| /external/XNNPACK/src/f32-gemm/gen/ |
| D | 4x2c4-wasmsimd.c | 122 const v128_t vacc0x01c2 = wasm_f32x4_add( in xnn_f32_gemm_ukernel_4x2c4__wasmsimd() local
136 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 0, 1, 4, 5), in xnn_f32_gemm_ukernel_4x2c4__wasmsimd()
137 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 2, 3, 6, 7)); in xnn_f32_gemm_ukernel_4x2c4__wasmsimd()
|
| D | 4x2c4-minmax-wasmsimd-arm.c | 124 const v128_t vacc0x01c2 = wasm_f32x4_add( in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_arm() local
138 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 0, 1, 4, 5), in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_arm()
139 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 2, 3, 6, 7)); in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_arm()
|
| D | 4x2c4-minmax-sse.c | 121 …const __m128 vacc0x01c2 = _mm_add_ps(_mm_unpacklo_ps(vacc0x0c4, vacc0x1c4), _mm_unpackhi_ps(vacc0x… in xnn_f32_gemm_minmax_ukernel_4x2c4__sse() local
126 …__m128 vacc01x01 = _mm_add_ps(_mm_movelh_ps(vacc0x01c2, vacc1x01c2), _mm_movehl_ps(vacc1x01c2, vac… in xnn_f32_gemm_minmax_ukernel_4x2c4__sse()
|
| D | 4x2c4-relu-wasmsimd.c | 122 const v128_t vacc0x01c2 = wasm_f32x4_add( in xnn_f32_gemm_relu_ukernel_4x2c4__wasmsimd() local
136 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 0, 1, 4, 5), in xnn_f32_gemm_relu_ukernel_4x2c4__wasmsimd()
137 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 2, 3, 6, 7)); in xnn_f32_gemm_relu_ukernel_4x2c4__wasmsimd()
|
| D | 4x2c4-minmax-wasmsimd-x86.c | 122 const v128_t vacc0x01c2 = wasm_f32x4_add( in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_x86() local
136 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 0, 1, 4, 5), in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_x86()
137 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 2, 3, 6, 7)); in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_x86()
|
| /external/XNNPACK/src/f32-igemm/gen/ |
| D | 4x2c4-wasmsimd.c | 142 const v128_t vacc0x01c2 = wasm_f32x4_add( in xnn_f32_igemm_ukernel_4x2c4__wasmsimd() local
156 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 0, 1, 4, 5), in xnn_f32_igemm_ukernel_4x2c4__wasmsimd()
157 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 2, 3, 6, 7)); in xnn_f32_igemm_ukernel_4x2c4__wasmsimd()
|
| D | 4x2c4-relu-wasmsimd.c | 142 const v128_t vacc0x01c2 = wasm_f32x4_add( in xnn_f32_igemm_relu_ukernel_4x2c4__wasmsimd() local
156 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 0, 1, 4, 5), in xnn_f32_igemm_relu_ukernel_4x2c4__wasmsimd()
157 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 2, 3, 6, 7)); in xnn_f32_igemm_relu_ukernel_4x2c4__wasmsimd()
|
| D | 4x2c4-minmax-sse.c | 141 …const __m128 vacc0x01c2 = _mm_add_ps(_mm_unpacklo_ps(vacc0x0c4, vacc0x1c4), _mm_unpackhi_ps(vacc0x… in xnn_f32_igemm_minmax_ukernel_4x2c4__sse() local
146 …__m128 vacc01x01 = _mm_add_ps(_mm_movelh_ps(vacc0x01c2, vacc1x01c2), _mm_movehl_ps(vacc1x01c2, vac… in xnn_f32_igemm_minmax_ukernel_4x2c4__sse()
|
| D | 4x2c4-minmax-wasmsimd-x86.c | 142 const v128_t vacc0x01c2 = wasm_f32x4_add( in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_x86() local
156 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 0, 1, 4, 5), in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_x86()
157 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 2, 3, 6, 7)); in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_x86()
|
| D | 4x2c4-minmax-wasmsimd-arm.c | 144 const v128_t vacc0x01c2 = wasm_f32x4_add( in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_arm() local
158 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 0, 1, 4, 5), in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_arm()
159 wasm_v32x4_shuffle(vacc0x01c2, vacc1x01c2, 2, 3, 6, 7)); in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_arm()
|