Re: [Xen-ia64-devel] [Patch] Stop all cpus at panic time

[Akio Takebe <takebe_akio@xxxxxxxxxxxxxx>]

Hi, Alex

>On Tue, 2007-12-04 at 23:48 +0900, Akio Takebe wrote:
>> Hi,
>> 
>> Current panic() of hypervisor doesn't stop all cpus.
>> So domains can work after hypervisor panic.
>> (but this issue happens at only using noreboot option)
>> If dom0 works after Hypervisor panic,
>> the system may get a serious problem.
>
>   This looks like a good 3.2 candidate.  I think there's still a minor
>issue though.  Our implementation of stop_this_cpu() removes the CPU
>from the online map, disables interrupts, then returns(!).  We should
>probably figure out if we still need the call to cpu_halt() ifdef'd out
>or at least put a busy loop in it's place.  Thanks,
Thank you for your review.
It's a good point.

I'm trying those choices.
smp_send_stop is also used by kexec.
I think it is important that INIT handler and kdump work fine at panic.
I tried INIT handler at panic, but INIT handler didn't work.
It seems to be a little broken.
I'll debug it soon.

Best Regards,

Akio Takebe


_______________________________________________
Xen-ia64-devel mailing list
Xen-ia64-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-ia64-devel