[X86][XOP] Add VPROT rotate by immediate intrinsics tests

[oota-llvm.git] / test / CodeGen / X86 / xop-intrinsics-x86_64.ll

diff --git a/test/CodeGen/X86/xop-intrinsics-x86_64.ll b/test/CodeGen/X86/xop-intrinsics-x86_64.ll
index e154e4a159b5b354b34a63ebe9c0cd7abc08b291..e96fed36d15c78d669e99dd2551e26ae641a46c9 100644 (file)
--- a/test/CodeGen/X86/xop-intrinsics-x86_64.ll
+++ b/test/CodeGen/X86/xop-intrinsics-x86_64.ll
@@ -8,14 +8,14 @@ define <2 x double> @test_int_x86_xop_vpermil2pd(<2 x double> %a0, <2 x double>
 define <2 x double> @test_int_x86_xop_vpermil2pd_mr(<2 x double> %a0, <2 x double>* %a1, <2 x double> %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpermil2pd
 define <2 x double> @test_int_x86_xop_vpermil2pd_mr(<2 x double> %a0, <2 x double>* %a1, <2 x double> %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpermil2pd

-  %vec = load <2 x double>* %a1
+  %vec = load <2 x double>, <2 x double>* %a1

   %res = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %vec, <2 x double> %a2, i8 1) ;  [#uses=1]
   ret <2 x double> %res
 }
 define <2 x double> @test_int_x86_xop_vpermil2pd_rm(<2 x double> %a0, <2 x double> %a1, <2 x double>* %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpermil2pd
   %res = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %vec, <2 x double> %a2, i8 1) ;  [#uses=1]
   ret <2 x double> %res
 }
 define <2 x double> @test_int_x86_xop_vpermil2pd_rm(<2 x double> %a0, <2 x double> %a1, <2 x double>* %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpermil2pd

-  %vec = load <2 x double>* %a2
+  %vec = load <2 x double>, <2 x double>* %a2

   %res = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %a1, <2 x double> %vec, i8 1) ;  [#uses=1]
   ret <2 x double> %res
 }
   %res = call <2 x double> @llvm.x86.xop.vpermil2pd(<2 x double> %a0, <2 x double> %a1, <2 x double> %vec, i8 1) ;  [#uses=1]
   ret <2 x double> %res
 }


@@ -31,7 +31,7 @@ define <4 x double> @test_int_x86_xop_vpermil2pd_256_mr(<4 x double> %a0, <4 x d
   ; CHECK-NOT: vmovaps
   ; CHECK: vpermil2pd
   ; CHECK: ymm
   ; CHECK-NOT: vmovaps
   ; CHECK: vpermil2pd
   ; CHECK: ymm

-  %vec = load <4 x double>* %a1
+  %vec = load <4 x double>, <4 x double>* %a1

   %res = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %vec, <4 x double> %a2, i8 2) ;
   ret <4 x double> %res
 }
   %res = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %vec, <4 x double> %a2, i8 2) ;
   ret <4 x double> %res
 }


@@ -39,7 +39,7 @@ define <4 x double> @test_int_x86_xop_vpermil2pd_256_rm(<4 x double> %a0, <4 x d
   ; CHECK-NOT: vmovaps
   ; CHECK: vpermil2pd
   ; CHECK: ymm
   ; CHECK-NOT: vmovaps
   ; CHECK: vpermil2pd
   ; CHECK: ymm

-  %vec = load <4 x double>* %a2
+  %vec = load <4 x double>, <4 x double>* %a2

   %res = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %a1, <4 x double> %vec, i8 2) ;
   ret <4 x double> %res
 }
   %res = call <4 x double> @llvm.x86.xop.vpermil2pd.256(<4 x double> %a0, <4 x double> %a1, <4 x double> %vec, i8 2) ;
   ret <4 x double> %res
 }


@@ -77,7 +77,7 @@ define <4 x i64> @test_int_x86_xop_vpcmov_256_mr(<4 x i64> %a0, <4 x i64>* %a1,
   ; CHECK-NOT: vmovaps
   ; CHECK: vpcmov
   ; CHECK: ymm
   ; CHECK-NOT: vmovaps
   ; CHECK: vpcmov
   ; CHECK: ymm

-  %vec = load <4 x i64>* %a1
+  %vec = load <4 x i64>, <4 x i64>* %a1

   %res = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %vec, <4 x i64> %a2) ;
   ret <4 x i64> %res
 }
   %res = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %vec, <4 x i64> %a2) ;
   ret <4 x i64> %res
 }


@@ -85,7 +85,7 @@ define <4 x i64> @test_int_x86_xop_vpcmov_256_rm(<4 x i64> %a0, <4 x i64> %a1, <
   ; CHECK-NOT: vmovaps
   ; CHECK: vpcmov
   ; CHECK: ymm
   ; CHECK-NOT: vmovaps
   ; CHECK: vpcmov
   ; CHECK: ymm

- %vec = load <4 x i64>* %a2
+ %vec = load <4 x i64>, <4 x i64>* %a2

  %res = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %a1, <4 x i64> %vec) ;
   ret <4 x i64> %res
 }
  %res = call <4 x i64> @llvm.x86.xop.vpcmov.256(<4 x i64> %a0, <4 x i64> %a1, <4 x i64> %vec) ;
   ret <4 x i64> %res
 }


@@ -99,7 +99,7 @@ define <16 x i8> @test_int_x86_xop_vpcomeqb(<16 x i8> %a0, <16 x i8> %a1) {
 define <16 x i8> @test_int_x86_xop_vpcomeqb_mem(<16 x i8> %a0, <16 x i8>* %a1) {
   ; CHECK-NOT: vmovaps
   ; CHECK:vpcomeqb
 define <16 x i8> @test_int_x86_xop_vpcomeqb_mem(<16 x i8> %a0, <16 x i8>* %a1) {
   ; CHECK-NOT: vmovaps
   ; CHECK:vpcomeqb

-  %vec = load <16 x i8>* %a1
+  %vec = load <16 x i8>, <16 x i8>* %a1

   %res = call <16 x i8> @llvm.x86.xop.vpcomeqb(<16 x i8> %a0, <16 x i8> %vec) ;
   ret <16 x i8> %res
 }
   %res = call <16 x i8> @llvm.x86.xop.vpcomeqb(<16 x i8> %a0, <16 x i8> %vec) ;
   ret <16 x i8> %res
 }


@@ -645,7 +645,7 @@ define <2 x i64> @test_int_x86_xop_vphsubdq(<4 x i32> %a0) {
 define <2 x i64> @test_int_x86_xop_vphsubdq_mem(<4 x i32>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vphsubdq
 define <2 x i64> @test_int_x86_xop_vphsubdq_mem(<4 x i32>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vphsubdq

-  %vec = load <4 x i32>* %a0
+  %vec = load <4 x i32>, <4 x i32>* %a0

   %res = call <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32> %vec) ;
   ret <2 x i64> %res
 }
   %res = call <2 x i64> @llvm.x86.xop.vphsubdq(<4 x i32> %vec) ;
   ret <2 x i64> %res
 }


@@ -659,7 +659,7 @@ define <4 x i32> @test_int_x86_xop_vphsubwd(<8 x i16> %a0) {
 define <4 x i32> @test_int_x86_xop_vphsubwd_mem(<8 x i16>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vphsubwd
 define <4 x i32> @test_int_x86_xop_vphsubwd_mem(<8 x i16>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vphsubwd

-  %vec = load <8 x i16>* %a0
+  %vec = load <8 x i16>, <8 x i16>* %a0

   %res = call <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16> %vec) ;
   ret <4 x i32> %res
 }
   %res = call <4 x i32> @llvm.x86.xop.vphsubwd(<8 x i16> %vec) ;
   ret <4 x i32> %res
 }


@@ -750,7 +750,7 @@ define <4 x i32> @test_int_x86_xop_vpmadcswd(<8 x i16> %a0, <8 x i16> %a1, <4 x
 define <4 x i32> @test_int_x86_xop_vpmadcswd_mem(<8 x i16> %a0, <8 x i16>* %a1, <4 x i32> %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpmadcswd
 define <4 x i32> @test_int_x86_xop_vpmadcswd_mem(<8 x i16> %a0, <8 x i16>* %a1, <4 x i32> %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpmadcswd

-  %vec = load <8 x i16>* %a1
+  %vec = load <8 x i16>, <8 x i16>* %a1

   %res = call <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16> %a0, <8 x i16> %vec, <4 x i32> %a2) ;
   ret <4 x i32> %res
 }
   %res = call <4 x i32> @llvm.x86.xop.vpmadcswd(<8 x i16> %a0, <8 x i16> %vec, <4 x i32> %a2) ;
   ret <4 x i32> %res
 }


@@ -764,14 +764,14 @@ define <16 x i8> @test_int_x86_xop_vpperm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8
 define <16 x i8> @test_int_x86_xop_vpperm_rm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8>* %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpperm
 define <16 x i8> @test_int_x86_xop_vpperm_rm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8>* %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpperm

-  %vec = load <16 x i8>* %a2
+  %vec = load <16 x i8>, <16 x i8>* %a2

   %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8> %vec) ;
   ret <16 x i8> %res
 }
 define <16 x i8> @test_int_x86_xop_vpperm_mr(<16 x i8> %a0, <16 x i8>* %a1, <16 x i8> %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpperm
   %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %a1, <16 x i8> %vec) ;
   ret <16 x i8> %res
 }
 define <16 x i8> @test_int_x86_xop_vpperm_mr(<16 x i8> %a0, <16 x i8>* %a1, <16 x i8> %a2) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpperm

-  %vec = load <16 x i8>* %a1
+  %vec = load <16 x i8>, <16 x i8>* %a1

   %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %vec, <16 x i8> %a2) ;
   ret <16 x i8> %res
 }
   %res = call <16 x i8> @llvm.x86.xop.vpperm(<16 x i8> %a0, <16 x i8> %vec, <16 x i8> %a2) ;
   ret <16 x i8> %res
 }


@@ -805,6 +805,34 @@ define <8 x i16> @test_int_x86_xop_vprotw(<8 x i16> %a0, <8 x i16> %a1) {
 }
 declare <8 x i16> @llvm.x86.xop.vprotw(<8 x i16>, <8 x i16>) nounwind readnone
 
 }
 declare <8 x i16> @llvm.x86.xop.vprotw(<8 x i16>, <8 x i16>) nounwind readnone
 

+define <16 x i8> @test_int_x86_xop_vprotbi(<16 x i8> %a0) {
+  ; CHECK: vprotb
+  %res = call <16 x i8> @llvm.x86.xop.vprotbi(<16 x i8> %a0, i8 1) ;
+  ret <16 x i8> %res
+}
+declare <16 x i8> @llvm.x86.xop.vprotbi(<16 x i8>, i8) nounwind readnone
+
+define <4 x i32> @test_int_x86_xop_vprotdi(<4 x i32> %a0) {
+  ; CHECK: vprotd
+  %res = call <4 x i32> @llvm.x86.xop.vprotdi(<4 x i32> %a0, i8 -2) ;
+  ret <4 x i32> %res
+}
+declare <4 x i32> @llvm.x86.xop.vprotdi(<4 x i32>, i8) nounwind readnone
+
+define <2 x i64> @test_int_x86_xop_vprotqi(<2 x i64> %a0) {
+  ; CHECK: vprotq
+  %res = call <2 x i64> @llvm.x86.xop.vprotqi(<2 x i64> %a0, i8 3) ;
+  ret <2 x i64> %res
+}
+declare <2 x i64> @llvm.x86.xop.vprotqi(<2 x i64>, i8) nounwind readnone
+
+define <8 x i16> @test_int_x86_xop_vprotwi(<8 x i16> %a0) {
+  ; CHECK: vprotw
+  %res = call <8 x i16> @llvm.x86.xop.vprotwi(<8 x i16> %a0, i8 -4) ;
+  ret <8 x i16> %res
+}
+declare <8 x i16> @llvm.x86.xop.vprotwi(<8 x i16>, i8) nounwind readnone
+

 define <16 x i8> @test_int_x86_xop_vpshab(<16 x i8> %a0, <16 x i8> %a1) {
   ; CHECK: vpshab
   %res = call <16 x i8> @llvm.x86.xop.vpshab(<16 x i8> %a0, <16 x i8> %a1) ;
 define <16 x i8> @test_int_x86_xop_vpshab(<16 x i8> %a0, <16 x i8> %a1) {
   ; CHECK: vpshab
   %res = call <16 x i8> @llvm.x86.xop.vpshab(<16 x i8> %a0, <16 x i8> %a1) ;


@@ -862,14 +890,14 @@ define <8 x i16> @test_int_x86_xop_vpshlw(<8 x i16> %a0, <8 x i16> %a1) {
 define <8 x i16> @test_int_x86_xop_vpshlw_rm(<8 x i16> %a0, <8 x i16>* %a1) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpshlw
 define <8 x i16> @test_int_x86_xop_vpshlw_rm(<8 x i16> %a0, <8 x i16>* %a1) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpshlw

-  %vec = load <8 x i16>* %a1
+  %vec = load <8 x i16>, <8 x i16>* %a1

   %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %a0, <8 x i16> %vec) ;
   ret <8 x i16> %res
 }
 define <8 x i16> @test_int_x86_xop_vpshlw_mr(<8 x i16>* %a0, <8 x i16> %a1) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpshlw
   %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %a0, <8 x i16> %vec) ;
   ret <8 x i16> %res
 }
 define <8 x i16> @test_int_x86_xop_vpshlw_mr(<8 x i16>* %a0, <8 x i16> %a1) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vpshlw

-  %vec = load <8 x i16>* %a0
+  %vec = load <8 x i16>, <8 x i16>* %a0

   %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %vec, <8 x i16> %a1) ;
   ret <8 x i16> %res
 }
   %res = call <8 x i16> @llvm.x86.xop.vpshlw(<8 x i16> %vec, <8 x i16> %a1) ;
   ret <8 x i16> %res
 }


@@ -884,7 +912,7 @@ define <4 x float> @test_int_x86_xop_vfrcz_ss(<4 x float> %a0) {
 define <4 x float> @test_int_x86_xop_vfrcz_ss_mem(float* %a0) {
   ; CHECK-NOT: mov
   ; CHECK: vfrczss
 define <4 x float> @test_int_x86_xop_vfrcz_ss_mem(float* %a0) {
   ; CHECK-NOT: mov
   ; CHECK: vfrczss

-  %elem = load float* %a0
+  %elem = load float, float* %a0

   %vec = insertelement <4 x float> undef, float %elem, i32 0
   %res = call <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float> %vec) ;
   ret <4 x float> %res
   %vec = insertelement <4 x float> undef, float %elem, i32 0
   %res = call <4 x float> @llvm.x86.xop.vfrcz.ss(<4 x float> %vec) ;
   ret <4 x float> %res


@@ -900,7 +928,7 @@ define <2 x double> @test_int_x86_xop_vfrcz_sd(<2 x double> %a0) {
 define <2 x double> @test_int_x86_xop_vfrcz_sd_mem(double* %a0) {
   ; CHECK-NOT: mov
   ; CHECK: vfrczsd
 define <2 x double> @test_int_x86_xop_vfrcz_sd_mem(double* %a0) {
   ; CHECK-NOT: mov
   ; CHECK: vfrczsd

-  %elem = load double* %a0
+  %elem = load double, double* %a0

   %vec = insertelement <2 x double> undef, double %elem, i32 0
   %res = call <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double> %vec) ;
   ret <2 x double> %res
   %vec = insertelement <2 x double> undef, double %elem, i32 0
   %res = call <2 x double> @llvm.x86.xop.vfrcz.sd(<2 x double> %vec) ;
   ret <2 x double> %res


@@ -915,7 +943,7 @@ define <2 x double> @test_int_x86_xop_vfrcz_pd(<2 x double> %a0) {
 define <2 x double> @test_int_x86_xop_vfrcz_pd_mem(<2 x double>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vfrczpd
 define <2 x double> @test_int_x86_xop_vfrcz_pd_mem(<2 x double>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vfrczpd

-  %vec = load <2 x double>* %a0
+  %vec = load <2 x double>, <2 x double>* %a0

   %res = call <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double> %vec) ;
   ret <2 x double> %res
 }
   %res = call <2 x double> @llvm.x86.xop.vfrcz.pd(<2 x double> %vec) ;
   ret <2 x double> %res
 }


@@ -931,7 +959,7 @@ define <4 x double> @test_int_x86_xop_vfrcz_pd_256_mem(<4 x double>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vfrczpd
   ; CHECK: ymm
   ; CHECK-NOT: vmovaps
   ; CHECK: vfrczpd
   ; CHECK: ymm

-  %vec = load <4 x double>* %a0
+  %vec = load <4 x double>, <4 x double>* %a0

   %res = call <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double> %vec) ;
   ret <4 x double> %res
 }
   %res = call <4 x double> @llvm.x86.xop.vfrcz.pd.256(<4 x double> %vec) ;
   ret <4 x double> %res
 }


@@ -945,7 +973,7 @@ define <4 x float> @test_int_x86_xop_vfrcz_ps(<4 x float> %a0) {
 define <4 x float> @test_int_x86_xop_vfrcz_ps_mem(<4 x float>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vfrczps
 define <4 x float> @test_int_x86_xop_vfrcz_ps_mem(<4 x float>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vfrczps

-  %vec = load <4 x float>* %a0
+  %vec = load <4 x float>, <4 x float>* %a0

   %res = call <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float> %vec) ;
   ret <4 x float> %res
 }
   %res = call <4 x float> @llvm.x86.xop.vfrcz.ps(<4 x float> %vec) ;
   ret <4 x float> %res
 }


@@ -961,7 +989,7 @@ define <8 x float> @test_int_x86_xop_vfrcz_ps_256_mem(<8 x float>* %a0) {
   ; CHECK-NOT: vmovaps
   ; CHECK: vfrczps
   ; CHECK: ymm
   ; CHECK-NOT: vmovaps
   ; CHECK: vfrczps
   ; CHECK: ymm

-  %vec = load <8 x float>* %a0
+  %vec = load <8 x float>, <8 x float>* %a0

   %res = call <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float> %vec) ;
   ret <8 x float> %res
 }
   %res = call <8 x float> @llvm.x86.xop.vfrcz.ps.256(<8 x float> %vec) ;
   ret <8 x float> %res
 }