On 26/11/18 12:59, Pavel Rappo wrote:
Yes - something that should be considered for a JDK 12+ fix perhaps. This fixOn 26 Nov 2018, at 12:12, Seán Coffey <sean.cof...@oracle.com> wrote: JDK-8213942 synchronization fix can be improvedLooks like a soft language for what actually is a race condition. I wonder at what point should we just rewrite the thing as Alan mentioned using "fully concurrent lookups"?
in its current guise is simple enough can can be backported. regards, Sean.
-Pavel