Since this is a bug fix, please open a Bug for it so that this can be tracked as a defect closure.
On Thu, Oct 13, 2016 at 4:18 PM, Brian Brooks <[email protected]> wrote: > A timer pool's tick starts at t0 (zero). Once the first period has passed, > the timer pool is scanned for any timers that have expired since t0 + 1. > > Current code does an atomic fetch increment on the tick, but uses the > previous tick during timer expiration processing. What is needed is the > previous tick + 1. > > The observable effect without this patch is that timers are expired one > tick > period (timer resolution) later than they should be. > > Signed-off-by: Brian Brooks <[email protected]> > Reviewed-by: Bill Fischofer <[email protected]> > --- > platform/linux-generic/odp_timer.c | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/platform/linux-generic/odp_timer.c > b/platform/linux-generic/odp_timer.c > index becea9d..b26ac6b 100644 > --- a/platform/linux-generic/odp_timer.c > +++ b/platform/linux-generic/odp_timer.c > @@ -691,7 +691,7 @@ static void timer_notify(odp_timer_pool *tp) > prev_tick = odp_atomic_fetch_inc_u64(&tp->cur_tick); > > /* Scan timer array, looking for timers to expire */ > - (void)odp_timer_pool_expire(tp, prev_tick); > + (void)odp_timer_pool_expire(tp, prev_tick + 1); > > /* Else skip scan of timers. cur_tick was updated and next itimer > * invocation will process older expiration ticks as well */ > -- > 2.7.4 > >
