From 37a18821cd8b0df5ac877659d2d397f0e268407c Mon Sep 17 00:00:00 2001
From: Chandler Carruth <chandlerc@gmail.com>
Date: Mon, 4 Aug 2014 01:14:24 +0000
Subject: [PATCH] [x86] Handle single input shuffles in the SSSE3 case more
 intelligently.

I spent some time looking into a better or more principled way to handle
this. For example, by detecting arbitrary "unneeded" ORs... But really,
there wasn't any point. We just shouldn't build blatantly wrong code so
late in the pipeline rather than adding more stages and logic later on
to fix it. Avoiding this is just too simple.

llvm-svn: 214680
---
 llvm/lib/Target/X86/X86ISelLowering.cpp         | 4 ++++
 llvm/test/CodeGen/X86/vector-shuffle-128-v16.ll | 4 ----
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/llvm/lib/Target/X86/X86ISelLowering.cpp b/llvm/lib/Target/X86/X86ISelLowering.cpp
index 5ae752c..94c19fb 100644
--- a/llvm/lib/Target/X86/X86ISelLowering.cpp
+++ b/llvm/lib/Target/X86/X86ISelLowering.cpp
@@ -7931,6 +7931,10 @@ static SDValue lowerV16I8VectorShuffle(SDValue Op, SDValue V1, SDValue V2,
       }
     V1 = DAG.getNode(X86ISD::PSHUFB, DL, MVT::v16i8, V1,
                      DAG.getNode(ISD::BUILD_VECTOR, DL, MVT::v16i8, V1Mask));
+    if (isSingleInputShuffleMask(Mask))
+      return V1; // Single inputs are easy.
+
+    // Otherwise, blend the two.
     V2 = DAG.getNode(X86ISD::PSHUFB, DL, MVT::v16i8, V2,
                      DAG.getNode(ISD::BUILD_VECTOR, DL, MVT::v16i8, V2Mask));
     return DAG.getNode(ISD::OR, DL, MVT::v16i8, V1, V2);
diff --git a/llvm/test/CodeGen/X86/vector-shuffle-128-v16.ll b/llvm/test/CodeGen/X86/vector-shuffle-128-v16.ll
index 49620e9..693a276 100644
--- a/llvm/test/CodeGen/X86/vector-shuffle-128-v16.ll
+++ b/llvm/test/CodeGen/X86/vector-shuffle-128-v16.ll
@@ -179,9 +179,7 @@ define <16 x i8> @shuffle_v16i8_03_02_01_00_07_06_05_04_11_10_09_08_15_14_13_12(
 ;
 ; SSSE3-LABEL: @shuffle_v16i8_03_02_01_00_07_06_05_04_11_10_09_08_15_14_13_12
 ; SSSE3:       # BB#0:
-; SSSE3-NEXT:    pshufb {{.*}} # xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
 ; SSSE3-NEXT:    pshufb {{.*}} # xmm0 = xmm0[3,2,1,0,7,6,5,4,11,10,9,8,15,14,13,12]
-; SSSE3-NEXT:    por %xmm1, %xmm0
 ; SSSE3-NEXT:    retq
   %shuffle = shufflevector <16 x i8> %a, <16 x i8> %b, <16 x i32> <i32 3, i32 2, i32 1, i32 0, i32 7, i32 6, i32 5, i32 4, i32 11, i32 10, i32 9, i32 8, i32 15, i32 14, i32 13, i32 12>
   ret <16 x i8> %shuffle
@@ -275,9 +273,7 @@ define <16 x i8> @trunc_v4i32_shuffle(<16 x i8> %a) {
 ;
 ; SSSE3-LABEL: @trunc_v4i32_shuffle
 ; SSSE3:       # BB#0:
-; SSSE3-NEXT:    pshufb {{.*}} # xmm1 = zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
 ; SSSE3-NEXT:    pshufb {{.*}} # xmm0 = xmm0[0,4,8,12],zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero
-; SSSE3-NEXT:    por %xmm1, %xmm0
 ; SSSE3-NEXT:    retq
   %shuffle = shufflevector <16 x i8> %a, <16 x i8> undef, <16 x i32> <i32 0, i32 4, i32 8, i32 12, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef, i32 undef>
   ret <16 x i8> %shuffle
-- 
2.7.4