We are seeing crashes with Java 1.5.0_10 that is used with AS 9.0 PE UR1 on a Sun Fire X2100 running Solaris 10. We see a JVM crash with the following stack trace


---- called from signal handler with signal 11 (SIGSEGV) ------
  [11] libjvm.so:estimate_path_freq(0x197d37a8), at 0xfea4e7f9
  [12] libjvm.so:IdealLoopTree::merge_many_backedges(0x199632f8, 0xf80a4d98), at
 0xfea965d7
  [13] libjvm.so:IdealLoopTree::beautify_loops(0x199632f8, 0xf80a4d98), at 0xfe9
4a610
  [14] libjvm.so:PhaseIdealLoop::PhaseIdealLoop(0xf80a4d98, 0xf80a4c34, 0x0, 0x1
), at 0xfe9892a6
  [15] libjvm.so:Compile::Optimize(0xf80a56d4), at 0xfe9cacaa
  [16] libjvm.so:Compile::Compile(0xf80a56d4, 0xf80a5b78, 0x8171990, 0x14420528,
 0xf6, 0x1), at 0xfe9c8bf8
  [17] libjvm.so:C2Compiler::compile_method(0x8171990, 0xf80a5b78, 0x14420528, 0
xf6), at 0xfe9c5c5b
  [18] libjvm.so:CompileBroker::invoke_compiler_on_method(0x8cd9230), at 0xfe9c6
16c
  [19] libjvm.so:CompileBroker::compiler_thread_loop(0xfede0000, 0xf80a5ca8, 0xf
ea03fd1, 0x81cd568, 0x81cd568, 0x81ce0a0), at 0xfea472ae
  [20] libjvm.so:compiler_thread_entry(0x81cd568, 0x81cd568), at 0xfea4a455
  [21] libjvm.so:JavaThread::thread_main_inner(0x81cd568), at 0xfea03fd1
  [22] libjvm.so:JavaThread::run(0x81cd568), at 0xfea03f75
  [23] libjvm.so:_start(0x81cd568), at 0xfecbeaf2
  [24] libc.so.1:_thr_setup(0xfe422800), at 0xfef3f9ae
  [25] libc.so.1:_lwp_start(), at 0xfef3fc90

I'm filing this as a new bug because I don't see any filed with this exact stack trace. It's possible that this issue could be related to CR 6301887 which is closed as a duplicate of CR 6297035.

I've attached a few hs_err logs, the core file, and related libraries to this report.