Merge tag 'amd-drm-next-6.16-2025-05-16' of https://gitlab.freedesktop.org/agd5f/linux into drm-next

amdgpu:
- Misc code cleanups
- UserQ fixes
- MALL reporting fix
- DP AUX fixes
- DCN 3.5 fixes
- DP MST fixes
- DC DMI quirks cleanup
- RAS fixes
- SR-IOV updates
- GC 9.5 updates
- Misc display fixes
- VCN 4.0.5 powergating race fix
- SMU 13.x updates
- Paritioning fixes
- VCN 5.0.1 SR-IOV updates
- JPEG 5.0.1 SR-IOV updates

amdkfd:
- Fix spurious warning in interrupt code
- XNACK fixes

radeon:
- CIK doorbell cleanup

Signed-off-by: Dave Airlie <[email protected]>

From: Alex Deucher <[email protected]>
Link: https://lore.kernel.org/r/[email protected]

Author: airlied

Documentation/gpu/amdgpu/amdgpu-glossary.rst

@@ -182,7 +182,7 @@ we have a dedicated glossary for Display Core at
     SMU/SMC
       System Management Unit / System Management Controller
 
-    SPI
+    SPI (AMDGPU)
       Shader Processor Input
 
     SRLC

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

@@ -281,6 +281,9 @@ static int convert_atom_mem_type_to_vram_type(struct amdgpu_device *adev,
 		case ATOM_DGPU_VRAM_TYPE_GDDR6:
 			vram_type = AMDGPU_VRAM_TYPE_GDDR6;
 			break;
+		case ATOM_DGPU_VRAM_TYPE_HBM3E:
+			vram_type = AMDGPU_VRAM_TYPE_HBM3E;
+			break;
 		default:
 			vram_type = AMDGPU_VRAM_TYPE_UNKNOWN;
 			break;

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

@@ -109,7 +109,7 @@ int amdgpu_unmap_static_csa(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	struct drm_exec exec;
 	int r;
 
-	drm_exec_init(&exec, DRM_EXEC_INTERRUPTIBLE_WAIT, 0);
+	drm_exec_init(&exec, 0, 0);
 	drm_exec_until_all_locked(&exec) {
 		r = amdgpu_vm_lock_pd(vm, &exec, 0);
 		if (likely(!r))

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

@@ -2105,6 +2105,7 @@ int amdgpu_debugfs_init(struct amdgpu_device *adev)
 	amdgpu_rap_debugfs_init(adev);
 	amdgpu_securedisplay_debugfs_init(adev);
 	amdgpu_fw_attestation_debugfs_init(adev);
+	amdgpu_psp_debugfs_init(adev);
 
 	debugfs_create_file("amdgpu_evict_vram", 0400, root, adev,
 			    &amdgpu_evict_vram_fops);

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

@@ -183,7 +183,7 @@ void amdgpu_eviction_fence_destroy(struct amdgpu_eviction_fence_mgr *evf_mgr)
 	dma_fence_wait(&ev_fence->base, false);
 
 	/* Last unref of ev_fence */
-	dma_fence_put(&evf_mgr->ev_fence->base);
+	dma_fence_put(&ev_fence->base);
 }
 
 int amdgpu_eviction_fence_attach(struct amdgpu_eviction_fence_mgr *evf_mgr,

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

@@ -62,6 +62,9 @@
  */
 #define AMDGPU_GMC_FAULT_TIMEOUT	5000ULL
 
+/* XNACK flags */
+#define AMDGPU_GMC_XNACK_FLAG_CHAIN BIT(0)
+
 struct firmware;
 
 enum amdgpu_memory_partition {
@@ -301,6 +304,7 @@ struct amdgpu_gmc {
 	struct amdgpu_xgmi xgmi;
 	struct amdgpu_irq_src	ecc_irq;
 	int noretry;
+	uint32_t xnack_flags;
 
 	uint32_t	vmid0_page_table_block_size;
 	uint32_t	vmid0_page_table_depth;

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

@@ -619,6 +619,10 @@ int amdgpu_irq_get(struct amdgpu_device *adev, struct amdgpu_irq_src *src,
 int amdgpu_irq_put(struct amdgpu_device *adev, struct amdgpu_irq_src *src,
 		   unsigned int type)
 {
+	/* When the threshold is reached,the interrupt source may not be enabled.return -EINVAL */
+	if (amdgpu_ras_is_rma(adev))
+		return -EINVAL;
+
 	if (!adev->irq.installed)
 		return -ENOENT;
 

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

@@ -1425,16 +1425,16 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 	mutex_init(&fpriv->bo_list_lock);
 	idr_init_base(&fpriv->bo_list_handles, 1);
 
+	r = amdgpu_userq_mgr_init(&fpriv->userq_mgr, file_priv, adev);
+	if (r)
+		DRM_WARN("Can't setup usermode queues, use legacy workload submission only\n");
+
 	r = amdgpu_eviction_fence_init(&fpriv->evf_mgr);
 	if (r)
 		goto error_vm;
 
 	amdgpu_ctx_mgr_init(&fpriv->ctx_mgr, adev);
 
-	r = amdgpu_userq_mgr_init(&fpriv->userq_mgr, file_priv, adev);
-	if (r)
-		DRM_WARN("Can't setup usermode queues, use legacy workload submission only\n");
-
 	file_priv->driver_priv = fpriv;
 	goto out_suspend;
 
@@ -1502,10 +1502,11 @@ void amdgpu_driver_postclose_kms(struct drm_device *dev,
 		amdgpu_bo_unreserve(pd);
 	}
 
-	fpriv->evf_mgr.fd_closing = true;
-	amdgpu_userq_mgr_fini(&fpriv->userq_mgr);
-	amdgpu_eviction_fence_destroy(&fpriv->evf_mgr);
-
+	if (!fpriv->evf_mgr.fd_closing) {
+		fpriv->evf_mgr.fd_closing = true;
+		amdgpu_userq_mgr_fini(&fpriv->userq_mgr);
+		amdgpu_eviction_fence_destroy(&fpriv->evf_mgr);
+	}
 	amdgpu_ctx_mgr_fini(&fpriv->ctx_mgr);
 	amdgpu_vm_fini(adev, &fpriv->vm);
 

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

@@ -1044,7 +1044,8 @@ static const char * const amdgpu_vram_names[] = {
 	"GDDR6",
 	"DDR5",
 	"LPDDR4",
-	"LPDDR5"
+	"LPDDR5",
+	"HBM3E"
 };
 
 /**

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

@@ -4186,6 +4186,110 @@ const struct attribute_group amdgpu_flash_attr_group = {
 	.is_visible = amdgpu_flash_attr_is_visible,
 };
 
+#if defined(CONFIG_DEBUG_FS)
+static int psp_read_spirom_debugfs_open(struct inode *inode, struct file *filp)
+{
+	struct amdgpu_device *adev = filp->f_inode->i_private;
+	struct spirom_bo *bo_triplet;
+	int ret;
+
+	/* serialize the open() file calling */
+	if (!mutex_trylock(&adev->psp.mutex))
+		return -EBUSY;
+
+	/*
+	 * make sure only one userpace process is alive for dumping so that
+	 * only one memory buffer of AMD_VBIOS_FILE_MAX_SIZE * 2 is consumed.
+	 * let's say the case where one process try opening the file while
+	 * another one has proceeded to read or release. In this way, eliminate
+	 * the use of mutex for read() or release() callback as well.
+	 */
+	if (adev->psp.spirom_dump_trip) {
+		mutex_unlock(&adev->psp.mutex);
+		return -EBUSY;
+	}
+
+	bo_triplet = kzalloc(sizeof(struct spirom_bo), GFP_KERNEL);
+	if (!bo_triplet) {
+		mutex_unlock(&adev->psp.mutex);
+		return -ENOMEM;
+	}
+
+	ret = amdgpu_bo_create_kernel(adev, AMD_VBIOS_FILE_MAX_SIZE_B * 2,
+				      AMDGPU_GPU_PAGE_SIZE,
+				      AMDGPU_GEM_DOMAIN_GTT,
+				      &bo_triplet->bo,
+				      &bo_triplet->mc_addr,
+				      &bo_triplet->cpu_addr);
+	if (ret)
+		goto rel_trip;
+
+	ret = psp_dump_spirom(&adev->psp, bo_triplet->mc_addr);
+	if (ret)
+		goto rel_bo;
+
+	adev->psp.spirom_dump_trip = bo_triplet;
+	mutex_unlock(&adev->psp.mutex);
+	return 0;
+rel_bo:
+	amdgpu_bo_free_kernel(&bo_triplet->bo, &bo_triplet->mc_addr,
+			      &bo_triplet->cpu_addr);
+rel_trip:
+	kfree(bo_triplet);
+	mutex_unlock(&adev->psp.mutex);
+	dev_err(adev->dev, "Trying IFWI dump fails, err = %d\n", ret);
+	return ret;
+}
+
+static ssize_t psp_read_spirom_debugfs_read(struct file *filp, char __user *buf, size_t size,
+					    loff_t *pos)
+{
+	struct amdgpu_device *adev = filp->f_inode->i_private;
+	struct spirom_bo *bo_triplet = adev->psp.spirom_dump_trip;
+
+	if (!bo_triplet)
+		return -EINVAL;
+
+	return simple_read_from_buffer(buf,
+				       size,
+				       pos, bo_triplet->cpu_addr,
+				       AMD_VBIOS_FILE_MAX_SIZE_B * 2);
+}
+
+static int psp_read_spirom_debugfs_release(struct inode *inode, struct file *filp)
+{
+	struct amdgpu_device *adev = filp->f_inode->i_private;
+	struct spirom_bo *bo_triplet = adev->psp.spirom_dump_trip;
+
+	if (bo_triplet) {
+		amdgpu_bo_free_kernel(&bo_triplet->bo, &bo_triplet->mc_addr,
+				      &bo_triplet->cpu_addr);
+		kfree(bo_triplet);
+	}
+
+	adev->psp.spirom_dump_trip = NULL;
+	return 0;
+}
+
+static const struct file_operations psp_dump_spirom_debugfs_ops = {
+	.owner = THIS_MODULE,
+	.open = psp_read_spirom_debugfs_open,
+	.read = psp_read_spirom_debugfs_read,
+	.release = psp_read_spirom_debugfs_release,
+	.llseek = default_llseek,
+};
+#endif
+
+void amdgpu_psp_debugfs_init(struct amdgpu_device *adev)
+{
+#if defined(CONFIG_DEBUG_FS)
+	struct drm_minor *minor = adev_to_drm(adev)->primary;
+
+	debugfs_create_file_size("psp_spirom_dump", 0444, minor->debugfs_root,
+				 adev, &psp_dump_spirom_debugfs_ops, AMD_VBIOS_FILE_MAX_SIZE_B * 2);
+#endif
+}
+
 const struct amd_ip_funcs psp_ip_funcs = {
 	.name = "psp",
 	.early_init = psp_early_init,