Lines Matching full:neon
8 declare <2 x float> @llvm.aarch64.neon.fmax.v2f32(<2 x float>, <2 x float>) nounwind readnone
9 declare <4 x float> @llvm.aarch64.neon.fmax.v4f32(<4 x float>, <4 x float>) nounwind readnone
10 declare <2 x i32> @llvm.aarch64.neon.smax.v2i32(<2 x i32>, <2 x i32>) nounwind readnone
11 declare <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32>, <4 x i32>) nounwind readnone
12 declare <2 x i32> @llvm.aarch64.neon.umax.v2i32(<2 x i32>, <2 x i32>) nounwind readnone
13 declare <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32>, <4 x i32>) nounwind readnone
14 declare <4 x i16> @llvm.aarch64.neon.smax.v4i16(<4 x i16>, <4 x i16>) nounwind readnone
15 declare <4 x i16> @llvm.aarch64.neon.umax.v4i16(<4 x i16>, <4 x i16>) nounwind readnone
17 declare <2 x float> @llvm.aarch64.neon.fmin.v2f32(<2 x float>, <2 x float>) nounwind readnone
18 declare <4 x float> @llvm.aarch64.neon.fmin.v4f32(<4 x float>, <4 x float>) nounwind readnone
19 declare <2 x i32> @llvm.aarch64.neon.smin.v2i32(<2 x i32>, <2 x i32>) nounwind readnone
20 declare <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32>, <4 x i32>) nounwind readnone
21 declare <2 x i32> @llvm.aarch64.neon.umin.v2i32(<2 x i32>, <2 x i32>) nounwind readnone
22 declare <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32>, <4 x i32>) nounwind readnone
23 declare <4 x i16> @llvm.aarch64.neon.smin.v4i16(<4 x i16>, <4 x i16>) nounwind readnone
24 declare <4 x i16> @llvm.aarch64.neon.umin.v4i16(<4 x i16>, <4 x i16>) nounwind readnone
26 declare <8 x i8> @llvm.aarch64.neon.sqshl.v8i8(<8 x i16>, <8 x i16>) nounwind readnone
27 declare <4 x i16> @llvm.aarch64.neon.sqshl.v4i16(<4 x i32>, <4 x i32>) nounwind readnone
28 declare <2 x i32> @llvm.aarch64.neon.sqshl.v2i32(<2 x i64>, <2 x i64>) nounwind readnone
30 declare <8 x i8> @llvm.aarch64.neon.sqshrun.v8i8(<8 x i16>, <8 x i16>) nounwind readnone
31 declare <4 x i16> @llvm.aarch64.neon.sqshrun.v4i16(<4 x i32>, <4 x i32>) nounwind readnone
32 declare <2 x i32> @llvm.aarch64.neon.sqshrun.v2i32(<2 x i64>, <2 x i64>) nounwind readnone
34 declare <2 x float> @llvm.aarch64.neon.frecpe.v2f32(<2 x float>) nounwind readnone
35 declare <4 x float> @llvm.aarch64.neon.frecpe.v4f32(<4 x float>) nounwind readnone
37 declare <2 x float> @llvm.aarch64.neon.frsqrte.v2f32(<2 x float>) nounwind readnone
38 declare <4 x float> @llvm.aarch64.neon.frsqrte.v4f32(<4 x float>) nounwind readnone
40 declare <2 x float> @llvm.aarch64.neon.frecps.v2f32(<2 x float>, <2 x float>) nounwind readnone
41 declare <4 x float> @llvm.aarch64.neon.frecps.v4f32(<4 x float>, <4 x float>) nounwind readnone
43 declare <2 x float> @llvm.aarch64.neon.frsqrts.v2f32(<2 x float>, <2 x float>) nounwind readnone
44 declare <4 x float> @llvm.aarch64.neon.frsqrts.v4f32(<4 x float>, <4 x float>) nounwind readnone
109 %1 = tail call <4 x float> @llvm.aarch64.neon.fmin.v4f32(<4 x float> %value, <4 x float> %high) nounwind readnone
110 %2 = tail call <4 x float> @llvm.aarch64.neon.fmax.v4f32(<4 x float> %1, <4 x float> %low) nounwind readnone
125 %a = tail call <4 x float> @llvm.aarch64.neon.fmin.v4f32(<4 x float> %_value, <4 x float> %_high) nounwind readnone
126 %b = tail call <4 x float> @llvm.aarch64.neon.fmax.v4f32(<4 x float> %a, <4 x float> %_low) nounwind readnone
135 %a = tail call <4 x float> @llvm.aarch64.neon.fmin.v4f32(<4 x float> %_value, <4 x float> %_high) nounwind readnone
136 %b = tail call <4 x float> @llvm.aarch64.neon.fmax.v4f32(<4 x float> %a, <4 x float> %_low) nounwind readnone
142 %1 = tail call <2 x float> @llvm.aarch64.neon.fmin.v2f32(<2 x float> %value, <2 x float> %high) nounwind readnone
143 %2 = tail call <2 x float> @llvm.aarch64.neon.fmax.v2f32(<2 x float> %1, <2 x float> %low) nounwind readnone
150 %a = tail call <2 x float> @llvm.aarch64.neon.fmin.v2f32(<2 x float> %value, <2 x float> %_high) nounwind readnone
151 %b = tail call <2 x float> @llvm.aarch64.neon.fmax.v2f32(<2 x float> %a, <2 x float> %_low) nounwind readnone
166 %1 = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %value, <4 x i32> %high) nounwind readnone
167 %2 = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %1, <4 x i32> %low) nounwind readnone
174 %1 = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %value, <4 x i32> %_high) nounwind readnone
175 %2 = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %1, <4 x i32> %_low) nounwind readnone
183 %a = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %_value, <4 x i32> %_high) nounwind readnone
184 %b = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %a, <4 x i32> %_low) nounwind readnone
193 %a = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %_value, <4 x i32> %_high) nounwind readnone
194 %b = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %a, <4 x i32> %_low) nounwind readnone
200 %1 = tail call <2 x i32> @llvm.aarch64.neon.smin.v2i32(<2 x i32> %value, <2 x i32> %high) nounwind readnone
201 %2 = tail call <2 x i32> @llvm.aarch64.neon.smax.v2i32(<2 x i32> %1, <2 x i32> %low) nounwind readnone
208 %a = tail call <2 x i32> @llvm.aarch64.neon.smin.v2i32(<2 x i32> %value, <2 x i32> %_high) nounwind readnone
209 %b = tail call <2 x i32> @llvm.aarch64.neon.smax.v2i32(<2 x i32> %a, <2 x i32> %_low) nounwind readnone
216 %1 = tail call <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32> %value, <4 x i32> %high) nounwind readnone
217 %2 = tail call <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32> %1, <4 x i32> %low) nounwind readnone
224 %1 = tail call <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32> %value, <4 x i32> %_high) nounwind readnone
225 %2 = tail call <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32> %1, <4 x i32> %_low) nounwind readnone
233 %a = tail call <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32> %_value, <4 x i32> %_high) nounwind readnone
234 %b = tail call <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32> %a, <4 x i32> %_low) nounwind readnone
243 %a = tail call <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32> %_value, <4 x i32> %_high) nounwind readnone
244 %b = tail call <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32> %a, <4 x i32> %_low) nounwind readnone
250 %1 = tail call <2 x i32> @llvm.aarch64.neon.umin.v2i32(<2 x i32> %value, <2 x i32> %high) nounwind readnone
251 %2 = tail call <2 x i32> @llvm.aarch64.neon.umax.v2i32(<2 x i32> %1, <2 x i32> %low) nounwind readnone
258 %a = tail call <2 x i32> @llvm.aarch64.neon.umin.v2i32(<2 x i32> %value, <2 x i32> %_high) nounwind readnone
259 %b = tail call <2 x i32> @llvm.aarch64.neon.umax.v2i32(<2 x i32> %a, <2 x i32> %_low) nounwind readnone
269 %1 = tail call <4 x float> @llvm.aarch64.neon.fmax.v4f32(<4 x float> %v1, <4 x float> %v2) nounwind readnone
275 %2 = tail call <4 x float> @llvm.aarch64.neon.fmax.v4f32(<4 x float> %v1, <4 x float> %1) nounwind readnone
282 %3 = tail call <4 x float> @llvm.aarch64.neon.fmax.v4f32(<4 x float> %1, <4 x float> %2) nounwind readnone
290 %3 = tail call <4 x float> @llvm.aarch64.neon.fmax.v4f32(<4 x float> %1, <4 x float> %2) nounwind readnone
296 %1 = tail call <2 x float> @llvm.aarch64.neon
302 %2 = tail call <2 x float> @llvm.aarch64.neon.fmax.v2f32(<2 x float> %v1, <2 x float> %1) nounwind readnone
318 %1 = tail call <4 x float> @llvm.aarch64.neon.fmin.v4f32(<4 x float> %v1, <4 x float> %v2) nounwind readnone
324 %2 = tail call <4 x float> @llvm.aarch64.neon.fmin.v4f32(<4 x float> %v1, <4 x float> %1) nounwind readnone
331 %3 = tail call <4 x float> @llvm.aarch64.neon.fmin.v4f32(<4 x float> %1, <4 x float> %2) nounwind readnone
339 %3 = tail call <4 x float> @llvm.aarch64.neon.fmin.v4f32(<4 x float> %1, <4 x float> %2) nounwind readnone
345 %1 = tail call <2 x float> @llvm.aarch64.neon.fmin.v2f32(<2 x float> %v1, <2 x float> %v2) nounwind readnone
351 %2 = tail call <2 x float> @llvm.aarch64.neon.fmin.v2f32(<2 x float> %v1, <2 x float> %1) nounwind readnone
375 %3 = tail call <2 x i32> @llvm.aarch64.neon.smax.v2i32(<2 x i32> %1, <2 x i32> %2) nounwind readnone
385 %5 = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %3, <4 x i32> %4) nounwind readnone
394 %3 = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
408 %3 = tail call <2 x i32> @llvm.aarch64.neon.smax.v2i32(<2 x i32> %1, <2 x i32> %2) nounwind readnone
418 %5 = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %3, <4 x i32> %4) nounwind readnone
427 %3 = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
439 %1 = tail call <2 x i32> @llvm.aarch64.neon.smax.v2i32(<2 x i32> %v1, <2 x i32> %v2) nounwind readnone
446 %3 = tail call <4 x i32 > @llvm.aarch64.neon.smax.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
452 %1 = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %v1, <4 x i32> %v2) nounwind readnone
473 %3 = tail call <2 x i32> @llvm.aarch64.neon.umax.v2i32(<2 x i32> %1, <2 x i32> %2) nounwind readnone
483 %5 = tail call <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32> %3, <4 x i32> %4) nounwind readnone
492 %3 = tail call <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
506 %3 = tail call <2 x i32> @llvm.aarch64.neon.umax.v2i32(<2 x i32> %1, <2 x i32> %2) nounwind readnone
516 %5 = tail call <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32> %3, <4 x i32> %4) nounwind readnone
525 %3 = tail call <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
537 %1 = tail call <2 x i32> @llvm.aarch64.neon.umax.v2i32(<2 x i32> %v1, <2 x i32> %v2) nounwind readnone
544 %3 = tail call <4 x i32 > @llvm.aarch64.neon.umax.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
550 %1 = tail call <4 x i32> @llvm.aarch64.neon.umax.v4i32(<4 x i32> %v1, <4 x i32> %v2) nounwind readnone
606 %3 = tail call <2 x i32> @llvm.aarch64.neon.smin.v2i32(<2 x i32> %1, <2 x i32> %2) nounwind readnone
616 %5 = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %3, <4 x i32> %4) nounwind readnone
625 %3 = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
639 %3 = tail call <2 x i32> @llvm.aarch64.neon.smin.v2i32(<2 x i32> %1, <2 x i32> %2) nounwind readnone
649 %5 = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %3, <4 x i32> %4) nounwind readnone
658 %3 = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
670 %1 = tail call <2 x i32> @llvm.aarch64.neon.smin.v2i32(<2 x i32> %v1, <2 x i32> %v2) nounwind readnone
677 %3 = tail call <4 x i32 > @llvm.aarch64.neon.smin.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
683 %1 = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %v1, <4 x i32> %v2) nounwind readnone
704 %3 = tail call <2 x i32> @llvm.aarch64.neon.umin.v2i32(<2 x i32> %1, <2 x i32> %2) nounwind readnone
714 %5 = tail call <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32> %3, <4 x i32> %4) nounwind readnone
723 %3 = tail call <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
737 %3 = tail call <2 x i32> @llvm.aarch64.neon.umin.v2i32(<2 x i32> %1, <2 x i32> %2) nounwind readnone
747 %5 = tail call <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32> %3, <4 x i32> %4) nounwind readnone
756 %3 = tail call <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
768 %1 = tail call <2 x i32> @llvm.aarch64.neon.umin.v2i32(<2 x i32> %v1, <2 x i32> %v2) nounwind readnone
775 %3 = tail call <4 x i32 > @llvm.aarch64.neon.umin.v4i32(<4 x i32> %1, <4 x i32> %2) nounwind readnone
781 %1 = tail call <4 x i32> @llvm.aarch64.neon.umin.v4i32(<4 x i32> %v1, <4 x i32> %v2) nounwind readnone
854 ; %r1 = tail call <4 x i16> @llvm.aarch64.neon.sqshrun.v4i16(<4 x i32> %_y3, <4 x i32> <i32 8, i32 8, i32 8, i32 8>) nounwind readnone
860 %r1 = tail call <4 x i32> @llvm.aarch64.neon.smax.v4i32(<4 x i32> %_y3, <4 x i32> %c0) nounwind readnone
861 %r2 = tail call <4 x i32> @llvm.aarch64.neon.smin.v4i32(<4 x i32> %r1, <4 x i32> %c255) nounwind readnone
872 %1 = tail call <2 x float> @llvm.aarch64.neon.frecpe.v2f32(<2 x float> %v) nounwind readnone
873 %2 = tail call <2 x float> @llvm.aarch64.neon.frecps.v2f32(<2 x float> %1, <2 x float> %v) nounwind readnone
875 %4 = tail call <2 x float> @llvm.aarch64.neon.frecps.v2f32(<2 x float> %3, <2 x float> %v) nounwind readnone
881 %1 = tail call <4 x float> @llvm.aarch64.neon.frecpe.v4f32(<4 x float> %v) nounwind readnone
882 %2 = tail call <4 x float> @llvm.aarch64.neon.frecps.v4f32(<4 x float> %1, <4 x float> %v) nounwind readnone
884 %4 = tail call <4 x float> @llvm.aarch64.neon.frecps.v4f32(<4 x float> %3, <4 x float> %v) nounwind readnone
903 %2 = tail call <2 x float> @llvm.aarch64.neon.frsqrte.v2f32(<2 x float> %1) nounwind readnone
905 %4 = tail call <2 x float> @llvm.aarch64.neon.frsqrts.v2f32(<2 x float> %1, <2 x float> %3) nounwind readnone
912 %1 = tail call <2 x float> @llvm.aarch64.neon.frsqrte.v2f32(<2 x float> %v) nounwind readnone
914 %3 = tail call <2 x float> @llvm.aarch64.neon.frsqrts.v2f32(<2 x float> %v, <2 x float> %2) nounwind readnone
921 %2 = tail call <4 x float> @llvm.aarch64.neon.frsqrte.v4f32(<4 x float> %1) nounwind readnone
923 %4 = tail call <4 x float> @llvm.aarch64.neon.frsqrts.v4f32(<4 x float> %1, <4 x float> %3) nounwind readnone
930 %1 = tail call <4 x float> @llvm.aarch64.neon.frsqrte.v4f32(<4 x float> %v) nounwind readnone
932 %3 = tail call <4 x float> @llvm.aarch64.neon.frsqrts.v4f32(<4 x float> %v, <4 x float> %2) nounwind readnone
968 ; %ym = call <4 x float> @llvm.aarch64.neon.ld4.v4f32(i8* %py2, i32 4) nounwind
973 ; %zm2 = call <4 x float> @llvm.aarch64.neon.ld4.v4f32(i8* %pz2, i32 4) nounwind
