linux-kernel - Re: [PATCH] [RFC] ARM: vdso: work around gcc-8 warning

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:   Thu, 24 Aug 2017 17:06:33 +0100
From:   Russell King - ARM Linux <linux@...linux.org.uk>
To:     Arnd Bergmann <arnd@...db.de>
Cc:     Ard Biesheuvel <ard.biesheuvel@...aro.org>,
        Nicolas Pitre <nico@...xnic.net>,
        Nathan Lynch <nathan_lynch@...tor.com>,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] [RFC] ARM: vdso: work around gcc-8 warning

On Wed, Aug 23, 2017 at 04:05:26PM +0200, Arnd Bergmann wrote:
> gcc-8 correctly points out that reading four bytes from a pointer to a
> 'char' variable is wrong
> 
> arch/arm/kernel/vdso.c: In function 'vdso_init':
> arch/arm/kernel/vdso.c:200:6: error: '__builtin_memcmp_eq' reading 4 bytes from a region of size 1 [-Werror=stringop-overflow=]
> 
> However, in this case the variable just stands for the beginning of the
> vdso and is not actually a 'char', so the code is doing what it is meant
> to do.
> 
> Not sure what the best solution for this is, changing the hack to
> declare the variable as 'int' instead makes the warning go away.

Well, one way to look at the code is an array of characters, so
we could say that it should be:

extern char vdso_start[], vdso_end[];

which is exactly the same construct that we use for _stext and
similar.

-- 
RMK's Patch system: http://www.armlinux.org.uk/developer/patches/
FTTC broadband for 0.8mile line in suburbia: sync at 8.8Mbps down 630kbps up
According to speedtest.net: 8.21Mbps down 510kbps up