Re: [PATCH] vp_vdpa: Fix return value check vp_vdpa_request_irq

[Jason Wang <jasowang@xxxxxxxxxx>]

On Fri, Mar 15, 2024 at 6:29 PM gavin.liu <gavin.liu@xxxxxxxxxxxxxxx> wrote:
>
> From: Yuxue Liu <yuxue.liu@xxxxxxxxxxxxxxx>
>
> In the vp_vdpa_set_status function, when setting the device status to
> VIRTIO_CONFIG_S_DRIVER_OK, the vp_vdpa_request_irq function may fail.
> In such cases, the device status should not be set to DRIVER_OK. Add
> exception printing to remind the user.
>
> Signed-off-by: Yuxue Liu <yuxue.liu@xxxxxxxxxxxxxxx>
> ---
>  drivers/vdpa/virtio_pci/vp_vdpa.c | 8 +++++++-
>  1 file changed, 7 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/vdpa/virtio_pci/vp_vdpa.c b/drivers/vdpa/virtio_pci/vp_vdpa.c
> index 281287fae89f..c1270b263867 100644
> --- a/drivers/vdpa/virtio_pci/vp_vdpa.c
> +++ b/drivers/vdpa/virtio_pci/vp_vdpa.c
> @@ -213,10 +213,16 @@ static void vp_vdpa_set_status(struct vdpa_device *vdpa, u8 status)
>         struct vp_vdpa *vp_vdpa = vdpa_to_vp(vdpa);
>         struct virtio_pci_modern_device *mdev = vp_vdpa_to_mdev(vp_vdpa);
>         u8 s = vp_vdpa_get_status(vdpa);
> +       struct pci_dev *pdev = mdev->pci_dev;
>
>         if (status & VIRTIO_CONFIG_S_DRIVER_OK &&
>             !(s & VIRTIO_CONFIG_S_DRIVER_OK)) {
> -               vp_vdpa_request_irq(vp_vdpa);
> +               if (vp_vdpa_request_irq(vp_vdpa)) {
> +                       WARN_ON(1);
> +                       dev_err(&pdev->dev,
> +                               "vp_vdpa : fail set status is DRIVER_OK\n");

I think having one of the WARN_ON or dev_err should be sufficient.

Thanks

> +                       return;
> +               }
>         }
>
>         vp_modern_set_status(mdev, status);
> --
> 2.43.0
>