reported on hs20/b04, 7/b123.
gc.gctests.gctest01.gctest01
gc/gctests/gctest01
http://sqeweb.sfbay.sun.com/nfs/results/vm/gtee/JDK7/NIGHTLY/VM/2011-02-13/Comp_Baseline/vm/linux-i586/server/comp/linux-i586_vm_server_comp_vm.gc.testlist/ResultDir/gctest01/hs_err_pid10855.log

I checked the history and it seems to happen most weekends when -Xcomp testing
is run (and this one with -XX:+DeoptimizeALot). The test itself passes but the
VM fails heap verification at exit, with (in the cases that i checked) apparently
a bad reference from an object in a survivor space into Eden (pointing sometimes,
but not always, above the top of Eden).

Here's a catalog of recent failures of this:-

gc/gctests/gctest01 Compiler Unknown linux-i586 2011-02-13
gc/gctests/gctest01 Compiler Unknown windows-amd64 2011-02-06
gc/gctests/gctest01 Compiler Unknown linux-amd64 2011-02-06
gc/gctests/gctest01 Compiler Unknown linux-amd64 2011-01-30
gc/gctests/gctest01 Compiler Unknown solaris-i586 2011-01-30
gc/gctests/gctest01 Compiler Unknown solaris-amd64 2011-01-16
gc/gctests/gctest01 Compiler Unknown solaris-amd64 2011-01-09
gc/gctests/gctest01 Compiler Unknown solaris-sparcv9 2011-01-02