Add a chapter on Compute Shaders

[spencer-lunarg]

Rendered view

Kind review @jeffbolznv - mainly around the shared memory races and any other stuff you think would be useful

[SaschaWillems]

Having a chapter explaining basic compute terms is a great addition 👍🏻

[jeffbolznv]

I think it's confusing to lead with this example, and would be better to start with examples of cross-thread communication and using barrier() first.

[spencer-lunarg]

can you provide some GLSL snippets here (in the comment), I know you are more of an expert in this area and probably better at providing some better examples to provide

[jeffbolznv]

You could use a reduction as a simple example of cross-thread communication:

    shared float temp[BLOCK_SIZE];

    temp[tid] = x;
    [[unroll]] for (uint s = BLOCK_SIZE / 2; s > 0; s >>= 1) {
        barrier();
        float other = temp[tid ^ s];
        barrier();
        temp[tid] += other;
    }

---

shared float my_var[BLOCK_SIZE];

void reduction(float x) {
    uint tid = gl_GlobalInvocationID;

    my_var[tid] = x;
    for (uint s = BLOCK_SIZE / 2; s > 0; s >>= 1) {
        barrier();
        float other = my_var[tid ^ s];
        barrier();
        my_var[tid] += other;
    }
}

[spencer-lunarg]

sorry, we can add this in a follow up, this example makes is confusing... was is suppose to be an example of having a race condition?

[gpx1000]

The chapter is a solid addition and fills a clear gap in the guide!

[charles-lunarg]

Lots of grammar nits & suggestions of things that can be improved.