Re: 3.2-rc2+: Reported regressions from 3.0 and 3.1

Linus Torvalds <[email protected]> Mon, 21 Nov 2011 14:11:54 -0800

Newsgroups	org.kernel.vger.kernel-testers,org.freedesktop.lists.dri-devel,org.kernel.vger.linux-acpi,org.kernel.vger.linux-kernel,org.kernel.vger.linux-scsi,org.kernel.vger.linux-wireless,org.kernel.vger.netdev
Message-ID	<CA+55aFwAELQieN=qR7Gz_MT5rHPp90H10uTMS+DOJjefk7q3ng@mail.gmail.com>

On Mon, Nov 21, 2011 at 1:49 PM, Rafael J. Wysocki <[email protected]> wrote:
>
> Subject    : [3.1 REGRESSION] Commit 5cec93c216db77c45f7ce970d46283bcb1933884 breaks the Chromium seccomp sandbox
> Submitter  : Nix <[email protected]>
> Date       : 2011-11-14 0:40
> Message-ID : [email protected]
> References : http://marc.info/?l=linux-kernel&m=132123396226377&w=2

So this should be fixed by commit 2b666859ec32 ("x86: Default to
vsyscall=native for now"), since we disabled the vsyscall emulation
because it broken UML too.

Of course, the chromium seccomp thing might re-surface with the
patches that enable the emulation (with better emulation), which Andy
is still working on, and that I was planning on merging for 3.3.

Andy, it migth be worth contacting Nix and having him test whether
your fixed emulation works for chromium too.

                      Linus