All of lore.kernel.org
 help / color / mirror / Atom feed
From: sashiko-bot@kernel.org
To: "Honglei Huang" <honghuan@amd.com>
Cc: dri-devel@lists.freedesktop.org
Subject: Re: [PATCH v8 4/4] drm/virtio: wire blob ioctl creation to userptr objects
Date: Fri, 18 Sep 2026 08:12:41 +0000	[thread overview]
Message-ID: <20260918081241.E5CE81F000FF@smtp.kernel.org> (raw)
In-Reply-To: <20260918075600.2197253-5-honghuan@amd.com>

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [High] drm/virtio: resource loss on resume for userptr blobs
--

commit 260e023045965d09441bfc27b3e47cf7b409d5b4
Author: Honglei Huang <honghuan@amd.com>

drm/virtio: wire blob ioctl creation to userptr objects

This commit integrates userptr into the blob resource creation ioctl.
It adds validation and wires up virtio_gpu_userptr_create() for guest
blob allocations.

> diff --git a/drivers/gpu/drm/virtio/virtgpu_ioctl.c b/drivers/gpu/drm/virtio/virtgpu_ioctl.c
> --- a/drivers/gpu/drm/virtio/virtgpu_ioctl.c
> +++ b/drivers/gpu/drm/virtio/virtgpu_ioctl.c
[ ... ]
> @@ -538,7 +561,9 @@ static int virtio_gpu_resource_create_blob_ioctl(struct drm_device *dev,
>  				      vfpriv->ctx_id, NULL, NULL);
>  	}
>  
> -	if (guest_blob)
> +	if (guest_blob && params.userptr)
> +		ret = virtio_gpu_userptr_create(vgdev, file, &params, &bo);

[Severity: High]
Does this code handle system hibernation properly for userptr objects?

When virtio_gpu_userptr_create() in drivers/gpu/drm/virtio/virtgpu_userptr.c
successfully allocates the object, it returns without saving bo->params or
calling virtio_gpu_add_object_to_restore_list():

	virtio_gpu_cmd_resource_create_blob(vgdev, &userptr->base, params, ents,
					    nents);

	*bo_ptr = &userptr->base;
	return 0;

Consequently, these objects are permanently lost when the system resumes from
hibernation because their host-side resources are not re-created. Subsequent
GPU commands referencing these resources will fail.

Additionally, if these userptr objects were added to the obj_restore_list, it
seems virtio_gpu_object_restore_all() in drivers/gpu/drm/virtio/virtgpu_object.c
would crash:

	if (bo->params.blob || bo->attached) {
		ret = virtio_gpu_object_shmem_init(vgdev, bo, &ents,
						   &nents);
		if (ret)
			break;
	}

This logic assumes all non-imported blobs are backed by drm_gem_shmem_object,
but userptr blobs are not.

> +	else if (guest_blob)
>  		ret = virtio_gpu_object_create(vgdev, &params, &bo, NULL);
>  	else if (!guest_blob && host3d_blob)
>  		ret = virtio_gpu_vram_create(vgdev, &params, &bo);

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260918075600.2197253-1-honghuan@amd.com?part=4

  reply	other threads:[~2026-09-18  8:12 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-09-18  7:55 [PATCH v8 0/4] virtio-gpu: Add userptr support for compute workloads Honglei Huang
2026-09-18  7:55 ` [PATCH v8 1/4] drm/virtio-gpu: Add VIRTIO_GPU_CAPSET_ROCM capability Honglei Huang
2026-09-18  8:07   ` sashiko-bot
2026-09-18  7:55 ` [PATCH v8 2/4] drm/virtgpu api: add blob userptr resource Honglei Huang
2026-09-18  7:55 ` [PATCH v8 3/4] drm/virtio: implement userptr support for zero-copy memory access Honglei Huang
2026-09-18  8:17   ` sashiko-bot
2026-09-18  7:56 ` [PATCH v8 4/4] drm/virtio: wire blob ioctl creation to userptr objects Honglei Huang
2026-09-18  8:12   ` sashiko-bot [this message]
  -- strict thread matches above, loose matches on Subject: below --
2026-09-18  9:59 [PATCH v8 0/4] virtio-gpu: Add userptr support for compute workloads Honglei Huang
2026-09-18  9:59 ` [PATCH v8 4/4] drm/virtio: wire blob ioctl creation to userptr objects Honglei Huang
2026-09-18 10:08   ` sashiko-bot
2026-09-19 14:28   ` Akihiko Odaki
2026-09-24  9:33     ` Huang, Honglei
2026-09-18  8:23 [PATCH v8 0/4] virtio-gpu: Add userptr support for compute workloads Honglei Huang
2026-09-18  8:23 ` [PATCH v8 4/4] drm/virtio: wire blob ioctl creation to userptr objects Honglei Huang
2026-09-18  6:56 [PATCH v8 0/4] virtio-gpu: Add userptr support for compute workloads Honglei Huang
2026-09-18  6:56 ` [PATCH v8 4/4] drm/virtio: wire blob ioctl creation to userptr objects Honglei Huang
2026-09-18  7:07   ` sashiko-bot
2026-09-18  5:59 [PATCH v8 0/4] virtio-gpu: Add userptr support for compute workloads Honglei Huang
2026-09-18  5:59 ` [PATCH v8 4/4] drm/virtio: wire blob ioctl creation to userptr objects Honglei Huang
2026-09-18  6:12   ` sashiko-bot

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20260918081241.E5CE81F000FF@smtp.kernel.org \
    --to=sashiko-bot@kernel.org \
    --cc=dri-devel@lists.freedesktop.org \
    --cc=honghuan@amd.com \
    --cc=sashiko-reviews@lists.linux.dev \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.