On Fri, 24 Feb 2023 20:08:21 GMT, Sergey Bylokhov <s...@openjdk.org> wrote:
>> There are lots of tests that rely on System.gc() actually triggering a gc. >> SFAIK all collectors we have today will obey it unless you use >> -XX:+DisableExplicitGC >> If you can make the test work with that then that would be interesting but >> calling just System.gc() is no worse than all those other tests. > > Many client tests generate OOM for that, we have a special Util, see the > usage of Util.generateOOME() or various implementation of generateOOME. > > If the OOM + sleep(to give the GC a chance to clean the weakrefs in case of > slow systems) does not work, then could it be considerred as a GC bug? > >>OutOfMemoryError: The Java Virtual Machine implementation has run out of >>either virtual or physical memory, and the automatic storage manager was >>unable to reclaim enough memory to satisfy an object creation request. > https://docs.oracle.com/javase/specs/jvms/se19/html/jvms-6.html#jvms-6.3 Yes, I saw this method, and [it periodically **calls `System.gc()`**](https://github.com/openjdk/jdk/blob/b4ea80731c6c0a0328a9801590ba5b081f08c3bd/test/jdk/javax/swing/regtesthelpers/Util.java#L102-L103). I still don't understand your concern. How is generating `OutOfMemoryError` better than calling `System.gc()`? Calling `System.gc()` conveys the intent in a clearer way. No OOME is required. ------------- PR: https://git.openjdk.org/jdk/pull/12594
