| /external/XNNPACK/src/f32-gemm/gen/ |
| D | 1x8-minmax-neonfma-lane-ld64.c | 48 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_gemm_minmax_ukernel_1x8__neonfma_lane_ld64() local
62 const float32x4_t va0 = vld1q_dup_f32(a0); a0 += 1; in xnn_f32_gemm_minmax_ukernel_1x8__neonfma_lane_ld64() local
|
| D | 1x8-minmax-neon-lane-ld64.c | 48 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64() local
62 const float32x4_t va0 = vld1q_dup_f32(a0); a0 += 1; in xnn_f32_gemm_minmax_ukernel_1x8__neon_lane_ld64() local
|
| D | 1x8-minmax-neon-dup-ld64.c | 48 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_gemm_minmax_ukernel_1x8__neon_dup_ld64() local
64 const float32x4_t va0 = vld1q_dup_f32(a0); a0 += 1; in xnn_f32_gemm_minmax_ukernel_1x8__neon_dup_ld64() local
|
| D | 1x8-minmax-neonfma-dup-ld64.c | 48 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_gemm_minmax_ukernel_1x8__neonfma_dup_ld64() local
64 const float32x4_t va0 = vld1q_dup_f32(a0); a0 += 1; in xnn_f32_gemm_minmax_ukernel_1x8__neonfma_dup_ld64() local
|
| D | 1x8-wasmsimd-splat.c | 48 const v128_t va0 = wasm_v128_load(a0); in xnn_f32_gemm_ukernel_1x8__wasmsimd_splat() local
85 const v128_t va0 = wasm_v128_load32_splat(a0); in xnn_f32_gemm_ukernel_1x8__wasmsimd_splat() local
|
| D | 1x8-wasmrelaxedsimd-fma-splat.c | 48 const v128_t va0 = wasm_v128_load(a0); in xnn_f32_gemm_ukernel_1x8__wasmrelaxedsimd_fma_splat() local
85 const v128_t va0 = wasm_v128_load32_splat(a0); in xnn_f32_gemm_ukernel_1x8__wasmrelaxedsimd_fma_splat() local
|
| D | 1x8-relu-wasmsimd-splat.c | 48 const v128_t va0 = wasm_v128_load(a0); in xnn_f32_gemm_relu_ukernel_1x8__wasmsimd_splat() local
85 const v128_t va0 = wasm_v128_load32_splat(a0); in xnn_f32_gemm_relu_ukernel_1x8__wasmsimd_splat() local
|
| D | 1x8-relu-wasmrelaxedsimd-fma-splat.c | 48 const v128_t va0 = wasm_v128_load(a0); in xnn_f32_gemm_relu_ukernel_1x8__wasmrelaxedsimd_fma_splat() local
85 const v128_t va0 = wasm_v128_load32_splat(a0); in xnn_f32_gemm_relu_ukernel_1x8__wasmrelaxedsimd_fma_splat() local
|
| D | 1x8-minmax-wasmrelaxedsimd-fma-splat.c | 50 const v128_t va0 = wasm_v128_load(a0); in xnn_f32_gemm_minmax_ukernel_1x8__wasmrelaxedsimd_fma_splat() local
87 const v128_t va0 = wasm_v128_load32_splat(a0); in xnn_f32_gemm_minmax_ukernel_1x8__wasmrelaxedsimd_fma_splat() local
|
| D | 1x8-minmax-wasmsimd-arm-splat.c | 50 const v128_t va0 = wasm_v128_load(a0); in xnn_f32_gemm_minmax_ukernel_1x8__wasmsimd_arm_splat() local
87 const v128_t va0 = wasm_v128_load32_splat(a0); in xnn_f32_gemm_minmax_ukernel_1x8__wasmsimd_arm_splat() local
|
| /external/XNNPACK/src/f32-gemm/gen-inc/ |
| D | 1x8inc-minmax-neon-lane-ld64.c | 50 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_gemminc_minmax_ukernel_1x8__neon_lane_ld64() local
64 const float32x4_t va0 = vld1q_dup_f32(a0); a0 += 1; in xnn_f32_gemminc_minmax_ukernel_1x8__neon_lane_ld64() local
|
| D | 1x8inc-minmax-neonfma-lane-ld64.c | 50 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_gemminc_minmax_ukernel_1x8__neonfma_lane_ld64() local
64 const float32x4_t va0 = vld1q_dup_f32(a0); a0 += 1; in xnn_f32_gemminc_minmax_ukernel_1x8__neonfma_lane_ld64() local
|
| D | 1x8inc-minmax-neon-dup-ld64.c | 50 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_gemminc_minmax_ukernel_1x8__neon_dup_ld64() local
66 const float32x4_t va0 = vld1q_dup_f32(a0); a0 += 1; in xnn_f32_gemminc_minmax_ukernel_1x8__neon_dup_ld64() local
|
| D | 1x8inc-minmax-neonfma-dup-ld64.c | 50 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_gemminc_minmax_ukernel_1x8__neonfma_dup_ld64() local
66 const float32x4_t va0 = vld1q_dup_f32(a0); a0 += 1; in xnn_f32_gemminc_minmax_ukernel_1x8__neonfma_dup_ld64() local
|
| /external/XNNPACK/src/f16-gemm/gen/ |
| D | 1x8-minmax-neonfp16arith-ld64.c | 49 const float16x4_t va0 = vld1_f16(a0); a0 += 4; in xnn_f16_gemm_minmax_ukernel_1x8__neonfp16arith_ld64() local
92 const float16x8_t va0 = vld1q_dup_f16(a0); a0 += 1; in xnn_f16_gemm_minmax_ukernel_1x8__neonfp16arith_ld64() local
|
| /external/XNNPACK/src/f32-igemm/gen/ |
| D | 1x8-minmax-neon-lane-ld64.c | 61 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64() local
76 const float32x4_t va0 = vld1q_dup_f32(a0); in xnn_f32_igemm_minmax_ukernel_1x8__neon_lane_ld64() local
|
| D | 1x8-minmax-neonfma-lane-ld64.c | 61 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_igemm_minmax_ukernel_1x8__neonfma_lane_ld64() local
76 const float32x4_t va0 = vld1q_dup_f32(a0); in xnn_f32_igemm_minmax_ukernel_1x8__neonfma_lane_ld64() local
|
| D | 1x8-minmax-neonfma-dup-ld64.c | 61 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_igemm_minmax_ukernel_1x8__neonfma_dup_ld64() local
78 const float32x4_t va0 = vld1q_dup_f32(a0); in xnn_f32_igemm_minmax_ukernel_1x8__neonfma_dup_ld64() local
|
| D | 1x8-minmax-neon-dup-ld64.c | 61 const float32x2_t va0 = vld1_f32(a0); a0 += 2; in xnn_f32_igemm_minmax_ukernel_1x8__neon_dup_ld64() local
78 const float32x4_t va0 = vld1q_dup_f32(a0); in xnn_f32_igemm_minmax_ukernel_1x8__neon_dup_ld64() local
|
| /external/XNNPACK/src/f16-gemm/gen-inc/ |
| D | 1x8inc-minmax-neonfp16arith-ld64.c | 51 const float16x4_t va0 = vld1_f16(a0); a0 += 4; in xnn_f16_gemminc_minmax_ukernel_1x8__neonfp16arith_ld64() local
94 const float16x8_t va0 = vld1q_dup_f16(a0); a0 += 1; in xnn_f16_gemminc_minmax_ukernel_1x8__neonfp16arith_ld64() local
|
| /external/XNNPACK/src/f16-igemm/gen/ |
| D | 1x8-minmax-neonfp16arith-ld64.c | 60 const float16x4_t va0 = vld1_f16(a0); a0 += 4; in xnn_f16_igemm_minmax_ukernel_1x8__neonfp16arith_ld64() local
101 const float16x8_t va0 = vld1q_dup_f16(a0); a0 += 1; in xnn_f16_igemm_minmax_ukernel_1x8__neonfp16arith_ld64() local
|
| /external/XNNPACK/src/qs8-gemm/gen/ |
| D | 1x4c2-xw-minmax-fp32-xop.c | 56 const __m128i va0 = _mm_loadl_epi64((const __m128i*) a0); in xnn_qs8_gemm_xw_minmax_fp32_ukernel_1x4c2__xop() local
81 const __m128i va0 = _mm_loadl_epi64((const __m128i*) a0); in xnn_qs8_gemm_xw_minmax_fp32_ukernel_1x4c2__xop() local
|
| D | 1x4c2-xw-minmax-fp32-avx.c | 51 const __m128i va0 = _mm_loadl_epi64((const __m128i*) a0); in xnn_qs8_gemm_xw_minmax_fp32_ukernel_1x4c2__avx() local
76 const __m128i va0 = _mm_loadl_epi64((const __m128i*) a0); in xnn_qs8_gemm_xw_minmax_fp32_ukernel_1x4c2__avx() local
|
| D | 1x4c2-xw-minmax-fp32-sse41.c | 51 const __m128i va0 = _mm_loadl_epi64((const __m128i*) a0); in xnn_qs8_gemm_xw_minmax_fp32_ukernel_1x4c2__sse41() local
76 const __m128i va0 = _mm_loadl_epi64((const __m128i*) a0); in xnn_qs8_gemm_xw_minmax_fp32_ukernel_1x4c2__sse41() local
|
| D | 1x4c2-xw-minmax-fp32-sse2.c | 51 const __m128i va0 = _mm_loadl_epi64((const __m128i*) a0); in xnn_qs8_gemm_xw_minmax_fp32_ukernel_1x4c2__sse2() local
76 const __m128i va0 = _mm_loadl_epi64((const __m128i*) a0); in xnn_qs8_gemm_xw_minmax_fp32_ukernel_1x4c2__sse2() local
|