Add a dereferenceable attribute

This attribute indicates that the parameter or return pointer is
dereferenceable. Practically speaking, loads from such a pointer within the
associated byte range are safe to speculatively execute. Such pointer
parameters are common in source languages (C++ references, for example).

llvm-svn: 213385

11 files changed, 177 insertions, 13 deletions

diff --git a/llvm/lib/Analysis/ValueTracking.cpp b/llvm/lib/Analysis/ValueTracking.cpp
index cedf076497e..e6d09f4e31f 100644
--- a/llvm/lib/Analysis/ValueTracking.cpp
+++ b/llvm/lib/Analysis/ValueTracking.cpp
@@ -2086,7 +2086,7 @@ bool llvm::isKnownNonNull(const Value *V, const TargetLibraryInfo *TLI) {
     return !GV->hasExternalWeakLinkage();
 
   if (ImmutableCallSite CS = V)
-    if (CS.paramHasAttr(0, Attribute::NonNull))
+    if (CS.isReturnNonNull())
       return true;
 
   // operator new never returns null.
diff --git a/llvm/lib/AsmParser/LLLexer.cpp b/llvm/lib/AsmParser/LLLexer.cpp
index 1e5bcdd930c..962298fcd54 100644
--- a/llvm/lib/AsmParser/LLLexer.cpp
+++ b/llvm/lib/AsmParser/LLLexer.cpp
@@ -612,6 +612,7 @@ lltok::Kind LLLexer::LexIdentifier() {
   KEYWORD(byval);
   KEYWORD(inalloca);
   KEYWORD(cold);
+  KEYWORD(dereferenceable);
   KEYWORD(inlinehint);
   KEYWORD(inreg);
   KEYWORD(jumptable);
diff --git a/llvm/lib/AsmParser/LLParser.cpp b/llvm/lib/AsmParser/LLParser.cpp
index be55ac6481f..ac6e0e512fe 100644
--- a/llvm/lib/AsmParser/LLParser.cpp
+++ b/llvm/lib/AsmParser/LLParser.cpp
@@ -1052,6 +1052,7 @@ bool LLParser::ParseFnAttributeValuePairs(AttrBuilder &B,
               "invalid use of attribute on a function");
       break;
     case lltok::kw_byval:
+    case lltok::kw_dereferenceable:
     case lltok::kw_inalloca:
     case lltok::kw_nest:
     case lltok::kw_noalias:
@@ -1212,6 +1213,16 @@ bool LLParser::ParseUInt32(unsigned &Val) {
   return false;
 }
 
+/// ParseUInt64
+///   ::= uint64
+bool LLParser::ParseUInt64(uint64_t &Val) {
+  if (Lex.getKind() != lltok::APSInt || Lex.getAPSIntVal().isSigned())
+    return TokError("expected integer");
+  Val = Lex.getAPSIntVal().getLimitedValue();
+  Lex.Lex();
+  return false;
+}
+
 /// ParseTLSModel
 ///   := 'localdynamic'
 ///   := 'initialexec'
@@ -1284,6 +1295,13 @@ bool LLParser::ParseOptionalParamAttrs(AttrBuilder &B) {
       continue;
     }
     case lltok::kw_byval:           B.addAttribute(Attribute::ByVal); break;
+    case lltok::kw_dereferenceable: {
+      uint64_t Bytes;
+      if (ParseOptionalDereferenceableBytes(Bytes))
+        return true;
+      B.addDereferenceableAttr(Bytes);
+      continue;
+    }
     case lltok::kw_inalloca:        B.addAttribute(Attribute::InAlloca); break;
     case lltok::kw_inreg:           B.addAttribute(Attribute::InReg); break;
     case lltok::kw_nest:            B.addAttribute(Attribute::Nest); break;
@@ -1341,6 +1359,13 @@ bool LLParser::ParseOptionalReturnAttrs(AttrBuilder &B) {
     switch (Token) {
     default:  // End of attributes.
       return HaveError;
+    case lltok::kw_dereferenceable: {
+      uint64_t Bytes;
+      if (ParseOptionalDereferenceableBytes(Bytes))
+        return true;
+      B.addDereferenceableAttr(Bytes);
+      continue;
+    }
     case lltok::kw_inreg:           B.addAttribute(Attribute::InReg); break;
     case lltok::kw_noalias:         B.addAttribute(Attribute::NoAlias); break;
     case lltok::kw_nonnull:         B.addAttribute(Attribute::NonNull); break;
@@ -1606,6 +1631,26 @@ bool LLParser::ParseOptionalAlignment(unsigned &Alignment) {
   return false;
 }
 
+/// ParseOptionalDereferenceableBytes
+///   ::= /* empty */
+///   ::= 'dereferenceable' '(' 4 ')'
+bool LLParser::ParseOptionalDereferenceableBytes(uint64_t &Bytes) {
+  Bytes = 0;
+  if (!EatIfPresent(lltok::kw_dereferenceable))
+    return false;
+  LocTy ParenLoc = Lex.getLoc();
+  if (!EatIfPresent(lltok::lparen))
+    return Error(ParenLoc, "expected '('");
+  LocTy DerefLoc = Lex.getLoc();
+  if (ParseUInt64(Bytes)) return true;
+  ParenLoc = Lex.getLoc();
+  if (!EatIfPresent(lltok::rparen))
+    return Error(ParenLoc, "expected ')'");
+  if (!Bytes)
+    return Error(DerefLoc, "dereferenceable bytes must be non-zero");
+  return false;
+}
+
 /// ParseOptionalCommaAlign
 ///   ::=
 ///   ::= ',' align 4
diff --git a/llvm/lib/AsmParser/LLParser.h b/llvm/lib/AsmParser/LLParser.h
index 2efb260d66b..7203bb245d0 100644
--- a/llvm/lib/AsmParser/LLParser.h
+++ b/llvm/lib/AsmParser/LLParser.h
@@ -202,6 +202,11 @@ namespace llvm {
       Loc = Lex.getLoc();
       return ParseUInt32(Val);
     }
+    bool ParseUInt64(uint64_t &Val);
+    bool ParseUInt64(uint64_t &Val, LocTy &Loc) {
+      Loc = Lex.getLoc();
+      return ParseUInt64(Val);
+    }
 
     bool ParseTLSModel(GlobalVariable::ThreadLocalMode &TLM);
     bool ParseOptionalThreadLocal(GlobalVariable::ThreadLocalMode &TLM);
@@ -219,6 +224,7 @@ namespace llvm {
     bool ParseOptionalDLLStorageClass(unsigned &DLLStorageClass);
     bool ParseOptionalCallingConv(CallingConv::ID &CC);
     bool ParseOptionalAlignment(unsigned &Alignment);
+    bool ParseOptionalDereferenceableBytes(uint64_t &Bytes);
     bool ParseScopeAndOrdering(bool isAtomic, SynchronizationScope &Scope,
                                AtomicOrdering &Ordering);
     bool ParseOrdering(AtomicOrdering &Ordering);
diff --git a/llvm/lib/AsmParser/LLToken.h b/llvm/lib/AsmParser/LLToken.h
index 534d82415fd..2f02606f93e 100644
--- a/llvm/lib/AsmParser/LLToken.h
+++ b/llvm/lib/AsmParser/LLToken.h
@@ -106,6 +106,7 @@ namespace lltok {
     kw_byval,
     kw_inalloca,
     kw_cold,
+    kw_dereferenceable,
     kw_inlinehint,
     kw_inreg,
     kw_jumptable,
diff --git a/llvm/lib/Bitcode/Reader/BitcodeReader.cpp b/llvm/lib/Bitcode/Reader/BitcodeReader.cpp
index 4a0484777cf..47a39539e20 100644
--- a/llvm/lib/Bitcode/Reader/BitcodeReader.cpp
+++ b/llvm/lib/Bitcode/Reader/BitcodeReader.cpp
@@ -588,6 +588,8 @@ static Attribute::AttrKind GetAttrFromCode(uint64_t Code) {
     return Attribute::NonLazyBind;
   case bitc::ATTR_KIND_NON_NULL:
     return Attribute::NonNull;
+  case bitc::ATTR_KIND_DEREFERENCEABLE:
+    return Attribute::Dereferenceable;
   case bitc::ATTR_KIND_NO_RED_ZONE:
     return Attribute::NoRedZone;
   case bitc::ATTR_KIND_NO_RETURN:
@@ -689,8 +691,10 @@ std::error_code BitcodeReader::ParseAttributeGroupBlock() {
             return EC;
           if (Kind == Attribute::Alignment)
             B.addAlignmentAttr(Record[++i]);
-          else
+          else if (Kind == Attribute::StackAlignment)
             B.addStackAlignmentAttr(Record[++i]);
+          else if (Kind == Attribute::Dereferenceable)
+            B.addDereferenceableAttr(Record[++i]);
         } else {                     // String attribute
           assert((Record[i] == 3 || Record[i] == 4) &&
                  "Invalid attribute group entry");
diff --git a/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp b/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp
index 67710472a15..b2e49486d70 100644
--- a/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp
+++ b/llvm/lib/Bitcode/Writer/BitcodeWriter.cpp
@@ -201,6 +201,8 @@ static uint64_t getAttrKindEncoding(Attribute::AttrKind Kind) {
     return bitc::ATTR_KIND_NON_LAZY_BIND;
   case Attribute::NonNull:
     return bitc::ATTR_KIND_NON_NULL;
+  case Attribute::Dereferenceable:
+    return bitc::ATTR_KIND_DEREFERENCEABLE;
   case Attribute::NoRedZone:
     return bitc::ATTR_KIND_NO_RED_ZONE;
   case Attribute::NoReturn:
diff --git a/llvm/lib/IR/AttributeImpl.h b/llvm/lib/IR/AttributeImpl.h
index 999a803c037..cc6d557ab4c 100644
--- a/llvm/lib/IR/AttributeImpl.h
+++ b/llvm/lib/IR/AttributeImpl.h
@@ -116,7 +116,8 @@ public:
   IntAttributeImpl(Attribute::AttrKind Kind, uint64_t Val)
       : EnumAttributeImpl(IntAttrEntry, Kind), Val(Val) {
     assert(
-        (Kind == Attribute::Alignment || Kind == Attribute::StackAlignment) &&
+        (Kind == Attribute::Alignment || Kind == Attribute::StackAlignment ||
+         Kind == Attribute::Dereferenceable) &&
         "Wrong kind for int attribute!");
   }
 
@@ -164,6 +165,7 @@ public:
 
   unsigned getAlignment() const;
   unsigned getStackAlignment() const;
+  uint64_t getDereferenceableBytes() const;
   std::string getAsString(bool InAttrGrp) const;
 
   typedef const Attribute *iterator;
diff --git a/llvm/lib/IR/Attributes.cpp b/llvm/lib/IR/Attributes.cpp
index 3ee622b31ba..04545ea919a 100644
--- a/llvm/lib/IR/Attributes.cpp
+++ b/llvm/lib/IR/Attributes.cpp
@@ -88,6 +88,12 @@ Attribute Attribute::getWithStackAlignment(LLVMContext &Context,
   return get(Context, StackAlignment, Align);
 }
 
+Attribute Attribute::getWithDereferenceableBytes(LLVMContext &Context,
+                                                uint64_t Bytes) {
+  assert(Bytes && "Bytes must be non-zero.");
+  return get(Context, Dereferenceable, Bytes);
+}
+
 //===----------------------------------------------------------------------===//
 // Attribute Accessor Methods
 //===----------------------------------------------------------------------===//
@@ -156,6 +162,14 @@ unsigned Attribute::getStackAlignment() const {
   return pImpl->getValueAsInt();
 }
 
+/// This returns the number of dereferenceable bytes.
+uint64_t Attribute::getDereferenceableBytes() const {
+  assert(hasAttribute(Attribute::Dereferenceable) &&
+         "Trying to get dereferenceable bytes from "
+         "non-dereferenceable attribute!");
+  return pImpl->getValueAsInt();
+}
+
 std::string Attribute::getAsString(bool InAttrGrp) const {
   if (!pImpl) return "";
 
@@ -263,6 +277,20 @@ std::string Attribute::getAsString(bool InAttrGrp) const {
     return Result;
   }
 
+  if (hasAttribute(Attribute::Dereferenceable)) {
+    std::string Result;
+    Result += "dereferenceable";
+    if (InAttrGrp) {
+      Result += "=";
+      Result += utostr(getValueAsInt());
+    } else {
+      Result += "(";
+      Result += utostr(getValueAsInt());
+      Result += ")";
+    }
+    return Result;
+  }
+
   // Convert target-dependent attributes to strings of the form:
   //
   //   "kind"
@@ -398,6 +426,8 @@ uint64_t AttributeImpl::getAttrMask(Attribute::AttrKind Val) {
   case Attribute::InAlloca:        return 1ULL << 43;
   case Attribute::NonNull:         return 1ULL << 44;
   case Attribute::JumpTable:       return 1ULL << 45;
+  case Attribute::Dereferenceable:
+    llvm_unreachable("dereferenceable attribute not supported in raw format");
   }
   llvm_unreachable("Unsupported attribute type");
 }
@@ -482,6 +512,13 @@ unsigned AttributeSetNode::getStackAlignment() const {
   return 0;
 }
 
+uint64_t AttributeSetNode::getDereferenceableBytes() const {
+  for (iterator I = begin(), E = end(); I != E; ++I)
+    if (I->hasAttribute(Attribute::Dereferenceable))
+      return I->getDereferenceableBytes();
+  return 0;
+}
+
 std::string AttributeSetNode::getAsString(bool InAttrGrp) const {
   std::string Str;
   for (iterator I = begin(), E = end(); I != E; ++I) {
@@ -515,6 +552,8 @@ uint64_t AttributeSetImpl::Raw(unsigned Index) const {
         Mask |= (Log2_32(ASN->getAlignment()) + 1) << 16;
       else if (Kind == Attribute::StackAlignment)
         Mask |= (Log2_32(ASN->getStackAlignment()) + 1) << 26;
+      else if (Kind == Attribute::Dereferenceable)
+        llvm_unreachable("dereferenceable not supported in bit mask");
       else
         Mask |= AttributeImpl::getAttrMask(Kind);
     }
@@ -620,6 +659,10 @@ AttributeSet AttributeSet::get(LLVMContext &C, unsigned Index,
     else if (Kind == Attribute::StackAlignment)
       Attrs.push_back(std::make_pair(Index, Attribute::
                               getWithStackAlignment(C, B.getStackAlignment())));
+    else if (Kind == Attribute::Dereferenceable)
+      Attrs.push_back(std::make_pair(Index,
+                                     Attribute::getWithDereferenceableBytes(C,
+                                       B.getDereferenceableBytes())));
     else
       Attrs.push_back(std::make_pair(Index, Attribute::get(C, Kind)));
   }
@@ -877,6 +920,11 @@ unsigned AttributeSet::getStackAlignment(unsigned Index) const {
   return ASN ? ASN->getStackAlignment() : 0;
 }
 
+uint64_t AttributeSet::getDereferenceableBytes(unsigned Index) const {
+  AttributeSetNode *ASN = getAttributes(Index);
+  return ASN ? ASN->getDereferenceableBytes() : 0;
+}
+
 std::string AttributeSet::getAsString(unsigned Index,
                                       bool InAttrGrp) const {
   AttributeSetNode *ASN = getAttributes(Index);
@@ -956,7 +1004,7 @@ void AttributeSet::dump() const {
 //===----------------------------------------------------------------------===//
 
 AttrBuilder::AttrBuilder(AttributeSet AS, unsigned Index)
-  : Attrs(0), Alignment(0), StackAlignment(0) {
+  : Attrs(0), Alignment(0), StackAlignment(0), DerefBytes(0) {
   AttributeSetImpl *pImpl = AS.pImpl;
   if (!pImpl) return;
 
@@ -973,13 +1021,14 @@ AttrBuilder::AttrBuilder(AttributeSet AS, unsigned Index)
 
 void AttrBuilder::clear() {
   Attrs.reset();
-  Alignment = StackAlignment = 0;
+  Alignment = StackAlignment = DerefBytes = 0;
 }
 
 AttrBuilder &AttrBuilder::addAttribute(Attribute::AttrKind Val) {
   assert((unsigned)Val < Attribute::EndAttrKinds && "Attribute out of range!");
   assert(Val != Attribute::Alignment && Val != Attribute::StackAlignment &&
-         "Adding alignment attribute without adding alignment value!");
+         Val != Attribute::Dereferenceable &&
+         "Adding integer attribute without adding a value!");
   Attrs[Val] = true;
   return *this;
 }
@@ -997,6 +1046,8 @@ AttrBuilder &AttrBuilder::addAttribute(Attribute Attr) {
     Alignment = Attr.getAlignment();
   else if (Kind == Attribute::StackAlignment)
     StackAlignment = Attr.getStackAlignment();
+  else if (Kind == Attribute::Dereferenceable)
+    DerefBytes = Attr.getDereferenceableBytes();
   return *this;
 }
 
@@ -1013,6 +1064,8 @@ AttrBuilder &AttrBuilder::removeAttribute(Attribute::AttrKind Val) {
     Alignment = 0;
   else if (Val == Attribute::StackAlignment)
     StackAlignment = 0;
+  else if (Val == Attribute::Dereferenceable)
+    DerefBytes = 0;
 
   return *this;
 }
@@ -1037,6 +1090,8 @@ AttrBuilder &AttrBuilder::removeAttributes(AttributeSet A, uint64_t Index) {
         Alignment = 0;
       else if (Kind == Attribute::StackAlignment)
         StackAlignment = 0;
+      else if (Kind == Attribute::Dereferenceable)
+        DerefBytes = 0;
     } else {
       assert(Attr.isStringAttribute() && "Invalid attribute type!");
       std::map<std::string, std::string>::iterator
@@ -1079,6 +1134,14 @@ AttrBuilder &AttrBuilder::addStackAlignmentAttr(unsigned Align) {
   return *this;
 }
 
+AttrBuilder &AttrBuilder::addDereferenceableAttr(uint64_t Bytes) {
+  if (Bytes == 0) return *this;
+
+  Attrs[Attribute::Dereferenceable] = true;
+  DerefBytes = Bytes;
+  return *this;
+}
+
 AttrBuilder &AttrBuilder::merge(const AttrBuilder &B) {
   // FIXME: What if both have alignments, but they don't match?!
   if (!Alignment)
@@ -1087,6 +1150,9 @@ AttrBuilder &AttrBuilder::merge(const AttrBuilder &B) {
   if (!StackAlignment)
     StackAlignment = B.StackAlignment;
 
+  if (!DerefBytes)
+    DerefBytes = B.DerefBytes;
+
   Attrs |= B.Attrs;
 
   for (td_const_iterator I = B.TargetDepAttrs.begin(),
@@ -1142,7 +1208,8 @@ bool AttrBuilder::operator==(const AttrBuilder &B) {
     if (B.TargetDepAttrs.find(I->first) == B.TargetDepAttrs.end())
       return false;
 
-  return Alignment == B.Alignment && StackAlignment == B.StackAlignment;
+  return Alignment == B.Alignment && StackAlignment == B.StackAlignment &&
+         DerefBytes == B.DerefBytes;
 }
 
 AttrBuilder &AttrBuilder::addRawValue(uint64_t Val) {
@@ -1151,6 +1218,8 @@ AttrBuilder &AttrBuilder::addRawValue(uint64_t Val) {
 
   for (Attribute::AttrKind I = Attribute::None; I != Attribute::EndAttrKinds;
        I = Attribute::AttrKind(I + 1)) {
+    if (I == Attribute::Dereferenceable)
+      continue;
     if (uint64_t A = (Val & AttributeImpl::getAttrMask(I))) {
       Attrs[I] = true;
  
@@ -1184,6 +1253,7 @@ AttributeSet AttributeFuncs::typeIncompatible(Type *Ty, uint64_t Index) {
       .addAttribute(Attribute::NoAlias)
       .addAttribute(Attribute::NoCapture)
       .addAttribute(Attribute::NonNull)
+      .addDereferenceableAttr(1) // the int here is ignored
       .addAttribute(Attribute::ReadNone)
       .addAttribute(Attribute::ReadOnly)
       .addAttribute(Attribute::StructRet)
diff --git a/llvm/lib/IR/Function.cpp b/llvm/lib/IR/Function.cpp
index 14435711bb6..de59b26ec92 100644
--- a/llvm/lib/IR/Function.cpp
+++ b/llvm/lib/IR/Function.cpp
@@ -77,11 +77,17 @@ unsigned Argument::getArgNo() const {
 }
 
 /// hasNonNullAttr - Return true if this argument has the nonnull attribute on
-/// it in its containing function.
+/// it in its containing function. Also returns true if at least one byte is
+/// known to be dereferenceable and the pointer is in addrspace(0).
 bool Argument::hasNonNullAttr() const {
   if (!getType()->isPointerTy()) return false;
-  return getParent()->getAttributes().
-    hasAttribute(getArgNo()+1, Attribute::NonNull);
+  if (getParent()->getAttributes().
+        hasAttribute(getArgNo()+1, Attribute::NonNull))
+    return true;
+  else if (getDereferenceableBytes() > 0 &&
+           getType()->getPointerAddressSpace() == 0)
+    return true;
+  return false;
 }
 
 /// hasByValAttr - Return true if this argument has the byval attribute on it
@@ -113,6 +119,12 @@ unsigned Argument::getParamAlignment() const {
 
 }
 
+uint64_t Argument::getDereferenceableBytes() const {
+  assert(getType()->isPointerTy() &&
+         "Only pointers have dereferenceable bytes");
+  return getParent()->getDereferenceableBytes(getArgNo()+1);
+}
+
 /// hasNestAttr - Return true if this argument has the nest attribute on
 /// it in its containing function.
 bool Argument::hasNestAttr() const {
diff --git a/llvm/lib/IR/Value.cpp b/llvm/lib/IR/Value.cpp
index 999685e15eb..efdd6a7ed6e 100644
--- a/llvm/lib/IR/Value.cpp
+++ b/llvm/lib/IR/Value.cpp
@@ -15,6 +15,7 @@
 #include "LLVMContextImpl.h"
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/SmallString.h"
+#include "llvm/IR/CallSite.h"
 #include "llvm/IR/Constant.h"
 #include "llvm/IR/Constants.h"
 #include "llvm/IR/DataLayout.h"
@@ -504,9 +505,29 @@ static bool isDereferenceablePointer(const Value *V, const DataLayout *DL,
   if (const GlobalVariable *GV = dyn_cast<GlobalVariable>(V))
     return !GV->hasExternalWeakLinkage();
 
-  // byval arguments are ok.
-  if (const Argument *A = dyn_cast<Argument>(V))
-    return A->hasByValAttr();
+  // byval arguments are okay. Arguments specifically marked as
+  // dereferenceable are okay too.
+  if (const Argument *A = dyn_cast<Argument>(V)) {
+    if (A->hasByValAttr())
+      return true;
+    else if (uint64_t Bytes = A->getDereferenceableBytes()) {
+      Type *Ty = V->getType()->getPointerElementType();
+      if (Ty->isSized() && DL && DL->getTypeStoreSize(Ty) <= Bytes)
+        return true;
+    }
+
+    return false;
+  }
+
+  // Return values from call sites specifically marked as dereferenceable are
+  // also okay.
+  if (ImmutableCallSite CS = V) {
+    if (uint64_t Bytes = CS.getDereferenceableBytes(0)) {
+      Type *Ty = V->getType()->getPointerElementType();
+      if (Ty->isSized() && DL && DL->getTypeStoreSize(Ty) <= Bytes)
+        return true;
+    }
+  }
 
   // For GEPs, determine if the indexing lands within the allocated object.
   if (const GEPOperator *GEP = dyn_cast<GEPOperator>(V)) {