Performance of memory var handles in hot loops

[John Rose]

On Apr 8, 2020, at 4:21 AM, Maurizio Cimadamore <maurizio.cimadamore at oracle.com> wrote:
> 
> These heroics work to some degree, but having things working correctly at the C2 level would be preferrable, as I suspect that we are also adding a lot of noise.

+100  I think this is covered by https://bugs.openjdk.java.net/browse/JDK-8223051

AFAIK we are pretty close to being able to tackle this, because
of recent JIT work on strip mining.  JIT folks would know better
than me.

— John