Fix race-condition in epoll1 poller's begin_worker() and pollset_kick()
which caused the designated poller to miss a kick in some cases
diff --git a/src/core/lib/iomgr/ev_epoll1_linux.c b/src/core/lib/iomgr/ev_epoll1_linux.c
index f579b16..bd7c955 100644
--- a/src/core/lib/iomgr/ev_epoll1_linux.c
+++ b/src/core/lib/iomgr/ev_epoll1_linux.c
@@ -600,6 +600,7 @@
}
gpr_mu_unlock(&neighbourhood->mu);
}
+
worker_insert(pollset, worker);
pollset->begin_refs--;
if (worker->kick_state == UNKICKED) {
@@ -628,7 +629,18 @@
pollset->shutting_down);
}
- return worker->kick_state == DESIGNATED_POLLER && !pollset->shutting_down;
+ /* We release pollset lock in this function at a couple of places:
+ * 1. Brielfly when assigning pollset to a neighbourhood
+ * 2. When doing gpr_cv_wait()
+ * It is possible that 'kicked_without_poller' was set to true during (1) and
+ * 'shutting_down' is set to true during (1) or (2). If either of them is
+ * true, this worker cannot do polling */
+
+ /* TODO(sreek): Perhaps there is a better way to handle kicked_without_poller
+ * case; especially when the worker is the DESIGNATED_POLLER */
+
+ return worker->kick_state == DESIGNATED_POLLER && !pollset->shutting_down &&
+ !pollset->kicked_without_poller;
}
static bool check_neighbourhood_for_available_poller(