#version 440

layout(local_size_x = 16, local_size_y = 16) in;
layout(rgba8, binding = 1) readonly uniform image2D inputImage;
layout(rgba16f, binding = 2) writeonly uniform image2D outputImage;

// There is no equivalent of gl_NumWorkGroups in HLSL. So instead pass the
// values in in a uniform buffer.
layout(std140, binding = 0) uniform numWorkGroupsBuf {
    uvec3 numWorkGroups;
};

void main()
{
    if (gl_GlobalInvocationID.x >= numWorkGroups.x || gl_GlobalInvocationID.y >= numWorkGroups.y)
        return;
    vec4 value = imageLoad(inputImage, ivec2(gl_GlobalInvocationID.xy));
    imageStore(outputImage, ivec2(gl_GlobalInvocationID.xy), value);
}