timeout: fix subprocess is never terminated #4315
Conversation
|
would it be possible to add a test for this? |
f490aaf to
7f03e29
Compare
Added |
49fc1ac to
edd81ed
Compare
2ac1cb7 to
5dadb70
Compare
5dadb70 to
6fade5a
Compare
6fade5a to
bf2b5f5
Compare
|
I'm not sure why the CI is failing randomly. |
bf2b5f5 to
6f42d5c
Compare
|
GNU testsuite comparison: |
6f42d5c to
b9d3c48
Compare
|
GNU testsuite comparison: |
b9d3c48 to
3678c6e
Compare
|
I have increased the timeout from 5 seconds to 10 seconds, and the failing tests are not related to this PR. |
|
GNU testsuite comparison: |
3678c6e to
46e94f1
Compare
|
GNU testsuite comparison: |
46e94f1 to
f406b56
Compare
|
I checked the GNU test logs and found that this should also fix #4073, which is caused by the split subprocess ( |
|
Indeed this reduces the log size from 86MB to just 2MB, that's great! Let's hope the CI becomes a bit more stable as a result. |
Closes #4176.
This should also fix #4073.
Follow GNU's timeout behavior, send a signal to the process group to make sure subprocesses are cleaned up, and also send SIGCONT to the child and the whole process group after that.