linux-kernel - Re: [PATCH v4] fs/namespace: defer RCU sync for MNT

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

Message-ID: <20250417153126.QrVXSjt-@linutronix.de>
Date: Thu, 17 Apr 2025 17:31:26 +0200
From: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To: Christian Brauner <brauner@...nel.org>
Cc: Ian Kent <raven@...maw.net>, Mark Brown <broonie@...nel.org>,
	Eric Chanudet <echanude@...hat.com>,
	Alexander Viro <viro@...iv.linux.org.uk>, Jan Kara <jack@...e.cz>,
	Clark Williams <clrkwllms@...nel.org>,
	Steven Rostedt <rostedt@...dmis.org>, Ian Kent <ikent@...hat.com>,
	linux-fsdevel@...r.kernel.org, linux-kernel@...r.kernel.org,
	linux-rt-devel@...ts.linux.dev,
	Alexander Larsson <alexl@...hat.com>,
	Lucas Karpinski <lkarpins@...hat.com>, Aishwarya.TCV@....com
Subject: Re: [PATCH v4] fs/namespace: defer RCU sync for MNT_DETACH umount

On 2025-04-17 17:28:20 [+0200], Christian Brauner wrote:
> >     So if there's some userspace process with a broken NFS server and it
> >     does umount(MNT_DETACH) it will end up hanging every other
> >     umount(MNT_DETACH) on the system because the dealyed_mntput_work
> >     workqueue (to my understanding) cannot make progress.
> 
> Ok, "to my understanding" has been updated after going back and reading
> the delayed work code. Luckily it's not as bad as I thought it is
> because it's queued on system_wq which is multi-threaded so it's at
> least not causing everyone with MNT_DETACH to get stuck. I'm still
> skeptical how safe this all is. 

I would (again) throw system_unbound_wq into the game because the former
will remain on the CPU on which has been enqueued (if speaking about
multi threading).

Sebastian