Re: [poll / rfc] kdb_stop_cpus

2011/6/4 Andriy Gapon <avg_at_freebsd.org>:
> on 03/06/2011 20:57 Robert N. M. Watson said the following:
>>
>> On 3 Jun 2011, at 16:13, Andriy Gapon wrote:
>>
>>> I wonder if anybody uses kdb_stop_cpus with non-default value. If, yes, I
>>> am very interested to learn about your usecase for it.
>>
>> The issue that prompted the sysctl was non-NMI IPIs being used to enter the
>> debugger or reboot following a core hanging with interrupts disabled. With
>> the switch to NMI IPIs in some of those circumstances, life is better -- at
>> least, on hardware that supports non-maskable IPIs. I seem to recall sparc64
>> doesn't, however?
>
> Seems to be so as Nathan has also pointed out for PPC.
> For this I also plan the following change:
>
> commit 458ebd9aca7e91fc6e0825c727c7220ab9f61016
>
>    generic_stop_cpus: move timeout detection code from under DIAGNOSTIC
>
>    ... and also increase it a bit.
>    IMO it's better to detect and report the (rather serious) condition and
>    allow a system to proceed somehow rather than be stuck in an endless
>    loop.
>
> diff --git a/sys/kern/subr_smp.c b/sys/kern/subr_smp.c
> index ae52f4b..4bd766b 100644
> --- a/sys/kern/subr_smp.c
> +++ b/sys/kern/subr_smp.c
> _at__at_ -232,12 +232,10 _at__at_ generic_stop_cpus(cpumask_t map, u_int type)
>                /* spin */
>                cpu_spinwait();
>                i++;
> -#ifdef DIAGNOSTIC
> -               if (i == 100000) {
> +               if (i == 100000000) {
>                        printf("timeout stopping cpus\n");
>                        break;
>                }
> -#endif
>        }
>
>        stopping_cpu = NOCPU;

I'd also add the ability, once the deadlock is detected, to break in
KDB, and put that under DIAGNOSTIC.
I had such a patch and I used it to debug some deadlocks on shutdown
code, but now it seems I can't find it anymore.

Attilio


-- 
Peace can only be achieved by understanding - A. Einstein