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: <20230703142218.362549-1-eperezma@redhat.com> <20230703105022-mutt-send-email-mst@kernel.org>
 <CAJaqyWf2F_yBLBjj1RiPeJ92_zfq8BSMz8Pak2Vg6QinN8jS1Q@mail.gmail.com>
 <20230704063646-mutt-send-email-mst@kernel.org> <CAJaqyWfdPpkD5pY4tfzQdOscLBcrDBhBqzWjMbY_ZKsoyiqGdA@mail.gmail.com>
 <20230704114159-mutt-send-email-mst@kernel.org> <CACGkMEtWjOMtsbgQ2sx=e1BkuRSyDmVfXDccCm-QSiSbacQyCA@mail.gmail.com>
 <20230705043940-mutt-send-email-mst@kernel.org> <CACGkMEufNZGvWMN9Shh6NPOZOe-vf0RomfS1DX6DtxJjvO7fNA@mail.gmail.com>
In-Reply-To: <CACGkMEufNZGvWMN9Shh6NPOZOe-vf0RomfS1DX6DtxJjvO7fNA@mail.gmail.com>
From:   Eugenio Perez Martin <eperezma@redhat.com>
Date:   Thu, 6 Jul 2023 09:05:56 +0200
Message-ID: <CAJaqyWcqNkzJXxsoz_Lk_X0CvNW24Ay2Ki6q02EB8iR=qpwsfg@mail.gmail.com>
Subject: Re: [PATCH] vdpa: reject F_ENABLE_AFTER_DRIVER_OK if backend does not
 support it
To:     Jason Wang <jasowang@redhat.com>
Cc:     "Michael S. Tsirkin" <mst@redhat.com>, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org,
        Shannon Nelson <shannon.nelson@amd.com>,
        virtualization@lists.linux-foundation.org, kvm@vger.kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Thu, Jul 6, 2023 at 3:55=E2=80=AFAM Jason Wang <jasowang@redhat.com> wro=
te:
>
> On Wed, Jul 5, 2023 at 4:41=E2=80=AFPM Michael S. Tsirkin <mst@redhat.com=
> wrote:
> >
> > On Wed, Jul 05, 2023 at 03:49:58PM +0800, Jason Wang wrote:
> > > On Tue, Jul 4, 2023 at 11:45=E2=80=AFPM Michael S. Tsirkin <mst@redha=
t.com> wrote:
> > > >
> > > > On Tue, Jul 04, 2023 at 01:36:11PM +0200, Eugenio Perez Martin wrot=
e:
> > > > > On Tue, Jul 4, 2023 at 12:38=E2=80=AFPM Michael S. Tsirkin <mst@r=
edhat.com> wrote:
> > > > > >
> > > > > > On Tue, Jul 04, 2023 at 12:25:32PM +0200, Eugenio Perez Martin =
wrote:
> > > > > > > On Mon, Jul 3, 2023 at 4:52=E2=80=AFPM Michael S. Tsirkin <ms=
t@redhat.com> wrote:
> > > > > > > >
> > > > > > > > On Mon, Jul 03, 2023 at 04:22:18PM +0200, Eugenio P=C3=A9re=
z wrote:
> > > > > > > > > With the current code it is accepted as long as userland =
send it.
> > > > > > > > >
> > > > > > > > > Although userland should not set a feature flag that has =
not been
> > > > > > > > > offered to it with VHOST_GET_BACKEND_FEATURES, the curren=
t code will not
> > > > > > > > > complain for it.
> > > > > > > > >
> > > > > > > > > Since there is no specific reason for any parent to rejec=
t that backend
> > > > > > > > > feature bit when it has been proposed, let's control it a=
t vdpa frontend
> > > > > > > > > level. Future patches may move this control to the parent=
 driver.
> > > > > > > > >
> > > > > > > > > Fixes: 967800d2d52e ("vdpa: accept VHOST_BACKEND_F_ENABLE=
_AFTER_DRIVER_OK backend feature")
> > > > > > > > > Signed-off-by: Eugenio P=C3=A9rez <eperezma@redhat.com>
> > > > > > > >
> > > > > > > > Please do send v3. And again, I don't want to send "after d=
river ok" hack
> > > > > > > > upstream at all, I merged it in next just to give it some t=
esting.
> > > > > > > > We want RING_ACCESS_AFTER_KICK or some such.
> > > > > > > >
> > > > > > >
> > > > > > > Current devices do not support that semantic.
> > > > > >
> > > > > > Which devices specifically access the ring after DRIVER_OK but =
before
> > > > > > a kick?
> > > > > >
> > > > >
> > > > > Previous versions of the QEMU LM series did a spurious kick to st=
art
> > > > > traffic at the LM destination [1]. When it was proposed, that spu=
rious
> > > > > kick was removed from the series because to check for descriptors
> > > > > after driver_ok, even without a kick, was considered work of the
> > > > > parent driver.
> > > > >
> > > > > I'm ok to go back to this spurious kick, but I'm not sure if the =
hw
> > > > > will read the ring before the kick actually. I can ask.
> > > > >
> > > > > Thanks!
> > > > >
> > > > > [1] https://lists.nongnu.org/archive/html/qemu-devel/2023-01/msg0=
2775.html
> > > >
> > > > Let's find out. We need to check for ENABLE_AFTER_DRIVER_OK too, no=
?
> > >
> > > My understanding is [1] assuming ACCESS_AFTER_KICK. This seems
> > > sub-optimal than assuming ENABLE_AFTER_DRIVER_OK.
> > >
> > > But this reminds me one thing, as the thread is going too long, I
> > > wonder if we simply assume ENABLE_AFTER_DRIVER_OK if RING_RESET is
> > > supported?
> > >
> > > Thanks
> >
> > I don't see what does one have to do with another ...
> >
> > I think with RING_RESET we had another solution, enable rings
> > mapping them to a zero page, then reset and re-enable later.
>
> As discussed before, this seems to have some problems:
>
> 1) The behaviour is not clarified in the document
> 2) zero is a valid IOVA
>

I think we're not on the same page here.

As I understood, rings mapped to a zero page means essentially an
avail ring whose avail_idx is always 0, offered to the device instead
of the guest's ring. Once all CVQ commands are processed, we use
RING_RESET to switch to the right ring, being guest's or SVQ vring.


> Thanks
>
> >
> > > >
> > > >
> > > >
> > > > > > > My plan was to convert
> > > > > > > it in vp_vdpa if needed, and reuse the current vdpa ops. Sorr=
y if I
> > > > > > > was not explicit enough.
> > > > > > >
> > > > > > > The only solution I can see to that is to trap & emulate in t=
he vdpa
> > > > > > > (parent?) driver, as talked in virtio-comment. But that compl=
icates
> > > > > > > the architecture:
> > > > > > > * Offer VHOST_BACKEND_F_RING_ACCESS_AFTER_KICK
> > > > > > > * Store vq enable state separately, at
> > > > > > > vdpa->config->set_vq_ready(true), but not transmit that enabl=
e to hw
> > > > > > > * Store the doorbell state separately, but do not configure i=
t to the
> > > > > > > device directly.
> > > > > > >
> > > > > > > But how to recover if the device cannot configure them at kic=
k time,
> > > > > > > for example?
> > > > > > >
> > > > > > > Maybe we can just fail if the parent driver does not support =
enabling
> > > > > > > the vq after DRIVER_OK? That way no new feature flag is neede=
d.
> > > > > > >
> > > > > > > Thanks!
> > > > > > >
> > > > > > > >
> > > > > > > > > ---
> > > > > > > > > Sent with Fixes: tag pointing to git.kernel.org/pub/scm/l=
inux/kernel/git/mst
> > > > > > > > > commit. Please let me know if I should send a v3 of [1] i=
nstead.
> > > > > > > > >
> > > > > > > > > [1] https://lore.kernel.org/lkml/20230609121244-mutt-send=
-email-mst@kernel.org/T/
> > > > > > > > > ---
> > > > > > > > >  drivers/vhost/vdpa.c | 7 +++++--
> > > > > > > > >  1 file changed, 5 insertions(+), 2 deletions(-)
> > > > > > > > >
> > > > > > > > > diff --git a/drivers/vhost/vdpa.c b/drivers/vhost/vdpa.c
> > > > > > > > > index e1abf29fed5b..a7e554352351 100644
> > > > > > > > > --- a/drivers/vhost/vdpa.c
> > > > > > > > > +++ b/drivers/vhost/vdpa.c
> > > > > > > > > @@ -681,18 +681,21 @@ static long vhost_vdpa_unlocked_ioc=
tl(struct file *filep,
> > > > > > > > >  {
> > > > > > > > >       struct vhost_vdpa *v =3D filep->private_data;
> > > > > > > > >       struct vhost_dev *d =3D &v->vdev;
> > > > > > > > > +     const struct vdpa_config_ops *ops =3D v->vdpa->conf=
ig;
> > > > > > > > >       void __user *argp =3D (void __user *)arg;
> > > > > > > > >       u64 __user *featurep =3D argp;
> > > > > > > > > -     u64 features;
> > > > > > > > > +     u64 features, parent_features =3D 0;
> > > > > > > > >       long r =3D 0;
> > > > > > > > >
> > > > > > > > >       if (cmd =3D=3D VHOST_SET_BACKEND_FEATURES) {
> > > > > > > > >               if (copy_from_user(&features, featurep, siz=
eof(features)))
> > > > > > > > >                       return -EFAULT;
> > > > > > > > > +             if (ops->get_backend_features)
> > > > > > > > > +                     parent_features =3D ops->get_backen=
d_features(v->vdpa);
> > > > > > > > >               if (features & ~(VHOST_VDPA_BACKEND_FEATURE=
S |
> > > > > > > > >                                BIT_ULL(VHOST_BACKEND_F_SU=
SPEND) |
> > > > > > > > >                                BIT_ULL(VHOST_BACKEND_F_RE=
SUME) |
> > > > > > > > > -                              BIT_ULL(VHOST_BACKEND_F_EN=
ABLE_AFTER_DRIVER_OK)))
> > > > > > > > > +                              parent_features))
> > > > > > > > >                       return -EOPNOTSUPP;
> > > > > > > > >               if ((features & BIT_ULL(VHOST_BACKEND_F_SUS=
PEND)) &&
> > > > > > > > >                    !vhost_vdpa_can_suspend(v))
> > > > > > > > > --
> > > > > > > > > 2.39.3
> > > > > > > >
> > > > > >
> > > >
> >
>