Move Instruction::isSafeToSpeculativelyExecute out of VMCore and

into Analysis as a standalone function, since there's no need for
it to be in VMCore. Also, update it to use isKnownNonZero and
other goodies available in Analysis, making it more precise,
enabling more aggressive optimization.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@146610 91177308-0d34-0410-b5e6-96231b3b80d8

include/llvm/Analysis/ValueTracking.h

@@ -156,6 +156,27 @@ namespace llvm {
   /// are lifetime markers.
   bool onlyUsedByLifetimeMarkers(const Value *V);
 
+  /// isSafeToSpeculativelyExecute - Return true if the instruction does not
+  /// have any effects besides calculating the result and does not have
+  /// undefined behavior.
+  ///
+  /// This method never returns true for an instruction that returns true for
+  /// mayHaveSideEffects; however, this method also does some other checks in
+  /// addition. It checks for undefined behavior, like dividing by zero or
+  /// loading from an invalid pointer (but not for undefined results, like a
+  /// shift with a shift amount larger than the width of the result). It checks
+  /// for malloc and alloca because speculatively executing them might cause a
+  /// memory leak. It also returns false for instructions related to control
+  /// flow, specifically terminators and PHI nodes.
+  ///
+  /// This method only looks at the instruction itself and its operands, so if
+  /// this method returns true, it is safe to move the instruction as long as
+  /// the correct dominance relationships for the operands and users hold.
+  /// However, this method can return true for instructions that read memory;
+  /// for such instructions, moving them may change the resulting value.
+  bool isSafeToSpeculativelyExecute(const Instruction *Inst,
+                                    const TargetData *TD = 0);
+
 } // end namespace llvm
 
 #endif

include/llvm/Instruction.h

@@ -244,26 +244,6 @@ public:
     return mayWriteToMemory() || mayThrow();
   }
 
-  /// isSafeToSpeculativelyExecute - Return true if the instruction does not
-  /// have any effects besides calculating the result and does not have
-  /// undefined behavior.
-  ///
-  /// This method never returns true for an instruction that returns true for
-  /// mayHaveSideEffects; however, this method also does some other checks in
-  /// addition. It checks for undefined behavior, like dividing by zero or
-  /// loading from an invalid pointer (but not for undefined results, like a
-  /// shift with a shift amount larger than the width of the result). It checks
-  /// for malloc and alloca because speculatively executing them might cause a
-  /// memory leak. It also returns false for instructions related to control
-  /// flow, specifically terminators and PHI nodes.
-  ///
-  /// This method only looks at the instruction itself and its operands, so if
-  /// this method returns true, it is safe to move the instruction as long as
-  /// the correct dominance relationships for the operands and users hold.
-  /// However, this method can return true for instructions that read memory;
-  /// for such instructions, moving them may change the resulting value.
-  bool isSafeToSpeculativelyExecute() const;
-
   /// clone() - Create a copy of 'this' instruction that is identical in all
   /// ways except the following:
   ///   * The instruction has no parent

lib/Analysis/LoopInfo.cpp

@@ -19,6 +19,7 @@
 #include "llvm/Instructions.h"
 #include "llvm/Analysis/Dominators.h"
 #include "llvm/Analysis/LoopIterator.h"
+#include "llvm/Analysis/ValueTracking.h"
 #include "llvm/Assembly/Writer.h"
 #include "llvm/Support/CFG.h"
 #include "llvm/Support/CommandLine.h"
@@ -95,7 +96,7 @@ bool Loop::makeLoopInvariant(Instruction *I, bool &Changed,
   // Test if the value is already loop-invariant.
   if (isLoopInvariant(I))
     return true;
-  if (!I->isSafeToSpeculativelyExecute())
+  if (!isSafeToSpeculativelyExecute(I))
     return false;
   if (I->mayReadFromMemory())
     return false;

lib/Analysis/PHITransAddr.cpp

@@ -12,6 +12,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "llvm/Analysis/PHITransAddr.h"
+#include "llvm/Analysis/ValueTracking.h"
 #include "llvm/Constants.h"
 #include "llvm/Instructions.h"
 #include "llvm/Analysis/Dominators.h"
@@ -27,7 +28,7 @@ static bool CanPHITrans(Instruction *Inst) {
     return true;
 
   if (isa<CastInst>(Inst) &&
-      Inst->isSafeToSpeculativelyExecute())
+      isSafeToSpeculativelyExecute(Inst))
     return true;
 
   if (Inst->getOpcode() == Instruction::Add &&
@@ -186,7 +187,7 @@ Value *PHITransAddr::PHITranslateSubExpr(Value *V, BasicBlock *CurBB,
   // operands need to be phi translated, and if so, reconstruct it.
 
   if (CastInst *Cast = dyn_cast<CastInst>(Inst)) {
-    if (!Cast->isSafeToSpeculativelyExecute()) return 0;
+    if (!isSafeToSpeculativelyExecute(Cast)) return 0;
     Value *PHIIn = PHITranslateSubExpr(Cast->getOperand(0), CurBB, PredBB, DT);
     if (PHIIn == 0) return 0;
     if (PHIIn == Cast->getOperand(0))
@@ -381,7 +382,7 @@ InsertPHITranslatedSubExpr(Value *InVal, BasicBlock *CurBB,
 
   // Handle cast of PHI translatable value.
   if (CastInst *Cast = dyn_cast<CastInst>(Inst)) {
-    if (!Cast->isSafeToSpeculativelyExecute()) return 0;
+    if (!isSafeToSpeculativelyExecute(Cast)) return 0;
     Value *OpVal = InsertPHITranslatedSubExpr(Cast->getOperand(0),
                                               CurBB, PredBB, DT, NewInsts);
     if (OpVal == 0) return 0;

lib/Analysis/ValueTracking.cpp

@@ -1875,3 +1875,64 @@ bool llvm::onlyUsedByLifetimeMarkers(const Value *V) {
   }
   return true;
 }
+
+bool llvm::isSafeToSpeculativelyExecute(const Instruction *Inst,
+                                        const TargetData *TD) {
+  for (unsigned i = 0, e = Inst->getNumOperands(); i != e; ++i)
+    if (Constant *C = dyn_cast<Constant>(Inst->getOperand(i)))
+      if (C->canTrap())
+        return false;
+
+  switch (Inst->getOpcode()) {
+  default:
+    return true;
+  case Instruction::UDiv:
+  case Instruction::URem:
+    // x / y is undefined if y == 0, but calcuations like x / 3 are safe.
+    return isKnownNonZero(Inst->getOperand(1), TD);
+  case Instruction::SDiv:
+  case Instruction::SRem: {
+    Value *Op = Inst->getOperand(1);
+    // x / y is undefined if y == 0
+    if (!isKnownNonZero(Op, TD))
+      return false;
+    // x / y might be undefined if y == -1
+    unsigned BitWidth = getBitWidth(Op->getType(), TD);
+    if (BitWidth == 0)
+      return false;
+    APInt KnownZero(BitWidth, 0);
+    APInt KnownOne(BitWidth, 0);
+    ComputeMaskedBits(Op, APInt::getAllOnesValue(BitWidth),
+                      KnownZero, KnownOne, TD);
+    return !!KnownZero;
+  }
+  case Instruction::Load: {
+    const LoadInst *LI = cast<LoadInst>(Inst);
+    if (!LI->isUnordered())
+      return false;
+    return LI->getPointerOperand()->isDereferenceablePointer();
+  }
+  case Instruction::Call:
+    return false; // The called function could have undefined behavior or
+                  // side-effects.
+                  // FIXME: We should special-case some intrinsics (bswap,
+                  // overflow-checking arithmetic, etc.)
+  case Instruction::VAArg:
+  case Instruction::Alloca:
+  case Instruction::Invoke:
+  case Instruction::PHI:
+  case Instruction::Store:
+  case Instruction::Ret:
+  case Instruction::Br:
+  case Instruction::IndirectBr:
+  case Instruction::Switch:
+  case Instruction::Unwind:
+  case Instruction::Unreachable:
+  case Instruction::Fence:
+  case Instruction::LandingPad:
+  case Instruction::AtomicRMW:
+  case Instruction::AtomicCmpXchg:
+  case Instruction::Resume:
+    return false; // Misc instructions which have effects
+  }
+}

lib/CodeGen/Analysis.cpp

@@ -12,6 +12,7 @@
 //===----------------------------------------------------------------------===//
 
 #include "llvm/CodeGen/Analysis.h"
+#include "llvm/Analysis/ValueTracking.h"
 #include "llvm/DerivedTypes.h"
 #include "llvm/Function.h"
 #include "llvm/Instructions.h"
@@ -234,7 +235,7 @@ bool llvm::isInTailCallPosition(ImmutableCallSite CS, Attributes CalleeRetAttr,
   // If I will have a chain, make sure no other instruction that will have a
   // chain interposes between I and the return.
   if (I->mayHaveSideEffects() || I->mayReadFromMemory() ||
-      !I->isSafeToSpeculativelyExecute())
+      !isSafeToSpeculativelyExecute(I))
     for (BasicBlock::const_iterator BBI = prior(prior(ExitBB->end())); ;
          --BBI) {
       if (&*BBI == I)
@@ -243,7 +244,7 @@ bool llvm::isInTailCallPosition(ImmutableCallSite CS, Attributes CalleeRetAttr,
       if (isa<DbgInfoIntrinsic>(BBI))
         continue;
       if (BBI->mayHaveSideEffects() || BBI->mayReadFromMemory() ||
-          !BBI->isSafeToSpeculativelyExecute())
+          !isSafeToSpeculativelyExecute(BBI))
         return false;
     }
 

lib/Transforms/Scalar/LICM.cpp

@@ -43,6 +43,7 @@
 #include "llvm/Analysis/LoopInfo.h"
 #include "llvm/Analysis/LoopPass.h"
 #include "llvm/Analysis/Dominators.h"
+#include "llvm/Analysis/ValueTracking.h"
 #include "llvm/Transforms/Utils/Local.h"
 #include "llvm/Transforms/Utils/SSAUpdater.h"
 #include "llvm/Target/TargetData.h"
@@ -591,7 +592,7 @@ void LICM::hoist(Instruction &I) {
 ///
 bool LICM::isSafeToExecuteUnconditionally(Instruction &Inst) {
   // If it is not a trapping instruction, it is always safe to hoist.
-  if (Inst.isSafeToSpeculativelyExecute())
+  if (isSafeToSpeculativelyExecute(&Inst))
     return true;
 
   return isGuaranteedToExecute(Inst);

lib/Transforms/Scalar/Sink.cpp

@@ -18,6 +18,7 @@
 #include "llvm/Analysis/Dominators.h"
 #include "llvm/Analysis/LoopInfo.h"
 #include "llvm/Analysis/AliasAnalysis.h"
+#include "llvm/Analysis/ValueTracking.h"
 #include "llvm/Assembly/Writer.h"
 #include "llvm/ADT/Statistic.h"
 #include "llvm/Support/CFG.h"
@@ -240,7 +241,7 @@ bool Sinking::SinkInstruction(Instruction *Inst,
   if (SuccToSinkTo->getUniquePredecessor() != ParentBlock) {
     // We cannot sink a load across a critical edge - there may be stores in
     // other code paths.
-    if (!Inst->isSafeToSpeculativelyExecute()) {
+    if (!isSafeToSpeculativelyExecute(Inst)) {
       DEBUG(dbgs() << " *** PUNTING: Wont sink load along critical edge.\n");
       return false;
     }

lib/Transforms/Utils/SimplifyCFG.cpp

@@ -257,7 +257,7 @@ static bool DominatesMergePoint(Value *V, BasicBlock *BB,
   // Okay, it looks like the instruction IS in the "condition".  Check to
   // see if it's a cheap instruction to unconditionally compute, and if it
   // only uses stuff defined outside of the condition.  If so, hoist it out.
-  if (!I->isSafeToSpeculativelyExecute())
+  if (!isSafeToSpeculativelyExecute(I))
     return false;
 
   unsigned Cost = 0;
@@ -1487,7 +1487,7 @@ bool llvm::FoldBranchToCommonDest(BranchInst *BI) {
   Instruction *BonusInst = 0;
   if (&*FrontIt != Cond &&
       FrontIt->hasOneUse() && *FrontIt->use_begin() == Cond &&
-      FrontIt->isSafeToSpeculativelyExecute()) {
+      isSafeToSpeculativelyExecute(FrontIt)) {
     BonusInst = &*FrontIt;
     ++FrontIt;
     

lib/VMCore/Instruction.cpp

@@ -391,59 +391,6 @@ bool Instruction::isCommutative(unsigned op) {
   }
 }
 
-bool Instruction::isSafeToSpeculativelyExecute() const {
-  for (unsigned i = 0, e = getNumOperands(); i != e; ++i)
-    if (Constant *C = dyn_cast<Constant>(getOperand(i)))
-      if (C->canTrap())
-        return false;
-
-  switch (getOpcode()) {
-  default:
-    return true;
-  case UDiv:
-  case URem: {
-    // x / y is undefined if y == 0, but calcuations like x / 3 are safe.
-    ConstantInt *Op = dyn_cast<ConstantInt>(getOperand(1));
-    return Op && !Op->isNullValue();
-  }
-  case SDiv:
-  case SRem: {
-    // x / y is undefined if y == 0, and might be undefined if y == -1,
-    // but calcuations like x / 3 are safe.
-    ConstantInt *Op = dyn_cast<ConstantInt>(getOperand(1));
-    return Op && !Op->isNullValue() && !Op->isAllOnesValue();
-  }
-  case Load: {
-    const LoadInst *LI = cast<LoadInst>(this);
-    if (!LI->isUnordered())
-      return false;
-    return LI->getPointerOperand()->isDereferenceablePointer();
-  }
-  case Call:
-    return false; // The called function could have undefined behavior or
-                  // side-effects.
-                  // FIXME: We should special-case some intrinsics (bswap,
-                  // overflow-checking arithmetic, etc.)
-  case VAArg:
-  case Alloca:
-  case Invoke:
-  case PHI:
-  case Store:
-  case Ret:
-  case Br:
-  case IndirectBr:
-  case Switch:
-  case Unwind:
-  case Unreachable:
-  case Fence:
-  case LandingPad:
-  case AtomicRMW:
-  case AtomicCmpXchg:
-  case Resume:
-    return false; // Misc instructions which have effects
-  }
-}
-
 Instruction *Instruction::clone() const {
   Instruction *New = clone_impl();
   New->SubclassOptionalData = SubclassOptionalData;

test/Transforms/LICM/speculate.ll

@@ -0,0 +1,167 @@
+; RUN: opt -S -licm < %s | FileCheck %s
+
+; UDiv is safe to speculate if the denominator is known non-zero.
+
+; CHECK: @safe_udiv
+; CHECK:      %div = udiv i64 %x, %or
+; CHECK-NEXT: br label %for.body
+
+define void @safe_udiv(i64 %x, i64 %m, i64 %n, i32* %p, i64* %q) nounwind {
+entry:
+  %or = or i64 %m, 1
+  br label %for.body
+
+for.body:                                         ; preds = %entry, %for.inc
+  %i.02 = phi i64 [ %inc, %for.inc ], [ 0, %entry ]
+  %arrayidx = getelementptr inbounds i32* %p, i64 %i.02
+  %0 = load i32* %arrayidx, align 4
+  %tobool = icmp eq i32 %0, 0
+  br i1 %tobool, label %for.inc, label %if.then
+
+if.then:                                          ; preds = %for.body
+  %div = udiv i64 %x, %or
+  %arrayidx1 = getelementptr inbounds i64* %q, i64 %i.02
+  store i64 %div, i64* %arrayidx1, align 8
+  br label %for.inc
+
+for.inc:                                          ; preds = %if.then, %for.body
+  %inc = add i64 %i.02, 1
+  %cmp = icmp slt i64 %inc, %n
+  br i1 %cmp, label %for.body, label %for.end
+
+for.end:                                          ; preds = %for.inc, %entry
+  ret void
+}
+
+; UDiv is unsafe to speculate if the denominator is not known non-zero.
+
+; CHECK: @unsafe_udiv
+; CHECK-NOT:  udiv
+; CHECK: for.body:
+
+define void @unsafe_udiv(i64 %x, i64 %m, i64 %n, i32* %p, i64* %q) nounwind {
+entry:
+  br label %for.body
+
+for.body:                                         ; preds = %entry, %for.inc
+  %i.02 = phi i64 [ %inc, %for.inc ], [ 0, %entry ]
+  %arrayidx = getelementptr inbounds i32* %p, i64 %i.02
+  %0 = load i32* %arrayidx, align 4
+  %tobool = icmp eq i32 %0, 0
+  br i1 %tobool, label %for.inc, label %if.then
+
+if.then:                                          ; preds = %for.body
+  %div = udiv i64 %x, %m
+  %arrayidx1 = getelementptr inbounds i64* %q, i64 %i.02
+  store i64 %div, i64* %arrayidx1, align 8
+  br label %for.inc
+
+for.inc:                                          ; preds = %if.then, %for.body
+  %inc = add i64 %i.02, 1
+  %cmp = icmp slt i64 %inc, %n
+  br i1 %cmp, label %for.body, label %for.end
+
+for.end:                                          ; preds = %for.inc, %entry
+  ret void
+}
+
+; SDiv is safe to speculate if the denominator is known non-zero and
+; known to have at least one zero bit.
+
+; CHECK: @safe_sdiv
+; CHECK:      %div = sdiv i64 %x, %or
+; CHECK-NEXT: br label %for.body
+
+define void @safe_sdiv(i64 %x, i64 %m, i64 %n, i32* %p, i64* %q) nounwind {
+entry:
+  %and = and i64 %m, -3
+  %or = or i64 %and, 1
+  br label %for.body
+
+for.body:                                         ; preds = %entry, %for.inc
+  %i.02 = phi i64 [ %inc, %for.inc ], [ 0, %entry ]
+  %arrayidx = getelementptr inbounds i32* %p, i64 %i.02
+  %0 = load i32* %arrayidx, align 4
+  %tobool = icmp eq i32 %0, 0
+  br i1 %tobool, label %for.inc, label %if.then
+
+if.then:                                          ; preds = %for.body
+  %div = sdiv i64 %x, %or
+  %arrayidx1 = getelementptr inbounds i64* %q, i64 %i.02
+  store i64 %div, i64* %arrayidx1, align 8
+  br label %for.inc
+
+for.inc:                                          ; preds = %if.then, %for.body
+  %inc = add i64 %i.02, 1
+  %cmp = icmp slt i64 %inc, %n
+  br i1 %cmp, label %for.body, label %for.end
+
+for.end:                                          ; preds = %for.inc, %entry
+  ret void
+}
+
+; SDiv is unsafe to speculate if the denominator is not known non-zero.
+
+; CHECK: @unsafe_sdiv_a
+; CHECK-NOT:  sdiv
+; CHECK: for.body:
+
+define void @unsafe_sdiv_a(i64 %x, i64 %m, i64 %n, i32* %p, i64* %q) nounwind {
+entry:
+  %or = or i64 %m, 1
+  br label %for.body
+
+for.body:                                         ; preds = %entry, %for.inc
+  %i.02 = phi i64 [ %inc, %for.inc ], [ 0, %entry ]
+  %arrayidx = getelementptr inbounds i32* %p, i64 %i.02
+  %0 = load i32* %arrayidx, align 4
+  %tobool = icmp eq i32 %0, 0
+  br i1 %tobool, label %for.inc, label %if.then
+
+if.then:                                          ; preds = %for.body
+  %div = sdiv i64 %x, %or
+  %arrayidx1 = getelementptr inbounds i64* %q, i64 %i.02
+  store i64 %div, i64* %arrayidx1, align 8
+  br label %for.inc
+
+for.inc:                                          ; preds = %if.then, %for.body
+  %inc = add i64 %i.02, 1
+  %cmp = icmp slt i64 %inc, %n
+  br i1 %cmp, label %for.body, label %for.end
+
+for.end:                                          ; preds = %for.inc, %entry
+  ret void
+}
+
+; SDiv is unsafe to speculate if the denominator is not known to have a zero bit.
+
+; CHECK: @unsafe_sdiv_b
+; CHECK-NOT:  sdiv
+; CHECK: for.body:
+
+define void @unsafe_sdiv_b(i64 %x, i64 %m, i64 %n, i32* %p, i64* %q) nounwind {
+entry:
+  %and = and i64 %m, -3
+  br label %for.body
+
+for.body:                                         ; preds = %entry, %for.inc
+  %i.02 = phi i64 [ %inc, %for.inc ], [ 0, %entry ]
+  %arrayidx = getelementptr inbounds i32* %p, i64 %i.02
+  %0 = load i32* %arrayidx, align 4
+  %tobool = icmp eq i32 %0, 0
+  br i1 %tobool, label %for.inc, label %if.then
+
+if.then:                                          ; preds = %for.body
+  %div = sdiv i64 %x, %and
+  %arrayidx1 = getelementptr inbounds i64* %q, i64 %i.02
+  store i64 %div, i64* %arrayidx1, align 8
+  br label %for.inc
+
+for.inc:                                          ; preds = %if.then, %for.body
+  %inc = add i64 %i.02, 1
+  %cmp = icmp slt i64 %inc, %n
+  br i1 %cmp, label %for.body, label %for.end
+
+for.end:                                          ; preds = %for.inc, %entry
+  ret void
+}