Merge tag 'drm-fixes-2024-10-18' of https://gitlab.freedesktop.org/drm/kernel

Pull drm fixes from Dave Airlie:
 "Weekly fixes, msm and xe are the two main ones, with a bunch of
  scattered fixes including a largish revert in mgag200, then amdgpu,
  vmwgfx and scattering of other minor ones.

  All seems pretty regular.

  msm:
   - Display:
      - move CRTC resource assignment to atomic_check otherwise to make
        consecutive calls to atomic_check() consistent
      - fix rounding / sign-extension issues with pclk calculation in
        case of DSC
      - cleanups to drop incorrect null checks in dpu snapshots
      - fix to use kvzalloc in dpu snapshot to avoid allocation issues
        in heavily loaded system cases
      - Fix to not program merge_3d block if dual LM is not being used
      - Fix to not flush merge_3d block if its not enabled otherwise
        this leads to false timeouts
   - GPU:
      - a7xx: add a fence wait before SMMU table update

  xe:
   - New workaround to Xe2 (Aradhya)
   - Fix unbalanced rpm put (Matthew Auld)
   - Remove fragile lock optimization (Matthew Brost)
   - Fix job release, delegating it to the drm scheduler (Matthew Brost)
   - Fix timestamp bit width for Xe2 (Lucas)
   - Fix external BO's dma-resv usag (Matthew Brost)
   - Fix returning success for timeout in wait_token (Nirmoy)
   - Initialize fence to avoid it being detected as signaled (Matthew
     Auld)
   - Improve cache flush for BMG (Matthew Auld)
   - Don't allow hflip for tile4 framebuffer on Xe2 (Juha-Pekka)

  amdgpu:
   - SR-IOV fix
   - CS chunk handling fix
   - MES fixes
   - SMU13 fixes

  amdkfd:
   - VRAM usage reporting fix

  radeon:
   - Fix possible_clones handling

  i915:
   - Two DP bandwidth related MST fixes

  ast:
   - Clear EDID on unplugged connectors

  host1x:
   - Fix boot on Tegra186
   - Set DMA parameters

  mgag200:
   - Revert VBLANK support

  panel:
   - himax-hx83192: Adjust power and gamma

  qaic:
   - Sgtable loop fixes

  vmwgfx:
   - Limit display layout allocatino size
   - Handle allocation errors in connector checks
   - Clean up KMS code for 2d-only setup
   - Report surface-check errors correctly
   - Remove NULL test around kvfree()"

* tag 'drm-fixes-2024-10-18' of https://gitlab.freedesktop.org/drm/kernel: (45 commits)
  drm/ast: vga: Clear EDID if no display is connected
  drm/ast: sil164: Clear EDID if no display is connected
  Revert "drm/mgag200: Add vblank support"
  drm/amdgpu/swsmu: default to fullscreen 3D profile for dGPUs
  drm/i915/display: Don't allow tile4 framebuffer to do hflip on display20 or greater
  drm/xe/bmg: improve cache flushing behaviour
  drm/xe/xe_sync: initialise ufence.signalled
  drm/xe/ufence: ufence can be signaled right after wait_woken
  drm/xe: Use bookkeep slots for external BO's in exec IOCTL
  drm/xe/query: Increase timestamp width
  drm/xe: Don't free job in TDR
  drm/xe: Take job list lock in xe_sched_add_pending_job
  drm/xe: fix unbalanced rpm put() with declare_wedged()
  drm/xe: fix unbalanced rpm put() with fence_fini()
  drm/xe/xe2lpg: Extend Wa_15016589081 for xe2lpg
  drm/i915/dp_mst: Don't require DSC hblank quirk for a non-DSC compatible mode
  drm/i915/dp_mst: Handle error during DSC BW overhead/slice calculation
  drm/msm/a6xx+: Insert a fence wait before SMMU table update
  drm/msm/dpu: don't always program merge_3d block
  drm/msm/dpu: Don't always set merge_3d pending flush
  ...

Author: torvalds

drivers/accel/qaic/qaic_control.c

@@ -496,7 +496,7 @@ static int encode_addr_size_pairs(struct dma_xfer *xfer, struct wrapper_list *wr
 	nents = sgt->nents;
 	nents_dma = nents;
 	*size = QAIC_MANAGE_EXT_MSG_LENGTH - msg_hdr_len - sizeof(**out_trans);
-	for_each_sgtable_sg(sgt, sg, i) {
+	for_each_sgtable_dma_sg(sgt, sg, i) {
 		*size -= sizeof(*asp);
 		/* Save 1K for possible follow-up transactions. */
 		if (*size < SZ_1K) {

drivers/accel/qaic/qaic_data.c

@@ -184,7 +184,7 @@ static int clone_range_of_sgt_for_slice(struct qaic_device *qdev, struct sg_tabl
 	nents = 0;
 
 	size = size ? size : PAGE_SIZE;
-	for (sg = sgt_in->sgl; sg; sg = sg_next(sg)) {
+	for_each_sgtable_dma_sg(sgt_in, sg, j) {
 		len = sg_dma_len(sg);
 
 		if (!len)
@@ -221,7 +221,7 @@ static int clone_range_of_sgt_for_slice(struct qaic_device *qdev, struct sg_tabl
 
 	/* copy relevant sg node and fix page and length */
 	sgn = sgf;
-	for_each_sgtable_sg(sgt, sg, j) {
+	for_each_sgtable_dma_sg(sgt, sg, j) {
 		memcpy(sg, sgn, sizeof(*sg));
 		if (sgn == sgf) {
 			sg_dma_address(sg) += offf;
@@ -301,7 +301,7 @@ static int encode_reqs(struct qaic_device *qdev, struct bo_slice *slice,
 	 * fence.
 	 */
 	dev_addr = req->dev_addr;
-	for_each_sgtable_sg(slice->sgt, sg, i) {
+	for_each_sgtable_dma_sg(slice->sgt, sg, i) {
 		slice->reqs[i].cmd = cmd;
 		slice->reqs[i].src_addr = cpu_to_le64(slice->dir == DMA_TO_DEVICE ?
 						      sg_dma_address(sg) : dev_addr);

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

@@ -265,7 +265,7 @@ static int amdgpu_cs_pass1(struct amdgpu_cs_parser *p,
 
 			/* Only a single BO list is allowed to simplify handling. */
 			if (p->bo_list)
-				ret = -EINVAL;
+				goto free_partial_kdata;
 
 			ret = amdgpu_cs_p1_bo_handles(p, p->chunks[i].kdata);
 			if (ret)

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

@@ -1635,11 +1635,9 @@ int amdgpu_gfx_sysfs_isolation_shader_init(struct amdgpu_device *adev)
 {
 	int r;
 
-	if (!amdgpu_sriov_vf(adev)) {
-		r = device_create_file(adev->dev, &dev_attr_enforce_isolation);
-		if (r)
-			return r;
-	}
+	r = device_create_file(adev->dev, &dev_attr_enforce_isolation);
+	if (r)
+		return r;
 
 	r = device_create_file(adev->dev, &dev_attr_run_cleaner_shader);
 	if (r)
@@ -1650,8 +1648,7 @@ int amdgpu_gfx_sysfs_isolation_shader_init(struct amdgpu_device *adev)
 
 void amdgpu_gfx_sysfs_isolation_shader_fini(struct amdgpu_device *adev)
 {
-	if (!amdgpu_sriov_vf(adev))
-		device_remove_file(adev->dev, &dev_attr_enforce_isolation);
+	device_remove_file(adev->dev, &dev_attr_enforce_isolation);
 	device_remove_file(adev->dev, &dev_attr_run_cleaner_shader);
 }
 

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

@@ -1203,8 +1203,10 @@ int amdgpu_mes_add_ring(struct amdgpu_device *adev, int gang_id,
 
 	r = amdgpu_ring_init(adev, ring, 1024, NULL, 0,
 			     AMDGPU_RING_PRIO_DEFAULT, NULL);
-	if (r)
+	if (r) {
+		amdgpu_mes_unlock(&adev->mes);
 		goto clean_up_memory;
+	}
 
 	amdgpu_mes_ring_to_queue_props(adev, ring, &qprops);
 
@@ -1237,7 +1239,6 @@ int amdgpu_mes_add_ring(struct amdgpu_device *adev, int gang_id,
 	amdgpu_ring_fini(ring);
 clean_up_memory:
 	kfree(ring);
-	amdgpu_mes_unlock(&adev->mes);
 	return r;
 }
 

drivers/gpu/drm/amd/amdgpu/mes_v12_0.c

@@ -621,7 +621,7 @@ static int mes_v12_0_set_hw_resources(struct amdgpu_mes *mes, int pipe)
 
 	if (amdgpu_mes_log_enable) {
 		mes_set_hw_res_pkt.enable_mes_event_int_logging = 1;
-		mes_set_hw_res_pkt.event_intr_history_gpu_mc_ptr = mes->event_log_gpu_addr;
+		mes_set_hw_res_pkt.event_intr_history_gpu_mc_ptr = mes->event_log_gpu_addr + pipe * AMDGPU_MES_LOG_BUFFER_SIZE;
 	}
 
 	return mes_v12_0_submit_pkt_and_poll_completion(mes, pipe,
@@ -1336,7 +1336,7 @@ static int mes_v12_0_sw_init(void *handle)
 	adev->mes.kiq_hw_fini = &mes_v12_0_kiq_hw_fini;
 	adev->mes.enable_legacy_queue_map = true;
 
-	adev->mes.event_log_size = AMDGPU_MES_LOG_BUFFER_SIZE;
+	adev->mes.event_log_size = adev->enable_uni_mes ? (AMDGPU_MAX_MES_PIPES * AMDGPU_MES_LOG_BUFFER_SIZE) : AMDGPU_MES_LOG_BUFFER_SIZE;
 
 	r = amdgpu_mes_init(adev);
 	if (r)

drivers/gpu/drm/amd/amdkfd/kfd_chardev.c

@@ -1148,7 +1148,7 @@ static int kfd_ioctl_alloc_memory_of_gpu(struct file *filep,
 
 		if (flags & KFD_IOC_ALLOC_MEM_FLAGS_AQL_QUEUE_MEM)
 			size >>= 1;
-		WRITE_ONCE(pdd->vram_usage, pdd->vram_usage + PAGE_ALIGN(size));
+		atomic64_add(PAGE_ALIGN(size), &pdd->vram_usage);
 	}
 
 	mutex_unlock(&p->mutex);
@@ -1219,7 +1219,7 @@ static int kfd_ioctl_free_memory_of_gpu(struct file *filep,
 		kfd_process_device_remove_obj_handle(
 			pdd, GET_IDR_HANDLE(args->handle));
 
-	WRITE_ONCE(pdd->vram_usage, pdd->vram_usage - size);
+	atomic64_sub(size, &pdd->vram_usage);
 
 err_unlock:
 err_pdd:
@@ -2347,7 +2347,7 @@ static int criu_restore_memory_of_gpu(struct kfd_process_device *pdd,
 	} else if (bo_bucket->alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_VRAM) {
 		bo_bucket->restored_offset = offset;
 		/* Update the VRAM usage count */
-		WRITE_ONCE(pdd->vram_usage, pdd->vram_usage + bo_bucket->size);
+		atomic64_add(bo_bucket->size, &pdd->vram_usage);
 	}
 	return 0;
 }

drivers/gpu/drm/amd/amdkfd/kfd_priv.h

@@ -775,7 +775,7 @@ struct kfd_process_device {
 	enum kfd_pdd_bound bound;
 
 	/* VRAM usage */
-	uint64_t vram_usage;
+	atomic64_t vram_usage;
 	struct attribute attr_vram;
 	char vram_filename[MAX_SYSFS_FILENAME_LEN];
 

drivers/gpu/drm/amd/amdkfd/kfd_process.c

@@ -332,7 +332,7 @@ static ssize_t kfd_procfs_show(struct kobject *kobj, struct attribute *attr,
 	} else if (strncmp(attr->name, "vram_", 5) == 0) {
 		struct kfd_process_device *pdd = container_of(attr, struct kfd_process_device,
 							      attr_vram);
-		return snprintf(buffer, PAGE_SIZE, "%llu\n", READ_ONCE(pdd->vram_usage));
+		return snprintf(buffer, PAGE_SIZE, "%llu\n", atomic64_read(&pdd->vram_usage));
 	} else if (strncmp(attr->name, "sdma_", 5) == 0) {
 		struct kfd_process_device *pdd = container_of(attr, struct kfd_process_device,
 							      attr_sdma);
@@ -1625,7 +1625,7 @@ struct kfd_process_device *kfd_create_process_device_data(struct kfd_node *dev,
 	pdd->bound = PDD_UNBOUND;
 	pdd->already_dequeued = false;
 	pdd->runtime_inuse = false;
-	pdd->vram_usage = 0;
+	atomic64_set(&pdd->vram_usage, 0);
 	pdd->sdma_past_activity_counter = 0;
 	pdd->user_gpu_id = dev->id;
 	atomic64_set(&pdd->evict_duration_counter, 0);

drivers/gpu/drm/amd/amdkfd/kfd_svm.c

@@ -405,6 +405,27 @@ static void svm_range_bo_release(struct kref *kref)
 		spin_lock(&svm_bo->list_lock);
 	}
 	spin_unlock(&svm_bo->list_lock);
+
+	if (mmget_not_zero(svm_bo->eviction_fence->mm)) {
+		struct kfd_process_device *pdd;
+		struct kfd_process *p;
+		struct mm_struct *mm;
+
+		mm = svm_bo->eviction_fence->mm;
+		/*
+		 * The forked child process takes svm_bo device pages ref, svm_bo could be
+		 * released after parent process is gone.
+		 */
+		p = kfd_lookup_process_by_mm(mm);
+		if (p) {
+			pdd = kfd_get_process_device_data(svm_bo->node, p);
+			if (pdd)
+				atomic64_sub(amdgpu_bo_size(svm_bo->bo), &pdd->vram_usage);
+			kfd_unref_process(p);
+		}
+		mmput(mm);
+	}
+
 	if (!dma_fence_is_signaled(&svm_bo->eviction_fence->base))
 		/* We're not in the eviction worker. Signal the fence. */
 		dma_fence_signal(&svm_bo->eviction_fence->base);
@@ -532,6 +553,7 @@ int
 svm_range_vram_node_new(struct kfd_node *node, struct svm_range *prange,
 			bool clear)
 {
+	struct kfd_process_device *pdd;
 	struct amdgpu_bo_param bp;
 	struct svm_range_bo *svm_bo;
 	struct amdgpu_bo_user *ubo;
@@ -623,6 +645,10 @@ svm_range_vram_node_new(struct kfd_node *node, struct svm_range *prange,
 	list_add(&prange->svm_bo_list, &svm_bo->range_list);
 	spin_unlock(&svm_bo->list_lock);
 
+	pdd = svm_range_get_pdd_by_node(prange, node);
+	if (pdd)
+		atomic64_add(amdgpu_bo_size(bo), &pdd->vram_usage);
+
 	return 0;
 
 reserve_bo_failed: