Home
last modified time | relevance | path

Searched refs:vidx89AB (Results 1 – 25 of 95) sorted by relevance

1234

/external/XNNPACK/src/f32-vsigmoid/gen/
Dvsigmoid-sse2-rr2-lut64-p2-div-x12.c58 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x12() local
79 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x12()
80 const uint64_t vidxAB = (uint64_t) _mm_cvtsi128_si64(_mm_unpackhi_epi64(vidx89AB, vidx89AB)); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x12()
111 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x12()
112 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x12()
113 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x12()
114 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x12()
Dvsigmoid-sse41-rr2-lut64-p2-div-x12.c58 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x12() local
75 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x12()
76 const uint64_t vidxAB = (uint64_t) _mm_extract_epi64(vidx89AB, 1); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x12()
101 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x12()
102 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x12()
103 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x12()
104 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x12()
Dvsigmoid-sse2-rr2-lut64-p2-div-x16.c62 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x16() local
84 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x16()
85 const uint64_t vidxAB = (uint64_t) _mm_cvtsi128_si64(_mm_unpackhi_epi64(vidx89AB, vidx89AB)); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x16()
125 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x16()
126 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x16()
127 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x16()
128 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x16()
Dvsigmoid-sse41-rr2-lut64-p2-div-x16.c62 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x16() local
80 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x16()
81 const uint64_t vidxAB = (uint64_t) _mm_extract_epi64(vidx89AB, 1); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x16()
113 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x16()
114 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x16()
115 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x16()
116 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x16()
Dvsigmoid-sse41-rr2-lut64-p2-div-x20.c66 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x20() local
85 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x20()
86 const uint64_t vidxAB = (uint64_t) _mm_extract_epi64(vidx89AB, 1); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x20()
125 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x20()
126 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x20()
127 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x20()
128 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x20()
Dvsigmoid-sse2-rr2-lut64-p2-div-x20.c66 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x20() local
89 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x20()
90 const uint64_t vidxAB = (uint64_t) _mm_cvtsi128_si64(_mm_unpackhi_epi64(vidx89AB, vidx89AB)); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x20()
139 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x20()
140 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x20()
141 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x20()
142 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x20()
Dvsigmoid-sse2-rr2-lut64-p2-div-x24.c70 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x24() local
94 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x24()
95 const uint64_t vidxAB = (uint64_t) _mm_cvtsi128_si64(_mm_unpackhi_epi64(vidx89AB, vidx89AB)); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x24()
153 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x24()
154 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x24()
155 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x24()
156 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_vsigmoid_ukernel__sse2_rr2_lut64_p2_div_x24()
Dvsigmoid-sse41-rr2-lut64-p2-div-x24.c70 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x24() local
90 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x24()
91 const uint64_t vidxAB = (uint64_t) _mm_extract_epi64(vidx89AB, 1); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x24()
137 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x24()
138 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x24()
139 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x24()
140 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_vsigmoid_ukernel__sse41_rr2_lut64_p2_div_x24()
Dvsigmoid-neonfma-rr1-lut64-p2-div-x12.c56 …const uint64x2_t vidx89AB = vreinterpretq_u64_s32(vandq_s32(vreinterpretq_s32_f32(vn89AB), vindex_… in xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut64_p2_div_x12() local
66 const uint64_t vidx89 = vgetq_lane_u64(vidx89AB, 0); in xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut64_p2_div_x12()
67 const uint64_t vidxAB = vgetq_lane_u64(vidx89AB, 1); in xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut64_p2_div_x12()
Dvsigmoid-wasmsimd-rr2-lut64-p2-div-x12.c57 const v128_t vidx89AB = wasm_i32x4_shl(wasm_v128_and(vn89AB, vindex_mask), 2); in xnn_f32_vsigmoid_ukernel__wasmsimd_rr2_lut64_p2_div_x12() local
73 const uint64_t vidx89 = wasm_i64x2_extract_lane(vidx89AB, 0); in xnn_f32_vsigmoid_ukernel__wasmsimd_rr2_lut64_p2_div_x12()
74 const uint64_t vidxAB = wasm_i64x2_extract_lane(vidx89AB, 1); in xnn_f32_vsigmoid_ukernel__wasmsimd_rr2_lut64_p2_div_x12()
Dvsigmoid-neonfma-rr1-lut2048-p1-div-x12.c55 …const uint64x2_t vidx89AB = vreinterpretq_u64_s32(vandq_s32(vreinterpretq_s32_f32(vn89AB), vindex_… in xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut2048_p1_div_x12() local
65 const uint64_t vidx89 = vgetq_lane_u64(vidx89AB, 0); in xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut2048_p1_div_x12()
66 const uint64_t vidxAB = vgetq_lane_u64(vidx89AB, 1); in xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut2048_p1_div_x12()
Dvsigmoid-neonfma-rr1-lut2048-p1-nr2recps-x12.c55 …const uint64x2_t vidx89AB = vreinterpretq_u64_s32(vandq_s32(vreinterpretq_s32_f32(vn89AB), vindex_… in xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut2048_p1_nr2recps_x12() local
65 const uint64_t vidx89 = vgetq_lane_u64(vidx89AB, 0); in xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut2048_p1_nr2recps_x12()
66 const uint64_t vidxAB = vgetq_lane_u64(vidx89AB, 1); in xnn_f32_vsigmoid_ukernel__neonfma_rr1_lut2048_p1_nr2recps_x12()
/external/XNNPACK/src/f32-velu/gen/
Dvelu-sse2-rr2-lut16-p3-x12.c62 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x12() local
84 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x12()
85 const uint64_t vidxAB = (uint64_t) _mm_cvtsi128_si64(_mm_unpackhi_epi64(vidx89AB, vidx89AB)); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x12()
116 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x12()
117 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x12()
118 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x12()
119 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x12()
Dvelu-sse41-rr2-lut16-p3-x12.c62 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x12() local
80 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x12()
81 const uint64_t vidxAB = (uint64_t) _mm_extract_epi64(vidx89AB, 1); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x12()
106 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x12()
107 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x12()
108 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x12()
109 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x12()
Dvelu-sse2-rr2-lut16-p3-x16.c65 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x16() local
89 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x16()
90 const uint64_t vidxAB = (uint64_t) _mm_cvtsi128_si64(_mm_unpackhi_epi64(vidx89AB, vidx89AB)); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x16()
130 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x16()
131 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x16()
132 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x16()
133 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x16()
Dvelu-sse41-rr2-lut16-p3-x16.c65 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x16() local
85 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x16()
86 const uint64_t vidxAB = (uint64_t) _mm_extract_epi64(vidx89AB, 1); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x16()
118 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x16()
119 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x16()
120 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x16()
121 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x16()
Dvelu-sse2-rr2-lut16-p3-x20.c68 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x20() local
94 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x20()
95 const uint64_t vidxAB = (uint64_t) _mm_cvtsi128_si64(_mm_unpackhi_epi64(vidx89AB, vidx89AB)); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x20()
144 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x20()
145 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x20()
146 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x20()
147 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x20()
Dvelu-sse41-rr2-lut16-p3-x20.c68 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x20() local
90 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x20()
91 const uint64_t vidxAB = (uint64_t) _mm_extract_epi64(vidx89AB, 1); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x20()
130 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x20()
131 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x20()
132 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x20()
133 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x20()
Dvelu-sse2-rr2-lut16-p3-x24.c71 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x24() local
99 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x24()
100 const uint64_t vidxAB = (uint64_t) _mm_cvtsi128_si64(_mm_unpackhi_epi64(vidx89AB, vidx89AB)); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x24()
158 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x24()
159 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x24()
160 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x24()
161 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_velu_ukernel__sse2_rr2_lut16_p3_x24()
Dvelu-sse41-rr2-lut16-p3-x24.c71 … const __m128i vidx89AB = _mm_slli_epi32(_mm_and_si128(_mm_castps_si128(vn89AB), vindex_mask), 2); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x24() local
95 const uint64_t vidx89 = (uint64_t) _mm_cvtsi128_si64(vidx89AB); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x24()
96 const uint64_t vidxAB = (uint64_t) _mm_extract_epi64(vidx89AB, 1); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x24()
142 const uint32_t vidx8 = (uint32_t) _mm_cvtsi128_si32(vidx89AB); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x24()
143 const uint32_t vidx9 = (uint32_t) _mm_extract_epi16(vidx89AB, 2); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x24()
144 const uint32_t vidxA = (uint32_t) _mm_extract_epi16(vidx89AB, 4); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x24()
145 const uint32_t vidxB = (uint32_t) _mm_extract_epi16(vidx89AB, 6); in xnn_f32_velu_ukernel__sse41_rr2_lut16_p3_x24()
Dvelu-wasmsimd-arm-rr2-lut16-p3-x12.c62 const v128_t vidx89AB = wasm_i32x4_shl(wasm_v128_and(vn89AB, vindex_mask), 2); in xnn_f32_velu_ukernel__wasmsimd_arm_rr2_lut16_p3_x12() local
79 const uint64_t vidx89 = wasm_i64x2_extract_lane(vidx89AB, 0); in xnn_f32_velu_ukernel__wasmsimd_arm_rr2_lut16_p3_x12()
80 const uint64_t vidxAB = wasm_i64x2_extract_lane(vidx89AB, 1); in xnn_f32_velu_ukernel__wasmsimd_arm_rr2_lut16_p3_x12()
/external/XNNPACK/src/f32-raddstoreexpminusmax/gen/
Dneonfma-rr1-lut64-p2-x12-acc2.c63 …const uint64x2_t vidx89AB = vreinterpretq_u64_s32(vandq_s32(vreinterpretq_s32_f32(vn89AB), vindex_… in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_rr1_lut64_p2_x12_acc2() local
64 const uint64_t vidx89 = vgetq_lane_u64(vidx89AB, 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_rr1_lut64_p2_x12_acc2()
65 const uint64_t vidxAB = vgetq_lane_u64(vidx89AB, 1); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_rr1_lut64_p2_x12_acc2()
Dneonfma-rr1-lut64-p2-x12-acc3.c64 …const uint64x2_t vidx89AB = vreinterpretq_u64_s32(vandq_s32(vreinterpretq_s32_f32(vn89AB), vindex_… in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_rr1_lut64_p2_x12_acc3() local
65 const uint64_t vidx89 = vgetq_lane_u64(vidx89AB, 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_rr1_lut64_p2_x12_acc3()
66 const uint64_t vidxAB = vgetq_lane_u64(vidx89AB, 1); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_rr1_lut64_p2_x12_acc3()
Dneonfma-rr1-lut64-p2-x12.c62 …const uint64x2_t vidx89AB = vreinterpretq_u64_s32(vandq_s32(vreinterpretq_s32_f32(vn89AB), vindex_… in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_rr1_lut64_p2_x12() local
63 const uint64_t vidx89 = vgetq_lane_u64(vidx89AB, 0); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_rr1_lut64_p2_x12()
64 const uint64_t vidxAB = vgetq_lane_u64(vidx89AB, 1); in xnn_f32_raddstoreexpminusmax_ukernel__neonfma_rr1_lut64_p2_x12()
Dneon-rr2-lut64-p2-x12.c63 …const uint64x2_t vidx89AB = vreinterpretq_u64_s32(vandq_s32(vreinterpretq_s32_f32(vn89AB), vindex_… in xnn_f32_raddstoreexpminusmax_ukernel__neon_rr2_lut64_p2_x12() local
64 const uint64_t vidx89 = vgetq_lane_u64(vidx89AB, 0); in xnn_f32_raddstoreexpminusmax_ukernel__neon_rr2_lut64_p2_x12()
65 const uint64_t vidxAB = vgetq_lane_u64(vidx89AB, 1); in xnn_f32_raddstoreexpminusmax_ukernel__neon_rr2_lut64_p2_x12()

1234