1#version 450
 2
 3#include "types.glsl"
 4#include "generic_binary_head.glsl"
 5
 6layout(local_size_x = 512, local_size_y = 1, local_size_z = 1) in;
 7
 8void main() {
 9    const uint idx = gl_GlobalInvocationID.x;
10    if (idx >= p.ne) {
11        return;
12    }
13
14    const uint offset = p.param3;
15    const uint src1_i = idx - offset;
16    const uint oz = src1_i / p.nb02;
17    const uint oy = (src1_i - (oz * p.nb02)) / p.nb01;
18    const uint ox = src1_i % p.nb01;
19
20    uint i00, i01, i02, i03;
21    get_indices(idx, i00, i01, i02, i03);
22
23    if (ox < p.ne10 && oy < p.ne11 && oz < p.ne12) {
24        data_d[get_doffset() + dst_idx(i00, i01, i02, i03)] = D_TYPE(FLOAT_TYPE(data_a[get_aoffset() + src0_idx(i00, i01, i02, i03)]) + FLOAT_TYPE(data_b[get_boffset() + ox + oy * p.ne10 + oz * p.ne10 * p.ne11]));
25    } else {
26        data_d[get_doffset() + dst_idx(i00, i01, i02, i03)] = D_TYPE(FLOAT_TYPE(data_a[get_aoffset() + src0_idx(i00, i01, i02, i03)]));
27    }
28}
29