var<workgroup> scratch: array<f32, 64>;

struct Params {
	count: u32,
	scale: f32,
}

@group(0) @binding(0) var<storage, read_write> params: Params;
struct DataBuffer {
	values: array<f32, 64>,
}

@group(0) @binding(1) var<storage, read_write> data: DataBuffer;

@compute @workgroup_size(64, 1, 1)
fn main(@builtin(local_invocation_index) __luma_work_local_index: u32, @builtin(global_invocation_id) __luma_work_global_id: vec3<u32>, @builtin(local_invocation_id) __luma_work_local_id: vec3<u32>, @builtin(workgroup_id) __luma_work_group_id: vec3<u32>) {
	let idx = __luma_work_local_index;
	let gid = __luma_work_global_id.x;
	let lid = __luma_work_local_id.x;
	let group_x = __luma_work_group_id.x;
	if ((gid < params.count)) {
		scratch[idx] = (data.values[idx] * params.scale);
		workgroupBarrier();
		data.values[idx] = (scratch[idx] + f32((lid + group_x)));
	}
}