linux-kernel - Re: [RFC PATCH 4/4] bpf: inhibit speculated out-of-bounds pointers

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: <20180104105958.vls2zxbxscqr46bm@salmiak>
Date:   Thu, 4 Jan 2018 10:59:58 +0000
From:   Mark Rutland <mark.rutland@....com>
To:     Peter Zijlstra <peterz@...radead.org>
Cc:     linux-kernel@...r.kernel.org, linux-arch@...r.kernel.org,
        Will Deacon <will.deacon@....com>
Subject: Re: [RFC PATCH 4/4] bpf: inhibit speculated out-of-bounds pointers

On Thu, Jan 04, 2018 at 12:45:29AM +0100, Peter Zijlstra wrote:
> On Wed, Jan 03, 2018 at 10:38:27PM +0000, Mark Rutland wrote:
> > Under speculation, CPUs may mis-predict branches in bounds checks. Thus,
> > memory accesses under a bounds check may be speculated even if the
> > bounds check fails, providing a primitive for building a side channel.
> > 
> > The EBPF map code has a number of such bounds-checks accesses in
> > map_lookup_elem implementations. This patch modifies these to use the
> > nospec helpers to inhibit such side channels.
> > 
> > The JITted lookup_elem implementations remain potentially vulnerable,
> > and are disabled (with JITted code falling back to the C
> > implementations).
> 
> Since this is now public, let me re-iterate that I don't particularly
> like this approach. If you have to kill the JIT, could we please keep
> that in the arch JIT implementation?

Hopefully, killing the JIT is a temporary bodge. I agree that we want the arch
backends to JIT safe sequences somehow; I just haven't figured out exactly what
we need to do to make that happen.

Thanks,
Mark.