netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Michael S. Tsirkin" <mst@redhat.com>
To: Xuan Zhuo <xuanzhuo@linux.alibaba.com>
Cc: netdev@vger.kernel.org, "David S. Miller" <davem@davemloft.net>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Jason Wang <jasowang@redhat.com>,
	Alexei Starovoitov <ast@kernel.org>,
	Daniel Borkmann <daniel@iogearbox.net>,
	Jesper Dangaard Brouer <hawk@kernel.org>,
	John Fastabend <john.fastabend@gmail.com>,
	virtualization@lists.linux-foundation.org, bpf@vger.kernel.org
Subject: Re: [PATCH net-next v2 00/21] virtio-net: support AF_XDP zero copy
Date: Thu, 9 Nov 2023 03:19:04 -0500	[thread overview]
Message-ID: <20231109031728-mutt-send-email-mst@kernel.org> (raw)
In-Reply-To: <20231107031227.100015-1-xuanzhuo@linux.alibaba.com>

On Tue, Nov 07, 2023 at 11:12:06AM +0800, Xuan Zhuo wrote:
> ## AF_XDP
> 
> XDP socket(AF_XDP) is an excellent bypass kernel network framework. The zero
> copy feature of xsk (XDP socket) needs to be supported by the driver. The
> performance of zero copy is very good. mlx5 and intel ixgbe already support
> this feature, This patch set allows virtio-net to support xsk's zerocopy xmit
> feature.
> 
> At present, we have completed some preparation:
> 
> 1. vq-reset (virtio spec and kernel code)
> 2. virtio-core premapped dma
> 3. virtio-net xdp refactor
> 
> So it is time for Virtio-Net to complete the support for the XDP Socket
> Zerocopy.
> 
> Virtio-net can not increase the queue num at will, so xsk shares the queue with
> kernel.
> 
> On the other hand, Virtio-Net does not support generate interrupt from driver
> manually, so when we wakeup tx xmit, we used some tips. If the CPU run by TX
> NAPI last time is other CPUs, use IPI to wake up NAPI on the remote CPU. If it
> is also the local CPU, then we wake up napi directly.
> 
> This patch set includes some refactor to the virtio-net to let that to support
> AF_XDP.
> 
> ## performance
> 
> ENV: Qemu with vhost-user(polling mode).
> Host CPU: Intel(R) Xeon(R) Platinum 8163 CPU @ 2.50GHz
> 
> ### virtio PMD in guest with testpmd
> 
> testpmd> show port stats all
> 
>  ######################## NIC statistics for port 0 ########################
>  RX-packets: 19531092064 RX-missed: 0     RX-bytes: 1093741155584
>  RX-errors: 0
>  RX-nombuf: 0
>  TX-packets: 5959955552 TX-errors: 0     TX-bytes: 371030645664
> 
> 
>  Throughput (since last show)
>  Rx-pps:   8861574     Rx-bps:  3969985208
>  Tx-pps:   8861493     Tx-bps:  3969962736
>  ############################################################################
> 
> ### AF_XDP PMD in guest with testpmd
> 
> testpmd> show port stats all
> 
>   ######################## NIC statistics for port 0  ########################
>   RX-packets: 68152727   RX-missed: 0          RX-bytes:  3816552712
>   RX-errors: 0
>   RX-nombuf:  0
>   TX-packets: 68114967   TX-errors: 33216      TX-bytes:  3814438152
> 
>   Throughput (since last show)
>   Rx-pps:      6333196          Rx-bps:   2837272088
>   Tx-pps:      6333227          Tx-bps:   2837285936
>   ############################################################################
> 
> But AF_XDP consumes more CPU for tx and rx napi(100% and 86%).
> 
> ## maintain
> 
> I am currently a reviewer for virtio-net. I commit to maintain AF_XDP support in
> virtio-net.
> 
> Please review.
> 
> Thanks.
> 
> v2
>     1. wakeup uses the way of GVE. No send ipi to wakeup napi on remote cpu.
>     2. remove rcu. Because we synchronize all operat, so the rcu is not needed.
>     3. split the commit "move to virtio_net.h" in last patch set. Just move the
>        struct/api to header when we use them.
>     4. add comments for some code
> 
> v1:
>     1. remove two virtio commits. Push this patchset to net-next
>     2. squash "virtio_net: virtnet_poll_tx support rescheduled" to xsk: support tx
>     3. fix some warnings
> 
> 
> Xuan Zhuo (21):
>   virtio_net: rename free_old_xmit_skbs to free_old_xmit
>   virtio_net: unify the code for recycling the xmit ptr
>   virtio_net: independent directory
>   virtio_net: move core structures to virtio_net.h
>   virtio_net: add prefix virtnet to all struct inside virtio_net.h
>   virtio_net: separate virtnet_rx_resize()
>   virtio_net: separate virtnet_tx_resize()
>   virtio_net: sq support premapped mode
>   virtio_net: xsk: bind/unbind xsk
>   virtio_net: xsk: prevent disable tx napi
>   virtio_net: move some api to header
>   virtio_net: xsk: tx: support tx
>   virtio_net: xsk: tx: support wakeup
>   virtio_net: xsk: tx: virtnet_free_old_xmit() distinguishes xsk buffer
>   virtio_net: xsk: tx: virtnet_sq_free_unused_buf() check xsk buffer
>   virtio_net: xsk: rx: introduce add_recvbuf_xsk()
>   virtio_net: xsk: rx: skip dma unmap when rq is bind with AF_XDP
>   virtio_net: xsk: rx: introduce receive_xsk() to recv xsk buffer
>   virtio_net: xsk: rx: virtnet_rq_free_unused_buf() check xsk buffer
>   virtio_net: update tx timeout record
>   virtio_net: xdp_features add NETDEV_XDP_ACT_XSK_ZEROCOPY
> 
>  MAINTAINERS                                 |   2 +-
>  drivers/net/Kconfig                         |   8 +-
>  drivers/net/Makefile                        |   2 +-
>  drivers/net/virtio/Kconfig                  |  13 +
>  drivers/net/virtio/Makefile                 |   8 +
>  drivers/net/{virtio_net.c => virtio/main.c} | 630 +++++++++-----------
>  drivers/net/virtio/virtio_net.h             | 346 +++++++++++
>  drivers/net/virtio/xsk.c                    | 498 ++++++++++++++++
>  drivers/net/virtio/xsk.h                    |  32 +
>  9 files changed, 1174 insertions(+), 365 deletions(-)
>  create mode 100644 drivers/net/virtio/Kconfig
>  create mode 100644 drivers/net/virtio/Makefile
>  rename drivers/net/{virtio_net.c => virtio/main.c} (92%)
>  create mode 100644 drivers/net/virtio/virtio_net.h
>  create mode 100644 drivers/net/virtio/xsk.c
>  create mode 100644 drivers/net/virtio/xsk.h


Too much code duplications for my taste. Would there maybe be less if
everything was in a single file?


> --
> 2.32.0.3.g01195cf9f


  parent reply	other threads:[~2023-11-09  8:19 UTC|newest]

Thread overview: 61+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-11-07  3:12 [PATCH net-next v2 00/21] virtio-net: support AF_XDP zero copy Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 01/21] virtio_net: rename free_old_xmit_skbs to free_old_xmit Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 02/21] virtio_net: unify the code for recycling the xmit ptr Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 03/21] virtio_net: independent directory Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 04/21] virtio_net: move core structures to virtio_net.h Xuan Zhuo
2023-11-09  6:03   ` Jason Wang
2023-11-07  3:12 ` [PATCH net-next v2 05/21] virtio_net: add prefix virtnet to all struct inside virtio_net.h Xuan Zhuo
2023-11-09  6:04   ` Jason Wang
2023-11-07  3:12 ` [PATCH net-next v2 06/21] virtio_net: separate virtnet_rx_resize() Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 07/21] virtio_net: separate virtnet_tx_resize() Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 08/21] virtio_net: sq support premapped mode Xuan Zhuo
2023-11-09  6:37   ` Jason Wang
2023-11-09 10:58     ` Xuan Zhuo
2023-11-14  3:26       ` Jason Wang
2023-11-14  3:28         ` Xuan Zhuo
2023-11-14  3:55           ` Jason Wang
2023-11-14  3:57             ` Xuan Zhuo
2023-11-14  4:27               ` Jason Wang
2023-11-14  4:45                 ` Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 09/21] virtio_net: xsk: bind/unbind xsk Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 10/21] virtio_net: xsk: prevent disable tx napi Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 11/21] virtio_net: move some api to header Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 12/21] virtio_net: xsk: tx: support tx Xuan Zhuo
2023-11-09  8:09   ` Michael S. Tsirkin
2023-11-09 11:06     ` Xuan Zhuo
2023-11-09 11:58       ` Michael S. Tsirkin
2023-11-10  1:51         ` Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 13/21] virtio_net: xsk: tx: support wakeup Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 14/21] virtio_net: xsk: tx: virtnet_free_old_xmit() distinguishes xsk buffer Xuan Zhuo
2023-11-09 11:11   ` Michael S. Tsirkin
2023-11-09 11:16     ` Xuan Zhuo
2023-11-09 11:59       ` Michael S. Tsirkin
2023-11-10  1:44         ` Xuan Zhuo
2023-11-10  5:32           ` Michael S. Tsirkin
2023-11-10  5:50             ` Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 15/21] virtio_net: xsk: tx: virtnet_sq_free_unused_buf() check " Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 16/21] virtio_net: xsk: rx: introduce add_recvbuf_xsk() Xuan Zhuo
2023-11-09  8:12   ` Michael S. Tsirkin
2023-11-09 11:11     ` Xuan Zhuo
2023-11-09 16:26       ` Maciej Fijalkowski
2023-11-10  2:38         ` Xuan Zhuo
2023-11-13 16:00           ` Maciej Fijalkowski
2023-11-14  3:16             ` Xuan Zhuo
2023-11-10  3:04     ` Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 17/21] virtio_net: xsk: rx: skip dma unmap when rq is bind with AF_XDP Xuan Zhuo
2023-11-09  8:15   ` Michael S. Tsirkin
2023-11-09 11:10     ` Xuan Zhuo
2023-11-09 12:00       ` Michael S. Tsirkin
2023-11-10  1:47         ` Xuan Zhuo
2023-11-10  5:33           ` Michael S. Tsirkin
2023-11-10  5:51             ` Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 18/21] virtio_net: xsk: rx: introduce receive_xsk() to recv xsk buffer Xuan Zhuo
2023-11-13 16:11   ` Maciej Fijalkowski
2023-11-14  3:43     ` Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 19/21] virtio_net: xsk: rx: virtnet_rq_free_unused_buf() check " Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 20/21] virtio_net: update tx timeout record Xuan Zhuo
2023-11-07  3:12 ` [PATCH net-next v2 21/21] virtio_net: xdp_features add NETDEV_XDP_ACT_XSK_ZEROCOPY Xuan Zhuo
2023-11-07 18:01 ` [PATCH net-next v2 00/21] virtio-net: support AF_XDP zero copy Jakub Kicinski
2023-11-08  5:49   ` Xuan Zhuo
2023-11-09  8:19 ` Michael S. Tsirkin [this message]
2023-11-09 10:37   ` Xuan Zhuo

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=20231109031728-mutt-send-email-mst@kernel.org \
    --to=mst@redhat.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=jasowang@redhat.com \
    --cc=john.fastabend@gmail.com \
    --cc=kuba@kernel.org \
    --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).