Re: [RFC][patch 00/12] clocksource / timekeeping rework V2

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

On Wed, 29 Jul 2009 09:10:31 -0600
dwalker@fifo99.com wrote:

> On Wed, 2009-07-29 at 15:41 +0200, Martin Schwidefsky wrote:
> > There is still more room for improvement. Some sore points are:
> > 
> > 1) The cycle_last value still is in the struct clocksource. It should
> >    be in the struct timekeeper but the check against cycles_last in
> > the
> >    read function of the TSC clock source makes it hard.
> > 2) read_persistent_clock returns seconds. With a really good initial
> >    time source this is not very precise. read_persistent_clock should
> >    return a struct timespec.
> > 3) xtime, raw_time, total_sleep_time, timekeeping_suspended, jiffies,
> >    the ntp state and probably a few other values may be better located
> >    in the struct timekeeper as well.
> 
> 
> You could also consolidate the clocksource_unregister() path and the
> clocksource_change_rating(0) path  , both are basically doing the same
> thing.. Neither one is heavily used..

I'm not quite sure I got this. If I look at the code:

/**
 * clocksource_change_rating - Change the rating of a registered clocksource
 */
void clocksource_change_rating(struct clocksource *cs, int rating)
{
        mutex_lock(&clocksource_mutex);
        cs->rating = rating;
        clocksource_select();
        mutex_unlock(&clocksource_mutex);
}
EXPORT_SYMBOL(clocksource_change_rating);

/**
 * clocksource_unregister - remove a registered clocksource
 */
void clocksource_unregister(struct clocksource *cs)
{
        mutex_lock(&clocksource_mutex);
        clocksource_dequeue_watchdog(cs);
        list_del(&cs->list);
        clocksource_select();
        mutex_unlock(&clocksource_mutex);
}
EXPORT_SYMBOL(clocksource_unregister);

the two functions do different things. What exactly is the idea you've
got in mind?

-- 
blue skies,
   Martin.

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