From: "Michael S. Tsirkin" <mst@redhat.com>
To: Jason Wang <jasowang@redhat.com>
Cc: Heng Qi <hengqi@linux.alibaba.com>,
virtio-comment@lists.oasis-open.org,
Yuri Benditovich <yuri.benditovich@daynix.com>,
Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Subject: [virtio-comment] Re: [PATCH RESEND] virtio-net: support setting coalescing params for multiple vqs
Date: Fri, 22 Dec 2023 03:26:20 -0500 [thread overview]
Message-ID: <20231222032455-mutt-send-email-mst@kernel.org> (raw)
In-Reply-To: <CACGkMEt_kvBbnpNt_cfB8KdKWM9WP8T_wnQBm13yKpe5JYVnAg@mail.gmail.com>
On Thu, Dec 21, 2023 at 11:37:17AM +0800, Jason Wang wrote:
> On Wed, Dec 20, 2023 at 10:40 PM Heng Qi <hengqi@linux.alibaba.com> wrote:
> >
> > Currently, when each time the driver attempts to update the coalescing parameters
> > for a vq, it needs to kick the device and wait for the ctrlq response to return.
> >
> > If a command can only update one vq parameters, when the parameters are updated
> > frequently (such as netdim), ctrlq on the device is kicked frequently, which will
> > increase the device CPU scheduling overhead, and the number and overhead of device
> > DMA will also increase.
> >
> > Merging multiple vq updated parameters into one command can effectively reduce
> > the number of kick devices and device DMA times.
> >
> > Test results show that this greatly improves the efficiency of the ctrlq in
> > responding to multiple vq coalescing parameter updates issued by the driver.
>
> So netdim is per virtqueue, to make use of this, you need to batch the
> netdim requests first. And if you do that, you can batch the commands
> as well. Then you get one kick for sevreal coal requests?
>
> Or are you saying you run out of ctrl vq?
>
> Thanks
I think Heng is saying ctrl vq implementation on this DPU is so bad,
you need to avoid that as much as possible.
--
MST
This publicly archived list offers a means to provide input to the
OASIS Virtual I/O Device (VIRTIO) TC.
In order to verify user consent to the Feedback License terms and
to minimize spam in the list archive, subscription is required
before posting.
Subscribe: virtio-comment-subscribe@lists.oasis-open.org
Unsubscribe: virtio-comment-unsubscribe@lists.oasis-open.org
List help: virtio-comment-help@lists.oasis-open.org
List archive: https://lists.oasis-open.org/archives/virtio-comment/
Feedback License: https://www.oasis-open.org/who/ipr/feedback_license.pdf
List Guidelines: https://www.oasis-open.org/policies-guidelines/mailing-lists
Committee: https://www.oasis-open.org/committees/virtio/
Join OASIS: https://www.oasis-open.org/join/
next prev parent reply other threads:[~2023-12-22 8:26 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-12-20 14:40 [virtio-comment] [PATCH RESEND] virtio-net: support setting coalescing params for multiple vqs Heng Qi
2023-12-21 3:37 ` [virtio-comment] " Jason Wang
2023-12-21 4:55 ` Heng Qi
2023-12-22 2:40 ` Jason Wang
2023-12-22 6:45 ` Heng Qi
2023-12-22 8:20 ` Michael S. Tsirkin
2023-12-26 2:17 ` Heng Qi
2023-12-26 3:32 ` Jason Wang
2023-12-26 9:18 ` Michael S. Tsirkin
2023-12-27 2:32 ` Jason Wang
2023-12-22 8:26 ` Michael S. Tsirkin [this message]
2023-12-21 6:56 ` Michael S. Tsirkin
2023-12-21 7:01 ` Heng Qi
2023-12-21 7:08 ` Michael S. Tsirkin
2023-12-21 7:19 ` Heng Qi
2023-12-21 7:48 ` Heng Qi
2023-12-22 8:24 ` Michael S. Tsirkin
2023-12-26 2:04 ` Heng Qi
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=20231222032455-mutt-send-email-mst@kernel.org \
--to=mst@redhat.com \
--cc=hengqi@linux.alibaba.com \
--cc=jasowang@redhat.com \
--cc=virtio-comment@lists.oasis-open.org \
--cc=xuanzhuo@linux.alibaba.com \
--cc=yuri.benditovich@daynix.com \
/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