[Linaro-mm-sig] Re: [PATCH 7/9] vfio/pci: Support mmap() of a VFIO DMABUF

On Thu, Apr 16, 2026 at 06:17:50AM -0700, Matt Evans wrote:

• dma_resv_lock(priv->dmabuf->resv, NULL); vdev = READ_ONCE(priv->vdev);

• if (READ_ONCE(priv->revoked) || !vdev) {

Why is this read once? It is inside the resv lock so it is stable?

• pr_debug_ratelimited("%s VA 0x%lx, pgoff 0x%lx: DMABUF revoked/cleaned up\n",
  

• 		     __func__, vmf->address, vma->vm_pgoff);
  

• dma_resv_unlock(priv->dmabuf->resv);
  

• return VM_FAULT_SIGBUS;
  

• }
• /* vdev is usable */
• if (!vfio_device_try_get_registration(&vdev->vdev)) {

• /*
  

•  * If vdev != NULL (above), the registration should
  

•  * already be >0 and so this try_get should never
  

•  * fail.
  

•  */
  

• dev_warn(&vdev->pdev->dev, "%s: Unexpected registration failure\n",
  

• 	 __func__);
  

• dma_resv_unlock(priv->dmabuf->resv);
  

• return VM_FAULT_SIGBUS;
  

• }
• dma_resv_unlock(priv->dmabuf->resv);
• scoped_guard(rwsem_read, &vdev->memory_lock) {

• if (!priv->revoked) {
  

• if (!READ_ONCE(priv->revoked)) {
  

Same here, it is not read once since you hold the memory_lock it is stable.

Jason