From the nightlies:

009-10-30T06:17:09.81] CLASSPATH="${test_work_dir}${PS}${CLASSPATH}"
[2009-10-30T06:17:09.81] # Actual: CLASSPATH=/export/local/24390.JDK7.NIGHTLY.VM+solaris-amd64_server_comp_nsk.stress.testlist/results/ResultDir/jck12a016:/export/local/common/testbase/6/vm/vm/bin/classes:/export/local/common/jdk/baseline/solaris-amd64/lib/tools.jar:/export/local/common/testbase/6/vm/vm/src/nsk/stress/share/classes:/net/sqenfs-1.sfbay/export1/comp/vm/testbase/jck/1.2/JCK-runtime-api-12a/classes:/net/sqenfs-1.sfbay/export1/comp/vm/testbase/jck/1.2/JCK-runtime-api-12a/javatest.jar
[2009-10-30T06:17:09.81] export CLASSPATH
[2009-10-30T06:17:09.81] ${JAVA} ${JAVA_OPTS} ${EXECUTE_CLASS} ${TEST_ARGS}
[2009-10-30T06:17:09.81] # Actual: /export/local/common/jdk/baseline/solaris-amd64/bin/java -d64 -server -Xcomp -XX:-PrintVMOptions -XX:CompileThreshold=100 -XX:-UseCompressedOops -XX:DefaultMaxRAMFraction=8 nsk.stress.jck12a.jck12a016.jck12a016 -stress:indulgent
[2009-10-30T06:17:09.81] Test count: 198
[2009-10-30T06:18:01.67] Thread count: 100
[2009-10-30T06:18:01.67] # To suppress the following error report, specify this argument
[2009-10-30T06:18:01.67] # after -XX: or in .hotspotrc: SuppressErrorAt=/synchronizer.cpp:1276
[2009-10-30T06:18:01.67] #
[2009-10-30T06:18:01.67] # A fatal error has been detected by the Java Runtime Environment:
[2009-10-30T06:18:01.67] #
[2009-10-30T06:18:01.67] # Internal Error (/tmp/jprt/P2/B/155909.cf231476/source/src/share/vm/runtime/synchronizer.cpp:1276), pid=3507, tid=35
[2009-10-30T06:18:01.67] # Error: assert(lock != mark->locker(),"must not re-lock the same lock")
[2009-10-30T06:18:01.67] #
[2009-10-30T06:18:01.67] # JRE version: 7.0-b74
[2009-10-30T06:18:01.67] # Java VM: Java HotSpot(TM) 64-Bit Server VM (17.0-b04-2009-10-29-155909.cf231476.main_to_comp-fastdebug compiled mode solaris-amd64 )
[2009-10-30T06:18:01.67] # An error report file with more information is saved as:
[2009-10-30T06:18:01.67] # /export/local/24390.JDK7.NIGHTLY.VM+solaris-amd64_server_comp_nsk.stress.testlist/results/ResultDir/jck12a016/hs_err_pid3507.log
[2009-10-30T06:18:01.67] #
[2009-10-30T06:18:01.67] # If you would like to submit a bug report, please visit:
[2009-10-30T06:18:01.67] # http://java.sun.com/webapps/bugreport/crash.jsp
[2009-10-30T06:18:01.67] #
[2009-10-30T06:18:01.67] Current thread is 35
[2009-10-30T06:18:01.67] Dumping core ...
[2009-10-30T06:18:01.67] # Test level exit status: 6
[2009-10-30T06:18:02.78]

The decoded log looks like this which points at an escape analysis problem:

Stack: [0xfffffd7fd187e000,0xfffffd7fd197e000], sp=0xfffffd7fd197af80, free space=3f30000000000000001k
Native frames: (J=compiled Java code, j=interpreted, Vv=VM code, C=native code)
V [libjvm.so+0x1f155bf];; __1cHVMErrorGreport6MpnMoutputStream__v_+0x68f
V [libjvm.so+0x1f1647d];; __1cHVMErrorOreport_and_die6M_v_+0x4fd
V [libjvm.so+0xaf1b6b];; __1cYreport_assertion_failure6Fpkci1_v_+0x5ab
V [libjvm.so+0x1d2f0fe];; __1cSObjectSynchronizerKslow_enter6FnGHandle_pnJBasicLock_pnGThread__v_+0x2ee
V [libjvm.so+0xb33d10];; __1cODeoptimizationOrelock_objects6FpnNGrowableArray4CpnLMonitorInfo___pnKJavaThrea
d__v_+0x4e0
V [libjvm.so+0xb2a7d6];; __1cODeoptimizationYfetch_unroll_info_helper6FpnKJavaThread__pn0ALUnrollBlock__+0x8
86

I reproduced this by running on a 2 cpu machine, prt-sol-x64-3, with this script where rerun.sh is the rerun script from the nightly failure:

#!/bin/sh

pchildren() {
  ptree $1 | nawk "\$1 == \"$1\" { print_it = 1; } { if (print_it) print \$1; }"
}

cleanup() {
    pchildren $pid | xargs kill -9
    exit 1
}

trap cleanup 2

rm -f core
while [ ! -f core ]; do
  sh rerun.sh > rerun.sh.log &
  pid=$!
  echo "Waiting for $pid"
  sleep 60
  pchildren $pid | xargs kill -9
done