On Sun, 9 Aug 2026 16:30:32 GMT, Paul Hübner <[email protected]> wrote:
>> TBH I'm not sure how we can use `@require vm.flagless` to tackle this >> problem. Thus I can withdraw this PR and hand over it to you. >> >> I believe `createLimitedTestJavaProcessBuilder` is used intentionally, but >> some oversights might be included, I'm not sure. >> Anyway I think we need to scrutinize whether the parameter (e.g. >> `--enable-preview` should be affected to spawning process. In >> TestG1HeapRegion.java, I think it does not need to propergate because it >> spawns debugger process, not debuggee. > >> TBH I'm not sure how we can use `@require vm.flagless` to tackle this >> problem. Thus I can withdraw this PR and hand over it to you. >> > Sorry, the flagless is something I wanted to bring up on the side. It does > not impact this problem directly, it's just an optimization to skip wasting > resources. > >> I believe `createLimitedTestJavaProcessBuilder` is used intentionally, but >> some oversights might be included, I'm not sure. Anyway I think we need to >> scrutinize whether the parameter (e.g. `--enable-preview` should be affected >> to spawning process. In TestG1HeapRegion.java, I think it does not need to >> propergate because it spawns debugger process, not debuggee. > > Sounds good, this can be done in a follow-up JBS issue I think. I'm going to > approve this PR, and I'll file the JBS issue ASAP (I'm in the process of > transferring my account to another email). @Arraying Thanks for your approval and suggestion! @plummercj Can you approve this PR and agree with Paul's plan? ------------- PR Comment: https://git.openjdk.org/jdk/pull/32262#issuecomment-5235016239
