A curious thing seems to happen when I attempt to run VBOTest with the latest release and latest CVS: the vm segfaults in the call to glClear at the beginning of render(). On a whim, I added a boolean to avoid the call on the first time through render and it seems to work fine.
I using Gentoo Linux 2.6.7 with NVidia GeForce3 and I have the latest version of the Nvidia drivers (same behavior with the prev version). I tried __GL_SINGLE_THREADED=1 but the vm just locks up before display even occurs. The same occurs with java 1.4.2 and 5rc.
Also, glxGears and quake3 are working fine, but JOGL also crashes (but I have not yet investigated that problem)
I'd appreciate any advice on how to diagnose this futher or pointers on where to get more information. Thanks!
Does other LWJGL OpenGL tests work fine? If so, it seems more like a problem with the VBO implementation.
- elias
the pbuffers test and the vs&vsh shaders appear to work flawlessly, so the problem does seem to be confined to VBOTest and VBOIndexTest.
For further investigation, installed java3d and tried some of the demos. Most appear to be working but segfault inside libGLcore.so.1 (nvidia lib) at exit.
Since no java gl libs appear to be working properly, I must have something odd in my system setup that is not interacting well with nvidia's drivers and the java vm.:cry: I was just hoping that someone had seen something similar and knew of a quick fix.
Just in case anyone else encounters this, my troubles appear to be a manifestation of:
http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=5073409
There's more discussion here:
http://www.javagaming.org/cgi-bin/JGNetForums/YaBB.cgi?board=2D;action=display;num=1095845978;start=