Some methods of the QOpenGLFunctions_3_2_Core class do not work as expected

[Alex7634]

Hello!

The glVertexAttribPointer, glDrawElements, QOpenGLVertexArrayObject.Binder methods cause the program to crash.
I was able to partially restore the functionality of the glVertexAttribPointer methods, glDrawElements by removing the null check in QOpenGLFunctions_3_2_Core.

The example (https://drive.google.com/file/d/1TL1i81L8NvztQx0VlHbcd7P9KkQASQ1k/view?usp=sharing) uses alternative methods of non-working methods.

I think that in methods void glVertexAttribPointer(int index, int size, int type, boolean normalized, int stride, java.nio.Buffer pointer) and void glDrawElements(int mode, int count, int type, java.nio.Buffer indices ) the problem is with the java.nio.Buffer argument.

In the QOpenGLVertexArrayObject.Binder method, I was unable to determine the cause.

[Alex7634]

Quite a lot of time has passed.

I understand that there is no interest in restoring the functionality of OpenGL functions in QtJambi.

[omix]

Hi, sorry for my late response. I am deeply busy with porting QtJambi to Android.
I am definitely interested in OpenGL functionality. Coupe of weeks ago, I took a look at your code. The only thing I can confirm is, that QtJambi translates java.nio.Buffer null to void* nullptr.

[Alex7634]

Thanks for the clarification on this matter.

Will wait.

[omix]

Is the only problem the Java null reference check throwing a NullPointerException?
I an not sure why there are null checks for Java Buffers. I will remove them.

[Alex7634]

Unfortunately my English is not good enough.

Is the only problem the Java null reference check throwing a NullPointerException?

This is not the only problem. The const void *ptr argument to the void QOpenGLFunctions::glVertexAttribPointer(GLuint indx, GLint size, GLenum type, GLboolean normalized, GLsizei stride, const void *ptr) function can have a non-zero value.
With a non-null value, this function does not work for me with the NullPointerException removed.

I an not sure why there are null checks for Java Buffers. I will remove them.

I am not a big expert in this matter, and I could be wrong. It seems to me that this is the standard script behavior for such arguments in a function. Everything is somewhat more complicated and the const void *ptr argument must be handled by the script in some other way.

[omix]

Solved