Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.35 as permitted sender) client-ip=23.128.96.35;
MIME-Version: 1.0
References: <1698350834-415881-1-git-send-email-steven.sistare@oracle.com>
In-Reply-To: <1698350834-415881-1-git-send-email-steven.sistare@oracle.com>
From:   Jason Wang <jasowang@redhat.com>
Date:   Fri, 27 Oct 2023 14:31:56 +0800
Message-ID: <CACGkMEuO6o3Ys9NcHJpa9w5EiS-ugsiaBbBKEHKHpPSrKBKJow@mail.gmail.com>
Subject: Re: [RFC] vdpa/mlx5: preserve CVQ vringh index
To:     Steve Sistare <steven.sistare@oracle.com>
Cc:     virtualization@lists.linux-foundation.org,
        linux-kernel@vger.kernel.org,
        "Michael S. Tsirkin" <mst@redhat.com>, Eli Cohen <elic@nvidia.com>,
        Si-Wei Liu <si-wei.liu@oracle.com>,
        Dragos Tatulea <dtatulea@nvidia.com>,
        Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable
Precedence: bulk

On Fri, Oct 27, 2023 at 4:07=E2=80=AFAM Steve Sistare <steven.sistare@oracl=
e.com> wrote:
>
> mlx5_vdpa does not preserve userland's view of vring base for the control
> queue in the following sequence:
>
> ioctl VHOST_SET_VRING_BASE
> ioctl VHOST_VDPA_SET_STATUS VIRTIO_CONFIG_S_DRIVER_OK
>   mlx5_vdpa_set_status()
>     setup_cvq_vring()
>       vringh_init_iotlb()
>         vringh_init_kern()
>           vrh->last_avail_idx =3D 0;
> ioctl VHOST_GET_VRING_BASE
>
> To fix, restore the value of cvq->vring.last_avail_idx after calling
> vringh_init_iotlb.
>
> Signed-off-by: Steve Sistare <steven.sistare@oracle.com>
> ---
>  drivers/vdpa/mlx5/net/mlx5_vnet.c |  7 ++++++-
>  drivers/vhost/vringh.c            | 30 ++++++++++++++++++++++++++++++
>  include/linux/vringh.h            |  2 ++
>  3 files changed, 38 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/vdpa/mlx5/net/mlx5_vnet.c b/drivers/vdpa/mlx5/net/ml=
x5_vnet.c
> index 946488b8989f..f64758143115 100644
> --- a/drivers/vdpa/mlx5/net/mlx5_vnet.c
> +++ b/drivers/vdpa/mlx5/net/mlx5_vnet.c
> @@ -2795,13 +2795,18 @@ static int setup_cvq_vring(struct mlx5_vdpa_dev *=
mvdev)
>         struct mlx5_control_vq *cvq =3D &mvdev->cvq;
>         int err =3D 0;
>
> -       if (mvdev->actual_features & BIT_ULL(VIRTIO_NET_F_CTRL_VQ))
> +       if (mvdev->actual_features & BIT_ULL(VIRTIO_NET_F_CTRL_VQ)) {
> +               u16 last_avail_idx =3D cvq->vring.last_avail_idx;
> +
>                 err =3D vringh_init_iotlb(&cvq->vring, mvdev->actual_feat=
ures,
>                                         MLX5_CVQ_MAX_ENT, false,
>                                         (struct vring_desc *)(uintptr_t)c=
vq->desc_addr,
>                                         (struct vring_avail *)(uintptr_t)=
cvq->driver_addr,
>                                         (struct vring_used *)(uintptr_t)c=
vq->device_addr);
>
> +               if (!err)
> +                       vringh_set_base_iotlb(&cvq->vring, last_avail_idx=
);

Btw, vringh_set_base_iotlb() deserves an independent patch and it
seems it is not specific to IOTLB, so we probably need an indirection
to have vringh_set_base() first.

Or I wonder if it's better to just introduce a new parameter to
vringh_init_iotlb()...

Thanks

> +       }
>         return err;
>  }
>
> diff --git a/drivers/vhost/vringh.c b/drivers/vhost/vringh.c
> index 7b8fd977f71c..799762c83007 100644
> --- a/drivers/vhost/vringh.c
> +++ b/drivers/vhost/vringh.c
> @@ -595,6 +595,24 @@ static inline void __vringh_notify_disable(struct vr=
ingh *vrh,
>         }
>  }
>
> +static inline int __vringh_set_base(struct vringh *vrh, u16 idx,
> +                           int (*putu16)(const struct vringh *vrh,
> +                               __virtio16 *p, u16 val))
> +{
> +    int ret;
> +
> +    ret =3D putu16(vrh, &vrh->vring.avail->idx, idx);
> +    if (ret)
> +        return ret;
> +
> +    ret =3D putu16(vrh, &vrh->vring.used->idx, idx);
> +    if (ret)
> +        return ret;
> +
> +    vrh->last_avail_idx =3D vrh->last_used_idx =3D idx;
> +    return 0;
> +}
> +
>  /* Userspace access helpers: in this case, addresses are really userspac=
e. */
>  static inline int getu16_user(const struct vringh *vrh, u16 *val, const =
__virtio16 *p)
>  {
> @@ -1456,6 +1474,18 @@ void vringh_set_iotlb(struct vringh *vrh, struct v=
host_iotlb *iotlb,
>  }
>  EXPORT_SYMBOL(vringh_set_iotlb);
>
> +/**
> + * vringh_set_base_iotlb - set avail_idx and used_idx
> + * @vrh: the vring
> + * @idx: the value to set
> + */
> +int vringh_set_base_iotlb(struct vringh *vrh, u16 idx)
> +{
> +    return __vringh_set_base(vrh, idx, putu16_iotlb);
> +}
> +EXPORT_SYMBOL(vringh_set_base_iotlb);
> +
> +
>  /**
>   * vringh_getdesc_iotlb - get next available descriptor from ring with
>   * IOTLB.
> diff --git a/include/linux/vringh.h b/include/linux/vringh.h
> index c3a8117dabe8..e9b8af4e6a5e 100644
> --- a/include/linux/vringh.h
> +++ b/include/linux/vringh.h
> @@ -306,6 +306,8 @@ int vringh_init_iotlb_va(struct vringh *vrh, u64 feat=
ures,
>                          struct vring_avail *avail,
>                          struct vring_used *used);
>
> +int vringh_set_base_iotlb(struct vringh *vrh, u16 idx);
> +
>  int vringh_getdesc_iotlb(struct vringh *vrh,
>                          struct vringh_kiov *riov,
>                          struct vringh_kiov *wiov,
> --
> 2.39.3
>