[Bug 813] New: Crash in JVM_handle_linux_signal

[bugzilla-daemon at icedtea.classpath.org]

http://icedtea.classpath.org/bugzilla/show_bug.cgi?id=813

             Bug #: 813
           Summary: Crash in JVM_handle_linux_signal
    Classification: Unclassified
           Product: IcedTea
           Version: unspecified
          Platform: x86
        OS/Version: Linux
            Status: NEW
          Severity: major
          Priority: P3
         Component: IcedTea7
        AssignedTo: ahughes at redhat.com
        ReportedBy: linuxhippy at gmail.com
                CC: unassigned at icedtea.classpath.org


>From time to time I get crashes with a useless hs_err file, stating the crash
happend in JVM_handle_linux_signal. There are a few other reports, all closed
as WONTFIX or INVALID with exact the same sympthoms.
I also got those crashes with OpenJDK-6, so this is no new regression.

Crashes with the same stack trace are:
https://bugs.launchpad.net/azureus/+bug/93769

I now found a way to reproduce the problem with
java-1.7.0-openjdk-1.7.0.0-0.1.20110803.1.fc16.i686 with a success rate of
about 50%, whereas the official oracle builds never fail (1.7.0u2).

I guess the crash isn't repsoduceable on any other system, as I think it might
be related to my version of mesa, but here is what I did:

java -Dsun.java2d.opengl=True -jar
/opt/oracle/jdk7/demo/jfc/Java2D/Java2Demo.jar 


Using gdb I was able to get a bit more insight into whats going wrong:


# A fatal error has been detected by the Java Runtime Environment:
#
#  SIGSEGV (0xb) at pc=0x00000001, pid=2549, tid=2091576128
#
# JRE version: 7.0
# Java VM: OpenJDK Server VM (21.0-b17 mixed mode linux-x86 )
# Problematic frame:
# C  0x00000001
Program received signal SIGSEGV, Segmentation fault.
__strlen_sse2_bsf () at ../sysdeps/i386/i686/multiarch/strlen-sse2-bsf.S:51
51        movdqu    (%edi), %xmm1
(gdb) bt
#0  __strlen_sse2_bsf () at ../sysdeps/i386/i686/multiarch/strlen-sse2-bsf.S:51
#1  0x0041facc in ?? () from
/usr/lib/jvm/java-1.7.0-openjdk-1.7.0.0/jre/lib/i386/server/libjvm.so
#2  0x003cee88 in ?? () from
/usr/lib/jvm/java-1.7.0-openjdk-1.7.0.0/jre/lib/i386/server/libjvm.so
#3  0x0069c9ae in ?? () from
/usr/lib/jvm/java-1.7.0-openjdk-1.7.0.0/jre/lib/i386/server/libjvm.so
#4  0x00434d70 in ?? () from
/usr/lib/jvm/java-1.7.0-openjdk-1.7.0.0/jre/lib/i386/server/libjvm.so
#5  0x007ee8f6 in ?? () from
/usr/lib/jvm/java-1.7.0-openjdk-1.7.0.0/jre/lib/i386/server/libjvm.so
#6  0x007f02a1 in ?? () from
/usr/lib/jvm/java-1.7.0-openjdk-1.7.0.0/jre/lib/i386/server/libjvm.so
#7  0x006a4610 in JVM_handle_linux_signal () from
/usr/lib/jvm/java-1.7.0-openjdk-1.7.0.0/jre/lib/i386/server/libjvm.so
#8  0x00699334 in ?? () from
/usr/lib/jvm/java-1.7.0-openjdk-1.7.0.0/jre/lib/i386/server/libjvm.so
#9  <signal handler called>
#10 0x00000001 in ?? ()

I wonder whats going wrong there, and why Oracle's version doesn't crash here?

-- 
Configure bugmail: http://icedtea.classpath.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.