[jdk18] RFR: 8278389: SuspendibleThreadSet::_suspend_all should be volatile/atomic

Thomas Schatzl tschatzl at openjdk.java.net
Fri Dec 10 14:07:21 UTC 2021


On Fri, 10 Dec 2021 13:51:33 GMT, Thomas Schatzl <tschatzl at openjdk.org> wrote:

> Hi all,
> 
>   can I have reviews for this change that makes `SuspendibleThreadSet::_suspend_all` volatile and all accesses to it use `Atomic` helpers?
> 
> I found that the only getter for that member is called `SuspendibleThreadSet::should_yield()`, which contradicts our naming rules; to keep the change minimal (and the naming discussion out of this PR because I do not feel that `_suspend_all` is particularly good) I introduced a private `suspend_all()` method.
> I can of course add some renaming in this change (but then for jdk19+ only).
> 
> Testing: gha, local compilation
> 
> Thanks,
>   Thomas

Closing to have proper GHA testing

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

PR: https://git.openjdk.java.net/jdk18/pull/9



More information about the hotspot-gc-dev mailing list