Received-SPF: pass (google.com: domain of linux-kernel+bounces-32783-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Precedence: bulk
MIME-Version: 1.0
References: <1705519403-255169-1-git-send-email-steven.sistare@oracle.com> <mhnsrwlvqjeftu5aa77iaowdk2wrq366yofjchbn5fwchgcbna@kcjtoj46fawi>
In-Reply-To: <mhnsrwlvqjeftu5aa77iaowdk2wrq366yofjchbn5fwchgcbna@kcjtoj46fawi>
From: Eugenio Perez Martin <eperezma@redhat.com>
Date: Mon, 22 Jan 2024 11:47:22 +0100
Message-ID: <CAJaqyWcLj7Y0rm3_jdfP8KgS4gkE9ns286XJywBGE0iijMireQ@mail.gmail.com>
Subject: Re: [PATCH V1] vdpa_sim: reset must not run
To: Stefano Garzarella <sgarzare@redhat.com>
Cc: Steve Sistare <steven.sistare@oracle.com>, virtualization@lists.linux-foundation.org, 
	linux-kernel@vger.kernel.org, "Michael S. Tsirkin" <mst@redhat.com>, 
	Jason Wang <jasowang@redhat.com>, Si-Wei Liu <si-wei.liu@oracle.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Mon, Jan 22, 2024 at 11:22=E2=80=AFAM Stefano Garzarella <sgarzare@redha=
t.com> wrote:
>
> On Wed, Jan 17, 2024 at 11:23:23AM -0800, Steve Sistare wrote:
> >vdpasim_do_reset sets running to true, which is wrong, as it allows
> >vdpasim_kick_vq to post work requests before the device has been
> >configured.  To fix, do not set running until VIRTIO_CONFIG_S_FEATURES_O=
K
> >is set.
> >
> >Fixes: 0c89e2a3a9d0 ("vdpa_sim: Implement suspend vdpa op")
> >Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
> >Reviewed-by: Eugenio P=C3=A9rez <eperezma@redhat.com>
> >---
> > drivers/vdpa/vdpa_sim/vdpa_sim.c | 3 ++-
> > 1 file changed, 2 insertions(+), 1 deletion(-)
> >
> >diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vd=
pa_sim.c
> >index be2925d0d283..6304cb0b4770 100644
> >--- a/drivers/vdpa/vdpa_sim/vdpa_sim.c
> >+++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c
> >@@ -160,7 +160,7 @@ static void vdpasim_do_reset(struct vdpasim *vdpasim=
, u32 flags)
> >               }
> >       }
> >
> >-      vdpasim->running =3D true;
> >+      vdpasim->running =3D false;
> >       spin_unlock(&vdpasim->iommu_lock);
> >
> >       vdpasim->features =3D 0;
> >@@ -483,6 +483,7 @@ static void vdpasim_set_status(struct vdpa_device *v=
dpa, u8 status)
> >
> >       mutex_lock(&vdpasim->mutex);
> >       vdpasim->status =3D status;
> >+      vdpasim->running =3D (status & VIRTIO_CONFIG_S_FEATURES_OK) !=3D =
0;
> >       mutex_unlock(&vdpasim->mutex);
>
> Should we do something similar also in vdpasim_resume() ?
>
> I mean something like this:
>
> diff --git a/drivers/vdpa/vdpa_sim/vdpa_sim.c b/drivers/vdpa/vdpa_sim/vdp=
a_sim.c
> index be2925d0d283..55e4633d5442 100644
> --- a/drivers/vdpa/vdpa_sim/vdpa_sim.c
> +++ b/drivers/vdpa/vdpa_sim/vdpa_sim.c
> @@ -520,7 +520,7 @@ static int vdpasim_resume(struct vdpa_device *vdpa)
>          int i;
>
>          mutex_lock(&vdpasim->mutex);
> -       vdpasim->running =3D true;
> +       vdpasim->running =3D (vdpasim->status & VIRTIO_CONFIG_S_FEATURES_=
OK) !=3D 0;
>
>          if (vdpasim->pending_kick) {
>                  /* Process pending descriptors */
>
> Thanks,
> Stefano
>

The suspend and resume operation should not be called before
DRIVER_OK, so maybe we should add that protection at
drivers/vhost/vdpa.c actually?

Thanks!