From: Stanislav Fomichev <stfomichev@gmail.com>
To: Eryk Kubanski <e.kubanski@partner.samsung.com>
Cc: "netdev@vger.kernel.org" <netdev@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"bjorn@kernel.org" <bjorn@kernel.org>,
"magnus.karlsson@intel.com" <magnus.karlsson@intel.com>,
"maciej.fijalkowski@intel.com" <maciej.fijalkowski@intel.com>,
"jonathan.lemon@gmail.com" <jonathan.lemon@gmail.com>
Subject: Re: Re: [PATCH bpf v2] xsk: Fix out of order segment free in __xsk_generic_xmit()
Date: Mon, 2 Jun 2025 08:28:51 -0700 [thread overview]
Message-ID: <aD3DM4elo_Xt82LE@mini-arch> (raw)
In-Reply-To: <20250602092754eucms1p1b99e467d1483531491c5b43b23495e14@eucms1p1>
On 06/02, Eryk Kubanski wrote:
> > I'm not sure I understand what's the issue here. If you're using the
> > same XSK from different CPUs, you should take care of the ordering
> > yourself on the userspace side?
>
> It's not a problem with user-space Completion Queue READER side.
> Im talking exclusively about kernel-space Completion Queue WRITE side.
>
> This problem can occur when multiple sockets are bound to the same
> umem, device, queue id. In this situation Completion Queue is shared.
> This means it can be accessed by multiple threads on kernel-side.
> Any use is indeed protected by spinlock, however any write sequence
> (Acquire write slot as writer, write to slot, submit write slot to reader)
> isn't atomic in any way and it's possible to submit not-yet-sent packet
> descriptors back to user-space as TX completed.
>
> Up untill now, all write-back operations had two phases, each phase
> locks the spinlock and unlocks it:
> 1) Acquire slot + Write descriptor (increase cached-writer by N + write values)
> 2) Submit slot to the reader (increase writer by N)
>
> Slot submission was solely based on the timing. Let's consider situation,
> where two different threads issue a syscall for two different AF_XDP sockets
> that are bound to the same umem, dev, queue-id.
>
> AF_XDP setup:
>
> kernel-space
>
> Write Read
> +--+ +--+
> | | | |
> | | | |
> | | | |
> Completion | | | | Fill
> Queue | | | | Queue
> | | | |
> | | | |
> | | | |
> | | | |
> +--+ +--+
> Read Write
> user-space
>
>
> +--------+ +--------+
> | AF_XDP | | AF_XDP |
> +--------+ +--------+
>
>
>
>
>
> Possible out-of-order scenario:
>
>
> writer cached_writer1 cached_writer2
> | | |
> | | |
> | | |
> | | |
> +--------------|--------|--------|--------|--------|--------|--------|----------------------------------------------+
> | | | | | | | | |
> Completion Queue | | | | | | | | |
> | | | | | | | | |
> +--------------|--------|--------|--------|--------|--------|--------|----------------------------------------------+
> | | |
> | | |
> |-----------------| |
> A) T1 syscall | |
> writes 2 | |
> descriptors |-----------------------------------|
> B) T2 syscall writes 4 descriptors
>
>
>
>
> Notes:
> 1) T1 and T2 AF_XDP sockets are two different sockets,
> __xsk_generic_xmit will obtain two different mutexes.
> 2) T1 and T2 can be executed simultaneously, there is no
> critical section whatsoever between them.
XSK represents a single queue and each queue is single producer single
consumer. The fact that you can dup a socket and call sendmsg from
different threads/processes does not lift that restriction. I think
if you add synchronization on the userspace (lock(); sendmsg();
unlock();), that should help, right?
next prev parent reply other threads:[~2025-06-02 15:28 UTC|newest]
Thread overview: 18+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <CGME20250530103506eucas1p1e4091678f4157b928ddfa6f6534a0009@eucas1p1.samsung.com>
2025-05-30 10:34 ` [PATCH bpf v2] xsk: Fix out of order segment free in __xsk_generic_xmit() e.kubanski
2025-05-30 11:56 ` Eryk Kubanski
2025-05-30 16:07 ` Stanislav Fomichev
2025-06-02 9:27 ` Eryk Kubanski
2025-06-02 15:28 ` Stanislav Fomichev [this message]
2025-06-02 15:58 ` Eryk Kubanski
2025-06-02 16:03 ` Maciej Fijalkowski
2025-06-02 16:18 ` Eryk Kubanski
2025-06-04 13:50 ` Maciej Fijalkowski
2025-06-04 14:15 ` Eryk Kubanski
2025-06-09 19:41 ` Maciej Fijalkowski
2025-06-10 9:35 ` Eryk Kubanski
[not found] ` <CGME20250530103506eucas1p1e4091678f4157b928ddfa6f6534a0009@eucms1p3>
2025-06-10 9:11 ` Eryk Kubanski
2025-06-11 13:10 ` Maciej Fijalkowski
2025-07-03 23:37 ` Jason Xing
2025-07-04 12:34 ` Maciej Fijalkowski
2025-07-04 15:29 ` Jason Xing
2025-06-04 14:41 ` kernel test robot
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=aD3DM4elo_Xt82LE@mini-arch \
--to=stfomichev@gmail.com \
--cc=bjorn@kernel.org \
--cc=e.kubanski@partner.samsung.com \
--cc=jonathan.lemon@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=maciej.fijalkowski@intel.com \
--cc=magnus.karlsson@intel.com \
--cc=netdev@vger.kernel.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).