RFR: 8306471: Add virtual threads support to JDWP ThreadReference.Stop and JDI ThreadReference.stop() [v10]

[Serguei Spitsyn]

On Thu, 11 May 2023 20:48:50 GMT, Chris Plummer <cjplummer at openjdk.org> wrote:

>> [JDK-8306034](https://bugs.openjdk.org/browse/JDK-8306034) (#13546) added some virtual thread support to JVMTI StopThread. This allows JDWP ThreadReference.Stop and JDI ThreadReference.stop() to have the same level support for virtual threads.
>> 
>> For your reference, here are the JVMTI spec changes that this pr is using as the basis for the JDWP and JDI spec updates.
>> 
>> https://github.com/openjdk/jdk/pull/13546/files#diff-cab9797fde25f344784e4473778cd35bfd9e19ba36f82c11b1a0674636958339
>> 
>> Mostly this is a spec update for JDWP and JDI, but there are also some minor changes needed to the ThreadReference.stop() handling of JDWP errors, and throwing the appropriate exceptions. Also some minor cleanup in jdb. The debug agent doesn't need changes since JVMTI errors are just passed through as the corresponding JDWP errors, and the code for this is already in place. These errors are not new nor need any special handling.
>> 
>> Our existing testing for ThreadReference.stop() is fairly weak:
>> 
>> - nsk/jdb/kill/kill001, which tests stop() when the thread is suspended at a breakpoint. It will get problem listed by [JDK-8306034](https://bugs.openjdk.org/browse/JDK-8306034). I have fixes for it already working and will push it separately.
>> - nsk/jdi/stop/stop001, which is problem listed and only tests when the thread is blocked in Object.wait()
>> - nsk/jdi/stop/stop002, which only tests that throwing an invalid exception fails properly
>> 
>> I decided to take stop002 and make it a more thorough test of ThreadReference.stop(). See the comment at the top of the test for a list of what is tested. As for reviewing this test, it's probably best to look at it as a completely new test rather than looking at diffs since so much has been changed and added.
>
> Chris Plummer has updated the pull request incrementally with one additional commit since the last revision:
> 
>   update linenumber for breakpoint. It changed with a recent commit that added a comment.

test/hotspot/jtreg/vmTestbase/nsk/jdi/ThreadReference/stop/stop002.java line 154:

> 152:                 log.complain("TEST #1 FAILED: expected IllegalArgumentException was not thrown");
> 153:                 tot_res = Consts.TEST_FAILED;
> 154:             } catch (InvalidTypeException ee) {

It is confusing that the log.complain at line 152 expects `IllegalArgumentException` but the really expected exception is `InvalidTypeException`. Is it intentional? If so then it is better to clarify it in the comment.

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

PR Review Comment: https://git.openjdk.org/jdk/pull/13548#discussion_r1191871355