[PATCH 00/19] DC Patches September 22, 2023

[PATCH 00/19] DC Patches September 22, 2023

[Wayne Lin]

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

- fix incorrect odm change detection logic
- add primary pipe check when building slice table for dcn3x
- add get primary dpp pipe resource interface
- add missing function pointer for DCN321 resource
- Rename DisableMinDispClkODM in dc_config
- Break after finding supported vlevel for repopulate
- Update OPP counter from new interface
- augment display clock in dc_cap structure
- Improve code style on bios_parser2
- skip audio config for virtual signal
- block MPO if it prevents pstate support
- Fix DP2.0 timing sync
- Improve x86 and dmub ips handshake
- remove guaranteed viewports limitation for odm
- add new windowed mpo odm minimal transition sequence
- reset stream slice count for new ODM policy
- determine fast update only before commit minimal transition state
- remove unused mmhub_reg_offsets

Cc: Daniel Wheeler <daniel.wheeler@amd.com>
---
Alex Hung (1):
  drm/amd/display: skip audio config for virtual signal

Alvin Lee (1):
  drm/amd/display: Break after finding supported vlevel for repopulate

Aric Cyr (1):
  drm/amd/display: 3.2.253

Dmytro Laktyushkin (1):
  drm/amd/display: block MPO if it prevents pstate support

Duncan Ma (1):
  drm/amd/display: Improve x86 and dmub ips handshake

Gabe Teeger (1):
  drm/amd/display: Rename DisableMinDispClkODM in dc_config

Hong-lu Cheng (1):
  drm/amd/display: Update OPP counter from new interface

Ilya Bakoulin (1):
  drm/amd/display: Fix DP2.0 timing sync

Max Tseng (1):
  drm/amd/display: augment display clock in dc_cap structure

Qingqing Zhuo (1):
  drm/amd/display: remove unused mmhub_reg_offsets

Rodrigo Siqueira (1):
  drm/amd/display: Improve code style on bios_parser2

Wenjing Liu (8):
  drm/amd/display: determine fast update only before commit minimal
    transition state
  drm/amd/display: reset stream slice count for new ODM policy
  drm/amd/display: add new windowed mpo odm minimal transition sequence
  drm/amd/display: remove guaranteed viewports limitation for odm
  drm/amd/display: add missing function pointer for DCN321 resource
  drm/amd/display: add get primary dpp pipe resource interface
  drm/amd/display: add primary pipe check when building slice table for
    dcn3x
  drm/amd/display: fix incorrect odm change detection logic

 .../drm/amd/display/dc/bios/bios_parser2.c    |  37 +--
 .../display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c  |  37 +++
 .../amd/display/dc/clk_mgr/dcn35/dcn35_smu.c  |  14 +-
 .../amd/display/dc/clk_mgr/dcn35/dcn35_smu.h  |   4 +-
 drivers/gpu/drm/amd/display/dc/core/dc.c      | 311 ++++++++++++------
 .../gpu/drm/amd/display/dc/core/dc_resource.c |  46 +++
 drivers/gpu/drm/amd/display/dc/dc.h           |  13 +-
 drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c  |  57 +++-
 .../gpu/drm/amd/display/dc/dce/dce_audio.c    |   4 +
 .../drm/amd/display/dc/dcn20/dcn20_hwseq.c    |  89 ++---
 .../drm/amd/display/dc/dcn314/dcn314_dccg.c   |   1 +
 .../amd/display/dc/dcn314/dcn314_resource.c   |   2 +
 .../amd/display/dc/dcn315/dcn315_resource.c   |   4 +-
 .../gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c |   1 +
 .../amd/display/dc/dcn321/dcn321_resource.c   |   1 +
 .../gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c |   1 +
 .../drm/amd/display/dc/dcn35/dcn35_hwseq.c    |  30 +-
 .../drm/amd/display/dc/dcn35/dcn35_hwseq.h    |   3 +
 .../gpu/drm/amd/display/dc/dcn35/dcn35_init.c |   4 +-
 .../drm/amd/display/dc/dcn35/dcn35_resource.c |  14 +-
 .../drm/amd/display/dc/dml/dcn32/dcn32_fpu.c  |  44 +--
 .../gpu/drm/amd/display/dc/inc/core_types.h   |  11 +
 .../gpu/drm/amd/display/dc/inc/hw/clk_mgr.h   |   2 +
 drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h  |   4 +
 .../gpu/drm/amd/display/dc/inc/hw_sequencer.h |   3 +-
 drivers/gpu/drm/amd/display/dc/inc/resource.h |  16 +
 26 files changed, 527 insertions(+), 226 deletions(-)

-- 
2.37.3

[PATCH 01/19] drm/amd/display: remove unused mmhub_reg_offsets

[Wayne Lin]

From: Qingqing Zhuo <qingqing.zhuo@amd.com>

[Why & How]
mmhub_reg_offsets never used in code. Remove it.

Reviewed-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Qingqing Zhuo <qingqing.zhuo@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c | 10 ----------
 1 file changed, 10 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c
index 961af9e7d5c7..06fd02084d7c 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c
@@ -205,16 +205,6 @@ enum dcn35_clk_src_array_id {
 	REG_STRUCT[id].reg_name = NBIO_BASE(regBIF_BX2_ ## reg_name ## _BASE_IDX) + \
 		regBIF_BX2_ ## reg_name
 
-/* MMHUB */
-#define MMHUB_BASE_INNER(seg) ctx->mmhub_reg_offsets[seg]
-
-#define MMHUB_BASE(seg) \
-	MMHUB_BASE_INNER(seg)
-
-#define MMHUB_SR(reg_name)\
-		REG_STRUCT.reg_name = MMHUB_BASE(reg ## reg_name ## _BASE_IDX) + \
-					reg ## reg_name
-
 #define bios_regs_init() \
 		( \
 		NBIO_SR(BIOS_SCRATCH_3),\
-- 
2.37.3

[PATCH 02/19] drm/amd/display: determine fast update only before commit minimal transition state

[Wayne Lin]

From: Wenjing Liu <wenjing.liu@amd.com>

[why]
commit minimal transition state would update current
state to new state with surface and stream update
applied. If we determine fast update only after we
have committed the minimal transition state based on
new state, we will skip committing the full new state.

[how]
determine fast update only earlier based on the
actaul currents state. Only skip full commit when
the transition between actual current state and new
state is fast update only.

Reviewed-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c | 6 ++++--
 1 file changed, 4 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 f91d0f6b0d7d..195ea6286c33 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -4307,8 +4307,11 @@ bool dc_update_planes_and_stream(struct dc *dc,
 	 */
 	bool force_minimal_pipe_splitting = 0;
 	bool is_plane_addition = 0;
+	bool is_fast_update_only;
 
 	populate_fast_updates(fast_update, srf_updates, surface_count, stream_update);
+	is_fast_update_only = fast_update_only(dc, fast_update, srf_updates,
+			surface_count, stream_update, stream);
 	force_minimal_pipe_splitting = could_mpcc_tree_change_for_active_pipes(
 			dc,
 			stream,
@@ -4360,8 +4363,7 @@ bool dc_update_planes_and_stream(struct dc *dc,
 	}
 
 	update_seamless_boot_flags(dc, context, surface_count, stream);
-	if (fast_update_only(dc, fast_update, srf_updates, surface_count, stream_update, stream) &&
-			!dc->debug.enable_legacy_fast_update) {
+	if (is_fast_update_only && !dc->debug.enable_legacy_fast_update) {
 		commit_planes_for_stream_fast(dc,
 				srf_updates,
 				surface_count,
-- 
2.37.3

[PATCH 03/19] drm/amd/display: reset stream slice count for new ODM policy

[Wayne Lin]

From: Wenjing Liu <wenjing.liu@amd.com>

[why]
ODM combine could prevent us from supporting more planes
we will reset ODM slice count back to 1 when all planes have
been removed to maximize the amount of planes supported when
new planes are added.

[how]
reset ODM slice count when all planes are removed.

Reviewed-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 5810cf78cf29..436892450936 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -2874,6 +2874,15 @@ bool dc_remove_plane_from_context(
 
 	stream_status->plane_states[stream_status->plane_count] = NULL;
 
+	if (stream_status->plane_count == 0 && dc->config.enable_windowed_mpo_odm)
+		/* ODM combine could prevent us from supporting more planes
+		 * we will reset ODM slice count back to 1 when all planes have
+		 * been removed to maximize the amount of planes supported when
+		 * new planes are added.
+		 */
+		resource_update_pipes_for_stream_with_slice_count(
+				context, dc->current_state, dc->res_pool, stream, 1);
+
 	return true;
 }
 
-- 
2.37.3

[PATCH 04/19] drm/amd/display: add new windowed mpo odm minimal transition sequence

[Wayne Lin]

From: Wenjing Liu <wenjing.liu@amd.com>

[why]
We do not have a programming sequence to support seamless transition
between MPC combine to ODM combine when plane count remains the same.
This can happen with windowed MPO ODM use cases when plane scaling
is changed over ODM Combine's capability.

[How]
We are adding a programming sequence specifically to handle the transition
between MPC combine and ODM combine during plane scaling change.

Reviewed-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c      | 252 ++++++++++++++----
 .../gpu/drm/amd/display/dc/inc/core_types.h   |  11 +
 2 files changed, 207 insertions(+), 56 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 195ea6286c33..e2dcb836a0f0 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -2965,6 +2965,34 @@ static void copy_stream_update_to_stream(struct dc *dc,
 	}
 }
 
+static void backup_plane_states_for_stream(
+		struct dc_plane_state plane_states[MAX_SURFACE_NUM],
+		struct dc_stream_state *stream)
+{
+	int i;
+	struct dc_stream_status *status = dc_stream_get_status(stream);
+
+	if (!status)
+		return;
+
+	for (i = 0; i < status->plane_count; i++)
+		plane_states[i] = *status->plane_states[i];
+}
+
+static void restore_plane_states_for_stream(
+		struct dc_plane_state plane_states[MAX_SURFACE_NUM],
+		struct dc_stream_state *stream)
+{
+	int i;
+	struct dc_stream_status *status = dc_stream_get_status(stream);
+
+	if (!status)
+		return;
+
+	for (i = 0; i < status->plane_count; i++)
+		*status->plane_states[i] = plane_states[i];
+}
+
 static bool update_planes_and_stream_state(struct dc *dc,
 		struct dc_surface_update *srf_updates, int surface_count,
 		struct dc_stream_state *stream,
@@ -2988,7 +3016,7 @@ static bool update_planes_and_stream_state(struct dc *dc,
 	}
 
 	context = dc->current_state;
-
+	backup_plane_states_for_stream(dc->current_state->scratch.plane_states, stream);
 	update_type = dc_check_update_surfaces_for_stream(
 			dc, srf_updates, surface_count, stream_update, stream_status);
 
@@ -3095,6 +3123,7 @@ static bool update_planes_and_stream_state(struct dc *dc,
 
 	*new_context = context;
 	*new_update_type = update_type;
+	backup_plane_states_for_stream(context->scratch.plane_states, stream);
 
 	return true;
 
@@ -3986,6 +4015,107 @@ static bool could_mpcc_tree_change_for_active_pipes(struct dc *dc,
 	return force_minimal_pipe_splitting;
 }
 
+struct pipe_split_policy_backup {
+	bool dynamic_odm_policy;
+	bool subvp_policy;
+	enum pipe_split_policy mpc_policy;
+};
+
+static void release_minimal_transition_state(struct dc *dc,
+		struct dc_state *context, struct pipe_split_policy_backup *policy)
+{
+	dc_release_state(context);
+	/* restore previous pipe split and odm policy */
+	if (!dc->config.is_vmin_only_asic)
+		dc->debug.pipe_split_policy = policy->mpc_policy;
+	dc->debug.enable_single_display_2to1_odm_policy = policy->dynamic_odm_policy;
+	dc->debug.force_disable_subvp = policy->subvp_policy;
+}
+
+static struct dc_state *create_minimal_transition_state(struct dc *dc,
+		struct dc_state *base_context, struct pipe_split_policy_backup *policy)
+{
+	struct dc_state *minimal_transition_context = dc_create_state(dc);
+	unsigned int i, j;
+
+	if (!dc->config.is_vmin_only_asic) {
+		policy->mpc_policy = dc->debug.pipe_split_policy;
+		dc->debug.pipe_split_policy = MPC_SPLIT_AVOID;
+	}
+	policy->dynamic_odm_policy = dc->debug.enable_single_display_2to1_odm_policy;
+	dc->debug.enable_single_display_2to1_odm_policy = false;
+	policy->subvp_policy = dc->debug.force_disable_subvp;
+	dc->debug.force_disable_subvp = true;
+
+	dc_resource_state_copy_construct(base_context, minimal_transition_context);
+
+	/* commit minimal state */
+	if (dc->res_pool->funcs->validate_bandwidth(dc, minimal_transition_context, false)) {
+		for (i = 0; i < minimal_transition_context->stream_count; i++) {
+			struct dc_stream_status *stream_status = &minimal_transition_context->stream_status[i];
+
+			for (j = 0; j < stream_status->plane_count; j++) {
+				struct dc_plane_state *plane_state = stream_status->plane_states[j];
+
+				/* force vsync flip when reconfiguring pipes to prevent underflow
+				 * and corruption
+				 */
+				plane_state->flip_immediate = false;
+			}
+		}
+	} else {
+		/* this should never happen */
+		release_minimal_transition_state(dc, minimal_transition_context, policy);
+		BREAK_TO_DEBUGGER();
+		minimal_transition_context = NULL;
+	}
+	return minimal_transition_context;
+}
+
+static bool commit_minimal_transition_state_for_windowed_mpo_odm(struct dc *dc,
+		struct dc_state *context,
+		struct dc_stream_state *stream)
+{
+	bool success = false;
+	struct dc_state *minimal_transition_context;
+	struct pipe_split_policy_backup policy;
+
+	/* commit based on new context */
+	minimal_transition_context = create_minimal_transition_state(dc,
+			context, &policy);
+	if (minimal_transition_context) {
+		if (dc->hwss.is_pipe_topology_transition_seamless(
+					dc, dc->current_state, minimal_transition_context) &&
+			dc->hwss.is_pipe_topology_transition_seamless(
+					dc, minimal_transition_context, context)) {
+			DC_LOG_DC("%s base = new state\n", __func__);
+			success = dc_commit_state_no_check(dc, minimal_transition_context) == DC_OK;
+		}
+		release_minimal_transition_state(dc, minimal_transition_context, &policy);
+	}
+
+	if (!success) {
+		/* commit based on current context */
+		restore_plane_states_for_stream(dc->current_state->scratch.plane_states, stream);
+		minimal_transition_context = create_minimal_transition_state(dc,
+				dc->current_state, &policy);
+		if (minimal_transition_context) {
+			if (dc->hwss.is_pipe_topology_transition_seamless(
+					dc, dc->current_state, minimal_transition_context) &&
+				dc->hwss.is_pipe_topology_transition_seamless(
+						dc, minimal_transition_context, context)) {
+				DC_LOG_DC("%s base = current state\n", __func__);
+				success = dc_commit_state_no_check(dc, minimal_transition_context) == DC_OK;
+			}
+			release_minimal_transition_state(dc, minimal_transition_context, &policy);
+		}
+		restore_plane_states_for_stream(context->scratch.plane_states, stream);
+	}
+
+	ASSERT(success);
+	return success;
+}
+
 /**
  * commit_minimal_transition_state - Create a transition pipe split state
  *
@@ -4007,23 +4137,14 @@ static bool could_mpcc_tree_change_for_active_pipes(struct dc *dc,
 static bool commit_minimal_transition_state(struct dc *dc,
 		struct dc_state *transition_base_context)
 {
-	struct dc_state *transition_context = dc_create_state(dc);
-	enum pipe_split_policy tmp_mpc_policy = 0;
-	bool temp_dynamic_odm_policy = 0;
-	bool temp_subvp_policy = 0;
+	struct dc_state *transition_context;
+	struct pipe_split_policy_backup policy;
 	enum dc_status ret = DC_ERROR_UNEXPECTED;
 	unsigned int i, j;
 	unsigned int pipe_in_use = 0;
 	bool subvp_in_use = false;
 	bool odm_in_use = false;
 
-	if (!transition_context)
-		return false;
-	/* Setup:
-	 * Store the current ODM and MPC config in some temp variables to be
-	 * restored after we commit the transition state.
-	 */
-
 	/* check current pipes in use*/
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
 		struct pipe_ctx *pipe = &transition_base_context->res_ctx.pipe_ctx[i];
@@ -4064,10 +4185,8 @@ static bool commit_minimal_transition_state(struct dc *dc,
 	 * Reduce the scenarios to use dc_commit_state_no_check in the stage of flip. Especially
 	 * enter/exit MPO when DCN still have enough resources.
 	 */
-	if (pipe_in_use != dc->res_pool->pipe_count && !subvp_in_use && !odm_in_use) {
-		dc_release_state(transition_context);
+	if (pipe_in_use != dc->res_pool->pipe_count && !subvp_in_use && !odm_in_use)
 		return true;
-	}
 
 	DC_LOG_DC("%s base = %s state, reason = %s\n", __func__,
 			dc->current_state == transition_base_context ? "current" : "new",
@@ -4076,49 +4195,13 @@ static bool commit_minimal_transition_state(struct dc *dc,
 			dc->debug.pipe_split_policy != MPC_SPLIT_AVOID ? "MPC in Use" :
 			"Unknown");
 
-	if (!dc->config.is_vmin_only_asic) {
-		tmp_mpc_policy = dc->debug.pipe_split_policy;
-		dc->debug.pipe_split_policy = MPC_SPLIT_AVOID;
-	}
-
-	temp_dynamic_odm_policy = dc->debug.enable_single_display_2to1_odm_policy;
-	dc->debug.enable_single_display_2to1_odm_policy = false;
-
-	temp_subvp_policy = dc->debug.force_disable_subvp;
-	dc->debug.force_disable_subvp = true;
-
-	dc_resource_state_copy_construct(transition_base_context, transition_context);
-
-	/* commit minimal state */
-	if (dc->res_pool->funcs->validate_bandwidth(dc, transition_context, false)) {
-		for (i = 0; i < transition_context->stream_count; i++) {
-			struct dc_stream_status *stream_status = &transition_context->stream_status[i];
-
-			for (j = 0; j < stream_status->plane_count; j++) {
-				struct dc_plane_state *plane_state = stream_status->plane_states[j];
-
-				/* force vsync flip when reconfiguring pipes to prevent underflow
-				 * and corruption
-				 */
-				plane_state->flip_immediate = false;
-			}
-		}
-
+	transition_context = create_minimal_transition_state(dc,
+			transition_base_context, &policy);
+	if (transition_context) {
 		ret = dc_commit_state_no_check(dc, transition_context);
+		release_minimal_transition_state(dc, transition_context, &policy);
 	}
 
-	/* always release as dc_commit_state_no_check retains in good case */
-	dc_release_state(transition_context);
-
-	/* TearDown:
-	 * Restore original configuration for ODM and MPO.
-	 */
-	if (!dc->config.is_vmin_only_asic)
-		dc->debug.pipe_split_policy = tmp_mpc_policy;
-
-	dc->debug.enable_single_display_2to1_odm_policy = temp_dynamic_odm_policy;
-	dc->debug.force_disable_subvp = temp_subvp_policy;
-
 	if (ret != DC_OK) {
 		/* this should never happen */
 		BREAK_TO_DEBUGGER();
@@ -4290,6 +4373,51 @@ static bool fast_update_only(struct dc *dc,
 			&& !full_update_required(dc, srf_updates, surface_count, stream_update, stream);
 }
 
+static bool should_commit_minimal_transition_for_windowed_mpo_odm(struct dc *dc,
+		struct dc_stream_state *stream,
+		struct dc_state *context)
+{
+	struct pipe_ctx *cur_pipe, *new_pipe;
+	bool cur_is_odm_in_use, new_is_odm_in_use;
+	struct dc_stream_status *cur_stream_status = stream_get_status(dc->current_state, stream);
+	struct dc_stream_status *new_stream_status = stream_get_status(context, stream);
+
+	if (!dc->debug.enable_single_display_2to1_odm_policy ||
+			!dc->config.enable_windowed_mpo_odm)
+		/* skip the check if windowed MPO ODM or dynamic ODM is turned
+		 * off.
+		 */
+		return false;
+
+	if (context == dc->current_state)
+		/* skip the check for fast update */
+		return false;
+
+	if (new_stream_status->plane_count != cur_stream_status->plane_count)
+		/* plane count changed, not a plane scaling update so not the
+		 * case we are looking for
+		 */
+		return false;
+
+	cur_pipe = resource_get_otg_master_for_stream(&dc->current_state->res_ctx, stream);
+	new_pipe = resource_get_otg_master_for_stream(&context->res_ctx, stream);
+	cur_is_odm_in_use = resource_get_odm_slice_count(cur_pipe) > 1;
+	new_is_odm_in_use = resource_get_odm_slice_count(new_pipe) > 1;
+	if (cur_is_odm_in_use == new_is_odm_in_use)
+		/* ODM state isn't changed, not the case we are looking for */
+		return false;
+
+	if (dc->hwss.is_pipe_topology_transition_seamless &&
+			dc->hwss.is_pipe_topology_transition_seamless(
+					dc, dc->current_state, context))
+		/* transition can be achieved without the need for committing
+		 * minimal transition state first
+		 */
+		return false;
+
+	return true;
+}
+
 bool dc_update_planes_and_stream(struct dc *dc,
 		struct dc_surface_update *srf_updates, int surface_count,
 		struct dc_stream_state *stream,
@@ -4362,6 +4490,19 @@ bool dc_update_planes_and_stream(struct dc *dc,
 		update_type = UPDATE_TYPE_FULL;
 	}
 
+	/* when windowed MPO ODM is supported, we need to handle a special case
+	 * where we can transition between ODM combine and MPC combine due to
+	 * plane scaling update. This transition will require us to commit
+	 * minimal transition state. The condition to trigger this update can't
+	 * be predicted by could_mpcc_tree_change_for_active_pipes because we
+	 * can only determine it after DML validation. Therefore we can't rely
+	 * on the existing commit minimal transition state sequence. Instead
+	 * we have to add additional handling here to handle this transition
+	 * with its own special sequence.
+	 */
+	if (should_commit_minimal_transition_for_windowed_mpo_odm(dc, stream, context))
+		commit_minimal_transition_state_for_windowed_mpo_odm(dc,
+				context, stream);
 	update_seamless_boot_flags(dc, context, surface_count, stream);
 	if (is_fast_update_only && !dc->debug.enable_legacy_fast_update) {
 		commit_planes_for_stream_fast(dc,
@@ -4376,7 +4517,6 @@ bool dc_update_planes_and_stream(struct dc *dc,
 				dc->hwss.is_pipe_topology_transition_seamless &&
 				!dc->hwss.is_pipe_topology_transition_seamless(
 						dc, dc->current_state, context)) {
-
 			DC_LOG_ERROR("performing non-seamless pipe topology transition with surface only update!\n");
 			BREAK_TO_DEBUGGER();
 		}
diff --git a/drivers/gpu/drm/amd/display/dc/inc/core_types.h b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
index 9f41e8d15065..c095ccbb80c4 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/core_types.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/core_types.h
@@ -563,6 +563,17 @@ struct dc_state {
 	struct {
 		unsigned int stutter_period_us;
 	} perf_params;
+
+	struct {
+		/* used to temporarily backup plane states of a stream during
+		 * dc update. The reason is that plane states are overwritten
+		 * with surface updates in dc update. Once they are overwritten
+		 * current state is no longer valid. We want to temporarily
+		 * store current value in plane states so we can still recover
+		 * a valid current state during dc update.
+		 */
+		struct dc_plane_state plane_states[MAX_SURFACE_NUM];
+	} scratch;
 };
 
 struct replay_context {
-- 
2.37.3

[PATCH 05/19] drm/amd/display: remove guaranteed viewports limitation for odm

[Wayne Lin]

From: Wenjing Liu <wenjing.liu@amd.com>

[why]
With the more generic hw minimal state transition sequence,
this limitation has been overcome.

Reviewed-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c      | 34 -------------------
 .../drm/amd/display/dc/dml/dcn32/dcn32_fpu.c  | 27 ---------------
 2 files changed, 61 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index e2dcb836a0f0..a2360dfdc83b 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -3948,40 +3948,6 @@ static bool could_mpcc_tree_change_for_active_pipes(struct dc *dc,
 				*is_plane_addition = true;
 			}
 		}
-		if (dc->config.enable_windowed_mpo_odm) {
-			const struct rect *guaranteed_viewport = &stream->src;
-			const struct rect *surface_src, *surface_dst;
-			bool are_cur_planes_guaranteed = true;
-			bool are_new_planes_guaranteed = true;
-
-			for (i = 0; i < cur_stream_status->plane_count; i++) {
-				surface_src = &cur_stream_status->plane_states[i]->src_rect;
-				surface_dst = &cur_stream_status->plane_states[i]->dst_rect;
-				if ((surface_src->height > surface_dst->height && surface_src->height > guaranteed_viewport->height) ||
-						(surface_src->width > surface_dst->width && surface_src->width > guaranteed_viewport->width))
-					are_cur_planes_guaranteed = false;
-			}
-
-			for (i = 0; i < surface_count; i++) {
-				if (srf_updates[i].scaling_info) {
-					surface_src = &srf_updates[i].scaling_info->src_rect;
-					surface_dst = &srf_updates[i].scaling_info->dst_rect;
-				} else {
-					surface_src = &srf_updates[i].surface->src_rect;
-					surface_dst = &srf_updates[i].surface->dst_rect;
-				}
-				if ((surface_src->height > surface_dst->height && surface_src->height > guaranteed_viewport->height) ||
-						(surface_src->width > surface_dst->width && surface_src->width > guaranteed_viewport->width))
-					are_new_planes_guaranteed = false;
-			}
-
-			if (are_cur_planes_guaranteed && !are_new_planes_guaranteed) {
-				force_minimal_pipe_splitting = true;
-				*is_plane_addition = true;
-			} else if (!are_cur_planes_guaranteed && are_new_planes_guaranteed) {
-				force_minimal_pipe_splitting = true;
-			}
-		}
 	}
 
 	for (i = 0; i < dc->res_pool->pipe_count; i++) {
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
index 1f53883d8f56..dcbd38bb3ed1 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
@@ -1267,8 +1267,6 @@ static bool should_allow_odm_power_optimization(struct dc *dc,
 {
 	struct dc_stream_state *stream = context->streams[0];
 	struct pipe_slice_table slice_table;
-	struct dc_plane_state *plane;
-	struct rect guaranteed_viewport;
 	int i;
 
 	/*
@@ -1333,31 +1331,6 @@ static bool should_allow_odm_power_optimization(struct dc *dc,
 		for (i = 0; i < slice_table.odm_combine_count; i++)
 			if (slice_table.odm_combines[i].slice_count > 1)
 				return false;
-
-		/* up to here we know that a plane with viewport equal to stream
-		 * src can be validated with single DPP pipe. Therefore any
-		 * planes with smaller or equal viewport is guaranteed to work
-		 * regardless of its position and scaling ratio. Also we know
-		 * any plane without downscale ratio greater than 1 should also
-		 * work. Up until DCN3x we still have software limitation that
-		 * doesn't implement a smooth transition between ODM combine and
-		 * MPC combine during plane resizing when we are crossing ODM
-		 * capability boundary. So we are adding this guaranteed
-		 * viewport condition to limit ODM power optimization support
-		 * for only the planes within the guaranteed viewport size. Such
-		 * planes can be supported with ODM power optimization without
-		 * ever the need to transition to MPC combine in any scaling
-		 * ratios and positions. Therefore we cover the software
-		 * limitation of this transition sequence.
-		 */
-		guaranteed_viewport = stream->src;
-		for (i = 0; i < context->stream_status[0].plane_count; i++) {
-			plane = context->stream_status[0].plane_states[i];
-
-			if ((plane->src_rect.height > plane->dst_rect.height && plane->src_rect.height > guaranteed_viewport.height) ||
-					(plane->src_rect.width > plane->dst_rect.width && plane->src_rect.width > guaranteed_viewport.width))
-				return false;
-		}
 	} else {
 		/*
 		 * the new ODM power optimization feature reduces software
-- 
2.37.3

[PATCH 06/19] drm/amd/display: Improve x86 and dmub ips handshake

[Wayne Lin]

From: Duncan Ma <duncan.ma@amd.com>

[Why]
There is a race condition between x86 and dmcub fw when attempting to exit
IPS2. Scenarios including exiting IPS2 before IPS2 has been entered. This
can cause unexpected hang when DMCUB attempt to exit while PMFW still
tries to enter IPS2.

[How]
A new design has been introduced to remove race conditions and improve the
handshake between x86 and DMCUB. An AON scratch register is borrowed from
PMFW to determine whether DMCUB has committed to IPS entry or not.

In the case when dmcub has committed IPS entry, x86 must poll until an exit
event occurred either from DMCUB(IPS1) or PMFW(IPS2). x86 will wait
upperbound of evaluation and IPS entry time to ensure IPS2 exit event has
been sent to PMFW.

Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Duncan Ma <duncan.ma@amd.com>
---
 .../display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c  | 37 ++++++++++++
 .../amd/display/dc/clk_mgr/dcn35/dcn35_smu.c  | 14 ++++-
 .../amd/display/dc/clk_mgr/dcn35/dcn35_smu.h  |  4 +-
 drivers/gpu/drm/amd/display/dc/dc.h           |  2 +
 drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c  | 57 +++++++++++++++----
 .../drm/amd/display/dc/dcn35/dcn35_hwseq.c    | 30 ++++++----
 .../drm/amd/display/dc/dcn35/dcn35_hwseq.h    |  3 +
 .../gpu/drm/amd/display/dc/dcn35/dcn35_init.c |  4 +-
 .../drm/amd/display/dc/dcn35/dcn35_resource.c |  2 +
 .../gpu/drm/amd/display/dc/inc/hw/clk_mgr.h   |  2 +
 .../gpu/drm/amd/display/dc/inc/hw_sequencer.h |  3 +-
 11 files changed, 130 insertions(+), 28 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c
index 819d273f011d..c9c7baf45be3 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_clk_mgr.c
@@ -738,6 +738,34 @@ static void dcn35_set_low_power_state(struct clk_mgr *clk_mgr_base)
 	}
 }
 
+static void dcn35_set_idle_state(struct clk_mgr *clk_mgr_base, bool allow_idle)
+{
+	struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
+	struct dc *dc = clk_mgr_base->ctx->dc;
+	uint32_t val = dcn35_smu_read_ips_scratch(clk_mgr);
+
+	if (dc->debug.disable_ips == 0) {
+		val |= DMUB_IPS1_ALLOW_MASK;
+		val |= DMUB_IPS2_ALLOW_MASK;
+	} else if (dc->debug.disable_ips == DMUB_IPS_DISABLE_IPS1) {
+		val = val & ~DMUB_IPS1_ALLOW_MASK;
+		val = val & ~DMUB_IPS2_ALLOW_MASK;
+	} else if (dc->debug.disable_ips == DMUB_IPS_DISABLE_IPS2) {
+		val |= DMUB_IPS1_ALLOW_MASK;
+		val = val & ~DMUB_IPS2_ALLOW_MASK;
+	} else if (dc->debug.disable_ips == DMUB_IPS_DISABLE_IPS2_Z10) {
+		val |= DMUB_IPS1_ALLOW_MASK;
+		val |= DMUB_IPS2_ALLOW_MASK;
+	}
+
+	if (!allow_idle) {
+		val = val & ~DMUB_IPS1_ALLOW_MASK;
+		val = val & ~DMUB_IPS2_ALLOW_MASK;
+	}
+
+	dcn35_smu_write_ips_scratch(clk_mgr, val);
+}
+
 static void dcn35_exit_low_power_state(struct clk_mgr *clk_mgr_base)
 {
 	struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
@@ -757,6 +785,13 @@ static bool dcn35_is_ips_supported(struct clk_mgr *clk_mgr_base)
 	return ips_supported;
 }
 
+static uint32_t dcn35_get_idle_state(struct clk_mgr *clk_mgr_base)
+{
+	struct clk_mgr_internal *clk_mgr = TO_CLK_MGR_INTERNAL(clk_mgr_base);
+
+	return dcn35_smu_read_ips_scratch(clk_mgr);
+}
+
 static void dcn35_init_clocks_fpga(struct clk_mgr *clk_mgr)
 {
 	dcn35_init_clocks(clk_mgr);
@@ -844,6 +879,8 @@ static struct clk_mgr_funcs dcn35_funcs = {
 	.set_low_power_state = dcn35_set_low_power_state,
 	.exit_low_power_state = dcn35_exit_low_power_state,
 	.is_ips_supported = dcn35_is_ips_supported,
+	.set_idle_state = dcn35_set_idle_state,
+	.get_idle_state = dcn35_get_idle_state
 };
 
 struct clk_mgr_funcs dcn35_fpga_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c
index f42c0a727dc8..cf74e69cb2a1 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.c
@@ -444,9 +444,9 @@ void dcn35_vbios_smu_enable_48mhz_tmdp_refclk_pwrdwn(struct clk_mgr_internal *cl
 			enable);
 }
 
-void dcn35_smu_exit_low_power_state(struct clk_mgr_internal *clk_mgr)
+int dcn35_smu_exit_low_power_state(struct clk_mgr_internal *clk_mgr)
 {
-	dcn35_smu_send_msg_with_param(
+	return dcn35_smu_send_msg_with_param(
 		clk_mgr,
 		VBIOSSMC_MSG_DispPsrExit,
 		0);
@@ -459,3 +459,13 @@ int dcn35_smu_get_ips_supported(struct clk_mgr_internal *clk_mgr)
 			VBIOSSMC_MSG_QueryIPS2Support,
 			0);
 }
+
+void dcn35_smu_write_ips_scratch(struct clk_mgr_internal *clk_mgr, uint32_t param)
+{
+	REG_WRITE(MP1_SMN_C2PMSG_71, param);
+}
+
+uint32_t dcn35_smu_read_ips_scratch(struct clk_mgr_internal *clk_mgr)
+{
+	return REG_READ(MP1_SMN_C2PMSG_71);
+}
diff --git a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h
index 4d441d6db8d0..8fb65a49351b 100644
--- a/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h
+++ b/drivers/gpu/drm/amd/display/dc/clk_mgr/dcn35/dcn35_smu.h
@@ -174,8 +174,10 @@ void dcn35_smu_set_zstate_support(struct clk_mgr_internal *clk_mgr, enum dcn_zst
 void dcn35_smu_set_dtbclk(struct clk_mgr_internal *clk_mgr, bool enable);
 void dcn35_vbios_smu_enable_48mhz_tmdp_refclk_pwrdwn(struct clk_mgr_internal *clk_mgr, bool enable);
 
-void dcn35_smu_exit_low_power_state(struct clk_mgr_internal *clk_mgr);
+int dcn35_smu_exit_low_power_state(struct clk_mgr_internal *clk_mgr);
 int dcn35_smu_get_ips_supported(struct clk_mgr_internal *clk_mgr);
 int dcn35_smu_get_dtbclk(struct clk_mgr_internal *clk_mgr);
 int dcn35_smu_get_dprefclk(struct clk_mgr_internal *clk_mgr);
+void dcn35_smu_write_ips_scratch(struct clk_mgr_internal *clk_mgr, uint32_t param);
+uint32_t dcn35_smu_read_ips_scratch(struct clk_mgr_internal *clk_mgr);
 #endif /* DAL_DC_35_SMU_H_ */
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index fd5040966ba0..ed9b72ac369f 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -965,6 +965,8 @@ struct dc_debug_options {
 	bool replay_skip_crtc_disabled;
 	bool ignore_pg;/*do nothing, let pmfw control it*/
 	bool psp_disabled_wa;
+	unsigned int ips2_eval_delay_us;
+	unsigned int ips2_entry_delay_us;
 };
 
 struct gpu_info_soc_bounding_box_v1_0;
diff --git a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c
index 91b3a4610957..973d7938f47b 100644
--- a/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c
+++ b/drivers/gpu/drm/amd/display/dc/dc_dmub_srv.c
@@ -1100,31 +1100,64 @@ void dc_dmub_srv_notify_idle(const struct dc *dc, bool allow_idle)
 
 	cmd.idle_opt_notify_idle.cntl_data.driver_idle = allow_idle;
 
-	dm_execute_dmub_cmd(dc->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
+	if (allow_idle) {
+		if (dc->hwss.set_idle_state)
+			dc->hwss.set_idle_state(dc, true);
+	}
 
-	if (allow_idle)
-		udelay(500);
+	dm_execute_dmub_cmd(dc->ctx, &cmd, DM_DMUB_WAIT_TYPE_WAIT);
 }
 
 void dc_dmub_srv_exit_low_power_state(const struct dc *dc)
 {
+	uint32_t allow_state = 0;
+	uint32_t commit_state = 0;
+
 	if (dc->debug.dmcub_emulation)
 		return;
 
 	if (!dc->idle_optimizations_allowed)
 		return;
 
-	// Tell PMFW to exit low power state
-	if (dc->clk_mgr->funcs->exit_low_power_state)
-		dc->clk_mgr->funcs->exit_low_power_state(dc->clk_mgr);
+	if (dc->hwss.get_idle_state &&
+		dc->hwss.set_idle_state &&
+		dc->clk_mgr->funcs->exit_low_power_state) {
+
+		allow_state = dc->hwss.get_idle_state(dc);
+		dc->hwss.set_idle_state(dc, false);
+
+		if (allow_state & DMUB_IPS2_ALLOW_MASK) {
+			// Wait for evaluation time
+			udelay(dc->debug.ips2_eval_delay_us);
+			commit_state = dc->hwss.get_idle_state(dc);
+			if (commit_state & DMUB_IPS2_COMMIT_MASK) {
+				// Tell PMFW to exit low power state
+				dc->clk_mgr->funcs->exit_low_power_state(dc->clk_mgr);
+
+				// Wait for IPS2 entry upper bound
+				udelay(dc->debug.ips2_entry_delay_us);
+				dc->clk_mgr->funcs->exit_low_power_state(dc->clk_mgr);
 
-	// Wait for dmcub to load up
-	dc_dmub_srv_is_hw_pwr_up(dc->ctx->dmub_srv, true);
+				do {
+					commit_state = dc->hwss.get_idle_state(dc);
+				} while (commit_state & DMUB_IPS2_COMMIT_MASK);
 
-	// Notify dmcub disallow idle
-	dc_dmub_srv_notify_idle(dc, false);
+				if (!dc_dmub_srv_is_hw_pwr_up(dc->ctx->dmub_srv, true))
+					ASSERT(0);
+
+				return;
+			}
+		}
+
+		dc_dmub_srv_notify_idle(dc, false);
+		if (allow_state & DMUB_IPS1_ALLOW_MASK) {
+			do {
+				commit_state = dc->hwss.get_idle_state(dc);
+			} while (commit_state & DMUB_IPS1_COMMIT_MASK);
+		}
+	}
 
-	// Confirm dmu is powered up
-	dc_dmub_srv_is_hw_pwr_up(dc->ctx->dmub_srv, true);
+	if (!dc_dmub_srv_is_hw_pwr_up(dc->ctx->dmub_srv, true))
+		ASSERT(0);
 }
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hwseq.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hwseq.c
index 06960fada059..03ef7820139a 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hwseq.c
@@ -647,18 +647,10 @@ bool dcn35_apply_idle_power_optimizations(struct dc *dc, bool enable)
 
 	// TODO: review other cases when idle optimization is allowed
 
-	if (!enable) {
-		// Tell PMFW to exit low power state
-		if (dc->clk_mgr->funcs->exit_low_power_state)
-			dc->clk_mgr->funcs->exit_low_power_state(dc->clk_mgr);
-
-		dc_dmub_srv_is_hw_pwr_up(dc->ctx->dmub_srv, true);
-	}
-
-	dc_dmub_srv_notify_idle(dc, enable);
-
 	if (!enable)
-		dc_dmub_srv_is_hw_pwr_up(dc->ctx->dmub_srv, true);
+		dc_dmub_srv_exit_low_power_state(dc);
+	else
+		dc_dmub_srv_notify_idle(dc, enable);
 
 	return true;
 }
@@ -1192,3 +1184,19 @@ void dcn35_optimize_bandwidth(
 			dc->hwss.root_clock_control(dc, &pg_update_state, false);
 	}
 }
+
+void dcn35_set_idle_state(const struct dc *dc, bool allow_idle)
+{
+	// TODO: Find a more suitable communcation
+	if (dc->clk_mgr->funcs->set_idle_state)
+		dc->clk_mgr->funcs->set_idle_state(dc->clk_mgr, allow_idle);
+}
+
+uint32_t dcn35_get_idle_state(const struct dc *dc)
+{
+	// TODO: Find a more suitable communcation
+	if (dc->clk_mgr->funcs->get_idle_state)
+		return dc->clk_mgr->funcs->get_idle_state(dc->clk_mgr);
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hwseq.h b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hwseq.h
index 7c0ff7b163a9..14bbdb0fa634 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hwseq.h
+++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_hwseq.h
@@ -79,4 +79,7 @@ void dcn35_dsc_pg_control(
 		struct dce_hwseq *hws,
 		unsigned int dsc_inst,
 		bool power_on);
+
+void dcn35_set_idle_state(const struct dc *dc, bool allow_idle);
+uint32_t dcn35_get_idle_state(const struct dc *dc);
 #endif /* __DC_HWSS_DCN35_H__ */
diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_init.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_init.c
index d68efe5c64a4..a9553a87b959 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_init.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_init.c
@@ -117,7 +117,9 @@ static const struct hw_sequencer_funcs dcn35_funcs = {
 	.calc_blocks_to_gate = dcn35_calc_blocks_to_gate,
 	.calc_blocks_to_ungate = dcn35_calc_blocks_to_ungate,
 	.block_power_control = dcn35_block_power_control,
-	.root_clock_control = dcn35_root_clock_control
+	.root_clock_control = dcn35_root_clock_control,
+	.set_idle_state = dcn35_set_idle_state,
+	.get_idle_state = dcn35_get_idle_state
 };
 
 static const struct hwseq_private_funcs dcn35_private_funcs = {
diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c
index 06fd02084d7c..67ff19d4116d 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c
@@ -748,6 +748,8 @@ static const struct dc_debug_options debug_defaults_drv = {
 	.ignore_pg = true,
 	.psp_disabled_wa = true,
 	.disable_ips = true,
+	.ips2_eval_delay_us = 200,
+	.ips2_entry_delay_us = 400
 };
 
 static const struct dc_panel_config panel_config_defaults = {
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h
index cb2dc3f75ae2..fa9614bcb160 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/clk_mgr.h
@@ -262,6 +262,8 @@ struct clk_mgr_funcs {
 	void (*set_low_power_state)(struct clk_mgr *clk_mgr);
 	void (*exit_low_power_state)(struct clk_mgr *clk_mgr);
 	bool (*is_ips_supported)(struct clk_mgr *clk_mgr);
+	void (*set_idle_state)(struct clk_mgr *clk_mgr, bool allow_idle);
+	uint32_t (*get_idle_state)(struct clk_mgr *clk_mgr);
 
 	void (*init_clocks)(struct clk_mgr *clk_mgr);
 
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 e0dd56182841..d4d59a916668 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw_sequencer.h
@@ -422,7 +422,8 @@ struct hw_sequencer_funcs {
 		struct pg_block_update *update_state, bool power_on);
 	void (*root_clock_control)(struct dc *dc,
 		struct pg_block_update *update_state, bool power_on);
-
+	void (*set_idle_state)(const struct dc *dc, bool allow_idle);
+	uint32_t (*get_idle_state)(const struct dc *dc);
 	void (*commit_subvp_config)(struct dc *dc, struct dc_state *context);
 	void (*enable_phantom_streams)(struct dc *dc, struct dc_state *context);
 	void (*subvp_pipe_control_lock)(struct dc *dc,
-- 
2.37.3

[PATCH 07/19] drm/amd/display: Fix DP2.0 timing sync

[Wayne Lin]

From: Ilya Bakoulin <ilya.bakoulin@amd.com>

[Why]
Triggering OTG sync before all OTG/HPO clock programming is complete
causes timing sync to fail and a subsequent P-state hang.

[How]
Move DTB clock programming earlier in the sequence to
enable_stream_timing.

Reviewed-by: Ariel Bernstein <eric.bernstein@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Ilya Bakoulin <ilya.bakoulin@amd.com>
---
 .../drm/amd/display/dc/dcn20/dcn20_hwseq.c    | 81 +++++++++++--------
 .../drm/amd/display/dc/dcn314/dcn314_dccg.c   |  1 +
 .../gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c |  1 +
 .../gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c |  1 +
 drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h  |  4 +
 5 files changed, 56 insertions(+), 32 deletions(-)

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 19ab08f5122e..f2b20319f0bb 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
@@ -670,6 +670,37 @@ static int calc_mpc_flow_ctrl_cnt(const struct dc_stream_state *stream,
 	return flow_ctrl_cnt;
 }
 
+static enum phyd32clk_clock_source get_phyd32clk_src(struct dc_link *link)
+{
+	switch (link->link_enc->transmitter) {
+	case TRANSMITTER_UNIPHY_A:
+		return PHYD32CLKA;
+	case TRANSMITTER_UNIPHY_B:
+		return PHYD32CLKB;
+	case TRANSMITTER_UNIPHY_C:
+		return PHYD32CLKC;
+	case TRANSMITTER_UNIPHY_D:
+		return PHYD32CLKD;
+	case TRANSMITTER_UNIPHY_E:
+		return PHYD32CLKE;
+	default:
+		return PHYD32CLKA;
+	}
+}
+
+static int get_odm_segment_count(struct pipe_ctx *pipe_ctx)
+{
+	struct pipe_ctx *odm_pipe = pipe_ctx->next_odm_pipe;
+	int count = 1;
+
+	while (odm_pipe != NULL) {
+		count++;
+		odm_pipe = odm_pipe->next_odm_pipe;
+	}
+
+	return count;
+}
+
 enum dc_status dcn20_enable_stream_timing(
 		struct pipe_ctx *pipe_ctx,
 		struct dc_state *context,
@@ -817,6 +848,23 @@ enum dc_status dcn20_enable_stream_timing(
 		if (pipe_ctx->stream_res.tg && pipe_ctx->stream_res.tg->funcs->phantom_crtc_post_enable)
 			pipe_ctx->stream_res.tg->funcs->phantom_crtc_post_enable(pipe_ctx->stream_res.tg);
 	}
+
+	if (dc->link_srv->dp_is_128b_132b_signal(pipe_ctx)) {
+		struct dccg *dccg = dc->res_pool->dccg;
+		struct timing_generator *tg = pipe_ctx->stream_res.tg;
+		struct dtbclk_dto_params dto_params = {0};
+
+		if (dccg->funcs->set_dtbclk_p_src)
+			dccg->funcs->set_dtbclk_p_src(dccg, DTBCLK0, tg->inst);
+
+		dto_params.otg_inst = tg->inst;
+		dto_params.pixclk_khz = pipe_ctx->stream->timing.pix_clk_100hz / 10;
+		dto_params.num_odm_segments = get_odm_segment_count(pipe_ctx);
+		dto_params.timing = &pipe_ctx->stream->timing;
+		dto_params.ref_dtbclk_khz = dc->clk_mgr->funcs->get_dtb_ref_clk_frequency(dc->clk_mgr);
+		dccg->funcs->set_dtbclk_dto(dccg, &dto_params);
+	}
+
 	return DC_OK;
 }
 
@@ -2659,37 +2707,6 @@ void dcn20_update_mpcc(struct dc *dc, struct pipe_ctx *pipe_ctx)
 	hubp->mpcc_id = mpcc_id;
 }
 
-static enum phyd32clk_clock_source get_phyd32clk_src(struct dc_link *link)
-{
-	switch (link->link_enc->transmitter) {
-	case TRANSMITTER_UNIPHY_A:
-		return PHYD32CLKA;
-	case TRANSMITTER_UNIPHY_B:
-		return PHYD32CLKB;
-	case TRANSMITTER_UNIPHY_C:
-		return PHYD32CLKC;
-	case TRANSMITTER_UNIPHY_D:
-		return PHYD32CLKD;
-	case TRANSMITTER_UNIPHY_E:
-		return PHYD32CLKE;
-	default:
-		return PHYD32CLKA;
-	}
-}
-
-static int get_odm_segment_count(struct pipe_ctx *pipe_ctx)
-{
-	struct pipe_ctx *odm_pipe = pipe_ctx->next_odm_pipe;
-	int count = 1;
-
-	while (odm_pipe != NULL) {
-		count++;
-		odm_pipe = odm_pipe->next_odm_pipe;
-	}
-
-	return count;
-}
-
 void dcn20_enable_stream(struct pipe_ctx *pipe_ctx)
 {
 	enum dc_lane_count lane_count =
@@ -2730,7 +2747,7 @@ void dcn20_enable_stream(struct pipe_ctx *pipe_ctx)
 		dto_params.ref_dtbclk_khz = dc->clk_mgr->funcs->get_dtb_ref_clk_frequency(dc->clk_mgr);
 		dccg->funcs->set_dtbclk_dto(dccg, &dto_params);
 	} else {
-		}
+	}
 	if (hws->funcs.calculate_dccg_k1_k2_values && dc->res_pool->dccg->funcs->set_pixel_rate_div) {
 		hws->funcs.calculate_dccg_k1_k2_values(pipe_ctx, &k1_div, &k2_div);
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c
index 173837bbc941..17a1174b8d80 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_dccg.c
@@ -375,6 +375,7 @@ static const struct dccg_funcs dccg314_funcs = {
 	.set_pixel_rate_div = dccg314_set_pixel_rate_div,
 	.trigger_dio_fifo_resync = dccg314_trigger_dio_fifo_resync,
 	.set_valid_pixel_rate = dccg314_set_valid_pixel_rate,
+	.set_dtbclk_p_src = dccg314_set_dtbclk_p_src
 };
 
 struct dccg *dccg314_create(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c
index 921f58c0c729..036d05468d76 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn32/dcn32_dccg.c
@@ -345,6 +345,7 @@ static const struct dccg_funcs dccg32_funcs = {
 	.otg_drop_pixel = dccg32_otg_drop_pixel,
 	.set_pixel_rate_div = dccg32_set_pixel_rate_div,
 	.trigger_dio_fifo_resync = dccg32_trigger_dio_fifo_resync,
+	.set_dtbclk_p_src = dccg32_set_dtbclk_p_src,
 };
 
 struct dccg *dccg32_create(
diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c
index 22137fde62bd..addedcfd1238 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_dccg.c
@@ -767,6 +767,7 @@ static const struct dccg_funcs dccg35_funcs = {
 	.set_valid_pixel_rate = dccg35_set_valid_pixel_rate,
 	.enable_symclk_se = dccg35_enable_symclk_se,
 	.disable_symclk_se = dccg35_disable_symclk_se,
+	.set_dtbclk_p_src = dccg35_set_dtbclk_p_src,
 };
 
 struct dccg *dccg35_create(
diff --git a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h
index 65bb7cd05385..13f12f2a3f81 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/hw/dccg.h
@@ -192,6 +192,10 @@ struct dccg_funcs {
 	void (*set_dp_dto)(
 			struct dccg *dccg,
 			const struct dp_dto_params *params);
+	void (*set_dtbclk_p_src)(
+			struct dccg *dccg,
+			enum streamclk_source src,
+			uint32_t otg_inst);
 };
 
 #endif //__DAL_DCCG_H__
-- 
2.37.3

[PATCH 08/19] drm/amd/display: block MPO if it prevents pstate support

[Wayne Lin]

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

This change adds a method to block mpo when it would cause us to
exit pstate support. We should block mpo by default where MPO
causes issues by preventing pstate change.

Reviewed-by: Jun Lei <jun.lei@amd.com>
Reviewed-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Dmytro Laktyushkin <dmytro.laktyushkin@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c        | 17 +++++++++++++++++
 drivers/gpu/drm/amd/display/dc/dc.h             |  6 ++++++
 .../drm/amd/display/dc/dcn315/dcn315_resource.c |  4 ++--
 3 files changed, 25 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 a2360dfdc83b..293489c41086 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -5420,3 +5420,20 @@ void dc_set_edp_power(const struct dc *dc, struct dc_link *edp_link,
 	edp_link->dc->link_srv->edp_set_panel_power(edp_link, powerOn);
 }
 
+/*
+ *****************************************************************************
+ * dc_get_power_profile_for_dc_state() - extracts power profile from dc state
+ *
+ * Called when DM wants to make power policy decisions based on dc_state
+ *
+ *****************************************************************************
+ */
+struct dc_power_profile dc_get_power_profile_for_dc_state(const struct dc_state *context)
+{
+	struct dc_power_profile profile = { 0 };
+
+	profile.power_level += !context->bw_ctx.bw.dcn.clk.p_state_change_support;
+
+	return profile;
+}
+
diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index ed9b72ac369f..faf897ac75d8 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -2407,6 +2407,12 @@ void dc_print_dmub_diagnostic_data(const struct dc *dc);
 
 void dc_query_current_properties(struct dc *dc, struct dc_current_properties *properties);
 
+struct dc_power_profile {
+	int power_level; /* Lower is better */
+};
+
+struct dc_power_profile dc_get_power_profile_for_dc_state(const struct dc_state *context);
+
 /* DSC Interfaces */
 #include "dc_dsc.h"
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c
index 8e3acb92385b..ed804439fb7e 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn315/dcn315_resource.c
@@ -137,8 +137,8 @@
 #define DCN3_15_MAX_DET_SIZE 384
 #define DCN3_15_CRB_SEGMENT_SIZE_KB 64
 #define DCN3_15_MAX_DET_SEGS (DCN3_15_MAX_DET_SIZE / DCN3_15_CRB_SEGMENT_SIZE_KB)
-/* Minimum 2 extra segments need to be in compbuf and claimable to guarantee seamless mpo transitions */
-#define MIN_RESERVED_DET_SEGS 2
+/* Minimum 3 extra segments need to be in compbuf and claimable to guarantee seamless mpo transitions */
+#define MIN_RESERVED_DET_SEGS 3
 
 enum dcn31_clk_src_array_id {
 	DCN31_CLK_SRC_PLL0,
-- 
2.37.3

[PATCH 09/19] drm/amd/display: skip audio config for virtual signal

[Wayne Lin]

From: Alex Hung <alex.hung@amd.com>

[WHY & HOW]
IGT's kms_hdmi_inject (subtest inject_audio) triggers a warning message
from dce_aud_az_configure. This can be fixed by checking Virtual signal
that does not need to configure AZ audio since it does not have any.

Reviewed-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Alex Hung <alex.hung@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dce/dce_audio.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
index c94a966c6612..140598f18bbd 100644
--- a/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
+++ b/drivers/gpu/drm/amd/display/dc/dce/dce_audio.c
@@ -407,6 +407,10 @@ void dce_aud_az_configure(
 	bool is_ac3_supported = false;
 	union audio_sample_rates sample_rate;
 	uint32_t strlen = 0;
+
+	if (signal == SIGNAL_TYPE_VIRTUAL)
+		return;
+
 	value = AZ_REG_READ(AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL);
 	set_reg_field_value(value, 1,
 			AZALIA_F0_CODEC_PIN_CONTROL_HOT_PLUG_CONTROL,
-- 
2.37.3

[PATCH 10/19] drm/amd/display: Improve code style on bios_parser2

[Wayne Lin]

From: Rodrigo Siqueira <rodrigo.siqueira@amd.com>

bios_parser2 uses space instead of tabs in multiple areas; this commit
converts those spaces into tabs. It also removes some other small
details.

Reviewed-by: Aurabindo Pillai <aurabindo.pillai@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Rodrigo Siqueira <rodrigo.siqueira@amd.com>
---
 .../drm/amd/display/dc/bios/bios_parser2.c    | 37 +++++++++----------
 1 file changed, 17 insertions(+), 20 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
index 2ec496be778a..7cdb1a8a0ba0 100644
--- a/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
+++ b/drivers/gpu/drm/amd/display/dc/bios/bios_parser2.c
@@ -355,7 +355,7 @@ static struct atom_display_object_path_v3 *get_bios_object_from_path_v3(struct b
 					&& id.enum_id == obj_id.enum_id)
 				return &bp->object_info_tbl.v1_5->display_path[i];
 		}
-        break;
+	break;
 
 	case OBJECT_TYPE_CONNECTOR:
 	case OBJECT_TYPE_GENERIC:
@@ -369,7 +369,7 @@ static struct atom_display_object_path_v3 *get_bios_object_from_path_v3(struct b
 					&& id.enum_id == obj_id.enum_id)
 				return &bp->object_info_tbl.v1_5->display_path[i];
 		}
-        break;
+	break;
 
 	default:
 		return NULL;
@@ -405,16 +405,16 @@ static enum bp_result bios_parser_get_i2c_info(struct dc_bios *dcb,
 	}
 
 	switch (bp->object_info_tbl.revision.minor) {
-	    case 4:
-	    default:
-	        object = get_bios_object(bp, id);
+	case 4:
+	default:
+		object = get_bios_object(bp, id);
 
-	        if (!object)
-				return BP_RESULT_BADINPUT;
+		if (!object)
+			return BP_RESULT_BADINPUT;
 
-	        offset = object->disp_recordoffset + bp->object_info_tbl_offset;
-	        break;
-	    case 5:
+		offset = object->disp_recordoffset + bp->object_info_tbl_offset;
+		break;
+	case 5:
 		object_path_v3 = get_bios_object_from_path_v3(bp, id);
 
 		if (!object_path_v3)
@@ -568,17 +568,16 @@ static enum bp_result bios_parser_get_hpd_info(
 		return BP_RESULT_BADINPUT;
 
 	switch (bp->object_info_tbl.revision.minor) {
-	    case 4:
-	    default:
-	        object = get_bios_object(bp, id);
+	case 4:
+	default:
+		object = get_bios_object(bp, id);
 
 		if (!object)
 			return BP_RESULT_BADINPUT;
 
-	        record = get_hpd_record(bp, object);
-
-	        break;
-	    case 5:
+		record = get_hpd_record(bp, object);
+		break;
+	case 5:
 		object_path_v3 = get_bios_object_from_path_v3(bp, id);
 
 		if (!object_path_v3)
@@ -1746,7 +1745,7 @@ static enum bp_result bios_parser_get_firmware_info(
 			case 2:
 			case 3:
 				result = get_firmware_info_v3_2(bp, info);
-			break;
+				break;
 			case 4:
 				result = get_firmware_info_v3_4(bp, info);
 				break;
@@ -2206,10 +2205,8 @@ static enum bp_result bios_parser_get_disp_connector_caps_info(
 {
 	struct bios_parser *bp = BP_FROM_DCB(dcb);
 	struct atom_display_object_path_v2 *object;
-
 	struct atom_display_object_path_v3 *object_path_v3;
 	struct atom_connector_caps_record *record_path_v3;
-
 	struct atom_disp_connector_caps_record *record = NULL;
 
 	if (!info)
-- 
2.37.3

[PATCH 11/19] drm/amd/display: augment display clock in dc_cap structure

[Wayne Lin]

From: Max Tseng <max.tseng@amd.com>

[Why]
Allow dc report maximum display clock possible at vmin

Reviewed-by: Wayne Lin <wayne.lin@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Max Tseng <max.tseng@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dc.h                     | 1 +
 drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c | 2 ++
 drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c   | 2 ++
 3 files changed, 5 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dc.h b/drivers/gpu/drm/amd/display/dc/dc.h
index faf897ac75d8..4065bb061c85 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -270,6 +270,7 @@ struct dc_caps {
 	uint16_t subvp_vertical_int_margin_us;
 	bool seamless_odm;
 	uint32_t max_v_total;
+	uint32_t max_disp_clock_khz_at_vmin;
 	uint8_t subvp_drr_vblank_start_margin_us;
 };
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c
index 4d0052ca742f..8b1691152f05 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn314/dcn314_resource.c
@@ -1912,6 +1912,8 @@ static bool dcn314_resource_construct(
 	dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
 	dc->caps.color.mpc.ocsc = 1;
 
+	dc->caps.max_disp_clock_khz_at_vmin = 694000;
+
 	/* Use pipe context based otg sync logic */
 	dc->config.use_pipe_ctx_sync_logic = true;
 
diff --git a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c
index 67ff19d4116d..10ae1b3da751 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn35/dcn35_resource.c
@@ -1832,6 +1832,8 @@ static bool dcn35_resource_construct(
 	dc->caps.color.mpc.ogam_rom_caps.hlg = 0;
 	dc->caps.color.mpc.ocsc = 1;
 
+	dc->caps.max_disp_clock_khz_at_vmin = 669154;
+
 	/* Use pipe context based otg sync logic */
 	dc->config.use_pipe_ctx_sync_logic = true;
 	/* read VBIOS LTTPR caps */
-- 
2.37.3

[PATCH 12/19] drm/amd/display: Update OPP counter from new interface

[Wayne Lin]

From: Hong-lu Cheng <hong-lu.cheng@amd.com>

[Description]
Change to new interface for notifying OS of cursor support changes

Reviewed-by: Gang Chu <gang.chu@amd.com>
Reviewed-by: Jaina Han <jaina.han@amd.com>
Reviewed-by: Jun Lei <jun.lei@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Hong-lu Cheng <hong-lu.cheng@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc.c b/drivers/gpu/drm/amd/display/dc/core/dc.c
index 293489c41086..07d960e7bac7 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc.c
@@ -3495,7 +3495,7 @@ static void wait_for_outstanding_hw_updates(struct dc *dc, const struct dc_state
  */
 	int pipe_idx;
 	int opp_inst;
-	int opp_count = dc->res_pool->pipe_count;
+	int opp_count = dc->res_pool->res_cap->num_opp;
 	struct hubp *hubp;
 	int mpcc_inst;
 	const struct pipe_ctx *pipe_ctx;
-- 
2.37.3

[PATCH 13/19] drm/amd/display: Break after finding supported vlevel for repopulate

[Wayne Lin]

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

[Description]
After repopulating pipes for a split / merge, we want to find the
lowest vlevel that supports the desired mpc combine level.

Reviewed-by: Nevenko Stupar <nevenko.stupar@amd.com>
Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Reviewed-by: Aric Cyr <aric.cyr@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Alvin Lee <alvin.lee2@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
index dcbd38bb3ed1..062633af77a0 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
@@ -2183,6 +2183,7 @@ bool dcn32_internal_validate_bw(struct dc *dc,
 					vba->VoltageLevel = i;
 					vlevel = i;
 					flags_valid = true;
+					break;
 				}
 			}
 
-- 
2.37.3

[PATCH 14/19] drm/amd/display: Rename DisableMinDispClkODM in dc_config

[Wayne Lin]

From: Gabe Teeger <gabe.teeger@amd.com>

[what and why]
The logic for DisableMinDispClkODM is reversed now that odm is
disabled by default. Renaming to EnableMinDispClkODM.

Reviewed-by: Charlene Liu <charlene.liu@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Gabe Teeger <gabe.teeger@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 4065bb061c85..53555ec3b4be 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -421,7 +421,7 @@ struct dc_config {
 	int sdpif_request_limit_words_per_umc;
 	bool use_old_fixed_vs_sequence;
 	bool dc_mode_clk_limit_support;
-	bool DisableMinDispClkODM;
+	bool EnableMinDispClkODM;
 	bool enable_auto_dpm_test_logs;
 };
 
-- 
2.37.3

[PATCH 15/19] drm/amd/display: add missing function pointer for DCN321 resource

[Wayne Lin]

From: Wenjing Liu <wenjing.liu@amd.com>

[why]
acquire_free_pipe_as_secondary_opp_head function pointer is not
assigned. This causes ODM mode to be disabled after enabling
windowed MPO ODM capability on DCN321.

[how]
assign the correct function to the function pointer.

Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
---
 drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c
index 7d40fb6156f3..f11ff6f071e2 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn321/dcn321_resource.c
@@ -1589,6 +1589,7 @@ static struct resource_funcs dcn321_res_pool_funcs = {
 	.calculate_wm_and_dlg = dcn32_calculate_wm_and_dlg,
 	.populate_dml_pipes = dcn32_populate_dml_pipes_from_context,
 	.acquire_free_pipe_as_secondary_dpp_pipe = dcn32_acquire_free_pipe_as_secondary_dpp_pipe,
+	.acquire_free_pipe_as_secondary_opp_head = dcn32_acquire_free_pipe_as_secondary_opp_head,
 	.release_pipe = dcn20_release_pipe,
 	.add_stream_to_ctx = dcn30_add_stream_to_ctx,
 	.add_dsc_to_stream_resource = dcn20_add_dsc_to_stream_resource,
-- 
2.37.3

[PATCH 16/19] drm/amd/display: add get primary dpp pipe resource interface

[Wayne Lin]

From: Wenjing Liu <wenjing.liu@amd.com>

[why]
Need to have a helper function to find the primary dp pipe of the plane
associated with a dpp pipe

Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
---
 drivers/gpu/drm/amd/display/dc/core/dc_resource.c | 14 ++++++++++++++
 drivers/gpu/drm/amd/display/dc/inc/resource.h     |  6 ++++++
 2 files changed, 20 insertions(+)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 436892450936..614a1fb08809 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1805,6 +1805,20 @@ struct pipe_ctx *resource_get_opp_head(const struct pipe_ctx *pipe_ctx)
 	return opp_head;
 }
 
+struct pipe_ctx *resource_get_primary_dpp_pipe(const struct pipe_ctx *dpp_pipe)
+{
+	struct pipe_ctx *pri_dpp_pipe = (struct pipe_ctx *) dpp_pipe;
+
+	ASSERT(resource_is_pipe_type(dpp_pipe, DPP_PIPE));
+	while (pri_dpp_pipe->prev_odm_pipe)
+		pri_dpp_pipe = pri_dpp_pipe->prev_odm_pipe;
+	while (pri_dpp_pipe->top_pipe &&
+			pri_dpp_pipe->top_pipe->plane_state == pri_dpp_pipe->plane_state)
+		pri_dpp_pipe = pri_dpp_pipe->top_pipe;
+	return pri_dpp_pipe;
+}
+
+
 int resource_get_mpc_slice_index(const struct pipe_ctx *pipe_ctx)
 {
 	struct pipe_ctx *split_pipe = pipe_ctx->top_pipe;
diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h
index 6777b07d0727..170d6ab81aef 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/resource.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h
@@ -409,6 +409,12 @@ struct pipe_ctx *resource_get_otg_master(const struct pipe_ctx *pipe_ctx);
  */
 struct pipe_ctx *resource_get_opp_head(const struct pipe_ctx *pipe_ctx);
 
+/*
+ * Get the DPP pipe allocated for MPC slice 0 and ODM slice 0 of the plane
+ * associated with dpp_pipe.
+ */
+struct pipe_ctx *resource_get_primary_dpp_pipe(const struct pipe_ctx *dpp_pipe);
+
 /*
  * Get the MPC slice index counting from 0 from left most slice
  * For example, if a DPP pipe is used as a secondary pipe in MPCC combine, MPC
-- 
2.37.3

[PATCH 17/19] drm/amd/display: add primary pipe check when building slice table for dcn3x

[Wayne Lin]

From: Wenjing Liu <wenjing.liu@amd.com>

[why]
There is a corner use case where the same plane pointer can be added into
dc state twice. In addition to plane state pointer check we also need to match
primary pipe before identifying the same plane.

Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
---
 .../gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c | 16 ++++++++++------
 1 file changed, 10 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
index 062633af77a0..7179c2b3b1b7 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn32/dcn32_fpu.c
@@ -1086,6 +1086,7 @@ struct pipe_slice_table {
 	int odm_combine_count;
 
 	struct {
+		struct pipe_ctx *pri_pipe;
 		struct dc_plane_state *plane;
 		int slice_count;
 	} mpc_combines[MAX_SURFACES];
@@ -1113,12 +1114,14 @@ static void update_slice_table_for_stream(struct pipe_slice_table *table,
 }
 
 static void update_slice_table_for_plane(struct pipe_slice_table *table,
-		struct dc_plane_state *plane, int diff)
+		struct pipe_ctx *dpp_pipe, struct dc_plane_state *plane, int diff)
 {
 	int i;
+	struct pipe_ctx *pri_dpp_pipe = resource_get_primary_dpp_pipe(dpp_pipe);
 
 	for (i = 0; i < table->mpc_combine_count; i++) {
-		if (table->mpc_combines[i].plane == plane) {
+		if (table->mpc_combines[i].plane == plane &&
+				table->mpc_combines[i].pri_pipe == pri_dpp_pipe) {
 			table->mpc_combines[i].slice_count += diff;
 			break;
 		}
@@ -1127,6 +1130,7 @@ static void update_slice_table_for_plane(struct pipe_slice_table *table,
 	if (i == table->mpc_combine_count) {
 		table->mpc_combine_count++;
 		table->mpc_combines[i].plane = plane;
+		table->mpc_combines[i].pri_pipe = pri_dpp_pipe;
 		table->mpc_combines[i].slice_count = diff;
 	}
 }
@@ -1154,7 +1158,7 @@ static void init_pipe_slice_table_from_context(
 				&context->res_ctx, dpp_pipes);
 		for (j = 0; j < count; j++)
 			if (dpp_pipes[j]->plane_state)
-				update_slice_table_for_plane(table,
+				update_slice_table_for_plane(table, dpp_pipes[j],
 						dpp_pipes[j]->plane_state, 1);
 	}
 }
@@ -1205,7 +1209,7 @@ static bool update_pipe_slice_table_with_split_flags(
 				/* merging DPP pipe of the first ODM slice means
 				 * reducing MPC slice count by 1
 				 */
-				update_slice_table_for_plane(table, pipe->plane_state, -1);
+				update_slice_table_for_plane(table, pipe, pipe->plane_state, -1);
 			updated = true;
 		}
 
@@ -1216,8 +1220,8 @@ static bool update_pipe_slice_table_with_split_flags(
 				update_slice_table_for_stream(
 						table, pipe->stream, split[i] - 1);
 			else if (!odm && resource_is_pipe_type(pipe, DPP_PIPE))
-				update_slice_table_for_plane(
-						table, pipe->plane_state, split[i] - 1);
+				update_slice_table_for_plane(table, pipe,
+						pipe->plane_state, split[i] - 1);
 			updated = true;
 		}
 	}
-- 
2.37.3

[PATCH 18/19] drm/amd/display: fix incorrect odm change detection logic

[Wayne Lin]

From: Wenjing Liu <wenjing.liu@amd.com>

[why]
The current ODM change detection only compares first two ODM slices.
If there are 4 ODM slices and the change is within the last two slices, the
logic fails to detect ODM change and cause us to skip ODM programming
unexpectedly.

[how]
Add a is ODM topology changed resource interface to check any ODM
topology changes with a more generic method.

Reviewed-by: Dillon Varone <dillon.varone@amd.com>
Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Wenjing Liu <wenjing.liu@amd.com>
---
 .../gpu/drm/amd/display/dc/core/dc_resource.c | 23 +++++++++++++++++++
 .../drm/amd/display/dc/dcn20/dcn20_hwseq.c    |  8 ++-----
 drivers/gpu/drm/amd/display/dc/inc/resource.h | 10 ++++++++
 3 files changed, 35 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
index 614a1fb08809..e9837c0ce7f6 100644
--- a/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
+++ b/drivers/gpu/drm/amd/display/dc/core/dc_resource.c
@@ -1925,6 +1925,29 @@ bool resource_is_pipe_topology_changed(const struct dc_state *state_a,
 	return false;
 }
 
+bool resource_is_odm_topology_changed(const struct pipe_ctx *otg_master_a,
+		const struct pipe_ctx *otg_master_b)
+{
+	const struct pipe_ctx *opp_head_a = otg_master_a;
+	const struct pipe_ctx *opp_head_b = otg_master_b;
+
+	if (!resource_is_pipe_type(otg_master_a, OTG_MASTER) ||
+			!resource_is_pipe_type(otg_master_b, OTG_MASTER))
+		return true;
+
+	while (opp_head_a && opp_head_b) {
+		if (opp_head_a->stream_res.opp != opp_head_b->stream_res.opp)
+			return true;
+		if ((opp_head_a->next_odm_pipe && !opp_head_b->next_odm_pipe) ||
+				(!opp_head_a->next_odm_pipe && opp_head_b->next_odm_pipe))
+			return true;
+		opp_head_a = opp_head_a->next_odm_pipe;
+		opp_head_b = opp_head_b->next_odm_pipe;
+	}
+
+	return false;
+}
+
 /*
  * Sample log:
  *    pipe topology update
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 f2b20319f0bb..5738a5f00b56 100644
--- a/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/dcn20/dcn20_hwseq.c
@@ -1430,13 +1430,9 @@ static void dcn20_detect_pipe_changes(struct pipe_ctx *old_pipe, struct pipe_ctx
 	}
 
 	/* Detect top pipe only changes */
-	if (!new_pipe->top_pipe && !new_pipe->prev_odm_pipe) {
+	if (resource_is_pipe_type(new_pipe, OTG_MASTER)) {
 		/* Detect odm changes */
-		if ((old_pipe->next_odm_pipe && new_pipe->next_odm_pipe
-			&& old_pipe->next_odm_pipe->pipe_idx != new_pipe->next_odm_pipe->pipe_idx)
-				|| (!old_pipe->next_odm_pipe && new_pipe->next_odm_pipe)
-				|| (old_pipe->next_odm_pipe && !new_pipe->next_odm_pipe)
-				|| old_pipe->stream_res.opp != new_pipe->stream_res.opp)
+		if (resource_is_odm_topology_changed(new_pipe, old_pipe))
 			new_pipe->update_flags.bits.odm = 1;
 
 		/* Detect global sync changes */
diff --git a/drivers/gpu/drm/amd/display/dc/inc/resource.h b/drivers/gpu/drm/amd/display/dc/inc/resource.h
index 170d6ab81aef..00230a7fbe25 100644
--- a/drivers/gpu/drm/amd/display/dc/inc/resource.h
+++ b/drivers/gpu/drm/amd/display/dc/inc/resource.h
@@ -447,6 +447,16 @@ int resource_get_odm_slice_index(const struct pipe_ctx *opp_head);
 bool resource_is_pipe_topology_changed(const struct dc_state *state_a,
 		const struct dc_state *state_b);
 
+/*
+ * determine if the two OTG master pipes have the same ODM topology
+ * return
+ * false - if pipes passed in are not OTG masters or ODM topology is
+ * changed.
+ * true - otherwise
+ */
+bool resource_is_odm_topology_changed(const struct pipe_ctx *otg_master_a,
+		const struct pipe_ctx *otg_master_b);
+
 /* log the pipe topology update in state */
 void resource_log_pipe_topology_update(struct dc *dc, struct dc_state *state);
 
-- 
2.37.3

[PATCH 19/19] drm/amd/display: 3.2.253

[Wayne Lin]

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

This version brings along the following:

- Fix incorrect odm change detection logic
- Add primary pipe check when building slice table for dcn3x
- Add get primary dpp pipe resource interface
- Add missing function pointer for DCN321 resource
- Rename DisableMinDispClkODM in dc_config
- Break after finding supported vlevel for repopulate
- Update OPP counter from new interface
- Augment display clock in dc_cap structure
- Improve code style on bios_parser2
- Skip audio config for virtual signal
- Block MPO if it prevents pstate support
- Fix DP2.0 timing sync
- Improve x86 and dmub ips handshake
- Remove guaranteed viewports limitation for odm
- Add new windowed mpo odm minimal transition sequence
- Reset stream slice count for new ODM policy
- Determine fast update only before commit minimal transition state
- Remove unused mmhub_reg_offsets

Acked-by: Wayne Lin <wayne.lin@amd.com>
Signed-off-by: Aric Cyr <aric.cyr@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 53555ec3b4be..b6002b11a745 100644
--- a/drivers/gpu/drm/amd/display/dc/dc.h
+++ b/drivers/gpu/drm/amd/display/dc/dc.h
@@ -47,7 +47,7 @@ struct aux_payload;
 struct set_config_cmd_payload;
 struct dmub_notification;
 
-#define DC_VER "3.2.252"
+#define DC_VER "3.2.253"
 
 #define MAX_SURFACES 3
 #define MAX_PLANES 6
-- 
2.37.3