Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1946389AbWJTLyR (ORCPT ); Fri, 20 Oct 2006 07:54:17 -0400 Received: (majordomo@vger.kernel.org) by vger.kernel.org id S1946390AbWJTLyQ (ORCPT ); Fri, 20 Oct 2006 07:54:16 -0400 Received: from mx1.suse.de ([195.135.220.2]:9197 "EHLO mx1.suse.de") by vger.kernel.org with ESMTP id S1946389AbWJTLyQ (ORCPT ); Fri, 20 Oct 2006 07:54:16 -0400 Date: Fri, 20 Oct 2006 13:54:14 +0200 From: Marcus Meissner To: Dave Jones , linux-kernel@vger.kernel.org, akpm@osdl.org Subject: Re: [PATCH] binfmt_elf: randomize PIE binaries Message-ID: <20061020115414.GA14448@suse.de> References: <20061019143547.GA8586@suse.de> <20061019202354.GI26530@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <20061019202354.GI26530@redhat.com> User-Agent: Mutt/1.5.9i Sender: linux-kernel-owner@vger.kernel.org X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1767 Lines: 43 On Thu, Oct 19, 2006 at 04:23:54PM -0400, Dave Jones wrote: > On Thu, Oct 19, 2006 at 04:35:47PM +0200, Marcus Meissner wrote: > > > --- linux-2.6.18/fs/binfmt_elf.c.xx 2006-10-19 11:21:49.000000000 +0200 > > +++ linux-2.6.18/fs/binfmt_elf.c 2006-10-19 11:24:58.000000000 +0200 > > @@ -856,7 +856,12 @@ static int load_elf_binary(struct linux_ > > * default mmap base, as well as whatever program they > > * might try to exec. This is because the brk will > > * follow the loader, and is not movable. */ > > - load_bias = ELF_PAGESTART(ELF_ET_DYN_BASE - vaddr); > > + if (current->flags & PF_RANDOMIZE) > > + load_bias = randomize_range(0, ELF_ET_DYN_BASE, > > + vaddr); > > + else > > + load_bias = ELF_ET_DYN_BASE - vaddr; > > + load_bias = ELF_PAGESTART(load_bias); > > } > > > > error = elf_map(bprm->file, load_bias + vaddr, elf_ppnt, > > > vaddr seems odd to be using as the len for randomize_range doesn't it ? > Should that be randomize_range(0, ELF_ET_DYN_BASE, ELF_ET_DYN_BASE - vaddr); > > maybe ? > Also, when we get to the elf_map(), we add vaddr back again, but in the > randomize_range case, we never subtracted it. hmm? Yes, the code was not really doing what I actually intended. (It was working, but the range always started at vaddr). I am using randomize_range (PAGE_SIZE, ELF_ET_DYN_BASE,0) (- vaddr) now after further comments from Arjan. Any other good ideas for a range to do the randomization in are welcome. Ciao, Marcus - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/