Lines Matching full:shuffle
50 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
51 %mul = mul <4 x i16> %shuffle, %b
61 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
62 %mul = mul <8 x i16> %shuffle, %b
72 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
73 %mul = mul <2 x i32> %shuffle, %b
83 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
84 %mul = mul <4 x i32> %shuffle, %b
94 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
95 %mul = mul <4 x i16> %shuffle, %b
105 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
106 %mul = mul <8 x i16> %shuffle, %b
116 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
117 %mul = mul <2 x i32> %shuffle, %b
127 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
128 %mul = mul <4 x i32> %shuffle, %b
138 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
139 %mul = mul <4 x i16> %shuffle, %b
149 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
150 %mul = mul <8 x i16> %shuffle, %b
160 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
161 %mul = mul <2 x i32> %shuffle, %b
171 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
172 %mul = mul <4 x i32> %shuffle, %b
182 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
183 %mul = mul <4 x i16> %shuffle, %b
193 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
194 %mul = mul <8 x i16> %shuffle, %b
204 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
205 %mul = mul <2 x i32> %shuffle, %b
215 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
216 %mul = mul <4 x i32> %shuffle, %b
226 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
227 %mul = mul <4 x i16> %shuffle, %a
236 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
237 %mul = mul <8 x i16> %shuffle, %a
246 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
247 %mul = mul <2 x i32> %shuffle, %a
256 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
257 %mul = mul <4 x i32> %shuffle, %a
266 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
267 %mul = mul <4 x i16> %shuffle, %a
276 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
277 %mul = mul <8 x i16> %shuffle, %a
286 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
287 %mul = mul <2 x i32> %shuffle, %a
296 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
297 %mul = mul <4 x i32> %shuffle, %a
306 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
307 %mul = mul <4 x i16> %shuffle, %a
316 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
317 %mul = mul <8 x i16> %shuffle, %a
326 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
327 %mul = mul <2 x i32> %shuffle, %a
336 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
337 %mul = mul <4 x i32> %shuffle, %a
346 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
347 %mul = mul <4 x i16> %shuffle, %a
356 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <8 x i32> <i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7, i32 7>
357 %mul = mul <8 x i16> %shuffle, %a
366 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
367 %mul = mul <2 x i32> %shuffle, %a
376 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
377 %mul = mul <4 x i32> %shuffle, %a
565 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
566 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
576 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
577 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
587 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
588 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
598 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
599 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
609 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
610 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
611 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
621 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
622 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
623 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
633 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
634 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
635 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
645 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
646 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
647 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
657 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
658 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
668 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
669 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
679 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
680 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
690 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
691 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
701 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
702 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
703 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
713 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
714 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
715 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
725 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
726 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
727 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
737 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
738 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
739 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
749 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
750 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
760 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
761 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
771 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
772 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
782 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
783 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
793 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
794 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
795 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
805 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
806 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
807 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
817 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
818 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
819 shuffle.i, <4 x i16> %shuffle)
829 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
830 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
831 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
841 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
842 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
852 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
853 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
863 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
864 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
874 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
875 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
885 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
886 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
887 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
897 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
898 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
899 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
909 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
910 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
911 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
921 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
922 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
923 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
933 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
934 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
943 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
944 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
953 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
954 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
963 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
964 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
973 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
974 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
975 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
984 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
985 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
986 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
995 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
996 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
997 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
1006 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
1007 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
1008 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
1017 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
1018 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
1027 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
1028 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
1037 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
1038 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
1047 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
1048 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
1057 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
1058 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
1059 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
1068 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
1069 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
1070 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
1079 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
1080 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
1081 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
1090 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
1091 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
1092 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
1101 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1102 %vqdmlal2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
1112 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
1113 %vqdmlal2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
1123 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
1124 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1125 %vqdmlal2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
1135 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
1136 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
1137 %vqdmlal2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
1147 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1148 %vqdmlsl2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
1158 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
1159 %vqdmlsl2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
1169 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
1170 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1171 %vqdmlsl2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
1181 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
1182 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
1183 %vqdmlsl2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
1193 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1194 %vqdmull2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
1203 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
1204 %vqdmull2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
1213 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1214 %vqdmull2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
1223 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
1224 %vqdmull2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
1233 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
1234 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1235 %vqdmull2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
1244 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
1245 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
1246 %vqdmull2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
1255 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
1256 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> <i32 7, i32 7, i32 7, i32 7>
1257 %vqdmull2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
1266 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
1267 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> <i32 3, i32 3>
1268 %vqdmull2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
1277 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1278 %vqdmulh2.i = tail call <4 x i16> @llvm.aarch64.neon.sqdmulh.v4i16(<4 x i16> %a, <4 x i16> %shuffle)
1287 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
1288 %vqdmulh2.i = tail call <8 x i16> @llvm.aarch64.neon.sqdmulh.v8i16(<8 x i16> %a, <8 x i16> %shuffle)
1297 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
1298 %vqdmulh2.i = tail call <2 x i32> @llvm.aarch64.neon.sqdmulh.v2i32(<2 x i32> %a, <2 x i32> %shuffle)
1307 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
1308 %vqdmulh2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmulh.v4i32(<4 x i32> %a, <4 x i32> %shuffle)
1317 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1318 %vqrdmulh2.i = tail call <4 x i16> @llvm.aarch64.neon.sqrdmulh.v4i16(<4 x i16> %a, <4 x i16> %shuffle)
1327 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> <i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3, i32 3>
1328 %vqrdmulh2.i = tail call <8 x i16> @llvm.aarch64.neon.sqrdmulh.v8i16(<8 x i16> %a, <8 x i16> %shuffle)
1337 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> <i32 1, i32 1>
1338 %vqrdmulh2.i = tail call <2 x i32> @llvm.aarch64.neon.sqrdmulh.v2i32(<2 x i32> %a, <2 x i32> %shuffle)
1347 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
1348 %vqrdmulh2.i = tail call <4 x i32> @llvm.aarch64.neon.sqrdmulh.v4i32(<4 x i32> %a, <4 x i32> %shuffle)
1357 %shuffle = shufflevector <2 x float> %v, <2 x float> undef, <2 x i32> <i32 1, i32 1>
1358 %mul = fmul <2 x float> %shuffle, %a
1380 %shuffle = shufflevector <2 x float> %v, <2 x float> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
1381 %mul = fmul <4 x float> %shuffle, %a
1390 %shuffle = shufflevector <1 x double> %v, <1 x double> undef, <2 x i32> zeroinitializer
1391 %mul = fmul <2 x double> %shuffle, %a
1400 %shuffle = shufflevector <4 x float> %v, <4 x float> undef, <2 x i32> <i32 3, i32 3>
1401 %mul = fmul <2 x float> %shuffle, %a
1423 %shuffle = shufflevector <4 x float> %v, <4 x float> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1424 %mul = fmul <4 x float> %shuffle, %a
1433 %shuffle = shufflevector <2 x double> %v, <2 x double> undef, <2 x i32> <i32 1, i32 1>
1434 %mul = fmul <2 x double> %shuffle, %a
1443 %shuffle = shufflevector <2 x float> %v, <2 x float> undef, <2 x i32> <i32 1, i32 1>
1444 %vmulx2.i = tail call <2 x float> @llvm.aarch64.neon.fmulx.v2f32(<2 x float> %a, <2 x float> %shuffle)
1453 %shuffle = shufflevector <2 x float> %v, <2 x float> undef, <4 x i32> <i32 1, i32 1, i32 1, i32 1>
1454 %vmulx2.i = tail call <4 x float> @llvm.aarch64.neon.fmulx.v4f32(<4 x float> %a, <4 x float> %shuffle)
1463 %shuffle = shufflevector <1 x double> %v, <1 x double> undef, <2 x i32> zeroinitializer
1464 %vmulx2.i = tail call <2 x double> @llvm.aarch64.neon.fmulx.v2f64(<2 x double> %a, <2 x double> %shuffle)
1473 %shuffle = shufflevector <4 x float> %v, <4 x float> undef, <2 x i32> <i32 3, i32 3>
1474 %vmulx2.i = tail call <2 x float> @llvm.aarch64.neon.fmulx.v2f32(<2 x float> %a, <2 x float> %shuffle)
1483 %shuffle = shufflevector <4 x float> %v, <4 x float> undef, <4 x i32> <i32 3, i32 3, i32 3, i32 3>
1484 %vmulx2.i = tail call <4 x float> @llvm.aarch64.neon.fmulx.v4f32(<4 x float> %a, <4 x float> %shuffle)
1493 %shuffle = shufflevector <2 x double> %v, <2 x double> undef, <2 x i32> <i32 1, i32 1>
1494 %vmulx2.i = tail call <2 x double> @llvm.aarch64.neon.fmulx.v2f64(<2 x double> %a, <2 x double> %shuffle)
1503 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
1504 %mul = mul <4 x i16> %shuffle, %b
1514 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> zeroinitializer
1515 %mul = mul <8 x i16> %shuffle, %b
1525 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
1526 %mul = mul <2 x i32> %shuffle, %b
1536 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> zeroinitializer
1537 %mul = mul <4 x i32> %shuffle, %b
1547 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
1548 %mul = mul <4 x i16> %shuffle, %b
1558 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <8 x i32> zeroinitializer
1559 %mul = mul <8 x i16> %shuffle, %b
1569 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
1570 %mul = mul <2 x i32> %shuffle, %b
1580 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <4 x i32> zeroinitializer
1581 %mul = mul <4 x i32> %shuffle, %b
1591 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
1592 %mul = mul <4 x i16> %shuffle, %b
1602 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> zeroinitializer
1603 %mul = mul <8 x i16> %shuffle, %b
1613 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
1614 %mul = mul <2 x i32> %shuffle, %b
1624 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> zeroinitializer
1625 %mul = mul <4 x i32> %shuffle, %b
1635 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
1636 %mul = mul <4 x i16> %shuffle, %b
1646 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <8 x i32> zeroinitializer
1647 %mul = mul <8 x i16> %shuffle, %b
1657 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
1658 %mul = mul <2 x i32> %shuffle, %b
1668 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <4 x i32> zeroinitializer
1669 %mul = mul <4 x i32> %shuffle, %b
1679 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
1680 %mul = mul <4 x i16> %shuffle, %a
1689 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> zeroinitializer
1690 %mul = mul <8 x i16> %shuffle, %a
1699 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
1700 %mul = mul <2 x i32> %shuffle, %a
1709 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> zeroinitializer
1710 %mul = mul <4 x i32> %shuffle, %a
1719 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
1720 %mul = mul <4 x i16> %shuffle, %a
1729 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> zeroinitializer
1730 %mul = mul <8 x i16> %shuffle, %a
1739 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
1740 %mul = mul <2 x i32> %shuffle, %a
1749 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> zeroinitializer
1750 %mul = mul <4 x i32> %shuffle, %a
1759 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
1760 %mul = mul <4 x i16> %shuffle, %a
1769 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <8 x i32> zeroinitializer
1770 %mul = mul <8 x i16> %shuffle, %a
1779 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
1780 %mul = mul <2 x i32> %shuffle, %a
1789 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <4 x i32> zeroinitializer
1790 %mul = mul <4 x i32> %shuffle, %a
1799 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
1800 %mul = mul <4 x i16> %shuffle, %a
1809 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <8 x i32> zeroinitializer
1810 %mul = mul <8 x i16> %shuffle, %a
1819 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
1820 %mul = mul <2 x i32> %shuffle, %a
1829 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <4 x i32> zeroinitializer
1830 %mul = mul <4 x i32> %shuffle, %a
1944 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
1945 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
1955 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
1956 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
1966 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
1967 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
1977 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
1978 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
1988 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
1989 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
1990 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2000 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2001 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2002 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2012 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2013 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2014 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2024 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2025 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2026 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2036 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2037 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
2047 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2048 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
2058 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2059 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
2069 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2070 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
2080 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2081 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2082 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2092 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2093 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2094 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2104 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2105 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2106 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2116 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2117 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2118 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2128 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2129 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
2139 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2140 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
2150 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2151 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
2161 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2162 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
2172 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2173 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2174 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2184 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2185 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2186 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2196 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2197 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2198 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2208 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2209 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2210 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2220 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2221 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
2231 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2232 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
2242 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2243 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
2253 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2254 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
2264 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2265 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2266 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2276 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2277 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2278 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2288 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2289 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2290 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2300 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2301 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2302 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2312 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2313 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
2322 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2323 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
2332 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2333 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
2342 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2343 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
2352 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2353 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2354 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2363 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2364 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2365 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2374 %shuffle
2375 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2376 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2385 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2386 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2387 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2396 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2397 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
2406 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2407 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
2416 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2417 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
2426 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2427 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
2436 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2437 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2438 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.smull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2447 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2448 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2449 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.smull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2458 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2459 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2460 %vmull2.i = tail call <4 x i32> @llvm.aarch64.neon.umull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2469 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2470 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2471 %vmull2.i = tail call <2 x i64> @llvm.aarch64.neon.umull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2480 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2481 %vqdmlal2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
2491 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2492 %vqdmlal2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
2502 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2503 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2504 %vqdmlal2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2514 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2515 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2516 %vqdmlal2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2526 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2527 %vqdmlsl2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %b, <4 x i16> %shuffle)
2537 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2538 %vqdmlsl2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %b, <2 x i32> %shuffle)
2548 %shuffle.i = shufflevector <8 x i16> %b, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2549 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2550 %vqdmlsl2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2560 %shuffle.i = shufflevector <4 x i32> %b, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2561 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2562 %vqdmlsl2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2572 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2573 %vqdmull2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
2582 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2583 %vqdmull2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
2592 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2593 %vqdmull2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %a, <4 x i16> %shuffle)
2602 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2603 %vqdmull2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %a, <2 x i32> %shuffle)
2612 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2613 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2614 %vqdmull2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2623 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2624 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2625 %vqdmull2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2634 %shuffle.i = shufflevector <8 x i16> %a, <8 x i16> undef, <4 x i32> <i32 4, i32 5, i32 6, i32 7>
2635 %shuffle = shufflevector <8 x i16> %v, <8 x i16> undef, <4 x i32> zeroinitializer
2636 %vqdmull2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmull.v4i32(<4 x i16> %shuffle.i, <4 x i16> %shuffle)
2645 %shuffle.i = shufflevector <4 x i32> %a, <4 x i32> undef, <2 x i32> <i32 2, i32 3>
2646 %shuffle = shufflevector <4 x i32> %v, <4 x i32> undef, <2 x i32> zeroinitializer
2647 %vqdmull2.i = tail call <2 x i64> @llvm.aarch64.neon.sqdmull.v2i64(<2 x i32> %shuffle.i, <2 x i32> %shuffle)
2656 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2657 %vqdmulh2.i = tail call <4 x i16> @llvm.aarch64.neon.sqdmulh.v4i16(<4 x i16> %a, <4 x i16> %shuffle)
2666 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> zeroinitializer
2667 %vqdmulh2.i = tail call <8 x i16> @llvm.aarch64.neon.sqdmulh.v8i16(<8 x i16> %a, <8 x i16> %shuffle)
2676 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2677 %vqdmulh2.i = tail call <2 x i32> @llvm.aarch64.neon.sqdmulh.v2i32(<2 x i32> %a, <2 x i32> %shuffle)
2686 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> zeroinitializer
2687 %vqdmulh2.i = tail call <4 x i32> @llvm.aarch64.neon.sqdmulh.v4i32(<4 x i32> %a, <4 x i32> %shuffle)
2696 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <4 x i32> zeroinitializer
2697 %vqrdmulh2.i = tail call <4 x i16> @llvm.aarch64.neon.sqrdmulh.v4i16(<4 x i16> %a, <4 x i16> %shuffle)
2706 %shuffle = shufflevector <4 x i16> %v, <4 x i16> undef, <8 x i32> zeroinitializer
2707 %vqrdmulh2.i = tail call <8 x i16> @llvm.aarch64.neon.sqrdmulh.v8i16(<8 x i16> %a, <8 x i16> %shuffle)
2716 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <2 x i32> zeroinitializer
2717 %vqrdmulh2.i = tail call <2 x i32> @llvm.aarch64.neon.sqrdmulh.v2i32(<2 x i32> %a, <2 x i32> %shuffle)
2726 %shuffle = shufflevector <2 x i32> %v, <2 x i32> undef, <4 x i32> zeroinitializer
2727 shuffle)
2736 %shuffle = shufflevector <2 x float> %v, <2 x float> undef, <2 x i32> zeroinitializer
2737 %mul = fmul <2 x float> %shuffle, %a
2746 %shuffle = shufflevector <2 x float> %v, <2 x float> undef, <4 x i32> zeroinitializer
2747 %mul = fmul <4 x float> %shuffle, %a
2756 %shuffle = shufflevector <4 x float> %v, <4 x float> undef, <2 x i32> zeroinitializer
2757 %mul = fmul <2 x float> %shuffle, %a
2779 %shuffle = shufflevector <4 x float> %v, <4 x float> undef, <4 x i32> zeroinitializer
2780 %mul = fmul <4 x float> %shuffle, %a
2789 %shuffle = shufflevector <2 x double> %v, <2 x double> undef, <2 x i32> zeroinitializer
2790 %mul = fmul <2 x double> %shuffle, %a
2799 %shuffle = shufflevector <2 x float> %v, <2 x float> undef, <2 x i32> zeroinitializer
2800 %vmulx2.i = tail call <2 x float> @llvm.aarch64.neon.fmulx.v2f32(<2 x float> %a, <2 x float> %shuffle)
2809 %shuffle = shufflevector <2 x float> %v, <2 x float> undef, <4 x i32> zeroinitializer
2810 %vmulx2.i = tail call <4 x float> @llvm.aarch64.neon.fmulx.v4f32(<4 x float> %a, <4 x float> %shuffle)
2819 %shuffle = shufflevector <1 x double> %v, <1 x double> undef, <2 x i32> zeroinitializer
2820 %vmulx2.i = tail call <2 x double> @llvm.aarch64.neon.fmulx.v2f64(<2 x double> %a, <2 x double> %shuffle)
2829 %shuffle = shufflevector <4 x float> %v, <4 x float> undef, <2 x i32> zeroinitializer
2830 %vmulx2.i = tail call <2 x float> @llvm.aarch64.neon.fmulx.v2f32(<2 x float> %a, <2 x float> %shuffle)
2839 %shuffle = shufflevector <4 x float> %v, <4 x float> undef, <4 x i32> zeroinitializer
2840 %vmulx2.i = tail call <4 x float> @llvm.aarch64.neon.fmulx.v4f32(<4 x float> %a, <4 x float> %shuffle)
2849 %shuffle = shufflevector <2 x double> %v, <2 x double> undef, <2 x i32> zeroinitializer
2850 %vmulx2.i = tail call <2 x double> @llvm.aarch64.neon.fmulx.v2f64(<2 x double> %a, <2 x double> %shuffle)
