Re: [PATCH 1/4] cputime: Generalize CONFIG_VIRT_CPU_ACCOUNTING

[Martin Schwidefsky <schwidefsky@de.ibm.com>]

On Thu, 16 Aug 2012 14:55:59 +0200
Frederic Weisbecker <fweisbec@gmail.com> wrote:

> On Thu, Aug 16, 2012 at 07:38:17PM +1000, Benjamin Herrenschmidt wrote:
> > On Thu, 2012-08-16 at 09:53 +0200, Martin Schwidefsky wrote:
> > > Hmm, ok. But then the description should be reworded not to be specific to
> > > the power architecture (the part of the message about "This also enables
> > > accounting of stolen time on logically-partitioned systems running on IBM
> > > POWER5-based machines."). 
> > 
> > Which is not very helpful to somebody running on a POWER6 or 7 (which
> > also support that option just fine :-)
> > 
> > So yes, the description should definitely be improved.
> 
> All right. How about something like the below?
> 
> diff --git a/init/Kconfig b/init/Kconfig
> index 894b073..5f5f8c2 100644
> --- a/init/Kconfig
> +++ b/init/Kconfig
> @@ -276,9 +276,9 @@ config VIRT_CPU_ACCOUNTING
>  	  accounting.  This is done by reading a CPU counter on each
>  	  kernel entry and exit and on transitions within the kernel
>  	  between system, softirq and hardirq state, so there is a
> -	  small performance impact.  This also enables accounting of
> -	  stolen time on logically-partitioned systems running on
> -	  IBM POWER5-based machines.
> +	  small performance impact.  In the case of IBM POWER > 5, this
> +	  also enables accounting of stolen time on logically-partitioned
> +	  systems.
> 
>  config BSD_PROCESS_ACCT
>  	bool "BSD Process Accounting"
> 

VIRT_CPU_ACCOUNTING will enable steal time for s390 as well.

-- 
blue skies,
   Martin.

"Reality continues to ruin my life." - Calvin.