8231585: java/lang/management/ThreadMXBean/MaxDepthForThreadInfoTest.java fails with java.lang.NullPointerException

[Daniil Titov]

Hi David, Serguei, and Chris,

Thank you for reviewing this change. The issue is not easy to reproduce. I tried to run the test (without the fix) with Graal on in Mach5 about 600 times and so far the failure didn't reproduce. I will continue running these tests in the next few days and will update the issue with the information about disappearing thread when it gets reproduced.

Best regards,
Daniil


On 4/18/20, 6:56 AM, "David Holmes" <david.holmes at oracle.com> wrote:

    Hi Daniil,

    On 18/04/2020 6:03 am, Daniil Titov wrote:
    > Please review the change that fixes intermittent failure of java/lang/management/ThreadMXBean/MaxDepthForThreadInfoTest.java
    > 
    > As David noticed (thank you, David, for this analysis) there is no guarantee that all threads found by getAllThreadIds() are still alive by the time we call getThreadInfo() so we have to allow for null array entries.
    > 
    > Testing:  Mach5 tests with Graal on passed 300 times.

    Fix looks good - thanks.

    I think it would be informative to actually determine which thread(s) is 
    disappearing (not as part of the test just as some diagnostic info to 
    add to the JBS issue). Can you reproduce the failure easily?

    Thanks,
    David

    > [1] http://cr.openjdk.java.net/~dtitov/8231585/webrev.01/
    > [2] https://bugs.openjdk.java.net/browse/JDK-8231585
    > 
    > Best regards,
    > Daniil
    > 
    >