Disable unrolling in the loop vectorizer when disabled in the pass manager

When unrolling is disabled in the pass manager, the loop vectorizer should also
not unroll loops. This will allow the -fno-unroll-loops option in Clang to
behave as expected (even for vectorizable loops). The loop vectorizer's
-force-vector-unroll option will (continue to) override the pass-manager
setting (including -force-vector-unroll=0 to force use of the internal
auto-selection logic).

In order to test this, I added a flag to opt (-disable-loop-unrolling) to force
disable unrolling through opt (the analog of -fno-unroll-loops in Clang). Also,
this fixes a small bug in opt where the loop vectorizer was enabled only after
the pass manager populated the queue of passes (the global_alias.ll test needed
a slight update to the RUN line as a result of this fix).

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@189499 91177308-0d34-0410-b5e6-96231b3b80d8

include/llvm/Transforms/Vectorize.h

@@ -114,7 +114,7 @@ createBBVectorizePass(const VectorizeConfig &C = VectorizeConfig());
 //
 // LoopVectorize - Create a loop vectorization pass.
 //
-Pass *createLoopVectorizePass();
+Pass *createLoopVectorizePass(bool NoUnrolling = false);
 
 //===----------------------------------------------------------------------===//
 //

lib/Transforms/IPO/PassManagerBuilder.cpp

@@ -196,7 +196,7 @@ void PassManagerBuilder::populateModulePassManager(PassManagerBase &MPM) {
   MPM.add(createLoopDeletionPass());          // Delete dead loops
 
   if (!LateVectorize && LoopVectorize)
-      MPM.add(createLoopVectorizePass());
+      MPM.add(createLoopVectorizePass(DisableUnrollLoops));
 
   if (!DisableUnrollLoops)
     MPM.add(createLoopUnrollPass());          // Unroll small loops
@@ -250,7 +250,7 @@ void PassManagerBuilder::populateModulePassManager(PassManagerBase &MPM) {
     // Add the various vectorization passes and relevant cleanup passes for
     // them since we are no longer in the middle of the main scalar pipeline.
     if (LoopVectorize) {
-      MPM.add(createLoopVectorizePass());
+      MPM.add(createLoopVectorizePass(DisableUnrollLoops));
 
       if (!DisableUnrollLoops)
         MPM.add(createLoopUnrollPass());    // Unroll small loops

lib/Transforms/Vectorize/LoopVectorize.cpp

@@ -761,9 +761,9 @@ struct LoopVectorizeHints {
   /// Vectorization unroll factor.
   unsigned Unroll;
 
-  LoopVectorizeHints(const Loop *L)
+  LoopVectorizeHints(const Loop *L, bool DisableUnrolling)
   : Width(VectorizationFactor)
-  , Unroll(VectorizationUnroll)
+  , Unroll(DisableUnrolling ? 1 : VectorizationUnroll)
   , LoopID(L->getLoopID()) {
     getHints(L);
     // The command line options override any loop metadata except for when
@@ -772,6 +772,9 @@ struct LoopVectorizeHints {
       Width = VectorizationFactor;
     if (VectorizationUnroll.getNumOccurrences() > 0)
       Unroll = VectorizationUnroll;
+
+    DEBUG(if (DisableUnrolling && Unroll == 1)
+            dbgs() << "LV: Unrolling disabled by the pass manager\n");
   }
 
   /// Return the loop vectorizer metadata prefix.
@@ -878,7 +881,8 @@ struct LoopVectorize : public LoopPass {
   /// Pass identification, replacement for typeid
   static char ID;
 
-  explicit LoopVectorize() : LoopPass(ID) {
+  explicit LoopVectorize(bool NoUnrolling = false)
+    : LoopPass(ID), DisableUnrolling(NoUnrolling) {
     initializeLoopVectorizePass(*PassRegistry::getPassRegistry());
   }
 
@@ -888,6 +892,7 @@ struct LoopVectorize : public LoopPass {
   TargetTransformInfo *TTI;
   DominatorTree *DT;
   TargetLibraryInfo *TLI;
+  bool DisableUnrolling;
 
   virtual bool runOnLoop(Loop *L, LPPassManager &LPM) {
     // We only vectorize innermost loops.
@@ -909,7 +914,7 @@ struct LoopVectorize : public LoopPass {
     DEBUG(dbgs() << "LV: Checking a loop in \"" <<
           L->getHeader()->getParent()->getName() << "\"\n");
 
-    LoopVectorizeHints Hints(L);
+    LoopVectorizeHints Hints(L, DisableUnrolling);
 
     if (Hints.Width == 1 && Hints.Unroll == 1) {
       DEBUG(dbgs() << "LV: Not vectorizing.\n");
@@ -4786,8 +4791,8 @@ INITIALIZE_PASS_DEPENDENCY(LoopSimplify)
 INITIALIZE_PASS_END(LoopVectorize, LV_NAME, lv_name, false, false)
 
 namespace llvm {
-  Pass *createLoopVectorizePass() {
+  Pass *createLoopVectorizePass(bool NoUnrolling) {
-    return new LoopVectorize();
+    return new LoopVectorize(NoUnrolling);
   }
 }
 

test/Transforms/LoopVectorize/X86/unroll-pm.ll

@@ -0,0 +1,31 @@
+; RUN: opt < %s -O2 -mtriple=x86_64-apple-macosx10.8.0 -mcpu=corei7-avx -force-vector-width=4 -S | FileCheck %s
+; RUN: opt < %s -O2 -mtriple=x86_64-apple-macosx10.8.0 -mcpu=corei7-avx -force-vector-width=4 -disable-loop-unrolling -S | FileCheck %s -check-prefix=CHECK-NOUNRL
+
+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-n8:16:32:64-S128"
+target triple = "x86_64-apple-macosx10.8.0"
+;CHECK-LABEL: @bar(
+;CHECK: store <4 x i32>
+;CHECK: store <4 x i32>
+;CHECK: ret
+;CHECK-NOUNRL-LABEL: @bar(
+;CHECK-NOUNRL: store <4 x i32>
+;CHECK-NOUNRL-NOT: store <4 x i32>
+;CHECK-NOUNRL: ret
+define i32 @bar(i32* nocapture %A, i32 %n) nounwind uwtable ssp {
+  %1 = icmp sgt i32 %n, 0
+  br i1 %1, label %.lr.ph, label %._crit_edge
+
+.lr.ph:                                           ; preds = %0, %.lr.ph
+  %indvars.iv = phi i64 [ %indvars.iv.next, %.lr.ph ], [ 0, %0 ]
+  %2 = getelementptr inbounds i32* %A, i64 %indvars.iv
+  %3 = load i32* %2, align 4
+  %4 = add nsw i32 %3, 6
+  store i32 %4, i32* %2, align 4
+  %indvars.iv.next = add i64 %indvars.iv, 1
+  %lftr.wideiv = trunc i64 %indvars.iv.next to i32
+  %exitcond = icmp eq i32 %lftr.wideiv, %n
+  br i1 %exitcond, label %._crit_edge, label %.lr.ph
+
+._crit_edge:                                      ; preds = %.lr.ph, %0
+  ret i32 undef
+}

test/Transforms/LoopVectorize/global_alias.ll

@@ -1,4 +1,4 @@
-; RUN: opt < %s -O3 -loop-vectorize -force-vector-unroll=1 -force-vector-width=4 -dce -instcombine -S | FileCheck %s
+; RUN: opt < %s -O1 -loop-vectorize -force-vector-unroll=1 -force-vector-width=4 -dce -instcombine -S | FileCheck %s
 
 target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:64:128-a0:0:64-n32-S64"
 

tools/opt/opt.cpp

@@ -135,6 +135,11 @@ UnitAtATime("funit-at-a-time",
             cl::desc("Enable IPO. This is same as llvm-gcc's -funit-at-a-time"),
             cl::init(true));
 
+static cl::opt<bool>
+DisableLoopUnrolling("disable-loop-unrolling",
+                     cl::desc("Disable loop unrolling in all relevant passes"),
+                     cl::init(false));
+
 static cl::opt<bool>
 DisableSimplifyLibCalls("disable-simplify-libcalls",
                         cl::desc("Disable simplify-libcalls"));
@@ -447,12 +452,13 @@ static void AddOptimizationPasses(PassManagerBase &MPM,FunctionPassManager &FPM,
     Builder.Inliner = createAlwaysInlinerPass();
   }
   Builder.DisableUnitAtATime = !UnitAtATime;
-  Builder.DisableUnrollLoops = OptLevel == 0;
+  Builder.DisableUnrollLoops = (DisableLoopUnrolling.getNumOccurrences() > 0) ?
+                               DisableLoopUnrolling : OptLevel == 0;
   
+  Builder.LoopVectorize = OptLevel > 1 && SizeLevel < 2;
+
   Builder.populateFunctionPassManager(FPM);
   Builder.populateModulePassManager(MPM);
-
-  Builder.LoopVectorize = OptLevel > 1 && SizeLevel < 2;
 }
 
 static void AddStandardCompilePasses(PassManagerBase &PM) {