From 768edf3cd037aab10391abc279f71470df8e3156 Mon Sep 17 00:00:00 2001
From: Hal Finkel <hfinkel@anl.gov>
Date: Sun, 24 Jun 2012 13:28:01 +0000
Subject: [PATCH] Allow controlling vectorization of boolean values separately
 from other integer types.

These are used as the result of comparisons, and often handled differently from larger integer types.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@159111 91177308-0d34-0410-b5e6-96231b3b80d8
---
 include/llvm/Transforms/Vectorize.h       |  3 +++
 lib/Transforms/Vectorize/BBVectorize.cpp  | 18 ++++++++++++++----
 test/Transforms/BBVectorize/simple-sel.ll |  3 +++
 3 files changed, 20 insertions(+), 4 deletions(-)

diff --git a/include/llvm/Transforms/Vectorize.h b/include/llvm/Transforms/Vectorize.h
index d73db66377c..953bf8679cc 100644
--- a/include/llvm/Transforms/Vectorize.h
+++ b/include/llvm/Transforms/Vectorize.h
@@ -28,6 +28,9 @@ struct VectorizeConfig {
   /// @brief The size of the native vector registers.
   unsigned VectorBits;
 
+  /// @brief Vectorize boolean values.
+  bool VectorizeBools;
+
   /// @brief Vectorize integer values.
   bool VectorizeInts;
 
diff --git a/lib/Transforms/Vectorize/BBVectorize.cpp b/lib/Transforms/Vectorize/BBVectorize.cpp
index 55e0da7ac0a..85187310dd8 100644
--- a/lib/Transforms/Vectorize/BBVectorize.cpp
+++ b/lib/Transforms/Vectorize/BBVectorize.cpp
@@ -76,6 +76,10 @@ MaxCandPairsForCycleCheck("bb-vectorize-max-cycle-check-pairs", cl::init(200),
   cl::Hidden, cl::desc("The maximum number of candidate pairs with which to use"
                        " a full cycle check"));
 
+static cl::opt<bool>
+NoBools("bb-vectorize-no-bools", cl::init(false), cl::Hidden,
+  cl::desc("Don't try to vectorize boolean (i1) values"));
+
 static cl::opt<bool>
 NoInts("bb-vectorize-no-ints", cl::init(false), cl::Hidden,
   cl::desc("Don't try to vectorize integer values"));
@@ -614,10 +618,15 @@ namespace {
         !(VectorType::isValidElementType(T2) || T2->isVectorTy()))
       return false;
 
-    if (!Config.VectorizeInts
-        && (T1->isIntOrIntVectorTy() || T2->isIntOrIntVectorTy()))
-      return false;
-
+    if (T1->getScalarSizeInBits() == 1 && T2->getScalarSizeInBits() == 1) {
+      if (!Config.VectorizeBools)
+        return false;
+    } else {
+      if (!Config.VectorizeInts
+          && (T1->isIntOrIntVectorTy() || T2->isIntOrIntVectorTy()))
+        return false;
+    }
+  
     if (!Config.VectorizeFloats
         && (T1->isFPOrFPVectorTy() || T2->isFPOrFPVectorTy()))
       return false;
@@ -1990,6 +1999,7 @@ llvm::vectorizeBasicBlock(Pass *P, BasicBlock &BB, const VectorizeConfig &C) {
 //===----------------------------------------------------------------------===//
 VectorizeConfig::VectorizeConfig() {
   VectorBits = ::VectorBits;
+  VectorizeBools = !::NoBools;
   VectorizeInts = !::NoInts;
   VectorizeFloats = !::NoFloats;
   VectorizePointers = !::NoPointers;
diff --git a/test/Transforms/BBVectorize/simple-sel.ll b/test/Transforms/BBVectorize/simple-sel.ll
index 7fb9c8b74c0..325792a5dca 100644
--- a/test/Transforms/BBVectorize/simple-sel.ll
+++ b/test/Transforms/BBVectorize/simple-sel.ll
@@ -1,5 +1,6 @@
 target datalayout = "e-p:64:64:64-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-s0:64:64-f80:128:128"
 ; RUN: opt < %s -bb-vectorize -bb-vectorize-req-chain-depth=3 -instcombine -gvn -S | FileCheck %s
+; RUN: opt < %s -bb-vectorize -bb-vectorize-req-chain-depth=3 -bb-vectorize-no-bools -instcombine -gvn -S | FileCheck %s -check-prefix=CHECK-NB
 
 ; Basic depth-3 chain with select
 define double @test1(double %A1, double %A2, double %B1, double %B2, i1 %C1, i1 %C2) {
@@ -30,6 +31,7 @@ define double @test1(double %A1, double %A2, double %B1, double %B2, i1 %C1, i1
 ; Basic depth-3 chain with select (and vect. compare)
 define double @test2(double %A1, double %A2, double %B1, double %B2) {
 ; CHECK: @test2
+; CHECK-NB: @test2
 ; CHECK: %X1.v.i1.1 = insertelement <2 x double> undef, double %B1, i32 0
 ; CHECK: %X1.v.i0.1 = insertelement <2 x double> undef, double %A1, i32 0
 ; CHECK: %X1.v.i1.2 = insertelement <2 x double> %X1.v.i1.1, double %B2, i32 1
@@ -43,6 +45,7 @@ define double @test2(double %A1, double %A2, double %B1, double %B2) {
 	%C1 = fcmp ogt double %X1, %A1
         %C2 = fcmp ogt double %X2, %A2
 ; CHECK: %C1 = fcmp ogt <2 x double> %X1, %X1.v.i0.2
+; CHECK-NB: fcmp ogt double
         %Z1 = select i1 %C1, double %Y1, double %B1
         %Z2 = select i1 %C2, double %Y2, double %B2
 ; CHECK: %Z1 = select <2 x i1> %C1, <2 x double> %Y1, <2 x double> %X1.v.i1.2
-- 
2.34.1