Merge tag 'amd-drm-next-6.9-2024-03-01' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amd-drm-next-6.9-2024-03-01:

amdgpu:
- GC 11.5.1 updates
- Misc display cleanups
- NBIO 7.9 updates
- Backlight fixes
- DMUB fixes
- MPO fixes
- atomfirmware table updates
- SR-IOV fixes
- VCN 4.x updates
- use RMW accessors for pci config registers
- PSR fixes
- Suspend/resume fixes
- RAS fixes
- ABM fixes
- Misc code cleanups
- SI DPM fix
- Revert freesync video

amdkfd:
- Misc cleanups
- Error handling fixes

radeon:
- use RMW accessors for pci config registers

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20240301204857.13960-1-alexander.deucher@amd.com
Signed-off-by: Dave Airlie <airlied@redhat.com>

Author: airlied

drivers/gpu/drm/amd/amdgpu/amdgpu.h

@@ -194,6 +194,7 @@ extern int amdgpu_emu_mode;
 extern uint amdgpu_smu_memory_pool_size;
 extern int amdgpu_smu_pptable_id;
 extern uint amdgpu_dc_feature_mask;
+extern uint amdgpu_freesync_vid_mode;
 extern uint amdgpu_dc_debug_mask;
 extern uint amdgpu_dc_visual_confirm;
 extern int amdgpu_dm_abm_level;

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.c

@@ -742,6 +742,11 @@ void amdgpu_amdkfd_debug_mem_fence(struct amdgpu_device *adev)
 	amdgpu_device_flush_hdp(adev, NULL);
 }
 
+bool amdgpu_amdkfd_is_fed(struct amdgpu_device *adev)
+{
+	return amdgpu_ras_get_fed_status(adev);
+}
+
 void amdgpu_amdkfd_ras_poison_consumption_handler(struct amdgpu_device *adev,
 	enum amdgpu_ras_block block, bool reset)
 {

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd.h

@@ -320,7 +320,7 @@ int amdgpu_amdkfd_gpuvm_map_gtt_bo_to_kernel(struct kgd_mem *mem,
 					     void **kptr, uint64_t *size);
 void amdgpu_amdkfd_gpuvm_unmap_gtt_bo_from_kernel(struct kgd_mem *mem);
 
-int amdgpu_amdkfd_map_gtt_bo_to_gart(struct amdgpu_device *adev, struct amdgpu_bo *bo);
+int amdgpu_amdkfd_map_gtt_bo_to_gart(struct amdgpu_bo *bo);
 
 int amdgpu_amdkfd_gpuvm_restore_process_bos(void *process_info,
 					    struct dma_fence __rcu **ef);
@@ -337,6 +337,7 @@ int amdgpu_amdkfd_get_tile_config(struct amdgpu_device *adev,
 				struct tile_config *config);
 void amdgpu_amdkfd_ras_poison_consumption_handler(struct amdgpu_device *adev,
 			enum amdgpu_ras_block block, bool reset);
+bool amdgpu_amdkfd_is_fed(struct amdgpu_device *adev);
 bool amdgpu_amdkfd_bo_mapped_to_dev(struct amdgpu_device *adev, struct kgd_mem *mem);
 void amdgpu_amdkfd_block_mmu_notifications(void *p);
 int amdgpu_amdkfd_criu_resume(void *p);

drivers/gpu/drm/amd/amdgpu/amdgpu_amdkfd_gpuvm.c

@@ -2189,13 +2189,12 @@ int amdgpu_amdkfd_gpuvm_sync_memory(
 
 /**
  * amdgpu_amdkfd_map_gtt_bo_to_gart - Map BO to GART and increment reference count
- * @adev: Device to which allocated BO belongs
  * @bo: Buffer object to be mapped
  *
  * Before return, bo reference count is incremented. To release the reference and unpin/
  * unmap the BO, call amdgpu_amdkfd_free_gtt_mem.
  */
-int amdgpu_amdkfd_map_gtt_bo_to_gart(struct amdgpu_device *adev, struct amdgpu_bo *bo)
+int amdgpu_amdkfd_map_gtt_bo_to_gart(struct amdgpu_bo *bo)
 {
 	int ret;
 

drivers/gpu/drm/amd/amdgpu/amdgpu_device.c

@@ -4040,8 +4040,10 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	 * early on during init and before calling to RREG32.
 	 */
 	adev->reset_domain = amdgpu_reset_create_reset_domain(SINGLE_DEVICE, "amdgpu-reset-dev");
-	if (!adev->reset_domain)
-		return -ENOMEM;
+	if (!adev->reset_domain) {
+		r = -ENOMEM;
+		goto unmap_memory;
+	}
 
 	/* detect hw virtualization here */
 	amdgpu_detect_virtualization(adev);
@@ -4051,20 +4053,20 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	r = amdgpu_device_get_job_timeout_settings(adev);
 	if (r) {
 		dev_err(adev->dev, "invalid lockup_timeout parameter syntax\n");
-		return r;
+		goto unmap_memory;
 	}
 
 	/* early init functions */
 	r = amdgpu_device_ip_early_init(adev);
 	if (r)
-		return r;
+		goto unmap_memory;
 
 	amdgpu_device_set_mcbp(adev);
 
 	/* Get rid of things like offb */
 	r = drm_aperture_remove_conflicting_pci_framebuffers(adev->pdev, &amdgpu_kms_driver);
 	if (r)
-		return r;
+		goto unmap_memory;
 
 	/* Enable TMZ based on IP_VERSION */
 	amdgpu_gmc_tmz_set(adev);
@@ -4074,7 +4076,7 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 	if (adev->gmc.xgmi.supported) {
 		r = adev->gfxhub.funcs->get_xgmi_info(adev);
 		if (r)
-			return r;
+			goto unmap_memory;
 	}
 
 	/* enable PCIE atomic ops */
@@ -4343,6 +4345,8 @@ int amdgpu_device_init(struct amdgpu_device *adev,
 failed:
 	amdgpu_vf_error_trans_all(adev);
 
+unmap_memory:
+	iounmap(adev->rmmio);
 	return r;
 }
 
@@ -5321,6 +5325,7 @@ int amdgpu_do_asic_reset(struct list_head *device_list_handle,
 	list_for_each_entry(tmp_adev, device_list_handle, reset_list) {
 		if (need_full_reset) {
 			/* post card */
+			amdgpu_ras_set_fed(tmp_adev, false);
 			r = amdgpu_device_asic_init(tmp_adev);
 			if (r) {
 				dev_warn(tmp_adev->dev, "asic atom init failed!");

drivers/gpu/drm/amd/amdgpu/amdgpu_discovery.c

@@ -1683,6 +1683,7 @@ static int amdgpu_discovery_set_common_ip_blocks(struct amdgpu_device *adev)
 	case IP_VERSION(11, 0, 3):
 	case IP_VERSION(11, 0, 4):
 	case IP_VERSION(11, 5, 0):
+	case IP_VERSION(11, 5, 1):
 		amdgpu_device_ip_block_add(adev, &soc21_common_ip_block);
 		break;
 	default:
@@ -1730,6 +1731,7 @@ static int amdgpu_discovery_set_gmc_ip_blocks(struct amdgpu_device *adev)
 	case IP_VERSION(11, 0, 3):
 	case IP_VERSION(11, 0, 4):
 	case IP_VERSION(11, 5, 0):
+	case IP_VERSION(11, 5, 1):
 		amdgpu_device_ip_block_add(adev, &gmc_v11_0_ip_block);
 		break;
 	default:
@@ -2003,6 +2005,7 @@ static int amdgpu_discovery_set_gc_ip_blocks(struct amdgpu_device *adev)
 	case IP_VERSION(11, 0, 3):
 	case IP_VERSION(11, 0, 4):
 	case IP_VERSION(11, 5, 0):
+	case IP_VERSION(11, 5, 1):
 		amdgpu_device_ip_block_add(adev, &gfx_v11_0_ip_block);
 		break;
 	default:
@@ -2137,6 +2140,7 @@ static int amdgpu_discovery_set_mm_ip_blocks(struct amdgpu_device *adev)
 			amdgpu_device_ip_block_add(adev, &jpeg_v4_0_3_ip_block);
 			break;
 		case IP_VERSION(4, 0, 5):
+		case IP_VERSION(4, 0, 6):
 			amdgpu_device_ip_block_add(adev, &vcn_v4_0_5_ip_block);
 			amdgpu_device_ip_block_add(adev, &jpeg_v4_0_5_ip_block);
 			break;
@@ -2182,6 +2186,7 @@ static int amdgpu_discovery_set_mes_ip_blocks(struct amdgpu_device *adev)
 	case IP_VERSION(11, 0, 3):
 	case IP_VERSION(11, 0, 4):
 	case IP_VERSION(11, 5, 0):
+	case IP_VERSION(11, 5, 1):
 		amdgpu_device_ip_block_add(adev, &mes_v11_0_ip_block);
 		adev->enable_mes = true;
 		adev->enable_mes_kiq = true;
@@ -2460,6 +2465,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 		adev->family = AMDGPU_FAMILY_GC_11_0_1;
 		break;
 	case IP_VERSION(11, 5, 0):
+	case IP_VERSION(11, 5, 1):
 		adev->family = AMDGPU_FAMILY_GC_11_5_0;
 		break;
 	default:
@@ -2479,6 +2485,7 @@ int amdgpu_discovery_set_ip_blocks(struct amdgpu_device *adev)
 	case IP_VERSION(11, 0, 1):
 	case IP_VERSION(11, 0, 4):
 	case IP_VERSION(11, 5, 0):
+	case IP_VERSION(11, 5, 1):
 		adev->flags |= AMD_IS_APU;
 		break;
 	default:

drivers/gpu/drm/amd/amdgpu/amdgpu_display.c

@@ -1350,14 +1350,6 @@ int amdgpu_display_modeset_create_props(struct amdgpu_device *adev)
 					 "dither",
 					 amdgpu_dither_enum_list, sz);
 
-	if (adev->dc_enabled) {
-		adev->mode_info.abm_level_property =
-			drm_property_create_range(adev_to_drm(adev), 0,
-						  "abm level", 0, 4);
-		if (!adev->mode_info.abm_level_property)
-			return -ENOMEM;
-	}
-
 	return 0;
 }
 

drivers/gpu/drm/amd/amdgpu/amdgpu_drv.c

@@ -199,6 +199,7 @@ int amdgpu_mes_kiq;
 int amdgpu_noretry = -1;
 int amdgpu_force_asic_type = -1;
 int amdgpu_tmz = -1; /* auto */
+uint amdgpu_freesync_vid_mode;
 int amdgpu_reset_method = -1; /* auto */
 int amdgpu_num_kcq = -1;
 int amdgpu_smartshift_bias;
@@ -883,6 +884,32 @@ module_param_named(damageclips, amdgpu_damage_clips, int, 0444);
 MODULE_PARM_DESC(tmz, "Enable TMZ feature (-1 = auto (default), 0 = off, 1 = on)");
 module_param_named(tmz, amdgpu_tmz, int, 0444);
 
+/**
+ * DOC: freesync_video (uint)
+ * Enable the optimization to adjust front porch timing to achieve seamless
+ * mode change experience when setting a freesync supported mode for which full
+ * modeset is not needed.
+ *
+ * The Display Core will add a set of modes derived from the base FreeSync
+ * video mode into the corresponding connector's mode list based on commonly
+ * used refresh rates and VRR range of the connected display, when users enable
+ * this feature. From the userspace perspective, they can see a seamless mode
+ * change experience when the change between different refresh rates under the
+ * same resolution. Additionally, userspace applications such as Video playback
+ * can read this modeset list and change the refresh rate based on the video
+ * frame rate. Finally, the userspace can also derive an appropriate mode for a
+ * particular refresh rate based on the FreeSync Mode and add it to the
+ * connector's mode list.
+ *
+ * Note: This is an experimental feature.
+ *
+ * The default value: 0 (off).
+ */
+MODULE_PARM_DESC(
+	freesync_video,
+	"Enable freesync modesetting optimization feature (0 = off (default), 1 = on)");
+module_param_named(freesync_video, amdgpu_freesync_vid_mode, uint, 0444);
+
 /**
  * DOC: reset_method (int)
  * GPU reset method (-1 = auto (default), 0 = legacy, 1 = mode0, 2 = mode1, 3 = mode2, 4 = baco)
@@ -2764,8 +2791,7 @@ static int amdgpu_pmops_runtime_idle(struct device *dev)
 {
 	struct drm_device *drm_dev = dev_get_drvdata(dev);
 	struct amdgpu_device *adev = drm_to_adev(drm_dev);
-	/* we don't want the main rpm_idle to call suspend - we want to autosuspend */
-	int ret = 1;
+	int ret;
 
 	if (adev->pm.rpm_mode == AMDGPU_RUNPM_NONE) {
 		pm_runtime_forbid(dev);

drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c

@@ -61,9 +61,7 @@ static struct kmem_cache *amdgpu_fence_slab;
 
 int amdgpu_fence_slab_init(void)
 {
-	amdgpu_fence_slab = kmem_cache_create(
-		"amdgpu_fence", sizeof(struct amdgpu_fence), 0,
-		SLAB_HWCACHE_ALIGN, NULL);
+	amdgpu_fence_slab = KMEM_CACHE(amdgpu_fence, SLAB_HWCACHE_ALIGN);
 	if (!amdgpu_fence_slab)
 		return -ENOMEM;
 	return 0;

drivers/gpu/drm/amd/amdgpu/amdgpu_gfx.c

@@ -304,11 +304,11 @@ static int amdgpu_gfx_kiq_acquire(struct amdgpu_device *adev,
 	return -EINVAL;
 }
 
-int amdgpu_gfx_kiq_init_ring(struct amdgpu_device *adev,
-			     struct amdgpu_ring *ring,
-			     struct amdgpu_irq_src *irq, int xcc_id)
+int amdgpu_gfx_kiq_init_ring(struct amdgpu_device *adev, int xcc_id)
 {
 	struct amdgpu_kiq *kiq = &adev->gfx.kiq[xcc_id];
+	struct amdgpu_irq_src *irq = &kiq->irq;
+	struct amdgpu_ring *ring = &kiq->ring;
 	int r = 0;
 
 	spin_lock_init(&kiq->ring_lock);