Segmentation fault when running/debugging on android (Xperia Z)

Hi everyone!

I’m trying to create a simple android test as described in the howto ( but i can’t get them working…
Also, i installed the android tests from the store, and i can’t start anyone of them, as they are launching and immediatly closing, returning to the main menu of the test app.

I’ve made several tests, also used the nightly libraries and tried to run it from eclipse, but all tries ended in the same error - a segmentation fault.
Here’s the complete log which was displayed by logcat:

And this is the error message from eclipse:

09-15 15:07:26.449: A/libc(13803): Fatal signal 11 (SIGSEGV) at 0x000092e6 (code=1), thread 13908 (Thread-1423)

My Phone: Sony Xperia Z, Qualcom Snapdragon Quadcore @1.5Ghz, OpenGL ES 2.0, 2Gb Memory

Has someone encountered this error and fixed it or is this a bug from jme3?

I’m getting the same error on my Xperia Tablet Z, which I believe is basically your device with a larger screen.

F/libc ( 893): Fatal signal 11 (SIGSEGV) at 0x000092e6 (code=1), thread 907 (Thread-8343)

A native crash is almost always a bug in the firmeware, wich can somethimes be triggerd by jme due to using functions that are rarly used in normal phone useage.
(At the very least the opengl driver on the phone should return the appropiate error in the error checking part, sigsevs are usually programmer faults)

While this itself is not much helpful, you might be able to check that you are usising the current firmeware version. If you using cyanogen or similar it might ber worth to test different ones out.

I don’t like these errors:
I/dalvikvm(13531): Could not find method com.jme3.util.NativeObjectManager.registerForCleanup, referenced from method
W/dalvikvm(13531): VFY: unable to resolve virtual method 10745: Lcom/jme3/util/NativeObjectManager;.registerForCleanup (Lcom/jme3/util/NativeObject;)V
D/dalvikvm(13531): VFY: replacing opcode 0x6e at 0x001a
If the method registerForCleanup() is not called when its supposed to be called, all sorts of weird things could happen. I would like those errors to be fixed first before any further debugging is done on this case.

There seems to be information online about this error as well: