[PATCH 00/15] DC Patches July 29, 2020

[PATCH 00/15] DC Patches July 29, 2020

[Eryk Brol]

This DC patchset brings improvements in multiple areas. In summary, we highlight:

* Fixes in DP Compliance, ABM, and others;
* Enhancements in DSC;
* Improvements in FSFT, DCN, and others.

Alvin Lee (1):
  drm/amd/display: Don't compare dppclk before updating DTO

Anthony Koo (1):
  drm/amd/display: [FW Promotion] Release 0.0.26

Aric Cyr (2):
  drm/amd/display: 3.2.96
  drm/amd/display: Fix DP Compliance tests 4.3.2.1 and 4.3.2.2

Dmytro Laktyushkin (1):
  drm/amd/display: Clean up global sync param retrieval

Eryk Brol (4):
  drm/amd/display: Rename bytes_pp to the correct bits_pp
  drm/amd/display: Fix naming of DSC Debugfs entry
  drm/amd/display: DSC Clock enable debugfs write entry
  drm/amd/display: DSC Slice width debugfs write entry

Jun Lei (1):
  drm/amd/display: Disable idle optimizations before programming DCN

Martin Tsai (1):
  drm/amd/display: Check lane status again after link training done

Reza Amini (1):
  drm/amd/display: Allow asic specific FSFT timing optimization

Stylon Wang (1):
  drm/amd/display: Fix dmesg warning from setting abm level

Wyatt Wood (1):
  drm/amd/display: Use hw lock mgr

hersen wu (1):
  drm/amd/display: dchubbub p-state warning during surface planes switch

 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |  32 ++-
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h |   6 +
 .../amd/display/amdgpu_dm/amdgpu_dm_debugfs.c | 193 +++++++++++++++++-
 .../display/amdgpu_dm/amdgpu_dm_mst_types.c   |  17 +-
 .../display/dc/clk_mgr/dcn10/rv1_clk_mgr.c    |  69 ++++++-
 .../display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c  |   7 +-
 drivers/gpu/drm/amd/display/dc/core/dc.c      |  18 +-
 .../gpu/drm/amd/display/dc/core/dc_link_dp.c  |  49 +++++
 .../gpu/drm/amd/display/dc/core/dc_stream.c   |  18 +-
 drivers/gpu/drm/amd/display/dc/dc.h           |   2 +-
 drivers/gpu/drm/amd/display/dc/dc_stream.h    |   4 +-
 drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c |   2 +-
 .../amd/display/dc/dcn10/dcn10_hw_sequencer.c |   4 +-
 .../gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c  |   2 +-
 .../drm/amd/display/dc/dcn20/dcn20_hwseq.c    |  27 +++
 .../drm/amd/display/dc/dcn20/dcn20_hwseq.h    |   5 +
 .../gpu/drm/amd/display/dc/dcn20/dcn20_init.c |   3 +
 .../drm/amd/display/dc/dcn20/dcn20_resource.c |  49 +----
 .../gpu/drm/amd/display/dc/dcn21/dcn21_init.c |   3 +
 .../drm/amd/display/dc/dml/display_mode_vba.c |  14 +-
 .../drm/amd/display/dc/dml/display_mode_vba.h |   7 +-
 drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h   |   2 +-
 .../gpu/drm/amd/display/dc/inc/hw_sequencer.h |   5 +
 .../gpu/drm/amd/display/dmub/inc/dmub_cmd.h   |   4 +-
 .../amd/display/include/link_service_types.h  |   2 +
 .../amd/display/modules/freesync/freesync.c   |   5 +-
 26 files changed, 451 insertions(+), 98 deletions(-)

-- 
2.25.1

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

[PATCH 01/15] drm/amd/display: Clean up global sync param retrieval

[Eryk Brol]

From: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>

[Why]
This change replaces older looping code in favor of these functions.

[How]
There are built in functions for extracting global sync params
during mode validation now.

Signed-off-by: Dmytro Laktyushkin <Dmytro.Laktyushkin@amd.com>
Reviewed-by: Eric Bernstein <Eric.Bernstein@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../drm/amd/display/dc/dcn20/dcn20_resource.c | 49 ++-----------------
 .../drm/amd/display/dc/dml/display_mode_vba.c | 14 +-----
 .../drm/amd/display/dc/dml/display_mode_vba.h |  7 +--
 3 files changed, 7 insertions(+), 63 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c
index 968a89bbcf24..2a5e7175926a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_resource.c
@@ -3069,8 +3069,7 @@ void dcn20_calculate_dlg_params(
 		int pipe_cnt,
 		int vlevel)
 {
-	int i, j, pipe_idx, pipe_idx_unsplit;
-	bool visited[MAX_PIPES] = { 0 };
+	int i, pipe_idx;
 
 	/* Writeback MCIF_WB arbitration parameters */
 	dc->res_pool->funcs->set_mcif_arb_params(dc, context, pipes, pipe_cnt);
@@ -3089,55 +3088,17 @@ void dcn20_calculate_dlg_params(
 	if (context->bw_ctx.bw.dcn.clk.dispclk_khz < dc->debug.min_disp_clk_khz)
 		context->bw_ctx.bw.dcn.clk.dispclk_khz = dc->debug.min_disp_clk_khz;
 
-	/*
-	 * An artifact of dml pipe split/odm is that pipes get merged back together for
-	 * calculation. Therefore we need to only extract for first pipe in ascending index order
-	 * and copy into the other split half.
-	 */
-	for (i = 0, pipe_idx = 0, pipe_idx_unsplit = 0; i < dc->res_pool->pipe_count; i++) {
-		if (!context->res_ctx.pipe_ctx[i].stream)
-			continue;
-
-		if (!visited[pipe_idx]) {
-			display_pipe_source_params_st *src = &pipes[pipe_idx].pipe.src;
-			display_pipe_dest_params_st *dst = &pipes[pipe_idx].pipe.dest;
-
-			dst->vstartup_start = context->bw_ctx.dml.vba.VStartup[pipe_idx_unsplit];
-			dst->vupdate_offset = context->bw_ctx.dml.vba.VUpdateOffsetPix[pipe_idx_unsplit];
-			dst->vupdate_width = context->bw_ctx.dml.vba.VUpdateWidthPix[pipe_idx_unsplit];
-			dst->vready_offset = context->bw_ctx.dml.vba.VReadyOffsetPix[pipe_idx_unsplit];
-			/*
-			 * j iterates inside pipes array, unlike i which iterates inside
-			 * pipe_ctx array
-			 */
-			if (src->is_hsplit)
-				for (j = pipe_idx + 1; j < pipe_cnt; j++) {
-					display_pipe_source_params_st *src_j = &pipes[j].pipe.src;
-					display_pipe_dest_params_st *dst_j = &pipes[j].pipe.dest;
-
-					if (src_j->is_hsplit && !visited[j]
-							&& src->hsplit_grp == src_j->hsplit_grp) {
-						dst_j->vstartup_start = context->bw_ctx.dml.vba.VStartup[pipe_idx_unsplit];
-						dst_j->vupdate_offset = context->bw_ctx.dml.vba.VUpdateOffsetPix[pipe_idx_unsplit];
-						dst_j->vupdate_width = context->bw_ctx.dml.vba.VUpdateWidthPix[pipe_idx_unsplit];
-						dst_j->vready_offset = context->bw_ctx.dml.vba.VReadyOffsetPix[pipe_idx_unsplit];
-						visited[j] = true;
-					}
-				}
-			visited[pipe_idx] = true;
-			pipe_idx_unsplit++;
-		}
-		pipe_idx++;
-	}
-
 	for (i = 0, pipe_idx = 0; i < dc->res_pool->pipe_count; i++) {
 		if (!context->res_ctx.pipe_ctx[i].stream)
 			continue;
+		pipes[pipe_idx].pipe.dest.vstartup_start = get_vstartup(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
+		pipes[pipe_idx].pipe.dest.vupdate_offset = get_vupdate_offset(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
+		pipes[pipe_idx].pipe.dest.vupdate_width = get_vupdate_width(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
+		pipes[pipe_idx].pipe.dest.vready_offset = get_vready_offset(&context->bw_ctx.dml, pipes, pipe_cnt, pipe_idx);
 		if (context->bw_ctx.bw.dcn.clk.dppclk_khz < pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000)
 			context->bw_ctx.bw.dcn.clk.dppclk_khz = pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000;
 		context->res_ctx.pipe_ctx[i].plane_res.bw.dppclk_khz =
 						pipes[pipe_idx].clks_cfg.dppclk_mhz * 1000;
-		ASSERT(visited[pipe_idx]);
 		context->res_ctx.pipe_ctx[i].pipe_dlg_param = pipes[pipe_idx].pipe.dest;
 		pipe_idx++;
 	}
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c
index 7916a7ea9336..b0064087b9bb 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.c
@@ -154,23 +154,11 @@ dml_get_pipe_attr_func(refcyc_per_meta_chunk_vblank_c_in_us, mode_lib->vba.TimeP
 dml_get_pipe_attr_func(refcyc_per_meta_chunk_flip_l_in_us, mode_lib->vba.TimePerMetaChunkFlip);
 dml_get_pipe_attr_func(refcyc_per_meta_chunk_flip_c_in_us, mode_lib->vba.TimePerChromaMetaChunkFlip);
 
+dml_get_pipe_attr_func(vstartup, mode_lib->vba.VStartup);
 dml_get_pipe_attr_func(vupdate_offset, mode_lib->vba.VUpdateOffsetPix);
 dml_get_pipe_attr_func(vupdate_width, mode_lib->vba.VUpdateWidthPix);
 dml_get_pipe_attr_func(vready_offset, mode_lib->vba.VReadyOffsetPix);
 
-unsigned int get_vstartup_calculated(
-		struct display_mode_lib *mode_lib,
-		const display_e2e_pipe_params_st *pipes,
-		unsigned int num_pipes,
-		unsigned int which_pipe)
-{
-	unsigned int which_plane;
-
-	recalculate_params(mode_lib, pipes, num_pipes);
-	which_plane = mode_lib->vba.pipe_plane[which_pipe];
-	return mode_lib->vba.VStartup[which_plane];
-}
-
 double get_total_immediate_flip_bytes(
 		struct display_mode_lib *mode_lib,
 		const display_e2e_pipe_params_st *pipes,
diff --git a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h
index 756d8eb1221c..21e5111ea7a0 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h
+++ b/drivers/gpu/drm/amd/display/dc/dml/display_mode_vba.h
@@ -98,16 +98,11 @@ dml_get_pipe_attr_decl(refcyc_per_meta_chunk_vblank_c_in_us);
 dml_get_pipe_attr_decl(refcyc_per_meta_chunk_flip_l_in_us);
 dml_get_pipe_attr_decl(refcyc_per_meta_chunk_flip_c_in_us);
 
+dml_get_pipe_attr_decl(vstartup);
 dml_get_pipe_attr_decl(vupdate_offset);
 dml_get_pipe_attr_decl(vupdate_width);
 dml_get_pipe_attr_decl(vready_offset);
 
-unsigned int get_vstartup_calculated(
-		struct display_mode_lib *mode_lib,
-		const display_e2e_pipe_params_st *pipes,
-		unsigned int num_pipes,
-		unsigned int which_pipe);
-
 double get_total_immediate_flip_bytes(
 		struct display_mode_lib *mode_lib,
 		const display_e2e_pipe_params_st *pipes,
-- 
2.25.1

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

[PATCH 02/15] drm/amd/display: Rename bytes_pp to the correct bits_pp

[Eryk Brol]

[Why]
Struct dcn_dsc_state is used for reading current state
and parameters of DSC on a pipe, the target rate parameter
uses bytes per pixel even though its reading BITS_PER_PIXEL
register.

[How]
Changing it to Bits Per Pixel for consistency.

Signed-off-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Mikita Lipski <Mikita.Lipski@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c | 2 +-
 drivers/gpu/drm/amd/display/dc/dcn10/dcn10_hw_sequencer.c | 4 +++-
 drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c          | 2 +-
 drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h               | 2 +-
 4 files changed, 6 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
index 8bbf54ffe01b..3dd7da16cc18 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
@@ -1204,7 +1204,7 @@ static ssize_t dp_dsc_bytes_per_pixel_read(struct file *f, char __user *buf,
 
 	snprintf(rd_buf_ptr, str_len,
 		"%d\n",
-		dsc_state.dsc_bytes_per_pixel);
+		dsc_state.dsc_bits_per_pixel);
 	rd_buf_ptr += str_len;
 
 	while (size) {
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 da0897fe3b54..a643927e272b 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
@@ -390,6 +390,8 @@ void dcn10_log_hw_state(struct dc *dc,
 	}
 	DTN_INFO("\n");
 
+	// dcn_dsc_state struct field bytes_per_pixel was renamed to bits_per_pixel
+	// TODO: Update golden log header to reflect this name change
 	DTN_INFO("DSC: CLOCK_EN  SLICE_WIDTH  Bytes_pp\n");
 	for (i = 0; i < pool->res_cap->num_dsc; i++) {
 		struct display_stream_compressor *dsc = pool->dscs[i];
@@ -400,7 +402,7 @@ void dcn10_log_hw_state(struct dc *dc,
 		dsc->inst,
 			s.dsc_clock_en,
 			s.dsc_slice_width,
-			s.dsc_bytes_per_pixel);
+			s.dsc_bits_per_pixel);
 		DTN_INFO("\n");
 	}
 	DTN_INFO("\n");
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c
index ba50214d6c32..79b640e202eb 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_dsc.c
@@ -156,7 +156,7 @@ static void dsc2_read_state(struct display_stream_compressor *dsc, struct dcn_ds
 
 	REG_GET(DSC_TOP_CONTROL, DSC_CLOCK_EN, &s->dsc_clock_en);
 	REG_GET(DSCC_PPS_CONFIG3, SLICE_WIDTH, &s->dsc_slice_width);
-	REG_GET(DSCC_PPS_CONFIG1, BITS_PER_PIXEL, &s->dsc_bytes_per_pixel);
+	REG_GET(DSCC_PPS_CONFIG1, BITS_PER_PIXEL, &s->dsc_bits_per_pixel);
 	REG_GET(DSCC_PPS_CONFIG3, SLICE_HEIGHT, &s->dsc_slice_height);
 	REG_GET(DSCC_PPS_CONFIG1, CHUNK_SIZE, &s->dsc_chunk_size);
 	REG_GET(DSCC_PPS_CONFIG2, PIC_WIDTH, &s->dsc_pic_width);
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h
index 5915994f9eb8..f520e13aee4c 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dsc.h
@@ -55,7 +55,7 @@ struct dsc_optc_config {
 struct dcn_dsc_state {
 	uint32_t dsc_clock_en;
 	uint32_t dsc_slice_width;
-	uint32_t dsc_bytes_per_pixel;
+	uint32_t dsc_bits_per_pixel;
 	uint32_t dsc_slice_height;
 	uint32_t dsc_pic_width;
 	uint32_t dsc_pic_height;
-- 
2.25.1

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

[PATCH 03/15] drm/amd/display: Fix naming of DSC Debugfs entry

[Eryk Brol]

[why]
Fix naming and return bits rather than bytes per pixel for
naming consistency. Because registers return Bytes per pixel,
but DSC Config structure is expecting bits per pixel as input.
So when returning the value convert from bytes into bits.

Signed-off-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Mikita Lipski <Mikita.Lipski@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c    | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
index 3dd7da16cc18..65e9ee64c70c 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
@@ -35,6 +35,7 @@
 #include "dmub/dmub_srv.h"
 #include "resource.h"
 #include "dsc.h"
+#include "dc_link_dp.h"
 
 struct dmub_debugfs_trace_header {
 	uint32_t entry_count;
@@ -1168,7 +1169,7 @@ static ssize_t dp_dsc_slice_height_read(struct file *f, char __user *buf,
 	return result;
 }
 
-static ssize_t dp_dsc_bytes_per_pixel_read(struct file *f, char __user *buf,
+static ssize_t dp_dsc_bits_per_pixel_read(struct file *f, char __user *buf,
 				    size_t size, loff_t *pos)
 {
 	char *rd_buf = NULL;
@@ -1479,9 +1480,9 @@ static const struct file_operations dp_dsc_slice_height_debugfs_fops = {
 	.llseek = default_llseek
 };
 
-static const struct file_operations dp_dsc_bytes_per_pixel_debugfs_fops = {
+static const struct file_operations dp_dsc_bits_per_pixel_debugfs_fops = {
 	.owner = THIS_MODULE,
-	.read = dp_dsc_bytes_per_pixel_read,
+	.read = dp_dsc_bits_per_pixel_read,
 	.llseek = default_llseek
 };
 
@@ -1572,7 +1573,7 @@ static const struct {
 		{"dsc_clock_en", &dp_dsc_clock_en_debugfs_fops},
 		{"dsc_slice_width", &dp_dsc_slice_width_debugfs_fops},
 		{"dsc_slice_height", &dp_dsc_slice_height_debugfs_fops},
-		{"dsc_bytes_per_pixel", &dp_dsc_bytes_per_pixel_debugfs_fops},
+		{"dsc_bits_per_pixel", &dp_dsc_bits_per_pixel_debugfs_fops},
 		{"dsc_pic_width", &dp_dsc_pic_width_debugfs_fops},
 		{"dsc_pic_height", &dp_dsc_pic_height_debugfs_fops},
 		{"dsc_chunk_size", &dp_dsc_chunk_size_debugfs_fops},
-- 
2.25.1

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

[PATCH 04/15] drm/amd/display: Don't compare dppclk before updating DTO

[Eryk Brol]

From: Alvin Lee <alvin.lee2@amd.com>

[Why]
In dcn3_update_clocks there are situations where dppclk is not
lowered (i.e. stays the same), but DTO still needs to be increased
before we program pipe frontend (i.e. in prepare_bandwidth). If we
don't program the new DTO value before we program the pipe,
we will underflow as soon as the pipe lock is released until the
next call to dcn3_update_clocks where the DTO is updated.

[How]
Remove dppclk check before programming new DTO value.

Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c   | 7 ++++---
 1 file changed, 4 insertions(+), 3 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c
index d8af56a58196..b0e9b0509568 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn30/dcn30_clk_mgr.c
@@ -323,9 +323,10 @@ static void dcn3_update_clocks(struct clk_mgr *clk_mgr_base,
 			/* if clock is being raised, increase refclk before lowering DTO */
 			if (update_dppclk || update_dispclk)
 				dcn20_update_clocks_update_dentist(clk_mgr);
-			/* always update dtos unless clock is lowered and not safe to lower */
-			if (new_clocks->dppclk_khz >= dc->current_state->bw_ctx.bw.dcn.clk.dppclk_khz)
-				dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower);
+			/* There is a check inside dcn20_update_clocks_update_dpp_dto which ensures
+			 * that we do not lower dto when it is not safe to lower. We do not need to
+			 * compare the current and new dppclk before calling this function.*/
+			dcn20_update_clocks_update_dpp_dto(clk_mgr, context, safe_to_lower);
 		}
 	}
 
-- 
2.25.1

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

[PATCH 05/15] drm/amd/display: Check lane status again after link training done

[Eryk Brol]

From: Martin Tsai <martin.tsai@amd.com>

[Why]
Some monitors could suffer symbol unlock but cannot send HPD IRQ to
notic source device to handle link loss. This makes monitor stuck in
abnormal status and causes black screen.

[How]
According to the suggestion from scalar vendor, to check lane status
again after link training done. That can improve the comaptibility
from current production monitors.

Signed-off-by: Martin Tsai <martin.tsai@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../gpu/drm/amd/display/dc/core/dc_link_dp.c  | 50 +++++++++++++++++++
 .../amd/display/include/link_service_types.h  |  2 +
 2 files changed, 52 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index 277b29590fb9..0447492d30ca 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -1133,6 +1133,45 @@ static inline enum link_training_result perform_link_training_int(
 	return status;
 }
 
+static enum link_training_result check_link_loss_status(
+	struct dc_link *link,
+	const struct link_training_settings *link_training_setting)
+{
+	enum link_training_result status = LINK_TRAINING_SUCCESS;
+	unsigned int lane01_status_address = DP_LANE0_1_STATUS;
+	union lane_status lane_status;
+	uint8_t dpcd_buf[4] = {0};
+	uint32_t lane;
+
+	core_link_read_dpcd(
+		link,
+		lane01_status_address,
+		(uint8_t *)(dpcd_buf),
+		sizeof(dpcd_buf));
+
+	/*parse lane status*/
+	for (lane = 0; lane < link->cur_link_settings.lane_count; lane++) {
+		/*
+		 * check lanes status
+		 */
+		lane_status.raw = get_nibble_at_index(&dpcd_buf[0], lane);
+
+		if (!lane_status.bits.CHANNEL_EQ_DONE_0 ||
+			!lane_status.bits.CR_DONE_0 ||
+			!lane_status.bits.SYMBOL_LOCKED_0) {
+			/* if one of the channel equalization, clock
+			 * recovery or symbol lock is dropped
+			 * consider it as (link has been
+			 * dropped) dp sink status has changed
+			 */
+			status = LINK_TRAINING_LINK_LOSS;
+			break;
+		}
+	}
+
+	return status;
+}
+
 static void initialize_training_settings(
 	 struct dc_link *link,
 	const struct dc_link_settings *link_setting,
@@ -1372,6 +1411,9 @@ static void print_status_message(
 	case LINK_TRAINING_LQA_FAIL:
 		lt_result = "LQA failed";
 		break;
+	case LINK_TRAINING_LINK_LOSS:
+		lt_result = "Link loss";
+		break;
 	default:
 		break;
 	}
@@ -1531,6 +1573,14 @@ enum link_training_result dc_link_dp_perform_link_training(
 				status);
 	}
 
+	/* delay 5ms after Main Link output idle pattern and then check
+	 * DPCD 0202h.
+	 */
+	if (link->connector_signal != SIGNAL_TYPE_EDP && status == LINK_TRAINING_SUCCESS) {
+		msleep(5);
+		status = check_link_loss_status(link, &lt_settings);
+	}
+
 	/* 6. print status message*/
 	print_status_message(link, &lt_settings, status);
 
diff --git a/drivers/gpu/drm/amd/display/include/link_service_types.h b/drivers/gpu/drm/amd/display/include/link_service_types.h
index 4869d4562e4d..550f46e9b95f 100644
--- a/drivers/gpu/drm/amd/display/include/link_service_types.h
+++ b/drivers/gpu/drm/amd/display/include/link_service_types.h
@@ -66,6 +66,8 @@ enum link_training_result {
 	/* other failure during EQ step */
 	LINK_TRAINING_EQ_FAIL_EQ,
 	LINK_TRAINING_LQA_FAIL,
+	/* one of the CR,EQ or symbol lock is dropped */
+	LINK_TRAINING_LINK_LOSS,
 };
 
 struct link_training_settings {
-- 
2.25.1

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

[PATCH 06/15] drm/amd/display: Fix dmesg warning from setting abm level

[Eryk Brol]

From: Stylon Wang <stylon.wang@amd.com>

[Why]
Setting abm level does not correctly update CRTC state. As a result
no surface update is added to dc stream state and triggers warning.

[How]
Correctly update CRTC state when setting abm level property.

CC: Stable <stable@vger.kernel.org>
Signed-off-by: Stylon Wang <stylon.wang@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 23 +++++++++++++++++++
 1 file changed, 23 insertions(+)

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 2b5e70f68e57..b1bc03594e5c 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -8625,6 +8625,29 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev,
 	if (ret)
 		goto fail;
 
+	/* Check connector changes */
+	for_each_oldnew_connector_in_state(state, connector, old_con_state, new_con_state, i) {
+		struct dm_connector_state *dm_old_con_state = to_dm_connector_state(old_con_state);
+		struct dm_connector_state *dm_new_con_state = to_dm_connector_state(new_con_state);
+
+		/* Skip connectors that are disabled or part of modeset already. */
+		if (!old_con_state->crtc && !new_con_state->crtc)
+			continue;
+
+		if (!new_con_state->crtc)
+			continue;
+
+		new_crtc_state = drm_atomic_get_crtc_state(state, new_con_state->crtc);
+		if (IS_ERR(new_crtc_state)) {
+			ret = PTR_ERR(new_crtc_state);
+			goto fail;
+		}
+
+		if (dm_old_con_state->abm_level !=
+		    dm_new_con_state->abm_level)
+			new_crtc_state->connectors_changed = true;
+	}
+
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	if (adev->asic_type >= CHIP_NAVI10) {
 		for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
-- 
2.25.1

[PATCH 06/15] drm/amd/display: Fix dmesg warning from setting abm level

[Eryk Brol]

From: Stylon Wang <stylon.wang@amd.com>

[Why]
Setting abm level does not correctly update CRTC state. As a result
no surface update is added to dc stream state and triggers warning.

[How]
Correctly update CRTC state when setting abm level property.

CC: Stable <stable@vger.kernel.org>
Signed-off-by: Stylon Wang <stylon.wang@amd.com>
Reviewed-by: Nicholas Kazlauskas <Nicholas.Kazlauskas@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c | 23 +++++++++++++++++++
 1 file changed, 23 insertions(+)

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 2b5e70f68e57..b1bc03594e5c 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -8625,6 +8625,29 @@ static int amdgpu_dm_atomic_check(struct drm_device *dev,
 	if (ret)
 		goto fail;
 
+	/* Check connector changes */
+	for_each_oldnew_connector_in_state(state, connector, old_con_state, new_con_state, i) {
+		struct dm_connector_state *dm_old_con_state = to_dm_connector_state(old_con_state);
+		struct dm_connector_state *dm_new_con_state = to_dm_connector_state(new_con_state);
+
+		/* Skip connectors that are disabled or part of modeset already. */
+		if (!old_con_state->crtc && !new_con_state->crtc)
+			continue;
+
+		if (!new_con_state->crtc)
+			continue;
+
+		new_crtc_state = drm_atomic_get_crtc_state(state, new_con_state->crtc);
+		if (IS_ERR(new_crtc_state)) {
+			ret = PTR_ERR(new_crtc_state);
+			goto fail;
+		}
+
+		if (dm_old_con_state->abm_level !=
+		    dm_new_con_state->abm_level)
+			new_crtc_state->connectors_changed = true;
+	}
+
 #if defined(CONFIG_DRM_AMD_DC_DCN)
 	if (adev->asic_type >= CHIP_NAVI10) {
 		for_each_oldnew_crtc_in_state(state, crtc, old_crtc_state, new_crtc_state, i) {
-- 
2.25.1

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

[PATCH 07/15] drm/amd/display: Disable idle optimizations before programming DCN

[Eryk Brol]

From: Jun Lei <jun.lei@amd.com>

[Why]
Programming DCN is explicitly forbidden during idle optimzations allowed
state. Existing implemenation relies on OS/DM, which is not robust. Instead
DC should sequence this.

Note that DC will not re-enter idle optimized state on its own, it is only
responsible for catching out of sequence calls. It is still DM
responsibility to sequence appropriate for optimized power, but this change
removes the requirement for DM to cover the .1% case.

[How]
 - elevate updates during idle optimized state to full updates
 - disable idle power optimizations prior to programming

Signed-off-by: Jun Lei <jun.lei@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c | 18 ++++++++++++++++--
 1 file changed, 16 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index ef0b5941bc50..92eb1ca1634f 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -1250,6 +1250,9 @@ static enum dc_status dc_commit_state_no_check(struct dc *dc, struct dc_state *c
 	int i, k, l;
 	struct dc_stream_state *dc_streams[MAX_STREAMS] = {0};
 
+#if defined(CONFIG_DRM_AMD_DC_DCN3_0)
+	dc_allow_idle_optimizations(dc, false);
+#endif
 
 	for (i = 0; i < context->stream_count; i++)
 		dc_streams[i] =  context->streams[i];
@@ -1838,6 +1841,11 @@ static enum surface_update_type check_update_surfaces_for_stream(
 	int i;
 	enum surface_update_type overall_type = UPDATE_TYPE_FAST;
 
+#if defined(CONFIG_DRM_AMD_DC_DCN3_0)
+	if (dc->idle_optimizations_allowed)
+		overall_type = UPDATE_TYPE_FULL;
+
+#endif
 	if (stream_status == NULL || stream_status->plane_count != surface_count)
 		overall_type = UPDATE_TYPE_FULL;
 
@@ -2306,8 +2314,14 @@ static void commit_planes_for_stream(struct dc *dc,
 		}
 	}
 
-	if (update_type == UPDATE_TYPE_FULL && dc->optimize_seamless_boot_streams == 0) {
-		dc->hwss.prepare_bandwidth(dc, context);
+	if (update_type == UPDATE_TYPE_FULL) {
+#if defined(CONFIG_DRM_AMD_DC_DCN3_0)
+		dc_allow_idle_optimizations(dc, false);
+
+#endif
+		if (dc->optimize_seamless_boot_streams == 0)
+			dc->hwss.prepare_bandwidth(dc, context);
+
 		context_clock_trace(dc, context);
 	}
 
-- 
2.25.1

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

[PATCH 08/15] drm/amd/display: Allow asic specific FSFT timing optimization

[Eryk Brol]

From: Reza Amini <Reza.Amini@amd.com>

[Why]
Each asic can optimize best based on its capabilities

[How]
Optimizing timing for a new pixel clock

Signed-off-by: Reza Amini <Reza.Amini@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../gpu/drm/amd/display/dc/core/dc_stream.c   | 18 ++++++-------
 drivers/gpu/drm/amd/display/dc/dc_stream.h    |  4 +--
 .../drm/amd/display/dc/dcn20/dcn20_hwseq.c    | 27 +++++++++++++++++++
 .../drm/amd/display/dc/dcn20/dcn20_hwseq.h    |  5 ++++
 .../gpu/drm/amd/display/dc/dcn20/dcn20_init.c |  3 +++
 .../gpu/drm/amd/display/dc/dcn21/dcn21_init.c |  3 +++
 .../gpu/drm/amd/display/dc/inc/hw_sequencer.h |  5 ++++
 .../amd/display/modules/freesync/freesync.c   |  5 +++-
 8 files changed, 57 insertions(+), 13 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
index 10d69ada88e3..0257a900fe2b 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_stream.c
@@ -246,20 +246,18 @@ struct dc_stream_status *dc_stream_get_status(
 
 #ifndef TRIM_FSFT
 /**
- * dc_optimize_timing() - dc to optimize timing
+ * dc_optimize_timing_for_fsft() - dc to optimize timing
  */
-bool dc_optimize_timing(
-	struct dc_crtc_timing *timing,
+bool dc_optimize_timing_for_fsft(
+	struct dc_stream_state *pStream,
 	unsigned int max_input_rate_in_khz)
 {
-	//optimization is expected to assing a value to these:
-	//timing->pix_clk_100hz
-	//timing->v_front_porch
-	//timing->v_total
-	//timing->fast_transport_output_rate_100hz;
-	timing->fast_transport_output_rate_100hz = timing->pix_clk_100hz;
+	struct dc  *dc;
 
-	return true;
+	dc = pStream->ctx->dc;
+
+	return (dc->hwss.optimize_timing_for_fsft &&
+		dc->hwss.optimize_timing_for_fsft(dc, &pStream->timing, max_input_rate_in_khz));
 }
 #endif
 
diff --git a/drivers/gpu/drm/amd/display/dc/dc_stream.h b/drivers/gpu/drm/amd/display/dc/dc_stream.h
index e4e85a159462..633442bc7ef2 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_stream.h
+++ b/drivers/gpu/drm/amd/display/dc/dc_stream.h
@@ -424,8 +424,8 @@ struct dc_stream_status *dc_stream_get_status(
 	struct dc_stream_state *dc_stream);
 
 #ifndef TRIM_FSFT
-bool dc_optimize_timing(
-	struct dc_crtc_timing *timing,
+bool dc_optimize_timing_for_fsft(
+	struct dc_stream_state *pStream,
 	unsigned int max_input_rate_in_khz);
 #endif
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
index 7725a406c16e..66180b4332f1 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
@@ -2498,3 +2498,30 @@ void dcn20_fpga_init_hw(struct dc *dc)
 		tg->funcs->tg_init(tg);
 	}
 }
+#ifndef TRIM_FSFT
+bool dcn20_optimize_timing_for_fsft(struct dc *dc,
+		struct dc_crtc_timing *timing,
+		unsigned int max_input_rate_in_khz)
+{
+	unsigned int old_v_front_porch;
+	unsigned int old_v_total;
+	unsigned int max_input_rate_in_100hz;
+	unsigned long long new_v_total;
+
+	max_input_rate_in_100hz = max_input_rate_in_khz * 10;
+	if (max_input_rate_in_100hz < timing->pix_clk_100hz)
+		return false;
+
+	old_v_total = timing->v_total;
+	old_v_front_porch = timing->v_front_porch;
+
+	timing->fast_transport_output_rate_100hz = timing->pix_clk_100hz;
+	timing->pix_clk_100hz = max_input_rate_in_100hz;
+
+	new_v_total = div_u64((unsigned long long)old_v_total * max_input_rate_in_100hz, timing->pix_clk_100hz);
+
+	timing->v_total = new_v_total;
+	timing->v_front_porch = old_v_front_porch + (timing->v_total - old_v_total);
+	return true;
+}
+#endif
diff --git a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.h b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.h
index 63ce763f148e..83220e34c1a9 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.h
@@ -132,5 +132,10 @@ int dcn20_init_sys_ctx(struct dce_hwseq *hws,
 		struct dc *dc,
 		struct dc_phy_addr_space_config *pa_config);
 
+#ifndef TRIM_FSFT
+bool dcn20_optimize_timing_for_fsft(struct dc *dc,
+		struct dc_crtc_timing *timing,
+		unsigned int max_input_rate_in_khz);
+#endif
 #endif /* __DC_HWSS_DCN20_H__ */
 
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 2380392b916e..3dde6f26de47 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_init.c
@@ -88,6 +88,9 @@ static const struct hw_sequencer_funcs dcn20_funcs = {
 	.set_backlight_level = dce110_set_backlight_level,
 	.set_abm_immediate_disable = dce110_set_abm_immediate_disable,
 	.set_pipe = dce110_set_pipe,
+#ifndef TRIM_FSFT
+	.optimize_timing_for_fsft = dcn20_optimize_timing_for_fsft,
+#endif
 };
 
 static const struct hwseq_private_funcs dcn20_private_funcs = {
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 177d0dc8927a..b187f71afa65 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn21/dcn21_init.c
@@ -92,6 +92,9 @@ static const struct hw_sequencer_funcs dcn21_funcs = {
 	.set_backlight_level = dcn21_set_backlight_level,
 	.set_abm_immediate_disable = dcn21_set_abm_immediate_disable,
 	.set_pipe = dcn21_set_pipe,
+#ifndef TRIM_FSFT
+	.optimize_timing_for_fsft = dcn20_optimize_timing_for_fsft,
+#endif
 };
 
 static const struct hwseq_private_funcs dcn21_private_funcs = {
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 720ce5e458d8..3c986717dcd5 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -116,6 +116,11 @@ struct hw_sequencer_funcs {
 	void (*set_static_screen_control)(struct pipe_ctx **pipe_ctx,
 			int num_pipes,
 			const struct dc_static_screen_params *events);
+#ifndef TRIM_FSFT
+	bool (*optimize_timing_for_fsft)(struct dc *dc,
+			struct dc_crtc_timing *timing,
+			unsigned int max_input_rate_in_khz);
+#endif
 
 	/* Stream Related */
 	void (*enable_stream)(struct pipe_ctx *pipe_ctx);
diff --git a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
index 7a2500fbf3f2..81820f3d6b3b 100644
--- a/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
+++ b/drivers/gpu/drm/amd/display/modules/freesync/freesync.c
@@ -829,10 +829,13 @@ void mod_freesync_build_vrr_infopacket(struct mod_freesync *mod_freesync,
 	switch (packet_type) {
 	case PACKET_TYPE_FS_V3:
 #ifndef TRIM_FSFT
+		// always populate with pixel rate.
 		build_vrr_infopacket_v3(
 				stream->signal, vrr,
 				stream->timing.flags.FAST_TRANSPORT,
-				stream->timing.fast_transport_output_rate_100hz,
+				(stream->timing.flags.FAST_TRANSPORT) ?
+						stream->timing.fast_transport_output_rate_100hz :
+						stream->timing.pix_clk_100hz,
 				app_tf, infopacket);
 #else
 		build_vrr_infopacket_v3(stream->signal, vrr, app_tf, infopacket);
-- 
2.25.1

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

[PATCH 09/15] drm/amd/display: DSC Clock enable debugfs write entry

[Eryk Brol]

[Why]
Need a mechanism to force enable DSC on any connector

[How]
Debugfs entry overwrites newly added connector's dsc preffered
settings structure and sets dsc_clock_en flag on it.
During the attomic commit, depending if connector is SST or
MST, we will enable DSC manually by overwriting stream's DSC flag.

Signed-off-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Mikita Lipski <Mikita.Lipski@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |  5 +-
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h |  5 ++
 .../amd/display/amdgpu_dm/amdgpu_dm_debugfs.c | 89 +++++++++++++++++++
 .../display/amdgpu_dm/amdgpu_dm_mst_types.c   | 11 ++-
 4 files changed, 107 insertions(+), 3 deletions(-)

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 b1bc03594e5c..5105e13a6088 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4613,7 +4613,7 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 							     dc_link_get_link_cap(aconnector->dc_link));
 
 #if defined(CONFIG_DRM_AMD_DC_DCN)
-		if (dsc_caps.is_dsc_supported)
+		if (dsc_caps.is_dsc_supported) {
 			if (dc_dsc_compute_config(aconnector->dc_link->ctx->dc->res_pool->dscs[0],
 						  &dsc_caps,
 						  aconnector->dc_link->ctx->dc->debug.dsc_min_slice_height_override,
@@ -4621,6 +4621,9 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 						  &stream->timing,
 						  &stream->timing.dsc_cfg))
 				stream->timing.flags.DSC = 1;
+			if (aconnector->dsc_settings.dsc_clock_en)
+				stream->timing.flags.DSC = 1;
+		}
 #endif
 	}
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
index 82cdf07b4bd0..6f0a81f6f3c4 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -342,6 +342,10 @@ struct amdgpu_display_manager {
 	struct amdgpu_encoder mst_encoders[AMDGPU_DM_MAX_CRTC];
 };
 
+struct dsc_preferred_settings {
+	bool dsc_clock_en;
+};
+
 struct amdgpu_dm_connector {
 
 	struct drm_connector base;
@@ -389,6 +393,7 @@ struct amdgpu_dm_connector {
 	uint32_t debugfs_dpcd_size;
 #endif
 	bool force_yuv420_output;
+	struct dsc_preferred_settings dsc_settings;
 };
 
 #define to_amdgpu_dm_connector(x) container_of(x, struct amdgpu_dm_connector, base)
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
index 65e9ee64c70c..d88620fa9d2b 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
@@ -998,6 +998,21 @@ static ssize_t dp_dpcd_data_read(struct file *f, char __user *buf,
 	return read_size - r;
 }
 
+/* function: read DSC status on the connector
+ *
+ * The read function: dp_dsc_clock_en_read
+ * returns current status of DSC clock on the connector.
+ * The return is a boolean flag: 1 or 0.
+ *
+ * Access it with the following command (you need to specify
+ * connector like DP-1):
+ *
+ *	cat /sys/kernel/debug/dri/0/DP-X/dsc_clock_en
+ *
+ * Expected output:
+ * 1 - means that DSC is currently enabled
+ * 0 - means that DSC is disabled
+ */
 static ssize_t dp_dsc_clock_en_read(struct file *f, char __user *buf,
 				    size_t size, loff_t *pos)
 {
@@ -1055,6 +1070,79 @@ static ssize_t dp_dsc_clock_en_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/* function: write force DSC on the connector
+ *
+ * The write function: dp_dsc_clock_en_write
+ * enables to force DSC on the connector.
+ * User can write to either force enable DSC
+ * on the next modeset or set it to driver default
+ *
+ * Writing DSC settings is done with the following command:
+ * - To force enable DSC (you need to specify
+ * connector like DP-1):
+ *
+ *	echo 0x1 > /sys/kernel/debug/dri/0/DP-X/dsc_clock_en
+ *
+ * - To return to default state set the flag to zero and
+ * let driver deal with DSC automatically
+ * (you need to specify connector like DP-1):
+ *
+ *	echo 0x0 > /sys/kernel/debug/dri/0/DP-X/dsc_clock_en
+ *
+ */
+static ssize_t dp_dsc_clock_en_write(struct file *f, const char __user *buf,
+				     size_t size, loff_t *pos)
+{
+	struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private;
+	struct pipe_ctx *pipe_ctx;
+	int i;
+	char *wr_buf = NULL;
+	uint32_t wr_buf_size = 42;
+	int max_param_num = 1;
+	long param[1] = {0};
+	uint8_t param_nums = 0;
+
+	if (size == 0)
+		return -EINVAL;
+
+	wr_buf = kcalloc(wr_buf_size, sizeof(char), GFP_KERNEL);
+
+	if (!wr_buf) {
+		DRM_DEBUG_DRIVER("no memory to allocate write buffer\n");
+		return -ENOSPC;
+	}
+
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
+					    (long *)param, buf,
+					    max_param_num,
+					    &param_nums)) {
+		kfree(wr_buf);
+		return -EINVAL;
+	}
+
+	if (param_nums <= 0) {
+		DRM_DEBUG_DRIVER("user data not be read\n");
+		kfree(wr_buf);
+		return -EINVAL;
+	}
+
+	for (i = 0; i < MAX_PIPES; i++) {
+		pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i];
+			if (pipe_ctx && pipe_ctx->stream &&
+			    pipe_ctx->stream->link == aconnector->dc_link)
+				break;
+	}
+
+	if (!pipe_ctx || !pipe_ctx->stream)
+		goto done;
+
+	aconnector->dsc_settings.dsc_clock_en = param[0];
+
+done:
+	kfree(wr_buf);
+	return size;
+}
+
 static ssize_t dp_dsc_slice_width_read(struct file *f, char __user *buf,
 				    size_t size, loff_t *pos)
 {
@@ -1465,6 +1553,7 @@ DEFINE_SHOW_ATTRIBUTE(hdcp_sink_capability);
 static const struct file_operations dp_dsc_clock_en_debugfs_fops = {
 	.owner = THIS_MODULE,
 	.read = dp_dsc_clock_en_read,
+	.write = dp_dsc_clock_en_write,
 	.llseek = default_llseek
 };
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 729ca38750d9..8250c37abe40 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -471,6 +471,7 @@ struct dsc_mst_fairness_params {
 	struct dc_dsc_bw_range bw_range;
 	bool compression_possible;
 	struct drm_dp_mst_port *port;
+	bool clock_overwrite;
 };
 
 struct dsc_mst_fairness_vars {
@@ -636,7 +637,9 @@ static void try_disable_dsc(struct drm_atomic_state *state,
 	int remaining_to_try = 0;
 
 	for (i = 0; i < count; i++) {
-		if (vars[i].dsc_enabled && vars[i].bpp_x16 == params[i].bw_range.max_target_bpp_x16) {
+		if (vars[i].dsc_enabled
+				&& vars[i].bpp_x16 == params[i].bw_range.max_target_bpp_x16
+				&& !params[i].clock_overwrite) {
 			kbps_increase[i] = params[i].bw_range.stream_kbps - params[i].bw_range.max_kbps;
 			tried[i] = false;
 			remaining_to_try += 1;
@@ -697,6 +700,7 @@ static bool compute_mst_dsc_configs_for_link(struct drm_atomic_state *state,
 	struct dsc_mst_fairness_vars vars[MAX_PIPES];
 	struct amdgpu_dm_connector *aconnector;
 	int count = 0;
+	bool debugfs_overwrite = false;
 
 	memset(params, 0, sizeof(params));
 
@@ -715,6 +719,9 @@ static bool compute_mst_dsc_configs_for_link(struct drm_atomic_state *state,
 		params[count].sink = stream->sink;
 		aconnector = (struct amdgpu_dm_connector *)stream->dm_stream_context;
 		params[count].port = aconnector->port;
+		params[count].clock_overwrite = aconnector->dsc_settings.dsc_clock_en;
+		if (params[count].clock_overwrite)
+			debugfs_overwrite = true;
 		params[count].compression_possible = stream->sink->dsc_caps.dsc_dec_caps.is_dsc_supported;
 		dc_dsc_get_policy_for_timing(params[count].timing, &dsc_policy);
 		if (!dc_dsc_compute_bandwidth_range(
@@ -740,7 +747,7 @@ static bool compute_mst_dsc_configs_for_link(struct drm_atomic_state *state,
 						 dm_mst_get_pbn_divider(dc_link)) < 0)
 			return false;
 	}
-	if (!drm_dp_mst_atomic_check(state)) {
+	if (!drm_dp_mst_atomic_check(state) && !debugfs_overwrite) {
 		set_dsc_configs_from_fairness_vars(params, vars, count);
 		return true;
 	}
-- 
2.25.1

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

[PATCH 10/15] drm/amd/display: [FW Promotion] Release 0.0.26

[Eryk Brol]

From: Anthony Koo <Anthony.Koo@amd.com>

Signed-off-by: Anthony Koo <Anthony.Koo@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
index e013875b89ed..8b27cf0f1d51 100644
--- a/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
+++ b/drivers/gpu/drm/amd/display/dmub/inc/dmub_cmd.h
@@ -36,10 +36,10 @@
 
 /* Firmware versioning. */
 #ifdef DMUB_EXPOSE_VERSION
-#define DMUB_FW_VERSION_GIT_HASH 0xe6d590b09
+#define DMUB_FW_VERSION_GIT_HASH 0x636e7b294
 #define DMUB_FW_VERSION_MAJOR 0
 #define DMUB_FW_VERSION_MINOR 0
-#define DMUB_FW_VERSION_REVISION 25
+#define DMUB_FW_VERSION_REVISION 26
 #define DMUB_FW_VERSION_UCODE ((DMUB_FW_VERSION_MAJOR << 24) | (DMUB_FW_VERSION_MINOR << 16) | DMUB_FW_VERSION_REVISION)
 #endif
 
-- 
2.25.1

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

[PATCH 11/15] drm/amd/display: dchubbub p-state warning during surface planes switch

[Eryk Brol]

From: hersen wu <hersenxs.wu@amd.com>

[Why]
ramp_up_dispclk_with_dpp is to change dispclk, dppclk and dprefclk
according to bandwidth requirement. call stack: rv1_update_clocks -->
update_clocks --> dcn10_prepare_bandwidth / dcn10_optimize_bandwidth
--> prepare_bandwidth / optimize_bandwidth. before change dcn hw,
prepare_bandwidth will be called first to allow enough clock,
watermark for change, after end of dcn hw change, optimize_bandwidth
is executed to lower clock to save power for new dcn hw settings.

below is sequence of commit_planes_for_stream:
step 1: prepare_bandwidth - raise clock to have enough bandwidth
step 2: lock_doublebuffer_enable
step 3: pipe_control_lock(true) - make dchubp register change will
not take effect right way
step 4: apply_ctx_for_surface - program dchubp
step 5: pipe_control_lock(false) - dchubp register change take effect
step 6: optimize_bandwidth --> dc_post_update_surfaces_to_stream
for full_date, optimize clock to save power

at end of step 1, dcn clocks (dprefclk, dispclk, dppclk) may be
changed for new dchubp configuration. but real dcn hub dchubps are
still running with old configuration until end of step 5. this need
clocks settings at step 1 should not less than that before step 1.
this is checked by two conditions: 1. if (should_set_clock(safe_to_lower
, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz) ||
new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz)
2. request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz

the second condition is based on new dchubp configuration. dppclk
for new dchubp may be different from dppclk before step 1.
for example, before step 1, dchubps are as below:
pipe 0: recout=(0,40,1920,980) viewport=(0,0,1920,979)
pipe 1: recout=(0,0,1920,1080) viewport=(0,0,1920,1080)
for dppclk for pipe0 need dppclk = dispclk

new dchubp pipe split configuration:
pipe 0: recout=(0,0,960,1080) viewport=(0,0,960,1080)
pipe 1: recout=(960,0,960,1080) viewport=(960,0,960,1080)
dppclk only needs dppclk = dispclk /2.

dispclk, dppclk are not lock by otg master lock. they take effect
after step 1. during this transition, dispclk are the same, but
dppclk is changed to half of previous clock for old dchubp
configuration between step 1 and step 6. This may cause p-state
warning intermittently.

[How]
for new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz, we
need make sure dppclk are not changed to less between step 1 and 6.
for new_clocks->dispclk_khz > clk_mgr_base->clks.dispclk_khz,
new display clock is raised, but we do not know ratio of
new_clocks->dispclk_khz and clk_mgr_base->clks.dispclk_khz,
new_clocks->dispclk_khz /2 does not guarantee equal or higher than
old dppclk. we could ignore power saving different between
dppclk = displck and dppclk = dispclk / 2 between step 1 and step 6.
as long as safe_to_lower = false, set dpclk = dispclk to simplify
condition check.

CC: Stable <stable@vger.kernel.org>
Signed-off-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../display/dc/clk_mgr/dcn10/rv1_clk_mgr.c    | 69 ++++++++++++++++++-
 1 file changed, 67 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c
index 3fab9296918a..e133edc587d3 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c
@@ -85,12 +85,77 @@ static int rv1_determine_dppclk_threshold(struct clk_mgr_internal *clk_mgr, stru
 	return disp_clk_threshold;
 }
 
-static void ramp_up_dispclk_with_dpp(struct clk_mgr_internal *clk_mgr, struct dc *dc, struct dc_clocks *new_clocks)
+static void ramp_up_dispclk_with_dpp(
+		struct clk_mgr_internal *clk_mgr,
+		struct dc *dc,
+		struct dc_clocks *new_clocks,
+		bool safe_to_lower)
 {
 	int i;
 	int dispclk_to_dpp_threshold = rv1_determine_dppclk_threshold(clk_mgr, new_clocks);
 	bool request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz;
 
+	/* this function is to change dispclk, dppclk and dprefclk according to
+	 * bandwidth requirement. Its call stack is rv1_update_clocks -->
+	 * update_clocks --> dcn10_prepare_bandwidth / dcn10_optimize_bandwidth
+	 * --> prepare_bandwidth / optimize_bandwidth. before change dcn hw,
+	 * prepare_bandwidth will be called first to allow enough clock,
+	 * watermark for change, after end of dcn hw change, optimize_bandwidth
+	 * is executed to lower clock to save power for new dcn hw settings.
+	 *
+	 * below is sequence of commit_planes_for_stream:
+	 *
+	 * step 1: prepare_bandwidth - raise clock to have enough bandwidth
+	 * step 2: lock_doublebuffer_enable
+	 * step 3: pipe_control_lock(true) - make dchubp register change will
+	 * not take effect right way
+	 * step 4: apply_ctx_for_surface - program dchubp
+	 * step 5: pipe_control_lock(false) - dchubp register change take effect
+	 * step 6: optimize_bandwidth --> dc_post_update_surfaces_to_stream
+	 * for full_date, optimize clock to save power
+	 *
+	 * at end of step 1, dcn clocks (dprefclk, dispclk, dppclk) may be
+	 * changed for new dchubp configuration. but real dcn hub dchubps are
+	 * still running with old configuration until end of step 5. this need
+	 * clocks settings at step 1 should not less than that before step 1.
+	 * this is checked by two conditions: 1. if (should_set_clock(safe_to_lower
+	 * , new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz) ||
+	 * new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz)
+	 * 2. request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz
+	 *
+	 * the second condition is based on new dchubp configuration. dppclk
+	 * for new dchubp may be different from dppclk before step 1.
+	 * for example, before step 1, dchubps are as below:
+	 * pipe 0: recout=(0,40,1920,980) viewport=(0,0,1920,979)
+	 * pipe 1: recout=(0,0,1920,1080) viewport=(0,0,1920,1080)
+	 * for dppclk for pipe0 need dppclk = dispclk
+	 *
+	 * new dchubp pipe split configuration:
+	 * pipe 0: recout=(0,0,960,1080) viewport=(0,0,960,1080)
+	 * pipe 1: recout=(960,0,960,1080) viewport=(960,0,960,1080)
+	 * dppclk only needs dppclk = dispclk /2.
+	 *
+	 * dispclk, dppclk are not lock by otg master lock. they take effect
+	 * after step 1. during this transition, dispclk are the same, but
+	 * dppclk is changed to half of previous clock for old dchubp
+	 * configuration between step 1 and step 6. This may cause p-state
+	 * warning intermittently.
+	 *
+	 * for new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz, we
+	 * need make sure dppclk are not changed to less between step 1 and 6.
+	 * for new_clocks->dispclk_khz > clk_mgr_base->clks.dispclk_khz,
+	 * new display clock is raised, but we do not know ratio of
+	 * new_clocks->dispclk_khz and clk_mgr_base->clks.dispclk_khz,
+	 * new_clocks->dispclk_khz /2 does not guarantee equal or higher than
+	 * old dppclk. we could ignore power saving different between
+	 * dppclk = displck and dppclk = dispclk / 2 between step 1 and step 6.
+	 * as long as safe_to_lower = false, set dpclk = dispclk to simplify
+	 * condition check.
+	 * todo: review this change for other asic.
+	 **/
+	if (!safe_to_lower)
+		request_dpp_div = false;
+
 	/* set disp clk to dpp clk threshold */
 
 	clk_mgr->funcs->set_dispclk(clk_mgr, dispclk_to_dpp_threshold);
@@ -209,7 +274,7 @@ static void rv1_update_clocks(struct clk_mgr *clk_mgr_base,
 	/* program dispclk on = as a w/a for sleep resume clock ramping issues */
 	if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)
 			|| new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz) {
-		ramp_up_dispclk_with_dpp(clk_mgr, dc, new_clocks);
+		ramp_up_dispclk_with_dpp(clk_mgr, dc, new_clocks, safe_to_lower);
 		clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz;
 		send_request_to_lower = true;
 	}
-- 
2.25.1

[PATCH 11/15] drm/amd/display: dchubbub p-state warning during surface planes switch

[Eryk Brol]

From: hersen wu <hersenxs.wu@amd.com>

[Why]
ramp_up_dispclk_with_dpp is to change dispclk, dppclk and dprefclk
according to bandwidth requirement. call stack: rv1_update_clocks -->
update_clocks --> dcn10_prepare_bandwidth / dcn10_optimize_bandwidth
--> prepare_bandwidth / optimize_bandwidth. before change dcn hw,
prepare_bandwidth will be called first to allow enough clock,
watermark for change, after end of dcn hw change, optimize_bandwidth
is executed to lower clock to save power for new dcn hw settings.

below is sequence of commit_planes_for_stream:
step 1: prepare_bandwidth - raise clock to have enough bandwidth
step 2: lock_doublebuffer_enable
step 3: pipe_control_lock(true) - make dchubp register change will
not take effect right way
step 4: apply_ctx_for_surface - program dchubp
step 5: pipe_control_lock(false) - dchubp register change take effect
step 6: optimize_bandwidth --> dc_post_update_surfaces_to_stream
for full_date, optimize clock to save power

at end of step 1, dcn clocks (dprefclk, dispclk, dppclk) may be
changed for new dchubp configuration. but real dcn hub dchubps are
still running with old configuration until end of step 5. this need
clocks settings at step 1 should not less than that before step 1.
this is checked by two conditions: 1. if (should_set_clock(safe_to_lower
, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz) ||
new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz)
2. request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz

the second condition is based on new dchubp configuration. dppclk
for new dchubp may be different from dppclk before step 1.
for example, before step 1, dchubps are as below:
pipe 0: recout=(0,40,1920,980) viewport=(0,0,1920,979)
pipe 1: recout=(0,0,1920,1080) viewport=(0,0,1920,1080)
for dppclk for pipe0 need dppclk = dispclk

new dchubp pipe split configuration:
pipe 0: recout=(0,0,960,1080) viewport=(0,0,960,1080)
pipe 1: recout=(960,0,960,1080) viewport=(960,0,960,1080)
dppclk only needs dppclk = dispclk /2.

dispclk, dppclk are not lock by otg master lock. they take effect
after step 1. during this transition, dispclk are the same, but
dppclk is changed to half of previous clock for old dchubp
configuration between step 1 and step 6. This may cause p-state
warning intermittently.

[How]
for new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz, we
need make sure dppclk are not changed to less between step 1 and 6.
for new_clocks->dispclk_khz > clk_mgr_base->clks.dispclk_khz,
new display clock is raised, but we do not know ratio of
new_clocks->dispclk_khz and clk_mgr_base->clks.dispclk_khz,
new_clocks->dispclk_khz /2 does not guarantee equal or higher than
old dppclk. we could ignore power saving different between
dppclk = displck and dppclk = dispclk / 2 between step 1 and step 6.
as long as safe_to_lower = false, set dpclk = dispclk to simplify
condition check.

CC: Stable <stable@vger.kernel.org>
Signed-off-by: Hersen Wu <hersenxs.wu@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../display/dc/clk_mgr/dcn10/rv1_clk_mgr.c    | 69 ++++++++++++++++++-
 1 file changed, 67 insertions(+), 2 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c
index 3fab9296918a..e133edc587d3 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn10/rv1_clk_mgr.c
@@ -85,12 +85,77 @@ static int rv1_determine_dppclk_threshold(struct clk_mgr_internal *clk_mgr, stru
 	return disp_clk_threshold;
 }
 
-static void ramp_up_dispclk_with_dpp(struct clk_mgr_internal *clk_mgr, struct dc *dc, struct dc_clocks *new_clocks)
+static void ramp_up_dispclk_with_dpp(
+		struct clk_mgr_internal *clk_mgr,
+		struct dc *dc,
+		struct dc_clocks *new_clocks,
+		bool safe_to_lower)
 {
 	int i;
 	int dispclk_to_dpp_threshold = rv1_determine_dppclk_threshold(clk_mgr, new_clocks);
 	bool request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz;
 
+	/* this function is to change dispclk, dppclk and dprefclk according to
+	 * bandwidth requirement. Its call stack is rv1_update_clocks -->
+	 * update_clocks --> dcn10_prepare_bandwidth / dcn10_optimize_bandwidth
+	 * --> prepare_bandwidth / optimize_bandwidth. before change dcn hw,
+	 * prepare_bandwidth will be called first to allow enough clock,
+	 * watermark for change, after end of dcn hw change, optimize_bandwidth
+	 * is executed to lower clock to save power for new dcn hw settings.
+	 *
+	 * below is sequence of commit_planes_for_stream:
+	 *
+	 * step 1: prepare_bandwidth - raise clock to have enough bandwidth
+	 * step 2: lock_doublebuffer_enable
+	 * step 3: pipe_control_lock(true) - make dchubp register change will
+	 * not take effect right way
+	 * step 4: apply_ctx_for_surface - program dchubp
+	 * step 5: pipe_control_lock(false) - dchubp register change take effect
+	 * step 6: optimize_bandwidth --> dc_post_update_surfaces_to_stream
+	 * for full_date, optimize clock to save power
+	 *
+	 * at end of step 1, dcn clocks (dprefclk, dispclk, dppclk) may be
+	 * changed for new dchubp configuration. but real dcn hub dchubps are
+	 * still running with old configuration until end of step 5. this need
+	 * clocks settings at step 1 should not less than that before step 1.
+	 * this is checked by two conditions: 1. if (should_set_clock(safe_to_lower
+	 * , new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz) ||
+	 * new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz)
+	 * 2. request_dpp_div = new_clocks->dispclk_khz > new_clocks->dppclk_khz
+	 *
+	 * the second condition is based on new dchubp configuration. dppclk
+	 * for new dchubp may be different from dppclk before step 1.
+	 * for example, before step 1, dchubps are as below:
+	 * pipe 0: recout=(0,40,1920,980) viewport=(0,0,1920,979)
+	 * pipe 1: recout=(0,0,1920,1080) viewport=(0,0,1920,1080)
+	 * for dppclk for pipe0 need dppclk = dispclk
+	 *
+	 * new dchubp pipe split configuration:
+	 * pipe 0: recout=(0,0,960,1080) viewport=(0,0,960,1080)
+	 * pipe 1: recout=(960,0,960,1080) viewport=(960,0,960,1080)
+	 * dppclk only needs dppclk = dispclk /2.
+	 *
+	 * dispclk, dppclk are not lock by otg master lock. they take effect
+	 * after step 1. during this transition, dispclk are the same, but
+	 * dppclk is changed to half of previous clock for old dchubp
+	 * configuration between step 1 and step 6. This may cause p-state
+	 * warning intermittently.
+	 *
+	 * for new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz, we
+	 * need make sure dppclk are not changed to less between step 1 and 6.
+	 * for new_clocks->dispclk_khz > clk_mgr_base->clks.dispclk_khz,
+	 * new display clock is raised, but we do not know ratio of
+	 * new_clocks->dispclk_khz and clk_mgr_base->clks.dispclk_khz,
+	 * new_clocks->dispclk_khz /2 does not guarantee equal or higher than
+	 * old dppclk. we could ignore power saving different between
+	 * dppclk = displck and dppclk = dispclk / 2 between step 1 and step 6.
+	 * as long as safe_to_lower = false, set dpclk = dispclk to simplify
+	 * condition check.
+	 * todo: review this change for other asic.
+	 **/
+	if (!safe_to_lower)
+		request_dpp_div = false;
+
 	/* set disp clk to dpp clk threshold */
 
 	clk_mgr->funcs->set_dispclk(clk_mgr, dispclk_to_dpp_threshold);
@@ -209,7 +274,7 @@ static void rv1_update_clocks(struct clk_mgr *clk_mgr_base,
 	/* program dispclk on = as a w/a for sleep resume clock ramping issues */
 	if (should_set_clock(safe_to_lower, new_clocks->dispclk_khz, clk_mgr_base->clks.dispclk_khz)
 			|| new_clocks->dispclk_khz == clk_mgr_base->clks.dispclk_khz) {
-		ramp_up_dispclk_with_dpp(clk_mgr, dc, new_clocks);
+		ramp_up_dispclk_with_dpp(clk_mgr, dc, new_clocks, safe_to_lower);
 		clk_mgr_base->clks.dispclk_khz = new_clocks->dispclk_khz;
 		send_request_to_lower = true;
 	}
-- 
2.25.1

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

Re: [PATCH 11/15] drm/amd/display: dchubbub p-state warning during surface planes switch

[Sasha Levin]

Hi

[This is an automated email]

This commit has been processed because it contains a -stable tag.
The stable tag indicates that it's relevant for the following trees: all

The bot has tested the following trees: v5.7.10, v5.4.53, v4.19.134, v4.14.189, v4.9.231, v4.4.231.

v5.7.10: Build OK!
v5.4.53: Build OK!
v4.19.134: Failed to apply! Possible dependencies:
    1e7e86c43f38 ("drm/amd/display: decouple front and backend pgm using dpms_off as backend enable flag")
    24f7dd7ea98d ("drm/amd/display: move pplib/smu notification to dccg block")
    4c5e8b541527 ("drm/amd/display: split dccg clock manager into asic folders")
    8c3db1284a01 ("drm/amdgpu: fill in amdgpu_dm_remove_sink_from_freesync_module")
    98e6436d3af5 ("drm/amd/display: Refactor FreeSync module")
    ad908423ef86 ("drm/amd/display: support 48 MHZ refclk off")
    c85e6e546edd ("drm/amd/display: Create new i2c resource")

v4.14.189: Failed to apply! Possible dependencies:
    1b0c0f9dc5ca ("drm/amdgpu: move userptr BOs to CPU domain during CS v2")
    1ed3d2567c80 ("drm/amdgpu: keep the MMU lock until the update ends v4")
    3fe89771cb0a ("drm/amdgpu: stop reserving the BO in the MMU callback v3")
    4562236b3bc0 ("drm/amd/dc: Add dc display driver (v2)")
    4c5e8b541527 ("drm/amd/display: split dccg clock manager into asic folders")
    60de1c1740f3 ("drm/amdgpu: use a rw_semaphore for MMU notifiers")
    9a18999640fa ("drm/amdgpu: move MMU notifier related defines to amdgpu_mn.h")
    9cca0b8e5df0 ("drm/amdgpu: move amdgpu_cs_sysvm_access_required into find_mapping")
    a216ab09955d ("drm/amdgpu: fix userptr put_page handling")
    b72cf4fca2bb ("drm/amdgpu: move taking mmap_sem into get_user_pages v2")
    ca666a3c298f ("drm/amdgpu: stop using BO status for user pages")

v4.9.231: Failed to apply! Possible dependencies:
    1cec20f0ea0e ("dma-buf: Restart reservation_object_wait_timeout_rcu() after writes")
    248a1d6f1ac4 ("drm/amd: fix include notation and remove -Iinclude/drm flag")
    4562236b3bc0 ("drm/amd/dc: Add dc display driver (v2)")
    4c5e8b541527 ("drm/amd/display: split dccg clock manager into asic folders")
    78010cd9736e ("dma-buf/fence: add an lockdep_assert_held()")
    f54d1867005c ("dma-buf: Rename struct fence to dma_fence")
    fedf54132d24 ("dma-buf: Restart reservation_object_get_fences_rcu() after writes")

v4.4.231: Failed to apply! Possible dependencies:
    0f477c6dea70 ("staging/android/sync: add sync_fence_create_dma")
    1f7371b2a5fa ("drm/amd/powerplay: add basic powerplay framework")
    248a1d6f1ac4 ("drm/amd: fix include notation and remove -Iinclude/drm flag")
    288912cb95d1 ("drm/amdgpu: use $(src) in Makefile (v2)")
    375fb53ec1be ("staging: android: replace explicit NULL comparison")
    395dec6f6bc5 ("Documentation: add doc for sync_file_get_fence()")
    4325198180e5 ("drm/amdgpu: remove GART page addr array")
    4562236b3bc0 ("drm/amd/dc: Add dc display driver (v2)")
    4c5e8b541527 ("drm/amd/display: split dccg clock manager into asic folders")
    62304fb1fc08 ("dma-buf/sync_file: de-stage sync_file")
    a1d29476d666 ("drm/amdgpu: optionally enable GART debugfs file")
    a8fe58cec351 ("drm/amd: add ACP driver support")
    b70f014d58b9 ("drm/amdgpu: change default sched jobs to 32")
    c784c82a3fd6 ("Documentation: add Sync File doc")
    d4cab38e153d ("staging/android: prepare sync_file for de-staging")
    d7fdb0ae9d11 ("staging/android: rename sync_fence to sync_file")
    f54d1867005c ("dma-buf: Rename struct fence to dma_fence")
    fac8434dab96 ("Documentation: Fix some grammar mistakes in sync_file.txt")
    fdba11f4079e ("drm/amdgpu: move all Kconfig options to amdgpu/Kconfig")


NOTE: The patch will not be queued to stable trees until it is upstream.

How should we proceed with this patch?

-- 
Thanks
Sasha

Re: [PATCH 11/15] drm/amd/display: dchubbub p-state warning during surface planes switch

[Sasha Levin]

Hi

[This is an automated email]

This commit has been processed because it contains a -stable tag.
The stable tag indicates that it's relevant for the following trees: all

The bot has tested the following trees: v5.7.10, v5.4.53, v4.19.134, v4.14.189, v4.9.231, v4.4.231.

v5.7.10: Build OK!
v5.4.53: Build OK!
v4.19.134: Failed to apply! Possible dependencies:
    1e7e86c43f38 ("drm/amd/display: decouple front and backend pgm using dpms_off as backend enable flag")
    24f7dd7ea98d ("drm/amd/display: move pplib/smu notification to dccg block")
    4c5e8b541527 ("drm/amd/display: split dccg clock manager into asic folders")
    8c3db1284a01 ("drm/amdgpu: fill in amdgpu_dm_remove_sink_from_freesync_module")
    98e6436d3af5 ("drm/amd/display: Refactor FreeSync module")
    ad908423ef86 ("drm/amd/display: support 48 MHZ refclk off")
    c85e6e546edd ("drm/amd/display: Create new i2c resource")

v4.14.189: Failed to apply! Possible dependencies:
    1b0c0f9dc5ca ("drm/amdgpu: move userptr BOs to CPU domain during CS v2")
    1ed3d2567c80 ("drm/amdgpu: keep the MMU lock until the update ends v4")
    3fe89771cb0a ("drm/amdgpu: stop reserving the BO in the MMU callback v3")
    4562236b3bc0 ("drm/amd/dc: Add dc display driver (v2)")
    4c5e8b541527 ("drm/amd/display: split dccg clock manager into asic folders")
    60de1c1740f3 ("drm/amdgpu: use a rw_semaphore for MMU notifiers")
    9a18999640fa ("drm/amdgpu: move MMU notifier related defines to amdgpu_mn.h")
    9cca0b8e5df0 ("drm/amdgpu: move amdgpu_cs_sysvm_access_required into find_mapping")
    a216ab09955d ("drm/amdgpu: fix userptr put_page handling")
    b72cf4fca2bb ("drm/amdgpu: move taking mmap_sem into get_user_pages v2")
    ca666a3c298f ("drm/amdgpu: stop using BO status for user pages")

v4.9.231: Failed to apply! Possible dependencies:
    1cec20f0ea0e ("dma-buf: Restart reservation_object_wait_timeout_rcu() after writes")
    248a1d6f1ac4 ("drm/amd: fix include notation and remove -Iinclude/drm flag")
    4562236b3bc0 ("drm/amd/dc: Add dc display driver (v2)")
    4c5e8b541527 ("drm/amd/display: split dccg clock manager into asic folders")
    78010cd9736e ("dma-buf/fence: add an lockdep_assert_held()")
    f54d1867005c ("dma-buf: Rename struct fence to dma_fence")
    fedf54132d24 ("dma-buf: Restart reservation_object_get_fences_rcu() after writes")

v4.4.231: Failed to apply! Possible dependencies:
    0f477c6dea70 ("staging/android/sync: add sync_fence_create_dma")
    1f7371b2a5fa ("drm/amd/powerplay: add basic powerplay framework")
    248a1d6f1ac4 ("drm/amd: fix include notation and remove -Iinclude/drm flag")
    288912cb95d1 ("drm/amdgpu: use $(src) in Makefile (v2)")
    375fb53ec1be ("staging: android: replace explicit NULL comparison")
    395dec6f6bc5 ("Documentation: add doc for sync_file_get_fence()")
    4325198180e5 ("drm/amdgpu: remove GART page addr array")
    4562236b3bc0 ("drm/amd/dc: Add dc display driver (v2)")
    4c5e8b541527 ("drm/amd/display: split dccg clock manager into asic folders")
    62304fb1fc08 ("dma-buf/sync_file: de-stage sync_file")
    a1d29476d666 ("drm/amdgpu: optionally enable GART debugfs file")
    a8fe58cec351 ("drm/amd: add ACP driver support")
    b70f014d58b9 ("drm/amdgpu: change default sched jobs to 32")
    c784c82a3fd6 ("Documentation: add Sync File doc")
    d4cab38e153d ("staging/android: prepare sync_file for de-staging")
    d7fdb0ae9d11 ("staging/android: rename sync_fence to sync_file")
    f54d1867005c ("dma-buf: Rename struct fence to dma_fence")
    fac8434dab96 ("Documentation: Fix some grammar mistakes in sync_file.txt")
    fdba11f4079e ("drm/amdgpu: move all Kconfig options to amdgpu/Kconfig")


NOTE: The patch will not be queued to stable trees until it is upstream.

How should we proceed with this patch?

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

[PATCH 12/15] drm/amd/display: Use hw lock mgr

[Eryk Brol]

From: Wyatt Wood <wyatt.wood@amd.com>

[Why]
Feature requires synchronization of dig, pipe, and cursor locking
between driver and fw.

[How]
Set flag to force psr to use hw lock mgr.

Signed-off-by: Wyatt Wood <wyatt.wood@amd.com>
Reviewed-by: Anthony Koo <Anthony.Koo@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
index 82e67bd81f2d..5167d6b8a48d 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
@@ -233,8 +233,8 @@ static bool dmub_psr_copy_settings(struct dmub_psr *dmub,
 	copy_settings_data->frame_cap_ind			= psr_context->psrFrameCaptureIndicationReq;
 	copy_settings_data->debug.bitfields.visual_confirm	= dc->dc->debug.visual_confirm == VISUAL_CONFIRM_PSR ?
 									true : false;
+	copy_settings_data->debug.bitfields.use_hw_lock_mgr	= 1;
 	copy_settings_data->init_sdp_deadline			= psr_context->sdpTransmitLineNumDeadline;
-	copy_settings_data->debug.bitfields.use_hw_lock_mgr	= 0;
 
 	dc_dmub_srv_cmd_queue(dc->dmub_srv, &cmd);
 	dc_dmub_srv_cmd_execute(dc->dmub_srv);
-- 
2.25.1

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

[PATCH 13/15] drm/amd/display: DSC Slice width debugfs write entry

[Eryk Brol]

[Why]
We need to be able to specify slice width for DSC on aconnector

[How]
Getting slice width parameter from debugfs entry, if it is
a valid the value is set in connector's dsc preffered settings
structure. Which then overwrites dsc_cfg structure's parameters
if DSC is decided to be enabled. Works for both SST and MST.

Signed-off-by: Eryk Brol <eryk.brol@amd.com>
Signed-off-by: Mikita Lipski <mikita.lipski@amd.com>
Reviewed-by: Mikita Lipski <Mikita.Lipski@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c |  4 +
 .../gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h |  1 +
 .../amd/display/amdgpu_dm/amdgpu_dm_debugfs.c | 93 +++++++++++++++++++
 .../display/amdgpu_dm/amdgpu_dm_mst_types.c   |  6 ++
 4 files changed, 104 insertions(+)

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 5105e13a6088..0327e785e45c 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.c
@@ -4621,8 +4621,12 @@ create_stream_for_sink(struct amdgpu_dm_connector *aconnector,
 						  &stream->timing,
 						  &stream->timing.dsc_cfg))
 				stream->timing.flags.DSC = 1;
+			/* Overwrite the stream flag if DSC is enabled through debugfs */
 			if (aconnector->dsc_settings.dsc_clock_en)
 				stream->timing.flags.DSC = 1;
+			if (stream->timing.flags.DSC && aconnector->dsc_settings.dsc_slice_width)
+				stream->timing.dsc_cfg.num_slices_h = DIV_ROUND_UP(stream->timing.h_addressable,
+										   aconnector->dsc_settings.dsc_slice_width);
 		}
 #endif
 	}
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
index 6f0a81f6f3c4..5b6f879a108c 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm.h
@@ -344,6 +344,7 @@ struct amdgpu_display_manager {
 
 struct dsc_preferred_settings {
 	bool dsc_clock_en;
+	uint32_t dsc_slice_width;
 };
 
 struct amdgpu_dm_connector {
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
index d88620fa9d2b..df8b84aecc4d 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_debugfs.c
@@ -1143,6 +1143,22 @@ static ssize_t dp_dsc_clock_en_write(struct file *f, const char __user *buf,
 	return size;
 }
 
+/* function: read DSC slice width parameter on the connector
+ *
+ * The read function: dp_dsc_slice_width_read
+ * returns dsc slice width used in the current configuration
+ * The return is an integer: 0 or other positive number
+ *
+ * Access the status with the following command:
+ *
+ *	cat /sys/kernel/debug/dri/0/DP-X/dsc_slice_width
+ *
+ * 0 - means that DSC is disabled
+ *
+ * Any other number more than zero represents the
+ * slice width currently used by DSC in pixels
+ *
+ */
 static ssize_t dp_dsc_slice_width_read(struct file *f, char __user *buf,
 				    size_t size, loff_t *pos)
 {
@@ -1200,6 +1216,82 @@ static ssize_t dp_dsc_slice_width_read(struct file *f, char __user *buf,
 	return result;
 }
 
+/* function: write DSC slice width parameter
+ *
+ * The write function: dp_dsc_slice_width_write
+ * overwrites automatically generated DSC configuration
+ * of slice width.
+ *
+ * The user has to write the slice width divisible by the
+ * picture width.
+ *
+ * Also the user has to write width in hexidecimal
+ * rather than in decimal.
+ *
+ * Writing DSC settings is done with the following command:
+ * - To force overwrite slice width: (example sets to 1920 pixels)
+ *
+ *	echo 0x780 > /sys/kernel/debug/dri/0/DP-X/dsc_slice_width
+ *
+ *  - To stop overwriting and let driver find the optimal size,
+ * set the width to zero:
+ *
+ *	echo 0x0 > /sys/kernel/debug/dri/0/DP-X/dsc_slice_width
+ *
+ */
+static ssize_t dp_dsc_slice_width_write(struct file *f, const char __user *buf,
+				     size_t size, loff_t *pos)
+{
+	struct amdgpu_dm_connector *aconnector = file_inode(f)->i_private;
+	struct pipe_ctx *pipe_ctx;
+	int i;
+	char *wr_buf = NULL;
+	uint32_t wr_buf_size = 42;
+	int max_param_num = 1;
+	long param[1] = {0};
+	uint8_t param_nums = 0;
+
+	if (size == 0)
+		return -EINVAL;
+
+	wr_buf = kcalloc(wr_buf_size, sizeof(char), GFP_KERNEL);
+
+	if (!wr_buf) {
+		DRM_DEBUG_DRIVER("no memory to allocate write buffer\n");
+		return -ENOSPC;
+	}
+
+	if (parse_write_buffer_into_params(wr_buf, wr_buf_size,
+					    (long *)param, buf,
+					    max_param_num,
+					    &param_nums)) {
+		kfree(wr_buf);
+		return -EINVAL;
+	}
+
+	if (param_nums <= 0) {
+		DRM_DEBUG_DRIVER("user data not be read\n");
+		kfree(wr_buf);
+		return -EINVAL;
+	}
+
+	for (i = 0; i < MAX_PIPES; i++) {
+		pipe_ctx = &aconnector->dc_link->dc->current_state->res_ctx.pipe_ctx[i];
+			if (pipe_ctx && pipe_ctx->stream &&
+			    pipe_ctx->stream->link == aconnector->dc_link)
+				break;
+	}
+
+	if (!pipe_ctx || !pipe_ctx->stream)
+		goto done;
+
+	aconnector->dsc_settings.dsc_slice_width = param[0];
+
+done:
+	kfree(wr_buf);
+	return size;
+}
+
 static ssize_t dp_dsc_slice_height_read(struct file *f, char __user *buf,
 				    size_t size, loff_t *pos)
 {
@@ -1560,6 +1652,7 @@ static const struct file_operations dp_dsc_clock_en_debugfs_fops = {
 static const struct file_operations dp_dsc_slice_width_debugfs_fops = {
 	.owner = THIS_MODULE,
 	.read = dp_dsc_slice_width_read,
+	.write = dp_dsc_slice_width_write,
 	.llseek = default_llseek
 };
 
diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
index 8250c37abe40..2c10352fa514 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_mst_types.c
@@ -472,6 +472,7 @@ struct dsc_mst_fairness_params {
 	bool compression_possible;
 	struct drm_dp_mst_port *port;
 	bool clock_overwrite;
+	uint32_t slice_width_overwrite;
 };
 
 struct dsc_mst_fairness_vars {
@@ -506,6 +507,10 @@ static void set_dsc_configs_from_fairness_vars(struct dsc_mst_fairness_params *p
 					&params[i].timing->dsc_cfg)) {
 			params[i].timing->flags.DSC = 1;
 			params[i].timing->dsc_cfg.bits_per_pixel = vars[i].bpp_x16;
+			if (params[i].slice_width_overwrite)
+				params[i].timing->dsc_cfg.num_slices_h = DIV_ROUND_UP(
+										params[i].timing->h_addressable,
+										params[i].slice_width_overwrite);
 		} else {
 			params[i].timing->flags.DSC = 0;
 		}
@@ -722,6 +727,7 @@ static bool compute_mst_dsc_configs_for_link(struct drm_atomic_state *state,
 		params[count].clock_overwrite = aconnector->dsc_settings.dsc_clock_en;
 		if (params[count].clock_overwrite)
 			debugfs_overwrite = true;
+		params[count].slice_width_overwrite = aconnector->dsc_settings.dsc_slice_width;
 		params[count].compression_possible = stream->sink->dsc_caps.dsc_dec_caps.is_dsc_supported;
 		dc_dsc_get_policy_for_timing(params[count].timing, &dsc_policy);
 		if (!dc_dsc_compute_bandwidth_range(
-- 
2.25.1

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

[PATCH 14/15] drm/amd/display: 3.2.96

[Eryk Brol]

From: Aric Cyr <aric.cyr@amd.com>

Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Reviewed-by: Aric Cyr <Aric.Cyr@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index f50ef4255020..ae0e27c67ef9 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -42,7 +42,7 @@
 #include "inc/hw/dmcu.h"
 #include "dml/display_mode_lib.h"
 
-#define DC_VER "3.2.95"
+#define DC_VER "3.2.96"
 
 #define MAX_SURFACES 3
 #define MAX_PLANES 6
-- 
2.25.1

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

[PATCH 15/15] drm/amd/display: Fix DP Compliance tests 4.3.2.1 and 4.3.2.2

[Eryk Brol]

From: Aric Cyr <aric.cyr@amd.com>

[Why]
Test expects that we also read HPD_IRQ_VECTOR when checking for
symbol loss as well lane status.

[How]
Read bytes 0x200-0x205 instead of just 0x202-0x205

Signed-off-by: Aric Cyr <aric.cyr@amd.com>
Reviewed-by: Jun Lei <Jun.Lei@amd.com>
Acked-by: Eryk Brol <eryk.brol@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
index 0447492d30ca..2bfa4e35c2cf 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c
@@ -1138,23 +1138,22 @@ static enum link_training_result check_link_loss_status(
 	const struct link_training_settings *link_training_setting)
 {
 	enum link_training_result status = LINK_TRAINING_SUCCESS;
-	unsigned int lane01_status_address = DP_LANE0_1_STATUS;
 	union lane_status lane_status;
-	uint8_t dpcd_buf[4] = {0};
+	uint8_t dpcd_buf[6] = {0};
 	uint32_t lane;
 
 	core_link_read_dpcd(
-		link,
-		lane01_status_address,
-		(uint8_t *)(dpcd_buf),
-		sizeof(dpcd_buf));
+			link,
+			DP_SINK_COUNT,
+			(uint8_t *)(dpcd_buf),
+			sizeof(dpcd_buf));
 
 	/*parse lane status*/
 	for (lane = 0; lane < link->cur_link_settings.lane_count; lane++) {
 		/*
 		 * check lanes status
 		 */
-		lane_status.raw = get_nibble_at_index(&dpcd_buf[0], lane);
+		lane_status.raw = get_nibble_at_index(&dpcd_buf[2], lane);
 
 		if (!lane_status.bits.CHANNEL_EQ_DONE_0 ||
 			!lane_status.bits.CR_DONE_0 ||
-- 
2.25.1

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

Re: [PATCH 12/15] drm/amd/display: Use hw lock mgr

[Brandon Wright]

[-- Attachment #1.1: Type: text/plain, Size: 1698 bytes --]

Just curious, but I noticed this new lock manager isn't being used because
of the definitive false return value in display/dc/dce/dmub_hw_lock_mgr.c:
should_use_dmub_lock.
Was this supposed to be enabled?

From: Wyatt Wood <wyatt.w...@amd.com>
>
> [Why]
> Feature requires synchronization of dig, pipe, and cursor locking
> between driver and fw.
>
> [How]
> Set flag to force psr to use hw lock mgr.
>
> Signed-off-by: Wyatt Wood <wyatt.w...@amd.com>
> Reviewed-by: Anthony Koo <anthony....@amd.com>
> Acked-by: Eryk Brol <eryk.b...@amd.com>
> ---
>  drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
> b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
> index 82e67bd81f2d..5167d6b8a48d 100644
> --- a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
> +++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c
> @@ -233,8 +233,8 @@ static bool dmub_psr_copy_settings(struct dmub_psr *dmub,
>         copy_settings_data->frame_cap_ind                       =
> psr_context->psrFrameCaptureIndicationReq;
>         copy_settings_data->debug.bitfields.visual_confirm      =
> dc->dc->debug.visual_confirm == VISUAL_CONFIRM_PSR ?
>                                                                         true :
> false;
> +       copy_settings_data->debug.bitfields.use_hw_lock_mgr     = 1;
>         copy_settings_data->init_sdp_deadline                   =
> psr_context->sdpTransmitLineNumDeadline;
> -       copy_settings_data->debug.bitfields.use_hw_lock_mgr     = 0;
>
>         dc_dmub_srv_cmd_queue(dc->dmub_srv, &cmd);
>         dc_dmub_srv_cmd_execute(dc->dmub_srv);
> --
> 2.25.1
>
>

[-- Attachment #1.2: Type: text/html, Size: 2457 bytes --]

[-- Attachment #2: Type: text/plain, Size: 154 bytes --]

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

RE: [PATCH 12/15] drm/amd/display: Use hw lock mgr

[Wood, Wyatt]

[-- Attachment #1.1: Type: text/plain, Size: 2157 bytes --]

The condition in which it’s enabled is hidden from upstream.

From: Brandon Wright <bearoso@gmail.com>
Sent: Saturday, August 8, 2020 10:11 PM
To: Brol, Eryk <Eryk.Brol@amd.com>; amd-gfx@lists.freedesktop.org; Wood, Wyatt <Wyatt.Wood@amd.com>
Subject: Re: [PATCH 12/15] drm/amd/display: Use hw lock mgr

Just curious, but I noticed this new lock manager isn't being used because of the definitive false return value in display/dc/dce/dmub_hw_lock_mgr.c:should_use_dmub_lock.
Was this supposed to be enabled?


From: Wyatt Wood <wyatt.w...@amd.com<mailto:wyatt.w...@amd.com>>



[Why]

Feature requires synchronization of dig, pipe, and cursor locking

between driver and fw.

[How]

Set flag to force psr to use hw lock mgr.



Signed-off-by: Wyatt Wood <wyatt.w...@amd.com<mailto:wyatt.w...@amd.com>>

Reviewed-by: Anthony Koo <anthony....@amd.com<mailto:anthony....@amd.com>>

Acked-by: Eryk Brol <eryk.b...@amd.com<mailto:eryk.b...@amd.com>>

---

 drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c | 2 +-

 1 file changed, 1 insertion(+), 1 deletion(-)



diff --git a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c

b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c

index 82e67bd81f2d..5167d6b8a48d 100644

--- a/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c

+++ b/drivers/gpu/drm/amd/display/dc/dce/dmub_psr.c

@@ -233,8 +233,8 @@ static bool dmub_psr_copy_settings(struct dmub_psr *dmub,

        copy_settings_data->frame_cap_ind                       =

psr_context->psrFrameCaptureIndicationReq;

        copy_settings_data->debug.bitfields.visual_confirm      =

dc->dc->debug.visual_confirm == VISUAL_CONFIRM_PSR ?

                                                                        true :

false;

+       copy_settings_data->debug.bitfields.use_hw_lock_mgr     = 1;

        copy_settings_data->init_sdp_deadline                   =

psr_context->sdpTransmitLineNumDeadline;

-       copy_settings_data->debug.bitfields.use_hw_lock_mgr     = 0;



        dc_dmub_srv_cmd_queue(dc->dmub_srv, &cmd);

        dc_dmub_srv_cmd_execute(dc->dmub_srv);

--

2.25.1

[-- Attachment #1.2: Type: text/html, Size: 8992 bytes --]

[-- Attachment #2: Type: text/plain, Size: 154 bytes --]

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