Right, it's a PAUSE; you can follow the mailing list thread I linked to to learn about the background. Also Gil Tene shared a very nice project so we can all experiment with it: - https://github.com/giltene/GilExamples/tree/master/SpinWaitTest
The Readme file contains links to experimental OpenJDK builds including the patch. On 28 January 2016 at 07:04, Ladislav Thon <[email protected]> wrote: >> Any idea how HotSpot would optimize spin-waiting? The patch contains >> just the method definition... > > IIUC, it's supposed to emit the PAUSE instruction on x86; can be a noop > (or any other architecture-specific insn) on other architectures. > > LT > >> >> Cheers >> Dan >> >> >> On Wed, Jan 27, 2016 at 6:41 PM, Sanne Grinovero <[email protected]> >> wrote: >>> Hello all, >>> I've been following the discussion about onSpinWait(), it seems JDK9 >>> might get it. >>> >>> That looks quite interesting, especially after having noticed that >>> sometimes we "burn CPU" while we could hint that the thread should >>> rather be descheduled. >>> >>> http://mail.openjdk.java.net/pipermail/core-libs-dev/2016-January/038361.html >>> >>> Thanks, >>> Sanne >>> _______________________________________________ >>> infinispan-dev mailing list >>> [email protected] >>> https://lists.jboss.org/mailman/listinfo/infinispan-dev >> _______________________________________________ >> infinispan-dev mailing list >> [email protected] >> https://lists.jboss.org/mailman/listinfo/infinispan-dev >> > > _______________________________________________ > infinispan-dev mailing list > [email protected] > https://lists.jboss.org/mailman/listinfo/infinispan-dev _______________________________________________ infinispan-dev mailing list [email protected] https://lists.jboss.org/mailman/listinfo/infinispan-dev
