That reminds me of a bug we fixed in the mlvm patches during JavaOne; it was a heap stomping problem. I don't think there was a workaround.
I think you have two options with this problem: - wait for the mlvm changes to percolate into JDK7 and the JDK7 standard builds - obtain or build a version of the JVM from the mlvm patches (or have you done this?) -- John On Jun 22, 2009, at 6:57 AM, Chanwit Kaewkasi wrote: > I see, Remi ! > Thank you for the explanation. > > Chanwit > > On Mon, Jun 22, 2009 at 2:43 PM, Rémi Forax<[email protected]> wrote: >> It's more a GC crash, with more memory you don't trigger the GC, that >> why it's working fine. _______________________________________________ mlvm-dev mailing list [email protected] http://mail.openjdk.java.net/mailman/listinfo/mlvm-dev
