Merge pull request #282 from correctcomputation/iss281

Major refactoring of array bounds code and some performance improvements

Author: Machiry

clang/include/clang/CConv/ABounds.h

@@ -44,13 +44,13 @@ class ABounds {
   virtual ~ABounds() { }
 
   virtual std::string mkString(AVarBoundsInfo *) = 0;
-  virtual bool areSame(ABounds *) = 0;
+  virtual bool areSame(ABounds *, AVarBoundsInfo *) = 0;
   virtual BoundsKey getBKey() = 0;
   virtual ABounds* makeCopy(BoundsKey NK) = 0;
 
   // Set that maintains all the bound keys that are used inin
+  // TODO: Is this still needed?
   static std::set<BoundsKey> KeysUsedInBounds;
-
   static bool isKeyUsedInBounds(BoundsKey ToCheck);
 
   static ABounds *getBoundsInfo(AVarBoundsInfo *AVBInfo,
@@ -66,7 +66,7 @@ class CountBound : public ABounds {
   virtual ~CountBound() { }
 
   std::string mkString(AVarBoundsInfo *ABI) override ;
-  bool areSame(ABounds *O) override;
+  bool areSame(ABounds *O, AVarBoundsInfo *ABI) override;
   BoundsKey getBKey() override;
   ABounds* makeCopy(BoundsKey NK) override;
 
@@ -88,7 +88,7 @@ class ByteBound : public ABounds {
   virtual ~ByteBound() { }
 
   std::string mkString(AVarBoundsInfo *ABI) override ;
-  bool areSame(ABounds *O) override;
+  bool areSame(ABounds *O, AVarBoundsInfo *ABI) override;
   BoundsKey getBKey() override;
   ABounds* makeCopy(BoundsKey NK) override;
 
@@ -111,7 +111,7 @@ class RangeBound : public ABounds {
   virtual ~RangeBound() { }
 
   std::string mkString(AVarBoundsInfo *ABI) override ;
-  bool areSame(ABounds *O) override;
+  bool areSame(ABounds *O, AVarBoundsInfo *ABI) override;
 
   BoundsKey getBKey() override {
     assert (false && "Not implemented.");

clang/include/clang/CConv/AVarBoundsInfo.h

@@ -89,40 +89,69 @@ enum BoundsPriority {
 };
 
 class AVarBoundsInfo;
-
+typedef std::map<ABounds::BoundsKind, std::set<BoundsKey>> BndsKindMap;
 // The main class that handles figuring out bounds of arr variables.
 class AvarBoundsInference {
 public:
-  AvarBoundsInference(AVarBoundsInfo *BoundsInfo) : BI(BoundsInfo) { }
+  AvarBoundsInference(AVarBoundsInfo *BoundsInfo) : BI(BoundsInfo) {
+    clearInferredBounds();
+  }
+
+  // Clear all possible inferred bounds for all the BoundsKeys
+  void clearInferredBounds() {
+    CurrIterInferBounds.clear();
+    BKsFailedFlowInference.clear();
+  }
 
   // Infer bounds for the given key from the set of given ARR atoms.
   // The flag FromPB requests the inference to use potential length variables.
   bool inferBounds(BoundsKey K, AVarGraph &BKGraph, bool FromPB = false);
+
+  // Get a consistent bound for all the arrays whose bounds have been
+  // inferred.
+  bool convergeInferredBounds();
 private:
-  bool inferPossibleBounds(BoundsKey K, ABounds *SB,
-                           AVarGraph &BKGraph,
-                           std::set<ABounds *> &EB);
+  // Find all the reachable variables form FromVarK that are visible
+  // in DstScope
+  bool getReachableBoundKeys(const ProgramVarScope *DstScope,
+                             BoundsKey FromVarK,
+                             std::set<BoundsKey> &PotK,
+                             AVarGraph &BKGraph,
+                             bool CheckImmediate = false);
 
   bool intersectBounds(std::set<ProgramVar *> &ProgVars,
                        ABounds::BoundsKind BK,
                        std::set<ABounds *> &CurrB);
 
-  bool getRelevantBounds(std::set<BoundsKey> &RBKeys,
-                         std::set<ABounds *> &ResBounds);
+  // Check if bounds specified by Bnds are declared bounds of K.
+  bool areDeclaredBounds(BoundsKey K,
+                         const std::pair<ABounds::BoundsKind,
+                                         std::set<BoundsKey>> &Bnds);
 
-  bool predictBounds(BoundsKey K, std::set<BoundsKey> &Neighbours,
-                     AVarGraph &BKGraph,
-                     ABounds **KB);
+  // Get all the bounds of the given array i.e., BK
+  bool getRelevantBounds(BoundsKey BK,
+                         BndsKindMap &ResBounds);
 
+  // Predict possible bounds for DstArrK from the bounds of  Neighbours.
+  // Return true if there is any change in the captured bounds information.
+  bool predictBounds(BoundsKey DstArrK, std::set<BoundsKey> &Neighbours,
+                     AVarGraph &BKGraph);
 
-  void mergeReachableProgramVars(std::set<ProgramVar *> &AllVars);
+
+  void mergeReachableProgramVars(std::set<BoundsKey> &AllVars);
 
   AVarBoundsInfo *BI;
+
+  // Potential Bounds for each bounds key inferred for the current iteration.
+  std::map<BoundsKey, BndsKindMap> CurrIterInferBounds;
+  // BoundsKey that failed the flow inference.
+  std::set<BoundsKey> BKsFailedFlowInference;
 };
 
 class AVarBoundsInfo {
 public:
-  AVarBoundsInfo() : ProgVarGraph(this), CtxSensProgVarGraph(this) {
+  AVarBoundsInfo() : ProgVarGraph(this), CtxSensProgVarGraph(this),
+                     RevCtxSensProgVarGraph(this) {
     BCount = 1;
     PVarInfo.clear();
     InProgramArrPtrBoundsKeys.clear();
@@ -163,6 +192,7 @@ class AVarBoundsInfo {
   BoundsKey getVariable(clang::FieldDecl *FD);
   BoundsKey getVariable(clang::FunctionDecl *FD);
   BoundsKey getConstKey(uint64_t value);
+  bool fetchAllConstKeys(uint64_t value, std::set<BoundsKey> &AllKeys);
 
   // Generate a random bounds key to be used for inference.
   BoundsKey getRandomBKey();
@@ -227,6 +257,8 @@ class AVarBoundsInfo {
                    const CVarSet &SrcCVarSet,
                    bool JsonFormat = false) const;
 
+  bool areSameProgramVar(BoundsKey B1, BoundsKey B2);
+
 private:
   friend class AvarBoundsInference;
 
@@ -238,8 +270,8 @@ class AVarBoundsInfo {
   BoundsKey BCount;
   // Map of VarKeys and corresponding program variables.
   std::map<BoundsKey, ProgramVar *> PVarInfo;
-  // Map of APSInt (constants) and corresponding VarKeys.
-  std::map<uint64_t, BoundsKey> ConstVarKeys;
+  // Map of APSInt (constants) and set of BoundKeys that correspond to it.
+  std::map<uint64_t, std::set<BoundsKey>> ConstVarKeys;
   // Map of BoundsKey and corresponding prioritized bounds information.
   // Note that although each PSL could have multiple ConstraintKeys Ex: **p.
   // Only the outer most pointer can have bounds.
@@ -272,9 +304,11 @@ class AVarBoundsInfo {
 
   // Graph of all program variables.
   AVarGraph ProgVarGraph;
-  // Graph that contains only edges between context-sensitive
-  // BoundsKey and corresponding original BoundsKey.
+  // Graph that contains only edges from normal BoundsKey to
+  // context-sensitive BoundsKey.
   AVarGraph CtxSensProgVarGraph;
+  // Same as above but in the reverse direction.
+  AVarGraph RevCtxSensProgVarGraph;
   // Stats on techniques used to find length for various variables.
   AVarBoundsStats BoundsInferStats;
   // This is the map of pointer variable bounds key and set of bounds key
@@ -305,19 +339,22 @@ class AVarBoundsInfo {
   // Check if the provided bounds key corresponds to function return.
   bool isFunctionReturn(BoundsKey BK);
 
-  // Of all teh pointer bounds key, find arr pointers.
+  // Of all the pointer bounds key, find arr pointers.
   void computerArrPointers(ProgramInfo *PI, std::set<BoundsKey> &Ret);
 
+  // Get all the array pointers that need bounds.
+  void getBoundsNeededArrPointers(const std::set<BoundsKey> &ArrPtrs,
+                                  std::set<BoundsKey> &AB);
+
   // Keep only highest priority bounds for all the provided BoundsKeys
   // returns true if any thing changed, else false.
   bool keepHighestPriorityBounds(std::set<BoundsKey> &ArrPtrs);
 
   // Perform worklist based inference on the requested array variables using
-  // the provided graph.
-  // The flag FromPB requests the algorithm to use potential length variables.
-  bool performWorkListInference(std::set<BoundsKey> &ArrNeededBounds,
+  // the provided graph and potential length variables.
+  bool performWorkListInference(const std::set<BoundsKey> &ArrNeededBounds,
                                 AVarGraph &BKGraph,
-                                bool FromPB = false);
+                                AvarBoundsInference &BI);
 
   void insertParamKey(ParamDeclType ParamDecl, BoundsKey NK);
 };

clang/include/clang/CConv/ArrayBoundsInferenceConsumer.h

@@ -22,6 +22,15 @@
 class LocalVarABVisitor;
 class ConstraintResolver;
 
+class AllocBasedBoundsInference : public ASTConsumer {
+public:
+  explicit AllocBasedBoundsInference(ProgramInfo &I, clang::ASTContext *C) : Info(I) { }
+  virtual void HandleTranslationUnit(ASTContext &Context);
+
+private:
+  ProgramInfo &Info;
+};
+
 // This class handles determining bounds of global array variables.
 // i.e., function parameters, structure fields and global variables.
 class GlobalABVisitor: public clang::RecursiveASTVisitor<GlobalABVisitor> {
@@ -96,7 +105,8 @@ class LengthVarInference : public StmtVisitor<LengthVarInference> {
   std::unique_ptr<CFG> Cfg;
 };
 
-void HandleArrayVariablesBoundsDetection(ASTContext *C, ProgramInfo &I);
+void HandleArrayVariablesBoundsDetection(ASTContext *C, ProgramInfo &I,
+                                         bool UseHeuristics = true);
 
 // Add constraints based on heuristics to the parameters of the
 // provided function.

clang/include/clang/CConv/ConstraintsGraph.h

@@ -114,6 +114,7 @@ class DataGraph :
       N = nullptr;
     }
     this->Nodes.clear();
+    invalidateBFSCache();
   }
 
   void removeEdge(Data Src, Data Dst) {
@@ -126,12 +127,14 @@ class DataGraph :
       (*NDst)->removeEdge(*E);
       delete E;
     }
+    invalidateBFSCache();
   }
 
   void addEdge(Data L, Data R) {
     NodeType *BL = this->findOrCreateNode(L);
     NodeType *BR = this->findOrCreateNode(R);
     BL->connectTo(*BR);
+    invalidateBFSCache();
   }
 
   bool getNeighbors(Data D, std::set<Data> &DataSet, bool Succ){
@@ -161,8 +164,16 @@ class DataGraph :
     auto *N = this->findNode(NodeType(Start));
     if (N == this->end())
       return;
-    for (auto TNode : llvm::breadth_first(*N))
-      Fn(TNode->getData());
+    // Insert into BFS cache.
+    if (BFSCache.find(Start) == BFSCache.end()) {
+      std::set<Data> ReachableNodes;
+      for (auto TNode : llvm::breadth_first(*N)) {
+        ReachableNodes.insert(TNode->getData());
+      }
+      BFSCache[Start] = ReachableNodes;
+    }
+    for (auto SN : BFSCache[Start])
+      Fn(SN);
   }
 
 protected:
@@ -183,6 +194,11 @@ class DataGraph :
   template <typename G>
   friend struct llvm::GraphTraits;
   friend class GraphVizOutputGraph;
+  std::map<Data, std::set<Data>> BFSCache;
+
+  void invalidateBFSCache() {
+    BFSCache.clear();
+  }
 };
 
 // Specialize the graph for the checked and pointer type constraint graphs. This

clang/include/clang/CConv/ProgramVar.h

@@ -19,6 +19,8 @@
 #include "clang/AST/ASTContext.h"
 #include "clang/CConv/PersistentSourceLoc.h"
 
+// Unique ID for a program variable or constant literal, both of
+// which could serve as bounds
 typedef uint32_t BoundsKey;
 
 // Class representing scope of a program variable.
@@ -27,9 +29,9 @@ class ProgramVarScope {
   enum ScopeKind {
     // Function scope.
     FunctionScopeKind,
-    // Function parameter scope.
+    // Parameters of a particular function.
     FunctionParamScopeKind,
-    // Context sensitive argument scope.
+    // All arguments to a particular call
     CtxFunctionArgScopeKind,
     // Struct scope.
     StructScopeKind,
@@ -215,6 +217,17 @@ class CtxFunctionArgScope : public FunctionParamScope {
                       const PersistentSourceLoc &CtxPSL) :
     FunctionParamScope(FN, IsSt) {
     PSL = CtxPSL;
+    std::string FileName = PSL.getFileName();
+    CtxIDStr = "";
+    if (!FileName.empty()) {
+      llvm::sys::fs::UniqueID UId;
+      if (llvm::sys::fs::getUniqueID(FileName, UId)) {
+        CtxIDStr = std::to_string(UId.getDevice()) + ":" +
+                   std::to_string(UId.getFile()) + ":";
+      }
+    }
+    CtxIDStr += std::to_string(PSL.getLineNo()) + ":" +
+                std::to_string(PSL.getColSNo());
     this->Kind = CtxFunctionArgScopeKind;
   }
 
@@ -261,16 +274,16 @@ class CtxFunctionArgScope : public FunctionParamScope {
   }
 
   std::string getStr() const {
-    return "CtxFuncArg_" + FName;
+    return FName + "_Ctx_" + CtxIDStr;
   }
 
   static const CtxFunctionArgScope *
   getCtxFunctionParamScope(const FunctionParamScope *FPS,
                            const PersistentSourceLoc &PSL);
 
 private:
-  PersistentSourceLoc PSL;
-
+  PersistentSourceLoc PSL; // source code location of this function call
+  std::string CtxIDStr;
   static std::set<CtxFunctionArgScope, PVSComp> AllCtxFnArgScopes;
 };
 
@@ -354,7 +367,7 @@ class ProgramVar {
   BoundsKey K;
   std::string VarName;
   const ProgramVarScope *VScope;
-  bool IsConstant;
+  bool IsConstant; // is a literal integer, not a variable
   // TODO: All the ProgramVars may not be used. We should try to figure out
   //  a way to free unused program vars.
   static std::set<ProgramVar *> AllProgramVars;

clang/include/clang/CConv/Utils.h

@@ -124,6 +124,9 @@ bool isFunctionAllocator(std::string FuncName);
 // Is the given variable built  in type?
 bool isPointerType(clang::ValueDecl *VD);
 
+// Is this a pointer or array type?
+bool isPtrOrArrayType(const clang::QualType &QT);
+
 // Check if provided type is a var arg type?
 bool isVarArgType(const std::string &TypeName);
 

clang/lib/CConv/ABounds.cpp

@@ -63,11 +63,10 @@ std::string CountBound::mkString(AVarBoundsInfo *ABI) {
   return "count(" + PV->mkString() + ")";
 }
 
-bool CountBound::areSame(ABounds *O) {
+bool CountBound::areSame(ABounds *O, AVarBoundsInfo *ABI) {
   if (O != nullptr) {
-    if (CountBound *OT = dyn_cast<CountBound>(O)) {
-      return OT->CountVar == CountVar;
-    }
+    if (CountBound *OT = dyn_cast<CountBound>(O))
+      return ABI->areSameProgramVar(this->CountVar, OT->CountVar);
   }
   return false;
 }
@@ -86,10 +85,10 @@ std::string ByteBound::mkString(AVarBoundsInfo *ABI) {
   return "byte_count(" + PV->mkString() + ")";
 }
 
-bool ByteBound::areSame(ABounds *O) {
+bool ByteBound::areSame(ABounds *O, AVarBoundsInfo *ABI) {
   if (O != nullptr) {
     if (ByteBound *BB = dyn_cast<ByteBound>(O)) {
-      return BB->ByteVar == ByteVar;
+      return ABI->areSameProgramVar(this->ByteVar, BB->ByteVar);
     }
   }
   return false;
@@ -111,10 +110,11 @@ std::string RangeBound::mkString(AVarBoundsInfo *ABI) {
   return "bounds(" + LBVar->mkString() + ", " + UBVar->mkString() + ")";
 }
 
-bool RangeBound::areSame(ABounds *O) {
+bool RangeBound::areSame(ABounds *O, AVarBoundsInfo *ABI) {
   if (O != nullptr) {
     if (RangeBound *RB = dyn_cast<RangeBound>(O)) {
-      return RB->LB == LB && RB->UB == UB;
+      return ABI->areSameProgramVar(this->LB, RB->LB) &&
+             ABI->areSameProgramVar(this->UB, RB->UB);
     }
   }
   return false;