[llvm] r274329 - AMDGPU: Add option to run the load/store vectorizer

[Matt Arsenault via llvm-commits]

Author: arsenm
Date: Thu Jun 30 22:33:52 2016
New Revision: 274329

URL: http://llvm.org/viewvc/llvm-project?rev=274329&view=rev
Log:
AMDGPU: Add option to run the load/store vectorizer

Modified:
    llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp

Modified: llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp?rev=274329&r1=274328&r2=274329&view=diff
==============================================================================
--- llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp (original)
+++ llvm/trunk/lib/Target/AMDGPU/AMDGPUTargetMachine.cpp Thu Jun 30 22:33:52 2016
@@ -39,6 +39,7 @@
 #include "llvm/Transforms/IPO.h"
 #include "llvm/Transforms/Scalar.h"
 #include "llvm/Transforms/Scalar/GVN.h"
+#include "llvm/Transforms/Vectorize.h"
 
 using namespace llvm;
 
@@ -59,6 +60,13 @@ static cl::opt<bool> EnableR600IfConvert
   cl::ReallyHidden,
   cl::init(true));
 
+// Option to disable vectorizer for tests.
+static cl::opt<bool> EnableLoadStoreVectorizer(
+  "amdgpu-load-store-vectorizer",
+  cl::desc("Enable load store vectorizer"),
+  cl::init(false),
+  cl::Hidden);
+
 extern "C" void LLVMInitializeAMDGPUTarget() {
   // Register the target
   RegisterTargetMachine<R600TargetMachine> X(TheAMDGPUTarget);
@@ -267,6 +275,7 @@ public:
   void addEarlyCSEOrGVNPass();
   void addStraightLineScalarOptimizationPasses();
   void addIRPasses() override;
+  void addCodeGenPrepare() override;
   bool addPreISel() override;
   bool addInstSelector() override;
   bool addGCPasses() override;
@@ -392,6 +401,13 @@ void AMDGPUPassConfig::addIRPasses() {
     addEarlyCSEOrGVNPass();
 }
 
+void AMDGPUPassConfig::addCodeGenPrepare() {
+  TargetPassConfig::addCodeGenPrepare();
+
+  if (EnableLoadStoreVectorizer)
+    addPass(createLoadStoreVectorizerPass());
+}
+
 bool AMDGPUPassConfig::addPreISel() {
   addPass(createFlattenCFGPass());
   return false;