RFR (XS): GCBasher OOMes with Shenandoah

[Aleksey Shipilev]

This seems to happen because our heap setting is too low (128M). Recent commit
in upstream bumped it to 256M for other collectors:
  http://hg.openjdk.java.net/jdk9/hs/hotspot/rev/5d25d00f0238

We should do it too:

diff -r f1f3863f4646 test/gc/stress/gcbasher/TestGCBasherWithShenandoah.java
--- a/test/gc/stress/gcbasher/TestGCBasherWithShenandoah.java	Mon Feb 13
16:52:53 2017 -0500
+++ b/test/gc/stress/gcbasher/TestGCBasherWithShenandoah.java	Tue Feb 14
09:04:12 2017 +0100
@@ -29,7 +29,7 @@
  * @key stress
  * @requires vm.flavor == "server"
  * @summary Stress the Shenandoah GC by trying to make old objects more likely
to be garbage than young objects.
- * @run main/othervm/timeout=200 -Xlog:gc*=info -Xmx128m -server
-XX:+UseShenandoahGC TestGCBasherWithG1 120000
+ * @run main/othervm/timeout=200 -Xlog:gc*=info -Xmx256m -server
-XX:+UseShenandoahGC TestGCBasherWithG1 120000
  */
 public class TestGCBasherWithShenandoah {
     public static void main(String[] args) throws IOException {

Ok?

Thanks,
-Aleksey