Merge tag 'amd-drm-fixes-6.3-2023-04-12' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.3-2023-04-12:

amdgpu:
- SMU13 fixes
- DP MST fix

Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230412215637.7881-1-alexander.deucher@amd.com

4 files changed, 135 insertions, 19 deletions

diff --git a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
index 1583157da355b2..efd025d8961e68 100644
--- a/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
+++ b/drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_helpers.c
@@ -177,6 +177,40 @@ void dm_helpers_dp_update_branch_info(
 	const struct dc_link *link)
 {}
 
+static void dm_helpers_construct_old_payload(
+			struct dc_link *link,
+			int pbn_per_slot,
+			struct drm_dp_mst_atomic_payload *new_payload,
+			struct drm_dp_mst_atomic_payload *old_payload)
+{
+	struct link_mst_stream_allocation_table current_link_table =
+									link->mst_stream_alloc_table;
+	struct link_mst_stream_allocation *dc_alloc;
+	int i;
+
+	*old_payload = *new_payload;
+
+	/* Set correct time_slots/PBN of old payload.
+	 * other fields (delete & dsc_enabled) in
+	 * struct drm_dp_mst_atomic_payload are don't care fields
+	 * while calling drm_dp_remove_payload()
+	 */
+	for (i = 0; i < current_link_table.stream_count; i++) {
+		dc_alloc =
+			&current_link_table.stream_allocations[i];
+
+		if (dc_alloc->vcp_id == new_payload->vcpi) {
+			old_payload->time_slots = dc_alloc->slot_count;
+			old_payload->pbn = dc_alloc->slot_count * pbn_per_slot;
+			break;
+		}
+	}
+
+	/* make sure there is an old payload*/
+	ASSERT(i != current_link_table.stream_count);
+
+}
+
 /*
  * Writes payload allocation table in immediate downstream device.
  */
@@ -188,7 +222,7 @@ bool dm_helpers_dp_mst_write_payload_allocation_table(
 {
 	struct amdgpu_dm_connector *aconnector;
 	struct drm_dp_mst_topology_state *mst_state;
-	struct drm_dp_mst_atomic_payload *payload;
+	struct drm_dp_mst_atomic_payload *target_payload, *new_payload, old_payload;
 	struct drm_dp_mst_topology_mgr *mst_mgr;
 
 	aconnector = (struct amdgpu_dm_connector *)stream->dm_stream_context;
@@ -204,17 +238,26 @@ bool dm_helpers_dp_mst_write_payload_allocation_table(
 	mst_state = to_drm_dp_mst_topology_state(mst_mgr->base.state);
 
 	/* It's OK for this to fail */
-	payload = drm_atomic_get_mst_payload_state(mst_state, aconnector->mst_output_port);
-	if (enable)
-		drm_dp_add_payload_part1(mst_mgr, mst_state, payload);
-	else
-		drm_dp_remove_payload(mst_mgr, mst_state, payload, payload);
+	new_payload = drm_atomic_get_mst_payload_state(mst_state, aconnector->mst_output_port);
+
+	if (enable) {
+		target_payload = new_payload;
+
+		drm_dp_add_payload_part1(mst_mgr, mst_state, new_payload);
+	} else {
+		/* construct old payload by VCPI*/
+		dm_helpers_construct_old_payload(stream->link, mst_state->pbn_div,
+						new_payload, &old_payload);
+		target_payload = &old_payload;
+
+		drm_dp_remove_payload(mst_mgr, mst_state, &old_payload, new_payload);
+	}
 
 	/* mst_mgr->->payloads are VC payload notify MST branch using DPCD or
 	 * AUX message. The sequence is slot 1-63 allocated sequence for each
 	 * stream. AMD ASIC stream slot allocation should follow the same
 	 * sequence. copy DRM MST allocation to dc */
-	fill_dc_mst_payload_table_from_drm(stream->link, enable, payload, proposed_table);
+	fill_dc_mst_payload_table_from_drm(stream->link, enable, target_payload, proposed_table);
 
 	return true;
 }
diff --git a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h
index f085cb97a62060..85a090b9e3d974 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h
+++ b/drivers/gpu/drm/amd/pm/swsmu/inc/smu_v13_0.h
@@ -61,6 +61,12 @@
 #define CTF_OFFSET_HOTSPOT		5
 #define CTF_OFFSET_MEM			5
 
+static const int pmfw_decoded_link_speed[5] = {1, 2, 3, 4, 5};
+static const int pmfw_decoded_link_width[7] = {0, 1, 2, 4, 8, 12, 16};
+
+#define DECODE_GEN_SPEED(gen_speed_idx)		(pmfw_decoded_link_speed[gen_speed_idx])
+#define DECODE_LANE_WIDTH(lane_width_idx)	(pmfw_decoded_link_width[lane_width_idx])
+
 struct smu_13_0_max_sustainable_clocks {
 	uint32_t display_clock;
 	uint32_t phy_clock;
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c
index 27448ffe60a439..a5c97d61e92a67 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_0_ppt.c
@@ -1144,8 +1144,8 @@ static int smu_v13_0_0_print_clk_levels(struct smu_context *smu,
 					(pcie_table->pcie_lane[i] == 5) ? "x12" :
 					(pcie_table->pcie_lane[i] == 6) ? "x16" : "",
 					pcie_table->clk_freq[i],
-					((gen_speed - 1) == pcie_table->pcie_gen[i]) &&
-					(lane_width == link_width[pcie_table->pcie_lane[i]]) ?
+					(gen_speed == DECODE_GEN_SPEED(pcie_table->pcie_gen[i])) &&
+					(lane_width == DECODE_LANE_WIDTH(link_width[pcie_table->pcie_lane[i]])) ?
 					"*" : "");
 		break;
 
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c
index 9e1967d8049e3a..4399416dd9b8f4 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_7_ppt.c
@@ -575,6 +575,14 @@ static int smu_v13_0_7_set_default_dpm_table(struct smu_context *smu)
 						     dpm_table);
 		if (ret)
 			return ret;
+
+		if (skutable->DriverReportedClocks.GameClockAc &&
+			(dpm_table->dpm_levels[dpm_table->count - 1].value >
+			skutable->DriverReportedClocks.GameClockAc)) {
+			dpm_table->dpm_levels[dpm_table->count - 1].value =
+				skutable->DriverReportedClocks.GameClockAc;
+			dpm_table->max = skutable->DriverReportedClocks.GameClockAc;
+		}
 	} else {
 		dpm_table->count = 1;
 		dpm_table->dpm_levels[0].value = smu->smu_table.boot_values.gfxclk / 100;
@@ -828,6 +836,57 @@ static int smu_v13_0_7_get_smu_metrics_data(struct smu_context *smu,
 	return ret;
 }
 
+static int smu_v13_0_7_get_dpm_ultimate_freq(struct smu_context *smu,
+					     enum smu_clk_type clk_type,
+					     uint32_t *min,
+					     uint32_t *max)
+{
+	struct smu_13_0_dpm_context *dpm_context =
+		smu->smu_dpm.dpm_context;
+	struct smu_13_0_dpm_table *dpm_table;
+
+	switch (clk_type) {
+	case SMU_MCLK:
+	case SMU_UCLK:
+		/* uclk dpm table */
+		dpm_table = &dpm_context->dpm_tables.uclk_table;
+		break;
+	case SMU_GFXCLK:
+	case SMU_SCLK:
+		/* gfxclk dpm table */
+		dpm_table = &dpm_context->dpm_tables.gfx_table;
+		break;
+	case SMU_SOCCLK:
+		/* socclk dpm table */
+		dpm_table = &dpm_context->dpm_tables.soc_table;
+		break;
+	case SMU_FCLK:
+		/* fclk dpm table */
+		dpm_table = &dpm_context->dpm_tables.fclk_table;
+		break;
+	case SMU_VCLK:
+	case SMU_VCLK1:
+		/* vclk dpm table */
+		dpm_table = &dpm_context->dpm_tables.vclk_table;
+		break;
+	case SMU_DCLK:
+	case SMU_DCLK1:
+		/* dclk dpm table */
+		dpm_table = &dpm_context->dpm_tables.dclk_table;
+		break;
+	default:
+		dev_err(smu->adev->dev, "Unsupported clock type!\n");
+		return -EINVAL;
+	}
+
+	if (min)
+		*min = dpm_table->min;
+	if (max)
+		*max = dpm_table->max;
+
+	return 0;
+}
+
 static int smu_v13_0_7_read_sensor(struct smu_context *smu,
 				   enum amd_pp_sensors sensor,
 				   void *data,
@@ -1074,8 +1133,8 @@ static int smu_v13_0_7_print_clk_levels(struct smu_context *smu,
 					(pcie_table->pcie_lane[i] == 5) ? "x12" :
 					(pcie_table->pcie_lane[i] == 6) ? "x16" : "",
 					pcie_table->clk_freq[i],
-					(gen_speed == pcie_table->pcie_gen[i]) &&
-					(lane_width == pcie_table->pcie_lane[i]) ?
+					(gen_speed == DECODE_GEN_SPEED(pcie_table->pcie_gen[i])) &&
+					(lane_width == DECODE_LANE_WIDTH(pcie_table->pcie_lane[i])) ?
 					"*" : "");
 		break;
 
@@ -1329,9 +1388,17 @@ static int smu_v13_0_7_populate_umd_state_clk(struct smu_context *smu)
 				&dpm_context->dpm_tables.fclk_table;
 	struct smu_umd_pstate_table *pstate_table =
 				&smu->pstate_table;
+	struct smu_table_context *table_context = &smu->smu_table;
+	PPTable_t *pptable = table_context->driver_pptable;
+	DriverReportedClocks_t driver_clocks =
+		pptable->SkuTable.DriverReportedClocks;
 
 	pstate_table->gfxclk_pstate.min = gfx_table->min;
-	pstate_table->gfxclk_pstate.peak = gfx_table->max;
+	if (driver_clocks.GameClockAc &&
+		(driver_clocks.GameClockAc < gfx_table->max))
+		pstate_table->gfxclk_pstate.peak = driver_clocks.GameClockAc;
+	else
+		pstate_table->gfxclk_pstate.peak = gfx_table->max;
 
 	pstate_table->uclk_pstate.min = mem_table->min;
 	pstate_table->uclk_pstate.peak = mem_table->max;
@@ -1348,12 +1415,12 @@ static int smu_v13_0_7_populate_umd_state_clk(struct smu_context *smu)
 	pstate_table->fclk_pstate.min = fclk_table->min;
 	pstate_table->fclk_pstate.peak = fclk_table->max;
 
-	/*
-	 * For now, just use the mininum clock frequency.
-	 * TODO: update them when the real pstate settings available
-	 */
-	pstate_table->gfxclk_pstate.standard = gfx_table->min;
-	pstate_table->uclk_pstate.standard = mem_table->min;
+	if (driver_clocks.BaseClockAc &&
+		driver_clocks.BaseClockAc < gfx_table->max)
+		pstate_table->gfxclk_pstate.standard = driver_clocks.BaseClockAc;
+	else
+		pstate_table->gfxclk_pstate.standard = gfx_table->max;
+	pstate_table->uclk_pstate.standard = mem_table->max;
 	pstate_table->socclk_pstate.standard = soc_table->min;
 	pstate_table->vclk_pstate.standard = vclk_table->min;
 	pstate_table->dclk_pstate.standard = dclk_table->min;
@@ -1676,7 +1743,7 @@ static const struct pptable_funcs smu_v13_0_7_ppt_funcs = {
 	.dpm_set_jpeg_enable = smu_v13_0_set_jpeg_enable,
 	.init_pptable_microcode = smu_v13_0_init_pptable_microcode,
 	.populate_umd_state_clk = smu_v13_0_7_populate_umd_state_clk,
-	.get_dpm_ultimate_freq = smu_v13_0_get_dpm_ultimate_freq,
+	.get_dpm_ultimate_freq = smu_v13_0_7_get_dpm_ultimate_freq,
 	.get_vbios_bootup_values = smu_v13_0_get_vbios_bootup_values,
 	.read_sensor = smu_v13_0_7_read_sensor,
 	.feature_is_enabled = smu_cmn_feature_is_enabled,