linux-kernel - Re: [PATCH] modules: mark __inittest/__exittest as __maybe

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]

Date:   Thu, 2 Feb 2017 11:19:04 +0100
From:   Arnd Bergmann <arnd@...db.de>
To:     Rusty Russell <rusty@...tcorp.com.au>
Cc:     Jessica Yu <jeyu@...hat.com>, Jiri Kosina <jikos@...nel.org>,
        Paul Gortmaker <paul.gortmaker@...driver.com>,
        Miroslav Benes <mbenes@...e.cz>,
        Anson Jacob <ansonjacob.aj@...il.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] modules: mark __inittest/__exittest as __maybe_unused

On Thu, Feb 2, 2017 at 10:25 AM, Rusty Russell <rusty@...tcorp.com.au> wrote:
> Arnd Bergmann <arnd@...db.de> writes:
>> clang warns about unused inline functions by default:
>>
>> arch/arm/crypto/aes-cipher-glue.c:68:1: warning: unused function '__inittest' [-Wunused-function]
>> arch/arm/crypto/aes-cipher-glue.c:69:1: warning: unused function '__exittest' [-Wunused-function]
>>
>> As these appear in every single module, let's just disable the warnings by marking the
>> two functions as __maybe_unused.
>
> Um, won't you have to do that to hundreds of kernel headers?  Why
> module.h?

clang specifically warns about inline functions that are defined in a
.c file but not used
there, but it is sensible enough to not warn about unused inline
functions that are defined
in a header.

In an ARM allmodconfig build, I currently see 178 .c files[1] that
have unused inline functions.
The proper way to deal with them is probably to move the warning into
the "make W=1"
level to hide it by default and then do one driver at a time.

The module.h definitions are special because the inline function is
defined through a
macro that gets evaluated by almost every loadable module, and we get
a warning for
every one of them, which the subsystem maintainers cannot deal with by
changing their
code locally.

    Arnd

[1] http://pastebin.com/pnHvbHQ3