[PATCH] x86, 32-bit: Fix invalid stack address while in softirq

[Robert Richter]

(Resending patch with [PATCH] in subject line and updated cc list.)

On 06.09.12 09:30:37, wyang1 wrote:
> Robert,
> 
> I agreed what you said, my patch more likes a workaround.
> 
> > So the proper fix I see is to fix kernel_stack_pointer() to return a
> > valid stack in case of an empty stack while in softirq. Something like
> > the patch below. Maybe it must be optimized a bit. I tested the patch
> > over night with no issues found. Please test it too.
> 
> I also tested the following patch over night. it is fine.:-)

Wei,

thanks for testing.

Ingo,

please take a look at this. Not sure if Linus want to look at this too
and if we need more optimization here.

Thanks,

-Robert