Avoid undefined behavior using CFlatData in CScript serialization

`&vch[vch.size()]` and even `&vch[0]` on vectors can cause assertion errors with VC in debug mode. This is the problem mentioned in #4239. The deeper problem with this is that we rely on undefined behavior. - Add `begin_ptr` and `end_ptr` functions that get the beginning and end pointer of vector in a reliable way that copes with empty vectors and doesn't reference outside the vector (see https://stackoverflow.com/questions/1339470/how-to-get-the-address-of-the-stdvector-buffer-start-most-elegantly/1339767#1339767). - Add a convenience constructor to CFlatData that wraps a vector. I added `begin_ptr` and `end_ptr` as separate functions as I imagine they will be useful in more places.
TrezarCoin · Jun 23, 2014 · fa126ef · fa126ef
1 parent 52d4abf
commit fa126ef
Show file tree

Hide file tree

Showing 2 changed files with 38 additions and 4 deletions.
diff --git a/src/script.h b/src/script.h
@@ -770,12 +770,12 @@ class CScriptCompressor
     void Serialize(Stream &s, int nType, int nVersion) const {
         std::vector<unsigned char> compr;
         if (Compress(compr)) {
-            s << CFlatData(&compr[0], &compr[compr.size()]);
+            s << CFlatData(compr);
             return;
         }
         unsigned int nSize = script.size() + nSpecialScripts;
         s << VARINT(nSize);
-        s << CFlatData(&script[0], &script[script.size()]);
+        s << CFlatData(script);
     }
 
     template<typename Stream>
@@ -784,13 +784,13 @@ class CScriptCompressor
         s >> VARINT(nSize);
         if (nSize < nSpecialScripts) {
             std::vector<unsigned char> vch(GetSpecialSize(nSize), 0x00);
-            s >> REF(CFlatData(&vch[0], &vch[vch.size()]));
+            s >> REF(CFlatData(vch));
             Decompress(nSize, vch);
             return;
         }
         nSize -= nSpecialScripts;
         script.resize(nSize);
-        s >> REF(CFlatData(&script[0], &script[script.size()]));
+        s >> REF(CFlatData(script));
     }
 };
 

diff --git a/src/serialize.h b/src/serialize.h
@@ -37,6 +37,34 @@ inline T& REF(const T& val)
     return const_cast<T&>(val);
 }
 
+/** Get begin pointer of vector (non-const version).
+ * @note These functions avoid the undefined case of indexing into an empty
+ * vector, as well as that of indexing after the end of the vector.
+ */
+template <class T, class TAl>
+inline T* begin_ptr(std::vector<T,TAl>& v)
+{
+    return v.empty() ? NULL : &v[0];
+}
+/** Get begin pointer of vector (const version) */
+template <class T, class TAl>
+inline const T* begin_ptr(const std::vector<T,TAl>& v)
+{
+    return v.empty() ? NULL : &v[0];
+}
+/** Get end pointer of vector (non-const version) */
+template <class T, class TAl>
+inline T* end_ptr(std::vector<T,TAl>& v)
+{
+    return v.empty() ? NULL : (&v[0] + v.size());
+}
+/** Get end pointer of vector (const version) */
+template <class T, class TAl>
+inline const T* end_ptr(const std::vector<T,TAl>& v)
+{
+    return v.empty() ? NULL : (&v[0] + v.size());
+}
+
 /////////////////////////////////////////////////////////////////
 //
 // Templates for serializing to anything that looks like a stream,
@@ -318,6 +346,12 @@ class CFlatData
     char* pend;
 public:
     CFlatData(void* pbeginIn, void* pendIn) : pbegin((char*)pbeginIn), pend((char*)pendIn) { }
+    template <class T, class TAl>
+    explicit CFlatData(std::vector<T,TAl> &v)
+    {
+        pbegin = (char*)begin_ptr(v);
+        pend = (char*)end_ptr(v);
+    }
     char* begin() { return pbegin; }
     const char* begin() const { return pbegin; }
     char* end() { return pend; }