvp cursor corruption is fixed*/
+ /* used to set cursor pitch, so must be aligned to power of 2 (HW actually supported 78x78) */
dc->caps.max_cursor_size = 64;
+ dc->caps.max_buffered_cursor_size = 64;
dc->caps.cursor_not_scaled = true;
dc->caps.min_horizontal_blanking_period = 80;
dc->caps.dmdata_alloc_size = 2048;
@@ -1853,9 +1899,9 @@ static bool dcn401_resource_construct(
dc->caps.subvp_vertical_int_margin_us = 30;
dc->caps.subvp_drr_vblank_start_margin_us = 100; // 100us margin
- dc->caps.max_slave_planes = 2;
- dc->caps.max_slave_yuv_planes = 2;
- dc->caps.max_slave_rgb_planes = 2;
+ dc->caps.max_slave_planes = 3;
+ dc->caps.max_slave_yuv_planes = 3;
+ dc->caps.max_slave_rgb_planes = 3;
dc->caps.post_blend_color_processing = true;
dc->caps.force_dp_tps4_for_cp2520 = true;
dc->caps.dp_hpo = true;
@@ -1864,6 +1910,7 @@ static bool dcn401_resource_construct(
dc->caps.extended_aux_timeout_support = true;
dc->caps.dmcub_support = true;
dc->caps.max_v_total = (1 << 15) - 1;
+ dc->caps.vtotal_limited_by_fp2 = true;
if (ASICREV_IS_GC_12_0_1_A0(dc->ctx->asic_id.hw_internal_rev))
dc->caps.dcc_plane_width_limit = 7680;
@@ -1882,8 +1929,8 @@ static bool dcn401_resource_construct(
dc->caps.color.dpp.gamma_corr = 1;
dc->caps.color.dpp.dgam_rom_for_yuv = 0;
- dc->caps.color.dpp.hw_3d_lut = 1;
- dc->caps.color.dpp.ogam_ram = 1;
+ dc->caps.color.dpp.hw_3d_lut = 0;
+ dc->caps.color.dpp.ogam_ram = 0;
// no OGAM ROM on DCN2 and later ASICs
dc->caps.color.dpp.ogam_rom_caps.srgb = 0;
dc->caps.color.dpp.ogam_rom_caps.bt2020 = 0;
@@ -1901,11 +1948,34 @@ static bool dcn401_resource_construct(
dc->caps.color.mpc.ogam_rom_caps.pq = 0;
dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
dc->caps.color.mpc.ocsc = 1;
+ dc->caps.color.mpc.preblend = true;
dc->config.use_spl = true;
dc->config.prefer_easf = true;
+
+ dc->config.dcn_sharpness_range.sdr_rgb_min = 0;
+ dc->config.dcn_sharpness_range.sdr_rgb_max = 1750;
+ dc->config.dcn_sharpness_range.sdr_rgb_mid = 750;
+ dc->config.dcn_sharpness_range.sdr_yuv_min = 0;
+ dc->config.dcn_sharpness_range.sdr_yuv_max = 3500;
+ dc->config.dcn_sharpness_range.sdr_yuv_mid = 1500;
+ dc->config.dcn_sharpness_range.hdr_rgb_min = 0;
+ dc->config.dcn_sharpness_range.hdr_rgb_max = 2750;
+ dc->config.dcn_sharpness_range.hdr_rgb_mid = 1500;
+
+ dc->config.dcn_override_sharpness_range.sdr_rgb_min = 0;
+ dc->config.dcn_override_sharpness_range.sdr_rgb_max = 3250;
+ dc->config.dcn_override_sharpness_range.sdr_rgb_mid = 1250;
+ dc->config.dcn_override_sharpness_range.sdr_yuv_min = 0;
+ dc->config.dcn_override_sharpness_range.sdr_yuv_max = 3500;
+ dc->config.dcn_override_sharpness_range.sdr_yuv_mid = 1500;
+ dc->config.dcn_override_sharpness_range.hdr_rgb_min = 0;
+ dc->config.dcn_override_sharpness_range.hdr_rgb_max = 2750;
+ dc->config.dcn_override_sharpness_range.hdr_rgb_mid = 1500;
+
dc->config.dc_mode_clk_limit_support = true;
dc->config.enable_windowed_mpo_odm = true;
dc->config.set_pipe_unlock_order = true; /* Need to ensure DET gets freed before allocating */
+
/* read VBIOS LTTPR caps */
{
if (ctx->dc_bios->funcs->get_lttpr_caps) {
@@ -2120,6 +2190,8 @@ static bool dcn401_resource_construct(
for (i = 0; i < dc->caps.max_planes; ++i)
dc->caps.planes[i] = plane_cap;
+ dc->caps.max_odm_combine_factor = 4;
+
dc->cap_funcs = cap_funcs;
if (dc->ctx->dc_bios->fw_info.oem_i2c_present) {
@@ -2138,7 +2210,6 @@ static bool dcn401_resource_construct(
dc->config.sdpif_request_limit_words_per_umc = 16;
dc->dml2_options.dcn_pipe_count = pool->base.pipe_count;
- dc->dml2_options.use_native_pstate_optimization = false;
dc->dml2_options.use_native_soc_bb_construction = true;
dc->dml2_options.minimize_dispclk_using_odm = true;
dc->dml2_options.map_dc_pipes_with_callbacks = true;
@@ -2169,10 +2240,12 @@ static bool dcn401_resource_construct(
dc->dml2_options.det_segment_size = DCN4_01_CRB_SEGMENT_SIZE_KB;
/* SPL */
- spl_init_easf_filter_coeffs();
- spl_init_blur_scale_coeffs();
dc->caps.scl_caps.sharpener_support = true;
+ /* init DC limited DML2 options */
+ memcpy(&dc->dml2_dc_power_options, &dc->dml2_options, sizeof(struct dml2_configuration_options));
+ dc->dml2_dc_power_options.use_clock_dc_limits = true;
+
return true;
create_fail:
diff --git a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h
index 19568c359669..2ae6831c31ef 100644
--- a/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h
+++ b/drivers/gpu/drm/amd/display/dc/resource/dcn401/dcn401_resource.h
@@ -22,9 +22,9 @@ struct resource_pool *dcn401_create_resource_pool(
enum dc_status dcn401_patch_unknown_plane_state(struct dc_plane_state *plane_state);
-bool dcn401_validate_bandwidth(struct dc *dc,
+enum dc_status dcn401_validate_bandwidth(struct dc *dc,
struct dc_state *context,
- bool fast_validate);
+ enum dc_validate_mode validate_mode);
void dcn401_prepare_mcache_programming(struct dc *dc, struct dc_state *context);
@@ -538,7 +538,8 @@ void dcn401_prepare_mcache_programming(struct dc *dc, struct dc_state *context);
SRI_ARR(OPTC_MEMORY_CONFIG, ODM, inst), \
SRI_ARR(OTG_DRR_CONTROL, OTG, inst), \
SRI_ARR(OTG_PSTATE_REGISTER, OTG, inst), \
- SRI_ARR(OTG_PIPE_UPDATE_STATUS, OTG, inst)
+ SRI_ARR(OTG_PIPE_UPDATE_STATUS, OTG, inst), \
+ SRI_ARR(INTERRUPT_DEST, OTG, inst)
/* HUBBUB */
#define HUBBUB_REG_LIST_DCN4_01_RI(id) \
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl.h b/drivers/gpu/drm/amd/display/dc/spl/dc_spl.h
deleted file mode 100644
index 02a2d6725ed5..000000000000
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl.h
+++ /dev/null
@@ -1,18 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-#ifndef __DC_SPL_H__
-#define __DC_SPL_H__
-
-#include "dc_spl_types.h"
-#define BLACK_OFFSET_RGB_Y 0x0
-#define BLACK_OFFSET_CBCR 0x8000
-
-/* SPL interfaces */
-
-bool spl_calculate_scaler_params(struct spl_in *spl_in, struct spl_out *spl_out);
-
-bool spl_get_number_of_taps(struct spl_in *spl_in, struct spl_out *spl_out);
-
-#endif /* __DC_SPL_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_filters.h b/drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_filters.h
deleted file mode 100644
index 48202bc4f81e..000000000000
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_filters.h
+++ /dev/null
@@ -1,22 +0,0 @@
-// SPDX-License-Identifier: MIT
-//
-// Copyright 2024 Advanced Micro Devices, Inc.
-
-#ifndef __DC_SPL_SCL_FILTERS_H__
-#define __DC_SPL_SCL_FILTERS_H__
-
-#include "dc_spl_types.h"
-
-const uint16_t *spl_get_filter_3tap_16p(struct spl_fixed31_32 ratio);
-const uint16_t *spl_get_filter_3tap_64p(struct spl_fixed31_32 ratio);
-const uint16_t *spl_get_filter_4tap_16p(struct spl_fixed31_32 ratio);
-const uint16_t *spl_get_filter_4tap_64p(struct spl_fixed31_32 ratio);
-const uint16_t *spl_get_filter_5tap_64p(struct spl_fixed31_32 ratio);
-const uint16_t *spl_get_filter_6tap_64p(struct spl_fixed31_32 ratio);
-const uint16_t *spl_get_filter_7tap_64p(struct spl_fixed31_32 ratio);
-const uint16_t *spl_get_filter_8tap_64p(struct spl_fixed31_32 ratio);
-const uint16_t *spl_get_filter_2tap_16p(void);
-const uint16_t *spl_get_filter_2tap_64p(void);
-const uint16_t *spl_dscl_get_filter_coeffs_64p(int taps, struct spl_fixed31_32 ratio);
-
-#endif /* __DC_SPL_SCL_FILTERS_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/spl/Makefile b/drivers/gpu/drm/amd/display/dc/sspl/Makefile
index 5edf3c6cf3e2..5e3e4aa13820 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/sspl/Makefile
@@ -25,7 +25,7 @@
SPL = dc_spl.o dc_spl_scl_filters.o dc_spl_scl_easf_filters.o dc_spl_isharp_filters.o dc_spl_filters.o spl_fixpt31_32.o spl_custom_float.o
-AMD_DAL_SPL = $(addprefix $(AMDDALPATH)/dc/spl/,$(SPL))
+AMD_DAL_SPL = $(addprefix $(AMDDALPATH)/dc/sspl/,$(SPL))
AMD_DISPLAY_FILES += $(AMD_DAL_SPL)
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl.c
index 73a65913cb12..55b929ca7982 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl.c
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl.c
@@ -3,14 +3,48 @@
// Copyright 2024 Advanced Micro Devices, Inc.
#include "dc_spl.h"
-#include "dc_spl_scl_filters.h"
#include "dc_spl_scl_easf_filters.h"
#include "dc_spl_isharp_filters.h"
#include "spl_debug.h"
-#define IDENTITY_RATIO(ratio) (spl_fixpt_u2d19(ratio) == (1 << 19))
+#define IDENTITY_RATIO(ratio) (spl_fixpt_u3d19(ratio) == (1 << 19))
#define MIN_VIEWPORT_SIZE 12
+static bool spl_is_yuv420(enum spl_pixel_format format)
+{
+ if ((format >= SPL_PIXEL_FORMAT_420BPP8) &&
+ (format <= SPL_PIXEL_FORMAT_420BPP10))
+ return true;
+
+ return false;
+}
+
+static bool spl_is_rgb8(enum spl_pixel_format format)
+{
+ if (format == SPL_PIXEL_FORMAT_ARGB8888)
+ return true;
+
+ return false;
+}
+
+static bool spl_is_video_format(enum spl_pixel_format format)
+{
+ if (format >= SPL_PIXEL_FORMAT_VIDEO_BEGIN
+ && format <= SPL_PIXEL_FORMAT_VIDEO_END)
+ return true;
+ else
+ return false;
+}
+
+static bool spl_is_subsampled_format(enum spl_pixel_format format)
+{
+ if (format >= SPL_PIXEL_FORMAT_SUBSAMPLED_BEGIN
+ && format <= SPL_PIXEL_FORMAT_SUBSAMPLED_END)
+ return true;
+ else
+ return false;
+}
+
static struct spl_rect intersect_rec(const struct spl_rect *r0, const struct spl_rect *r1)
{
struct spl_rect rec;
@@ -41,6 +75,21 @@ static struct spl_rect shift_rec(const struct spl_rect *rec_in, int x, int y)
return rec_out;
}
+static void spl_opp_adjust_rect(struct spl_rect *rec, const struct spl_opp_adjust *adjust)
+{
+ if ((rec->x + adjust->x) >= 0)
+ rec->x += adjust->x;
+
+ if ((rec->y + adjust->y) >= 0)
+ rec->y += adjust->y;
+
+ if ((rec->width + adjust->width) >= 1)
+ rec->width += adjust->width;
+
+ if ((rec->height + adjust->height) >= 1)
+ rec->height += adjust->height;
+}
+
static struct spl_rect calculate_plane_rec_in_timing_active(
struct spl_in *spl_in,
const struct spl_rect *rec_in)
@@ -137,15 +186,37 @@ static struct spl_rect calculate_mpc_slice_in_timing_active(
struct spl_in *spl_in,
struct spl_rect *plane_clip_rec)
{
- int mpc_slice_count = spl_in->basic_in.mpc_combine_h;
- int mpc_slice_idx = spl_in->basic_in.mpc_combine_v;
+ bool use_recout_width_aligned =
+ spl_in->basic_in.num_h_slices_recout_width_align.use_recout_width_aligned;
+ int mpc_slice_count =
+ spl_in->basic_in.num_h_slices_recout_width_align.num_slices_recout_width.mpc_num_h_slices;
+ int recout_width_align =
+ spl_in->basic_in.num_h_slices_recout_width_align.num_slices_recout_width.mpc_recout_width_align;
+ int mpc_slice_idx = spl_in->basic_in.mpc_h_slice_index;
int epimo = mpc_slice_count - plane_clip_rec->width % mpc_slice_count - 1;
struct spl_rect mpc_rec;
- mpc_rec.width = plane_clip_rec->width / mpc_slice_count;
- mpc_rec.x = plane_clip_rec->x + mpc_rec.width * mpc_slice_idx;
- mpc_rec.height = plane_clip_rec->height;
- mpc_rec.y = plane_clip_rec->y;
+ if (spl_in->basic_in.custom_width != 0) {
+ mpc_rec.width = spl_in->basic_in.custom_width;
+ mpc_rec.x = spl_in->basic_in.custom_x;
+ mpc_rec.height = plane_clip_rec->height;
+ mpc_rec.y = plane_clip_rec->y;
+ } else if (use_recout_width_aligned) {
+ mpc_rec.width = recout_width_align;
+ if ((mpc_rec.width * (mpc_slice_idx + 1)) > plane_clip_rec->width) {
+ mpc_rec.width = plane_clip_rec->width % recout_width_align;
+ mpc_rec.x = plane_clip_rec->x + recout_width_align * mpc_slice_idx;
+ } else
+ mpc_rec.x = plane_clip_rec->x + mpc_rec.width * mpc_slice_idx;
+ mpc_rec.height = plane_clip_rec->height;
+ mpc_rec.y = plane_clip_rec->y;
+
+ } else {
+ mpc_rec.width = plane_clip_rec->width / mpc_slice_count;
+ mpc_rec.x = plane_clip_rec->x + mpc_rec.width * mpc_slice_idx;
+ mpc_rec.height = plane_clip_rec->height;
+ mpc_rec.y = plane_clip_rec->y;
+ }
SPL_ASSERT(mpc_slice_count == 1 ||
spl_in->basic_out.view_format != SPL_VIEW_3D_SIDE_BY_SIDE ||
mpc_rec.width % 2 == 0);
@@ -153,7 +224,7 @@ static struct spl_rect calculate_mpc_slice_in_timing_active(
/* extra pixels in the division remainder need to go to pipes after
* the extra pixel index minus one(epimo) defined here as:
*/
- if (mpc_slice_idx > epimo) {
+ if (mpc_slice_idx > epimo && spl_in->basic_in.custom_width == 0) {
mpc_rec.x += mpc_slice_idx - epimo - 1;
mpc_rec.width += 1;
}
@@ -186,10 +257,10 @@ static struct spl_rect calculate_odm_slice_in_timing_active(struct spl_in *spl_i
odm_rec.x = odm_slice_width * odm_slice_idx;
odm_rec.width = is_last_odm_slice ?
- /* last slice width is the reminder of h_active */
- h_active - odm_slice_width * (odm_slice_count - 1) :
- /* odm slice width is the floor of h_active / count */
- odm_slice_width;
+ /* last slice width is the reminder of h_active */
+ h_active - odm_slice_width * (odm_slice_count - 1) :
+ /* odm slice width is the floor of h_active / count */
+ odm_slice_width;
odm_rec.y = 0;
odm_rec.height = v_active;
@@ -391,8 +462,7 @@ static void spl_calculate_scaling_ratios(struct spl_in *spl_in,
spl_scratch->scl_data.ratios.horz_c = spl_scratch->scl_data.ratios.horz;
spl_scratch->scl_data.ratios.vert_c = spl_scratch->scl_data.ratios.vert;
- if (spl_in->basic_in.format == SPL_PIXEL_FORMAT_420BPP8
- || spl_in->basic_in.format == SPL_PIXEL_FORMAT_420BPP10) {
+ if (spl_is_yuv420(spl_in->basic_in.format)) {
spl_scratch->scl_data.ratios.horz_c.value /= 2;
spl_scratch->scl_data.ratios.vert_c.value /= 2;
}
@@ -529,23 +599,6 @@ static void spl_calculate_init_and_vp(bool flip_scan_dir,
*vp_offset = src_size - *vp_offset - *vp_size;
}
-static bool spl_is_yuv420(enum spl_pixel_format format)
-{
- if ((format >= SPL_PIXEL_FORMAT_420BPP8) &&
- (format <= SPL_PIXEL_FORMAT_420BPP10))
- return true;
-
- return false;
-}
-
-static bool spl_is_rgb8(enum spl_pixel_format format)
-{
- if (format == SPL_PIXEL_FORMAT_ARGB8888)
- return true;
-
- return false;
-}
-
/*Calculate inits and viewport */
static void spl_calculate_inits_and_viewports(struct spl_in *spl_in,
struct spl_scratch *spl_scratch)
@@ -556,8 +609,7 @@ static void spl_calculate_inits_and_viewports(struct spl_in *spl_in,
struct spl_rect recout_clip_in_recout_dst;
struct spl_rect overlap_in_active_timing;
struct spl_rect odm_slice = calculate_odm_slice_in_timing_active(spl_in);
- int vpc_div = (spl_in->basic_in.format == SPL_PIXEL_FORMAT_420BPP8
- || spl_in->basic_in.format == SPL_PIXEL_FORMAT_420BPP10) ? 2 : 1;
+ int vpc_div = spl_is_subsampled_format(spl_in->basic_in.format) ? 2 : 1;
bool orthogonal_rotation, flip_vert_scan_dir, flip_horz_scan_dir;
struct spl_fixed31_32 init_adj_h = spl_fixpt_zero;
struct spl_fixed31_32 init_adj_v = spl_fixpt_zero;
@@ -585,12 +637,7 @@ static void spl_calculate_inits_and_viewports(struct spl_in *spl_in,
&flip_vert_scan_dir,
&flip_horz_scan_dir);
- if (orthogonal_rotation) {
- spl_swap(src.width, src.height);
- spl_swap(flip_vert_scan_dir, flip_horz_scan_dir);
- }
-
- if (spl_is_yuv420(spl_in->basic_in.format)) {
+ if (spl_is_subsampled_format(spl_in->basic_in.format)) {
/* this gives the direction of the cositing (negative will move
* left, right otherwise)
*/
@@ -598,15 +645,15 @@ static void spl_calculate_inits_and_viewports(struct spl_in *spl_in,
switch (spl_in->basic_in.cositing) {
- case CHROMA_COSITING_LEFT:
- init_adj_h = spl_fixpt_zero;
+ case CHROMA_COSITING_TOPLEFT:
+ init_adj_h = spl_fixpt_from_fraction(sign, 4);
init_adj_v = spl_fixpt_from_fraction(sign, 4);
break;
- case CHROMA_COSITING_NONE:
+ case CHROMA_COSITING_LEFT:
init_adj_h = spl_fixpt_from_fraction(sign, 4);
- init_adj_v = spl_fixpt_from_fraction(sign, 4);
+ init_adj_v = spl_fixpt_zero;
break;
- case CHROMA_COSITING_TOPLEFT:
+ case CHROMA_COSITING_NONE:
default:
init_adj_h = spl_fixpt_zero;
init_adj_v = spl_fixpt_zero;
@@ -614,6 +661,12 @@ static void spl_calculate_inits_and_viewports(struct spl_in *spl_in,
}
}
+ if (orthogonal_rotation) {
+ spl_swap(src.width, src.height);
+ spl_swap(flip_vert_scan_dir, flip_horz_scan_dir);
+ spl_swap(init_adj_h, init_adj_v);
+ }
+
spl_calculate_init_and_vp(
flip_horz_scan_dir,
recout_clip_in_recout_dst.x,
@@ -678,7 +731,7 @@ static void spl_handle_3d_recout(struct spl_in *spl_in, struct spl_rect *recout)
* since 3d is special and needs to calculate vp as if there is no recout offset
* This may break with rotation, good thing we aren't mixing hw rotation and 3d
*/
- if (spl_in->basic_in.mpc_combine_v) {
+ if (spl_in->basic_in.mpc_h_slice_index) {
SPL_ASSERT(spl_in->basic_in.rotation == SPL_ROTATION_ANGLE_0 ||
(spl_in->basic_out.view_format != SPL_VIEW_3D_TOP_AND_BOTTOM &&
spl_in->basic_out.view_format != SPL_VIEW_3D_SIDE_BY_SIDE));
@@ -689,31 +742,15 @@ static void spl_handle_3d_recout(struct spl_in *spl_in, struct spl_rect *recout)
}
}
-static void spl_clamp_viewport(struct spl_rect *viewport)
+static void spl_clamp_viewport(struct spl_rect *viewport, int min_viewport_size)
{
+ if (min_viewport_size == 0)
+ min_viewport_size = MIN_VIEWPORT_SIZE;
/* Clamp minimum viewport size */
- if (viewport->height < MIN_VIEWPORT_SIZE)
- viewport->height = MIN_VIEWPORT_SIZE;
- if (viewport->width < MIN_VIEWPORT_SIZE)
- viewport->width = MIN_VIEWPORT_SIZE;
-}
-
-static bool spl_dscl_is_420_format(enum spl_pixel_format format)
-{
- if (format == SPL_PIXEL_FORMAT_420BPP8 ||
- format == SPL_PIXEL_FORMAT_420BPP10)
- return true;
- else
- return false;
-}
-
-static bool spl_dscl_is_video_format(enum spl_pixel_format format)
-{
- if (format >= SPL_PIXEL_FORMAT_VIDEO_BEGIN
- && format <= SPL_PIXEL_FORMAT_VIDEO_END)
- return true;
- else
- return false;
+ if (viewport->height < min_viewport_size)
+ viewport->height = min_viewport_size;
+ if (viewport->width < min_viewport_size)
+ viewport->width = min_viewport_size;
}
static enum scl_mode spl_get_dscl_mode(const struct spl_in *spl_in,
@@ -732,8 +769,8 @@ static enum scl_mode spl_get_dscl_mode(const struct spl_in *spl_in,
&& !enable_isharp)
return SCL_MODE_SCALING_444_BYPASS;
- if (!spl_dscl_is_420_format(pixel_format)) {
- if (spl_dscl_is_video_format(pixel_format))
+ if (!spl_is_subsampled_format(pixel_format)) {
+ if (spl_is_video_format(pixel_format))
return SCL_MODE_SCALING_444_YCBCR_ENABLE;
else
return SCL_MODE_SCALING_444_RGB_ENABLE;
@@ -744,32 +781,20 @@ static enum scl_mode spl_get_dscl_mode(const struct spl_in *spl_in,
* Do not bypass UV at 1:1 for cositing to be applied
*/
if (!enable_isharp) {
- if (data->ratios.horz.value == one && data->ratios.vert.value == one)
+ if (data->ratios.horz.value == one && data->ratios.vert.value == one && !spl_in->basic_out.always_scale)
return SCL_MODE_SCALING_420_LUMA_BYPASS;
}
return SCL_MODE_SCALING_420_YCBCR_ENABLE;
}
-static bool spl_choose_lls_policy(enum spl_pixel_format format,
- enum spl_transfer_func_type tf_type,
- enum spl_transfer_func_predefined tf_predefined_type,
+static void spl_choose_lls_policy(enum spl_pixel_format format,
enum linear_light_scaling *lls_pref)
{
- if (spl_is_yuv420(format)) {
+ if (spl_is_subsampled_format(format))
*lls_pref = LLS_PREF_NO;
- if ((tf_type == SPL_TF_TYPE_PREDEFINED) ||
- (tf_type == SPL_TF_TYPE_DISTRIBUTED_POINTS))
- return true;
- } else { /* RGB or YUV444 */
- if ((tf_type == SPL_TF_TYPE_PREDEFINED) ||
- (tf_type == SPL_TF_TYPE_BYPASS)) {
- *lls_pref = LLS_PREF_YES;
- return true;
- }
- }
- *lls_pref = LLS_PREF_NO;
- return false;
+ else /* RGB or YUV444 */
+ *lls_pref = LLS_PREF_YES;
}
/* Enable EASF ?*/
@@ -778,7 +803,6 @@ static bool enable_easf(struct spl_in *spl_in, struct spl_scratch *spl_scratch)
int vratio = 0;
int hratio = 0;
bool skip_easf = false;
- bool lls_enable_easf = true;
if (spl_in->disable_easf)
skip_easf = true;
@@ -794,17 +818,13 @@ static bool enable_easf(struct spl_in *spl_in, struct spl_scratch *spl_scratch)
skip_easf = true;
/*
- * If lls_pref is LLS_PREF_DONT_CARE, then use pixel format and transfer
- * function to determine whether to use LINEAR or NONLINEAR scaling
+ * If lls_pref is LLS_PREF_DONT_CARE, then use pixel format
+ * to determine whether to use LINEAR or NONLINEAR scaling
*/
if (spl_in->lls_pref == LLS_PREF_DONT_CARE)
- lls_enable_easf = spl_choose_lls_policy(spl_in->basic_in.format,
- spl_in->basic_in.tf_type, spl_in->basic_in.tf_predefined_type,
+ spl_choose_lls_policy(spl_in->basic_in.format,
&spl_in->lls_pref);
- if (!lls_enable_easf)
- skip_easf = true;
-
/* Check for linear scaling or EASF preferred */
if (spl_in->lls_pref != LLS_PREF_YES && !spl_in->prefer_easf)
skip_easf = true;
@@ -815,7 +835,7 @@ static bool enable_easf(struct spl_in *spl_in, struct spl_scratch *spl_scratch)
/* Check if video is in fullscreen mode */
static bool spl_is_video_fullscreen(struct spl_in *spl_in)
{
- if (spl_is_yuv420(spl_in->basic_in.format) && spl_in->is_fullscreen)
+ if (spl_is_video_format(spl_in->basic_in.format) && spl_in->is_fullscreen)
return true;
return false;
}
@@ -846,10 +866,10 @@ static bool spl_get_isharp_en(struct spl_in *spl_in,
* Apply sharpness to RGB and YUV (NV12/P010)
* surfaces based on policy setting
*/
- if (!spl_is_yuv420(spl_in->basic_in.format) &&
+ if (!spl_is_video_format(spl_in->basic_in.format) &&
(spl_in->sharpen_policy == SHARPEN_YUV))
return enable_isharp;
- else if ((spl_is_yuv420(spl_in->basic_in.format) && !fullscreen) &&
+ else if ((spl_is_video_format(spl_in->basic_in.format) && !fullscreen) &&
(spl_in->sharpen_policy == SHARPEN_RGB_FULLSCREEN_YUV))
return enable_isharp;
else if (!spl_in->is_fullscreen &&
@@ -869,8 +889,12 @@ static bool spl_get_isharp_en(struct spl_in *spl_in,
/* Calculate number of tap with adaptive scaling off */
static void spl_get_taps_non_adaptive_scaler(
- struct spl_scratch *spl_scratch, const struct spl_taps *in_taps)
+ struct spl_scratch *spl_scratch,
+ const struct spl_taps *in_taps,
+ bool is_subsampled)
{
+ bool check_max_downscale = false;
+
if (in_taps->h_taps == 0) {
if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.horz) > 1)
spl_scratch->scl_data.taps.h_taps = spl_min(2 * spl_fixpt_ceil(
@@ -882,8 +906,8 @@ static void spl_get_taps_non_adaptive_scaler(
if (in_taps->v_taps == 0) {
if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert) > 1)
- spl_scratch->scl_data.taps.v_taps = spl_min(spl_fixpt_ceil(spl_fixpt_mul_int(
- spl_scratch->scl_data.ratios.vert, 2)), 8);
+ spl_scratch->scl_data.taps.v_taps = spl_min(2 * spl_fixpt_ceil(
+ spl_scratch->scl_data.ratios.vert), 8);
else
spl_scratch->scl_data.taps.v_taps = 4;
} else
@@ -891,8 +915,8 @@ static void spl_get_taps_non_adaptive_scaler(
if (in_taps->v_taps_c == 0) {
if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert_c) > 1)
- spl_scratch->scl_data.taps.v_taps_c = spl_min(spl_fixpt_ceil(spl_fixpt_mul_int(
- spl_scratch->scl_data.ratios.vert_c, 2)), 8);
+ spl_scratch->scl_data.taps.v_taps_c = spl_min(2 * spl_fixpt_ceil(
+ spl_scratch->scl_data.ratios.vert_c), 8);
else
spl_scratch->scl_data.taps.v_taps_c = 4;
} else
@@ -910,13 +934,31 @@ static void spl_get_taps_non_adaptive_scaler(
else
spl_scratch->scl_data.taps.h_taps_c = in_taps->h_taps_c;
+
+ /*
+ * Max downscale supported is 6.0x. Add ASSERT to catch if go beyond that
+ */
+ check_max_downscale = spl_fixpt_le(spl_scratch->scl_data.ratios.horz,
+ spl_fixpt_from_fraction(6, 1));
+ SPL_ASSERT(check_max_downscale);
+ check_max_downscale = spl_fixpt_le(spl_scratch->scl_data.ratios.vert,
+ spl_fixpt_from_fraction(6, 1));
+ SPL_ASSERT(check_max_downscale);
+ check_max_downscale = spl_fixpt_le(spl_scratch->scl_data.ratios.horz_c,
+ spl_fixpt_from_fraction(6, 1));
+ SPL_ASSERT(check_max_downscale);
+ check_max_downscale = spl_fixpt_le(spl_scratch->scl_data.ratios.vert_c,
+ spl_fixpt_from_fraction(6, 1));
+ SPL_ASSERT(check_max_downscale);
+
+
if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.horz))
spl_scratch->scl_data.taps.h_taps = 1;
if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert))
spl_scratch->scl_data.taps.v_taps = 1;
- if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.horz_c))
+ if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.horz_c) && !is_subsampled)
spl_scratch->scl_data.taps.h_taps_c = 1;
- if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert_c))
+ if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert_c) && !is_subsampled)
spl_scratch->scl_data.taps.v_taps_c = 1;
}
@@ -928,16 +970,16 @@ static bool spl_get_optimal_number_of_taps(
bool *enable_isharp)
{
int num_part_y, num_part_c;
- int max_taps_y, max_taps_c;
- int min_taps_y, min_taps_c;
+ unsigned int max_taps_y, max_taps_c;
+ unsigned int min_taps_y, min_taps_c;
enum lb_memory_config lb_config;
- bool skip_easf = false;
- bool is_ycbcr = spl_dscl_is_video_format(spl_in->basic_in.format);
+ bool skip_easf = false;
+ bool is_subsampled = spl_is_subsampled_format(spl_in->basic_in.format);
if (spl_scratch->scl_data.viewport.width > spl_scratch->scl_data.h_active &&
max_downscale_src_width != 0 &&
spl_scratch->scl_data.viewport.width > max_downscale_src_width) {
- spl_get_taps_non_adaptive_scaler(spl_scratch, in_taps);
+ spl_get_taps_non_adaptive_scaler(spl_scratch, in_taps, is_subsampled);
*enable_easf_v = false;
*enable_easf_h = false;
*enable_isharp = false;
@@ -946,7 +988,7 @@ static bool spl_get_optimal_number_of_taps(
/* Disable adaptive scaler and sharpener when integer scaling is enabled */
if (spl_in->scaling_quality.integer_scaling) {
- spl_get_taps_non_adaptive_scaler(spl_scratch, in_taps);
+ spl_get_taps_non_adaptive_scaler(spl_scratch, in_taps, is_subsampled);
*enable_easf_v = false;
*enable_easf_h = false;
*enable_isharp = false;
@@ -961,10 +1003,11 @@ static bool spl_get_optimal_number_of_taps(
* From programming guide: taps = min{ ceil(2*H_RATIO,1), 8} for downscaling
* taps = 4 for upscaling
*/
- if (skip_easf)
- spl_get_taps_non_adaptive_scaler(spl_scratch, in_taps);
+ if (skip_easf) {
+ spl_get_taps_non_adaptive_scaler(spl_scratch, in_taps, is_subsampled);
+ }
else {
- if (spl_is_yuv420(spl_in->basic_in.format)) {
+ if (spl_is_video_format(spl_in->basic_in.format)) {
spl_scratch->scl_data.taps.h_taps = 6;
spl_scratch->scl_data.taps.v_taps = 6;
spl_scratch->scl_data.taps.h_taps_c = 4;
@@ -982,8 +1025,7 @@ static bool spl_get_optimal_number_of_taps(
min_taps_c = spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert_c);
/* Use LB_MEMORY_CONFIG_3 for 4:2:0 */
- if ((spl_in->basic_in.format == SPL_PIXEL_FORMAT_420BPP8)
- || (spl_in->basic_in.format == SPL_PIXEL_FORMAT_420BPP10))
+ if (spl_is_yuv420(spl_in->basic_in.format))
lb_config = LB_MEMORY_CONFIG_3;
else
lb_config = LB_MEMORY_CONFIG_0;
@@ -992,12 +1034,18 @@ static bool spl_get_optimal_number_of_taps(
lb_config, &num_part_y, &num_part_c);
/* MAX_V_TAPS = MIN (NUM_LINES - MAX(CEILING(V_RATIO,1)-2, 0), 8) */
if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert) > 2)
- max_taps_y = num_part_y - (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert) - 2);
+ if ((spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert) - 2) > num_part_y)
+ max_taps_y = 0;
+ else
+ max_taps_y = num_part_y - (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert) - 2);
else
max_taps_y = num_part_y;
if (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert_c) > 2)
- max_taps_c = num_part_c - (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert_c) - 2);
+ if ((spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert_c) - 2) > num_part_c)
+ max_taps_c = 0;
+ else
+ max_taps_c = num_part_c - (spl_fixpt_ceil(spl_scratch->scl_data.ratios.vert_c) - 2);
else
max_taps_c = num_part_c;
@@ -1039,13 +1087,11 @@ static bool spl_get_optimal_number_of_taps(
if (spl_scratch->scl_data.taps.h_taps_c == 5)
spl_scratch->scl_data.taps.h_taps_c = 4;
- if (spl_is_yuv420(spl_in->basic_in.format)) {
- if ((spl_scratch->scl_data.taps.h_taps <= 4) ||
- (spl_scratch->scl_data.taps.h_taps_c <= 3)) {
+ if (spl_is_video_format(spl_in->basic_in.format)) {
+ if (spl_scratch->scl_data.taps.h_taps <= 4) {
*enable_easf_v = false;
*enable_easf_h = false;
- } else if ((spl_scratch->scl_data.taps.v_taps <= 3) ||
- (spl_scratch->scl_data.taps.v_taps_c <= 3)) {
+ } else if (spl_scratch->scl_data.taps.v_taps <= 3) {
*enable_easf_v = false;
*enable_easf_h = true;
} else {
@@ -1085,11 +1131,10 @@ static bool spl_get_optimal_number_of_taps(
(IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert))) {
spl_scratch->scl_data.taps.h_taps = 1;
spl_scratch->scl_data.taps.v_taps = 1;
-
- if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.horz_c) && !is_ycbcr)
+ if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.horz_c) && !is_subsampled)
spl_scratch->scl_data.taps.h_taps_c = 1;
- if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert_c) && !is_ycbcr)
+ if (IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert_c) && !is_subsampled)
spl_scratch->scl_data.taps.v_taps_c = 1;
*enable_easf_v = false;
@@ -1103,13 +1148,14 @@ static bool spl_get_optimal_number_of_taps(
(IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert)))
spl_scratch->scl_data.taps.v_taps = 1;
- if ((!*enable_easf_h) && !is_ycbcr &&
+ if ((!*enable_easf_h) && !is_subsampled &&
(IDENTITY_RATIO(spl_scratch->scl_data.ratios.horz_c)))
spl_scratch->scl_data.taps.h_taps_c = 1;
- if ((!*enable_easf_v) && !is_ycbcr &&
+ if ((!*enable_easf_v) && !is_subsampled &&
(IDENTITY_RATIO(spl_scratch->scl_data.ratios.vert_c)))
spl_scratch->scl_data.taps.v_taps_c = 1;
+
}
}
return true;
@@ -1118,7 +1164,7 @@ static bool spl_get_optimal_number_of_taps(
static void spl_set_black_color_data(enum spl_pixel_format format,
struct scl_black_color *scl_black_color)
{
- bool ycbcr = spl_dscl_is_video_format(format);
+ bool ycbcr = spl_is_video_format(format);
if (ycbcr) {
scl_black_color->offset_rgb_y = BLACK_OFFSET_RGB_Y;
scl_black_color->offset_rgb_cbcr = BLACK_OFFSET_CBCR;
@@ -1260,7 +1306,7 @@ static void spl_set_easf_data(struct spl_scratch *spl_scratch, struct spl_out *s
if (enable_easf_v) {
dscl_prog_data->easf_v_en = true;
dscl_prog_data->easf_v_ring = 0;
- dscl_prog_data->easf_v_sharp_factor = 0;
+ dscl_prog_data->easf_v_sharp_factor = 1;
dscl_prog_data->easf_v_bf1_en = 1; // 1-bit, BF1 calculation enable, 0=disable, 1=enable
dscl_prog_data->easf_v_bf2_mode = 0xF; // 4-bit, BF2 calculation mode
/* 2-bit, BF3 chroma mode correction calculation mode */
@@ -1424,7 +1470,7 @@ static void spl_set_easf_data(struct spl_scratch *spl_scratch, struct spl_out *s
if (enable_easf_h) {
dscl_prog_data->easf_h_en = true;
dscl_prog_data->easf_h_ring = 0;
- dscl_prog_data->easf_h_sharp_factor = 0;
+ dscl_prog_data->easf_h_sharp_factor = 1;
dscl_prog_data->easf_h_bf1_en =
1; // 1-bit, BF1 calculation enable, 0=disable, 1=enable
dscl_prog_data->easf_h_bf2_mode =
@@ -1585,7 +1631,7 @@ static void spl_set_easf_data(struct spl_scratch *spl_scratch, struct spl_out *s
0x0; // fp1.5.10, C3 coefficient
}
- if (spl_is_yuv420(format)) { /* TODO: 0 = RGB, 1 = YUV */
+ if (spl_is_subsampled_format(format)) { /* TODO: 0 = RGB, 1 = YUV */
dscl_prog_data->easf_matrix_mode = 1;
/*
* 2-bit, BF3 chroma mode correction calculation mode
@@ -1768,6 +1814,8 @@ static bool spl_calculate_number_of_taps(struct spl_in *spl_in, struct spl_scrat
spl_calculate_recout(spl_in, spl_scratch, spl_out);
/* depends on pixel format */
spl_calculate_scaling_ratios(spl_in, spl_scratch, spl_out);
+ /* Adjust recout for opp if needed */
+ spl_opp_adjust_rect(&spl_scratch->scl_data.recout, &spl_in->basic_in.opp_recout_adjust);
/* depends on scaling ratios and recout, does not calculate offset yet */
spl_calculate_viewport_size(spl_in, spl_scratch);
@@ -1779,7 +1827,7 @@ static bool spl_calculate_number_of_taps(struct spl_in *spl_in, struct spl_scrat
}
/* Calculate scaler parameters */
-bool spl_calculate_scaler_params(struct spl_in *spl_in, struct spl_out *spl_out)
+bool SPL_NAMESPACE(spl_calculate_scaler_params(struct spl_in *spl_in, struct spl_out *spl_out))
{
bool res = false;
bool enable_easf_v = false;
@@ -1804,7 +1852,7 @@ bool spl_calculate_scaler_params(struct spl_in *spl_in, struct spl_out *spl_out)
// Handle 3d recout
spl_handle_3d_recout(spl_in, &spl_scratch.scl_data.recout);
// Clamp
- spl_clamp_viewport(&spl_scratch.scl_data.viewport);
+ spl_clamp_viewport(&spl_scratch.scl_data.viewport, spl_in->min_viewport_size);
// Save all calculated parameters in dscl_prog_data structure to program hw registers
spl_set_dscl_prog_data(spl_in, &spl_scratch, spl_out, enable_easf_v, enable_easf_h, enable_isharp);
@@ -1844,7 +1892,7 @@ bool spl_calculate_scaler_params(struct spl_in *spl_in, struct spl_out *spl_out)
}
/* External interface to get number of taps only */
-bool spl_get_number_of_taps(struct spl_in *spl_in, struct spl_out *spl_out)
+bool SPL_NAMESPACE(spl_get_number_of_taps(struct spl_in *spl_in, struct spl_out *spl_out))
{
bool res = false;
bool enable_easf_v = false;
@@ -1859,3 +1907,4 @@ bool spl_get_number_of_taps(struct spl_in *spl_in, struct spl_out *spl_out)
spl_set_taps_data(dscl_prog_data, data);
return res;
}
+
diff --git a/drivers/gpu/drm/amd/display/dc/sspl/dc_spl.h b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl.h
new file mode 100644
index 000000000000..d621c42a237e
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl.h
@@ -0,0 +1,27 @@
+// SPDX-License-Identifier: MIT
+//
+// Copyright 2024 Advanced Micro Devices, Inc.
+
+#ifndef __DC_SPL_H__
+#define __DC_SPL_H__
+
+#include "dc_spl_types.h"
+#define BLACK_OFFSET_RGB_Y 0x0
+#define BLACK_OFFSET_CBCR 0x8000
+
+#ifndef SPL_PFX_
+#define SPL_PFX_
+#endif
+
+#define SPL_EXPAND2(a, b) a##b
+#define SPL_EXPAND(a, b) SPL_EXPAND2(a, b)
+#define SPL_NAMESPACE(symbol) SPL_EXPAND(SPL_PFX_, symbol)
+
+
+/* SPL interfaces */
+
+bool SPL_NAMESPACE(spl_calculate_scaler_params(struct spl_in *spl_in, struct spl_out *spl_out));
+
+bool SPL_NAMESPACE(spl_get_number_of_taps(struct spl_in *spl_in, struct spl_out *spl_out));
+
+#endif /* __DC_SPL_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_filters.c b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_filters.c
index 99238644e0a1..99238644e0a1 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_filters.c
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_filters.c
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_filters.h b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_filters.h
index 20439cdbdb10..20439cdbdb10 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_filters.h
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_filters.h
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.c b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_isharp_filters.c
index e0572252c640..12acdd34e6a6 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.c
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_isharp_filters.c
@@ -11,232 +11,6 @@
// LUT content is packed as 4-bytes into one DWORD/entry
// A_start = 0.000000
// A_end = 10.000000
-// A_gain = 2.000000
-// B_start = 11.000000
-// B_end = 86.000000
-// C_start = 40.000000
-// C_end = 64.000000
-//========================================
-static const uint32_t filter_isharp_1D_lut_0[ISHARP_LUT_TABLE_SIZE] = {
-0x02010000,
-0x0A070503,
-0x1614100D,
-0x1C1B1918,
-0x22211F1E,
-0x27262423,
-0x2A2A2928,
-0x2D2D2C2B,
-0x302F2F2E,
-0x31313030,
-0x31313131,
-0x31313131,
-0x30303031,
-0x292D2F2F,
-0x191D2125,
-0x050A0F14,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-0x00000000,
-};
-//========================================
-// Delta Gain 1DLUT
-// LUT content is packed as 4-bytes into one DWORD/entry
-// A_start = 0.000000
-// A_end = 10.000000
-// A_gain = 0.500000
-// B_start = 11.000000
-// B_end = 127.000000
-// C_start = 96.000000
-// C_end = 127.000000
-//========================================
-
-static const uint32_t filter_isharp_1D_lut_0p5x[ISHARP_LUT_TABLE_SIZE] = {
-0x00000000,
-0x02020101,
-0x06050403,
-0x07070606,
-0x09080808,
-0x0A0A0A09,
-0x0C0B0B0B,
-0x0D0D0C0C,
-0x0E0E0D0D,
-0x0F0F0E0E,
-0x100F0F0F,
-0x10101010,
-0x11111010,
-0x11111111,
-0x11111111,
-0x11111111,
-0x11111111,
-0x11111111,
-0x11111111,
-0x10101111,
-0x10101010,
-0x0F0F0F10,
-0x0E0E0F0F,
-0x0D0D0E0E,
-0x0C0C0D0D,
-0x0B0B0B0C,
-0x090A0A0A,
-0x08080809,
-0x06060707,
-0x04050506,
-0x02030304,
-0x00010102,
-};
-//========================================
-// Delta Gain 1DLUT
-// LUT content is packed as 4-bytes into one DWORD/entry
-// A_start = 0.000000
-// A_end = 10.000000
-// A_gain = 1.000000
-// B_start = 11.000000
-// B_end = 127.000000
-// C_start = 96.000000
-// C_end = 127.000000
-//========================================
-static const uint32_t filter_isharp_1D_lut_1p0x[ISHARP_LUT_TABLE_SIZE] = {
-0x01000000,
-0x05040302,
-0x0B0A0806,
-0x0E0E0D0C,
-0x1211100F,
-0x15141312,
-0x17171615,
-0x1A191918,
-0x1C1B1B1A,
-0x1E1D1D1C,
-0x1F1F1E1E,
-0x2020201F,
-0x21212121,
-0x22222222,
-0x23232222,
-0x23232323,
-0x23232323,
-0x22222323,
-0x22222222,
-0x21212121,
-0x1F202020,
-0x1E1E1F1F,
-0x1C1D1D1E,
-0x1A1B1B1C,
-0x1819191A,
-0x15161717,
-0x12131415,
-0x0F101112,
-0x0C0D0E0E,
-0x08090A0B,
-0x04050607,
-0x00010203,
-};
-//========================================
-// Delta Gain 1DLUT
-// LUT content is packed as 4-bytes into one DWORD/entry
-// A_start = 0.000000
-// A_end = 10.000000
-// A_gain = 1.500000
-// B_start = 11.000000
-// B_end = 127.000000
-// C_start = 96.000000
-// C_end = 127.000000
-//========================================
-static const uint32_t filter_isharp_1D_lut_1p5x[ISHARP_LUT_TABLE_SIZE] = {
-0x01010000,
-0x07050402,
-0x110F0C0A,
-0x16141312,
-0x1B191817,
-0x1F1E1D1C,
-0x23222120,
-0x26262524,
-0x2A292827,
-0x2C2C2B2A,
-0x2F2E2E2D,
-0x3130302F,
-0x32323131,
-0x33333332,
-0x34343433,
-0x34343434,
-0x34343434,
-0x33343434,
-0x32333333,
-0x31313232,
-0x2F303031,
-0x2D2E2E2F,
-0x2A2B2C2C,
-0x2728292A,
-0x24252626,
-0x20212223,
-0x1C1D1E1F,
-0x1718191B,
-0x12131416,
-0x0C0E0F10,
-0x0608090B,
-0x00020305
-};
-//========================================
-// Delta Gain 1DLUT
-// LUT content is packed as 4-bytes into one DWORD/entry
-// A_start = 0.000000
-// A_end = 10.000000
-// A_gain = 2.000000
-// B_start = 11.000000
-// B_end = 127.000000
-// C_start = 40.000000
-// C_end = 127.000000
-//========================================
-static const uint32_t filter_isharp_1D_lut_2p0x[ISHARP_LUT_TABLE_SIZE] = {
-0x02010000,
-0x0A070503,
-0x1614100D,
-0x1D1B1A18,
-0x2322201F,
-0x29282625,
-0x2F2D2C2B,
-0x33323130,
-0x38373534,
-0x3B3A3938,
-0x3E3E3D3C,
-0x4140403F,
-0x43424241,
-0x44444443,
-0x45454545,
-0x46454545,
-0x45454546,
-0x45454545,
-0x43444444,
-0x41424243,
-0x3F404041,
-0x3C3D3E3E,
-0x38393A3B,
-0x34353738,
-0x30313233,
-0x2B2C2D2F,
-0x25262829,
-0x1F202223,
-0x181A1B1D,
-0x10121416,
-0x080B0D0E,
-0x00020406,
-};
-//========================================
-// Delta Gain 1DLUT
-// LUT content is packed as 4-bytes into one DWORD/entry
-// A_start = 0.000000
-// A_end = 10.000000
// A_gain = 3.000000
// B_start = 11.000000
// B_end = 127.000000
@@ -278,52 +52,6 @@ static const uint32_t filter_isharp_1D_lut_3p0x[ISHARP_LUT_TABLE_SIZE] = {
0x0003060A,
};
-//========================================
-// Wide scaler coefficients
-//========================================================
-// <using> gen_scaler_coeffs.m
-// <date> 15-Dec-2021
-// <coeffDescrip> 6t_64p_LanczosEd_p_1_p_10qb_
-// <num_taps> 6
-// <num_phases> 64
-// <CoefType> LanczosEd
-// <CoefQuant> S1.10
-//========================================================
-static const uint16_t filter_isharp_wide_6tap_64p[198] = {
-0x0000, 0x0000, 0x0400, 0x0000, 0x0000, 0x0000,
-0x0003, 0x0FF3, 0x0400, 0x000D, 0x0FFD, 0x0000,
-0x0006, 0x0FE7, 0x03FE, 0x001C, 0x0FF9, 0x0000,
-0x0009, 0x0FDB, 0x03FC, 0x002B, 0x0FF5, 0x0000,
-0x000C, 0x0FD0, 0x03F9, 0x003A, 0x0FF1, 0x0000,
-0x000E, 0x0FC5, 0x03F5, 0x004A, 0x0FED, 0x0001,
-0x0011, 0x0FBB, 0x03F0, 0x005A, 0x0FE9, 0x0001,
-0x0013, 0x0FB2, 0x03EB, 0x006A, 0x0FE5, 0x0001,
-0x0015, 0x0FA9, 0x03E4, 0x007B, 0x0FE1, 0x0002,
-0x0017, 0x0FA1, 0x03DD, 0x008D, 0x0FDC, 0x0002,
-0x0018, 0x0F99, 0x03D4, 0x00A0, 0x0FD8, 0x0003,
-0x001A, 0x0F92, 0x03CB, 0x00B2, 0x0FD3, 0x0004,
-0x001B, 0x0F8C, 0x03C1, 0x00C6, 0x0FCE, 0x0004,
-0x001C, 0x0F86, 0x03B7, 0x00D9, 0x0FC9, 0x0005,
-0x001D, 0x0F80, 0x03AB, 0x00EE, 0x0FC4, 0x0006,
-0x001E, 0x0F7C, 0x039F, 0x0101, 0x0FBF, 0x0007,
-0x001F, 0x0F78, 0x0392, 0x0115, 0x0FBA, 0x0008,
-0x001F, 0x0F74, 0x0385, 0x012B, 0x0FB5, 0x0008,
-0x0020, 0x0F71, 0x0376, 0x0140, 0x0FB0, 0x0009,
-0x0020, 0x0F6E, 0x0367, 0x0155, 0x0FAB, 0x000B,
-0x0020, 0x0F6C, 0x0357, 0x016B, 0x0FA6, 0x000C,
-0x0020, 0x0F6A, 0x0347, 0x0180, 0x0FA2, 0x000D,
-0x0020, 0x0F69, 0x0336, 0x0196, 0x0F9D, 0x000E,
-0x0020, 0x0F69, 0x0325, 0x01AB, 0x0F98, 0x000F,
-0x001F, 0x0F68, 0x0313, 0x01C3, 0x0F93, 0x0010,
-0x001F, 0x0F69, 0x0300, 0x01D8, 0x0F8F, 0x0011,
-0x001E, 0x0F69, 0x02ED, 0x01EF, 0x0F8B, 0x0012,
-0x001D, 0x0F6A, 0x02D9, 0x0205, 0x0F87, 0x0014,
-0x001D, 0x0F6C, 0x02C5, 0x021A, 0x0F83, 0x0015,
-0x001C, 0x0F6E, 0x02B1, 0x0230, 0x0F7F, 0x0016,
-0x001B, 0x0F70, 0x029C, 0x0247, 0x0F7B, 0x0017,
-0x001A, 0x0F72, 0x0287, 0x025D, 0x0F78, 0x0018,
-0x0019, 0x0F75, 0x0272, 0x0272, 0x0F75, 0x0019
-};
// Blur and scale coefficients
//========================================================
// <using> gen_BlurScale_coeffs.m
@@ -456,9 +184,113 @@ static const uint16_t filter_isharp_bs_3tap_64p[99] = {
};
/* Converted Blur & Scale coeff tables from S1.10 to S1.12 */
-static uint16_t filter_isharp_bs_4tap_in_6_64p_s1_12[198];
-static uint16_t filter_isharp_bs_4tap_64p_s1_12[132];
-static uint16_t filter_isharp_bs_3tap_64p_s1_12[99];
+static const uint16_t filter_isharp_bs_4tap_in_6_64p_s1_12[198] = {
+0x0000, 0x0394, 0x08dc, 0x0390, 0x0000, 0x0000,
+0x0000, 0x0378, 0x08dc, 0x03ac, 0x0000, 0x0000,
+0x0000, 0x035c, 0x08d8, 0x03c8, 0x0004, 0x0000,
+0x0000, 0x0340, 0x08d4, 0x03e8, 0x0004, 0x0000,
+0x0000, 0x0324, 0x08d0, 0x0404, 0x0008, 0x0000,
+0x0000, 0x0308, 0x08cc, 0x0420, 0x000c, 0x0000,
+0x0000, 0x02ec, 0x08c8, 0x0440, 0x000c, 0x0000,
+0x0000, 0x02d4, 0x08c0, 0x045c, 0x0010, 0x0000,
+0x0000, 0x02b8, 0x08b8, 0x047c, 0x0014, 0x0000,
+0x0000, 0x02a0, 0x08b0, 0x0498, 0x0018, 0x0000,
+0x0000, 0x0288, 0x08a8, 0x04b4, 0x001c, 0x0000,
+0x0000, 0x0270, 0x08a0, 0x04d0, 0x0020, 0x0000,
+0x0000, 0x0258, 0x0894, 0x04f0, 0x0024, 0x0000,
+0x0000, 0x0240, 0x0888, 0x050c, 0x002c, 0x0000,
+0x0000, 0x0228, 0x087c, 0x052c, 0x0030, 0x0000,
+0x0000, 0x0214, 0x0870, 0x0544, 0x0038, 0x0000,
+0x0000, 0x01fc, 0x0860, 0x0568, 0x003c, 0x0000,
+0x0000, 0x01e8, 0x0854, 0x0580, 0x0044, 0x0000,
+0x0000, 0x01d0, 0x0844, 0x05a0, 0x004c, 0x0000,
+0x0000, 0x01bc, 0x0834, 0x05bc, 0x0054, 0x0000,
+0x0000, 0x01a8, 0x0824, 0x05d8, 0x005c, 0x0000,
+0x0000, 0x0194, 0x0810, 0x05f8, 0x0064, 0x0000,
+0x0000, 0x0180, 0x0800, 0x0614, 0x006c, 0x0000,
+0x0000, 0x0170, 0x07ec, 0x0630, 0x0074, 0x0000,
+0x0000, 0x015c, 0x07d8, 0x064c, 0x0080, 0x0000,
+0x0000, 0x014c, 0x07c4, 0x0668, 0x0088, 0x0000,
+0x0000, 0x0138, 0x07b0, 0x0684, 0x0094, 0x0000,
+0x0000, 0x0128, 0x0798, 0x06a0, 0x00a0, 0x0000,
+0x0000, 0x0118, 0x0784, 0x06bc, 0x00a8, 0x0000,
+0x0000, 0x0108, 0x076c, 0x06d8, 0x00b4, 0x0000,
+0x0000, 0x00fc, 0x0754, 0x06ec, 0x00c4, 0x0000,
+0x0000, 0x00ec, 0x073c, 0x0708, 0x00d0, 0x0000,
+0x0000, 0x00dc, 0x0724, 0x0724, 0x00dc, 0x0000,
+};
+
+static const uint16_t filter_isharp_bs_4tap_64p_s1_12[132] = {
+0x0394, 0x08dc, 0x0390, 0x0000,
+0x0378, 0x08dc, 0x03ac, 0x0000,
+0x035c, 0x08d8, 0x03c8, 0x0004,
+0x0340, 0x08d4, 0x03e8, 0x0004,
+0x0324, 0x08d0, 0x0404, 0x0008,
+0x0308, 0x08cc, 0x0420, 0x000c,
+0x02ec, 0x08c8, 0x0440, 0x000c,
+0x02d4, 0x08c0, 0x045c, 0x0010,
+0x02b8, 0x08b8, 0x047c, 0x0014,
+0x02a0, 0x08b0, 0x0498, 0x0018,
+0x0288, 0x08a8, 0x04b4, 0x001c,
+0x0270, 0x08a0, 0x04d0, 0x0020,
+0x0258, 0x0894, 0x04f0, 0x0024,
+0x0240, 0x0888, 0x050c, 0x002c,
+0x0228, 0x087c, 0x052c, 0x0030,
+0x0214, 0x0870, 0x0544, 0x0038,
+0x01fc, 0x0860, 0x0568, 0x003c,
+0x01e8, 0x0854, 0x0580, 0x0044,
+0x01d0, 0x0844, 0x05a0, 0x004c,
+0x01bc, 0x0834, 0x05bc, 0x0054,
+0x01a8, 0x0824, 0x05d8, 0x005c,
+0x0194, 0x0810, 0x05f8, 0x0064,
+0x0180, 0x0800, 0x0614, 0x006c,
+0x0170, 0x07ec, 0x0630, 0x0074,
+0x015c, 0x07d8, 0x064c, 0x0080,
+0x014c, 0x07c4, 0x0668, 0x0088,
+0x0138, 0x07b0, 0x0684, 0x0094,
+0x0128, 0x0798, 0x06a0, 0x00a0,
+0x0118, 0x0784, 0x06bc, 0x00a8,
+0x0108, 0x076c, 0x06d8, 0x00b4,
+0x00fc, 0x0754, 0x06ec, 0x00c4,
+0x00ec, 0x073c, 0x0708, 0x00d0,
+0x00dc, 0x0724, 0x0724, 0x00dc,
+};
+
+static const uint16_t filter_isharp_bs_3tap_64p_s1_12[99] = {
+0x0800, 0x0800, 0x0000,
+0x07d8, 0x0818, 0x0010,
+0x07b0, 0x082c, 0x0024,
+0x0788, 0x0844, 0x0034,
+0x0760, 0x0858, 0x0048,
+0x0738, 0x0870, 0x0058,
+0x0710, 0x0884, 0x006c,
+0x06e8, 0x0898, 0x0080,
+0x06c0, 0x08a8, 0x0098,
+0x0698, 0x08bc, 0x00ac,
+0x0670, 0x08cc, 0x00c4,
+0x0648, 0x08e0, 0x00d8,
+0x0620, 0x08f0, 0x00f0,
+0x05f8, 0x0900, 0x0108,
+0x05d0, 0x0910, 0x0120,
+0x05a8, 0x0920, 0x0138,
+0x0584, 0x0928, 0x0154,
+0x055c, 0x0938, 0x016c,
+0x0534, 0x0944, 0x0188,
+0x0510, 0x094c, 0x01a4,
+0x04e8, 0x0958, 0x01c0,
+0x04c4, 0x0960, 0x01dc,
+0x049c, 0x096c, 0x01f8,
+0x0478, 0x0970, 0x0218,
+0x0454, 0x0978, 0x0234,
+0x042c, 0x0980, 0x0254,
+0x0408, 0x0988, 0x0270,
+0x03e4, 0x098c, 0x0290,
+0x03c0, 0x0990, 0x02b0,
+0x039c, 0x0994, 0x02d0,
+0x037c, 0x0990, 0x02f4,
+0x0358, 0x0994, 0x0314,
+0x0334, 0x0998, 0x0334,
+};
/* Pre-generated 1DLUT for given setup and sharpness level */
struct isharp_1D_lut_pregen filter_isharp_1D_lut_pregen[NUM_SHARPNESS_SETUPS] = {
@@ -509,47 +341,6 @@ struct scale_ratio_to_sharpness_level_adj sharpness_level_adj[NUM_SHARPNESS_ADJ_
{1, 1, 5},
};
-const uint32_t *spl_get_filter_isharp_1D_lut_0(void)
-{
- return filter_isharp_1D_lut_0;
-}
-const uint32_t *spl_get_filter_isharp_1D_lut_0p5x(void)
-{
- return filter_isharp_1D_lut_0p5x;
-}
-const uint32_t *spl_get_filter_isharp_1D_lut_1p0x(void)
-{
- return filter_isharp_1D_lut_1p0x;
-}
-const uint32_t *spl_get_filter_isharp_1D_lut_1p5x(void)
-{
- return filter_isharp_1D_lut_1p5x;
-}
-const uint32_t *spl_get_filter_isharp_1D_lut_2p0x(void)
-{
- return filter_isharp_1D_lut_2p0x;
-}
-const uint32_t *spl_get_filter_isharp_1D_lut_3p0x(void)
-{
- return filter_isharp_1D_lut_3p0x;
-}
-const uint16_t *spl_get_filter_isharp_wide_6tap_64p(void)
-{
- return filter_isharp_wide_6tap_64p;
-}
-uint16_t *spl_get_filter_isharp_bs_4tap_in_6_64p(void)
-{
- return filter_isharp_bs_4tap_in_6_64p_s1_12;
-}
-uint16_t *spl_get_filter_isharp_bs_4tap_64p(void)
-{
- return filter_isharp_bs_4tap_64p_s1_12;
-}
-uint16_t *spl_get_filter_isharp_bs_3tap_64p(void)
-{
- return filter_isharp_bs_3tap_64p_s1_12;
-}
-
static unsigned int spl_calculate_sharpness_level_adj(struct spl_fixed31_32 ratio)
{
int j;
@@ -589,7 +380,7 @@ static unsigned int spl_calculate_sharpness_level_adj(struct spl_fixed31_32 rati
}
static unsigned int spl_calculate_sharpness_level(struct spl_fixed31_32 ratio,
- int discrete_sharpness_level, enum system_setup setup,
+ unsigned int discrete_sharpness_level, enum system_setup setup,
struct spl_sharpness_range sharpness_range,
enum scale_to_sharpness_policy scale_to_sharpness_policy)
{
@@ -720,24 +511,29 @@ uint32_t *spl_get_pregen_filter_isharp_1D_lut(enum system_setup setup)
return filter_isharp_1D_lut_pregen[setup].value;
}
-void spl_init_blur_scale_coeffs(void)
+const uint16_t *spl_dscl_get_blur_scale_coeffs_64p(int taps)
{
- convert_filter_s1_10_to_s1_12(filter_isharp_bs_3tap_64p,
- filter_isharp_bs_3tap_64p_s1_12, 3);
- convert_filter_s1_10_to_s1_12(filter_isharp_bs_4tap_64p,
- filter_isharp_bs_4tap_64p_s1_12, 4);
- convert_filter_s1_10_to_s1_12(filter_isharp_bs_4tap_in_6_64p,
- filter_isharp_bs_4tap_in_6_64p_s1_12, 6);
+ if (taps == 3)
+ return filter_isharp_bs_3tap_64p_s1_12;
+ else if (taps == 4)
+ return filter_isharp_bs_4tap_64p_s1_12;
+ else if (taps == 6)
+ return filter_isharp_bs_4tap_in_6_64p_s1_12;
+ else {
+ /* should never happen, bug */
+ SPL_BREAK_TO_DEBUGGER();
+ return NULL;
+ }
}
-uint16_t *spl_dscl_get_blur_scale_coeffs_64p(int taps)
+const uint16_t *spl_dscl_get_blur_scale_coeffs_64p_s1_10(int taps)
{
if (taps == 3)
- return spl_get_filter_isharp_bs_3tap_64p();
+ return filter_isharp_bs_3tap_64p;
else if (taps == 4)
- return spl_get_filter_isharp_bs_4tap_64p();
+ return filter_isharp_bs_4tap_64p;
else if (taps == 6)
- return spl_get_filter_isharp_bs_4tap_in_6_64p();
+ return filter_isharp_bs_4tap_in_6_64p;
else {
/* should never happen, bug */
SPL_BREAK_TO_DEBUGGER();
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_isharp_filters.h
index 89af91e19b6c..f5e3d3ecc913 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_isharp_filters.h
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_isharp_filters.h
@@ -7,18 +7,6 @@
#include "dc_spl_types.h"
-const uint32_t *spl_get_filter_isharp_1D_lut_0(void);
-const uint32_t *spl_get_filter_isharp_1D_lut_0p5x(void);
-const uint32_t *spl_get_filter_isharp_1D_lut_1p0x(void);
-const uint32_t *spl_get_filter_isharp_1D_lut_1p5x(void);
-const uint32_t *spl_get_filter_isharp_1D_lut_2p0x(void);
-const uint32_t *spl_get_filter_isharp_1D_lut_3p0x(void);
-uint16_t *spl_get_filter_isharp_bs_4tap_in_6_64p(void);
-uint16_t *spl_get_filter_isharp_bs_4tap_64p(void);
-uint16_t *spl_get_filter_isharp_bs_3tap_64p(void);
-const uint16_t *spl_get_filter_isharp_wide_6tap_64p(void);
-uint16_t *spl_dscl_get_blur_scale_coeffs_64p(int taps);
-
#define NUM_SHARPNESS_ADJ_LEVELS 6
struct scale_ratio_to_sharpness_level_adj {
unsigned int ratio_numer;
@@ -40,11 +28,15 @@ enum system_setup {
NUM_SHARPNESS_SETUPS
};
-void spl_init_blur_scale_coeffs(void);
void spl_set_blur_scale_data(struct dscl_prog_data *dscl_prog_data,
const struct spl_scaler_data *data);
void spl_build_isharp_1dlut_from_reference_curve(struct spl_fixed31_32 ratio, enum system_setup setup,
struct adaptive_sharpness sharpness, enum scale_to_sharpness_policy scale_to_sharpness_policy);
uint32_t *spl_get_pregen_filter_isharp_1D_lut(enum system_setup setup);
+
+// public API
+const uint16_t *spl_dscl_get_blur_scale_coeffs_64p(int taps);
+const uint16_t *spl_dscl_get_blur_scale_coeffs_64p_s1_10(int taps);
+
#endif /* __DC_SPL_ISHARP_FILTERS_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_easf_filters.c b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_easf_filters.c
index 09bf82f7d468..0d1bd81ff04a 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_easf_filters.c
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_easf_filters.c
@@ -1136,32 +1136,871 @@ static const uint16_t easf_filter_6tap_64p_ratio_1_00[198] = {
};
/* Converted scaler coeff tables from S1.10 to S1.12 */
-static uint16_t easf_filter_3tap_64p_ratio_0_30_s1_12[99];
-static uint16_t easf_filter_3tap_64p_ratio_0_40_s1_12[99];
-static uint16_t easf_filter_3tap_64p_ratio_0_50_s1_12[99];
-static uint16_t easf_filter_3tap_64p_ratio_0_60_s1_12[99];
-static uint16_t easf_filter_3tap_64p_ratio_0_70_s1_12[99];
-static uint16_t easf_filter_3tap_64p_ratio_0_80_s1_12[99];
-static uint16_t easf_filter_3tap_64p_ratio_0_90_s1_12[99];
-static uint16_t easf_filter_3tap_64p_ratio_1_00_s1_12[99];
-static uint16_t easf_filter_4tap_64p_ratio_0_30_s1_12[132];
-static uint16_t easf_filter_4tap_64p_ratio_0_40_s1_12[132];
-static uint16_t easf_filter_4tap_64p_ratio_0_50_s1_12[132];
-static uint16_t easf_filter_4tap_64p_ratio_0_60_s1_12[132];
-static uint16_t easf_filter_4tap_64p_ratio_0_70_s1_12[132];
-static uint16_t easf_filter_4tap_64p_ratio_0_80_s1_12[132];
-static uint16_t easf_filter_4tap_64p_ratio_0_90_s1_12[132];
-static uint16_t easf_filter_4tap_64p_ratio_1_00_s1_12[132];
-static uint16_t easf_filter_6tap_64p_ratio_0_30_s1_12[198];
-static uint16_t easf_filter_6tap_64p_ratio_0_40_s1_12[198];
-static uint16_t easf_filter_6tap_64p_ratio_0_50_s1_12[198];
-static uint16_t easf_filter_6tap_64p_ratio_0_60_s1_12[198];
-static uint16_t easf_filter_6tap_64p_ratio_0_70_s1_12[198];
-static uint16_t easf_filter_6tap_64p_ratio_0_80_s1_12[198];
-static uint16_t easf_filter_6tap_64p_ratio_0_90_s1_12[198];
-static uint16_t easf_filter_6tap_64p_ratio_1_00_s1_12[198];
-
-struct scale_ratio_to_reg_value_lookup easf_v_bf3_mode_lookup[] = {
+static const uint16_t easf_filter_3tap_64p_ratio_0_30_s1_12[99] = {
+0x0800, 0x0800, 0x0000,
+0x07d8, 0x0818, 0x0010,
+0x07b0, 0x082c, 0x0024,
+0x0788, 0x0844, 0x0034,
+0x0760, 0x0858, 0x0048,
+0x0738, 0x0870, 0x0058,
+0x0710, 0x0884, 0x006c,
+0x06e8, 0x0898, 0x0080,
+0x06c0, 0x08a8, 0x0098,
+0x0698, 0x08bc, 0x00ac,
+0x0670, 0x08cc, 0x00c4,
+0x0648, 0x08e0, 0x00d8,
+0x0620, 0x08f0, 0x00f0,
+0x05f8, 0x0900, 0x0108,
+0x05d0, 0x0910, 0x0120,
+0x05a8, 0x0920, 0x0138,
+0x0584, 0x0928, 0x0154,
+0x055c, 0x0938, 0x016c,
+0x0534, 0x0944, 0x0188,
+0x0510, 0x094c, 0x01a4,
+0x04e8, 0x0958, 0x01c0,
+0x04c4, 0x0960, 0x01dc,
+0x049c, 0x096c, 0x01f8,
+0x0478, 0x0970, 0x0218,
+0x0454, 0x0978, 0x0234,
+0x042c, 0x0980, 0x0254,
+0x0408, 0x0988, 0x0270,
+0x03e4, 0x098c, 0x0290,
+0x03c0, 0x0990, 0x02b0,
+0x039c, 0x0994, 0x02d0,
+0x037c, 0x0990, 0x02f4,
+0x0358, 0x0994, 0x0314,
+0x0334, 0x0998, 0x0334,
+};
+
+static const uint16_t easf_filter_3tap_64p_ratio_0_40_s1_12[99] = {
+0x0800, 0x0800, 0x0000,
+0x07d8, 0x0818, 0x0010,
+0x07ac, 0x0838, 0x001c,
+0x0784, 0x0850, 0x002c,
+0x075c, 0x0868, 0x003c,
+0x0734, 0x0880, 0x004c,
+0x0708, 0x0898, 0x0060,
+0x06e0, 0x08b0, 0x0070,
+0x06b8, 0x08c4, 0x0084,
+0x068c, 0x08dc, 0x0098,
+0x0664, 0x08f0, 0x00ac,
+0x063c, 0x0900, 0x00c4,
+0x0614, 0x0914, 0x00d8,
+0x05e8, 0x0928, 0x00f0,
+0x05c0, 0x093c, 0x0104,
+0x0598, 0x094c, 0x011c,
+0x0570, 0x095c, 0x0134,
+0x0548, 0x0968, 0x0150,
+0x0520, 0x0978, 0x0168,
+0x04f8, 0x0984, 0x0184,
+0x04d0, 0x0990, 0x01a0,
+0x04ac, 0x0998, 0x01bc,
+0x0484, 0x09a4, 0x01d8,
+0x045c, 0x09b0, 0x01f4,
+0x0438, 0x09b8, 0x0210,
+0x0410, 0x09c0, 0x0230,
+0x03ec, 0x09c4, 0x0250,
+0x03c8, 0x09c8, 0x0270,
+0x03a4, 0x09cc, 0x0290,
+0x0380, 0x09d0, 0x02b0,
+0x035c, 0x09d4, 0x02d0,
+0x0338, 0x09d4, 0x02f4,
+0x0314, 0x09d8, 0x0314,
+};
+
+static const uint16_t easf_filter_3tap_64p_ratio_0_50_s1_12[99] = {
+0x0800, 0x0800, 0x0000,
+0x07d4, 0x0824, 0x0008,
+0x07a8, 0x0844, 0x0014,
+0x077c, 0x0868, 0x001c,
+0x0750, 0x0888, 0x0028,
+0x0724, 0x08a8, 0x0034,
+0x06f8, 0x08c8, 0x0040,
+0x06cc, 0x08e4, 0x0050,
+0x06a0, 0x0904, 0x005c,
+0x0674, 0x0920, 0x006c,
+0x0648, 0x093c, 0x007c,
+0x061c, 0x0954, 0x0090,
+0x05f0, 0x0970, 0x00a0,
+0x05c4, 0x0988, 0x00b4,
+0x0598, 0x09a0, 0x00c8,
+0x056c, 0x09b8, 0x00dc,
+0x0540, 0x09cc, 0x00f4,
+0x0518, 0x09e0, 0x0108,
+0x04ec, 0x09f4, 0x0120,
+0x04c0, 0x0a08, 0x0138,
+0x0498, 0x0a18, 0x0150,
+0x046c, 0x0a28, 0x016c,
+0x0444, 0x0a34, 0x0188,
+0x041c, 0x0a40, 0x01a4,
+0x03f4, 0x0a4c, 0x01c0,
+0x03cc, 0x0a58, 0x01dc,
+0x03a4, 0x0a60, 0x01fc,
+0x037c, 0x0a68, 0x021c,
+0x0354, 0x0a70, 0x023c,
+0x0330, 0x0a74, 0x025c,
+0x030c, 0x0a78, 0x027c,
+0x02e8, 0x0a78, 0x02a0,
+0x02c4, 0x0a78, 0x02c4,
+};
+
+static const uint16_t easf_filter_3tap_64p_ratio_0_60_s1_12[99] = {
+0x0800, 0x0800, 0x0000,
+0x07d0, 0x082c, 0x0004,
+0x07a0, 0x0858, 0x0008,
+0x0770, 0x0884, 0x000c,
+0x0740, 0x08ac, 0x0014,
+0x0710, 0x08d4, 0x001c,
+0x06e0, 0x0900, 0x0020,
+0x06b0, 0x0924, 0x002c,
+0x0680, 0x094c, 0x0034,
+0x0650, 0x0970, 0x0040,
+0x0620, 0x0994, 0x004c,
+0x05f0, 0x09b8, 0x0058,
+0x05c0, 0x09dc, 0x0064,
+0x0590, 0x09fc, 0x0074,
+0x0560, 0x0a1c, 0x0084,
+0x0530, 0x0a3c, 0x0094,
+0x0500, 0x0a5c, 0x00a4,
+0x04d4, 0x0a74, 0x00b8,
+0x04a4, 0x0a90, 0x00cc,
+0x0474, 0x0aac, 0x00e0,
+0x0448, 0x0ac0, 0x00f8,
+0x041c, 0x0ad4, 0x0110,
+0x03f0, 0x0ae8, 0x0128,
+0x03c4, 0x0afc, 0x0140,
+0x0398, 0x0b0c, 0x015c,
+0x036c, 0x0b1c, 0x0178,
+0x0344, 0x0b28, 0x0194,
+0x031c, 0x0b30, 0x01b4,
+0x02f4, 0x0b38, 0x01d4,
+0x02cc, 0x0b40, 0x01f4,
+0x02a4, 0x0b48, 0x0214,
+0x0280, 0x0b48, 0x0238,
+0x025c, 0x0b48, 0x025c,
+};
+
+static const uint16_t easf_filter_3tap_64p_ratio_0_70_s1_12[99] = {
+0x0800, 0x0800, 0x0000,
+0x07cc, 0x0834, 0x0000,
+0x0794, 0x086c, 0x0000,
+0x0760, 0x08a0, 0x0000,
+0x072c, 0x08d4, 0x0000,
+0x06f4, 0x090c, 0x0000,
+0x06c0, 0x093c, 0x0004,
+0x0688, 0x0970, 0x0008,
+0x0654, 0x09a0, 0x000c,
+0x061c, 0x09d4, 0x0010,
+0x05e8, 0x0a00, 0x0018,
+0x05b4, 0x0a30, 0x001c,
+0x057c, 0x0a60, 0x0024,
+0x0548, 0x0a88, 0x0030,
+0x0514, 0x0ab4, 0x0038,
+0x04e0, 0x0adc, 0x0044,
+0x04ac, 0x0b00, 0x0054,
+0x0478, 0x0b28, 0x0060,
+0x0444, 0x0b4c, 0x0070,
+0x0414, 0x0b6c, 0x0080,
+0x03e0, 0x0b8c, 0x0094,
+0x03b0, 0x0ba8, 0x00a8,
+0x0380, 0x0bc4, 0x00bc,
+0x0354, 0x0bd8, 0x00d4,
+0x0324, 0x0bf0, 0x00ec,
+0x02f8, 0x0c04, 0x0104,
+0x02cc, 0x0c14, 0x0120,
+0x02a0, 0x0c24, 0x013c,
+0x0278, 0x0c30, 0x0158,
+0x0250, 0x0c38, 0x0178,
+0x0228, 0x0c40, 0x0198,
+0x0204, 0x0c40, 0x01bc,
+0x01dc, 0x0c48, 0x01dc,
+};
+
+static const uint16_t easf_filter_3tap_64p_ratio_0_80_s1_12[99] = {
+0x0800, 0x0800, 0x0000,
+0x07c4, 0x0840, 0x3ffc,
+0x0788, 0x0880, 0x3ff8,
+0x0748, 0x08c8, 0x3ff0,
+0x070c, 0x0904, 0x3ff0,
+0x06d0, 0x0944, 0x3fec,
+0x0690, 0x0988, 0x3fe8,
+0x0654, 0x09c4, 0x3fe8,
+0x0618, 0x0a04, 0x3fe4,
+0x05d8, 0x0a44, 0x3fe4,
+0x059c, 0x0a80, 0x3fe4,
+0x0560, 0x0ab8, 0x3fe8,
+0x0524, 0x0af4, 0x3fe8,
+0x04e8, 0x0b2c, 0x3fec,
+0x04b0, 0x0b5c, 0x3ff4,
+0x0474, 0x0b94, 0x3ff8,
+0x043c, 0x0bc4, 0x0000,
+0x0404, 0x0bf4, 0x0008,
+0x03cc, 0x0c20, 0x0014,
+0x0394, 0x0c4c, 0x0020,
+0x0360, 0x0c74, 0x002c,
+0x032c, 0x0c98, 0x003c,
+0x02f8, 0x0cbc, 0x004c,
+0x02c8, 0x0cdc, 0x005c,
+0x0298, 0x0cf8, 0x0070,
+0x0268, 0x0d14, 0x0084,
+0x023c, 0x0d28, 0x009c,
+0x0210, 0x0d3c, 0x00b4,
+0x01e4, 0x0d4c, 0x00d0,
+0x01bc, 0x0d58, 0x00ec,
+0x0194, 0x0d60, 0x010c,
+0x0170, 0x0d64, 0x012c,
+0x014c, 0x0d68, 0x014c,
+};
+
+static const uint16_t easf_filter_3tap_64p_ratio_0_90_s1_12[99] = {
+0x0800, 0x0800, 0x0000,
+0x07b8, 0x0850, 0x3ff8,
+0x0770, 0x08a0, 0x3ff0,
+0x0728, 0x08f0, 0x3fe8,
+0x06e4, 0x093c, 0x3fe0,
+0x069c, 0x0988, 0x3fdc,
+0x0654, 0x09d8, 0x3fd4,
+0x060c, 0x0a28, 0x3fcc,
+0x05c8, 0x0a70, 0x3fc8,
+0x0580, 0x0abc, 0x3fc4,
+0x053c, 0x0b08, 0x3fbc,
+0x04f8, 0x0b50, 0x3fb8,
+0x04b4, 0x0b94, 0x3fb8,
+0x0470, 0x0bdc, 0x3fb4,
+0x0430, 0x0c1c, 0x3fb4,
+0x03ec, 0x0c60, 0x3fb4,
+0x03b0, 0x0c9c, 0x3fb4,
+0x0370, 0x0cd8, 0x3fb8,
+0x0334, 0x0d10, 0x3fbc,
+0x02f8, 0x0d48, 0x3fc0,
+0x02c0, 0x0d78, 0x3fc8,
+0x0288, 0x0da8, 0x3fd0,
+0x0254, 0x0dd4, 0x3fd8,
+0x0220, 0x0dfc, 0x3fe4,
+0x01ec, 0x0e20, 0x3ff4,
+0x01bc, 0x0e44, 0x0000,
+0x0190, 0x0e5c, 0x0014,
+0x0164, 0x0e74, 0x0028,
+0x0138, 0x0e8c, 0x003c,
+0x0114, 0x0e98, 0x0054,
+0x00ec, 0x0ea4, 0x0070,
+0x00cc, 0x0ea8, 0x008c,
+0x00a8, 0x0eb0, 0x00a8,
+};
+
+static const uint16_t easf_filter_3tap_64p_ratio_1_00_s1_12[99] = {
+0x0800, 0x0800, 0x0000,
+0x07ac, 0x085c, 0x3ff8,
+0x0754, 0x08bc, 0x3ff0,
+0x0700, 0x091c, 0x3fe4,
+0x06ac, 0x0978, 0x3fdc,
+0x0658, 0x09d8, 0x3fd0,
+0x0604, 0x0a34, 0x3fc8,
+0x05b0, 0x0a94, 0x3fbc,
+0x0560, 0x0aec, 0x3fb4,
+0x0510, 0x0b44, 0x3fac,
+0x04c0, 0x0ba0, 0x3fa0,
+0x0470, 0x0bf8, 0x3f98,
+0x0424, 0x0c4c, 0x3f90,
+0x03d8, 0x0ca0, 0x3f88,
+0x0390, 0x0cf0, 0x3f80,
+0x0348, 0x0d3c, 0x3f7c,
+0x0300, 0x0d8c, 0x3f74,
+0x02c0, 0x0dd0, 0x3f70,
+0x027c, 0x0e14, 0x3f70,
+0x0240, 0x0e54, 0x3f6c,
+0x0204, 0x0e90, 0x3f6c,
+0x01c8, 0x0ecc, 0x3f6c,
+0x0190, 0x0f00, 0x3f70,
+0x015c, 0x0f30, 0x3f74,
+0x012c, 0x0f58, 0x3f7c,
+0x00fc, 0x0f80, 0x3f84,
+0x00d0, 0x0fa0, 0x3f90,
+0x00a8, 0x0fbc, 0x3f9c,
+0x0080, 0x0fd4, 0x3fac,
+0x005c, 0x0fe8, 0x3fbc,
+0x003c, 0x0ff4, 0x3fd0,
+0x001c, 0x0ffc, 0x3fe8,
+0x0000, 0x1000, 0x0000,
+};
+
+static const uint16_t easf_filter_4tap_64p_ratio_0_30_s1_12[132] = {
+0x0410, 0x07e0, 0x0410, 0x0000,
+0x03f8, 0x07dc, 0x0428, 0x0004,
+0x03e0, 0x07d8, 0x043c, 0x000c,
+0x03c8, 0x07d4, 0x0450, 0x0014,
+0x03ac, 0x07d0, 0x046c, 0x0018,
+0x0394, 0x07cc, 0x0480, 0x0020,
+0x037c, 0x07c8, 0x0494, 0x0028,
+0x0368, 0x07c0, 0x04a8, 0x0030,
+0x0350, 0x07b8, 0x04c0, 0x0038,
+0x0338, 0x07b4, 0x04d4, 0x0040,
+0x0320, 0x07ac, 0x04e8, 0x004c,
+0x0308, 0x07a4, 0x0500, 0x0054,
+0x02f4, 0x079c, 0x0514, 0x005c,
+0x02dc, 0x0794, 0x0528, 0x0068,
+0x02c4, 0x0788, 0x0544, 0x0070,
+0x02b0, 0x0780, 0x0554, 0x007c,
+0x029c, 0x0774, 0x0568, 0x0088,
+0x0284, 0x076c, 0x057c, 0x0094,
+0x0270, 0x0760, 0x0594, 0x009c,
+0x025c, 0x0754, 0x05a8, 0x00a8,
+0x0248, 0x0748, 0x05b8, 0x00b8,
+0x0230, 0x073c, 0x05d0, 0x00c4,
+0x021c, 0x0730, 0x05e4, 0x00d0,
+0x020c, 0x0724, 0x05f4, 0x00dc,
+0x01f8, 0x0714, 0x0608, 0x00ec,
+0x01e4, 0x0708, 0x061c, 0x00f8,
+0x01d0, 0x06f8, 0x0630, 0x0108,
+0x01c0, 0x06e8, 0x0640, 0x0118,
+0x01ac, 0x06dc, 0x0654, 0x0124,
+0x0198, 0x06cc, 0x0668, 0x0134,
+0x0188, 0x06bc, 0x0678, 0x0144,
+0x0178, 0x06ac, 0x0688, 0x0154,
+0x0168, 0x0698, 0x0698, 0x0168,
+};
+
+static const uint16_t easf_filter_4tap_64p_ratio_0_40_s1_12[132] = {
+0x03ec, 0x0824, 0x03f0, 0x0000,
+0x03d4, 0x0824, 0x0404, 0x0004,
+0x03b8, 0x0820, 0x0420, 0x0008,
+0x03a0, 0x081c, 0x0438, 0x000c,
+0x0388, 0x0818, 0x0450, 0x0010,
+0x036c, 0x0814, 0x0468, 0x0018,
+0x0354, 0x0810, 0x0480, 0x001c,
+0x033c, 0x080c, 0x0494, 0x0024,
+0x0324, 0x0804, 0x04b0, 0x0028,
+0x030c, 0x07fc, 0x04c8, 0x0030,
+0x02f4, 0x07f4, 0x04e0, 0x0038,
+0x02dc, 0x07ec, 0x04f8, 0x0040,
+0x02c4, 0x07e4, 0x0510, 0x0048,
+0x02b0, 0x07dc, 0x0524, 0x0050,
+0x0298, 0x07d0, 0x0540, 0x0058,
+0x0280, 0x07c8, 0x0558, 0x0060,
+0x026c, 0x07bc, 0x0570, 0x0068,
+0x0254, 0x07b0, 0x0588, 0x0074,
+0x0240, 0x07a4, 0x05a0, 0x007c,
+0x022c, 0x0798, 0x05b4, 0x0088,
+0x0214, 0x078c, 0x05cc, 0x0094,
+0x0200, 0x077c, 0x05e4, 0x00a0,
+0x01ec, 0x0770, 0x05f8, 0x00ac,
+0x01d8, 0x0760, 0x0610, 0x00b8,
+0x01c4, 0x0750, 0x0628, 0x00c4,
+0x01b4, 0x0744, 0x0638, 0x00d0,
+0x01a0, 0x0734, 0x064c, 0x00e0,
+0x018c, 0x0720, 0x0668, 0x00ec,
+0x017c, 0x0710, 0x0678, 0x00fc,
+0x016c, 0x0700, 0x068c, 0x0108,
+0x0158, 0x06ec, 0x06a4, 0x0118,
+0x0148, 0x06dc, 0x06b4, 0x0128,
+0x0138, 0x06c8, 0x06c8, 0x0138,
+};
+
+static const uint16_t easf_filter_4tap_64p_ratio_0_50_s1_12[132] = {
+0x0394, 0x08d8, 0x0394, 0x0000,
+0x0378, 0x08d4, 0x03b4, 0x0000,
+0x035c, 0x08d4, 0x03d0, 0x0000,
+0x0340, 0x08d4, 0x03ec, 0x0000,
+0x0324, 0x08d0, 0x0408, 0x0004,
+0x0308, 0x08cc, 0x0428, 0x0004,
+0x02f0, 0x08c8, 0x0444, 0x0004,
+0x02d4, 0x08c0, 0x0464, 0x0008,
+0x02b8, 0x08bc, 0x0484, 0x0008,
+0x02a0, 0x08b4, 0x04a0, 0x000c,
+0x0288, 0x08ac, 0x04bc, 0x0010,
+0x026c, 0x08a4, 0x04dc, 0x0014,
+0x0254, 0x0898, 0x04fc, 0x0018,
+0x023c, 0x0890, 0x0518, 0x001c,
+0x0224, 0x0884, 0x0538, 0x0020,
+0x020c, 0x0878, 0x0554, 0x0028,
+0x01f8, 0x086c, 0x0570, 0x002c,
+0x01e0, 0x085c, 0x0590, 0x0034,
+0x01c8, 0x084c, 0x05b4, 0x0038,
+0x01b4, 0x0840, 0x05cc, 0x0040,
+0x01a0, 0x0830, 0x05e8, 0x0048,
+0x018c, 0x081c, 0x0608, 0x0050,
+0x0178, 0x080c, 0x0624, 0x0058,
+0x0164, 0x07f8, 0x0644, 0x0060,
+0x0150, 0x07e4, 0x0660, 0x006c,
+0x0140, 0x07d0, 0x067c, 0x0074,
+0x012c, 0x07bc, 0x0698, 0x0080,
+0x011c, 0x07a8, 0x06b0, 0x008c,
+0x010c, 0x0790, 0x06cc, 0x0098,
+0x00fc, 0x077c, 0x06e4, 0x00a4,
+0x00ec, 0x0764, 0x0700, 0x00b0,
+0x00dc, 0x074c, 0x0718, 0x00c0,
+0x00cc, 0x0734, 0x0734, 0x00cc,
+};
+
+static const uint16_t easf_filter_4tap_64p_ratio_0_60_s1_12[132] = {
+0x0320, 0x09bc, 0x0324, 0x0000,
+0x0300, 0x09c0, 0x0344, 0x3ffc,
+0x02e0, 0x09c0, 0x0364, 0x3ffc,
+0x02c4, 0x09c0, 0x0384, 0x3ff8,
+0x02a4, 0x09bc, 0x03ac, 0x3ff4,
+0x0288, 0x09b8, 0x03cc, 0x3ff4,
+0x0268, 0x09b4, 0x03f4, 0x3ff0,
+0x024c, 0x09b0, 0x0414, 0x3ff0,
+0x0230, 0x09a8, 0x043c, 0x3fec,
+0x0214, 0x09a0, 0x0460, 0x3fec,
+0x01f8, 0x0994, 0x0488, 0x3fec,
+0x01e0, 0x098c, 0x04a8, 0x3fec,
+0x01c4, 0x0980, 0x04d0, 0x3fec,
+0x01ac, 0x0970, 0x04f8, 0x3fec,
+0x0194, 0x0964, 0x051c, 0x3fec,
+0x017c, 0x0954, 0x0544, 0x3fec,
+0x0164, 0x0944, 0x0568, 0x3ff0,
+0x0150, 0x0934, 0x058c, 0x3ff0,
+0x0138, 0x0920, 0x05b4, 0x3ff4,
+0x0124, 0x090c, 0x05d8, 0x3ff8,
+0x0110, 0x08f8, 0x05fc, 0x3ffc,
+0x00fc, 0x08e0, 0x0624, 0x0000,
+0x00e8, 0x08c8, 0x064c, 0x0004,
+0x00d8, 0x08b0, 0x0670, 0x0008,
+0x00c4, 0x0898, 0x0694, 0x0010,
+0x00b4, 0x087c, 0x06bc, 0x0014,
+0x00a4, 0x0860, 0x06e0, 0x001c,
+0x0094, 0x0844, 0x0704, 0x0024,
+0x0088, 0x0828, 0x0724, 0x002c,
+0x0078, 0x080c, 0x0748, 0x0034,
+0x006c, 0x07ec, 0x0768, 0x0040,
+0x0060, 0x07cc, 0x078c, 0x0048,
+0x0054, 0x07ac, 0x07ac, 0x0054,
+};
+
+static const uint16_t easf_filter_4tap_64p_ratio_0_70_s1_12[132] = {
+0x028c, 0x0ae4, 0x0290, 0x0000,
+0x0268, 0x0ae8, 0x02b4, 0x3ffc,
+0x0248, 0x0ae8, 0x02d8, 0x3ff8,
+0x0224, 0x0ae8, 0x0304, 0x3ff0,
+0x0204, 0x0ae4, 0x032c, 0x3fec,
+0x01e4, 0x0ae0, 0x0354, 0x3fe8,
+0x01c4, 0x0adc, 0x037c, 0x3fe4,
+0x01a4, 0x0ad4, 0x03a8, 0x3fe0,
+0x0188, 0x0acc, 0x03d0, 0x3fdc,
+0x016c, 0x0ac0, 0x03fc, 0x3fd8,
+0x0150, 0x0ab4, 0x042c, 0x3fd0,
+0x0134, 0x0aa4, 0x045c, 0x3fcc,
+0x0118, 0x0a94, 0x048c, 0x3fc8,
+0x0100, 0x0a84, 0x04b4, 0x3fc8,
+0x00e8, 0x0a70, 0x04e4, 0x3fc4,
+0x00d0, 0x0a5c, 0x0514, 0x3fc0,
+0x00bc, 0x0a48, 0x0540, 0x3fbc,
+0x00a4, 0x0a30, 0x0570, 0x3fbc,
+0x0090, 0x0a14, 0x05a4, 0x3fb8,
+0x007c, 0x09fc, 0x05d0, 0x3fb8,
+0x006c, 0x09e0, 0x05fc, 0x3fb8,
+0x0058, 0x09c0, 0x0634, 0x3fb4,
+0x0048, 0x09a0, 0x0664, 0x3fb4,
+0x0038, 0x0980, 0x0690, 0x3fb8,
+0x002c, 0x0960, 0x06bc, 0x3fb8,
+0x001c, 0x093c, 0x06f0, 0x3fb8,
+0x0010, 0x0918, 0x071c, 0x3fbc,
+0x0004, 0x08f4, 0x074c, 0x3fbc,
+0x3ff8, 0x08cc, 0x077c, 0x3fc0,
+0x3ff0, 0x08a4, 0x07a8, 0x3fc4,
+0x3fe8, 0x087c, 0x07d0, 0x3fcc,
+0x3fe0, 0x0854, 0x07fc, 0x3fd0,
+0x3fd8, 0x0828, 0x0828, 0x3fd8,
+};
+
+static const uint16_t easf_filter_4tap_64p_ratio_0_80_s1_12[132] = {
+0x01d4, 0x0c54, 0x01d8, 0x0000,
+0x01b0, 0x0c58, 0x01fc, 0x3ffc,
+0x0188, 0x0c58, 0x0228, 0x3ff8,
+0x0164, 0x0c54, 0x0258, 0x3ff0,
+0x0140, 0x0c50, 0x0284, 0x3fec,
+0x0120, 0x0c48, 0x02b4, 0x3fe4,
+0x0100, 0x0c40, 0x02e0, 0x3fe0,
+0x00e0, 0x0c34, 0x0314, 0x3fd8,
+0x00c0, 0x0c28, 0x0344, 0x3fd4,
+0x00a4, 0x0c18, 0x0378, 0x3fcc,
+0x0088, 0x0c04, 0x03ac, 0x3fc8,
+0x0070, 0x0bf0, 0x03e0, 0x3fc0,
+0x0054, 0x0bdc, 0x0418, 0x3fb8,
+0x0040, 0x0bc4, 0x0448, 0x3fb4,
+0x0028, 0x0ba8, 0x0484, 0x3fac,
+0x0014, 0x0b8c, 0x04bc, 0x3fa4,
+0x0000, 0x0b6c, 0x04f4, 0x3fa0,
+0x3fec, 0x0b4c, 0x0530, 0x3f98,
+0x3fdc, 0x0b28, 0x0568, 0x3f94,
+0x3fcc, 0x0b04, 0x05a4, 0x3f8c,
+0x3fc0, 0x0adc, 0x05dc, 0x3f88,
+0x3fb0, 0x0ab4, 0x0618, 0x3f84,
+0x3fa4, 0x0a88, 0x0658, 0x3f7c,
+0x3f9c, 0x0a5c, 0x0690, 0x3f78,
+0x3f90, 0x0a30, 0x06cc, 0x3f74,
+0x3f88, 0x0a00, 0x0708, 0x3f70,
+0x3f80, 0x09d0, 0x0740, 0x3f70,
+0x3f7c, 0x09a0, 0x0778, 0x3f6c,
+0x3f74, 0x096c, 0x07b8, 0x3f68,
+0x3f70, 0x0938, 0x07f0, 0x3f68,
+0x3f6c, 0x0904, 0x0828, 0x3f68,
+0x3f6c, 0x08cc, 0x0860, 0x3f68,
+0x3f68, 0x0898, 0x0898, 0x3f68,
+};
+
+static const uint16_t easf_filter_4tap_64p_ratio_0_90_s1_12[132] = {
+0x00fc, 0x0e0c, 0x00f8, 0x0000,
+0x00d0, 0x0e0c, 0x0128, 0x3ffc,
+0x00ac, 0x0e0c, 0x0150, 0x3ff8,
+0x0084, 0x0e04, 0x0184, 0x3ff4,
+0x0064, 0x0dfc, 0x01b0, 0x3ff0,
+0x0040, 0x0df0, 0x01e4, 0x3fec,
+0x0020, 0x0de0, 0x0218, 0x3fe8,
+0x0004, 0x0dd0, 0x024c, 0x3fe0,
+0x3fe8, 0x0db8, 0x0284, 0x3fdc,
+0x3fcc, 0x0da0, 0x02c0, 0x3fd4,
+0x3fb4, 0x0d84, 0x02fc, 0x3fcc,
+0x3fa0, 0x0d68, 0x0334, 0x3fc4,
+0x3f88, 0x0d48, 0x0370, 0x3fc0,
+0x3f78, 0x0d24, 0x03ac, 0x3fb8,
+0x3f64, 0x0cfc, 0x03f0, 0x3fb0,
+0x3f54, 0x0cd4, 0x0434, 0x3fa4,
+0x3f48, 0x0ca8, 0x0474, 0x3f9c,
+0x3f3c, 0x0c78, 0x04b8, 0x3f94,
+0x3f30, 0x0c48, 0x04fc, 0x3f8c,
+0x3f28, 0x0c14, 0x0540, 0x3f84,
+0x3f20, 0x0be0, 0x0588, 0x3f78,
+0x3f18, 0x0ba8, 0x05d0, 0x3f70,
+0x3f14, 0x0b70, 0x0614, 0x3f68,
+0x3f10, 0x0b34, 0x065c, 0x3f60,
+0x3f0c, 0x0af8, 0x06a8, 0x3f54,
+0x3f0c, 0x0abc, 0x06ec, 0x3f4c,
+0x3f0c, 0x0a7c, 0x0734, 0x3f44,
+0x3f0c, 0x0a38, 0x0780, 0x3f3c,
+0x3f0c, 0x09f8, 0x07c8, 0x3f34,
+0x3f10, 0x09b4, 0x080c, 0x3f30,
+0x3f14, 0x0970, 0x0854, 0x3f28,
+0x3f18, 0x092c, 0x089c, 0x3f20,
+0x3f1c, 0x08e4, 0x08e4, 0x3f1c,
+};
+
+static const uint16_t easf_filter_4tap_64p_ratio_1_00_s1_12[132] = {
+0x0000, 0x1000, 0x0000, 0x0000,
+0x3fd8, 0x0ffc, 0x002c, 0x0000,
+0x3fb4, 0x0ff8, 0x0054, 0x0000,
+0x3f90, 0x0fec, 0x0088, 0x3ffc,
+0x3f70, 0x0fdc, 0x00b8, 0x3ffc,
+0x3f54, 0x0fc8, 0x00ec, 0x3ff8,
+0x3f38, 0x0fb0, 0x0120, 0x3ff8,
+0x3f20, 0x0f94, 0x0158, 0x3ff4,
+0x3f0c, 0x0f70, 0x0194, 0x3ff0,
+0x3ef8, 0x0f4c, 0x01d4, 0x3fe8,
+0x3ee4, 0x0f24, 0x0214, 0x3fe4,
+0x3ed8, 0x0ef8, 0x0250, 0x3fe0,
+0x3ec8, 0x0ec8, 0x0298, 0x3fd8,
+0x3ec0, 0x0e94, 0x02dc, 0x3fd0,
+0x3eb4, 0x0e5c, 0x0328, 0x3fc8,
+0x3eac, 0x0e24, 0x0370, 0x3fc0,
+0x3ea8, 0x0de4, 0x03bc, 0x3fb8,
+0x3ea4, 0x0da4, 0x0408, 0x3fb0,
+0x3ea4, 0x0d64, 0x0454, 0x3fa4,
+0x3ea4, 0x0d20, 0x04a4, 0x3f98,
+0x3ea4, 0x0cd8, 0x04f4, 0x3f90,
+0x3ea4, 0x0c8c, 0x054c, 0x3f84,
+0x3ea8, 0x0c40, 0x05a0, 0x3f78,
+0x3eb0, 0x0bf4, 0x05f0, 0x3f6c,
+0x3eb4, 0x0ba4, 0x0648, 0x3f60,
+0x3ebc, 0x0b54, 0x069c, 0x3f54,
+0x3ec4, 0x0b00, 0x06f4, 0x3f48,
+0x3ecc, 0x0ab0, 0x0748, 0x3f3c,
+0x3ed4, 0x0a58, 0x07a4, 0x3f30,
+0x3ee0, 0x0a04, 0x07f8, 0x3f24,
+0x3ee8, 0x09b0, 0x0850, 0x3f18,
+0x3ef4, 0x0958, 0x08a8, 0x3f0c,
+0x3f00, 0x0900, 0x0900, 0x3f00,
+};
+
+static const uint16_t easf_filter_6tap_64p_ratio_0_30_s1_12[198] = {
+0x012c, 0x0400, 0x05a4, 0x0404, 0x012c, 0x0000,
+0x0124, 0x03f4, 0x05a4, 0x040c, 0x0138, 0x0000,
+0x011c, 0x03e8, 0x05a4, 0x0418, 0x0140, 0x0000,
+0x0114, 0x03dc, 0x05a0, 0x0424, 0x0148, 0x0004,
+0x010c, 0x03d4, 0x05a0, 0x042c, 0x0150, 0x0004,
+0x0100, 0x03c8, 0x05a0, 0x0438, 0x015c, 0x0004,
+0x00f8, 0x03bc, 0x05a0, 0x0440, 0x0164, 0x0008,
+0x00f0, 0x03b0, 0x059c, 0x044c, 0x0170, 0x0008,
+0x00e8, 0x03a4, 0x059c, 0x0458, 0x0178, 0x0008,
+0x00e0, 0x0398, 0x0598, 0x0460, 0x0184, 0x000c,
+0x00d8, 0x038c, 0x0594, 0x0470, 0x018c, 0x000c,
+0x00d0, 0x0380, 0x0594, 0x0474, 0x0198, 0x0010,
+0x00cc, 0x0374, 0x0590, 0x0480, 0x01a0, 0x0010,
+0x00c4, 0x0368, 0x058c, 0x0488, 0x01ac, 0x0014,
+0x00bc, 0x035c, 0x058c, 0x0494, 0x01b4, 0x0014,
+0x00b4, 0x034c, 0x0588, 0x04a0, 0x01c0, 0x0018,
+0x00ac, 0x0340, 0x0584, 0x04a8, 0x01cc, 0x001c,
+0x00a8, 0x0334, 0x0580, 0x04b4, 0x01d4, 0x001c,
+0x00a0, 0x0328, 0x057c, 0x04bc, 0x01e0, 0x0020,
+0x0098, 0x031c, 0x0578, 0x04c4, 0x01ec, 0x0024,
+0x0094, 0x0310, 0x0574, 0x04cc, 0x01f8, 0x0024,
+0x008c, 0x0304, 0x0570, 0x04d8, 0x0200, 0x0028,
+0x0088, 0x02f8, 0x0568, 0x04e0, 0x020c, 0x002c,
+0x0080, 0x02ec, 0x0564, 0x04e8, 0x0218, 0x0030,
+0x007c, 0x02e0, 0x0560, 0x04ec, 0x0224, 0x0034,
+0x0078, 0x02d4, 0x0558, 0x04f8, 0x0230, 0x0034,
+0x0070, 0x02c8, 0x0554, 0x0500, 0x023c, 0x0038,
+0x006c, 0x02bc, 0x054c, 0x050c, 0x0244, 0x003c,
+0x0064, 0x02b0, 0x0548, 0x0514, 0x0250, 0x0040,
+0x0060, 0x02a4, 0x0540, 0x051c, 0x025c, 0x0044,
+0x005c, 0x0298, 0x053c, 0x0520, 0x0268, 0x0048,
+0x0058, 0x028c, 0x0534, 0x0524, 0x0274, 0x0050,
+0x0054, 0x0280, 0x052c, 0x052c, 0x0280, 0x0054,
+};
+
+static const uint16_t easf_filter_6tap_64p_ratio_0_40_s1_12[198] = {
+0x00a0, 0x0418, 0x068c, 0x041c, 0x00a0, 0x0000,
+0x0098, 0x0408, 0x068c, 0x0428, 0x00ac, 0x0000,
+0x0090, 0x03f8, 0x068c, 0x043c, 0x00b4, 0x3ffc,
+0x0088, 0x03e8, 0x068c, 0x044c, 0x00bc, 0x3ffc,
+0x0084, 0x03d8, 0x068c, 0x0458, 0x00c4, 0x3ffc,
+0x007c, 0x03c8, 0x0688, 0x046c, 0x00d0, 0x3ff8,
+0x0074, 0x03b8, 0x0688, 0x047c, 0x00d8, 0x3ff8,
+0x006c, 0x03a8, 0x0684, 0x048c, 0x00e4, 0x3ff8,
+0x0064, 0x0398, 0x0684, 0x049c, 0x00ec, 0x3ff8,
+0x0060, 0x0388, 0x0680, 0x04a8, 0x00f8, 0x3ff8,
+0x0058, 0x0378, 0x0680, 0x04b8, 0x0104, 0x3ff4,
+0x0054, 0x0368, 0x067c, 0x04c8, 0x010c, 0x3ff4,
+0x004c, 0x0358, 0x0678, 0x04d8, 0x0118, 0x3ff4,
+0x0048, 0x0348, 0x0674, 0x04e4, 0x0124, 0x3ff4,
+0x0040, 0x0338, 0x0670, 0x04f4, 0x0130, 0x3ff4,
+0x003c, 0x0328, 0x0668, 0x0504, 0x013c, 0x3ff4,
+0x0038, 0x0318, 0x0664, 0x0510, 0x0148, 0x3ff4,
+0x0034, 0x0308, 0x065c, 0x0520, 0x0154, 0x3ff4,
+0x002c, 0x02f8, 0x0658, 0x0530, 0x0160, 0x3ff4,
+0x0028, 0x02e8, 0x0654, 0x053c, 0x016c, 0x3ff4,
+0x0024, 0x02d8, 0x064c, 0x054c, 0x0178, 0x3ff4,
+0x0020, 0x02c8, 0x0644, 0x055c, 0x0184, 0x3ff4,
+0x001c, 0x02b8, 0x0640, 0x0568, 0x0190, 0x3ff4,
+0x0018, 0x02a8, 0x0638, 0x0574, 0x01a0, 0x3ff4,
+0x0014, 0x0298, 0x0630, 0x0584, 0x01ac, 0x3ff4,
+0x0014, 0x0288, 0x0624, 0x0590, 0x01bc, 0x3ff4,
+0x0010, 0x0278, 0x061c, 0x059c, 0x01c8, 0x3ff8,
+0x000c, 0x0268, 0x0614, 0x05ac, 0x01d4, 0x3ff8,
+0x0008, 0x0258, 0x060c, 0x05b8, 0x01e4, 0x3ff8,
+0x0008, 0x024c, 0x0600, 0x05bc, 0x01f4, 0x3ffc,
+0x0004, 0x023c, 0x05f8, 0x05cc, 0x0200, 0x3ffc,
+0x0004, 0x022c, 0x05ec, 0x05d4, 0x0210, 0x0000,
+0x0000, 0x021c, 0x05e4, 0x05e4, 0x021c, 0x0000,
+};
+
+static const uint16_t easf_filter_6tap_64p_ratio_0_50_s1_12[198] = {
+0x0000, 0x041c, 0x07cc, 0x0418, 0x0000, 0x0000,
+0x3ff8, 0x0404, 0x07cc, 0x0434, 0x0008, 0x3ffc,
+0x3ff4, 0x03ec, 0x07cc, 0x044c, 0x000c, 0x3ffc,
+0x3ff0, 0x03d8, 0x07cc, 0x0460, 0x0014, 0x3ff8,
+0x3fe8, 0x03c0, 0x07cc, 0x0478, 0x001c, 0x3ff8,
+0x3fe4, 0x03ac, 0x07c8, 0x0490, 0x0024, 0x3ff4,
+0x3fe0, 0x0394, 0x07c8, 0x04a4, 0x002c, 0x3ff4,
+0x3fdc, 0x0380, 0x07c4, 0x04bc, 0x0034, 0x3ff0,
+0x3fd8, 0x0368, 0x07c0, 0x04d4, 0x0040, 0x3fec,
+0x3fd4, 0x0350, 0x07bc, 0x04ec, 0x0048, 0x3fec,
+0x3fd0, 0x033c, 0x07b8, 0x0504, 0x0050, 0x3fe8,
+0x3fcc, 0x0324, 0x07b4, 0x051c, 0x005c, 0x3fe4,
+0x3fc8, 0x0310, 0x07ac, 0x0530, 0x0068, 0x3fe4,
+0x3fc4, 0x02fc, 0x07a8, 0x0548, 0x0070, 0x3fe0,
+0x3fc4, 0x02e4, 0x07a0, 0x055c, 0x007c, 0x3fe0,
+0x3fc0, 0x02d0, 0x0798, 0x0574, 0x0088, 0x3fdc,
+0x3fc0, 0x02b8, 0x0790, 0x058c, 0x0094, 0x3fd8,
+0x3fbc, 0x02a4, 0x0788, 0x05a0, 0x00a0, 0x3fd8,
+0x3fbc, 0x0290, 0x077c, 0x05b8, 0x00ac, 0x3fd4,
+0x3fbc, 0x027c, 0x0774, 0x05c8, 0x00b8, 0x3fd4,
+0x3fb8, 0x0268, 0x0768, 0x05e0, 0x00c8, 0x3fd0,
+0x3fb8, 0x0250, 0x0760, 0x05f8, 0x00d4, 0x3fcc,
+0x3fb8, 0x023c, 0x0754, 0x0608, 0x00e4, 0x3fcc,
+0x3fb8, 0x0228, 0x0748, 0x0620, 0x00f0, 0x3fc8,
+0x3fb8, 0x0214, 0x073c, 0x0630, 0x0100, 0x3fc8,
+0x3fb8, 0x0204, 0x072c, 0x0644, 0x0110, 0x3fc4,
+0x3fb8, 0x01f0, 0x0720, 0x0658, 0x011c, 0x3fc4,
+0x3fb8, 0x01dc, 0x0710, 0x0670, 0x012c, 0x3fc0,
+0x3fb8, 0x01c8, 0x0704, 0x0680, 0x013c, 0x3fc0,
+0x3fb8, 0x01b8, 0x06f4, 0x0690, 0x014c, 0x3fc0,
+0x3fb8, 0x01a4, 0x06e4, 0x06a4, 0x0160, 0x3fbc,
+0x3fb8, 0x0194, 0x06d4, 0x06b4, 0x0170, 0x3fbc,
+0x3fbc, 0x0180, 0x06c4, 0x06c4, 0x0180, 0x3fbc,
+};
+
+static const uint16_t easf_filter_6tap_64p_ratio_0_60_s1_12[198] = {
+0x3f64, 0x03ec, 0x0960, 0x03ec, 0x3f64, 0x0000,
+0x3f64, 0x03cc, 0x0960, 0x0408, 0x3f68, 0x0000,
+0x3f60, 0x03ac, 0x0960, 0x042c, 0x3f6c, 0x3ffc,
+0x3f60, 0x038c, 0x0960, 0x0448, 0x3f70, 0x3ffc,
+0x3f60, 0x0370, 0x095c, 0x046c, 0x3f70, 0x3ff8,
+0x3f5c, 0x0350, 0x0958, 0x048c, 0x3f78, 0x3ff8,
+0x3f5c, 0x0334, 0x0954, 0x04ac, 0x3f7c, 0x3ff4,
+0x3f5c, 0x0314, 0x0950, 0x04cc, 0x3f80, 0x3ff4,
+0x3f5c, 0x02f8, 0x0948, 0x04f0, 0x3f84, 0x3ff0,
+0x3f5c, 0x02d8, 0x0944, 0x050c, 0x3f8c, 0x3ff0,
+0x3f60, 0x02bc, 0x093c, 0x052c, 0x3f90, 0x3fec,
+0x3f60, 0x02a0, 0x0930, 0x0550, 0x3f98, 0x3fe8,
+0x3f60, 0x0284, 0x0928, 0x056c, 0x3fa0, 0x3fe8,
+0x3f64, 0x0268, 0x091c, 0x058c, 0x3fa8, 0x3fe4,
+0x3f64, 0x024c, 0x0910, 0x05b0, 0x3fb0, 0x3fe0,
+0x3f64, 0x0230, 0x0904, 0x05d0, 0x3fbc, 0x3fdc,
+0x3f68, 0x0214, 0x08f8, 0x05ec, 0x3fc4, 0x3fdc,
+0x3f6c, 0x01fc, 0x08e8, 0x060c, 0x3fcc, 0x3fd8,
+0x3f6c, 0x01e0, 0x08dc, 0x062c, 0x3fd8, 0x3fd4,
+0x3f70, 0x01c8, 0x08cc, 0x0648, 0x3fe4, 0x3fd0,
+0x3f74, 0x01b0, 0x08bc, 0x0664, 0x3ff0, 0x3fcc,
+0x3f74, 0x0194, 0x08a8, 0x068c, 0x3ffc, 0x3fc8,
+0x3f78, 0x017c, 0x0898, 0x06a8, 0x0008, 0x3fc4,
+0x3f7c, 0x0168, 0x0884, 0x06c0, 0x0018, 0x3fc0,
+0x3f80, 0x0150, 0x0870, 0x06dc, 0x0024, 0x3fc0,
+0x3f84, 0x0138, 0x085c, 0x06f8, 0x0034, 0x3fbc,
+0x3f88, 0x0120, 0x0848, 0x0718, 0x0040, 0x3fb8,
+0x3f8c, 0x010c, 0x0830, 0x0734, 0x0050, 0x3fb4,
+0x3f90, 0x00f8, 0x081c, 0x074c, 0x0060, 0x3fb0,
+0x3f94, 0x00e4, 0x0800, 0x0768, 0x0074, 0x3fac,
+0x3f98, 0x00d0, 0x07e8, 0x0784, 0x0084, 0x3fa8,
+0x3f9c, 0x00bc, 0x07d4, 0x079c, 0x0094, 0x3fa4,
+0x3fa0, 0x00a8, 0x07b8, 0x07b8, 0x00a8, 0x3fa0,
+};
+
+static const uint16_t easf_filter_6tap_64p_ratio_0_70_s1_12[198] = {
+0x3f00, 0x0368, 0x0b30, 0x0368, 0x3f00, 0x0000,
+0x3f04, 0x0340, 0x0b30, 0x0390, 0x3efc, 0x0000,
+0x3f08, 0x0318, 0x0b2c, 0x03bc, 0x3ef8, 0x0000,
+0x3f0c, 0x02f0, 0x0b28, 0x03e4, 0x3ef8, 0x0000,
+0x3f10, 0x02c8, 0x0b24, 0x0410, 0x3ef4, 0x0000,
+0x3f14, 0x02a0, 0x0b1c, 0x043c, 0x3ef4, 0x0000,
+0x3f1c, 0x027c, 0x0b14, 0x0464, 0x3ef0, 0x0000,
+0x3f20, 0x0254, 0x0b0c, 0x0490, 0x3ef0, 0x0000,
+0x3f24, 0x0230, 0x0b00, 0x04bc, 0x3ef0, 0x0000,
+0x3f2c, 0x020c, 0x0af4, 0x04e4, 0x3ef0, 0x0000,
+0x3f30, 0x01e8, 0x0ae8, 0x0510, 0x3ef0, 0x0000,
+0x3f38, 0x01c8, 0x0ad8, 0x0534, 0x3ef4, 0x0000,
+0x3f40, 0x01a4, 0x0ac8, 0x0564, 0x3ef4, 0x3ffc,
+0x3f44, 0x0184, 0x0ab4, 0x0590, 0x3ef8, 0x3ffc,
+0x3f4c, 0x0164, 0x0aa4, 0x05b8, 0x3efc, 0x3ff8,
+0x3f50, 0x0144, 0x0a90, 0x05e8, 0x3efc, 0x3ff8,
+0x3f58, 0x0124, 0x0a78, 0x0610, 0x3f04, 0x3ff8,
+0x3f60, 0x0108, 0x0a64, 0x0638, 0x3f08, 0x3ff4,
+0x3f64, 0x00e8, 0x0a4c, 0x066c, 0x3f0c, 0x3ff0,
+0x3f6c, 0x00cc, 0x0a34, 0x0690, 0x3f14, 0x3ff0,
+0x3f70, 0x00b4, 0x0a18, 0x06bc, 0x3f1c, 0x3fec,
+0x3f78, 0x0098, 0x0a00, 0x06e8, 0x3f20, 0x3fe8,
+0x3f80, 0x007c, 0x09e4, 0x0710, 0x3f2c, 0x3fe4,
+0x3f84, 0x0064, 0x09c8, 0x0738, 0x3f34, 0x3fe4,
+0x3f8c, 0x004c, 0x09a8, 0x0764, 0x3f3c, 0x3fe0,
+0x3f90, 0x0034, 0x098c, 0x078c, 0x3f48, 0x3fdc,
+0x3f98, 0x0020, 0x096c, 0x07b0, 0x3f54, 0x3fd8,
+0x3f9c, 0x0008, 0x094c, 0x07dc, 0x3f60, 0x3fd4,
+0x3fa4, 0x3ff4, 0x0928, 0x0808, 0x3f6c, 0x3fcc,
+0x3fa8, 0x3fe0, 0x0908, 0x082c, 0x3f7c, 0x3fc8,
+0x3fb0, 0x3fcc, 0x08e4, 0x0854, 0x3f88, 0x3fc4,
+0x3fb4, 0x3fbc, 0x08c0, 0x0878, 0x3f98, 0x3fc0,
+0x3fbc, 0x3fac, 0x0898, 0x0898, 0x3fac, 0x3fbc,
+};
+
+static const uint16_t easf_filter_6tap_64p_ratio_0_80_s1_12[198] = {
+0x3efc, 0x0284, 0x0d00, 0x0284, 0x3efc, 0x0000,
+0x3f04, 0x0254, 0x0d00, 0x02b4, 0x3ef0, 0x0004,
+0x3f10, 0x0224, 0x0cf8, 0x02e8, 0x3ee8, 0x0004,
+0x3f18, 0x01f4, 0x0cf4, 0x0318, 0x3ee0, 0x0008,
+0x3f24, 0x01c8, 0x0ce8, 0x034c, 0x3ed8, 0x0008,
+0x3f30, 0x019c, 0x0ce0, 0x037c, 0x3ecc, 0x000c,
+0x3f38, 0x0170, 0x0cd0, 0x03b8, 0x3ec4, 0x000c,
+0x3f44, 0x0144, 0x0cc4, 0x03e8, 0x3ebc, 0x0010,
+0x3f4c, 0x011c, 0x0cb4, 0x0420, 0x3eb4, 0x0010,
+0x3f58, 0x00f4, 0x0ca0, 0x0458, 0x3eac, 0x0010,
+0x3f60, 0x00cc, 0x0c8c, 0x048c, 0x3ea8, 0x0014,
+0x3f6c, 0x00a8, 0x0c74, 0x04c4, 0x3ea0, 0x0014,
+0x3f74, 0x0084, 0x0c5c, 0x04fc, 0x3e9c, 0x0014,
+0x3f7c, 0x0060, 0x0c44, 0x0534, 0x3e94, 0x0018,
+0x3f88, 0x0040, 0x0c28, 0x0568, 0x3e90, 0x0018,
+0x3f90, 0x0020, 0x0c08, 0x05a4, 0x3e8c, 0x0018,
+0x3f98, 0x0000, 0x0bec, 0x05dc, 0x3e88, 0x0018,
+0x3fa0, 0x3fe4, 0x0bcc, 0x0614, 0x3e84, 0x0018,
+0x3fac, 0x3fc4, 0x0ba8, 0x064c, 0x3e84, 0x0018,
+0x3fb4, 0x3fac, 0x0b84, 0x0684, 0x3e80, 0x0018,
+0x3fb8, 0x3f90, 0x0b60, 0x06c0, 0x3e80, 0x0018,
+0x3fc0, 0x3f78, 0x0b38, 0x06f8, 0x3e80, 0x0018,
+0x3fc8, 0x3f60, 0x0b14, 0x072c, 0x3e80, 0x0018,
+0x3fd0, 0x3f4c, 0x0ae8, 0x0760, 0x3e84, 0x0018,
+0x3fd8, 0x3f34, 0x0ac0, 0x079c, 0x3e84, 0x0014,
+0x3fdc, 0x3f20, 0x0a94, 0x07d4, 0x3e88, 0x0014,
+0x3fe4, 0x3f10, 0x0a68, 0x0808, 0x3e8c, 0x0010,
+0x3fe8, 0x3f00, 0x0a38, 0x0840, 0x3e90, 0x0010,
+0x3fec, 0x3ef0, 0x0a0c, 0x0874, 0x3e98, 0x000c,
+0x3ff4, 0x3ee0, 0x09d8, 0x08a8, 0x3ea0, 0x000c,
+0x3ff8, 0x3ed0, 0x09ac, 0x08dc, 0x3ea8, 0x0008,
+0x3ffc, 0x3ec4, 0x0978, 0x0914, 0x3eb0, 0x0004,
+0x0000, 0x3eb8, 0x0948, 0x0948, 0x3eb8, 0x0000,
+};
+
+static const uint16_t easf_filter_6tap_64p_ratio_0_90_s1_12[198] = {
+0x3f60, 0x0154, 0x0e9c, 0x0150, 0x3f60, 0x0000,
+0x3f6c, 0x011c, 0x0e9c, 0x018c, 0x3f50, 0x0000,
+0x3f7c, 0x00ec, 0x0e94, 0x01bc, 0x3f44, 0x0004,
+0x3f88, 0x00b8, 0x0e8c, 0x01f8, 0x3f34, 0x0008,
+0x3f94, 0x0088, 0x0e80, 0x0234, 0x3f28, 0x0008,
+0x3fa0, 0x005c, 0x0e74, 0x026c, 0x3f18, 0x000c,
+0x3fac, 0x0030, 0x0e60, 0x02b0, 0x3f08, 0x000c,
+0x3fb8, 0x0004, 0x0e50, 0x02e8, 0x3efc, 0x0010,
+0x3fc4, 0x3fdc, 0x0e38, 0x0328, 0x3eec, 0x0014,
+0x3fd0, 0x3fb4, 0x0e20, 0x0368, 0x3ee0, 0x0014,
+0x3fd8, 0x3f90, 0x0e04, 0x03ac, 0x3ed0, 0x0018,
+0x3fe4, 0x3f6c, 0x0de8, 0x03e8, 0x3ec4, 0x001c,
+0x3fec, 0x3f4c, 0x0dc8, 0x042c, 0x3eb4, 0x0020,
+0x3ff4, 0x3f2c, 0x0da4, 0x0474, 0x3ea8, 0x0020,
+0x0000, 0x3f0c, 0x0d80, 0x04b8, 0x3e98, 0x0024,
+0x0008, 0x3ef0, 0x0d58, 0x04fc, 0x3e8c, 0x0028,
+0x000c, 0x3ed8, 0x0d30, 0x0540, 0x3e80, 0x002c,
+0x0014, 0x3ec0, 0x0d04, 0x0588, 0x3e74, 0x002c,
+0x001c, 0x3ea8, 0x0cd8, 0x05cc, 0x3e68, 0x0030,
+0x0020, 0x3e94, 0x0ca8, 0x0614, 0x3e5c, 0x0034,
+0x0028, 0x3e80, 0x0c78, 0x065c, 0x3e50, 0x0034,
+0x002c, 0x3e6c, 0x0c44, 0x06a4, 0x3e48, 0x0038,
+0x0030, 0x3e5c, 0x0c0c, 0x06f0, 0x3e3c, 0x003c,
+0x0034, 0x3e50, 0x0bd8, 0x0734, 0x3e34, 0x003c,
+0x0038, 0x3e44, 0x0ba0, 0x0778, 0x3e2c, 0x0040,
+0x003c, 0x3e38, 0x0b64, 0x07c4, 0x3e24, 0x0040,
+0x0040, 0x3e2c, 0x0b28, 0x0808, 0x3e20, 0x0044,
+0x0040, 0x3e24, 0x0aec, 0x0850, 0x3e1c, 0x0044,
+0x0044, 0x3e1c, 0x0aac, 0x0898, 0x3e18, 0x0044,
+0x0044, 0x3e18, 0x0a70, 0x08d8, 0x3e14, 0x0048,
+0x0044, 0x3e14, 0x0a2c, 0x0924, 0x3e10, 0x0048,
+0x0048, 0x3e10, 0x09ec, 0x0964, 0x3e10, 0x0048,
+0x0048, 0x3e10, 0x09a8, 0x09a8, 0x3e10, 0x0048,
+};
+
+static const uint16_t easf_filter_6tap_64p_ratio_1_00_s1_12[198] = {
+0x0000, 0x0000, 0x1000, 0x0000, 0x0000, 0x0000,
+0x000c, 0x3fcc, 0x1000, 0x0034, 0x3ff4, 0x0000,
+0x0018, 0x3f9c, 0x0ff8, 0x0070, 0x3fe4, 0x0000,
+0x0024, 0x3f6c, 0x0ff0, 0x00ac, 0x3fd4, 0x0000,
+0x0030, 0x3f40, 0x0fe4, 0x00e8, 0x3fc4, 0x0000,
+0x0038, 0x3f14, 0x0fd4, 0x0128, 0x3fb4, 0x0004,
+0x0044, 0x3eec, 0x0fc0, 0x0168, 0x3fa4, 0x0004,
+0x004c, 0x3ec8, 0x0fac, 0x01a8, 0x3f94, 0x0004,
+0x0054, 0x3ea4, 0x0f90, 0x01ec, 0x3f84, 0x0008,
+0x005c, 0x3e84, 0x0f74, 0x0234, 0x3f70, 0x0008,
+0x0060, 0x3e64, 0x0f50, 0x0280, 0x3f60, 0x000c,
+0x0068, 0x3e48, 0x0f2c, 0x02c8, 0x3f4c, 0x0010,
+0x006c, 0x3e30, 0x0f04, 0x0318, 0x3f38, 0x0010,
+0x0070, 0x3e18, 0x0edc, 0x0364, 0x3f24, 0x0014,
+0x0074, 0x3e00, 0x0eac, 0x03b8, 0x3f10, 0x0018,
+0x0078, 0x3df0, 0x0e7c, 0x0404, 0x3efc, 0x001c,
+0x007c, 0x3de0, 0x0e48, 0x0454, 0x3ee8, 0x0020,
+0x007c, 0x3dd0, 0x0e14, 0x04ac, 0x3ed4, 0x0020,
+0x0080, 0x3dc4, 0x0dd8, 0x0500, 0x3ec0, 0x0024,
+0x0080, 0x3db8, 0x0d9c, 0x0554, 0x3eac, 0x002c,
+0x0080, 0x3db0, 0x0d5c, 0x05ac, 0x3e98, 0x0030,
+0x0080, 0x3da8, 0x0d1c, 0x0600, 0x3e88, 0x0034,
+0x0080, 0x3da4, 0x0cd8, 0x0658, 0x3e74, 0x0038,
+0x0080, 0x3da4, 0x0c94, 0x06ac, 0x3e60, 0x003c,
+0x007c, 0x3da0, 0x0c4c, 0x070c, 0x3e4c, 0x0040,
+0x007c, 0x3da4, 0x0c00, 0x0760, 0x3e3c, 0x0044,
+0x0078, 0x3da4, 0x0bb4, 0x07bc, 0x3e2c, 0x0048,
+0x0074, 0x3da8, 0x0b64, 0x0814, 0x3e1c, 0x0050,
+0x0074, 0x3db0, 0x0b14, 0x0868, 0x3e0c, 0x0054,
+0x0070, 0x3db8, 0x0ac4, 0x08c0, 0x3dfc, 0x0058,
+0x006c, 0x3dc0, 0x0a70, 0x091c, 0x3dec, 0x005c,
+0x0068, 0x3dc8, 0x0a1c, 0x0974, 0x3de0, 0x0060,
+0x0064, 0x3dd4, 0x09c8, 0x09c8, 0x3dd4, 0x0064,
+};
+
+static struct scale_ratio_to_reg_value_lookup easf_v_bf3_mode_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x0000},
{5, 10, 0x0000},
@@ -1173,7 +2012,7 @@ struct scale_ratio_to_reg_value_lookup easf_v_bf3_mode_lookup[] = {
{-1, -1, 0x0002},
};
-struct scale_ratio_to_reg_value_lookup easf_h_bf3_mode_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_h_bf3_mode_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x0000},
{5, 10, 0x0000},
@@ -1185,7 +2024,7 @@ struct scale_ratio_to_reg_value_lookup easf_h_bf3_mode_lookup[] = {
{-1, -1, 0x0002},
};
-struct scale_ratio_to_reg_value_lookup easf_reducer_gain6_6tap_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_reducer_gain6_6tap_lookup[] = {
{3, 10, 0x4100},
{4, 10, 0x4100},
{5, 10, 0x4100},
@@ -1197,7 +2036,7 @@ struct scale_ratio_to_reg_value_lookup easf_reducer_gain6_6tap_lookup[] = {
{-1, -1, 0x4100},
};
-struct scale_ratio_to_reg_value_lookup easf_reducer_gain4_6tap_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_reducer_gain4_6tap_lookup[] = {
{3, 10, 0x4000},
{4, 10, 0x4000},
{5, 10, 0x4000},
@@ -1209,7 +2048,7 @@ struct scale_ratio_to_reg_value_lookup easf_reducer_gain4_6tap_lookup[] = {
{-1, -1, 0x4000},
};
-struct scale_ratio_to_reg_value_lookup easf_gain_ring6_6tap_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_gain_ring6_6tap_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x251F},
{5, 10, 0x291F},
@@ -1221,7 +2060,7 @@ struct scale_ratio_to_reg_value_lookup easf_gain_ring6_6tap_lookup[] = {
{-1, -1, 0xA640},
};
-struct scale_ratio_to_reg_value_lookup easf_gain_ring4_6tap_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_gain_ring4_6tap_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x9600},
{5, 10, 0xA460},
@@ -1233,7 +2072,7 @@ struct scale_ratio_to_reg_value_lookup easf_gain_ring4_6tap_lookup[] = {
{-1, -1, 0xB058},
};
-struct scale_ratio_to_reg_value_lookup easf_reducer_gain6_4tap_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_reducer_gain6_4tap_lookup[] = {
{3, 10, 0x4100},
{4, 10, 0x4100},
{5, 10, 0x4100},
@@ -1245,7 +2084,7 @@ struct scale_ratio_to_reg_value_lookup easf_reducer_gain6_4tap_lookup[] = {
{-1, -1, 0x4100},
};
-struct scale_ratio_to_reg_value_lookup easf_reducer_gain4_4tap_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_reducer_gain4_4tap_lookup[] = {
{3, 10, 0x4000},
{4, 10, 0x4000},
{5, 10, 0x4000},
@@ -1257,7 +2096,7 @@ struct scale_ratio_to_reg_value_lookup easf_reducer_gain4_4tap_lookup[] = {
{-1, -1, 0x4000},
};
-struct scale_ratio_to_reg_value_lookup easf_gain_ring6_4tap_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_gain_ring6_4tap_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x0000},
{5, 10, 0x0000},
@@ -1269,7 +2108,7 @@ struct scale_ratio_to_reg_value_lookup easf_gain_ring6_4tap_lookup[] = {
{-1, -1, 0x0000},
};
-struct scale_ratio_to_reg_value_lookup easf_gain_ring4_4tap_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_gain_ring4_4tap_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x0000},
{5, 10, 0x0000},
@@ -1281,7 +2120,7 @@ struct scale_ratio_to_reg_value_lookup easf_gain_ring4_4tap_lookup[] = {
{-1, -1, 0xAC00},
};
-struct scale_ratio_to_reg_value_lookup easf_3tap_dntilt_uptilt_offset_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_3tap_dntilt_uptilt_offset_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x0000},
{5, 10, 0x0000},
@@ -1293,7 +2132,7 @@ struct scale_ratio_to_reg_value_lookup easf_3tap_dntilt_uptilt_offset_lookup[] =
{-1, -1, 0xA8D8},
};
-struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt_maxval_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt_maxval_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x0000},
{5, 10, 0x0000},
@@ -1305,7 +2144,7 @@ struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt_maxval_lookup[] = {
{-1, -1, 0x3ADB},
};
-struct scale_ratio_to_reg_value_lookup easf_3tap_dntilt_slope_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_3tap_dntilt_slope_lookup[] = {
{3, 10, 0x3800},
{4, 10, 0x3800},
{5, 10, 0x3800},
@@ -1317,7 +2156,7 @@ struct scale_ratio_to_reg_value_lookup easf_3tap_dntilt_slope_lookup[] = {
{-1, -1, 0x3B66},
};
-struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt1_slope_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt1_slope_lookup[] = {
{3, 10, 0x3800},
{4, 10, 0x3800},
{5, 10, 0x3800},
@@ -1329,7 +2168,7 @@ struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt1_slope_lookup[] = {
{-1, -1, 0x2F20},
};
-struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt2_slope_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt2_slope_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x0000},
{5, 10, 0x0000},
@@ -1341,7 +2180,7 @@ struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt2_slope_lookup[] = {
{-1, -1, 0x1F00},
};
-struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt2_offset_lookup[] = {
+static struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt2_offset_lookup[] = {
{3, 10, 0x0000},
{4, 10, 0x0000},
{5, 10, 0x0000},
@@ -1353,61 +2192,7 @@ struct scale_ratio_to_reg_value_lookup easf_3tap_uptilt2_offset_lookup[] = {
{-1, -1, 0x9E00},
};
-void spl_init_easf_filter_coeffs(void)
-{
- convert_filter_s1_10_to_s1_12(easf_filter_3tap_64p_ratio_0_30,
- easf_filter_3tap_64p_ratio_0_30_s1_12, 3);
- convert_filter_s1_10_to_s1_12(easf_filter_3tap_64p_ratio_0_40,
- easf_filter_3tap_64p_ratio_0_40_s1_12, 3);
- convert_filter_s1_10_to_s1_12(easf_filter_3tap_64p_ratio_0_50,
- easf_filter_3tap_64p_ratio_0_50_s1_12, 3);
- convert_filter_s1_10_to_s1_12(easf_filter_3tap_64p_ratio_0_60,
- easf_filter_3tap_64p_ratio_0_60_s1_12, 3);
- convert_filter_s1_10_to_s1_12(easf_filter_3tap_64p_ratio_0_70,
- easf_filter_3tap_64p_ratio_0_70_s1_12, 3);
- convert_filter_s1_10_to_s1_12(easf_filter_3tap_64p_ratio_0_80,
- easf_filter_3tap_64p_ratio_0_80_s1_12, 3);
- convert_filter_s1_10_to_s1_12(easf_filter_3tap_64p_ratio_0_90,
- easf_filter_3tap_64p_ratio_0_90_s1_12, 3);
- convert_filter_s1_10_to_s1_12(easf_filter_3tap_64p_ratio_1_00,
- easf_filter_3tap_64p_ratio_1_00_s1_12, 3);
-
- convert_filter_s1_10_to_s1_12(easf_filter_4tap_64p_ratio_0_30,
- easf_filter_4tap_64p_ratio_0_30_s1_12, 4);
- convert_filter_s1_10_to_s1_12(easf_filter_4tap_64p_ratio_0_40,
- easf_filter_4tap_64p_ratio_0_40_s1_12, 4);
- convert_filter_s1_10_to_s1_12(easf_filter_4tap_64p_ratio_0_50,
- easf_filter_4tap_64p_ratio_0_50_s1_12, 4);
- convert_filter_s1_10_to_s1_12(easf_filter_4tap_64p_ratio_0_60,
- easf_filter_4tap_64p_ratio_0_60_s1_12, 4);
- convert_filter_s1_10_to_s1_12(easf_filter_4tap_64p_ratio_0_70,
- easf_filter_4tap_64p_ratio_0_70_s1_12, 4);
- convert_filter_s1_10_to_s1_12(easf_filter_4tap_64p_ratio_0_80,
- easf_filter_4tap_64p_ratio_0_80_s1_12, 4);
- convert_filter_s1_10_to_s1_12(easf_filter_4tap_64p_ratio_0_90,
- easf_filter_4tap_64p_ratio_0_90_s1_12, 4);
- convert_filter_s1_10_to_s1_12(easf_filter_4tap_64p_ratio_1_00,
- easf_filter_4tap_64p_ratio_1_00_s1_12, 4);
-
- convert_filter_s1_10_to_s1_12(easf_filter_6tap_64p_ratio_0_30,
- easf_filter_6tap_64p_ratio_0_30_s1_12, 6);
- convert_filter_s1_10_to_s1_12(easf_filter_6tap_64p_ratio_0_40,
- easf_filter_6tap_64p_ratio_0_40_s1_12, 6);
- convert_filter_s1_10_to_s1_12(easf_filter_6tap_64p_ratio_0_50,
- easf_filter_6tap_64p_ratio_0_50_s1_12, 6);
- convert_filter_s1_10_to_s1_12(easf_filter_6tap_64p_ratio_0_60,
- easf_filter_6tap_64p_ratio_0_60_s1_12, 6);
- convert_filter_s1_10_to_s1_12(easf_filter_6tap_64p_ratio_0_70,
- easf_filter_6tap_64p_ratio_0_70_s1_12, 6);
- convert_filter_s1_10_to_s1_12(easf_filter_6tap_64p_ratio_0_80,
- easf_filter_6tap_64p_ratio_0_80_s1_12, 6);
- convert_filter_s1_10_to_s1_12(easf_filter_6tap_64p_ratio_0_90,
- easf_filter_6tap_64p_ratio_0_90_s1_12, 6);
- convert_filter_s1_10_to_s1_12(easf_filter_6tap_64p_ratio_1_00,
- easf_filter_6tap_64p_ratio_1_00_s1_12, 6);
-}
-
-uint16_t *spl_get_easf_filter_3tap_64p(struct spl_fixed31_32 ratio)
+static const uint16_t *spl_get_easf_filter_3tap_64p(struct spl_fixed31_32 ratio)
{
if (ratio.value < spl_fixpt_from_fraction(3, 10).value)
return easf_filter_3tap_64p_ratio_0_30_s1_12;
@@ -1427,7 +2212,7 @@ uint16_t *spl_get_easf_filter_3tap_64p(struct spl_fixed31_32 ratio)
return easf_filter_3tap_64p_ratio_1_00_s1_12;
}
-uint16_t *spl_get_easf_filter_4tap_64p(struct spl_fixed31_32 ratio)
+static const uint16_t *spl_get_easf_filter_4tap_64p(struct spl_fixed31_32 ratio)
{
if (ratio.value < spl_fixpt_from_fraction(3, 10).value)
return easf_filter_4tap_64p_ratio_0_30_s1_12;
@@ -1447,7 +2232,7 @@ uint16_t *spl_get_easf_filter_4tap_64p(struct spl_fixed31_32 ratio)
return easf_filter_4tap_64p_ratio_1_00_s1_12;
}
-uint16_t *spl_get_easf_filter_6tap_64p(struct spl_fixed31_32 ratio)
+static const uint16_t *spl_get_easf_filter_6tap_64p(struct spl_fixed31_32 ratio)
{
if (ratio.value < spl_fixpt_from_fraction(3, 10).value)
return easf_filter_6tap_64p_ratio_0_30_s1_12;
@@ -1467,7 +2252,7 @@ uint16_t *spl_get_easf_filter_6tap_64p(struct spl_fixed31_32 ratio)
return easf_filter_6tap_64p_ratio_1_00_s1_12;
}
-uint16_t *spl_dscl_get_easf_filter_coeffs_64p(int taps, struct spl_fixed31_32 ratio)
+const uint16_t *spl_dscl_get_easf_filter_coeffs_64p(int taps, struct spl_fixed31_32 ratio)
{
if (taps == 6)
return spl_get_easf_filter_6tap_64p(ratio);
@@ -1482,6 +2267,81 @@ uint16_t *spl_dscl_get_easf_filter_coeffs_64p(int taps, struct spl_fixed31_32 ra
}
}
+static const uint16_t *spl_get_easf_filter_3tap_64p_s1_10(struct spl_fixed31_32 ratio)
+{
+ if (ratio.value < spl_fixpt_from_fraction(3, 10).value)
+ return easf_filter_3tap_64p_ratio_0_30;
+ else if (ratio.value < spl_fixpt_from_fraction(4, 10).value)
+ return easf_filter_3tap_64p_ratio_0_40;
+ else if (ratio.value < spl_fixpt_from_fraction(5, 10).value)
+ return easf_filter_3tap_64p_ratio_0_50;
+ else if (ratio.value < spl_fixpt_from_fraction(6, 10).value)
+ return easf_filter_3tap_64p_ratio_0_60;
+ else if (ratio.value < spl_fixpt_from_fraction(7, 10).value)
+ return easf_filter_3tap_64p_ratio_0_70;
+ else if (ratio.value < spl_fixpt_from_fraction(8, 10).value)
+ return easf_filter_3tap_64p_ratio_0_80;
+ else if (ratio.value < spl_fixpt_from_fraction(9, 10).value)
+ return easf_filter_3tap_64p_ratio_0_90;
+ else
+ return easf_filter_3tap_64p_ratio_1_00;
+}
+
+static const uint16_t *spl_get_easf_filter_4tap_64p_s1_10(struct spl_fixed31_32 ratio)
+{
+ if (ratio.value < spl_fixpt_from_fraction(3, 10).value)
+ return easf_filter_4tap_64p_ratio_0_30;
+ else if (ratio.value < spl_fixpt_from_fraction(4, 10).value)
+ return easf_filter_4tap_64p_ratio_0_40;
+ else if (ratio.value < spl_fixpt_from_fraction(5, 10).value)
+ return easf_filter_4tap_64p_ratio_0_50;
+ else if (ratio.value < spl_fixpt_from_fraction(6, 10).value)
+ return easf_filter_4tap_64p_ratio_0_60;
+ else if (ratio.value < spl_fixpt_from_fraction(7, 10).value)
+ return easf_filter_4tap_64p_ratio_0_70;
+ else if (ratio.value < spl_fixpt_from_fraction(8, 10).value)
+ return easf_filter_4tap_64p_ratio_0_80;
+ else if (ratio.value < spl_fixpt_from_fraction(9, 10).value)
+ return easf_filter_4tap_64p_ratio_0_90;
+ else
+ return easf_filter_4tap_64p_ratio_1_00;
+}
+
+static const uint16_t *spl_get_easf_filter_6tap_64p_s1_10(struct spl_fixed31_32 ratio)
+{
+ if (ratio.value < spl_fixpt_from_fraction(3, 10).value)
+ return easf_filter_6tap_64p_ratio_0_30;
+ else if (ratio.value < spl_fixpt_from_fraction(4, 10).value)
+ return easf_filter_6tap_64p_ratio_0_40;
+ else if (ratio.value < spl_fixpt_from_fraction(5, 10).value)
+ return easf_filter_6tap_64p_ratio_0_50;
+ else if (ratio.value < spl_fixpt_from_fraction(6, 10).value)
+ return easf_filter_6tap_64p_ratio_0_60;
+ else if (ratio.value < spl_fixpt_from_fraction(7, 10).value)
+ return easf_filter_6tap_64p_ratio_0_70;
+ else if (ratio.value < spl_fixpt_from_fraction(8, 10).value)
+ return easf_filter_6tap_64p_ratio_0_80;
+ else if (ratio.value < spl_fixpt_from_fraction(9, 10).value)
+ return easf_filter_6tap_64p_ratio_0_90;
+ else
+ return easf_filter_6tap_64p_ratio_1_00;
+}
+
+const uint16_t *spl_dscl_get_easf_filter_coeffs_64p_s1_10(int taps, struct spl_fixed31_32 ratio)
+{
+ if (taps == 6)
+ return spl_get_easf_filter_6tap_64p_s1_10(ratio);
+ else if (taps == 4)
+ return spl_get_easf_filter_4tap_64p_s1_10(ratio);
+ else if (taps == 3)
+ return spl_get_easf_filter_3tap_64p_s1_10(ratio);
+ else {
+ /* should never happen, bug */
+ SPL_BREAK_TO_DEBUGGER();
+ return NULL;
+ }
+}
+
void spl_set_filters_data(struct dscl_prog_data *dscl_prog_data,
const struct spl_scaler_data *data, bool enable_easf_v,
bool enable_easf_h)
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_easf_filters.h b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_easf_filters.h
index 8bb2b8108e38..321ae22a04d4 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_easf_filters.h
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_easf_filters.h
@@ -13,11 +13,6 @@ struct scale_ratio_to_reg_value_lookup {
const uint32_t reg_value;
};
-void spl_init_easf_filter_coeffs(void);
-uint16_t *spl_get_easf_filter_3tap_64p(struct spl_fixed31_32 ratio);
-uint16_t *spl_get_easf_filter_4tap_64p(struct spl_fixed31_32 ratio);
-uint16_t *spl_get_easf_filter_6tap_64p(struct spl_fixed31_32 ratio);
-uint16_t *spl_dscl_get_easf_filter_coeffs_64p(int taps, struct spl_fixed31_32 ratio);
void spl_set_filters_data(struct dscl_prog_data *dscl_prog_data,
const struct spl_scaler_data *data, bool enable_easf_v,
bool enable_easf_h);
@@ -35,4 +30,8 @@ uint32_t spl_get_3tap_uptilt1_slope(int taps, struct spl_fixed31_32 ratio);
uint32_t spl_get_3tap_uptilt2_slope(int taps, struct spl_fixed31_32 ratio);
uint32_t spl_get_3tap_uptilt2_offset(int taps, struct spl_fixed31_32 ratio);
+/* public API */
+const uint16_t *spl_dscl_get_easf_filter_coeffs_64p(int taps, struct spl_fixed31_32 ratio);
+const uint16_t *spl_dscl_get_easf_filter_coeffs_64p_s1_10(int taps, struct spl_fixed31_32 ratio);
+
#endif /* __DC_SPL_SCL_EASF_FILTERS_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_filters.c b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_filters.c
index b02c7b0b262b..5e52bdf1ad44 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_scl_filters.c
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_filters.c
@@ -4,194 +4,6 @@
#include "spl_debug.h"
#include "dc_spl_scl_filters.h"
-//=========================================
-// <num_taps> = 2
-// <num_phases> = 16
-// <scale_ratio> = 0.833333 (input/output)
-// <sharpness> = 0
-// <CoefType> = ModifiedLanczos
-// <CoefQuant> = s1.10
-// <CoefOut> = s1.12
-//=========================================
-static const uint16_t filter_2tap_16p[18] = {
- 0x1000, 0x0000,
- 0x0FF0, 0x0010,
- 0x0FB0, 0x0050,
- 0x0F34, 0x00CC,
- 0x0E68, 0x0198,
- 0x0D44, 0x02BC,
- 0x0BC4, 0x043C,
- 0x09FC, 0x0604,
- 0x0800, 0x0800
-};
-
-//=========================================
-// <num_taps> = 3
-// <num_phases> = 16
-// <scale_ratio> = 0.83333 (input/output)
-// <sharpness> = 0
-// <CoefType> = ModifiedLanczos
-// <CoefQuant> = 1.10
-// <CoefOut> = 1.12
-//=========================================
-static const uint16_t filter_3tap_16p_upscale[27] = {
- 0x0804, 0x07FC, 0x0000,
- 0x06AC, 0x0978, 0x3FDC,
- 0x055C, 0x0AF0, 0x3FB4,
- 0x0420, 0x0C50, 0x3F90,
- 0x0300, 0x0D88, 0x3F78,
- 0x0200, 0x0E90, 0x3F70,
- 0x0128, 0x0F5C, 0x3F7C,
- 0x007C, 0x0FD8, 0x3FAC,
- 0x0000, 0x1000, 0x0000
-};
-
-//=========================================
-// <num_taps> = 3
-// <num_phases> = 16
-// <scale_ratio> = 1.16666 (input/output)
-// <sharpness> = 0
-// <CoefType> = ModifiedLanczos
-// <CoefQuant> = 1.10
-// <CoefOut> = 1.12
-//=========================================
-static const uint16_t filter_3tap_16p_116[27] = {
- 0x0804, 0x07FC, 0x0000,
- 0x0700, 0x0914, 0x3FEC,
- 0x0604, 0x0A1C, 0x3FE0,
- 0x050C, 0x0B14, 0x3FE0,
- 0x041C, 0x0BF4, 0x3FF0,
- 0x0340, 0x0CB0, 0x0010,
- 0x0274, 0x0D3C, 0x0050,
- 0x01C0, 0x0D94, 0x00AC,
- 0x0128, 0x0DB4, 0x0124
-};
-
-//=========================================
-// <num_taps> = 3
-// <num_phases> = 16
-// <scale_ratio> = 1.49999 (input/output)
-// <sharpness> = 0
-// <CoefType> = ModifiedLanczos
-// <CoefQuant> = 1.10
-// <CoefOut> = 1.12
-//=========================================
-static const uint16_t filter_3tap_16p_149[27] = {
- 0x0804, 0x07FC, 0x0000,
- 0x0730, 0x08CC, 0x0004,
- 0x0660, 0x098C, 0x0014,
- 0x0590, 0x0A3C, 0x0034,
- 0x04C4, 0x0AD4, 0x0068,
- 0x0400, 0x0B54, 0x00AC,
- 0x0348, 0x0BB0, 0x0108,
- 0x029C, 0x0BEC, 0x0178,
- 0x0200, 0x0C00, 0x0200
-};
-
-//=========================================
-// <num_taps> = 3
-// <num_phases> = 16
-// <scale_ratio> = 1.83332 (input/output)
-// <sharpness> = 0
-// <CoefType> = ModifiedLanczos
-// <CoefQuant> = 1.10
-// <CoefOut> = 1.12
-//=========================================
-static const uint16_t filter_3tap_16p_183[27] = {
- 0x0804, 0x07FC, 0x0000,
- 0x0754, 0x0880, 0x002C,
- 0x06A8, 0x08F0, 0x0068,
- 0x05FC, 0x0954, 0x00B0,
- 0x0550, 0x09AC, 0x0104,
- 0x04A8, 0x09F0, 0x0168,
- 0x0408, 0x0A20, 0x01D8,
- 0x036C, 0x0A40, 0x0254,
- 0x02DC, 0x0A48, 0x02DC
-};
-
-//=========================================
-// <num_taps> = 4
-// <num_phases> = 16
-// <scale_ratio> = 0.83333 (input/output)
-// <sharpness> = 0
-// <CoefType> = ModifiedLanczos
-// <CoefQuant> = 1.10
-// <CoefOut> = 1.12
-//=========================================
-static const uint16_t filter_4tap_16p_upscale[36] = {
- 0x0000, 0x1000, 0x0000, 0x0000,
- 0x3F74, 0x0FDC, 0x00B4, 0x3FFC,
- 0x3F0C, 0x0F70, 0x0194, 0x3FF0,
- 0x3ECC, 0x0EC4, 0x0298, 0x3FD8,
- 0x3EAC, 0x0DE4, 0x03B8, 0x3FB8,
- 0x3EA4, 0x0CD8, 0x04F4, 0x3F90,
- 0x3EB8, 0x0BA0, 0x0644, 0x3F64,
- 0x3ED8, 0x0A54, 0x07A0, 0x3F34,
- 0x3F00, 0x08FC, 0x0900, 0x3F04
-};
-
-//=========================================
-// <num_taps> = 4
-// <num_phases> = 16
-// <scale_ratio> = 1.16666 (input/output)
-// <sharpness> = 0
-// <CoefType> = ModifiedLanczos
-// <CoefQuant> = 1.10
-// <CoefOut> = 1.12
-//=========================================
-static const uint16_t filter_4tap_16p_116[36] = {
- 0x01A8, 0x0CB4, 0x01A4, 0x0000,
- 0x0110, 0x0CB0, 0x0254, 0x3FEC,
- 0x0090, 0x0C80, 0x031C, 0x3FD4,
- 0x0024, 0x0C2C, 0x03F4, 0x3FBC,
- 0x3FD8, 0x0BAC, 0x04DC, 0x3FA0,
- 0x3F9C, 0x0B14, 0x05CC, 0x3F84,
- 0x3F70, 0x0A60, 0x06C4, 0x3F6C,
- 0x3F5C, 0x098C, 0x07BC, 0x3F5C,
- 0x3F54, 0x08AC, 0x08AC, 0x3F54
-};
-
-//=========================================
-// <num_taps> = 4
-// <num_phases> = 16
-// <scale_ratio> = 1.49999 (input/output)
-// <sharpness> = 0
-// <CoefType> = ModifiedLanczos
-// <CoefQuant> = 1.10
-// <CoefOut> = 1.12
-//=========================================
-static const uint16_t filter_4tap_16p_149[36] = {
- 0x02B8, 0x0A90, 0x02B8, 0x0000,
- 0x0230, 0x0A90, 0x0350, 0x3FF0,
- 0x01B8, 0x0A78, 0x03F0, 0x3FE0,
- 0x0148, 0x0A48, 0x049C, 0x3FD4,
- 0x00E8, 0x0A00, 0x054C, 0x3FCC,
- 0x0098, 0x09A0, 0x0600, 0x3FC8,
- 0x0054, 0x0928, 0x06B4, 0x3FD0,
- 0x001C, 0x08A4, 0x0760, 0x3FE0,
- 0x3FFC, 0x0804, 0x0804, 0x3FFC
-};
-
-//=========================================
-// <num_taps> = 4
-// <num_phases> = 16
-// <scale_ratio> = 1.83332 (input/output)
-// <sharpness> = 0
-// <CoefType> = ModifiedLanczos
-// <CoefQuant> = 1.10
-// <CoefOut> = 1.12
-//=========================================
-static const uint16_t filter_4tap_16p_183[36] = {
- 0x03B0, 0x08A0, 0x03B0, 0x0000,
- 0x0348, 0x0898, 0x041C, 0x0004,
- 0x02DC, 0x0884, 0x0490, 0x0010,
- 0x0278, 0x0864, 0x0500, 0x0024,
- 0x021C, 0x0838, 0x0570, 0x003C,
- 0x01C8, 0x07FC, 0x05E0, 0x005C,
- 0x0178, 0x07B8, 0x064C, 0x0084,
- 0x0130, 0x076C, 0x06B0, 0x00B4,
- 0x00F0, 0x0714, 0x0710, 0x00EC
-};
//=========================================
// <num_taps> = 2
@@ -1318,19 +1130,7 @@ static const uint16_t filter_8tap_64p_183[264] = {
0x3FD4, 0x3F84, 0x0214, 0x0694, 0x0694, 0x0214, 0x3F84, 0x3FD4
};
-const uint16_t *spl_get_filter_3tap_16p(struct spl_fixed31_32 ratio)
-{
- if (ratio.value < spl_fixpt_one.value)
- return filter_3tap_16p_upscale;
- else if (ratio.value < spl_fixpt_from_fraction(4, 3).value)
- return filter_3tap_16p_116;
- else if (ratio.value < spl_fixpt_from_fraction(5, 3).value)
- return filter_3tap_16p_149;
- else
- return filter_3tap_16p_183;
-}
-
-const uint16_t *spl_get_filter_3tap_64p(struct spl_fixed31_32 ratio)
+static const uint16_t *spl_get_filter_3tap_64p(struct spl_fixed31_32 ratio)
{
if (ratio.value < spl_fixpt_one.value)
return filter_3tap_64p_upscale;
@@ -1342,19 +1142,7 @@ const uint16_t *spl_get_filter_3tap_64p(struct spl_fixed31_32 ratio)
return filter_3tap_64p_183;
}
-const uint16_t *spl_get_filter_4tap_16p(struct spl_fixed31_32 ratio)
-{
- if (ratio.value < spl_fixpt_one.value)
- return filter_4tap_16p_upscale;
- else if (ratio.value < spl_fixpt_from_fraction(4, 3).value)
- return filter_4tap_16p_116;
- else if (ratio.value < spl_fixpt_from_fraction(5, 3).value)
- return filter_4tap_16p_149;
- else
- return filter_4tap_16p_183;
-}
-
-const uint16_t *spl_get_filter_4tap_64p(struct spl_fixed31_32 ratio)
+static const uint16_t *spl_get_filter_4tap_64p(struct spl_fixed31_32 ratio)
{
if (ratio.value < spl_fixpt_one.value)
return filter_4tap_64p_upscale;
@@ -1366,7 +1154,7 @@ const uint16_t *spl_get_filter_4tap_64p(struct spl_fixed31_32 ratio)
return filter_4tap_64p_183;
}
-const uint16_t *spl_get_filter_5tap_64p(struct spl_fixed31_32 ratio)
+static const uint16_t *spl_get_filter_5tap_64p(struct spl_fixed31_32 ratio)
{
if (ratio.value < spl_fixpt_one.value)
return filter_5tap_64p_upscale;
@@ -1378,7 +1166,7 @@ const uint16_t *spl_get_filter_5tap_64p(struct spl_fixed31_32 ratio)
return filter_5tap_64p_183;
}
-const uint16_t *spl_get_filter_6tap_64p(struct spl_fixed31_32 ratio)
+static const uint16_t *spl_get_filter_6tap_64p(struct spl_fixed31_32 ratio)
{
if (ratio.value < spl_fixpt_one.value)
return filter_6tap_64p_upscale;
@@ -1390,7 +1178,7 @@ const uint16_t *spl_get_filter_6tap_64p(struct spl_fixed31_32 ratio)
return filter_6tap_64p_183;
}
-const uint16_t *spl_get_filter_7tap_64p(struct spl_fixed31_32 ratio)
+static const uint16_t *spl_get_filter_7tap_64p(struct spl_fixed31_32 ratio)
{
if (ratio.value < spl_fixpt_one.value)
return filter_7tap_64p_upscale;
@@ -1402,7 +1190,7 @@ const uint16_t *spl_get_filter_7tap_64p(struct spl_fixed31_32 ratio)
return filter_7tap_64p_183;
}
-const uint16_t *spl_get_filter_8tap_64p(struct spl_fixed31_32 ratio)
+static const uint16_t *spl_get_filter_8tap_64p(struct spl_fixed31_32 ratio)
{
if (ratio.value < spl_fixpt_one.value)
return filter_8tap_64p_upscale;
@@ -1414,12 +1202,7 @@ const uint16_t *spl_get_filter_8tap_64p(struct spl_fixed31_32 ratio)
return filter_8tap_64p_183;
}
-const uint16_t *spl_get_filter_2tap_16p(void)
-{
- return filter_2tap_16p;
-}
-
-const uint16_t *spl_get_filter_2tap_64p(void)
+static const uint16_t *spl_get_filter_2tap_64p(void)
{
return filter_2tap_64p;
}
@@ -1448,4 +1231,3 @@ const uint16_t *spl_dscl_get_filter_coeffs_64p(int taps, struct spl_fixed31_32 r
return NULL;
}
}
-
diff --git a/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_filters.h b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_filters.h
new file mode 100644
index 000000000000..c315a438d064
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_scl_filters.h
@@ -0,0 +1,13 @@
+// SPDX-License-Identifier: MIT
+//
+// Copyright 2024 Advanced Micro Devices, Inc.
+
+#ifndef __DC_SPL_SCL_FILTERS_H__
+#define __DC_SPL_SCL_FILTERS_H__
+
+#include "dc_spl_types.h"
+
+/* public API */
+const uint16_t *spl_dscl_get_filter_coeffs_64p(int taps, struct spl_fixed31_32 ratio);
+
+#endif /* __DC_SPL_SCL_FILTERS_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_types.h
index 55d557df4aa5..23d254dea18f 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/dc_spl_types.h
+++ b/drivers/gpu/drm/amd/display/dc/sspl/dc_spl_types.h
@@ -63,13 +63,13 @@ enum spl_pixel_format {
SPL_PIXEL_FORMAT_420BPP8,
SPL_PIXEL_FORMAT_420BPP10,
/*end of pixel format definition*/
- SPL_PIXEL_FORMAT_INVALID,
- SPL_PIXEL_FORMAT_422BPP8,
- SPL_PIXEL_FORMAT_422BPP10,
SPL_PIXEL_FORMAT_GRPH_BEGIN = SPL_PIXEL_FORMAT_INDEX8,
SPL_PIXEL_FORMAT_GRPH_END = SPL_PIXEL_FORMAT_FP16,
+ SPL_PIXEL_FORMAT_SUBSAMPLED_BEGIN = SPL_PIXEL_FORMAT_420BPP8,
+ SPL_PIXEL_FORMAT_SUBSAMPLED_END = SPL_PIXEL_FORMAT_420BPP10,
SPL_PIXEL_FORMAT_VIDEO_BEGIN = SPL_PIXEL_FORMAT_420BPP8,
SPL_PIXEL_FORMAT_VIDEO_END = SPL_PIXEL_FORMAT_420BPP10,
+ SPL_PIXEL_FORMAT_INVALID,
SPL_PIXEL_FORMAT_UNKNOWN
};
@@ -427,6 +427,14 @@ struct spl_out {
// SPL inputs
+// opp extra adjustment for rect
+struct spl_opp_adjust {
+ int x;
+ int y;
+ int width;
+ int height;
+};
+
// Basic input information
struct basic_in {
enum spl_pixel_format format; // Pixel Format
@@ -436,8 +444,15 @@ struct basic_in {
struct spl_rect clip_rect; // Clip rect
enum spl_rotation_angle rotation; // Rotation
bool horizontal_mirror; // Horizontal mirror
- int mpc_combine_h; // MPC Horizontal Combine Factor (split_count)
- int mpc_combine_v; // MPC Vertical Combine Factor (split_idx)
+ struct { // previous mpc_combine_h - split count
+ bool use_recout_width_aligned;
+ union {
+ int mpc_num_h_slices;
+ int mpc_recout_width_align;
+ } num_slices_recout_width;
+ } num_h_slices_recout_width_align;
+ int mpc_h_slice_index; // previous mpc_combine_v - split_idx
+ struct spl_opp_adjust opp_recout_adjust;
// Inputs for adaptive scaler - TODO
enum spl_transfer_func_type tf_type; /* Transfer function type */
enum spl_transfer_func_predefined tf_predefined_type; /* Transfer function predefined type */
@@ -445,6 +460,8 @@ struct basic_in {
enum spl_color_space color_space; // Color Space
unsigned int max_luminance; // Max Luminance TODO: Is determined in dc_hw_sequencer.c is_sdr
bool film_grain_applied; // Film Grain Applied // TODO: To check from where to get this?
+ int custom_width; // Width for non-standard segmentation - used when != 0
+ int custom_x; // Start x for non-standard segmentation - used when custom_width != 0
};
// Basic output information
@@ -465,6 +482,10 @@ enum sharpness_setting {
SHARPNESS_ZERO,
SHARPNESS_CUSTOM
};
+enum sharpness_range_source {
+ SHARPNESS_RANGE_DCN = 0,
+ SHARPNESS_RANGE_DCN_OVERRIDE
+};
struct spl_sharpness_range {
int sdr_rgb_min;
int sdr_rgb_max;
@@ -478,7 +499,7 @@ struct spl_sharpness_range {
};
struct adaptive_sharpness {
bool enable;
- int sharpness_level;
+ unsigned int sharpness_level;
struct spl_sharpness_range sharpness_range;
};
enum linear_light_scaling { // convert it in translation logic
@@ -529,6 +550,7 @@ struct spl_in {
bool is_hdr_on;
int h_active;
int v_active;
+ int min_viewport_size;
int sdr_white_level_nits;
enum sharpen_policy sharpen_policy;
};
diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_custom_float.c b/drivers/gpu/drm/amd/display/dc/sspl/spl_custom_float.c
index be2f34d034c5..be2f34d034c5 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/spl_custom_float.c
+++ b/drivers/gpu/drm/amd/display/dc/sspl/spl_custom_float.c
diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_custom_float.h b/drivers/gpu/drm/amd/display/dc/sspl/spl_custom_float.h
index cdc4e107b9de..cdc4e107b9de 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/spl_custom_float.h
+++ b/drivers/gpu/drm/amd/display/dc/sspl/spl_custom_float.h
diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_debug.h b/drivers/gpu/drm/amd/display/dc/sspl/spl_debug.h
index a6f6132df241..a6f6132df241 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/spl_debug.h
+++ b/drivers/gpu/drm/amd/display/dc/sspl/spl_debug.h
diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.c b/drivers/gpu/drm/amd/display/dc/sspl/spl_fixpt31_32.c
index 131f1e3949d3..ebf0287417e0 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.c
+++ b/drivers/gpu/drm/amd/display/dc/sspl/spl_fixpt31_32.c
@@ -29,8 +29,6 @@ static inline unsigned long long spl_complete_integer_division_u64(
{
unsigned long long result;
- SPL_ASSERT(divisor);
-
result = spl_div64_u64_rem(dividend, divisor, remainder);
return result;
@@ -196,8 +194,6 @@ struct spl_fixed31_32 spl_fixpt_recip(struct spl_fixed31_32 arg)
* Good idea to use Newton's method
*/
- SPL_ASSERT(arg.value);
-
return spl_fixpt_from_fraction(
spl_fixpt_one.value,
arg.value);
@@ -346,7 +342,7 @@ struct spl_fixed31_32 spl_fixpt_exp(struct spl_fixed31_32 arg)
if (m > 0)
return spl_fixpt_shl(
spl_fixed31_32_exp_from_taylor_series(r),
- (unsigned char)m);
+ (unsigned int)m);
else
return spl_fixpt_div_int(
spl_fixed31_32_exp_from_taylor_series(r),
diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.h b/drivers/gpu/drm/amd/display/dc/sspl/spl_fixpt31_32.h
index ed2647f9a099..9f349ffe9148 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/spl_fixpt31_32.h
+++ b/drivers/gpu/drm/amd/display/dc/sspl/spl_fixpt31_32.h
@@ -189,7 +189,7 @@ static inline struct spl_fixed31_32 spl_fixpt_clamp(
* @brief
* result = arg << shift
*/
-static inline struct spl_fixed31_32 spl_fixpt_shl(struct spl_fixed31_32 arg, unsigned char shift)
+static inline struct spl_fixed31_32 spl_fixpt_shl(struct spl_fixed31_32 arg, unsigned int shift)
{
SPL_ASSERT(((arg.value >= 0) && (arg.value <= LLONG_MAX >> shift)) ||
((arg.value < 0) && (arg.value >= ~(LLONG_MAX >> shift))));
@@ -203,7 +203,7 @@ static inline struct spl_fixed31_32 spl_fixpt_shl(struct spl_fixed31_32 arg, uns
* @brief
* result = arg >> shift
*/
-static inline struct spl_fixed31_32 spl_fixpt_shr(struct spl_fixed31_32 arg, unsigned char shift)
+static inline struct spl_fixed31_32 spl_fixpt_shr(struct spl_fixed31_32 arg, unsigned int shift)
{
bool negative = arg.value < 0;
diff --git a/drivers/gpu/drm/amd/display/dc/spl/spl_os_types.h b/drivers/gpu/drm/amd/display/dc/sspl/spl_os_types.h
index 2e6ba71960ac..2e6ba71960ac 100644
--- a/drivers/gpu/drm/amd/display/dc/spl/spl_os_types.h
+++ b/drivers/gpu/drm/amd/display/dc/sspl/spl_os_types.h
diff --git a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h
index b353c4ceb60d..0bafb6710761 100644
--- a/drivers/gpu/drm/amd/display/dmub/dmub_srv.h
+++ b/drivers/gpu/drm/amd/display/dmub/dmub_srv.h
@@ -51,8 +51,8 @@
* for the cache windows.
*
* The call to dmub_srv_hw_init() programs the DMCUB registers to prepare
- * for command submission. Commands can be queued via dmub_srv_cmd_queue()
- * and executed via dmub_srv_cmd_execute().
+ * for command submission. Commands can be queued via dmub_srv_fb_cmd_queue()
+ * and executed via dmub_srv_fb_cmd_execute().
*
* If the queue is full the dmub_srv_wait_for_idle() call can be used to
* wait until the queue has been cleared.
@@ -69,6 +69,9 @@
#define DMUB_PC_SNAPSHOT_COUNT 10
+/* Default tracebuffer size if meta is absent. */
+#define DMUB_TRACE_BUFFER_SIZE (64 * 1024)
+
/* Forward declarations */
struct dmub_srv;
struct dmub_srv_common_regs;
@@ -111,6 +114,7 @@ enum dmub_asic {
DMUB_ASIC_DCN321,
DMUB_ASIC_DCN35,
DMUB_ASIC_DCN351,
+ DMUB_ASIC_DCN36,
DMUB_ASIC_DCN401,
DMUB_ASIC_MAX,
};
@@ -125,7 +129,9 @@ enum dmub_window_id {
DMUB_WINDOW_5_TRACEBUFF,
DMUB_WINDOW_6_FW_STATE,
DMUB_WINDOW_7_SCRATCH_MEM,
+ DMUB_WINDOW_IB_MEM,
DMUB_WINDOW_SHARED_STATE,
+ DMUB_WINDOW_LSDMA_BUFFER,
DMUB_WINDOW_TOTAL,
};
@@ -138,6 +144,7 @@ enum dmub_notification_type {
DMUB_NOTIFICATION_SET_CONFIG_REPLY,
DMUB_NOTIFICATION_DPIA_NOTIFICATION,
DMUB_NOTIFICATION_HPD_SENSE_NOTIFY,
+ DMUB_NOTIFICATION_FUSED_IO,
DMUB_NOTIFICATION_MAX
};
@@ -166,6 +173,13 @@ enum dmub_srv_power_state_type {
DMUB_POWER_STATE_D3 = 8
};
+/* enum dmub_inbox_cmd_interface type - defines default interface for host->dmub commands */
+enum dmub_inbox_cmd_interface_type {
+ DMUB_CMD_INTERFACE_DEFAULT = 0,
+ DMUB_CMD_INTERFACE_FB = 1,
+ DMUB_CMD_INTERFACE_REG = 2,
+};
+
/**
* struct dmub_region - dmub hw memory region
* @base: base address for region, must be 256 byte aligned
@@ -309,7 +323,7 @@ struct dmub_srv_hw_params {
* @timeout_occured: Indicates a timeout occured on any message from driver to dmub
* @timeout_cmd: first cmd sent from driver that timed out - subsequent timeouts are not stored
*/
-struct dmub_srv_debug {
+struct dmub_timeout_info {
bool timeout_occured;
union dmub_rb_cmd timeout_cmd;
unsigned long long timestamp;
@@ -336,13 +350,29 @@ struct dmub_diagnostic_data {
uint32_t outbox1_wptr;
uint32_t outbox1_size;
uint32_t gpint_datain0;
- struct dmub_srv_debug timeout_info;
+ struct dmub_timeout_info timeout_info;
uint8_t is_dmcub_enabled : 1;
uint8_t is_dmcub_soft_reset : 1;
uint8_t is_dmcub_secure_reset : 1;
uint8_t is_traceport_en : 1;
uint8_t is_cw0_enabled : 1;
uint8_t is_cw6_enabled : 1;
+ uint8_t is_pwait : 1;
+};
+
+struct dmub_srv_inbox {
+ /* generic status */
+ uint64_t num_submitted;
+ uint64_t num_reported;
+ union {
+ /* frame buffer mailbox status */
+ struct dmub_rb rb;
+ /* register mailbox status */
+ struct {
+ bool is_pending;
+ bool is_multi_pending;
+ };
+ };
};
/**
@@ -418,6 +448,8 @@ struct dmub_srv_hw_funcs {
uint32_t (*emul_get_inbox1_rptr)(struct dmub_srv *dmub);
+ uint32_t (*emul_get_inbox1_wptr)(struct dmub_srv *dmub);
+
void (*emul_set_inbox1_wptr)(struct dmub_srv *dmub, uint32_t wptr_offset);
bool (*is_supported)(struct dmub_srv *dmub);
@@ -452,24 +484,27 @@ struct dmub_srv_hw_funcs {
void (*send_inbox0_cmd)(struct dmub_srv *dmub, union dmub_inbox0_data_register data);
uint32_t (*get_current_time)(struct dmub_srv *dmub);
- void (*get_diagnostic_data)(struct dmub_srv *dmub, struct dmub_diagnostic_data *dmub_oca);
+ void (*get_diagnostic_data)(struct dmub_srv *dmub);
bool (*should_detect)(struct dmub_srv *dmub);
void (*init_reg_offsets)(struct dmub_srv *dmub, struct dc_context *ctx);
void (*subvp_save_surf_addr)(struct dmub_srv *dmub, const struct dc_plane_address *addr, uint8_t subvp_index);
+
void (*send_reg_inbox0_cmd_msg)(struct dmub_srv *dmub,
union dmub_rb_cmd *cmd);
uint32_t (*read_reg_inbox0_rsp_int_status)(struct dmub_srv *dmub);
void (*read_reg_inbox0_cmd_rsp)(struct dmub_srv *dmub,
union dmub_rb_cmd *cmd);
void (*write_reg_inbox0_rsp_int_ack)(struct dmub_srv *dmub);
+ void (*clear_reg_inbox0_rsp_int_ack)(struct dmub_srv *dmub);
+ void (*enable_reg_inbox0_rsp_int)(struct dmub_srv *dmub, bool enable);
+
uint32_t (*read_reg_outbox0_rdy_int_status)(struct dmub_srv *dmub);
void (*write_reg_outbox0_rdy_int_ack)(struct dmub_srv *dmub);
void (*read_reg_outbox0_msg)(struct dmub_srv *dmub, uint32_t *msg);
void (*write_reg_outbox0_rsp)(struct dmub_srv *dmub, uint32_t *rsp);
uint32_t (*read_reg_outbox0_rsp_int_status)(struct dmub_srv *dmub);
- void (*enable_reg_inbox0_rsp_int)(struct dmub_srv *dmub, bool enable);
void (*enable_reg_outbox0_rdy_int)(struct dmub_srv *dmub, bool enable);
};
@@ -489,6 +524,7 @@ struct dmub_srv_create_params {
enum dmub_asic asic;
uint32_t fw_version;
bool is_virtual;
+ enum dmub_inbox_cmd_interface_type inbox_type;
};
/**
@@ -506,6 +542,7 @@ struct dmub_srv {
uint32_t fw_version;
bool is_virtual;
struct dmub_fb scratch_mem_fb;
+ struct dmub_fb ib_mem_gart;
volatile struct dmub_shared_state_feature_block *shared_state;
volatile const struct dmub_fw_state *fw_state;
@@ -515,11 +552,11 @@ struct dmub_srv {
struct dmub_srv_dcn32_regs *regs_dcn32;
struct dmub_srv_dcn35_regs *regs_dcn35;
const struct dmub_srv_dcn401_regs *regs_dcn401;
-
struct dmub_srv_base_funcs funcs;
struct dmub_srv_hw_funcs hw_funcs;
- struct dmub_rb inbox1_rb;
+ struct dmub_srv_inbox inbox1;
uint32_t inbox1_last_wptr;
+ struct dmub_srv_inbox reg_inbox0;
/**
* outbox1_rb is accessed without locks (dal & dc)
* and to be used only in dmub_srv_stat_get_notification()
@@ -539,9 +576,11 @@ struct dmub_srv {
struct dmub_fw_meta_info meta_info;
struct dmub_feature_caps feature_caps;
struct dmub_visual_confirm_color visual_confirm_color;
+ enum dmub_inbox_cmd_interface_type inbox_type;
enum dmub_srv_power_state_type power_state;
- struct dmub_srv_debug debug;
+ struct dmub_diagnostic_data debug;
+ struct dmub_fb lsdma_rb_fb;
};
/**
@@ -563,22 +602,11 @@ struct dmub_notification {
struct aux_reply_data aux_reply;
enum dp_hpd_status hpd_status;
enum set_config_status sc_status;
- /**
- * DPIA notification command.
- */
- struct dmub_rb_cmd_dpia_notification dpia_notification;
struct dmub_rb_cmd_hpd_sense_notify_data hpd_sense_notify;
+ struct dmub_cmd_fused_request fused_request;
};
};
-/* enum dmub_ips_mode - IPS mode identifier */
-enum dmub_ips_mode {
- DMUB_IPS_MODE_IPS1_MAX = 0,
- DMUB_IPS_MODE_IPS2,
- DMUB_IPS_MODE_IPS1_RCG,
- DMUB_IPS_MODE_IPS1_ONO2_ON
-};
-
/**
* DMUB firmware version helper macro - useful for checking if the version
* of a firmware to know if feature or functionality is supported or present.
@@ -696,19 +724,7 @@ enum dmub_status dmub_srv_hw_init(struct dmub_srv *dmub,
enum dmub_status dmub_srv_hw_reset(struct dmub_srv *dmub);
/**
- * dmub_srv_sync_inbox1() - sync sw state with hw state
- * @dmub: the dmub service
- *
- * Sync sw state with hw state when resume from S0i3
- *
- * Return:
- * DMUB_STATUS_OK - success
- * DMUB_STATUS_INVALID - unspecified error
- */
-enum dmub_status dmub_srv_sync_inbox1(struct dmub_srv *dmub);
-
-/**
- * dmub_srv_cmd_queue() - queues a command to the DMUB
+ * dmub_srv_fb_cmd_queue() - queues a command to the DMUB
* @dmub: the dmub service
* @cmd: the command to queue
*
@@ -720,11 +736,11 @@ enum dmub_status dmub_srv_sync_inbox1(struct dmub_srv *dmub);
* DMUB_STATUS_QUEUE_FULL - no remaining room in queue
* DMUB_STATUS_INVALID - unspecified error
*/
-enum dmub_status dmub_srv_cmd_queue(struct dmub_srv *dmub,
+enum dmub_status dmub_srv_fb_cmd_queue(struct dmub_srv *dmub,
const union dmub_rb_cmd *cmd);
/**
- * dmub_srv_cmd_execute() - Executes a queued sequence to the dmub
+ * dmub_srv_fb_cmd_execute() - Executes a queued sequence to the dmub
* @dmub: the dmub service
*
* Begins execution of queued commands on the dmub.
@@ -733,7 +749,7 @@ enum dmub_status dmub_srv_cmd_queue(struct dmub_srv *dmub,
* DMUB_STATUS_OK - success
* DMUB_STATUS_INVALID - unspecified error
*/
-enum dmub_status dmub_srv_cmd_execute(struct dmub_srv *dmub);
+enum dmub_status dmub_srv_fb_cmd_execute(struct dmub_srv *dmub);
/**
* dmub_srv_wait_for_hw_pwr_up() - Waits for firmware hardware power up is completed
@@ -792,6 +808,23 @@ enum dmub_status dmub_srv_wait_for_phy_init(struct dmub_srv *dmub,
uint32_t timeout_us);
/**
+ * dmub_srv_wait_for_pending() - Re-entrant wait for messages currently pending
+ * @dmub: the dmub service
+ * @timeout_us: the maximum number of microseconds to wait
+ *
+ * Waits until the commands queued prior to this call are complete.
+ * If interfaces remain busy due to additional work being submitted
+ * concurrently, this function will not continue to wait.
+ *
+ * Return:
+ * DMUB_STATUS_OK - success
+ * DMUB_STATUS_TIMEOUT - wait for buffer to flush timed out
+ * DMUB_STATUS_INVALID - unspecified error
+ */
+enum dmub_status dmub_srv_wait_for_pending(struct dmub_srv *dmub,
+ uint32_t timeout_us);
+
+/**
* dmub_srv_wait_for_idle() - Waits for the DMUB to be idle
* @dmub: the dmub service
* @timeout_us: the maximum number of microseconds to wait
@@ -889,15 +922,12 @@ enum dmub_status dmub_srv_get_fw_boot_status(struct dmub_srv *dmub,
enum dmub_status dmub_srv_get_fw_boot_option(struct dmub_srv *dmub,
union dmub_fw_boot_options *option);
-enum dmub_status dmub_srv_cmd_with_reply_data(struct dmub_srv *dmub,
- union dmub_rb_cmd *cmd);
-
enum dmub_status dmub_srv_set_skip_panel_power_sequence(struct dmub_srv *dmub,
bool skip);
bool dmub_srv_get_outbox0_msg(struct dmub_srv *dmub, struct dmcub_trace_buf_entry *entry);
-bool dmub_srv_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data);
+bool dmub_srv_get_diagnostic_data(struct dmub_srv *dmub);
bool dmub_srv_should_detect(struct dmub_srv *dmub);
@@ -956,26 +986,6 @@ enum dmub_status dmub_srv_clear_inbox0_ack(struct dmub_srv *dmub);
void dmub_srv_subvp_save_surf_addr(struct dmub_srv *dmub, const struct dc_plane_address *addr, uint8_t subvp_index);
/**
- * dmub_srv_send_reg_inbox0_cmd() - send a dmub command and wait for the command
- * being processed by DMUB.
- * @dmub: The dmub service
- * @cmd: The dmub command being sent. If with_replay is true, the function will
- * update cmd with replied data.
- * @with_reply: true if DMUB reply needs to be copied back to cmd. false if the
- * cmd doesn't need to be replied.
- * @timeout_us: timeout in microseconds.
- *
- * Return:
- * DMUB_STATUS_OK - success
- * DMUB_STATUS_TIMEOUT - DMUB fails to process the command within the timeout
- * interval.
- */
-enum dmub_status dmub_srv_send_reg_inbox0_cmd(
- struct dmub_srv *dmub,
- union dmub_rb_cmd *cmd,
- bool with_reply, uint32_t timeout_us);
-
-/**
* dmub_srv_set_power_state() - Track DC power state in dmub_srv
* @dmub: The dmub service
* @power_state: DC power state setting
@@ -987,4 +997,71 @@ enum dmub_status dmub_srv_send_reg_inbox0_cmd(
*/
void dmub_srv_set_power_state(struct dmub_srv *dmub, enum dmub_srv_power_state_type dmub_srv_power_state);
+/**
+ * dmub_srv_reg_cmd_execute() - Executes provided command to the dmub
+ * @dmub: the dmub service
+ * @cmd: the command packet to be executed
+ *
+ * Executes a single command for the dmub.
+ *
+ * Return:
+ * DMUB_STATUS_OK - success
+ * DMUB_STATUS_INVALID - unspecified error
+ */
+enum dmub_status dmub_srv_reg_cmd_execute(struct dmub_srv *dmub, union dmub_rb_cmd *cmd);
+
+
+/**
+ * dmub_srv_cmd_get_response() - Copies return data for command into buffer
+ * @dmub: the dmub service
+ * @cmd_rsp: response buffer
+ *
+ * Copies return data for command into buffer
+ */
+void dmub_srv_cmd_get_response(struct dmub_srv *dmub,
+ union dmub_rb_cmd *cmd_rsp);
+
+/**
+ * dmub_srv_sync_inboxes() - Sync inbox state
+ * @dmub: the dmub service
+ *
+ * Sync inbox state
+ *
+ * Return:
+ * DMUB_STATUS_OK - success
+ * DMUB_STATUS_INVALID - unspecified error
+ */
+enum dmub_status dmub_srv_sync_inboxes(struct dmub_srv *dmub);
+
+/**
+ * dmub_srv_wait_for_inbox_free() - Waits for space in the DMUB inbox to free up
+ * @dmub: the dmub service
+ * @timeout_us: the maximum number of microseconds to wait
+ * @num_free_required: number of free entries required
+ *
+ * Waits until the DMUB buffer is freed to the specified number.
+ * The maximum wait time is given in microseconds to prevent spinning
+ * forever.
+ *
+ * Return:
+ * DMUB_STATUS_OK - success
+ * DMUB_STATUS_TIMEOUT - wait for buffer to flush timed out
+ * DMUB_STATUS_INVALID - unspecified error
+ */
+enum dmub_status dmub_srv_wait_for_inbox_free(struct dmub_srv *dmub,
+ uint32_t timeout_us,
+ uint32_t num_free_required);
+
+/**
+ * dmub_srv_update_inbox_status() - Updates pending status for inbox & reg inbox0
+ * @dmub: the dmub service
+ *
+ * Return:
+ * DMUB_STATUS_OK - success
+ * DMUB_STATUS_TIMEOUT - wait for buffer to flush timed out
+ * DMUB_STATUS_HW_FAILURE - issue with HW programming
+ * DMUB_STATUS_INVALID - unspecified error
+ */
+enum dmub_status dmub_srv_update_inbox_status(struct dmub_srv *dmub);
+
#endif /* _DMUB_SRV_H_ */
diff --git a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
index b800a507d1e0..c587b3441e07 100644
--- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
+++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
@@ -36,6 +36,9 @@
//<DMUB_TYPES>==================================================================
/* Basic type definitions. */
+#ifdef __forceinline
+#undef __forceinline
+#endif
#define __forceinline inline
/**
@@ -101,6 +104,14 @@
*/
#define DMUB_MAX_FPO_STREAMS 4
+/* Define to ensure that the "common" members always appear in the same
+ * order in different structs for back compat purposes
+ */
+#define COMMON_STREAM_STATIC_SUB_STATE \
+ struct dmub_fams2_cmd_legacy_stream_static_state legacy; \
+ struct dmub_fams2_cmd_subvp_stream_static_state subvp; \
+ struct dmub_fams2_cmd_drr_stream_static_state drr;
+
/* Maximum number of streams on any ASIC. */
#define DMUB_MAX_STREAMS 6
@@ -161,6 +172,13 @@
#endif
/**
+ * OS/FW agnostic memcmp
+ */
+#ifndef dmub_memcmp
+#define dmub_memcmp(lhs, rhs, bytes) memcmp((lhs), (rhs), (bytes))
+#endif
+
+/**
* OS/FW agnostic udelay
*/
#ifndef dmub_udelay
@@ -281,6 +299,31 @@ union dmub_addr {
} u; /*<< Low/high bit access */
uint64_t quad_part; /*<< 64 bit address */
};
+
+/* Flattened structure containing SOC BB parameters stored in the VBIOS
+ * It is not practical to store the entire bounding box in VBIOS since the bounding box struct can gain new parameters.
+ * This also prevents alighment issues when new parameters are added to the SoC BB.
+ * The following parameters should be added since these values can't be obtained elsewhere:
+ * -dml2_soc_power_management_parameters
+ * -dml2_soc_vmin_clock_limits
+ */
+struct dmub_soc_bb_params {
+ uint32_t dram_clk_change_blackout_ns;
+ uint32_t dram_clk_change_read_only_ns;
+ uint32_t dram_clk_change_write_only_ns;
+ uint32_t fclk_change_blackout_ns;
+ uint32_t g7_ppt_blackout_ns;
+ uint32_t stutter_enter_plus_exit_latency_ns;
+ uint32_t stutter_exit_latency_ns;
+ uint32_t z8_stutter_enter_plus_exit_latency_ns;
+ uint32_t z8_stutter_exit_latency_ns;
+ uint32_t z8_min_idle_time_ns;
+ uint32_t type_b_dram_clk_change_blackout_ns;
+ uint32_t type_b_ppt_blackout_ns;
+ uint32_t vmin_limit_dispclk_khz;
+ uint32_t vmin_limit_dcfclk_khz;
+ uint32_t g7_temperature_read_blackout_ns;
+};
#pragma pack(pop)
/**
@@ -431,7 +474,68 @@ union replay_debug_flags {
*/
uint32_t enable_ips_residency_profiling : 1;
- uint32_t reserved : 20;
+ /**
+ * 0x1000 (bit 12)
+ * @enable_coasting_vtotal_check: Enable Coasting_vtotal_check
+ */
+ uint32_t enable_coasting_vtotal_check : 1;
+ /**
+ * 0x2000 (bit 13)
+ * @enable_visual_confirm_debug: Enable Visual Confirm Debug
+ */
+ uint32_t enable_visual_confirm_debug : 1;
+
+ uint32_t reserved : 18;
+ } bitfields;
+
+ uint32_t u32All;
+};
+
+/**
+ * Flags record error state.
+ */
+union replay_visual_confirm_error_state_flags {
+ struct {
+ /**
+ * 0x1 (bit 0) - Desync Error flag.
+ */
+ uint32_t desync_error : 1;
+
+ /**
+ * 0x2 (bit 1) - State Transition Error flag.
+ */
+ uint32_t state_transition_error : 1;
+
+ /**
+ * 0x4 (bit 2) - Crc Error flag
+ */
+ uint32_t crc_error : 1;
+
+ /**
+ * 0x8 (bit 3) - Reserved
+ */
+ uint32_t reserved_3 : 1;
+
+ /**
+ * 0x10 (bit 4) - Incorrect Coasting vtotal checking --> use debug flag to control DPCD write.
+ * Added new debug flag to control DPCD.
+ */
+ uint32_t incorrect_vtotal_in_static_screen : 1;
+
+ /**
+ * 0x20 (bit 5) - No doubled Refresh Rate.
+ */
+ uint32_t no_double_rr : 1;
+
+ /**
+ * Reserved bit 6-7
+ */
+ uint32_t reserved_6_7 : 2;
+
+ /**
+ * Reserved bit 9-31
+ */
+ uint32_t reserved_9_31 : 24;
} bitfields;
uint32_t u32All;
@@ -475,11 +579,28 @@ union replay_hw_flags {
* Use TPS3 signal when restore main link.
*/
uint32_t force_wakeup_by_tps3 : 1;
+ /**
+ * @is_alpm_initialized: Indicates whether ALPM is initialized
+ */
+ uint32_t is_alpm_initialized : 1;
+
+ /**
+ * @alpm_mode: Indicates ALPM mode selected
+ */
+ uint32_t alpm_mode : 2;
} bitfields;
uint32_t u32All;
};
+union fw_assisted_mclk_switch_version {
+ struct {
+ uint8_t minor : 5;
+ uint8_t major : 3;
+ };
+ uint8_t ver;
+};
+
/**
* DMUB feature capabilities.
* After DMUB init, driver will query FW capabilities prior to enabling certain features.
@@ -659,6 +780,21 @@ enum dmub_ips_disable_type {
DMUB_IPS_DISABLE_IPS2_Z10 = 4,
DMUB_IPS_DISABLE_DYNAMIC = 5,
DMUB_IPS_RCG_IN_ACTIVE_IPS2_IN_OFF = 6,
+ DMUB_IPS_DISABLE_Z8_RETENTION = 7,
+};
+
+enum dmub_ips_rcg_disable_type {
+ DMUB_IPS_RCG_ENABLE = 0,
+ DMUB_IPS0_RCG_DISABLE = 1,
+ DMUB_IPS1_RCG_DISABLE = 2,
+ DMUB_IPS_RCG_DISABLE = 3
+};
+
+enum dmub_ips_in_vpb_disable_type {
+ DMUB_IPS_VPB_RCG_ONLY = 0, // Legacy behaviour
+ DMUB_IPS_VPB_DISABLE_ALL = 1,
+ DMUB_IPS_VPB_ENABLE_IPS1_AND_RCG = 2,
+ DMUB_IPS_VPB_ENABLE_ALL = 3 // Enable IPS1 Z8, IPS1 and RCG
};
#define DMUB_IPS1_ALLOW_MASK 0x00000001
@@ -666,6 +802,17 @@ enum dmub_ips_disable_type {
#define DMUB_IPS1_COMMIT_MASK 0x00000004
#define DMUB_IPS2_COMMIT_MASK 0x00000008
+enum dmub_ips_comand_type {
+ /**
+ * Start/stop IPS residency measurements for a given IPS mode
+ */
+ DMUB_CMD__IPS_RESIDENCY_CNTL = 0,
+ /**
+ * Query IPS residency information for a given IPS mode
+ */
+ DMUB_CMD__IPS_QUERY_RESIDENCY_INFO = 1,
+};
+
/**
* union dmub_fw_boot_options - Boot option definitions for SCRATCH14
*/
@@ -735,13 +882,14 @@ enum dmub_shared_state_feature_id {
/**
* struct dmub_shared_state_ips_fw - Firmware signals for IPS.
*/
-union dmub_shared_state_ips_fw_signals {
+ union dmub_shared_state_ips_fw_signals {
struct {
- uint32_t ips1_commit : 1; /**< 1 if in IPS1 */
+ uint32_t ips1_commit : 1; /**< 1 if in IPS1 or IPS0 RCG */
uint32_t ips2_commit : 1; /**< 1 if in IPS2 */
uint32_t in_idle : 1; /**< 1 if DMCUB is in idle */
uint32_t detection_required : 1; /**< 1 if detection is required */
- uint32_t reserved_bits : 28; /**< Reversed */
+ uint32_t ips1z8_commit: 1; /**< 1 if in IPS1 Z8 Retention */
+ uint32_t reserved_bits : 27; /**< Reversed */
} bits;
uint32_t all;
};
@@ -749,14 +897,19 @@ union dmub_shared_state_ips_fw_signals {
/**
* struct dmub_shared_state_ips_signals - Firmware signals for IPS.
*/
-union dmub_shared_state_ips_driver_signals {
+ union dmub_shared_state_ips_driver_signals {
struct {
uint32_t allow_pg : 1; /**< 1 if PG is allowed */
uint32_t allow_ips1 : 1; /**< 1 is IPS1 is allowed */
uint32_t allow_ips2 : 1; /**< 1 is IPS1 is allowed */
uint32_t allow_z10 : 1; /**< 1 if Z10 is allowed */
uint32_t allow_idle: 1; /**< 1 if driver is allowing idle */
- uint32_t reserved_bits : 27; /**< Reversed bits */
+ uint32_t allow_ips0_rcg : 1; /**< 1 is IPS0 RCG is allowed */
+ uint32_t allow_ips1_rcg : 1; /**< 1 is IPS1 RCG is allowed */
+ uint32_t allow_ips1z8 : 1; /**< 1 is IPS1 Z8 Retention is allowed */
+ uint32_t allow_dynamic_ips1 : 1; /**< 1 if IPS1 is allowed in dynamic use cases such as VPB */
+ uint32_t allow_dynamic_ips1_z8: 1; /**< 1 if IPS1 z8 ret is allowed in dynamic use cases such as VPB */
+ uint32_t reserved_bits : 22; /**< Reversed bits */
} bits;
uint32_t all;
};
@@ -785,7 +938,9 @@ struct dmub_shared_state_ips_fw {
uint32_t ips1_exit_count; /**< Exit counter for IPS1 */
uint32_t ips2_entry_count; /**< Entry counter for IPS2 */
uint32_t ips2_exit_count; /**< Exit counter for IPS2 */
- uint32_t reserved[55]; /**< Reversed, to be updated when adding new fields. */
+ uint32_t ips1_z8ret_entry_count; /**< Entry counter for IPS1 Z8 Retention */
+ uint32_t ips1_z8ret_exit_count; /**< Exit counter for IPS1 Z8 Retention */
+ uint32_t reserved[53]; /**< Reversed, to be updated when adding new fields. */
}; /* 248-bytes, fixed */
/**
@@ -1173,6 +1328,10 @@ enum dmub_gpint_command {
* DESC: Setup debug configs.
*/
DMUB_GPINT__SETUP_DEBUG_MODE = 136,
+ /**
+ * DESC: Initiates IPS wake sequence.
+ */
+ DMUB_GPINT__IPS_DEBUG_WAKE = 137,
};
/**
@@ -1252,6 +1411,16 @@ enum dmub_inbox0_command {
#define DMUB_RB_SIZE (DMUB_RB_CMD_SIZE * DMUB_RB_MAX_ENTRY)
/**
+ * Maximum number of items in the DMUB REG INBOX0 internal ringbuffer.
+ */
+#define DMUB_REG_INBOX0_RB_MAX_ENTRY 16
+
+/**
+ * Ringbuffer size in bytes.
+ */
+#define DMUB_REG_INBOX0_RB_SIZE (DMUB_RB_CMD_SIZE * DMUB_REG_INBOX0_RB_MAX_ENTRY)
+
+/**
* REG_SET mask for reg offload.
*/
#define REG_SET_MASK 0xFFFF
@@ -1387,6 +1556,21 @@ enum dmub_cmd_type {
*/
DMUB_CMD__PSP = 88,
+ /**
+ * Command type used for all Fused IO commands.
+ */
+ DMUB_CMD__FUSED_IO = 89,
+
+ /**
+ * Command type used for all LSDMA commands.
+ */
+ DMUB_CMD__LSDMA = 90,
+
+ /**
+ * Command type use for all IPS commands.
+ */
+ DMUB_CMD__IPS = 91,
+
DMUB_CMD__VBIOS = 128,
};
@@ -1418,6 +1602,10 @@ enum dmub_out_cmd_type {
* Command type used for HPD redetect notification
*/
DMUB_OUT_CMD__HPD_SENSE_NOTIFY = 6,
+ /**
+ * Command type used for Fused IO notification
+ */
+ DMUB_OUT_CMD__FUSED_IO = 7,
};
/* DMUB_CMD__DPIA command sub-types. */
@@ -1444,7 +1632,8 @@ struct dmub_cmd_header {
unsigned int sub_type : 8; /**< command sub type */
unsigned int ret_status : 1; /**< 1 if returned data, 0 otherwise */
unsigned int multi_cmd_pending : 1; /**< 1 if multiple commands chained together */
- unsigned int reserved0 : 6; /**< reserved bits */
+ unsigned int is_reg_based : 1; /**< 1 if register based mailbox cmd, 0 if FB based cmd */
+ unsigned int reserved0 : 5; /**< reserved bits */
unsigned int payload_bytes : 6; /* payload excluding header - up to 60 bytes */
unsigned int reserved1 : 2; /**< reserved bits */
};
@@ -1792,6 +1981,121 @@ struct dmub_rb_cmd_fams2_flip {
struct dmub_fams2_flip_info flip_info;
};
+struct dmub_cmd_lsdma_data {
+ union {
+ struct lsdma_init_data {
+ union dmub_addr gpu_addr_base;
+ uint32_t ring_size;
+ } init_data;
+ struct lsdma_tiled_copy_data {
+ uint32_t src_addr_lo;
+ uint32_t src_addr_hi;
+ uint32_t dst_addr_lo;
+ uint32_t dst_addr_hi;
+
+ uint32_t src_x : 16;
+ uint32_t src_y : 16;
+
+ uint32_t src_width : 16;
+ uint32_t src_height : 16;
+
+ uint32_t dst_x : 16;
+ uint32_t dst_y : 16;
+
+ uint32_t dst_width : 16;
+ uint32_t dst_height : 16;
+
+ uint32_t rect_x : 16;
+ uint32_t rect_y : 16;
+
+ uint32_t src_swizzle_mode : 5;
+ uint32_t src_mip_max : 5;
+ uint32_t src_mip_id : 5;
+ uint32_t dst_mip_max : 5;
+ uint32_t dst_swizzle_mode : 5;
+ uint32_t dst_mip_id : 5;
+ uint32_t tmz : 1;
+ uint32_t dcc : 1;
+
+ uint32_t data_format : 6;
+ uint32_t padding1 : 4;
+ uint32_t dst_element_size : 3;
+ uint32_t num_type : 3;
+ uint32_t src_element_size : 3;
+ uint32_t write_compress : 2;
+ uint32_t cache_policy_dst : 2;
+ uint32_t cache_policy_src : 2;
+ uint32_t read_compress : 2;
+ uint32_t src_dim : 2;
+ uint32_t dst_dim : 2;
+ uint32_t max_uncom : 1;
+
+ uint32_t max_com : 2;
+ uint32_t padding : 30;
+ } tiled_copy_data;
+ struct lsdma_linear_copy_data {
+ uint32_t count : 30;
+ uint32_t cache_policy_dst : 2;
+
+ uint32_t tmz : 1;
+ uint32_t cache_policy_src : 2;
+ uint32_t padding : 29;
+
+ uint32_t src_lo;
+ uint32_t src_hi;
+ uint32_t dst_lo;
+ uint32_t dst_hi;
+ } linear_copy_data;
+ struct lsdma_reg_write_data {
+ uint32_t reg_addr;
+ uint32_t reg_data;
+ } reg_write_data;
+ struct lsdma_pio_copy_data {
+ union {
+ struct {
+ uint32_t byte_count : 26;
+ uint32_t src_loc : 1;
+ uint32_t dst_loc : 1;
+ uint32_t src_addr_inc : 1;
+ uint32_t dst_addr_inc : 1;
+ uint32_t overlap_disable : 1;
+ uint32_t constant_fill : 1;
+ } fields;
+ uint32_t raw;
+ } packet;
+ uint32_t src_lo;
+ uint32_t src_hi;
+ uint32_t dst_lo;
+ uint32_t dst_hi;
+ } pio_copy_data;
+ struct lsdma_pio_constfill_data {
+ union {
+ struct {
+ uint32_t byte_count : 26;
+ uint32_t src_loc : 1;
+ uint32_t dst_loc : 1;
+ uint32_t src_addr_inc : 1;
+ uint32_t dst_addr_inc : 1;
+ uint32_t overlap_disable : 1;
+ uint32_t constant_fill : 1;
+ } fields;
+ uint32_t raw;
+ } packet;
+ uint32_t dst_lo;
+ uint32_t dst_hi;
+ uint32_t data;
+ } pio_constfill_data;
+
+ uint32_t all[14];
+ } u;
+
+};
+
+struct dmub_rb_cmd_lsdma {
+ struct dmub_cmd_header header;
+ struct dmub_cmd_lsdma_data lsdma_data;
+};
+
struct dmub_optc_state_v2 {
uint32_t v_total_min;
uint32_t v_total_max;
@@ -1823,52 +2127,33 @@ enum fams2_stream_type {
FAMS2_STREAM_TYPE_SUBVP = 4,
};
-/* dynamic stream state */
-struct dmub_fams2_legacy_stream_dynamic_state {
- uint8_t force_allow_at_vblank;
- uint8_t pad[3];
-};
+struct dmub_rect16 {
+ /**
+ * Dirty rect x offset.
+ */
+ uint16_t x;
-struct dmub_fams2_subvp_stream_dynamic_state {
- uint16_t viewport_start_hubp_vline;
- uint16_t viewport_height_hubp_vlines;
- uint16_t viewport_start_c_hubp_vline;
- uint16_t viewport_height_c_hubp_vlines;
- uint16_t phantom_viewport_height_hubp_vlines;
- uint16_t phantom_viewport_height_c_hubp_vlines;
- uint16_t microschedule_start_otg_vline;
- uint16_t mall_start_otg_vline;
- uint16_t mall_start_hubp_vline;
- uint16_t mall_start_c_hubp_vline;
- uint8_t force_allow_at_vblank_only;
- uint8_t pad[3];
-};
+ /**
+ * Dirty rect y offset.
+ */
+ uint16_t y;
-struct dmub_fams2_drr_stream_dynamic_state {
- uint16_t stretched_vtotal;
- uint8_t use_cur_vtotal;
- uint8_t pad;
-};
+ /**
+ * Dirty rect width.
+ */
+ uint16_t width;
-struct dmub_fams2_stream_dynamic_state {
- uint64_t ref_tick;
- uint32_t cur_vtotal;
- uint16_t adjusted_allow_end_otg_vline;
- uint8_t pad[2];
- struct dmub_optc_position ref_otg_pos;
- struct dmub_optc_position target_otg_pos;
- union {
- struct dmub_fams2_legacy_stream_dynamic_state legacy;
- struct dmub_fams2_subvp_stream_dynamic_state subvp;
- struct dmub_fams2_drr_stream_dynamic_state drr;
- } sub_state;
+ /**
+ * Dirty rect height.
+ */
+ uint16_t height;
};
/* static stream state */
struct dmub_fams2_legacy_stream_static_state {
uint8_t vactive_det_fill_delay_otg_vlines;
uint8_t programming_delay_otg_vlines;
-};
+}; //v0
struct dmub_fams2_subvp_stream_static_state {
uint16_t vratio_numerator;
@@ -1887,14 +2172,61 @@ struct dmub_fams2_subvp_stream_static_state {
uint8_t phantom_otg_inst;
uint8_t phantom_pipe_mask;
uint8_t phantom_plane_pipe_masks[DMUB_MAX_PHANTOM_PLANES]; // phantom pipe mask per plane (for flip passthrough)
-};
+}; //v0
struct dmub_fams2_drr_stream_static_state {
uint16_t nom_stretched_vtotal;
uint8_t programming_delay_otg_vlines;
uint8_t only_stretch_if_required;
uint8_t pad[2];
-};
+}; //v0
+
+struct dmub_fams2_cmd_legacy_stream_static_state {
+ uint16_t vactive_det_fill_delay_otg_vlines;
+ uint16_t programming_delay_otg_vlines;
+}; //v1
+
+struct dmub_fams2_cmd_subvp_stream_static_state {
+ uint16_t vratio_numerator;
+ uint16_t vratio_denominator;
+ uint16_t phantom_vtotal;
+ uint16_t phantom_vactive;
+ uint16_t programming_delay_otg_vlines;
+ uint16_t prefetch_to_mall_otg_vlines;
+ union {
+ struct {
+ uint8_t is_multi_planar : 1;
+ uint8_t is_yuv420 : 1;
+ } bits;
+ uint8_t all;
+ } config;
+ uint8_t phantom_otg_inst;
+ uint8_t phantom_pipe_mask;
+ uint8_t pad0;
+ uint8_t phantom_plane_pipe_masks[DMUB_MAX_PHANTOM_PLANES]; // phantom pipe mask per plane (for flip passthrough)
+ uint8_t pad1[4 - (DMUB_MAX_PHANTOM_PLANES % 4)];
+}; //v1
+
+struct dmub_fams2_cmd_drr_stream_static_state {
+ uint16_t nom_stretched_vtotal;
+ uint16_t programming_delay_otg_vlines;
+ uint8_t only_stretch_if_required;
+ uint8_t pad[3];
+}; //v1
+
+union dmub_fams2_stream_static_sub_state {
+ struct dmub_fams2_legacy_stream_static_state legacy;
+ struct dmub_fams2_subvp_stream_static_state subvp;
+ struct dmub_fams2_drr_stream_static_state drr;
+}; //v0
+
+union dmub_fams2_cmd_stream_static_sub_state {
+ COMMON_STREAM_STATIC_SUB_STATE
+}; //v1
+
+union dmub_fams2_stream_static_sub_state_v2 {
+ COMMON_STREAM_STATIC_SUB_STATE
+}; //v2
struct dmub_fams2_stream_static_state {
enum fams2_stream_type type;
@@ -1924,13 +2256,45 @@ struct dmub_fams2_stream_static_state {
uint8_t pipe_mask; // pipe mask for the whole config
uint8_t num_planes;
uint8_t plane_pipe_masks[DMUB_MAX_PLANES]; // pipe mask per plane (for flip passthrough)
- uint8_t pad[DMUB_MAX_PLANES % 4];
+ uint8_t pad[4 - (DMUB_MAX_PLANES % 4)];
+ union dmub_fams2_stream_static_sub_state sub_state;
+}; //v0
+
+struct dmub_fams2_cmd_stream_static_base_state {
+ enum fams2_stream_type type;
+ uint32_t otg_vline_time_ns;
+ uint32_t otg_vline_time_ticks;
+ uint16_t htotal;
+ uint16_t vtotal; // nominal vtotal
+ uint16_t vblank_start;
+ uint16_t vblank_end;
+ uint16_t max_vtotal;
+ uint16_t allow_start_otg_vline;
+ uint16_t allow_end_otg_vline;
+ uint16_t drr_keepout_otg_vline; // after this vline, vtotal cannot be changed
+ uint16_t scheduling_delay_otg_vlines; // min time to budget for ready to microschedule start
+ uint16_t contention_delay_otg_vlines; // time to budget for contention on execution
+ uint16_t vline_int_ack_delay_otg_vlines; // min time to budget for vertical interrupt firing
+ uint16_t allow_to_target_delay_otg_vlines; // time from allow vline to target vline
union {
- struct dmub_fams2_legacy_stream_static_state legacy;
- struct dmub_fams2_subvp_stream_static_state subvp;
- struct dmub_fams2_drr_stream_static_state drr;
- } sub_state;
-};
+ struct {
+ uint8_t is_drr : 1; // stream is DRR enabled
+ uint8_t clamp_vtotal_min : 1; // clamp vtotal to min instead of nominal
+ uint8_t min_ttu_vblank_usable : 1; // if min ttu vblank is above wm, no force pstate is needed in blank
+ } bits;
+ uint8_t all;
+ } config;
+ uint8_t otg_inst;
+ uint8_t pipe_mask; // pipe mask for the whole config
+ uint8_t num_planes;
+ uint8_t plane_pipe_masks[DMUB_MAX_PLANES]; // pipe mask per plane (for flip passthrough)
+ uint8_t pad[4 - (DMUB_MAX_PLANES % 4)];
+}; //v1
+
+struct dmub_fams2_stream_static_state_v1 {
+ struct dmub_fams2_cmd_stream_static_base_state base;
+ union dmub_fams2_stream_static_sub_state_v2 sub_state;
+}; //v1
/**
* enum dmub_fams2_allow_delay_check_mode - macroscheduler mode for breaking on excessive
@@ -1970,7 +2334,16 @@ struct dmub_cmd_fams2_global_config {
union dmub_cmd_fams2_config {
struct dmub_cmd_fams2_global_config global;
- struct dmub_fams2_stream_static_state stream;
+ struct dmub_fams2_stream_static_state stream; //v0
+ union {
+ struct dmub_fams2_cmd_stream_static_base_state base;
+ union dmub_fams2_cmd_stream_static_sub_state sub_state;
+ } stream_v1; //v1
+};
+
+struct dmub_fams2_config_v2 {
+ struct dmub_cmd_fams2_global_config global;
+ struct dmub_fams2_stream_static_state_v1 stream_v1[DMUB_MAX_STREAMS]; //v1
};
/**
@@ -1982,6 +2355,22 @@ struct dmub_rb_cmd_fams2 {
};
/**
+ * Indirect buffer descriptor
+ */
+struct dmub_ib_data {
+ union dmub_addr src; // location of indirect buffer in memory
+ uint16_t size; // indirect buffer size in bytes
+};
+
+/**
+ * DMUB rb command definition for commands passed over indirect buffer
+ */
+struct dmub_rb_cmd_ib {
+ struct dmub_cmd_header header;
+ struct dmub_ib_data ib_data;
+};
+
+/**
* enum dmub_cmd_idle_opt_type - Idle optimization command type.
*/
enum dmub_cmd_idle_opt_type {
@@ -2004,6 +2393,11 @@ enum dmub_cmd_idle_opt_type {
* DCN hardware notify power state.
*/
DMUB_CMD__IDLE_OPT_SET_DC_POWER_STATE = 3,
+
+ /**
+ * DCN notify to release HW.
+ */
+ DMUB_CMD__IDLE_OPT_RELEASE_HW = 4,
};
/**
@@ -2149,7 +2543,8 @@ struct dmub_dig_transmitter_control_data_v1_7 {
uint8_t connobj_id; /**< Connector Object Id defined in ObjectId.h */
uint8_t HPO_instance; /**< HPO instance (0: inst0, 1: inst1) */
uint8_t reserved1; /**< For future use */
- uint8_t reserved2[3]; /**< For future use */
+ uint8_t skip_phy_ssc_reduction;
+ uint8_t reserved2[2]; /**< For future use */
uint32_t reserved3[11]; /**< For future use */
};
@@ -2496,7 +2891,11 @@ enum dp_hpd_type {
/**
* DP HPD short pulse
*/
- DP_IRQ
+ DP_IRQ = 1,
+ /**
+ * Failure to acquire DP HPD state
+ */
+ DP_NONE_HPD = 2
};
/**
@@ -2763,6 +3162,7 @@ enum dmub_cmd_fams_type {
DMUB_CMD__FAMS2_CONFIG = 4,
DMUB_CMD__FAMS2_DRR_UPDATE = 5,
DMUB_CMD__FAMS2_FLIP = 6,
+ DMUB_CMD__FAMS2_IB_CONFIG = 7,
};
/**
@@ -3005,6 +3405,12 @@ struct dmub_cmd_psr_copy_settings_data {
* Some panels request main link off before xth vertical line
*/
uint16_t poweroff_before_vertical_line;
+ /**
+ * Some panels cannot handle idle pattern during PSR entry.
+ * To power down phy before disable stream to avoid sending
+ * idle pattern.
+ */
+ uint8_t power_down_phy_before_disable_stream;
};
/**
@@ -3463,6 +3869,12 @@ struct dmub_rb_cmd_psr_set_power_opt {
struct dmub_cmd_psr_set_power_opt_data psr_set_power_opt_data;
};
+enum dmub_alpm_mode {
+ ALPM_AUXWAKE = 0,
+ ALPM_AUXLESS = 1,
+ ALPM_UNSUPPORTED = 2,
+};
+
/**
* Definition of Replay Residency GPINT command.
* Bit[0] - Residency mode for Revision 0
@@ -3592,6 +4004,17 @@ enum dmub_cmd_replay_general_subtype {
*/
REPLAY_GENERAL_CMD_DISABLED_ADAPTIVE_SYNC_SDP,
REPLAY_GENERAL_CMD_DISABLED_DESYNC_ERROR_DETECTION,
+ REPLAY_GENERAL_CMD_UPDATE_ERROR_STATUS,
+ REPLAY_GENERAL_CMD_SET_LOW_RR_ACTIVATE,
+};
+
+struct dmub_alpm_auxless_data {
+ uint16_t lfps_setup_ns;
+ uint16_t lfps_period_ns;
+ uint16_t lfps_silence_ns;
+ uint16_t lfps_t1_t2_override_us;
+ short lfps_t1_t2_offset_us;
+ uint8_t lttpr_count;
};
/**
@@ -3625,6 +4048,14 @@ struct dmub_cmd_replay_copy_settings_data {
*/
uint8_t digbe_inst;
/**
+ * @hpo_stream_enc_inst: HPO stream encoder instance
+ */
+ uint8_t hpo_stream_enc_inst;
+ /**
+ * @hpo_link_enc_inst: HPO link encoder instance
+ */
+ uint8_t hpo_link_enc_inst;
+ /**
* AUX HW instance.
*/
uint8_t aux_inst;
@@ -3664,6 +4095,15 @@ struct dmub_cmd_replay_copy_settings_data {
* Use FSM state for Replay power up/down
*/
uint8_t use_phy_fsm;
+ /**
+ * Use for AUX-less ALPM LFPS wake operation
+ */
+ struct dmub_alpm_auxless_data auxless_alpm_data;
+
+ /**
+ * @pad: Align structure to 4 byte boundary.
+ */
+ uint8_t pad[2];
};
/**
@@ -3719,6 +4159,18 @@ struct dmub_rb_cmd_replay_enable_data {
* This does not support HDMI/DP2 for now.
*/
uint8_t phy_rate;
+ /**
+ * @hpo_stream_enc_inst: HPO stream encoder instance
+ */
+ uint8_t hpo_stream_enc_inst;
+ /**
+ * @hpo_link_enc_inst: HPO link encoder instance
+ */
+ uint8_t hpo_link_enc_inst;
+ /**
+ * @pad: Align structure to 4 byte boundary.
+ */
+ uint8_t pad[2];
};
/**
@@ -4212,6 +4664,42 @@ enum dmub_cmd_abm_type {
* Get the current ACE curve.
*/
DMUB_CMD__ABM_GET_ACE_CURVE = 10,
+
+ /**
+ * Get current histogram data
+ */
+ DMUB_CMD__ABM_GET_HISTOGRAM_DATA = 11,
+};
+
+/**
+ * LSDMA command sub-types.
+ */
+enum dmub_cmd_lsdma_type {
+ /**
+ * Initialize parameters for LSDMA.
+ * Ring buffer is mapped to the ring buffer
+ */
+ DMUB_CMD__LSDMA_INIT_CONFIG = 0,
+ /**
+ * LSDMA copies data from source to destination linearly
+ */
+ DMUB_CMD__LSDMA_LINEAR_COPY = 1,
+ /**
+ * Send the tiled-to-tiled copy command
+ */
+ DMUB_CMD__LSDMA_TILED_TO_TILED_COPY = 2,
+ /**
+ * Send the poll reg write command
+ */
+ DMUB_CMD__LSDMA_POLL_REG_WRITE = 3,
+ /**
+ * Send the pio copy command
+ */
+ DMUB_CMD__LSDMA_PIO_COPY = 4,
+ /**
+ * Send the pio constfill command
+ */
+ DMUB_CMD__LSDMA_PIO_CONSTFILL = 5,
};
struct abm_ace_curve {
@@ -4806,6 +5294,20 @@ enum dmub_abm_ace_curve_type {
};
/**
+ * enum dmub_abm_histogram_type - Histogram type.
+ */
+enum dmub_abm_histogram_type {
+ /**
+ * ACE curve as defined by the SW layer.
+ */
+ ABM_HISTOGRAM_TYPE__SW = 0,
+ /**
+ * ACE curve as defined by the SW to HW translation interface layer.
+ */
+ ABM_HISTOGRAM_TYPE__SW_IF = 1,
+};
+
+/**
* Definition of a DMUB_CMD__ABM_GET_ACE_CURVE command.
*/
struct dmub_rb_cmd_abm_get_ace_curve {
@@ -4841,6 +5343,41 @@ struct dmub_rb_cmd_abm_get_ace_curve {
};
/**
+ * Definition of a DMUB_CMD__ABM_GET_HISTOGRAM command.
+ */
+struct dmub_rb_cmd_abm_get_histogram {
+ /**
+ * Command header.
+ */
+ struct dmub_cmd_header header;
+
+ /**
+ * Address where Histogram should be copied.
+ */
+ union dmub_addr dest;
+
+ /**
+ * Type of Histogram being queried.
+ */
+ enum dmub_abm_histogram_type histogram_type;
+
+ /**
+ * Indirect buffer length.
+ */
+ uint16_t bytes;
+
+ /**
+ * eDP panel instance.
+ */
+ uint8_t panel_inst;
+
+ /**
+ * Explicit padding to 4 byte boundary.
+ */
+ uint8_t pad;
+};
+
+/**
* Definition of a DMUB_CMD__ABM_SAVE_RESTORE command.
*/
struct dmub_rb_cmd_abm_save_restore {
@@ -5204,6 +5741,64 @@ struct dmub_rb_cmd_get_usbc_cable_id {
} data;
};
+enum dmub_cmd_fused_io_sub_type {
+ DMUB_CMD__FUSED_IO_EXECUTE = 0,
+ DMUB_CMD__FUSED_IO_ABORT = 1,
+};
+
+enum dmub_cmd_fused_request_type {
+ FUSED_REQUEST_READ,
+ FUSED_REQUEST_WRITE,
+ FUSED_REQUEST_POLL,
+};
+
+enum dmub_cmd_fused_request_status {
+ FUSED_REQUEST_STATUS_SUCCESS,
+ FUSED_REQUEST_STATUS_BEGIN,
+ FUSED_REQUEST_STATUS_SUBMIT,
+ FUSED_REQUEST_STATUS_REPLY,
+ FUSED_REQUEST_STATUS_POLL,
+ FUSED_REQUEST_STATUS_ABORTED,
+ FUSED_REQUEST_STATUS_FAILED = 0x80,
+ FUSED_REQUEST_STATUS_INVALID,
+ FUSED_REQUEST_STATUS_BUSY,
+ FUSED_REQUEST_STATUS_TIMEOUT,
+ FUSED_REQUEST_STATUS_POLL_TIMEOUT,
+};
+
+struct dmub_cmd_fused_request {
+ uint8_t status;
+ uint8_t type : 2;
+ uint8_t _reserved0 : 3;
+ uint8_t poll_mask_msb : 3; // Number of MSB to zero out from last byte before comparing
+ uint8_t identifier;
+ uint8_t _reserved1;
+ uint32_t timeout_us;
+ union dmub_cmd_fused_request_location {
+ struct dmub_cmd_fused_request_location_i2c {
+ uint8_t is_aux : 1; // False
+ uint8_t ddc_line : 3;
+ uint8_t over_aux : 1;
+ uint8_t _reserved0 : 3;
+ uint8_t address;
+ uint8_t offset;
+ uint8_t length;
+ } i2c;
+ struct dmub_cmd_fused_request_location_aux {
+ uint32_t is_aux : 1; // True
+ uint32_t ddc_line : 3;
+ uint32_t address : 20;
+ uint32_t length : 8; // Automatically split into 16B transactions
+ } aux;
+ } u;
+ uint8_t buffer[0x30]; // Read: out, write: in, poll: expected
+};
+
+struct dmub_rb_cmd_fused_io {
+ struct dmub_cmd_header header;
+ struct dmub_cmd_fused_request request;
+};
+
/**
* Command type of a DMUB_CMD__SECURE_DISPLAY command
*/
@@ -5312,6 +5907,59 @@ struct dmub_rb_cmd_assr_enable {
};
/**
+ * Current definition of "ips_mode" from driver
+ */
+enum ips_residency_mode {
+ IPS_RESIDENCY__IPS1_MAX,
+ IPS_RESIDENCY__IPS2,
+ IPS_RESIDENCY__IPS1_RCG,
+ IPS_RESIDENCY__IPS1_ONO2_ON,
+};
+
+#define NUM_IPS_HISTOGRAM_BUCKETS 16
+
+/**
+ * IPS residency statistics to be sent to driver - subset of struct dmub_ips_residency_stats
+ */
+struct dmub_ips_residency_info {
+ uint32_t residency_millipercent;
+ uint32_t entry_counter;
+ uint32_t histogram[NUM_IPS_HISTOGRAM_BUCKETS];
+ uint64_t total_time_us;
+ uint64_t total_inactive_time_us;
+};
+
+/**
+ * Data passed from driver to FW in a DMUB_CMD__IPS_RESIDENCY_CNTL command.
+ */
+struct dmub_cmd_ips_residency_cntl_data {
+ uint8_t panel_inst;
+ uint8_t start_measurement;
+ uint8_t padding[2]; // align to 4-byte boundary
+};
+
+struct dmub_rb_cmd_ips_residency_cntl {
+ struct dmub_cmd_header header;
+ struct dmub_cmd_ips_residency_cntl_data cntl_data;
+};
+
+/**
+ * Data passed from FW to driver in a DMUB_CMD__IPS_QUERY_RESIDENCY_INFO command.
+ */
+struct dmub_cmd_ips_query_residency_info_data {
+ union dmub_addr dest;
+ uint32_t size;
+ uint32_t ips_mode;
+ uint8_t panel_inst;
+ uint8_t padding[3]; // align to 4-byte boundary
+};
+
+struct dmub_rb_cmd_ips_query_residency_info {
+ struct dmub_cmd_header header;
+ struct dmub_cmd_ips_query_residency_info_data info_data;
+};
+
+/**
* union dmub_rb_cmd - DMUB inbox command.
*/
union dmub_rb_cmd {
@@ -5482,6 +6130,11 @@ union dmub_rb_cmd {
struct dmub_rb_cmd_abm_get_ace_curve abm_get_ace_curve;
/**
+ * Definition of a DMUB_CMD__ABM_GET_HISTOGRAM command.
+ */
+ struct dmub_rb_cmd_abm_get_histogram abm_get_histogram;
+
+ /**
* Definition of a DMUB_CMD__ABM_SET_EVENT command.
*/
struct dmub_rb_cmd_abm_set_event abm_set_event;
@@ -5612,11 +6265,25 @@ union dmub_rb_cmd {
* Definition of a DMUB_CMD__PSP_ASSR_ENABLE command.
*/
struct dmub_rb_cmd_assr_enable assr_enable;
+
struct dmub_rb_cmd_fams2 fams2_config;
+ struct dmub_rb_cmd_ib ib_fams2_config;
+
struct dmub_rb_cmd_fams2_drr_update fams2_drr_update;
struct dmub_rb_cmd_fams2_flip fams2_flip;
+
+ struct dmub_rb_cmd_fused_io fused_io;
+
+ /**
+ * Definition of a DMUB_CMD__LSDMA command.
+ */
+ struct dmub_rb_cmd_lsdma lsdma;
+
+ struct dmub_rb_cmd_ips_residency_cntl ips_residency_cntl;
+
+ struct dmub_rb_cmd_ips_query_residency_info ips_query_residency_info;
};
/**
@@ -5647,6 +6314,7 @@ union dmub_rb_out_cmd {
* HPD sense notification command.
*/
struct dmub_rb_cmd_hpd_sense_notify hpd_sense_notify;
+ struct dmub_rb_cmd_fused_io fused_io;
};
#pragma pack(pop)
@@ -5694,6 +6362,45 @@ static inline bool dmub_rb_empty(struct dmub_rb *rb)
}
/**
+ * @brief gets number of outstanding requests in the RB
+ *
+ * @param rb DMUB Ringbuffer
+ * @return true if full
+ */
+static inline uint32_t dmub_rb_num_outstanding(struct dmub_rb *rb)
+{
+ uint32_t data_count;
+
+ if (rb->wrpt >= rb->rptr)
+ data_count = rb->wrpt - rb->rptr;
+ else
+ data_count = rb->capacity - (rb->rptr - rb->wrpt);
+
+ return data_count / DMUB_RB_CMD_SIZE;
+}
+
+/**
+ * @brief gets number of free buffers in the RB
+ *
+ * @param rb DMUB Ringbuffer
+ * @return true if full
+ */
+static inline uint32_t dmub_rb_num_free(struct dmub_rb *rb)
+{
+ uint32_t data_count;
+
+ if (rb->wrpt >= rb->rptr)
+ data_count = rb->wrpt - rb->rptr;
+ else
+ data_count = rb->capacity - (rb->rptr - rb->wrpt);
+
+ /* +1 because 1 entry is always unusable */
+ data_count += DMUB_RB_CMD_SIZE;
+
+ return (rb->capacity - data_count) / DMUB_RB_CMD_SIZE;
+}
+
+/**
* @brief Checks if the ringbuffer is full
*
* @param rb DMUB Ringbuffer
@@ -5709,6 +6416,7 @@ static inline bool dmub_rb_full(struct dmub_rb *rb)
else
data_count = rb->capacity - (rb->rptr - rb->wrpt);
+ /* -1 because 1 entry is always unusable */
return (data_count == (rb->capacity - DMUB_RB_CMD_SIZE));
}
diff --git a/drivers/gpu/drm/amd/display/dmub/src/Makefile b/drivers/gpu/drm/amd/display/dmub/src/Makefile
index a00b9e992292..468b768c11ae 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/Makefile
+++ b/drivers/gpu/drm/amd/display/dmub/src/Makefile
@@ -26,6 +26,7 @@ DMUB += dmub_dcn31.o dmub_dcn314.o dmub_dcn315.o dmub_dcn316.o
DMUB += dmub_dcn32.o
DMUB += dmub_dcn35.o
DMUB += dmub_dcn351.o
+DMUB += dmub_dcn36.o
DMUB += dmub_dcn401.o
AMD_DAL_DMUB = $(addprefix $(AMDDALPATH)/dmub/src/,$(DMUB))
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c
index e500ca9ae09c..73888c1bea93 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.c
@@ -414,63 +414,66 @@ uint32_t dmub_dcn20_get_current_time(struct dmub_srv *dmub)
return REG_READ(DMCUB_TIMER_CURRENT);
}
-void dmub_dcn20_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data)
+void dmub_dcn20_get_diagnostic_data(struct dmub_srv *dmub)
{
uint32_t is_dmub_enabled, is_soft_reset, is_sec_reset;
uint32_t is_traceport_enabled, is_cw0_enabled, is_cw6_enabled;
+ struct dmub_timeout_info timeout = {0};
- if (!dmub || !diag_data)
+ if (!dmub)
return;
- memset(diag_data, 0, sizeof(*diag_data));
-
- diag_data->dmcub_version = dmub->fw_version;
-
- diag_data->scratch[0] = REG_READ(DMCUB_SCRATCH0);
- diag_data->scratch[1] = REG_READ(DMCUB_SCRATCH1);
- diag_data->scratch[2] = REG_READ(DMCUB_SCRATCH2);
- diag_data->scratch[3] = REG_READ(DMCUB_SCRATCH3);
- diag_data->scratch[4] = REG_READ(DMCUB_SCRATCH4);
- diag_data->scratch[5] = REG_READ(DMCUB_SCRATCH5);
- diag_data->scratch[6] = REG_READ(DMCUB_SCRATCH6);
- diag_data->scratch[7] = REG_READ(DMCUB_SCRATCH7);
- diag_data->scratch[8] = REG_READ(DMCUB_SCRATCH8);
- diag_data->scratch[9] = REG_READ(DMCUB_SCRATCH9);
- diag_data->scratch[10] = REG_READ(DMCUB_SCRATCH10);
- diag_data->scratch[11] = REG_READ(DMCUB_SCRATCH11);
- diag_data->scratch[12] = REG_READ(DMCUB_SCRATCH12);
- diag_data->scratch[13] = REG_READ(DMCUB_SCRATCH13);
- diag_data->scratch[14] = REG_READ(DMCUB_SCRATCH14);
- diag_data->scratch[15] = REG_READ(DMCUB_SCRATCH15);
-
- diag_data->undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
- diag_data->inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
- diag_data->data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
-
- diag_data->inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
- diag_data->inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
- diag_data->inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
-
- diag_data->inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
- diag_data->inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
- diag_data->inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
+ /* timeout data filled externally, cache before resetting memory */
+ timeout = dmub->debug.timeout_info;
+ memset(&dmub->debug, 0, sizeof(dmub->debug));
+ dmub->debug.timeout_info = timeout;
+
+ dmub->debug.dmcub_version = dmub->fw_version;
+
+ dmub->debug.scratch[0] = REG_READ(DMCUB_SCRATCH0);
+ dmub->debug.scratch[1] = REG_READ(DMCUB_SCRATCH1);
+ dmub->debug.scratch[2] = REG_READ(DMCUB_SCRATCH2);
+ dmub->debug.scratch[3] = REG_READ(DMCUB_SCRATCH3);
+ dmub->debug.scratch[4] = REG_READ(DMCUB_SCRATCH4);
+ dmub->debug.scratch[5] = REG_READ(DMCUB_SCRATCH5);
+ dmub->debug.scratch[6] = REG_READ(DMCUB_SCRATCH6);
+ dmub->debug.scratch[7] = REG_READ(DMCUB_SCRATCH7);
+ dmub->debug.scratch[8] = REG_READ(DMCUB_SCRATCH8);
+ dmub->debug.scratch[9] = REG_READ(DMCUB_SCRATCH9);
+ dmub->debug.scratch[10] = REG_READ(DMCUB_SCRATCH10);
+ dmub->debug.scratch[11] = REG_READ(DMCUB_SCRATCH11);
+ dmub->debug.scratch[12] = REG_READ(DMCUB_SCRATCH12);
+ dmub->debug.scratch[13] = REG_READ(DMCUB_SCRATCH13);
+ dmub->debug.scratch[14] = REG_READ(DMCUB_SCRATCH14);
+ dmub->debug.scratch[15] = REG_READ(DMCUB_SCRATCH15);
+
+ dmub->debug.undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
+ dmub->debug.inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
+ dmub->debug.data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
+
+ dmub->debug.inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
+ dmub->debug.inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
+ dmub->debug.inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
+
+ dmub->debug.inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
+ dmub->debug.inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
+ dmub->debug.inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
REG_GET(DMCUB_CNTL, DMCUB_ENABLE, &is_dmub_enabled);
- diag_data->is_dmcub_enabled = is_dmub_enabled;
+ dmub->debug.is_dmcub_enabled = is_dmub_enabled;
REG_GET(DMCUB_CNTL, DMCUB_SOFT_RESET, &is_soft_reset);
- diag_data->is_dmcub_soft_reset = is_soft_reset;
+ dmub->debug.is_dmcub_soft_reset = is_soft_reset;
REG_GET(DMCUB_SEC_CNTL, DMCUB_SEC_RESET_STATUS, &is_sec_reset);
- diag_data->is_dmcub_secure_reset = is_sec_reset;
+ dmub->debug.is_dmcub_secure_reset = is_sec_reset;
REG_GET(DMCUB_CNTL, DMCUB_TRACEPORT_EN, &is_traceport_enabled);
- diag_data->is_traceport_en = is_traceport_enabled;
+ dmub->debug.is_traceport_en = is_traceport_enabled;
REG_GET(DMCUB_REGION3_CW0_TOP_ADDRESS, DMCUB_REGION3_CW0_ENABLE, &is_cw0_enabled);
- diag_data->is_cw0_enabled = is_cw0_enabled;
+ dmub->debug.is_cw0_enabled = is_cw0_enabled;
REG_GET(DMCUB_REGION3_CW6_TOP_ADDRESS, DMCUB_REGION3_CW6_ENABLE, &is_cw6_enabled);
- diag_data->is_cw6_enabled = is_cw6_enabled;
- diag_data->timeout_info = dmub->debug;
+ dmub->debug.is_cw6_enabled = is_cw6_enabled;
}
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.h b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.h
index de287b101848..42c1fb4bc73f 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.h
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn20.h
@@ -247,6 +247,6 @@ bool dmub_dcn20_use_cached_trace_buffer(struct dmub_srv *dmub);
uint32_t dmub_dcn20_get_current_time(struct dmub_srv *dmub);
-void dmub_dcn20_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *dmub_oca);
+void dmub_dcn20_get_diagnostic_data(struct dmub_srv *dmub);
#endif /* _DMUB_DCN20_H_ */
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c
index d9f31b191c69..3f38db752b84 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.c
@@ -83,8 +83,8 @@ static inline void dmub_dcn31_translate_addr(const union dmub_addr *addr_in,
void dmub_dcn31_reset(struct dmub_srv *dmub)
{
union dmub_gpint_data_register cmd;
- const uint32_t timeout = 100;
- uint32_t in_reset, scratch, i, pwait_mode;
+ const uint32_t timeout = 100000;
+ uint32_t in_reset, is_enabled, scratch, i, pwait_mode;
REG_GET(DMCUB_CNTL2, DMCUB_SOFT_RESET, &in_reset);
@@ -108,7 +108,7 @@ void dmub_dcn31_reset(struct dmub_srv *dmub)
}
for (i = 0; i < timeout; ++i) {
- scratch = dmub->hw_funcs.get_gpint_response(dmub);
+ scratch = REG_READ(DMCUB_SCRATCH7);
if (scratch == DMUB_GPINT__STOP_FW_RESPONSE)
break;
@@ -125,9 +125,14 @@ void dmub_dcn31_reset(struct dmub_srv *dmub)
/* Force reset in case we timed out, DMCUB is likely hung. */
}
- REG_UPDATE(DMCUB_CNTL2, DMCUB_SOFT_RESET, 1);
- REG_UPDATE(DMCUB_CNTL, DMCUB_ENABLE, 0);
- REG_UPDATE(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET, 1);
+ REG_GET(DMCUB_CNTL, DMCUB_ENABLE, &is_enabled);
+
+ if (is_enabled) {
+ REG_UPDATE(DMCUB_CNTL2, DMCUB_SOFT_RESET, 1);
+ REG_UPDATE(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET, 1);
+ REG_UPDATE(DMCUB_CNTL, DMCUB_ENABLE, 0);
+ }
+
REG_WRITE(DMCUB_INBOX1_RPTR, 0);
REG_WRITE(DMCUB_INBOX1_WPTR, 0);
REG_WRITE(DMCUB_OUTBOX1_RPTR, 0);
@@ -371,6 +376,7 @@ void dmub_dcn31_enable_dmub_boot_options(struct dmub_srv *dmub, const struct dmu
boot_options.bits.usb4_cm_version = params->usb4_cm_version;
boot_options.bits.dpia_hpd_int_enable_supported = params->dpia_hpd_int_enable_supported;
boot_options.bits.power_optimization = params->power_optimization;
+ boot_options.bits.lower_hbr3_phy_ssc = params->lower_hbr3_phy_ssc;
boot_options.bits.sel_mux_phy_c_d_phy_f_g = (dmub->asic == DMUB_ASIC_DCN31B) ? 1 : 0;
@@ -408,69 +414,75 @@ uint32_t dmub_dcn31_get_current_time(struct dmub_srv *dmub)
return REG_READ(DMCUB_TIMER_CURRENT);
}
-void dmub_dcn31_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data)
+void dmub_dcn31_get_diagnostic_data(struct dmub_srv *dmub)
{
- uint32_t is_dmub_enabled, is_soft_reset, is_sec_reset;
+ uint32_t is_dmub_enabled, is_soft_reset, is_sec_reset, is_pwait;
uint32_t is_traceport_enabled, is_cw0_enabled, is_cw6_enabled;
+ struct dmub_timeout_info timeout = {0};
- if (!dmub || !diag_data)
+ if (!dmub)
return;
- memset(diag_data, 0, sizeof(*diag_data));
-
- diag_data->dmcub_version = dmub->fw_version;
-
- diag_data->scratch[0] = REG_READ(DMCUB_SCRATCH0);
- diag_data->scratch[1] = REG_READ(DMCUB_SCRATCH1);
- diag_data->scratch[2] = REG_READ(DMCUB_SCRATCH2);
- diag_data->scratch[3] = REG_READ(DMCUB_SCRATCH3);
- diag_data->scratch[4] = REG_READ(DMCUB_SCRATCH4);
- diag_data->scratch[5] = REG_READ(DMCUB_SCRATCH5);
- diag_data->scratch[6] = REG_READ(DMCUB_SCRATCH6);
- diag_data->scratch[7] = REG_READ(DMCUB_SCRATCH7);
- diag_data->scratch[8] = REG_READ(DMCUB_SCRATCH8);
- diag_data->scratch[9] = REG_READ(DMCUB_SCRATCH9);
- diag_data->scratch[10] = REG_READ(DMCUB_SCRATCH10);
- diag_data->scratch[11] = REG_READ(DMCUB_SCRATCH11);
- diag_data->scratch[12] = REG_READ(DMCUB_SCRATCH12);
- diag_data->scratch[13] = REG_READ(DMCUB_SCRATCH13);
- diag_data->scratch[14] = REG_READ(DMCUB_SCRATCH14);
- diag_data->scratch[15] = REG_READ(DMCUB_SCRATCH15);
-
- diag_data->undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
- diag_data->inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
- diag_data->data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
-
- diag_data->inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
- diag_data->inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
- diag_data->inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
-
- diag_data->inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
- diag_data->inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
- diag_data->inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
-
- diag_data->outbox1_rptr = REG_READ(DMCUB_OUTBOX1_RPTR);
- diag_data->outbox1_wptr = REG_READ(DMCUB_OUTBOX1_WPTR);
- diag_data->outbox1_size = REG_READ(DMCUB_OUTBOX1_SIZE);
+ /* timeout data filled externally, cache before resetting memory */
+ timeout = dmub->debug.timeout_info;
+ memset(&dmub->debug, 0, sizeof(dmub->debug));
+ dmub->debug.timeout_info = timeout;
+
+ dmub->debug.dmcub_version = dmub->fw_version;
+
+ dmub->debug.scratch[0] = REG_READ(DMCUB_SCRATCH0);
+ dmub->debug.scratch[1] = REG_READ(DMCUB_SCRATCH1);
+ dmub->debug.scratch[2] = REG_READ(DMCUB_SCRATCH2);
+ dmub->debug.scratch[3] = REG_READ(DMCUB_SCRATCH3);
+ dmub->debug.scratch[4] = REG_READ(DMCUB_SCRATCH4);
+ dmub->debug.scratch[5] = REG_READ(DMCUB_SCRATCH5);
+ dmub->debug.scratch[6] = REG_READ(DMCUB_SCRATCH6);
+ dmub->debug.scratch[7] = REG_READ(DMCUB_SCRATCH7);
+ dmub->debug.scratch[8] = REG_READ(DMCUB_SCRATCH8);
+ dmub->debug.scratch[9] = REG_READ(DMCUB_SCRATCH9);
+ dmub->debug.scratch[10] = REG_READ(DMCUB_SCRATCH10);
+ dmub->debug.scratch[11] = REG_READ(DMCUB_SCRATCH11);
+ dmub->debug.scratch[12] = REG_READ(DMCUB_SCRATCH12);
+ dmub->debug.scratch[13] = REG_READ(DMCUB_SCRATCH13);
+ dmub->debug.scratch[14] = REG_READ(DMCUB_SCRATCH14);
+ dmub->debug.scratch[15] = REG_READ(DMCUB_SCRATCH15);
+
+ dmub->debug.undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
+ dmub->debug.inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
+ dmub->debug.data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
+
+ dmub->debug.inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
+ dmub->debug.inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
+ dmub->debug.inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
+
+ dmub->debug.inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
+ dmub->debug.inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
+ dmub->debug.inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
+
+ dmub->debug.outbox1_rptr = REG_READ(DMCUB_OUTBOX1_RPTR);
+ dmub->debug.outbox1_wptr = REG_READ(DMCUB_OUTBOX1_WPTR);
+ dmub->debug.outbox1_size = REG_READ(DMCUB_OUTBOX1_SIZE);
REG_GET(DMCUB_CNTL, DMCUB_ENABLE, &is_dmub_enabled);
- diag_data->is_dmcub_enabled = is_dmub_enabled;
+ dmub->debug.is_dmcub_enabled = is_dmub_enabled;
+
+ REG_GET(DMCUB_CNTL, DMCUB_PWAIT_MODE_STATUS, &is_pwait);
+ dmub->debug.is_pwait = is_pwait;
REG_GET(DMCUB_CNTL2, DMCUB_SOFT_RESET, &is_soft_reset);
- diag_data->is_dmcub_soft_reset = is_soft_reset;
+ dmub->debug.is_dmcub_soft_reset = is_soft_reset;
REG_GET(DMCUB_SEC_CNTL, DMCUB_SEC_RESET_STATUS, &is_sec_reset);
- diag_data->is_dmcub_secure_reset = is_sec_reset;
+ dmub->debug.is_dmcub_secure_reset = is_sec_reset;
REG_GET(DMCUB_CNTL, DMCUB_TRACEPORT_EN, &is_traceport_enabled);
- diag_data->is_traceport_en = is_traceport_enabled;
+ dmub->debug.is_traceport_en = is_traceport_enabled;
REG_GET(DMCUB_REGION3_CW0_TOP_ADDRESS, DMCUB_REGION3_CW0_ENABLE, &is_cw0_enabled);
- diag_data->is_cw0_enabled = is_cw0_enabled;
+ dmub->debug.is_cw0_enabled = is_cw0_enabled;
REG_GET(DMCUB_REGION3_CW6_TOP_ADDRESS, DMCUB_REGION3_CW6_ENABLE, &is_cw6_enabled);
- diag_data->is_cw6_enabled = is_cw6_enabled;
- diag_data->timeout_info = dmub->debug;
+ dmub->debug.is_cw6_enabled = is_cw6_enabled;
}
bool dmub_dcn31_should_detect(struct dmub_srv *dmub)
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.h b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.h
index eccdab4986ce..1c43ef2bca66 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.h
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn31.h
@@ -251,7 +251,7 @@ void dmub_dcn31_set_outbox0_rptr(struct dmub_srv *dmub, uint32_t rptr_offset);
uint32_t dmub_dcn31_get_current_time(struct dmub_srv *dmub);
-void dmub_dcn31_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data);
+void dmub_dcn31_get_diagnostic_data(struct dmub_srv *dmub);
bool dmub_dcn31_should_detect(struct dmub_srv *dmub);
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c
index 9600b7f858b0..e7056205b050 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.c
@@ -417,73 +417,75 @@ uint32_t dmub_dcn32_get_current_time(struct dmub_srv *dmub)
return REG_READ(DMCUB_TIMER_CURRENT);
}
-void dmub_dcn32_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data)
+void dmub_dcn32_get_diagnostic_data(struct dmub_srv *dmub)
{
uint32_t is_dmub_enabled, is_soft_reset, is_sec_reset;
uint32_t is_traceport_enabled, is_cw0_enabled, is_cw6_enabled;
+ struct dmub_timeout_info timeout = {0};
- if (!dmub || !diag_data)
+ if (!dmub)
return;
- memset(diag_data, 0, sizeof(*diag_data));
-
- diag_data->dmcub_version = dmub->fw_version;
-
- diag_data->scratch[0] = REG_READ(DMCUB_SCRATCH0);
- diag_data->scratch[1] = REG_READ(DMCUB_SCRATCH1);
- diag_data->scratch[2] = REG_READ(DMCUB_SCRATCH2);
- diag_data->scratch[3] = REG_READ(DMCUB_SCRATCH3);
- diag_data->scratch[4] = REG_READ(DMCUB_SCRATCH4);
- diag_data->scratch[5] = REG_READ(DMCUB_SCRATCH5);
- diag_data->scratch[6] = REG_READ(DMCUB_SCRATCH6);
- diag_data->scratch[7] = REG_READ(DMCUB_SCRATCH7);
- diag_data->scratch[8] = REG_READ(DMCUB_SCRATCH8);
- diag_data->scratch[9] = REG_READ(DMCUB_SCRATCH9);
- diag_data->scratch[10] = REG_READ(DMCUB_SCRATCH10);
- diag_data->scratch[11] = REG_READ(DMCUB_SCRATCH11);
- diag_data->scratch[12] = REG_READ(DMCUB_SCRATCH12);
- diag_data->scratch[13] = REG_READ(DMCUB_SCRATCH13);
- diag_data->scratch[14] = REG_READ(DMCUB_SCRATCH14);
- diag_data->scratch[15] = REG_READ(DMCUB_SCRATCH15);
- diag_data->scratch[16] = REG_READ(DMCUB_SCRATCH16);
-
- diag_data->undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
- diag_data->inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
- diag_data->data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
-
- diag_data->inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
- diag_data->inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
- diag_data->inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
-
- diag_data->inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
- diag_data->inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
- diag_data->inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
-
- diag_data->outbox1_rptr = REG_READ(DMCUB_OUTBOX1_RPTR);
- diag_data->outbox1_wptr = REG_READ(DMCUB_OUTBOX1_WPTR);
- diag_data->outbox1_size = REG_READ(DMCUB_OUTBOX1_SIZE);
+ /* timeout data filled externally, cache before resetting memory */
+ timeout = dmub->debug.timeout_info;
+ memset(&dmub->debug, 0, sizeof(dmub->debug));
+ dmub->debug.timeout_info = timeout;
+
+ dmub->debug.dmcub_version = dmub->fw_version;
+
+ dmub->debug.scratch[0] = REG_READ(DMCUB_SCRATCH0);
+ dmub->debug.scratch[1] = REG_READ(DMCUB_SCRATCH1);
+ dmub->debug.scratch[2] = REG_READ(DMCUB_SCRATCH2);
+ dmub->debug.scratch[3] = REG_READ(DMCUB_SCRATCH3);
+ dmub->debug.scratch[4] = REG_READ(DMCUB_SCRATCH4);
+ dmub->debug.scratch[5] = REG_READ(DMCUB_SCRATCH5);
+ dmub->debug.scratch[6] = REG_READ(DMCUB_SCRATCH6);
+ dmub->debug.scratch[7] = REG_READ(DMCUB_SCRATCH7);
+ dmub->debug.scratch[8] = REG_READ(DMCUB_SCRATCH8);
+ dmub->debug.scratch[9] = REG_READ(DMCUB_SCRATCH9);
+ dmub->debug.scratch[10] = REG_READ(DMCUB_SCRATCH10);
+ dmub->debug.scratch[11] = REG_READ(DMCUB_SCRATCH11);
+ dmub->debug.scratch[12] = REG_READ(DMCUB_SCRATCH12);
+ dmub->debug.scratch[13] = REG_READ(DMCUB_SCRATCH13);
+ dmub->debug.scratch[14] = REG_READ(DMCUB_SCRATCH14);
+ dmub->debug.scratch[15] = REG_READ(DMCUB_SCRATCH15);
+ dmub->debug.scratch[16] = REG_READ(DMCUB_SCRATCH16);
+
+ dmub->debug.undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
+ dmub->debug.inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
+ dmub->debug.data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
+
+ dmub->debug.inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
+ dmub->debug.inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
+ dmub->debug.inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
+
+ dmub->debug.inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
+ dmub->debug.inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
+ dmub->debug.inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
+
+ dmub->debug.outbox1_rptr = REG_READ(DMCUB_OUTBOX1_RPTR);
+ dmub->debug.outbox1_wptr = REG_READ(DMCUB_OUTBOX1_WPTR);
+ dmub->debug.outbox1_size = REG_READ(DMCUB_OUTBOX1_SIZE);
REG_GET(DMCUB_CNTL, DMCUB_ENABLE, &is_dmub_enabled);
- diag_data->is_dmcub_enabled = is_dmub_enabled;
+ dmub->debug.is_dmcub_enabled = is_dmub_enabled;
REG_GET(DMCUB_CNTL2, DMCUB_SOFT_RESET, &is_soft_reset);
- diag_data->is_dmcub_soft_reset = is_soft_reset;
+ dmub->debug.is_dmcub_soft_reset = is_soft_reset;
REG_GET(DMCUB_SEC_CNTL, DMCUB_SEC_RESET_STATUS, &is_sec_reset);
- diag_data->is_dmcub_secure_reset = is_sec_reset;
+ dmub->debug.is_dmcub_secure_reset = is_sec_reset;
REG_GET(DMCUB_CNTL, DMCUB_TRACEPORT_EN, &is_traceport_enabled);
- diag_data->is_traceport_en = is_traceport_enabled;
+ dmub->debug.is_traceport_en = is_traceport_enabled;
REG_GET(DMCUB_REGION3_CW0_TOP_ADDRESS, DMCUB_REGION3_CW0_ENABLE, &is_cw0_enabled);
- diag_data->is_cw0_enabled = is_cw0_enabled;
+ dmub->debug.is_cw0_enabled = is_cw0_enabled;
REG_GET(DMCUB_REGION3_CW6_TOP_ADDRESS, DMCUB_REGION3_CW6_ENABLE, &is_cw6_enabled);
- diag_data->is_cw6_enabled = is_cw6_enabled;
+ dmub->debug.is_cw6_enabled = is_cw6_enabled;
- diag_data->gpint_datain0 = REG_READ(DMCUB_GPINT_DATAIN0);
-
- diag_data->timeout_info = dmub->debug;
+ dmub->debug.gpint_datain0 = REG_READ(DMCUB_GPINT_DATAIN0);
}
void dmub_dcn32_configure_dmub_in_system_memory(struct dmub_srv *dmub)
{
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.h b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.h
index 29c1132951af..1a229450c53d 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.h
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn32.h
@@ -254,7 +254,7 @@ void dmub_dcn32_set_outbox0_rptr(struct dmub_srv *dmub, uint32_t rptr_offset);
uint32_t dmub_dcn32_get_current_time(struct dmub_srv *dmub);
-void dmub_dcn32_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data);
+void dmub_dcn32_get_diagnostic_data(struct dmub_srv *dmub);
void dmub_dcn32_configure_dmub_in_system_memory(struct dmub_srv *dmub);
void dmub_dcn32_send_inbox0_cmd(struct dmub_srv *dmub, union dmub_inbox0_data_register data);
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c
index e5e77bd3c31e..2228d62adc7e 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.c
@@ -88,23 +88,19 @@ static inline void dmub_dcn35_translate_addr(const union dmub_addr *addr_in,
void dmub_dcn35_reset(struct dmub_srv *dmub)
{
union dmub_gpint_data_register cmd;
- const uint32_t timeout = 100;
+ const uint32_t timeout = 100000;
uint32_t in_reset, is_enabled, scratch, i, pwait_mode;
REG_GET(DMCUB_CNTL2, DMCUB_SOFT_RESET, &in_reset);
+ REG_GET(DMCUB_CNTL, DMCUB_ENABLE, &is_enabled);
- if (in_reset == 0) {
+ if (in_reset == 0 && is_enabled != 0) {
cmd.bits.status = 1;
cmd.bits.command_code = DMUB_GPINT__STOP_FW;
cmd.bits.param = 0;
dmub->hw_funcs.set_gpint(dmub, cmd);
- /**
- * Timeout covers both the ACK and the wait
- * for remaining work to finish.
- */
-
for (i = 0; i < timeout; ++i) {
if (dmub->hw_funcs.is_gpint_acked(dmub, cmd))
break;
@@ -113,7 +109,7 @@ void dmub_dcn35_reset(struct dmub_srv *dmub)
}
for (i = 0; i < timeout; ++i) {
- scratch = dmub->hw_funcs.get_gpint_response(dmub);
+ scratch = REG_READ(DMCUB_SCRATCH7);
if (scratch == DMUB_GPINT__STOP_FW_RESPONSE)
break;
@@ -130,11 +126,9 @@ void dmub_dcn35_reset(struct dmub_srv *dmub)
/* Force reset in case we timed out, DMCUB is likely hung. */
}
- REG_GET(DMCUB_CNTL, DMCUB_ENABLE, &is_enabled);
-
if (is_enabled) {
REG_UPDATE(DMCUB_CNTL2, DMCUB_SOFT_RESET, 1);
- REG_UPDATE(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET, 1);
+ udelay(1);
REG_UPDATE(DMCUB_CNTL, DMCUB_ENABLE, 0);
}
@@ -160,11 +154,7 @@ void dmub_dcn35_reset_release(struct dmub_srv *dmub)
LONO_SOCCLK_GATE_DISABLE, 1,
LONO_DMCUBCLK_GATE_DISABLE, 1);
- REG_UPDATE(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET, 1);
- udelay(1);
REG_UPDATE_2(DMCUB_CNTL, DMCUB_ENABLE, 1, DMCUB_TRACEPORT_EN, 1);
- REG_UPDATE(DMCUB_CNTL2, DMCUB_SOFT_RESET, 1);
- udelay(1);
REG_UPDATE(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET, 0);
REG_UPDATE(DMCUB_CNTL2, DMCUB_SOFT_RESET, 0);
}
@@ -462,66 +452,72 @@ uint32_t dmub_dcn35_get_current_time(struct dmub_srv *dmub)
return REG_READ(DMCUB_TIMER_CURRENT);
}
-void dmub_dcn35_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data)
+void dmub_dcn35_get_diagnostic_data(struct dmub_srv *dmub)
{
- uint32_t is_dmub_enabled, is_soft_reset;
+ uint32_t is_dmub_enabled, is_soft_reset, is_pwait;
uint32_t is_traceport_enabled, is_cw6_enabled;
+ struct dmub_timeout_info timeout = {0};
- if (!dmub || !diag_data)
+ if (!dmub)
return;
- memset(diag_data, 0, sizeof(*diag_data));
-
- diag_data->dmcub_version = dmub->fw_version;
-
- diag_data->scratch[0] = REG_READ(DMCUB_SCRATCH0);
- diag_data->scratch[1] = REG_READ(DMCUB_SCRATCH1);
- diag_data->scratch[2] = REG_READ(DMCUB_SCRATCH2);
- diag_data->scratch[3] = REG_READ(DMCUB_SCRATCH3);
- diag_data->scratch[4] = REG_READ(DMCUB_SCRATCH4);
- diag_data->scratch[5] = REG_READ(DMCUB_SCRATCH5);
- diag_data->scratch[6] = REG_READ(DMCUB_SCRATCH6);
- diag_data->scratch[7] = REG_READ(DMCUB_SCRATCH7);
- diag_data->scratch[8] = REG_READ(DMCUB_SCRATCH8);
- diag_data->scratch[9] = REG_READ(DMCUB_SCRATCH9);
- diag_data->scratch[10] = REG_READ(DMCUB_SCRATCH10);
- diag_data->scratch[11] = REG_READ(DMCUB_SCRATCH11);
- diag_data->scratch[12] = REG_READ(DMCUB_SCRATCH12);
- diag_data->scratch[13] = REG_READ(DMCUB_SCRATCH13);
- diag_data->scratch[14] = REG_READ(DMCUB_SCRATCH14);
- diag_data->scratch[15] = REG_READ(DMCUB_SCRATCH15);
- diag_data->scratch[16] = REG_READ(DMCUB_SCRATCH16);
-
- diag_data->undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
- diag_data->inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
- diag_data->data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
-
- diag_data->inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
- diag_data->inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
- diag_data->inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
-
- diag_data->inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
- diag_data->inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
- diag_data->inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
-
- diag_data->outbox1_rptr = REG_READ(DMCUB_OUTBOX1_RPTR);
- diag_data->outbox1_wptr = REG_READ(DMCUB_OUTBOX1_WPTR);
- diag_data->outbox1_size = REG_READ(DMCUB_OUTBOX1_SIZE);
+ /* timeout data filled externally, cache before resetting memory */
+ timeout = dmub->debug.timeout_info;
+ memset(&dmub->debug, 0, sizeof(dmub->debug));
+ dmub->debug.timeout_info = timeout;
+
+ dmub->debug.dmcub_version = dmub->fw_version;
+
+ dmub->debug.scratch[0] = REG_READ(DMCUB_SCRATCH0);
+ dmub->debug.scratch[1] = REG_READ(DMCUB_SCRATCH1);
+ dmub->debug.scratch[2] = REG_READ(DMCUB_SCRATCH2);
+ dmub->debug.scratch[3] = REG_READ(DMCUB_SCRATCH3);
+ dmub->debug.scratch[4] = REG_READ(DMCUB_SCRATCH4);
+ dmub->debug.scratch[5] = REG_READ(DMCUB_SCRATCH5);
+ dmub->debug.scratch[6] = REG_READ(DMCUB_SCRATCH6);
+ dmub->debug.scratch[7] = REG_READ(DMCUB_SCRATCH7);
+ dmub->debug.scratch[8] = REG_READ(DMCUB_SCRATCH8);
+ dmub->debug.scratch[9] = REG_READ(DMCUB_SCRATCH9);
+ dmub->debug.scratch[10] = REG_READ(DMCUB_SCRATCH10);
+ dmub->debug.scratch[11] = REG_READ(DMCUB_SCRATCH11);
+ dmub->debug.scratch[12] = REG_READ(DMCUB_SCRATCH12);
+ dmub->debug.scratch[13] = REG_READ(DMCUB_SCRATCH13);
+ dmub->debug.scratch[14] = REG_READ(DMCUB_SCRATCH14);
+ dmub->debug.scratch[15] = REG_READ(DMCUB_SCRATCH15);
+ dmub->debug.scratch[16] = REG_READ(DMCUB_SCRATCH16);
+
+ dmub->debug.undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
+ dmub->debug.inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
+ dmub->debug.data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
+
+ dmub->debug.inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
+ dmub->debug.inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
+ dmub->debug.inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
+
+ dmub->debug.inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
+ dmub->debug.inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
+ dmub->debug.inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
+
+ dmub->debug.outbox1_rptr = REG_READ(DMCUB_OUTBOX1_RPTR);
+ dmub->debug.outbox1_wptr = REG_READ(DMCUB_OUTBOX1_WPTR);
+ dmub->debug.outbox1_size = REG_READ(DMCUB_OUTBOX1_SIZE);
REG_GET(DMCUB_CNTL, DMCUB_ENABLE, &is_dmub_enabled);
- diag_data->is_dmcub_enabled = is_dmub_enabled;
+ dmub->debug.is_dmcub_enabled = is_dmub_enabled;
+
+ REG_GET(DMCUB_CNTL, DMCUB_PWAIT_MODE_STATUS, &is_pwait);
+ dmub->debug.is_pwait = is_pwait;
REG_GET(DMCUB_CNTL2, DMCUB_SOFT_RESET, &is_soft_reset);
- diag_data->is_dmcub_soft_reset = is_soft_reset;
+ dmub->debug.is_dmcub_soft_reset = is_soft_reset;
REG_GET(DMCUB_CNTL, DMCUB_TRACEPORT_EN, &is_traceport_enabled);
- diag_data->is_traceport_en = is_traceport_enabled;
+ dmub->debug.is_traceport_en = is_traceport_enabled;
REG_GET(DMCUB_REGION3_CW6_TOP_ADDRESS, DMCUB_REGION3_CW6_ENABLE, &is_cw6_enabled);
- diag_data->is_cw6_enabled = is_cw6_enabled;
+ dmub->debug.is_cw6_enabled = is_cw6_enabled;
- diag_data->gpint_datain0 = REG_READ(DMCUB_GPINT_DATAIN0);
- diag_data->timeout_info = dmub->debug;
+ dmub->debug.gpint_datain0 = REG_READ(DMCUB_GPINT_DATAIN0);
}
void dmub_dcn35_configure_dmub_in_system_memory(struct dmub_srv *dmub)
{
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.h b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.h
index 686e97c00ccc..39fcb7275da5 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.h
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn35.h
@@ -269,7 +269,7 @@ void dmub_dcn35_set_outbox0_rptr(struct dmub_srv *dmub, uint32_t rptr_offset);
uint32_t dmub_dcn35_get_current_time(struct dmub_srv *dmub);
-void dmub_dcn35_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data);
+void dmub_dcn35_get_diagnostic_data(struct dmub_srv *dmub);
void dmub_dcn35_configure_dmub_in_system_memory(struct dmub_srv *dmub);
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn36.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn36.c
new file mode 100644
index 000000000000..b1ce09d48920
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn36.c
@@ -0,0 +1,34 @@
+/* SPDX-License-Identifier: MIT */
+/* Copyright 2025 Advanced Micro Devices, Inc. */
+
+#include "../dmub_srv.h"
+#include "dmub_reg.h"
+#include "dmub_dcn36.h"
+
+#include "dcn/dcn_3_6_0_offset.h"
+#include "dcn/dcn_3_6_0_sh_mask.h"
+
+#define BASE_INNER(seg) ctx->dcn_reg_offsets[seg]
+#define CTX dmub
+#define REGS dmub->regs_dcn35
+#define REG_OFFSET_EXP(reg_name) BASE(reg##reg_name##_BASE_IDX) + reg##reg_name
+
+void dmub_srv_dcn36_regs_init(struct dmub_srv *dmub, struct dc_context *ctx)
+{
+ struct dmub_srv_dcn35_regs *regs = dmub->regs_dcn35;
+#define REG_STRUCT regs
+
+#define DMUB_SR(reg) REG_STRUCT->offset.reg = REG_OFFSET_EXP(reg);
+ DMUB_DCN35_REGS()
+ DMCUB_INTERNAL_REGS()
+#undef DMUB_SR
+
+#define DMUB_SF(reg, field) REG_STRUCT->mask.reg##__##field = FD_MASK(reg, field);
+ DMUB_DCN35_FIELDS()
+#undef DMUB_SF
+
+#define DMUB_SF(reg, field) REG_STRUCT->shift.reg##__##field = FD_SHIFT(reg, field);
+ DMUB_DCN35_FIELDS()
+#undef DMUB_SF
+#undef REG_STRUCT
+}
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn36.h b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn36.h
new file mode 100644
index 000000000000..57850550f682
--- /dev/null
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn36.h
@@ -0,0 +1,13 @@
+/* SPDX-License-Identifier: MIT */
+/* Copyright 2025 Advanced Micro Devices, Inc. */
+
+#ifndef _DMUB_DCN36_H_
+#define _DMUB_DCN36_H_
+
+#include "dmub_dcn35.h"
+
+struct dmub_srv;
+
+void dmub_srv_dcn36_regs_init(struct dmub_srv *dmub, struct dc_context *ctx);
+
+#endif /* _DMUB_DCN36_H_ */
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.c
index 39a8cb6d7523..b31adbd0d685 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.c
@@ -63,44 +63,45 @@ static inline void dmub_dcn401_translate_addr(const union dmub_addr *addr_in,
void dmub_dcn401_reset(struct dmub_srv *dmub)
{
union dmub_gpint_data_register cmd;
- const uint32_t timeout = 30;
- uint32_t in_reset, scratch, i;
+ const uint32_t timeout_us = 1 * 1000 * 1000; //1s
+ const uint32_t poll_delay_us = 1; //1us
+ uint32_t i = 0;
+ uint32_t enabled, in_reset, scratch, pwait_mode;
- REG_GET(DMCUB_CNTL2, DMCUB_SOFT_RESET, &in_reset);
+ REG_GET(DMCUB_CNTL,
+ DMCUB_ENABLE, &enabled);
+ REG_GET(DMCUB_CNTL2,
+ DMCUB_SOFT_RESET, &in_reset);
- if (in_reset == 0) {
+ if (enabled && in_reset == 0) {
cmd.bits.status = 1;
cmd.bits.command_code = DMUB_GPINT__STOP_FW;
cmd.bits.param = 0;
dmub->hw_funcs.set_gpint(dmub, cmd);
- /**
- * Timeout covers both the ACK and the wait
- * for remaining work to finish.
- *
- * This is mostly bound by the PHY disable sequence.
- * Each register check will be greater than 1us, so
- * don't bother using udelay.
- */
-
- for (i = 0; i < timeout; ++i) {
- if (dmub->hw_funcs.is_gpint_acked(dmub, cmd))
+ for (; i < timeout_us; i++) {
+ scratch = dmub->hw_funcs.get_gpint_response(dmub);
+ if (scratch == DMUB_GPINT__STOP_FW_RESPONSE)
break;
+
+ udelay(poll_delay_us);
}
- for (i = 0; i < timeout; ++i) {
- scratch = dmub->hw_funcs.get_gpint_response(dmub);
- if (scratch == DMUB_GPINT__STOP_FW_RESPONSE)
+ for (; i < timeout_us; i++) {
+ REG_GET(DMCUB_CNTL, DMCUB_PWAIT_MODE_STATUS, &pwait_mode);
+ if (pwait_mode & (1 << 0))
break;
+
+ udelay(poll_delay_us);
}
+ }
- /* Force reset in case we timed out, DMCUB is likely hung. */
+ if (i >= timeout_us) {
+ /* timeout should never occur */
+ BREAK_TO_DEBUGGER();
}
- REG_UPDATE(DMCUB_CNTL2, DMCUB_SOFT_RESET, 1);
- REG_UPDATE(DMCUB_CNTL, DMCUB_ENABLE, 0);
- REG_UPDATE(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET, 1);
REG_WRITE(DMCUB_INBOX1_RPTR, 0);
REG_WRITE(DMCUB_INBOX1_WPTR, 0);
REG_WRITE(DMCUB_OUTBOX1_RPTR, 0);
@@ -131,7 +132,10 @@ void dmub_dcn401_backdoor_load(struct dmub_srv *dmub,
dmub_dcn401_get_fb_base_offset(dmub, &fb_base, &fb_offset);
+ /* reset and disable DMCUB and MMHUBBUB DMUIF */
REG_UPDATE(DMCUB_SEC_CNTL, DMCUB_SEC_RESET, 1);
+ REG_UPDATE(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET, 1);
+ REG_UPDATE(DMCUB_CNTL, DMCUB_ENABLE, 0);
dmub_dcn401_translate_addr(&cw0->offset, fb_base, fb_offset, &offset);
@@ -151,6 +155,7 @@ void dmub_dcn401_backdoor_load(struct dmub_srv *dmub,
DMCUB_REGION3_CW1_TOP_ADDRESS, cw1->region.top,
DMCUB_REGION3_CW1_ENABLE, 1);
+ /* release DMCUB reset only to prevent premature execution */
REG_UPDATE_2(DMCUB_SEC_CNTL, DMCUB_SEC_RESET, 0, DMCUB_MEM_UNIT_ID,
0x20);
}
@@ -161,7 +166,10 @@ void dmub_dcn401_backdoor_load_zfb_mode(struct dmub_srv *dmub,
{
union dmub_addr offset;
+ /* reset and disable DMCUB and MMHUBBUB DMUIF */
REG_UPDATE(DMCUB_SEC_CNTL, DMCUB_SEC_RESET, 1);
+ REG_UPDATE(MMHUBBUB_SOFT_RESET, DMUIF_SOFT_RESET, 1);
+ REG_UPDATE(DMCUB_CNTL, DMCUB_ENABLE, 0);
offset = cw0->offset;
@@ -181,6 +189,7 @@ void dmub_dcn401_backdoor_load_zfb_mode(struct dmub_srv *dmub,
DMCUB_REGION3_CW1_TOP_ADDRESS, cw1->region.top,
DMCUB_REGION3_CW1_ENABLE, 1);
+ /* release DMCUB reset only to prevent premature execution */
REG_UPDATE_2(DMCUB_SEC_CNTL, DMCUB_SEC_RESET, 0, DMCUB_MEM_UNIT_ID,
0x20);
}
@@ -402,72 +411,78 @@ uint32_t dmub_dcn401_get_current_time(struct dmub_srv *dmub)
return REG_READ(DMCUB_TIMER_CURRENT);
}
-void dmub_dcn401_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data)
+void dmub_dcn401_get_diagnostic_data(struct dmub_srv *dmub)
{
- uint32_t is_dmub_enabled, is_soft_reset, is_sec_reset;
+ uint32_t is_dmub_enabled, is_soft_reset, is_sec_reset, is_pwait;
uint32_t is_traceport_enabled, is_cw0_enabled, is_cw6_enabled;
+ struct dmub_timeout_info timeout = {0};
- if (!dmub || !diag_data)
+ if (!dmub)
return;
- memset(diag_data, 0, sizeof(*diag_data));
-
- diag_data->dmcub_version = dmub->fw_version;
-
- diag_data->scratch[0] = REG_READ(DMCUB_SCRATCH0);
- diag_data->scratch[1] = REG_READ(DMCUB_SCRATCH1);
- diag_data->scratch[2] = REG_READ(DMCUB_SCRATCH2);
- diag_data->scratch[3] = REG_READ(DMCUB_SCRATCH3);
- diag_data->scratch[4] = REG_READ(DMCUB_SCRATCH4);
- diag_data->scratch[5] = REG_READ(DMCUB_SCRATCH5);
- diag_data->scratch[6] = REG_READ(DMCUB_SCRATCH6);
- diag_data->scratch[7] = REG_READ(DMCUB_SCRATCH7);
- diag_data->scratch[8] = REG_READ(DMCUB_SCRATCH8);
- diag_data->scratch[9] = REG_READ(DMCUB_SCRATCH9);
- diag_data->scratch[10] = REG_READ(DMCUB_SCRATCH10);
- diag_data->scratch[11] = REG_READ(DMCUB_SCRATCH11);
- diag_data->scratch[12] = REG_READ(DMCUB_SCRATCH12);
- diag_data->scratch[13] = REG_READ(DMCUB_SCRATCH13);
- diag_data->scratch[14] = REG_READ(DMCUB_SCRATCH14);
- diag_data->scratch[15] = REG_READ(DMCUB_SCRATCH15);
- diag_data->scratch[16] = REG_READ(DMCUB_SCRATCH16);
-
- diag_data->undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
- diag_data->inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
- diag_data->data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
-
- diag_data->inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
- diag_data->inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
- diag_data->inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
-
- diag_data->inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
- diag_data->inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
- diag_data->inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
-
- diag_data->outbox1_rptr = REG_READ(DMCUB_OUTBOX1_RPTR);
- diag_data->outbox1_wptr = REG_READ(DMCUB_OUTBOX1_WPTR);
- diag_data->outbox1_size = REG_READ(DMCUB_OUTBOX1_SIZE);
+ /* timeout data filled externally, cache before resetting memory */
+ timeout = dmub->debug.timeout_info;
+ memset(&dmub->debug, 0, sizeof(dmub->debug));
+ dmub->debug.timeout_info = timeout;
+
+ dmub->debug.dmcub_version = dmub->fw_version;
+
+ dmub->debug.scratch[0] = REG_READ(DMCUB_SCRATCH0);
+ dmub->debug.scratch[1] = REG_READ(DMCUB_SCRATCH1);
+ dmub->debug.scratch[2] = REG_READ(DMCUB_SCRATCH2);
+ dmub->debug.scratch[3] = REG_READ(DMCUB_SCRATCH3);
+ dmub->debug.scratch[4] = REG_READ(DMCUB_SCRATCH4);
+ dmub->debug.scratch[5] = REG_READ(DMCUB_SCRATCH5);
+ dmub->debug.scratch[6] = REG_READ(DMCUB_SCRATCH6);
+ dmub->debug.scratch[7] = REG_READ(DMCUB_SCRATCH7);
+ dmub->debug.scratch[8] = REG_READ(DMCUB_SCRATCH8);
+ dmub->debug.scratch[9] = REG_READ(DMCUB_SCRATCH9);
+ dmub->debug.scratch[10] = REG_READ(DMCUB_SCRATCH10);
+ dmub->debug.scratch[11] = REG_READ(DMCUB_SCRATCH11);
+ dmub->debug.scratch[12] = REG_READ(DMCUB_SCRATCH12);
+ dmub->debug.scratch[13] = REG_READ(DMCUB_SCRATCH13);
+ dmub->debug.scratch[14] = REG_READ(DMCUB_SCRATCH14);
+ dmub->debug.scratch[15] = REG_READ(DMCUB_SCRATCH15);
+ dmub->debug.scratch[16] = REG_READ(DMCUB_SCRATCH16);
+
+ dmub->debug.undefined_address_fault_addr = REG_READ(DMCUB_UNDEFINED_ADDRESS_FAULT_ADDR);
+ dmub->debug.inst_fetch_fault_addr = REG_READ(DMCUB_INST_FETCH_FAULT_ADDR);
+ dmub->debug.data_write_fault_addr = REG_READ(DMCUB_DATA_WRITE_FAULT_ADDR);
+
+ dmub->debug.inbox1_rptr = REG_READ(DMCUB_INBOX1_RPTR);
+ dmub->debug.inbox1_wptr = REG_READ(DMCUB_INBOX1_WPTR);
+ dmub->debug.inbox1_size = REG_READ(DMCUB_INBOX1_SIZE);
+
+ dmub->debug.inbox0_rptr = REG_READ(DMCUB_INBOX0_RPTR);
+ dmub->debug.inbox0_wptr = REG_READ(DMCUB_INBOX0_WPTR);
+ dmub->debug.inbox0_size = REG_READ(DMCUB_INBOX0_SIZE);
+
+ dmub->debug.outbox1_rptr = REG_READ(DMCUB_OUTBOX1_RPTR);
+ dmub->debug.outbox1_wptr = REG_READ(DMCUB_OUTBOX1_WPTR);
+ dmub->debug.outbox1_size = REG_READ(DMCUB_OUTBOX1_SIZE);
REG_GET(DMCUB_CNTL, DMCUB_ENABLE, &is_dmub_enabled);
- diag_data->is_dmcub_enabled = is_dmub_enabled;
+ dmub->debug.is_dmcub_enabled = is_dmub_enabled;
+
+ REG_GET(DMCUB_CNTL, DMCUB_PWAIT_MODE_STATUS, &is_pwait);
+ dmub->debug.is_pwait = is_pwait;
REG_GET(DMCUB_CNTL2, DMCUB_SOFT_RESET, &is_soft_reset);
- diag_data->is_dmcub_soft_reset = is_soft_reset;
+ dmub->debug.is_dmcub_soft_reset = is_soft_reset;
REG_GET(DMCUB_SEC_CNTL, DMCUB_SEC_RESET_STATUS, &is_sec_reset);
- diag_data->is_dmcub_secure_reset = is_sec_reset;
+ dmub->debug.is_dmcub_secure_reset = is_sec_reset;
REG_GET(DMCUB_CNTL, DMCUB_TRACEPORT_EN, &is_traceport_enabled);
- diag_data->is_traceport_en = is_traceport_enabled;
+ dmub->debug.is_traceport_en = is_traceport_enabled;
REG_GET(DMCUB_REGION3_CW0_TOP_ADDRESS, DMCUB_REGION3_CW0_ENABLE, &is_cw0_enabled);
- diag_data->is_cw0_enabled = is_cw0_enabled;
+ dmub->debug.is_cw0_enabled = is_cw0_enabled;
REG_GET(DMCUB_REGION3_CW6_TOP_ADDRESS, DMCUB_REGION3_CW6_ENABLE, &is_cw6_enabled);
- diag_data->is_cw6_enabled = is_cw6_enabled;
+ dmub->debug.is_cw6_enabled = is_cw6_enabled;
- diag_data->gpint_datain0 = REG_READ(DMCUB_GPINT_DATAIN0);
- diag_data->timeout_info = dmub->debug;
+ dmub->debug.gpint_datain0 = REG_READ(DMCUB_GPINT_DATAIN0);
}
void dmub_dcn401_configure_dmub_in_system_memory(struct dmub_srv *dmub)
{
@@ -501,28 +516,69 @@ void dmub_dcn401_send_reg_inbox0_cmd_msg(struct dmub_srv *dmub,
union dmub_rb_cmd *cmd)
{
uint32_t *dwords = (uint32_t *)cmd;
-
+ int32_t payload_size_bytes = cmd->cmd_common.header.payload_bytes;
+ uint32_t msg_index;
static_assert(sizeof(*cmd) == 64, "DMUB command size mismatch");
- REG_WRITE(DMCUB_REG_INBOX0_MSG0, dwords[0]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG1, dwords[1]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG2, dwords[2]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG3, dwords[3]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG4, dwords[4]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG5, dwords[5]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG6, dwords[6]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG7, dwords[7]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG8, dwords[8]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG9, dwords[9]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG10, dwords[10]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG11, dwords[11]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG12, dwords[12]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG13, dwords[13]);
- REG_WRITE(DMCUB_REG_INBOX0_MSG14, dwords[14]);
+ /* read remaining data based on payload size */
+ for (msg_index = 0; msg_index < 15; msg_index++) {
+ if (payload_size_bytes <= msg_index * 4) {
+ break;
+ }
+
+ switch (msg_index) {
+ case 0:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG0, dwords[msg_index + 1]);
+ break;
+ case 1:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG1, dwords[msg_index + 1]);
+ break;
+ case 2:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG2, dwords[msg_index + 1]);
+ break;
+ case 3:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG3, dwords[msg_index + 1]);
+ break;
+ case 4:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG4, dwords[msg_index + 1]);
+ break;
+ case 5:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG5, dwords[msg_index + 1]);
+ break;
+ case 6:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG6, dwords[msg_index + 1]);
+ break;
+ case 7:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG7, dwords[msg_index + 1]);
+ break;
+ case 8:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG8, dwords[msg_index + 1]);
+ break;
+ case 9:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG9, dwords[msg_index + 1]);
+ break;
+ case 10:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG10, dwords[msg_index + 1]);
+ break;
+ case 11:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG11, dwords[msg_index + 1]);
+ break;
+ case 12:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG12, dwords[msg_index + 1]);
+ break;
+ case 13:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG13, dwords[msg_index + 1]);
+ break;
+ case 14:
+ REG_WRITE(DMCUB_REG_INBOX0_MSG14, dwords[msg_index + 1]);
+ break;
+ }
+ }
+
/* writing to INBOX RDY register will trigger DMUB REG INBOX0 RDY
* interrupt.
*/
- REG_WRITE(DMCUB_REG_INBOX0_RDY, dwords[15]);
+ REG_WRITE(DMCUB_REG_INBOX0_RDY, dwords[0]);
}
uint32_t dmub_dcn401_read_reg_inbox0_rsp_int_status(struct dmub_srv *dmub)
@@ -540,30 +596,39 @@ void dmub_dcn401_read_reg_inbox0_cmd_rsp(struct dmub_srv *dmub,
static_assert(sizeof(*cmd) == 64, "DMUB command size mismatch");
- dwords[0] = REG_READ(DMCUB_REG_INBOX0_MSG0);
- dwords[1] = REG_READ(DMCUB_REG_INBOX0_MSG1);
- dwords[2] = REG_READ(DMCUB_REG_INBOX0_MSG2);
- dwords[3] = REG_READ(DMCUB_REG_INBOX0_MSG3);
- dwords[4] = REG_READ(DMCUB_REG_INBOX0_MSG4);
- dwords[5] = REG_READ(DMCUB_REG_INBOX0_MSG5);
- dwords[6] = REG_READ(DMCUB_REG_INBOX0_MSG6);
- dwords[7] = REG_READ(DMCUB_REG_INBOX0_MSG7);
- dwords[8] = REG_READ(DMCUB_REG_INBOX0_MSG8);
- dwords[9] = REG_READ(DMCUB_REG_INBOX0_MSG9);
- dwords[10] = REG_READ(DMCUB_REG_INBOX0_MSG10);
- dwords[11] = REG_READ(DMCUB_REG_INBOX0_MSG11);
- dwords[12] = REG_READ(DMCUB_REG_INBOX0_MSG12);
- dwords[13] = REG_READ(DMCUB_REG_INBOX0_MSG13);
- dwords[14] = REG_READ(DMCUB_REG_INBOX0_MSG14);
- dwords[15] = REG_READ(DMCUB_REG_INBOX0_RSP);
+ dwords[0] = REG_READ(DMCUB_REG_INBOX0_RSP);
+ dwords[1] = REG_READ(DMCUB_REG_INBOX0_MSG0);
+ dwords[2] = REG_READ(DMCUB_REG_INBOX0_MSG1);
+ dwords[3] = REG_READ(DMCUB_REG_INBOX0_MSG2);
+ dwords[4] = REG_READ(DMCUB_REG_INBOX0_MSG3);
+ dwords[5] = REG_READ(DMCUB_REG_INBOX0_MSG4);
+ dwords[6] = REG_READ(DMCUB_REG_INBOX0_MSG5);
+ dwords[7] = REG_READ(DMCUB_REG_INBOX0_MSG6);
+ dwords[8] = REG_READ(DMCUB_REG_INBOX0_MSG7);
+ dwords[9] = REG_READ(DMCUB_REG_INBOX0_MSG8);
+ dwords[10] = REG_READ(DMCUB_REG_INBOX0_MSG9);
+ dwords[11] = REG_READ(DMCUB_REG_INBOX0_MSG10);
+ dwords[12] = REG_READ(DMCUB_REG_INBOX0_MSG11);
+ dwords[13] = REG_READ(DMCUB_REG_INBOX0_MSG12);
+ dwords[14] = REG_READ(DMCUB_REG_INBOX0_MSG13);
+ dwords[15] = REG_READ(DMCUB_REG_INBOX0_MSG14);
}
void dmub_dcn401_write_reg_inbox0_rsp_int_ack(struct dmub_srv *dmub)
{
REG_UPDATE(HOST_INTERRUPT_CSR, HOST_REG_INBOX0_RSP_INT_ACK, 1);
+}
+
+void dmub_dcn401_clear_reg_inbox0_rsp_int_ack(struct dmub_srv *dmub)
+{
REG_UPDATE(HOST_INTERRUPT_CSR, HOST_REG_INBOX0_RSP_INT_ACK, 0);
}
+void dmub_dcn401_enable_reg_inbox0_rsp_int(struct dmub_srv *dmub, bool enable)
+{
+ REG_UPDATE(HOST_INTERRUPT_CSR, HOST_REG_INBOX0_RSP_INT_EN, enable ? 1:0);
+}
+
void dmub_dcn401_write_reg_outbox0_rdy_int_ack(struct dmub_srv *dmub)
{
REG_UPDATE(HOST_INTERRUPT_CSR, HOST_REG_OUTBOX0_RDY_INT_ACK, 1);
@@ -588,11 +653,6 @@ uint32_t dmub_dcn401_read_reg_outbox0_rsp_int_status(struct dmub_srv *dmub)
return status;
}
-void dmub_dcn401_enable_reg_inbox0_rsp_int(struct dmub_srv *dmub, bool enable)
-{
- REG_UPDATE(HOST_INTERRUPT_CSR, HOST_REG_INBOX0_RSP_INT_EN, enable ? 1:0);
-}
-
void dmub_dcn401_enable_reg_outbox0_rdy_int(struct dmub_srv *dmub, bool enable)
{
REG_UPDATE(HOST_INTERRUPT_CSR, HOST_REG_OUTBOX0_RDY_INT_EN, enable ? 1:0);
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.h b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.h
index 4c8843b79695..88c3a44d67d9 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.h
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_dcn401.h
@@ -169,7 +169,8 @@ struct dmub_srv;
DMUB_SF(HOST_INTERRUPT_CSR, HOST_REG_INBOX0_RSP_INT_EN) \
DMUB_SF(HOST_INTERRUPT_CSR, HOST_REG_OUTBOX0_RDY_INT_ACK) \
DMUB_SF(HOST_INTERRUPT_CSR, HOST_REG_OUTBOX0_RDY_INT_STAT) \
- DMUB_SF(HOST_INTERRUPT_CSR, HOST_REG_OUTBOX0_RDY_INT_EN)
+ DMUB_SF(HOST_INTERRUPT_CSR, HOST_REG_OUTBOX0_RDY_INT_EN) \
+ DMUB_SF(DMCUB_CNTL, DMCUB_PWAIT_MODE_STATUS)
struct dmub_srv_dcn401_reg_offset {
#define DMUB_SR(reg) uint32_t reg;
@@ -263,7 +264,7 @@ void dmub_dcn401_set_outbox0_rptr(struct dmub_srv *dmub, uint32_t rptr_offset);
uint32_t dmub_dcn401_get_current_time(struct dmub_srv *dmub);
-void dmub_dcn401_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data);
+void dmub_dcn401_get_diagnostic_data(struct dmub_srv *dmub);
void dmub_dcn401_configure_dmub_in_system_memory(struct dmub_srv *dmub);
void dmub_dcn401_send_inbox0_cmd(struct dmub_srv *dmub, union dmub_inbox0_data_register data);
@@ -276,11 +277,13 @@ uint32_t dmub_dcn401_read_reg_inbox0_rsp_int_status(struct dmub_srv *dmub);
void dmub_dcn401_read_reg_inbox0_cmd_rsp(struct dmub_srv *dmub,
union dmub_rb_cmd *cmd);
void dmub_dcn401_write_reg_inbox0_rsp_int_ack(struct dmub_srv *dmub);
+void dmub_dcn401_clear_reg_inbox0_rsp_int_ack(struct dmub_srv *dmub);
+void dmub_dcn401_enable_reg_inbox0_rsp_int(struct dmub_srv *dmub, bool enable);
+
void dmub_dcn401_write_reg_outbox0_rdy_int_ack(struct dmub_srv *dmub);
void dmub_dcn401_read_reg_outbox0_msg(struct dmub_srv *dmub, uint32_t *msg);
void dmub_dcn401_write_reg_outbox0_rsp(struct dmub_srv *dmub, uint32_t *msg);
uint32_t dmub_dcn401_read_reg_outbox0_rsp_int_status(struct dmub_srv *dmub);
-void dmub_dcn401_enable_reg_inbox0_rsp_int(struct dmub_srv *dmub, bool enable);
void dmub_dcn401_enable_reg_outbox0_rdy_int(struct dmub_srv *dmub, bool enable);
uint32_t dmub_dcn401_read_reg_outbox0_rdy_int_status(struct dmub_srv *dmub);
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c
index a3f3ff5d49ac..b17a19400c06 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv.c
@@ -38,6 +38,7 @@
#include "dmub_dcn32.h"
#include "dmub_dcn35.h"
#include "dmub_dcn351.h"
+#include "dmub_dcn36.h"
#include "dmub_dcn401.h"
#include "os_types.h"
/*
@@ -61,13 +62,15 @@
/* Default state size if meta is absent. */
#define DMUB_FW_STATE_SIZE (64 * 1024)
-/* Default tracebuffer size if meta is absent. */
-#define DMUB_TRACE_BUFFER_SIZE (64 * 1024)
-
-
/* Default scratch mem size. */
#define DMUB_SCRATCH_MEM_SIZE (1024)
+/* Default indirect buffer size. */
+#define DMUB_IB_MEM_SIZE (1280)
+
+/* Default LSDMA ring buffer size. */
+#define DMUB_LSDMA_RB_SIZE (64 * 1024)
+
/* Number of windows in use. */
#define DMUB_NUM_WINDOWS (DMUB_WINDOW_TOTAL)
/* Base addresses. */
@@ -160,6 +163,9 @@ static bool dmub_srv_hw_setup(struct dmub_srv *dmub, enum dmub_asic asic)
{
struct dmub_srv_hw_funcs *funcs = &dmub->hw_funcs;
+ /* default to specifying now inbox type */
+ enum dmub_inbox_cmd_interface_type default_inbox_type = DMUB_CMD_INTERFACE_DEFAULT;
+
switch (asic) {
case DMUB_ASIC_DCN20:
case DMUB_ASIC_DCN21:
@@ -318,6 +324,7 @@ static bool dmub_srv_hw_setup(struct dmub_srv *dmub, enum dmub_asic asic)
case DMUB_ASIC_DCN35:
case DMUB_ASIC_DCN351:
+ case DMUB_ASIC_DCN36:
dmub->regs_dcn35 = &dmub_srv_dcn35_regs;
funcs->configure_dmub_in_system_memory = dmub_dcn35_configure_dmub_in_system_memory;
funcs->send_inbox0_cmd = dmub_dcn35_send_inbox0_cmd;
@@ -355,7 +362,9 @@ static bool dmub_srv_hw_setup(struct dmub_srv *dmub, enum dmub_asic asic)
funcs->init_reg_offsets = dmub_srv_dcn35_regs_init;
if (asic == DMUB_ASIC_DCN351)
- funcs->init_reg_offsets = dmub_srv_dcn351_regs_init;
+ funcs->init_reg_offsets = dmub_srv_dcn351_regs_init;
+ if (asic == DMUB_ASIC_DCN36)
+ funcs->init_reg_offsets = dmub_srv_dcn36_regs_init;
funcs->is_hw_powered_up = dmub_dcn35_is_hw_powered_up;
funcs->should_detect = dmub_dcn35_should_detect;
@@ -395,10 +404,15 @@ static bool dmub_srv_hw_setup(struct dmub_srv *dmub, enum dmub_asic asic)
funcs->get_current_time = dmub_dcn401_get_current_time;
funcs->get_diagnostic_data = dmub_dcn401_get_diagnostic_data;
+
funcs->send_reg_inbox0_cmd_msg = dmub_dcn401_send_reg_inbox0_cmd_msg;
funcs->read_reg_inbox0_rsp_int_status = dmub_dcn401_read_reg_inbox0_rsp_int_status;
funcs->read_reg_inbox0_cmd_rsp = dmub_dcn401_read_reg_inbox0_cmd_rsp;
funcs->write_reg_inbox0_rsp_int_ack = dmub_dcn401_write_reg_inbox0_rsp_int_ack;
+ funcs->clear_reg_inbox0_rsp_int_ack = dmub_dcn401_clear_reg_inbox0_rsp_int_ack;
+ funcs->enable_reg_inbox0_rsp_int = dmub_dcn401_enable_reg_inbox0_rsp_int;
+ default_inbox_type = DMUB_CMD_INTERFACE_FB; // still default to FB for now
+
funcs->write_reg_outbox0_rdy_int_ack = dmub_dcn401_write_reg_outbox0_rdy_int_ack;
funcs->read_reg_outbox0_msg = dmub_dcn401_read_reg_outbox0_msg;
funcs->write_reg_outbox0_rsp = dmub_dcn401_write_reg_outbox0_rsp;
@@ -411,6 +425,20 @@ static bool dmub_srv_hw_setup(struct dmub_srv *dmub, enum dmub_asic asic)
return false;
}
+ /* set default inbox type if not overriden */
+ if (dmub->inbox_type == DMUB_CMD_INTERFACE_DEFAULT) {
+ if (default_inbox_type != DMUB_CMD_INTERFACE_DEFAULT) {
+ /* use default inbox type as specified by DCN rev */
+ dmub->inbox_type = default_inbox_type;
+ } else if (funcs->send_reg_inbox0_cmd_msg) {
+ /* prefer reg as default inbox type if present */
+ dmub->inbox_type = DMUB_CMD_INTERFACE_REG;
+ } else {
+ /* use fb as fallback */
+ dmub->inbox_type = DMUB_CMD_INTERFACE_FB;
+ }
+ }
+
return true;
}
@@ -426,6 +454,7 @@ enum dmub_status dmub_srv_create(struct dmub_srv *dmub,
dmub->asic = params->asic;
dmub->fw_version = params->fw_version;
dmub->is_virtual = params->is_virtual;
+ dmub->inbox_type = params->inbox_type;
/* Setup asic dependent hardware funcs. */
if (!dmub_srv_hw_setup(dmub, params->asic)) {
@@ -536,7 +565,9 @@ enum dmub_status
window_sizes[DMUB_WINDOW_5_TRACEBUFF] = trace_buffer_size;
window_sizes[DMUB_WINDOW_6_FW_STATE] = fw_state_size;
window_sizes[DMUB_WINDOW_7_SCRATCH_MEM] = DMUB_SCRATCH_MEM_SIZE;
+ window_sizes[DMUB_WINDOW_IB_MEM] = DMUB_IB_MEM_SIZE;
window_sizes[DMUB_WINDOW_SHARED_STATE] = max(DMUB_FW_HEADER_SHARED_STATE_SIZE, shared_state_size);
+ window_sizes[DMUB_WINDOW_LSDMA_BUFFER] = DMUB_LSDMA_RB_SIZE;
out->fb_size =
dmub_srv_calc_regions_for_memory_type(params, out, window_sizes, DMUB_WINDOW_MEMORY_TYPE_FB);
@@ -622,6 +653,7 @@ enum dmub_status dmub_srv_hw_init(struct dmub_srv *dmub,
struct dmub_fb *tracebuff_fb = params->fb[DMUB_WINDOW_5_TRACEBUFF];
struct dmub_fb *fw_state_fb = params->fb[DMUB_WINDOW_6_FW_STATE];
struct dmub_fb *scratch_mem_fb = params->fb[DMUB_WINDOW_7_SCRATCH_MEM];
+ struct dmub_fb *ib_mem_gart = params->fb[DMUB_WINDOW_IB_MEM];
struct dmub_fb *shared_state_fb = params->fb[DMUB_WINDOW_SHARED_STATE];
struct dmub_rb_init_params rb_params, outbox0_rb_params;
@@ -632,7 +664,7 @@ enum dmub_status dmub_srv_hw_init(struct dmub_srv *dmub,
return DMUB_STATUS_INVALID;
if (!inst_fb || !stack_fb || !data_fb || !bios_fb || !mail_fb ||
- !tracebuff_fb || !fw_state_fb || !scratch_mem_fb) {
+ !tracebuff_fb || !fw_state_fb || !scratch_mem_fb || !ib_mem_gart) {
ASSERT(0);
return DMUB_STATUS_INVALID;
}
@@ -695,7 +727,7 @@ enum dmub_status dmub_srv_hw_init(struct dmub_srv *dmub,
inbox1.base = cw4.region.base;
inbox1.top = cw4.region.base + DMUB_RB_SIZE;
outbox1.base = inbox1.top;
- outbox1.top = cw4.region.top;
+ outbox1.top = inbox1.top + DMUB_RB_SIZE;
cw5.offset.quad_part = tracebuff_fb->gpu_addr;
cw5.region.base = DMUB_CW5_BASE;
@@ -708,7 +740,7 @@ enum dmub_status dmub_srv_hw_init(struct dmub_srv *dmub,
cw6.region.base = DMUB_CW6_BASE;
cw6.region.top = cw6.region.base + fw_state_fb->size;
- dmub->fw_state = fw_state_fb->cpu_addr;
+ dmub->fw_state = (void *)((uintptr_t)(fw_state_fb->cpu_addr) + DMUB_DEBUG_FW_STATE_OFFSET);
region6.offset.quad_part = shared_state_fb->gpu_addr;
region6.region.base = DMUB_CW6_BASE;
@@ -718,6 +750,8 @@ enum dmub_status dmub_srv_hw_init(struct dmub_srv *dmub,
dmub->scratch_mem_fb = *scratch_mem_fb;
+ dmub->ib_mem_gart = *ib_mem_gart;
+
if (dmub->hw_funcs.setup_windows)
dmub->hw_funcs.setup_windows(dmub, &cw2, &cw3, &cw4, &cw5, &cw6, &region6);
@@ -737,7 +771,7 @@ enum dmub_status dmub_srv_hw_init(struct dmub_srv *dmub,
rb_params.ctx = dmub;
rb_params.base_address = mail_fb->cpu_addr;
rb_params.capacity = DMUB_RB_SIZE;
- dmub_rb_init(&dmub->inbox1_rb, &rb_params);
+ dmub_rb_init(&dmub->inbox1.rb, &rb_params);
// Initialize outbox1 ring buffer
rb_params.ctx = dmub;
@@ -768,27 +802,6 @@ enum dmub_status dmub_srv_hw_init(struct dmub_srv *dmub,
return DMUB_STATUS_OK;
}
-enum dmub_status dmub_srv_sync_inbox1(struct dmub_srv *dmub)
-{
- if (!dmub->sw_init)
- return DMUB_STATUS_INVALID;
-
- if (dmub->hw_funcs.get_inbox1_rptr && dmub->hw_funcs.get_inbox1_wptr) {
- uint32_t rptr = dmub->hw_funcs.get_inbox1_rptr(dmub);
- uint32_t wptr = dmub->hw_funcs.get_inbox1_wptr(dmub);
-
- if (rptr > dmub->inbox1_rb.capacity || wptr > dmub->inbox1_rb.capacity) {
- return DMUB_STATUS_HW_FAILURE;
- } else {
- dmub->inbox1_rb.rptr = rptr;
- dmub->inbox1_rb.wrpt = wptr;
- dmub->inbox1_last_wptr = dmub->inbox1_rb.wrpt;
- }
- }
-
- return DMUB_STATUS_OK;
-}
-
enum dmub_status dmub_srv_hw_reset(struct dmub_srv *dmub)
{
if (!dmub->sw_init)
@@ -799,8 +812,13 @@ enum dmub_status dmub_srv_hw_reset(struct dmub_srv *dmub)
/* mailboxes have been reset in hw, so reset the sw state as well */
dmub->inbox1_last_wptr = 0;
- dmub->inbox1_rb.wrpt = 0;
- dmub->inbox1_rb.rptr = 0;
+ dmub->inbox1.rb.wrpt = 0;
+ dmub->inbox1.rb.rptr = 0;
+ dmub->inbox1.num_reported = 0;
+ dmub->inbox1.num_submitted = 0;
+ dmub->reg_inbox0.num_reported = 0;
+ dmub->reg_inbox0.num_submitted = 0;
+ dmub->reg_inbox0.is_pending = 0;
dmub->outbox0_rb.wrpt = 0;
dmub->outbox0_rb.rptr = 0;
dmub->outbox1_rb.wrpt = 0;
@@ -811,7 +829,7 @@ enum dmub_status dmub_srv_hw_reset(struct dmub_srv *dmub)
return DMUB_STATUS_OK;
}
-enum dmub_status dmub_srv_cmd_queue(struct dmub_srv *dmub,
+enum dmub_status dmub_srv_fb_cmd_queue(struct dmub_srv *dmub,
const union dmub_rb_cmd *cmd)
{
if (!dmub->hw_init)
@@ -820,18 +838,20 @@ enum dmub_status dmub_srv_cmd_queue(struct dmub_srv *dmub,
if (dmub->power_state != DMUB_POWER_STATE_D0)
return DMUB_STATUS_POWER_STATE_D3;
- if (dmub->inbox1_rb.rptr > dmub->inbox1_rb.capacity ||
- dmub->inbox1_rb.wrpt > dmub->inbox1_rb.capacity) {
+ if (dmub->inbox1.rb.rptr > dmub->inbox1.rb.capacity ||
+ dmub->inbox1.rb.wrpt > dmub->inbox1.rb.capacity) {
return DMUB_STATUS_HW_FAILURE;
}
- if (dmub_rb_push_front(&dmub->inbox1_rb, cmd))
+ if (dmub_rb_push_front(&dmub->inbox1.rb, cmd)) {
+ dmub->inbox1.num_submitted++;
return DMUB_STATUS_OK;
+ }
return DMUB_STATUS_QUEUE_FULL;
}
-enum dmub_status dmub_srv_cmd_execute(struct dmub_srv *dmub)
+enum dmub_status dmub_srv_fb_cmd_execute(struct dmub_srv *dmub)
{
struct dmub_rb flush_rb;
@@ -846,13 +866,13 @@ enum dmub_status dmub_srv_cmd_execute(struct dmub_srv *dmub)
* been flushed to framebuffer memory. Otherwise DMCUB might
* read back stale, fully invalid or partially invalid data.
*/
- flush_rb = dmub->inbox1_rb;
+ flush_rb = dmub->inbox1.rb;
flush_rb.rptr = dmub->inbox1_last_wptr;
dmub_rb_flush_pending(&flush_rb);
- dmub->hw_funcs.set_inbox1_wptr(dmub, dmub->inbox1_rb.wrpt);
+ dmub->hw_funcs.set_inbox1_wptr(dmub, dmub->inbox1.rb.wrpt);
- dmub->inbox1_last_wptr = dmub->inbox1_rb.wrpt;
+ dmub->inbox1_last_wptr = dmub->inbox1.rb.wrpt;
return DMUB_STATUS_OK;
}
@@ -910,26 +930,84 @@ enum dmub_status dmub_srv_wait_for_auto_load(struct dmub_srv *dmub,
return DMUB_STATUS_TIMEOUT;
}
+static void dmub_srv_update_reg_inbox0_status(struct dmub_srv *dmub)
+{
+ if (dmub->reg_inbox0.is_pending) {
+ dmub->reg_inbox0.is_pending = dmub->hw_funcs.read_reg_inbox0_rsp_int_status &&
+ !dmub->hw_funcs.read_reg_inbox0_rsp_int_status(dmub);
+
+ if (!dmub->reg_inbox0.is_pending) {
+ /* ack the rsp interrupt */
+ if (dmub->hw_funcs.write_reg_inbox0_rsp_int_ack)
+ dmub->hw_funcs.write_reg_inbox0_rsp_int_ack(dmub);
+
+ /* only update the reported count if commands aren't being batched */
+ if (!dmub->reg_inbox0.is_pending && !dmub->reg_inbox0.is_multi_pending) {
+ dmub->reg_inbox0.num_reported = dmub->reg_inbox0.num_submitted;
+ }
+ }
+ }
+}
+
+enum dmub_status dmub_srv_wait_for_pending(struct dmub_srv *dmub,
+ uint32_t timeout_us)
+{
+ uint32_t i;
+ const uint32_t polling_interval_us = 1;
+ struct dmub_srv_inbox scratch_reg_inbox0 = dmub->reg_inbox0;
+ struct dmub_srv_inbox scratch_inbox1 = dmub->inbox1;
+ const volatile struct dmub_srv_inbox *reg_inbox0 = &dmub->reg_inbox0;
+ const volatile struct dmub_srv_inbox *inbox1 = &dmub->inbox1;
+
+ if (!dmub->hw_init ||
+ !dmub->hw_funcs.get_inbox1_wptr)
+ return DMUB_STATUS_INVALID;
+
+ for (i = 0; i <= timeout_us; i += polling_interval_us) {
+ scratch_inbox1.rb.wrpt = dmub->hw_funcs.get_inbox1_wptr(dmub);
+ scratch_inbox1.rb.rptr = dmub->hw_funcs.get_inbox1_rptr(dmub);
+
+ scratch_reg_inbox0.is_pending = scratch_reg_inbox0.is_pending &&
+ dmub->hw_funcs.read_reg_inbox0_rsp_int_status &&
+ !dmub->hw_funcs.read_reg_inbox0_rsp_int_status(dmub);
+
+ if (scratch_inbox1.rb.rptr > dmub->inbox1.rb.capacity)
+ return DMUB_STATUS_HW_FAILURE;
+
+ /* check current HW state first, but use command submission vs reported as a fallback */
+ if ((dmub_rb_empty(&scratch_inbox1.rb) ||
+ inbox1->num_reported >= scratch_inbox1.num_submitted) &&
+ (!scratch_reg_inbox0.is_pending ||
+ reg_inbox0->num_reported >= scratch_reg_inbox0.num_submitted))
+ return DMUB_STATUS_OK;
+
+ udelay(polling_interval_us);
+ }
+
+ return DMUB_STATUS_TIMEOUT;
+}
+
enum dmub_status dmub_srv_wait_for_idle(struct dmub_srv *dmub,
uint32_t timeout_us)
{
- uint32_t i, rptr;
+ enum dmub_status status;
+ uint32_t i;
+ const uint32_t polling_interval_us = 1;
if (!dmub->hw_init)
return DMUB_STATUS_INVALID;
- for (i = 0; i <= timeout_us; ++i) {
- rptr = dmub->hw_funcs.get_inbox1_rptr(dmub);
+ for (i = 0; i < timeout_us; i += polling_interval_us) {
+ status = dmub_srv_update_inbox_status(dmub);
- if (rptr > dmub->inbox1_rb.capacity)
- return DMUB_STATUS_HW_FAILURE;
+ if (status != DMUB_STATUS_OK)
+ return status;
- dmub->inbox1_rb.rptr = rptr;
-
- if (dmub_rb_empty(&dmub->inbox1_rb))
+ /* check for idle */
+ if (dmub_rb_empty(&dmub->inbox1.rb) && !dmub->reg_inbox0.is_pending)
return DMUB_STATUS_OK;
- udelay(1);
+ udelay(polling_interval_us);
}
return DMUB_STATUS_TIMEOUT;
@@ -1040,35 +1118,6 @@ enum dmub_status dmub_srv_set_skip_panel_power_sequence(struct dmub_srv *dmub,
return DMUB_STATUS_OK;
}
-enum dmub_status dmub_srv_cmd_with_reply_data(struct dmub_srv *dmub,
- union dmub_rb_cmd *cmd)
-{
- enum dmub_status status = DMUB_STATUS_OK;
-
- // Queue command
- status = dmub_srv_cmd_queue(dmub, cmd);
-
- if (status != DMUB_STATUS_OK)
- return status;
-
- // Execute command
- status = dmub_srv_cmd_execute(dmub);
-
- if (status != DMUB_STATUS_OK)
- return status;
-
- // Wait for DMUB to process command
- status = dmub_srv_wait_for_idle(dmub, 100000);
-
- if (status != DMUB_STATUS_OK)
- return status;
-
- // Copy data back from ring buffer into command
- dmub_rb_get_return_data(&dmub->inbox1_rb, cmd);
-
- return status;
-}
-
static inline bool dmub_rb_out_trace_buffer_front(struct dmub_rb *rb,
void *entry)
{
@@ -1099,11 +1148,11 @@ bool dmub_srv_get_outbox0_msg(struct dmub_srv *dmub, struct dmcub_trace_buf_entr
return dmub_rb_out_trace_buffer_front(&dmub->outbox0_rb, (void *)entry);
}
-bool dmub_srv_get_diagnostic_data(struct dmub_srv *dmub, struct dmub_diagnostic_data *diag_data)
+bool dmub_srv_get_diagnostic_data(struct dmub_srv *dmub)
{
- if (!dmub || !dmub->hw_funcs.get_diagnostic_data || !diag_data)
+ if (!dmub || !dmub->hw_funcs.get_diagnostic_data)
return false;
- dmub->hw_funcs.get_diagnostic_data(dmub, diag_data);
+ dmub->hw_funcs.get_diagnostic_data(dmub);
return true;
}
@@ -1160,46 +1209,162 @@ void dmub_srv_subvp_save_surf_addr(struct dmub_srv *dmub, const struct dc_plane_
}
}
-enum dmub_status dmub_srv_send_reg_inbox0_cmd(
- struct dmub_srv *dmub,
- union dmub_rb_cmd *cmd,
- bool with_reply, uint32_t timeout_us)
+void dmub_srv_set_power_state(struct dmub_srv *dmub, enum dmub_srv_power_state_type dmub_srv_power_state)
{
- uint32_t rsp_ready = 0;
- uint32_t i;
+ if (!dmub || !dmub->hw_init)
+ return;
+
+ dmub->power_state = dmub_srv_power_state;
+}
+
+enum dmub_status dmub_srv_reg_cmd_execute(struct dmub_srv *dmub, union dmub_rb_cmd *cmd)
+{
+ uint32_t num_pending = 0;
+ if (!dmub->hw_init)
+ return DMUB_STATUS_INVALID;
+
+ if (dmub->power_state != DMUB_POWER_STATE_D0)
+ return DMUB_STATUS_POWER_STATE_D3;
+
+ if (!dmub->hw_funcs.send_reg_inbox0_cmd_msg ||
+ !dmub->hw_funcs.clear_reg_inbox0_rsp_int_ack)
+ return DMUB_STATUS_INVALID;
+
+ if (dmub->reg_inbox0.num_submitted >= dmub->reg_inbox0.num_reported)
+ num_pending = dmub->reg_inbox0.num_submitted - dmub->reg_inbox0.num_reported;
+ else
+ /* num_submitted wrapped */
+ num_pending = DMUB_REG_INBOX0_RB_MAX_ENTRY -
+ (dmub->reg_inbox0.num_reported - dmub->reg_inbox0.num_submitted);
+
+ if (num_pending >= DMUB_REG_INBOX0_RB_MAX_ENTRY)
+ return DMUB_STATUS_QUEUE_FULL;
+
+ /* clear last rsp ack and send message */
+ dmub->hw_funcs.clear_reg_inbox0_rsp_int_ack(dmub);
dmub->hw_funcs.send_reg_inbox0_cmd_msg(dmub, cmd);
- for (i = 0; i < timeout_us; i++) {
- rsp_ready = dmub->hw_funcs.read_reg_inbox0_rsp_int_status(dmub);
- if (rsp_ready)
- break;
- udelay(1);
+ dmub->reg_inbox0.num_submitted++;
+ dmub->reg_inbox0.is_pending = true;
+ dmub->reg_inbox0.is_multi_pending = cmd->cmd_common.header.multi_cmd_pending;
+
+ return DMUB_STATUS_OK;
+}
+
+void dmub_srv_cmd_get_response(struct dmub_srv *dmub,
+ union dmub_rb_cmd *cmd_rsp)
+{
+ if (dmub) {
+ if (dmub->inbox_type == DMUB_CMD_INTERFACE_REG &&
+ dmub->hw_funcs.read_reg_inbox0_cmd_rsp) {
+ dmub->hw_funcs.read_reg_inbox0_cmd_rsp(dmub, cmd_rsp);
+ } else {
+ dmub_rb_get_return_data(&dmub->inbox1.rb, cmd_rsp);
+ }
}
- if (rsp_ready == 0)
- return DMUB_STATUS_TIMEOUT;
+}
- if (with_reply)
- dmub->hw_funcs.read_reg_inbox0_cmd_rsp(dmub, cmd);
+static enum dmub_status dmub_srv_sync_reg_inbox0(struct dmub_srv *dmub)
+{
+ if (!dmub || !dmub->sw_init)
+ return DMUB_STATUS_INVALID;
- dmub->hw_funcs.write_reg_inbox0_rsp_int_ack(dmub);
+ dmub->reg_inbox0.is_pending = 0;
+ dmub->reg_inbox0.is_multi_pending = 0;
- /* wait for rsp int status is cleared to initial state before exit */
- for (; i <= timeout_us; i++) {
- rsp_ready = dmub->hw_funcs.read_reg_inbox0_rsp_int_status(dmub);
- if (rsp_ready == 0)
- break;
- udelay(1);
+ return DMUB_STATUS_OK;
+}
+
+static enum dmub_status dmub_srv_sync_inbox1(struct dmub_srv *dmub)
+{
+ if (!dmub->sw_init)
+ return DMUB_STATUS_INVALID;
+
+ if (dmub->hw_funcs.get_inbox1_rptr && dmub->hw_funcs.get_inbox1_wptr) {
+ uint32_t rptr = dmub->hw_funcs.get_inbox1_rptr(dmub);
+ uint32_t wptr = dmub->hw_funcs.get_inbox1_wptr(dmub);
+
+ if (rptr > dmub->inbox1.rb.capacity || wptr > dmub->inbox1.rb.capacity) {
+ return DMUB_STATUS_HW_FAILURE;
+ } else {
+ dmub->inbox1.rb.rptr = rptr;
+ dmub->inbox1.rb.wrpt = wptr;
+ dmub->inbox1_last_wptr = dmub->inbox1.rb.wrpt;
+ }
}
- ASSERT(rsp_ready == 0);
return DMUB_STATUS_OK;
}
-void dmub_srv_set_power_state(struct dmub_srv *dmub, enum dmub_srv_power_state_type dmub_srv_power_state)
+enum dmub_status dmub_srv_sync_inboxes(struct dmub_srv *dmub)
{
- if (!dmub || !dmub->hw_init)
- return;
+ enum dmub_status status;
- dmub->power_state = dmub_srv_power_state;
+ status = dmub_srv_sync_reg_inbox0(dmub);
+ if (status != DMUB_STATUS_OK)
+ return status;
+
+ status = dmub_srv_sync_inbox1(dmub);
+ if (status != DMUB_STATUS_OK)
+ return status;
+
+ return DMUB_STATUS_OK;
+}
+
+enum dmub_status dmub_srv_wait_for_inbox_free(struct dmub_srv *dmub,
+ uint32_t timeout_us,
+ uint32_t num_free_required)
+{
+ enum dmub_status status;
+ uint32_t i;
+ const uint32_t polling_interval_us = 1;
+
+ if (!dmub->hw_init)
+ return DMUB_STATUS_INVALID;
+
+ for (i = 0; i < timeout_us; i += polling_interval_us) {
+ status = dmub_srv_update_inbox_status(dmub);
+
+ if (status != DMUB_STATUS_OK)
+ return status;
+
+ /* check for space in inbox1 */
+ if (dmub_rb_num_free(&dmub->inbox1.rb) >= num_free_required)
+ return DMUB_STATUS_OK;
+
+ udelay(polling_interval_us);
+ }
+
+ return DMUB_STATUS_TIMEOUT;
+}
+
+enum dmub_status dmub_srv_update_inbox_status(struct dmub_srv *dmub)
+{
+ uint32_t rptr;
+
+ if (!dmub->hw_init)
+ return DMUB_STATUS_INVALID;
+
+ if (dmub->power_state != DMUB_POWER_STATE_D0)
+ return DMUB_STATUS_POWER_STATE_D3;
+
+ /* update inbox1 state */
+ rptr = dmub->hw_funcs.get_inbox1_rptr(dmub);
+
+ if (rptr > dmub->inbox1.rb.capacity)
+ return DMUB_STATUS_HW_FAILURE;
+
+ if (dmub->inbox1.rb.rptr > rptr) {
+ /* rb wrapped */
+ dmub->inbox1.num_reported += (rptr + dmub->inbox1.rb.capacity - dmub->inbox1.rb.rptr) / DMUB_RB_CMD_SIZE;
+ } else {
+ dmub->inbox1.num_reported += (rptr - dmub->inbox1.rb.rptr) / DMUB_RB_CMD_SIZE;
+ }
+ dmub->inbox1.rb.rptr = rptr;
+
+ /* update reg_inbox0 */
+ dmub_srv_update_reg_inbox0_status(dmub);
+
+ return DMUB_STATUS_OK;
}
diff --git a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv_stat.c b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv_stat.c
index cce887cefc01..567c5b1aeb7a 100644
--- a/drivers/gpu/drm/amd/display/dmub/src/dmub_srv_stat.c
+++ b/drivers/gpu/drm/amd/display/dmub/src/dmub_srv_stat.c
@@ -95,23 +95,6 @@ enum dmub_status dmub_srv_stat_get_notification(struct dmub_srv *dmub,
case DMUB_OUT_CMD__DPIA_NOTIFICATION:
notify->type = DMUB_NOTIFICATION_DPIA_NOTIFICATION;
notify->link_index = cmd.dpia_notification.payload.header.instance;
-
- if (cmd.dpia_notification.payload.header.type == DPIA_NOTIFY__BW_ALLOCATION) {
-
- notify->dpia_notification.payload.data.dpia_bw_alloc.estimated_bw =
- cmd.dpia_notification.payload.data.dpia_bw_alloc.estimated_bw;
- notify->dpia_notification.payload.data.dpia_bw_alloc.allocated_bw =
- cmd.dpia_notification.payload.data.dpia_bw_alloc.allocated_bw;
-
- if (cmd.dpia_notification.payload.data.dpia_bw_alloc.bits.bw_request_failed)
- notify->result = DPIA_BW_REQ_FAILED;
- else if (cmd.dpia_notification.payload.data.dpia_bw_alloc.bits.bw_request_succeeded)
- notify->result = DPIA_BW_REQ_SUCCESS;
- else if (cmd.dpia_notification.payload.data.dpia_bw_alloc.bits.est_bw_changed)
- notify->result = DPIA_EST_BW_CHANGED;
- else if (cmd.dpia_notification.payload.data.dpia_bw_alloc.bits.bw_alloc_cap_changed)
- notify->result = DPIA_BW_ALLOC_CAPS_CHANGED;
- }
break;
case DMUB_OUT_CMD__HPD_SENSE_NOTIFY:
notify->type = DMUB_NOTIFICATION_HPD_SENSE_NOTIFY;
@@ -119,6 +102,10 @@ enum dmub_status dmub_srv_stat_get_notification(struct dmub_srv *dmub,
&cmd.hpd_sense_notify.data,
sizeof(cmd.hpd_sense_notify.data));
break;
+ case DMUB_OUT_CMD__FUSED_IO:
+ notify->type = DMUB_NOTIFICATION_FUSED_IO;
+ dmub_memcpy(&notify->fused_request, &cmd.fused_io.request, sizeof(cmd.fused_io.request));
+ break;
default:
notify->type = DMUB_NOTIFICATION_NO_DATA;
break;
diff --git a/drivers/gpu/drm/amd/display/include/dal_asic_id.h b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
index 090230d29df8..5fc29164e4b4 100644
--- a/drivers/gpu/drm/amd/display/include/dal_asic_id.h
+++ b/drivers/gpu/drm/amd/display/include/dal_asic_id.h
@@ -257,6 +257,7 @@ enum {
#define ASICREV_IS_GC_11_0_2(eChipRev) (eChipRev >= GC_11_0_2_A0 && eChipRev < GC_11_0_3_A0)
#define ASICREV_IS_GC_11_0_3(eChipRev) (eChipRev >= GC_11_0_3_A0 && eChipRev < GC_11_UNKNOWN)
#define ASICREV_IS_GC_11_0_4(eChipRev) (eChipRev >= GC_11_0_4_A0 && eChipRev < GC_11_UNKNOWN)
+#define ASICREV_IS_DCN36(eChipRev) ((eChipRev) >= 0x50 && (eChipRev) < 0xC0)
#define AMDGPU_FAMILY_GC_12_0_0 152 /* GC 12.0.0 */
diff --git a/drivers/gpu/drm/amd/display/include/dal_types.h b/drivers/gpu/drm/amd/display/include/dal_types.h
index 654387cf057f..a021d12acd74 100644
--- a/drivers/gpu/drm/amd/display/include/dal_types.h
+++ b/drivers/gpu/drm/amd/display/include/dal_types.h
@@ -63,6 +63,7 @@ enum dce_version {
DCN_VERSION_3_21,
DCN_VERSION_3_5,
DCN_VERSION_3_51,
+ DCN_VERSION_3_6,
DCN_VERSION_4_01,
DCN_VERSION_MAX
};
diff --git a/drivers/gpu/drm/amd/display/include/gpio_service_interface.h b/drivers/gpu/drm/amd/display/include/gpio_service_interface.h
index 7e3240e73c1f..63813009a3a6 100644
--- a/drivers/gpu/drm/amd/display/include/gpio_service_interface.h
+++ b/drivers/gpu/drm/amd/display/include/gpio_service_interface.h
@@ -86,6 +86,9 @@ enum dc_irq_source dal_irq_get_source(
enum dc_irq_source dal_irq_get_rx_source(
const struct gpio *irq);
+enum dc_irq_source dal_irq_get_read_request(
+ const struct gpio *irq);
+
enum gpio_result dal_irq_setup_hpd_filter(
struct gpio *irq,
struct gpio_hpd_config *config);
diff --git a/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h b/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h
index 813463ffe15c..cc467031651d 100644
--- a/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h
+++ b/drivers/gpu/drm/amd/display/include/grph_object_ctrl_defs.h
@@ -424,7 +424,7 @@ struct integrated_info {
/*
* DFS-bypass flag
*/
-/* Copy of SYS_INFO_GPUCAPS__ENABEL_DFS_BYPASS from atombios.h */
+/* Copy of SYS_INFO_GPUCAPS__ENABLE_DFS_BYPASS from atombios.h */
enum {
DFS_BYPASS_ENABLE = 0x10
};
diff --git a/drivers/gpu/drm/amd/display/include/link_service_types.h b/drivers/gpu/drm/amd/display/include/link_service_types.h
index 1867aac57cf2..da74ed66c8f9 100644
--- a/drivers/gpu/drm/amd/display/include/link_service_types.h
+++ b/drivers/gpu/drm/amd/display/include/link_service_types.h
@@ -89,6 +89,8 @@ struct link_training_settings {
bool enhanced_framing;
enum lttpr_mode lttpr_mode;
+ bool lttpr_early_tps2;
+
/* disallow different lanes to have different lane settings */
bool disallow_per_lane_settings;
/* dpcd lane settings will always use the same hw lane settings
diff --git a/drivers/gpu/drm/amd/display/include/logger_interface.h b/drivers/gpu/drm/amd/display/include/logger_interface.h
index 058f882d5bdd..4c01514b926c 100644
--- a/drivers/gpu/drm/amd/display/include/logger_interface.h
+++ b/drivers/gpu/drm/amd/display/include/logger_interface.h
@@ -40,11 +40,6 @@ struct dc_state;
*
*/
-void pre_surface_trace(
- struct dc *dc,
- const struct dc_plane_state *const *plane_states,
- int surface_count);
-
void update_surface_trace(
struct dc *dc,
const struct dc_surface_update *updates,
diff --git a/drivers/gpu/drm/amd/display/include/logger_types.h b/drivers/gpu/drm/amd/display/include/logger_types.h
index 4d68c1c6e210..177acb0574f1 100644
--- a/drivers/gpu/drm/amd/display/include/logger_types.h
+++ b/drivers/gpu/drm/amd/display/include/logger_types.h
@@ -32,6 +32,7 @@
#define DC_LOG_WARNING(...) drm_warn((DC_LOGGER)->dev, __VA_ARGS__)
#define DC_LOG_DEBUG(...) drm_dbg((DC_LOGGER)->dev, __VA_ARGS__)
#define DC_LOG_DC(...) drm_dbg((DC_LOGGER)->dev, __VA_ARGS__)
+#define DC_LOG_INFO(...) drm_info((DC_LOGGER)->dev, __VA_ARGS__)
#define DC_LOG_SURFACE(...) pr_debug("[SURFACE]:"__VA_ARGS__)
#define DC_LOG_HW_HOTPLUG(...) drm_dbg((DC_LOGGER)->dev, __VA_ARGS__)
#define DC_LOG_HW_LINK_TRAINING(...) pr_debug("[HW_LINK_TRAINING]:"__VA_ARGS__)
diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
index f980a84dceef..71efd2770c99 100644
--- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
+++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
@@ -122,6 +122,17 @@ static unsigned int calc_duration_in_us_from_v_total(
return duration_in_us;
}
+static unsigned int calc_max_hardware_v_total(const struct dc_stream_state *stream)
+{
+ unsigned int max_hw_v_total = stream->ctx->dc->caps.max_v_total;
+
+ if (stream->ctx->dc->caps.vtotal_limited_by_fp2) {
+ max_hw_v_total -= stream->timing.v_front_porch + 1;
+ }
+
+ return max_hw_v_total;
+}
+
unsigned int mod_freesync_calc_v_total_from_refresh(
const struct dc_stream_state *stream,
unsigned int refresh_in_uhz)
@@ -136,7 +147,7 @@ unsigned int mod_freesync_calc_v_total_from_refresh(
((unsigned int)(div64_u64((1000000000ULL * 1000000),
refresh_in_uhz)));
- if (MICRO_HZ_TO_HZ(refresh_in_uhz) <= stream->timing.min_refresh_in_uhz) {
+ if (refresh_in_uhz <= stream->timing.min_refresh_in_uhz) {
/* When the target refresh rate is the minimum panel refresh rate,
* round down the vtotal value to avoid stretching vblank over
* panel's vtotal boundary.
@@ -144,6 +155,14 @@ unsigned int mod_freesync_calc_v_total_from_refresh(
v_total = div64_u64(div64_u64(((unsigned long long)(
frame_duration_in_ns) * (stream->timing.pix_clk_100hz / 10)),
stream->timing.h_total), 1000000);
+ } else if (refresh_in_uhz >= stream->timing.max_refresh_in_uhz) {
+ /* When the target refresh rate is the maximum panel refresh rate
+ * round up the vtotal value to prevent off-by-one error causing
+ * v_total_min to be below the panel's lower bound
+ */
+ v_total = div64_u64(div64_u64(((unsigned long long)(
+ frame_duration_in_ns) * (stream->timing.pix_clk_100hz / 10)),
+ stream->timing.h_total) + (1000000 - 1), 1000000);
} else {
v_total = div64_u64(div64_u64(((unsigned long long)(
frame_duration_in_ns) * (stream->timing.pix_clk_100hz / 10)),
@@ -541,43 +560,6 @@ static bool vrr_settings_require_update(struct core_freesync *core_freesync,
return false;
}
-bool mod_freesync_get_vmin_vmax(struct mod_freesync *mod_freesync,
- const struct dc_stream_state *stream,
- unsigned int *vmin,
- unsigned int *vmax)
-{
- *vmin = stream->adjust.v_total_min;
- *vmax = stream->adjust.v_total_max;
-
- return true;
-}
-
-bool mod_freesync_get_v_position(struct mod_freesync *mod_freesync,
- struct dc_stream_state *stream,
- unsigned int *nom_v_pos,
- unsigned int *v_pos)
-{
- struct core_freesync *core_freesync = NULL;
- struct crtc_position position;
-
- if (mod_freesync == NULL)
- return false;
-
- core_freesync = MOD_FREESYNC_TO_CORE(mod_freesync);
-
- if (dc_stream_get_crtc_position(core_freesync->dc, &stream, 1,
- &position.vertical_count,
- &position.nominal_vcount)) {
-
- *nom_v_pos = position.nominal_vcount;
- *v_pos = position.vertical_count;
-
- return true;
- }
-
- return false;
-}
-
static void build_vrr_infopacket_data_v1(const struct mod_vrr_params *vrr,
struct dc_info_packet *infopacket,
bool freesync_on_desktop)
@@ -1016,7 +998,7 @@ void mod_freesync_build_vrr_params(struct mod_freesync *mod_freesync,
if (stream->ctx->dc->caps.max_v_total != 0 && stream->timing.h_total != 0) {
min_hardware_refresh_in_uhz = div64_u64((stream->timing.pix_clk_100hz * 100000000ULL),
- (stream->timing.h_total * (long long)stream->ctx->dc->caps.max_v_total));
+ (stream->timing.h_total * (long long)calc_max_hardware_v_total(stream)));
}
/* Limit minimum refresh rate to what can be supported by hardware */
min_refresh_in_uhz = min_hardware_refresh_in_uhz > in_config->min_refresh_in_uhz ?
@@ -1280,28 +1262,6 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync,
}
}
-void mod_freesync_get_settings(struct mod_freesync *mod_freesync,
- const struct mod_vrr_params *vrr,
- unsigned int *v_total_min, unsigned int *v_total_max,
- unsigned int *event_triggers,
- unsigned int *window_min, unsigned int *window_max,
- unsigned int *lfc_mid_point_in_us,
- unsigned int *inserted_frames,
- unsigned int *inserted_duration_in_us)
-{
- if (mod_freesync == NULL)
- return;
-
- if (vrr->supported) {
- *v_total_min = vrr->adjust.v_total_min;
- *v_total_max = vrr->adjust.v_total_max;
- *event_triggers = 0;
- *lfc_mid_point_in_us = vrr->btr.mid_point_in_us;
- *inserted_frames = vrr->btr.frames_to_insert;
- *inserted_duration_in_us = vrr->btr.inserted_duration_in_us;
- }
-}
-
unsigned long long mod_freesync_calc_nominal_field_rate(
const struct dc_stream_state *stream)
{
@@ -1317,85 +1277,7 @@ unsigned long long mod_freesync_calc_nominal_field_rate(
return nominal_field_rate_in_uhz;
}
-unsigned long long mod_freesync_calc_field_rate_from_timing(
- unsigned int vtotal, unsigned int htotal, unsigned int pix_clk)
-{
- unsigned long long field_rate_in_uhz = 0;
- unsigned int total = htotal * vtotal;
-
- /* Calculate nominal field rate for stream, rounded up to nearest integer */
- field_rate_in_uhz = pix_clk;
- field_rate_in_uhz *= 1000000ULL;
-
- field_rate_in_uhz = div_u64(field_rate_in_uhz, total);
-
- return field_rate_in_uhz;
-}
-
bool mod_freesync_get_freesync_enabled(struct mod_vrr_params *pVrr)
{
return (pVrr->state != VRR_STATE_UNSUPPORTED) && (pVrr->state != VRR_STATE_DISABLED);
}
-
-bool mod_freesync_is_valid_range(uint32_t min_refresh_cap_in_uhz,
- uint32_t max_refresh_cap_in_uhz,
- uint32_t nominal_field_rate_in_uhz)
-{
-
- /* Typically nominal refresh calculated can have some fractional part.
- * Allow for some rounding error of actual video timing by taking floor
- * of caps and request. Round the nominal refresh rate.
- *
- * Dividing will convert everything to units in Hz although input
- * variable name is in uHz!
- *
- * Also note, this takes care of rounding error on the nominal refresh
- * so by rounding error we only expect it to be off by a small amount,
- * such as < 0.1 Hz. i.e. 143.9xxx or 144.1xxx.
- *
- * Example 1. Caps Min = 40 Hz, Max = 144 Hz
- * Request Min = 40 Hz, Max = 144 Hz
- * Nominal = 143.5x Hz rounded to 144 Hz
- * This function should allow this as valid request
- *
- * Example 2. Caps Min = 40 Hz, Max = 144 Hz
- * Request Min = 40 Hz, Max = 144 Hz
- * Nominal = 144.4x Hz rounded to 144 Hz
- * This function should allow this as valid request
- *
- * Example 3. Caps Min = 40 Hz, Max = 144 Hz
- * Request Min = 40 Hz, Max = 144 Hz
- * Nominal = 120.xx Hz rounded to 120 Hz
- * This function should return NOT valid since the requested
- * max is greater than current timing's nominal
- *
- * Example 4. Caps Min = 40 Hz, Max = 120 Hz
- * Request Min = 40 Hz, Max = 120 Hz
- * Nominal = 144.xx Hz rounded to 144 Hz
- * This function should return NOT valid since the nominal
- * is greater than the capability's max refresh
- */
- nominal_field_rate_in_uhz =
- div_u64(nominal_field_rate_in_uhz + 500000, 1000000);
- min_refresh_cap_in_uhz /= 1000000;
- max_refresh_cap_in_uhz /= 1000000;
-
- /* Check nominal is within range */
- if (nominal_field_rate_in_uhz > max_refresh_cap_in_uhz ||
- nominal_field_rate_in_uhz < min_refresh_cap_in_uhz)
- return false;
-
- /* If nominal is less than max, limit the max allowed refresh rate */
- if (nominal_field_rate_in_uhz < max_refresh_cap_in_uhz)
- max_refresh_cap_in_uhz = nominal_field_rate_in_uhz;
-
- /* Check min is within range */
- if (min_refresh_cap_in_uhz > max_refresh_cap_in_uhz)
- return false;
-
- /* For variable range, check for at least 10 Hz range */
- if (nominal_field_rate_in_uhz - min_refresh_cap_in_uhz < 10)
- return false;
-
- return true;
-}
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h
index 55c7d873175f..a37634942b07 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp.h
@@ -386,6 +386,7 @@ enum mod_hdcp_status mod_hdcp_write_repeater_auth_ack(struct mod_hdcp *hdcp);
enum mod_hdcp_status mod_hdcp_write_stream_manage(struct mod_hdcp *hdcp);
enum mod_hdcp_status mod_hdcp_write_content_type(struct mod_hdcp *hdcp);
enum mod_hdcp_status mod_hdcp_clear_cp_irq_status(struct mod_hdcp *hdcp);
+enum mod_hdcp_status mod_hdcp_write_poll_read_lc_fw(struct mod_hdcp *hdcp);
/* hdcp version helpers */
static inline uint8_t is_dp_hdcp(struct mod_hdcp *hdcp)
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c
index 1d41dd58f6bc..bb8ae80b37f8 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_execution.c
@@ -452,21 +452,12 @@ out:
return status;
}
-static enum mod_hdcp_status locality_check(struct mod_hdcp *hdcp,
+static enum mod_hdcp_status locality_check_sw(struct mod_hdcp *hdcp,
struct mod_hdcp_event_context *event_ctx,
struct mod_hdcp_transition_input_hdcp2 *input)
{
enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
- if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
- event_ctx->unexpected_event = 1;
- goto out;
- }
-
- if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp2_prepare_lc_init,
- &input->lc_init_prepare, &status,
- hdcp, "lc_init_prepare"))
- goto out;
if (!mod_hdcp_execute_and_set(mod_hdcp_write_lc_init,
&input->lc_init_write, &status,
hdcp, "lc_init_write"))
@@ -482,6 +473,48 @@ static enum mod_hdcp_status locality_check(struct mod_hdcp *hdcp,
&input->l_prime_read, &status,
hdcp, "l_prime_read"))
goto out;
+out:
+ return status;
+}
+
+static enum mod_hdcp_status locality_check_fw(struct mod_hdcp *hdcp,
+ struct mod_hdcp_event_context *event_ctx,
+ struct mod_hdcp_transition_input_hdcp2 *input)
+{
+ enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+
+ if (!mod_hdcp_execute_and_set(mod_hdcp_write_poll_read_lc_fw,
+ &input->l_prime_read, &status,
+ hdcp, "l_prime_read"))
+ goto out;
+
+out:
+ return status;
+}
+
+static enum mod_hdcp_status locality_check(struct mod_hdcp *hdcp,
+ struct mod_hdcp_event_context *event_ctx,
+ struct mod_hdcp_transition_input_hdcp2 *input)
+{
+ enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+ const bool use_fw = hdcp->config.ddc.funcs.atomic_write_poll_read_i2c
+ && hdcp->config.ddc.funcs.atomic_write_poll_read_aux
+ && !hdcp->connection.link.adjust.hdcp2.force_sw_locality_check;
+
+ if (event_ctx->event != MOD_HDCP_EVENT_CALLBACK) {
+ event_ctx->unexpected_event = 1;
+ goto out;
+ }
+
+ if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp2_prepare_lc_init,
+ &input->lc_init_prepare, &status,
+ hdcp, "lc_init_prepare"))
+ goto out;
+
+ status = (use_fw ? locality_check_fw : locality_check_sw)(hdcp, event_ctx, input);
+ if (status != MOD_HDCP_STATUS_SUCCESS)
+ goto out;
+
if (!mod_hdcp_execute_and_set(mod_hdcp_hdcp2_validate_l_prime,
&input->l_prime_validation, &status,
hdcp, "l_prime_validation"))
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_transition.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_transition.c
index c5f6c11de7e5..89ffb89e1932 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_transition.c
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp2_transition.c
@@ -184,17 +184,28 @@ enum mod_hdcp_status mod_hdcp_hdcp2_transition(struct mod_hdcp *hdcp,
callback_in_ms(0, output);
set_state_id(hdcp, output, H2_A2_LOCALITY_CHECK);
break;
- case H2_A2_LOCALITY_CHECK:
+ case H2_A2_LOCALITY_CHECK: {
+ const bool use_fw = hdcp->config.ddc.funcs.atomic_write_poll_read_i2c
+ && !adjust->hdcp2.force_sw_locality_check;
+
+ /*
+ * 1A-05: consider disconnection after LC init a failure
+ * 1A-13-1: consider invalid l' a failure
+ * 1A-13-2: consider l' timeout a failure
+ */
if (hdcp->state.stay_count > 10 ||
input->lc_init_prepare != PASS ||
- input->lc_init_write != PASS ||
- input->l_prime_available_poll != PASS ||
- input->l_prime_read != PASS) {
- /*
- * 1A-05: consider disconnection after LC init a failure
- * 1A-13-1: consider invalid l' a failure
- * 1A-13-2: consider l' timeout a failure
- */
+ (!use_fw && input->lc_init_write != PASS) ||
+ (!use_fw && input->l_prime_available_poll != PASS)) {
+ fail_and_restart_in_ms(0, &status, output);
+ break;
+ } else if (input->l_prime_read != PASS) {
+ if (use_fw && hdcp->config.debug.lc_enable_sw_fallback) {
+ adjust->hdcp2.force_sw_locality_check = true;
+ callback_in_ms(0, output);
+ break;
+ }
+
fail_and_restart_in_ms(0, &status, output);
break;
} else if (input->l_prime_validation != PASS) {
@@ -205,6 +216,7 @@ enum mod_hdcp_status mod_hdcp_hdcp2_transition(struct mod_hdcp *hdcp,
callback_in_ms(0, output);
set_state_id(hdcp, output, H2_A3_EXCHANGE_KS_AND_TEST_FOR_REPEATER);
break;
+ }
case H2_A3_EXCHANGE_KS_AND_TEST_FOR_REPEATER:
if (input->eks_prepare != PASS ||
input->eks_write != PASS) {
@@ -498,14 +510,25 @@ enum mod_hdcp_status mod_hdcp_hdcp2_dp_transition(struct mod_hdcp *hdcp,
callback_in_ms(0, output);
set_state_id(hdcp, output, D2_A2_LOCALITY_CHECK);
break;
- case D2_A2_LOCALITY_CHECK:
+ case D2_A2_LOCALITY_CHECK: {
+ const bool use_fw = hdcp->config.ddc.funcs.atomic_write_poll_read_aux
+ && !adjust->hdcp2.force_sw_locality_check;
+
if (hdcp->state.stay_count > 10 ||
input->lc_init_prepare != PASS ||
- input->lc_init_write != PASS ||
- input->l_prime_read != PASS) {
+ (!use_fw && input->lc_init_write != PASS)) {
/* 1A-12: consider invalid l' a failure */
fail_and_restart_in_ms(0, &status, output);
break;
+ } else if (input->l_prime_read != PASS) {
+ if (use_fw && hdcp->config.debug.lc_enable_sw_fallback) {
+ adjust->hdcp2.force_sw_locality_check = true;
+ callback_in_ms(0, output);
+ break;
+ }
+
+ fail_and_restart_in_ms(0, &status, output);
+ break;
} else if (input->l_prime_validation != PASS) {
callback_in_ms(0, output);
increment_stay_counter(hdcp);
@@ -514,6 +537,7 @@ enum mod_hdcp_status mod_hdcp_hdcp2_dp_transition(struct mod_hdcp *hdcp,
callback_in_ms(0, output);
set_state_id(hdcp, output, D2_A34_EXCHANGE_KS_AND_TEST_FOR_REPEATER);
break;
+ }
case D2_A34_EXCHANGE_KS_AND_TEST_FOR_REPEATER:
if (input->eks_prepare != PASS ||
input->eks_write != PASS) {
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c
index 6e064e6ae949..2e6408579194 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_ddc.c
@@ -688,3 +688,76 @@ enum mod_hdcp_status mod_hdcp_clear_cp_irq_status(struct mod_hdcp *hdcp)
return MOD_HDCP_STATUS_INVALID_OPERATION;
}
+
+static bool write_stall_read_lc_fw_aux(struct mod_hdcp *hdcp)
+{
+ struct mod_hdcp_message_hdcp2 *hdcp2 = &hdcp->auth.msg.hdcp2;
+
+ struct mod_hdcp_atomic_op_aux write = {
+ hdcp_dpcd_addrs[MOD_HDCP_MESSAGE_ID_WRITE_LC_INIT],
+ hdcp2->lc_init + 1,
+ sizeof(hdcp2->lc_init) - 1,
+ };
+ struct mod_hdcp_atomic_op_aux stall = { 0, NULL, 0, };
+ struct mod_hdcp_atomic_op_aux read = {
+ hdcp_dpcd_addrs[MOD_HDCP_MESSAGE_ID_READ_LC_SEND_L_PRIME],
+ hdcp2->lc_l_prime + 1,
+ sizeof(hdcp2->lc_l_prime) - 1,
+ };
+
+ hdcp2->lc_l_prime[0] = HDCP_2_2_LC_SEND_LPRIME;
+
+ return hdcp->config.ddc.funcs.atomic_write_poll_read_aux(
+ hdcp->config.ddc.handle,
+ &write,
+ &stall,
+ &read,
+ 16 * 1000,
+ 0
+ );
+}
+
+static bool write_poll_read_lc_fw_i2c(struct mod_hdcp *hdcp)
+{
+ struct mod_hdcp_message_hdcp2 *hdcp2 = &hdcp->auth.msg.hdcp2;
+ uint8_t expected_rxstatus[2] = { sizeof(hdcp2->lc_l_prime) };
+
+ hdcp->buf[0] = hdcp_i2c_offsets[MOD_HDCP_MESSAGE_ID_WRITE_LC_INIT];
+ memmove(&hdcp->buf[1], hdcp2->lc_init, sizeof(hdcp2->lc_init));
+
+ struct mod_hdcp_atomic_op_i2c write = {
+ HDCP_I2C_ADDR,
+ 0,
+ hdcp->buf,
+ sizeof(hdcp2->lc_init) + 1,
+ };
+ struct mod_hdcp_atomic_op_i2c poll = {
+ HDCP_I2C_ADDR,
+ hdcp_i2c_offsets[MOD_HDCP_MESSAGE_ID_READ_RXSTATUS],
+ expected_rxstatus,
+ sizeof(expected_rxstatus),
+ };
+ struct mod_hdcp_atomic_op_i2c read = {
+ HDCP_I2C_ADDR,
+ hdcp_i2c_offsets[MOD_HDCP_MESSAGE_ID_READ_LC_SEND_L_PRIME],
+ hdcp2->lc_l_prime,
+ sizeof(hdcp2->lc_l_prime),
+ };
+
+ return hdcp->config.ddc.funcs.atomic_write_poll_read_i2c(
+ hdcp->config.ddc.handle,
+ &write,
+ &poll,
+ &read,
+ 20 * 1000,
+ 6
+ );
+}
+
+enum mod_hdcp_status mod_hdcp_write_poll_read_lc_fw(struct mod_hdcp *hdcp)
+{
+ const bool success = (is_dp_hdcp(hdcp) ? write_stall_read_lc_fw_aux : write_poll_read_lc_fw_i2c)(hdcp);
+
+ return success ? MOD_HDCP_STATUS_SUCCESS : MOD_HDCP_STATUS_DDC_FAILURE;
+}
+
diff --git a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c
index 8c137d7c032e..e58e7b93810b 100644
--- a/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c
+++ b/drivers/gpu/drm/amd/display/modules/hdcp/hdcp_psp.c
@@ -368,6 +368,9 @@ enum mod_hdcp_status mod_hdcp_hdcp1_enable_encryption(struct mod_hdcp *hdcp)
struct mod_hdcp_display *display = get_first_active_display(hdcp);
enum mod_hdcp_status status = MOD_HDCP_STATUS_SUCCESS;
+ if (!display)
+ return MOD_HDCP_STATUS_DISPLAY_NOT_FOUND;
+
mutex_lock(&psp->hdcp_context.mutex);
hdcp_cmd = (struct ta_hdcp_shared_memory *)psp->hdcp_context.context.mem_context.shared_buf;
memset(hdcp_cmd, 0, sizeof(struct ta_hdcp_shared_memory));
diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h
index cc3dc9b589f6..57916ed98c86 100644
--- a/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h
+++ b/drivers/gpu/drm/amd/display/modules/inc/mod_freesync.h
@@ -110,25 +110,6 @@ struct mod_vrr_params {
struct mod_freesync *mod_freesync_create(struct dc *dc);
void mod_freesync_destroy(struct mod_freesync *mod_freesync);
-bool mod_freesync_get_vmin_vmax(struct mod_freesync *mod_freesync,
- const struct dc_stream_state *stream,
- unsigned int *vmin,
- unsigned int *vmax);
-
-bool mod_freesync_get_v_position(struct mod_freesync *mod_freesync,
- struct dc_stream_state *stream,
- unsigned int *nom_v_pos,
- unsigned int *v_pos);
-
-void mod_freesync_get_settings(struct mod_freesync *mod_freesync,
- const struct mod_vrr_params *vrr,
- unsigned int *v_total_min, unsigned int *v_total_max,
- unsigned int *event_triggers,
- unsigned int *window_min, unsigned int *window_max,
- unsigned int *lfc_mid_point_in_us,
- unsigned int *inserted_frames,
- unsigned int *inserted_duration_in_us);
-
void mod_freesync_build_vrr_infopacket(struct mod_freesync *mod_freesync,
const struct dc_stream_state *stream,
const struct mod_vrr_params *vrr,
@@ -155,13 +136,6 @@ void mod_freesync_handle_v_update(struct mod_freesync *mod_freesync,
unsigned long long mod_freesync_calc_nominal_field_rate(
const struct dc_stream_state *stream);
-unsigned long long mod_freesync_calc_field_rate_from_timing(
- unsigned int vtotal, unsigned int htotal, unsigned int pix_clk);
-
-bool mod_freesync_is_valid_range(uint32_t min_refresh_cap_in_uhz,
- uint32_t max_refresh_cap_in_uhz,
- uint32_t nominal_field_rate_in_uhz);
-
unsigned int mod_freesync_calc_v_total_from_refresh(
const struct dc_stream_state *stream,
unsigned int refresh_in_uhz);
diff --git a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
index a4d344a4db9e..c42468bb70ac 100644
--- a/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
+++ b/drivers/gpu/drm/amd/display/modules/inc/mod_hdcp.h
@@ -133,9 +133,22 @@ enum mod_hdcp_display_disable_option {
MOD_HDCP_DISPLAY_DISABLE_ENCRYPTION,
};
+struct mod_hdcp_atomic_op_i2c {
+ uint8_t address;
+ uint8_t offset;
+ uint8_t *data;
+ uint32_t size;
+};
+
+struct mod_hdcp_atomic_op_aux {
+ uint32_t address;
+ uint8_t *data;
+ uint32_t size;
+};
+
struct mod_hdcp_ddc {
void *handle;
- struct {
+ struct mod_hdcp_ddc_funcs {
bool (*read_i2c)(void *handle,
uint32_t address,
uint8_t offset,
@@ -153,6 +166,22 @@ struct mod_hdcp_ddc {
uint32_t address,
const uint8_t *data,
uint32_t size);
+ bool (*atomic_write_poll_read_i2c)(
+ void *handle,
+ const struct mod_hdcp_atomic_op_i2c *write,
+ const struct mod_hdcp_atomic_op_i2c *poll,
+ struct mod_hdcp_atomic_op_i2c *read,
+ uint32_t poll_timeout_us,
+ uint8_t poll_mask_msb
+ );
+ bool (*atomic_write_poll_read_aux)(
+ void *handle,
+ const struct mod_hdcp_atomic_op_aux *write,
+ const struct mod_hdcp_atomic_op_aux *poll,
+ struct mod_hdcp_atomic_op_aux *read,
+ uint32_t poll_timeout_us,
+ uint8_t poll_mask_msb
+ );
} funcs;
};
@@ -185,7 +214,8 @@ struct mod_hdcp_link_adjustment_hdcp2 {
uint8_t force_type : 2;
uint8_t force_no_stored_km : 1;
uint8_t increase_h_prime_timeout: 1;
- uint8_t reserved : 3;
+ uint8_t force_sw_locality_check : 1;
+ uint8_t reserved : 2;
};
struct mod_hdcp_link_adjustment {
@@ -272,6 +302,10 @@ struct mod_hdcp_display_query {
struct mod_hdcp_config {
struct mod_hdcp_psp psp;
struct mod_hdcp_ddc ddc;
+ struct {
+ uint8_t lc_enable_sw_fallback : 1;
+ uint8_t reserved : 7;
+ } debug;
uint8_t index;
};
diff --git a/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c b/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c
index a344e2e49b0e..b3d55cac3569 100644
--- a/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c
+++ b/drivers/gpu/drm/amd/display/modules/info_packet/info_packet.c
@@ -383,10 +383,10 @@ void mod_build_vsc_infopacket(const struct dc_stream_state *stream,
colorimetryFormat = ColorimetryYCC_DP_ITU709;
else if (cs == COLOR_SPACE_ADOBERGB)
colorimetryFormat = ColorimetryYCC_DP_AdobeYCC;
- else if (cs == COLOR_SPACE_2020_YCBCR)
+ else if (cs == COLOR_SPACE_2020_YCBCR_LIMITED)
colorimetryFormat = ColorimetryYCC_DP_ITU2020YCbCr;
- if (cs == COLOR_SPACE_2020_YCBCR && tf == TRANSFER_FUNC_GAMMA_22)
+ if (cs == COLOR_SPACE_2020_YCBCR_LIMITED && tf == TRANSFER_FUNC_GAMMA_22)
colorimetryFormat = ColorimetryYCC_DP_ITU709;
break;
diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c
index 95838c7ab054..29ccd3532d13 100644
--- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c
+++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c
@@ -996,9 +996,9 @@ void set_replay_coasting_vtotal(struct dc_link *link,
link->replay_settings.coasting_vtotal_table[type] = vtotal;
}
-void set_replay_ips_full_screen_video_src_vtotal(struct dc_link *link, uint16_t vtotal)
+void set_replay_low_rr_full_screen_video_src_vtotal(struct dc_link *link, uint16_t vtotal)
{
- link->replay_settings.abm_with_ips_on_full_screen_video_pseudo_vtotal = vtotal;
+ link->replay_settings.low_rr_full_screen_video_pseudo_vtotal = vtotal;
}
void calculate_replay_link_off_frame_count(struct dc_link *link,
@@ -1039,3 +1039,8 @@ bool fill_custom_backlight_caps(unsigned int config_no, struct dm_acpi_atif_back
memcpy(caps->data_points, custom_backlight_profiles[config_no].data_points, data_points_size);
return true;
}
+
+void reset_replay_dsync_error_count(struct dc_link *link)
+{
+ link->replay_settings.replay_desync_error_fail_count = 0;
+}
diff --git a/drivers/gpu/drm/amd/display/modules/power/power_helpers.h b/drivers/gpu/drm/amd/display/modules/power/power_helpers.h
index cac302e8fa10..391209a3bf29 100644
--- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.h
+++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.h
@@ -62,7 +62,7 @@ void set_replay_defer_update_coasting_vtotal(struct dc_link *link,
uint32_t vtotal);
void update_replay_coasting_vtotal_from_defer(struct dc_link *link,
enum replay_coasting_vtotal_type type);
-void set_replay_ips_full_screen_video_src_vtotal(struct dc_link *link, uint16_t vtotal);
+void set_replay_low_rr_full_screen_video_src_vtotal(struct dc_link *link, uint16_t vtotal);
void calculate_replay_link_off_frame_count(struct dc_link *link,
uint16_t vtotal, uint16_t htotal);
@@ -78,4 +78,7 @@ bool psr_su_set_dsc_slice_height(struct dc *dc, struct dc_link *link,
bool fill_custom_backlight_caps(unsigned int config_no,
struct dm_acpi_atif_backlight_caps *caps);
+void reset_replay_dsync_error_count(struct dc_link *link);
+void change_replay_to_psr(struct dc_link *link);
+void change_psr_to_replay(struct dc_link *link);
#endif /* MODULES_POWER_POWER_HELPERS_H_ */