Why does SphereCollisionShape give Fatal signal 11?

It worked in galaxy nexus, but now when using Samsung S4 it suddenly gives fatal signal 11 upon start. I havn’t changed anything in my code.
I have installed other games on my galaxy nexus, made by myself with jme3, and they all work fine. Just this one that doesn’t co/op.
I found another thread (only one) about fatal signal 11. And it seems like the problem was that you attach the geometry directly to the geometry.
I, on the other hand, does not attach anything to this controller. I simply get the position of it and adjust my geometries by that.

Anyhow heres (part of) my code:
[java] public class PlayerBody extends RigidBodyControl{

public PlayerBody(Person person){
	super(new SphereCollisionShape(0.8f),0.2f);

}
}[/java]

Anyone has any idea of what might be causing this?

Native bullet (which is used for android due to jbullets slowness there) still isn’t finished. Especially the collision code still is buggy and theres few checks or fallbacks to avoid native crashes. Is the sphere maybe colliding already at the beginning? Do you add or remove any object twice? Bullet doesn’t like that. Generally you give way too few info on whats actually happening or what part of your code fails. If you don’t know, use System.out.println’s until you find out to where it gets before crashing.

What seems to be happening is that the game starts twice.
[java] public MainActivity() throws Exception{

	Log.d("adde", "Starting");
	appClass = "start.MainClass";
	eglConfigType = ConfigType.FASTEST;
	exitDialogTitle = "Exit?";
	exitDialogMessage = "Do you wish to exit?";
	eglConfigVerboseLogging = true;
	screenOrientation = ActivityInfo.SCREEN_ORIENTATION_LANDSCAPE;
	mouseEventsInvertX = false;
	mouseEventsInvertY = false;
	
	
	if (thais != null){
		//throw new Exception("Multiple instance of class MainActivity is NOT ALLOWED.");
	} 
	thais = this;
}[/java] 

First this code is run once.
Then
[java]public MainGame(){
Log.d(“adde”, “New maingamee created”);
app = this;
}[/java]
But it never reaches to simpleInitApp()
Instead it runs
[java]public MainActivity() throws Exception{
Log.d(“adde”, “Starting”);
appClass = “start.MainClass”;…[/java]
again.
MainClass extends MainGame btw.
But as for this problem I have found no way of solving it. Why would it startup twice?

Hm throw and catcha exception in the doubl eclaled method, (in the catch write it to a file, at the end you should have two stacktraces written down with the path they were called from) Maybee this helps further.

I can uncomment that throwing of exception that I made but that doesn’t help. It will just throw the exception then. If I throw, it why catch it at the same place? I could just use println instead then…
Not sure I understand what u ment…

@Addez said: I can uncomment that throwing of exception that I made but that doesn't help. It will just throw the exception then. If I throw, it why catch it at the same place? I could just use println instead then.. Not sure I understand what u ment..

He just wants you to be able to see the calling stack trace for who is calling the method both times. Generating and printing an exception is an easy why to do this. You don’t even have to throw it.

new Throwable().printStackTrace();

…for example, will dump the current stack trace to System.out.

Now I’v fixed the problem with double initiation. It will start twice, but the first one close, then I reset thais (this variable) and then the new one starts.
The thing that makes the game restart is this line:
[java]screenOrientation = ActivityInfo.SCREEN_ORIENTATION_LANDSCAPE;[/java]
If I remove this, the game starts only once.
But get fatal signal 1 between creating SimpleApplication instance and the simpleInitApp() .

On the other hand if I keep the screenOrientation line and let it start twice. Then it does not even create a SimpleApplication object and does not run a simpleInitApp at all either.

Changing my own exception throwing with this
[java]new Throwable().printStackTrace();[/java]
Will also only print fatal signal 11 and nothing more.
Man Im so stuck on this…

@Addez said: Changing my own exception throwing with this [java]new Throwable().printStackTrace();[/java] Will also only print fatal signal 11 and nothing more. Man Im so stuck on this..

Note: that idea was not to replace any existing exception handling you are doing. It purely a way to get DEBUG info on what the calling stack trace is. It will show you who called what methods and so on that led to where you are. It’s a good way to hunt down why things happen twice when they aren’t supposed to, etc…

But anyway, I have no idea what a println does on an Android device. That was just an example. You might have to print the trace to a file or something.

Well, before the fatal signal I get this:
[java]06-10 11:50:56.339: I/OGLESShaderRenderer(27444): INFO OGLESShaderRenderer 11:50:56 GL_EXTENSIONS:
GL_AMD_compressed_ATC_texture GL_AMD_performance_monitor GL_AMD_program_binary_Z400
GL_EXT_debug_labelGL_EXT_debug_markerGL_EXT_robustness GL_EXT_texture_format_BGRA8888
GL_EXT_texture_type_2_10_10_10_REV GL_NV_fence GL_OES_compressed_ETC1_RGB8_texture GL_OES_depth_texture
GL_OES_depth24 GL_OES_EGL_image GL_OES_EGL_image_external GL_OES_element_index_uint
GL_OES_fbo_render_mipmap GL_OES_fragment_precision_high GL_OES_get_program_binary
GL_OES_packed_depth_stencil GL_OES_depth_texture_cube_map GL_OES_rgb8_rgba8 GL_OES_standard_derivatives
GL_OES_texture_3D GL_OES_texture_float GL_OES_texture_half_float GL_OES_texture_half_float_linear
GL_OES_texture_npot GL_OES_vertex_half_float GL_OES_vertex_type_10_10_10_2 GL_OES_vertex_array_object
GL_QCOM_alpha_test GL_QCOM_binning_control GL_QCOM_driver_control GL_QCOM_perfmon_global_mode
GL_QCOM_extended_get GL_QCOM_extended_get2 GL_QCOM_tiled_rendering GL_QCOM_writeonly_rendering
GL_EXT_sRGB GL_EXT_texture_filter_anisotropic GL_EXT_color_buffer_float GL_EXT_color_buffer_half_float
06-10 11:50:56.349: A/libc(27444): Fatal signal 11 (SIGSEGV) at 0x000092e6 (code=1), thread 27461 (Thread-2405)
[/java]

Could it be the renderer that has a bug?