From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S932685Ab1JCWDS (ORCPT <rfc822;w@1wt.eu>);
	Mon, 3 Oct 2011 18:03:18 -0400
Received: from mail-pz0-f42.google.com ([209.85.210.42]:43646 "EHLO
	mail-pz0-f42.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org
	with ESMTP id S1751736Ab1JCWDJ (ORCPT
	<rfc822;linux-kernel@vger.kernel.org>);
	Mon, 3 Oct 2011 18:03:09 -0400
Date: Mon, 3 Oct 2011 15:03:05 -0700
From: Andrew Morton <akpm00@gmail.com>
To: Jiri Kosina <jkosina@suse.cz>
Cc: Josh Boyer <jwboyer@redhat.com>, Nicolas Pitre <nicolas.pitre@linaro.org>,
        Andrew Morton <akpm00@gmail.com>, Ingo Molnar <mingo@elte.hu>,
        hongjiu.lu@intel.com, linux-kernel@vger.kernel.org,
        Russell King <rmk@arm.linux.org.uk>
Subject: Re: [PATCH v2] binfmt_elf: Fix PIE execution with randomization
 disabled (was Re: [RFC PATCH] binfmt_elf: Fix PIE execution with
 randomization disabled)
Message-Id: <20111003150305.936fc46e.akpm00@gmail.com>
In-Reply-To: <alpine.LNX.2.00.1110031709230.31654@pobox.suse.cz>
References: <20110929195359.GJ16720@zod.bos.redhat.com>
	<20110929141929.43df799d.akpm00@gmail.com>
	<alpine.LFD.2.02.1109291958210.9106@xanadu.home>
	<20110930021629.GL16720@zod.bos.redhat.com>
	<alpine.LNX.2.00.1110031651400.31654@pobox.suse.cz>
	<20111003150345.GP16720@zod.bos.redhat.com>
	<alpine.LNX.2.00.1110031709230.31654@pobox.suse.cz>
X-Mailer: Sylpheed 3.0.2 (GTK+ 2.20.1; x86_64-pc-linux-gnu)
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On Mon, 3 Oct 2011 17:11:47 +0200 (CEST)
Jiri Kosina <jkosina@suse.cz> wrote:

> 
> From: Jiri Kosina <jkosina@suse.cz>
> Subject: [PATCH] binfmt_elf: fix PIE execution with randomization disabled
> 
> The case of address space randomization being disabled in runtime through
> randomize_va_space sysctl is not treated properly in load_elf_binary(),
> resulting in SIGKILL coming at exec() time for certain PIE-linked binaries
> in case the randomization has been disabled at runtime prior to calling
> exec().
> 
> Handle the randomize_va_space == 0 case the same way as if we were not
> supporting .text randomization at all.
> 
> Based on original patch by H.J. Lu <hongjiu.lu@intel.com> and
> Josh Boyer <jwboyer@redhat.com>
> 
> Cc: Ingo Molnar <mingo@elte.hu>
> Cc: Jiri Kosina <jkosina@suse.cz>
> Cc: Nicolas Pitre <nicolas.pitre@linaro.org>
> Cc: Russell King <rmk@arm.linux.org.uk>
> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
> ---
>  fs/binfmt_elf.c |    5 ++++-
>  1 files changed, 4 insertions(+), 1 deletions(-)
> 
> diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
> index dd0fdfc..bb11fe4 100644
> --- a/fs/binfmt_elf.c
> +++ b/fs/binfmt_elf.c
> @@ -795,7 +795,10 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs)
>  			 * might try to exec.  This is because the brk will
>  			 * follow the loader, and is not movable.  */
>  #if defined(CONFIG_X86) || defined(CONFIG_ARM)
> -			load_bias = 0;
> +			if (current->flags & PF_RANDOMIZE)
> +				load_bias = 0;
> +			else
> +				load_bias = ELF_PAGESTART(ELF_ET_DYN_BASE - vaddr);
>  #else
>  			load_bias = ELF_PAGESTART(ELF_ET_DYN_BASE - vaddr);
>  #endif

Guys, it took several people several days and 10+ emails to work out
what's happening in there, and the first attempt to fix it was buggy. 
This is all a huuuuge signal that the code is unobvious, hard to
understand, hard to maintain.

Please, let's get a good code comment in there while it's fresh in your
minds.  So the next person who comes along doesn't have the same amount
of difficulty?