| /external/XNNPACK/src/f32-gemm/gen/ |
| D | 4x2-minmax-neon-lane-ld64.c | 71 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_gemm_minmax_ukernel_4x2__neon_lane_ld64() local
90 const float32x2_t va3 = vld1_dup_f32(a3); a3 += 1; in xnn_f32_gemm_minmax_ukernel_4x2__neon_lane_ld64() local
|
| D | 4x2-minmax-neonfma-lane-ld64.c | 71 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_gemm_minmax_ukernel_4x2__neonfma_lane_ld64() local
112 const float32x2_t va3 = vld1_dup_f32(a3); a3 += 1; in xnn_f32_gemm_minmax_ukernel_4x2__neonfma_lane_ld64() local
|
| D | 4x8-minmax-neon-lane-ld64.c | 75 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_gemm_minmax_ukernel_4x8__neon_lane_ld64() local
104 const float32x4_t va3 = vld1q_dup_f32(a3); a3 += 1; in xnn_f32_gemm_minmax_ukernel_4x8__neon_lane_ld64() local
|
| D | 4x2c4-wasmsimd.c | 78 const v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_ukernel_4x2c4__wasmsimd() local
101 const v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_ukernel_4x2c4__wasmsimd() local
|
| D | 4x8s4-minmax-wasmsimd-arm.c | 80 v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_minmax_ukernel_4x8s4__wasmsimd_arm() local
159 const v128_t va3 = wasm_v32x4_load_splat(a3); in xnn_f32_gemm_minmax_ukernel_4x8s4__wasmsimd_arm() local
|
| D | 4x8-minmax-neonfma-lane-ld64.c | 75 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_gemm_minmax_ukernel_4x8__neonfma_lane_ld64() local
104 const float32x4_t va3 = vld1q_dup_f32(a3); a3 += 1; in xnn_f32_gemm_minmax_ukernel_4x8__neonfma_lane_ld64() local
|
| D | 4x8s4-minmax-sse.c | 78 __m128 va3 = _mm_loadu_ps(a3); in xnn_f32_gemm_minmax_ukernel_4x8s4__sse() local
157 const __m128 va3 = _mm_load1_ps(a3); in xnn_f32_gemm_minmax_ukernel_4x8s4__sse() local
|
| D | 4x8s4-minmax-wasmsimd-x86.c | 78 v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_minmax_ukernel_4x8s4__wasmsimd_x86() local
157 const v128_t va3 = wasm_v32x4_load_splat(a3); in xnn_f32_gemm_minmax_ukernel_4x8s4__wasmsimd_x86() local
|
| D | 4x2c4-minmax-sse.c | 78 const __m128 va3 = _mm_loadu_ps(a3); in xnn_f32_gemm_minmax_ukernel_4x2c4__sse() local
101 const __m128 va3 = _mm_loadu_ps(a3); in xnn_f32_gemm_minmax_ukernel_4x2c4__sse() local
|
| D | 4x2c4-minmax-wasmsimd-x86.c | 78 const v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_x86() local
101 const v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_x86() local
|
| D | 4x2c4-relu-wasmsimd.c | 78 const v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_relu_ukernel_4x2c4__wasmsimd() local
101 const v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_relu_ukernel_4x2c4__wasmsimd() local
|
| D | 4x2c4-minmax-wasmsimd-arm.c | 80 const v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_arm() local
103 const v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemm_minmax_ukernel_4x2c4__wasmsimd_arm() local
|
| /external/XNNPACK/src/f32-igemm/gen/ |
| D | 4x2-minmax-neon-lane-ld64.c | 93 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_igemm_minmax_ukernel_4x2__neon_lane_ld64() local
112 const float32x2_t va3 = vld1_dup_f32(a3); in xnn_f32_igemm_minmax_ukernel_4x2__neon_lane_ld64() local
|
| D | 4x4-minmax-neon-lane-ld64.c | 93 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_igemm_minmax_ukernel_4x4__neon_lane_ld64() local
113 const float32x4_t va3 = vld1q_dup_f32(a3); in xnn_f32_igemm_minmax_ukernel_4x4__neon_lane_ld64() local
|
| D | 4x4-minmax-neonfma-lane-ld64.c | 93 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_igemm_minmax_ukernel_4x4__neonfma_lane_ld64() local
113 const float32x4_t va3 = vld1q_dup_f32(a3); in xnn_f32_igemm_minmax_ukernel_4x4__neonfma_lane_ld64() local
|
| D | 4x2-minmax-neonfma-lane-ld64.c | 93 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_igemm_minmax_ukernel_4x2__neonfma_lane_ld64() local
134 const float32x2_t va3 = vld1_dup_f32(a3); in xnn_f32_igemm_minmax_ukernel_4x2__neonfma_lane_ld64() local
|
| D | 4x8s4-minmax-wasmsimd-arm.c | 102 v128_t va3 = wasm_v128_load(a3); in xnn_f32_igemm_minmax_ukernel_4x8s4__wasmsimd_arm() local
185 const v128_t va3 = wasm_v32x4_load_splat(a3); in xnn_f32_igemm_minmax_ukernel_4x8s4__wasmsimd_arm() local
|
| D | 4x8-minmax-neon-lane-ld64.c | 97 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_igemm_minmax_ukernel_4x8__neon_lane_ld64() local
127 const float32x4_t va3 = vld1q_dup_f32(a3); in xnn_f32_igemm_minmax_ukernel_4x8__neon_lane_ld64() local
|
| D | 4x8s4-minmax-sse.c | 100 __m128 va3 = _mm_loadu_ps(a3); in xnn_f32_igemm_minmax_ukernel_4x8s4__sse() local
183 const __m128 va3 = _mm_load1_ps(a3); in xnn_f32_igemm_minmax_ukernel_4x8s4__sse() local
|
| D | 4x8s4-minmax-wasmsimd-x86.c | 100 v128_t va3 = wasm_v128_load(a3); in xnn_f32_igemm_minmax_ukernel_4x8s4__wasmsimd_x86() local
183 const v128_t va3 = wasm_v32x4_load_splat(a3); in xnn_f32_igemm_minmax_ukernel_4x8s4__wasmsimd_x86() local
|
| /external/XNNPACK/src/f32-gemm/gen-inc/ |
| D | 4x8s4inc-minmax-wasmsimd-x86.c | 80 v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemminc_minmax_ukernel_4x8s4__wasmsimd_x86() local
159 const v128_t va3 = wasm_v32x4_load_splat(a3); in xnn_f32_gemminc_minmax_ukernel_4x8s4__wasmsimd_x86() local
|
| D | 4x8s4inc-minmax-wasmsimd-arm.c | 82 v128_t va3 = wasm_v128_load(a3); in xnn_f32_gemminc_minmax_ukernel_4x8s4__wasmsimd_arm() local
161 const v128_t va3 = wasm_v32x4_load_splat(a3); in xnn_f32_gemminc_minmax_ukernel_4x8s4__wasmsimd_arm() local
|
| D | 4x8inc-minmax-neonfma-lane-ld64.c | 77 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_gemminc_minmax_ukernel_4x8__neonfma_lane_ld64() local
106 const float32x4_t va3 = vld1q_dup_f32(a3); a3 += 1; in xnn_f32_gemminc_minmax_ukernel_4x8__neonfma_lane_ld64() local
|
| D | 4x8s4inc-minmax-sse.c | 80 __m128 va3 = _mm_loadu_ps(a3); in xnn_f32_gemminc_minmax_ukernel_4x8s4__sse() local
159 const __m128 va3 = _mm_load1_ps(a3); in xnn_f32_gemminc_minmax_ukernel_4x8s4__sse() local
|
| D | 4x8inc-minmax-neon-lane-ld64.c | 77 const float32x2_t va3 = vld1_f32(a3); a3 += 2; in xnn_f32_gemminc_minmax_ukernel_4x8__neon_lane_ld64() local
106 const float32x4_t va3 = vld1q_dup_f32(a3); a3 += 1; in xnn_f32_gemminc_minmax_ukernel_4x8__neon_lane_ld64() local
|