Re: [PATCH] 2.6.17-rt1 : fix x86_64 oops

[Dipankar Sarma <dipankar@in.ibm.com>]

On Mon, Jul 03, 2006 at 10:27:50PM +0530, Dipankar Sarma wrote:
> On Thu, Jun 29, 2006 at 10:11:45PM +0200, Ingo Molnar wrote:
> > 
> > * Paul E. McKenney <paulmck@us.ibm.com> wrote:
> > > OK, I ran this with both torture types (rcu and rcu_bh) on i386 with 
> > > CONFIG_PREEMPT=y on 2.6.17-mm4 and didn't see any "scheduling while 
> > > atomic" oopses -- or any other oopses, for that matter.
> > > 
> > > Here is the .config file I used.  What am I missing here?
> > 
> > hm, i'm seeing some other types of crashes too - so rcutorture could 
> > just have been collateral damage. It was on i386, an allyesconfig 
> > bzImage kernel.
> 
> With 2.6.17-rt5 I see this -
> 
> Starting pass 0
> Unable to handle kernel paging request at ffffffff88006bd0 RIP:
> <ffffffff802597d5>{rcu_process_callbacks+107}
> rcutorture: --- End of test: SUCCESS: nreaders=8 stat_interval=1PGD 203027 PUD 205027 PMD 21eb18067 PTE 21829f163
> Oops: 0000 [1] PREEMPT SMP
> CPU 1
> 
> I have been able to reproduce a similar looking oopse with 2.6.16-rt29.
> 2.6.16-rt20 works fine. I will try to track it down to the exact
> release as far as I can.

OK, it looks as if rt20 is fine but rt21 is broken. So something
that got in rt21 is causing this oops.

Ingo, do you have a suspect ?

Thanks
Dipankar