Re: two lockups with 2.4.25

[Oleg Drokin <green@linuxhacker.ru>]

Hello!

On Mon, May 03, 2004 at 09:51:08AM -0300, Marcelo Tosatti wrote:
> > > Chris Stromsoe <cbs@cts.ucla.edu> wrote:
> > > CS> Pid: 26885, comm:               sophie
> > > >>>EIP; c0110ed7 <flush_tlb_others+9b/bc>   <=====
> > > >>>EDX; 01000000 Before first symbol
> > > >>>ESI; f4762cc0 <_end+343c85cc/3896e90c>
> > > >>>EDI; 081a1d28 Before first symbol
> > > >>>EBP; e806fe94 <_end+27cd57a0/3896e90c>
> > > CS> Trace; c011100f <flush_tlb_page+6f/7c>
> > > CS> Trace; c01259b7 <do_wp_page+223/284>
> > > CS> Trace; c01260de <handle_mm_fault+82/b8>
> > > CS> Trace; c01132f9 <do_page_fault+1a1/4ed>
> > > CS> Trace; c0113158 <do_page_fault+0/4ed>
> > > CS> Trace; c0224ce1 <__kfree_skb+129/134>
> > > CS> Trace; c01145e3 <schedule+45b/520>
> > > CS> Trace; c0106fc4 <error_code+34/3c>
> > > This backtrace is suspiciously similar to a backtrace from NMI WD I had
> > > not so long ago. Also on 2.5.25 Are your boxes SMP?
> AFAICS this backtrace is fine -- flush_tlb_others() does invplg and thats it. No
> locking involved, yes?

Well, the trace looks fine, except the box was hung at the time of the
trace capturing. And most of the trace is the same as I had reported with
NMI watchdog. This is the only thing I was going to say.
So far the box I experienced NMI WD oops on works and does not hang.
And I posted previous oops on lkml.
If something will happen to it again, I won't keep it secret for sure ;)

Bye,
    Oleg