7 files changed, 26 insertions, 29 deletions

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
index 5feaba6a77de..6d291aa6386b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_chardev.c
@@ -1950,7 +1950,7 @@ static int criu_checkpoint(struct file *filep,
 {
 	int ret;
 	uint32_t num_devices, num_bos, num_objects;
-	uint64_t priv_size, priv_offset = 0;
+	uint64_t priv_size, priv_offset = 0, bo_priv_offset;
 
 	if (!args->devices || !args->bos || !args->priv_data)
 		return -EINVAL;
@@ -1994,38 +1994,34 @@ static int criu_checkpoint(struct file *filep,
 	if (ret)
 		goto exit_unlock;
 
-	ret = criu_checkpoint_bos(p, num_bos, (uint8_t __user *)args->bos,
-			    (uint8_t __user *)args->priv_data, &priv_offset);
-	if (ret)
-		goto exit_unlock;
+	/* Leave room for BOs in the private data. They need to be restored
+	 * before events, but we checkpoint them last to simplify the error
+	 * handling.
+	 */
+	bo_priv_offset = priv_offset;
+	priv_offset += num_bos * sizeof(struct kfd_criu_bo_priv_data);
 
 	if (num_objects) {
 		ret = kfd_criu_checkpoint_queues(p, (uint8_t __user *)args->priv_data,
 						 &priv_offset);
 		if (ret)
-			goto close_bo_fds;
+			goto exit_unlock;
 
 		ret = kfd_criu_checkpoint_events(p, (uint8_t __user *)args->priv_data,
 						 &priv_offset);
 		if (ret)
-			goto close_bo_fds;
+			goto exit_unlock;
 
 		ret = kfd_criu_checkpoint_svm(p, (uint8_t __user *)args->priv_data, &priv_offset);
 		if (ret)
-			goto close_bo_fds;
+			goto exit_unlock;
 	}
 
-close_bo_fds:
-	if (ret) {
-		/* If IOCTL returns err, user assumes all FDs opened in criu_dump_bos are closed */
-		uint32_t i;
-		struct kfd_criu_bo_bucket *bo_buckets = (struct kfd_criu_bo_bucket *) args->bos;
-
-		for (i = 0; i < num_bos; i++) {
-			if (bo_buckets[i].alloc_flags & KFD_IOC_ALLOC_MEM_FLAGS_VRAM)
-				close_fd(bo_buckets[i].dmabuf_fd);
-		}
-	}
+	/* This must be the last thing in this function that can fail.
+	 * Otherwise we leak dmabuf file descriptors.
+	 */
+	ret = criu_checkpoint_bos(p, num_bos, (uint8_t __user *)args->bos,
+			   (uint8_t __user *)args->priv_data, &bo_priv_offset);
 
 exit_unlock:
 	mutex_unlock(&p->mutex);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
index 267dd69737fa..af01ba061e1b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
@@ -1111,7 +1111,7 @@ static int kfd_parse_subtype_cache(struct crat_subtype_cache *cache,
 			props->cache_latency = cache->cache_latency;
 
 			memcpy(props->sibling_map, cache->sibling_map,
-					sizeof(props->sibling_map));
+					CRAT_SIBLINGMAP_SIZE);
 
 			/* set the sibling_map_size as 32 for CRAT from ACPI */
 			props->sibling_map_size = CRAT_SIBLINGMAP_SIZE;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index ee8e5f8b007d..c8dd948966c2 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -495,7 +495,10 @@ static int kfd_gws_init(struct kfd_dev *kfd)
 		(KFD_GC_VERSION(kfd) == IP_VERSION(9, 4, 1)
 			&& kfd->mec2_fw_version >= 0x30)   ||
 		(KFD_GC_VERSION(kfd) == IP_VERSION(9, 4, 2)
-			&& kfd->mec2_fw_version >= 0x28))))
+			&& kfd->mec2_fw_version >= 0x28) ||
+		(KFD_GC_VERSION(kfd) >= IP_VERSION(10, 3, 0)
+			&& KFD_GC_VERSION(kfd) < IP_VERSION(11, 0, 0)
+			&& kfd->mec2_fw_version >= 0x6b))))
 		ret = amdgpu_amdkfd_alloc_gws(kfd->adev,
 				kfd->adev->gds.gws_size, &kfd->gws);
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_events.c b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
index 83e3ce9f6049..729d26d648af 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_events.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_events.c
@@ -506,6 +506,7 @@ int kfd_criu_restore_event(struct file *devkfd,
 		ret = create_other_event(p, ev, &ev_priv->event_id);
 		break;
 	}
+	mutex_unlock(&p->event_mutex);
 
 exit:
 	if (ret)
@@ -513,8 +514,6 @@ exit:
 
 	kfree(ev_priv);
 
-	mutex_unlock(&p->event_mutex);
-
 	return ret;
 }
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
index c9a95a816912..10048ce16aea 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
@@ -28,7 +28,6 @@
 #include "amdgpu_sync.h"
 #include "amdgpu_object.h"
 #include "amdgpu_vm.h"
-#include "amdgpu_mn.h"
 #include "amdgpu_res_cursor.h"
 #include "kfd_priv.h"
 #include "kfd_svm.h"
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
index afe7c4998676..814f99888ab1 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
@@ -26,7 +26,7 @@
 #include "amdgpu_sync.h"
 #include "amdgpu_object.h"
 #include "amdgpu_vm.h"
-#include "amdgpu_mn.h"
+#include "amdgpu_hmm.h"
 #include "amdgpu.h"
 #include "amdgpu_xgmi.h"
 #include "kfd_priv.h"
@@ -1596,9 +1596,9 @@ static int svm_range_validate_and_map(struct mm_struct *mm,
 		next = min(vma->vm_end, end);
 		npages = (next - addr) >> PAGE_SHIFT;
 		WRITE_ONCE(p->svms.faulting_task, current);
-		r = amdgpu_hmm_range_get_pages(&prange->notifier, mm, NULL,
-					       addr, npages, &hmm_range,
-					       readonly, true, owner);
+		r = amdgpu_hmm_range_get_pages(&prange->notifier, addr, npages,
+					       readonly, owner, NULL,
+					       &hmm_range);
 		WRITE_ONCE(p->svms.faulting_task, NULL);
 		if (r) {
 			pr_debug("failed %d to get svm range pages\n", r);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index 1d9b90d979c4..ef9c6fdfb88d 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -1723,7 +1723,7 @@ static int fill_in_l2_l3_pcache(struct kfd_cache_properties **props_ext,
 /* kfd_fill_cache_non_crat_info - Fill GPU cache info using kfd_gpu_cache_info
  * tables
  */
-void kfd_fill_cache_non_crat_info(struct kfd_topology_device *dev, struct kfd_dev *kdev)
+static void kfd_fill_cache_non_crat_info(struct kfd_topology_device *dev, struct kfd_dev *kdev)
 {
 	struct kfd_gpu_cache_info *pcache_info = NULL;
 	int i, j, k;