Re: RT and Cascade interrupts

[john cooper <john.cooper@timesys.com>]

Ingo Molnar wrote:
> * Daniel Walker <dwalker@mvista.com> wrote:
> 
> 
>>	It seems like the cascade interrupts could run in threads, but 
>>i386 doesn't, and I know ARM crashed with cascades in threads. You may 
>>have a bit of a slow down, but it seems possible. Does anyone have 
>>some reasoning for why we aren't running the cascades in threads?
> 
> 
> are the x86 cascade interrupts real ones in fact? Normally they should 
> never trigger directly. (except on ARM which has a completely different 
> notion of cascade irq)

I just caught this thread in the corner of my eye.

I'm seeing the BUG assert in kernel/timers.c:cascade()
kick in (tmp->base is somehow 0) during a test which
creates a few tasks of priority higher than ksoftirqd.
This race doesn't happen if ksoftirqd's priority is
elevated (eg: chrt -f -p 75 2) so the -RT patch might
be opening up a window here.

I'm in the process of investigating this and see a few
potential suspects but was wondering if anyone else has
seen this behavior?

-john

-- 
john.cooper@timesys.com