Merge tag 'amd-drm-fixes-6.9-2024-05-10' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.9-2024-05-10:

amdgpu:
- DCN 3.5 fix
- MST DSC fixes
- S0i3 fix
- S4 fix
- HDP MMIO mapping fix
- Fix a regression in visible vram handling

amdkfd:
- Spatial partition fix

Signed-off-by: Dave Airlie <airlied@redhat.com>

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240510171110.1394940-1-alexander.deucher@amd.com

7 files changed, 51 insertions, 18 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 109fe557a02bc8..29c197c0001871 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -427,7 +427,7 @@ bool amdgpu_res_cpu_visible(struct amdgpu_device *adev,
 
 	amdgpu_res_first(res, 0, res->size, &cursor);
 	while (cursor.remaining) {
-		if ((cursor.start + cursor.size) >= adev->gmc.visible_vram_size)
+		if ((cursor.start + cursor.size) > adev->gmc.visible_vram_size)
 			return false;
 		amdgpu_res_next(&cursor, cursor.size);
 	}
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 55aa74cbc5325e..1e6cc0bfc4328d 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -1139,7 +1139,7 @@ static int kfd_ioctl_alloc_memory_of_gpu(struct file *filep,
 			goto err_unlock;
 		}
 		offset = dev->adev->rmmio_remap.bus_addr;
-		if (!offset) {
+		if (!offset || (PAGE_SIZE > 4096)) {
 			err = -ENOMEM;
 			goto err_unlock;
 		}
@@ -2307,7 +2307,7 @@ static int criu_restore_memory_of_gpu(struct kfd_process_device *pdd,
 			return -EINVAL;
 		}
 		offset = pdd->dev->adev->rmmio_remap.bus_addr;
-		if (!offset) {
+		if (!offset || (PAGE_SIZE > 4096)) {
 			pr_err("amdgpu_amdkfd_get_mmio_remap_phys_addr failed\n");
 			return -ENOMEM;
 		}
@@ -3349,6 +3349,9 @@ static int kfd_mmio_mmap(struct kfd_node *dev, struct kfd_process *process,
 	if (vma->vm_end - vma->vm_start != PAGE_SIZE)
 		return -EINVAL;
 
+	if (PAGE_SIZE > 4096)
+		return -EINVAL;
+
 	address = dev->adev->rmmio_remap.bus_addr;
 
 	vm_flags_set(vma, VM_IO | VM_DONTCOPY | VM_DONTEXPAND | VM_NORESERVE |
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index c51f131eaa2fbc..bc9eb847ecfe76 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -1997,9 +1997,8 @@ int kfd_topology_add_device(struct kfd_node *gpu)
 			HSA_CAP_ASIC_REVISION_MASK);
 
 	dev->node_props.location_id = pci_dev_id(gpu->adev->pdev);
-	/* On multi-partition nodes, node id = location_id[31:28] */
-	if (gpu->kfd->num_nodes > 1)
-		dev->node_props.location_id |= (dev->gpu->node_id << 28);
+	if (KFD_GC_VERSION(dev->gpu->kfd) == IP_VERSION(9, 4, 3))
+		dev->node_props.location_id |= dev->gpu->node_id;
 
 	dev->node_props.domain = pci_domain_nr(gpu->adev->pdev->bus);
 	dev->node_props.max_engine_clk_fcompute =
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 941e96f100f4e5..cb31a699c66229 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
@@ -1219,8 +1219,10 @@ static bool is_dsc_need_re_compute(
 	if (dc_link->type != dc_connection_mst_branch)
 		return false;
 
-	if (!(dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_SUPPORT ||
-		dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_PASSTHROUGH_SUPPORT))
+	/* add a check for older MST DSC with no virtual DPCDs */
+	if (needs_dsc_aux_workaround(dc_link)  &&
+		(!(dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_SUPPORT ||
+		dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_PASSTHROUGH_SUPPORT)))
 		return false;
 
 	for (i = 0; i < MAX_PIPES; i++)
@@ -1240,7 +1242,15 @@ static bool is_dsc_need_re_compute(
 			continue;
 
 		aconnector = (struct amdgpu_dm_connector *) stream->dm_stream_context;
-		if (!aconnector)
+		if (!aconnector || !aconnector->dsc_aux)
+			continue;
+
+		/*
+		 *	check if cached virtual MST DSC caps are available and DSC is supported
+		 *	as per specifications in their Virtual DPCD registers.
+		*/
+		if (!(aconnector->dc_sink->dsc_caps.dsc_dec_caps.is_dsc_supported ||
+			aconnector->dc_link->dpcd_caps.dsc_caps.dsc_basic_caps.fields.dsc_support.DSC_PASSTHROUGH_SUPPORT))
 			continue;
 
 		stream_on_link[new_stream_on_link_num] = aconnector;
diff --git a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c
index 21e0eef3269b10..53e40d3c48d4bc 100644
--- a/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c
+++ b/drivers/gpu/drm/amd/display/dc/dml/dcn35/dcn35_fpu.c
@@ -195,9 +195,9 @@ struct _vcs_dpi_soc_bounding_box_st dcn3_5_soc = {
 	.dcn_downspread_percent = 0.5,
 	.gpuvm_min_page_size_bytes = 4096,
 	.hostvm_min_page_size_bytes = 4096,
-	.do_urgent_latency_adjustment = 0,
+	.do_urgent_latency_adjustment = 1,
 	.urgent_latency_adjustment_fabric_clock_component_us = 0,
-	.urgent_latency_adjustment_fabric_clock_reference_mhz = 0,
+	.urgent_latency_adjustment_fabric_clock_reference_mhz = 3000,
 };
 
 void dcn35_build_wm_range_table_fpu(struct clk_mgr *clk_mgr)
diff --git a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c
index a5560b3fc39ba9..9067ca78f85111 100644
--- a/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c
+++ b/drivers/gpu/drm/amd/display/dc/hwss/dcn35/dcn35_hwseq.c
@@ -638,22 +638,43 @@ void dcn35_power_down_on_boot(struct dc *dc)
 
 bool dcn35_apply_idle_power_optimizations(struct dc *dc, bool enable)
 {
-	struct dc_link *edp_links[MAX_NUM_EDP];
-	int i, edp_num;
 	if (dc->debug.dmcub_emulation)
 		return true;
 
 	if (enable) {
-		dc_get_edp_links(dc, edp_links, &edp_num);
-		if (edp_num == 0 || edp_num > 1)
-			return false;
+		uint32_t num_active_edp = 0;
+		int i;
 
 		for (i = 0; i < dc->current_state->stream_count; ++i) {
 			struct dc_stream_state *stream = dc->current_state->streams[i];
+			struct dc_link *link = stream->link;
+			bool is_psr = link && !link->panel_config.psr.disable_psr &&
+				      (link->psr_settings.psr_version == DC_PSR_VERSION_1 ||
+				       link->psr_settings.psr_version == DC_PSR_VERSION_SU_1);
+			bool is_replay = link && link->replay_settings.replay_feature_enabled;
+
+			/* Ignore streams that disabled. */
+			if (stream->dpms_off)
+				continue;
+
+			/* Active external displays block idle optimizations. */
+			if (!dc_is_embedded_signal(stream->signal))
+				return false;
+
+			/* If not PWRSEQ0 can't enter idle optimizations */
+			if (link && link->link_index != 0)
+				return false;
 
-			if (!stream->dpms_off && !dc_is_embedded_signal(stream->signal))
+			/* Check for panel power features required for idle optimizations. */
+			if (!is_psr && !is_replay)
 				return false;
+
+			num_active_edp += 1;
 		}
+
+		/* If more than one active eDP then disallow. */
+		if (num_active_edp > 1)
+			return false;
 	}
 
 	// TODO: review other cases when idle optimization is allowed
diff --git a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c
index 949131bd1ecb21..4abfcd32747d3a 100644
--- a/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c
+++ b/drivers/gpu/drm/amd/pm/swsmu/smu13/smu_v13_0_4_ppt.c
@@ -226,7 +226,7 @@ static int smu_v13_0_4_system_features_control(struct smu_context *smu, bool en)
 	struct amdgpu_device *adev = smu->adev;
 	int ret = 0;
 
-	if (!en && !adev->in_s0ix) {
+	if (!en && adev->in_s4) {
 		/* Adds a GFX reset as workaround just before sending the
 		 * MP1_UNLOAD message to prevent GC/RLC/PMFW from entering
 		 * an invalid state.