Passing an array of vec4 (GLfloat quad) as uniform to openGL shaders

Question

I want to pass an array of vec4's to the fragment shader of OpenGL as a uniform and it seems a lot harder than I hoped for. It came to a point where I am trying to pass an array of GLfloat's and reconstruct the vec4s in the shader, but it still doesn't work

I tried declaring in the vshader:

#extension GL_NV_shader_buffer_load : enable
uniform GLfloat* points;

and from the code, call:

glUniform1fv(points, pointsLen*4, points);

thinking that i am passing floats in groups of 1, and i have 4 times the number of vec4s.
Sadly, this vshader doesn't pass compilation to a GLSL program.

How do i fix this? Either by fixing this or doing it in a completely different way :)

Answer 1

NVIDIA's shader_buffer_load doesn't work that way. When you declare a uniform pointer, that means that you are telling the compiler that you will provide a GPU address where the data for that pointer is stored. That GPU address comes from a buffer object.

So you need to put your data in a buffer object and then make it resident (so that you can get the GPU address). You don't give it a CPU pointer to an array of data.

If you just want to have an array of data, you don't need shader_buffer_load at all. GLSL allows for uniform arrays. You just have to put an explicit size on them.