linux-kernel - Re: [PATCH] x86,nmi: Fix section mismatch warnings on 32-bit

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, 06 Jun 2012 12:14:33 +0100
From:	"Jan Beulich" <JBeulich@...e.com>
To:	"Don Zickus" <dzickus@...hat.com>
Cc:	<x86@...nel.org>, <zhong@...ux.vnet.ibm.com>,
	"LKML" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] x86,nmi: Fix section mismatch warnings on 32-bit

>>> On 05.06.12 at 20:35, Don Zickus <dzickus@...hat.com> wrote:
> On Tue, Jun 05, 2012 at 04:42:20PM +0100, Jan Beulich wrote:
>> >>> On 05.06.12 at 16:47, Don Zickus <dzickus@...hat.com> wrote:
>> > On Tue, Jun 05, 2012 at 09:12:13AM +0100, Jan Beulich wrote:
>> >> So before applying a change like this, I'd really like to understand
>> >> the issue itself. Could you simply attach the nmi_selftest.o file
>> >> that you get without your patch here (and indicate which
>> >> upstream version or revision the corresponding source file was
>> >> from, to be able to re-construct what the compiler actually did)?
>> > 
>> > I could but it might be easier for you to re-create it on your end.
>> > Using gcc-4.4.4 on an x86_64 box and v3.5-rc1, I just set
>> > CONFIG_DEBUG_NMI_SELFTEST to 'y' and compile with
>> > 
>> > make ARCH=i386 O=out/i686
>> > 
>> > and it easily reproduces for me.
>> 
>> I don't think I have a plain 4.4.4 around anywhere, nor is it clear
>> whether you used a plain one (and how it was configured). For
>> tricky, highly version dependent things like this (where a compiler
>> bug also can't be excluded), having the _exact_ object file the
>> problem was seen with would be much easier than trying to hunt
>> down a compiler version that I can reproduce this with.
> 
> I didn't think it would be compiler dependent as I do not know what
> compiler the reporter was using.  I used a RHEL-6 4.4.4 compiler (which
> you probably don't have :^) ).

Indeed, somehow I failed to see the obvious - it's commit
72b3fb24713755cf9740b403e95aa67ceedf3509 that causes
these problems. Instantiating static data like this just doesn't
play with any of the pointers passed being into .init.*.

I'd suggest either open coding register_nmi_handler() (with
the static data put into __initdata), or further abstracting it
by allowing an optional fifth argument (specifying the section
annotation if needed).

Jan

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/