Lines Matching refs:xop
1 ; RUN: llc -O3 -disable-peephole -mtriple=x86_64-unknown-unknown -mattr=+avx,+xop < %s | FileCheck …
15 %2 = call <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double> %a0)
18 declare <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double>) nounwind readnone
24 %2 = call <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double> %a0)
27 declare <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double>) nounwind readnone
33 %2 = call <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float> %a0)
36 declare <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float>) nounwind readnone
42 %2 = call <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float> %a0)
45 declare <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float>) nounwind readnone
51 %2 = call <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double> %a0)
54 declare <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double>) nounwind readnone
60 %2 = call <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float> %a0)
63 declare <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float>) nounwind readnone
69 %2 = call <2 x i64> @llvm.x86.xop.vpcmov(<2 x i64> %a0, <2 x i64> %a1, <2 x i64> %a2)
76 %2 = call <2 x i64> @llvm.x86.xop.vpcmov(<2 x i64> %a0, <2 x i64> %a2, <2 x i64> %a1)
79 declare <2 x i64> @llvm.x86.xop.vpcmov(<2 x i64>, <2 x i64>, <2 x i64>) nounwind readnone
85 %2 = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %a1, <4 x i64> %a2)
92 %2 = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %a2, <4 x i64> %a1)
95 declare <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64>, <4 x i64>, <4 x i64>) nounwind readnone
101 %2 = call <16 x i8> @llvm.x86.xop.vpcomb(<16 x i8> %a0, <16 x i8> %a1, i8 0)
104 declare <16 x i8> @llvm.x86.xop.vpcomb(<16 x i8>, <16 x i8>, i8) nounwind readnone
110 %2 = call <4 x i32> @llvm.x86.xop.vpcomd(<4 x i32> %a0, <4 x i32> %a1, i8 0)
113 declare <4 x i32> @llvm.x86.xop.vpcomd(<4 x i32>, <4 x i32>, i8) nounwind readnone
119 %2 = call <2 x i64> @llvm.x86.xop.vpcomq(<2 x i64> %a0, <2 x i64> %a1, i8 0)
122 declare <2 x i64> @llvm.x86.xop.vpcomq(<2 x i64>, <2 x i64>, i8) nounwind readnone
128 %2 = call <16 x i8> @llvm.x86.xop.vpcomub(<16 x i8> %a0, <16 x i8> %a1, i8 0)
131 declare <16 x i8> @llvm.x86.xop.vpcomub(<16 x i8>, <16 x i8>, i8) nounwind readnone
137 %2 = call <4 x i32> @llvm.x86.xop.vpcomud(<4 x i32> %a0, <4 x i32> %a1, i8 0)
140 declare <4 x i32> @llvm.x86.xop.vpcomud(<4 x i32>, <4 x i32>, i8) nounwind readnone
146 %2 = call <2 x i64> @llvm.x86.xop.vpcomuq(<2 x i64> %a0, <2 x i64> %a1, i8 0)
149 declare <2 x i64> @llvm.x86.xop.vpcomuq(<2 x i64>, <2 x i64>, i8) nounwind readnone
155 %2 = call <8 x i16> @llvm.x86.xop.vpcomuw(<8 x i16> %a0, <8 x i16> %a1, i8 0)
158 declare <8 x i16> @llvm.x86.xop.vpcomuw(<8 x i16>, <8 x i16>, i8) nounwind readnone
164 %2 = call <8 x i16> @llvm.x86.xop.vpcomw(<8 x i16> %a0, <8 x i16> %a1, i8 0)
167 declare <8 x i16> @llvm.x86.xop.vpcomw(<8 x i16>, <8 x i16>, i8) nounwind readnone
173 …%2 = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %a1, <2 x double> %…
180 …%2 = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %a2, <2 x double> %…
183 declare <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double>, <2 x double>, <2 x double>, i8) nounwin…
189 …%2 = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %a1, <4 x doubl…
196 …%2 = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %a2, <4 x doubl…
199 declare <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double>, <4 x double>, <4 x double>, i8) nou…
205 …%2 = call <4 x float> @llvm.x86.xop.vpermil2ps(<4 x float> %a0, <4 x float> %a1, <4 x float> %a2, …
212 …%2 = call <4 x float> @llvm.x86.xop.vpermil2ps(<4 x float> %a0, <4 x float> %a2, <4 x float> %a1, …
215 declare <4 x float> @llvm.x86.xop.vpermil2ps(<4 x float>, <4 x float>, <4 x float>, i8) nounwind re…
221 …%2 = call <8 x float> @llvm.x86.xop.vpermil2ps.256(<8 x float> %a0, <8 x float> %a1, <8 x float> %…
228 …%2 = call <8 x float> @llvm.x86.xop.vpermil2ps.256(<8 x float> %a0, <8 x float> %a2, <8 x float> %…
231 declare <8 x float> @llvm.x86.xop.vpermil2ps.256(<8 x float>, <8 x float>, <8 x float>, i8) nounwin…
237 %2 = call <4 x i32> @llvm.x86.xop.vphaddbd(<16 x i8> %a0)
240 declare <4 x i32> @llvm.x86.xop.vphaddbd(<16 x i8>) nounwind readnone
246 %2 = call <2 x i64> @llvm.x86.xop.vphaddbq(<16 x i8> %a0)
249 declare <2 x i64> @llvm.x86.xop.vphaddbq(<16 x i8>) nounwind readnone
255 %2 = call <8 x i16> @llvm.x86.xop.vphaddbw(<16 x i8> %a0)
258 declare <8 x i16> @llvm.x86.xop.vphaddbw(<16 x i8>) nounwind readnone
264 %2 = call <2 x i64> @llvm.x86.xop.vphadddq(<4 x i32> %a0)
267 declare <2 x i64> @llvm.x86.xop.vphadddq(<4 x i32>) nounwind readnone
273 %2 = call <4 x i32> @llvm.x86.xop.vphaddubd(<16 x i8> %a0)
276 declare <4 x i32> @llvm.x86.xop.vphaddubd(<16 x i8>) nounwind readnone
282 %2 = call <2 x i64> @llvm.x86.xop.vphaddubq(<16 x i8> %a0)
285 declare <2 x i64> @llvm.x86.xop.vphaddubq(<16 x i8>) nounwind readnone
291 %2 = call <8 x i16> @llvm.x86.xop.vphaddubw(<16 x i8> %a0)
294 declare <8 x i16> @llvm.x86.xop.vphaddubw(<16 x i8>) nounwind readnone
300 %2 = call <2 x i64> @llvm.x86.xop.vphaddudq(<4 x i32> %a0)
303 declare <2 x i64> @llvm.x86.xop.vphaddudq(<4 x i32>) nounwind readnone
309 %2 = call <4 x i32> @llvm.x86.xop.vphadduwd(<8 x i16> %a0)
312 declare <4 x i32> @llvm.x86.xop.vphadduwd(<8 x i16>) nounwind readnone
318 %2 = call <2 x i64> @llvm.x86.xop.vphadduwq(<8 x i16> %a0)
321 declare <2 x i64> @llvm.x86.xop.vphadduwq(<8 x i16>) nounwind readnone
327 %2 = call <4 x i32> @llvm.x86.xop.vphaddwd(<8 x i16> %a0)
330 declare <4 x i32> @llvm.x86.xop.vphaddwd(<8 x i16>) nounwind readnone
336 %2 = call <2 x i64> @llvm.x86.xop.vphaddwq(<8 x i16> %a0)
339 declare <2 x i64> @llvm.x86.xop.vphaddwq(<8 x i16>) nounwind readnone
345 %2 = call <8 x i16> @llvm.x86.xop.vphsubbw(<16 x i8> %a0)
348 declare <8 x i16> @llvm.x86.xop.vphsubbw(<16 x i8>) nounwind readnone
354 %2 = call <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32> %a0)
357 declare <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32>) nounwind readnone
363 %2 = call <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16> %a0)
366 declare <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16>) nounwind readnone
372 %2 = call <4 x i32> @llvm.x86.xop.vpmacsdd(<4 x i32> %a0, <4 x i32> %a1, <4 x i32> %a2)
375 declare <4 x i32> @llvm.x86.xop.vpmacsdd(<4 x i32>, <4 x i32>, <4 x i32>) nounwind readnone
381 %2 = call <2 x i64> @llvm.x86.xop.vpmacsdqh(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2)
384 declare <2 x i64> @llvm.x86.xop.vpmacsdqh(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
390 %2 = call <2 x i64> @llvm.x86.xop.vpmacsdql(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2)
393 declare <2 x i64> @llvm.x86.xop.vpmacsdql(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
399 %2 = call <4 x i32> @llvm.x86.xop.vpmacssdd(<4 x i32> %a0, <4 x i32> %a1, <4 x i32> %a2)
402 declare <4 x i32> @llvm.x86.xop.vpmacssdd(<4 x i32>, <4 x i32>, <4 x i32>) nounwind readnone
408 %2 = call <2 x i64> @llvm.x86.xop.vpmacssdqh(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2)
411 declare <2 x i64> @llvm.x86.xop.vpmacssdqh(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
417 %2 = call <2 x i64> @llvm.x86.xop.vpmacssdql(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2)
420 declare <2 x i64> @llvm.x86.xop.vpmacssdql(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
426 %2 = call <4 x i32> @llvm.x86.xop.vpmacsswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2)
429 declare <4 x i32> @llvm.x86.xop.vpmacsswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
435 %2 = call <8 x i16> @llvm.x86.xop.vpmacssww(<8 x i16> %a0, <8 x i16> %a1, <8 x i16> %a2)
438 declare <8 x i16> @llvm.x86.xop.vpmacssww(<8 x i16>, <8 x i16>, <8 x i16>) nounwind readnone
444 %2 = call <4 x i32> @llvm.x86.xop.vpmacswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2)
447 declare <4 x i32> @llvm.x86.xop.vpmacswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
453 %2 = call <8 x i16> @llvm.x86.xop.vpmacsww(<8 x i16> %a0, <8 x i16> %a1, <8 x i16> %a2)
456 declare <8 x i16> @llvm.x86.xop.vpmacsww(<8 x i16>, <8 x i16>, <8 x i16>) nounwind readnone
462 %2 = call <4 x i32> @llvm.x86.xop.vpmadcsswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2)
465 declare <4 x i32> @llvm.x86.xop.vpmadcsswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
471 %2 = call <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2)
474 declare <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
480 %2 = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8> %a2)
487 %2 = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %a2, <16 x i8> %a1)
490 declare <16 x i8> @llvm.x86.xop.vpperm(<16 x i8>, <16 x i8>, <16 x i8>) nounwind readnone
496 %2 = call <16 x i8> @llvm.x86.xop.vprotbi(<16 x i8> %a0, i8 7)
499 declare <16 x i8> @llvm.x86.xop.vprotbi(<16 x i8>, i8) nounwind readnone
505 %2 = call <16 x i8> @llvm.x86.xop.vprotb(<16 x i8> %a0, <16 x i8> %a1)
512 %2 = call <16 x i8> @llvm.x86.xop.vprotb(<16 x i8> %a1, <16 x i8> %a0)
515 declare <16 x i8> @llvm.x86.xop.vprotb(<16 x i8>, <16 x i8>) nounwind readnone
521 %2 = call <4 x i32> @llvm.x86.xop.vprotdi(<4 x i32> %a0, i8 7)
524 declare <4 x i32> @llvm.x86.xop.vprotdi(<4 x i32>, i8) nounwind readnone
530 %2 = call <4 x i32> @llvm.x86.xop.vprotd(<4 x i32> %a0, <4 x i32> %a1)
537 %2 = call <4 x i32> @llvm.x86.xop.vprotd(<4 x i32> %a1, <4 x i32> %a0)
540 declare <4 x i32> @llvm.x86.xop.vprotd(<4 x i32>, <4 x i32>) nounwind readnone
546 %2 = call <2 x i64> @llvm.x86.xop.vprotqi(<2 x i64> %a0, i8 7)
549 declare <2 x i64> @llvm.x86.xop.vprotqi(<2 x i64>, i8) nounwind readnone
555 %2 = call <2 x i64> @llvm.x86.xop.vprotq(<2 x i64> %a0, <2 x i64> %a1)
562 %2 = call <2 x i64> @llvm.x86.xop.vprotq(<2 x i64> %a1, <2 x i64> %a0)
565 declare <2 x i64> @llvm.x86.xop.vprotq(<2 x i64>, <2 x i64>) nounwind readnone
571 %2 = call <8 x i16> @llvm.x86.xop.vprotwi(<8 x i16> %a0, i8 7)
574 declare <8 x i16> @llvm.x86.xop.vprotwi(<8 x i16>, i8) nounwind readnone
580 %2 = call <8 x i16> @llvm.x86.xop.vprotw(<8 x i16> %a0, <8 x i16> %a1)
587 %2 = call <8 x i16> @llvm.x86.xop.vprotw(<8 x i16> %a1, <8 x i16> %a0)
590 declare <8 x i16> @llvm.x86.xop.vprotw(<8 x i16>, <8 x i16>) nounwind readnone
596 %2 = call <16 x i8> @llvm.x86.xop.vpshab(<16 x i8> %a0, <16 x i8> %a1)
603 %2 = call <16 x i8> @llvm.x86.xop.vpshab(<16 x i8> %a1, <16 x i8> %a0)
606 declare <16 x i8> @llvm.x86.xop.vpshab(<16 x i8>, <16 x i8>) nounwind readnone
612 %2 = call <4 x i32> @llvm.x86.xop.vpshad(<4 x i32> %a0, <4 x i32> %a1)
619 %2 = call <4 x i32> @llvm.x86.xop.vpshad(<4 x i32> %a1, <4 x i32> %a0)
622 declare <4 x i32> @llvm.x86.xop.vpshad(<4 x i32>, <4 x i32>) nounwind readnone
628 %2 = call <2 x i64> @llvm.x86.xop.vpshaq(<2 x i64> %a0, <2 x i64> %a1)
635 %2 = call <2 x i64> @llvm.x86.xop.vpshaq(<2 x i64> %a1, <2 x i64> %a0)
638 declare <2 x i64> @llvm.x86.xop.vpshaq(<2 x i64>, <2 x i64>) nounwind readnone
644 %2 = call <8 x i16> @llvm.x86.xop.vpshaw(<8 x i16> %a0, <8 x i16> %a1)
651 %2 = call <8 x i16> @llvm.x86.xop.vpshaw(<8 x i16> %a1, <8 x i16> %a0)
654 declare <8 x i16> @llvm.x86.xop.vpshaw(<8 x i16>, <8 x i16>) nounwind readnone
660 %2 = call <16 x i8> @llvm.x86.xop.vpshlb(<16 x i8> %a0, <16 x i8> %a1)
667 %2 = call <16 x i8> @llvm.x86.xop.vpshlb(<16 x i8> %a1, <16 x i8> %a0)
670 declare <16 x i8> @llvm.x86.xop.vpshlb(<16 x i8>, <16 x i8>) nounwind readnone
676 %2 = call <4 x i32> @llvm.x86.xop.vpshld(<4 x i32> %a0, <4 x i32> %a1)
683 %2 = call <4 x i32> @llvm.x86.xop.vpshld(<4 x i32> %a1, <4 x i32> %a0)
686 declare <4 x i32> @llvm.x86.xop.vpshld(<4 x i32>, <4 x i32>) nounwind readnone
692 %2 = call <2 x i64> @llvm.x86.xop.vpshlq(<2 x i64> %a0, <2 x i64> %a1)
699 %2 = call <2 x i64> @llvm.x86.xop.vpshlq(<2 x i64> %a1, <2 x i64> %a0)
702 declare <2 x i64> @llvm.x86.xop.vpshlq(<2 x i64>, <2 x i64>) nounwind readnone
708 %2 = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %a0, <8 x i16> %a1)
715 %2 = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %a1, <8 x i16> %a0)
718 declare <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16>, <8 x i16>) nounwind readnone