svn commit: r328110 - head/sys/kern

[Andriy Gapon]

On 18/01/2018 09:38, Wojciech Macek wrote:
> Author: wma
> Date: Thu Jan 18 07:38:54 2018
> New Revision: 328110
> URL: https://svnweb.freebsd.org/changeset/base/328110
> 
> Log:
>   KDB: restart only CPUs stopped by KDB
>   
>   There is a case when not all CPUs went online. In that situation,
>   restart only APs which were operational before entering KDB.

What is the context here?
I mean, what is the state of those CPUs that are not online?

Also, it seems you allow for the situation where a CPU that was not online at
the time of kdb_trap becomes online (and running) while kdb is active?
If that's so, then it can mess up the system big time.

I think that this is not a right solution.

P.S.
While not a recipe for a solution, these musing may be of interest to you:
https://lists.freebsd.org/pipermail/freebsd-arch/2011-June/011373.html

>   Created by:            Wojciech Macek <wma at semihalf.com>
>   Obtained from:         Semihalf
>   Reviewed by:           nwhitehorn
>   Differential revision: https://reviews.freebsd.org/D13949
>   Sponsored by:          QCM Technologies
> 
> Modified:
>   head/sys/kern/subr_kdb.c
> 
> Modified: head/sys/kern/subr_kdb.c
> ==============================================================================
> --- head/sys/kern/subr_kdb.c	Thu Jan 18 06:28:21 2018	(r328109)
> +++ head/sys/kern/subr_kdb.c	Thu Jan 18 07:38:54 2018	(r328110)
> @@ -673,6 +673,7 @@ kdb_trap(int type, int code, struct trapframe *tf)
>  #ifdef SMP
>  	if (!SCHEDULER_STOPPED()) {
>  		other_cpus = all_cpus;
> +		CPU_NAND(&other_cpus, &stopped_cpus);
>  		CPU_CLR(PCPU_GET(cpuid), &other_cpus);
>  		stop_cpus_hard(other_cpus);
>  		did_stop_cpus = 1;
> @@ -707,8 +708,9 @@ kdb_trap(int type, int code, struct trapframe *tf)
>  	kdb_active--;
>  
>  #ifdef SMP
> +	CPU_AND(&other_cpus, &stopped_cpus);
>  	if (did_stop_cpus)
> -		restart_cpus(stopped_cpus);
> +		restart_cpus(other_cpus);
>  #endif
>  
>  	intr_restore(intr);
> 


-- 
Andriy Gapon