Merge tag 'amd-drm-fixes-5.7-2020-04-29' of...

Merge tag 'amd-drm-fixes-5.7-2020-04-29' of git://people.freedesktop.org/~agd5f/linux into drm-fixes

amd-drm-fixes-5.7-2020-04-29:

amdgpu:
- Fix a green screen on resume issue
- PM fixes for SR-IOV
- SDMA fix for navi
- Renoir display fixes
- Cursor and pageflip stuttering fixes
- Misc additional display fixes

UAPI:
- Add additional DCC tiling flags for navi1x
  Used by: https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/4697Signed-off-by: Dave Airlie <airlied@redhat.com>
From: Alex Deucher <alexdeucher@gmail.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20200429212008.4306-1-alexander.deucher@amd.com

drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c

@@ -85,9 +85,10 @@
  * - 3.34.0 - Non-DC can flip correctly between buffers with different pitches
  * - 3.35.0 - Add drm_amdgpu_info_device::tcc_disabled_mask
  * - 3.36.0 - Allow reading more status registers on si/cik
+ * - 3.37.0 - L2 is invalidated before SDMA IBs, needed for correctness
  */
 #define KMS_DRIVER_MAJOR	3
-#define KMS_DRIVER_MINOR	36
+#define KMS_DRIVER_MINOR	37
 #define KMS_DRIVER_PATCHLEVEL	0
 
 int amdgpu_vram_limit = 0;

drivers/gpu/drm/amd/amdgpu/navi10_sdma_pkt_open.h

@@ -73,6 +73,22 @@
 #define SDMA_OP_AQL_COPY  0
 #define SDMA_OP_AQL_BARRIER_OR  0
 
+#define SDMA_GCR_RANGE_IS_PA		(1 << 18)
+#define SDMA_GCR_SEQ(x)			(((x) & 0x3) << 16)
+#define SDMA_GCR_GL2_WB			(1 << 15)
+#define SDMA_GCR_GL2_INV		(1 << 14)
+#define SDMA_GCR_GL2_DISCARD		(1 << 13)
+#define SDMA_GCR_GL2_RANGE(x)		(((x) & 0x3) << 11)
+#define SDMA_GCR_GL2_US			(1 << 10)
+#define SDMA_GCR_GL1_INV		(1 << 9)
+#define SDMA_GCR_GLV_INV		(1 << 8)
+#define SDMA_GCR_GLK_INV		(1 << 7)
+#define SDMA_GCR_GLK_WB			(1 << 6)
+#define SDMA_GCR_GLM_INV		(1 << 5)
+#define SDMA_GCR_GLM_WB			(1 << 4)
+#define SDMA_GCR_GL1_RANGE(x)		(((x) & 0x3) << 2)
+#define SDMA_GCR_GLI_INV(x)		(((x) & 0x3) << 0)
+
 /*define for op field*/
 #define SDMA_PKT_HEADER_op_offset 0
 #define SDMA_PKT_HEADER_op_mask   0x000000FF

drivers/gpu/drm/amd/amdgpu/sdma_v5_0.c

@@ -382,6 +382,18 @@ static void sdma_v5_0_ring_emit_ib(struct amdgpu_ring *ring,
 	unsigned vmid = AMDGPU_JOB_GET_VMID(job);
 	uint64_t csa_mc_addr = amdgpu_sdma_get_csa_mc_addr(ring, vmid);
 
+	/* Invalidate L2, because if we don't do it, we might get stale cache
+	 * lines from previous IBs.
+	 */
+	amdgpu_ring_write(ring, SDMA_PKT_HEADER_OP(SDMA_OP_GCR_REQ));
+	amdgpu_ring_write(ring, 0);
+	amdgpu_ring_write(ring, (SDMA_GCR_GL2_INV |
+				 SDMA_GCR_GL2_WB |
+				 SDMA_GCR_GLM_INV |
+				 SDMA_GCR_GLM_WB) << 16);
+	amdgpu_ring_write(ring, 0xffffff80);
+	amdgpu_ring_write(ring, 0xffff);
+
 	/* An IB packet must end on a 8 DW boundary--the next dword
 	 * must be on a 8-dword boundary. Our IB packet below is 6
 	 * dwords long, thus add x number of NOPs, such that, in
@@ -1595,7 +1607,7 @@ static const struct amdgpu_ring_funcs sdma_v5_0_ring_funcs = {
 		SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 +
 		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 6 * 2 +
 		10 + 10 + 10, /* sdma_v5_0_ring_emit_fence x3 for user fence, vm fence */
-	.emit_ib_size = 7 + 6, /* sdma_v5_0_ring_emit_ib */
+	.emit_ib_size = 5 + 7 + 6, /* sdma_v5_0_ring_emit_ib */
 	.emit_ib = sdma_v5_0_ring_emit_ib,
 	.emit_fence = sdma_v5_0_ring_emit_fence,
 	.emit_pipeline_sync = sdma_v5_0_ring_emit_pipeline_sync,

drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c

@@ -3340,7 +3340,8 @@ fill_plane_dcc_attributes(struct amdgpu_device *adev,
 			  const union dc_tiling_info *tiling_info,
 			  const uint64_t info,
 			  struct dc_plane_dcc_param *dcc,
-			  struct dc_plane_address *address)
+			  struct dc_plane_address *address,
+			  bool force_disable_dcc)
 {
 	struct dc *dc = adev->dm.dc;
 	struct dc_dcc_surface_param input;
@@ -3352,6 +3353,9 @@ fill_plane_dcc_attributes(struct amdgpu_device *adev,
 	memset(&input, 0, sizeof(input));
 	memset(&output, 0, sizeof(output));
 
+	if (force_disable_dcc)
+		return 0;
+
 	if (!offset)
 		return 0;
 
@@ -3401,7 +3405,8 @@ fill_plane_buffer_attributes(struct amdgpu_device *adev,
 			     union dc_tiling_info *tiling_info,
 			     struct plane_size *plane_size,
 			     struct dc_plane_dcc_param *dcc,
-			     struct dc_plane_address *address)
+			     struct dc_plane_address *address,
+			     bool force_disable_dcc)
 {
 	const struct drm_framebuffer *fb = &afb->base;
 	int ret;
@@ -3507,7 +3512,8 @@ fill_plane_buffer_attributes(struct amdgpu_device *adev,
 
 		ret = fill_plane_dcc_attributes(adev, afb, format, rotation,
 						plane_size, tiling_info,
-						tiling_flags, dcc, address);
+						tiling_flags, dcc, address,
+						force_disable_dcc);
 		if (ret)
 			return ret;
 	}
@@ -3599,7 +3605,8 @@ fill_dc_plane_info_and_addr(struct amdgpu_device *adev,
 			    const struct drm_plane_state *plane_state,
 			    const uint64_t tiling_flags,
 			    struct dc_plane_info *plane_info,
-			    struct dc_plane_address *address)
+			    struct dc_plane_address *address,
+			    bool force_disable_dcc)
 {
 	const struct drm_framebuffer *fb = plane_state->fb;
 	const struct amdgpu_framebuffer *afb =
@@ -3681,7 +3688,8 @@ fill_dc_plane_info_and_addr(struct amdgpu_device *adev,
 					   plane_info->rotation, tiling_flags,
 					   &plane_info->tiling_info,
 					   &plane_info->plane_size,
-					   &plane_info->dcc, address);
+					   &plane_info->dcc, address,
+					   force_disable_dcc);
 	if (ret)
 		return ret;
 
@@ -3704,6 +3712,7 @@ static int fill_dc_plane_attributes(struct amdgpu_device *adev,
 	struct dc_plane_info plane_info;
 	uint64_t tiling_flags;
 	int ret;
+	bool force_disable_dcc = false;
 
 	ret = fill_dc_scaling_info(plane_state, &scaling_info);
 	if (ret)
@@ -3718,9 +3727,11 @@ static int fill_dc_plane_attributes(struct amdgpu_device *adev,
 	if (ret)
 		return ret;
 
+	force_disable_dcc = adev->asic_type == CHIP_RAVEN && adev->in_suspend;
 	ret = fill_dc_plane_info_and_addr(adev, plane_state, tiling_flags,
 					  &plane_info,
-					  &dc_plane_state->address);
+					  &dc_plane_state->address,
+					  force_disable_dcc);
 	if (ret)
 		return ret;
 
@@ -5342,6 +5353,7 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane,
 	uint64_t tiling_flags;
 	uint32_t domain;
 	int r;
+	bool force_disable_dcc = false;
 
 	dm_plane_state_old = to_dm_plane_state(plane->state);
 	dm_plane_state_new = to_dm_plane_state(new_state);
@@ -5400,11 +5412,13 @@ static int dm_plane_helper_prepare_fb(struct drm_plane *plane,
 			dm_plane_state_old->dc_state != dm_plane_state_new->dc_state) {
 		struct dc_plane_state *plane_state = dm_plane_state_new->dc_state;
 
+		force_disable_dcc = adev->asic_type == CHIP_RAVEN && adev->in_suspend;
 		fill_plane_buffer_attributes(
 			adev, afb, plane_state->format, plane_state->rotation,
 			tiling_flags, &plane_state->tiling_info,
 			&plane_state->plane_size, &plane_state->dcc,
-			&plane_state->address);
+			&plane_state->address,
+			force_disable_dcc);
 	}
 
 	return 0;
@@ -6676,7 +6690,12 @@ static void amdgpu_dm_commit_planes(struct drm_atomic_state *state,
 		fill_dc_plane_info_and_addr(
 			dm->adev, new_plane_state, tiling_flags,
 			&bundle->plane_infos[planes_count],
-			&bundle->flip_addrs[planes_count].address);
+			&bundle->flip_addrs[planes_count].address,
+			false);
+
+		DRM_DEBUG_DRIVER("plane: id=%d dcc_en=%d\n",
+				 new_plane_state->plane->index,
+				 bundle->plane_infos[planes_count].dcc.enable);
 
 		bundle->surface_updates[planes_count].plane_info =
 			&bundle->plane_infos[planes_count];
@@ -8096,7 +8115,8 @@ dm_determine_update_type_for_commit(struct amdgpu_display_manager *dm,
 				ret = fill_dc_plane_info_and_addr(
 					dm->adev, new_plane_state, tiling_flags,
 					plane_info,
-					&flip_addr->address);
+					&flip_addr->address,
+					false);
 				if (ret)
 					goto cleanup;
 

drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c

@@ -2908,6 +2908,12 @@ bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd
 					sizeof(hpd_irq_dpcd_data),
 					"Status: ");
 
+		for (i = 0; i < MAX_PIPES; i++) {
+			pipe_ctx = &link->dc->current_state->res_ctx.pipe_ctx[i];
+			if (pipe_ctx && pipe_ctx->stream && pipe_ctx->stream->link == link)
+				link->dc->hwss.blank_stream(pipe_ctx);
+		}
+
 		for (i = 0; i < MAX_PIPES; i++) {
 			pipe_ctx = &link->dc->current_state->res_ctx.pipe_ctx[i];
 			if (pipe_ctx && pipe_ctx->stream && pipe_ctx->stream->link == link)
@@ -2927,6 +2933,12 @@ bool dc_link_handle_hpd_rx_irq(struct dc_link *link, union hpd_irq_data *out_hpd
 		if (pipe_ctx->stream->signal == SIGNAL_TYPE_DISPLAY_PORT_MST)
 			dc_link_reallocate_mst_payload(link);
 
+		for (i = 0; i < MAX_PIPES; i++) {
+			pipe_ctx = &link->dc->current_state->res_ctx.pipe_ctx[i];
+			if (pipe_ctx && pipe_ctx->stream && pipe_ctx->stream->link == link)
+				link->dc->hwss.unblank_stream(pipe_ctx, &previous_link_settings);
+		}
+
 		status = false;
 		if (out_link_loss)
 			*out_link_loss = true;
@@ -4227,6 +4239,21 @@ void dp_set_fec_enable(struct dc_link *link, bool enable)
 void dpcd_set_source_specific_data(struct dc_link *link)
 {
 	const uint32_t post_oui_delay = 30; // 30ms
+	uint8_t dspc = 0;
+	enum dc_status ret = DC_ERROR_UNEXPECTED;
+
+	ret = core_link_read_dpcd(link, DP_DOWN_STREAM_PORT_COUNT, &dspc,
+				  sizeof(dspc));
+
+	if (ret != DC_OK) {
+		DC_LOG_ERROR("Error in DP aux read transaction,"
+			     " not writing source specific data\n");
+		return;
+	}
+
+	/* Return if OUI unsupported */
+	if (!(dspc & DP_OUI_SUPPORT))
+		return;
 
 	if (!link->dc->vendor_signature.is_valid) {
 		struct dpcd_amd_signature amd_signature;

drivers/gpu/drm/amd/display/dc/core/dc_stream.c

@@ -231,34 +231,6 @@ struct dc_stream_status *dc_stream_get_status(
 	return dc_stream_get_status_from_state(dc->current_state, stream);
 }
 
-static void delay_cursor_until_vupdate(struct pipe_ctx *pipe_ctx, struct dc *dc)
-{
-#if defined(CONFIG_DRM_AMD_DC_DCN)
-	unsigned int vupdate_line;
-	unsigned int lines_to_vupdate, us_to_vupdate, vpos, nvpos;
-	struct dc_stream_state *stream = pipe_ctx->stream;
-	unsigned int us_per_line;
-
-	if (stream->ctx->asic_id.chip_family == FAMILY_RV &&
-			ASICREV_IS_RAVEN(stream->ctx->asic_id.hw_internal_rev)) {
-
-		vupdate_line = dc->hwss.get_vupdate_offset_from_vsync(pipe_ctx);
-		if (!dc_stream_get_crtc_position(dc, &stream, 1, &vpos, &nvpos))
-			return;
-
-		if (vpos >= vupdate_line)
-			return;
-
-		us_per_line = stream->timing.h_total * 10000 / stream->timing.pix_clk_100hz;
-		lines_to_vupdate = vupdate_line - vpos;
-		us_to_vupdate = lines_to_vupdate * us_per_line;
-
-		/* 70 us is a conservative estimate of cursor update time*/
-		if (us_to_vupdate < 70)
-			udelay(us_to_vupdate);
-	}
-#endif
-}
 
 /**
  * dc_stream_set_cursor_attributes() - Update cursor attributes and set cursor surface address
@@ -298,9 +270,7 @@ bool dc_stream_set_cursor_attributes(
 
 		if (!pipe_to_program) {
 			pipe_to_program = pipe_ctx;
-
-			delay_cursor_until_vupdate(pipe_ctx, dc);
-			dc->hwss.pipe_control_lock(dc, pipe_to_program, true);
+			dc->hwss.cursor_lock(dc, pipe_to_program, true);
 		}
 
 		dc->hwss.set_cursor_attribute(pipe_ctx);
@@ -309,7 +279,7 @@ bool dc_stream_set_cursor_attributes(
 	}
 
 	if (pipe_to_program)
-		dc->hwss.pipe_control_lock(dc, pipe_to_program, false);
+		dc->hwss.cursor_lock(dc, pipe_to_program, false);
 
 	return true;
 }
@@ -349,16 +319,14 @@ bool dc_stream_set_cursor_position(
 
 		if (!pipe_to_program) {
 			pipe_to_program = pipe_ctx;
-
-			delay_cursor_until_vupdate(pipe_ctx, dc);
-			dc->hwss.pipe_control_lock(dc, pipe_to_program, true);
+			dc->hwss.cursor_lock(dc, pipe_to_program, true);
 		}
 
 		dc->hwss.set_cursor_position(pipe_ctx);
 	}
 
 	if (pipe_to_program)
-		dc->hwss.pipe_control_lock(dc, pipe_to_program, false);
+		dc->hwss.cursor_lock(dc, pipe_to_program, false);
 
 	return true;
 }

drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c

@@ -2757,6 +2757,7 @@ static const struct hw_sequencer_funcs dce110_funcs = {
 	.disable_plane = dce110_power_down_fe,
 	.pipe_control_lock = dce_pipe_control_lock,
 	.interdependent_update_lock = NULL,
+	.cursor_lock = dce_pipe_control_lock,
 	.prepare_bandwidth = dce110_prepare_bandwidth,
 	.optimize_bandwidth = dce110_optimize_bandwidth,
 	.set_drr = set_drr,

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c

@@ -1625,6 +1625,16 @@ void dcn10_pipe_control_lock(
 		hws->funcs.verify_allow_pstate_change_high(dc);
 }
 
+void dcn10_cursor_lock(struct dc *dc, struct pipe_ctx *pipe, bool lock)
+{
+	/* cursor lock is per MPCC tree, so only need to lock one pipe per stream */
+	if (!pipe || pipe->top_pipe)
+		return;
+
+	dc->res_pool->mpc->funcs->cursor_lock(dc->res_pool->mpc,
+			pipe->stream_res.opp->inst, lock);
+}
+
 static bool wait_for_reset_trigger_to_occur(
 	struct dc_context *dc_ctx,
 	struct timing_generator *tg)

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h

@@ -49,6 +49,7 @@ void dcn10_pipe_control_lock(
 	struct dc *dc,
 	struct pipe_ctx *pipe,
 	bool lock);
+void dcn10_cursor_lock(struct dc *dc, struct pipe_ctx *pipe, bool lock);
 void dcn10_blank_pixel_data(
 		struct dc *dc,
 		struct pipe_ctx *pipe_ctx,

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_init.c

@@ -50,6 +50,7 @@ static const struct hw_sequencer_funcs dcn10_funcs = {
 	.disable_audio_stream = dce110_disable_audio_stream,
 	.disable_plane = dcn10_disable_plane,
 	.pipe_control_lock = dcn10_pipe_control_lock,
+	.cursor_lock = dcn10_cursor_lock,
 	.interdependent_update_lock = dcn10_lock_all_pipes,
 	.prepare_bandwidth = dcn10_prepare_bandwidth,
 	.optimize_bandwidth = dcn10_optimize_bandwidth,

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.c

@@ -223,6 +223,9 @@ struct mpcc *mpc1_insert_plane(
 	REG_SET(MPCC_TOP_SEL[mpcc_id], 0, MPCC_TOP_SEL, dpp_id);
 	REG_SET(MPCC_OPP_ID[mpcc_id], 0, MPCC_OPP_ID, tree->opp_id);
 
+	/* Configure VUPDATE lock set for this MPCC to map to the OPP */
+	REG_SET(MPCC_UPDATE_LOCK_SEL[mpcc_id], 0, MPCC_UPDATE_LOCK_SEL, tree->opp_id);
+
 	/* update mpc tree mux setting */
 	if (tree->opp_list == insert_above_mpcc) {
 		/* insert the toppest mpcc */
@@ -318,6 +321,7 @@ void mpc1_remove_mpcc(
 		REG_SET(MPCC_TOP_SEL[mpcc_id], 0, MPCC_TOP_SEL, 0xf);
 		REG_SET(MPCC_BOT_SEL[mpcc_id], 0, MPCC_BOT_SEL, 0xf);
 		REG_SET(MPCC_OPP_ID[mpcc_id],  0, MPCC_OPP_ID,  0xf);
+		REG_SET(MPCC_UPDATE_LOCK_SEL[mpcc_id], 0, MPCC_UPDATE_LOCK_SEL, 0xf);
 
 		/* mark this mpcc as not in use */
 		mpc10->mpcc_in_use_mask &= ~(1 << mpcc_id);
@@ -328,6 +332,7 @@ void mpc1_remove_mpcc(
 		REG_SET(MPCC_TOP_SEL[mpcc_id], 0, MPCC_TOP_SEL, 0xf);
 		REG_SET(MPCC_BOT_SEL[mpcc_id], 0, MPCC_BOT_SEL, 0xf);
 		REG_SET(MPCC_OPP_ID[mpcc_id],  0, MPCC_OPP_ID,  0xf);
+		REG_SET(MPCC_UPDATE_LOCK_SEL[mpcc_id], 0, MPCC_UPDATE_LOCK_SEL, 0xf);
 	}
 }
 
@@ -361,6 +366,7 @@ void mpc1_mpc_init(struct mpc *mpc)
 		REG_SET(MPCC_TOP_SEL[mpcc_id], 0, MPCC_TOP_SEL, 0xf);
 		REG_SET(MPCC_BOT_SEL[mpcc_id], 0, MPCC_BOT_SEL, 0xf);
 		REG_SET(MPCC_OPP_ID[mpcc_id],  0, MPCC_OPP_ID,  0xf);
+		REG_SET(MPCC_UPDATE_LOCK_SEL[mpcc_id], 0, MPCC_UPDATE_LOCK_SEL, 0xf);
 
 		mpc1_init_mpcc(&(mpc->mpcc_array[mpcc_id]), mpcc_id);
 	}
@@ -381,6 +387,7 @@ void mpc1_mpc_init_single_inst(struct mpc *mpc, unsigned int mpcc_id)
 	REG_SET(MPCC_TOP_SEL[mpcc_id], 0, MPCC_TOP_SEL, 0xf);
 	REG_SET(MPCC_BOT_SEL[mpcc_id], 0, MPCC_BOT_SEL, 0xf);
 	REG_SET(MPCC_OPP_ID[mpcc_id],  0, MPCC_OPP_ID,  0xf);
+	REG_SET(MPCC_UPDATE_LOCK_SEL[mpcc_id], 0, MPCC_UPDATE_LOCK_SEL, 0xf);
 
 	mpc1_init_mpcc(&(mpc->mpcc_array[mpcc_id]), mpcc_id);
 
@@ -453,6 +460,13 @@ void mpc1_read_mpcc_state(
 			MPCC_BUSY, &s->busy);
 }
 
+void mpc1_cursor_lock(struct mpc *mpc, int opp_id, bool lock)
+{
+	struct dcn10_mpc *mpc10 = TO_DCN10_MPC(mpc);
+
+	REG_SET(CUR[opp_id], 0, CUR_VUPDATE_LOCK_SET, lock ? 1 : 0);
+}
+
 static const struct mpc_funcs dcn10_mpc_funcs = {
 	.read_mpcc_state = mpc1_read_mpcc_state,
 	.insert_plane = mpc1_insert_plane,
@@ -464,6 +478,7 @@ static const struct mpc_funcs dcn10_mpc_funcs = {
 	.assert_mpcc_idle_before_connect = mpc1_assert_mpcc_idle_before_connect,
 	.init_mpcc_list_from_hw = mpc1_init_mpcc_list_from_hw,
 	.update_blending = mpc1_update_blending,
+	.cursor_lock = mpc1_cursor_lock,
 	.set_denorm = NULL,
 	.set_denorm_clamp = NULL,
 	.set_output_csc = NULL,

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_mpc.h

@@ -39,11 +39,12 @@
 	SRII(MPCC_BG_G_Y, MPCC, inst),\
 	SRII(MPCC_BG_R_CR, MPCC, inst),\
 	SRII(MPCC_BG_B_CB, MPCC, inst),\
-	SRII(MPCC_BG_B_CB, MPCC, inst),\
-	SRII(MPCC_SM_CONTROL, MPCC, inst)
+	SRII(MPCC_SM_CONTROL, MPCC, inst),\
+	SRII(MPCC_UPDATE_LOCK_SEL, MPCC, inst)
 
 #define MPC_OUT_MUX_COMMON_REG_LIST_DCN1_0(inst) \
-	SRII(MUX, MPC_OUT, inst)
+	SRII(MUX, MPC_OUT, inst),\
+	VUPDATE_SRII(CUR, VUPDATE_LOCK_SET, inst)
 
 #define MPC_COMMON_REG_VARIABLE_LIST \
 	uint32_t MPCC_TOP_SEL[MAX_MPCC]; \
@@ -55,7 +56,9 @@
 	uint32_t MPCC_BG_R_CR[MAX_MPCC]; \
 	uint32_t MPCC_BG_B_CB[MAX_MPCC]; \
 	uint32_t MPCC_SM_CONTROL[MAX_MPCC]; \
-	uint32_t MUX[MAX_OPP];
+	uint32_t MUX[MAX_OPP]; \
+	uint32_t MPCC_UPDATE_LOCK_SEL[MAX_MPCC]; \
+	uint32_t CUR[MAX_OPP];
 
 #define MPC_COMMON_MASK_SH_LIST_DCN1_0(mask_sh)\
 	SF(MPCC0_MPCC_TOP_SEL, MPCC_TOP_SEL, mask_sh),\
@@ -78,7 +81,8 @@
 	SF(MPCC0_MPCC_SM_CONTROL, MPCC_SM_FIELD_ALT, mask_sh),\
 	SF(MPCC0_MPCC_SM_CONTROL, MPCC_SM_FORCE_NEXT_FRAME_POL, mask_sh),\
 	SF(MPCC0_MPCC_SM_CONTROL, MPCC_SM_FORCE_NEXT_TOP_POL, mask_sh),\
-	SF(MPC_OUT0_MUX, MPC_OUT_MUX, mask_sh)
+	SF(MPC_OUT0_MUX, MPC_OUT_MUX, mask_sh),\
+	SF(MPCC0_MPCC_UPDATE_LOCK_SEL, MPCC_UPDATE_LOCK_SEL, mask_sh)
 
 #define MPC_REG_FIELD_LIST(type) \
 	type MPCC_TOP_SEL;\
@@ -101,7 +105,9 @@
 	type MPCC_SM_FIELD_ALT;\
 	type MPCC_SM_FORCE_NEXT_FRAME_POL;\
 	type MPCC_SM_FORCE_NEXT_TOP_POL;\
-	type MPC_OUT_MUX;
+	type MPC_OUT_MUX;\
+	type MPCC_UPDATE_LOCK_SEL;\
+	type CUR_VUPDATE_LOCK_SET;
 
 struct dcn_mpc_registers {
 	MPC_COMMON_REG_VARIABLE_LIST
@@ -192,4 +198,6 @@ void mpc1_read_mpcc_state(
 		int mpcc_inst,
 		struct mpcc_state *s);
 
+void mpc1_cursor_lock(struct mpc *mpc, int opp_id, bool lock);
+
 #endif

drivers/gpu/drm/amd/display/dc/dcn10/dcn10_resource.c

@@ -181,6 +181,14 @@ enum dcn10_clk_src_array_id {
 	.reg_name[id] = BASE(mm ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
 					mm ## block ## id ## _ ## reg_name
 
+#define VUPDATE_SRII(reg_name, block, id)\
+	.reg_name[id] = BASE(mm ## reg_name ## 0 ## _ ## block ## id ## _BASE_IDX) + \
+					mm ## reg_name ## 0 ## _ ## block ## id
+
+/* set field/register/bitfield name */
+#define SFRB(field_name, reg_name, bitfield, post_fix)\
+	.field_name = reg_name ## __ ## bitfield ## post_fix
+
 /* NBIO */
 #define NBIO_BASE_INNER(seg) \
 	NBIF_BASE__INST0_SEG ## seg
@@ -419,11 +427,13 @@ static const struct dcn_mpc_registers mpc_regs = {
 };
 
 static const struct dcn_mpc_shift mpc_shift = {
-	MPC_COMMON_MASK_SH_LIST_DCN1_0(__SHIFT)
+	MPC_COMMON_MASK_SH_LIST_DCN1_0(__SHIFT),\
+	SFRB(CUR_VUPDATE_LOCK_SET, CUR0_VUPDATE_LOCK_SET0, CUR0_VUPDATE_LOCK_SET, __SHIFT)
 };
 
 static const struct dcn_mpc_mask mpc_mask = {
-	MPC_COMMON_MASK_SH_LIST_DCN1_0(_MASK),
+	MPC_COMMON_MASK_SH_LIST_DCN1_0(_MASK),\
+	SFRB(CUR_VUPDATE_LOCK_SET, CUR0_VUPDATE_LOCK_SET0, CUR0_VUPDATE_LOCK_SET, _MASK)
 };
 
 #define tg_regs(id)\

drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c

@@ -2294,6 +2294,7 @@ void dcn20_fpga_init_hw(struct dc *dc)
 
 	REG_UPDATE(DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_REFDIV, 2);
 	REG_UPDATE(DCHUBBUB_GLOBAL_TIMER_CNTL, DCHUBBUB_GLOBAL_TIMER_ENABLE, 1);
+	if (REG(REFCLK_CNTL))
 		REG_WRITE(REFCLK_CNTL, 0);
 	//
 

drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c

@@ -52,6 +52,7 @@ static const struct hw_sequencer_funcs dcn20_funcs = {
 	.disable_plane = dcn20_disable_plane,
 	.pipe_control_lock = dcn20_pipe_control_lock,
 	.interdependent_update_lock = dcn10_lock_all_pipes,
+	.cursor_lock = dcn10_cursor_lock,
 	.prepare_bandwidth = dcn20_prepare_bandwidth,
 	.optimize_bandwidth = dcn20_optimize_bandwidth,
 	.update_bandwidth = dcn20_update_bandwidth,

drivers/gpu/drm/amd/display/dc/dcn20/dcn20_mpc.c

@@ -545,6 +545,7 @@ const struct mpc_funcs dcn20_mpc_funcs = {
 	.mpc_init = mpc1_mpc_init,
 	.mpc_init_single_inst = mpc1_mpc_init_single_inst,
 	.update_blending = mpc2_update_blending,
+	.cursor_lock = mpc1_cursor_lock,
 	.get_mpcc_for_dpp = mpc2_get_mpcc_for_dpp,
 	.wait_for_idle = mpc2_assert_idle_mpcc,
 	.assert_mpcc_idle_before_connect = mpc2_assert_mpcc_idle_before_connect,

drivers/gpu/drm/amd/display/dc/dcn20/dcn20_mpc.h

@@ -179,7 +179,8 @@
 	SF(MPC_OUT0_DENORM_CLAMP_G_Y, MPC_OUT_DENORM_CLAMP_MAX_G_Y, mask_sh),\
 	SF(MPC_OUT0_DENORM_CLAMP_G_Y, MPC_OUT_DENORM_CLAMP_MIN_G_Y, mask_sh),\
 	SF(MPC_OUT0_DENORM_CLAMP_B_CB, MPC_OUT_DENORM_CLAMP_MAX_B_CB, mask_sh),\
-	SF(MPC_OUT0_DENORM_CLAMP_B_CB, MPC_OUT_DENORM_CLAMP_MIN_B_CB, mask_sh)
+	SF(MPC_OUT0_DENORM_CLAMP_B_CB, MPC_OUT_DENORM_CLAMP_MIN_B_CB, mask_sh),\
+	SF(CUR_VUPDATE_LOCK_SET0, CUR_VUPDATE_LOCK_SET, mask_sh)
 
 /*
  *	DCN2 MPC_OCSC debug status register:

drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c

@@ -508,6 +508,10 @@ enum dcn20_clk_src_array_id {
 	.block ## _ ## reg_name[id] = BASE(mm ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
 					mm ## block ## id ## _ ## reg_name
 
+#define VUPDATE_SRII(reg_name, block, id)\
+	.reg_name[id] = BASE(mm ## reg_name ## _ ## block ## id ## _BASE_IDX) + \
+					mm ## reg_name ## _ ## block ## id
+
 /* NBIO */
 #define NBIO_BASE_INNER(seg) \
 	NBIO_BASE__INST0_SEG ## seg

drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c

@@ -53,6 +53,7 @@ static const struct hw_sequencer_funcs dcn21_funcs = {
 	.disable_plane = dcn20_disable_plane,
 	.pipe_control_lock = dcn20_pipe_control_lock,
 	.interdependent_update_lock = dcn10_lock_all_pipes,
+	.cursor_lock = dcn10_cursor_lock,
 	.prepare_bandwidth = dcn20_prepare_bandwidth,
 	.optimize_bandwidth = dcn20_optimize_bandwidth,
 	.update_bandwidth = dcn20_update_bandwidth,

drivers/gpu/drm/amd/display/dc/dcn21/dcn21_resource.c

@@ -284,7 +284,7 @@ struct _vcs_dpi_soc_bounding_box_st dcn2_1_soc = {
 	.dram_channel_width_bytes = 4,
 	.fabric_datapath_to_dcn_data_return_bytes = 32,
 	.dcn_downspread_percent = 0.5,
-	.downspread_percent = 0.5,
+	.downspread_percent = 0.38,
 	.dram_page_open_time_ns = 50.0,
 	.dram_rw_turnaround_time_ns = 17.5,
 	.dram_return_buffer_per_channel_bytes = 8192,
@@ -340,6 +340,10 @@ struct _vcs_dpi_soc_bounding_box_st dcn2_1_soc = {
 	.block ## _ ## reg_name[id] = BASE(mm ## block ## id ## _ ## reg_name ## _BASE_IDX) + \
 					mm ## block ## id ## _ ## reg_name
 
+#define VUPDATE_SRII(reg_name, block, id)\
+	.reg_name[id] = BASE(mm ## reg_name ## _ ## block ## id ## _BASE_IDX) + \
+					mm ## reg_name ## _ ## block ## id
+
 /* NBIO */
 #define NBIO_BASE_INNER(seg) \
 	NBIF0_BASE__INST0_SEG ## seg
@@ -1374,63 +1378,48 @@ static void update_bw_bounding_box(struct dc *dc, struct clk_bw_params *bw_param
 {
 	struct dcn21_resource_pool *pool = TO_DCN21_RES_POOL(dc->res_pool);
 	struct clk_limit_table *clk_table = &bw_params->clk_table;
-	unsigned int i, j, k;
-	int closest_clk_lvl;
+	struct _vcs_dpi_voltage_scaling_st clock_limits[DC__VOLTAGE_STATES];
+	unsigned int i, j, closest_clk_lvl;
 
 	// Default clock levels are used for diags, which may lead to overclocking.
-	if (!IS_FPGA_MAXIMUS_DC(dc->ctx->dce_environment) && !IS_DIAG_DC(dc->ctx->dce_environment)) {
+	if (!IS_DIAG_DC(dc->ctx->dce_environment)) {
 		dcn2_1_ip.max_num_otg = pool->base.res_cap->num_timing_generator;
 		dcn2_1_ip.max_num_dpp = pool->base.pipe_count;
 		dcn2_1_soc.num_chans = bw_params->num_channels;
 
-		/* Vmin: leave lowest DCN clocks, override with dcfclk, fclk, memclk from fuse */
-		dcn2_1_soc.clock_limits[0].state = 0;
-		dcn2_1_soc.clock_limits[0].dcfclk_mhz = clk_table->entries[0].dcfclk_mhz;
-		dcn2_1_soc.clock_limits[0].fabricclk_mhz = clk_table->entries[0].fclk_mhz;
-		dcn2_1_soc.clock_limits[0].socclk_mhz = clk_table->entries[0].socclk_mhz;
-		dcn2_1_soc.clock_limits[0].dram_speed_mts = clk_table->entries[0].memclk_mhz * 2;
-
-		/*
-		 * Other levels: find closest DCN clocks that fit the given clock limit using dcfclk
-		 * as indicator
-		 */
-
-		closest_clk_lvl = -1;
-		/* index currently being filled */
-		k = 1;
-		for (i = 1; i < clk_table->num_entries; i++) {
-			/* loop backwards, skip duplicate state*/
-			for (j = dcn2_1_soc.num_states - 1; j >= k; j--) {
+		ASSERT(clk_table->num_entries);
+		for (i = 0; i < clk_table->num_entries; i++) {
+			/* loop backwards*/
+			for (closest_clk_lvl = 0, j = dcn2_1_soc.num_states - 1; j >= 0; j--) {
 				if ((unsigned int) dcn2_1_soc.clock_limits[j].dcfclk_mhz <= clk_table->entries[i].dcfclk_mhz) {
 					closest_clk_lvl = j;
 					break;
 				}
 			}
 
-			/* if found a lvl that fits, use the DCN clks from it, if not, go to next clk limit*/
-			if (closest_clk_lvl != -1) {
-				dcn2_1_soc.clock_limits[k].state = i;
-				dcn2_1_soc.clock_limits[k].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz;
-				dcn2_1_soc.clock_limits[k].fabricclk_mhz = clk_table->entries[i].fclk_mhz;
-				dcn2_1_soc.clock_limits[k].socclk_mhz = clk_table->entries[i].socclk_mhz;
-				dcn2_1_soc.clock_limits[k].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2;
+			clock_limits[i].state = i;
+			clock_limits[i].dcfclk_mhz = clk_table->entries[i].dcfclk_mhz;
+			clock_limits[i].fabricclk_mhz = clk_table->entries[i].fclk_mhz;
+			clock_limits[i].socclk_mhz = clk_table->entries[i].socclk_mhz;
+			clock_limits[i].dram_speed_mts = clk_table->entries[i].memclk_mhz * 2;
 
-				dcn2_1_soc.clock_limits[k].dispclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].dispclk_mhz;
-				dcn2_1_soc.clock_limits[k].dppclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].dppclk_mhz;
-				dcn2_1_soc.clock_limits[k].dram_bw_per_chan_gbps = dcn2_1_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps;
-				dcn2_1_soc.clock_limits[k].dscclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].dscclk_mhz;
-				dcn2_1_soc.clock_limits[k].dtbclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].dtbclk_mhz;
-				dcn2_1_soc.clock_limits[k].phyclk_d18_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz;
-				dcn2_1_soc.clock_limits[k].phyclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].phyclk_mhz;
-				k++;
-			}
+			clock_limits[i].dispclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].dispclk_mhz;
+			clock_limits[i].dppclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].dppclk_mhz;
+			clock_limits[i].dram_bw_per_chan_gbps = dcn2_1_soc.clock_limits[closest_clk_lvl].dram_bw_per_chan_gbps;
+			clock_limits[i].dscclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].dscclk_mhz;
+			clock_limits[i].dtbclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].dtbclk_mhz;
+			clock_limits[i].phyclk_d18_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].phyclk_d18_mhz;
+			clock_limits[i].phyclk_mhz = dcn2_1_soc.clock_limits[closest_clk_lvl].phyclk_mhz;
 		}
-		dcn2_1_soc.num_states = k;
-	}
-
+		for (i = 0; i < clk_table->num_entries; i++)
+			dcn2_1_soc.clock_limits[i] = clock_limits[i];
+		if (clk_table->num_entries) {
+			dcn2_1_soc.num_states = clk_table->num_entries;
 			/* duplicate last level */
 			dcn2_1_soc.clock_limits[dcn2_1_soc.num_states] = dcn2_1_soc.clock_limits[dcn2_1_soc.num_states - 1];
 			dcn2_1_soc.clock_limits[dcn2_1_soc.num_states].state = dcn2_1_soc.num_states;
+		}
+	}
 
 	dml_init_instance(&dc->dml, &dcn2_1_soc, &dcn2_1_ip, DML_PROJECT_DCN21);
 }

drivers/gpu/drm/amd/display/dc/inc/hw/mpc.h

@@ -210,6 +210,22 @@ struct mpc_funcs {
 		struct mpcc_blnd_cfg *blnd_cfg,
 		int mpcc_id);
 
+	/*
+	 * Lock cursor updates for the specified OPP.
+	 * OPP defines the set of MPCC that are locked together for cursor.
+	 *
+	 * Parameters:
+	 * [in] 	mpc		- MPC context.
+	 * [in]     opp_id	- The OPP to lock cursor updates on
+	 * [in]		lock	- lock/unlock the OPP
+	 *
+	 * Return:  void
+	 */
+	void (*cursor_lock)(
+			struct mpc *mpc,
+			int opp_id,
+			bool lock);
+
 	struct mpcc* (*get_mpcc_for_dpp)(
 			struct mpc_tree *tree,
 			int dpp_id);

drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h

@@ -86,6 +86,7 @@ struct hw_sequencer_funcs {
 			struct dc_state *context, bool lock);
 	void (*set_flip_control_gsl)(struct pipe_ctx *pipe_ctx,
 			bool flip_immediate);
+	void (*cursor_lock)(struct dc *dc, struct pipe_ctx *pipe, bool lock);
 
 	/* Timing Related */
 	void (*get_position)(struct pipe_ctx **pipe_ctx, int num_pipes,

drivers/gpu/drm/amd/powerplay/amd_powerplay.c

@@ -1435,7 +1435,8 @@ static int pp_get_asic_baco_capability(void *handle, bool *cap)
 	if (!hwmgr)
 		return -EINVAL;
 
-	if (!hwmgr->pm_en || !hwmgr->hwmgr_func->get_asic_baco_capability)
+	if (!(hwmgr->not_vf && amdgpu_dpm) ||
+		!hwmgr->hwmgr_func->get_asic_baco_capability)
 		return 0;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1452,8 +1453,7 @@ static int pp_get_asic_baco_state(void *handle, int *state)
 	if (!hwmgr)
 		return -EINVAL;
 
-	if (!(hwmgr->not_vf && amdgpu_dpm) ||
-		!hwmgr->hwmgr_func->get_asic_baco_state)
+	if (!hwmgr->pm_en || !hwmgr->hwmgr_func->get_asic_baco_state)
 		return 0;
 
 	mutex_lock(&hwmgr->smu_lock);
@@ -1470,7 +1470,8 @@ static int pp_set_asic_baco_state(void *handle, int state)
 	if (!hwmgr)
 		return -EINVAL;
 
-	if (!hwmgr->pm_en || !hwmgr->hwmgr_func->set_asic_baco_state)
+	if (!(hwmgr->not_vf && amdgpu_dpm) ||
+		!hwmgr->hwmgr_func->set_asic_baco_state)
 		return 0;
 
 	mutex_lock(&hwmgr->smu_lock);

include/uapi/drm/amdgpu_drm.h

@@ -346,6 +346,10 @@ struct drm_amdgpu_gem_userptr {
 #define AMDGPU_TILING_DCC_PITCH_MAX_MASK		0x3FFF
 #define AMDGPU_TILING_DCC_INDEPENDENT_64B_SHIFT		43
 #define AMDGPU_TILING_DCC_INDEPENDENT_64B_MASK		0x1
+#define AMDGPU_TILING_DCC_INDEPENDENT_128B_SHIFT	44
+#define AMDGPU_TILING_DCC_INDEPENDENT_128B_MASK		0x1
+#define AMDGPU_TILING_SCANOUT_SHIFT			63
+#define AMDGPU_TILING_SCANOUT_MASK			0x1
 
 /* Set/Get helpers for tiling flags. */
 #define AMDGPU_TILING_SET(field, value) \