RFR: 8295099: vmTestbase/nsk/stress/strace/strace013.java failed with "TestFailure: wrong lengths of stack traces: strace013Thread0: NNN strace013Thread83: MMM"

[David Holmes]

On Tue, 8 Nov 2022 12:25:55 GMT, Thomas Stuefe <stuefe at openjdk.org> wrote:

>> I changed things to remove the "k-2" slack, but also note that if a spurious wakeup were ever to occur then the test could again fail. I think the slack was misguided and should have been 3 - but in that case I fail to see the point of this test. If we get rare failures in the future I would advocate for the removal of this test.
>
>> I changed things to remove the "k-2" slack, but also note that if a spurious wakeup were ever to occur then the test could again fail. I think the slack was misguided and should have been 3 - but in that case I fail to see the point of this test. If we get rare failures in the future I would advocate for the removal of this test.
> 
> Seems reasonable. Lets see how your fix works.

@tstuefe  and @lmesnik  I have completely redone the fix and just corrected the slack value from 2 to 3. I also kept some comment fixes and added extra commentary.

If we want to fix these tests they should probably all be done together, though I doubt the others can be made more accurate.

Thanks

Thanks.

-------------

PR: https://git.openjdk.org/jdk/pull/11011