Lines Matching refs:lane
415 //Vector add: vadd -> Vr[i]:=Va[i]+Vb[i], Vr, Va, Vb have equal lane sizes.
434 //Vector long add: vaddl -> Vr[i]:=Va[i]+Vb[i], Va, Vb have equal lane sizes, result is a 128 bit vector of lanes that are twice the width.
523 //multiply lane
1185 //Loads of a single vector or lane. Perform loads and stores of a single vector of some type.
1211 //Load a single lane from memory
1212 uint8x16_t vld1q_lane_u8(__transfersize(1) uint8_t const * ptr, uint8x16_t vec, __constrange(0,15) int lane); //VLD1.8 {d0[0]}, [r0]
1213 uint16x8_t vld1q_lane_u16(__transfersize(1) uint16_t const * ptr, uint16x8_t vec, __constrange(0,7) int lane); // VLD1.16 {d0[0]}, [r0]
1214 uint32x4_t vld1q_lane_u32(__transfersize(1) uint32_t const * ptr, uint32x4_t vec, __constrange(0,3) int lane); // VLD1.32 {d0[0]}, [r0]
1215 uint64x2_t vld1q_lane_u64(__transfersize(1) uint64_t const * ptr, uint64x2_t vec, __constrange(0,1) int lane); // VLD1.64 {d0}, [r0]
1216 int8x16_t vld1q_lane_s8(__transfersize(1) int8_t const * ptr, int8x16_t vec, __constrange(0,15) int lane); //VLD1.8 {d0[0]}, [r0]
1217 int16x8_t vld1q_lane_s16(__transfersize(1) int16_t const * ptr, int16x8_t vec, __constrange(0,7) int lane); //VLD1.16 {d0[0]}, [r0]
1218 int32x4_t vld1q_lane_s32(__transfersize(1) int32_t const * ptr, int32x4_t vec, __constrange(0,3) int lane); //VLD1.32 {d0[0]}, [r0]
1219 float16x8_t vld1q_lane_f16(__transfersize(1) __fp16 const * ptr, float16x8_t vec, __constrange(0,7) int lane); //VLD1.16 {d0[0]}, [r0]
1220 float32x4_t vld1q_lane_f32(__transfersize(1) float32_t const * ptr, float32x4_t vec, __constrange(0,3) int lane); // VLD1.32 {d0[0]}, [r0]
1221 int64x2_t vld1q_lane_s64(__transfersize(1) int64_t const * ptr, int64x2_t vec, __constrange(0,1) int lane); //VLD1.64 {d0}, [r0]
1222 poly8x16_t vld1q_lane_p8(__transfersize(1) poly8_t const * ptr, poly8x16_t vec, __constrange(0,15) int lane); //VLD1.8 {d0[0]}, [r0]
1223 poly16x8_t vld1q_lane_p16(__transfersize(1) poly16_t const * ptr, poly16x8_t vec, __constrange(0,7) int lane); // VLD1.16 {d0[0]}, [r0]
1224 uint8x8_t vld1_lane_u8(__transfersize(1) uint8_t const * ptr, uint8x8_t vec, __constrange(0,7) int lane); //VLD1.8 {d0[0]}, [r0]
1225 uint16x4_t vld1_lane_u16(__transfersize(1) uint16_t const * ptr, uint16x4_t vec, __constrange(0,3) int lane); //VLD1.16 {d0[0]}, [r0]
1226 uint32x2_t vld1_lane_u32(__transfersize(1) uint32_t const * ptr, uint32x2_t vec, __constrange(0,1) int lane); //VLD1.32 {d0[0]}, [r0]
1227 uint64x1_t vld1_lane_u64(__transfersize(1) uint64_t const * ptr, uint64x1_t vec, __constrange(0,0) int lane); //VLD1.64 {d0}, [r0]
1228 int8x8_t vld1_lane_s8(__transfersize(1) int8_t const * ptr, int8x8_t vec, __constrange(0,7) int lane); // VLD1.8{d0[0]}, [r0]
1229 int16x4_t vld1_lane_s16(__transfersize(1) int16_t const * ptr, int16x4_t vec, __constrange(0,3) int lane); //VLD1.16 {d0[0]}, [r0]
1230 int32x2_t vld1_lane_s32(__transfersize(1) int32_t const * ptr, int32x2_t vec, __constrange(0,1) int lane); //VLD1.32 {d0[0]}, [r0]
1231 float16x4_t vld1q_lane_f16(__transfersize(1) __fp16 const * ptr, float16x4_t vec, __constrange(0,3) int lane); //VLD1.16 {d0[0]}, [r0]
1232 float32x2_t vld1_lane_f32(__transfersize(1) float32_t const * ptr, float32x2_t vec, __constrange(0,1) int lane); // VLD1.32 {d0[0]}, [r0]
1233 int64x1_t vld1_lane_s64(__transfersize(1) int64_t const * ptr, int64x1_t vec, __constrange(0,0) int lane); //VLD1.64 {d0}, [r0]
1234 poly8x8_t vld1_lane_p8(__transfersize(1) poly8_t const * ptr, poly8x8_t vec, __constrange(0,7) int lane); //VLD1.8 {d0[0]}, [r0]
1235 poly16x4_t vld1_lane_p16(__transfersize(1) poly16_t const * ptr, poly16x4_t vec, __constrange(0,3) int lane); //VLD1.16 {d0[0]}, [r0]
1261 //Store a single vector or lane. Stores all lanes or a single lane of a vector.
1287 //Store a lane of a vector into memory
1393 //Load a single lane of N-element structure from memory
1395 uint16x8x2_t vld2q_lane_u16_ptr(__transfersize(2) uint16_t const * ptr, uint16x8x2_t * src, __constrange(0,7) int lane); // VLD2.16 {d0[0], d2[0]}, [r0]
1396 uint32x4x2_t vld2q_lane_u32_ptr(__transfersize(2) uint32_t const * ptr, uint32x4x2_t * src, __constrange(0,3) int lane); // VLD2.32 {d0[0], d2[0]}, [r0]
1397 int16x8x2_t vld2q_lane_s16_ptr(__transfersize(2) int16_t const * ptr, int16x8x2_t * src, __constrange(0,7) int lane); // VLD2.16 {d0[0], d2[0]}, [r0]
1398 int32x4x2_t vld2q_lane_s32_ptr(__transfersize(2) int32_t const * ptr, int32x4x2_t * src, __constrange(0,3) int lane); // VLD2.32 {d0[0], d2[0]}, [r0]
1399 float16x8x2_t vld2q_lane_f16_ptr(__transfersize(2) __fp16 const * ptr, float16x8x2_t * src, __constrange(0,7) int lane); // VLD2.16 {d0[0], d2[0]}, [r0]
1400 float32x4x2_t vld2q_lane_f32_ptr(__transfersize(2) float32_t const * ptr, float32x4x2_t * src, __constrange(0,3) int lane); // VLD2.32 {d0[0], d2[0]}, [r0]
1401 poly16x8x2_t vld2q_lane_p16_ptr(__transfersize(2) poly16_t const * ptr, poly16x8x2_t * src, __constrange(0,7) int lane); // VLD2.16 {d0[0], d2[0]}, [r0]
1402 uint8x8x2_t vld2_lane_u8_ptr(__transfersize(2) uint8_t const * ptr, uint8x8x2_t * src, __constrange(0,7) int lane); //VLD2.8 {d0[0], d1[0]}, [r0]
1403 uint16x4x2_t vld2_lane_u16_ptr(__transfersize(2) uint16_t const * ptr, uint16x4x2_t * src, __constrange(0,3) int lane); // VLD2.16 {d0[0], d1[0]}, [r0]
1404 uint32x2x2_t vld2_lane_u32_ptr(__transfersize(2) uint32_t const * ptr, uint32x2x2_t * src, __constrange(0,1) int lane); // VLD2.32 {d0[0], d1[0]}, [r0]
1405 int8x8x2_t vld2_lane_s8_ptr(__transfersize(2) int8_t const * ptr, int8x8x2_t * src, __constrange(0,7) int lane); //VLD2.8 {d0[0], d1[0]}, [r0]
1406 int16x4x2_t vld2_lane_s16_ptr(__transfersize(2) int16_t const * ptr, int16x4x2_t * src, __constrange(0,3) int lane); //VLD2.16 {d0[0], d1[0]}, [r0]
1407 int32x2x2_t vld2_lane_s32_ptr(__transfersize(2) int32_t const * ptr, int32x2x2_t * src, __constrange(0,1) int lane); //VLD2.32 {d0[0], d1[0]}, [r0]
1408 //float16x4x2_t vld2_lane_f16_ptr(__transfersize(2) __fp16 const * ptr, float16x4x2_t * src, __constrange(0,3) int lane); // VLD2.16 {d0[0], d1[0]}, [r0]
1409 float32x2x2_t vld2_lane_f32_ptr(__transfersize(2) float32_t const * ptr, float32x2x2_t * src, __constrange(0,1) int lane); // VLD2.32 {d0[0], d1[0]}, [r0]
1410 poly8x8x2_t vld2_lane_p8_ptr(__transfersize(2) poly8_t const * ptr, poly8x8x2_t * src, __constrange(0,7) int lane); //VLD2.8 {d0[0], d1[0]}, [r0]
1411 poly16x4x2_t vld2_lane_p16_ptr(__transfersize(2) poly16_t const * ptr, poly16x4x2_t * src, __constrange(0,3) int lane); // VLD2.16 {d0[0], d1[0]}, [r0]
1412 uint16x8x3_t vld3q_lane_u16_ptr(__transfersize(3) uint16_t const * ptr, uint16x8x3_t * src, __constrange(0,7) int lane); // VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
1413 uint32x4x3_t vld3q_lane_u32_ptr(__transfersize(3) uint32_t const * ptr, uint32x4x3_t * src, __constrange(0,3) int lane); // VLD3.32 {d0[0], d2[0], d4[0]}, [r0]
1414 int16x8x3_t vld3q_lane_s16_ptr(__transfersize(3) int16_t const * ptr, int16x8x3_t * src, __constrange(0,7) int lane); // VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
1415 int32x4x3_t vld3q_lane_s32_ptr(__transfersize(3) int32_t const * ptr, int32x4x3_t * src, __constrange(0,3) int lane); // VLD3.32 {d0[0], d2[0], d4[0]}, [r0]
1416 float16x8x3_t vld3q_lane_f16_ptr(__transfersize(3) __fp16 const * ptr, float16x8x3_t * src, __constrange(0,7) int lane); // VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
1417 float32x4x3_t vld3q_lane_f32_ptr(__transfersize(3) float32_t const * ptr, float32x4x3_t * src, __constrange(0,3) int lane); // VLD3.32 {d0[0], d2[0], d4[0]}, [r0]
1418 poly16x8x3_t vld3q_lane_p16_ptr(__transfersize(3) poly16_t const * ptr, poly16x8x3_t * src, __constrange(0,7) int lane); // VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
1419 uint8x8x3_t vld3_lane_u8_ptr(__transfersize(3) uint8_t const * ptr, uint8x8x3_t * src, __constrange(0,7) int lane); //VLD3.8 {d0[0], d1[0], d2[0]}, [r0]
1420 uint16x4x3_t vld3_lane_u16_ptr(__transfersize(3) uint16_t const * ptr, uint16x4x3_t * src, __constrange(0,3) int lane); // VLD3.16 {d0[0], d1[0], d2[0]}, [r0]
1421 uint32x2x3_t vld3_lane_u32_ptr(__transfersize(3) uint32_t const * ptr, uint32x2x3_t * src, __constrange(0,1) int lane); // VLD3.32 {d0[0], d1[0], d2[0]}, [r0]
1422 int8x8x3_t vld3_lane_s8_ptr(__transfersize(3) int8_t const * ptr, int8x8x3_t * src, __constrange(0,7) int lane); //VLD3.8 {d0[0], d1[0], d2[0]}, [r0]
1423 int16x4x3_t vld3_lane_s16_ptr(__transfersize(3) int16_t const * ptr, int16x4x3_t * src, __constrange(0,3) int lane); //VLD3.16 {d0[0], d1[0], d2[0]}, [r0]
1424 int32x2x3_t vld3_lane_s32_ptr(__transfersize(3) int32_t const * ptr, int32x2x3_t * src, __constrange(0,1) int lane); //VLD3.32 {d0[0], d1[0], d2[0]}, [r0]
1425 float16x4x3_t vld3_lane_f16_ptr(__transfersize(3) __fp16 const * ptr, float16x4x3_t * src, __constrange(0,3) int lane); // VLD3.16 {d0[0], d1[0], d2[0]}, [r0]
1426 float32x2x3_t vld3_lane_f32_ptr(__transfersize(3) float32_t const * ptr, float32x2x3_t * src, __constrange(0,1) int lane); // VLD3.32 {d0[0], d1[0], d2[0]}, [r0]
1427 poly8x8x3_t vld3_lane_p8_ptr(__transfersize(3) poly8_t const * ptr, poly8x8x3_t * src, __constrange(0,7) int lane); //VLD3.8 {d0[0], d1[0], d2[0]}, [r0]
1428 poly16x4x3_t vld3_lane_p16_ptr(__transfersize(3) poly16_t const * ptr, poly16x4x3_t * src, __constrange(0,3) int lane); // VLD3.16 {d0[0], d1[0], d2[0]}, [r0]
1429 uint16x8x4_t vld4q_lane_u16_ptr(__transfersize(4) uint16_t const * ptr, uint16x8x4_t * src, __constrange(0,7) int lane); // VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
1430 uint32x4x4_t vld4q_lane_u32_ptr(__transfersize(4) uint32_t const * ptr, uint32x4x4_t * src, __constrange(0,3) int lane); // VLD4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
1431 int16x8x4_t vld4q_lane_s16_ptr(__transfersize(4) int16_t const * ptr, int16x8x4_t * src, __constrange(0,7) int lane); // VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
1432 int32x4x4_t vld4q_lane_s32_ptr(__transfersize(4) int32_t const * ptr, int32x4x4_t * src, __constrange(0,3) int lane); // VLD4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
1433 float16x8x4_t vld4q_lane_f16_ptr(__transfersize(4) __fp16 const * ptr, float16x8x4_t * src, __constrange(0,7) int lane); // VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
1434 float32x4x4_t vld4q_lane_f32_ptr(__transfersize(4) float32_t const * ptr, float32x4x4_t * src, __constrange(0,3) int lane); // VLD4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
1435 poly16x8x4_t vld4q_lane_p16_ptr(__transfersize(4) poly16_t const * ptr, poly16x8x4_t * src, __constrange(0,7) int lane); // VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
1436 uint8x8x4_t vld4_lane_u8_ptr(__transfersize(4) uint8_t const * ptr, uint8x8x4_t * src, __constrange(0,7) int lane); //VLD4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1437 uint16x4x4_t vld4_lane_u16_ptr(__transfersize(4) uint16_t const * ptr, uint16x4x4_t * src, __constrange(0,3) int lane); // VLD4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1438 uint32x2x4_t vld4_lane_u32_ptr(__transfersize(4) uint32_t const * ptr, uint32x2x4_t * src, __constrange(0,1) int lane); // VLD4.32 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1439 int8x8x4_t vld4_lane_s8_ptr(__transfersize(4) int8_t const * ptr, int8x8x4_t * src, __constrange(0,7) int lane); //VLD4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1440 int16x4x4_t vld4_lane_s16_ptr(__transfersize(4) int16_t const * ptr, int16x4x4_t * src, __constrange(0,3) int lane); //VLD4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1441 int32x2x4_t vld4_lane_s32_ptr(__transfersize(4) int32_t const * ptr, int32x2x4_t * src, __constrange(0,1) int lane); //VLD4.32 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1442 lane); // VLD4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1443 float32x2x4_t vld4_lane_f32_ptr(__transfersize(4) float32_t const * ptr, float32x2x4_t * src, __constrange(0,1) int lane); // VLD4.32 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1444 poly8x8x4_t vld4_lane_p8_ptr(__transfersize(4) poly8_t const * ptr, poly8x8x4_t * src, __constrange(0,7) int lane); //VLD4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1445 poly16x4x4_t vld4_lane_p16_ptr(__transfersize(4) poly16_t const * ptr, poly16x4x4_t * src, __constrange(0,3) int lane); // VLD4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
1513 //Store a single lane of N-element structure to memory
1514 void vst2q_lane_u16_ptr(__transfersize(2) uint16_t * ptr, uint16x8x2_t * val, __constrange(0,7) int lane); // VST2.16{d0[0], d2[0]}, [r0]
1515 void vst2q_lane_u32_ptr(__transfersize(2) uint32_t * ptr, uint32x4x2_t * val, __constrange(0,3) int lane); // VST2.32{d0[0], d2[0]}, [r0]
1516 void vst2q_lane_s16_ptr(__transfersize(2) int16_t * ptr, int16x8x2_t * val, __constrange(0,7) int lane); // VST2.16{d0[0], d2[0]}, [r0]
1517 void vst2q_lane_s32_ptr(__transfersize(2) int32_t * ptr, int32x4x2_t * val, __constrange(0,3) int lane); // VST2.32{d0[0], d2[0]}, [r0]
1518 void vst2q_lane_f16_ptr(__transfersize(2) __fp16 * ptr, float16x8x2_t * val, __constrange(0,7) int lane); // VST2.16{d0[0], d2[0]}, [r0]
1519 void vst2q_lane_f32_ptr(__transfersize(2) float32_t * ptr, float32x4x2_t * val, __constrange(0,3) int lane); //VST2.32 {d0[0], d2[0]}, [r0]
1520 void vst2q_lane_p16_ptr(__transfersize(2) poly16_t * ptr, poly16x8x2_t * val, __constrange(0,7) int lane); // VST2.16{d0[0], d2[0]}, [r0]
1521 void vst2_lane_u8_ptr(__transfersize(2) uint8_t * ptr, uint8x8x2_t * val, __constrange(0,7) int lane); // VST2.8{d0[0], d1[0]}, [r0]
1522 void vst2_lane_u16_ptr(__transfersize(2) uint16_t * ptr, uint16x4x2_t * val, __constrange(0,3) int lane); // VST2.16{d0[0], d1[0]}, [r0]
1523 void vst2_lane_u32_ptr(__transfersize(2) uint32_t * ptr, uint32x2x2_t * val, __constrange(0,1) int lane); // VST2.32{d0[0], d1[0]}, [r0]
1524 void vst2_lane_s8_ptr(__transfersize(2) int8_t * ptr, int8x8x2_t * val, __constrange(0,7) int lane); // VST2.8 {d0[0],d1[0]}, [r0]
1525 void vst2_lane_s16_ptr(__transfersize(2) int16_t * ptr, int16x4x2_t * val, __constrange(0,3) int lane); // VST2.16{d0[0], d1[0]}, [r0]
1526 void vst2_lane_s32_ptr(__transfersize(2) int32_t * ptr, int32x2x2_t * val, __constrange(0,1) int lane); // VST2.32{d0[0], d1[0]}, [r0]
1527 void vst2_lane_f16_ptr(__transfersize(2) __fp16 * ptr, float16x4x2_t * val, __constrange(0,3) int lane); // VST2.16{d0[0], d1[0]}, [r0]
1528 void vst2_lane_f32_ptr(__transfersize(2) float32_t * ptr, float32x2x2_t * val, __constrange(0,1) int lane); // VST2.32{d0[0], d1[0]}, [r0]
1529 void vst2_lane_p8_ptr(__transfersize(2) poly8_t * ptr, poly8x8x2_t * val, __constrange(0,7) int lane); // VST2.8{d0[0], d1[0]}, [r0]
1530 void vst2_lane_p16_ptr(__transfersize(2) poly16_t * ptr, poly16x4x2_t * val, __constrange(0,3) int lane); // VST2.16{d0[0], d1[0]}, [r0]
1531 void vst3q_lane_u16_ptr(__transfersize(3) uint16_t * ptr, uint16x8x3_t * val, __constrange(0,7) int lane); // VST3.16{d0[0], d2[0], d4[0]}, [r0]
1532 void vst3q_lane_u32_ptr(__transfersize(3) uint32_t * ptr, uint32x4x3_t * val, __constrange(0,3) int lane); // VST3.32{d0[0], d2[0], d4[0]}, [r0]
1533 void vst3q_lane_s16_ptr(__transfersize(3) int16_t * ptr, int16x8x3_t * val, __constrange(0,7) int lane); // VST3.16{d0[0], d2[0], d4[0]}, [r0]
1534 void vst3q_lane_s32_ptr(__transfersize(3) int32_t * ptr, int32x4x3_t * val, __constrange(0,3) int lane); // VST3.32{d0[0], d2[0], d4[0]}, [r0]
1535 void vst3q_lane_f16_ptr(__transfersize(3) __fp16 * ptr, float16x8x3_t * val, __constrange(0,7) int lane); // VST3.16{d0[0], d2[0], d4[0]}, [r0]
1536 void vst3q_lane_f32_ptr(__transfersize(3) float32_t * ptr, float32x4x3_t * val, __constrange(0,3) int lane); //VST3.32 {d0[0], d2[0], d4[0]}, [r0]
1537 void vst3q_lane_p16_ptr(__transfersize(3) poly16_t * ptr, poly16x8x3_t * val, __constrange(0,7) int lane); // VST3.16{d0[0], d2[0], d4[0]}, [r0]
1538 void vst3_lane_u8_ptr(__transfersize(3) uint8_t * ptr, uint8x8x3_t * val, __constrange(0,7) int lane); // VST3.8{d0[0], d1[0], d2[0]}, [r0]
1539 void vst3_lane_u16_ptr(__transfersize(3) uint16_t * ptr, uint16x4x3_t * val, __constrange(0,3) int lane); // VST3.16{d0[0], d1[0], d2[0]}, [r0]
1540 void vst3_lane_u32_ptr(__transfersize(3) uint32_t * ptr, uint32x2x3_t * val, __constrange(0,1) int lane); // VST3.32{d0[0], d1[0], d2[0]}, [r0]
1541 void vst3_lane_s8_ptr(__transfersize(3) int8_t * ptr, int8x8x3_t * val, __constrange(0,7) int lane); // VST3.8 {d0[0],d1[0], d2[0]}, [r0]
1542 void vst3_lane_s16_ptr(__transfersize(3) int16_t * ptr, int16x4x3_t * val, __constrange(0,3) int lane); // VST3.16{d0[0], d1[0], d2[0]}, [r0]
1543 void vst3_lane_s32_ptr(__transfersize(3) int32_t * ptr, int32x2x3_t * val, __constrange(0,1) int lane); // VST3.32{d0[0], d1[0], d2[0]}, [r0]
1544 void vst3_lane_f16_ptr(__transfersize(3) __fp16 * ptr, float16x4x3_t * val, __constrange(0,3) int lane); // VST3.16{d0[0], d1[0], d2[0]}, [r0]
1545 void vst3_lane_f32_ptr(__transfersize(3) float32_t * ptr, float32x2x3_t * val, __constrange(0,1) int lane); // VST3.32{d0[0], d1[0], d2[0]}, [r0]
1546 void vst3_lane_p8_ptr(__transfersize(3) poly8_t * ptr, poly8x8x3_t * val, __constrange(0,7) int lane); // VST3.8{d0[0], d1[0], d2[0]}, [r0]
1547 void vst3_lane_p16_ptr(__transfersize(3) poly16_t * ptr, poly16x4x3_t * val, __constrange(0,3) int lane); // VST3.16{d0[0], d1[0], d2[0]}, [r0]
1548 void vst4q_lane_u16_ptr(__transfersize(4) uint16_t * ptr, uint16x8x4_t * val, __constrange(0,7) int lane); // VST4.16{d0[0], d2[0], d4[0], d6[0]}, [r0]
1549 void vst4q_lane_u32_ptr(__transfersize(4) uint32_t * ptr, uint32x4x4_t * val, __constrange(0,3) int lane); // VST4.32{d0[0], d2[0], d4[0], d6[0]}, [r0]
1550 void vst4q_lane_s16_ptr(__transfersize(4) int16_t * ptr, int16x8x4_t * val, __constrange(0,7) int lane); // VST4.16{d0[0], d2[0], d4[0], d6[0]}, [r0]
1551 void vst4q_lane_s32_ptr(__transfersize(4) int32_t * ptr, int32x4x4_t * val, __constrange(0,3) int lane); // VST4.32{d0[0], d2[0], d4[0], d6[0]}, [r0]
1552 void vst4q_lane_f16_ptr(__transfersize(4) __fp16 * ptr, float16x8x4_t * val, __constrange(0,7) int lane); // VST4.16{d0[0], d2[0], d4[0], d6[0]}, [r0]
1553 void vst4q_lane_f32_ptr(__transfersize(4) float32_t * ptr, float32x4x4_t * val, __constrange(0,3) int lane); //VST4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
1554 void vst4q_lane_p16_ptr(__transfersize(4) poly16_t * ptr, poly16x8x4_t * val, __constrange(0,7) int lane); // VST4.16{d0[0], d2[0], d4[0], d6[0]}, [r0]
1555 void vst4_lane_u8_ptr(__transfersize(4) uint8_t * ptr, uint8x8x4_t * val, __constrange(0,7) int lane); // VST4.8{d0[0], d1[0], d2[0], d3[0]}, [r0]
1556 void vst4_lane_u16_ptr(__transfersize(4) uint16_t * ptr, uint16x4x4_t * val, __constrange(0,3) int lane); // VST4.16{d0[0], d1[0], d2[0], d3[0]}, [r0]
1557 void vst4_lane_u32_ptr(__transfersize(4) uint32_t * ptr, uint32x2x4_t * val, __constrange(0,1) int lane); // VST4.32{d0[0], d1[0], d2[0], d3[0]}, [r0]
1558 void vst4_lane_s8_ptr(__transfersize(4) int8_t * ptr, int8x8x4_t * val, __constrange(0,7) int lane); // VST4.8 {d0[0],d1[0], d2[0], d3[0]}, [r0]
1559 void vst4_lane_s16_ptr(__transfersize(4) int16_t * ptr, int16x4x4_t * val, __constrange(0,3) int lane); // VST4.16{d0[0], d1[0], d2[0], d3[0]}, [r0]
1560 void vst4_lane_s32_ptr(__transfersize(4) int32_t * ptr, int32x2x4_t * val, __constrange(0,1) int lane); // VST4.32{d0[0], d1[0], d2[0], d3[0]}, [r0]
1561 void vst4_lane_f16_ptr(__transfersize(4) __fp16 * ptr, float16x4x4_t * val, __constrange(0,3) int lane); // VST4.16{d0[0], d1[0], d2[0], d3[0]}, [r0]
1562 void vst4_lane_f32_ptr(__transfersize(4) float32_t * ptr, float32x2x4_t * val, __constrange(0,1) int lane); // VST4.32{d0[0], d1[0], d2[0], d3[0]}, [r0]
1563 void vst4_lane_p8_ptr(__transfersize(4) poly8_t * ptr, poly8x8x4_t * val, __constrange(0,7) int lane); // VST4.8{d0[0], d1[0], d2[0], d3[0]}, [r0]
1564 void vst4_lane_p16_ptr(__transfersize(4) poly16_t * ptr, poly16x4x4_t * val, __constrange(0,3) int lane); // VST4.16{d0[0], d1[0], d2[0], d3[0]}, [r0]
1565 //Extract lanes from a vector and put into a register. These intrinsics extract a single lane (element) from a vector.
1566 uint8_t vget_lane_u8(uint8x8_t vec, __constrange(0,7) int lane); // VMOV.U8 r0, d0[0]
1567 uint16_t vget_lane_u16(uint16x4_t vec, __constrange(0,3) int lane); // VMOV.U16 r0, d0[0]
1568 uint32_t vget_lane_u32(uint32x2_t vec, __constrange(0,1) int lane); // VMOV.32 r0, d0[0]
1569 int8_t vget_lane_s8(int8x8_t vec, __constrange(0,7) int lane); // VMOV.S8 r0, d0[0]
1570 int16_t vget_lane_s16(int16x4_t vec, __constrange(0,3) int lane); // VMOV.S16 r0, d0[0]
1571 int32_t vget_lane_s32(int32x2_t vec, __constrange(0,1) int lane); // VMOV.32 r0, d0[0]
1572 poly8_t vget_lane_p8(poly8x8_t vec, __constrange(0,7) int lane); // VMOV.U8 r0, d0[0]
1573 poly16_t vget_lane_p16(poly16x4_t vec, __constrange(0,3) int lane); // VMOV.U16 r0, d0[0]
1574 float32_t vget_lane_f32(float32x2_t vec, __constrange(0,1) int lane); // VMOV.32 r0, d0[0]
1575 uint8_t vgetq_lane_u8(uint8x16_t vec, __constrange(0,15) int lane); // VMOV.U8 r0, d0[0]
1576 uint16_t vgetq_lane_u16(uint16x8_t vec, __constrange(0,7) int lane); // VMOV.U16 r0, d0[0]
1577 uint32_t vgetq_lane_u32(uint32x4_t vec, __constrange(0,3) int lane); // VMOV.32 r0, d0[0]
1578 int8_t vgetq_lane_s8(int8x16_t vec, __constrange(0,15) int lane); // VMOV.S8 r0, d0[0]
1579 int16_t vgetq_lane_s16(int16x8_t vec, __constrange(0,7) int lane); // VMOV.S16 r0, d0[0]
1580 int32_t vgetq_lane_s32(int32x4_t vec, __constrange(0,3) int lane); // VMOV.32 r0, d0[0]
1581 poly8_t vgetq_lane_p8(poly8x16_t vec, __constrange(0,15) int lane); // VMOV.U8 r0, d0[0]
1582 lane); // VMOV.U16 r0, d0[0]
1583 float32_t vgetq_lane_f32(float32x4_t vec, __constrange(0,3) int lane); // VMOV.32 r0, d0[0]
1584 int64_t vget_lane_s64(int64x1_t vec, __constrange(0,0) int lane); // VMOV r0,r0,d0
1585 uint64_t vget_lane_u64(uint64x1_t vec, __constrange(0,0) int lane); // VMOV r0,r0,d0
1586 int64_t vgetq_lane_s64(int64x2_t vec, __constrange(0,1) int lane); // VMOV r0,r0,d0
1587 uint64_t vgetq_lane_u64(uint64x2_t vec, __constrange(0,1) int lane); // VMOV r0,r0,d0
1588 //Load a single lane of a vector from a literal. These intrinsics set a single lane (element) within a vector.
1589 uint8x8_t vset_lane_u8(uint8_t value, uint8x8_t vec, __constrange(0,7) int lane); // VMOV.8 d0[0],r0
1590 uint16x4_t vset_lane_u16(uint16_t value, uint16x4_t vec, __constrange(0,3) int lane); // VMOV.16 d0[0],r0
1591 uint32x2_t vset_lane_u32(uint32_t value, uint32x2_t vec, __constrange(0,1) int lane); // VMOV.32 d0[0],r0
1592 int8x8_t vset_lane_s8(int8_t value, int8x8_t vec, __constrange(0,7) int lane); // VMOV.8 d0[0],r0
1593 int16x4_t vset_lane_s16(int16_t value, int16x4_t vec, __constrange(0,3) int lane); // VMOV.16 d0[0],r0
1594 int32x2_t vset_lane_s32(int32_t value, int32x2_t vec, __constrange(0,1) int lane); // VMOV.32 d0[0],r0
1595 poly8x8_t vset_lane_p8(poly8_t value, poly8x8_t vec, __constrange(0,7) int lane); // VMOV.8 d0[0],r0
1596 poly16x4_t vset_lane_p16(poly16_t value, poly16x4_t vec, __constrange(0,3) int lane); // VMOV.16 d0[0],r0
1597 float32x2_t vset_lane_f32(float32_t value, float32x2_t vec, __constrange(0,1) int lane); // VMOV.32 d0[0],r0
1598 uint8x16_t vsetq_lane_u8(uint8_t value, uint8x16_t vec, __constrange(0,15) int lane); // VMOV.8 d0[0],r0
1599 uint16x8_t vsetq_lane_u16(uint16_t value, uint16x8_t vec, __constrange(0,7) int lane); // VMOV.16 d0[0],r0
1600 uint32x4_t vsetq_lane_u32(uint32_t value, uint32x4_t vec, __constrange(0,3) int lane); // VMOV.32 d0[0],r0
1601 int8x16_t vsetq_lane_s8(int8_t value, int8x16_t vec, __constrange(0,15) int lane); // VMOV.8 d0[0],r0
1602 int16x8_t vsetq_lane_s16(int16_t value, int16x8_t vec, __constrange(0,7) int lane); // VMOV.16 d0[0],r0
1603 int32x4_t vsetq_lane_s32(int32_t value, int32x4_t vec, __constrange(0,3) int lane); // VMOV.32 d0[0],r0
1604 poly8x16_t vsetq_lane_p8(poly8_t value, poly8x16_t vec, __constrange(0,15) int lane); // VMOV.8 d0[0],r0
1605 poly16x8_t vsetq_lane_p16(poly16_t value, poly16x8_t vec, __constrange(0,7) int lane); // VMOV.16 d0[0],r0
1606 float32x4_t vsetq_lane_f32(float32_t value, float32x4_t vec, __constrange(0,3) int lane); // VMOV.32 d0[0],r0
1607 int64x1_t vset_lane_s64(int64_t value, int64x1_t vec, __constrange(0,0) int lane); // VMOV d0,r0,r0
1608 uint64x1_t vset_lane_u64(uint64_t value, uint64x1_t vec, __constrange(0,0) int lane); // VMOV d0,r0,r0
1609 int64x2_t vsetq_lane_s64(int64_t value, int64x2_t vec, __constrange(0,1) int lane); // VMOV d0,r0,r0
1610 uint64x2_t vsetq_lane_u64(uint64_t value, uint64x2_t vec, __constrange(0,1) int lane); // VMOV d0,r0,r0
1670 //Load all lanes of the vector to the value of a lane of a vector
1671 uint8x8_t vdup_lane_u8(uint8x8_t vec, __constrange(0,7) int lane); // VDUP.8 d0,d0[0]
1672 uint16x4_t vdup_lane_u16(uint16x4_t vec, __constrange(0,3) int lane); // VDUP.16 d0,d0[0]
1673 uint32x2_t vdup_lane_u32(uint32x2_t vec, __constrange(0,1) int lane); // VDUP.32 d0,d0[0]
1674 int8x8_t vdup_lane_s8(int8x8_t vec, __constrange(0,7) int lane); // VDUP.8 d0,d0[0]
1675 int16x4_t vdup_lane_s16(int16x4_t vec, __constrange(0,3) int lane); // VDUP.16 d0,d0[0]
1676 int32x2_t vdup_lane_s32(int32x2_t vec, __constrange(0,1) int lane); // VDUP.32 d0,d0[0]
1677 poly8x8_t vdup_lane_p8(poly8x8_t vec, __constrange(0,7) int lane); // VDUP.8 d0,d0[0]
1678 poly16x4_t vdup_lane_p16(poly16x4_t vec, __constrange(0,3) int lane); // VDUP.16 d0,d0[0]
1679 float32x2_t vdup_lane_f32(float32x2_t vec, __constrange(0,1) int lane); // VDUP.32 d0,d0[0]
1680 uint8x16_t vdupq_lane_u8(uint8x8_t vec, __constrange(0,7) int lane); // VDUP.8 q0,d0[0]
1681 uint16x8_t vdupq_lane_u16(uint16x4_t vec, __constrange(0,3) int lane); // VDUP.16 q0,d0[0]
1682 uint32x4_t vdupq_lane_u32(uint32x2_t vec, __constrange(0,1) int lane); // VDUP.32 q0,d0[0]
1683 int8x16_t vdupq_lane_s8(int8x8_t vec, __constrange(0,7) int lane); // VDUP.8 q0,d0[0]
1684 int16x8_t vdupq_lane_s16(int16x4_t vec, __constrange(0,3) int lane); // VDUP.16 q0,d0[0]
1685 int32x4_t vdupq_lane_s32(int32x2_t vec, __constrange(0,1) int lane); // VDUP.32 q0,d0[0]
1686 poly8x16_t vdupq_lane_p8(poly8x8_t vec, __constrange(0,7) int lane); // VDUP.8 q0,d0[0]
1687 poly16x8_t vdupq_lane_p16(poly16x4_t vec, __constrange(0,3) int lane); // VDUP.16 q0,d0[0]
1688 float32x4_t vdupq_lane_f32(float32x2_t vec, __constrange(0,1) int lane); // VDUP.32 q0,d0[0]
1689 int64x1_t vdup_lane_s64(int64x1_t vec, __constrange(0,0) int lane); // VMOV d0,d0
1690 uint64x1_t vdup_lane_u64(uint64x1_t vec, __constrange(0,0) int lane); // VMOV d0,d0
1691 int64x2_t vdupq_lane_s64(int64x1_t vec, __constrange(0,0) int lane); // VMOV q0,q0
1692 uint64x2_t vdupq_lane_u64(uint64x1_t vec, __constrange(0,0) int lane); // VMOV q0,q0
2267 #define _NEON2SSE_SWITCH16(NAME, a, b, LANE) \
2268 switch(LANE) \
2289 #define _NEON2SSE_SWITCH8(NAME, vec, LANE, p) \
2290 switch(LANE) \
2303 #define _NEON2SSE_SWITCH4(NAME, case0, case1, case2, case3, vec, LANE, p) \
2304 switch(LANE) \
2313 _NEON2SSE_INLINE __m128i _MM_ALIGNR_EPI8(__m128i a, __m128i b, int LANE)
2315 _NEON2SSE_SWITCH16(_mm_alignr_epi8, a, _NEON2SSE_COMMA b, LANE)
2318 _NEON2SSE_INLINE __m128i _MM_INSERT_EPI16(__m128i vec, int p, const int LANE)
2320 _NEON2SSE_SWITCH8(_mm_insert_epi16, vec, LANE, _NEON2SSE_COMMA p)
2323 _NEON2SSE_INLINE int _MM_EXTRACT_EPI16(__m128i vec, const int LANE)
2325 _NEON2SSE_SWITCH8(_mm_extract_epi16, vec, LANE,)
2329 _NEON2SSE_INLINE int _MM_EXTRACT_EPI32(__m128i vec, const int LANE)
2331 _NEON2SSE_SWITCH4(_mm_extract_epi32, 0,1,2,3, vec, LANE,)
2334 _NEON2SSE_INLINE int _MM_EXTRACT_PS(__m128 vec, const int LANE)
2336 _NEON2SSE_SWITCH4(_mm_extract_ps, 0,1,2,3, vec, LANE,)
2339 _NEON2SSE_INLINE int _MM_EXTRACT_EPI8(__m128i vec, const int LANE)
2341 _NEON2SSE_SWITCH16(_mm_extract_epi8, vec, , LANE)
2344 _NEON2SSE_INLINE __m128i _MM_INSERT_EPI32(__m128i vec, int p, const int LANE)
2346 _NEON2SSE_SWITCH4(_mm_insert_epi32, 0, 1, 2, 3, vec, LANE, _NEON2SSE_COMMA p)
2349 _NEON2SSE_INLINE __m128i _MM_INSERT_EPI8(__m128i vec, int p, const int LANE)
2351 _NEON2SSE_SWITCH16(_mm_insert_epi8, vec, _NEON2SSE_COMMA p, LANE)
2356 _NEON2SSE_INLINE __m128i _MM_INSERT_EPI64(__m128i vec, int p, const int LANE)
2358 switch(LANE) {
2368 _NEON2SSE_INLINE int64_t _MM_EXTRACT_EPI64(__m128i val, const int LANE)
2370 if (LANE ==0) return _mm_extract_epi64(val, 0);
2375 _NEON2SSE_INLINE __m128 _MM_INSERT_PS(__m128 vec, __m128 p, const int LANE)
2377 _NEON2SSE_SWITCH4(_mm_insert_ps, 0, 16, 32, 48, vec, LANE, _NEON2SSE_COMMA p)
2454 _NEON2SSE_INLINE int _MM_EXTRACT_EPI32(__m128i vec, const int LANE)
2458 return tmp[LANE];
2461 _NEON2SSE_INLINE int _MM_EXTRACT_EPI8(__m128i vec, const int LANE)
2465 return (int)tmp[LANE];
2468 _NEON2SSE_INLINE int _MM_EXTRACT_PS(__m128 vec, const int LANE)
2472 return tmp[LANE];
2475 _NEON2SSE_INLINE __m128i _MM_INSERT_EPI32(__m128i vec, int p, const int LANE)
2480 pvec[LANE] = p;
2481 mask[LANE] = 0x0;
2487 _NEON2SSE_INLINE __m128i _MM_INSERT_EPI8(__m128i vec, int p, const int LANE)
2492 pvec[LANE] = (int8_t)p;
2493 mask[LANE] = 0x0;
2499 _NEON2SSE_INLINE __m128 _MM_INSERT_PS(__m128 vec, __m128 p, const int LANE)
2503 mask[LANE >> 4] = 0x0; //here the LANE is not actural lane, need to deal with it
2674 _NEON2SSE_INLINE __m128i _MM_INSERT_EPI64_32(__m128i vec, int p, const int LANE)
2679 pvec[LANE] = p;
2680 mask[LANE] = 0x0;
2686 _NEON2SSE_INLINE int64_t _MM_EXTRACT_EPI64_32(__m128i val, const int LANE)
2690 return tmp[LANE];
2835 //Va, Vb have equal lane sizes, result is a 128 bit vector of lanes that are twice the width.
4417 //Va, Vb have equal lane sizes, result is a 128 bit vector of lanes that are twice the width.
9279 //******* Lane load functions - insert the data at vector's given position (lane) *************************
9281 uint8x16_t vld1q_lane_u8(__transfersize(1) uint8_t const * ptr, uint8x16_t vec, __constrange(0,15) int lane); // VLD1.8 {d0[0]}, [r0]
9282 #define vld1q_lane_u8(ptr, vec, lane) _MM_INSERT_EPI8(vec, *(ptr), lane)
9284 uint16x8_t vld1q_lane_u16(__transfersize(1) uint16_t const * ptr, uint16x8_t vec, __constrange(0,7) int lane); // VLD1.16 {d0[0]}, [r0]
9285 #define vld1q_lane_u16(ptr, vec, lane) _MM_INSERT_EPI16(vec, *(ptr), lane)
9287 uint32x4_t vld1q_lane_u32(__transfersize(1) uint32_t const * ptr, uint32x4_t vec, __constrange(0,3) int lane); // VLD1.32 {d0[0]}, [r0]
9288 #define vld1q_lane_u32(ptr, vec, lane) _MM_INSERT_EPI32(vec, *(ptr), lane)
9290 uint64x2_t vld1q_lane_u64(__transfersize(1) uint64_t const * ptr, uint64x2_t vec, __constrange(0,1) int lane); // VLD1.64 {d0}, [r0]
9291 #define vld1q_lane_u64(ptr, vec, lane) _MM_INSERT_EPI64(vec, *(ptr), lane); // _p;
9294 int8x16_t vld1q_lane_s8(__transfersize(1) int8_t const * ptr, int8x16_t vec, __constrange(0,15) int lane); // VLD1.8 {d0[0]}, [r0]
9295 #define vld1q_lane_s8(ptr, vec, lane) _MM_INSERT_EPI8(vec, *(ptr), lane)
9297 int16x8_t vld1q_lane_s16(__transfersize(1) int16_t const * ptr, int16x8_t vec, __constrange(0,7) int lane); // VLD1.16 {d0[0]}, [r0]
9298 #define vld1q_lane_s16(ptr, vec, lane) _MM_INSERT_EPI16(vec, *(ptr), lane)
9300 int32x4_t vld1q_lane_s32(__transfersize(1) int32_t const * ptr, int32x4_t vec, __constrange(0,3) int lane); // VLD1.32 {d0[0]}, [r0]
9301 #define vld1q_lane_s32(ptr, vec, lane) _MM_INSERT_EPI32(vec, *(ptr), lane)
9303 float16x8_t vld1q_lane_f16(__transfersize(1) __fp16 const * ptr, float16x8_t vec, __constrange(0,7) int lane); // VLD1.16 {d0[0]}, [r0]
9306 float32x4_t vld1q_lane_f32(__transfersize(1) float32_t const * ptr, float32x4_t vec, __constrange(0,3) int lane); // VLD1.32 {d0[0]}, [r0]
9307 _NEON2SSE_INLINE float32x4_t vld1q_lane_f32(__transfersize(1) float32_t const * ptr, float32x4_t vec, __constrange(0,3) int lane)
9312 return _MM_INSERT_PS(vec, p, _INSERTPS_NDX(0, lane));
9315 int64x2_t vld1q_lane_s64(__transfersize(1) int64_t const * ptr, int64x2_t vec, __constrange(0,1) int lane); // VLD1.64 {d0}, [r0]
9316 #define vld1q_lane_s64(ptr, vec, lane) _MM_INSERT_EPI64(vec, *(ptr), lane)
9318 poly8x16_t vld1q_lane_p8(__transfersize(1) poly8_t const * ptr, poly8x16_t vec, __constrange(0,15) int lane); // VLD1.8 {d0[0]}, [r0]
9319 #define vld1q_lane_p8(ptr, vec, lane) _MM_INSERT_EPI8(vec, *(ptr), lane)
9321 poly16x8_t vld1q_lane_p16(__transfersize(1) poly16_t const * ptr, poly16x8_t vec, __constrange(0,7) int lane); // VLD1.16 {d0[0]}, [r0]
9322 #define vld1q_lane_p16(ptr, vec, lane) _MM_INSERT_EPI16(vec, *(ptr), lane)
9324 uint8x8_t vld1_lane_u8(__transfersize(1) uint8_t const * ptr, uint8x8_t vec, __constrange(0,7) int lane); // VLD1.8 {d0[0]}, [r0]
9325 _NEON2SSE_INLINE uint8x8_t vld1_lane_u8(__transfersize(1) uint8_t const * ptr, uint8x8_t vec, __constrange(0,7) int lane)
9329 res.m64_u8[lane] = *(ptr);
9333 uint16x4_t vld1_lane_u16(__transfersize(1) uint16_t const * ptr, uint16x4_t vec, __constrange(0,3) int lane); // VLD1.16 {d0[0]}, [r0]
9334 _NEON2SSE_INLINE uint16x4_t vld1_lane_u16(__transfersize(1) uint16_t const * ptr, uint16x4_t vec, __constrange(0,3) int lane)
9338 res.m64_u16[lane] = *(ptr);
9342 uint32x2_t vld1_lane_u32(__transfersize(1) uint32_t const * ptr, uint32x2_t vec, __constrange(0,1) int lane); // VLD1.32 {d0[0]}, [r0]
9343 _NEON2SSE_INLINE uint32x2_t vld1_lane_u32(__transfersize(1) uint32_t const * ptr, uint32x2_t vec, __constrange(0,1) int lane)
9347 res.m64_u32[lane] = *(ptr);
9351 uint64x1_t vld1_lane_u64(__transfersize(1) uint64_t const * ptr, uint64x1_t vec, __constrange(0,0) int lane); // VLD1.64 {d0}, [r0]
9352 _NEON2SSE_INLINE uint64x1_t vld1_lane_u64(__transfersize(1) uint64_t const * ptr, uint64x1_t vec, __constrange(0,0) int lane)
9360 int8x8_t vld1_lane_s8(__transfersize(1) int8_t const * ptr, int8x8_t vec, __constrange(0,7) int lane); // VLD1.8 {d0[0]}, [r0]
9361 #define vld1_lane_s8(ptr, vec, lane) vld1_lane_u8((uint8_t*)ptr, vec, lane)
9363 int16x4_t vld1_lane_s16(__transfersize(1) int16_t const * ptr, int16x4_t vec, __constrange(0,3) int lane); // VLD1.16 {d0[0]}, [r0]
9364 #define vld1_lane_s16(ptr, vec, lane) vld1_lane_u16((uint16_t*)ptr, vec, lane)
9366 int32x2_t vld1_lane_s32(__transfersize(1) int32_t const * ptr, int32x2_t vec, __constrange(0,1) int lane); // VLD1.32 {d0[0]}, [r0]
9367 #define vld1_lane_s32(ptr, vec, lane) vld1_lane_u32((uint32_t*)ptr, vec, lane)
9369 float16x4_t vld1_lane_f16(__transfersize(1) __fp16 const * ptr, float16x4_t vec, __constrange(0,3) int lane); // VLD1.16 {d0[0]}, [r0]
9372 float32x2_t vld1_lane_f32(__transfersize(1) float32_t const * ptr, float32x2_t vec, __constrange(0,1) int lane); // VLD1.32 {d0[0]}, [r0]
9373 _NEON2SSE_INLINE float32x2_t vld1_lane_f32(__transfersize(1) float32_t const * ptr, float32x2_t vec, __constrange(0,1) int lane)
9377 res.m64_f32[lane] = *(ptr);
9381 int64x1_t vld1_lane_s64(__transfersize(1) int64_t const * ptr, int64x1_t vec, __constrange(0,0) int lane); // VLD1.64 {d0}, [r0]
9382 #define vld1_lane_s64(ptr, vec, lane) vld1_lane_u64((uint64_t*)ptr, vec, lane)
9384 poly8x8_t vld1_lane_p8(__transfersize(1) poly8_t const * ptr, poly8x8_t vec, __constrange(0,7) int lane); // VLD1.8 {d0[0]}, [r0]
9387 poly16x4_t vld1_lane_p16(__transfersize(1) poly16_t const * ptr, poly16x4_t vec, __constrange(0,3) int lane); // VLD1.16 {d0[0]}, [r0]
9626 //***********Store a lane of a vector into memory (extract given lane) *********************
9628 void vst1q_lane_u8(__transfersize(1) uint8_t * ptr, uint8x16_t val, __constrange(0,15) int lane); // VST1.8 {d0[0]}, [r0]
9629 #define vst1q_lane_u8(ptr, val, lane) *(ptr) = _MM_EXTRACT_EPI8 (val, lane)
9631 void vst1q_lane_u16(__transfersize(1) uint16_t * ptr, uint16x8_t val, __constrange(0,7) int lane); // VST1.16 {d0[0]}, [r0]
9632 #define vst1q_lane_u16(ptr, val, lane) *(ptr) = _MM_EXTRACT_EPI16 (val, lane)
9634 void vst1q_lane_u32(__transfersize(1) uint32_t * ptr, uint32x4_t val, __constrange(0,3) int lane); // VST1.32 {d0[0]}, [r0]
9635 #define vst1q_lane_u32(ptr, val, lane) *(ptr) = _MM_EXTRACT_EPI32 (val, lane)
9637 void vst1q_lane_u64(__transfersize(1) uint64_t * ptr, uint64x2_t val, __constrange(0,1) int lane); // VST1.64 {d0}, [r0]
9638 #define vst1q_lane_u64(ptr, val, lane) *(ptr) = _MM_EXTRACT_EPI64 (val, lane)
9640 void vst1q_lane_s8(__transfersize(1) int8_t * ptr, int8x16_t val, __constrange(0,15) int lane); // VST1.8 {d0[0]}, [r0]
9641 #define vst1q_lane_s8(ptr, val, lane) *(ptr) = _MM_EXTRACT_EPI8 (val, lane)
9643 void vst1q_lane_s16(__transfersize(1) int16_t * ptr, int16x8_t val, __constrange(0,7) int lane); // VST1.16 {d0[0]}, [r0]
9644 #define vst1q_lane_s16(ptr, val, lane) *(ptr) = _MM_EXTRACT_EPI16 (val, lane)
9646 void vst1q_lane_s32(__transfersize(1) int32_t * ptr, int32x4_t val, __constrange(0,3) int lane); // VST1.32 {d0[0]}, [r0]
9647 #define vst1q_lane_s32(ptr, val, lane) *(ptr) = _MM_EXTRACT_EPI32 (val, lane)
9649 void vst1q_lane_s64(__transfersize(1) int64_t * ptr, int64x2_t val, __constrange(0,1) int lane); // VST1.64 {d0}, [r0]
9650 #define vst1q_lane_s64(ptr, val, lane) *(ptr) = _MM_EXTRACT_EPI64 (val, lane)
9652 void vst1q_lane_f16(__transfersize(1) __fp16 * ptr, float16x8_t val, __constrange(0,7) int lane); // VST1.16 {d0[0]}, [r0]
9655 void vst1q_lane_f32(__transfersize(1) float32_t * ptr, float32x4_t val, __constrange(0,3) int lane); // VST1.32 {d0[0]}, [r0]
9656 _NEON2SSE_INLINE void vst1q_lane_f32(__transfersize(1) float32_t * ptr, float32x4_t val, __constrange(0,3) int lane)
9659 ilane = _MM_EXTRACT_PS(val,lane);
9663 void vst1q_lane_p8(__transfersize(1) poly8_t * ptr, poly8x16_t val, __constrange(0,15) int lane); // VST1.8 {d0[0]}, [r0]
9666 void vst1q_lane_p16(__transfersize(1) poly16_t * ptr, poly16x8_t val, __constrange(0,7) int lane); // VST1.16 {d0[0]}, [r0]
9669 void vst1_lane_u8(__transfersize(1) uint8_t * ptr, uint8x8_t val, __constrange(0,7) int lane); // VST1.8 {d0[0]}, [r0]
9670 _NEON2SSE_INLINE void vst1_lane_u8(__transfersize(1) uint8_t * ptr, uint8x8_t val, __constrange(0,7) int lane)
9672 *(ptr) = val.m64_u8[lane];
9675 void vst1_lane_u16(__transfersize(1) uint16_t * ptr, uint16x4_t val, __constrange(0,3) int lane); // VST1.16 {d0[0]}, [r0]
9676 _NEON2SSE_INLINE void vst1_lane_u16(__transfersize(1) uint16_t * ptr, uint16x4_t val, __constrange(0,3) int lane)
9678 *(ptr) = val.m64_u16[lane];
9681 void vst1_lane_u32(__transfersize(1) uint32_t * ptr, uint32x2_t val, __constrange(0,1) int lane); // VST1.32 {d0[0]}, [r0]
9682 _NEON2SSE_INLINE void vst1_lane_u32(__transfersize(1) uint32_t * ptr, uint32x2_t val, __constrange(0,1) int lane)
9684 *(ptr) = val.m64_u32[lane];
9687 void vst1_lane_u64(__transfersize(1) uint64_t * ptr, uint64x1_t val, __constrange(0,0) int lane); // VST1.64 {d0}, [r0]
9688 _NEON2SSE_INLINE void vst1_lane_u64(__transfersize(1) uint64_t * ptr, uint64x1_t val, __constrange(0,0) int lane)
9693 void vst1_lane_s8(__transfersize(1) int8_t * ptr, int8x8_t val, __constrange(0,7) int lane); // VST1.8 {d0[0]}, [r0]
9694 #define vst1_lane_s8(ptr, val, lane) vst1_lane_u8((uint8_t*)ptr, val, lane)
9696 void vst1_lane_s16(__transfersize(1) int16_t * ptr, int16x4_t val, __constrange(0,3) int lane); // VST1.16 {d0[0]}, [r0]
9697 #define vst1_lane_s16(ptr, val, lane) vst1_lane_u16((uint16_t*)ptr, val, lane)
9699 void vst1_lane_s32(__transfersize(1) int32_t * ptr, int32x2_t val, __constrange(0,1) int lane); // VST1.32 {d0[0]}, [r0]
9700 #define vst1_lane_s32(ptr, val, lane) vst1_lane_u32((uint32_t*)ptr, val, lane)
9703 void vst1_lane_s64(__transfersize(1) int64_t * ptr, int64x1_t val, __constrange(0,0) int lane); // VST1.64 {d0}, [r0]
9704 #define vst1_lane_s64(ptr, val, lane) vst1_lane_u64((uint64_t*)ptr, val, lane)
9707 void vst1_lane_f16(__transfersize(1) __fp16 * ptr, float16x4_t val, __constrange(0,3) int lane); // VST1.16 {d0[0]}, [r0]
9710 void vst1_lane_f32(__transfersize(1) float32_t * ptr, float32x2_t val, __constrange(0,1) int lane); // VST1.32 {d0[0]}, [r0]
9711 _NEON2SSE_INLINE void vst1_lane_f32(__transfersize(1) float32_t * ptr, float32x2_t val, __constrange(0,1) int lane)
9713 *(ptr) = val.m64_f32[lane];
9716 void vst1_lane_p8(__transfersize(1) poly8_t * ptr, poly8x8_t val, __constrange(0,7) int lane); // VST1.8 {d0[0]}, [r0]
9719 void vst1_lane_p16(__transfersize(1) poly16_t * ptr, poly16x4_t val, __constrange(0,3) int lane); // VST1.16 {d0[0]}, [r0]
10640 //*******************Lane loads for an N-element structures ***********************
10642 //********************** Lane pairs ************************************************
10643 //does vld1_lane_xx ptr[0] to src->val[0] at lane positon and ptr[1] to src->val[1] at lane positon
10649 //uint16x8x2_t vld2q_lane_u16(__transfersize(2) uint16_t const * ptr, uint16x8x2_t src,__constrange(0,7) int lane);// VLD2.16 {d0[0], d2[0]}, [r0]
10650 _NEON2SSE_INLINE uint16x8x2_t vld2q_lane_u16_ptr(__transfersize(2) uint16_t const * ptr, uint16x8x2_t* src,__constrange(0,7) int lane) // VLD2.16 {d0[0], d2[0]}, [r0]
10653 v.val[0] = vld1q_lane_s16 (ptr, src->val[0], lane);
10654 v.val[1] = vld1q_lane_s16 ((ptr + 1), src->val[1], lane);
10657 #define vld2q_lane_u16(ptr, src, lane) vld2q_lane_u16_ptr(ptr, &src, lane)
10659 //uint32x4x2_t vld2q_lane_u32(__transfersize(2) uint32_t const * ptr, uint32x4x2_t src,__constrange(0,3) int lane);// VLD2.32 {d0[0], d2[0]}, [r0]
10660 _NEON2SSE_INLINE uint32x4x2_t vld2q_lane_u32_ptr(__transfersize(2) uint32_t const * ptr, uint32x4x2_t* src,__constrange(0,3) int lane) // VLD2.32 {d0[0], d2[0]}, [r0]
10663 v.val[0] = _MM_INSERT_EPI32 (src->val[0], ptr[0], lane);
10664 v.val[1] = _MM_INSERT_EPI32 (src->val[1], ptr[1], lane);
10667 #define vld2q_lane_u32(ptr, src, lane) vld2q_lane_u32_ptr(ptr, &src, lane)
10669 //int16x8x2_t vld2q_lane_s16(__transfersize(2) int16_t const * ptr, int16x8x2_t src, __constrange(0,7)int lane);// VLD2.16 {d0[0], d2[0]}, [r0]
10670 _NEON2SSE_INLINE int16x8x2_t vld2q_lane_s16_ptr(__transfersize(2) int16_t const * ptr, int16x8x2_t* src, __constrange(0,7) int lane)
10673 v.val[0] = vld1q_lane_s16 (ptr, src->val[0], lane);
10674 v.val[1] = vld1q_lane_s16 ((ptr + 1), src->val[1], lane);
10677 #define vld2q_lane_s16(ptr, src, lane) vld2q_lane_s16_ptr(ptr, &src, lane)
10679 //int32x4x2_t vld2q_lane_s32(__transfersize(2) int32_t const * ptr, int32x4x2_t src, __constrange(0,3)int lane);// VLD2.32 {d0[0], d2[0]}, [r0]
10680 _NEON2SSE_INLINE int32x4x2_t vld2q_lane_s32_ptr(__transfersize(2) int32_t const * ptr, int32x4x2_t* src, __constrange(0,3) int lane)
10683 v.val[0] = _MM_INSERT_EPI32 (src->val[0], ptr[0], lane);
10684 v.val[1] = _MM_INSERT_EPI32 (src->val[1], ptr[1], lane);
10687 #define vld2q_lane_s32(ptr, src, lane) vld2q_lane_s32_ptr(ptr, &src, lane)
10689 //float16x8x2_t vld2q_lane_f16(__transfersize(2) __fp16 const * ptr, float16x8x2_t src, __constrange(0,7)int lane);// VLD2.16 {d0[0], d2[0]}, [r0]
10692 //float32x4x2_t vld2q_lane_f32_ptr(__transfersize(2) float32_t const * ptr, float32x4x2_t src,__constrange(0,3) int lane);// VLD2.32 {d0[0], d2[0]}, [r0]
10693 _NEON2SSE_INLINE float32x4x2_t vld2q_lane_f32_ptr(__transfersize(2) float32_t const * ptr, float32x4x2_t* src,__constrange(0,3) int lane) // VLD2.32 {d0[0], d2[0]}, [r0]
10696 v.val[0] = vld1q_lane_f32(ptr, src->val[0], lane);
10697 v.val[1] = vld1q_lane_f32((ptr + 1), src->val[1], lane);
10700 #define vld2q_lane_f32(ptr,src,lane) vld2q_lane_f32_ptr(ptr,&src,lane)
10702 //poly16x8x2_t vld2q_lane_p16(__transfersize(2) poly16_t const * ptr, poly16x8x2_t src,__constrange(0,7) int lane);// VLD2.16 {d0[0], d2[0]}, [r0]
10705 //uint8x8x2_t vld2_lane_u8(__transfersize(2) uint8_t const * ptr, uint8x8x2_t src, __constrange(0,7) int lane);// VLD2.8 {d0[0], d1[0]}, [r0]
10706 _NEON2SSE_INLINE uint8x8x2_t vld2_lane_u8_ptr(__transfersize(2) uint8_t const * ptr, uint8x8x2_t* src, __constrange(0,7) int lane) // VLD2.8 {d0[0], d1[0]}, [r0]
10709 v.val[0] = vld1_lane_u8(ptr, src->val[0], lane);
10710 v.val[1] = vld1_lane_u8((ptr + 1), src->val[1], lane);
10713 #define vld2_lane_u8(ptr, src, lane) vld2_lane_u8_ptr(ptr, &src, lane)
10715 //uint16x4x2_t vld2_lane_u16(__transfersize(2) uint16_t const * ptr, uint16x4x2_t src, __constrange(0,3)int lane);// VLD2.16 {d0[0], d1[0]}, [r0]
10716 _NEON2SSE_INLINE uint16x4x2_t vld2_lane_u16_ptr(__transfersize(2) uint16_t const * ptr, uint16x4x2_t* src, __constrange(0,3) int lane)
10719 v.val[0] = vld1_lane_u16(ptr, src->val[0], lane);
10720 v.val[1] = vld1_lane_u16((ptr + 1), src->val[1], lane);
10723 #define vld2_lane_u16(ptr, src, lane) vld2_lane_u16_ptr(ptr, &src, lane)
10725 //uint32x2x2_t vld2_lane_u32(__transfersize(2) uint32_t const * ptr, uint32x2x2_t src, __constrange(0,1)int lane);// VLD2.32 {d0[0], d1[0]}, [r0]
10726 _NEON2SSE_INLINE uint32x2x2_t vld2_lane_u32_ptr(__transfersize(2) uint32_t const * ptr, uint32x2x2_t* src, __constrange(0,1) int lane)
10729 v.val[0] = vld1_lane_u32(ptr, src->val[0], lane);
10730 v.val[1] = vld1_lane_u32((ptr + 1), src->val[1], lane);
10733 #define vld2_lane_u32(ptr, src, lane) vld2_lane_u32_ptr(ptr, &src, lane)
10735 //int8x8x2_t vld2_lane_s8(__transfersize(2) int8_t const * ptr, int8x8x2_t src, __constrange(0,7) int lane);// VLD2.8 {d0[0], d1[0]}, [r0]
10736 int8x8x2_t vld2_lane_s8_ptr(__transfersize(2) int8_t const * ptr, int8x8x2_t * src, __constrange(0,7) int lane); // VLD2.8 {d0[0], d1[0]}, [r0]
10737 #define vld2_lane_s8(ptr, src, lane) vld2_lane_u8(( uint8_t*) ptr, src, lane)
10739 //int16x4x2_t vld2_lane_s16(__transfersize(2) int16_t const * ptr, int16x4x2_t src, __constrange(0,3) int lane);// VLD2.16 {d0[0], d1[0]}, [r0]
10740 int16x4x2_t vld2_lane_s16_ptr(__transfersize(2) int16_t const * ptr, int16x4x2_t * src, __constrange(0,3) int lane); // VLD2.16 {d0[0], d1[0]}, [r0]
10741 #define vld2_lane_s16(ptr, src, lane) vld2_lane_u16(( uint16_t*) ptr, src, lane)
10743 //int32x2x2_t vld2_lane_s32(__transfersize(2) int32_t const * ptr, int32x2x2_t src, __constrange(0,1) int lane);// VLD2.32 {d0[0], d1[0]}, [r0]
10744 int32x2x2_t vld2_lane_s32_ptr(__transfersize(2) int32_t const * ptr, int32x2x2_t * src, __constrange(0,1) int lane); // VLD2.32 {d0[0], d1[0]}, [r0]
10745 #define vld2_lane_s32(ptr, src, lane) vld2_lane_u32(( uint32_t*) ptr, src, lane)
10747 //float16x4x2_t vld2_lane_f16(__transfersize(2) __fp16 const * ptr, float16x4x2_t src, __constrange(0,3) int lane); // VLD2.16 {d0[0], d1[0]}, [r0]
10750 float32x2x2_t vld2_lane_f32_ptr(__transfersize(2) float32_t const * ptr, float32x2x2_t * src,__constrange(0,1) int lane); // VLD2.32 {d0[0], d1[0]}, [r0]
10751 _NEON2SSE_INLINE float32x2x2_t vld2_lane_f32_ptr(__transfersize(2) float32_t const * ptr, float32x2x2_t * src,__constrange(0,1) int lane)
10754 v.val[0] = vld1_lane_f32(ptr, src->val[0], lane);
10755 v.val[1] = vld1_lane_f32((ptr + 1), src->val[1], lane);
10758 #define vld2_lane_f32(ptr, src, lane) vld2_lane_f32_ptr(ptr, &src, lane)
10760 //poly8x8x2_t vld2_lane_p8(__transfersize(2) poly8_t const * ptr, poly8x8x2_t src, __constrange(0,7) int lane);// VLD2.8 {d0[0], d1[0]}, [r0]
10761 poly8x8x2_t vld2_lane_p8_ptr(__transfersize(2) poly8_t const * ptr, poly8x8x2_t * src, __constrange(0,7) int lane); // VLD2.8 {d0[0], d1[0]}, [r0]
10764 //poly16x4x2_t vld2_lane_p16(__transfersize(2) poly16_t const * ptr, poly16x4x2_t src, __constrange(0,3)int lane);// VLD2.16 {d0[0], d1[0]}, [r0]
10765 poly16x4x2_t vld2_lane_p16_ptr(__transfersize(2) poly16_t const * ptr, poly16x4x2_t * src, __constrange(0,3) int lane); // VLD2.16 {d0[0], d1[0]}, [r0]
10768 //*********** Lane triplets **********************
10770 //does vld1_lane_xx ptr[0] to src->val[0], ptr[1] to src->val[1] and ptr[2] to src->val[2] at lane positon
10773 //uint16x8x3_t vld3q_lane_u16(__transfersize(3) uint16_t const * ptr, uint16x8x3_t src,__constrange(0,7) int lane);// VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
10774 _NEON2SSE_INLINE uint16x8x3_t vld3q_lane_u16_ptr(__transfersize(3) uint16_t const * ptr, uint16x8x3_t* src,__constrange(0,7) int lane) // VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
10777 v.val[0] = _MM_INSERT_EPI16 ( src->val[0], ptr[0], lane);
10778 v.val[1] = _MM_INSERT_EPI16 ( src->val[1], ptr[1], lane);
10779 v.val[2] = _MM_INSERT_EPI16 ( src->val[2], ptr[2], lane);
10782 #define vld3q_lane_u16(ptr, src, lane) vld3q_lane_u16_ptr(ptr, &src, lane)
10784 //uint32x4x3_t vld3q_lane_u32(__transfersize(3) uint32_t const * ptr, uint32x4x3_t src,__constrange(0,3) int lane);// VLD3.32 {d0[0], d2[0], d4[0]}, [r0]
10785 _NEON2SSE_INLINE uint32x4x3_t vld3q_lane_u32_ptr(__transfersize(3) uint32_t const * ptr, uint32x4x3_t* src,__constrange(0,3) int lane) // VLD3.32 {d0[0], d2[0], d4[0]}, [r0]
10788 v.val[0] = _MM_INSERT_EPI32 ( src->val[0], ptr[0], lane);
10789 v.val[1] = _MM_INSERT_EPI32 ( src->val[1], ptr[1], lane);
10790 v.val[2] = _MM_INSERT_EPI32 ( src->val[2], ptr[2], lane);
10793 #define vld3q_lane_u32(ptr, src, lane) vld3q_lane_u32_ptr(ptr, &src, lane)
10795 //int16x8x3_t vld3q_lane_s16(__transfersize(3) int16_t const * ptr, int16x8x3_t src, __constrange(0,7)int lane);// VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
10796 _NEON2SSE_INLINE int16x8x3_t vld3q_lane_s16_ptr(__transfersize(3) int16_t const * ptr, int16x8x3_t* src, __constrange(0,7) int lane) // VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
10799 v.val[0] = _MM_INSERT_EPI16 ( src->val[0], ptr[0], lane);
10800 v.val[1] = _MM_INSERT_EPI16 ( src->val[1], ptr[1], lane);
10801 v.val[2] = _MM_INSERT_EPI16 ( src->val[2], ptr[2], lane);
10804 #define vld3q_lane_s16(ptr, src, lane) vld3q_lane_s16_ptr(ptr, &src, lane)
10806 //int32x4x3_t vld3q_lane_s32(__transfersize(3) int32_t const * ptr, int32x4x3_t src, __constrange(0,3)int lane);// VLD3.32 {d0[0], d2[0], d4[0]}, [r0]
10807 _NEON2SSE_INLINE int32x4x3_t vld3q_lane_s32_ptr(__transfersize(3) int32_t const * ptr, int32x4x3_t* src, __constrange(0,3) int lane) // VLD3.32 {d0[0], d2[0], d4[0]}, [r0]
10810 v.val[0] = _MM_INSERT_EPI32 ( src->val[0], ptr[0], lane);
10811 v.val[1] = _MM_INSERT_EPI32 ( src->val[1], ptr[1], lane);
10812 v.val[2] = _MM_INSERT_EPI32 ( src->val[2], ptr[2], lane);
10815 #define vld3q_lane_s32(ptr, src, lane) vld3q_lane_s32_ptr(ptr, &src, lane)
10817 float16x8x3_t vld3q_lane_f16_ptr(__transfersize(3) __fp16 const * ptr, float16x8x3_t * src, __constrange(0,7) int lane); // VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
10819 #define vld3q_lane_f16(ptr, src, lane) vld3q_lane_f16_ptr(ptr, &src, lane)
10822 //float32x4x3_t vld3q_lane_f32(__transfersize(3) float32_t const * ptr, float32x4x3_t src,__constrange(0,3) int lane);// VLD3.32 {d0[0], d2[0], d4[0]}, [r0]
10823 _NEON2SSE_INLINE float32x4x3_t vld3q_lane_f32_ptr(__transfersize(3) float32_t const * ptr, float32x4x3_t* src,__constrange(0,3) int lane) // VLD3.32 {d0[0], d2[0], d4[0]}, [r0]
10826 v.val[0] = vld1q_lane_f32(&ptr[0], src->val[0], lane);
10827 v.val[1] = vld1q_lane_f32(&ptr[1], src->val[1], lane);
10828 v.val[2] = vld1q_lane_f32(&ptr[2], src->val[2], lane);
10831 #define vld3q_lane_f32(ptr,src,lane) vld3q_lane_f32_ptr(ptr,&src,lane)
10833 poly16x8x3_t vld3q_lane_p16_ptr(__transfersize(3) poly16_t const * ptr, poly16x8x3_t * src,__constrange(0,7) int lane); // VLD3.16 {d0[0], d2[0], d4[0]}, [r0]
10836 //uint8x8x3_t vld3_lane_u8(__transfersize(3) uint8_t const * ptr, uint8x8x3_t src, __constrange(0,7) int lane);// VLD3.8 {d0[0], d1[0], d2[0]}, [r0]
10837 _NEON2SSE_INLINE uint8x8x3_t vld3_lane_u8_ptr(__transfersize(3) uint8_t const * ptr, uint8x8x3_t* src, __constrange(0,7) int lane) // VLD3.8 {d0[0], d1[0], d2[0]}, [r0]
10840 v.val[0] = vld1_lane_u8(ptr, src->val[0], lane);
10841 v.val[1] = vld1_lane_u8((ptr + 1), src->val[1], lane);
10842 v.val[2] = vld1_lane_u8((ptr + 2), src->val[2], lane);
10845 #define vld3_lane_u8(ptr, src, lane) vld3_lane_u8_ptr(ptr, &src, lane)
10847 //uint16x4x3_t vld3_lane_u16(__transfersize(3) uint16_t const * ptr, uint16x4x3_t src, __constrange(0,3)int lane);// VLD3.16 {d0[0], d1[0], d2[0]}, [r0]
10848 _NEON2SSE_INLINE uint16x4x3_t vld3_lane_u16_ptr(__transfersize(3) uint16_t const * ptr, uint16x4x3_t* src, __constrange(0,3) int lane) // VLD3.16 {d0[0], d1[0], d2[0]}, [r0]
10851 v.val[0] = vld1_lane_u16(ptr, src->val[0], lane);
10852 v.val[1] = vld1_lane_u16((ptr + 1), src->val[1], lane);
10853 v.val[2] = vld1_lane_u16((ptr + 2), src->val[2], lane);
10856 #define vld3_lane_u16(ptr, src, lane) vld3_lane_u16_ptr(ptr, &src, lane)
10858 //uint32x2x3_t vld3_lane_u32(__transfersize(3) uint32_t const * ptr, uint32x2x3_t src, __constrange(0,1)int lane);// VLD3.32 {d0[0], d1[0], d2[0]}, [r0]
10859 _NEON2SSE_INLINE uint32x2x3_t vld3_lane_u32_ptr(__transfersize(3) uint32_t const * ptr, uint32x2x3_t* src, __constrange(0,1) int lane) // VLD3.32 {d0[0], d1[0], d2[0]}, [r0]
10863 v.val[0] = vld1_lane_u32(ptr, src->val[0], lane);;
10864 v.val[1] = vld1_lane_u32((ptr + 1), src->val[1], lane);;
10865 v.val[2] = vld1_lane_u32((ptr + 2), src->val[2], lane);;
10868 #define vld3_lane_u32(ptr, src, lane) vld3_lane_u32_ptr(ptr, &src, lane)
10870 int8x8x3_t vld3_lane_s8_ptr(__transfersize(3) int8_t const * ptr, int8x8x3_t * src, __constrange(0,7) int lane); // VLD3.8 {d0[0], d1[0], d2[0]}, [r0]
10871 #define vld3_lane_s8(ptr, src, lane) vld3_lane_u8_ptr(( uint8_t*) ptr, &src, lane)
10873 int16x4x3_t vld3_lane_s16_ptr(__transfersize(3) int16_t const * ptr, int16x4x3_t * src, __constrange(0,3) int lane); // VLD3.16 {d0[0], d1[0], d2[0]}, [r0]
10874 #define vld3_lane_s16(ptr, src, lane) vld3_lane_u16_ptr(( uint16_t*) ptr, &src, lane)
10876 int32x2x3_t vld3_lane_s32_ptr(__transfersize(3) int32_t const * ptr, int32x2x3_t * src, __constrange(0,1) int lane); // VLD3.32 {d0[0], d1[0], d2[0]}, [r0]
10877 #define vld3_lane_s32(ptr, src, lane) vld3_lane_u32_ptr(( uint32_t*) ptr, &src, lane)
10879 float16x4x3_t vld3_lane_f16_ptr(__transfersize(3) __fp16 const * ptr, float16x4x3_t * src, __constrange(0,3) int lane); // VLD3.16 {d0[0], d1[0], d2[0]}, [r0]
10882 //float32x2x3_t vld3_lane_f32(__transfersize(3) float32_t const * ptr, float32x2x3_t src,__constrange(0,1) int lane);// VLD3.32 {d0[0], d1[0], d2[0]}, [r0]
10883 _NEON2SSE_INLINE float32x2x3_t vld3_lane_f32_ptr(__transfersize(3) float32_t const * ptr, float32x2x3_t* src,__constrange(0,1) int lane) // VLD3.32 {d0[0], d1[0], d2[0]}, [r0]
10886 v.val[0] = vld1_lane_f32(ptr, src->val[0], lane);
10887 v.val[1] = vld1_lane_f32((ptr + 1), src->val[1], lane);
10888 v.val[2] = vld1_lane_f32((ptr + 2), src->val[2], lane);
10891 #define vld3_lane_f32(ptr,src,lane) vld3_lane_f32_ptr(ptr,&src,lane)
10893 //poly8x8x3_t vld3_lane_p8_ptr(__transfersize(3) poly8_t const * ptr, poly8x8x3_t * src, __constrange(0,7) int lane); // VLD3.8 {d0[0], d1[0], d2[0]}, [r0]
10896 //poly16x4x3_t vld3_lane_p16(__transfersize(3) poly16_t const * ptr, poly16x4x3_t * src, __constrange(0,3) int lane); // VLD3.16 {d0[0], d1[0], d2[0]}, [r0]
10899 //******************* Lane Quadruples load ***************************
10901 lane positon
10904 //uint16x8x4_t vld4q_lane_u16(__transfersize(4) uint16_t const * ptr, uint16x8x4_t src,__constrange(0,7) int lane)// VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10905 _NEON2SSE_INLINE uint16x8x4_t vld4q_lane_u16_ptr(__transfersize(4) uint16_t const * ptr, uint16x8x4_t* src,__constrange(0,7) int lane)
10908 v.val[0] = _MM_INSERT_EPI16 ( src->val[0], ptr[0], lane);
10909 v.val[1] = _MM_INSERT_EPI16 ( src->val[1], ptr[1], lane);
10910 v.val[2] = _MM_INSERT_EPI16 ( src->val[2], ptr[2], lane);
10911 v.val[3] = _MM_INSERT_EPI16 ( src->val[3], ptr[3], lane);
10914 #define vld4q_lane_u16(ptr, src, lane) vld4q_lane_u16_ptr(ptr, &src, lane)
10916 //uint32x4x4_t vld4q_lane_u32(__transfersize(4) uint32_t const * ptr, uint32x4x4_t src,__constrange(0,3) int lane)// VLD4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10917 _NEON2SSE_INLINE uint32x4x4_t vld4q_lane_u32_ptr(__transfersize(4) uint32_t const * ptr, uint32x4x4_t* src,__constrange(0,3) int lane)
10920 v.val[0] = _MM_INSERT_EPI32 ( src->val[0], ptr[0], lane);
10921 v.val[1] = _MM_INSERT_EPI32 ( src->val[1], ptr[1], lane);
10922 v.val[2] = _MM_INSERT_EPI32 ( src->val[2], ptr[2], lane);
10923 v.val[3] = _MM_INSERT_EPI32 ( src->val[3], ptr[3], lane);
10926 #define vld4q_lane_u32(ptr, src, lane) vld4q_lane_u32_ptr(ptr, &src, lane)
10928 //int16x8x4_t vld4q_lane_s16(__transfersize(4) int16_t const * ptr, int16x8x4_t src, __constrange(0,7)int lane);// VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10929 int16x8x4_t vld4q_lane_s16_ptr(__transfersize(4) int16_t const * ptr, int16x8x4_t * src, __constrange(0,7) int lane); // VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10930 #define vld4q_lane_s16(ptr, src, lane) vld4q_lane_u16(( uint16_t*) ptr, src, lane)
10932 //int32x4x4_t vld4q_lane_s32(__transfersize(4) int32_t const * ptr, int32x4x4_t src, __constrange(0,3)int lane);// VLD4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10933 int32x4x4_t vld4q_lane_s32_ptr(__transfersize(4) int32_t const * ptr, int32x4x4_t * src, __constrange(0,3) int lane); // VLD4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10934 #define vld4q_lane_s32(ptr, src, lane) vld4q_lane_u32(( uint32_t*) ptr, src, lane)
10936 //float16x8x4_t vld4q_lane_f16(__transfersize(4) __fp16 const * ptr, float16x8x4_t src, __constrange(0,7)int lane);// VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10937 float16x8x4_t vld4q_lane_f16_ptr(__transfersize(4) __fp16 const * ptr, float16x8x4_t * src, __constrange(0,7) int lane); // VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10940 //float32x4x4_t vld4q_lane_f32(__transfersize(4) float32_t const * ptr, float32x4x4_t src,__constrange(0,3) int lane)// VLD4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10941 _NEON2SSE_INLINE float32x4x4_t vld4q_lane_f32_ptr(__transfersize(4) float32_t const * ptr, float32x4x4_t* src,__constrange(0,3) int lane)
10944 v.val[0] = vld1q_lane_f32(&ptr[0], src->val[0], lane);
10945 v.val[1] = vld1q_lane_f32(&ptr[1], src->val[1], lane);
10946 v.val[2] = vld1q_lane_f32(&ptr[2], src->val[2], lane);
10947 v.val[3] = vld1q_lane_f32(&ptr[3], src->val[3], lane);
10950 #define vld4q_lane_f32(ptr,val,lane) vld4q_lane_f32_ptr(ptr,&val,lane)
10952 //poly16x8x4_t vld4q_lane_p16(__transfersize(4) poly16_t const * ptr, poly16x8x4_t src,__constrange(0,7) int lane);// VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10953 poly16x8x4_t vld4q_lane_p16_ptr(__transfersize(4) poly16_t const * ptr, poly16x8x4_t * src,__constrange(0,7) int lane); // VLD4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
10956 //uint8x8x4_t vld4_lane_u8(__transfersize(4) uint8_t const * ptr, uint8x8x4_t src, __constrange(0,7) int lane)// VLD4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]
10957 _NEON2SSE_INLINE uint8x8x4_t vld4_lane_u8_ptr(__transfersize(4) uint8_t const * ptr, uint8x8x4_t* src, __constrange(0,7) int lane)
10960 v.val[0] = vld1_lane_u8(ptr, src->val[0], lane);
10961 v.val[1] = vld1_lane_u8((ptr + 1), src->val[1], lane);
10962 v.val[2] = vld1_lane_u8((ptr + 2), src->val[2], lane);
10963 v.val[3] = vld1_lane_u8((ptr + 3), src->val[3], lane);
10966 #define vld4_lane_u8(ptr, src, lane) vld4_lane_u8_ptr(ptr, &src, lane)
10968 //uint16x4x4_t vld4_lane_u16(__transfersize(4) uint16_t const * ptr, uint16x4x4_t src, __constrange(0,3)int lane)// VLD4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
10969 _NEON2SSE_INLINE uint16x4x4_t vld4_lane_u16_ptr(__transfersize(4) uint16_t const * ptr, uint16x4x4_t* src, __constrange(0,3) int lane)
10972 v.val[0] = vld1_lane_u16(ptr, src->val[0], lane);
10973 v.val[1] = vld1_lane_u16((ptr + 1), src->val[1], lane);
10974 v.val[2] = vld1_lane_u16((ptr + 2), src->val[2], lane);
10975 v.val[3] = vld1_lane_u16((ptr + 3), src->val[3], lane);
10978 #define vld4_lane_u16(ptr, src, lane) vld4_lane_u16_ptr(ptr, &src, lane)
10980 //uint32x2x4_t vld4_lane_u32(__transfersize(4) uint32_t const * ptr, uint32x2x4_t src, __constrange(0,1)int lane)// VLD4.32 {d0[0], d1[0], d2[0], d3[0]}, [r0]
10981 _NEON2SSE_INLINE uint32x2x4_t vld4_lane_u32_ptr(__transfersize(4) uint32_t const * ptr, uint32x2x4_t* src, __constrange(0,1) int lane)
10984 v.val[0] = vld1_lane_u32(ptr, src->val[0], lane);
10985 v.val[1] = vld1_lane_u32((ptr + 1), src->val[1], lane);
10986 v.val[2] = vld1_lane_u32((ptr + 2), src->val[2], lane);
10987 v.val[3] = vld1_lane_u32((ptr + 3), src->val[3], lane);
10990 #define vld4_lane_u32(ptr, src, lane) vld4_lane_u32_ptr(ptr, &src, lane)
10992 //int8x8x4_t vld4_lane_s8(__transfersize(4) int8_t const * ptr, int8x8x4_t src, __constrange(0,7) int lane);// VLD4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]
10993 int8x8x4_t vld4_lane_s8_ptr(__transfersize(4) int8_t const * ptr, int8x8x4_t * src, __constrange(0,7) int lane);
10994 #define vld4_lane_s8(ptr,src,lane) vld4_lane_u8((uint8_t*)ptr,src,lane)
10996 //int16x4x4_t vld4_lane_s16(__transfersize(4) int16_t const * ptr, int16x4x4_t src, __constrange(0,3) int lane);// VLD4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
10997 int16x4x4_t vld4_lane_s16_ptr(__transfersize(4) int16_t const * ptr, int16x4x4_t * src, __constrange(0,3) int lane);
10998 #define vld4_lane_s16(ptr,src,lane) vld4_lane_u16((uint16_t*)ptr,src,lane)
11000 //int32x2x4_t vld4_lane_s32(__transfersize(4) int32_t const * ptr, int32x2x4_t src, __constrange(0,1) int lane);// VLD4.32 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11001 int32x2x4_t vld4_lane_s32_ptr(__transfersize(4) int32_t const * ptr, int32x2x4_t * src, __constrange(0,1) int lane);
11002 #define vld4_lane_s32(ptr,src,lane) vld4_lane_u32((uint32_t*)ptr,src,lane)
11004 //float16x4x4_t vld4_lane_f16(__transfersize(4) __fp16 const * ptr, float16x4x4_t src, __constrange(0,3)int lane);// VLD4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11005 float16x4x4_t vld4_lane_f16_ptr(__transfersize(4) __fp16 const * ptr, float16x4x4_t * src, __constrange(0,3) int lane);
11008 //float32x2x4_t vld4_lane_f32(__transfersize(4) float32_t const * ptr, float32x2x4_t src,__constrange(0,1) int lane)// VLD4.32 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11009 _NEON2SSE_INLINE float32x2x4_t vld4_lane_f32_ptr(__transfersize(4) float32_t const * ptr, float32x2x4_t* src,__constrange(0,1) int lane)
11013 v.val[0] = vld1_lane_f32(ptr, src->val[0], lane);
11014 v.val[1] = vld1_lane_f32((ptr + 1), src->val[1], lane);
11015 v.val[2] = vld1_lane_f32((ptr + 2), src->val[2], lane);
11016 v.val[3] = vld1_lane_f32((ptr + 3), src->val[3], lane);
11019 #define vld4_lane_f32(ptr,src,lane) vld4_lane_f32_ptr(ptr,&src,lane)
11021 //poly8x8x4_t vld4_lane_p8(__transfersize(4) poly8_t const * ptr, poly8x8x4_t src, __constrange(0,7) int lane);// VLD4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11022 poly8x8x4_t vld4_lane_p8_ptr(__transfersize(4) poly8_t const * ptr, poly8x8x4_t * src, __constrange(0,7) int lane);
11025 //poly16x4x4_t vld4_lane_p16(__transfersize(4) poly16_t const * ptr, poly16x4x4_t src, __constrange(0,3)int lane);// VLD4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11026 poly16x4x4_t vld4_lane_p16_ptr(__transfersize(4) poly16_t const * ptr, poly16x4x4_t * src, __constrange(0,3) int lane);
11602 //*********** Store a lane of a vector into memory (extract given lane) for a couple of vectors *********************
11604 //void vst2q_lane_u16(__transfersize(2) uint16_t * ptr, uint16x8x2_t val, __constrange(0,7) int lane)// VST2.16 {d0[0], d2[0]}, [r0]
11605 _NEON2SSE_INLINE void vst2q_lane_u16_ptr(__transfersize(2) uint16_t * ptr, uint16x8x2_t* val, __constrange(0,7) int lane)
11607 vst1q_lane_s16(ptr, val->val[0], lane);
11608 vst1q_lane_s16((ptr + 1), val->val[1], lane);
11610 #define vst2q_lane_u16(ptr, val, lane) vst2q_lane_u16_ptr(ptr, &val, lane)
11612 //void vst2q_lane_u32(__transfersize(2) uint32_t * ptr, uint32x4x2_t val, __constrange(0,3) int lane)// VST2.32 {d0[0], d2[0]}, [r0]
11613 _NEON2SSE_INLINE void vst2q_lane_u32_ptr(__transfersize(2) uint32_t* ptr, uint32x4x2_t* val, __constrange(0,3) int lane)
11615 vst1q_lane_u32(ptr, val->val[0], lane);
11616 vst1q_lane_u32((ptr + 1), val->val[1], lane);
11618 #define vst2q_lane_u32(ptr, val, lane) vst2q_lane_u32_ptr(ptr, &val, lane)
11620 //void vst2q_lane_s16(__transfersize(2) int16_t * ptr, int16x8x2_t val, __constrange(0,7) int lane);// VST2.16 {d0[0], d2[0]}, [r0]
11621 void vst2q_lane_s16_ptr(__transfersize(2) int16_t * ptr, int16x8x2_t * val, __constrange(0,7) int lane);
11622 #define vst2q_lane_s16(ptr, val, lane) vst2q_lane_u16((uint16_t*)ptr, val, lane)
11624 //void vst2q_lane_s32(__transfersize(2) int32_t * ptr, int32x4x2_t val, __constrange(0,3) int lane);// VST2.32 {d0[0], d2[0]}, [r0]
11625 void vst2q_lane_s32_ptr(__transfersize(2) int32_t * ptr, int32x4x2_t * val, __constrange(0,3) int lane);
11626 #define vst2q_lane_s32(ptr, val, lane) vst2q_lane_u32((uint32_t*)ptr, val, lane)
11628 //void vst2q_lane_f16(__transfersize(2) __fp16 * ptr, float16x8x2_t val, __constrange(0,7) int lane);// VST2.16 {d0[0], d2[0]}, [r0]
11629 void vst2q_lane_f16_ptr(__transfersize(2) __fp16 * ptr, float16x8x2_t * val, __constrange(0,7) int lane);
11632 //void vst2q_lane_f32(__transfersize(2) float32_t * ptr, float32x4x2_t val, __constrange(0,3) int lane)// VST2.32 {d0[0], d2[0]}, [r0]
11633 _NEON2SSE_INLINE void vst2q_lane_f32_ptr(__transfersize(2) float32_t* ptr, float32x4x2_t* val, __constrange(0,3) int lane)
11635 vst1q_lane_f32(ptr, val->val[0], lane);
11636 vst1q_lane_f32((ptr + 1), val->val[1], lane);
11638 #define vst2q_lane_f32(ptr,src,lane) vst2q_lane_f32_ptr(ptr,&src,lane)
11640 //void vst2q_lane_p16(__transfersize(2) poly16_t * ptr, poly16x8x2_t val, __constrange(0,7) int lane);// VST2.16 {d0[0], d2[0]}, [r0]
11641 void vst2q_lane_p16_ptr(__transfersize(2) poly16_t * ptr, poly16x8x2_t * val, __constrange(0,7) int lane);
11644 //void vst2_lane_u8(__transfersize(2) uint8_t * ptr, uint8x8x2_t val, __constrange(0,7) int lane);// VST2.8 {d0[0], d1[0]}, [r0]
11645 void vst2_lane_u8_ptr(__transfersize(2) uint8_t * ptr, uint8x8x2_t * val, __constrange(0,7) int lane); // VST2.8 {d0[0], d1[0]}, [r0]
11646 _NEON2SSE_INLINE void vst2_lane_u8_ptr(__transfersize(2) uint8_t * ptr, uint8x8x2_t* val, __constrange(0,7) int lane) // VST2.8 {d0[0], d1[0]}, [r0]
11648 *(ptr) = val->val[0].m64_u8[lane];
11649 *(ptr + 1) = val->val[1].m64_u8[lane];
11651 #define vst2_lane_u8(ptr, val, lane) vst2_lane_u8_ptr(ptr, &val, lane)
11653 //void vst2_lane_u16(__transfersize(2) uint16_t * ptr, uint16x4x2_t val, __constrange(0,3) int lane);// VST2.16 {d0[0], d1[0]}, [r0]
11654 void vst2_lane_u16_ptr(__transfersize(2) uint16_t * ptr, uint16x4x2_t * val, __constrange(0,3) int lane); // VST2.16 {d0[0], d1[0]}, [r0]
11655 _NEON2SSE_INLINE void vst2_lane_u16_ptr(__transfersize(2) uint16_t * ptr, uint16x4x2_t * val, __constrange(0,3) int lane)
11657 *(ptr) = val->val[0].m64_u16[lane];
11658 *(ptr + 1) = val->val[1].m64_u16[lane];
11660 #define vst2_lane_u16(ptr, val, lane) vst2_lane_u16_ptr(ptr, &val, lane)
11662 //void vst2_lane_u32(__transfersize(2) uint32_t * ptr, uint32x2x2_t val, __constrange(0,1) int lane);// VST2.32 {d0[0], d1[0]}, [r0]
11663 void vst2_lane_u32_ptr(__transfersize(2) uint32_t * ptr, uint32x2x2_t * val, __constrange(0,1) int lane); // VST2.32 {d0[0], d1[0]}, [r0]
11664 _NEON2SSE_INLINE void vst2_lane_u32_ptr(__transfersize(2) uint32_t * ptr, uint32x2x2_t * val, __constrange(0,1) int lane)
11666 *(ptr) = val->val[0].m64_u32[lane];
11667 *(ptr + 1) = val->val[1].m64_u32[lane];
11669 #define vst2_lane_u32(ptr, val, lane) vst2_lane_u32_ptr(ptr, &val, lane)
11671 //void vst2_lane_s8(__transfersize(2) int8_t * ptr, int8x8x2_t val, __constrange(0,7) int lane);// VST2.8 {d0[0], d1[0]}, [r0]
11672 void vst2_lane_s8_ptr(__transfersize(2) int8_t * ptr, int8x8x2_t * val, __constrange(0,7) int lane);
11673 #define vst2_lane_s8(ptr, val, lane) vst2_lane_u8((uint8_t*)ptr, val, lane)
11675 //void vst2_lane_s16(__transfersize(2) int16_t * ptr, int16x4x2_t val, __constrange(0,3) int lane);// VST2.16 {d0[0], d1[0]}, [r0]
11676 void vst2_lane_s16_ptr(__transfersize(2) int16_t * ptr, int16x4x2_t * val, __constrange(0,3) int lane);
11677 #define vst2_lane_s16(ptr, val, lane) vst2_lane_u16((uint16_t*)ptr, val, lane)
11679 //void vst2_lane_s32(__transfersize(2) int32_t * ptr, int32x2x2_t val, __constrange(0,1) int lane);// VST2.32 {d0[0], d1[0]}, [r0]
11680 void vst2_lane_s32_ptr(__transfersize(2) int32_t * ptr, int32x2x2_t * val, __constrange(0,1) int lane);
11681 #define vst2_lane_s32(ptr, val, lane) vst2_lane_u32((uint32_t*)ptr, val, lane)
11683 //void vst2_lane_f16(__transfersize(2) __fp16 * ptr, float16x4x2_t val, __constrange(0,3) int lane); // VST2.16 {d0[0], d1[0]}, [r0]
11686 void vst2_lane_f32_ptr(__transfersize(2) float32_t * ptr, float32x2x2_t * val, __constrange(0,1) int lane); // VST2.32 {d0[0], d1[0]}, [r0]
11687 _NEON2SSE_INLINE void vst2_lane_f32_ptr(__transfersize(2) float32_t * ptr, float32x2x2_t * val, __constrange(0,1) int lane)
11689 *(ptr) = val->val[0].m64_f32[lane];
11690 *(ptr + 1) = val->val[1].m64_f32[lane];
11692 #define vst2_lane_f32(ptr,src,lane) vst2_lane_f32_ptr(ptr,&src,lane)
11694 //void vst2_lane_p8(__transfersize(2) poly8_t * ptr, poly8x8x2_t val, __constrange(0,7) int lane);// VST2.8 {d0[0], d1[0]}, [r0]
11697 //void vst2_lane_p16(__transfersize(2) poly16_t * ptr, poly16x4x2_t val, __constrange(0,3) int lane);// VST2.16 {d0[0], d1[0]}, [r0]
11702 //void vst3q_lane_u16(__transfersize(3) uint16_t * ptr, uint16x8x3_t val, __constrange(0,7) int lane)// VST3.16 {d0[0], d2[0], d4[0]}, [r0]
11703 _NEON2SSE_INLINE void vst3q_lane_u16_ptr(__transfersize(3) uint16_t * ptr, uint16x8x3_t* val, __constrange(0,7) int lane)
11705 vst2q_lane_u16_ptr(ptr, (uint16x8x2_t*)val, lane);
11706 vst1q_lane_u16((ptr + 2), val->val[2], lane);
11708 #define vst3q_lane_u16(ptr, val, lane) vst3q_lane_u16_ptr(ptr, &val, lane)
11710 //void vst3q_lane_u32(__transfersize(3) uint32_t * ptr, uint32x4x3_t val, __constrange(0,3) int lane)// VST3.32 {d0[0], d2[0], d4[0]}, [r0]
11711 _NEON2SSE_INLINE void vst3q_lane_u32_ptr(__transfersize(3) uint32_t * ptr, uint32x4x3_t* val, __constrange(0,3) int lane)
11713 vst2q_lane_u32_ptr(ptr, (uint32x4x2_t*)val, lane);
11714 vst1q_lane_u32((ptr + 2), val->val[2], lane);
11716 #define vst3q_lane_u32(ptr, val, lane) vst3q_lane_u32_ptr(ptr, &val, lane)
11718 //void vst3q_lane_s16(__transfersize(3) int16_t * ptr, int16x8x3_t val, __constrange(0,7) int lane);// VST3.16 {d0[0], d2[0], d4[0]}, [r0]
11719 void vst3q_lane_s16_ptr(__transfersize(3) int16_t * ptr, int16x8x3_t * val, __constrange(0,7) int lane);
11720 #define vst3q_lane_s16(ptr, val, lane) vst3q_lane_u16((uint16_t *)ptr, val, lane)
11722 //void vst3q_lane_s32(__transfersize(3) int32_t * ptr, int32x4x3_t val, __constrange(0,3) int lane);// VST3.32 {d0[0], d2[0], d4[0]}, [r0]
11723 void vst3q_lane_s32_ptr(__transfersize(3) int32_t * ptr, int32x4x3_t * val, __constrange(0,3) int lane);
11724 #define vst3q_lane_s32(ptr, val, lane) vst3q_lane_u32((uint32_t *)ptr, val, lane)
11726 //void vst3q_lane_f16(__transfersize(3) __fp16 * ptr, float16x8x3_t val, __constrange(0,7) int lane);// VST3.16 {d0[0], d2[0], d4[0]}, [r0]
11727 void vst3q_lane_f16_ptr(__transfersize(3) __fp16 * ptr, float16x8x3_t * val, __constrange(0,7) int lane);
11730 //void vst3q_lane_f32(__transfersize(3) float32_t * ptr, float32x4x3_t val, __constrange(0,3) int lane)// VST3.32 {d0[0], d2[0], d4[0]}, [r0]
11731 _NEON2SSE_INLINE void vst3q_lane_f32_ptr(__transfersize(3) float32_t * ptr, float32x4x3_t* val, __constrange(0,3) int lane)
11733 vst1q_lane_f32(ptr, val->val[0], lane);
11734 vst1q_lane_f32((ptr + 1), val->val[1], lane);
11735 vst1q_lane_f32((ptr + 2), val->val[2], lane);
11737 #define vst3q_lane_f32(ptr,val,lane) vst3q_lane_f32_ptr(ptr,&val,lane)
11739 //void vst3q_lane_p16(__transfersize(3) poly16_t * ptr, poly16x8x3_t val, __constrange(0,7) int lane);// VST3.16 {d0[0], d2[0], d4[0]}, [r0]
11740 void vst3q_lane_p16_ptr(__transfersize(3) poly16_t * ptr, poly16x8x3_t * val, __constrange(0,7) int lane);
11743 //void vst3_lane_u8(__transfersize(3) uint8_t * ptr, uint8x8x3_t val, __constrange(0,7) int lane)// VST3.8 {d0[0], d1[0], d2[0]}, [r0]
11744 _NEON2SSE_INLINE void vst3_lane_u8_ptr(__transfersize(3) uint8_t * ptr, uint8x8x3_t* val, __constrange(0,7) int lane)
11746 *(ptr) = val->val[0].m64_u8[lane];
11747 *(ptr + 1) = val->val[1].m64_u8[lane];
11748 *(ptr + 2) = val->val[2].m64_u8[lane];
11750 #define vst3_lane_u8(ptr, val, lane) vst3_lane_u8_ptr(ptr, &val, lane)
11752 //void vst3_lane_u16(__transfersize(3) uint16_t * ptr, uint16x4x3_t val, __constrange(0,3) int lane)// VST3.16 {d0[0], d1[0], d2[0]}, [r0]
11753 _NEON2SSE_INLINE void vst3_lane_u16_ptr(__transfersize(3) uint16_t * ptr, uint16x4x3_t* val, __constrange(0,3) int lane)
11755 *(ptr) = val->val[0].m64_u16[lane];
11756 *(ptr + 1) = val->val[1].m64_u16[lane];
11757 *(ptr + 2) = val->val[2].m64_u16[lane];
11759 #define vst3_lane_u16(ptr, val, lane) vst3_lane_u16_ptr(ptr, &val, lane)
11761 //void vst3_lane_u32(__transfersize(3) uint32_t * ptr, uint32x2x3_t val, __constrange(0,1) int lane)// VST3.32 {d0[0], d1[0], d2[0]}, [r0]
11762 _NEON2SSE_INLINE void vst3_lane_u32_ptr(__transfersize(3) uint32_t * ptr, uint32x2x3_t* val, __constrange(0,1) int lane)
11764 *(ptr) = val->val[0].m64_u32[lane];
11765 *(ptr + 1) = val->val[1].m64_u32[lane];
11766 *(ptr + 2) = val->val[2].m64_u32[lane];
11768 #define vst3_lane_u32(ptr, val, lane) vst3_lane_u32_ptr(ptr, &val, lane)
11770 //void vst3_lane_s8(__transfersize(3) int8_t * ptr, int8x8x3_t val, __constrange(0,7) int lane);// VST3.8 {d0[0], d1[0], d2[0]}, [r0]
11771 void vst3_lane_s8_ptr(__transfersize(3) int8_t * ptr, int8x8x3_t * val, __constrange(0,7) int lane);
11772 #define vst3_lane_s8(ptr, val, lanelane)
11774 //void vst3_lane_s16(__transfersize(3) int16_t * ptr, int16x4x3_t val, __constrange(0,3) int lane);// VST3.16 {d0[0], d1[0], d2[0]}, [r0]
11775 void vst3_lane_s16_ptr(__transfersize(3) int16_t * ptr, int16x4x3_t * val, __constrange(0,3) int lane);
11776 #define vst3_lane_s16(ptr, val, lane) vst3_lane_u16((uint16_t *)ptr, val, lane)
11778 //void vst3_lane_s32(__transfersize(3) int32_t * ptr, int32x2x3_t val, __constrange(0,1) int lane);// VST3.32 {d0[0], d1[0], d2[0]}, [r0]
11779 void vst3_lane_s32_ptr(__transfersize(3) int32_t * ptr, int32x2x3_t * val, __constrange(0,1) int lane);
11780 #define vst3_lane_s32(ptr, val, lane) vst3_lane_u32((uint32_t *)ptr, val, lane)
11782 //void vst3_lane_f16(__transfersize(3) __fp16 * ptr, float16x4x3_t val, __constrange(0,3) int lane);// VST3.16 {d0[0], d1[0], d2[0]}, [r0]
11783 void vst3_lane_f16_ptr(__transfersize(3) __fp16 * ptr, float16x4x3_t * val, __constrange(0,3) int lane);
11786 //void vst3_lane_f32(__transfersize(3) float32_t * ptr, float32x2x3_t val, __constrange(0,1) int lane)// VST3.32 {d0[0], d1[0], d2[0]}, [r0]
11787 void vst3_lane_f32_ptr(__transfersize(3) float32_t * ptr, float32x2x3_t * val, __constrange(0,1) int lane);
11788 _NEON2SSE_INLINE void vst3_lane_f32_ptr(__transfersize(3) float32_t * ptr, float32x2x3_t * val, __constrange(0,1) int lane)
11790 *(ptr) = val->val[0].m64_f32[lane];
11791 *(ptr + 1) = val->val[1].m64_f32[lane];
11792 *(ptr + 2) = val->val[2].m64_f32[lane];
11794 #define vst3_lane_f32(ptr,val,lane) vst3_lane_f32_ptr(ptr,&val,lane)
11796 //void vst3_lane_p8(__transfersize(3) poly8_t * ptr, poly8x8x3_t val, __constrange(0,7) int lane);// VST3.8 {d0[0], d1[0], d2[0]}, [r0]
11797 void vst3_lane_p8_ptr(__transfersize(3) poly8_t * ptr, poly8x8x3_t * val, __constrange(0,7) int lane);
11800 //void vst3_lane_p16(__transfersize(3) poly16_t * ptr, poly16x4x3_t val, __constrange(0,3) int lane);// VST3.16 {d0[0], d1[0], d2[0]}, [r0]
11801 void vst3_lane_p16_ptr(__transfersize(3) poly16_t * ptr, poly16x4x3_t * val, __constrange(0,3) int lane);
11806 //void vst4q_lane_u16(__transfersize(4) uint16_t * ptr, uint16x8x4_t val, __constrange(0,7) int lane)// VST4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
11807 _NEON2SSE_INLINE void vst4q_lane_u16_ptr(__transfersize(4) uint16_t * ptr, uint16x8x4_t* val4, __constrange(0,7) int lane)
11809 vst2q_lane_u16_ptr(ptr, (uint16x8x2_t*)val4->val, lane);
11810 vst2q_lane_u16_ptr((ptr + 2),((uint16x8x2_t*)val4->val + 1), lane);
11812 #define vst4q_lane_u16(ptr, val, lane) vst4q_lane_u16_ptr(ptr, &val, lane)
11814 //void vst4q_lane_u32(__transfersize(4) uint32_t * ptr, uint32x4x4_t val, __constrange(0,3) int lane)// VST4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
11815 _NEON2SSE_INLINE void vst4q_lane_u32_ptr(__transfersize(4) uint32_t * ptr, uint32x4x4_t* val4, __constrange(0,3) int lane)
11817 vst2q_lane_u32_ptr(ptr, (uint32x4x2_t*)val4->val, lane);
11818 vst2q_lane_u32_ptr((ptr + 2), ((uint32x4x2_t*)val4->val + 1), lane);
11820 #define vst4q_lane_u32(ptr, val, lane) vst4q_lane_u32_ptr(ptr, &val, lane)
11822 //void vst4q_lane_s16(__transfersize(4) int16_t * ptr, int16x8x4_t val, __constrange(0,7) int lane);// VST4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
11823 void vst4q_lane_s16_ptr(__transfersize(4) int16_t * ptr, int16x8x4_t * val, __constrange(0,7) int lane);
11824 #define vst4q_lane_s16(ptr,val,lane) vst4q_lane_u16((uint16_t *)ptr,val,lane)
11826 //void vst4q_lane_s32(__transfersize(4) int32_t * ptr, int32x4x4_t val, __constrange(0,3) int lane);// VST4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
11827 void vst4q_lane_s32_ptr(__transfersize(4) int32_t * ptr, int32x4x4_t * val, __constrange(0,3) int lane);
11828 #define vst4q_lane_s32(ptr,val,lane) vst4q_lane_u32((uint32_t *)ptr,val,lane)
11830 //void vst4q_lane_f16(__transfersize(4) __fp16 * ptr, float16x8x4_t val, __constrange(0,7) int lane);// VST4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
11831 void vst4q_lane_f16_ptr(__transfersize(4) __fp16 * ptr, float16x8x4_t * val, __constrange(0,7) int lane);
11834 //void vst4q_lane_f32(__transfersize(4) float32_t * ptr, float32x4x4_t val, __constrange(0,3) int lane)// VST4.32 {d0[0], d2[0], d4[0], d6[0]}, [r0]
11835 _NEON2SSE_INLINE void vst4q_lane_f32_ptr(__transfersize(4) float32_t * ptr, float32x4x4_t* val, __constrange(0,3) int lane)
11837 vst1q_lane_f32(ptr, val->val[0], lane);
11838 vst1q_lane_f32((ptr + 1), val->val[1], lane);
11839 vst1q_lane_f32((ptr + 2), val->val[2], lane);
11840 vst1q_lane_f32((ptr + 3), val->val[3], lane);
11842 #define vst4q_lane_f32(ptr,val,lane) vst4q_lane_f32_ptr(ptr,&val,lane)
11844 //void vst4q_lane_p16(__transfersize(4) poly16_t * ptr, poly16x8x4_t val, __constrange(0,7) int lane);// VST4.16 {d0[0], d2[0], d4[0], d6[0]}, [r0]
11845 void vst4q_lane_p16_ptr(__transfersize(4) poly16_t * ptr, poly16x8x4_t * val, __constrange(0,7) int lane);
11848 //void vst4_lane_u8(__transfersize(4) uint8_t * ptr, uint8x8x4_t val, __constrange(0,7) int lane)// VST4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11849 _NEON2SSE_INLINE void vst4_lane_u8_ptr(__transfersize(4) uint8_t * ptr, uint8x8x4_t* val, __constrange(0,7) int lane)
11851 *(ptr) = val->val[0].m64_u8[lane];
11852 *(ptr + 1) = val->val[1].m64_u8[lane];
11853 *(ptr + 2) = val->val[2].m64_u8[lane];
11854 *(ptr + 3) = val->val[3].m64_u8[lane];
11856 #define vst4_lane_u8(ptr, val, lane) vst4_lane_u8_ptr(ptr, &val, lane)
11858 //void vst4_lane_u16(__transfersize(4) uint16_t * ptr, uint16x4x4_t val, __constrange(0,3) int lane)// VST4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11859 _NEON2SSE_INLINE void vst4_lane_u16_ptr(__transfersize(4) uint16_t * ptr, uint16x4x4_t* val, __constrange(0,3) int lane)
11861 *(ptr) = val->val[0].m64_u16[lane];
11862 *(ptr + 1) = val->val[1].m64_u16[lane];
11863 *(ptr + 2) = val->val[2].m64_u16[lane];
11864 *(ptr + 3) = val->val[3].m64_u16[lane];
11866 #define vst4_lane_u16(ptr, val, lane) vst4_lane_u16_ptr(ptr, &val, lane)
11868 //void vst4_lane_u32(__transfersize(4) uint32_t * ptr, uint32x2x4_t val, __constrange(0,1) int lane)// VST4.32 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11869 _NEON2SSE_INLINE void vst4_lane_u32_ptr(__transfersize(4) uint32_t * ptr, uint32x2x4_t* val, __constrange(0,1) int lane)
11871 *(ptr) = val->val[0].m64_u32[lane];
11872 *(ptr + 1) = val->val[1].m64_u32[lane];
11873 *(ptr + 2) = val->val[2].m64_u32[lane];
11874 *(ptr + 3) = val->val[3].m64_u32[lane];
11876 #define vst4_lane_u32(ptr, val, lane) vst4_lane_u32_ptr(ptr, &val, lane)
11878 //void vst4_lane_s8(__transfersize(4) int8_t * ptr, int8x8x4_t val, __constrange(0,7) int lane)// VST4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11879 #define vst4_lane_s8(ptr, val, lane) vst4_lane_u8((uint8_t*)ptr, val, lane)
11881 //void vst4_lane_s16(__transfersize(4) int16_t * ptr, int16x4x4_t val, __constrange(0,3) int lane)// VST4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11882 #define vst4_lane_s16(ptr, val, lane) vst4_lane_u16((uint16_t*)ptr, val, lane)
11884 //void vst4_lane_s32(__transfersize(4) int32_t * ptr, int32x2x4_t val, __constrange(0,1) int lane)// VST4.32 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11885 #define vst4_lane_s32(ptr, val, lane) vst4_lane_u32((uint32_t*)ptr, val, lane)
11887 //void vst4_lane_f16(__transfersize(4) __fp16 * ptr, float16x4x4_t val, __constrange(0,3) int lane);// VST4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11888 void vst4_lane_f16_ptr(__transfersize(4) __fp16 * ptr, float16x4x4_t * val, __constrange(0,3) int lane);
11891 void vst4_lane_f32_ptr(__transfersize(4) float32_t * ptr, float32x2x4_t * val, __constrange(0,1) int lane); // VST4.32 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11892 _NEON2SSE_INLINE void vst4_lane_f32_ptr(__transfersize(4) float32_t * ptr, float32x2x4_t* val, __constrange(0,1) int lane)
11894 *(ptr) = val->val[0].m64_f32[lane];
11895 *(ptr + 1) = val->val[1].m64_f32[lane];
11896 *(ptr + 2) = val->val[2].m64_f32[lane];
11897 *(ptr + 3) = val->val[3].m64_f32[lane];
11899 #define vst4_lane_f32(ptr,val,lane) vst4_lane_f32_ptr(ptr,&val,lane)
11901 //void vst4_lane_p8(__transfersize(4) poly8_t * ptr, poly8x8x4_t val, __constrange(0,7) int lane);// VST4.8 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11902 void vst4_lane_p8_ptr(__transfersize(4) poly8_t * ptr, poly8x8x4_t * val, __constrange(0,7) int lane);
11905 //void vst4_lane_p16(__transfersize(4) poly16_t * ptr, poly16x4x4_t val, __constrange(0,3) int lane);// VST4.16 {d0[0], d1[0], d2[0], d3[0]}, [r0]
11906 void vst4_lane_p16_ptr(__transfersize(4) poly16_t * ptr, poly16x4x4_t * val, __constrange(0,3) int lane);
11912 //These intrinsics extract a single lane (element) from a vector.
11913 uint8_t vget_lane_u8(uint8x8_t vec, __constrange(0,7) int lane); // VMOV.U8 r0, d0[0]
11914 #define vget_lane_u8(vec, lane) vec.m64_u8[lane]
11916 uint16_t vget_lane_u16(uint16x4_t vec, __constrange(0,3) int lane); // VMOV.s16 r0, d0[0]
11917 #define vget_lane_u16(vec, lane) vec.m64_u16[lane]
11920 uint32_t vget_lane_u32(uint32x2_t vec, __constrange(0,1) int lane); // VMOV.32 r0, d0[0]
11921 #define vget_lane_u32(vec, lane) vec.m64_u32[lane]
11923 int8_t vget_lane_s8(int8x8_t vec, __constrange(0,7) int lane); // VMOV.S8 r0, d0[0]
11924 #define vget_lane_s8(vec, lane) vec.m64_i8[lane]
11926 int16_t vget_lane_s16(int16x4_t vec, __constrange(0,3) int lane); // VMOV.S16 r0, d0[0]
11927 #define vget_lane_s16(vec, lane) vec.m64_i16[lane]
11929 int32_t vget_lane_s32(int32x2_t vec, __constrange(0,1) int lane); // VMOV.32 r0, d0[0]
11930 #define vget_lane_s32(vec, lane) vec.m64_i32[lane]
11932 poly8_t vget_lane_p8(poly8x8_t vec, __constrange(0,7) int lane); // VMOV.U8 r0, d0[0]
11935 poly16_t vget_lane_p16(poly16x4_t vec, __constrange(0,3) int lane); // VMOV.s16 r0, d0[0]
11938 float32_t vget_lane_f32(float32x2_t vec, __constrange(0,1) int lane); // VMOV.32 r0, d0[0]
11939 #define vget_lane_f32(vec, lane) vec.m64_f32[lane]
11941 uint8_t vgetq_lane_u8(uint8x16_t vec, __constrange(0,15) int lane); // VMOV.U8 r0, d0[0]
11944 uint16_t vgetq_lane_u16(uint16x8_t vec, __constrange(0,7) int lane); // VMOV.s16 r0, d0[0]
11947 uint32_t vgetq_lane_u32(uint32x4_t vec, __constrange(0,3) int lane); // VMOV.32 r0, d0[0]
11950 int8_t vgetq_lane_s8(int8x16_t vec, __constrange(0,15) int lane); // VMOV.S8 r0, d0[0]
11953 int16_t vgetq_lane_s16(int16x8_t vec, __constrange(0,7) int lane); // VMOV.S16 r0, d0[0]
11956 int32_t vgetq_lane_s32(int32x4_t vec, __constrange(0,3) int lane); // VMOV.32 r0, d0[0]
11959 poly8_t vgetq_lane_p8(poly8x16_t vec, __constrange(0,15) int lane); // VMOV.U8 r0, d0[0]
11962 poly16_t vgetq_lane_p16(poly16x8_t vec, __constrange(0,7) int lane); // VMOV.s16 r0, d0[0]
11965 float32_t vgetq_lane_f32(float32x4_t vec, __constrange(0,3) int lane); // VMOV.32 r0, d0[0]
11966 _NEON2SSE_INLINE float32_t vgetq_lane_f32(float32x4_t vec, __constrange(0,3) int lane)
11969 ilane = _MM_EXTRACT_PS(vec,lane);
11973 int64_t vget_lane_s64(int64x1_t vec, __constrange(0,0) int lane); // VMOV r0,r0,d0
11974 #define vget_lane_s64(vec, lane) vec.m64_i64[0]
11976 uint64_t vget_lane_u64(uint64x1_t vec, __constrange(0,0) int lane); // VMOV r0,r0,d0
11977 #define vget_lane_u64(vec, lane) vec.m64_u64[0]
11980 int64_t vgetq_lane_s64(int64x2_t vec, __constrange(0,1) int lane); // VMOV r0,r0,d0
11983 uint64_t vgetq_lane_u64(uint64x2_t vec, __constrange(0,1) int lane); // VMOV r0,r0,d0
11988 //These intrinsics set a single lane (element) within a vector.
11991 uint8x8_t vset_lane_u8(uint8_t value, uint8x8_t vec, __constrange(0,7) int lane); // VMOV.8 d0[0],r0
11992 _NEON2SSE_INLINE uint8x8_t vset_lane_u8(uint8_t value, uint8x8_t vec, __constrange(0,7) int lane)
11996 return vld1_lane_u8(&val, vec, lane);
11999 uint16x4_t vset_lane_u16(uint16_t value, uint16x4_t vec, __constrange(0,3) int lane); // VMOV.16 d0[0],r0
12000 _NEON2SSE_INLINE uint16x4_t vset_lane_u16(uint16_t value, uint16x4_t vec, __constrange(0,3) int lane)
12004 return vld1_lane_u16(&val, vec, lane);
12007 uint32x2_t vset_lane_u32(uint32_t value, uint32x2_t vec, __constrange(0,1) int lane); // VMOV.32 d0[0],r0
12008 _NEON2SSE_INLINE uint32x2_t vset_lane_u32(uint32_t value, uint32x2_t vec, __constrange(0,1) int lane)
12012 return vld1_lane_u32(&val, vec, lane);
12015 int8x8_t vset_lane_s8(int8_t value, int8x8_t vec, __constrange(0,7) int lane); // VMOV.8 d0[0],r0
12016 _NEON2SSE_INLINE int8x8_t vset_lane_s8(int8_t value, int8x8_t vec, __constrange(0,7) int lane)
12020 return vld1_lane_s8(&val, vec, lane);
12023 int16x4_t vset_lane_s16(int16_t value, int16x4_t vec, __constrange(0,3) int lane); // VMOV.16 d0[0],r0
12024 _NEON2SSE_INLINE int16x4_t vset_lane_s16(int16_t value, int16x4_t vec, __constrange(0,3) int lane)
12028 return vld1_lane_s16(&val, vec, lane);
12031 int32x2_t vset_lane_s32(int32_t value, int32x2_t vec, __constrange(0,1) int lane); // VMOV.32 d0[0],r0
12032 _NEON2SSE_INLINE int32x2_t vset_lane_s32(int32_t value, int32x2_t vec, __constrange(0,1) int lane)
12036 return vld1_lane_s32(&val, vec, lane);
12039 poly8x8_t vset_lane_p8(poly8_t value, poly8x8_t vec, __constrange(0,7) int lane); // VMOV.8 d0[0],r0
12042 poly16x4_t vset_lane_p16(poly16_t value, poly16x4_t vec, __constrange(0,3) int lane); // VMOV.16 d0[0],r0
12045 float32x2_t vset_lane_f32(float32_t value, float32x2_t vec, __constrange(0,1) int lane); // VMOV.32 d0[0],r0
12046 _NEON2SSE_INLINE float32x2_t vset_lane_f32(float32_t value, float32x2_t vec, __constrange(0,1) int lane)
12050 return vld1_lane_f32(&val, vec, lane);
12053 uint8x16_t vsetq_lane_u8(uint8_t value, uint8x16_t vec, __constrange(0,15) int lane); // VMOV.8 d0[0],r0
12054 _NEON2SSE_INLINE uint8x16_t vsetq_lane_u8(uint8_t value, uint8x16_t vec, __constrange(0,15) int lane)
12058 return vld1q_lane_u8(&val, vec, lane);
12061 uint16x8_t vsetq_lane_u16(uint16_t value, uint16x8_t vec, __constrange(0,7) int lane); // VMOV.16 d0[0],r0
12062 _NEON2SSE_INLINE uint16x8_t vsetq_lane_u16(uint16_t value, uint16x8_t vec, __constrange(0,7) int lane)
12066 return vld1q_lane_u16(&val, vec, lane);
12069 uint32x4_t vsetq_lane_u32(uint32_t value, uint32x4_t vec, __constrange(0,3) int lane); // VMOV.32 d0[0],r0
12070 _NEON2SSE_INLINE uint32x4_t vsetq_lane_u32(uint32_t value, uint32x4_t vec, __constrange(0,3) int lane)
12074 return vld1q_lane_u32(&val, vec, lane);
12077 int8x16_t vsetq_lane_s8(int8_t value, int8x16_t vec, __constrange(0,15) int lane); // VMOV.8 d0[0],r0
12078 _NEON2SSE_INLINE int8x16_t vsetq_lane_s8(int8_t value, int8x16_t vec, __constrange(0,15) int lane)
12082 return vld1q_lane_s8(&val, vec, lane);
12085 int16x8_t vsetq_lane_s16(int16_t value, int16x8_t vec, __constrange(0,7) int lane); // VMOV.16 d0[0],r0
12086 _NEON2SSE_INLINE int16x8_t vsetq_lane_s16(int16_t value, int16x8_t vec, __constrange(0,7) int lane)
12090 return vld1q_lane_s16(&val, vec, lane);
12093 int32x4_t vsetq_lane_s32(int32_t value, int32x4_t vec, __constrange(0,3) int lane); // VMOV.32 d0[0],r0
12094 _NEON2SSE_INLINE int32x4_t vsetq_lane_s32(int32_t value, int32x4_t vec, __constrange(0,3) int lane)
12098 return vld1q_lane_s32(&val, vec, lane);
12101 poly8x16_t vsetq_lane_p8(poly8_t value, poly8x16_t vec, __constrange(0,15) int lane); // VMOV.8 d0[0],r0
12104 poly16x8_t vsetq_lane_p16(poly16_t value, poly16x8_t vec, __constrange(0,7) int lane); // VMOV.16 d0[0],r0
12107 float32x4_t vsetq_lane_f32(float32_t value, float32x4_t vec, __constrange(0,3) int lane); // VMOV.32 d0[0],r0
12108 _NEON2SSE_INLINE float32x4_t vsetq_lane_f32(float32_t value, float32x4_t vec, __constrange(0,3) int lane)
12112 return vld1q_lane_f32(&val, vec, lane);
12115 int64x1_t vset_lane_s64(int64_t value, int64x1_t vec, __constrange(0,0) int lane); // VMOV d0,r0,r0
12116 _NEON2SSE_INLINE int64x1_t vset_lane_s64(int64_t value, int64x1_t vec, __constrange(0,0) int lane)
12120 return vld1_lane_s64(&val, vec, lane);
12123 uint64x1_t vset_lane_u64(uint64_t value, uint64x1_t vec, __constrange(0,0) int lane); // VMOV d0,r0,r0
12124 _NEON2SSE_INLINE uint64x1_t vset_lane_u64(uint64_t value, uint64x1_t vec, __constrange(0,0) int lane)
12128 return vld1_lane_u64(&val, vec, lane);
12131 int64x2_t vsetq_lane_s64(int64_t value, int64x2_t vec, __constrange(0,1) int lane); // VMOV d0,r0,r0
12132 _NEON2SSE_INLINE int64x2_t vsetq_lane_s64(int64_t value, int64x2_t vec, __constrange(0,1) int lane)
12136 return vld1q_lane_s64(&val, vec, lane);
12139 uint64x2_t vsetq_lane_u64(uint64_t value, uint64x2_t vec, __constrange(0,1) int lane); // VMOV d0,r0,r0
12390 //**************Set all lanes to the value of one lane of a vector *************
12392 //here shuffle is better solution than lane extraction followed by set1 function
12393 uint8x8_t vdup_lane_u8(uint8x8_t vec, __constrange(0,7) int lane); // VDUP.8 d0,d0[0]
12394 _NEON2SSE_INLINE uint8x8_t vdup_lane_u8(uint8x8_t vec, __constrange(0,7) int lane)
12399 valane = vec.m64_u8[lane];
12406 uint16x4_t vdup_lane_u16(uint16x4_t vec, __constrange(0,3) int lane); // VDUP.16 d0,d0[0]
12407 _NEON2SSE_INLINE uint16x4_t vdup_lane_u16(uint16x4_t vec, __constrange(0,3) int lane)
12411 valane = vec.m64_u16[lane];
12419 uint32x2_t vdup_lane_u32(uint32x2_t vec, __constrange(0,1) int lane); // VDUP.32 d0,d0[0]
12420 _NEON2SSE_INLINE uint32x2_t vdup_lane_u32(uint32x2_t vec, __constrange(0,1) int lane)
12423 res.m64_u32[0] = vec.m64_u32[lane];
12428 int8x8_t vdup_lane_s8(int8x8_t vec, __constrange(0,7) int lane); // VDUP.8 d0,d0[0]
12431 int16x4_t vdup_lane_s16(int16x4_t vec, __constrange(0,3) int lane); // VDUP.16 d0,d0[0]
12434 int32x2_t vdup_lane_s32(int32x2_t vec, __constrange(0,1) int lane); // VDUP.32 d0,d0[0]
12437 poly8x8_t vdup_lane_p8(poly8x8_t vec, __constrange(0,7) int lane); // VDUP.8 d0,d0[0]
12440 poly16x4_t vdup_lane_p16(poly16x4_t vec, __constrange(0,3) int lane); // VDUP.16 d0,d0[0]
12443 float32x2_t vdup_lane_f32(float32x2_t vec, __constrange(0,1) int lane); // VDUP.32 d0,d0[0]
12444 _NEON2SSE_INLINE float32x2_t vdup_lane_f32(float32x2_t vec, __constrange(0,1) int lane)
12447 res.m64_f32[0] = vec.m64_f32[lane];
12452 uint8x16_t vdupq_lane_u8(uint8x8_t vec, __constrange(0,7) int lane); // VDUP.8 q0,d0[0]
12453 _NEON2SSE_INLINE uint8x16_t vdupq_lane_u8(uint8x8_t vec, __constrange(0,7) int lane) // VDUP.8 q0,d0[0]
12455 _NEON2SSE_ALIGN_16 int8_t lanemask8[16] = {lane, lane, lane, lane, lane, lane, lane, lane, lane, lane, lane, lane, lane, lane, lane, lane};
12459 uint16x8_t vdupq_lane_u16(uint16x4_t vec, __constrange(0,3) int lane); // VDUP.16 q0,d0[0]
12460 _NEON2SSE_INLINE uint16x8_t vdupq_lane_u16(uint16x4_t vec, __constrange(0,3) int lane) // VDUP.16 q0,d0[0]
12463 const int8_t lane16 = ((int8_t) lane) << 1;
12469 uint32x4_t vdupq_lane_u32(uint32x2_t vec, __constrange(0,1) int lane); // VDUP.32 q0,d0[0]
12470 #define vdupq_lane_u32(vec, lane) _mm_shuffle_epi32 (_pM128i(vec), lane | (lane << 2) | (lane << 4) | (lane << 6))
12472 int8x16_t vdupq_lane_s8(int8x8_t vec, __constrange(0,7) int lane); // VDUP.8 q0,d0[0]
12475 int16x8_t vdupq_lane_s16(int16x4_t vec, __constrange(0,3) int lane); // VDUP.16 q0,d0[0]
12478 int32x4_t vdupq_lane_s32(int32x2_t vec, __constrange(0,1) int lane); // VDUP.32 q0,d0[0]
12481 poly8x16_t vdupq_lane_p8(poly8x8_t vec, __constrange(0,7) int lane); // VDUP.8 q0,d0[0]
12484 poly16x8_t vdupq_lane_p16(poly16x4_t vec, __constrange(0,3) int lane); // VDUP.16 q0,d0[0]
12487 float32x4_t vdupq_lane_f32(float32x2_t vec, __constrange(0,1) int lane); // VDUP.32 q0,d0[0]
12488 #define vdupq_lane_f32(vec, lane) _mm_load1_ps((vec.m64_f32 + lane))
12490 int64x1_t vdup_lane_s64(int64x1_t vec, __constrange(0,0) int lane); // VMOV d0,d0
12491 #define vdup_lane_s64(vec,lane) vec
12493 uint64x1_t vdup_lane_u64(uint64x1_t vec, __constrange(0,0) int lane); // VMOV d0,d0
12494 #define vdup_lane_u64(vec,lane) vec
12496 int64x2_t vdupq_lane_s64(int64x1_t vec, __constrange(0,0) int lane); // VMOV q0,q0
12497 _NEON2SSE_INLINE int64x2_t vdupq_lane_s64(int64x1_t vec, __constrange(0,0) int lane)
12504 uint64x2_t vdupq_lane_u64(uint64x1_t vec, __constrange(0,0) int lane); // VMOV q0,q0
13642 //********** Vector multiply lane *****************************
