Re: [RFC PATCH 16/22 -v2] add get_monotonic_cycles

[Mathieu Desnoyers]

* Steven Rostedt (rostedt@xxxxxxxxxxx) wrote:
> 
> 
> On Tue, 15 Jan 2008, Mathieu Desnoyers wrote:
> > >
> > > Signed-off-by: Steven Rostedt <srostedt@xxxxxxxxxx>
> > > ---
> > >  include/linux/clocksource.h |    3 ++
> > >  kernel/time/timekeeping.c   |   48 ++++++++++++++++++++++++++++++++++++++++++++
> > >  2 files changed, 51 insertions(+)
> > >
> > > Index: linux-compile-i386.git/kernel/time/timekeeping.c
> > > ===================================================================
> > > --- linux-compile-i386.git.orig/kernel/time/timekeeping.c	2008-01-09 14:27:26.000000000 -0500
> > > +++ linux-compile-i386.git/kernel/time/timekeeping.c	2008-01-09 14:34:40.000000000 -0500
> > > @@ -103,6 +103,54 @@ static inline void __get_realtime_clock_
> > >  	timespec_add_ns(ts, nsecs);
> > >  }
> > >
> > > +cycle_t notrace get_monotonic_cycles(void)
> > > +{
> > > +	cycle_t cycle_now, cycle_delta, cycle_raw, cycle_last;
> > > +
> > > +	do {
> > > +		/*
> > > +		 * cycle_raw and cycle_last can change on
> > > +		 * another CPU and we need the delta calculation
> > > +		 * of cycle_now and cycle_last happen atomic, as well
> > > +		 * as the adding to cycle_raw. We don't need to grab
> > > +		 * any locks, we just keep trying until get all the
> > > +		 * calculations together in one state.
> > > +		 *
> > > +		 * In fact, we __cant__ grab any locks. This
> > > +		 * function is called from the latency_tracer which can
> > > +		 * be called anywhere. To grab any locks (including
> > > +		 * seq_locks) we risk putting ourselves into a deadlock.
> > > +		 */
> >
> > I wonder what makes the compiler read the clock->cycle_raw and
> > clock->cycle_last variables twice ? I guess some memory barriers could
> > be welcome here ?
> 
> We need both cycle_raw and cycle_last to be the same from the time we read
> the clock source to the time we calculate cycle_delta. If either one
> changes then delta is bogus.
> 

Ok, but what actually insures that the clock->cycle_* reads won't be
reordered across the clocksource_read() ?


> Also, it just occurred to me that this is an old patch. I thought I
> renamed cycle_raw to cycle_monotonic. But I must have lost that patch :-/
> 
> >
> > > +		cycle_raw = clock->cycle_raw;
> > > +		cycle_last = clock->cycle_last;
> > > +
> > > +		/* read clocksource: */
> > > +		cycle_now = clocksource_read(clock);
> > > +
> > > +		/* calculate the delta since the last update_wall_time: */
> > > +		cycle_delta = (cycle_now - cycle_last) & clock->mask;
> > > +
> > > +	} while (cycle_raw != clock->cycle_raw ||
> > > +		 cycle_last != clock->cycle_last);
> > > +
> > > +	return cycle_raw + cycle_delta;
> > > +}
> 
> -- Steve
> 

-- 
Mathieu Desnoyers
Computer Engineering Ph.D. Student, Ecole Polytechnique de Montreal
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/