RFR: 8256732: Zero: broken +ZeroTLAB exposes badly initialized memory [v4]

[Andrew Haley]

On Tue, 16 Mar 2021 09:10:15 GMT, David Holmes <dholmes at openjdk.org> wrote:

>> Marked as reviewed by dholmes (Reviewer).
>
> Sorry clicked too soon and hadn't added this comment. :)
> 
> I still think there is some underlying bug here with regards to ZeroTLAB but otherwise the changes you have made seem okay, and the proof as always is in the testing, so I've approved it.
> 
> Cheers,
> David

ZeroTLAB seems to do nothing useful for performance on any platform, even AArch64 which has bulk memory zeroing instructions. Can we not just nuke it?

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

PR: https://git.openjdk.java.net/jdk/pull/1343