IDEA crashes all the time on Mac OS X

Hello!

i tried to use the 8xxx series but after starting it, it crashes. Or once it started, crashes a few minutes later.
i am using 10.5.3 with Java Update for Leopard (Java 6). Same problem was using 10.5.2, i hoped that 10.5.3 will resolve the problem. I was wrong.

Since this problem appeared, the 7.x version crashes too, not so regularly.
I can't program or develop anything, i can't use it for an hour either.
I don't use any plugins at all, permission repaired, no other application crashes but idea.

The error report of Apple print the description to the screen attached as rtf file...


Where can i find the IDEA error log? I would copy it here.


Any help is appreciated!
Thanks in advance!

Edited by: Fazekas Imre on Jun 2, 2008 12:40 PM



Attachment(s):
IdeaProblem.rtf
23 comments
Comment actions Permalink

Hello Fazekas,

This problem is a pure native crash inside the JDK code. There's nothing
we can do to fix this problem. You should run some hardware checks to verify
that the memory on your machine is OK.

i tried to use the 8xxx serites but after starting it, it crashes. Or
once it started, crashes a few minutes later.

i'm using 10.5.3 with Java Update for Leopard (Java 6). Same problem
was using 10.5.2, i hoped that 10.5.3 will resolve the problem. I was
wrong.

Since this problem appeared, the 7.x version creshes too, not so
regularly.
I can't program or develop anything, i can't use it for an hour
either.
I don't use any plugins at all, permission repaired, no other
application crashes but idea.
The error report of Apple print the description to the screen attached
as rtf file...

Where can i find the IDEA error log? I would copy it here.

Any help is appreciated! Thanks in advance!

Edited by: Fazekas Imre on May 31, 2008 11:21 PM

--
Dmitry Jemerov
Development Lead
JetBrains, Inc.
http://www.jetbrains.com/
"Develop with Pleasure!"


0
Comment actions Permalink

OK, thanks. I will check the memory in a few hours.
By the way, is it sure that it will be memory problem? no other program craches.

0
Comment actions Permalink

The program named memtest ran perfectly, no error signed.
I'm trying to test with Apple Hardware Test.

0
Comment actions Permalink

OK, thanks. I will check the memory in a few hours.
By the way, is it sure that it will be memory problem? no other program craches.


It might not be, but it's certainly not a problem in IDEA. Try submitting an error report to Apple.

For what it's worth, I've be using IDEA an my Macbook (Leopard 10.5.3 32 bits, Java 5) without any problems.

0
Comment actions Permalink

As a workaround, you can edit the file Diana-8418.app/Contents/Info.plist:

change the value of the JVMVersion key from "1.5+" to "1.5*".

That will force Idea to use the latest available 1.5 version.

I've found the application responsiveness to be better with 1.5 anyway.

0
Comment actions Permalink

Thanks for the replies, i really appreciate it! :)

The Apple's hardware test ran correctly without error message. Everything seems to be all right.
Could the Java 6 be the source/cause of these crashes?
(Thinking back of my previous system, i used tiger with Java 5.0 without any problem.)


I sent all crash log to the Apple. I send 3 logs a day at least, so i'm sure they have put me on the spam list. :)

Edited by: Fazekas Imre on Jun 2, 2008 5:58 PM

0
Comment actions Permalink

Hello Fazekas,

Thanks for the replies, i really appreciate it! :)

The Apple's hardware test ran correctly without error message.
Everything seems to be all right.
Could the Java 6 be the source/cause of these crashes?
(Thinking back of my previous system, i used tiger with Java 5.0
without any problem.)


As far as I could see from your error report, IDEA actually runs under Java
5 (5.0_13) on your machine. So Java 6 should not be a source of errors.

--
Dmitry Jemerov
Development Lead
JetBrains, Inc.
http://www.jetbrains.com/
"Develop with Pleasure!"


0
Comment actions Permalink

The Memtest 4.2 ran correctly too without error, so the memory is OK.
I don't have any other idea to resolve this problem?
Should i set the idea to use Java 6 at startup?
How can i achieve this (on MacOSX)?

Edited by: Fazekas Imre on Jun 2, 2008 8:23 PM

0
Comment actions Permalink

If you really want to try it, leave the JVMVersion set at 1.5+ (see message above).

Then open your Java Preferences (/Applications/Utilities/Java), and in the bottom pane drag Java 6 to the top. Hit Save and (re)start Idea.

0
Comment actions Permalink

You could try to install a fresh Mac OS X in parallel to your existing version.

0
Comment actions Permalink

I changed to Java 6 JRE as default, and started the idea 7.03, worked fined for 2 days.

After having downloaded and started the latest Selena, the EAP build crashed 3 times while reading the Java 6 API classes.
I gave up to start it and getting used the version 7.03 and it crashed too in a moment.

I don't have any idea what is going on. I haven't tried to use Java 5 as project JDK, Java 6 is indispensable in my projects. :(

0
Comment actions Permalink

I'm panicing now, and crying...
I reinstalled the Leopard. I upgraded the system to 10.5.3 including Java Update 1 for Leopard.
The result is the same. The idea 8xxx series craches all the time. i can't even start it. The 7.03 can be started but crashes in a hour surely.
I'm using a lot of Java-based program and no other program crashes.

To my despair, the netbeans works fine!!!! I don't want to use netbeans...

And don't have any idea... Anyone?

0
Comment actions Permalink

When the idea starts at first time, requests a password for fslogger.
If i don't give the right pass 3 times, the idea starts and it is stable and doesn't crash!

What is this fslogger exactly?

0
Comment actions Permalink

Hello Fazekas,

When the idea starts at first time, requests a password for fslogger.
If i don't give the right pass 3 times, the idea starts and it is
stable and doesn't crash!

What is this fslogger exactly?


fslogger lets IDEA monitor file system changes in order to refresh files
and directories more efficiently after external changes.

--
Dmitry Jemerov
Development Lead
JetBrains, Inc.
http://www.jetbrains.com/
"Develop with Pleasure!"


0
Comment actions Permalink

It was stable for a half hour.
it crashed again. :(
The techtool pro 4.6 found the system hardware error-free, the apple hardware tester and the memtester too.
I'm out of ideas.

Why can the netbeans run nicely?
(The eclipse is running nicely too.)

Voila, my latest crash report generated by Leopard...

Edited by: Fazekas Imre on Jun 7, 2008 9:06 PM



Attachment(s):
mds_2008-06-07-173906_imre-fazekass-macbook-pro-2.crash.zip
0
Comment actions Permalink

I filled a bugreport (ID: 5993735) to Apple and wrote a topic into their Java devlist and i received this response by mail:

"
From the top of this crashed stack, I believe this is an issue with a JNI library used by IntelliJ (libyjpagent.jnilib). It appears that one of their instrumentation calls is throwing a C++ exception while attempting to create a new "YSet" object, and that is somehow triggering a hard crash.

Thread 22 Crashed:
0 libSystem.B.dylib 0x93e0e2ee __semwait_signal_nocancel + 10
1 libSystem.B.dylib 0x93e001b7 usleep$NOCANCEL$UNIX2003 + 61
2 libSystem.B.dylib 0x93e219bb abort + 85
3 libstdc++.6.dylib 0x962e0005 0x96298000 + 294917
4 libstdc++.6.dylib 0x962de10c __gxx_personality_v0 + 1108
5 libstdc++.6.dylib 0x962de14b std::terminate() + 29
6 libstdc++.6.dylib 0x962de261 __cxa_throw + 101
7 libstdc++.6.dylib 0x962de5d8 operator new(unsigned long) + 100
8 libstdc++.6.dylib 0x962de689 operator new[](unsigned long) + 17
9 libyjpagent.jnilib 0x0115d627 YSet::init() + 135 10 libyjpagent.jnilib 0x0115c121 instrumentForTracingAllocationJ2EE(MethodIDProvider const*, YVector&, YVector]]>&, bool, bool, bool) + 3089
11 libyjpagent.jnilib 0x01131388 MethodIDProvider::getMethodID(char const, char const, char const*) const + 1416
12 libyjpagent.jnilib 0x011316e1 class_file_load_hook_handler(_jvmtiEnv, JNIEnv_, _jclass, _jobject, char const, _jobject, long, unsigned char const, long, unsigned char**) + 81
13 libjvm.dylib 0x0064bc5d JVM_RaiseSignal + 236285
14 libjvm.dylib 0x004619b8 0x3b9000 + 690616
15 libjvm.dylib 0x0046d00f 0x3b9000 + 737295
16 libjvm.dylib 0x0046ce2d 0x3b9000 + 736813
17 libjvm.dylib 0x005fa36f JVM_DefineClassWithSource + 223
18 libjava.jnilib 0x00301c6d Java_java_lang_ClassLoader_defineClass1 + 400
19 ??? 0x039f09b1 0 + 60754353
20 ??? 0x039eaaaa 0 + 60730026
21 ??? 0x039eaaaa 0 + 60730026
"

0
Comment actions Permalink

The IDEA crash on my Ubuntu 8.04 too under Virtaulbox 1.6.2.
It's a virtual mashine. Does it mean something?
I try to test it by bootcamp-based XP install today.

Edited by: Fazekas Imre on Jun 8, 2008 10:57 AM

0
Comment actions Permalink

As workaround please remove -agentlib:yjpagent from Info.plist
Fazekas Imre wrote:

I filled a bugreport (ID: 5993735) to Apple and wrote a topic into their Java devlist and i received this response by mail:

"
From the top of this crashed stack, I believe this is an issue with a JNI library used by IntelliJ (libyjpagent.jnilib). It appears that one of their instrumentation calls is throwing a C++ exception while attempting to create a new "YSet" object, and that is somehow triggering a hard crash.

Thread 22 Crashed:
0 libSystem.B.dylib 0x93e0e2ee __semwait_signal_nocancel + 10
1 libSystem.B.dylib 0x93e001b7 usleep$NOCANCEL$UNIX2003 + 61
2 libSystem.B.dylib 0x93e219bb abort + 85
3 libstdc++.6.dylib 0x962e0005 0x96298000 + 294917
4 libstdc++.6.dylib 0x962de10c __gxx_personality_v0 + 1108
5 libstdc++.6.dylib 0x962de14b std::terminate() + 29
6 libstdc++.6.dylib 0x962de261 __cxa_throw + 101
7 libstdc++.6.dylib 0x962de5d8 operator new(unsigned long) + 100
8 libstdc++.6.dylib 0x962de689 operator new[](unsigned long) + 17
9 libyjpagent.jnilib 0x0115d627 YSet<int>::init() + 135
10 libyjpagent.jnilib 0x0115c121 instrumentForTracingAllocationJ2EE(MethodIDProvider const*, YVector<unsigned char>&, YVector<unsigned char>&, bool, bool, bool) + 3089
11 libyjpagent.jnilib 0x01131388 MethodIDProvider::getMethodID(char const, char const, char const*) const + 1416
12 libyjpagent.jnilib 0x011316e1 class_file_load_hook_handler(_jvmtiEnv, JNIEnv_, _jclass, _jobject, char const, _jobject, long, unsigned char const, long, unsigned char**) + 81
13 libjvm.dylib 0x0064bc5d JVM_RaiseSignal + 236285
14 libjvm.dylib 0x004619b8 0x3b9000 + 690616
15 libjvm.dylib 0x0046d00f 0x3b9000 + 737295
16 libjvm.dylib 0x0046ce2d 0x3b9000 + 736813
17 libjvm.dylib 0x005fa36f JVM_DefineClassWithSource + 223
18 libjava.jnilib 0x00301c6d Java_java_lang_ClassLoader_defineClass1 + 400
19 ??? 0x039f09b1 0 + 60754353
20 ??? 0x039eaaaa 0 + 60730026
21 ??? 0x039eaaaa 0 + 60730026
"



--
Best regards,
Maxim Mossienko
IntelliJ Labs / JetBrains Inc.
http://www.intellij.com
"Develop with pleasure!"

0
Comment actions Permalink

I removed the -agentlib:yjpagent=disablej2ee substring.
Same result, sorry.

I installed the bootcamp based XP, get back to you with the result.
If it crash under windows, i will take my computer into the Apple shop, despite of all hardware test good results.

0
Comment actions Permalink

IDEA is crashing on Windows too. It doesn't seems to be an IDEA problem to me.
Could any other problem that memory hardware issue? I ran all available memory testing program without error...

0
Comment actions Permalink

The memory in my macbook pro was replaced by a new one and it's working well. So, it was memory problem.

Sorry to all IDEA developers, it was a very tricky situation.
I don't know why why the memory testers and hardware detectors haven't marked any possible problems... It's quite unintelligible.

Edited by: Fazekas Imre on Jun 10, 2008 2:46 PM

0
Comment actions Permalink

My Mac version will hang all the time as well. I have swapped out memory but I see the same problem. (It happens with large projects and saving files) I've lost many hours of work due to these crashes.

I have not run into these problems with any other application on my Mac.

0
Comment actions Permalink

Here too, but I've only encountered those problems when exiting IntelliJ. Killing it helps, but it's far from great ofcourse.

0

Please sign in to leave a comment.