v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v_01.x + m0;
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v_01.x + m2;
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v2 + m4;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v2 + m6;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v2 + m8;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v_01.x + m2;
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v6 + m6;
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v2 + m0;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v6 + m8;
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v2 + m6;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v2 + m2;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v_01.x + m4;
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v_01.x + m0;
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v_01.x + m2;
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v6 + m6;
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v_01.x + m4;
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v_01.x + m8;
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v6 + m2;
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v2 + m2;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v_01.x + m0;
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v_01.x + m0;
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v6 + m4;
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v_01.x + m4;
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v_01.x + m4;
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v2 + m8;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v6 + m0;
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v_01.x + m0;
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v_01.x + m2;
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v2 + m4;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v2 + m6;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_01 = v_01 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_89.x < v_01.x);
v1 = v_01.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1617;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v2 + m8;
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_23.x = v2;
v_23 = v_23 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_23.x + v_1011.x < v_23.x);
v2 = v_23.x;
v3 = v_23.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_1819;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+0]]
o0 = v_01.x + m2;
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + b
v_01 = v_01 + v_1011 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_01.x + v_1011.x < v_01.x);
v1 = v_01.y;
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
// a = a + m[sigma[r][2*i+1]]
// skip since adding 0u does nothing
xor = v_1011 ^ v_2021;
v_1011 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v10 = v_1011.x;
- v11 = v_1011.y;
+
+
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+0]]
// skip since adding 0u does nothing
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x >> 24u) | (xor.y << 8u), (xor.y >> 24u) | (xor.x << 8u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + b
v_67.x = v6;
v_67 = v_67 + v_89 + select(vec2<u32>(0u), vec2<u32>(0u, 1u), v_67.x + v_89.x < v_67.x);
v6 = v_67.x;
v7 = v_67.y;
- v8 = v_89.x;
- v9 = v_89.y;
+
+
// a = a + m[sigma[r][2*i+1]]
o0 = v6 + m6;
xor = v_89 ^ v_1819;
v_89 = vec2((xor.x << 1u) | (xor.y >> 31u), (xor.y << 1u) | (xor.x >> 31u));
- v8 = v_89.x;
- v9 = v_89.y;
+
+