RFR: 8369238: Allow virtual thread preemption on some common class initialization paths [v3]

[Patricio Chilano Mateo]

On Wed, 29 Oct 2025 06:11:56 GMT, David Holmes <dholmes at openjdk.org> wrote:

>> Not sure if I like that idea because then we might have to change other callers along the way for this new convention and everybody's already confused by TRAPS so then they'd be confused by a new TRAPS too.
>
> I don't think we would have to. There are a handful of methods that now declare TRAPS but the only exception they should ever encounter is the PreemptException. It would be easier to understand the code if this was evident in their use of TRAPS. Also note it is purely documentation - the definition of `PREEMPTABLE_TRAPS` is exactly the same as `TRAPS`

If we want to avoid confusion with other users of `ObjectLocker` maybe we can leave it as it is and subclass it with a preemptable version? https://github.com/pchilano/jdk/compare/JDK-8369238...pchilano:jdk:PreemptableObjectLocker (this version could also use `PREEMPTABLE_TRAPS`)

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

PR Review Comment: https://git.openjdk.org/jdk/pull/27802#discussion_r2475339482