Re: [PATCH v3] virtio-gpu: reject requests with short/truncated control headers

Akihiko Odaki <[email protected]> Tue, 4 Aug 2026 14:57:47 +0900
Newsgroups gmane.comp.emulators.qemu
Message-ID <[email protected]>
On 2026/08/04 2:09, Ankur Saini wrote:
> A short control request can leave command data partially initialized.
> For the common header, guest-controlled flags can then cause stale fence
> metadata to be returned to the guest.
> 
> The command fill helpers detect a short copy but only log and return,
> leaving the malformed request without an error or completion. Make
> VIRTIO_GPU_FILL_CMD() clear the partially copied object and complete the
> request with ERR_INVALID_PARAMETER. Make VUGPU_FILL_CMD() report the same
> error through the existing vhost-user-gpu dispatcher. This also rejects
> truncated type-specific commands.

One small correction: this says the malformed request is left without 
completion, but virtio_gpu_simple_process_cmd() and vg_process_cmd() 
actually complete them, though its response type is 
VIRTIO_GPU_RESP_OK_NODATA.

With that addressed,

Reviewed-by: Akihiko Odaki <[email protected]>

Regards,
Akihiko Odaki

> 
> The vhost-user-gpu common header is copied outside VUGPU_FILL_CMD(), so
> clear it and complete the request directly when that copy is short.
> 
> Fixes: CVE-2026-18054
> Resolves: https://gitlab.com/qemu-project/qemu/-/work_items/4094
> Reported-by: Ankur Saini <[email protected]>
> Suggested-by: Akihiko Odaki <[email protected]>
> Signed-off-by: Ankur Saini <[email protected]>
> ---
> Changes in v3:
> - Handle short copies in VIRTIO_GPU_FILL_CMD() and VUGPU_FILL_CMD(),
>    also rejecting truncated type-specific commands.
> - Clear partially copied QEMU command data before completing the error.
> - Keep vhost-user-gpu's explicit common-header check and drop the
>    redundant virtio_gpu_process_cmdq() check.
> - Rebase onto current master.
> - Link to v2: https://lore.kernel.org/qemu-devel/[email protected]
> 
> Changes in v2:
> - Also reject short/truncated control headers in vhost-user-gpu.
> - Expand the commit message to describe both affected paths.
> - Link to v1: https://lore.kernel.org/qemu-devel/[email protected]
> ---
>   contrib/vhost-user-gpu/vhost-user-gpu.c | 21 ++++++++++++---------
>   contrib/vhost-user-gpu/vugpu.h          |  1 +
>   include/hw/virtio/virtio-gpu.h          |  3 +++
>   3 files changed, 16 insertions(+), 9 deletions(-)
> 
> diff --git a/contrib/vhost-user-gpu/vhost-user-gpu.c b/contrib/vhost-user-gpu/vhost-user-gpu.c
> index bb41758e34..8149834745 100644
> --- a/contrib/vhost-user-gpu/vhost-user-gpu.c
> +++ b/contrib/vhost-user-gpu/vhost-user-gpu.c
> @@ -924,16 +924,19 @@ vg_handle_ctrl(VuDev *dev, int qidx)
>           if (len != sizeof(cmd->cmd_hdr)) {
>               g_warning("%s: command size incorrect %zu vs %zu\n",
>                         __func__, len, sizeof(cmd->cmd_hdr));
> -        }
> -
> -        virtio_gpu_ctrl_hdr_bswap(&cmd->cmd_hdr);
> -        g_debug("%d %s\n", cmd->cmd_hdr.type,
> -                vg_cmd_to_string(cmd->cmd_hdr.type));
> -
> -        if (vg->virgl) {
> -            vg_virgl_process_cmd(vg, cmd);
> +            memset(&cmd->cmd_hdr, 0, sizeof(cmd->cmd_hdr));
> +            vg_ctrl_response_nodata(
> +                vg, cmd, VIRTIO_GPU_RESP_ERR_INVALID_PARAMETER);
>           } else {
> -            vg_process_cmd(vg, cmd);
> +            virtio_gpu_ctrl_hdr_bswap(&cmd->cmd_hdr);
> +            g_debug("%d %s\n", cmd->cmd_hdr.type,
> +                    vg_cmd_to_string(cmd->cmd_hdr.type));
> +
> +            if (vg->virgl) {
> +                vg_virgl_process_cmd(vg, cmd);
> +            } else {
> +                vg_process_cmd(vg, cmd);
> +            }
>           }
>   
>           if (cmd->state != VG_CMD_STATE_FINISHED) {
> diff --git a/contrib/vhost-user-gpu/vugpu.h b/contrib/vhost-user-gpu/vugpu.h
> index 2374eb90cb..aaf2870cb2 100644
> --- a/contrib/vhost-user-gpu/vugpu.h
> +++ b/contrib/vhost-user-gpu/vugpu.h
> @@ -179,6 +179,7 @@ struct virtio_gpu_ctrl_command {
>           if (vugpufillcmd_s_ != sizeof(out)) {                   \
>               g_critical("%s: command size incorrect %zu vs %zu", \
>                          __func__, vugpufillcmd_s_, sizeof(out)); \
> +            cmd->error = VIRTIO_GPU_RESP_ERR_INVALID_PARAMETER; \
>               return;                                             \
>           }                                                       \
>       } while (0)
> diff --git a/include/hw/virtio/virtio-gpu.h b/include/hw/virtio/virtio-gpu.h
> index 2f60c72078..f965defa6b 100644
> --- a/include/hw/virtio/virtio-gpu.h
> +++ b/include/hw/virtio/virtio-gpu.h
> @@ -315,6 +315,9 @@ struct VirtIOGPURutabaga {
>               qemu_log_mask(LOG_GUEST_ERROR,                              \
>                             "%s: command size incorrect %zu vs %zu\n",    \
>                             __func__, virtiogpufillcmd_s_, sizeof(out));  \
> +            memset(&out, 0, sizeof(out));                               \
> +            virtio_gpu_ctrl_response_nodata(                            \
> +                g, cmd, VIRTIO_GPU_RESP_ERR_INVALID_PARAMETER);         \
>               return;                                                     \
>           }                                                               \
>       } while (0)
> 
> ---
> base-commit: b428fe036233cbd15d37e3c027ab6ca4d3661a80
> change-id: 20260728-virtio-gpu-short-header-476aa1dae4f7
> 
> Best regards,