[PATCH 7/9] drm/amd/display: Disconnect pipe separetely when disable pipe split

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



From: Alvin Lee <alvin.lee2@xxxxxxx>

[Why]
When changing pixel formats for HDR (e.g. ARGB -> FP16)
there are configurations that change from 2 pipes to 1 pipe.
In these cases, it seems that disconnecting MPCC and doing
a surface update at the same time(after unlocking) causes
some registers to be updated slightly faster than others
after unlocking (e.g. if the pixel format is updated to FP16
before the new surface address is programmed, we get
corruption on the screen because the pixel formats aren't
matching). We separate disconnecting MPCC from the rest
of  the  pipe programming sequence to prevent this.

[How]
Move MPCC disconnect into separate operation than the
rest of the pipe programming.

Signed-off-by: Alvin Lee <alvin.lee2@xxxxxxx>
Reviewed-by: Aric Cyr <Aric.Cyr@xxxxxxx>
Acked-by: Qingqing Zhuo <qingqing.zhuo@xxxxxxx>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c      |  10 ++
 .../amd/display/dc/dcn10/dcn10_hw_sequencer.c | 146 ++++++++++++++++++
 .../amd/display/dc/dcn10/dcn10_hw_sequencer.h |   6 +
 .../gpu/drm/amd/display/dc/dcn10/dcn10_init.c |   2 +
 .../gpu/drm/amd/display/dc/dcn20/dcn20_init.c |   2 +
 .../gpu/drm/amd/display/dc/dcn21/dcn21_init.c |   2 +
 .../gpu/drm/amd/display/dc/dcn30/dcn30_init.c |   2 +
 .../gpu/drm/amd/display/dc/inc/hw_sequencer.h |   4 +
 8 files changed, 174 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 5aa3b89471c3..ebbb8182228d 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -2300,6 +2300,7 @@ static void commit_planes_for_stream(struct dc *dc,
 		enum surface_update_type update_type,
 		struct dc_state *context)
 {
+	bool mpcc_disconnected = false;
 	int i, j;
 	struct pipe_ctx *top_pipe_to_program = NULL;
 
@@ -2330,6 +2331,15 @@ static void commit_planes_for_stream(struct dc *dc,
 		context_clock_trace(dc, context);
 	}
 
+	if (update_type != UPDATE_TYPE_FAST && dc->hwss.interdependent_update_lock &&
+		dc->hwss.disconnect_pipes && dc->hwss.wait_for_pending_cleared){
+		dc->hwss.interdependent_update_lock(dc, context, true);
+		mpcc_disconnected = dc->hwss.disconnect_pipes(dc, context);
+		dc->hwss.interdependent_update_lock(dc, context, false);
+		if (mpcc_disconnected)
+			dc->hwss.wait_for_pending_cleared(dc, context);
+	}
+
 	for (j = 0; j < dc->res_pool->pipe_count; j++) {
 		struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[j];
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
index 57cd52789606..95e9d05f884b 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c
@@ -2752,6 +2752,152 @@ static struct pipe_ctx *dcn10_find_top_pipe_for_stream(
 	return NULL;
 }
 
+bool dcn10_disconnect_pipes(
+		struct dc *dc,
+		struct dc_state *context)
+{
+		bool found_stream = false;
+		int i, j;
+		struct dce_hwseq *hws = dc->hwseq;
+		struct dc_state *old_ctx = dc->current_state;
+		bool mpcc_disconnected = false;
+		struct pipe_ctx *old_pipe;
+		struct pipe_ctx *new_pipe;
+		DC_LOGGER_INIT(dc->ctx->logger);
+
+		/* Set pipe update flags and lock pipes */
+		for (i = 0; i < dc->res_pool->pipe_count; i++) {
+			old_pipe = &dc->current_state->res_ctx.pipe_ctx[i];
+			new_pipe = &context->res_ctx.pipe_ctx[i];
+			new_pipe->update_flags.raw = 0;
+
+			if (!old_pipe->plane_state && !new_pipe->plane_state)
+				continue;
+
+			if (old_pipe->plane_state && !new_pipe->plane_state)
+				new_pipe->update_flags.bits.disable = 1;
+
+			/* Check for scl update */
+			if (memcmp(&old_pipe->plane_res.scl_data, &new_pipe->plane_res.scl_data, sizeof(struct scaler_data)))
+					new_pipe->update_flags.bits.scaler = 1;
+
+			/* Check for vp update */
+			if (memcmp(&old_pipe->plane_res.scl_data.viewport, &new_pipe->plane_res.scl_data.viewport, sizeof(struct rect))
+					|| memcmp(&old_pipe->plane_res.scl_data.viewport_c,
+						&new_pipe->plane_res.scl_data.viewport_c, sizeof(struct rect)))
+				new_pipe->update_flags.bits.viewport = 1;
+
+		}
+
+		if (!IS_DIAG_DC(dc->ctx->dce_environment)) {
+			/* Disconnect mpcc here only if losing pipe split*/
+			for (i = 0; i < dc->res_pool->pipe_count; i++) {
+				if (context->res_ctx.pipe_ctx[i].update_flags.bits.disable &&
+					old_ctx->res_ctx.pipe_ctx[i].top_pipe) {
+
+					/* Find the top pipe in the new ctx for the bottom pipe that we
+					 * want to remove by comparing the streams. If both pipes are being
+					 * disabled then do it in the regular pipe programming sequence
+					 */
+					for (j = 0; j < dc->res_pool->pipe_count; j++) {
+						if (old_ctx->res_ctx.pipe_ctx[i].top_pipe->stream == context->res_ctx.pipe_ctx[j].stream &&
+							!context->res_ctx.pipe_ctx[j].top_pipe &&
+							!context->res_ctx.pipe_ctx[j].update_flags.bits.disable) {
+							found_stream = true;
+							break;
+						}
+					}
+
+					// Disconnect if the top pipe lost it's pipe split
+					if (found_stream && !context->res_ctx.pipe_ctx[j].bottom_pipe) {
+						hws->funcs.plane_atomic_disconnect(dc, &dc->current_state->res_ctx.pipe_ctx[i]);
+						DC_LOG_DC("Reset mpcc for pipe %d\n", dc->current_state->res_ctx.pipe_ctx[i].pipe_idx);
+						mpcc_disconnected = true;
+					}
+				}
+				found_stream = false;
+			}
+		}
+
+		if (mpcc_disconnected) {
+			for (i = 0; i < dc->res_pool->pipe_count; i++) {
+				struct pipe_ctx *pipe_ctx = &context->res_ctx.pipe_ctx[i];
+				struct pipe_ctx *old_pipe = &dc->current_state->res_ctx.pipe_ctx[i];
+				struct dc_plane_state *plane_state = pipe_ctx->plane_state;
+				struct hubp *hubp = pipe_ctx->plane_res.hubp;
+
+				if (!pipe_ctx || !plane_state || !pipe_ctx->stream)
+					continue;
+
+				// Only update scaler and viewport here if we lose a pipe split.
+				// This is to prevent half the screen from being black when we
+				// unlock after disconnecting MPCC.
+				if (!(old_pipe && !pipe_ctx->top_pipe &&
+					!pipe_ctx->bottom_pipe && old_pipe->bottom_pipe))
+					continue;
+
+				if (pipe_ctx->update_flags.raw || pipe_ctx->plane_state->update_flags.raw || pipe_ctx->stream->update_flags.raw) {
+					if (pipe_ctx->update_flags.bits.scaler ||
+						plane_state->update_flags.bits.scaling_change ||
+						plane_state->update_flags.bits.position_change ||
+						plane_state->update_flags.bits.per_pixel_alpha_change ||
+						pipe_ctx->stream->update_flags.bits.scaling) {
+
+						pipe_ctx->plane_res.scl_data.lb_params.alpha_en = pipe_ctx->plane_state->per_pixel_alpha;
+						ASSERT(pipe_ctx->plane_res.scl_data.lb_params.depth == LB_PIXEL_DEPTH_30BPP);
+						/* scaler configuration */
+						pipe_ctx->plane_res.dpp->funcs->dpp_set_scaler(
+						pipe_ctx->plane_res.dpp, &pipe_ctx->plane_res.scl_data);
+					}
+
+					if (pipe_ctx->update_flags.bits.viewport ||
+						(context == dc->current_state && plane_state->update_flags.bits.position_change) ||
+						(context == dc->current_state && plane_state->update_flags.bits.scaling_change) ||
+						(context == dc->current_state && pipe_ctx->stream->update_flags.bits.scaling)) {
+
+						hubp->funcs->mem_program_viewport(
+							hubp,
+							&pipe_ctx->plane_res.scl_data.viewport,
+							&pipe_ctx->plane_res.scl_data.viewport_c);
+					}
+				}
+			}
+		}
+	return mpcc_disconnected;
+}
+
+void dcn10_wait_for_pending_cleared(struct dc *dc,
+		struct dc_state *context)
+{
+		struct pipe_ctx *pipe_ctx;
+		struct timing_generator *tg;
+		int i;
+
+		for (i = 0; i < dc->res_pool->pipe_count; i++) {
+			pipe_ctx = &context->res_ctx.pipe_ctx[i];
+			tg = pipe_ctx->stream_res.tg;
+
+			/*
+			 * Only wait for top pipe's tg penindg bit
+			 * Also skip if pipe is disabled.
+			 */
+			if (pipe_ctx->top_pipe ||
+			    !pipe_ctx->stream || !pipe_ctx->plane_state ||
+			    !tg->funcs->is_tg_enabled(tg))
+				continue;
+
+			/*
+			 * Wait for VBLANK then VACTIVE to ensure we get VUPDATE.
+			 * For some reason waiting for OTG_UPDATE_PENDING cleared
+			 * seems to not trigger the update right away, and if we
+			 * lock again before VUPDATE then we don't get a separated
+			 * operation.
+			 */
+			pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VBLANK);
+			pipe_ctx->stream_res.tg->funcs->wait_for_state(pipe_ctx->stream_res.tg, CRTC_STATE_VACTIVE);
+		}
+}
+
 void dcn10_apply_ctx_for_surface(
 		struct dc *dc,
 		const struct dc_stream_state *stream,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
index 6d891166da8a..e5691e499023 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.h
@@ -194,6 +194,12 @@ void dcn10_get_surface_visual_confirm_color(
 void dcn10_get_hdr_visual_confirm_color(
 		struct pipe_ctx *pipe_ctx,
 		struct tg_color *color);
+bool dcn10_disconnect_pipes(
+		struct dc *dc,
+		struct dc_state *context);
+
+void dcn10_wait_for_pending_cleared(struct dc *dc,
+		struct dc_state *context);
 void dcn10_set_hdr_multiplier(struct pipe_ctx *pipe_ctx);
 void dcn10_verify_allow_pstate_change_high(struct dc *dc);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_init.c b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_init.c
index 5c98b71c1d47..a1d1559bb5d7 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_init.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn10/dcn10_init.c
@@ -34,6 +34,8 @@ static const struct hw_sequencer_funcs dcn10_funcs = {
 	.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
 	.apply_ctx_for_surface = dcn10_apply_ctx_for_surface,
 	.post_unlock_program_front_end = dcn10_post_unlock_program_front_end,
+	.disconnect_pipes = dcn10_disconnect_pipes,
+	.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
 	.update_plane_addr = dcn10_update_plane_addr,
 	.update_dchub = dcn10_update_dchub,
 	.update_pending_status = dcn10_update_pending_status,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c
index 3dde6f26de47..966e1790b9bf 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c
@@ -34,6 +34,8 @@ static const struct hw_sequencer_funcs dcn20_funcs = {
 	.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
 	.apply_ctx_for_surface = NULL,
 	.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+	.disconnect_pipes = dcn10_disconnect_pipes,
+	.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
 	.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
 	.update_plane_addr = dcn20_update_plane_addr,
 	.update_dchub = dcn10_update_dchub,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c
index b187f71afa65..2ba880c3943c 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c
@@ -35,6 +35,8 @@ static const struct hw_sequencer_funcs dcn21_funcs = {
 	.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
 	.apply_ctx_for_surface = NULL,
 	.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+	.disconnect_pipes = dcn10_disconnect_pipes,
+	.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
 	.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
 	.update_plane_addr = dcn20_update_plane_addr,
 	.update_dchub = dcn10_update_dchub,
diff --git a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_init.c b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_init.c
index 9afee7160490..19daa456e3bf 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_init.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn30/dcn30_init.c
@@ -35,6 +35,8 @@ static const struct hw_sequencer_funcs dcn30_funcs = {
 	.apply_ctx_to_hw = dce110_apply_ctx_to_hw,
 	.apply_ctx_for_surface = NULL,
 	.program_front_end_for_ctx = dcn20_program_front_end_for_ctx,
+	.disconnect_pipes = dcn10_disconnect_pipes,
+	.wait_for_pending_cleared = dcn10_wait_for_pending_cleared,
 	.post_unlock_program_front_end = dcn20_post_unlock_program_front_end,
 	.update_plane_addr = dcn20_update_plane_addr,
 	.update_dchub = dcn10_update_dchub,
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
index 3c986717dcd5..64c1be818b0e 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -67,6 +67,10 @@ struct hw_sequencer_funcs {
 			int num_planes, struct dc_state *context);
 	void (*program_front_end_for_ctx)(struct dc *dc,
 			struct dc_state *context);
+	bool (*disconnect_pipes)(struct dc *dc,
+			struct dc_state *context);
+	void (*wait_for_pending_cleared)(struct dc *dc,
+			struct dc_state *context);
 	void (*post_unlock_program_front_end)(struct dc *dc,
 			struct dc_state *context);
 	void (*update_plane_addr)(const struct dc *dc,
-- 
2.17.1

_______________________________________________
amd-gfx mailing list
amd-gfx@xxxxxxxxxxxxxxxxxxxxx
https://lists.freedesktop.org/mailman/listinfo/amd-gfx



[Index of Archives]     [Linux USB Devel]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]

  Powered by Linux