[X86] Run XOP vector rotation tests with/without AVX2

I noticed this while reviewing D77152 - by only testing bdver4 we weren't checking an XOP target that only had AVX1

Author: RKSimon

llvm/test/CodeGen/X86/rotate_vec.ll

@@ -1,5 +1,6 @@
 ; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py
-; RUN: llc < %s -mtriple=x86_64-unknown -mcpu=bdver4 | FileCheck %s --check-prefixes=CHECK,XOP
+; RUN: llc < %s -mtriple=x86_64-unknown -mcpu=bdver2 | FileCheck %s --check-prefixes=CHECK,XOP,XOPAVX1
+; RUN: llc < %s -mtriple=x86_64-unknown -mcpu=bdver4 | FileCheck %s --check-prefixes=CHECK,XOP,XOPAVX2
 ; RUN: llc < %s -mtriple=x86_64-unknown -mcpu=skylake-avx512 | FileCheck %s --check-prefixes=CHECK,AVX512
 
 define <4 x i32> @rot_v4i32_splat(<4 x i32> %x) {
@@ -77,10 +78,20 @@ define <4 x i32> @rot_v4i32_non_splat_2masks(<4 x i32> %x) {
 }
 
 define <4 x i32> @rot_v4i32_zero_non_splat(<4 x i32> %x) {
-; CHECK-LABEL: rot_v4i32_zero_non_splat:
-; CHECK:       # %bb.0:
-; CHECK-NEXT:    vbroadcastss %xmm0, %xmm0
-; CHECK-NEXT:    retq
+; XOPAVX1-LABEL: rot_v4i32_zero_non_splat:
+; XOPAVX1:       # %bb.0:
+; XOPAVX1-NEXT:    vpermilps {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; XOPAVX1-NEXT:    retq
+;
+; XOPAVX2-LABEL: rot_v4i32_zero_non_splat:
+; XOPAVX2:       # %bb.0:
+; XOPAVX2-NEXT:    vbroadcastss %xmm0, %xmm0
+; XOPAVX2-NEXT:    retq
+;
+; AVX512-LABEL: rot_v4i32_zero_non_splat:
+; AVX512:       # %bb.0:
+; AVX512-NEXT:    vbroadcastss %xmm0, %xmm0
+; AVX512-NEXT:    retq
   %1 = call <4 x i32> @llvm.fshl.v4i32(<4 x i32> %x, <4 x i32> %x, <4 x i32> <i32 0, i32 1, i32 2, i32 3>)
   %2 = shufflevector <4 x i32> %1, <4 x i32> undef, <4 x i32> zeroinitializer
   ret <4 x i32> %2
@@ -97,12 +108,19 @@ define <4 x i32> @rot_v4i32_allsignbits(<4 x i32> %x, <4 x i32> %y) {
 }
 
 define <4 x i32> @rot_v4i32_mask_ashr0(<4 x i32> %a0) {
-; XOP-LABEL: rot_v4i32_mask_ashr0:
-; XOP:       # %bb.0:
-; XOP-NEXT:    vpsravd {{.*}}(%rip), %xmm0, %xmm0
-; XOP-NEXT:    vprotd $1, %xmm0, %xmm0
-; XOP-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
-; XOP-NEXT:    retq
+; XOPAVX1-LABEL: rot_v4i32_mask_ashr0:
+; XOPAVX1:       # %bb.0:
+; XOPAVX1-NEXT:    vpshad {{.*}}(%rip), %xmm0, %xmm0
+; XOPAVX1-NEXT:    vprotd $1, %xmm0, %xmm0
+; XOPAVX1-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; XOPAVX1-NEXT:    retq
+;
+; XOPAVX2-LABEL: rot_v4i32_mask_ashr0:
+; XOPAVX2:       # %bb.0:
+; XOPAVX2-NEXT:    vpsravd {{.*}}(%rip), %xmm0, %xmm0
+; XOPAVX2-NEXT:    vprotd $1, %xmm0, %xmm0
+; XOPAVX2-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; XOPAVX2-NEXT:    retq
 ;
 ; AVX512-LABEL: rot_v4i32_mask_ashr0:
 ; AVX512:       # %bb.0:
@@ -118,13 +136,21 @@ define <4 x i32> @rot_v4i32_mask_ashr0(<4 x i32> %a0) {
 }
 
 define <4 x i32> @rot_v4i32_mask_ashr1(<4 x i32> %a0) {
-; XOP-LABEL: rot_v4i32_mask_ashr1:
-; XOP:       # %bb.0:
-; XOP-NEXT:    vpsrad $25, %xmm0, %xmm0
-; XOP-NEXT:    vprotd $1, %xmm0, %xmm0
-; XOP-NEXT:    vpbroadcastd %xmm0, %xmm0
-; XOP-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
-; XOP-NEXT:    retq
+; XOPAVX1-LABEL: rot_v4i32_mask_ashr1:
+; XOPAVX1:       # %bb.0:
+; XOPAVX1-NEXT:    vpsrad $25, %xmm0, %xmm0
+; XOPAVX1-NEXT:    vprotd $1, %xmm0, %xmm0
+; XOPAVX1-NEXT:    vpshufd {{.*#+}} xmm0 = xmm0[0,0,0,0]
+; XOPAVX1-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; XOPAVX1-NEXT:    retq
+;
+; XOPAVX2-LABEL: rot_v4i32_mask_ashr1:
+; XOPAVX2:       # %bb.0:
+; XOPAVX2-NEXT:    vpsrad $25, %xmm0, %xmm0
+; XOPAVX2-NEXT:    vprotd $1, %xmm0, %xmm0
+; XOPAVX2-NEXT:    vpbroadcastd %xmm0, %xmm0
+; XOPAVX2-NEXT:    vpand {{.*}}(%rip), %xmm0, %xmm0
+; XOPAVX2-NEXT:    retq
 ;
 ; AVX512-LABEL: rot_v4i32_mask_ashr1:
 ; AVX512:       # %bb.0: