/aosp_15_r20/external/XNNPACK/src/qu8-avgpool/ |
H A D | 9p8x-minmax-scalar-c1.c | 102 const uint32_t vsum23 = vi2 + vi3; in xnn_qu8_avgpool_minmax_ukernel_9p8x__scalar_c1() local 173 const uint32_t vsum23 = vi2 + vi3; in xnn_qu8_avgpool_minmax_ukernel_9p8x__scalar_c1() local 265 const uint32_t vsum23 = vi2 + vi3; in xnn_qu8_avgpool_minmax_ukernel_9p8x__scalar_c1() local
|
H A D | 9x-minmax-neon-c8.c | 130 const uint16x8_t vsum23 = vaddl_u8(vi2, vi3); in xnn_qu8_avgpool_minmax_ukernel_9x__neon_c8() local 203 const uint16x8_t vsum23 = vaddl_u8(vi2, vi3); in xnn_qu8_avgpool_minmax_ukernel_9x__neon_c8() local
|
H A D | 9x-minmax-sse2-c8.c | 135 const __m128i vsum23 = _mm_add_epi16(vxi2, vxi3); in xnn_qu8_avgpool_minmax_ukernel_9x__sse2_c8() local 210 const __m128i vsum23 = _mm_add_epi16(vxi2, vxi3); in xnn_qu8_avgpool_minmax_ukernel_9x__sse2_c8() local
|
/aosp_15_r20/external/XNNPACK/src/f32-avgpool/ |
H A D | 9p8x-minmax-scalar-c1.c | 95 const float vsum23 = vi2 + vi3; in xnn_f32_avgpool_minmax_ukernel_9p8x__scalar_c1() local 164 const float vsum23 = vi2 + vi3; in xnn_f32_avgpool_minmax_ukernel_9p8x__scalar_c1() local 254 const float vsum23 = vi2 + vi3; in xnn_f32_avgpool_minmax_ukernel_9p8x__scalar_c1() local
|
H A D | 9p8x-minmax-wasm-c1.c | 95 const float vsum23 = vi2 + vi3; in xnn_f32_avgpool_minmax_ukernel_9p8x__wasm_c1() local 164 const float vsum23 = vi2 + vi3; in xnn_f32_avgpool_minmax_ukernel_9p8x__wasm_c1() local 254 const float vsum23 = vi2 + vi3; in xnn_f32_avgpool_minmax_ukernel_9p8x__wasm_c1() local
|
H A D | 9x-minmax-sse-c4.c | 128 const __m128 vsum23 = _mm_add_ps(vi2, vi3); in xnn_f32_avgpool_minmax_ukernel_9x__sse_c4() local 156 const __m128 vsum23 = _mm_add_ps(vi2, vi3); in xnn_f32_avgpool_minmax_ukernel_9x__sse_c4() local
|
H A D | 9x-minmax-wasmsimd-arm-c4.c | 128 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_avgpool_minmax_ukernel_9x__wasmsimd_arm_c4() local 157 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_avgpool_minmax_ukernel_9x__wasmsimd_arm_c4() local
|
H A D | 9x-minmax-wasmsimd-x86-c4.c | 128 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_avgpool_minmax_ukernel_9x__wasmsimd_x86_c4() local 157 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_avgpool_minmax_ukernel_9x__wasmsimd_x86_c4() local
|
H A D | 9x-minmax-neon-c4.c | 119 const float32x4_t vsum23 = vaddq_f32(vi2, vi3); in xnn_f32_avgpool_minmax_ukernel_9x__neon_c4() local 147 const float32x4_t vsum23 = vaddq_f32(vi2, vi3); in xnn_f32_avgpool_minmax_ukernel_9x__neon_c4() local
|
/aosp_15_r20/external/XNNPACK/src/f32-gavgpool-cw/ |
H A D | neon-x4.c | 75 const float32x4_t vsum23 = vpaddq_f32(vsum2, vsum3); in xnn_f32_gavgpool_cw_ukernel__neon_x4() local 80 const float32x4_t vsum23 = vcombine_f32(vadd_f32(vget_low_f32(vsum2), vget_high_f32(vsum2)), in xnn_f32_gavgpool_cw_ukernel__neon_x4() local
|
/aosp_15_r20/external/XNNPACK/src/f32-gavgpool/ |
H A D | 7x-minmax-wasmsimd-arm-c4.c | 72 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_gavgpool_minmax_ukernel_7x__wasmsimd_arm_c4() local 99 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_gavgpool_minmax_ukernel_7x__wasmsimd_arm_c4() local
|
H A D | 7x-minmax-wasmsimd-x86-c4.c | 72 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_gavgpool_minmax_ukernel_7x__wasmsimd_x86_c4() local 99 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_gavgpool_minmax_ukernel_7x__wasmsimd_x86_c4() local
|
H A D | 7x-minmax-neon-c4.c | 65 const float32x4_t vsum23 = vaddq_f32(vi2, vi3); in xnn_f32_gavgpool_minmax_ukernel_7x__neon_c4() local 91 const float32x4_t vsum23 = vaddq_f32(vi2, vi3); in xnn_f32_gavgpool_minmax_ukernel_7x__neon_c4() local
|
H A D | 7x-minmax-sse-c4.c | 72 const __m128 vsum23 = _mm_add_ps(vi2, vi3); in xnn_f32_gavgpool_minmax_ukernel_7x__sse_c4() local 99 const __m128 vsum23 = _mm_add_ps(vi2, vi3); in xnn_f32_gavgpool_minmax_ukernel_7x__sse_c4() local
|
/aosp_15_r20/external/XNNPACK/src/f16-avgpool/ |
H A D | 9x-minmax-f16c-c8.c | 130 …const __m256 vsum23 = _mm256_cvtph_ps(_mm256_cvtps_ph(_mm256_add_ps(vi2, vi3), _MM_FROUND_NO_EXC)); in xnn_f16_avgpool_minmax_ukernel_9x__f16c_c8() local 159 …const __m256 vsum23 = _mm256_cvtph_ps(_mm256_cvtps_ph(_mm256_add_ps(vi2, vi3), _MM_FROUND_NO_EXC)); in xnn_f16_avgpool_minmax_ukernel_9x__f16c_c8() local
|
H A D | 9x-minmax-neonfp16arith-c8.c | 119 const float16x8_t vsum23 = vaddq_f16(vi2, vi3); in xnn_f16_avgpool_minmax_ukernel_9x__neonfp16arith_c8() local 147 const float16x8_t vsum23 = vaddq_f16(vi2, vi3); in xnn_f16_avgpool_minmax_ukernel_9x__neonfp16arith_c8() local
|
/aosp_15_r20/external/XNNPACK/src/f32-pavgpool/ |
H A D | 9x-minmax-wasmsimd-x86-c4.c | 131 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_pavgpool_minmax_ukernel_9x__wasmsimd_x86_c4() local 160 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_pavgpool_minmax_ukernel_9x__wasmsimd_x86_c4() local
|
H A D | 9x-minmax-neon-c4.c | 121 const float32x4_t vsum23 = vaddq_f32(vi2, vi3); in xnn_f32_pavgpool_minmax_ukernel_9x__neon_c4() local 149 const float32x4_t vsum23 = vaddq_f32(vi2, vi3); in xnn_f32_pavgpool_minmax_ukernel_9x__neon_c4() local
|
H A D | 9x-minmax-wasmsimd-arm-c4.c | 131 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_pavgpool_minmax_ukernel_9x__wasmsimd_arm_c4() local 160 const v128_t vsum23 = wasm_f32x4_add(vi2, vi3); in xnn_f32_pavgpool_minmax_ukernel_9x__wasmsimd_arm_c4() local
|
H A D | 9x-minmax-sse-c4.c | 131 const __m128 vsum23 = _mm_add_ps(vi2, vi3); in xnn_f32_pavgpool_minmax_ukernel_9x__sse_c4() local 159 const __m128 vsum23 = _mm_add_ps(vi2, vi3); in xnn_f32_pavgpool_minmax_ukernel_9x__sse_c4() local
|
/aosp_15_r20/external/XNNPACK/src/f16-pavgpool/ |
H A D | 9x-minmax-avx2-c8.c | 133 …const __m256 vsum23 = _mm256_cvtph_ps(_mm256_cvtps_ph(_mm256_add_ps(vi2, vi3), _MM_FROUND_NO_EXC)); in xnn_f16_pavgpool_minmax_ukernel_9x__avx2_c8() local 162 …const __m256 vsum23 = _mm256_cvtph_ps(_mm256_cvtps_ph(_mm256_add_ps(vi2, vi3), _MM_FROUND_NO_EXC)); in xnn_f16_pavgpool_minmax_ukernel_9x__avx2_c8() local
|
H A D | 9x-minmax-neonfp16arith-c8.c | 121 const float16x8_t vsum23 = vaddq_f16(vi2, vi3); in xnn_f16_pavgpool_minmax_ukernel_9x__neonfp16arith_c8() local 149 const float16x8_t vsum23 = vaddq_f16(vi2, vi3); in xnn_f16_pavgpool_minmax_ukernel_9x__neonfp16arith_c8() local
|
/aosp_15_r20/external/pytorch/aten/src/ATen/native/quantized/cpu/qnnpack/src/q8gavgpool/ |
H A D | up8x7-neon.c | 87 const int16x8_t vsum23 = vreinterpretq_s16_u16(vaddl_u8(vi2, vi3)); in pytorch_q8gavgpool_ukernel_up8x7__neon() local 157 const int16x8_t vsum23 = vreinterpretq_s16_u16(vaddl_u8(vi2, vi3)); in pytorch_q8gavgpool_ukernel_up8x7__neon() local
|
/aosp_15_r20/external/pytorch/aten/src/ATen/native/quantized/cpu/qnnpack/src/q8avgpool/ |
H A D | up8x9-neon.c | 105 const uint16x8_t vsum23 = vaddl_u8(vi2, vi3); in pytorch_q8avgpool_ukernel_up8x9__neon() local 183 const uint16x8_t vsum23 = vaddl_u8(vi2, vi3); in pytorch_q8avgpool_ukernel_up8x9__neon() local
|
H A D | up8x9-sse2.c | 103 const __m128i vsum23 = _mm_add_epi16(vxi2, vxi3); in pytorch_q8avgpool_ukernel_up8x9__sse2() local 185 const __m128i vsum23 = _mm_add_epi16(vxi2, vxi3); in pytorch_q8avgpool_ukernel_up8x9__sse2() local
|