drm/virtio: use kvmalloc for large allocations

commit ea86f3defd55f141a44146e66cbf8ffb683d60da upstream.

We observed that some of virtio_gpu_object_shmem_init() allocations
can be rather costly - order 6 - which can be difficult to fulfill
under memory pressure conditions. Switch to kvmalloc_array() in
virtio_gpu_object_shmem_init() and let the kernel vmalloc the entries
array.

Signed-off-by: Sergey Senozhatsky <senozhatsky@chromium.org>
Link: http://patchwork.freedesktop.org/patch/msgid/20201105014744.1662226-1-senozhatsky@chromium.org
Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
Signed-off-by: Doug Horn <doughorn@google.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

diff --git a/drivers/gpu/drm/virtio/virtgpu_vq.c b/drivers/gpu/drm/virtio/virtgpu_vq.c
index 92022a83bbd5e684b775f7d96ca318f627497cbf..bb46e7a0f1b5dee8eba51da9c24b9fb9bf01db4d 100644 (file)
--- a/drivers/gpu/drm/virtio/virtgpu_vq.c
+++ b/drivers/gpu/drm/virtio/virtgpu_vq.c
@@ -992,8 +992,9 @@ int virtio_gpu_object_attach(struct virtio_gpu_device *vgdev,
        }
 
        /* gets freed when the ring has consumed it */
-       ents = kmalloc_array(nents, sizeof(struct virtio_gpu_mem_entry),
-                            GFP_KERNEL);
+       ents = kvmalloc_array(nents,
+                             sizeof(struct virtio_gpu_mem_entry),
+                             GFP_KERNEL);
        if (!ents) {
                DRM_ERROR("failed to allocate ent list\n");
                return -ENOMEM;