Uploaded image for project: 'C++ Standard Library'
  1. C++ Standard Library
  2. STDCXX-618

purify reports free memory read in 21.string.append test

    XMLWordPrintableJSON

Details

    • Bug
    • Status: Open
    • Minor
    • Resolution: Unresolved
    • 4.2.0
    • 4.2.2
    • 21. Strings
    • None
    • Incorrect Behavior

    Description

      While investigating purify errors in 21.string.append.cpp, I ran into bunch FMR failures [this same failure occurs in other tests also]. They point to a legitimate bug in basic_string::replace() and possibly others like vector::insert(). I'm not really sure what the best approach for fixing this is because the issue itself is a little weird. There is also the issue of exception safety which makes it a little more interesting.

      #include <cassert>
      #include <string>
      #include <vector>
      
      template <class _IteratorT>
      struct dumb_input_iterator
      {
          typedef typename _IteratorT::value_type  value_type;
          typedef typename _IteratorT::difference_type difference_type;
          typedef typename _IteratorT::pointer pointer;
          typedef typename _IteratorT::reference reference;
          typedef std::input_iterator_tag iterator_category;
      
          dumb_input_iterator (const _IteratorT &__rhs)
              : _C_iter (__rhs) {}
      
          reference operator* () const {
              return _C_iter.operator*();
          }
      
          pointer operator-> () const {
              return _C_iter.operator->();
          }
      
          dumb_input_iterator& operator++ () {
              return _C_iter.operator++(), *this;
          }
      
          dumb_input_iterator operator++ (int) {
              dumb_input_iterator __tmp(*this);
              _C_iter.operator++();
              return __tmp;
          }
      
          bool
          operator== (const dumb_input_iterator &__rhs) const {
              return _C_iter == __rhs._C_iter;
          }
      
          bool
          operator!= (const dumb_input_iterator &__rhs) const {
              return !(_C_iter == __rhs._C_iter);
          }
      
      private:
      
          _IteratorT _C_iter;
      };
      
      int main()
      {
      
          // i'm not exactly sure why you'd do this, but the following code
          // creates input iterators on top of bidirectional iterators. when
          // passing an input iterator to these [and possibly other] methods
          // it is assumed that the iterator does not refer to elements inside
          // the container, which is wrong.
      
          size_t n;
      
          typedef std::string string;
          string s(16, 'a');
      
          // loop in an attempt to ensure internal buffer reallocation
          for (n = 0; n < 8; ++n)
          {
              size_t i = s.size ();
      
              const dumb_input_iterator<string::iterator> db (s.begin ());
              const dumb_input_iterator<string::iterator> de (s.end   ());
      
              s.append (db, de);
              // s.append (s.begin (), s.end ()); // works as expected
      
              for (; i < s.size (); ++i)
                  assert (s[i] == 'a');
          }
      
          typedef std::vector<int> vector;
          vector v(16, 1);
      
          // loop in an attempt to ensure internal buffer reallocation
          for (n = 0; n < 8; ++n)
          {
              size_t i = v.size ();
      
              const dumb_input_iterator<vector::iterator> db (v.begin ());
              const dumb_input_iterator<vector::iterator> de (v.end   ());
      
              v.insert (v.end (), db, de);
              // v.insert (v.end (), v.begin (), v.end ()); // works as expected
      
              for (; i < v.size (); ++i)
                  assert (v[i] == 1);
          }
      
          return 0;
      }
      

      Attachments

        1. 21.string.append.log
          55 kB
          Travis Vitek

        Activity

          People

            Unassigned Unassigned
            vitek Travis Vitek
            Votes:
            0 Vote for this issue
            Watchers:
            0 Start watching this issue

            Dates

              Created:
              Updated:

              Time Tracking

                Estimated:
                Original Estimate - 4h
                4h
                Remaining:
                Remaining Estimate - 4h
                4h
                Logged:
                Time Spent - Not Specified
                Not Specified