linux-kernel - Re: [PATCH] irqchip/qcom-irq-combiner: Rename driver structure

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: <aPX-WBsUUXXmpqdg@hovoldconsulting.com>
Date: Mon, 20 Oct 2025 11:18:16 +0200
From: Johan Hovold <johan@...nel.org>
To: Thomas Gleixner <tglx@...utronix.de>
Cc: linux-kernel@...r.kernel.org
Subject: Re: [PATCH] irqchip/qcom-irq-combiner: Rename driver structure

On Fri, Oct 17, 2025 at 03:16:19PM +0200, Thomas Gleixner wrote:
> On Fri, Oct 17 2025 at 07:52, Johan Hovold wrote:
> > The "_probe" suffix of the driver structure name prevents modpost from
> > warning about section mismatches so replace it to catch any future
> > issues like the recently fixed probe function being incorrectly marked
> > as __init.
> 
> I take this grudgingly, but not without pointing out again that this is
> just a horrible hack. Why can't this stuff be properly annotated and
> recorded in a throwaway section so that modpost can act on reliable data
> instead of using string matching on arbitrary function names?

This apparently dates back to 2006, but some of these magic patterns
were at least dropped a few years ago with commit e1dc1bfe5b27
("modpost: remove more symbol patterns from the section check
whitelist").

With the irqchip driver fixed and soon hopefully the stm clocksource
driver [1], we can retire also the special handling of the "_probe"
suffix. I've just sent a patch. [2]

Johan


[1] https://lore.kernel.org/lkml/20251017054943.7195-1-johan@kernel.org/
[2] https://lore.kernel.org/lkml/20251020091613.22562-1-johan@kernel.org/