From: Jakub Kicinski <kuba@kernel.org>
To: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Cc: Christoph Hellwig <hch@infradead.org>,
virtualization@lists.linux-foundation.org,
"David S. Miller" <davem@davemloft.net>,
Eric Dumazet <edumazet@google.com>,
Paolo Abeni <pabeni@redhat.com>,
Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
Jesper Dangaard Brouer <hawk@kernel.org>,
John Fastabend <john.fastabend@gmail.com>,
netdev@vger.kernel.org, bpf@vger.kernel.org,
"Michael S. Tsirkin" <mst@redhat.com>,
Jason Wang <jasowang@redhat.com>,
Pavel Begunkov <asml.silence@gmail.com>
Subject: Re: [PATCH vhost v11 05/10] virtio_ring: introduce virtqueue_dma_dev()
Date: Tue, 1 Aug 2023 08:45:10 -0700 [thread overview]
Message-ID: <20230801084510.1c2460b9@kernel.org> (raw)
In-Reply-To: <1690858650.8698683-2-xuanzhuo@linux.alibaba.com>
On Tue, 1 Aug 2023 10:57:30 +0800 Xuan Zhuo wrote:
> > You have this working and benchmarked or this is just and idea?
>
> This is not just an idea. I said that has been used on large scale.
>
> This is the library for the APP to use the AF_XDP. We has open it.
> https://gitee.com/anolis/libxudp
>
> This is the Alibaba version of the nginx. That has been opened, that supported
> to work with the libray to use AF_XDP.
> http://tengine.taobao.org/
>
> I supported this on our kernel release Anolis/Alinux.
Interesting!
> The work was done about 2 years ago. You know, I pushed the first version to
> enable AF_XDP on virtio-net about two years ago. I never thought the job would
> be so difficult.
Me neither, but it is what it is.
> The nic (virtio-net) of AliYun can reach 24,000,000PPS.
> So I think there is no different with the real HW on the performance.
>
> With the AF_XDP, the UDP pps is seven times that of the kernel udp stack.
UDP pps or QUIC pps? UDP with or without GSO?
Do you have measurements of how much it saves in real world workloads?
I'm asking mostly out of curiosity, not to question the use case.
> > What about io_uring zero copy w/ pre-registered buffers.
> > You'll get csum offload, GSO, all the normal perf features.
>
> We tried io-uring, but it was not suitable for our scenario.
>
> Yes, now the AF_XDP does not support the csum offload and GSO.
> This is indeed a small problem.
Can you say more about io-uring suitability? It can do zero copy
and recently-ish Pavel optimized it quite a bit.
next prev parent reply other threads:[~2023-08-01 15:45 UTC|newest]
Thread overview: 89+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-07-10 3:42 [PATCH vhost v11 00/10] virtio core prepares for AF_XDP Xuan Zhuo
2023-07-10 3:42 ` [PATCH vhost v11 01/10] virtio_ring: check use_dma_api before unmap desc for indirect Xuan Zhuo
2023-07-10 3:42 ` [PATCH vhost v11 02/10] virtio_ring: put mapping error check in vring_map_one_sg Xuan Zhuo
2023-07-10 3:42 ` [PATCH vhost v11 03/10] virtio_ring: introduce virtqueue_set_premapped() Xuan Zhuo
[not found] ` <CACGkMEumF73qbByV3K1+fdgnXBXqu-YS2yas+Vy_=Dn+yjy-cw@mail.gmail.com>
2023-07-12 8:35 ` Xuan Zhuo
2023-07-13 11:14 ` Christoph Hellwig
2023-07-13 14:47 ` Michael S. Tsirkin
2023-07-20 6:22 ` Christoph Hellwig
2023-07-13 14:52 ` Michael S. Tsirkin
2023-07-10 3:42 ` [PATCH vhost v11 04/10] virtio_ring: support add premapped buf Xuan Zhuo
2023-07-12 8:31 ` Jason Wang
2023-07-12 8:33 ` Xuan Zhuo
2023-07-10 3:42 ` [PATCH vhost v11 05/10] virtio_ring: introduce virtqueue_dma_dev() Xuan Zhuo
2023-07-12 8:33 ` Jason Wang
2023-07-13 11:15 ` Christoph Hellwig
2023-07-13 14:51 ` Michael S. Tsirkin
2023-07-20 6:22 ` Christoph Hellwig
2023-07-20 6:45 ` Xuan Zhuo
2023-07-20 6:57 ` Christoph Hellwig
2023-07-20 7:34 ` Xuan Zhuo
2023-07-24 20:05 ` Michael S. Tsirkin
2023-07-20 17:21 ` Michael S. Tsirkin
2023-07-24 16:43 ` Christoph Hellwig
2023-07-25 2:13 ` Xuan Zhuo
2023-07-25 7:34 ` Michael S. Tsirkin
2023-07-25 11:07 ` Xuan Zhuo
2023-07-28 6:02 ` Xuan Zhuo
2023-07-28 15:03 ` Jakub Kicinski
2023-07-31 1:23 ` Jason Wang
2023-07-31 15:46 ` Jakub Kicinski
2023-08-01 2:03 ` Xuan Zhuo
2023-08-01 2:36 ` Jakub Kicinski
2023-08-01 2:57 ` Xuan Zhuo
2023-08-01 15:45 ` Jakub Kicinski [this message]
2023-08-02 1:36 ` Xuan Zhuo
2023-08-02 11:12 ` Pavel Begunkov
2023-07-31 2:34 ` Xuan Zhuo
2023-08-01 16:17 ` Michael S. Tsirkin
2023-08-02 1:49 ` Xuan Zhuo
2023-08-07 6:14 ` Xuan Zhuo
2023-08-08 2:26 ` Jason Wang
2023-08-08 2:47 ` Xuan Zhuo
2023-08-08 3:08 ` Jason Wang
2023-08-08 3:09 ` Xuan Zhuo
2023-08-08 3:49 ` Jason Wang
[not found] ` <1691466855.2312648-3-xuanzhuo@linux.alibaba.com>
2023-08-08 3:59 ` Jason Wang
2023-08-10 1:56 ` Xuan Zhuo
2023-08-10 6:37 ` Jason Wang
2023-08-10 6:40 ` Michael S. Tsirkin
2023-08-10 6:39 ` Michael S. Tsirkin
2023-08-10 6:47 ` Xuan Zhuo
2023-07-10 3:42 ` [PATCH vhost v11 06/10] virtio_ring: skip unmap for premapped Xuan Zhuo
2023-07-13 3:50 ` Jason Wang
2023-07-13 4:02 ` Xuan Zhuo
2023-07-13 4:21 ` Jason Wang
2023-07-13 5:45 ` Xuan Zhuo
2023-07-10 3:42 ` [PATCH vhost v11 07/10] virtio_ring: correct the expression of the description of virtqueue_resize() Xuan Zhuo
2023-07-10 3:42 ` [PATCH vhost v11 08/10] virtio_ring: separate the logic of reset/enable from virtqueue_resize Xuan Zhuo
2023-07-10 3:42 ` [PATCH vhost v11 09/10] virtio_ring: introduce virtqueue_reset() Xuan Zhuo
2023-07-10 3:42 ` [PATCH vhost v11 10/10] virtio_net: merge dma operation for one page Xuan Zhuo
2023-07-10 9:40 ` Michael S. Tsirkin
2023-07-10 10:18 ` Xuan Zhuo
2023-07-10 11:59 ` Michael S. Tsirkin
2023-07-10 12:38 ` Xuan Zhuo
2023-07-11 2:36 ` Jason Wang
2023-07-11 2:40 ` Xuan Zhuo
2023-07-11 2:58 ` Jason Wang
2023-07-12 7:54 ` Xuan Zhuo
2023-07-12 8:32 ` Xuan Zhuo
2023-07-12 8:37 ` Jason Wang
2023-07-12 8:38 ` Xuan Zhuo
2023-07-14 10:37 ` Michael S. Tsirkin
2023-07-19 3:21 ` Xuan Zhuo
2023-07-19 8:55 ` Michael S. Tsirkin
2023-07-19 9:38 ` Jason Wang
2023-07-19 9:51 ` Michael S. Tsirkin
2023-07-20 2:26 ` Xuan Zhuo
2023-07-20 2:24 ` Xuan Zhuo
2023-07-13 4:20 ` Jason Wang
2023-07-13 5:53 ` Xuan Zhuo
2023-07-13 6:51 ` Xuan Zhuo
2023-07-14 3:56 ` Jason Wang
2023-07-20 6:23 ` Christoph Hellwig
2023-07-20 7:41 ` Jason Wang
2023-07-20 8:21 ` Christoph Hellwig
2023-07-13 7:00 ` Xuan Zhuo
2023-07-14 3:57 ` Jason Wang
2023-07-14 3:58 ` Xuan Zhuo
2023-07-14 5:45 ` 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=20230801084510.1c2460b9@kernel.org \
--to=kuba@kernel.org \
--cc=asml.silence@gmail.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=hawk@kernel.org \
--cc=hch@infradead.org \
--cc=jasowang@redhat.com \
--cc=john.fastabend@gmail.com \
--cc=mst@redhat.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=virtualization@lists.linux-foundation.org \
--cc=xuanzhuo@linux.alibaba.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;
as well as URLs for NNTP newsgroup(s).