ByteBuffer.allocateDirect().asFloatBuffer() vs BufferUtils.createFloatBuffer()

Question

I just solved a bug and I'm not sure why. Creating a 4x4 projection matrix in LWJGL, for use in a vertex shader..

This line causes problems. It fails silently and my mat4 in the shader is stuck as all zeros (as if it was never written).

FloatBuffer mProj = ByteBuffer.allocate(4*16).asFloatBuffer();

This works as expected.

FloatBuffer mProj = BufferUtils.createFloatBuffer(16);

As a sanity-check, I confirmed that my floats are 4 bytes. So what's the difference?

Sotirios Delimanolis · Accepted Answer · 2014-06-10 00:34:52Z

3

The only difference between those two is possibly the byte order. You can set it as

FloatBuffer mProj = ByteBuffer.allocate(4 * 16).order(ByteOrder.nativeOrder()).asFloatBuffer();

answered Jun 10, 2014 at 0:34

281k62 gold badges718 silver badges744 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

That's exactly it, thank you! I had looked at the documentation for ByteBuffer.asFloatBuffer, but it made no mention of byte order there.