Re: 3.1+ iwlwifi lockup

["Guy, Wey-Yi" <wey-yi.w.guy@intel.com>]

On Tue, 2011-11-22 at 19:34 -0800, Dave Jones wrote:
> On Tue, Nov 22, 2011 at 06:29:40PM -0800, Guy, Wey-Yi wrote:
>  
>  > On Tue, 2011-11-22 at 18:45 -0800, Dave Jones wrote:
>  > > On Tue, Nov 22, 2011 at 12:15:02PM -0800, wwguy wrote:
>  > >  > Hi Dave,
>  > >  > 
>  > >  > Sorry for taking so long.
>  > >  > it is a bogus command from the firmware which driver try to process it
>  > >  > and cause problem. I am not sure how it happen. But could you please try
>  > >  > the patch I attach here.
>  > >  > 
>  > >  > For now I am just ignore the command, but not sure it is good enough,
>  > >  > might has to reload the firmware, but not sure.
>  > > 
>  > > With Linus' current tree and this patch, I got a panic pretty quickly,
>  > > but with a different trace.
>  > > 
>  > Are you seeing the problem without this patch? I guess the patch I done
>  > here is not the right thing to do and I try to address the problem too
>  > aggressively
> 
> I've been running Linus' current tree for the whole time and haven't seen
> any iwlwifi traces since my initial report on the 11th.
> 
> I can't reproduce the bug, even by doing the same operation
> (an scp of a dozen photographs).
> 
> It seems to happen almost at random, as if it's somehow timing related.
> I'm running the patched kernel again now, and it's been fine for the last hour.
> 
Agree, that is why I could not re-produce here after many try. Could you
share your configuration, such as band/channel, 32/64 bit system, 20/40
MHz, traffic type, open/security, power management, ..., 

Thanks
Wey
>