Re: [virtio-dev] Re: [PATCH v1] virtio-mmio: Specify wait needed in driver during reset

[Jason Wang <jasowang@redhat.com>]

在 2021/8/17 下午8:48, Srivatsa Vaddagiri 写道:
> * Srivatsa Vaddagiri <quic_svaddagi@quicinc.com> [2021-08-17 15:33:33]:
>
> > Handling reset is the only open issue we have, as guest triggering reset has
> > currently no provision to poll for reset completion and hypervisor itself cannot
> > handle reset completly. This is where we observed discrepancy between PCI vs
> > MMIO in handling reset, which we wanted to address with this discussion.
> >
> > I think the option we discussed earlier of a new feature bit seems less
> > intrusive than incrementing MMIO version?
> >
> > https://lists.oasis-open.org/archives/virtio-dev/202107/msg00168.html
> Alternately would it be fine to merge a change like this to Linux virtio_mmio
> driver w/o requiring any further changes to spec?
>
> diff --git a/drivers/virtio/virtio_mmio.c b/drivers/virtio/virtio_mmio.c
> index 56128b9..9db81e6 100644
> --- a/drivers/virtio/virtio_mmio.c
> +++ b/drivers/virtio/virtio_mmio.c
> @@ -262,6 +262,9 @@ static void vm_reset(struct virtio_device *vdev)
>   
>   	/* 0 status means a reset. */
>   	writel(0, vm_dev->base + VIRTIO_MMIO_STATUS);
> +
> +	while (vm_get_status(vdev))
> +		msleep(1);
>   }
>   
>   


I think we still need some clarification on the spec.

Note that doing this, with an untrusted device the system tool like 
modprobe may block at this forever

We probably need a way to fail the reset.

VDUSE meets similar issue.

Thanks


>