RFR: 8307653: Adjust delay time and gc log argument in TestAbortOnVMOperationTimeout [v2]

[Guoxiong Li]

On Wed, 10 May 2023 16:03:43 GMT, Daniel D. Daugherty <dcubed at openjdk.org> wrote:

>> I see the text `-J-Dtest.timeout.factor=4.0` in the log.
>
> When you get a timeout with 'timeout=480', what 'elapsed time' values are in the '.jtr' file.
> For my 'fastdebug' run, I see:
> 
> $ grep 'elapsed time' build/macosx-x86_64-normal-server-fastdebug/test-support/jtreg_open_test_hotspot_jtreg_runtime_Safepoint_TestAbortOnVMOperationTimeout_java/runtime/Safepoint/TestAbortOnVMOperationTimeout.jtr
> elapsed time (seconds): 3.458
> elapsed time (seconds): 3.455
> elapsed time (seconds): 10.176

I can't reproduce the failed test by using `@run driver/timeout=480` now. The test passes locally now with `timeout=480`. The `elapsed time` values, which belong to the current passed test, are shown below.


elapsed time (seconds): 115.718
elapsed time (seconds): 115.699
elapsed time (seconds): 555.602


I try to recall my previous steps:
- The test failed locally with `@run driver`,  without `timeout` option.
- I changed it to `@run driver/timeout=900`. It passed.
- I changed it to `@run driver/timeout=480`. It failed.
- I changed it to `@run driver/timeout=600`. It passed.

So I push the code with `@run driver/timeout=600`.

Now I try `@run driver/timeout=480` locally, it can't reproduce. So strange! Such intermittent fail may be related to [JDK-8296819](https://bugs.openjdk.org/browse/JDK-8296819)?

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

PR Review Comment: https://git.openjdk.org/jdk/pull/13878#discussion_r1190529074