Merge pull request #47 from eddyb/fix-pointer-range-loops

Backport 4 commits to optimize pointer range loops better.

Author: alexcrichton

lib/Analysis/InstructionSimplify.cpp

@@ -3092,6 +3092,16 @@ static Value *SimplifyICmpInst(unsigned Predicate, Value *LHS, Value *RHS,
   if (LHS->getType()->isPointerTy())
     if (auto *C = computePointerICmp(Q.DL, Q.TLI, Q.DT, Pred, Q.CxtI, LHS, RHS))
       return C;
+  if (auto *CLHS = dyn_cast<PtrToIntOperator>(LHS))
+    if (auto *CRHS = dyn_cast<PtrToIntOperator>(RHS))
+      if (Q.DL.getTypeSizeInBits(CLHS->getPointerOperandType()) ==
+              Q.DL.getTypeSizeInBits(CLHS->getType()) &&
+          Q.DL.getTypeSizeInBits(CRHS->getPointerOperandType()) ==
+              Q.DL.getTypeSizeInBits(CRHS->getType()))
+        if (auto *C = computePointerICmp(Q.DL, Q.TLI, Q.DT, Pred, Q.CxtI,
+                                         CLHS->getPointerOperand(),
+                                         CRHS->getPointerOperand()))
+          return C;
 
   if (GetElementPtrInst *GLHS = dyn_cast<GetElementPtrInst>(LHS)) {
     if (GEPOperator *GRHS = dyn_cast<GEPOperator>(RHS)) {
@@ -3587,6 +3597,32 @@ static Value *SimplifyGEPInst(Type *SrcTy, ArrayRef<Value *> Ops,
     }
   }
 
+  if (Q.DL.getTypeAllocSize(LastType) == 1 &&
+      all_of(Ops.slice(1).drop_back(1),
+             [](Value *Idx) { return match(Idx, m_Zero()); })) {
+    unsigned PtrWidth =
+        Q.DL.getPointerSizeInBits(Ops[0]->getType()->getPointerAddressSpace());
+    if (Q.DL.getTypeSizeInBits(Ops.back()->getType()) == PtrWidth) {
+      APInt BasePtrOffset(PtrWidth, 0);
+      Value *StrippedBasePtr =
+          Ops[0]->stripAndAccumulateInBoundsConstantOffsets(Q.DL,
+                                                            BasePtrOffset);
+
+      // gep (gep V, C), (sub 0, V) -> C
+      if (match(Ops.back(),
+                m_Sub(m_Zero(), m_PtrToInt(m_Specific(StrippedBasePtr))))) {
+        auto *CI = ConstantInt::get(GEPTy->getContext(), BasePtrOffset);
+        return ConstantExpr::getIntToPtr(CI, GEPTy);
+      }
+      // gep (gep V, C), (xor V, -1) -> C-1
+      if (match(Ops.back(),
+                m_Xor(m_PtrToInt(m_Specific(StrippedBasePtr)), m_AllOnes()))) {
+        auto *CI = ConstantInt::get(GEPTy->getContext(), BasePtrOffset - 1);
+        return ConstantExpr::getIntToPtr(CI, GEPTy);
+      }
+    }
+  }
+
   // Check to see if this is constant foldable.
   for (unsigned i = 0, e = Ops.size(); i != e; ++i)
     if (!isa<Constant>(Ops[i]))

lib/Transforms/InstCombine/InstructionCombining.cpp

@@ -1898,6 +1898,25 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
     }
   }
 
+  if (!GEP.isInBounds()) {
+    unsigned PtrWidth =
+        DL.getPointerSizeInBits(PtrOp->getType()->getPointerAddressSpace());
+    APInt BasePtrOffset(PtrWidth, 0);
+    Value *UnderlyingPtrOp =
+            PtrOp->stripAndAccumulateInBoundsConstantOffsets(DL,
+                                                             BasePtrOffset);
+    if (auto *AI = dyn_cast<AllocaInst>(UnderlyingPtrOp)) {
+      if (GEP.accumulateConstantOffset(DL, BasePtrOffset) &&
+          BasePtrOffset.isNonNegative()) {
+        APInt AllocSize(PtrWidth, DL.getTypeAllocSize(AI->getAllocatedType()));
+        if (BasePtrOffset.ule(AllocSize)) {
+          return GetElementPtrInst::CreateInBounds(
+              PtrOp, makeArrayRef(Ops).slice(1), GEP.getName());
+        }
+      }
+    }
+  }
+
   return nullptr;
 }
 

test/Transforms/InstCombine/getelementptr.ll

@@ -366,7 +366,7 @@ define i32 @test21() {
         %rval = load i32, i32* %pbobel
         ret i32 %rval
 ; CHECK-LABEL: @test21(
-; CHECK: getelementptr %intstruct, %intstruct* %pbob1, i64 0, i32 0
+; CHECK: getelementptr inbounds %intstruct, %intstruct* %pbob1, i64 0, i32 0
 }
 
 
@@ -540,8 +540,8 @@ define i8* @test32(i8* %v) {
 	%G = load i8*, i8** %F
 	ret i8* %G
 ; CHECK-LABEL: @test32(
-; CHECK: %D = getelementptr [4 x i8*], [4 x i8*]* %A, i64 0, i64 1
-; CHECK: %F = getelementptr [4 x i8*], [4 x i8*]* %A, i64 0, i64 2
+; CHECK: %D = getelementptr inbounds [4 x i8*], [4 x i8*]* %A, i64 0, i64 1
+; CHECK: %F = getelementptr inbounds [4 x i8*], [4 x i8*]* %A, i64 0, i64 2
 }
 
 ; PR3290

test/Transforms/InstSimplify/compare.ll

@@ -205,6 +205,19 @@ define i1 @gep16(i8* %ptr, i32 %a) {
 ; CHECK-NEXT: ret i1 false
 }
 
+define i1 @gep17() {
+; CHECK-LABEL: @gep17(
+  %alloca = alloca i32, align 4
+  %bc = bitcast i32* %alloca to [4 x i8]*
+  %gep1 = getelementptr inbounds i32, i32* %alloca, i32 1
+  %pti1 = ptrtoint i32* %gep1 to i32
+  %gep2 = getelementptr inbounds [4 x i8], [4 x i8]* %bc, i32 0, i32 1
+  %pti2 = ptrtoint i8* %gep2 to i32
+  %cmp = icmp ugt i32 %pti1, %pti2
+  ret i1 %cmp
+; CHECK-NEXT: ret i1 true
+}
+
 define i1 @zext(i32 %x) {
 ; CHECK-LABEL: @zext(
   %e1 = zext i32 %x to i64