RFR: 8351375: nsk/jvmti/ tests should fail when nsk_jvmti_setFailStatus() is called

[Serguei Spitsyn]

On Thu, 13 Mar 2025 19:02:08 GMT, Leonid Mesnik <lmesnik at openjdk.org> wrote:

> The nsk_jvmti_setFailStatus() sometimes is called after test check results. In these cases the warning logs are generated and hide the real failure reasons. Also, I think it is a error-prone way to set and check error, since check might be just forgotten. Also, the test execution after failure might be incorrect and also make failure analysis harder.
> So I think it makes sense to always fail when nsk_jvmti_setFailStatus() is called.
> 
> If this is going to work I'll rename it later and add add optional message to be more informative.

The fix looks okay in general. I've posted one comment about typos and one question.

test/hotspot/jtreg/vmTestbase/nsk/share/jvmti/hotswap/HotSwap.cpp line 157:

> 155:       NSK_DISPLAY0("CompiledMethodLoad event recieved in dead phase");
> 156:       return;
> 157:     }

Nit: Typos at L148: `GetMethodNamme` => `GetMethodName`, `is work` => `works`, `phasem` => `phase`
A suggestion is to reformulate the comment as below:

    // GetMethodName works in live phase only so just exit if the event is generated too late


Also, I wonder if we want to abort/fail in all cases when `phase == JVMTI_PHASE_DEAD`.

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

PR Review: https://git.openjdk.org/jdk/pull/24040#pullrequestreview-2700233519
PR Review Comment: https://git.openjdk.org/jdk/pull/24040#discussion_r2004307114