RFR: JDK-8292561: Make "ReplayCompiles" a diagnostic product switch

Thomas Stuefe stuefe at openjdk.org
Sat Aug 20 05:53:30 UTC 2022


On Fri, 19 Aug 2022 23:42:21 GMT, Vladimir Kozlov <kvn at openjdk.org> wrote:

> Current ciReplay tests requires `vm.debug == true` - that is why they can use debug flags.

What I meant was I first attempted to remove the `vm.debug == true` restriction and run the tests on the release VM. But then found the missing CICrashAt option to be the issue that made them fail with "did not crash, expected it to crash" errors. So I assumed that was the only issue, and if that were solved, we could just reuse all tests for the release VM.

> I don't think we need to modify them. For new tests which you want to run in product VM we can use new NMT feature you are proposing.

Okay.

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

PR: https://git.openjdk.org/jdk/pull/9935


More information about the hotspot-compiler-dev mailing list