[X86][AVX512] Add support for lowering shuffles to VPERMILPD

llvm-svn: 274450

2 files changed, 37 insertions, 52 deletions

diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index 02f92080431..f5557d340f6 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -11721,6 +11721,17 @@ static SDValue lowerV8F64VectorShuffle(const SDLoc &DL, ArrayRef<int> Mask,
     // Use low duplicate instructions for masks that match their pattern.
     if (isShuffleEquivalent(V1, V2, Mask, {0, 0, 2, 2, 4, 4, 6, 6}))
       return DAG.getNode(X86ISD::MOVDDUP, DL, MVT::v8f64, V1);
+
+    if (!is128BitLaneCrossingShuffleMask(MVT::v8f64, Mask)) {
+      // Non-half-crossing single input shuffles can be lowered with an
+      // interleaved permutation.
+      unsigned VPERMILPMask = (Mask[0] == 1) | ((Mask[1] == 1) << 1) |
+                              ((Mask[2] == 3) << 2) | ((Mask[3] == 3) << 3) |
+                              ((Mask[4] == 5) << 4) | ((Mask[5] == 5) << 5) |
+                              ((Mask[6] == 7) << 6) | ((Mask[7] == 7) << 7);
+      return DAG.getNode(X86ISD::VPERMILPI, DL, MVT::v8f64, V1,
+                         DAG.getConstant(VPERMILPMask, DL, MVT::i8));
+    }
   }
 
   if (SDValue Shuf128 =
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-512-v8.ll b/llvm/test/CodeGen/X86/vector-shuffle-512-v8.ll
index affd6bdb67b..e9b46b95061 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-512-v8.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-512-v8.ll
@@ -554,14 +554,12 @@ define <8 x double> @shuffle_v8f64_00234467(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_00234467:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [0,0,2,3,4,4,6,7]
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,0,2,3,4,4,6,7]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_00234467:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [0,0,0,0,2,0,3,0,4,0,4,0,6,0,7,0]
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,0,2,3,4,4,6,7]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 0, i32 2, i32 3, i32 4, i32 4, i32 6, i32 7>
   ret <8 x double> %shuffle
@@ -586,14 +584,12 @@ define <8 x double> @shuffle_v8f64_10325476(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_10325476:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,3,2,5,4,7,6]
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,3,2,5,4,7,6]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_10325476:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,0,0,3,0,2,0,5,0,4,0,7,0,6,0]
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,3,2,5,4,7,6]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 1, i32 0, i32 3, i32 2, i32 5, i32 4, i32 7, i32 6>
   ret <8 x double> %shuffle
@@ -603,14 +599,12 @@ define <8 x double> @shuffle_v8f64_11335577(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_11335577:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,1,3,3,5,5,7,7]
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,1,3,3,5,5,7,7]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_11335577:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,1,0,3,0,3,0,5,0,5,0,7,0,7,0]
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,1,3,3,5,5,7,7]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 1, i32 1, i32 3, i32 3, i32 5, i32 5, i32 7, i32 7>
   ret <8 x double> %shuffle
@@ -620,14 +614,12 @@ define <8 x double> @shuffle_v8f64_10235467(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_10235467:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,2,3,5,4,6,7]
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,2,3,5,4,6,7]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_10235467:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,0,0,2,0,3,0,5,0,4,0,6,0,7,0]
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,2,3,5,4,6,7]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 1, i32 0, i32 2, i32 3, i32 5, i32 4, i32 6, i32 7>
   ret <8 x double> %shuffle
@@ -637,14 +629,12 @@ define <8 x double> @shuffle_v8f64_10225466(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_10225466:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,2,2,5,4,6,6]
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,2,2,5,4,6,6]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_10225466:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,0,0,2,0,2,0,5,0,4,0,6,0,6,0]
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,2,2,5,4,6,6]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 1, i32 0, i32 2, i32 2, i32 5, i32 4, i32 6, i32 6>
   ret <8 x double> %shuffle
@@ -807,14 +797,12 @@ define <8 x double> @shuffle_v8f64_10324567(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_10324567:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,3,2,4,5,6,7]
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,3,2,4,5,6,7]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_10324567:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,0,0,3,0,2,0,4,0,5,0,6,0,7,0]
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,3,2,4,5,6,7]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 1, i32 0, i32 3, i32 2, i32 4, i32 5, i32 6, i32 7>
   ret <8 x double> %shuffle
@@ -824,14 +812,12 @@ define <8 x double> @shuffle_v8f64_11334567(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_11334567:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,1,3,3,4,5,6,7]
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,1,3,3,4,5,6,7]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_11334567:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [1,0,1,0,3,0,3,0,4,0,5,0,6,0,7,0]
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,1,3,3,4,5,6,7]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 1, i32 1, i32 3, i32 3, i32 4, i32 5, i32 6, i32 7>
   ret <8 x double> %shuffle
@@ -841,14 +827,12 @@ define <8 x double> @shuffle_v8f64_01235467(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_01235467:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [0,1,2,3,5,4,6,7]
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,1,2,3,5,4,6,7]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_01235467:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [0,0,1,0,2,0,3,0,5,0,4,0,6,0,7,0]
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,1,2,3,5,4,6,7]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 5, i32 4, i32 6, i32 7>
   ret <8 x double> %shuffle
@@ -858,14 +842,12 @@ define <8 x double> @shuffle_v8f64_01235466(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_01235466:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [0,1,2,3,5,4,6,6]
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,1,2,3,5,4,6,6]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_01235466:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = [0,0,1,0,2,0,3,0,5,0,4,0,6,0,6,0]
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,1,2,3,5,4,6,6]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 1, i32 2, i32 3, i32 5, i32 4, i32 6, i32 6>
   ret <8 x double> %shuffle
@@ -909,14 +891,12 @@ define <8 x double> @shuffle_v8f64_103245uu(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_103245uu:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <1,0,3,2,4,5,u,u>
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,3,2,4,5,6,6]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_103245uu:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <1,0,0,0,3,0,2,0,4,0,5,0,u,u,u,u>
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,0,3,2,4,5,6,6]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 1, i32 0, i32 3, i32 2, i32 4, i32 5, i32 undef, i32 undef>
   ret <8 x double> %shuffle
@@ -926,14 +906,12 @@ define <8 x double> @shuffle_v8f64_1133uu67(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_1133uu67:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <1,1,3,3,u,u,6,7>
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,1,3,3,4,4,6,7]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_1133uu67:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <1,0,1,0,3,0,3,0,u,u,u,u,6,0,7,0>
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[1,1,3,3,4,4,6,7]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 1, i32 1, i32 3, i32 3, i32 undef, i32 undef, i32 6, i32 7>
   ret <8 x double> %shuffle
@@ -943,14 +921,12 @@ define <8 x double> @shuffle_v8f64_0uu354uu(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_0uu354uu:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <0,u,u,3,5,4,u,u>
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,0,2,3,5,4,6,6]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_0uu354uu:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <0,0,u,u,u,u,3,0,5,0,4,0,u,u,u,u>
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,0,2,3,5,4,6,6]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 0, i32 undef, i32 undef, i32 3, i32 5, i32 4, i32 undef, i32 undef>
   ret <8 x double> %shuffle
@@ -960,14 +936,12 @@ define <8 x double> @shuffle_v8f64_uuu3uu66(<8 x double> %a, <8 x double> %b) {
 ;
 ; AVX512F-LABEL: shuffle_v8f64_uuu3uu66:
 ; AVX512F:       # BB#0:
-; AVX512F-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <u,u,u,3,u,u,6,6>
-; AVX512F-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,0,2,3,4,4,6,6]
 ; AVX512F-NEXT:    retq
 ;
 ; AVX512F-32-LABEL: shuffle_v8f64_uuu3uu66:
 ; AVX512F-32:       # BB#0:
-; AVX512F-32-NEXT:    vmovdqa64 {{.*#+}} zmm1 = <u,u,u,u,u,u,3,0,u,u,u,u,6,0,6,0>
-; AVX512F-32-NEXT:    vpermpd %zmm0, %zmm1, %zmm0
+; AVX512F-32-NEXT:    vpermilpd {{.*#+}} zmm0 = zmm0[0,0,2,3,4,4,6,6]
 ; AVX512F-32-NEXT:    retl
   %shuffle = shufflevector <8 x double> %a, <8 x double> %b, <8 x i32> <i32 undef, i32 undef, i32 undef, i32 3, i32 undef, i32 undef, i32 6, i32 6>
   ret <8 x double> %shuffle