Add a loop rerolling pass

This adds a loop rerolling pass: the opposite of (partial) loop unrolling. The
transformation aims to take loops like this:

for (int i = 0; i < 3200; i += 5) {
  a[i]     += alpha * b[i];
  a[i + 1] += alpha * b[i + 1];
  a[i + 2] += alpha * b[i + 2];
  a[i + 3] += alpha * b[i + 3];
  a[i + 4] += alpha * b[i + 4];
}

and turn them into this:

for (int i = 0; i < 3200; ++i) {
  a[i] += alpha * b[i];
}

and loops like this:

for (int i = 0; i < 500; ++i) {
  x[3*i] = foo(0);
  x[3*i+1] = foo(0);
  x[3*i+2] = foo(0);
}

and turn them into this:

for (int i = 0; i < 1500; ++i) {
  x[i] = foo(0);
}

There are two motivations for this transformation:

  1. Code-size reduction (especially relevant, obviously, when compiling for
code size).

  2. Providing greater choice to the loop vectorizer (and generic unroller) to
choose the unrolling factor (and a better ability to vectorize). The loop
vectorizer can take vector lengths and register pressure into account when
choosing an unrolling factor, for example, and a pre-unrolled loop limits that
choice. This is especially problematic if the manual unrolling was optimized
for a machine different from the current target.

The current implementation is limited to single basic-block loops only. The
rerolling recognition should work regardless of how the loop iterations are
intermixed within the loop body (subject to dependency and side-effect
constraints), but the significant restriction is that the order of the
instructions in each iteration must be identical. This seems sufficient to
capture all current use cases.

This pass is not currently enabled by default at any optimization level.

llvm-svn: 194939

Author: Hal Finkel

llvm/include/llvm-c/Transforms/Scalar.h

@@ -65,6 +65,9 @@ void LLVMAddLoopIdiomPass(LLVMPassManagerRef PM);
 /** See llvm::createLoopRotatePass function. */
 void LLVMAddLoopRotatePass(LLVMPassManagerRef PM);
 
+/** See llvm::createLoopRerollPass function. */
+void LLVMAddLoopRerollPass(LLVMPassManagerRef PM);
+
 /** See llvm::createLoopUnrollPass function. */
 void LLVMAddLoopUnrollPass(LLVMPassManagerRef PM);
 

llvm/include/llvm/InitializePasses.h

@@ -162,6 +162,7 @@ void initializeLoopRotatePass(PassRegistry&);
 void initializeLoopSimplifyPass(PassRegistry&);
 void initializeLoopStrengthReducePass(PassRegistry&);
 void initializeGlobalMergePass(PassRegistry&);
+void initializeLoopRerollPass(PassRegistry&);
 void initializeLoopUnrollPass(PassRegistry&);
 void initializeLoopUnswitchPass(PassRegistry&);
 void initializeLoopIdiomRecognizePass(PassRegistry&);

llvm/include/llvm/LinkAllPasses.h

@@ -91,6 +91,7 @@ namespace {
       (void) llvm::createLoopExtractorPass();
       (void) llvm::createLoopSimplifyPass();
       (void) llvm::createLoopStrengthReducePass();
+      (void) llvm::createLoopRerollPass();
       (void) llvm::createLoopUnrollPass();
       (void) llvm::createLoopUnswitchPass();
       (void) llvm::createLoopIdiomPass();

llvm/include/llvm/Transforms/Scalar.h

@@ -143,6 +143,12 @@ Pass *createLoopInstSimplifyPass();
 Pass *createLoopUnrollPass(int Threshold = -1, int Count = -1,
                            int AllowPartial = -1, int Runtime = -1);
 
+//===----------------------------------------------------------------------===//
+//
+// LoopReroll - This pass is a simple loop rerolling pass.
+//
+Pass *createLoopRerollPass();
+
 //===----------------------------------------------------------------------===//
 //
 // LoopRotate - This pass is a simple loop rotating pass.

llvm/lib/Transforms/IPO/PassManagerBuilder.cpp

@@ -54,6 +54,10 @@ static cl::opt<bool> UseNewSROA("use-new-sroa",
   cl::init(true), cl::Hidden,
   cl::desc("Enable the new, experimental SROA pass"));
 
+static cl::opt<bool>
+RunLoopRerolling("reroll-loops", cl::Hidden,
+                 cl::desc("Run the loop rerolling pass"));
+
 PassManagerBuilder::PassManagerBuilder() {
     OptLevel = 2;
     SizeLevel = 0;
@@ -216,6 +220,8 @@ void PassManagerBuilder::populateModulePassManager(PassManagerBase &MPM) {
 
   addExtensionsToPM(EP_ScalarOptimizerLate, MPM);
 
+  if (RunLoopRerolling)
+    MPM.add(createLoopRerollPass());
   if (SLPVectorize)
     MPM.add(createSLPVectorizerPass());   // Vectorize parallel scalar chains.
 

llvm/lib/Transforms/Scalar/CMakeLists.txt

@@ -16,6 +16,7 @@ add_llvm_library(LLVMScalarOpts
   LoopInstSimplify.cpp
   LoopRotation.cpp
   LoopStrengthReduce.cpp
+  LoopRerollPass.cpp
   LoopUnrollPass.cpp
   LoopUnswitch.cpp
   LowerAtomic.cpp