Merge branch 'drm-next-5.1' of git://people.freedesktop.org/~agd5f/linux into drm-next

ttm:
- Replace ref/unref naming with get/put

amdgpu:
- Revert DC clang fix, causes a segfault with some compiler versions
- SR-IOV fix
- PCIE fix for vega20
- Misc DC fixes

Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexdeucher@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20190201062345.7304-1-alexander.deucher@amd.com

27 files changed, 239 insertions, 154 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
index dac187454b33..0d90672d0e58 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_xgmi.c
@@ -91,10 +91,6 @@ int amdgpu_xgmi_update_topology(struct amdgpu_hive_info *hive, struct amdgpu_dev
 			"XGMI: Set topology failure on device %llx, hive %llx, ret %d",
 			adev->gmc.xgmi.node_id,
 			adev->gmc.xgmi.hive_id, ret);
-	else
-		dev_info(adev->dev, "XGMI: Set topology for node %d, hive 0x%llx.\n",
-			 adev->gmc.xgmi.physical_node_id,
-				 adev->gmc.xgmi.hive_id);
 
 	return ret;
 }
@@ -160,6 +156,9 @@ int amdgpu_xgmi_add_device(struct amdgpu_device *adev)
 			break;
 	}
 
+	dev_info(adev->dev, "XGMI: Add node %d, hive 0x%llx.\n",
+		 adev->gmc.xgmi.physical_node_id, adev->gmc.xgmi.hive_id);
+
 	mutex_unlock(&hive->hive_lock);
 exit:
 	return ret;
diff --git a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
index 9c082f9aea1a..600259b4e291 100644
--- a/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gmc_v9_0.c
@@ -965,7 +965,11 @@ static int gmc_v9_0_sw_init(void *handle)
 		 * vm size is 256TB (48bit), maximum size of Vega10,
 		 * block size 512 (9bit)
 		 */
-		amdgpu_vm_adjust_size(adev, 256 * 1024, 9, 3, 48);
+		/* sriov restrict max_pfn below AMDGPU_GMC_HOLE */
+		if (amdgpu_sriov_vf(adev))
+			amdgpu_vm_adjust_size(adev, 256 * 1024, 9, 3, 47);
+		else
+			amdgpu_vm_adjust_size(adev, 256 * 1024, 9, 3, 48);
 		break;
 	default:
 		break;
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 407f733a47ea..1923f47d3dd6 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4658,8 +4658,10 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 	flip = kzalloc(sizeof(*flip), GFP_KERNEL);
 	full = kzalloc(sizeof(*full), GFP_KERNEL);
 
-	if (!flip || !full)
+	if (!flip || !full) {
 		dm_error("Failed to allocate update bundles\n");
+		goto cleanup;
+	}
 
 	/* update planes when needed */
 	for_each_oldnew_plane_in_state(state, plane, old_plane_state, new_plane_state, i) {
@@ -4883,6 +4885,10 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 						     dc_state);
 		mutex_unlock(&dm->dc_lock);
 	}
+
+cleanup:
+	kfree(flip);
+	kfree(full);
 }
 
 /*
@@ -4917,10 +4923,25 @@ static int amdgpu_dm_atomic_commit(struct drm_device *dev,
 	 */
 	for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
 		struct dm_crtc_state *dm_old_crtc_state = to_dm_crtc_state(old_crtc_state);
+		struct dm_crtc_state *dm_new_crtc_state = to_dm_crtc_state(new_crtc_state);
 		struct amdgpu_crtc *acrtc = to_amdgpu_crtc(crtc);
 
-		if (drm_atomic_crtc_needs_modeset(new_crtc_state) && dm_old_crtc_state->stream)
+		if (drm_atomic_crtc_needs_modeset(new_crtc_state)
+		    && dm_old_crtc_state->stream) {
+			/*
+			 * If the stream is removed and CRC capture was
+			 * enabled on the CRTC the extra vblank reference
+			 * needs to be dropped since CRC capture will be
+			 * disabled.
+			 */
+			if (!dm_new_crtc_state->stream
+			    && dm_new_crtc_state->crc_enabled) {
+				drm_crtc_vblank_put(crtc);
+				dm_new_crtc_state->crc_enabled = false;
+			}
+
 			manage_dm_interrupts(adev, acrtc, false);
+		}
 	}
 	/*
 	 * Add check here for SoC's that support hardware cursor plane, to
@@ -5152,6 +5173,10 @@ static void amdgpu_dm_atomic_commit_tail(struct drm_atomic_state *state)
 			continue;
 
 		manage_dm_interrupts(adev, acrtc, true);
+
+		/* The stream has changed so CRC capture needs to re-enabled. */
+		if (dm_new_crtc_state->crc_enabled)
+			amdgpu_dm_crtc_set_crc_source(crtc, "auto");
 	}
 
 	/* update planes when needed per crtc*/
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
index f088ac585978..a10e3a50d9ef 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_crc.c
@@ -64,8 +64,10 @@ amdgpu_dm_crtc_verify_crc_source(struct drm_crtc *crtc, const char *src_name,
 
 int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name)
 {
+	struct amdgpu_device *adev = crtc->dev->dev_private;
 	struct dm_crtc_state *crtc_state = to_dm_crtc_state(crtc->state);
 	struct dc_stream_state *stream_state = crtc_state->stream;
+	bool enable;
 
 	enum amdgpu_dm_pipe_crc_source source = dm_parse_crc_source(src_name);
 
@@ -80,29 +82,33 @@ int amdgpu_dm_crtc_set_crc_source(struct drm_crtc *crtc, const char *src_name)
 		return -EINVAL;
 	}
 
-	/* When enabling CRC, we should also disable dithering. */
-	if (source == AMDGPU_DM_PIPE_CRC_SOURCE_AUTO) {
-		if (dc_stream_configure_crc(stream_state->ctx->dc,
-					    stream_state,
-					    true, true)) {
-			crtc_state->crc_enabled = true;
-			dc_stream_set_dither_option(stream_state,
-						    DITHER_OPTION_TRUN8);
-		}
-		else
-			return -EINVAL;
-	} else {
-		if (dc_stream_configure_crc(stream_state->ctx->dc,
-					    stream_state,
-					    false, false)) {
-			crtc_state->crc_enabled = false;
-			dc_stream_set_dither_option(stream_state,
-						    DITHER_OPTION_DEFAULT);
-		}
-		else
-			return -EINVAL;
+	enable = (source == AMDGPU_DM_PIPE_CRC_SOURCE_AUTO);
+
+	mutex_lock(&adev->dm.dc_lock);
+	if (!dc_stream_configure_crc(stream_state->ctx->dc, stream_state,
+				     enable, enable)) {
+		mutex_unlock(&adev->dm.dc_lock);
+		return -EINVAL;
 	}
 
+	/* When enabling CRC, we should also disable dithering. */
+	dc_stream_set_dither_option(stream_state,
+				    enable ? DITHER_OPTION_TRUN8
+					   : DITHER_OPTION_DEFAULT);
+
+	mutex_unlock(&adev->dm.dc_lock);
+
+	/*
+	 * Reading the CRC requires the vblank interrupt handler to be
+	 * enabled. Keep a reference until CRC capture stops.
+	 */
+	if (!crtc_state->crc_enabled && enable)
+		drm_crtc_vblank_get(crtc);
+	else if (crtc_state->crc_enabled && !enable)
+		drm_crtc_vblank_put(crtc);
+
+	crtc_state->crc_enabled = enable;
+
 	/* Reset crc_skipped on dm state */
 	crtc_state->crc_skip_count = 0;
 	return 0;
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 7f64be06679d..b39766bd2840 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
@@ -263,6 +263,13 @@ bool dm_helpers_dp_mst_write_payload_allocation_table(
 	return true;
 }
 
+/*
+ * poll pending down reply before clear payload allocation table
+ */
+void dm_helpers_dp_mst_poll_pending_down_reply(
+	struct dc_context *ctx,
+	const struct dc_link *link)
+{}
 
 /*
  * Clear payload allocation table before enable MST DP link.
diff --git a/drivers/gpu/drm/amd/display/dc/calcs/Makefile b/drivers/gpu/drm/amd/display/dc/calcs/Makefile
index dc85a3c088af..95f332ee3e7e 100644
--- a/drivers/gpu/drm/amd/display/dc/calcs/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/calcs/Makefile
@@ -30,7 +30,7 @@ else ifneq ($(call cc-option, -mstack-alignment=16),)
 	cc_stack_align := -mstack-alignment=16
 endif
 
-calcs_ccflags := -mhard-float -msse -msse2 $(cc_stack_align)
+calcs_ccflags := -mhard-float -msse $(cc_stack_align)
 
 CFLAGS_dcn_calcs.o := $(calcs_ccflags)
 CFLAGS_dcn_calc_auto.o := $(calcs_ccflags)
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 1dabafc12cfe..e22be0aefd1b 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -1463,11 +1463,13 @@ static void commit_planes_do_stream_update(struct dc *dc,
 					stream_update->adjust->v_total_min,
 					stream_update->adjust->v_total_max);
 
-			if (stream_update->periodic_fn_vsync_delta &&
-					pipe_ctx->stream_res.tg->funcs->program_vline_interrupt)
+			if (stream_update->vline0_config && pipe_ctx->stream_res.tg->funcs->program_vline_interrupt)
 				pipe_ctx->stream_res.tg->funcs->program_vline_interrupt(
-					pipe_ctx->stream_res.tg, &pipe_ctx->stream->timing,
-					pipe_ctx->stream->periodic_fn_vsync_delta);
+					pipe_ctx->stream_res.tg, VLINE0, stream->vline0_config);
+
+			if (stream_update->vline1_config && pipe_ctx->stream_res.tg->funcs->program_vline_interrupt)
+				pipe_ctx->stream_res.tg->funcs->program_vline_interrupt(
+					pipe_ctx->stream_res.tg, VLINE1, stream->vline1_config);
 
 			if ((stream_update->hdr_static_metadata && !stream->use_dynamic_meta) ||
 					stream_update->vrr_infopacket ||
diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link.c b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
index 3dd5f2717b53..8ff5d42587c2 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link.c
@@ -1467,6 +1467,11 @@ static enum dc_status enable_link_dp_mst(
 	if (link->cur_link_settings.lane_count != LANE_COUNT_UNKNOWN)
 		return DC_OK;
 
+	/* to make sure the pending down rep can be processed
+	 * before clear payload table
+	 */
+	dm_helpers_dp_mst_poll_pending_down_reply(link->ctx, link);
+
 	/* clear payload table */
 	dm_helpers_dp_mst_clear_payload_allocation_table(link->ctx, link);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index 7bb1da18c1ba..0de6d7f377a6 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -45,6 +45,11 @@ struct freesync_context {
 	bool dummy;
 };
 
+struct vline_config {
+	unsigned int start_line;
+	unsigned int end_line;
+};
+
 struct dc_stream_state {
 	// sink is deprecated, new code should not reference
 	// this pointer
@@ -85,8 +90,6 @@ struct dc_stream_state {
 	uint8_t qs_bit;
 	uint8_t qy_bit;
 
-	unsigned long long periodic_fn_vsync_delta;
-
 	/* TODO: custom INFO packets */
 	/* TODO: ABM info (DMCU) */
 	/* PSR info */
@@ -96,6 +99,9 @@ struct dc_stream_state {
 	/* DMCU info */
 	unsigned int abm_level;
 
+	struct vline_config vline0_config;
+	struct vline_config vline1_config;
+
 	/* from core_stream struct */
 	struct dc_context *ctx;
 
@@ -143,7 +149,9 @@ struct dc_stream_update {
 	struct dc_info_packet *hdr_static_metadata;
 	unsigned int *abm_level;
 
-	unsigned long long *periodic_fn_vsync_delta;
+	struct vline_config *vline0_config;
+	struct vline_config *vline1_config;
+
 	struct dc_crtc_timing_adjust *adjust;
 	struct dc_info_packet *vrr_infopacket;
 	struct dc_info_packet *vsc_infopacket;
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
index eb019d404928..1d4f9b48ed7d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.c
@@ -92,68 +92,26 @@ static void optc1_disable_stereo(struct timing_generator *optc)
 		OTG_3D_STRUCTURE_STEREO_SEL_OVR, 0);
 }
 
-static uint32_t get_start_vline(struct timing_generator *optc, const struct dc_crtc_timing *dc_crtc_timing)
-{
-	struct dc_crtc_timing patched_crtc_timing;
-	int vesa_sync_start;
-	int asic_blank_end;
-	int vertical_line_start;
-
-	patched_crtc_timing = *dc_crtc_timing;
-	optc1_apply_front_porch_workaround(optc, &patched_crtc_timing);
-
-	vesa_sync_start = patched_crtc_timing.v_addressable +
-			patched_crtc_timing.v_border_bottom +
-			patched_crtc_timing.v_front_porch;
-
-	asic_blank_end = (patched_crtc_timing.v_total -
-			vesa_sync_start -
-			patched_crtc_timing.v_border_top);
-
-	vertical_line_start = asic_blank_end - optc->dlg_otg_param.vstartup_start + 1;
-	if (vertical_line_start < 0)
-		vertical_line_start = 0;
-
-	return vertical_line_start;
-}
-
 void optc1_program_vline_interrupt(
 		struct timing_generator *optc,
-		const struct dc_crtc_timing *dc_crtc_timing,
-		unsigned long long vsync_delta)
+		enum vline_select vline,
+		struct vline_config vline_config)
 {
-
 	struct optc *optc1 = DCN10TG_FROM_TG(optc);
 
-	unsigned long long req_delta_tens_of_usec = div64_u64((vsync_delta + 9999), 10000);
-	unsigned long long pix_clk_hundreds_khz = div64_u64((dc_crtc_timing->pix_clk_100hz + 999), 1000);
-	uint32_t req_delta_lines = (uint32_t) div64_u64(
-			(req_delta_tens_of_usec * pix_clk_hundreds_khz + dc_crtc_timing->h_total - 1),
-								dc_crtc_timing->h_total);
-
-	uint32_t vsync_line = get_start_vline(optc, dc_crtc_timing);
-	uint32_t start_line = 0;
-	uint32_t end_line = 0;
-
-	if (req_delta_lines != 0)
-		req_delta_lines--;
-
-	if (req_delta_lines > vsync_line)
-		start_line = dc_crtc_timing->v_total - (req_delta_lines - vsync_line) + 2;
-	else
-		start_line = vsync_line - req_delta_lines;
-
-	end_line = start_line + 2;
-
-	if (start_line >= dc_crtc_timing->v_total)
-		start_line = start_line % dc_crtc_timing->v_total;
-
-	if (end_line >= dc_crtc_timing->v_total)
-		end_line = end_line % dc_crtc_timing->v_total;
-
-	REG_SET_2(OTG_VERTICAL_INTERRUPT0_POSITION, 0,
-			OTG_VERTICAL_INTERRUPT0_LINE_START, start_line,
-			OTG_VERTICAL_INTERRUPT0_LINE_END, end_line);
+	switch (vline) {
+	case VLINE0:
+		REG_SET_2(OTG_VERTICAL_INTERRUPT0_POSITION, 0,
+				OTG_VERTICAL_INTERRUPT0_LINE_START, vline_config.start_line,
+				OTG_VERTICAL_INTERRUPT0_LINE_END, vline_config.end_line);
+		break;
+	case VLINE1:
+		REG_SET(OTG_VERTICAL_INTERRUPT1_POSITION, 0,
+					OTG_VERTICAL_INTERRUPT1_LINE_START, vline_config.start_line);
+		break;
+	default:
+		break;
+	}
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
index 8bacf0b6e27e..8eb71c0160a7 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_optc.h
@@ -67,6 +67,8 @@
 	SRI(OTG_CLOCK_CONTROL, OTG, inst),\
 	SRI(OTG_VERTICAL_INTERRUPT0_CONTROL, OTG, inst),\
 	SRI(OTG_VERTICAL_INTERRUPT0_POSITION, OTG, inst),\
+	SRI(OTG_VERTICAL_INTERRUPT1_CONTROL, OTG, inst),\
+	SRI(OTG_VERTICAL_INTERRUPT1_POSITION, OTG, inst),\
 	SRI(OTG_VERTICAL_INTERRUPT2_CONTROL, OTG, inst),\
 	SRI(OTG_VERTICAL_INTERRUPT2_POSITION, OTG, inst),\
 	SRI(OPTC_INPUT_CLOCK_CONTROL, ODM, inst),\
@@ -135,6 +137,8 @@ struct dcn_optc_registers {
 	uint32_t OTG_CLOCK_CONTROL;
 	uint32_t OTG_VERTICAL_INTERRUPT0_CONTROL;
 	uint32_t OTG_VERTICAL_INTERRUPT0_POSITION;
+	uint32_t OTG_VERTICAL_INTERRUPT1_CONTROL;
+	uint32_t OTG_VERTICAL_INTERRUPT1_POSITION;
 	uint32_t OTG_VERTICAL_INTERRUPT2_CONTROL;
 	uint32_t OTG_VERTICAL_INTERRUPT2_POSITION;
 	uint32_t OPTC_INPUT_CLOCK_CONTROL;
@@ -227,6 +231,8 @@ struct dcn_optc_registers {
 	SF(OTG0_OTG_VERTICAL_INTERRUPT0_CONTROL, OTG_VERTICAL_INTERRUPT0_INT_ENABLE, mask_sh),\
 	SF(OTG0_OTG_VERTICAL_INTERRUPT0_POSITION, OTG_VERTICAL_INTERRUPT0_LINE_START, mask_sh),\
 	SF(OTG0_OTG_VERTICAL_INTERRUPT0_POSITION, OTG_VERTICAL_INTERRUPT0_LINE_END, mask_sh),\
+	SF(OTG0_OTG_VERTICAL_INTERRUPT1_CONTROL, OTG_VERTICAL_INTERRUPT1_INT_ENABLE, mask_sh),\
+	SF(OTG0_OTG_VERTICAL_INTERRUPT1_POSITION, OTG_VERTICAL_INTERRUPT1_LINE_START, mask_sh),\
 	SF(OTG0_OTG_VERTICAL_INTERRUPT2_CONTROL, OTG_VERTICAL_INTERRUPT2_INT_ENABLE, mask_sh),\
 	SF(OTG0_OTG_VERTICAL_INTERRUPT2_POSITION, OTG_VERTICAL_INTERRUPT2_LINE_START, mask_sh),\
 	SF(ODM0_OPTC_INPUT_CLOCK_CONTROL, OPTC_INPUT_CLK_EN, mask_sh),\
@@ -361,6 +367,8 @@ struct dcn_optc_registers {
 	type OTG_VERTICAL_INTERRUPT0_INT_ENABLE;\
 	type OTG_VERTICAL_INTERRUPT0_LINE_START;\
 	type OTG_VERTICAL_INTERRUPT0_LINE_END;\
+	type OTG_VERTICAL_INTERRUPT1_INT_ENABLE;\
+	type OTG_VERTICAL_INTERRUPT1_LINE_START;\
 	type OTG_VERTICAL_INTERRUPT2_INT_ENABLE;\
 	type OTG_VERTICAL_INTERRUPT2_LINE_START;\
 	type OPTC_INPUT_CLK_EN;\
@@ -476,8 +484,8 @@ void optc1_program_timing(
 	bool use_vbios);
 
 void optc1_program_vline_interrupt(struct timing_generator *optc,
-		const struct dc_crtc_timing *dc_crtc_timing,
-		unsigned long long vsync_delta);
+		enum vline_select vline,
+		struct vline_config vline_config);
 
 void optc1_program_global_sync(
 		struct timing_generator *optc);
diff --git a/drivers/gpu/drm/amd/display/dc/dm_helpers.h b/drivers/gpu/drm/amd/display/dc/dm_helpers.h
index 5d4527d03045..e81b24374bcb 100644
--- a/drivers/gpu/drm/amd/display/dc/dm_helpers.h
+++ b/drivers/gpu/drm/amd/display/dc/dm_helpers.h
@@ -58,6 +58,13 @@ bool dm_helpers_dp_mst_write_payload_allocation_table(
 		bool enable);
 
 /*
+ * poll pending down reply before clear payload allocation table
+ */
+void dm_helpers_dp_mst_poll_pending_down_reply(
+	struct dc_context *ctx,
+	const struct dc_link *link);
+
+/*
  * Clear payload allocation table before enable MST DP link.
  */
 void dm_helpers_dp_mst_clear_payload_allocation_table(
diff --git a/drivers/gpu/drm/amd/display/dc/dml/Makefile b/drivers/gpu/drm/amd/display/dc/dml/Makefile
index 33c7d7588712..d97ca6528f9d 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/Makefile
+++ b/drivers/gpu/drm/amd/display/dc/dml/Makefile
@@ -30,7 +30,7 @@ else ifneq ($(call cc-option, -mstack-alignment=16),)
 	cc_stack_align := -mstack-alignment=16
 endif
 
-dml_ccflags := -mhard-float -msse -msse2 $(cc_stack_align)
+dml_ccflags := -mhard-float -msse $(cc_stack_align)
 
 CFLAGS_display_mode_lib.o := $(dml_ccflags)
 CFLAGS_display_pipe_clocks.o := $(dml_ccflags)
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
index af700c7dac50..df64cf73ceb9 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/timing_generator.h
@@ -134,6 +134,15 @@ struct dc_crtc_timing;
 
 struct drr_params;
 
+struct vline_config;
+
+
+enum vline_select {
+	VLINE0,
+	VLINE1,
+	VLINE2
+};
+
 struct timing_generator_funcs {
 	bool (*validate_timing)(struct timing_generator *tg,
 							const struct dc_crtc_timing *timing);
@@ -141,8 +150,8 @@ struct timing_generator_funcs {
 							const struct dc_crtc_timing *timing,
 							bool use_vbios);
 	void (*program_vline_interrupt)(struct timing_generator *optc,
-			const struct dc_crtc_timing *dc_crtc_timing,
-			unsigned long long vsync_delta);
+			enum vline_select vline,
+			struct vline_config vline_config);
 	bool (*enable_crtc)(struct timing_generator *tg);
 	bool (*disable_crtc)(struct timing_generator *tg);
 	bool (*is_counter_moving)(struct timing_generator *tg);
diff --git a/drivers/gpu/drm/amd/display/dc/irq_types.h b/drivers/gpu/drm/amd/display/dc/irq_types.h
index 0b5f3a278c22..d0ccd81ad5b4 100644
--- a/drivers/gpu/drm/amd/display/dc/irq_types.h
+++ b/drivers/gpu/drm/amd/display/dc/irq_types.h
@@ -144,6 +144,14 @@ enum dc_irq_source {
 	DC_IRQ_SOURCE_DC5_VLINE0,
 	DC_IRQ_SOURCE_DC6_VLINE0,
 
+	DC_IRQ_SOURCE_DC1_VLINE1,
+	DC_IRQ_SOURCE_DC2_VLINE1,
+	DC_IRQ_SOURCE_DC3_VLINE1,
+	DC_IRQ_SOURCE_DC4_VLINE1,
+	DC_IRQ_SOURCE_DC5_VLINE1,
+	DC_IRQ_SOURCE_DC6_VLINE1,
+
+
 	DAL_IRQ_SOURCES_NUMBER
 };
 
diff --git a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
index 7b49a9a13a4a..da022ca79b56 100644
--- a/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
+++ b/drivers/gpu/drm/amd/powerplay/hwmgr/vega20_hwmgr.c
@@ -771,6 +771,47 @@ static int vega20_init_smc_table(struct pp_hwmgr *hwmgr)
 	return 0;
 }
 
+static int vega20_override_pcie_parameters(struct pp_hwmgr *hwmgr)
+{
+	struct amdgpu_device *adev = (struct amdgpu_device *)(hwmgr->adev);
+	uint32_t pcie_speed = 0, pcie_width = 0, pcie_arg;
+	int ret;
+
+	if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN4)
+		pcie_speed = 16;
+	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN3)
+		pcie_speed = 8;
+	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN2)
+		pcie_speed = 5;
+	else if (adev->pm.pcie_gen_mask & CAIL_PCIE_LINK_SPEED_SUPPORT_GEN1)
+		pcie_speed = 2;
+
+	if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X32)
+		pcie_width = 32;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X16)
+		pcie_width = 16;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X12)
+		pcie_width = 12;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X8)
+		pcie_width = 8;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X4)
+		pcie_width = 4;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X2)
+		pcie_width = 2;
+	else if (adev->pm.pcie_mlw_mask & CAIL_PCIE_LINK_WIDTH_SUPPORT_X1)
+		pcie_width = 1;
+
+	pcie_arg = pcie_width | (pcie_speed << 8);
+
+	ret = smum_send_msg_to_smc_with_parameter(hwmgr,
+			PPSMC_MSG_OverridePcieParameters, pcie_arg);
+	PP_ASSERT_WITH_CODE(!ret,
+		"[OverridePcieParameters] Attempt to override pcie params failed!",
+		return ret);
+
+	return 0;
+}
+
 static int vega20_set_allowed_featuresmask(struct pp_hwmgr *hwmgr)
 {
 	struct vega20_hwmgr *data =
@@ -1570,6 +1611,11 @@ static int vega20_enable_dpm_tasks(struct pp_hwmgr *hwmgr)
 			"[EnableDPMTasks] Failed to initialize SMC table!",
 			return result);
 
+	result = vega20_override_pcie_parameters(hwmgr);
+	PP_ASSERT_WITH_CODE(!result,
+			"[EnableDPMTasks] Failed to override pcie parameters!",
+			return result);
+
 	result = vega20_run_btc(hwmgr);
 	PP_ASSERT_WITH_CODE(!result,
 			"[EnableDPMTasks] Failed to run btc!",
diff --git a/drivers/gpu/drm/ast/ast_main.c b/drivers/gpu/drm/ast/ast_main.c
index 373700c05a00..2854399856ba 100644
--- a/drivers/gpu/drm/ast/ast_main.c
+++ b/drivers/gpu/drm/ast/ast_main.c
@@ -639,13 +639,9 @@ int ast_dumb_create(struct drm_file *file,
 
 static void ast_bo_unref(struct ast_bo **bo)
 {
-	struct ttm_buffer_object *tbo;
-
 	if ((*bo) == NULL)
 		return;
-
-	tbo = &((*bo)->bo);
-	ttm_bo_unref(&tbo);
+	ttm_bo_put(&((*bo)->bo));
 	*bo = NULL;
 }
 
diff --git a/drivers/gpu/drm/mgag200/mgag200_main.c b/drivers/gpu/drm/mgag200/mgag200_main.c
index 79d54103d470..163255099779 100644
--- a/drivers/gpu/drm/mgag200/mgag200_main.c
+++ b/drivers/gpu/drm/mgag200/mgag200_main.c
@@ -33,7 +33,7 @@ int mgag200_framebuffer_init(struct drm_device *dev,
 			     struct drm_gem_object *obj)
 {
 	int ret;
-	
+
 	drm_helper_mode_fill_fb_struct(dev, &gfb->base, mode_cmd);
 	gfb->obj = obj;
 	ret = drm_framebuffer_init(dev, &gfb->base, &mga_fb_funcs);
@@ -318,13 +318,9 @@ int mgag200_dumb_create(struct drm_file *file,
 
 static void mgag200_bo_unref(struct mgag200_bo **bo)
 {
-	struct ttm_buffer_object *tbo;
-
 	if ((*bo) == NULL)
 		return;
-
-	tbo = &((*bo)->bo);
-	ttm_bo_unref(&tbo);
+	ttm_bo_put(&((*bo)->bo));
 	*bo = NULL;
 }
 
diff --git a/drivers/gpu/drm/nouveau/nouveau_bo.h b/drivers/gpu/drm/nouveau/nouveau_bo.h
index 73c48440d4d7..846f4bdec0de 100644
--- a/drivers/gpu/drm/nouveau/nouveau_bo.h
+++ b/drivers/gpu/drm/nouveau/nouveau_bo.h
@@ -61,12 +61,14 @@ nouveau_bo_ref(struct nouveau_bo *ref, struct nouveau_bo **pnvbo)
 		return -EINVAL;
 	prev = *pnvbo;
 
-	*pnvbo = ref ? nouveau_bo(ttm_bo_reference(&ref->bo)) : NULL;
-	if (prev) {
-		struct ttm_buffer_object *bo = &prev->bo;
-
-		ttm_bo_unref(&bo);
+	if (ref) {
+		ttm_bo_get(&ref->bo);
+		*pnvbo = nouveau_bo(&ref->bo);
+	} else {
+		*pnvbo = NULL;
 	}
+	if (prev)
+		ttm_bo_put(&prev->bo);
 
 	return 0;
 }
diff --git a/drivers/gpu/drm/nouveau/nouveau_gem.c b/drivers/gpu/drm/nouveau/nouveau_gem.c
index b56524d343c3..fb028e3b5f51 100644
--- a/drivers/gpu/drm/nouveau/nouveau_gem.c
+++ b/drivers/gpu/drm/nouveau/nouveau_gem.c
@@ -41,7 +41,6 @@ nouveau_gem_object_del(struct drm_gem_object *gem)
 {
 	struct nouveau_bo *nvbo = nouveau_gem_object(gem);
 	struct nouveau_drm *drm = nouveau_bdev(nvbo->bo.bdev);
-	struct ttm_buffer_object *bo = &nvbo->bo;
 	struct device *dev = drm->dev->dev;
 	int ret;
 
@@ -56,7 +55,7 @@ nouveau_gem_object_del(struct drm_gem_object *gem)
 
 	/* reset filp so nouveau_bo_del_ttm() can test for it */
 	gem->filp = NULL;
-	ttm_bo_unref(&bo);
+	ttm_bo_put(&nvbo->bo);
 
 	pm_runtime_mark_last_busy(dev);
 	pm_runtime_put_autosuspend(dev);
diff --git a/drivers/gpu/drm/ttm/ttm_bo.c b/drivers/gpu/drm/ttm/ttm_bo.c
index de088c8070fb..3f56647cdb35 100644
--- a/drivers/gpu/drm/ttm/ttm_bo.c
+++ b/drivers/gpu/drm/ttm/ttm_bo.c
@@ -679,15 +679,6 @@ void ttm_bo_put(struct ttm_buffer_object *bo)
 }
 EXPORT_SYMBOL(ttm_bo_put);
 
-void ttm_bo_unref(struct ttm_buffer_object **p_bo)
-{
-	struct ttm_buffer_object *bo = *p_bo;
-
-	*p_bo = NULL;
-	ttm_bo_put(bo);
-}
-EXPORT_SYMBOL(ttm_bo_unref);
-
 int ttm_bo_lock_delayed_workqueue(struct ttm_bo_device *bdev)
 {
 	return cancel_delayed_work_sync(&bdev->wq);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
index 7ce1c2f87d9a..5d5c2bce01f3 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_bo.c
@@ -534,7 +534,6 @@ static void vmw_user_bo_release(struct ttm_base_object **p_base)
 {
 	struct vmw_user_buffer_object *vmw_user_bo;
 	struct ttm_base_object *base = *p_base;
-	struct ttm_buffer_object *bo;
 
 	*p_base = NULL;
 
@@ -543,8 +542,7 @@ static void vmw_user_bo_release(struct ttm_base_object **p_base)
 
 	vmw_user_bo = container_of(base, struct vmw_user_buffer_object,
 				   prime.base);
-	bo = &vmw_user_bo->vbo.base;
-	ttm_bo_unref(&bo);
+	ttm_bo_put(&vmw_user_bo->vbo.base);
 }
 
 
@@ -597,7 +595,6 @@ int vmw_user_bo_alloc(struct vmw_private *dev_priv,
 		      struct ttm_base_object **p_base)
 {
 	struct vmw_user_buffer_object *user_bo;
-	struct ttm_buffer_object *tmp;
 	int ret;
 
 	user_bo = kzalloc(sizeof(*user_bo), GFP_KERNEL);
@@ -614,7 +611,7 @@ int vmw_user_bo_alloc(struct vmw_private *dev_priv,
 	if (unlikely(ret != 0))
 		return ret;
 
-	tmp = ttm_bo_reference(&user_bo->vbo.base);
+	ttm_bo_get(&user_bo->vbo.base);
 	ret = ttm_prime_object_init(tfile,
 				    size,
 				    &user_bo->prime,
@@ -623,7 +620,7 @@ int vmw_user_bo_alloc(struct vmw_private *dev_priv,
 				    &vmw_user_bo_release,
 				    &vmw_user_bo_ref_obj_release);
 	if (unlikely(ret != 0)) {
-		ttm_bo_unref(&tmp);
+		ttm_bo_put(&user_bo->vbo.base);
 		goto out_no_base_object;
 	}
 
@@ -911,7 +908,7 @@ int vmw_user_bo_lookup(struct ttm_object_file *tfile,
 
 	vmw_user_bo = container_of(base, struct vmw_user_buffer_object,
 				   prime.base);
-	(void)ttm_bo_reference(&vmw_user_bo->vbo.base);
+	ttm_bo_get(&vmw_user_bo->vbo.base);
 	if (p_base)
 		*p_base = base;
 	else
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c
index 48d1380a952e..70dab55e7888 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_cmdbuf.c
@@ -765,7 +765,7 @@ static bool vmw_cmdbuf_try_alloc(struct vmw_cmdbuf_man *man,
 
 	if (info->done)
 		return true;
- 
+
 	memset(info->node, 0, sizeof(*info->node));
 	spin_lock(&man->lock);
 	ret = drm_mm_insert_node(&man->mm, info->node, info->page_size);
@@ -1276,8 +1276,10 @@ int vmw_cmdbuf_set_pool_size(struct vmw_cmdbuf_man *man,
 	return 0;
 
 out_no_map:
-	if (man->using_mob)
-		ttm_bo_unref(&man->cmd_space);
+	if (man->using_mob) {
+		ttm_bo_put(man->cmd_space);
+		man->cmd_space = NULL;
+	}
 
 	return ret;
 }
@@ -1380,7 +1382,8 @@ void vmw_cmdbuf_remove_pool(struct vmw_cmdbuf_man *man)
 	(void) vmw_cmdbuf_idle(man, false, 10*HZ);
 	if (man->using_mob) {
 		(void) ttm_bo_kunmap(&man->map_obj);
-		ttm_bo_unref(&man->cmd_space);
+		ttm_bo_put(man->cmd_space);
+		man->cmd_space = NULL;
 	} else {
 		dma_free_coherent(&man->dev_priv->dev->pdev->dev,
 				  man->size, man->map, man->handle);
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
index cd607ba9c2fe..accb2fafe2f1 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_drv.h
@@ -1337,18 +1337,15 @@ static inline void vmw_bo_unreference(struct vmw_buffer_object **buf)
 
 	*buf = NULL;
 	if (tmp_buf != NULL) {
-		struct ttm_buffer_object *bo = &tmp_buf->base;
-
-		ttm_bo_unref(&bo);
+		ttm_bo_put(&tmp_buf->base);
 	}
 }
 
 static inline struct vmw_buffer_object *
 vmw_bo_reference(struct vmw_buffer_object *buf)
 {
-	if (ttm_bo_reference(&buf->base))
-		return buf;
-	return NULL;
+	ttm_bo_get(&buf->base);
+	return buf;
 }
 
 static inline struct ttm_mem_global *vmw_mem_glob(struct vmw_private *dev_priv)
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
index 7ed179d30ec5..d83cc66e1210 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_mob.c
@@ -300,7 +300,8 @@ out_no_setup:
 						 &batch->otables[i]);
 	}
 
-	ttm_bo_unref(&batch->otable_bo);
+	ttm_bo_put(batch->otable_bo);
+	batch->otable_bo = NULL;
 out_no_bo:
 	return ret;
 }
@@ -365,7 +366,8 @@ static void vmw_otable_batch_takedown(struct vmw_private *dev_priv,
 	vmw_bo_fence_single(bo, NULL);
 	ttm_bo_unreserve(bo);
 
-	ttm_bo_unref(&batch->otable_bo);
+	ttm_bo_put(batch->otable_bo);
+	batch->otable_bo = NULL;
 }
 
 /*
@@ -463,7 +465,8 @@ static int vmw_mob_pt_populate(struct vmw_private *dev_priv,
 
 out_unreserve:
 	ttm_bo_unreserve(mob->pt_bo);
-	ttm_bo_unref(&mob->pt_bo);
+	ttm_bo_put(mob->pt_bo);
+	mob->pt_bo = NULL;
 
 	return ret;
 }
@@ -580,8 +583,10 @@ static void vmw_mob_pt_setup(struct vmw_mob *mob,
  */
 void vmw_mob_destroy(struct vmw_mob *mob)
 {
-	if (mob->pt_bo)
-		ttm_bo_unref(&mob->pt_bo);
+	if (mob->pt_bo) {
+		ttm_bo_put(mob->pt_bo);
+		mob->pt_bo = NULL;
+	}
 	kfree(mob);
 }
 
@@ -698,8 +703,10 @@ int vmw_mob_bind(struct vmw_private *dev_priv,
 
 out_no_cmd_space:
 	vmw_fifo_resource_dec(dev_priv);
-	if (pt_set_up)
-		ttm_bo_unref(&mob->pt_bo);
+	if (pt_set_up) {
+		ttm_bo_put(mob->pt_bo);
+		mob->pt_bo = NULL;
+	}
 
 	return -ENOMEM;
 }
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
index 3025bfc001a1..a7c30e567f09 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_resource.c
@@ -461,7 +461,8 @@ vmw_resource_check_buffer(struct ww_acquire_ctx *ticket,
 	}
 
 	INIT_LIST_HEAD(&val_list);
-	val_buf->bo = ttm_bo_reference(&res->backup->base);
+	ttm_bo_get(&res->backup->base);
+	val_buf->bo = &res->backup->base;
 	val_buf->num_shared = 0;
 	list_add_tail(&val_buf->head, &val_list);
 	ret = ttm_eu_reserve_buffers(ticket, &val_list, interruptible, NULL);
@@ -484,7 +485,8 @@ vmw_resource_check_buffer(struct ww_acquire_ctx *ticket,
 out_no_validate:
 	ttm_eu_backoff_reservation(ticket, &val_list);
 out_no_reserve:
-	ttm_bo_unref(&val_buf->bo);
+	ttm_bo_put(val_buf->bo);
+	val_buf->bo = NULL;
 	if (backup_dirty)
 		vmw_bo_unreference(&res->backup);
 
@@ -544,7 +546,8 @@ vmw_resource_backoff_reservation(struct ww_acquire_ctx *ticket,
 	INIT_LIST_HEAD(&val_list);
 	list_add_tail(&val_buf->head, &val_list);
 	ttm_eu_backoff_reservation(ticket, &val_list);
-	ttm_bo_unref(&val_buf->bo);
+	ttm_bo_put(val_buf->bo);
+	val_buf->bo = NULL;
 }
 
 /**
diff --git a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c
index b3f547fc5d3d..e9944ac2e057 100644
--- a/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c
+++ b/drivers/gpu/drm/vmwgfx/vmwgfx_validation.c
@@ -628,8 +628,10 @@ void vmw_validation_unref_lists(struct vmw_validation_context *ctx)
 	struct vmw_validation_bo_node *entry;
 	struct vmw_validation_res_node *val;
 
-	list_for_each_entry(entry, &ctx->bo_list, base.head)
-		ttm_bo_unref(&entry->base.bo);
+	list_for_each_entry(entry, &ctx->bo_list, base.head) {
+		ttm_bo_put(entry->base.bo);
+		entry->base.bo = NULL;
+	}
 
 	list_splice_init(&ctx->resource_ctx_list, &ctx->resource_list);
 	list_for_each_entry(val, &ctx->resource_list, head)