Lines Matching refs:xop
2 ; RUN: llc < %s -mtriple=x86_64-unknown-unknown -mattr=+avx,+fma4,+xop | FileCheck %s
9 …%res = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %a1, <2 x i64> %a…
18 …%res = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %vec, <2 x i64> %…
27 …%res = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %a1, <2 x i64> %v…
30 declare <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double>, <2 x double>, <2 x i64>, i8) nounwind r…
37 …%res = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %a1, <4 x i64…
46 …%res = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %vec, <4 x i6…
55 …%res = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %a1, <4 x i64…
58 declare <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double>, <4 x double>, <4 x i64>, i8) nounwi…
65 …%res = call <4 x float> @llvm.x86.xop.vpermil2ps(<4 x float> %a0, <4 x float> %a1, <4 x i32> %a2, …
68 declare <4 x float> @llvm.x86.xop.vpermil2ps(<4 x float>, <4 x float>, <4 x i32>, i8) nounwind read…
75 …%res = call <8 x float> @llvm.x86.xop.vpermil2ps.256(<8 x float> %a0, <8 x float> %a1, <8 x i32> %…
78 declare <8 x float> @llvm.x86.xop.vpermil2ps.256(<8 x float>, <8 x float>, <8 x i32>, i8) nounwind …
85 %res = call <2 x i64> @llvm.x86.xop.vpcmov(<2 x i64> %a0, <2 x i64> %a1, <2 x i64> %a2) ;
88 declare <2 x i64> @llvm.x86.xop.vpcmov(<2 x i64>, <2 x i64>, <2 x i64>) nounwind readnone
95 %res = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %a1, <4 x i64> %a2) ;
104 %res = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %vec, <4 x i64> %a2) ;
113 %res = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %a1, <4 x i64> %vec) ;
116 declare <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64>, <4 x i64>, <4 x i64>) nounwind readnone
123 %res = call <4 x i32> @llvm.x86.xop.vphaddbd(<16 x i8> %a0) ;
126 declare <4 x i32> @llvm.x86.xop.vphaddbd(<16 x i8>) nounwind readnone
133 %res = call <2 x i64> @llvm.x86.xop.vphaddbq(<16 x i8> %a0) ;
136 declare <2 x i64> @llvm.x86.xop.vphaddbq(<16 x i8>) nounwind readnone
143 %res = call <8 x i16> @llvm.x86.xop.vphaddbw(<16 x i8> %a0) ;
146 declare <8 x i16> @llvm.x86.xop.vphaddbw(<16 x i8>) nounwind readnone
153 %res = call <2 x i64> @llvm.x86.xop.vphadddq(<4 x i32> %a0) ;
156 declare <2 x i64> @llvm.x86.xop.vphadddq(<4 x i32>) nounwind readnone
163 %res = call <4 x i32> @llvm.x86.xop.vphaddubd(<16 x i8> %a0) ;
166 declare <4 x i32> @llvm.x86.xop.vphaddubd(<16 x i8>) nounwind readnone
173 %res = call <2 x i64> @llvm.x86.xop.vphaddubq(<16 x i8> %a0) ;
176 declare <2 x i64> @llvm.x86.xop.vphaddubq(<16 x i8>) nounwind readnone
183 %res = call <8 x i16> @llvm.x86.xop.vphaddubw(<16 x i8> %a0) ;
186 declare <8 x i16> @llvm.x86.xop.vphaddubw(<16 x i8>) nounwind readnone
193 %res = call <2 x i64> @llvm.x86.xop.vphaddudq(<4 x i32> %a0) ;
196 declare <2 x i64> @llvm.x86.xop.vphaddudq(<4 x i32>) nounwind readnone
203 %res = call <4 x i32> @llvm.x86.xop.vphadduwd(<8 x i16> %a0) ;
206 declare <4 x i32> @llvm.x86.xop.vphadduwd(<8 x i16>) nounwind readnone
213 %res = call <2 x i64> @llvm.x86.xop.vphadduwq(<8 x i16> %a0) ;
216 declare <2 x i64> @llvm.x86.xop.vphadduwq(<8 x i16>) nounwind readnone
223 %res = call <4 x i32> @llvm.x86.xop.vphaddwd(<8 x i16> %a0) ;
226 declare <4 x i32> @llvm.x86.xop.vphaddwd(<8 x i16>) nounwind readnone
233 %res = call <2 x i64> @llvm.x86.xop.vphaddwq(<8 x i16> %a0) ;
236 declare <2 x i64> @llvm.x86.xop.vphaddwq(<8 x i16>) nounwind readnone
243 %res = call <8 x i16> @llvm.x86.xop.vphsubbw(<16 x i8> %a0) ;
246 declare <8 x i16> @llvm.x86.xop.vphsubbw(<16 x i8>) nounwind readnone
253 %res = call <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32> %a0) ;
262 %res = call <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32> %vec) ;
265 declare <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32>) nounwind readnone
272 %res = call <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16> %a0) ;
281 %res = call <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16> %vec) ;
284 declare <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16>) nounwind readnone
291 %res = call <4 x i32> @llvm.x86.xop.vpmacsdd(<4 x i32> %a0, <4 x i32> %a1, <4 x i32> %a2) ;
294 declare <4 x i32> @llvm.x86.xop.vpmacsdd(<4 x i32>, <4 x i32>, <4 x i32>) nounwind readnone
301 %res = call <2 x i64> @llvm.x86.xop.vpmacsdqh(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2) ;
304 declare <2 x i64> @llvm.x86.xop.vpmacsdqh(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
311 %res = call <2 x i64> @llvm.x86.xop.vpmacsdql(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2) ;
314 declare <2 x i64> @llvm.x86.xop.vpmacsdql(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
321 %res = call <4 x i32> @llvm.x86.xop.vpmacssdd(<4 x i32> %a0, <4 x i32> %a1, <4 x i32> %a2) ;
324 declare <4 x i32> @llvm.x86.xop.vpmacssdd(<4 x i32>, <4 x i32>, <4 x i32>) nounwind readnone
331 %res = call <2 x i64> @llvm.x86.xop.vpmacssdqh(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2) ;
334 declare <2 x i64> @llvm.x86.xop.vpmacssdqh(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
341 %res = call <2 x i64> @llvm.x86.xop.vpmacssdql(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2) ;
344 declare <2 x i64> @llvm.x86.xop.vpmacssdql(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
351 %res = call <4 x i32> @llvm.x86.xop.vpmacsswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2) ;
354 declare <4 x i32> @llvm.x86.xop.vpmacsswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
361 %res = call <8 x i16> @llvm.x86.xop.vpmacssww(<8 x i16> %a0, <8 x i16> %a1, <8 x i16> %a2) ;
364 declare <8 x i16> @llvm.x86.xop.vpmacssww(<8 x i16>, <8 x i16>, <8 x i16>) nounwind readnone
371 %res = call <4 x i32> @llvm.x86.xop.vpmacswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2) ;
374 declare <4 x i32> @llvm.x86.xop.vpmacswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
381 %res = call <8 x i16> @llvm.x86.xop.vpmacsww(<8 x i16> %a0, <8 x i16> %a1, <8 x i16> %a2) ;
384 declare <8 x i16> @llvm.x86.xop.vpmacsww(<8 x i16>, <8 x i16>, <8 x i16>) nounwind readnone
391 %res = call <4 x i32> @llvm.x86.xop.vpmadcsswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2) ;
394 declare <4 x i32> @llvm.x86.xop.vpmadcsswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
401 %res = call <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2) ;
410 %res = call <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16> %a0, <8 x i16> %vec, <4 x i32> %a2) ;
413 declare <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
420 %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8> %a2) ;
429 %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8> %vec) ;
438 %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %vec, <16 x i8> %a2) ;
441 declare <16 x i8> @llvm.x86.xop.vpperm(<16 x i8>, <16 x i8>, <16 x i8>) nounwind readnone
448 %res = call <16 x i8> @llvm.x86.xop.vprotb(<16 x i8> %a0, <16 x i8> %a1) ;
451 declare <16 x i8> @llvm.x86.xop.vprotb(<16 x i8>, <16 x i8>) nounwind readnone
458 %res = call <4 x i32> @llvm.x86.xop.vprotd(<4 x i32> %a0, <4 x i32> %a1) ;
461 declare <4 x i32> @llvm.x86.xop.vprotd(<4 x i32>, <4 x i32>) nounwind readnone
468 %res = call <2 x i64> @llvm.x86.xop.vprotq(<2 x i64> %a0, <2 x i64> %a1) ;
471 declare <2 x i64> @llvm.x86.xop.vprotq(<2 x i64>, <2 x i64>) nounwind readnone
478 %res = call <8 x i16> @llvm.x86.xop.vprotw(<8 x i16> %a0, <8 x i16> %a1) ;
481 declare <8 x i16> @llvm.x86.xop.vprotw(<8 x i16>, <8 x i16>) nounwind readnone
488 %res = call <16 x i8> @llvm.x86.xop.vprotbi(<16 x i8> %a0, i8 1) ;
491 declare <16 x i8> @llvm.x86.xop.vprotbi(<16 x i8>, i8) nounwind readnone
498 %res = call <4 x i32> @llvm.x86.xop.vprotdi(<4 x i32> %a0, i8 -2) ;
501 declare <4 x i32> @llvm.x86.xop.vprotdi(<4 x i32>, i8) nounwind readnone
508 %res = call <2 x i64> @llvm.x86.xop.vprotqi(<2 x i64> %a0, i8 3) ;
511 declare <2 x i64> @llvm.x86.xop.vprotqi(<2 x i64>, i8) nounwind readnone
518 %res = call <8 x i16> @llvm.x86.xop.vprotwi(<8 x i16> %a0, i8 -4) ;
521 declare <8 x i16> @llvm.x86.xop.vprotwi(<8 x i16>, i8) nounwind readnone
528 %res = call <16 x i8> @llvm.x86.xop.vpshab(<16 x i8> %a0, <16 x i8> %a1) ;
531 declare <16 x i8> @llvm.x86.xop.vpshab(<16 x i8>, <16 x i8>) nounwind readnone
538 %res = call <4 x i32> @llvm.x86.xop.vpshad(<4 x i32> %a0, <4 x i32> %a1) ;
541 declare <4 x i32> @llvm.x86.xop.vpshad(<4 x i32>, <4 x i32>) nounwind readnone
548 %res = call <2 x i64> @llvm.x86.xop.vpshaq(<2 x i64> %a0, <2 x i64> %a1) ;
551 declare <2 x i64> @llvm.x86.xop.vpshaq(<2 x i64>, <2 x i64>) nounwind readnone
558 %res = call <8 x i16> @llvm.x86.xop.vpshaw(<8 x i16> %a0, <8 x i16> %a1) ;
561 declare <8 x i16> @llvm.x86.xop.vpshaw(<8 x i16>, <8 x i16>) nounwind readnone
568 %res = call <16 x i8> @llvm.x86.xop.vpshlb(<16 x i8> %a0, <16 x i8> %a1) ;
571 declare <16 x i8> @llvm.x86.xop.vpshlb(<16 x i8>, <16 x i8>) nounwind readnone
578 %res = call <4 x i32> @llvm.x86.xop.vpshld(<4 x i32> %a0, <4 x i32> %a1) ;
581 declare <4 x i32> @llvm.x86.xop.vpshld(<4 x i32>, <4 x i32>) nounwind readnone
588 %res = call <2 x i64> @llvm.x86.xop.vpshlq(<2 x i64> %a0, <2 x i64> %a1) ;
591 declare <2 x i64> @llvm.x86.xop.vpshlq(<2 x i64>, <2 x i64>) nounwind readnone
598 %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %a0, <8 x i16> %a1) ;
607 %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %a0, <8 x i16> %vec) ;
616 %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %vec, <8 x i16> %a1) ;
619 declare <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16>, <8 x i16>) nounwind readnone
626 %res = call <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float> %a0) ;
636 %res = call <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float> %vec) ;
639 declare <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float>) nounwind readnone
646 %res = call <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double> %a0) ;
656 %res = call <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double> %vec) ;
659 declare <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double>) nounwind readnone
666 %res = call <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double> %a0) ;
675 %res = call <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double> %vec) ;
678 declare <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double>) nounwind readnone
685 %res = call <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double> %a0) ;
694 %res = call <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double> %vec) ;
697 declare <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double>) nounwind readnone
704 %res = call <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float> %a0) ;
713 %res = call <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float> %vec) ;
716 declare <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float>) nounwind readnone
723 %res = call <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float> %a0) ;
732 %res = call <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float> %vec) ;
735 declare <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float>) nounwind readnone
742 %res = call <16 x i8> @llvm.x86.xop.vpcomb(<16 x i8> %a0, <16 x i8> %a1, i8 0) ;
745 declare <16 x i8> @llvm.x86.xop.vpcomb(<16 x i8>, <16 x i8>, i8) nounwind readnone
752 %res = call <8 x i16> @llvm.x86.xop.vpcomw(<8 x i16> %a0, <8 x i16> %a1, i8 0) ;
755 declare <8 x i16> @llvm.x86.xop.vpcomw(<8 x i16>, <8 x i16>, i8) nounwind readnone
762 %res = call <4 x i32> @llvm.x86.xop.vpcomd(<4 x i32> %a0, <4 x i32> %a1, i8 0) ;
765 declare <4 x i32> @llvm.x86.xop.vpcomd(<4 x i32>, <4 x i32>, i8) nounwind readnone
772 %res = call <2 x i64> @llvm.x86.xop.vpcomq(<2 x i64> %a0, <2 x i64> %a1, i8 0) ;
775 declare <2 x i64> @llvm.x86.xop.vpcomq(<2 x i64>, <2 x i64>, i8) nounwind readnone
782 %res = call <16 x i8> @llvm.x86.xop.vpcomub(<16 x i8> %a0, <16 x i8> %a1, i8 0) ;
785 declare <16 x i8> @llvm.x86.xop.vpcomub(<16 x i8>, <16 x i8>, i8) nounwind readnone
792 %res = call <8 x i16> @llvm.x86.xop.vpcomuw(<8 x i16> %a0, <8 x i16> %a1, i8 0) ;
795 declare <8 x i16> @llvm.x86.xop.vpcomuw(<8 x i16>, <8 x i16>, i8) nounwind readnone
802 %res = call <4 x i32> @llvm.x86.xop.vpcomud(<4 x i32> %a0, <4 x i32> %a1, i8 0) ;
805 declare <4 x i32> @llvm.x86.xop.vpcomud(<4 x i32>, <4 x i32>, i8) nounwind readnone
812 %res = call <2 x i64> @llvm.x86.xop.vpcomuq(<2 x i64> %a0, <2 x i64> %a1, i8 0) ;
815 declare <2 x i64> @llvm.x86.xop.vpcomuq(<2 x i64>, <2 x i64>, i8) nounwind readnone