| /external/XNNPACK/src/f32-gemm/gen/ |
| D | 4x2c4-wasmsimd.c | 128 const v128_t vacc2x01c2 = wasm_f32x4_add( in xnn_f32_gemm_ukernel_4x2c4__wasmsimd() local
139 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 0, 1, 4, 5), in xnn_f32_gemm_ukernel_4x2c4__wasmsimd()
140 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 2, 3, 6, 7)); in xnn_f32_gemm_ukernel_4x2c4__wasmsimd()
|
| D | 4x2c4-minmax-wasmsimd-arm.c | 130 const v128_t vacc2x01c2 = wasm_f32x4_add( in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_arm() local
141 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 0, 1, 4, 5), in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_arm()
142 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 2, 3, 6, 7)); in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_arm()
|
| D | 4x2c4-minmax-sse.c | 123 …const __m128 vacc2x01c2 = _mm_add_ps(_mm_unpacklo_ps(vacc2x0c4, vacc2x1c4), _mm_unpackhi_ps(vacc2x… in xnn_f32_gemm_minmax_ukernel_4x2c4__sse() local
127 …__m128 vacc23x01 = _mm_add_ps(_mm_movelh_ps(vacc2x01c2, vacc3x01c2), _mm_movehl_ps(vacc3x01c2, vac… in xnn_f32_gemm_minmax_ukernel_4x2c4__sse()
|
| D | 4x2c4-relu-wasmsimd.c | 128 const v128_t vacc2x01c2 = wasm_f32x4_add( in xnn_f32_gemm_relu_ukernel_4x2c4__wasmsimd() local
139 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 0, 1, 4, 5), in xnn_f32_gemm_relu_ukernel_4x2c4__wasmsimd()
140 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 2, 3, 6, 7)); in xnn_f32_gemm_relu_ukernel_4x2c4__wasmsimd()
|
| D | 4x2c4-minmax-wasmsimd-x86.c | 128 const v128_t vacc2x01c2 = wasm_f32x4_add( in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_x86() local
139 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 0, 1, 4, 5), in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_x86()
140 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 2, 3, 6, 7)); in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_x86()
|
| /external/XNNPACK/src/f32-igemm/gen/ |
| D | 4x2c4-wasmsimd.c | 148 const v128_t vacc2x01c2 = wasm_f32x4_add( in xnn_f32_igemm_ukernel_4x2c4__wasmsimd() local
159 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 0, 1, 4, 5), in xnn_f32_igemm_ukernel_4x2c4__wasmsimd()
160 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 2, 3, 6, 7)); in xnn_f32_igemm_ukernel_4x2c4__wasmsimd()
|
| D | 4x2c4-relu-wasmsimd.c | 148 const v128_t vacc2x01c2 = wasm_f32x4_add( in xnn_f32_igemm_relu_ukernel_4x2c4__wasmsimd() local
159 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 0, 1, 4, 5), in xnn_f32_igemm_relu_ukernel_4x2c4__wasmsimd()
160 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 2, 3, 6, 7)); in xnn_f32_igemm_relu_ukernel_4x2c4__wasmsimd()
|
| D | 4x2c4-minmax-sse.c | 143 …const __m128 vacc2x01c2 = _mm_add_ps(_mm_unpacklo_ps(vacc2x0c4, vacc2x1c4), _mm_unpackhi_ps(vacc2x… in xnn_f32_igemm_minmax_ukernel_4x2c4__sse() local
147 …__m128 vacc23x01 = _mm_add_ps(_mm_movelh_ps(vacc2x01c2, vacc3x01c2), _mm_movehl_ps(vacc3x01c2, vac… in xnn_f32_igemm_minmax_ukernel_4x2c4__sse()
|
| D | 4x2c4-minmax-wasmsimd-x86.c | 148 const v128_t vacc2x01c2 = wasm_f32x4_add( in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_x86() local
159 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 0, 1, 4, 5), in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_x86()
160 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 2, 3, 6, 7)); in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_x86()
|
| D | 4x2c4-minmax-wasmsimd-arm.c | 150 const v128_t vacc2x01c2 = wasm_f32x4_add( in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_arm() local
161 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 0, 1, 4, 5), in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_arm()
162 wasm_v32x4_shuffle(vacc2x01c2, vacc3x01c2, 2, 3, 6, 7)); in xnn_f32_igemm_minmax_ukernel_4x2c4__wasmsimd_arm()
|