MFC 261991:

Upgrade our copy of llvm/clang to 3.4 release.  This version supports
all of the features in the current working draft of the upcoming C++
standard, provisionally named C++1y.

The code generator's performance is greatly increased, and the loop
auto-vectorizer is now enabled at -Os and -O2 in addition to -O3.  The
PowerPC backend has made several major improvements to code generation
quality and compile time, and the X86, SPARC, ARM32, Aarch64 and SystemZ
backends have all seen major feature work.

Release notes for llvm and clang can be found here:
<http://llvm.org/releases/3.4/docs/ReleaseNotes.html>
<http://llvm.org/releases/3.4/tools/clang/docs/ReleaseNotes.html>

MFC 262121 (by emaste):

Update lldb for clang/llvm 3.4 import

This commit largely restores the lldb source to the upstream r196259
snapshot with the addition of threaded inferior support and a few bug
fixes.

Specific upstream lldb revisions restored include:
   SVN      git
  181387  779e6ac
  181703  7bef4e2
  182099  b31044e
  182650  f2dcf35
  182683  0d91b80
  183862  15c1774
  183929  99447a6
  184177  0b2934b
  184948  4dc3761
  184954  007e7bc
  186990  eebd175

Sponsored by:	DARPA, AFRL

MFC 262186 (by emaste):

Fix mismerge in r262121

A break statement was lost in the merge.  The error had no functional
impact, but restore it to reduce the diff against upstream.

MFC 262303:

Pull in r197521 from upstream clang trunk (by rdivacky):

  Use the integrated assembler by default on FreeBSD/ppc and ppc64.

Requested by:	jhibbits

MFC 262611:

Pull in r196874 from upstream llvm trunk:

  Fix a crash that occurs when PWD is invalid.

  MCJIT needs to be able to run in hostile environments, even when PWD
  is invalid. There's no need to crash MCJIT in this case.

  The obvious fix is to simply leave MCContext's CompilationDir empty
  when PWD can't be determined. This way, MCJIT clients,
  and other clients that link with LLVM don't need a valid working directory.

  If we do want to guarantee valid CompilationDir, that should be done
  only for clients of getCompilationDir(). This is as simple as checking
  for an empty string.

  The only current use of getCompilationDir is EmitGenDwarfInfo, which
  won't conceivably run with an invalid working dir. However, in the
  purely hypothetically and untestable case that this happens, the
  AT_comp_dir will be omitted from the compilation_unit DIE.

This should help fix assertions occurring with ports-mgmt/tinderbox,
when it is using jails, and sometimes invalidates clang's current
working directory.

Reported by:	decke

MFC 262809:

Pull in r203007 from upstream clang trunk:

  Don't produce an alias between destructors with different calling conventions.

  Fixes pr19007.

(Please note that is an LLVM PR identifier, not a FreeBSD one.)

This should fix Firefox and/or libxul crashes (due to problems with
regparm/stdcall calling conventions) on i386.

Reported by:	multiple users on freebsd-current
PR:		bin/187103

MFC 263048:

Repair recognition of "CC" as an alias for the C++ compiler, since it
was silently broken by upstream for a Windows-specific use-case.

Apparently some versions of CMake still rely on this archaic feature...

Reported by:	rakuco

MFC 263049:

Garbage collect the old way of adding the libstdc++ include directories
in clang's InitHeaderSearch.cpp.  This has been superseded by David
Chisnall's commit in r255321.

Moreover, if libc++ is used, the libstdc++ include directories should
not be in the search path at all.  These directories are now only used
if you pass -stdlib=libstdc++.

1 files changed, 70 insertions, 44 deletions

diff --git a/contrib/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp b/contrib/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
index ec10751..191a101 100644
--- a/contrib/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
+++ b/contrib/llvm/lib/Transforms/InstCombine/InstructionCombining.cpp
@@ -699,7 +699,10 @@ Instruction *InstCombiner::FoldOpIntoPhi(Instruction &I) {
       Value *TrueVInPred = TrueV->DoPHITranslation(PhiTransBB, ThisBB);
       Value *FalseVInPred = FalseV->DoPHITranslation(PhiTransBB, ThisBB);
       Value *InV = 0;
-      if (Constant *InC = dyn_cast<Constant>(PN->getIncomingValue(i)))
+      // Beware of ConstantExpr:  it may eventually evaluate to getNullValue,
+      // even if currently isNullValue gives false.
+      Constant *InC = dyn_cast<Constant>(PN->getIncomingValue(i));
+      if (InC && !isa<ConstantExpr>(InC))
         InV = InC->isNullValue() ? FalseVInPred : TrueVInPred;
       else
         InV = Builder->CreateSelect(PN->getIncomingValue(i),
@@ -755,19 +758,25 @@ Instruction *InstCombiner::FoldOpIntoPhi(Instruction &I) {
   return ReplaceInstUsesWith(I, NewPN);
 }
 
-/// FindElementAtOffset - Given a type and a constant offset, determine whether
-/// or not there is a sequence of GEP indices into the type that will land us at
-/// the specified offset.  If so, fill them into NewIndices and return the
-/// resultant element type, otherwise return null.
-Type *InstCombiner::FindElementAtOffset(Type *Ty, int64_t Offset,
-                                          SmallVectorImpl<Value*> &NewIndices) {
-  if (!TD) return 0;
-  if (!Ty->isSized()) return 0;
+/// FindElementAtOffset - Given a pointer type and a constant offset, determine
+/// whether or not there is a sequence of GEP indices into the pointed type that
+/// will land us at the specified offset.  If so, fill them into NewIndices and
+/// return the resultant element type, otherwise return null.
+Type *InstCombiner::FindElementAtOffset(Type *PtrTy, int64_t Offset,
+                                        SmallVectorImpl<Value*> &NewIndices) {
+  assert(PtrTy->isPtrOrPtrVectorTy());
+
+  if (!TD)
+    return 0;
+
+  Type *Ty = PtrTy->getPointerElementType();
+  if (!Ty->isSized())
+    return 0;
 
   // Start with the index over the outer type.  Note that the type size
   // might be zero (even if the offset isn't zero) if the indexed type
   // is something like [0 x {int, int}]
-  Type *IntPtrTy = TD->getIntPtrType(Ty->getContext());
+  Type *IntPtrTy = TD->getIntPtrType(PtrTy);
   int64_t FirstIdx = 0;
   if (int64_t TySize = TD->getTypeAllocSize(Ty)) {
     FirstIdx = Offset/TySize;
@@ -1176,6 +1185,22 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
         GetElementPtrInst::Create(Src->getOperand(0), Indices, GEP.getName());
   }
 
+  // Canonicalize (gep i8* X, -(ptrtoint Y)) to (sub (ptrtoint X), (ptrtoint Y))
+  // The GEP pattern is emitted by the SCEV expander for certain kinds of
+  // pointer arithmetic.
+  if (TD && GEP.getNumIndices() == 1 &&
+      match(GEP.getOperand(1), m_Neg(m_PtrToInt(m_Value())))) {
+    unsigned AS = GEP.getPointerAddressSpace();
+    if (GEP.getType() == Builder->getInt8PtrTy(AS) &&
+        GEP.getOperand(1)->getType()->getScalarSizeInBits() ==
+        TD->getPointerSizeInBits(AS)) {
+      Operator *Index = cast<Operator>(GEP.getOperand(1));
+      Value *PtrToInt = Builder->CreatePtrToInt(PtrOp, Index->getType());
+      Value *NewSub = Builder->CreateSub(PtrToInt, Index->getOperand(1));
+      return CastInst::Create(Instruction::IntToPtr, NewSub, GEP.getType());
+    }
+  }
+
   // Handle gep(bitcast x) and gep(gep x, 0, 0, 0).
   Value *StrippedPtr = PtrOp->stripPointerCasts();
   PointerType *StrippedPtrTy = dyn_cast<PointerType>(StrippedPtr->getType());
@@ -1231,13 +1256,12 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
       // %t = getelementptr i32* bitcast ([2 x i32]* %str to i32*), i32 %V
       // into:  %t1 = getelementptr [2 x i32]* %str, i32 0, i32 %V; bitcast
       Type *SrcElTy = StrippedPtrTy->getElementType();
-      Type *ResElTy=cast<PointerType>(PtrOp->getType())->getElementType();
+      Type *ResElTy = PtrOp->getType()->getPointerElementType();
       if (TD && SrcElTy->isArrayTy() &&
-          TD->getTypeAllocSize(cast<ArrayType>(SrcElTy)->getElementType()) ==
+          TD->getTypeAllocSize(SrcElTy->getArrayElementType()) ==
           TD->getTypeAllocSize(ResElTy)) {
-        Value *Idx[2];
-        Idx[0] = Constant::getNullValue(Type::getInt32Ty(GEP.getContext()));
-        Idx[1] = GEP.getOperand(1);
+        Type *IdxType = TD->getIntPtrType(GEP.getType());
+        Value *Idx[2] = { Constant::getNullValue(IdxType), GEP.getOperand(1) };
         Value *NewGEP = GEP.isInBounds() ?
           Builder->CreateInBoundsGEP(StrippedPtr, Idx, GEP.getName()) :
           Builder->CreateGEP(StrippedPtr, Idx, GEP.getName());
@@ -1261,7 +1285,7 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
 
           // Earlier transforms ensure that the index has type IntPtrType, which
           // considerably simplifies the logic by eliminating implicit casts.
-          assert(Idx->getType() == TD->getIntPtrType(GEP.getContext()) &&
+          assert(Idx->getType() == TD->getIntPtrType(GEP.getType()) &&
                  "Index not cast to pointer width?");
 
           bool NSW;
@@ -1287,8 +1311,8 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
         // Check that changing to the array element type amounts to dividing the
         // index by a scale factor.
         uint64_t ResSize = TD->getTypeAllocSize(ResElTy);
-        uint64_t ArrayEltSize =
-          TD->getTypeAllocSize(cast<ArrayType>(SrcElTy)->getElementType());
+        uint64_t ArrayEltSize
+          = TD->getTypeAllocSize(SrcElTy->getArrayElementType());
         if (ResSize && ArrayEltSize % ResSize == 0) {
           Value *Idx = GEP.getOperand(1);
           unsigned BitWidth = Idx->getType()->getPrimitiveSizeInBits();
@@ -1296,7 +1320,7 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
 
           // Earlier transforms ensure that the index has type IntPtrType, which
           // considerably simplifies the logic by eliminating implicit casts.
-          assert(Idx->getType() == TD->getIntPtrType(GEP.getContext()) &&
+          assert(Idx->getType() == TD->getIntPtrType(GEP.getType()) &&
                  "Index not cast to pointer width?");
 
           bool NSW;
@@ -1304,9 +1328,11 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
             // Successfully decomposed Idx as NewIdx * Scale, form a new GEP.
             // If the multiplication NewIdx * Scale may overflow then the new
             // GEP may not be "inbounds".
-            Value *Off[2];
-            Off[0] = Constant::getNullValue(Type::getInt32Ty(GEP.getContext()));
-            Off[1] = NewIdx;
+            Value *Off[2] = {
+              Constant::getNullValue(TD->getIntPtrType(GEP.getType())),
+              NewIdx
+            };
+
             Value *NewGEP = GEP.isInBounds() && NSW ?
               Builder->CreateInBoundsGEP(StrippedPtr, Off, GEP.getName()) :
               Builder->CreateGEP(StrippedPtr, Off, GEP.getName());
@@ -1318,15 +1344,20 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
     }
   }
 
+  if (!TD)
+    return 0;
+
   /// See if we can simplify:
   ///   X = bitcast A* to B*
   ///   Y = gep X, <...constant indices...>
   /// into a gep of the original struct.  This is important for SROA and alias
   /// analysis of unions.  If "A" is also a bitcast, wait for A/X to be merged.
   if (BitCastInst *BCI = dyn_cast<BitCastInst>(PtrOp)) {
-    APInt Offset(TD ? TD->getPointerSizeInBits() : 1, 0);
-    if (TD &&
-        !isa<BitCastInst>(BCI->getOperand(0)) &&
+    Value *Operand = BCI->getOperand(0);
+    PointerType *OpType = cast<PointerType>(Operand->getType());
+    unsigned OffsetBits = TD->getPointerTypeSizeInBits(OpType);
+    APInt Offset(OffsetBits, 0);
+    if (!isa<BitCastInst>(Operand) &&
         GEP.accumulateConstantOffset(*TD, Offset) &&
         StrippedPtrTy->getAddressSpace() == GEP.getPointerAddressSpace()) {
 
@@ -1335,8 +1366,7 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
       if (!Offset) {
         // If the bitcast is of an allocation, and the allocation will be
         // converted to match the type of the cast, don't touch this.
-        if (isa<AllocaInst>(BCI->getOperand(0)) ||
-            isAllocationFn(BCI->getOperand(0), TLI)) {
+        if (isa<AllocaInst>(Operand) || isAllocationFn(Operand, TLI)) {
           // See if the bitcast simplifies, if so, don't nuke this GEP yet.
           if (Instruction *I = visitBitCast(*BCI)) {
             if (I != BCI) {
@@ -1347,19 +1377,17 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
             return &GEP;
           }
         }
-        return new BitCastInst(BCI->getOperand(0), GEP.getType());
+        return new BitCastInst(Operand, GEP.getType());
       }
 
       // Otherwise, if the offset is non-zero, we need to find out if there is a
       // field at Offset in 'A's type.  If so, we can pull the cast through the
       // GEP.
       SmallVector<Value*, 8> NewIndices;
-      Type *InTy =
-        cast<PointerType>(BCI->getOperand(0)->getType())->getElementType();
-      if (FindElementAtOffset(InTy, Offset.getSExtValue(), NewIndices)) {
+      if (FindElementAtOffset(OpType, Offset.getSExtValue(), NewIndices)) {
         Value *NGEP = GEP.isInBounds() ?
-          Builder->CreateInBoundsGEP(BCI->getOperand(0), NewIndices) :
-          Builder->CreateGEP(BCI->getOperand(0), NewIndices);
+          Builder->CreateInBoundsGEP(Operand, NewIndices) :
+          Builder->CreateGEP(Operand, NewIndices);
 
         if (NGEP->getType() == GEP.getType())
           return ReplaceInstUsesWith(GEP, NGEP);
@@ -1372,8 +1400,6 @@ Instruction *InstCombiner::visitGetElementPtrInst(GetElementPtrInst &GEP) {
   return 0;
 }
 
-
-
 static bool
 isAllocSiteRemovable(Instruction *AI, SmallVectorImpl<WeakVH> &Users,
                      const TargetLibraryInfo *TLI) {
@@ -2042,7 +2068,7 @@ Instruction *InstCombiner::visitLandingPadInst(LandingPadInst &LI) {
         continue;
       // If Filter is a subset of LFilter, i.e. every element of Filter is also
       // an element of LFilter, then discard LFilter.
-      SmallVector<Value *, 16>::iterator J = NewClauses.begin() + j;
+      SmallVectorImpl<Value *>::iterator J = NewClauses.begin() + j;
       // If Filter is empty then it is a subset of LFilter.
       if (!FElts) {
         // Discard LFilter.
@@ -2209,7 +2235,7 @@ static bool AddReachableCodeToWorklist(BasicBlock *BB,
       // DCE instruction if trivially dead.
       if (isInstructionTriviallyDead(Inst, TLI)) {
         ++NumDeadInst;
-        DEBUG(errs() << "IC: DCE: " << *Inst << '\n');
+        DEBUG(dbgs() << "IC: DCE: " << *Inst << '\n');
         Inst->eraseFromParent();
         continue;
       }
@@ -2217,7 +2243,7 @@ static bool AddReachableCodeToWorklist(BasicBlock *BB,
       // ConstantProp instruction if trivially constant.
       if (!Inst->use_empty() && isa<Constant>(Inst->getOperand(0)))
         if (Constant *C = ConstantFoldInstruction(Inst, TD, TLI)) {
-          DEBUG(errs() << "IC: ConstFold to: " << *C << " from: "
+          DEBUG(dbgs() << "IC: ConstFold to: " << *C << " from: "
                        << *Inst << '\n');
           Inst->replaceAllUsesWith(C);
           ++NumConstProp;
@@ -2293,7 +2319,7 @@ static bool AddReachableCodeToWorklist(BasicBlock *BB,
 bool InstCombiner::DoOneIteration(Function &F, unsigned Iteration) {
   MadeIRChange = false;
 
-  DEBUG(errs() << "\n\nINSTCOMBINE ITERATION #" << Iteration << " on "
+  DEBUG(dbgs() << "\n\nINSTCOMBINE ITERATION #" << Iteration << " on "
                << F.getName() << "\n");
 
   {
@@ -2338,7 +2364,7 @@ bool InstCombiner::DoOneIteration(Function &F, unsigned Iteration) {
 
     // Check to see if we can DCE the instruction.
     if (isInstructionTriviallyDead(I, TLI)) {
-      DEBUG(errs() << "IC: DCE: " << *I << '\n');
+      DEBUG(dbgs() << "IC: DCE: " << *I << '\n');
       EraseInstFromFunction(*I);
       ++NumDeadInst;
       MadeIRChange = true;
@@ -2348,7 +2374,7 @@ bool InstCombiner::DoOneIteration(Function &F, unsigned Iteration) {
     // Instruction isn't dead, see if we can constant propagate it.
     if (!I->use_empty() && isa<Constant>(I->getOperand(0)))
       if (Constant *C = ConstantFoldInstruction(I, TD, TLI)) {
-        DEBUG(errs() << "IC: ConstFold to: " << *C << " from: " << *I << '\n');
+        DEBUG(dbgs() << "IC: ConstFold to: " << *C << " from: " << *I << '\n');
 
         // Add operands to the worklist.
         ReplaceInstUsesWith(*I, C);
@@ -2396,13 +2422,13 @@ bool InstCombiner::DoOneIteration(Function &F, unsigned Iteration) {
     std::string OrigI;
 #endif
     DEBUG(raw_string_ostream SS(OrigI); I->print(SS); OrigI = SS.str(););
-    DEBUG(errs() << "IC: Visiting: " << OrigI << '\n');
+    DEBUG(dbgs() << "IC: Visiting: " << OrigI << '\n');
 
     if (Instruction *Result = visit(*I)) {
       ++NumCombined;
       // Should we replace the old instruction with a new one?
       if (Result != I) {
-        DEBUG(errs() << "IC: Old = " << *I << '\n'
+        DEBUG(dbgs() << "IC: Old = " << *I << '\n'
                      << "    New = " << *Result << '\n');
 
         if (!I->getDebugLoc().isUnknown())
@@ -2431,7 +2457,7 @@ bool InstCombiner::DoOneIteration(Function &F, unsigned Iteration) {
         EraseInstFromFunction(*I);
       } else {
 #ifndef NDEBUG
-        DEBUG(errs() << "IC: Mod = " << OrigI << '\n'
+        DEBUG(dbgs() << "IC: Mod = " << OrigI << '\n'
                      << "    New = " << *I << '\n');
 #endif