On Tue, 3 Nov 2020 23:16:50 GMT, Alex Menkov <amen...@openjdk.org> wrote:
> - Fixed synchronization logic in resexhausted001; > - On Windows JVM cannot produce OOM caused by thread creation failure. > Massive thread creation causes big system (not VM) memory consumption, as a > result test host dramatically slows down up to hang. Specifying small heap > size we can get RESOURCE_EXHAUSTED_JAVA_HEAP, but this is not the goal of > resexhausted001. > So I disabled resexhausted001 on Windows. > Had to implement runtime check (instead of using @requires) because > resexhausted001 is also called by resexhausted004. > - Additionally resexhausted001-resexhausted003 test were improved to verify > JVMTI generates ResourceExhausted event with correct flag. Hi Alex, The fix looks good to me. Thanks, Serguei ------------- Marked as reviewed by sspitsyn (Reviewer). PR: https://git.openjdk.java.net/jdk/pull/1046
