------- Comment #24 from andreast at gcc dot gnu dot org 2008-02-03 22:08 ------- I have the same results on shrintel except the TestClosureGC.jar, this one passes on x86_64 and x86 multilib. But I know this one is hairy.
Nevertheless, do you run a parallel check in libjava? Me is wondering because the failing part is this usleep and if it hangs you can not reliably tell where it hung except you run a single test run. (Means you have problems to investigate where it 'hanged'.) I do not get any mentionable problems on x86/x86_64. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=35035
