linux-kernel - Re: [PATCH v2] printk_legacy_map: use LD_WAIT_CONFIG instead of LD_WAIT

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: <20251024105222.RZPI5xWT@linutronix.de>
Date: Fri, 24 Oct 2025 12:52:22 +0200
From: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To: Oleg Nesterov <oleg@...hat.com>
Cc: Petr Mladek <pmladek@...e.com>, John Ogness <john.ogness@...utronix.de>,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2] printk_legacy_map: use LD_WAIT_CONFIG instead of
 LD_WAIT_SLEEP

On 2025-10-24 12:40:56 [+0200], Oleg Nesterov wrote:
> On 10/23, Sebastian Andrzej Siewior wrote:
> >
> > On 2025-10-23 12:32:34 [+0200], Oleg Nesterov wrote:
> > > printk_legacy_map is used on !PREEMPT_RT to avoid false positives from
> > > CONFIG_PROVE_RAW_LOCK_NESTING about raw_spinlock/spinlock nesting.
> >
> > Could we please get rid of CONFIG_PROVE_RAW_LOCK_NESTING here? This is
> > lockdep internal implementation and has nothing to do with printk or
> > anything.
> 
> OK, but let me ensure I didn't miss something (again ;).
> 
> I mentioned CONFIG_PROVE_RAW_LOCK_NESTING in the changelog because if
> CONFIG_PROVE_RAW_LOCK_NESTING=n, then LD_WAIT_CONFIG == LD_WAIT_SPIN
> and lockdep will not complain if spinlock_t nests inside raw_spinlock_t.
> 
> IOW, without CONFIG_PROVE_RAW_LOCK_NESTING printk_legacy_map is not
> really needed.

This is correct but CONFIG_PROVE_RAW_LOCK_NESTING is kind of a must and
should not be an option. This just verifies what we have in
	Documentation/locking/locktypes.rst
and this is not an option.
As I explained it earlier in this (or other) thread I removed it and
brought it back because on non-RT architectures spilled warnings and
people did not want to apply a larger pile of patches because "it did
work" and "I don't use RT" and I have so much time available for
arguing.

It is not an option architectures which support RT. It is one on
architectures which do not support RT so they don't have warning and may
debug & fix them if they want to.

> Oleg.

Sebastian