From: "Björn Töpel" <bjorn.topel@intel.com>
To: "Maxim Mikityanskiy" <maximmi@nvidia.com>,
"Björn Töpel" <bjorn.topel@gmail.com>,
ast@kernel.org, daniel@iogearbox.net, netdev@vger.kernel.org,
bpf@vger.kernel.org
Cc: magnus.karlsson@intel.com, maciej.fijalkowski@intel.com,
kuba@kernel.org, jonathan.lemon@gmail.com, davem@davemloft.net,
hawk@kernel.org, john.fastabend@gmail.com,
ciara.loftus@intel.com, weqaar.a.janjua@intel.com
Subject: Re: [PATCH bpf-next v2 0/8] Introduce bpf_redirect_xsk() helper
Date: Wed, 20 Jan 2021 14:27:58 +0100 [thread overview]
Message-ID: <9fd68fd5-6a9f-3c36-1b83-4ba587387f5d@intel.com> (raw)
In-Reply-To: <7dcee85b-b3ef-947f-f433-03ad7066c5dd@nvidia.com>
On 2021-01-20 14:15, Maxim Mikityanskiy wrote:
> On 2021-01-19 17:50, Björn Töpel wrote:
>> This series extends bind() for XDP sockets, so that the bound socket
>> is added to the netdev_rx_queue _rx array in the netdevice. We call
>> this to register the socket. To redirect packets to the registered
>> socket, a new BPF helper is used: bpf_redirect_xsk().
>>
>> For shared XDP sockets, only the first bound socket is
>> registered. Users that need more complex setup has to use XSKMAP and
>> bpf_redirect_map().
>>
>> Now, why would one use bpf_redirect_xsk() over the regular
>> bpf_redirect_map() helper?
>>
>> * Better performance!
>> * Convenience; Most user use one socket per queue. This scenario is
>> what registered sockets support. There is no need to create an
>> XSKMAP. This can also reduce complexity from containerized setups,
>> where users might what to use XDP sockets without CAP_SYS_ADMIN
>> capabilities.
>>
>> The first patch restructures xdp_do_redirect() a bit, to make it
>> easier to add the new helper. This restructure also give us a slight
>> performance benefit. The following three patches extends bind() and
>> adds the new helper. After that, two libbpf patches that selects XDP
>> program based on what kernel is running. Finally, selftests for the new
>> functionality is added.
>>
>> Note that the libbpf "auto-selection" is based on kernel version, so
>> it is hard coded to the "-next" version (5.12). If you would like to
>> try this is out, you will need to change the libbpf patch locally!
>>
>> Thanks to Maciej and Magnus for the internal review/comments!
>>
>> Performance (rxdrop, zero-copy)
>>
>> Baseline
>> Two cores: 21.3 Mpps
>> One core: 24.5 Mpps
>
> Two cores is slower? It used to be faster all the time, didn't it?
>
Up until busy-polling. Note that I'm using a busy-poll napi budget of 512.
>> Patched
>> Two cores, bpf_redirect_map: 21.7 Mpps + 2%
>> One core, bpf_redirect_map: 24.9 Mpps + 2%
>>
>> Two cores, bpf_redirect_xsk: 24.0 Mpps +13%
>
> Nice, impressive improvement!
>
Thanks! Getting rid of the queue/netdev checks really payed off!
Björn
next prev parent reply other threads:[~2021-01-20 20:10 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-01-19 15:50 [PATCH bpf-next v2 0/8] Introduce bpf_redirect_xsk() helper Björn Töpel
2021-01-19 15:50 ` [PATCH bpf-next v2 1/8] xdp: restructure redirect actions Björn Töpel
2021-01-20 12:44 ` Toke Høiland-Jørgensen
2021-01-20 13:40 ` Björn Töpel
2021-01-20 14:52 ` Toke Høiland-Jørgensen
2021-01-20 15:49 ` Björn Töpel
2021-01-20 16:30 ` Toke Høiland-Jørgensen
2021-01-20 17:26 ` Björn Töpel
2021-01-19 15:50 ` [PATCH bpf-next v2 2/8] xsk: remove explicit_free parameter from __xsk_rcv() Björn Töpel
2021-01-19 15:50 ` [PATCH bpf-next v2 3/8] xsk: fold xp_assign_dev and __xp_assign_dev Björn Töpel
2021-01-19 15:50 ` [PATCH bpf-next v2 4/8] xsk: register XDP sockets at bind(), and add new AF_XDP BPF helper Björn Töpel
2021-01-20 8:25 ` kernel test robot
2021-01-20 8:41 ` Björn Töpel
2021-01-20 8:50 ` kernel test robot
2021-01-20 12:50 ` Toke Høiland-Jørgensen
2021-01-20 13:25 ` Björn Töpel
2021-01-20 14:54 ` Toke Høiland-Jørgensen
2021-01-20 15:18 ` Björn Töpel
2021-01-20 17:29 ` Toke Høiland-Jørgensen
2021-01-20 18:22 ` Björn Töpel
2021-01-20 20:26 ` Toke Høiland-Jørgensen
2021-01-20 21:15 ` Alexei Starovoitov
2021-01-21 8:18 ` Björn Töpel
2021-01-19 15:50 ` [PATCH bpf-next v2 5/8] libbpf, xsk: select AF_XDP BPF program based on kernel version Björn Töpel
2021-01-20 12:52 ` Toke Høiland-Jørgensen
2021-01-20 13:25 ` Björn Töpel
2021-01-20 14:49 ` Björn Töpel
2021-01-20 15:11 ` Toke Høiland-Jørgensen
2021-01-20 15:27 ` Björn Töpel
2021-01-20 17:30 ` Toke Høiland-Jørgensen
2021-01-20 18:25 ` Alexei Starovoitov
2021-01-20 18:30 ` Björn Töpel
2021-01-20 14:56 ` Toke Høiland-Jørgensen
2021-01-19 15:50 ` [PATCH bpf-next v2 6/8] libbpf, xsk: select bpf_redirect_xsk(), if supported Björn Töpel
2021-01-19 15:50 ` [PATCH bpf-next v2 7/8] selftest/bpf: add XDP socket tests for bpf_redirect_{xsk, map}() Björn Töpel
2021-01-21 7:39 ` Andrii Nakryiko
2021-01-21 12:31 ` Björn Töpel
2021-01-19 15:50 ` [PATCH bpf-next v2 8/8] selftest/bpf: remove a lot of ifobject casting in xdpxceiver Björn Töpel
2021-01-20 13:15 ` [PATCH bpf-next v2 0/8] Introduce bpf_redirect_xsk() helper Maxim Mikityanskiy
2021-01-20 13:27 ` Björn Töpel [this message]
2021-01-20 15:57 ` Jesper Dangaard Brouer
2021-01-20 16:19 ` Maciej Fijalkowski
2021-01-21 17:01 ` Jesper Dangaard Brouer
2021-01-22 8:59 ` Magnus Karlsson
2021-01-22 9:45 ` Maciej Fijalkowski
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=9fd68fd5-6a9f-3c36-1b83-4ba587387f5d@intel.com \
--to=bjorn.topel@intel.com \
--cc=ast@kernel.org \
--cc=bjorn.topel@gmail.com \
--cc=bpf@vger.kernel.org \
--cc=ciara.loftus@intel.com \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=hawk@kernel.org \
--cc=john.fastabend@gmail.com \
--cc=jonathan.lemon@gmail.com \
--cc=kuba@kernel.org \
--cc=maciej.fijalkowski@intel.com \
--cc=magnus.karlsson@intel.com \
--cc=maximmi@nvidia.com \
--cc=netdev@vger.kernel.org \
--cc=weqaar.a.janjua@intel.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).