linux-hardening - Re: [PATCH 2/2] gcc-plugins: disable plugins when gmp.h is unavailable

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20240313154111-53729221-b8fb-4969-ac4d-ea74861f6c84@linutronix.de>
Date: Wed, 13 Mar 2024 15:42:17 +0100
From: Thomas Weißschuh <thomas.weissschuh@...utronix.de>
To: Masahiro Yamada <masahiroy@...nel.org>
Cc: Kees Cook <keescook@...omium.org>, linux-hardening@...r.kernel.org, 
	linux-kernel@...r.kernel.org, linux-kbuild@...r.kernel.org
Subject: Re: [PATCH 2/2] gcc-plugins: disable plugins when gmp.h is
 unavailable

On Wed, Mar 13, 2024 at 12:47:59PM +0900, Masahiro Yamada wrote:
> On Wed, Mar 13, 2024 at 6:53 AM Kees Cook <keescook@...omium.org> wrote:
> >
> > On Tue, Mar 12, 2024 at 04:03:30PM +0100, Thomas Weißschuh wrote:
> > > The header gmp.h is meant to be picked up from the host system.
> > >
> > > When it is unavailable the plugin build fails:
> > >
> > > In file included from ../crosstools/gcc-13.2.0-nolibc/i386-linux/bin/../lib/gcc/i386-linux/13.2.0/plugin/include/gcc-plugin.h:28,
> > >                  from ../scripts/gcc-plugins/gcc-common.h:7,
> > >                  from ../scripts/gcc-plugins/stackleak_plugin.c:30:
> > > ../crosstools/gcc-13.2.0-nolibc/i386-linux/bin/../lib/gcc/i386-linux/13.2.0/plugin/include/system.h:703:10: fatal error: gmp.h: No such file or directory
> > >   703 | #include <gmp.h>
> > >       |          ^~~~~~~
> > >
> > > Signed-off-by: Thomas Weißschuh <thomas.weissschuh@...utronix.de>
> > > ---
> > >  scripts/gcc-plugins/Kconfig | 1 +
> > >  1 file changed, 1 insertion(+)
> > >
> > > diff --git a/scripts/gcc-plugins/Kconfig b/scripts/gcc-plugins/Kconfig
> > > index e383cda05367..a664fb5cdde5 100644
> > > --- a/scripts/gcc-plugins/Kconfig
> > > +++ b/scripts/gcc-plugins/Kconfig
> > > @@ -10,6 +10,7 @@ menuconfig GCC_PLUGINS
> > >       depends on HAVE_GCC_PLUGINS
> > >       depends on CC_IS_GCC
> > >       depends on $(success,test -e $(shell,$(CC) -print-file-name=plugin)/include/plugin-version.h)
> > > +     depends on $(host-cc-option,-include gmp.h)
> >
> > Why does the prior depends not fail? That's where plugin detection is
> > happening.
> >
> > --
> > Kees Cook
> 
> This patch set should be rejected.
> 
> 
> It was already discussed in the past.
> Just install a proper package, then gcc-plugin will work.
> 
> https://lore.kernel.org/all/CAHk-=wjjiYjCp61gdAMpDOsUBU-A2hFFKJoVx5VAC7yV4K6WYg@mail.gmail.com/

Thanks for the pointer.

As there is already consensus I agree that the patches should not be used.

Thomas