Hi Sebastian, On Wed, Apr 30, 2014 at 01:23:34PM +0100, Sebastian Andrzej Siewior wrote: > On ARM one has to call register_current_timer_delay() in order to use > the (quick) timer based calibrarion instead of the a little slower loop > based delay. > The timer function specified in register_current_timer_delay() is also > used by read_current_timer() which would otherwise return -ENXIO. > And read_current_timer() timer is used by get_cycles() which in turn is > used by random_get_entropy(). That means each sub-architecture returned > here 0 if register_current_timer_delay() was no performed. > > The parameters for for sched_clock_register() and > register_current_timer_delay() are mostly the same. Instead of calling > register_current_timer_delay() each time just after (or before) > sched_clock_register() I thought it might easier by doing it once at > sched_clock time since the parameter are the same. That means each ARM sub > arch that working sched-clock would also have a working random_get_entropy() > implementation. > > Any comments?
As long as sched_clock is guaranteed to be a fixed frequency, always-on clocksource then this could work, but it removes the flexibility of having a separate delay clock and sched clock (is this useful?). Looking at your patch, I noticed that we need to extend the register_current_timer_delay function to deal with clocks that aren't as wide as cycle_t, otherwise we don't delay() for long enough when the clock overflows (this is potentially already an issue for architected timers < 64-bit). Could you cook a patch for that please? Will -- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
