linux-kernel - Re: v4.10: kernel stack frame pointer .. has bad value (null)

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]

Date:   Wed, 22 Feb 2017 15:15:21 -0600
From:   Josh Poimboeuf <jpoimboe@...hat.com>
To:     "H. Peter Anvin" <hpa@...or.com>
Cc:     Pavel Machek <pavel@....cz>,
        kernel list <linux-kernel@...r.kernel.org>, mingo@...nel.org,
        luto@...nel.org, bp@...en8.de, brgerst@...il.com,
        dvlasenk@...hat.com, torvalds@...ux-foundation.org,
        peterz@...radead.org, tglx@...utronix.de
Subject: Re: v4.10: kernel stack frame pointer .. has bad value (null)

On Wed, Feb 22, 2017 at 12:51:11PM -0800, H. Peter Anvin wrote:
> On 02/22/17 08:45, Josh Poimboeuf wrote:
> >>
> >> FWIW, it would be really darned nice to not have all those zeroes in a
> >> 32-bit stack frame dump.
> > 
> > Yeah, I'll fix that.
> > 
> >> Is not a zero stack frame pointer value an end of stack token?
> > 
> > There's no end of stack "token" per se, though any frame pointer value
> > outside the bounds of the stack will terminate the stack trace (and that
> > still happened here).
> > 
> 
> Well, my understanding is that at least gdb and perhaps other unwinders
> consider a zero stack frame pointer to be an indicator that the stack
> has reached its end.  That's why I'm wondering if this is possible in
> this case or if it is unlikely because of the value.

I'm not sure I follow your question.  The frame pointer was zero, and
that did cause the unwinder to stop the stack trace.  The warning was
because it ended in an unexpected place.

> > The warning is because the stack trace didn't make it all the way to the
> > "end" location of the stack (right before the syscall pt_regs location).
> > The warning is part of the effort to ensure reliable stacks.
> 
> It would be useful to get an understanding why...

Agreed...

-- 
Josh