On Wed, 17 Mar 2021 16:40:47 GMT, Roger Riggs <rri...@openjdk.org> wrote:
>> The failures happened in tiers 6 and 8. The system may be overloaded so even >> 100ms may not be enough for the child process to start sleeping. From the >> error log, the child process tried to spawn a thread (probably one of those >> usually started during VM bootstrap) at around 118ms >> >> [0.118s][warning][os,thread] Failed to start thread - _beginthreadex failed >> (EACCES) for attributes: stacksize: default, flags: >> >> The test runs 4 times. Each time it checks only STDOUT or STDERR, but not >> both. So I think we can use the other stream to signal to the main process >> that the child process is ready. That would be more reliable than an >> arbitrary wait time. > > That complicates the test and the child quite a bit for minimal gain. Arbitrary time out has been a reliable source of intermittent failures. Since we have spent a lot of time analyzing this failure, I think it's worthwhile to fix it properly, which doesn't seem that complicated. That's better than the same bug happening again a year later and a different set of people would spend hours to analyze it again. ------------- PR: https://git.openjdk.java.net/jdk/pull/3049