On 8/27/20 7:51 AM, Andrew Hughes wrote: > Looks ok to me. Not having a g1_write_barrier_pre_helper function > which is actually Shenandoah-only should add a little clarity :-) Pushed: http://hg.openjdk.java.net/aarch64-port/jdk8u-shenandoah/hotspot/rev/e0a34665279f -- Thanks, -Aleksey