GP-3014 - Add deduping to DynamicHash gatherFirstLevelVars

Ghidra/Features/Decompiler/src/decompile/cpp/dynamic.cc

@@ -364,7 +364,7 @@ void DynamicHash::pieceTogetherHash(const Varnode *root,uint4 method)
   hash <<= 4;
   hash |= method;		// 4-bits
   hash <<= 7;
-  hash |= (uint8)op->code();	// 7-bits
+  hash |= (uint8)transtable[op->code()];	// 7-bits
   hash <<= 5;
   hash |= (uint8)(slot & 0x1f);	// 5-bits
   
@@ -607,6 +607,25 @@ PcodeOp *DynamicHash::findOp(const Funcdata *fd,const Address &addr,uint8 h)
   return oplist2[pos];
 }
 
+/// Otherwise preserve the order of the list.
+/// \param varlist is the given list of Varnodes to check
+void DynamicHash::dedupVarnodes(vector<Varnode *> &varlist)
+
+{
+  if (varlist.size() < 2) return;
+  vector<Varnode *> resList;
+  for(int4 i=0;i<varlist.size();++i) {
+    Varnode *vn = varlist[i];
+    if (!vn->isMark()) {
+      vn->setMark();
+      resList.push_back(vn);
+    }
+  }
+  for(int4 i=0;i<resList.size();++i)
+    resList[i]->clearMark();
+  varlist.swap(resList);
+}
+
 /// \brief Get the Varnodes immediately attached to PcodeOps at the given address
 ///
 /// Varnodes can be either inputs or outputs to the PcodeOps. The op-code, slot, and
@@ -619,7 +638,7 @@ PcodeOp *DynamicHash::findOp(const Funcdata *fd,const Address &addr,uint8 h)
 void DynamicHash::gatherFirstLevelVars(vector<Varnode *> &varlist,const Funcdata *fd,const Address &addr,uint8 h)
 
 {
-  OpCode opc = getOpCodeFromHash(h);
+  uint4 opcVal = getOpCodeFromHash(h);
   int4 slot = getSlotFromHash(h);
   bool isnotattached = getIsNotAttached(h);
   PcodeOpTree::const_iterator iter = fd->beginOp(addr);
@@ -629,7 +648,7 @@ void DynamicHash::gatherFirstLevelVars(vector<Varnode *> &varlist,const Funcdata
     PcodeOp *op = (*iter).second;
     ++iter;
     if (op->isDead()) continue;
-    if (op->code() != opc) continue;
+    if (transtable[op->code()] != opcVal) continue;
     if (slot <0) {
       Varnode *vn = op->getOut();
       if (vn != (Varnode *)0) {
@@ -655,6 +674,7 @@ void DynamicHash::gatherFirstLevelVars(vector<Varnode *> &varlist,const Funcdata
       varlist.push_back(vn);
     }
   }
+  dedupVarnodes(varlist);
 }
 
 /// \brief Place all PcodeOps at the given address in the provided container
@@ -697,11 +717,11 @@ uint4 DynamicHash::getMethodFromHash(uint8 h)
 
 /// The hash encodes the op-code of the p-code op attached to the root Varnode
 /// \param h is the hash value
-/// \return the op-code
-OpCode DynamicHash::getOpCodeFromHash(uint8 h)
+/// \return the op-code as an integer
+uint4 DynamicHash::getOpCodeFromHash(uint8 h)
 
 {
-  return (OpCode)((h>>37)&0x7f);
+  return (h>>37)&0x7f;
 }
 
 /// The hash encodes the position of the root Varnode within the list of hash collisions

Ghidra/Features/Decompiler/src/decompile/cpp/dynamic.hh

@@ -77,6 +77,7 @@ class DynamicHash {
   void gatherUnmarkedOp(void);		///< Mark any new PcodeOps in the sub-graph
   void pieceTogetherHash(const Varnode *root,uint4 method);	///< Clean-up and piece together formal hash value
   static void moveOffSkip(const PcodeOp *&op,int4 &slot);	///< Convert given PcodeOp to a non-skip op by following data-flow
+  static void dedupVarnodes(vector<Varnode *> &varlist);	///< Remove any duplicate Varnodes in given list
 public:
   void clear(void);			///< Called for each additional hash (after the first)
   void calcHash(const Varnode *root,uint4 method);	///< Calculate the hash for given Varnode and method
@@ -92,7 +93,7 @@ public:
   static void gatherOpsAtAddress(vector<PcodeOp *> &opList,const Funcdata *fd,const Address &addr);
   static int4 getSlotFromHash(uint8 h);			///< Retrieve the encoded slot from a hash
   static uint4 getMethodFromHash(uint8 h);		///< Retrieve the encoded method from a hash
-  static OpCode getOpCodeFromHash(uint8 h);		///< Retrieve the encoded op-code from a hash
+  static uint4 getOpCodeFromHash(uint8 h);		///< Retrieve the encoded op-code from a hash
   static uint4 getPositionFromHash(uint8 h);		///< Retrieve the encoded position from a hash
   static uint4 getTotalFromHash(uint8 h);		///< Retrieve the encoded collision total from a hash
   static bool getIsNotAttached(uint8 h);		///< Retrieve the attachment boolean from a hash

Ghidra/Framework/SoftwareModeling/src/main/java/ghidra/program/model/pcode/DynamicHash.java

@@ -415,7 +415,7 @@ public class DynamicHash {
 		hash <<= 4;
 		hash |= method;			// 4-bits
 		hash <<= 7;
-		hash |= op.getOpcode();	// 7-bits
+		hash |= transtable[op.getOpcode()];	// 7-bits
 		hash <<= 5;
 		hash |= slot & 0x1f;	// 5-bits
 
@@ -705,6 +705,21 @@ public class DynamicHash {
 		}
 	}
 
+	private static void dedupVarnodes(ArrayList<Varnode> varlist) {
+		if (varlist.size() < 2) {
+			return;
+		}
+		ArrayList<Varnode> resList = new ArrayList<>();
+		HashSet<Varnode> hashSet = new HashSet<>();
+		for (Varnode vn : varlist) {
+			if (hashSet.add(vn)) {
+				resList.add(vn);
+			}
+		}
+		varlist.clear();
+		varlist.addAll(resList);
+	}
+
 	public static void gatherFirstLevelVars(ArrayList<Varnode> varlist, PcodeSyntaxTree fd,
 			Address addr, long h) {
 		int opc = getOpCodeFromHash(h);
@@ -714,7 +729,7 @@ public class DynamicHash {
 
 		while (iter.hasNext()) {
 			PcodeOp op = iter.next();
-			if (op.getOpcode() != opc) {
+			if (transtable[op.getOpcode()] != opc) {
 				continue;
 			}
 			if (slot < 0) {
@@ -745,6 +760,7 @@ public class DynamicHash {
 				varlist.add(vn);
 			}
 		}
+		dedupVarnodes(varlist);
 	}
 
 	public static int getSlotFromHash(long h) {