Re: 3.5.0-rc3 (approximately) oops on cifs mount

[Jeff Layton <jlayton@xxxxxxxxxx>]

On Tue, 3 Jul 2012 11:59:02 -0700
Andy Lutomirski <luto@xxxxxxxxxxxxxx> wrote:

> On Mon, Jul 2, 2012 at 5:28 PM, Andy Lutomirski <luto@xxxxxxxxxxxxxx> wrote:
> > On Mon, Jul 2, 2012 at 5:21 PM, Steve French <smfrench@xxxxxxxxx> wrote:
> >> Any recreation information?
> >
> > I think it was a result of trying to mount with an incorrect password,
> > which surprises me a bit.  I'll try to reproduce in a little bit.
> 
> Correction: it was not a wrong password -- I mounted with the correct
> password.  So far, I can't reproduce it.
> 
> --Andy
> 

It crashed here:

    BUG_ON(cpu >= nr_cpu_ids && cpu != WORK_CPU_UNBOUND);

Very odd. Seems almost like this must be some sort of memory
corruption. The work is initialized just a little bit before it gets
queued. The only thing I can figure is that something raced in and
clobbered the work struct beforehand.

One thing you could do -- maybe disassemble get_work_gcwq() and post
the results here? Maybe from that we can see what the "cpu" value
actually was by figuring out the register that holds it?

-- 
Jeff Layton <jlayton@xxxxxxxxxx>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/