Lines Matching refs:xop
2 ; RUN: llc < %s -fast-isel -mtriple=i686-unknown-unknown -mattr=+avx,+fma4,+xop | FileCheck %s
3 ; RUN: llc < %s -fast-isel -mtriple=x86_64-unknown-unknown -mattr=+avx,+fma4,+xop | FileCheck %s
5 ; NOTE: This should use IR equivalent to what is generated by clang/test/CodeGen/xop-builtins.c
15 %res = call <8 x i16> @llvm.x86.xop.vpmacssww(<8 x i16> %arg0, <8 x i16> %arg1, <8 x i16> %arg2)
19 declare <8 x i16> @llvm.x86.xop.vpmacssww(<8 x i16>, <8 x i16>, <8 x i16>) nounwind readnone
29 %res = call <8 x i16> @llvm.x86.xop.vpmacsww(<8 x i16> %arg0, <8 x i16> %arg1, <8 x i16> %arg2)
33 declare <8 x i16> @llvm.x86.xop.vpmacsww(<8 x i16>, <8 x i16>, <8 x i16>) nounwind readnone
43 %res = call <4 x i32> @llvm.x86.xop.vpmacsswd(<8 x i16> %arg0, <8 x i16> %arg1, <4 x i32> %arg2)
47 declare <4 x i32> @llvm.x86.xop.vpmacsswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
57 %res = call <4 x i32> @llvm.x86.xop.vpmacswd(<8 x i16> %arg0, <8 x i16> %arg1, <4 x i32> %arg2)
61 declare <4 x i32> @llvm.x86.xop.vpmacswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
71 %res = call <4 x i32> @llvm.x86.xop.vpmacssdd(<4 x i32> %arg0, <4 x i32> %arg1, <4 x i32> %arg2)
75 declare <4 x i32> @llvm.x86.xop.vpmacssdd(<4 x i32>, <4 x i32>, <4 x i32>) nounwind readnone
85 %res = call <4 x i32> @llvm.x86.xop.vpmacsdd(<4 x i32> %arg0, <4 x i32> %arg1, <4 x i32> %arg2)
89 declare <4 x i32> @llvm.x86.xop.vpmacsdd(<4 x i32>, <4 x i32>, <4 x i32>) nounwind readnone
98 %res = call <2 x i64> @llvm.x86.xop.vpmacssdql(<4 x i32> %arg0, <4 x i32> %arg1, <2 x i64> %a2)
101 declare <2 x i64> @llvm.x86.xop.vpmacssdql(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
110 %res = call <2 x i64> @llvm.x86.xop.vpmacsdql(<4 x i32> %arg0, <4 x i32> %arg1, <2 x i64> %a2)
113 declare <2 x i64> @llvm.x86.xop.vpmacsdql(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
122 %res = call <2 x i64> @llvm.x86.xop.vpmacssdqh(<4 x i32> %arg0, <4 x i32> %arg1, <2 x i64> %a2)
125 declare <2 x i64> @llvm.x86.xop.vpmacssdqh(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
134 %res = call <2 x i64> @llvm.x86.xop.vpmacsdqh(<4 x i32> %arg0, <4 x i32> %arg1, <2 x i64> %a2)
137 declare <2 x i64> @llvm.x86.xop.vpmacsdqh(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
147 %res = call <4 x i32> @llvm.x86.xop.vpmadcsswd(<8 x i16> %arg0, <8 x i16> %arg1, <4 x i32> %arg2)
151 declare <4 x i32> @llvm.x86.xop.vpmadcsswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
161 %res = call <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16> %arg0, <8 x i16> %arg1, <4 x i32> %arg2)
165 declare <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
173 %res = call <8 x i16> @llvm.x86.xop.vphaddbw(<16 x i8> %arg0)
177 declare <8 x i16> @llvm.x86.xop.vphaddbw(<16 x i8>) nounwind readnone
185 %res = call <4 x i32> @llvm.x86.xop.vphaddbd(<16 x i8> %arg0)
189 declare <4 x i32> @llvm.x86.xop.vphaddbd(<16 x i8>) nounwind readnone
197 %res = call <2 x i64> @llvm.x86.xop.vphaddbq(<16 x i8> %arg0)
200 declare <2 x i64> @llvm.x86.xop.vphaddbq(<16 x i8>) nounwind readnone
208 %res = call <4 x i32> @llvm.x86.xop.vphaddwd(<8 x i16> %arg0)
212 declare <4 x i32> @llvm.x86.xop.vphaddwd(<8 x i16>) nounwind readnone
220 %res = call <2 x i64> @llvm.x86.xop.vphaddwq(<8 x i16> %arg0)
223 declare <2 x i64> @llvm.x86.xop.vphaddwq(<8 x i16>) nounwind readnone
231 %res = call <2 x i64> @llvm.x86.xop.vphadddq(<4 x i32> %arg0)
234 declare <2 x i64> @llvm.x86.xop.vphadddq(<4 x i32>) nounwind readnone
242 %res = call <8 x i16> @llvm.x86.xop.vphaddubw(<16 x i8> %arg0)
246 declare <8 x i16> @llvm.x86.xop.vphaddubw(<16 x i8>) nounwind readnone
254 %res = call <4 x i32> @llvm.x86.xop.vphaddubd(<16 x i8> %arg0)
258 declare <4 x i32> @llvm.x86.xop.vphaddubd(<16 x i8>) nounwind readnone
266 %res = call <2 x i64> @llvm.x86.xop.vphaddubq(<16 x i8> %arg0)
269 declare <2 x i64> @llvm.x86.xop.vphaddubq(<16 x i8>) nounwind readnone
277 %res = call <4 x i32> @llvm.x86.xop.vphadduwd(<8 x i16> %arg0)
281 declare <4 x i32> @llvm.x86.xop.vphadduwd(<8 x i16>) nounwind readnone
290 %res = call <2 x i64> @llvm.x86.xop.vphadduwq(<8 x i16> %arg0)
293 declare <2 x i64> @llvm.x86.xop.vphadduwq(<8 x i16>) nounwind readnone
301 %res = call <2 x i64> @llvm.x86.xop.vphaddudq(<4 x i32> %arg0)
304 declare <2 x i64> @llvm.x86.xop.vphaddudq(<4 x i32>) nounwind readnone
312 %res = call <8 x i16> @llvm.x86.xop.vphsubbw(<16 x i8> %arg0)
316 declare <8 x i16> @llvm.x86.xop.vphsubbw(<16 x i8>) nounwind readnone
324 %res = call <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16> %arg0)
328 declare <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16>) nounwind readnone
336 %res = call <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32> %arg0)
339 declare <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32>) nounwind readnone
350 %res = call <2 x i64> @llvm.x86.xop.vpcmov(<2 x i64> %a0, <2 x i64> %a1, <2 x i64> %a2)
353 declare <2 x i64> @llvm.x86.xop.vpcmov(<2 x i64>, <2 x i64>, <2 x i64>) nounwind readnone
365 %res = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %a1, <4 x i64> %a2)
368 declare <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64>, <4 x i64>, <4 x i64>) nounwind readnone
378 %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %arg0, <16 x i8> %arg1, <16 x i8> %arg2)
382 declare <16 x i8> @llvm.x86.xop.vpperm(<16 x i8>, <16 x i8>, <16 x i8>) nounwind readnone
482 %res = call <16 x i8> @llvm.x86.xop.vpshlb(<16 x i8> %arg0, <16 x i8> %arg1)
486 declare <16 x i8> @llvm.x86.xop.vpshlb(<16 x i8>, <16 x i8>) nounwind readnone
495 %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %arg0, <8 x i16> %arg1)
499 declare <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16>, <8 x i16>) nounwind readnone
508 %res = call <4 x i32> @llvm.x86.xop.vpshld(<4 x i32> %arg0, <4 x i32> %arg1)
512 declare <4 x i32> @llvm.x86.xop.vpshld(<4 x i32>, <4 x i32>) nounwind readnone
519 %res = call <2 x i64> @llvm.x86.xop.vpshlq(<2 x i64> %a0, <2 x i64> %a1)
522 declare <2 x i64> @llvm.x86.xop.vpshlq(<2 x i64>, <2 x i64>) nounwind readnone
531 %res = call <16 x i8> @llvm.x86.xop.vpshab(<16 x i8> %arg0, <16 x i8> %arg1)
535 declare <16 x i8> @llvm.x86.xop.vpshab(<16 x i8>, <16 x i8>) nounwind readnone
544 %res = call <8 x i16> @llvm.x86.xop.vpshaw(<8 x i16> %arg0, <8 x i16> %arg1)
548 declare <8 x i16> @llvm.x86.xop.vpshaw(<8 x i16>, <8 x i16>) nounwind readnone
557 %res = call <4 x i32> @llvm.x86.xop.vpshad(<4 x i32> %arg0, <4 x i32> %arg1)
561 declare <4 x i32> @llvm.x86.xop.vpshad(<4 x i32>, <4 x i32>) nounwind readnone
568 %res = call <2 x i64> @llvm.x86.xop.vpshaq(<2 x i64> %a0, <2 x i64> %a1)
571 declare <2 x i64> @llvm.x86.xop.vpshaq(<2 x i64>, <2 x i64>) nounwind readnone
676 …%res = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %a1, <2 x i64> %a…
679 declare <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double>, <2 x double>, <2 x i64>, i8) nounwind r…
686 …%res = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %a1, <4 x i64…
689 declare <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double>, <4 x double>, <4 x i64>, i8) nounwi…
697 …%res = call <4 x float> @llvm.x86.xop.vpermil2ps(<4 x float> %a0, <4 x float> %a1, <4 x i32> %arg2…
700 declare <4 x float> @llvm.x86.xop.vpermil2ps(<4 x float>, <4 x float>, <4 x i32>, i8) nounwind read…
708 …%res = call <8 x float> @llvm.x86.xop.vpermil2ps.256(<8 x float> %a0, <8 x float> %a1, <8 x i32> %…
711 declare <8 x float> @llvm.x86.xop.vpermil2ps.256(<8 x float>, <8 x float>, <8 x i32>, i8) nounwind …
718 %res = call <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float> %a0)
721 declare <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float>) nounwind readnone
728 %res = call <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double> %a0)
731 declare <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double>) nounwind readnone
738 %res = call <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float> %a0)
741 declare <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float>) nounwind readnone
748 %res = call <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double> %a0)
751 declare <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double>) nounwind readnone
758 %res = call <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float> %a0)
761 declare <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float>) nounwind readnone
768 %res = call <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double> %a0)
771 declare <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double>) nounwind readnone