linux-kernel - Re: [PATCH] x86/KASLR: use the right memcpy

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

Date:   Wed, 31 May 2017 09:43:15 +0800
From:   Baoquan He <bhe@...hat.com>
To:     Arnd Bergmann <arnd@...db.de>
Cc:     Ingo Molnar <mingo@...hat.com>,
        the arch/x86 maintainers <x86@...nel.org>,
        "H. Peter Anvin" <hpa@...or.com>,
        Thomas Gleixner <tglx@...utronix.de>,
        Kees Cook <keescook@...omium.org>,
        Thomas Garnier <thgarnie@...gle.com>,
        Dave Jiang <dave.jiang@...el.com>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] x86/KASLR: use the right memcpy

On 05/30/17 at 04:24pm, Arnd Bergmann wrote:
> On Tue, May 30, 2017 at 3:56 PM, Baoquan He <bhe@...hat.com> wrote:
> > On 05/30/17 at 11:14am, Arnd Bergmann wrote:
> >> The decompressor has its own implementation of the string functions,
> >> but has to include the right header to get those, while implicitly
> >> including linux/string.h may result in a link error:
> >>
> >> arch/x86/boot/compressed/kaslr.o: In function `choose_random_location':
> >> kaslr.c:(.text+0xf51): undefined reference to `_mmx_memcpy'
> >>
> >> This has appeared now as kaslr started using memcpy. Other files in the
> >> decompressor already do the same thing.
> >>
> >> Fixes: d52e7d5a952c ("x86/KASLR: Parse all 'memmap=' boot option entries")
> >> Signed-off-by: Arnd Bergmann <arnd@...db.de>
> >
> > Thanks for this fix, Arnd!
> >
> > The linking error didn't happen when I tested the patch of d52e7d5a952c.
> > Could you tell in what condition it will be triggered? Not sure if I
> > should wait for this fix being merged and do a back porting, or can
> > defer it if it's not risky.
> 
> It only happens on 32-bit kernels with CONFIG_X86_USE_3DNOW, maybe
> there are additional requirements.

Checked code again, in commit d52e7d5a952c "#include "../boot.h" is
removed. Not sure if that removal caused the 32-bit kernel link error.
While I didn't see boot/string.h is included into the boot/boot.h.