Re: [PATCH 2/2] ksefltest: pidfd: Fix wait_states: Test terminated by timeout

[lizhijian@xxxxxxxxxxx]

On 04/12/2021 01:03, Shuah Khan wrote:
> On 10/29/21 2:32 AM, Christian Brauner wrote:
>> On Fri, Oct 29, 2021 at 10:45:28AM +0800, Li Zhijian wrote:
>>> 0Day/LKP observed that the kselftest blocks foever since one of the
>>> pidfd_wait doesn't terminate in 1 of 30 runs. After digging into
>>> the source, we found that it blocks at:
>>> ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WCONTINUED, NULL), 0);
>>>
>>> we can reproduce it by:
>>> $ while true; do make run_tests -C pidfd; done
>>>
>>> a delay to ensure that the parent can see child process WCONTINUED.
>>>
>>> CC: Christian Brauner <christian@xxxxxxxxxx>
>>> CC: Shuah Khan <shuah@xxxxxxxxxx>
>>> CC: Philip Li <philip.li@xxxxxxxxx>
>>> Reported-by: kernel test robot <lkp@xxxxxxxxx>
>>> Signed-off-by: Li Zhijian <lizhijian@xxxxxxxxxxxxxx>
>>> ---
>>
>> Not a fan of the usleep() solution but if it fixes it it's fine for
>> a test, I think.
>> Acked-by: Christian Brauner <christian.brauner@xxxxxxxxxx>
>>
>
> I don't like introducing usleep() which will increase the kselftest
> run-time. Every little bit adds up if we allow usleep() in tests.

Thanks for your comments.

how about introduce a pipe to communicate between child and parent.


 From d68c4629dd60a1e22cb83b771d38e899352ff9a9 Mon Sep 17 00:00:00 2001
From: Li Zhijian <lizhijian@xxxxxxxxxxxxxx>
Date: Tue, 26 Oct 2021 16:39:56 +0800
Subject: [PATCH] ksefltest: pidfd: Fix wait_states: Test terminated by timeout

0Day/LKP observed that the kselftest blocks foever since one of the
pidfd_wait doesn't terminate in 1 of 30 runs. After digging into
the source, we found that it blocks at:
ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WCONTINUED, NULL), 0);

we can reproduce it by:
$ while true; do make run_tests -C pidfd; done

Introduce a blocking read in child process to make sure the parent can
check its WCONTINUED.

CC: Philip Li <philip.li@xxxxxxxxx>
Reported-by: kernel test robot <lkp@xxxxxxxxx>
Signed-off-by: Li Zhijian <lizhijian@xxxxxxxxxxxxxx>
---
  tools/testing/selftests/pidfd/pidfd_wait.c | 11 +++++++++--
  1 file changed, 9 insertions(+), 2 deletions(-)

diff --git a/tools/testing/selftests/pidfd/pidfd_wait.c b/tools/testing/selftests/pidfd/pidfd_wait.c
index be2943f072f6..d5c0ffa26c32 100644
--- a/tools/testing/selftests/pidfd/pidfd_wait.c
+++ b/tools/testing/selftests/pidfd/pidfd_wait.c
@@ -96,21 +96,26 @@ TEST(wait_states)
                 .flags = CLONE_PIDFD | CLONE_PARENT_SETTID,
                 .exit_signal = SIGCHLD,
         };
-       int ret;
+       int ret, pfd[2];
         pid_t pid;
         siginfo_t info = {
                 .si_signo = 0,
         };
-
+       ASSERT_EQ(pipe(pfd), 0);
         pid = sys_clone3(&args);
         ASSERT_GE(pid, 0);

         if (pid == 0) {
+               char buf[2];
+               close(pfd[1]);
                 kill(getpid(), SIGSTOP);
+               ASSERT_EQ(read(pfd[0], buf, 1), 1);
+               close(pfd[0]);
                 kill(getpid(), SIGSTOP);
                 exit(EXIT_SUCCESS);
         }

+       close(pfd[0]);
         ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WSTOPPED, NULL), 0);
         ASSERT_EQ(info.si_signo, SIGCHLD);
         ASSERT_EQ(info.si_code, CLD_STOPPED);
@@ -119,6 +124,8 @@ TEST(wait_states)
         ASSERT_EQ(sys_pidfd_send_signal(pidfd, SIGCONT, NULL, 0), 0);

         ASSERT_EQ(sys_waitid(P_PIDFD, pidfd, &info, WCONTINUED, NULL), 0);
+       ASSERT_EQ(write(pfd[1], "C", 1), 1);
+       close(pfd[1]);
         ASSERT_EQ(info.si_signo, SIGCHLD);
         ASSERT_EQ(info.si_code, CLD_CONTINUED);
         ASSERT_EQ(info.si_pid, parent_tid);
-- 
2.33.0