Merge tag 'amd-drm-fixes-6.5-2023-07-20' of https://gitlab.freedesktop.org/agd5f/linux into drm-fixes

amd-drm-fixes-6.5-2023-07-20:

amdgpu:
- More PCIe DPM fixes for Intel platforms
- DCN3.0.1 fixes
- Virtual display timer fix
- Async flip fix
- SMU13 clock reporting fixes
- Add missing PSP firmware declaration
- DP MST fix
- DCN3.1.x fixes
- Slab out of bounds fix

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

From: Alex Deucher <alexander.deucher@amd.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20230720133456.7826-1-alexander.deucher@amd.com

Author: airlied

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

@@ -1709,7 +1709,8 @@ int amdgpu_amdkfd_gpuvm_alloc_memory_of_gpu(
 			alloc_flags |= (flags & KFD_IOC_ALLOC_MEM_FLAGS_PUBLIC) ?
 			AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED : 0;
 		}
-		xcp_id = fpriv->xcp_id == ~0 ? 0 : fpriv->xcp_id;
+		xcp_id = fpriv->xcp_id == AMDGPU_XCP_NO_PARTITION ?
+					0 : fpriv->xcp_id;
 	} else if (flags & KFD_IOC_ALLOC_MEM_FLAGS_GTT) {
 		domain = alloc_domain = AMDGPU_GEM_DOMAIN_GTT;
 		alloc_flags = 0;

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

@@ -1229,13 +1229,13 @@ int amdgpu_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 		pasid = 0;
 	}
 
-	r = amdgpu_vm_init(adev, &fpriv->vm);
+	r = amdgpu_xcp_open_device(adev, fpriv, file_priv);
 	if (r)
 		goto error_pasid;
 
-	r = amdgpu_xcp_open_device(adev, fpriv, file_priv);
+	r = amdgpu_vm_init(adev, &fpriv->vm, fpriv->xcp_id);
 	if (r)
-		goto error_vm;
+		goto error_pasid;
 
 	r = amdgpu_vm_set_pasid(adev, &fpriv->vm, pasid);
 	if (r)

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

@@ -1382,7 +1382,7 @@ int amdgpu_mes_self_test(struct amdgpu_device *adev)
 		goto error_pasid;
 	}
 
-	r = amdgpu_vm_init(adev, vm);
+	r = amdgpu_vm_init(adev, vm, -1);
 	if (r) {
 		DRM_ERROR("failed to initialize vm\n");
 		goto error_pasid;

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

@@ -55,8 +55,9 @@ static enum hrtimer_restart amdgpu_vkms_vblank_simulate(struct hrtimer *timer)
 		DRM_WARN("%s: vblank timer overrun\n", __func__);
 
 	ret = drm_crtc_handle_vblank(crtc);
+	/* Don't queue timer again when vblank is disabled. */
 	if (!ret)
-		DRM_ERROR("amdgpu_vkms failure on handling vblank");
+		return HRTIMER_NORESTART;
 
 	return HRTIMER_RESTART;
 }
@@ -81,7 +82,7 @@ static void amdgpu_vkms_disable_vblank(struct drm_crtc *crtc)
 {
 	struct amdgpu_crtc *amdgpu_crtc = to_amdgpu_crtc(crtc);
 
-	hrtimer_cancel(&amdgpu_crtc->vblank_timer);
+	hrtimer_try_to_cancel(&amdgpu_crtc->vblank_timer);
 }
 
 static bool amdgpu_vkms_get_vblank_timestamp(struct drm_crtc *crtc,

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

@@ -2121,13 +2121,14 @@ long amdgpu_vm_wait_idle(struct amdgpu_vm *vm, long timeout)
  *
  * @adev: amdgpu_device pointer
  * @vm: requested vm
+ * @xcp_id: GPU partition selection id
  *
  * Init @vm fields.
  *
  * Returns:
  * 0 for success, error for failure.
  */
-int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm)
+int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, int32_t xcp_id)
 {
 	struct amdgpu_bo *root_bo;
 	struct amdgpu_bo_vm *root;
@@ -2177,7 +2178,7 @@ int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm)
 	vm->evicting = false;
 
 	r = amdgpu_vm_pt_create(adev, vm, adev->vm_manager.root_level,
-				false, &root);
+				false, &root, xcp_id);
 	if (r)
 		goto error_free_delayed;
 	root_bo = &root->bo;

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

@@ -392,7 +392,7 @@ int amdgpu_vm_set_pasid(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 			u32 pasid);
 
 long amdgpu_vm_wait_idle(struct amdgpu_vm *vm, long timeout);
-int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm);
+int amdgpu_vm_init(struct amdgpu_device *adev, struct amdgpu_vm *vm, int32_t xcp_id);
 int amdgpu_vm_make_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm);
 void amdgpu_vm_release_compute(struct amdgpu_device *adev, struct amdgpu_vm *vm);
 void amdgpu_vm_fini(struct amdgpu_device *adev, struct amdgpu_vm *vm);
@@ -475,7 +475,8 @@ void amdgpu_vm_get_memory(struct amdgpu_vm *vm,
 int amdgpu_vm_pt_clear(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 		       struct amdgpu_bo_vm *vmbo, bool immediate);
 int amdgpu_vm_pt_create(struct amdgpu_device *adev, struct amdgpu_vm *vm,
-			int level, bool immediate, struct amdgpu_bo_vm **vmbo);
+			int level, bool immediate, struct amdgpu_bo_vm **vmbo,
+			int32_t xcp_id);
 void amdgpu_vm_pt_free_root(struct amdgpu_device *adev, struct amdgpu_vm *vm);
 bool amdgpu_vm_pt_is_root_clean(struct amdgpu_device *adev,
 				struct amdgpu_vm *vm);

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

@@ -498,11 +498,12 @@ int amdgpu_vm_pt_clear(struct amdgpu_device *adev, struct amdgpu_vm *vm,
  * @level: the page table level
  * @immediate: use a immediate update
  * @vmbo: pointer to the buffer object pointer
+ * @xcp_id: GPU partition id
  */
 int amdgpu_vm_pt_create(struct amdgpu_device *adev, struct amdgpu_vm *vm,
-			int level, bool immediate, struct amdgpu_bo_vm **vmbo)
+			int level, bool immediate, struct amdgpu_bo_vm **vmbo,
+			int32_t xcp_id)
 {
-	struct amdgpu_fpriv *fpriv = container_of(vm, struct amdgpu_fpriv, vm);
 	struct amdgpu_bo_param bp;
 	struct amdgpu_bo *bo;
 	struct dma_resv *resv;
@@ -535,7 +536,7 @@ int amdgpu_vm_pt_create(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 
 	bp.type = ttm_bo_type_kernel;
 	bp.no_wait_gpu = immediate;
-	bp.xcp_id_plus1 = fpriv->xcp_id == ~0 ? 0 : fpriv->xcp_id + 1;
+	bp.xcp_id_plus1 = xcp_id + 1;
 
 	if (vm->root.bo)
 		bp.resv = vm->root.bo->tbo.base.resv;
@@ -561,7 +562,7 @@ int amdgpu_vm_pt_create(struct amdgpu_device *adev, struct amdgpu_vm *vm,
 	bp.type = ttm_bo_type_kernel;
 	bp.resv = bo->tbo.base.resv;
 	bp.bo_ptr_size = sizeof(struct amdgpu_bo);
-	bp.xcp_id_plus1 = fpriv->xcp_id == ~0 ? 0 : fpriv->xcp_id + 1;
+	bp.xcp_id_plus1 = xcp_id + 1;
 
 	r = amdgpu_bo_create(adev, &bp, &(*vmbo)->shadow);
 
@@ -606,7 +607,8 @@ static int amdgpu_vm_pt_alloc(struct amdgpu_device *adev,
 		return 0;
 
 	amdgpu_vm_eviction_unlock(vm);
-	r = amdgpu_vm_pt_create(adev, vm, cursor->level, immediate, &pt);
+	r = amdgpu_vm_pt_create(adev, vm, cursor->level, immediate, &pt,
+				vm->root.bo->xcp_id);
 	amdgpu_vm_eviction_lock(vm);
 	if (r)
 		return r;

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

@@ -363,7 +363,7 @@ int amdgpu_xcp_open_device(struct amdgpu_device *adev,
 	if (!adev->xcp_mgr)
 		return 0;
 
-	fpriv->xcp_id = ~0;
+	fpriv->xcp_id = AMDGPU_XCP_NO_PARTITION;
 	for (i = 0; i < MAX_XCP; ++i) {
 		if (!adev->xcp_mgr->xcp[i].ddev)
 			break;
@@ -381,7 +381,7 @@ int amdgpu_xcp_open_device(struct amdgpu_device *adev,
 		}
 	}
 
-	fpriv->vm.mem_id = fpriv->xcp_id == ~0 ? -1 :
+	fpriv->vm.mem_id = fpriv->xcp_id == AMDGPU_XCP_NO_PARTITION ? -1 :
 				adev->xcp_mgr->xcp[fpriv->xcp_id].mem_id;
 	return 0;
 }

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

@@ -37,6 +37,8 @@
 #define AMDGPU_XCP_FL_NONE 0
 #define AMDGPU_XCP_FL_LOCKED (1 << 0)
 
+#define AMDGPU_XCP_NO_PARTITION (~0)
+
 struct amdgpu_fpriv;
 
 enum AMDGPU_XCP_IP_BLOCK {

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

@@ -68,7 +68,7 @@ static void aqua_vanjaram_set_xcp_id(struct amdgpu_device *adev,
 	enum AMDGPU_XCP_IP_BLOCK ip_blk;
 	uint32_t inst_mask;
 
-	ring->xcp_id = ~0;
+	ring->xcp_id = AMDGPU_XCP_NO_PARTITION;
 	if (adev->xcp_mgr->mode == AMDGPU_XCP_MODE_NONE)
 		return;
 
@@ -177,7 +177,7 @@ static int aqua_vanjaram_select_scheds(
 	u32 sel_xcp_id;
 	int i;
 
-	if (fpriv->xcp_id == ~0) {
+	if (fpriv->xcp_id == AMDGPU_XCP_NO_PARTITION) {
 		u32 least_ref_cnt = ~0;
 
 		fpriv->xcp_id = 0;