cbuffer cbuffer_data : register(b0) { uint4 data[3]; }; matrix data_load_1(uint offset) { const uint scalar_offset = ((offset + 0u)) / 4; uint4 ubo_load_1 = data[scalar_offset / 4]; uint2 ubo_load = ((scalar_offset & 2) ? ubo_load_1.zw : ubo_load_1.xy); vector ubo_load_xz = vector(f16tof32(ubo_load & 0xFFFF)); float16_t ubo_load_y = f16tof32(ubo_load[0] >> 16); const uint scalar_offset_1 = ((offset + 8u)) / 4; uint4 ubo_load_3 = data[scalar_offset_1 / 4]; uint2 ubo_load_2 = ((scalar_offset_1 & 2) ? ubo_load_3.zw : ubo_load_3.xy); vector ubo_load_2_xz = vector(f16tof32(ubo_load_2 & 0xFFFF)); float16_t ubo_load_2_y = f16tof32(ubo_load_2[0] >> 16); const uint scalar_offset_2 = ((offset + 16u)) / 4; uint4 ubo_load_5 = data[scalar_offset_2 / 4]; uint2 ubo_load_4 = ((scalar_offset_2 & 2) ? ubo_load_5.zw : ubo_load_5.xy); vector ubo_load_4_xz = vector(f16tof32(ubo_load_4 & 0xFFFF)); float16_t ubo_load_4_y = f16tof32(ubo_load_4[0] >> 16); const uint scalar_offset_3 = ((offset + 24u)) / 4; uint4 ubo_load_7 = data[scalar_offset_3 / 4]; uint2 ubo_load_6 = ((scalar_offset_3 & 2) ? ubo_load_7.zw : ubo_load_7.xy); vector ubo_load_6_xz = vector(f16tof32(ubo_load_6 & 0xFFFF)); float16_t ubo_load_6_y = f16tof32(ubo_load_6[0] >> 16); return matrix(vector(ubo_load_xz[0], ubo_load_y, ubo_load_xz[1]), vector(ubo_load_2_xz[0], ubo_load_2_y, ubo_load_2_xz[1]), vector(ubo_load_4_xz[0], ubo_load_4_y, ubo_load_4_xz[1]), vector(ubo_load_6_xz[0], ubo_load_6_y, ubo_load_6_xz[1])); } void main() { uint2 ubo_load_8 = data[2].xy; vector ubo_load_8_xz = vector(f16tof32(ubo_load_8 & 0xFFFF)); float16_t ubo_load_8_y = f16tof32(ubo_load_8[0] >> 16); const vector x = mul(data_load_1(0u), vector(ubo_load_8_xz[0], ubo_load_8_y, ubo_load_8_xz[1])); return; }