diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
index 0767559947b5f..cda61f7924d1e 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_mn.c
@@ -272,7 +272,7 @@ static void amdgpu_mn_invalidate_node(struct amdgpu_mn_node *node,
 
 	list_for_each_entry(bo, &node->bos, mn_list) {
 
-		if (!amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm, start, end))
+		if (!amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm, start, end, NULL))
 			continue;
 
 		r = dma_resv_wait_timeout_rcu(amdkcl_ttm_resvp(&bo->tbo),
@@ -361,7 +361,7 @@ static int amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn,
 
 			if (amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm,
 							 range->start,
-							 end))
+							 end, NULL))
 				amdgpu_amdkfd_evict_userptr(mem, range->mm);
 		}
 	}
@@ -474,7 +474,7 @@ static void amdgpu_mn_invalidate_range_start_hsa(struct mmu_notifier *mn,
 			struct kgd_mem *mem = bo->kfd_bo;
 
 			if (amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm,
-							 start, end))
+							 start, end, NULL))
 				amdgpu_amdkfd_evict_userptr(mem, mm);
 		}
 	}
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
index 8bae747023516..eb732cc94016d 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.c
@@ -1456,7 +1456,7 @@ struct mm_struct *amdgpu_ttm_tt_get_usermm(struct ttm_tt *ttm)
  *
  */
 bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
-				  unsigned long end)
+				  unsigned long end, unsigned long *userptr)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
 	unsigned long size;
@@ -1471,6 +1471,8 @@ bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
 	if (gtt->userptr > end || gtt->userptr + size <= start)
 		return false;
 
+	if (userptr)
+		*userptr = gtt->userptr;
 	return true;
 }
 
@@ -1494,7 +1496,7 @@ bool amdgpu_ttm_tt_is_userptr(struct ttm_tt *ttm)
  *
  */
 bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
-				  unsigned long end)
+				  unsigned long end, unsigned long *userptr)
 {
 	struct amdgpu_ttm_tt *gtt = (void *)ttm;
 	struct amdgpu_ttm_gup_task_list *entry;
@@ -1524,6 +1526,8 @@ bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
 
 	atomic_inc(&gtt->mmu_invalidations);
 
+	if (userptr)
+		*userptr = gtt->userptr;
 	return true;
 }
 
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
index ff9989ed0d7e1..b8062f60eae68 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ttm.h
@@ -193,7 +193,7 @@ int amdgpu_ttm_tt_set_userptr(struct ttm_buffer_object *bo,
 bool amdgpu_ttm_tt_has_userptr(struct ttm_tt *ttm);
 struct mm_struct *amdgpu_ttm_tt_get_usermm(struct ttm_tt *ttm);
 bool amdgpu_ttm_tt_affect_userptr(struct ttm_tt *ttm, unsigned long start,
-				  unsigned long end);
+				  unsigned long end, unsigned long *userptr);
 #ifdef HAVE_AMDKCL_HMM_MIRROR_ENABLED
 bool amdgpu_ttm_tt_is_userptr(struct ttm_tt *ttm);
 #else
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
index c89eaa870bce3..4192879e67bad 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
@@ -50,7 +50,9 @@ static bool
 svm_range_cpu_invalidate_pagetables(struct mmu_interval_notifier *mni,
 				    const struct mmu_notifier_range *range,
 				    unsigned long cur_seq);
-
+static int
+svm_range_check_vm(struct kfd_process *p, uint64_t start, uint64_t last,
+		   uint64_t *bo_s, uint64_t *bo_l);
 static const struct mmu_interval_notifier_ops svm_range_mn_ops = {
 	.invalidate = svm_range_cpu_invalidate_pagetables,
 };
@@ -2304,6 +2306,7 @@ svm_range_best_restore_location(struct svm_range *prange,
 
 	return -1;
 }
+
 static int
 svm_range_get_range_boundaries(struct kfd_process *p, int64_t addr,
 				unsigned long *start, unsigned long *last)
@@ -2351,8 +2354,59 @@ svm_range_get_range_boundaries(struct kfd_process *p, int64_t addr,
 		  vma->vm_end >> PAGE_SHIFT, *last);
 
 	return 0;
+}
+
+static int
+svm_range_check_vm_userptr(struct kfd_process *p, uint64_t start, uint64_t last,
+			   uint64_t *bo_s, uint64_t *bo_l)
+{
+	struct amdgpu_bo_va_mapping *mapping;
+	struct interval_tree_node *node;
+	struct amdgpu_bo *bo = NULL;
+	unsigned long userptr;
+	uint32_t i;
+	int r;
 
+	for (i = 0; i < p->n_pdds; i++) {
+		struct amdgpu_vm *vm;
+
+		if (!p->pdds[i]->drm_priv)
+			continue;
+
+		vm = drm_priv_to_vm(p->pdds[i]->drm_priv);
+		r = amdgpu_bo_reserve(vm->root.bo, false);
+		if (r)
+			return r;
+
+		/* Check userptr by searching entire vm->va interval tree */
+		node = interval_tree_iter_first(&vm->va, 0, ~0ULL);
+		while (node) {
+			mapping = container_of((struct rb_node *)node,
+					       struct amdgpu_bo_va_mapping, rb);
+			bo = mapping->bo_va->base.bo;
+
+			if (!amdgpu_ttm_tt_affect_userptr(bo->tbo.ttm,
+							 start << PAGE_SHIFT,
+							 last << PAGE_SHIFT,
+							 &userptr)) {
+				node = interval_tree_iter_next(node, 0, ~0ULL);
+				continue;
+			}
+
+			pr_debug("[0x%llx 0x%llx] already userptr mapped\n",
+				 start, last);
+			if (bo_s && bo_l) {
+				*bo_s = userptr >> PAGE_SHIFT;
+				*bo_l = *bo_s + bo->tbo.ttm->num_pages - 1;
+			}
+			amdgpu_bo_unreserve(vm->root.bo);
+			return -EADDRINUSE;
+		}
+		amdgpu_bo_unreserve(vm->root.bo);
+	}
+	return 0;
 }
+
 static struct
 svm_range *svm_range_create_unregistered_range(struct amdgpu_device *adev,
 						struct kfd_process *p,
@@ -2362,10 +2416,26 @@ svm_range *svm_range_create_unregistered_range(struct amdgpu_device *adev,
 	struct svm_range *prange = NULL;
 	unsigned long start, last;
 	uint32_t gpuid, gpuidx;
+	uint64_t bo_s = 0;
+	uint64_t bo_l = 0;
+	int r;
 
 	if (svm_range_get_range_boundaries(p, addr, &start, &last))
 		return NULL;
 
+	r = svm_range_check_vm(p, start, last, &bo_s, &bo_l);
+	if (r != -EADDRINUSE)
+		r = svm_range_check_vm_userptr(p, start, last, &bo_s, &bo_l);
+
+	if (r == -EADDRINUSE) {
+		if (addr >= bo_s && addr <= bo_l)
+			return NULL;
+
+		/* Create one page svm range if 2MB range overlapping */
+		start = addr;
+		last = addr;
+	}
+
 	prange = svm_range_new(&p->svms, start, last);
 	if (!prange) {
 		pr_debug("Failed to create prange in address [0x%llx]\n", addr);
@@ -2668,6 +2738,8 @@ int svm_range_list_init(struct kfd_process *p)
  * @p: current kfd_process
  * @start: range start address, in pages
  * @last: range last address, in pages
+ * @bo_s: mapping start address in pages if address range already mapped
+ * @bo_l: mapping last address in pages if address range already mapped
  *
  * The purpose is to avoid virtual address ranges already allocated by
  * kfd_ioctl_alloc_memory_of_gpu ioctl.
@@ -2682,8 +2754,11 @@ int svm_range_list_init(struct kfd_process *p)
  * a signal. Release all buffer reservations and return to user-space.
  */
 static int
-svm_range_check_vm(struct kfd_process *p, uint64_t start, uint64_t last)
+svm_range_check_vm(struct kfd_process *p, uint64_t start, uint64_t last,
+		   uint64_t *bo_s, uint64_t *bo_l)
 {
+	struct amdgpu_bo_va_mapping *mapping;
+	struct interval_tree_node *node;
 	uint32_t i;
 	int r;
 
@@ -2697,8 +2772,17 @@ svm_range_check_vm(struct kfd_process *p, uint64_t start, uint64_t last)
 		r = amdgpu_bo_reserve(vm->root.bo, false);
 		if (r)
 			return r;
-		if (interval_tree_iter_first(&vm->va, start, last)) {
-			pr_debug("Range [0x%llx 0x%llx] already mapped\n", start, last);
+
+		node = interval_tree_iter_first(&vm->va, start, last);
+		if (node) {
+			pr_debug("range [0x%llx 0x%llx] already TTM mapped\n",
+				 start, last);
+			mapping = container_of((struct rb_node *)node,
+					       struct amdgpu_bo_va_mapping, rb);
+			if (bo_s && bo_l) {
+				*bo_s = mapping->start;
+				*bo_l = mapping->last;
+			}
 			amdgpu_bo_unreserve(vm->root.bo);
 			return -EADDRINUSE;
 		}
@@ -2739,7 +2823,8 @@ svm_range_is_valid(struct kfd_process *p, uint64_t start, uint64_t size)
 		start = min(end, vma->vm_end);
 	} while (start < end);
 
-	return svm_range_check_vm(p, start_unchg, (end - 1) >> PAGE_SHIFT);
+	return svm_range_check_vm(p, start_unchg, (end - 1) >> PAGE_SHIFT, NULL,
+				  NULL);
 }
 
 /**