Re: [PATCH 0/2] [GIT PULL] tracing: various bug fixes

[Ingo Molnar <mingo@elte.hu>]

* Steven Rostedt <rostedt@goodmis.org> wrote:

> 
> On Wed, 22 Apr 2009, Steven Rostedt wrote:
> 
> > 
> > On Wed, 22 Apr 2009, Jeremy Fitzhardinge wrote:
> > 
> > > > 
> > > >    
> > > > > Do you have CONFIG_PARAVIRT on?  When I disabled it, I have yet to
> > > > > reproduce the bug. But I've only rebooted a few times. I'm going to
> > > > > continue to reboot to see if I can trigger it.
> > > > >     
> > > > 
> > > > 
> > > > Yes it is enabled.
> > > > 
> > > > 
> > > > 
> > > >   
> > > > > I'm thinking that the paravirt alternative code may have clobbered a
> > > > > register in either the enable or disabling of interrupts. This might cause
> > > > > a strange value to go into the hardirqs_enabled flag.
> > > > >     
> > > > 
> > > > 
> > > > 
> > > > Ok I will try it without PARAVIRT and tell you if I can reproduce it.
> > > >   
> > > 
> > > Interesting.  What code is generated for native_irq_enable/disable?
> > 
> > Note, when I take my standard config and enable PARAVIRT and LOCKDEP it 
> > still boots up fine. It may be a combination of some configs :-/
> > 
> > I'm currently doing a "config bisect" to find out what configs are needed.
> 
> I just finished the config bisect, and it came down to 
> CONFIG_DEBUG_PAGEALLOC (set is bad, not set is good). I'm sure 
> this is not the only config that is at issue. It is probably a 
> combination of configs
> :-/
> 
> I'll try others to see if things change.

It could also be the timing of an incoming IRQ to hit a particular 
race window. PAGEALLOC might trigger the right kind of delays to 
make it trigger.

	Ingo