linux-kernel - Re: [PATCH 3/4] ubsan: Force -Wno-maybe-uninitialized only for GCC

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]

Date:   Mon, 5 Oct 2020 23:03:46 -0700
From:   Kees Cook <keescook@...omium.org>
To:     Nathan Chancellor <natechancellor@...il.com>
Cc:     Andrew Morton <akpm@...ux-foundation.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Ard Biesheuvel <ardb@...nel.org>,
        Arnd Bergmann <arnd@...db.de>,
        Masahiro Yamada <masahiroy@...nel.org>,
        Michal Marek <michal.lkml@...kovi.net>,
        Nick Desaulniers <ndesaulniers@...gle.com>,
        Marco Elver <elver@...gle.com>,
        Randy Dunlap <rdunlap@...radead.org>,
        Dmitry Vyukov <dvyukov@...gle.com>,
        George Popescu <georgepope@...roid.com>,
        Herbert Xu <herbert@...dor.apana.org.au>,
        Peter Oberparleiter <oberpar@...ux.ibm.com>,
        Andrey Ryabinin <aryabinin@...tuozzo.com>,
        clang-built-linux@...glegroups.com, linux-kbuild@...r.kernel.org,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH 3/4] ubsan: Force -Wno-maybe-uninitialized only for GCC

On Sun, Oct 04, 2020 at 12:16:14AM -0700, Nathan Chancellor wrote:
> On Fri, Oct 02, 2020 at 03:15:26PM -0700, Kees Cook wrote:
> > Clang handles 'maybe-uninitialized' better in the face of using UBSAN,
> > so do not make this universally disabled for UBSAN builds.
> > 
> > Signed-off-by: Kees Cook <keescook@...omium.org>
> 
> Well this patch is not strictly necessary because Clang does not support
> -Wmaybe-uninitialized anyways :) its flags are -Wuninitialized and
> -Wsometimes-uninitialized so the warning stays enabled for UBSAN as it
> stands.

Ah, yes. Heh. Well... perhaps I can just drop this patch.

> However, something like this could still worthwhile because it would
> save us one call to cc-disable-warning (yay micro optimizations).
> 
> Maybe it just does not need to have a whole new symbol, just make it
> 
> ubsan-cflags-$(CONFIG_CC_IS_GCC)
> 
> instead of
> 
> ubsan-cflags-$(CONFIG_UBSAN)

If it gets kept, I'd still like it gated on CONFIG_UBSAN in some way
(e.g. the patch has an implicit depends due to the "if UBSAN" section).

But yes, this patch is rather a no-op.

-- 
Kees Cook