linux-kernel - Re: [PATCH v2] export linux/a.out.h

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

Message-Id: <1213629335.26255.792.camel@pmac.infradead.org>
Date:	Mon, 16 Jun 2008 16:15:35 +0100
From:	David Woodhouse <dwmw2@...radead.org>
To:	Andi Kleen <andi@...stfloor.org>
Cc:	Peter Korsgaard <jacmet@...site.dk>, linux-kernel@...r.kernel.org,
	sam@...nborg.org, dhowells@...hat.com,
	"Kirill A. Shutemov" <k.shutemov@...il.com>, rth@...ddle.net
Subject: Re: [PATCH v2] export linux/a.out.h

On Mon, 2008-06-16 at 16:40 +0200, Andi Kleen wrote:
> >  We should get rid of CONFIG_ARCH_SUPPORTS_AOUT entirely. The only
> > remaining users of it are fs/exec.c, which uses it entirely redundantly:
> >	#if defined(__alpha__) && defined(CONFIG_ARCH_SUPPORTS_AOUT)
> >
> > ...and (indirectly, by virtue of being the only other file that includes
> > <linux/a.out.h> when ARCH_SUPPORTS_AOUT might be unset) fs/binfmt_elf.c,
> > which still has unnecessary references to 'struct exec' even though Andi
> > removed the support for a.out interpreters a few months ago. That can
> > just go away.
> 
> I didn't think the Alpha /sbin/loader stuff was an a.out interpreter for ELF.
>
> Also pure a.out is still supported.
> 
> It might be ok to remove it anyways, but you should check with someone
> who knows Alpha first. cc Richard.

The Alpha /sbin/loader support, and the a.out-interpreter-for-ELF, are
two completely separate issues -- you're right that /sbin/loader isn't
an a.out interpreter for ELF files.

They're the two places which rely on '#ifdef CONFIG_ARCH_SUPPORTS_AOUT'
in <linux/a.out.h>, because they include <linux/a.out.h> on platforms
which might not have <asm/a.out.h>.

In the Alpha case in fs/exec.c, we can simply include <linux/a.out.h>
only #ifdef __alpha__ -- while noting that that code could probably do
with a little more cleanup and turning into a properly registered binfmt
rather than a special case, if possible.

In the case of fs/binfmt_elf.c, we can simply remove the inclusion of
<linux/a.out.h> and all references to 'struct exec'. You removed the
rest of the a.out interpreter support already; that last part probably
should have gone away at the same time.

Proper a.out support in fs/binfmt_aout.c can stay as it is; you won't be
building binfmt_aout.c unless your platform supports a.out anyway.

And then we can remove the #ifdef CONFIG_ARCH_SUPPORTS_AOUT in
<linux/a.out.h>, and export it to userspace again, which was the point
of Peter's patch that started this thread.

-- 
dwmw2

--
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/