@@ -28,8 +28,7 @@ extern "C" {
*/
typedef struct odp_barrier_t {
int count;
- odp_atomic_int_t in;
- odp_atomic_int_t out;
+ odp_atomic_int_t bar;
} odp_barrier_t;
@@ -11,41 +11,35 @@
void odp_barrier_init_count(odp_barrier_t *barrier, int count)
{
barrier->count = count;
- barrier->in = 0;
- barrier->out = count - 1;
- odp_sync_stores();
+ barrier->bar = 0;
+ odp_sync_stores();
}
+/*
+ * Efficient barrier_sync -
+ *
+ * Barriers are initialized with a count of the number of callers
+ * that must sync on the barrier before any may proceed.
+ *
+ * To avoid race conditions and to permit the barrier to be fully
+ * reusable, the barrier value cycles between 0..2*count-1. When
+ * synchronizing the wasless variable simply tracks which half of
+ * the cycle the barrier was in upon entry. Exit is when the
+ * barrier crosses to the other half of the cycle.
+ */
void odp_barrier_sync(odp_barrier_t *barrier)
{
int count;
+ int wasless;
- odp_sync_stores();
-
- count = odp_atomic_fetch_inc_int(&barrier->in);
-
- if (count == barrier->count - 1) {
- /* If last thread, release others */
- barrier->in = 0;
- odp_sync_stores();
-
- /* Wait for others to exit */
- while (barrier->out)
- odp_spin();
-
- /* Ready, reset out counter */
- barrier->out = barrier->count - 1;
- odp_sync_stores();
+ wasless = barrier->bar < barrier->count;
+ count = odp_atomic_fetch_inc_int(&barrier->bar);
+ if (count == 2*barrier->count-1) {
+ barrier->bar = 0;
} else {
- /* Wait for the last thread*/
- while (barrier->in)
- odp_spin();
-
- /* Ready */
- odp_atomic_dec_int(&barrier->out);
- odp_mem_barrier();
+ while ((barrier->bar < barrier->count) == wasless)
+ odp_spin();
}
}
--
Adding odp_sync_stores() to odp_barrier_init() per Petri's comments. Signed-off-by: Bill Fischofer <bill.fischofer@linaro.org> --- include/odp_barrier.h | 3 +- platform/linux-generic/source/odp_barrier.c | 46 +++++++++++++---------------- 2 files changed, 21 insertions(+), 28 deletions(-) 1.8.3.2