Skip to contentSkip to conversation

selftests/bpf: fix possible hang in sockopt_inherit

PATCH Applied Stanislav Fomichev 3 participants 3 messages Jul 15, 2020 – Jul 16, 2020

Subsystems (5)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 selftests/bpf: fix possible hang in sockopt_inherit

Andrii reported that sockopt_inherit occasionally hangs up on 5.5 kernel [0]. This can happen if server_thread runs faster than the main thread. In that case, pthread_cond_wait will wait forever because pthread_cond_signal was executed before the main thread was blocking. Let's move pthread_mutex_lock up a bit to make sure server_thread runs strictly after the main thread goes to sleep.

(Not sure why this is 5.5 specific, maybe scheduling is less deterministic? But I was able to confirm that it does indeed happen in a VM.)

[0] https://lore.kernel.org/bpf/CAEf4BzY0-bVNHmCkMFPgObs=isUAyg-dFzGDY7QWYkmm7rmTSg@mail.gmail.com/

---
 tools/testing/selftests/bpf/prog_tests/sockopt_inherit.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
tools/testing/selftests/bpf/prog_tests/sockopt_inherit.c +1 -2
index 8547ecbdc61f..ec281b0363b8 100644
@@ -193,11 +193,10 @@ static void run_test(int cgroup_fd)
 	if (CHECK_FAIL(server_fd < 0))
 		goto close_bpf_object;
 
+	pthread_mutex_lock(&server_started_mtx);
 	if (CHECK_FAIL(pthread_create(&tid, NULL, server_thread,
 				      (void *)&server_fd)))
 		goto close_server_fd;
-
-	pthread_mutex_lock(&server_started_mtx);
 	pthread_cond_wait(&server_started, &server_started_mtx);
 	pthread_mutex_unlock(&server_started_mtx);
 
Signature · 1 line
2.27.0.389.gc38d7665816-goog
Acked-by on v1 1/1
#1 lore raw
Quoted 12 lines · Stanislav Fomichev

Andrii reported that sockopt_inherit occasionally hangs up on 5.5 kernel [0]. This can happen if server_thread runs faster than the main thread. In that case, pthread_cond_wait will wait forever because pthread_cond_signal was executed before the main thread was blocking. Let's move pthread_mutex_lock up a bit to make sure server_thread runs strictly after the main thread goes to sleep.

(Not sure why this is 5.5 specific, maybe scheduling is less deterministic? But I was able to confirm that it does indeed happen in a VM.)

Quoted 5 lines

[0] https://lore.kernel.org/bpf/CAEf4BzY0-bVNHmCkMFPgObs=isUAyg-dFzGDY7QWYkmm7rmTSg@mail.gmail.com/

---

Great, thanks for figuring this out! Hopefully this is it.

Quoted 23 lines
 tools/testing/selftests/bpf/prog_tests/sockopt_inherit.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
tools/testing/selftests/bpf/prog_tests/sockopt_inherit.c +1 -3
index 8547ecbdc61f..ec281b0363b8 100644
@@ -193,11 +193,10 @@ static void run_test(int cgroup_fd)
        if (CHECK_FAIL(server_fd < 0))
                goto close_bpf_object;
 
+       pthread_mutex_lock(&server_started_mtx);
        if (CHECK_FAIL(pthread_create(&tid, NULL, server_thread,
                                      (void *)&server_fd)))
                goto close_server_fd;
-
-       pthread_mutex_lock(&server_started_mtx);
        pthread_cond_wait(&server_started, &server_started_mtx);
        pthread_mutex_unlock(&server_started_mtx);
 
--

2.27.0.389.gc38d7665816-goog

maintainer Applied
#2 lore raw
In reply to Stanislav Fomichev
Quoted 7 lines · Stanislav Fomichev

Andrii reported that sockopt_inherit occasionally hangs up on 5.5 kernel [0]. This can happen if server_thread runs faster than the main thread. In that case, pthread_cond_wait will wait forever because pthread_cond_signal was executed before the main thread was blocking. Let's move pthread_mutex_lock up a bit to make sure server_thread runs strictly after the main thread goes to sleep.

Quoted 8 lines

(Not sure why this is 5.5 specific, maybe scheduling is less deterministic? But I was able to confirm that it does indeed happen in a VM.)

[0] https://lore.kernel.org/bpf/CAEf4BzY0-bVNHmCkMFPgObs=isUAyg-dFzGDY7QWYkmm7rmTSg@mail.gmail.com/

Applied, thanks!