rp->rtort_pipe_count is modified concurrently by rcu_torture_writer(). To
prevent a data race, reuse i which is read using READ_ONCE before instead
of re-reading it.
Signed-off-by: linke li <lilinke99@xxxxxx>
Reviewed-by: Joel Fernandes (Google) <joel@xxxxxxxxxxxxxxxxx>
---
v1 -> v2: change the patch to focus on data race in rcu_torture_pipe_update_one ; update change log
kernel/rcu/rcutorture.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/kernel/rcu/rcutorture.c b/kernel/rcu/rcutorture.c
index 7567ca8e743c..53415121be39 100644
--- a/kernel/rcu/rcutorture.c
+++ b/kernel/rcu/rcutorture.c
@@ -466,7 +466,7 @@ rcu_torture_pipe_update_one(struct rcu_torture *rp)
i = RCU_TORTURE_PIPE_LEN;
atomic_inc(&rcu_torture_wcount[i]);
WRITE_ONCE(rp->rtort_pipe_count, i + 1);
- if (rp->rtort_pipe_count >= RCU_TORTURE_PIPE_LEN) {
+ if (i + 1 >= RCU_TORTURE_PIPE_LEN) {
rp->rtort_mbtest = 0;
return true;
}
--
2.39.3 (Apple Git-145)
