linux-kernel - Re: [PATCH 2/3] objtool: Ignore retpoline alternatives

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: <20180111172303.GI6176@hirez.programming.kicks-ass.net>
Date:   Thu, 11 Jan 2018 18:23:03 +0100
From:   Peter Zijlstra <peterz@...radead.org>
To:     David Woodhouse <dwmw2@...radead.org>
Cc:     Jiri Kosina <jikos@...nel.org>,
        Josh Poimboeuf <jpoimboe@...hat.com>, x86@...nel.org,
        linux-kernel@...r.kernel.org, Thomas Gleixner <tglx@...utronix.de>,
        Andi Kleen <ak@...ux.intel.com>, Paul Turner <pjt@...gle.com>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Greg Kroah-Hartman <gregkh@...ux-foundation.org>,
        Tim Chen <tim.c.chen@...ux.intel.com>,
        Dave Hansen <dave.hansen@...el.com>,
        Kees Cook <keescook@...gle.com>,
        Andy Lutomirski <luto@...capital.net>,
        gnomes@...rguk.ukuu.org.uk
Subject: Re: [PATCH 2/3] objtool: Ignore retpoline alternatives

On Thu, Jan 11, 2018 at 05:19:23PM +0000, David Woodhouse wrote:
> On Thu, 2018-01-11 at 18:05 +0100, Peter Zijlstra wrote:
> > On Thu, Jan 11, 2018 at 06:01:23PM +0100, Jiri Kosina wrote:
> > > On Thu, 11 Jan 2018, Josh Poimboeuf wrote:
> > > 
> > > > I think I heard that retpolines won't be ported to anything older than 
> > > > GCC 4.9, so maybe it's safe to use '%='.  I don't remember when it was 
> > > > introduced into GCC though.
> > > 
> > > I'm afraid we'll have to backport retpolines in some form to 4.3.x at 
> > > least. I'd be surprised if we'd be the only ones on this planet :)
> > 
> > So upstream code is going to require 4.5 at least, and 4.4 has %=.
> > Backport effort will just have to cope or backport more GCC bits, that
> > is, if you're backporting retpoline to 4.3 also backport asm-goto.
> 
> Again, the RSB thing is for more than just retpoline; it's needed for
> IBRS support too and that *doesn't* necessarily require a newer
> compiler at all.

I think the asm-goto was exactly for IBRS, because without that you get
to sprinkle LFENCE all over the place.