diff options
| author | Takashi Iwai <tiwai@suse.de> | 2024-11-18 16:34:54 +0300 |
|---|---|---|
| committer | Takashi Iwai <tiwai@suse.de> | 2024-11-18 16:34:54 +0300 |
| commit | 1a7585c3a4504705a97c1560ff67d589b693115d (patch) | |
| tree | 8938dbb5a270dddae76e2075c2a42e31008c4714 /drivers/gpu | |
| parent | 9b4662d0df9f4433f1828904ba5e8733c1ad5158 (diff) | |
| parent | 82ff5abc2edcfba0c0f1a1be807795e2876f46e9 (diff) | |
| download | linux-1a7585c3a4504705a97c1560ff67d589b693115d.tar.xz | |
Merge tag 'asoc-v6.13' of https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-next
ASoC: Updates for v6.13
This release was mainly about new drivers, there's a very large batch of
new drivers and devices including quite a few from newer vendors which
is great to see. Other than the new drivers and the usual routine fixes
and enhancements the bulk of the work has been Morimoto-san's continuing
work on simplifiying APIs, plus a few other bits:
- More API simplifications from Morimoto-san.
- Renaming of the sh directory to Renesas to reflect the focus on other
architectures.
- Factoring out of some of the common code for Realtek devices.
- Support for Allwinner H616, AMD ACP 6.3 systems, AWInic AW88081,
Cirrus Logic CS32L84, Everest ES8328, Iron Devices SMA1307, Longsoon
I2S, NeoFidelity NTP8918 and NTP8835, Philips UDA1342, Qualcomm
SM8750, RealTek RT721, and ST Microelectronics STM32MP25.
Diffstat (limited to 'drivers/gpu')
73 files changed, 804 insertions, 349 deletions
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c index f85ace0384d2..7dd55ed57c1d 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_acpi.c @@ -147,6 +147,7 @@ static union acpi_object *amdgpu_atif_call(struct amdgpu_atif *atif, struct acpi_buffer *params) { acpi_status status; + union acpi_object *obj; union acpi_object atif_arg_elements[2]; struct acpi_object_list atif_arg; struct acpi_buffer buffer = { ACPI_ALLOCATE_BUFFER, NULL }; @@ -169,16 +170,24 @@ static union acpi_object *amdgpu_atif_call(struct amdgpu_atif *atif, status = acpi_evaluate_object(atif->handle, NULL, &atif_arg, &buffer); + obj = (union acpi_object *)buffer.pointer; - /* Fail only if calling the method fails and ATIF is supported */ - if (ACPI_FAILURE(status) && status != AE_NOT_FOUND) { + /* Fail if calling the method fails */ + if (ACPI_FAILURE(status)) { DRM_DEBUG_DRIVER("failed to evaluate ATIF got %s\n", acpi_format_exception(status)); - kfree(buffer.pointer); + kfree(obj); return NULL; } - return buffer.pointer; + if (obj->type != ACPI_TYPE_BUFFER) { + DRM_DEBUG_DRIVER("bad object returned from ATIF: %d\n", + obj->type); + kfree(obj); + return NULL; + } + + return obj; } /** diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c index cbef720de779..9da4414de617 100644 --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_debugfs.c @@ -402,7 +402,7 @@ static ssize_t amdgpu_debugfs_gprwave_read(struct file *f, char __user *buf, siz int r; uint32_t *data, x; - if (size & 0x3 || *pos & 0x3) + if (size > 4096 || size & 0x3 || *pos & 0x3) return -EINVAL; r = pm_runtime_get_sync(adev_to_drm(adev)->dev); @@ -1648,7 +1648,7 @@ int amdgpu_debugfs_regs_init(struct amdgpu_device *adev) for (i = 0; i < ARRAY_SIZE(debugfs_regs); i++) { ent = debugfs_create_file(debugfs_regs_names[i], - S_IFREG | 0444, root, + S_IFREG | 0400, root, adev, debugfs_regs[i]); if (!i && !IS_ERR_OR_NULL(ent)) i_size_write(ent->d_inode, adev->rmmio_size); @@ -2100,11 +2100,11 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev) amdgpu_securedisplay_debugfs_init(adev); amdgpu_fw_attestation_debugfs_init(adev); - debugfs_create_file("amdgpu_evict_vram", 0444, root, adev, + debugfs_create_file("amdgpu_evict_vram", 0400, root, adev, &amdgpu_evict_vram_fops); - debugfs_create_file("amdgpu_evict_gtt", 0444, root, adev, + debugfs_create_file("amdgpu_evict_gtt", 0400, root, adev, &amdgpu_evict_gtt_fops); - debugfs_create_file("amdgpu_test_ib", 0444, root, adev, + debugfs_create_file("amdgpu_test_ib", 0400, root, adev, &amdgpu_debugfs_test_ib_fops); debugfs_create_file("amdgpu_vm_info", 0444, root, adev, &amdgpu_debugfs_vm_info_fops); diff --git a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c index 5e8833e4fed2..ccfd2a4b4acc 100644 --- a/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c +++ b/drivers/gpu/drm/amd/amdgpu/aqua_vanjaram.c @@ -482,7 +482,7 @@ static bool __aqua_vanjaram_is_valid_mode(struct amdgpu_xcp_mgr *xcp_mgr, case AMDGPU_SPX_PARTITION_MODE: return adev->gmc.num_mem_partitions == 1 && num_xcc > 0; case AMDGPU_DPX_PARTITION_MODE: - return adev->gmc.num_mem_partitions != 8 && (num_xcc % 4) == 0; + return adev->gmc.num_mem_partitions <= 2 && (num_xcc % 4) == 0; case AMDGPU_TPX_PARTITION_MODE: return (adev->gmc.num_mem_partitions == 1 || adev->gmc.num_mem_partitions == 3) && diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c index a8763496aed3..9288f37a3cc5 100644 --- a/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c +++ b/drivers/gpu/drm/amd/amdgpu/sdma_v7_0.c @@ -51,6 +51,12 @@ MODULE_FIRMWARE("amdgpu/sdma_7_0_1.bin"); #define SDMA0_HYP_DEC_REG_END 0x589a #define SDMA1_HYP_DEC_REG_OFFSET 0x20 +/*define for compression field for sdma7*/ +#define SDMA_PKT_CONSTANT_FILL_HEADER_compress_offset 0 +#define SDMA_PKT_CONSTANT_FILL_HEADER_compress_mask 0x00000001 +#define SDMA_PKT_CONSTANT_FILL_HEADER_compress_shift 16 +#define SDMA_PKT_CONSTANT_FILL_HEADER_COMPRESS(x) (((x) & SDMA_PKT_CONSTANT_FILL_HEADER_compress_mask) << SDMA_PKT_CONSTANT_FILL_HEADER_compress_shift) + static const struct amdgpu_hwip_reg_entry sdma_reg_list_7_0[] = { SOC15_REG_ENTRY_STR(GC, 0, regSDMA0_STATUS_REG), SOC15_REG_ENTRY_STR(GC, 0, regSDMA0_STATUS1_REG), @@ -1724,7 +1730,8 @@ static void sdma_v7_0_emit_fill_buffer(struct amdgpu_ib *ib, uint64_t dst_offset, uint32_t byte_count) { - ib->ptr[ib->length_dw++] = SDMA_PKT_COPY_LINEAR_HEADER_OP(SDMA_OP_CONST_FILL); + ib->ptr[ib->length_dw++] = SDMA_PKT_CONSTANT_FILL_HEADER_OP(SDMA_OP_CONST_FILL) | + SDMA_PKT_CONSTANT_FILL_HEADER_COMPRESS(1); ib->ptr[ib->length_dw++] = lower_32_bits(dst_offset); ib->ptr[ib->length_dw++] = upper_32_bits(dst_offset); ib->ptr[ib->length_dw++] = src_data; 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 6b5e2206e687..07e9ce99694f 100644 --- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c +++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c @@ -8374,7 +8374,8 @@ static void manage_dm_interrupts(struct amdgpu_device *adev, 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) { + DC_PSR_VERSION_UNSUPPORTED || + !(adev->flags & AMD_IS_APU)) { timing = &acrtc_state->stream->timing; /* at least 2 frames */ @@ -9428,6 +9429,7 @@ static void amdgpu_dm_commit_streams(struct drm_atomic_state *state, bool mode_set_reset_required = false; u32 i; struct dc_commit_streams_params params = {dc_state->streams, dc_state->stream_count}; + bool set_backlight_level = false; /* Disable writeback */ for_each_old_connector_in_state(state, connector, old_con_state, i) { @@ -9547,6 +9549,7 @@ static void amdgpu_dm_commit_streams(struct drm_atomic_state *state, acrtc->hw_mode = new_crtc_state->mode; crtc->hwmode = new_crtc_state->mode; mode_set_reset_required = true; + set_backlight_level = true; } else if (modereset_required(new_crtc_state)) { drm_dbg_atomic(dev, "Atomic commit: RESET. crtc id %d:[%p]\n", @@ -9598,6 +9601,19 @@ static void amdgpu_dm_commit_streams(struct drm_atomic_state *state, acrtc->otg_inst = status->primary_otg_inst; } } + + /* During boot up and resume the DC layer will reset the panel brightness + * to fix a flicker issue. + * It will cause the dm->actual_brightness is not the current panel brightness + * level. (the dm->brightness is the correct panel level) + * So we set the backlight level with dm->brightness value after set mode + */ + if (set_backlight_level) { + for (i = 0; i < dm->num_of_edps; i++) { + if (dm->backlight_dev[i]) + amdgpu_dm_backlight_set_level(dm, i, dm->brightness[i]); + } + } } static void dm_set_writeback(struct amdgpu_display_manager *dm, 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 069e0195e50a..eea317dcbe8c 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 @@ -44,6 +44,7 @@ #include "dm_helpers.h" #include "ddc_service_types.h" +#include "clk_mgr.h" static u32 edid_extract_panel_id(struct edid *edid) { @@ -1121,6 +1122,8 @@ bool dm_helpers_dp_handle_test_pattern_request( struct pipe_ctx *pipe_ctx = NULL; struct amdgpu_dm_connector *aconnector = link->priv; struct drm_device *dev = aconnector->base.dev; + struct dc_state *dc_state = ctx->dc->current_state; + struct clk_mgr *clk_mgr = ctx->dc->clk_mgr; int i; for (i = 0; i < MAX_PIPES; i++) { @@ -1221,6 +1224,16 @@ bool dm_helpers_dp_handle_test_pattern_request( pipe_ctx->stream->test_pattern.type = test_pattern; pipe_ctx->stream->test_pattern.color_space = test_pattern_color_space; + /* Temp W/A for compliance test failure */ + dc_state->bw_ctx.bw.dcn.clk.p_state_change_support = false; + dc_state->bw_ctx.bw.dcn.clk.dramclk_khz = clk_mgr->dc_mode_softmax_enabled ? + clk_mgr->bw_params->dc_mode_softmax_memclk : clk_mgr->bw_params->max_memclk_mhz; + dc_state->bw_ctx.bw.dcn.clk.idle_dramclk_khz = dc_state->bw_ctx.bw.dcn.clk.dramclk_khz; + ctx->dc->clk_mgr->funcs->update_clocks( + ctx->dc->clk_mgr, + dc_state, + false); + dc_link_dp_set_test_pattern( (struct dc_link *) link, test_pattern, 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 0d8498ab9b23..be8fbb04ad98 100644 --- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c +++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c @@ -3127,7 +3127,9 @@ static enum bp_result bios_parser_get_vram_info( struct atom_data_revision revision; // vram info moved to umc_info for DCN4x - if (info && DATA_TABLES(umc_info)) { + if (dcb->ctx->dce_version >= DCN_VERSION_4_01 && + dcb->ctx->dce_version < DCN_VERSION_MAX && + info && DATA_TABLES(umc_info)) { header = GET_IMAGE(struct atom_common_table_header, DATA_TABLES(umc_info)); 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 11c904ae2958..c4c52173ef22 100644 --- a/drivers/gpu/drm/amd/display/dc/dml2/dml2_policy.c +++ b/drivers/gpu/drm/amd/display/dc/dml2/dml2_policy.c @@ -303,6 +303,7 @@ void build_unoptimized_policy_settings(enum dml_project_id project, struct dml_m if (project == dml_project_dcn35 || project == dml_project_dcn351) { policy->DCCProgrammingAssumesScanDirectionUnknownFinal = false; + policy->EnhancedPrefetchScheduleAccelerationFinal = 0; policy->AllowForPStateChangeOrStutterInVBlankFinal = dml_prefetch_support_uclk_fclk_and_stutter_if_possible; /*new*/ policy->UseOnlyMaxPrefetchModes = 1; } 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 3cd52e7a9c77..95838c7ab054 100644 --- a/drivers/gpu/drm/amd/display/modules/power/power_helpers.c +++ b/drivers/gpu/drm/amd/display/modules/power/power_helpers.c @@ -841,6 +841,8 @@ bool is_psr_su_specific_panel(struct dc_link *link) isPSRSUSupported = false; else if (dpcd_caps->sink_dev_id_str[1] == 0x08 && dpcd_caps->sink_dev_id_str[0] == 0x03) isPSRSUSupported = false; + else if (dpcd_caps->sink_dev_id_str[1] == 0x08 && dpcd_caps->sink_dev_id_str[0] == 0x01) + isPSRSUSupported = false; else if (dpcd_caps->psr_info.force_psrsu_cap == 0x1) isPSRSUSupported = true; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c index 9ad9cf7a9c98..ee1bcfaae3e3 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c +++ b/drivers/gpu/drm/amd/pm/swsmu/amdgpu_smu.c @@ -1234,6 +1234,14 @@ static void smu_init_xgmi_plpd_mode(struct smu_context *smu) } } +static bool smu_is_workload_profile_available(struct smu_context *smu, + u32 profile) +{ + if (profile >= PP_SMC_POWER_PROFILE_COUNT) + return false; + return smu->workload_map && smu->workload_map[profile].valid_mapping; +} + static int smu_sw_init(void *handle) { struct amdgpu_device *adev = (struct amdgpu_device *)handle; @@ -1251,25 +1259,33 @@ static int smu_sw_init(void *handle) smu->watermarks_bitmap = 0; smu->power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; + smu->user_dpm_profile.user_workload_mask = 0; atomic_set(&smu->smu_power.power_gate.vcn_gated, 1); atomic_set(&smu->smu_power.power_gate.jpeg_gated, 1); atomic_set(&smu->smu_power.power_gate.vpe_gated, 1); atomic_set(&smu->smu_power.power_gate.umsch_mm_gated, 1); - smu->workload_prority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT] = 0; - smu->workload_prority[PP_SMC_POWER_PROFILE_FULLSCREEN3D] = 1; - smu->workload_prority[PP_SMC_POWER_PROFILE_POWERSAVING] = 2; - smu->workload_prority[PP_SMC_POWER_PROFILE_VIDEO] = 3; - smu->workload_prority[PP_SMC_POWER_PROFILE_VR] = 4; - smu->workload_prority[PP_SMC_POWER_PROFILE_COMPUTE] = 5; - smu->workload_prority[PP_SMC_POWER_PROFILE_CUSTOM] = 6; - - if (smu->is_apu) - smu->workload_mask = 1 << smu->workload_prority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT]; - else - smu->workload_mask = 1 << smu->workload_prority[PP_SMC_POWER_PROFILE_FULLSCREEN3D]; + smu->workload_priority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT] = 0; + smu->workload_priority[PP_SMC_POWER_PROFILE_FULLSCREEN3D] = 1; + smu->workload_priority[PP_SMC_POWER_PROFILE_POWERSAVING] = 2; + smu->workload_priority[PP_SMC_POWER_PROFILE_VIDEO] = 3; + smu->workload_priority[PP_SMC_POWER_PROFILE_VR] = 4; + smu->workload_priority[PP_SMC_POWER_PROFILE_COMPUTE] = 5; + smu->workload_priority[PP_SMC_POWER_PROFILE_CUSTOM] = 6; + + if (smu->is_apu || + !smu_is_workload_profile_available(smu, PP_SMC_POWER_PROFILE_FULLSCREEN3D)) { + smu->driver_workload_mask = + 1 << smu->workload_priority[PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT]; + } else { + smu->driver_workload_mask = + 1 << smu->workload_priority[PP_SMC_POWER_PROFILE_FULLSCREEN3D]; + smu->default_power_profile_mode = PP_SMC_POWER_PROFILE_FULLSCREEN3D; + } + smu->workload_mask = smu->driver_workload_mask | + smu->user_dpm_profile.user_workload_mask; smu->workload_setting[0] = PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; smu->workload_setting[1] = PP_SMC_POWER_PROFILE_FULLSCREEN3D; smu->workload_setting[2] = PP_SMC_POWER_PROFILE_POWERSAVING; @@ -2339,17 +2355,20 @@ static int smu_switch_power_profile(void *handle, return -EINVAL; if (!en) { - smu->workload_mask &= ~(1 << smu->workload_prority[type]); + smu->driver_workload_mask &= ~(1 << smu->workload_priority[type]); index = fls(smu->workload_mask); index = index > 0 && index <= WORKLOAD_POLICY_MAX ? index - 1 : 0; workload[0] = smu->workload_setting[index]; } else { - smu->workload_mask |= (1 << smu->workload_prority[type]); + smu->driver_workload_mask |= (1 << smu->workload_priority[type]); index = fls(smu->workload_mask); index = index <= WORKLOAD_POLICY_MAX ? index - 1 : 0; workload[0] = smu->workload_setting[index]; } + smu->workload_mask = smu->driver_workload_mask | + smu->user_dpm_profile.user_workload_mask; + if (smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_MANUAL && smu_dpm_ctx->dpm_level != AMD_DPM_FORCED_LEVEL_PERF_DETERMINISM) smu_bump_power_profile_mode(smu, workload, 0); @@ -3040,12 +3059,23 @@ static int smu_set_power_profile_mode(void *handle, uint32_t param_size) { struct smu_context *smu = handle; + int ret; if (!smu->pm_enabled || !smu->adev->pm.dpm_enabled || !smu->ppt_funcs->set_power_profile_mode) return -EOPNOTSUPP; - return smu_bump_power_profile_mode(smu, param, param_size); + if (smu->user_dpm_profile.user_workload_mask & + (1 << smu->workload_priority[param[param_size]])) + return 0; + + smu->user_dpm_profile.user_workload_mask = + (1 << smu->workload_priority[param[param_size]]); + smu->workload_mask = smu->user_dpm_profile.user_workload_mask | + smu->driver_workload_mask; + ret = smu_bump_power_profile_mode(smu, param, param_size); + + return ret; } static int smu_get_fan_control_mode(void *handle, u32 *fan_mode) diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h index b44a185d07e8..d60d9a12a47e 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/amdgpu_smu.h @@ -240,6 +240,7 @@ struct smu_user_dpm_profile { /* user clock state information */ uint32_t clk_mask[SMU_CLK_COUNT]; uint32_t clk_dependency; + uint32_t user_workload_mask; }; #define SMU_TABLE_INIT(tables, table_id, s, a, d) \ @@ -557,7 +558,8 @@ struct smu_context { bool disable_uclk_switch; uint32_t workload_mask; - uint32_t workload_prority[WORKLOAD_POLICY_MAX]; + uint32_t driver_workload_mask; + uint32_t workload_priority[WORKLOAD_POLICY_MAX]; uint32_t workload_setting[WORKLOAD_POLICY_MAX]; uint32_t power_profile_mode; uint32_t default_power_profile_mode; diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0.h index ee457a6f0813..c2fd0a4a13e5 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/pmfw_if/smu14_driver_if_v14_0.h @@ -25,7 +25,7 @@ #define SMU14_DRIVER_IF_V14_0_H //Increment this version if SkuTable_t or BoardTable_t change -#define PPTABLE_VERSION 0x18 +#define PPTABLE_VERSION 0x1B #define NUM_GFXCLK_DPM_LEVELS 16 #define NUM_SOCCLK_DPM_LEVELS 8 @@ -145,7 +145,7 @@ typedef enum { } FEATURE_BTC_e; // Debug Overrides Bitmask -#define DEBUG_OVERRIDE_DISABLE_VOLT_LINK_VCN_FCLK 0x00000001 +#define DEBUG_OVERRIDE_NOT_USE 0x00000001 #define DEBUG_OVERRIDE_DISABLE_VOLT_LINK_DCN_FCLK 0x00000002 #define DEBUG_OVERRIDE_DISABLE_VOLT_LINK_MP0_FCLK 0x00000004 #define DEBUG_OVERRIDE_DISABLE_VOLT_LINK_VCN_DCFCLK 0x00000008 @@ -161,6 +161,7 @@ typedef enum { #define DEBUG_OVERRIDE_ENABLE_SOC_VF_BRINGUP_MODE 0x00002000 #define DEBUG_OVERRIDE_ENABLE_PER_WGP_RESIENCY 0x00004000 #define DEBUG_OVERRIDE_DISABLE_MEMORY_VOLTAGE_SCALING 0x00008000 +#define DEBUG_OVERRIDE_DFLL_BTC_FCW_LOG 0x00010000 // VR Mapping Bit Defines #define VR_MAPPING_VR_SELECT_MASK 0x01 @@ -391,6 +392,21 @@ typedef struct { EccInfo_t EccInfo[24]; } EccInfoTable_t; +#define EPCS_HIGH_POWER 600 +#define EPCS_NORMAL_POWER 450 +#define EPCS_LOW_POWER 300 +#define EPCS_SHORTED_POWER 150 +#define EPCS_NO_BOOTUP 0 + +typedef enum{ + EPCS_SHORTED_LIMIT, + EPCS_LOW_POWER_LIMIT, + EPCS_NORMAL_POWER_LIMIT, + EPCS_HIGH_POWER_LIMIT, + EPCS_NOT_CONFIGURED, + EPCS_STATUS_COUNT, +} EPCS_STATUS_e; + //D3HOT sequences typedef enum { BACO_SEQUENCE, @@ -662,7 +678,7 @@ typedef enum { } PP_GRTAVFS_FW_SEP_FUSE_e; #define PP_NUM_RTAVFS_PWL_ZONES 5 - +#define PP_NUM_PSM_DIDT_PWL_ZONES 3 // VBIOS or PPLIB configures telemetry slope and offset. Only slope expected to be set for SVI3 // Slope Q1.7, Offset Q1.2 @@ -746,10 +762,10 @@ typedef struct { uint16_t Padding; //Frequency changes - int16_t GfxclkFmin; // MHz - int16_t GfxclkFmax; // MHz - uint16_t UclkFmin; // MHz - uint16_t UclkFmax; // MHz + int16_t GfxclkFoffset; + uint16_t Padding1; + uint16_t UclkFmin; + uint16_t UclkFmax; uint16_t FclkFmin; uint16_t FclkFmax; @@ -770,19 +786,23 @@ typedef struct { uint8_t MaxOpTemp; uint8_t AdvancedOdModeEnabled; - uint8_t Padding1[3]; + uint8_t Padding2[3]; uint16_t GfxVoltageFullCtrlMode; uint16_t SocVoltageFullCtrlMode; uint16_t GfxclkFullCtrlMode; uint16_t UclkFullCtrlMode; uint16_t FclkFullCtrlMode; - uint16_t Padding2; + uint16_t Padding3; int16_t GfxEdc; int16_t GfxPccLimitControl; - uint32_t Spare[10]; + uint16_t GfxclkFmaxVmax; + uint8_t GfxclkFmaxVmaxTemperature; + uint8_t Padding4[1]; + + uint32_t Spare[9]; uint32_t MmHubPadding[8]; // SMU internal use. Adding here instead of external as a workaround } OverDriveTable_t; @@ -802,8 +822,8 @@ typedef struct { uint16_t VddSocVmax; //gfxclk - int16_t GfxclkFmin; // MHz - int16_t GfxclkFmax; // MHz + int16_t GfxclkFoffset; + uint16_t Padding; //uclk uint16_t UclkFmin; // MHz uint16_t UclkFmax; // MHz @@ -828,7 +848,7 @@ typedef struct { uint8_t FanZeroRpmEnable; //temperature uint8_t MaxOpTemp; - uint8_t Padding[2]; + uint8_t Padding1[2]; //Full Ctrl uint16_t GfxVoltageFullCtrlMode; @@ -839,7 +859,7 @@ typedef struct { //EDC int16_t GfxEdc; int16_t GfxPccLimitControl; - int16_t Padding1; + int16_t Padding2; uint32_t Spare[5]; } OverDriveLimits_t; @@ -987,8 +1007,9 @@ typedef struct { uint16_t BaseClockDc; uint16_t GameClockDc; uint16_t BoostClockDc; - - uint32_t Reserved[4]; + uint16_t MaxReportedClock; + uint16_t Padding; + uint32_t Reserved[3]; } DriverReportedClocks_t; typedef struct { @@ -1132,7 +1153,7 @@ typedef struct { uint32_t DcModeMaxFreq [PPCLK_COUNT ]; // In MHz uint16_t GfxclkAibFmax; - uint16_t GfxclkFreqCap; + uint16_t GfxDpmPadding; //GFX Idle Power Settings uint16_t GfxclkFgfxoffEntry; // Entry in RLC stage (PLL), in Mhz @@ -1172,8 +1193,7 @@ typedef struct { uint32_t DvoFmaxLowScaler; //Unitless float // GFX DCS - uint16_t DcsGfxOffVoltage; //Voltage in mV(Q2) applied to VDDGFX when entering DCS GFXOFF phase - uint16_t PaddingDcs; + uint32_t PaddingDcs; uint16_t DcsMinGfxOffTime; //Minimum amount of time PMFW shuts GFX OFF as part of GFX DCS phase uint16_t DcsMaxGfxOffTime; //Maximum amount of time PMFW can shut GFX OFF as part of GFX DCS phase at a stretch. @@ -1205,8 +1225,7 @@ typedef struct { uint16_t DalDcModeMaxUclkFreq; uint8_t PaddingsMem[2]; //FCLK Section - uint16_t FclkDpmDisallowPstateFreq; //Frequency which FW will target when indicated that display config cannot support P-state. Set to 0 use FW calculated value - uint16_t PaddingFclk; + uint32_t PaddingFclk; // Link DPM Settings uint8_t PcieGenSpeed[NUM_LINK_LEVELS]; ///< 0:PciE-gen1 1:PciE-gen2 2:PciE-gen3 3:PciE-gen4 4:PciE-gen5 @@ -1215,12 +1234,19 @@ typedef struct { // SECTION: VDD_GFX AVFS uint8_t OverrideGfxAvfsFuses; - uint8_t GfxAvfsPadding[3]; + uint8_t GfxAvfsPadding[1]; + uint16_t DroopGBStDev; uint32_t SocHwRtAvfsFuses[PP_GRTAVFS_HW_FUSE_COUNT]; //new added for Soc domain uint32_t GfxL2HwRtAvfsFuses[PP_GRTAVFS_HW_FUSE_COUNT]; //see fusedoc for encoding //uint32_t GfxSeHwRtAvfsFuses[PP_GRTAVFS_HW_FUSE_COUNT]; - uint32_t spare_HwRtAvfsFuses[PP_GRTAVFS_HW_FUSE_COUNT]; + + uint16_t PsmDidt_Vcross[PP_NUM_PSM_DIDT_PWL_ZONES-1]; + uint32_t PsmDidt_StaticDroop_A[PP_NUM_PSM_DIDT_PWL_ZONES]; + uint32_t PsmDidt_StaticDroop_B[PP_NUM_PSM_DIDT_PWL_ZONES]; + uint32_t PsmDidt_DynDroop_A[PP_NUM_PSM_DIDT_PWL_ZONES]; + uint32_t PsmDidt_DynDroop_B[PP_NUM_PSM_DIDT_PWL_ZONES]; + uint32_t spare_HwRtAvfsFuses[19]; uint32_t SocCommonRtAvfs[PP_GRTAVFS_FW_COMMON_FUSE_COUNT]; uint32_t GfxCommonRtAvfs[PP_GRTAVFS_FW_COMMON_FUSE_COUNT]; @@ -1246,11 +1272,7 @@ typedef struct { uint32_t dGbV_dT_vmin; uint32_t dGbV_dT_vmax; - //Unused: PMFW-9370 - uint32_t V2F_vmin_range_low; - uint32_t V2F_vmin_range_high; - uint32_t V2F_vmax_range_low; - uint32_t V2F_vmax_range_high; + uint32_t PaddingV2F[4]; AvfsDcBtcParams_t DcBtcGfxParams; QuadraticInt_t SSCurve_GFX; @@ -1327,18 +1349,18 @@ typedef struct { uint16_t PsmDidtReleaseTimer; uint32_t PsmDidtStallPattern; //Will be written to both pattern 1 and didt_static_level_prog // CAC EDC - uint32_t Leakage_C0; // in IEEE float - uint32_t Leakage_C1; // in IEEE float - uint32_t Leakage_C2; // in IEEE float - uint32_t Leakage_C3; // in IEEE float - uint32_t Leakage_C4; // in IEEE float - uint32_t Leakage_C5; // in IEEE float - uint32_t GFX_CLK_SCALAR; // in IEEE float - uint32_t GFX_CLK_INTERCEPT; // in IEEE float - uint32_t GFX_CAC_M; // in IEEE float - uint32_t GFX_CAC_B; // in IEEE float - uint32_t VDD_GFX_CurrentLimitGuardband; // in IEEE float - uint32_t DynToTotalCacScalar; // in IEEE + uint32_t CacEdcCacLeakageC0; + uint32_t CacEdcCacLeakageC1; + uint32_t CacEdcCacLeakageC2; + uint32_t CacEdcCacLeakageC3; + uint32_t CacEdcCacLeakageC4; + uint32_t CacEdcCacLeakageC5; + uint32_t CacEdcGfxClkScalar; + uint32_t CacEdcGfxClkIntercept; + uint32_t CacEdcCac_m; + uint32_t CacEdcCac_b; + uint32_t CacEdcCurrLimitGuardband; + uint32_t CacEdcDynToTotalCacRatio; // GFX EDC XVMIN uint32_t XVmin_Gfx_EdcThreshScalar; uint32_t XVmin_Gfx_EdcEnableFreq; @@ -1467,7 +1489,7 @@ typedef struct { uint8_t VddqOffEnabled; uint8_t PaddingUmcFlags[2]; - uint32_t PostVoltageSetBacoDelay; // in microseconds. Amount of time FW will wait after power good is established or PSI0 command is issued + uint32_t Paddign1; uint32_t BacoEntryDelay; // in milliseconds. Amount of time FW will wait to trigger BACO entry after receiving entry notification from OS uint8_t FuseWritePowerMuxPresent; @@ -1530,7 +1552,7 @@ typedef struct { int16_t FuzzyFan_ErrorSetDelta; int16_t FuzzyFan_ErrorRateSetDelta; int16_t FuzzyFan_PwmSetDelta; - uint16_t FuzzyFan_Reserved; + uint16_t FanPadding2; uint16_t FwCtfLimit[TEMP_COUNT]; @@ -1547,9 +1569,10 @@ typedef struct { uint16_t FanSpare[1]; uint8_t FanIntakeSensorSupport; uint8_t FanIntakePadding; - uint32_t FanAmbientPerfBoostThreshold; uint32_t FanSpare2[12]; + uint32_t ODFeatureCtrlMask; + uint16_t TemperatureLimit_Hynix; // In degrees Celsius. Memory temperature limit associated with Hynix uint16_t TemperatureLimit_Micron; // In degrees Celsius. Memory temperature limit associated with Micron uint16_t TemperatureFwCtfLimit_Hynix; @@ -1637,7 +1660,7 @@ typedef struct { uint16_t AverageDclk0Frequency ; uint16_t AverageVclk1Frequency ; uint16_t AverageDclk1Frequency ; - uint16_t PCIeBusy ; + uint16_t AveragePCIeBusy ; uint16_t dGPU_W_MAX ; uint16_t padding ; @@ -1665,12 +1688,12 @@ typedef struct { uint16_t AverageGfxActivity ; uint16_t AverageUclkActivity ; - uint16_t Vcn0ActivityPercentage ; + uint16_t AverageVcn0ActivityPercentage; uint16_t Vcn1ActivityPercentage ; uint32_t EnergyAccumulator; uint16_t AverageSocketPower; - uint16_t MovingAverageTotalBoardPower; + uint16_t AverageTotalBoardPower; uint16_t AvgTemperature[TEMP_COUNT]; uint16_t AvgTemperatureFanIntake; @@ -1684,7 +1707,8 @@ typedef struct { uint8_t ThrottlingPercentage[THROTTLER_COUNT]; - uint8_t padding1[3]; + uint8_t VmaxThrottlingPercentage; + uint8_t padding1[2]; //metrics for D3hot entry/exit and driver ARM msgs uint32_t D3HotEntryCountPerMode[D3HOT_SEQUENCE_COUNT]; @@ -1693,7 +1717,7 @@ typedef struct { uint16_t ApuSTAPMSmartShiftLimit; uint16_t ApuSTAPMLimit; - uint16_t MovingAvgApuSocketPower; + uint16_t AvgApuSocketPower; uint16_t AverageUclkActivity_MAX; @@ -1823,6 +1847,17 @@ typedef struct { #define TABLE_TRANSFER_FAILED 0xFF #define TABLE_TRANSFER_PENDING 0xAB +#define TABLE_PPT_FAILED 0x100 +#define TABLE_TDC_FAILED 0x200 +#define TABLE_TEMP_FAILED 0x400 +#define TABLE_FAN_TARGET_TEMP_FAILED 0x800 +#define TABLE_FAN_STOP_TEMP_FAILED 0x1000 +#define TABLE_FAN_START_TEMP_FAILED 0x2000 +#define TABLE_FAN_PWM_MIN_FAILED 0x4000 +#define TABLE_ACOUSTIC_TARGET_RPM_FAILED 0x8000 +#define TABLE_ACOUSTIC_LIMIT_RPM_FAILED 0x10000 +#define TABLE_MGPU_ACOUSTIC_TARGET_RPM_FAILED 0x20000 + // Table types #define TABLE_PPTABLE 0 #define TABLE_COMBO_PPTABLE 1 @@ -1849,5 +1884,6 @@ typedef struct { #define IH_INTERRUPT_CONTEXT_ID_THERMAL_THROTTLING 0x7 #define IH_INTERRUPT_CONTEXT_ID_FAN_ABNORMAL 0x8 #define IH_INTERRUPT_CONTEXT_ID_FAN_RECOVERY 0x9 +#define IH_INTERRUPT_CONTEXT_ID_DYNAMIC_TABLE 0xA #endif diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h index 46b456590a08..727d5b405435 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h +++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v14_0.h @@ -28,7 +28,7 @@ #define SMU14_DRIVER_IF_VERSION_INV 0xFFFFFFFF #define SMU14_DRIVER_IF_VERSION_SMU_V14_0_0 0x7 #define SMU14_DRIVER_IF_VERSION_SMU_V14_0_1 0x6 -#define SMU14_DRIVER_IF_VERSION_SMU_V14_0_2 0x26 +#define SMU14_DRIVER_IF_VERSION_SMU_V14_0_2 0x2E #define FEATURE_MASK(feature) (1ULL << feature) diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c index c0f6b59369b7..31fe512028f4 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/arcturus_ppt.c @@ -1455,7 +1455,6 @@ static int arcturus_set_power_profile_mode(struct smu_context *smu, return -EINVAL; } - if ((profile_mode == PP_SMC_POWER_PROFILE_CUSTOM) && (smu->smc_fw_version >= 0x360d00)) { if (size != 10) @@ -1523,14 +1522,14 @@ static int arcturus_set_power_profile_mode(struct smu_context *smu, ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - 1 << workload_type, + smu->workload_mask, NULL); if (ret) { dev_err(smu->adev->dev, "Fail to set workload type %d\n", workload_type); return ret; } - smu->power_profile_mode = profile_mode; + smu_cmn_assign_power_profile(smu); return 0; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c index 16af1a329621..12223f507977 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/navi10_ppt.c @@ -2081,10 +2081,13 @@ static int navi10_set_power_profile_mode(struct smu_context *smu, long *input, u smu->power_profile_mode); if (workload_type < 0) return -EINVAL; + ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - 1 << workload_type, NULL); + smu->workload_mask, NULL); if (ret) dev_err(smu->adev->dev, "[%s] Failed to set work load mask!", __func__); + else + smu_cmn_assign_power_profile(smu); return ret; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c index 9c3c48297cba..3b7b2ec8319a 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/sienna_cichlid_ppt.c @@ -1786,10 +1786,13 @@ static int sienna_cichlid_set_power_profile_mode(struct smu_context *smu, long * smu->power_profile_mode); if (workload_type < 0) return -EINVAL; + ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - 1 << workload_type, NULL); + smu->workload_mask, NULL); if (ret) dev_err(smu->adev->dev, "[%s] Failed to set work load mask!", __func__); + else + smu_cmn_assign_power_profile(smu); return ret; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c index 22737b11b1bf..952ee22cbc90 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu11/vangogh_ppt.c @@ -242,7 +242,9 @@ static int vangogh_tables_init(struct smu_context *smu) goto err0_out; smu_table->metrics_time = 0; - smu_table->gpu_metrics_table_size = max(sizeof(struct gpu_metrics_v2_3), sizeof(struct gpu_metrics_v2_2)); + smu_table->gpu_metrics_table_size = sizeof(struct gpu_metrics_v2_2); + smu_table->gpu_metrics_table_size = max(smu_table->gpu_metrics_table_size, sizeof(struct gpu_metrics_v2_3)); + smu_table->gpu_metrics_table_size = max(smu_table->gpu_metrics_table_size, sizeof(struct gpu_metrics_v2_4)); smu_table->gpu_metrics_table = kzalloc(smu_table->gpu_metrics_table_size, GFP_KERNEL); if (!smu_table->gpu_metrics_table) goto err1_out; @@ -1077,7 +1079,7 @@ static int vangogh_set_power_profile_mode(struct smu_context *smu, long *input, } ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_ActiveProcessNotify, - 1 << workload_type, + smu->workload_mask, NULL); if (ret) { dev_err_once(smu->adev->dev, "Fail to set workload type %d\n", @@ -1085,7 +1087,7 @@ static int vangogh_set_power_profile_mode(struct smu_context *smu, long *input, return ret; } - smu->power_profile_mode = profile_mode; + smu_cmn_assign_power_profile(smu); return 0; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c index cc0504b063fa..62316a6707ef 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu12/renoir_ppt.c @@ -890,14 +890,14 @@ static int renoir_set_power_profile_mode(struct smu_context *smu, long *input, u } ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_ActiveProcessNotify, - 1 << workload_type, + smu->workload_mask, NULL); if (ret) { dev_err_once(smu->adev->dev, "Fail to set workload type %d\n", workload_type); return ret; } - smu->power_profile_mode = profile_mode; + smu_cmn_assign_power_profile(smu); return 0; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c index cb923e33fd6f..5dd7ceca64fe 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c @@ -2567,12 +2567,22 @@ static int smu_v13_0_0_set_power_profile_mode(struct smu_context *smu, workload_mask |= 1 << workload_type; } + smu->workload_mask |= workload_mask; ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - workload_mask, + smu->workload_mask, NULL); - if (!ret) - smu->workload_mask = workload_mask; + if (!ret) { + smu_cmn_assign_power_profile(smu); + if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_POWERSAVING) { + workload_type = smu_cmn_to_asic_specific_index(smu, + CMN2ASIC_MAPPING_WORKLOAD, + PP_SMC_POWER_PROFILE_FULLSCREEN3D); + smu->power_profile_mode = smu->workload_mask & (1 << workload_type) + ? PP_SMC_POWER_PROFILE_FULLSCREEN3D + : PP_SMC_POWER_PROFILE_BOOTUP_DEFAULT; + } + } return ret; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c index b891a5e0a396..9d0b19419de0 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c @@ -2499,13 +2499,14 @@ static int smu_v13_0_7_set_power_profile_mode(struct smu_context *smu, long *inp smu->power_profile_mode); if (workload_type < 0) return -EINVAL; + ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, - 1 << workload_type, NULL); + smu->workload_mask, NULL); if (ret) dev_err(smu->adev->dev, "[%s] Failed to set work load mask!", __func__); else - smu->workload_mask = (1 << workload_type); + smu_cmn_assign_power_profile(smu); return ret; } diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c index 5899d01fa73d..1aa13d32ceb2 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu14/smu_v14_0_2_ppt.c @@ -367,54 +367,6 @@ static int smu_v14_0_2_store_powerplay_table(struct smu_context *smu) return 0; } -#ifndef atom_smc_dpm_info_table_14_0_0 -struct atom_smc_dpm_info_table_14_0_0 { - struct atom_common_table_header table_header; - BoardTable_t BoardTable; -}; -#endif - -static int smu_v14_0_2_append_powerplay_table(struct smu_context *smu) -{ - struct smu_table_context *table_context = &smu->smu_table; - PPTable_t *smc_pptable = table_context->driver_pptable; - struct atom_smc_dpm_info_table_14_0_0 *smc_dpm_table; - BoardTable_t *BoardTable = &smc_pptable->BoardTable; - int index, ret; - - index = get_index_into_master_table(atom_master_list_of_data_tables_v2_1, - smc_dpm_info); - - ret = amdgpu_atombios_get_data_table(smu->adev, index, NULL, NULL, NULL, - (uint8_t **)&smc_dpm_table); - if (ret) - return ret; - - memcpy(BoardTable, &smc_dpm_table->BoardTable, sizeof(BoardTable_t)); - - return 0; -} - -#if 0 -static int smu_v14_0_2_get_pptable_from_pmfw(struct smu_context *smu, - void **table, - uint32_t *size) -{ - struct smu_table_context *smu_table = &smu->smu_table; - void *combo_pptable = smu_table->combo_pptable; - int ret = 0; - - ret = smu_cmn_get_combo_pptable(smu); - if (ret) - return ret; - - *table = combo_pptable; - *size = sizeof(struct smu_14_0_powerplay_table); - - return 0; -} -#endif - static int smu_v14_0_2_get_pptable_from_pmfw(struct smu_context *smu, void **table, uint32_t *size) @@ -436,16 +388,12 @@ static int smu_v14_0_2_get_pptable_from_pmfw(struct smu_context *smu, static int smu_v14_0_2_setup_pptable(struct smu_context *smu) { struct smu_table_context *smu_table = &smu->smu_table; - struct amdgpu_device *adev = smu->adev; int ret = 0; if (amdgpu_sriov_vf(smu->adev)) return 0; - if (!adev->scpm_enabled) - ret = smu_v14_0_setup_pptable(smu); - else - ret = smu_v14_0_2_get_pptable_from_pmfw(smu, + ret = smu_v14_0_2_get_pptable_from_pmfw(smu, &smu_table->power_play_table, &smu_table->power_play_table_size); if (ret) @@ -455,16 +403,6 @@ static int smu_v14_0_2_setup_pptable(struct smu_context *smu) if (ret) return ret; - /* - * With SCPM enabled, the operation below will be handled - * by PSP. Driver involvment is unnecessary and useless. - */ - if (!adev->scpm_enabled) { - ret = smu_v14_0_2_append_powerplay_table(smu); - if (ret) - return ret; - } - ret = smu_v14_0_2_check_powerplay_table(smu); if (ret) return ret; @@ -1077,12 +1015,9 @@ static void smu_v14_0_2_get_od_setting_limits(struct smu_context *smu, switch (od_feature_bit) { case PP_OD_FEATURE_GFXCLK_FMIN: - od_min_setting = overdrive_lowerlimits->GfxclkFmin; - od_max_setting = overdrive_upperlimits->GfxclkFmin; - break; case PP_OD_FEATURE_GFXCLK_FMAX: - od_min_setting = overdrive_lowerlimits->GfxclkFmax; - od_max_setting = overdrive_upperlimits->GfxclkFmax; + od_min_setting = overdrive_lowerlimits->GfxclkFoffset; + od_max_setting = overdrive_upperlimits->GfxclkFoffset; break; case PP_OD_FEATURE_UCLK_FMIN: od_min_setting = overdrive_lowerlimits->UclkFmin; @@ -1269,10 +1204,16 @@ static int smu_v14_0_2_print_clk_levels(struct smu_context *smu, PP_OD_FEATURE_GFXCLK_BIT)) break; - size += sysfs_emit_at(buf, size, "OD_SCLK:\n"); - size += sysfs_emit_at(buf, size, "0: %uMhz\n1: %uMhz\n", - od_table->OverDriveTable.GfxclkFmin, - od_table->OverDriveTable.GfxclkFmax); + PPTable_t *pptable = smu->smu_table.driver_pptable; + const OverDriveLimits_t * const overdrive_upperlimits = + &pptable->SkuTable.OverDriveLimitsBasicMax; + const OverDriveLimits_t * const overdrive_lowerlimits = + &pptable->SkuTable.OverDriveLimitsBasicMin; + + size += sysfs_emit_at(buf, size, "OD_SCLK_OFFSET:\n"); + size += sysfs_emit_at(buf, size, "0: %dMhz\n1: %uMhz\n", + overdrive_lowerlimits->GfxclkFoffset, + overdrive_upperlimits->GfxclkFoffset); break; case SMU_OD_MCLK: @@ -1414,7 +1355,7 @@ static int smu_v14_0_2_print_clk_levels(struct smu_context *smu, PP_OD_FEATURE_GFXCLK_FMAX, NULL, &max_value); - size += sysfs_emit_at(buf, size, "SCLK: %7uMhz %10uMhz\n", + size += sysfs_emit_at(buf, size, "SCLK_OFFSET: %7dMhz %10uMhz\n", min_value, max_value); } @@ -1796,7 +1737,7 @@ static int smu_v14_0_2_set_power_profile_mode(struct smu_context *smu, DpmActivityMonitorCoeffInt_t *activity_monitor = &(activity_monitor_external.DpmActivityMonitorCoeffInt); int workload_type, ret = 0; - + uint32_t current_profile_mode = smu->power_profile_mode; smu->power_profile_mode = input[size]; if (smu->power_profile_mode >= PP_SMC_POWER_PROFILE_COUNT) { @@ -1854,6 +1795,11 @@ static int smu_v14_0_2_set_power_profile_mode(struct smu_context *smu, } } + if (smu->power_profile_mode == PP_SMC_POWER_PROFILE_COMPUTE) + smu_v14_0_deep_sleep_control(smu, false); + else if (current_profile_mode == PP_SMC_POWER_PROFILE_COMPUTE) + smu_v14_0_deep_sleep_control(smu, true); + /* conv PP_SMC_POWER_PROFILE* to WORKLOAD_PPLIB_*_BIT */ workload_type = smu_cmn_to_asic_specific_index(smu, CMN2ASIC_MAPPING_WORKLOAD, @@ -1861,12 +1807,11 @@ static int smu_v14_0_2_set_power_profile_mode(struct smu_context *smu, if (workload_type < 0) return -EINVAL; - ret = smu_cmn_send_smc_msg_with_param(smu, - SMU_MSG_SetWorkloadMask, - 1 << workload_type, - NULL); + ret = smu_cmn_send_smc_msg_with_param(smu, SMU_MSG_SetWorkloadMask, + smu->workload_mask, NULL); + if (!ret) - smu->workload_mask = 1 << workload_type; + smu_cmn_assign_power_profile(smu); return ret; } @@ -2158,7 +2103,7 @@ static ssize_t smu_v14_0_2_get_gpu_metrics(struct smu_context *smu, gpu_metrics->average_gfx_activity = metrics->AverageGfxActivity; gpu_metrics->average_umc_activity = metrics->AverageUclkActivity; - gpu_metrics->average_mm_activity = max(metrics->Vcn0ActivityPercentage, + gpu_metrics->average_mm_activity = max(metrics->AverageVcn0ActivityPercentage, metrics->Vcn1ActivityPercentage); gpu_metrics->average_socket_power = metrics->AverageSocketPower; @@ -2217,8 +2162,7 @@ static void smu_v14_0_2_dump_od_table(struct smu_context *smu, { struct amdgpu_device *adev = smu->adev; - dev_dbg(adev->dev, "OD: Gfxclk: (%d, %d)\n", od_table->OverDriveTable.GfxclkFmin, - od_table->OverDriveTable.GfxclkFmax); + dev_dbg(adev->dev, "OD: Gfxclk offset: (%d)\n", od_table->OverDriveTable.GfxclkFoffset); dev_dbg(adev->dev, "OD: Uclk: (%d, %d)\n", od_table->OverDriveTable.UclkFmin, od_table->OverDriveTable.UclkFmax); } @@ -2309,10 +2253,8 @@ static int smu_v14_0_2_set_default_od_settings(struct smu_context *smu) memcpy(user_od_table, boot_od_table, sizeof(OverDriveTableExternal_t)); - user_od_table->OverDriveTable.GfxclkFmin = - user_od_table_bak.OverDriveTable.GfxclkFmin; - user_od_table->OverDriveTable.GfxclkFmax = - user_od_table_bak.OverDriveTable.GfxclkFmax; + user_od_table->OverDriveTable.GfxclkFoffset = + user_od_table_bak.OverDriveTable.GfxclkFoffset; user_od_table->OverDriveTable.UclkFmin = user_od_table_bak.OverDriveTable.UclkFmin; user_od_table->OverDriveTable.UclkFmax = @@ -2441,22 +2383,6 @@ static int smu_v14_0_2_od_edit_dpm_table(struct smu_context *smu, } switch (input[i]) { - case 0: - smu_v14_0_2_get_od_setting_limits(smu, - PP_OD_FEATURE_GFXCLK_FMIN, - &minimum, - &maximum); - if (input[i + 1] < minimum || - input[i + 1] > maximum) { - dev_info(adev->dev, "GfxclkFmin (%ld) must be within [%u, %u]!\n", - input[i + 1], minimum, maximum); - return -EINVAL; - } - - od_table->OverDriveTable.GfxclkFmin = input[i + 1]; - od_table->OverDriveTable.FeatureCtrlMask |= 1U << PP_OD_FEATURE_GFXCLK_BIT; - break; - case 1: smu_v14_0_2_get_od_setting_limits(smu, PP_OD_FEATURE_GFXCLK_FMAX, @@ -2469,7 +2395,7 @@ static int smu_v14_0_2_od_edit_dpm_table(struct smu_context *smu, return -EINVAL; } - od_table->OverDriveTable.GfxclkFmax = input[i + 1]; + od_table->OverDriveTable.GfxclkFoffset = input[i + 1]; od_table->OverDriveTable.FeatureCtrlMask |= 1U << PP_OD_FEATURE_GFXCLK_BIT; break; @@ -2480,13 +2406,6 @@ static int smu_v14_0_2_od_edit_dpm_table(struct smu_context *smu, } } - if (od_table->OverDriveTable.GfxclkFmin > od_table->OverDriveTable.GfxclkFmax) { - dev_err(adev->dev, - "Invalid setting: GfxclkFmin(%u) is bigger than GfxclkFmax(%u)\n", - (uint32_t)od_table->OverDriveTable.GfxclkFmin, - (uint32_t)od_table->OverDriveTable.GfxclkFmax); - return -EINVAL; - } break; case PP_OD_EDIT_MCLK_VDDC_TABLE: @@ -2817,7 +2736,6 @@ static const struct pptable_funcs smu_v14_0_2_ppt_funcs = { .check_fw_status = smu_v14_0_check_fw_status, .setup_pptable = smu_v14_0_2_setup_pptable, .check_fw_version = smu_v14_0_check_fw_version, - .write_pptable = smu_cmn_write_pptable, .set_driver_table_location = smu_v14_0_set_driver_table_location, .system_features_control = smu_v14_0_system_features_control, .set_allowed_mask = smu_v14_0_set_allowed_mask, diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c index 91ad434bcdae..bdfc5e617333 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.c @@ -1138,6 +1138,14 @@ int smu_cmn_set_mp1_state(struct smu_context *smu, return ret; } +void smu_cmn_assign_power_profile(struct smu_context *smu) +{ + uint32_t index; + index = fls(smu->workload_mask); + index = index > 0 && index <= WORKLOAD_POLICY_MAX ? index - 1 : 0; + smu->power_profile_mode = smu->workload_setting[index]; +} + bool smu_cmn_is_audio_func_enabled(struct amdgpu_device *adev) { struct pci_dev *p = NULL; diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h index 1de685defe85..8a801e389659 100644 --- a/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h +++ b/drivers/gpu/drm/amd/pm/swsmu/smu_cmn.h @@ -130,6 +130,8 @@ void smu_cmn_init_soft_gpu_metrics(void *table, uint8_t frev, uint8_t crev); int smu_cmn_set_mp1_state(struct smu_context *smu, enum pp_mp1_state mp1_state); +void smu_cmn_assign_power_profile(struct smu_context *smu); + /* * Helper function to make sysfs_emit_at() happy. Align buf to * the current page boundary and record the offset. diff --git a/drivers/gpu/drm/bridge/aux-bridge.c b/drivers/gpu/drm/bridge/aux-bridge.c index b29980f95379..295e9d031e2d 100644 --- a/drivers/gpu/drm/bridge/aux-bridge.c +++ b/drivers/gpu/drm/bridge/aux-bridge.c @@ -58,9 +58,10 @@ int drm_aux_bridge_register(struct device *parent) adev->id = ret; adev->name = "aux_bridge"; adev->dev.parent = parent; - adev->dev.of_node = of_node_get(parent->of_node); adev->dev.release = drm_aux_bridge_release; + device_set_of_node_from_dev(&adev->dev, parent); + ret = auxiliary_device_init(adev); if (ret) { ida_free(&drm_aux_bridge_ida, adev->id); diff --git a/drivers/gpu/drm/bridge/tc358767.c b/drivers/gpu/drm/bridge/tc358767.c index 290e2532fab1..f3afdab55c11 100644 --- a/drivers/gpu/drm/bridge/tc358767.c +++ b/drivers/gpu/drm/bridge/tc358767.c @@ -2391,6 +2391,7 @@ static int tc_probe_bridge_endpoint(struct tc_data *tc) if (tc->pre_emphasis[0] < 0 || tc->pre_emphasis[0] > 2 || tc->pre_emphasis[1] < 0 || tc->pre_emphasis[1] > 2) { dev_err(dev, "Incorrect Pre-Emphasis setting, use either 0=0dB 1=3.5dB 2=6dB\n"); + of_node_put(node); return -EINVAL; } } diff --git a/drivers/gpu/drm/drm_panel_orientation_quirks.c b/drivers/gpu/drm/drm_panel_orientation_quirks.c index 0830cae9a4d0..2d84d7ea1ab7 100644 --- a/drivers/gpu/drm/drm_panel_orientation_quirks.c +++ b/drivers/gpu/drm/drm_panel_orientation_quirks.c @@ -403,7 +403,6 @@ static const struct dmi_system_id orientation_data[] = { }, { /* Lenovo Yoga Tab 3 X90F */ .matches = { DMI_MATCH(DMI_SYS_VENDOR, "Intel Corporation"), - DMI_MATCH(DMI_PRODUCT_NAME, "CHERRYVIEW D1 PLATFORM"), DMI_MATCH(DMI_PRODUCT_VERSION, "Blade3-10A-001"), }, .driver_data = (void *)&lcd1600x2560_rightside_up, diff --git a/drivers/gpu/drm/i915/Kconfig b/drivers/gpu/drm/i915/Kconfig index faa253b27664..14ac351fd76d 100644 --- a/drivers/gpu/drm/i915/Kconfig +++ b/drivers/gpu/drm/i915/Kconfig @@ -123,9 +123,8 @@ config DRM_I915_USERPTR config DRM_I915_GVT_KVMGT tristate "Enable KVM host support Intel GVT-g graphics virtualization" depends on DRM_I915 - depends on X86 + depends on KVM_X86 depends on 64BIT - depends on KVM depends on VFIO select DRM_I915_GVT select KVM_EXTERNAL_WRITE_TRACKING diff --git a/drivers/gpu/drm/imagination/pvr_context.c b/drivers/gpu/drm/imagination/pvr_context.c index eded5e955cc0..4cb3494c0bb2 100644 --- a/drivers/gpu/drm/imagination/pvr_context.c +++ b/drivers/gpu/drm/imagination/pvr_context.c @@ -17,10 +17,14 @@ #include <drm/drm_auth.h> #include <drm/drm_managed.h> + +#include <linux/bug.h> #include <linux/errno.h> #include <linux/kernel.h> +#include <linux/list.h> #include <linux/sched.h> #include <linux/slab.h> +#include <linux/spinlock.h> #include <linux/string.h> #include <linux/types.h> #include <linux/xarray.h> @@ -354,6 +358,10 @@ int pvr_context_create(struct pvr_file *pvr_file, struct drm_pvr_ioctl_create_co return err; } + spin_lock(&pvr_dev->ctx_list_lock); + list_add_tail(&ctx->file_link, &pvr_file->contexts); + spin_unlock(&pvr_dev->ctx_list_lock); + return 0; err_destroy_fw_obj: @@ -380,6 +388,11 @@ pvr_context_release(struct kref *ref_count) container_of(ref_count, struct pvr_context, ref_count); struct pvr_device *pvr_dev = ctx->pvr_dev; + WARN_ON(in_interrupt()); + spin_lock(&pvr_dev->ctx_list_lock); + list_del(&ctx->file_link); + spin_unlock(&pvr_dev->ctx_list_lock); + xa_erase(&pvr_dev->ctx_ids, ctx->ctx_id); pvr_context_destroy_queues(ctx); pvr_fw_object_destroy(ctx->fw_obj); @@ -437,11 +450,30 @@ pvr_context_destroy(struct pvr_file *pvr_file, u32 handle) */ void pvr_destroy_contexts_for_file(struct pvr_file *pvr_file) { + struct pvr_device *pvr_dev = pvr_file->pvr_dev; struct pvr_context *ctx; unsigned long handle; xa_for_each(&pvr_file->ctx_handles, handle, ctx) pvr_context_destroy(pvr_file, handle); + + spin_lock(&pvr_dev->ctx_list_lock); + ctx = list_first_entry(&pvr_file->contexts, struct pvr_context, file_link); + + while (!list_entry_is_head(ctx, &pvr_file->contexts, file_link)) { + list_del_init(&ctx->file_link); + + if (pvr_context_get_if_referenced(ctx)) { + spin_unlock(&pvr_dev->ctx_list_lock); + + pvr_vm_unmap_all(ctx->vm_ctx); + + pvr_context_put(ctx); + spin_lock(&pvr_dev->ctx_list_lock); + } + ctx = list_first_entry(&pvr_file->contexts, struct pvr_context, file_link); + } + spin_unlock(&pvr_dev->ctx_list_lock); } /** @@ -451,6 +483,7 @@ void pvr_destroy_contexts_for_file(struct pvr_file *pvr_file) void pvr_context_device_init(struct pvr_device *pvr_dev) { xa_init_flags(&pvr_dev->ctx_ids, XA_FLAGS_ALLOC1); + spin_lock_init(&pvr_dev->ctx_list_lock); } /** diff --git a/drivers/gpu/drm/imagination/pvr_context.h b/drivers/gpu/drm/imagination/pvr_context.h index 0c7b97dfa6ba..07afa179cdf4 100644 --- a/drivers/gpu/drm/imagination/pvr_context.h +++ b/drivers/gpu/drm/imagination/pvr_context.h @@ -85,6 +85,9 @@ struct pvr_context { /** @compute: Transfer queue. */ struct pvr_queue *transfer; } queues; + + /** @file_link: pvr_file PVR context list link. */ + struct list_head file_link; }; static __always_inline struct pvr_queue * @@ -124,6 +127,24 @@ pvr_context_get(struct pvr_context *ctx) } /** + * pvr_context_get_if_referenced() - Take an additional reference on a still + * referenced context. + * @ctx: Context pointer. + * + * Call pvr_context_put() to release. + * + * Returns: + * * True on success, or + * * false if no context pointer passed, or the context wasn't still + * * referenced. + */ +static __always_inline bool +pvr_context_get_if_referenced(struct pvr_context *ctx) +{ + return ctx != NULL && kref_get_unless_zero(&ctx->ref_count) != 0; +} + +/** * pvr_context_lookup() - Lookup context pointer from handle and file. * @pvr_file: Pointer to pvr_file structure. * @handle: Context handle. diff --git a/drivers/gpu/drm/imagination/pvr_device.h b/drivers/gpu/drm/imagination/pvr_device.h index b574e23d484b..6d0dfacb677b 100644 --- a/drivers/gpu/drm/imagination/pvr_device.h +++ b/drivers/gpu/drm/imagination/pvr_device.h @@ -23,6 +23,7 @@ #include <linux/kernel.h> #include <linux/math.h> #include <linux/mutex.h> +#include <linux/spinlock_types.h> #include <linux/timer.h> #include <linux/types.h> #include <linux/wait.h> @@ -293,6 +294,12 @@ struct pvr_device { /** @sched_wq: Workqueue for schedulers. */ struct workqueue_struct *sched_wq; + + /** + * @ctx_list_lock: Lock to be held when accessing the context list in + * struct pvr_file. + */ + spinlock_t ctx_list_lock; }; /** @@ -344,6 +351,9 @@ struct pvr_file { * This array is used to allocate handles returned to userspace. */ struct xarray vm_ctx_handles; + + /** @contexts: PVR context list. */ + struct list_head contexts; }; /** diff --git a/drivers/gpu/drm/imagination/pvr_drv.c b/drivers/gpu/drm/imagination/pvr_drv.c index 1a0cb7aa9cea..fb17196e05f4 100644 --- a/drivers/gpu/drm/imagination/pvr_drv.c +++ b/drivers/gpu/drm/imagination/pvr_drv.c @@ -28,6 +28,7 @@ #include <linux/export.h> #include <linux/fs.h> #include <linux/kernel.h> +#include <linux/list.h> #include <linux/mod_devicetable.h> #include <linux/module.h> #include <linux/moduleparam.h> @@ -1326,6 +1327,8 @@ pvr_drm_driver_open(struct drm_device *drm_dev, struct drm_file *file) */ pvr_file->pvr_dev = pvr_dev; + INIT_LIST_HEAD(&pvr_file->contexts); + xa_init_flags(&pvr_file->ctx_handles, XA_FLAGS_ALLOC1); xa_init_flags(&pvr_file->free_list_handles, XA_FLAGS_ALLOC1); xa_init_flags(&pvr_file->hwrt_handles, XA_FLAGS_ALLOC1); diff --git a/drivers/gpu/drm/imagination/pvr_vm.c b/drivers/gpu/drm/imagination/pvr_vm.c index 97c0f772ed65..7bd6ba4c6e8a 100644 --- a/drivers/gpu/drm/imagination/pvr_vm.c +++ b/drivers/gpu/drm/imagination/pvr_vm.c @@ -14,6 +14,7 @@ #include <drm/drm_gem.h> #include <drm/drm_gpuvm.h> +#include <linux/bug.h> #include <linux/container_of.h> #include <linux/err.h> #include <linux/errno.h> @@ -597,12 +598,26 @@ err_free: } /** - * pvr_vm_context_release() - Teardown a VM context. - * @ref_count: Pointer to reference counter of the VM context. + * pvr_vm_unmap_all() - Unmap all mappings associated with a VM context. + * @vm_ctx: Target VM context. * * This function ensures that no mappings are left dangling by unmapping them * all in order of ascending device-virtual address. */ +void +pvr_vm_unmap_all(struct pvr_vm_context *vm_ctx) +{ + WARN_ON(pvr_vm_unmap(vm_ctx, vm_ctx->gpuvm_mgr.mm_start, + vm_ctx->gpuvm_mgr.mm_range)); +} + +/** + * pvr_vm_context_release() - Teardown a VM context. + * @ref_count: Pointer to reference counter of the VM context. + * + * This function also ensures that no mappings are left dangling by calling + * pvr_vm_unmap_all. + */ static void pvr_vm_context_release(struct kref *ref_count) { @@ -612,8 +627,7 @@ pvr_vm_context_release(struct kref *ref_count) if (vm_ctx->fw_mem_ctx_obj) pvr_fw_object_destroy(vm_ctx->fw_mem_ctx_obj); - WARN_ON(pvr_vm_unmap(vm_ctx, vm_ctx->gpuvm_mgr.mm_start, - vm_ctx->gpuvm_mgr.mm_range)); + pvr_vm_unmap_all(vm_ctx); pvr_mmu_context_destroy(vm_ctx->mmu_ctx); drm_gem_private_object_fini(&vm_ctx->dummy_gem); diff --git a/drivers/gpu/drm/imagination/pvr_vm.h b/drivers/gpu/drm/imagination/pvr_vm.h index f2a6463f2b05..79406243617c 100644 --- a/drivers/gpu/drm/imagination/pvr_vm.h +++ b/drivers/gpu/drm/imagination/pvr_vm.h @@ -39,6 +39,7 @@ int pvr_vm_map(struct pvr_vm_context *vm_ctx, struct pvr_gem_object *pvr_obj, u64 pvr_obj_offset, u64 device_addr, u64 size); int pvr_vm_unmap(struct pvr_vm_context *vm_ctx, u64 device_addr, u64 size); +void pvr_vm_unmap_all(struct pvr_vm_context *vm_ctx); dma_addr_t pvr_vm_get_page_table_root_addr(struct pvr_vm_context *vm_ctx); struct dma_resv *pvr_vm_get_dma_resv(struct pvr_vm_context *vm_ctx); diff --git a/drivers/gpu/drm/mediatek/mtk_crtc.c b/drivers/gpu/drm/mediatek/mtk_crtc.c index 175b00e5a253..eb0e1233ad04 100644 --- a/drivers/gpu/drm/mediatek/mtk_crtc.c +++ b/drivers/gpu/drm/mediatek/mtk_crtc.c @@ -127,9 +127,8 @@ static void mtk_crtc_destroy(struct drm_crtc *crtc) mtk_mutex_put(mtk_crtc->mutex); #if IS_REACHABLE(CONFIG_MTK_CMDQ) - cmdq_pkt_destroy(&mtk_crtc->cmdq_client, &mtk_crtc->cmdq_handle); - if (mtk_crtc->cmdq_client.chan) { + cmdq_pkt_destroy(&mtk_crtc->cmdq_client, &mtk_crtc->cmdq_handle); mbox_free_channel(mtk_crtc->cmdq_client.chan); mtk_crtc->cmdq_client.chan = NULL; } @@ -913,6 +912,7 @@ static int mtk_crtc_init_comp_planes(struct drm_device *drm_dev, BIT(pipe), mtk_crtc_plane_type(mtk_crtc->layer_nr, num_planes), mtk_ddp_comp_supported_rotations(comp), + mtk_ddp_comp_get_blend_modes(comp), mtk_ddp_comp_get_formats(comp), mtk_ddp_comp_get_num_formats(comp), i); if (ret) diff --git a/drivers/gpu/drm/mediatek/mtk_ddp_comp.c b/drivers/gpu/drm/mediatek/mtk_ddp_comp.c index be66d94be361..edc6417639e6 100644 --- a/drivers/gpu/drm/mediatek/mtk_ddp_comp.c +++ b/drivers/gpu/drm/mediatek/mtk_ddp_comp.c @@ -363,6 +363,7 @@ static const struct mtk_ddp_comp_funcs ddp_ovl = { .layer_config = mtk_ovl_layer_config, .bgclr_in_on = mtk_ovl_bgclr_in_on, .bgclr_in_off = mtk_ovl_bgclr_in_off, + .get_blend_modes = mtk_ovl_get_blend_modes, .get_formats = mtk_ovl_get_formats, .get_num_formats = mtk_ovl_get_num_formats, }; @@ -416,6 +417,7 @@ static const struct mtk_ddp_comp_funcs ddp_ovl_adaptor = { .disconnect = mtk_ovl_adaptor_disconnect, .add = mtk_ovl_adaptor_add_comp, .remove = mtk_ovl_adaptor_remove_comp, + .get_blend_modes = mtk_ovl_adaptor_get_blend_modes, .get_formats = mtk_ovl_adaptor_get_formats, .get_num_formats = mtk_ovl_adaptor_get_num_formats, .mode_valid = mtk_ovl_adaptor_mode_valid, diff --git a/drivers/gpu/drm/mediatek/mtk_ddp_comp.h b/drivers/gpu/drm/mediatek/mtk_ddp_comp.h index ecf6dc283cd7..39720b27f4e9 100644 --- a/drivers/gpu/drm/mediatek/mtk_ddp_comp.h +++ b/drivers/gpu/drm/mediatek/mtk_ddp_comp.h @@ -80,6 +80,7 @@ struct mtk_ddp_comp_funcs { void (*ctm_set)(struct device *dev, struct drm_crtc_state *state); struct device * (*dma_dev_get)(struct device *dev); + u32 (*get_blend_modes)(struct device *dev); const u32 *(*get_formats)(struct device *dev); size_t (*get_num_formats)(struct device *dev); void (*connect)(struct device *dev, struct device *mmsys_dev, unsigned int next); @@ -267,6 +268,15 @@ static inline struct device *mtk_ddp_comp_dma_dev_get(struct mtk_ddp_comp *comp) } static inline +u32 mtk_ddp_comp_get_blend_modes(struct mtk_ddp_comp *comp) +{ + if (comp->funcs && comp->funcs->get_blend_modes) + return comp->funcs->get_blend_modes(comp->dev); + + return 0; +} + +static inline const u32 *mtk_ddp_comp_get_formats(struct mtk_ddp_comp *comp) { if (comp->funcs && comp->funcs->get_formats) diff --git a/drivers/gpu/drm/mediatek/mtk_disp_drv.h b/drivers/gpu/drm/mediatek/mtk_disp_drv.h index 082ac18fe04a..04154db9085c 100644 --- a/drivers/gpu/drm/mediatek/mtk_disp_drv.h +++ b/drivers/gpu/drm/mediatek/mtk_disp_drv.h @@ -103,6 +103,7 @@ void mtk_ovl_register_vblank_cb(struct device *dev, void mtk_ovl_unregister_vblank_cb(struct device *dev); void mtk_ovl_enable_vblank(struct device *dev); void mtk_ovl_disable_vblank(struct device *dev); +u32 mtk_ovl_get_blend_modes(struct device *dev); const u32 *mtk_ovl_get_formats(struct device *dev); size_t mtk_ovl_get_num_formats(struct device *dev); @@ -131,6 +132,7 @@ void mtk_ovl_adaptor_start(struct device *dev); void mtk_ovl_adaptor_stop(struct device *dev); unsigned int mtk_ovl_adaptor_layer_nr(struct device *dev); struct device *mtk_ovl_adaptor_dma_dev_get(struct device *dev); +u32 mtk_ovl_adaptor_get_blend_modes(struct device *dev); const u32 *mtk_ovl_adaptor_get_formats(struct device *dev); size_t mtk_ovl_adaptor_get_num_formats(struct device *dev); enum drm_mode_status mtk_ovl_adaptor_mode_valid(struct device *dev, diff --git a/drivers/gpu/drm/mediatek/mtk_disp_ovl.c b/drivers/gpu/drm/mediatek/mtk_disp_ovl.c index 89b439dcf3a6..e0c0bb01f65a 100644 --- a/drivers/gpu/drm/mediatek/mtk_disp_ovl.c +++ b/drivers/gpu/drm/mediatek/mtk_disp_ovl.c @@ -65,8 +65,8 @@ #define OVL_CON_CLRFMT_RGB (1 << 12) #define OVL_CON_CLRFMT_ARGB8888 (2 << 12) #define OVL_CON_CLRFMT_RGBA8888 (3 << 12) -#define OVL_CON_CLRFMT_ABGR8888 (OVL_CON_CLRFMT_RGBA8888 | OVL_CON_BYTE_SWAP) -#define OVL_CON_CLRFMT_BGRA8888 (OVL_CON_CLRFMT_ARGB8888 | OVL_CON_BYTE_SWAP) +#define OVL_CON_CLRFMT_ABGR8888 (OVL_CON_CLRFMT_ARGB8888 | OVL_CON_BYTE_SWAP) +#define OVL_CON_CLRFMT_BGRA8888 (OVL_CON_CLRFMT_RGBA8888 | OVL_CON_BYTE_SWAP) #define OVL_CON_CLRFMT_UYVY (4 << 12) #define OVL_CON_CLRFMT_YUYV (5 << 12) #define OVL_CON_MTX_YUV_TO_RGB (6 << 16) @@ -146,6 +146,7 @@ struct mtk_disp_ovl_data { bool fmt_rgb565_is_0; bool smi_id_en; bool supports_afbc; + const u32 blend_modes; const u32 *formats; size_t num_formats; bool supports_clrfmt_ext; @@ -214,6 +215,13 @@ void mtk_ovl_disable_vblank(struct device *dev) writel_relaxed(0x0, ovl->regs + DISP_REG_OVL_INTEN); } +u32 mtk_ovl_get_blend_modes(struct device *dev) +{ + struct mtk_disp_ovl *ovl = dev_get_drvdata(dev); + + return ovl->data->blend_modes; +} + const u32 *mtk_ovl_get_formats(struct device *dev) { struct mtk_disp_ovl *ovl = dev_get_drvdata(dev); @@ -386,14 +394,27 @@ void mtk_ovl_layer_off(struct device *dev, unsigned int idx, DISP_REG_OVL_RDMA_CTRL(idx)); } -static unsigned int ovl_fmt_convert(struct mtk_disp_ovl *ovl, unsigned int fmt, - unsigned int blend_mode) +static unsigned int mtk_ovl_fmt_convert(struct mtk_disp_ovl *ovl, + struct mtk_plane_state *state) { - /* The return value in switch "MEM_MODE_INPUT_FORMAT_XXX" - * is defined in mediatek HW data sheet. - * The alphabet order in XXX is no relation to data - * arrangement in memory. + unsigned int fmt = state->pending.format; + unsigned int blend_mode = DRM_MODE_BLEND_COVERAGE; + + /* + * For the platforms where OVL_CON_CLRFMT_MAN is defined in the hardware data sheet + * and supports premultiplied color formats, such as OVL_CON_CLRFMT_PARGB8888. + * + * Check blend_modes in the driver data to see if premultiplied mode is supported. + * If not, use coverage mode instead to set it to the supported color formats. + * + * Current DRM assumption is that alpha is default premultiplied, so the bitmask of + * blend_modes must include BIT(DRM_MODE_BLEND_PREMULTI). Otherwise, mtk_plane_init() + * will get an error return from drm_plane_create_blend_mode_property() and + * state->base.pixel_blend_mode should not be used. */ + if (ovl->data->blend_modes & BIT(DRM_MODE_BLEND_PREMULTI)) + blend_mode = state->base.pixel_blend_mode; + switch (fmt) { default: case DRM_FORMAT_RGB565: @@ -471,20 +492,26 @@ void mtk_ovl_layer_config(struct device *dev, unsigned int idx, return; } - con = ovl_fmt_convert(ovl, fmt, blend_mode); + con = mtk_ovl_fmt_convert(ovl, state); if (state->base.fb) { - con |= OVL_CON_AEN; con |= state->base.alpha & OVL_CON_ALPHA; - } - /* CONST_BLD must be enabled for XRGB formats although the alpha channel - * can be ignored, or OVL will still read the value from memory. - * For RGB888 related formats, whether CONST_BLD is enabled or not won't - * affect the result. Therefore we use !has_alpha as the condition. - */ - if ((state->base.fb && !state->base.fb->format->has_alpha) || - blend_mode == DRM_MODE_BLEND_PIXEL_NONE) - ignore_pixel_alpha = OVL_CONST_BLEND; + /* + * For blend_modes supported SoCs, always enable alpha blending. + * For blend_modes unsupported SoCs, enable alpha blending when has_alpha is set. + */ + if (blend_mode || state->base.fb->format->has_alpha) + con |= OVL_CON_AEN; + + /* + * Although the alpha channel can be ignored, CONST_BLD must be enabled + * for XRGB format, otherwise OVL will still read the value from memory. + * For RGB888 related formats, whether CONST_BLD is enabled or not won't + * affect the result. Therefore we use !has_alpha as the condition. + */ + if (blend_mode == DRM_MODE_BLEND_PIXEL_NONE || !state->base.fb->format->has_alpha) + ignore_pixel_alpha = OVL_CONST_BLEND; + } if (pending->rotation & DRM_MODE_REFLECT_Y) { con |= OVL_CON_VIRT_FLIP; @@ -663,6 +690,9 @@ static const struct mtk_disp_ovl_data mt8192_ovl_driver_data = { .layer_nr = 4, .fmt_rgb565_is_0 = true, .smi_id_en = true, + .blend_modes = BIT(DRM_MODE_BLEND_PREMULTI) | + BIT(DRM_MODE_BLEND_COVERAGE) | + BIT(DRM_MODE_BLEND_PIXEL_NONE), .formats = mt8173_formats, .num_formats = ARRAY_SIZE(mt8173_formats), }; @@ -673,6 +703,9 @@ static const struct mtk_disp_ovl_data mt8192_ovl_2l_driver_data = { .layer_nr = 2, .fmt_rgb565_is_0 = true, .smi_id_en = true, + .blend_modes = BIT(DRM_MODE_BLEND_PREMULTI) | + BIT(DRM_MODE_BLEND_COVERAGE) | + BIT(DRM_MODE_BLEND_PIXEL_NONE), .formats = mt8173_formats, .num_formats = ARRAY_SIZE(mt8173_formats), }; @@ -684,6 +717,9 @@ static const struct mtk_disp_ovl_data mt8195_ovl_driver_data = { .fmt_rgb565_is_0 = true, .smi_id_en = true, .supports_afbc = true, + .blend_modes = BIT(DRM_MODE_BLEND_PREMULTI) | + BIT(DRM_MODE_BLEND_COVERAGE) | + BIT(DRM_MODE_BLEND_PIXEL_NONE), .formats = mt8195_formats, .num_formats = ARRAY_SIZE(mt8195_formats), .supports_clrfmt_ext = true, diff --git a/drivers/gpu/drm/mediatek/mtk_disp_ovl_adaptor.c b/drivers/gpu/drm/mediatek/mtk_disp_ovl_adaptor.c index c6768210b08b..bf2546c4681a 100644 --- a/drivers/gpu/drm/mediatek/mtk_disp_ovl_adaptor.c +++ b/drivers/gpu/drm/mediatek/mtk_disp_ovl_adaptor.c @@ -400,6 +400,13 @@ void mtk_ovl_adaptor_disable_vblank(struct device *dev) mtk_ethdr_disable_vblank(ovl_adaptor->ovl_adaptor_comp[OVL_ADAPTOR_ETHDR0]); } +u32 mtk_ovl_adaptor_get_blend_modes(struct device *dev) +{ + struct mtk_disp_ovl_adaptor *ovl_adaptor = dev_get_drvdata(dev); + + return mtk_ethdr_get_blend_modes(ovl_adaptor->ovl_adaptor_comp[OVL_ADAPTOR_ETHDR0]); +} + const u32 *mtk_ovl_adaptor_get_formats(struct device *dev) { struct mtk_disp_ovl_adaptor *ovl_adaptor = dev_get_drvdata(dev); diff --git a/drivers/gpu/drm/mediatek/mtk_dp.c b/drivers/gpu/drm/mediatek/mtk_dp.c index d8796a904eca..f2bee617f063 100644 --- a/drivers/gpu/drm/mediatek/mtk_dp.c +++ b/drivers/gpu/drm/mediatek/mtk_dp.c @@ -145,6 +145,89 @@ struct mtk_dp_data { u16 audio_m_div2_bit; }; +static const struct mtk_dp_efuse_fmt mt8188_dp_efuse_fmt[MTK_DP_CAL_MAX] = { + [MTK_DP_CAL_GLB_BIAS_TRIM] = { + .idx = 0, + .shift = 10, + .mask = 0x1f, + .min_val = 1, + .max_val = 0x1e, + .default_val = 0xf, + }, + [MTK_DP_CAL_CLKTX_IMPSE] = { + .idx = 0, + .shift = 15, + .mask = 0xf, + .min_val = 1, + .max_val = 0xe, + .default_val = 0x8, + }, + [MTK_DP_CAL_LN_TX_IMPSEL_PMOS_0] = { + .idx = 1, + .shift = 0, + .mask = 0xf, + .min_val = 1, + .max_val = 0xe, + .default_val = 0x8, + }, + [MTK_DP_CAL_LN_TX_IMPSEL_PMOS_1] = { + .idx = 1, + .shift = 8, + .mask = 0xf, + .min_val = 1, + .max_val = 0xe, + .default_val = 0x8, + }, + [MTK_DP_CAL_LN_TX_IMPSEL_PMOS_2] = { + .idx = 1, + .shift = 16, + .mask = 0xf, + .min_val = 1, + .max_val = 0xe, + .default_val = 0x8, + }, + [MTK_DP_CAL_LN_TX_IMPSEL_PMOS_3] = { + .idx = 1, + .shift = 24, + .mask = 0xf, + .min_val = 1, + .max_val = 0xe, + .default_val = 0x8, + }, + [MTK_DP_CAL_LN_TX_IMPSEL_NMOS_0] = { + .idx = 1, + .shift = 4, + .mask = 0xf, + .min_val = 1, + .max_val = 0xe, + .default_val = 0x8, + }, + [MTK_DP_CAL_LN_TX_IMPSEL_NMOS_1] = { + .idx = 1, + .shift = 12, + .mask = 0xf, + .min_val = 1, + .max_val = 0xe, + .default_val = 0x8, + }, + [MTK_DP_CAL_LN_TX_IMPSEL_NMOS_2] = { + .idx = 1, + .shift = 20, + .mask = 0xf, + .min_val = 1, + .max_val = 0xe, + .default_val = 0x8, + }, + [MTK_DP_CAL_LN_TX_IMPSEL_NMOS_3] = { + .idx = 1, + .shift = 28, + .mask = 0xf, + .min_val = 1, + .max_val = 0xe, + .default_val = 0x8, + }, +}; + static const struct mtk_dp_efuse_fmt mt8195_edp_efuse_fmt[MTK_DP_CAL_MAX] = { [MTK_DP_CAL_GLB_BIAS_TRIM] = { .idx = 3, @@ -2771,7 +2854,7 @@ static SIMPLE_DEV_PM_OPS(mtk_dp_pm_ops, mtk_dp_suspend, mtk_dp_resume); static const struct mtk_dp_data mt8188_dp_data = { .bridge_type = DRM_MODE_CONNECTOR_DisplayPort, .smc_cmd = MTK_DP_SIP_ATF_VIDEO_UNMUTE, - .efuse_fmt = mt8195_dp_efuse_fmt, + .efuse_fmt = mt8188_dp_efuse_fmt, .audio_supported = true, .audio_pkt_in_hblank_area = true, .audio_m_div2_bit = MT8188_AUDIO_M_CODE_MULT_DIV_SEL_DP_ENC0_P0_DIV_2, diff --git a/drivers/gpu/drm/mediatek/mtk_ethdr.c b/drivers/gpu/drm/mediatek/mtk_ethdr.c index d1d9cf8b10e1..0f22e7d337cb 100644 --- a/drivers/gpu/drm/mediatek/mtk_ethdr.c +++ b/drivers/gpu/drm/mediatek/mtk_ethdr.c @@ -145,6 +145,13 @@ static irqreturn_t mtk_ethdr_irq_handler(int irq, void *dev_id) return IRQ_HANDLED; } +u32 mtk_ethdr_get_blend_modes(struct device *dev) +{ + return BIT(DRM_MODE_BLEND_PREMULTI) | + BIT(DRM_MODE_BLEND_COVERAGE) | + BIT(DRM_MODE_BLEND_PIXEL_NONE); +} + void mtk_ethdr_layer_config(struct device *dev, unsigned int idx, struct mtk_plane_state *state, struct cmdq_pkt *cmdq_pkt) diff --git a/drivers/gpu/drm/mediatek/mtk_ethdr.h b/drivers/gpu/drm/mediatek/mtk_ethdr.h index 81af9edea3f7..a72aeee46829 100644 --- a/drivers/gpu/drm/mediatek/mtk_ethdr.h +++ b/drivers/gpu/drm/mediatek/mtk_ethdr.h @@ -13,6 +13,7 @@ void mtk_ethdr_clk_disable(struct device *dev); void mtk_ethdr_config(struct device *dev, unsigned int w, unsigned int h, unsigned int vrefresh, unsigned int bpc, struct cmdq_pkt *cmdq_pkt); +u32 mtk_ethdr_get_blend_modes(struct device *dev); void mtk_ethdr_layer_config(struct device *dev, unsigned int idx, struct mtk_plane_state *state, struct cmdq_pkt *cmdq_pkt); diff --git a/drivers/gpu/drm/mediatek/mtk_plane.c b/drivers/gpu/drm/mediatek/mtk_plane.c index 7d2cb4e0fafa..8a48b3b0a956 100644 --- a/drivers/gpu/drm/mediatek/mtk_plane.c +++ b/drivers/gpu/drm/mediatek/mtk_plane.c @@ -320,8 +320,8 @@ static const struct drm_plane_helper_funcs mtk_plane_helper_funcs = { int mtk_plane_init(struct drm_device *dev, struct drm_plane *plane, unsigned long possible_crtcs, enum drm_plane_type type, - unsigned int supported_rotations, const u32 *formats, - size_t num_formats, unsigned int plane_idx) + unsigned int supported_rotations, const u32 blend_modes, + const u32 *formats, size_t num_formats, unsigned int plane_idx) { int err; @@ -366,12 +366,11 @@ int mtk_plane_init(struct drm_device *dev, struct drm_plane *plane, if (err) DRM_ERROR("failed to create property: alpha\n"); - err = drm_plane_create_blend_mode_property(plane, - BIT(DRM_MODE_BLEND_PREMULTI) | - BIT(DRM_MODE_BLEND_COVERAGE) | - BIT(DRM_MODE_BLEND_PIXEL_NONE)); - if (err) - DRM_ERROR("failed to create property: blend_mode\n"); + if (blend_modes) { + err = drm_plane_create_blend_mode_property(plane, blend_modes); + if (err) + DRM_ERROR("failed to create property: blend_mode\n"); + } drm_plane_helper_add(plane, &mtk_plane_helper_funcs); diff --git a/drivers/gpu/drm/mediatek/mtk_plane.h b/drivers/gpu/drm/mediatek/mtk_plane.h index 5b177eac67b7..3b13b89989c7 100644 --- a/drivers/gpu/drm/mediatek/mtk_plane.h +++ b/drivers/gpu/drm/mediatek/mtk_plane.h @@ -48,6 +48,6 @@ to_mtk_plane_state(struct drm_plane_state *state) int mtk_plane_init(struct drm_device *dev, struct drm_plane *plane, unsigned long possible_crtcs, enum drm_plane_type type, - unsigned int supported_rotations, const u32 *formats, - size_t num_formats, unsigned int plane_idx); + unsigned int supported_rotations, const u32 blend_modes, + const u32 *formats, size_t num_formats, unsigned int plane_idx); #endif diff --git a/drivers/gpu/drm/panthor/panthor_device.c b/drivers/gpu/drm/panthor/panthor_device.c index 4082c8f2951d..6fbff516c1c1 100644 --- a/drivers/gpu/drm/panthor/panthor_device.c +++ b/drivers/gpu/drm/panthor/panthor_device.c @@ -390,11 +390,15 @@ int panthor_device_mmap_io(struct panthor_device *ptdev, struct vm_area_struct * { u64 offset = (u64)vma->vm_pgoff << PAGE_SHIFT; + if ((vma->vm_flags & VM_SHARED) == 0) + return -EINVAL; + switch (offset) { case DRM_PANTHOR_USER_FLUSH_ID_MMIO_OFFSET: if (vma->vm_end - vma->vm_start != PAGE_SIZE || (vma->vm_flags & (VM_WRITE | VM_EXEC))) return -EINVAL; + vm_flags_clear(vma, VM_MAYWRITE); break; diff --git a/drivers/gpu/drm/panthor/panthor_fw.c b/drivers/gpu/drm/panthor/panthor_fw.c index ef232c0c2049..4e2d3a02ea06 100644 --- a/drivers/gpu/drm/panthor/panthor_fw.c +++ b/drivers/gpu/drm/panthor/panthor_fw.c @@ -487,6 +487,7 @@ static int panthor_fw_load_section_entry(struct panthor_device *ptdev, struct panthor_fw_binary_iter *iter, u32 ehdr) { + ssize_t vm_pgsz = panthor_vm_page_size(ptdev->fw->vm); struct panthor_fw_binary_section_entry_hdr hdr; struct panthor_fw_section *section; u32 section_size; @@ -515,8 +516,7 @@ static int panthor_fw_load_section_entry(struct panthor_device *ptdev, return -EINVAL; } - if ((hdr.va.start & ~PAGE_MASK) != 0 || - (hdr.va.end & ~PAGE_MASK) != 0) { + if (!IS_ALIGNED(hdr.va.start, vm_pgsz) || !IS_ALIGNED(hdr.va.end, vm_pgsz)) { drm_err(&ptdev->base, "Firmware corrupted, virtual addresses not page aligned: 0x%x-0x%x\n", hdr.va.start, hdr.va.end); return -EINVAL; diff --git a/drivers/gpu/drm/panthor/panthor_gem.c b/drivers/gpu/drm/panthor/panthor_gem.c index 38f560864879..be97d56bc011 100644 --- a/drivers/gpu/drm/panthor/panthor_gem.c +++ b/drivers/gpu/drm/panthor/panthor_gem.c @@ -44,8 +44,7 @@ void panthor_kernel_bo_destroy(struct panthor_kernel_bo *bo) to_panthor_bo(bo->obj)->exclusive_vm_root_gem != panthor_vm_root_gem(vm))) goto out_free_bo; - ret = panthor_vm_unmap_range(vm, bo->va_node.start, - panthor_kernel_bo_size(bo)); + ret = panthor_vm_unmap_range(vm, bo->va_node.start, bo->va_node.size); if (ret) goto out_free_bo; @@ -95,10 +94,16 @@ panthor_kernel_bo_create(struct panthor_device *ptdev, struct panthor_vm *vm, } bo = to_panthor_bo(&obj->base); - size = obj->base.size; kbo->obj = &obj->base; bo->flags = bo_flags; + /* The system and GPU MMU page size might differ, which becomes a + * problem for FW sections that need to be mapped at explicit address + * since our PAGE_SIZE alignment might cover a VA range that's + * expected to be used for another section. + * Make sure we never map more than we need. + */ + size = ALIGN(size, panthor_vm_page_size(vm)); ret = panthor_vm_alloc_va(vm, gpu_va, size, &kbo->va_node); if (ret) goto err_put_obj; diff --git a/drivers/gpu/drm/panthor/panthor_mmu.c b/drivers/gpu/drm/panthor/panthor_mmu.c index 3cd2bce59edc..7db2edb3374c 100644 --- a/drivers/gpu/drm/panthor/panthor_mmu.c +++ b/drivers/gpu/drm/panthor/panthor_mmu.c @@ -826,6 +826,14 @@ void panthor_vm_idle(struct panthor_vm *vm) mutex_unlock(&ptdev->mmu->as.slots_lock); } +u32 panthor_vm_page_size(struct panthor_vm *vm) +{ + const struct io_pgtable *pgt = io_pgtable_ops_to_pgtable(vm->pgtbl_ops); + u32 pg_shift = ffs(pgt->cfg.pgsize_bitmap) - 1; + + return 1u << pg_shift; +} + static void panthor_vm_stop(struct panthor_vm *vm) { drm_sched_stop(&vm->sched, NULL); @@ -1025,12 +1033,13 @@ int panthor_vm_alloc_va(struct panthor_vm *vm, u64 va, u64 size, struct drm_mm_node *va_node) { + ssize_t vm_pgsz = panthor_vm_page_size(vm); int ret; - if (!size || (size & ~PAGE_MASK)) + if (!size || !IS_ALIGNED(size, vm_pgsz)) return -EINVAL; - if (va != PANTHOR_VM_KERNEL_AUTO_VA && (va & ~PAGE_MASK)) + if (va != PANTHOR_VM_KERNEL_AUTO_VA && !IS_ALIGNED(va, vm_pgsz)) return -EINVAL; mutex_lock(&vm->mm_lock); @@ -1571,7 +1580,9 @@ panthor_vm_pool_get_vm(struct panthor_vm_pool *pool, u32 handle) { struct panthor_vm *vm; + xa_lock(&pool->xa); vm = panthor_vm_get(xa_load(&pool->xa, handle)); + xa_unlock(&pool->xa); return vm; } @@ -2366,11 +2377,12 @@ panthor_vm_bind_prepare_op_ctx(struct drm_file *file, const struct drm_panthor_vm_bind_op *op, struct panthor_vm_op_ctx *op_ctx) { + ssize_t vm_pgsz = panthor_vm_page_size(vm); struct drm_gem_object *gem; int ret; /* Aligned on page size. */ - if ((op->va | op->size) & ~PAGE_MASK) + if (!IS_ALIGNED(op->va | op->size, vm_pgsz)) return -EINVAL; switch (op->flags & DRM_PANTHOR_VM_BIND_OP_TYPE_MASK) { diff --git a/drivers/gpu/drm/panthor/panthor_mmu.h b/drivers/gpu/drm/panthor/panthor_mmu.h index 6788771071e3..8d21e83d8aba 100644 --- a/drivers/gpu/drm/panthor/panthor_mmu.h +++ b/drivers/gpu/drm/panthor/panthor_mmu.h @@ -30,6 +30,7 @@ panthor_vm_get_bo_for_va(struct panthor_vm *vm, u64 va, u64 *bo_offset); int panthor_vm_active(struct panthor_vm *vm); void panthor_vm_idle(struct panthor_vm *vm); +u32 panthor_vm_page_size(struct panthor_vm *vm); int panthor_vm_as(struct panthor_vm *vm); int panthor_vm_flush_all(struct panthor_vm *vm); diff --git a/drivers/gpu/drm/panthor/panthor_sched.c b/drivers/gpu/drm/panthor/panthor_sched.c index aee362abb710..9929e22f4d8d 100644 --- a/drivers/gpu/drm/panthor/panthor_sched.c +++ b/drivers/gpu/drm/panthor/panthor_sched.c @@ -589,10 +589,11 @@ struct panthor_group { * @timedout: True when a timeout occurred on any of the queues owned by * this group. * - * Timeouts can be reported by drm_sched or by the FW. In any case, any - * timeout situation is unrecoverable, and the group becomes useless. - * We simply wait for all references to be dropped so we can release the - * group object. + * Timeouts can be reported by drm_sched or by the FW. If a reset is required, + * and the group can't be suspended, this also leads to a timeout. In any case, + * any timeout situation is unrecoverable, and the group becomes useless. We + * simply wait for all references to be dropped so we can release the group + * object. */ bool timedout; @@ -2640,6 +2641,12 @@ void panthor_sched_suspend(struct panthor_device *ptdev) csgs_upd_ctx_init(&upd_ctx); while (slot_mask) { u32 csg_id = ffs(slot_mask) - 1; + struct panthor_csg_slot *csg_slot = &sched->csg_slots[csg_id]; + + /* We consider group suspension failures as fatal and flag the + * group as unusable by setting timedout=true. + */ + csg_slot->group->timedout = true; csgs_upd_ctx_queue_reqs(ptdev, &upd_ctx, csg_id, CSG_STATE_TERMINATE, @@ -3409,6 +3416,11 @@ panthor_job_create(struct panthor_file *pfile, goto err_put_job; } + if (!group_can_run(job->group)) { + ret = -EINVAL; + goto err_put_job; + } + if (job->queue_idx >= job->group->queue_count || !job->group->queues[job->queue_idx]) { ret = -EINVAL; diff --git a/drivers/gpu/drm/scheduler/sched_main.c b/drivers/gpu/drm/scheduler/sched_main.c index eaef20f41786..e97c6c60bc96 100644 --- a/drivers/gpu/drm/scheduler/sched_main.c +++ b/drivers/gpu/drm/scheduler/sched_main.c @@ -1276,10 +1276,11 @@ int drm_sched_init(struct drm_gpu_scheduler *sched, sched->own_submit_wq = false; } else { #ifdef CONFIG_LOCKDEP - sched->submit_wq = alloc_ordered_workqueue_lockdep_map(name, 0, + sched->submit_wq = alloc_ordered_workqueue_lockdep_map(name, + WQ_MEM_RECLAIM, &drm_sched_lockdep_map); #else - sched->submit_wq = alloc_ordered_workqueue(name, 0); + sched->submit_wq = alloc_ordered_workqueue(name, WQ_MEM_RECLAIM); #endif if (!sched->submit_wq) return -ENOMEM; diff --git a/drivers/gpu/drm/tegra/drm.c b/drivers/gpu/drm/tegra/drm.c index c9eb329665ec..34d22ba210b0 100644 --- a/drivers/gpu/drm/tegra/drm.c +++ b/drivers/gpu/drm/tegra/drm.c @@ -1153,8 +1153,8 @@ static int host1x_drm_probe(struct host1x_device *dev) if (host1x_drm_wants_iommu(dev) && device_iommu_mapped(dma_dev)) { tegra->domain = iommu_paging_domain_alloc(dma_dev); - if (!tegra->domain) { - err = -ENOMEM; + if (IS_ERR(tegra->domain)) { + err = PTR_ERR(tegra->domain); goto free; } diff --git a/drivers/gpu/drm/tests/drm_connector_test.c b/drivers/gpu/drm/tests/drm_connector_test.c index 15e36a8db685..6bba97d0be88 100644 --- a/drivers/gpu/drm/tests/drm_connector_test.c +++ b/drivers/gpu/drm/tests/drm_connector_test.c @@ -996,7 +996,7 @@ static void drm_test_drm_hdmi_compute_mode_clock_rgb(struct kunit *test) unsigned long long rate; struct drm_device *drm = &priv->drm; - mode = drm_display_mode_from_cea_vic(drm, 16); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 16); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_FALSE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); @@ -1017,7 +1017,7 @@ static void drm_test_drm_hdmi_compute_mode_clock_rgb_10bpc(struct kunit *test) unsigned long long rate; struct drm_device *drm = &priv->drm; - mode = drm_display_mode_from_cea_vic(drm, 16); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 16); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_FALSE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); @@ -1038,7 +1038,7 @@ static void drm_test_drm_hdmi_compute_mode_clock_rgb_10bpc_vic_1(struct kunit *t unsigned long long rate; struct drm_device *drm = &priv->drm; - mode = drm_display_mode_from_cea_vic(drm, 1); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 1); KUNIT_ASSERT_NOT_NULL(test, mode); rate = drm_hdmi_compute_mode_clock(mode, 10, HDMI_COLORSPACE_RGB); @@ -1056,7 +1056,7 @@ static void drm_test_drm_hdmi_compute_mode_clock_rgb_12bpc(struct kunit *test) unsigned long long rate; struct drm_device *drm = &priv->drm; - mode = drm_display_mode_from_cea_vic(drm, 16); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 16); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_FALSE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); @@ -1077,7 +1077,7 @@ static void drm_test_drm_hdmi_compute_mode_clock_rgb_12bpc_vic_1(struct kunit *t unsigned long long rate; struct drm_device *drm = &priv->drm; - mode = drm_display_mode_from_cea_vic(drm, 1); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 1); KUNIT_ASSERT_NOT_NULL(test, mode); rate = drm_hdmi_compute_mode_clock(mode, 12, HDMI_COLORSPACE_RGB); @@ -1095,7 +1095,7 @@ static void drm_test_drm_hdmi_compute_mode_clock_rgb_double(struct kunit *test) unsigned long long rate; struct drm_device *drm = &priv->drm; - mode = drm_display_mode_from_cea_vic(drm, 6); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 6); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_TRUE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); @@ -1118,7 +1118,7 @@ static void drm_test_connector_hdmi_compute_mode_clock_yuv420_valid(struct kunit unsigned long long rate; unsigned int vic = *(unsigned int *)test->param_value; - mode = drm_display_mode_from_cea_vic(drm, vic); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, vic); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_FALSE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); @@ -1155,7 +1155,7 @@ static void drm_test_connector_hdmi_compute_mode_clock_yuv420_10_bpc(struct kuni drm_hdmi_compute_mode_clock_yuv420_vic_valid_tests[0]; unsigned long long rate; - mode = drm_display_mode_from_cea_vic(drm, vic); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, vic); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_FALSE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); @@ -1180,7 +1180,7 @@ static void drm_test_connector_hdmi_compute_mode_clock_yuv420_12_bpc(struct kuni drm_hdmi_compute_mode_clock_yuv420_vic_valid_tests[0]; unsigned long long rate; - mode = drm_display_mode_from_cea_vic(drm, vic); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, vic); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_FALSE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); @@ -1203,7 +1203,7 @@ static void drm_test_connector_hdmi_compute_mode_clock_yuv422_8_bpc(struct kunit struct drm_device *drm = &priv->drm; unsigned long long rate; - mode = drm_display_mode_from_cea_vic(drm, 16); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 16); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_FALSE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); @@ -1225,7 +1225,7 @@ static void drm_test_connector_hdmi_compute_mode_clock_yuv422_10_bpc(struct kuni struct drm_device *drm = &priv->drm; unsigned long long rate; - mode = drm_display_mode_from_cea_vic(drm, 16); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 16); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_FALSE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); @@ -1247,7 +1247,7 @@ static void drm_test_connector_hdmi_compute_mode_clock_yuv422_12_bpc(struct kuni struct drm_device *drm = &priv->drm; unsigned long long rate; - mode = drm_display_mode_from_cea_vic(drm, 16); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 16); KUNIT_ASSERT_NOT_NULL(test, mode); KUNIT_ASSERT_FALSE(test, mode->flags & DRM_MODE_FLAG_DBLCLK); diff --git a/drivers/gpu/drm/tests/drm_hdmi_state_helper_test.c b/drivers/gpu/drm/tests/drm_hdmi_state_helper_test.c index 34ee95d41f29..294773342e71 100644 --- a/drivers/gpu/drm/tests/drm_hdmi_state_helper_test.c +++ b/drivers/gpu/drm/tests/drm_hdmi_state_helper_test.c @@ -441,7 +441,7 @@ static void drm_test_check_broadcast_rgb_auto_cea_mode_vic_1(struct kunit *test) ctx = drm_kunit_helper_acquire_ctx_alloc(test); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ctx); - mode = drm_display_mode_from_cea_vic(drm, 1); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 1); KUNIT_ASSERT_NOT_NULL(test, mode); drm = &priv->drm; @@ -555,7 +555,7 @@ static void drm_test_check_broadcast_rgb_full_cea_mode_vic_1(struct kunit *test) ctx = drm_kunit_helper_acquire_ctx_alloc(test); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ctx); - mode = drm_display_mode_from_cea_vic(drm, 1); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 1); KUNIT_ASSERT_NOT_NULL(test, mode); drm = &priv->drm; @@ -671,7 +671,7 @@ static void drm_test_check_broadcast_rgb_limited_cea_mode_vic_1(struct kunit *te ctx = drm_kunit_helper_acquire_ctx_alloc(test); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ctx); - mode = drm_display_mode_from_cea_vic(drm, 1); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 1); KUNIT_ASSERT_NOT_NULL(test, mode); drm = &priv->drm; @@ -1263,7 +1263,7 @@ static void drm_test_check_output_bpc_format_vic_1(struct kunit *test) ctx = drm_kunit_helper_acquire_ctx_alloc(test); KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ctx); - mode = drm_display_mode_from_cea_vic(drm, 1); + mode = drm_kunit_display_mode_from_cea_vic(test, drm, 1); KUNIT_ASSERT_NOT_NULL(test, mode); /* diff --git a/drivers/gpu/drm/tests/drm_kunit_helpers.c b/drivers/gpu/drm/tests/drm_kunit_helpers.c index aa62719dab0e..04a6b8cc62ac 100644 --- a/drivers/gpu/drm/tests/drm_kunit_helpers.c +++ b/drivers/gpu/drm/tests/drm_kunit_helpers.c @@ -3,6 +3,7 @@ #include <drm/drm_atomic.h> #include <drm/drm_atomic_helper.h> #include <drm/drm_drv.h> +#include <drm/drm_edid.h> #include <drm/drm_fourcc.h> #include <drm/drm_kunit_helpers.h> #include <drm/drm_managed.h> @@ -311,6 +312,47 @@ drm_kunit_helper_create_crtc(struct kunit *test, } EXPORT_SYMBOL_GPL(drm_kunit_helper_create_crtc); +static void kunit_action_drm_mode_destroy(void *ptr) +{ + struct drm_display_mode *mode = ptr; + + drm_mode_destroy(NULL, mode); +} + +/** + * drm_kunit_display_mode_from_cea_vic() - return a mode for CEA VIC + for a KUnit test + * @test: The test context object + * @dev: DRM device + * @video_code: CEA VIC of the mode + * + * Creates a new mode matching the specified CEA VIC for a KUnit test. + * + * Resources will be cleaned up automatically. + * + * Returns: A new drm_display_mode on success or NULL on failure + */ +struct drm_display_mode * +drm_kunit_display_mode_from_cea_vic(struct kunit *test, struct drm_device *dev, + u8 video_code) +{ + struct drm_display_mode *mode; + int ret; + + mode = drm_display_mode_from_cea_vic(dev, video_code); + if (!mode) + return NULL; + + ret = kunit_add_action_or_reset(test, + kunit_action_drm_mode_destroy, + mode); + if (ret) + return NULL; + + return mode; +} +EXPORT_SYMBOL_GPL(drm_kunit_display_mode_from_cea_vic); + MODULE_AUTHOR("Maxime Ripard <maxime@cerno.tech>"); MODULE_DESCRIPTION("KUnit test suite helper functions"); MODULE_LICENSE("GPL"); diff --git a/drivers/gpu/drm/xe/display/xe_display.c b/drivers/gpu/drm/xe/display/xe_display.c index 75736faf2a80..c6e0c8d77a70 100644 --- a/drivers/gpu/drm/xe/display/xe_display.c +++ b/drivers/gpu/drm/xe/display/xe_display.c @@ -309,18 +309,7 @@ static void xe_display_flush_cleanup_work(struct xe_device *xe) } /* TODO: System and runtime suspend/resume sequences will be sanitized as a follow-up. */ -void xe_display_pm_runtime_suspend(struct xe_device *xe) -{ - if (!xe->info.probe_display) - return; - - if (xe->d3cold.allowed) - xe_display_pm_suspend(xe, true); - - intel_hpd_poll_enable(xe); -} - -void xe_display_pm_suspend(struct xe_device *xe, bool runtime) +static void __xe_display_pm_suspend(struct xe_device *xe, bool runtime) { struct intel_display *display = &xe->display; bool s2idle = suspend_to_idle(); @@ -353,28 +342,38 @@ void xe_display_pm_suspend(struct xe_device *xe, bool runtime) intel_opregion_suspend(display, s2idle ? PCI_D1 : PCI_D3cold); intel_dmc_suspend(xe); + + if (runtime && has_display(xe)) + intel_hpd_poll_enable(xe); } -void xe_display_pm_suspend_late(struct xe_device *xe) +void xe_display_pm_suspend(struct xe_device *xe) +{ + __xe_display_pm_suspend(xe, false); +} + +void xe_display_pm_runtime_suspend(struct xe_device *xe) { - bool s2idle = suspend_to_idle(); if (!xe->info.probe_display) return; - intel_power_domains_suspend(xe, s2idle); + if (xe->d3cold.allowed) { + __xe_display_pm_suspend(xe, true); + return; + } - intel_display_power_suspend_late(xe); + intel_hpd_poll_enable(xe); } -void xe_display_pm_runtime_resume(struct xe_device *xe) +void xe_display_pm_suspend_late(struct xe_device *xe) { + bool s2idle = suspend_to_idle(); if (!xe->info.probe_display) return; - intel_hpd_poll_disable(xe); + intel_power_domains_suspend(xe, s2idle); - if (xe->d3cold.allowed) - xe_display_pm_resume(xe, true); + intel_display_power_suspend_late(xe); } void xe_display_pm_resume_early(struct xe_device *xe) @@ -387,7 +386,7 @@ void xe_display_pm_resume_early(struct xe_device *xe) intel_power_domains_resume(xe); } -void xe_display_pm_resume(struct xe_device *xe, bool runtime) +static void __xe_display_pm_resume(struct xe_device *xe, bool runtime) { struct intel_display *display = &xe->display; @@ -411,9 +410,11 @@ void xe_display_pm_resume(struct xe_device *xe, bool runtime) intel_display_driver_resume(xe); drm_kms_helper_poll_enable(&xe->drm); intel_display_driver_enable_user_access(xe); - intel_hpd_poll_disable(xe); } + if (has_display(xe)) + intel_hpd_poll_disable(xe); + intel_opregion_resume(display); intel_fbdev_set_suspend(&xe->drm, FBINFO_STATE_RUNNING, false); @@ -421,6 +422,26 @@ void xe_display_pm_resume(struct xe_device *xe, bool runtime) intel_power_domains_enable(xe); } +void xe_display_pm_resume(struct xe_device *xe) +{ + __xe_display_pm_resume(xe, false); +} + +void xe_display_pm_runtime_resume(struct xe_device *xe) +{ + if (!xe->info.probe_display) + return; + + if (xe->d3cold.allowed) { + __xe_display_pm_resume(xe, true); + return; + } + + intel_hpd_init(xe); + intel_hpd_poll_disable(xe); +} + + static void display_device_remove(struct drm_device *dev, void *arg) { struct xe_device *xe = arg; diff --git a/drivers/gpu/drm/xe/display/xe_display.h b/drivers/gpu/drm/xe/display/xe_display.h index 53d727fd792b..bed55fd26f30 100644 --- a/drivers/gpu/drm/xe/display/xe_display.h +++ b/drivers/gpu/drm/xe/display/xe_display.h @@ -34,10 +34,10 @@ void xe_display_irq_enable(struct xe_device *xe, u32 gu_misc_iir); void xe_display_irq_reset(struct xe_device *xe); void xe_display_irq_postinstall(struct xe_device *xe, struct xe_gt *gt); -void xe_display_pm_suspend(struct xe_device *xe, bool runtime); +void xe_display_pm_suspend(struct xe_device *xe); void xe_display_pm_suspend_late(struct xe_device *xe); void xe_display_pm_resume_early(struct xe_device *xe); -void xe_display_pm_resume(struct xe_device *xe, bool runtime); +void xe_display_pm_resume(struct xe_device *xe); void xe_display_pm_runtime_suspend(struct xe_device *xe); void xe_display_pm_runtime_resume(struct xe_device *xe); @@ -65,10 +65,10 @@ static inline void xe_display_irq_enable(struct xe_device *xe, u32 gu_misc_iir) static inline void xe_display_irq_reset(struct xe_device *xe) {} static inline void xe_display_irq_postinstall(struct xe_device *xe, struct xe_gt *gt) {} -static inline void xe_display_pm_suspend(struct xe_device *xe, bool runtime) {} +static inline void xe_display_pm_suspend(struct xe_device *xe) {} static inline void xe_display_pm_suspend_late(struct xe_device *xe) {} static inline void xe_display_pm_resume_early(struct xe_device *xe) {} -static inline void xe_display_pm_resume(struct xe_device *xe, bool runtime) {} +static inline void xe_display_pm_resume(struct xe_device *xe) {} static inline void xe_display_pm_runtime_suspend(struct xe_device *xe) {} static inline void xe_display_pm_runtime_resume(struct xe_device *xe) {} diff --git a/drivers/gpu/drm/xe/regs/xe_gt_regs.h b/drivers/gpu/drm/xe/regs/xe_gt_regs.h index 00ad34ed73a5..bd604b9f08e4 100644 --- a/drivers/gpu/drm/xe/regs/xe_gt_regs.h +++ b/drivers/gpu/drm/xe/regs/xe_gt_regs.h @@ -517,7 +517,7 @@ * [4-6] RSVD * [7] Disabled */ -#define CCS_MODE XE_REG(0x14804) +#define CCS_MODE XE_REG(0x14804, XE_REG_OPTION_MASKED) #define CCS_MODE_CSLICE_0_3_MASK REG_GENMASK(11, 0) /* 3 bits per cslice */ #define CCS_MODE_CSLICE_MASK 0x7 /* CCS0-3 + rsvd */ #define CCS_MODE_CSLICE_WIDTH ilog2(CCS_MODE_CSLICE_MASK + 1) diff --git a/drivers/gpu/drm/xe/xe_device.c b/drivers/gpu/drm/xe/xe_device.c index 0a9ffc19e92f..a1987b554a8d 100644 --- a/drivers/gpu/drm/xe/xe_device.c +++ b/drivers/gpu/drm/xe/xe_device.c @@ -87,10 +87,6 @@ static int xe_file_open(struct drm_device *dev, struct drm_file *file) mutex_init(&xef->exec_queue.lock); xa_init_flags(&xef->exec_queue.xa, XA_FLAGS_ALLOC1); - spin_lock(&xe->clients.lock); - xe->clients.count++; - spin_unlock(&xe->clients.lock); - file->driver_priv = xef; kref_init(&xef->refcount); @@ -107,17 +103,12 @@ static int xe_file_open(struct drm_device *dev, struct drm_file *file) static void xe_file_destroy(struct kref *ref) { struct xe_file *xef = container_of(ref, struct xe_file, refcount); - struct xe_device *xe = xef->xe; xa_destroy(&xef->exec_queue.xa); mutex_destroy(&xef->exec_queue.lock); xa_destroy(&xef->vm.xa); mutex_destroy(&xef->vm.lock); - spin_lock(&xe->clients.lock); - xe->clients.count--; - spin_unlock(&xe->clients.lock); - xe_drm_client_put(xef->client); kfree(xef->process_name); kfree(xef); @@ -333,7 +324,6 @@ struct xe_device *xe_device_create(struct pci_dev *pdev, xe->info.force_execlist = xe_modparam.force_execlist; spin_lock_init(&xe->irq.lock); - spin_lock_init(&xe->clients.lock); init_waitqueue_head(&xe->ufence_wq); @@ -890,7 +880,7 @@ void xe_device_l2_flush(struct xe_device *xe) spin_lock(>->global_invl_lock); xe_mmio_write32(gt, XE2_GLOBAL_INVAL, 0x1); - if (xe_mmio_wait32(gt, XE2_GLOBAL_INVAL, 0x1, 0x0, 150, NULL, true)) + if (xe_mmio_wait32(gt, XE2_GLOBAL_INVAL, 0x1, 0x0, 500, NULL, true)) xe_gt_err_once(gt, "Global invalidation timeout\n"); spin_unlock(>->global_invl_lock); diff --git a/drivers/gpu/drm/xe/xe_device.h b/drivers/gpu/drm/xe/xe_device.h index 894f04770454..34620ef855c0 100644 --- a/drivers/gpu/drm/xe/xe_device.h +++ b/drivers/gpu/drm/xe/xe_device.h @@ -178,4 +178,18 @@ void xe_device_declare_wedged(struct xe_device *xe); struct xe_file *xe_file_get(struct xe_file *xef); void xe_file_put(struct xe_file *xef); +/* + * Occasionally it is seen that the G2H worker starts running after a delay of more than + * a second even after being queued and activated by the Linux workqueue subsystem. This + * leads to G2H timeout error. The root cause of issue lies with scheduling latency of + * Lunarlake Hybrid CPU. Issue disappears if we disable Lunarlake atom cores from BIOS + * and this is beyond xe kmd. + * + * TODO: Drop this change once workqueue scheduling delay issue is fixed on LNL Hybrid CPU. + */ +#define LNL_FLUSH_WORKQUEUE(wq__) \ + flush_workqueue(wq__) +#define LNL_FLUSH_WORK(wrk__) \ + flush_work(wrk__) + #endif diff --git a/drivers/gpu/drm/xe/xe_device_types.h b/drivers/gpu/drm/xe/xe_device_types.h index 09d731a9125c..687f3a9039bb 100644 --- a/drivers/gpu/drm/xe/xe_device_types.h +++ b/drivers/gpu/drm/xe/xe_device_types.h @@ -353,15 +353,6 @@ struct xe_device { struct workqueue_struct *wq; } sriov; - /** @clients: drm clients info */ - struct { - /** @clients.lock: Protects drm clients info */ - spinlock_t lock; - - /** @clients.count: number of drm clients */ - u64 count; - } clients; - /** @usm: unified memory state */ struct { /** @usm.asid: convert a ASID to VM */ diff --git a/drivers/gpu/drm/xe/xe_exec.c b/drivers/gpu/drm/xe/xe_exec.c index f23ac1e2ed88..756b492f13b0 100644 --- a/drivers/gpu/drm/xe/xe_exec.c +++ b/drivers/gpu/drm/xe/xe_exec.c @@ -132,12 +132,16 @@ int xe_exec_ioctl(struct drm_device *dev, void *data, struct drm_file *file) if (XE_IOCTL_DBG(xe, !q)) return -ENOENT; - if (XE_IOCTL_DBG(xe, q->flags & EXEC_QUEUE_FLAG_VM)) - return -EINVAL; + if (XE_IOCTL_DBG(xe, q->flags & EXEC_QUEUE_FLAG_VM)) { + err = -EINVAL; + goto err_exec_queue; + } if (XE_IOCTL_DBG(xe, args->num_batch_buffer && - q->width != args->num_batch_buffer)) - return -EINVAL; + q->width != args->num_batch_buffer)) { + err = -EINVAL; + goto err_exec_queue; + } if (XE_IOCTL_DBG(xe, q->ops->reset_status(q))) { err = -ECANCELED; @@ -220,6 +224,7 @@ retry: fence = xe_sync_in_fence_get(syncs, num_syncs, q, vm); if (IS_ERR(fence)) { err = PTR_ERR(fence); + xe_vm_unlock(vm); goto err_unlock_list; } for (i = 0; i < num_syncs; i++) diff --git a/drivers/gpu/drm/xe/xe_exec_queue.c b/drivers/gpu/drm/xe/xe_exec_queue.c index d098d2dd1b2d..fd0f3b3c9101 100644 --- a/drivers/gpu/drm/xe/xe_exec_queue.c +++ b/drivers/gpu/drm/xe/xe_exec_queue.c @@ -260,8 +260,14 @@ void xe_exec_queue_fini(struct xe_exec_queue *q) { int i; + /* + * Before releasing our ref to lrc and xef, accumulate our run ticks + */ + xe_exec_queue_update_run_ticks(q); + for (i = 0; i < q->width; ++i) xe_lrc_put(q->lrc[i]); + __xe_exec_queue_free(q); } diff --git a/drivers/gpu/drm/xe/xe_force_wake.c b/drivers/gpu/drm/xe/xe_force_wake.c index b263fff15273..7d9fc489dcb8 100644 --- a/drivers/gpu/drm/xe/xe_force_wake.c +++ b/drivers/gpu/drm/xe/xe_force_wake.c @@ -115,9 +115,15 @@ static int __domain_wait(struct xe_gt *gt, struct xe_force_wake_domain *domain, XE_FORCE_WAKE_ACK_TIMEOUT_MS * USEC_PER_MSEC, &value, true); if (ret) - xe_gt_notice(gt, "Force wake domain %d failed to ack %s (%pe) reg[%#x] = %#x\n", - domain->id, str_wake_sleep(wake), ERR_PTR(ret), - domain->reg_ack.addr, value); + xe_gt_err(gt, "Force wake domain %d failed to ack %s (%pe) reg[%#x] = %#x\n", + domain->id, str_wake_sleep(wake), ERR_PTR(ret), + domain->reg_ack.addr, value); + if (value == ~0) { + xe_gt_err(gt, + "Force wake domain %d: %s. MMIO unreliable (forcewake register returns 0xFFFFFFFF)!\n", + domain->id, str_wake_sleep(wake)); + ret = -EIO; + } return ret; } diff --git a/drivers/gpu/drm/xe/xe_ggtt.c b/drivers/gpu/drm/xe/xe_ggtt.c index 2895f154654c..ff19eca5d358 100644 --- a/drivers/gpu/drm/xe/xe_ggtt.c +++ b/drivers/gpu/drm/xe/xe_ggtt.c @@ -397,6 +397,16 @@ static void ggtt_invalidate_gt_tlb(struct xe_gt *gt) static void xe_ggtt_invalidate(struct xe_ggtt *ggtt) { + struct xe_device *xe = tile_to_xe(ggtt->tile); + + /* + * XXX: Barrier for GGTT pages. Unsure exactly why this required but + * without this LNL is having issues with the GuC reading scratch page + * vs. correct GGTT page. Not particularly a hot code path so blindly + * do a mmio read here which results in GuC reading correct GGTT page. + */ + xe_mmio_read32(xe_root_mmio_gt(xe), VF_CAP_REG); + /* Each GT in a tile has its own TLB to cache GGTT lookups */ ggtt_invalidate_gt_tlb(ggtt->tile->primary_gt); ggtt_invalidate_gt_tlb(ggtt->tile->media_gt); diff --git a/drivers/gpu/drm/xe/xe_gt_ccs_mode.c b/drivers/gpu/drm/xe/xe_gt_ccs_mode.c index d2e4dc3aaf61..ffcbd05671fc 100644 --- a/drivers/gpu/drm/xe/xe_gt_ccs_mode.c +++ b/drivers/gpu/drm/xe/xe_gt_ccs_mode.c @@ -68,6 +68,12 @@ static void __xe_gt_apply_ccs_mode(struct xe_gt *gt, u32 num_engines) } } + /* + * Mask bits need to be set for the register. Though only Xe2+ + * platforms require setting of mask bits, it won't harm for older + * platforms as these bits are unused there. + */ + mode |= CCS_MODE_CSLICE_0_3_MASK << 16; xe_mmio_write32(gt, CCS_MODE, mode); xe_gt_dbg(gt, "CCS_MODE=%x config:%08x, num_engines:%d, num_slices:%d\n", @@ -133,9 +139,10 @@ ccs_mode_store(struct device *kdev, struct device_attribute *attr, } /* CCS mode can only be updated when there are no drm clients */ - spin_lock(&xe->clients.lock); - if (xe->clients.count) { - spin_unlock(&xe->clients.lock); + mutex_lock(&xe->drm.filelist_mutex); + if (!list_empty(&xe->drm.filelist)) { + mutex_unlock(&xe->drm.filelist_mutex); + xe_gt_dbg(gt, "Rejecting compute mode change as there are active drm clients\n"); return -EBUSY; } @@ -146,7 +153,7 @@ ccs_mode_store(struct device *kdev, struct device_attribute *attr, xe_gt_reset_async(gt); } - spin_unlock(&xe->clients.lock); + mutex_unlock(&xe->drm.filelist_mutex); return count; } diff --git a/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c b/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c index 8250ef71e685..afdb477ecf83 100644 --- a/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c +++ b/drivers/gpu/drm/xe/xe_gt_sriov_pf_config.c @@ -387,6 +387,8 @@ static void pf_release_ggtt(struct xe_tile *tile, struct xe_ggtt_node *node) * the xe_ggtt_clear() called by below xe_ggtt_remove_node(). */ xe_ggtt_node_remove(node, false); + } else { + xe_ggtt_node_fini(node); } } @@ -442,7 +444,7 @@ static int pf_provision_vf_ggtt(struct xe_gt *gt, unsigned int vfid, u64 size) config->ggtt_region = node; return 0; err: - xe_ggtt_node_fini(node); + pf_release_ggtt(tile, node); return err; } diff --git a/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c b/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c index bbb9e411d21f..9d82ea30f4df 100644 --- a/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c +++ b/drivers/gpu/drm/xe/xe_gt_tlb_invalidation.c @@ -72,6 +72,8 @@ static void xe_gt_tlb_fence_timeout(struct work_struct *work) struct xe_device *xe = gt_to_xe(gt); struct xe_gt_tlb_invalidation_fence *fence, *next; + LNL_FLUSH_WORK(>->uc.guc.ct.g2h_worker); + spin_lock_irq(>->tlb_invalidation.pending_lock); list_for_each_entry_safe(fence, next, >->tlb_invalidation.pending_fences, link) { diff --git a/drivers/gpu/drm/xe/xe_guc_ct.c b/drivers/gpu/drm/xe/xe_guc_ct.c index d16eb9ab49fb..9c505d3517cd 100644 --- a/drivers/gpu/drm/xe/xe_guc_ct.c +++ b/drivers/gpu/drm/xe/xe_guc_ct.c @@ -897,6 +897,15 @@ retry_same_fence: ret = wait_event_timeout(ct->g2h_fence_wq, g2h_fence.done, HZ); + if (!ret) { + LNL_FLUSH_WORK(&ct->g2h_worker); + if (g2h_fence.done) { + xe_gt_warn(gt, "G2H fence %u, action %04x, done\n", + g2h_fence.seqno, action[0]); + ret = 1; + } + } + /* * Ensure we serialize with completion side to prevent UAF with fence going out of scope on * the stack, since we have no clue if it will fire after the timeout before we can erase diff --git a/drivers/gpu/drm/xe/xe_guc_submit.c b/drivers/gpu/drm/xe/xe_guc_submit.c index 8a9254e5af6e..4f5d00aea716 100644 --- a/drivers/gpu/drm/xe/xe_guc_submit.c +++ b/drivers/gpu/drm/xe/xe_guc_submit.c @@ -745,8 +745,6 @@ static void guc_exec_queue_free_job(struct drm_sched_job *drm_job) { struct xe_sched_job *job = to_xe_sched_job(drm_job); - xe_exec_queue_update_run_ticks(job->q); - trace_xe_sched_job_free(job); xe_sched_job_put(job); } @@ -916,12 +914,22 @@ static void xe_guc_exec_queue_lr_cleanup(struct work_struct *w) static bool check_timeout(struct xe_exec_queue *q, struct xe_sched_job *job) { struct xe_gt *gt = guc_to_gt(exec_queue_to_guc(q)); - u32 ctx_timestamp = xe_lrc_ctx_timestamp(q->lrc[0]); - u32 ctx_job_timestamp = xe_lrc_ctx_job_timestamp(q->lrc[0]); + u32 ctx_timestamp, ctx_job_timestamp; u32 timeout_ms = q->sched_props.job_timeout_ms; u32 diff; u64 running_time_ms; + if (!xe_sched_job_started(job)) { + xe_gt_warn(gt, "Check job timeout: seqno=%u, lrc_seqno=%u, guc_id=%d, not started", + xe_sched_job_seqno(job), xe_sched_job_lrc_seqno(job), + q->guc->id); + + return xe_sched_invalidate_job(job, 2); + } + + ctx_timestamp = xe_lrc_ctx_timestamp(q->lrc[0]); + ctx_job_timestamp = xe_lrc_ctx_job_timestamp(q->lrc[0]); + /* * Counter wraps at ~223s at the usual 19.2MHz, be paranoid catch * possible overflows with a high timeout. @@ -1049,10 +1057,6 @@ guc_exec_queue_timedout_job(struct drm_sched_job *drm_job) exec_queue_killed_or_banned_or_wedged(q) || exec_queue_destroyed(q); - /* Job hasn't started, can't be timed out */ - if (!skip_timeout_check && !xe_sched_job_started(job)) - goto rearm; - /* * XXX: Sampling timeout doesn't work in wedged mode as we have to * modify scheduling state to read timestamp. We could read the @@ -1726,8 +1730,13 @@ void xe_guc_submit_stop(struct xe_guc *guc) mutex_lock(&guc->submission_state.lock); - xa_for_each(&guc->submission_state.exec_queue_lookup, index, q) + xa_for_each(&guc->submission_state.exec_queue_lookup, index, q) { + /* Prevent redundant attempts to stop parallel queues */ + if (q->guc->id != index) + continue; + guc_exec_queue_stop(guc, q); + } mutex_unlock(&guc->submission_state.lock); @@ -1765,8 +1774,13 @@ int xe_guc_submit_start(struct xe_guc *guc) mutex_lock(&guc->submission_state.lock); atomic_dec(&guc->submission_state.stopped); - xa_for_each(&guc->submission_state.exec_queue_lookup, index, q) + xa_for_each(&guc->submission_state.exec_queue_lookup, index, q) { + /* Prevent redundant attempts to start parallel queues */ + if (q->guc->id != index) + continue; + guc_exec_queue_start(q); + } mutex_unlock(&guc->submission_state.lock); wake_up_all(&guc->ct.wq); diff --git a/drivers/gpu/drm/xe/xe_pm.c b/drivers/gpu/drm/xe/xe_pm.c index 7cf2160fe040..33eb039053e4 100644 --- a/drivers/gpu/drm/xe/xe_pm.c +++ b/drivers/gpu/drm/xe/xe_pm.c @@ -123,7 +123,7 @@ int xe_pm_suspend(struct xe_device *xe) for_each_gt(gt, xe, id) xe_gt_suspend_prepare(gt); - xe_display_pm_suspend(xe, false); + xe_display_pm_suspend(xe); /* FIXME: Super racey... */ err = xe_bo_evict_all(xe); @@ -133,7 +133,7 @@ int xe_pm_suspend(struct xe_device *xe) for_each_gt(gt, xe, id) { err = xe_gt_suspend(gt); if (err) { - xe_display_pm_resume(xe, false); + xe_display_pm_resume(xe); goto err; } } @@ -187,7 +187,7 @@ int xe_pm_resume(struct xe_device *xe) for_each_gt(gt, xe, id) xe_gt_resume(gt); - xe_display_pm_resume(xe, false); + xe_display_pm_resume(xe); err = xe_bo_restore_user(xe); if (err) diff --git a/drivers/gpu/drm/xe/xe_sync.c b/drivers/gpu/drm/xe/xe_sync.c index c6cf227ead40..2e72c06fd40d 100644 --- a/drivers/gpu/drm/xe/xe_sync.c +++ b/drivers/gpu/drm/xe/xe_sync.c @@ -54,8 +54,9 @@ static struct xe_user_fence *user_fence_create(struct xe_device *xe, u64 addr, { struct xe_user_fence *ufence; u64 __user *ptr = u64_to_user_ptr(addr); + u64 __maybe_unused prefetch_val; - if (!access_ok(ptr, sizeof(*ptr))) + if (get_user(prefetch_val, ptr)) return ERR_PTR(-EFAULT); ufence = kzalloc(sizeof(*ufence), GFP_KERNEL); diff --git a/drivers/gpu/drm/xe/xe_wait_user_fence.c b/drivers/gpu/drm/xe/xe_wait_user_fence.c index f5deb81eba01..5b4264ea38bd 100644 --- a/drivers/gpu/drm/xe/xe_wait_user_fence.c +++ b/drivers/gpu/drm/xe/xe_wait_user_fence.c @@ -155,6 +155,13 @@ int xe_wait_user_fence_ioctl(struct drm_device *dev, void *data, } if (!timeout) { + LNL_FLUSH_WORKQUEUE(xe->ordered_wq); + err = do_compare(addr, args->value, args->mask, + args->op); + if (err <= 0) { + drm_dbg(&xe->drm, "LNL_FLUSH_WORKQUEUE resolved ufence timeout\n"); + break; + } err = -ETIME; break; } |