Bugzilla – Bug 141042
recent java upgrade crashes some apps.
Last modified: 2006-02-03 22:13:18 UTC
recent JAVA upgrade renders azureus useless in x86_64 , before the upgrade, everthing is working fine. Starting Azureus... Java exec found in PATH. Verifying... Suitable java version found [java = 1.5.0_06] Configuring environment... Loading Azureus: java -Xms16m -Xmx128m -cp "/home/cristian/azureus/Azureus2.jar:/home/cristian/azureus/swt.jar:/home/cristian/azureus/swt-mozilla.jar:/home/cristian/azureus/swt-pi.jar" -Djava.library.path="/home/cristian/azureus" -Dazureus.install.path="/home/cristian/azureus" org.gudy.azureus2.ui.swt.Main '' # # An unexpected error has been detected by HotSpot Virtual Machine: # # SIGSEGV (0xb) at pc=0x00002aaaaae4a6c0, pid=8701, tid=46912501738240 # # Java VM: Java HotSpot(TM) 64-Bit Server VM (1.5.0_06-b05 mixed mode) # Problematic frame: # C [libc.so.6+0x6f6c0] index+0x50 # # An error report file with more information is saved as hs_err_pid8701.log # # If you would like to submit a bug report, please visit: # http://java.sun.com/webapps/bugreport/crash.jsp # ./azureus: line 107: 8701 Abortado ${JAVA_PROGRAM_DIR}java -Xms16m -Xmx128m -cp "${CLASSPATH}" -Djava.library.path="${PROGRAM_DIR}" -Dazureus.install.path="${PROGRAM_DIR}" org.gudy.azureus2.ui.swt.Main "$@" Azureus TERMINATED. complete log of the hard crash attached. ps: tested latest cvs of Azureus and I got the same result.
Created attachment 61769 [details] complete log of the crash
Thanks a lot for the bug report ... Did the previous Java version definitely work with this particular kernel version (the log says 2.6.13-15.7)? If not, it seems to be the same or a similar problem described in Bug #139758 and Bug #115562 .
Daniel : the bugs you mentioned are private; I can't access them. I 'll look into this issue later. thanks.
Same thing here: My IDEA JavaIDE can't find any classes now. And if I try Yast Update, I get conflicts on Java. Everything was fine before the recent security update.
How can I back out of the recent Java security patch?
Daniel: can you comment about bug #139758 and bug #115562??? we don't have access to those bugs..and I have no idea what are you talking about.. :-( so..If I understand correctly.. this can be either a bug in JAVA or Kernel ?
The recent Java scecurity path (from SUSE update) also broke Tomcat5.
Byrce: oops. that's serious. did you find a workaround or something ?? I don't knwo if SUn has released any upgrade these days, gonna check it. Daniel.. any comment??
same problem reproduced with latest jre 1.6 beta http://download.java.net/jdk6/binaries/
Regarding the other two Bugs: They are SLES bugs, I will ask the assignees if it is possible to open them for the public. It's the same type of error: On 64Bit platforms the JIT segfaults with the newest Kernel (there however it is IBM Java, which uses the same source-base though). Christian: So far I couldn't recreate the bug. Could you try whether it works when the JIT is disabled? The option for that is -Xint ... that would be great, thanks. Bryce: Could you create an attachement with the error message? This seems to be another problem. Furthermore: What YAST conflicts do you get? Regarding rolling back the update: Go to YAST, search for the Java packages. For each package you want to downgrade, select the version tab below and select the desired version ... at the end: Accept. Thanks for your reports.
Our group here at Sun is actively investigating this problem and would like to know if you can help identify the cause. Right now the crash log looks more like a kernel issue. Could you please run your application after setting the environment variable LD_ASSUME_KERNEL to 2.4.1 and see if the issue goes away. In addition, if you have any other hs_err_pid files please attach them to the report. The one currently attached has "error occurred during error reporting" in the "Native frames" section. If there are any other, more complete, hs_err_pid files, that would help. Thanks, Sun Microsystems Inc.
ranjith.mandala@sun.com : I have sumitted my information to the SUN bugtracker long time ago, no respose at all. and No, setting LD_ASSUME_KERNEL to 2.4.1 doesn't solve the issue tested "mustang" and the SUSE version, same result.
dbornkessel@suse.de : using the -Xint I got the same crash :-(
Cristian, Could you please attach the complete hs_err_pid file or/and core file generated during the crash? Thanks, Ranjith Sun Microsystems Inc.
$ export LD_ASSUME_KERNEL=2.4.1 $ cd azureus $./azureus ( startup script changed to use -Xint ) (azureus fires up, for just a few seconds) # # An unexpected error has been detected by HotSpot Virtual Machine: # # SIGSEGV (0xb) at pc=0x00002aaaaaec94b0, pid=15789, tid=16384 # # Java VM: Java HotSpot(TM) 64-Bit Server VM (1.5.0_06-b05 interpreted mode) # Problematic frame: # C [libc.so.6+0x6e4b0] index+0x50 # # An error report file with more information is saved as hs_err_pid15789.log # # If you would like to submit a bug report, please visit: # http://java.sun.com/webapps/bugreport/crash.jsp # ./azureus: line 107: 15789 Abortado ${JAVA_PROGRAM_DIR}java -Xint -Xms16m -Xmx128m -cp "${CLASSPATH}" -Djava.library.path="${PROGRAM_DIR}" -Dazureus.install.path="${PROGRAM_DIR}" org.gudy.azureus2.ui.swt.Main "$@" Azureus TERMINATED. once again, I attach the crash log, should be similiar though..
Created attachment 66104 [details] Debug log debug log attached
Cristian, Is there any core file that was generated during the crash? Thanks, Ranjith Sun Microsystems Inc.
if somebody tell me HOW to intentioanlly produce a core file and where is located ..maybe I can.. ( found no core file in the directory at least )
Executing `ulimit -c unlimited` before invoking the software in question should work.
got a core file.. :-) but it's 111MB uncompressed and 9.4 MB compressed. here is http://www.onfocus.cl/judas/core.22909.bz2 ps: Torrents are not mission critical ..:-) but this problem should be crashing other really critical things.
a second core dump and all the info is here: http://www.onfocus.cl/judas/java-bug.tar.bz2 this time using java MUSTANG 1.6.0-rc-fastdebug-b69-debug hope this will be useful...
Cristian, Thanks for the log and the core files. We have opened up a new bug for this issue ( Bug Id: 6380404 ). You can monitor this bug on the Java Bug Database at http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=6380404. It may take a day or two before this bug shows up in this external database. Thanks, Ranjith Sun Microsystems Inc.
weird ... I have tracked down this problem.. seems to be a bug in the eclipse swt libs. updated to the latest version and the problem is gone. I wonder why the latest java update triggered the problem.