Date: Fri, 05 Dec 2014 19:59:35 -0500 From: Daniel Micay <danielmicay@...il.com> To: oss-security@...ts.openwall.com Subject: Re: Offset2lib: bypassing full ASLR on 64bit Linux On 05/12/14 05:15 PM, Reed Loden wrote: > On Fri, Dec 5, 2014 at 7:09 AM, Daniel Micay <danielmicay@...il.com> wrote: > >> >> Mozilla has no excuse for not enabling PIE for Firefox, because 99% of >> the code is in dynamic libraries already. It has no performance impact. >> > > For the record, Mozilla tried it several months ago and had to back it out. > > "Nautilus (the file manager) can't open PIE executables, which makes > distributing PIE executable essentially impossible." > > https://bugzilla.mozilla.org/show_bug.cgi?id=857628#c6 (which caused > https://bugzilla.mozilla.org/show_bug.cgi?id=1076892) > > ~reed I don't really see how this would prevent Mozilla from shipping a browser with ASLR. The Tor browser has been shipping a fork of Firefox built as a position independent executable for ages. It doesn't impact users because they're either starting it via a .desktop file or the command-line. The support for desktop icons in Nautilus is deprecated / disabled by default with only a hidden dconf preference to enable it. If you really want to support the workflow of opening up the file manager, navigating to the binary and double-clicking it then using a wrapper script is a quite obvious solution. The issue was already reported earlier by Mozilla, and the claim that it's a blocking issue didn't make sense then either: https://bugzilla.gnome.org/show_bug.cgi?id=737849 Chromium has used features like PIE, SSP and full RELRO for years while Firefox doesn't enable any of it. I don't see it as much different from how Chromium ships with an industry leading sandbox + features like JIT hardening while Firefox doesn't have any of it. Even Internet Explorer and Safari are shipping with decent sandboxing. There are multiple cases of remote code execution discovered in every 6 week cycle and no industry standard exploit mitigations in place. It's too bad that this doesn't least lead to any civil / criminal liability due to negligence, especially when it's advertised as being more secure / private than competitors. I find it hard to believe that there's any attention to security when even the tiny amount of effort involved in enabling year / decade old exploit mitigations is too much. Every project shipping a network-facing or setuid binary without PIE has some explaining to do. Download attachment "signature.asc" of type "application/pgp-signature" (820 bytes)
Powered by blists - more mailing lists
Please check out the Open Source Software Security Wiki, which is counterpart to this mailing list.
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.