linux-kernel - Re: [RFC v3 3/7] vsock: batch buffers in tx

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAJaqyWdSNVyWeHyeZgzhczL0+bcvYShwJEfzHpjmy45cSSre3Q@mail.gmail.com>
Date:   Thu, 1 Sep 2022 11:00:03 +0200
From:   Eugenio Perez Martin <eperezma@...hat.com>
To:     Guo Zhi <qtxuning1999@...u.edu.cn>
Cc:     Jason Wang <jasowang@...hat.com>,
        Stefano Garzarella <sgarzare@...hat.com>,
        Michael Tsirkin <mst@...hat.com>,
        netdev <netdev@...r.kernel.org>,
        linux-kernel <linux-kernel@...r.kernel.org>,
        kvm list <kvm@...r.kernel.org>,
        virtualization <virtualization@...ts.linux-foundation.org>
Subject: Re: [RFC v3 3/7] vsock: batch buffers in tx

On Thu, Sep 1, 2022 at 7:55 AM Guo Zhi <qtxuning1999@...u.edu.cn> wrote:
>
> Vsock uses buffers in order, and for tx driver doesn't have to
> know the length of the buffer. So we can do a batch for vsock if
> in order negotiated, only write one used ring for a batch of buffers
>
> Signed-off-by: Guo Zhi <qtxuning1999@...u.edu.cn>
> ---
>  drivers/vhost/vsock.c | 12 ++++++++++--
>  1 file changed, 10 insertions(+), 2 deletions(-)
>
> diff --git a/drivers/vhost/vsock.c b/drivers/vhost/vsock.c
> index 368330417bde..e08fbbb5439e 100644
> --- a/drivers/vhost/vsock.c
> +++ b/drivers/vhost/vsock.c
> @@ -497,7 +497,7 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work *work)
>         struct vhost_vsock *vsock = container_of(vq->dev, struct vhost_vsock,
>                                                  dev);
>         struct virtio_vsock_pkt *pkt;
> -       int head, pkts = 0, total_len = 0;
> +       int head, pkts = 0, total_len = 0, add = 0;
>         unsigned int out, in;
>         bool added = false;
>
> @@ -551,10 +551,18 @@ static void vhost_vsock_handle_tx_kick(struct vhost_work *work)
>                 else
>                         virtio_transport_free_pkt(pkt);
>
> -               vhost_add_used(vq, head, 0);
> +               if (!vhost_has_feature(vq, VIRTIO_F_IN_ORDER)) {
> +                       vhost_add_used(vq, head, 0);
> +               } else {
> +                       vq->heads[add].id = head;
> +                       vq->heads[add++].len = 0;

Knowing that the descriptors are used in order we can save a few
memory writes to the vq->heads[] array. vhost.c is checking for the
feature in_order anyway.

Thanks!

> +               }
>                 added = true;
>         } while(likely(!vhost_exceeds_weight(vq, ++pkts, total_len)));
>
> +       /* If in order feature negotiaged, we can do a batch to increase performance */
> +       if (vhost_has_feature(vq, VIRTIO_F_IN_ORDER) && added)
> +               vhost_add_used_n(vq, vq->heads, add);
>  no_more_replies:
>         if (added)
>                 vhost_signal(&vsock->dev, vq);
> --
> 2.17.1
>