Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
MIME-Version: 1.0
References: <20230609092127.170673-1-eperezma@redhat.com> <20230609092127.170673-3-eperezma@redhat.com>
 <20230609121244-mutt-send-email-mst@kernel.org>
In-Reply-To: <20230609121244-mutt-send-email-mst@kernel.org>
From:   Eugenio Perez Martin <eperezma@redhat.com>
Date:   Mon, 3 Jul 2023 10:21:20 +0200
Message-ID: <CAJaqyWcaLe6TX4VuqMkh_po=C17MVX-tVq-XqHkMWYLTXh-sHQ@mail.gmail.com>
Subject: Re: [PATCH v2 2/4] vdpa: accept VHOST_BACKEND_F_ENABLE_AFTER_DRIVER_OK
 backend feature
To:     "Michael S. Tsirkin" <mst@redhat.com>
Cc:     Jason Wang <jasowang@redhat.com>,
        Zhu Lingshan <lingshan.zhu@intel.com>,
        virtualization@lists.linux-foundation.org,
        "Gonglei (Arei)" <arei.gonglei@huawei.com>, longpeng2@huawei.com,
        Lei Yang <leiyang@redhat.com>,
        Parav Pandit <parav@mellanox.com>,
        Gautam Dawar <gdawar@xilinx.com>,
        Dragos Tatulea <dtatulea@nvidia.com>,
        Laurent Vivier <lvivier@redhat.com>, linux-api@vger.kernel.org,
        Liuxiangdong <liuxiangdong5@huawei.com>,
        linux-kernel@vger.kernel.org, alvaro.karsz@solid-run.com,
        Shannon Nelson <snelson@pensando.io>,
        Stefano Garzarella <sgarzare@redhat.com>,
        Cindy Lu <lulu@redhat.com>, si-wei.liu@oracle.com,
        Harpreet Singh Anand <hanand@xilinx.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Fri, Jun 9, 2023 at 6:13=E2=80=AFPM Michael S. Tsirkin <mst@redhat.com> =
wrote:
>
> On Fri, Jun 09, 2023 at 11:21:25AM +0200, Eugenio P=C3=A9rez wrote:
> > Accepting VHOST_BACKEND_F_ENABLE_AFTER_DRIVER_OK backend feature if
> > userland sets it.
> >
> > Signed-off-by: Eugenio P=C3=A9rez <eperezma@redhat.com>
> > Acked-by: Shannon Nelson <shannon.nelson@amd.com>
>
> I don't get it, so all vdpa devices accept this automatically?
> Should this not be up to the parent?
>

At the moment I don't see a reason why if a parent offers this
feature, it could reject it afterwards. However I think we can add a
fail if userland acks the backend feature but the parent does not
offer it however.

Would it work to add such fail in vdpa frontend and move it to the
backend if and when any parent driver needs it in the future?

Thanks!

> > ---
> >  drivers/vhost/vdpa.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c
> > index bf77924d5b60..a3204406b73d 100644
> > --- a/drivers/vhost/vdpa.c
> > +++ b/drivers/vhost/vdpa.c
> > @@ -680,7 +680,8 @@ static long vhost_vdpa_unlocked_ioctl(struct file *=
filep,
> >                       return -EFAULT;
> >               if (features & ~(VHOST_VDPA_BACKEND_FEATURES |
> >                                BIT_ULL(VHOST_BACKEND_F_SUSPEND) |
> > -                              BIT_ULL(VHOST_BACKEND_F_RESUME)))
> > +                              BIT_ULL(VHOST_BACKEND_F_RESUME) |
> > +                              BIT_ULL(VHOST_BACKEND_F_ENABLE_AFTER_DRI=
VER_OK)))
> >                       return -EOPNOTSUPP;
> >               if ((features & BIT_ULL(VHOST_BACKEND_F_SUSPEND)) &&
> >                    !vhost_vdpa_can_suspend(v))
> > --
> > 2.31.1
>