STI, HLT in acpi_cpu_idle_c1

[Gerrit Nagelhout]

John Baldwin wrote:
> As per an errata that Brian pointed out, try using this patch 
> perhaps.  You 
> can leave your isr_problem one in and hopefully the counter 
> will stay at zero 
> if it works.  It adds some dummy reads of the local APIC to 
> EOI's for IPIs.
> 

This patch doesn't seem to help any.  My isr problem detection code
still found a problem 10 times in about 17 hours of stress testing.
I know that the HACK I added to make the system work is probably
incorrect, but what kind of side effects would you expect to see from
just calling EOI, and possibly not handling the interrupt?
Thanks,

Gerrit