linux-kernel - Re: [PATCH v3] nf_conntrack: sysctl: expose gc worker scan interval via sysctl

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <aRsuU57juCvsMBKE@strlen.de>
Date: Mon, 17 Nov 2025 15:16:51 +0100
From: Florian Westphal <fw@...len.de>
To: Vimal Agrawal <avimalin@...il.com>
Cc: vimal.agrawal@...hos.com, linux-kernel@...r.kernel.org,
	pablo@...filter.org, netfilter-devel@...r.kernel.org,
	anirudh.gupta@...hos.com
Subject: Re: [PATCH v3] nf_conntrack: sysctl: expose gc worker scan interval
 via sysctl

Vimal Agrawal <avimalin@...il.com> wrote:
> How about we keep only the minimum expiry time out of all (one which
> is going to expire next) and if there are listeners then we just
> schedule gc_worker to that minimum time (and do this only if there are
> ctnetlink listeners in userspace)? so that we don't delay even if
> there is 1 such low value timer expiring in the near future.
> Do you think it will cause too frequent wake ups for gc_worker?

Yes, I am worried about 1k extra wakeups/s in worst case
(we always have exactly one flow which has 1 jiffy remaining).

Plus we might have per-net workers in the future, so it could get
worse.