I have a very simple OpenGL program where the positions are embedded in the GLSL code. The vertex shader looks like this:
#version 440 core
uniform float offsetX;
uniform float offsetY;
out VS_OUT
{
vec2 tc;
} vs_out;
void main()
{
const vec2[4] position = vec2[4](vec2(-0.5, -0.5),
vec2(0.5, -0.5),
vec2(-0.5, 0.5),
vec2(0.5, 0.5));
vs_out.tc = position[gl_VertexID].xy + vec2(0.5);
vec2 pos = position[gl_VertexID].xy + vec2(offsetX, offsetY);
gl_Position = vec4(pos, 0.0, 1.0);
}
To run this, to me it seems like I should basically be able to just be able to call glDrawArrays from my application.
However, this is not the case. I my application I have the following code (reused from an earlier example):
// Create buffer
glCreateBuffers(1, &buffer);
glNamedBufferStorage(buffer, sizeof(vertices), vertices, bufferFlags);
// Create vao
glCreateVertexArrays(1, &vao);
// Enable attributes
glEnableVertexArrayAttrib(vao, positionIndex);
glEnableVertexArrayAttrib(vao, texCoordIndex);
// Set up formats for attributes
glVertexArrayAttribFormat(vao, positionIndex, positionSize, GL_FLOAT, GL_FALSE, firstElementOffset);
glVertexArrayAttribFormat(vao, texCoordIndex, texCoordSize, GL_FLOAT, GL_FALSE, positionSize * sizeof(GLfloat));
// Make attributes use binding 0
glVertexArrayAttribBinding(vao, positionIndex, bufferBindingIndex);
glVertexArrayAttribBinding(vao, texCoordIndex, bufferBindingIndex);
// Bind "buffer" to vertex buffer binding point 0
glVertexArrayVertexBuffer(vao, bufferBindingIndex, buffer, firstElementOffset, elementStride);
None of that actually seems necessary because my shaders are now responsible for generating the position and texture coordinates. However, if I comment out the call to glVertexArrayVertexBuffer my program crashes.
Am I required to have a buffer bound to binding point 0 (even if I do not use any of that buffer's data) before calling glDraw*?
