From: Guo Zhi <qtxuning1999@sjtu.edu.cn>
To: jasowang <jasowang@redhat.com>
Cc: eperezma <eperezma@redhat.com>, sgarzare <sgarzare@redhat.com>,
Michael Tsirkin <mst@redhat.com>, netdev <netdev@vger.kernel.org>,
linux-kernel <linux-kernel@vger.kernel.org>,
kvm list <kvm@vger.kernel.org>,
virtualization <virtualization@lists.linux-foundation.org>
Subject: Re: [RFC v3 3/7] vsock: batch buffers in tx
Date: Thu, 8 Sep 2022 16:41:14 +0800 (CST) [thread overview]
Message-ID: <1010358496.165709.1662626474492.JavaMail.zimbra@sjtu.edu.cn> (raw)
In-Reply-To: <ff96c12e-95cb-be57-9b5b-2da08b0630c6@redhat.com>
----- Original Message -----
> From: "jasowang" <jasowang@redhat.com>
> To: "Guo Zhi" <qtxuning1999@sjtu.edu.cn>, "eperezma" <eperezma@redhat.com>, "sgarzare" <sgarzare@redhat.com>, "Michael
> Tsirkin" <mst@redhat.com>
> Cc: "netdev" <netdev@vger.kernel.org>, "linux-kernel" <linux-kernel@vger.kernel.org>, "kvm list" <kvm@vger.kernel.org>,
> "virtualization" <virtualization@lists.linux-foundation.org>
> Sent: Wednesday, September 7, 2022 12:27:40 PM
> Subject: Re: [RFC v3 3/7] vsock: batch buffers in tx
> 在 2022/9/1 13:54, Guo Zhi 写道:
>> 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@sjtu.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);
>
>
> I'd do this step by step.
>
> 1) switch to use vhost_add_used_n() for vsock, less copy_to_user()
> better performance
> 2) do in-order on top
>
>
LGTM!, I think it is the correct way.
>> + } else {
>> + vq->heads[add].id = head;
>> + vq->heads[add++].len = 0;
>
>
> How can we guarantee that we are in the boundary of the heads array?
>
> Btw in the case of in-order we don't need to record the heads, instead
> we just need to know the head of the last buffer, it reduces the stress
> on the cache.
>
> Thanks
>
Yeah, I will change this and only copy last head for in order feature.
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);
next prev parent reply other threads:[~2022-09-08 8:41 UTC|newest]
Thread overview: 22+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-09-01 5:54 [RFC v3 0/7] In order support for virtio_ring, vhost and vsock Guo Zhi
2022-09-01 5:54 ` [RFC v3 1/7] vhost: expose used buffers Guo Zhi
2022-09-01 8:55 ` Eugenio Perez Martin
2022-09-07 4:21 ` Jason Wang
2022-09-08 8:43 ` Guo Zhi
2022-09-01 5:54 ` [RFC v3 2/7] vhost_test: batch used buffer Guo Zhi
2022-09-01 5:54 ` [RFC v3 3/7] vsock: batch buffers in tx Guo Zhi
2022-09-01 9:00 ` Eugenio Perez Martin
2022-09-07 4:27 ` Jason Wang
2022-09-08 8:41 ` Guo Zhi [this message]
2022-09-01 5:54 ` [RFC v3 4/7] vsock: announce VIRTIO_F_IN_ORDER in vsock Guo Zhi
2022-09-01 5:54 ` [RFC v3 5/7] virtio: unmask F_NEXT flag in desc_extra Guo Zhi
2022-09-01 6:07 ` Xuan Zhuo
2022-09-01 6:14 ` Guo Zhi
2022-09-01 5:54 ` [RFC v3 6/7] virtio: in order support for virtio_ring Guo Zhi
2022-09-01 6:10 ` Xuan Zhuo
2022-09-07 5:38 ` Jason Wang
2022-09-08 8:47 ` Guo Zhi
2022-09-08 9:36 ` Guo Zhi
2022-10-03 11:32 ` Guo Zhi
2022-09-01 5:54 ` [RFC v3 7/7] virtio: announce VIRTIO_F_IN_ORDER support Guo Zhi
2022-09-07 4:13 ` [RFC v3 0/7] In order support for virtio_ring, vhost and vsock Jason Wang
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1010358496.165709.1662626474492.JavaMail.zimbra@sjtu.edu.cn \
--to=qtxuning1999@sjtu.edu.cn \
--cc=eperezma@redhat.com \
--cc=jasowang@redhat.com \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mst@redhat.com \
--cc=netdev@vger.kernel.org \
--cc=sgarzare@redhat.com \
--cc=virtualization@lists.linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).