From: "Irlanki Sandeep" <irlanki.s@samsung.com>
To: "'Jakub Kicinski'" <kuba@kernel.org>
Cc: <netdev@vger.kernel.org>, <bpf@vger.kernel.org>,
<edumazet@google.com>, <davem@davemloft.net>, <pabeni@redhat.com>,
<ncardwell@google.com>, <ast@kernel.org>, <daniel@iogearbox.net>,
<andrii@kernel.org>, <corbet@lwn.net>,
<linux-kernel@vger.kernel.org>, <lorenzo@google.com>,
<maze@google.com>, <sporeba@google.com>, <motomuman@google.com>,
<madhan.raj@samsung.com>, <srihari.k@samsung.com>,
<g.pokhra@samsung.com>, <r.kumawat@samsung.com>,
<raj.kumars@samsung.com>, <ts413.lee@samsung.com>,
<ramyashree.s@samsung.com>, <daeil2.hwang@samsung.com>,
"'CPGS'" <cpgs@samsung.com>
Subject: RE: [PATCH net-next v6] tcp: add BPF kfuncs to set per-socket ECN mode and AccECN option
Date: Wed, 14 Oct 2026 13:19:31 +0530 [thread overview]
Message-ID: <117401dd5bb0$8f05b2f0$ad1118d0$@samsung.com> (raw)
In-Reply-To: <20261007170028.318acb6b@kernel.org>
On Wed, 7 Oct 2026 17:00:28 -0700, Jakub Kicinski <kuba@kernel.org> wrote:
> On Thu, 1 Oct 2026 18:57:59 +0530 Irlanki Sandeep wrote:
> > Suggested-by: Jakub Kicinski <kuba@kernel.org>
>
> This needs an ACK from TCP maintainers. If you respin please make sure
> to drop this tag, this tag is for cases where someone suggested sending
> the patch as a whole. Not for changes requested during normal course of
> review.
Understood, thanks for clarifying the tag semantics. I will drop the
Suggested-by tag in v7 if a respin is needed.
> There was a false positive from sparse regarding the use of kfunc,
> I kicked off the AI review manually, please check the results here
> in an hour or so:
>
https://netdev-ai.bots.linux.dev/sashiko/#/patchset/20261001132759.1076145-1
-irlanki.s@samsung.com
Thanks for triggering the manual AI review. The results show 0 Critical,
0 High, 1 Medium, and 1 Low:
1. [Medium] SYN arrival race before LISTEN_CB completes:
The review identifies that inet_csk_listen_start() hashes the socket
before tcp_call_bpf(BPF_SOCK_OPS_TCP_LISTEN_CB) runs. A SYN arriving on
another CPU in this window will see ecn_mode as UNSPEC and fall back to
the sysctl. This is a preexisting limitation affecting any
BPF_SOCK_OPS_TCP_LISTEN_CB program that configures listener behavior,
not specific to this patch. I will document this fallback behavior in
the kernel-doc and commit message for v7.
2. [Low] Selftest scope (API validation vs. packet inspection):
The selftest validates the kfunc API layer: argument validation, return
codes, verifier safety (sk_is_tcp() guard), and struct field
mutation/inheritance. It does not include packet-level inspection. This
is comparable in scope to progs/setget_sockopt.c, which similarly tests
bpf_setsockopt/bpf_getsockopt round-trips without packet-level wire
verification.
Awaiting feedback from Eric and the TCP maintainers.
Thanks,
Sandeep
prev parent reply other threads:[~2026-10-14 7:49 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <CGME20261001132701epcas5p2da9b0fa7ca3bcef1ff84a62f255bc573@epcas5p2.samsung.com>
2026-10-01 13:27 ` [PATCH net-next v6] tcp: add BPF kfuncs to set per-socket ECN mode and AccECN option Irlanki Sandeep
2026-10-08 0:00 ` Jakub Kicinski
2026-10-14 7:49 ` Irlanki Sandeep [this message]
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='117401dd5bb0$8f05b2f0$ad1118d0$@samsung.com' \
--to=irlanki.s@samsung.com \
--cc=andrii@kernel.org \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=corbet@lwn.net \
--cc=cpgs@samsung.com \
--cc=daeil2.hwang@samsung.com \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=edumazet@google.com \
--cc=g.pokhra@samsung.com \
--cc=kuba@kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=lorenzo@google.com \
--cc=madhan.raj@samsung.com \
--cc=maze@google.com \
--cc=motomuman@google.com \
--cc=ncardwell@google.com \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.com \
--cc=r.kumawat@samsung.com \
--cc=raj.kumars@samsung.com \
--cc=ramyashree.s@samsung.com \
--cc=sporeba@google.com \
--cc=srihari.k@samsung.com \
--cc=ts413.lee@samsung.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.