Bug 663300 - [abrt] java-1.6.0-openjdk-1:1.6.0.0-49.1.9.3.fc14: os::abort: Process /usr/lib/jvm/java-1.6.0-openjdk-1.6.0.0/jre/bin/java was killed by signal 6 (SIGABRT)
Summary: [abrt] java-1.6.0-openjdk-1:1.6.0.0-49.1.9.3.fc14: os::abort: Process /usr/li...
Keywords:
Status: CLOSED NOTABUG
Alias: None
Product: Fedora
Classification: Fedora
Component: java-1.6.0-openjdk
Version: 14
Hardware: i686
OS: Unspecified
low
medium
Target Milestone: ---
Assignee: Deepak Bhole
QA Contact: Fedora Extras Quality Assurance
URL:
Whiteboard: abrt_hash:995012aa53cef41ffc87940c395...
: 784894 (view as bug list)
Depends On:
Blocks:
TreeView+ depends on / blocked
 
Reported: 2010-12-15 10:34 UTC by Jean-frederic Clere
Modified: 2012-09-10 19:59 UTC (History)
11 users (show)

Fixed In Version:
Clone Of:
Environment:
Last Closed: 2012-07-26 16:05:34 UTC
Type: ---
Embargoed:


Attachments (Terms of Use)
File: backtrace (30.14 KB, text/plain)
2010-12-15 10:34 UTC, Jean-frederic Clere
no flags Details
File: backtrace (46.70 KB, text/plain)
2012-06-06 15:51 UTC, Alex Wajda
no flags Details
open-jdk error log (131.02 KB, text/x-log)
2012-09-05 09:10 UTC, Gang Li
no flags Details

Description Jean-frederic Clere 2010-12-15 10:34:02 UTC
abrt version: 1.1.14
architecture: i686
Attached file: backtrace
cmdline: java -Dprogram.name=run.sh -server -Xms1303m -Xmx1303m -XX:MaxPermSize=256m -Dorg.jboss.resolver.warning=true -Dsun.rmi.dgc.client.gcInterval=3600000 -Dsun.rmi.dgc.server.gcInterval=3600000 -Dsun.lang.ClassLoader.allowArraySyntax=true -agentlib:jdwp=transport=dt_socket,address=8000,server=y,suspend=n -Djava.net.preferIPv4Stack=true -Djava.library.path=/home/jfclere/JBPAPP_5_1/build/output/jboss-5.1.0.Branch/native/lib -Djava.endorsed.dirs=/home/jfclere/JBPAPP_5_1/build/output/jboss-5.1.0.Branch/lib/endorsed -classpath /home/jfclere/JBPAPP_5_1/build/output/jboss-5.1.0.Branch/bin/run.jar org.jboss.Main -b localhost -c default
component: java-1.6.0-openjdk
crash_function: os::abort
executable: /usr/lib/jvm/java-1.6.0-openjdk-1.6.0.0/jre/bin/java
kernel: 2.6.35.9-64.fc14.i686
package: java-1.6.0-openjdk-1:1.6.0.0-49.1.9.3.fc14
rating: 4
reason: Process /usr/lib/jvm/java-1.6.0-openjdk-1.6.0.0/jre/bin/java was killed by signal 6 (SIGABRT)
release: Fedora release 14 (Laughlin)
time: 1292408261
uid: 500

How to reproduce
-----
1. starting AS5.1 with  JAVA_OPTS="$JAVA_OPTS -agentlib:jdwp=transport=dt_socket,address=8000,server=y,suspend=n"
2.
3.

Comment 1 Jean-frederic Clere 2010-12-15 10:34:05 UTC
Created attachment 468830 [details]
File: backtrace

Comment 2 Deepak Bhole 2010-12-15 17:40:16 UTC
Is this error consistent, intermittent or one time?

Comment 3 Jean-frederic Clere 2010-12-16 07:30:41 UTC
It seems I am not able to reproduce it today.

Comment 4 Jean-frederic Clere 2010-12-21 10:38:43 UTC
I have it again today. Do you need something?

Comment 5 Deepak Bhole 2010-12-21 19:41:45 UTC
Yes, do you have an hs_err* file for the crash? If so, can you please attach it? It should contain some more information.

Comment 6 Jean-frederic Clere 2011-01-13 15:45:58 UTC
Package: java-1.6.0-openjdk-1:1.6.0.0-49.1.9.3.fc14
Architecture: i686
OS Release: Fedora release 14 (Laughlin)


How to reproduce
-----
1. start tomcat twice the second cores.
2.
3.

Comment 7 Deepak Bhole 2011-01-13 16:00:31 UTC
Hi, is there an hs_err* file you can attach for this issue?

Comment 8 Jean-frederic Clere 2011-01-14 07:16:31 UTC
No it wasn't created with his crash.

Comment 9 abrt-bot 2012-03-21 18:02:41 UTC
*** Bug 784894 has been marked as a duplicate of this bug. ***

Comment 10 abrt-bot 2012-03-30 13:25:19 UTC
Backtrace analysis found this bug to be similar to some already closed bugs from other components. You might want to check those bugs for additional information.

Bugs which were found to be similar to this bug: 
  java-1.6.0-openjdk: bug #575886, bug #604857, bug #605907, bug #638011, bug #679687, bug #708311
  java-1.7.0-openjdk: bug #784894

This comment is automatically generated.

Comment 11 Alex Wajda 2012-06-06 15:51:07 UTC
happens everytime when stopping JVM in debug mode 

backtrace_rating: 4
Package: java-1.7.0-openjdk-devel-1.7.0.3-2.1.fc17.7
Architecture: x86_64
OS Release: Fedora release 17 (Beefy Miracle)

Comment 12 Alex Wajda 2012-06-06 15:51:16 UTC
Created attachment 589946 [details]
File: backtrace

Comment 13 Deepak Bhole 2012-06-06 16:26:58 UTC
(In reply to comment #11)
> happens everytime when stopping JVM in debug mode 
> 
> backtrace_rating: 4
> Package: java-1.7.0-openjdk-devel-1.7.0.3-2.1.fc17.7
> Architecture: x86_64
> OS Release: Fedora release 17 (Beefy Miracle)

Hi.

Are you sure there isn't another debug session running? Your trace looks different and it appears to be happening due to network resource conflict.

Comment 14 Alex Wajda 2012-06-06 19:00:18 UTC
I am absolutelu sure that there is no other debug sessions. I tried a million times and spend the whole day trying to investigate this problem - no use.

It happens every time when you run Java in debug mode and then when you stop it it crashes with the following message:

"Disconnected from the target VM, address: '127.0.0.1:8000', transport: 'socket'
/opt/tomcat7/bin/catalina.sh stop
ERROR: transport error 202: connect failed: Connection refused
ERROR: JDWP Transport dt_socket failed to initialize, TRANSPORT_INIT(510)
JDWP exit error AGENT_ERROR_TRANSPORT_INIT(197): No transports initialized [../../../src/share/back/debugInit.c:741]
FATAL ERROR in native method: JDWP No transports initialized, jvmtiError=AGENT_ERROR_TRANSPORT_INIT(197)
/opt/tomcat7/bin/catalina.sh: line 439:  2203 Aborted                 (core dumped) "/usr/lib/jvm/java-1.7.0/bin/java" -Xdebug -Xrunjdwp:transport=dt_socket,address=8000,suspend=y,server=n -Xms512m -Xmx1024m -XX:MaxPermSize=256m -Djava.util.logging.manager=org.apache.juli.ClassLoaderLogManager -Djava.endorsed.dirs="/opt/tomcat7/endorsed" -classpath "/opt/tomcat7/bin/bootstrap.jar:/opt/tomcat7/bin/tomcat-juli.jar" -Dcatalina.base="/home/awajda/.IntelliJIdea11/system/tomcat/Run_HI_(1)_HI-search" -Dcatalina.home="/opt/tomcat7" -Djava.io.tmpdir="/opt/tomcat7/temp" org.apache.catalina.startup.Bootstrap stop"

The Java process is left running and "kill" is the only way to stop it.

I'm having the same problem on both my laptops (both running F17): one was upgraded from F16 and the other one has a clean F17 installation.

Comment 15 Alex Wajda 2012-06-06 19:04:20 UTC
The strange thing is that the debug itself is working and the first message in the Tomcat logs says:
/opt/tomcat7/bin/catalina.sh run
Connected to the target VM, address: '127.0.0.1:8000', transport: 'socket'
Jun 06, 2012 8:51:36 PM org.apache.catalina.core.AprLifecycleListener init

So the port is correct, host is correct, debug is working, but just does not stop property and ends up hanging.

Comment 16 Omair Majid 2012-06-06 19:06:55 UTC
(In reply to comment #14)
> "/usr/lib/jvm/java-1.7.0/bin/java" -Xdebug
> -Xrunjdwp:transport=dt_socket,address=8000,suspend=y,server=n

This line (part of the command you ran) looks strange. Did you mean "server=y" and "suspend=n" ?

Comment 17 Alex Wajda 2012-06-06 21:16:01 UTC
No, that part is set automatically by IDE (IDEA in my case). I don't know why server=n, but I don't think it related to the problem. suspend=y is also Ok as IDE connects to it as soon as JVM is up.

Regarding the problem - I seem to have found a workaround. Presumably it is caused by the IDE issue, but nevertheless JVM shouldn't been crashing with fatal error. So it's rather inadequate error handling sort of issue of JVM.

I happened to have JAVA_OPTS variable to be set twice in my IDE Tomcat run/debug profile. Like this:
JAVA_OPTS=-Xdebug -Xrunjdwp:transport=dt_socket,address=127.0.0.1:53225,suspend=y,server=n
JAVA_OPTS=-Xdebug -Xrunjdwp:transport=dt_socket,address=127.0.0.1:53225,suspend=y,server=n -Xms512m -Xmx1024m -XX:MaxPermSize=256m

The second one supposed to be overriding the first one, and it worked well on other OS's, but for some weird reason on my fresh F17 this causes the JVM to crash on shutdown presumably because IDE analyses those two variables separately(?) and tries to connect/disconnect twice to/from the same address, I don't know... that's all my guessing.

To work it around I got rid of overriding JAVA_OPTS and replaced the 2nd one with the CATALINA_OPTS=-Xms512m -Xmx1024m -XX:MaxPermSize=256m. For Tomcat it is essentially the same as in the catalina.sh script $JAVA_OPTS is directly followed by $CATALINA_OPTS resulting in the same sequence of JVM command line arguments, but the JVM crash gone know.

Comment 18 Roman Kennke 2012-06-06 22:44:02 UTC
Port 8000 is most likely taken by tomcat itself. In the 'workaround' you use another port, that's probably why it works. It should not crash though, but print out that the address is already in use.

Comment 19 Florian Furlanetto 2012-07-05 08:31:48 UTC
Starting up Glassfish 3.1.2 in debug mode from Netbeans 7.1.2 

backtrace_rating: 4
Package: java-1.7.0-openjdk-devel-1.7.0.3-2.2.1.fc17.8
Architecture: x86_64
OS Release: Fedora release 17 (Beefy Miracle)

Comment 20 Florian Furlanetto 2012-07-12 07:24:57 UTC
Starting Glassfish 3.1.2 in debug mode from Netbeans 7.1.2

backtrace_rating: 4
Package: java-1.7.0-openjdk-devel-1.7.0.3-2.2.1.fc17.8
Architecture: x86_64
OS Release: Fedora release 17 (Beefy Miracle)

Comment 21 Deepak Bhole 2012-07-12 17:08:07 UTC
Hi Florian, have you tried it with another port as mentioned in comment 18?

Comment 22 fred 2012-07-25 13:06:21 UTC
ERROR: transport error 202: bind failed: Adresse déjà utilisée --->>>> adress already in use!
ERROR: JDWP Transport dt_socket failed to initialize, TRANSPORT_INIT(510)
JDWP exit error AGENT_ERROR_TRANSPORT_INIT(197): No transports initialized [../../../src/share/back/debugInit.c:741]
FATAL ERROR in native method: JDWP No transports initialized, jvmtiError=AGENT_ERROR_TRANSPORT_INIT(197)


backtrace_rating: 4
Package: java-1.7.0-openjdk-devel-1.7.0.3-2.2.1.fc17.8
Architecture: x86_64
OS Release: Fedora release 17 (Beefy Miracle)

Comment 23 Deepak Bhole 2012-07-26 16:05:34 UTC
Looks like the JDWP port you are attempting to use is already in use by another (Java?) process. Please check the output of 'netstat -ant' to confirm and kill that process first.

Comment 24 Gang Li 2012-09-05 09:10:44 UTC
Created attachment 609917 [details]
open-jdk error log

When i use eclipse, the eclipse exit with an error report on java(open jdk), and a error log reported. See attachment.

Comment 25 Deepak Bhole 2012-09-10 19:59:28 UTC
(In reply to comment #24)
> Created attachment 609917 [details]
> open-jdk error log
> 
> When i use eclipse, the eclipse exit with an error report on java(open jdk),
> and a error log reported. See attachment.

That trace seems unrelated to this issue. Please open a separate bug against the eclipse component for that as it is coming form native code outside of OpenJDK's control.


Note You need to log in before you can comment on or make changes to this bug.