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 …
133 %res = call <4 x i32> @llvm.x86.xop.vphaddbd(<16 x i8> %a0) ;
136 declare <4 x i32> @llvm.x86.xop.vphaddbd(<16 x i8>) nounwind readnone
143 %res = call <2 x i64> @llvm.x86.xop.vphaddbq(<16 x i8> %a0) ;
146 declare <2 x i64> @llvm.x86.xop.vphaddbq(<16 x i8>) nounwind readnone
153 %res = call <8 x i16> @llvm.x86.xop.vphaddbw(<16 x i8> %a0) ;
156 declare <8 x i16> @llvm.x86.xop.vphaddbw(<16 x i8>) nounwind readnone
163 %res = call <2 x i64> @llvm.x86.xop.vphadddq(<4 x i32> %a0) ;
166 declare <2 x i64> @llvm.x86.xop.vphadddq(<4 x i32>) nounwind readnone
173 %res = call <4 x i32> @llvm.x86.xop.vphaddubd(<16 x i8> %a0) ;
176 declare <4 x i32> @llvm.x86.xop.vphaddubd(<16 x i8>) nounwind readnone
183 %res = call <2 x i64> @llvm.x86.xop.vphaddubq(<16 x i8> %a0) ;
186 declare <2 x i64> @llvm.x86.xop.vphaddubq(<16 x i8>) nounwind readnone
193 %res = call <8 x i16> @llvm.x86.xop.vphaddubw(<16 x i8> %a0) ;
196 declare <8 x i16> @llvm.x86.xop.vphaddubw(<16 x i8>) nounwind readnone
203 %res = call <2 x i64> @llvm.x86.xop.vphaddudq(<4 x i32> %a0) ;
206 declare <2 x i64> @llvm.x86.xop.vphaddudq(<4 x i32>) nounwind readnone
213 %res = call <4 x i32> @llvm.x86.xop.vphadduwd(<8 x i16> %a0) ;
216 declare <4 x i32> @llvm.x86.xop.vphadduwd(<8 x i16>) nounwind readnone
223 %res = call <2 x i64> @llvm.x86.xop.vphadduwq(<8 x i16> %a0) ;
226 declare <2 x i64> @llvm.x86.xop.vphadduwq(<8 x i16>) nounwind readnone
233 %res = call <4 x i32> @llvm.x86.xop.vphaddwd(<8 x i16> %a0) ;
236 declare <4 x i32> @llvm.x86.xop.vphaddwd(<8 x i16>) nounwind readnone
243 %res = call <2 x i64> @llvm.x86.xop.vphaddwq(<8 x i16> %a0) ;
246 declare <2 x i64> @llvm.x86.xop.vphaddwq(<8 x i16>) nounwind readnone
253 %res = call <8 x i16> @llvm.x86.xop.vphsubbw(<16 x i8> %a0) ;
256 declare <8 x i16> @llvm.x86.xop.vphsubbw(<16 x i8>) nounwind readnone
263 %res = call <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32> %a0) ;
272 %res = call <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32> %vec) ;
275 declare <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32>) nounwind readnone
282 %res = call <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16> %a0) ;
291 %res = call <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16> %vec) ;
294 declare <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16>) nounwind readnone
301 %res = call <4 x i32> @llvm.x86.xop.vpmacsdd(<4 x i32> %a0, <4 x i32> %a1, <4 x i32> %a2) ;
304 declare <4 x i32> @llvm.x86.xop.vpmacsdd(<4 x i32>, <4 x i32>, <4 x i32>) nounwind readnone
311 %res = call <2 x i64> @llvm.x86.xop.vpmacsdqh(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2) ;
314 declare <2 x i64> @llvm.x86.xop.vpmacsdqh(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
321 %res = call <2 x i64> @llvm.x86.xop.vpmacsdql(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2) ;
324 declare <2 x i64> @llvm.x86.xop.vpmacsdql(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
331 %res = call <4 x i32> @llvm.x86.xop.vpmacssdd(<4 x i32> %a0, <4 x i32> %a1, <4 x i32> %a2) ;
334 declare <4 x i32> @llvm.x86.xop.vpmacssdd(<4 x i32>, <4 x i32>, <4 x i32>) nounwind readnone
341 %res = call <2 x i64> @llvm.x86.xop.vpmacssdqh(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2) ;
344 declare <2 x i64> @llvm.x86.xop.vpmacssdqh(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
351 %res = call <2 x i64> @llvm.x86.xop.vpmacssdql(<4 x i32> %a0, <4 x i32> %a1, <2 x i64> %a2) ;
354 declare <2 x i64> @llvm.x86.xop.vpmacssdql(<4 x i32>, <4 x i32>, <2 x i64>) nounwind readnone
361 %res = call <4 x i32> @llvm.x86.xop.vpmacsswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2) ;
364 declare <4 x i32> @llvm.x86.xop.vpmacsswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
371 %res = call <8 x i16> @llvm.x86.xop.vpmacssww(<8 x i16> %a0, <8 x i16> %a1, <8 x i16> %a2) ;
374 declare <8 x i16> @llvm.x86.xop.vpmacssww(<8 x i16>, <8 x i16>, <8 x i16>) nounwind readnone
381 %res = call <4 x i32> @llvm.x86.xop.vpmacswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2) ;
384 declare <4 x i32> @llvm.x86.xop.vpmacswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
391 %res = call <8 x i16> @llvm.x86.xop.vpmacsww(<8 x i16> %a0, <8 x i16> %a1, <8 x i16> %a2) ;
394 declare <8 x i16> @llvm.x86.xop.vpmacsww(<8 x i16>, <8 x i16>, <8 x i16>) nounwind readnone
401 %res = call <4 x i32> @llvm.x86.xop.vpmadcsswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2) ;
404 declare <4 x i32> @llvm.x86.xop.vpmadcsswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
411 %res = call <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16> %a0, <8 x i16> %a1, <4 x i32> %a2) ;
420 %res = call <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16> %a0, <8 x i16> %vec, <4 x i32> %a2) ;
423 declare <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16>, <8 x i16>, <4 x i32>) nounwind readnone
430 %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8> %a2) ;
439 %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8> %vec) ;
448 %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %vec, <16 x i8> %a2) ;
451 declare <16 x i8> @llvm.x86.xop.vpperm(<16 x i8>, <16 x i8>, <16 x i8>) nounwind readnone
458 %res = call <16 x i8> @llvm.x86.xop.vpshab(<16 x i8> %a0, <16 x i8> %a1) ;
461 declare <16 x i8> @llvm.x86.xop.vpshab(<16 x i8>, <16 x i8>) nounwind readnone
468 %res = call <4 x i32> @llvm.x86.xop.vpshad(<4 x i32> %a0, <4 x i32> %a1) ;
471 declare <4 x i32> @llvm.x86.xop.vpshad(<4 x i32>, <4 x i32>) nounwind readnone
478 %res = call <2 x i64> @llvm.x86.xop.vpshaq(<2 x i64> %a0, <2 x i64> %a1) ;
481 declare <2 x i64> @llvm.x86.xop.vpshaq(<2 x i64>, <2 x i64>) nounwind readnone
488 %res = call <8 x i16> @llvm.x86.xop.vpshaw(<8 x i16> %a0, <8 x i16> %a1) ;
491 declare <8 x i16> @llvm.x86.xop.vpshaw(<8 x i16>, <8 x i16>) nounwind readnone
498 %res = call <16 x i8> @llvm.x86.xop.vpshlb(<16 x i8> %a0, <16 x i8> %a1) ;
501 declare <16 x i8> @llvm.x86.xop.vpshlb(<16 x i8>, <16 x i8>) nounwind readnone
508 %res = call <4 x i32> @llvm.x86.xop.vpshld(<4 x i32> %a0, <4 x i32> %a1) ;
511 declare <4 x i32> @llvm.x86.xop.vpshld(<4 x i32>, <4 x i32>) nounwind readnone
518 %res = call <2 x i64> @llvm.x86.xop.vpshlq(<2 x i64> %a0, <2 x i64> %a1) ;
521 declare <2 x i64> @llvm.x86.xop.vpshlq(<2 x i64>, <2 x i64>) nounwind readnone
528 %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %a0, <8 x i16> %a1) ;
537 %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %a0, <8 x i16> %vec) ;
546 %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %vec, <8 x i16> %a1) ;
549 declare <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16>, <8 x i16>) nounwind readnone
556 %res = call <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float> %a0) ;
566 %res = call <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float> %vec) ;
569 declare <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float>) nounwind readnone
576 %res = call <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double> %a0) ;
586 %res = call <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double> %vec) ;
589 declare <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double>) nounwind readnone
596 %res = call <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double> %a0) ;
605 %res = call <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double> %vec) ;
608 declare <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double>) nounwind readnone
615 %res = call <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double> %a0) ;
624 %res = call <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double> %vec) ;
627 declare <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double>) nounwind readnone
634 %res = call <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float> %a0) ;
643 %res = call <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float> %vec) ;
646 declare <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float>) nounwind readnone
653 %res = call <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float> %a0) ;
662 %res = call <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float> %vec) ;
665 declare <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float>) nounwind readnone