a weird problem

This is a discussion on a weird problem within the C++ Programming forums, part of the General Programming Boards category; hi guys and gals, here i've got a problem and ask for your kindly help. Code: #include <cstdlib> #include <cmath> ...

  1. #1
    C/C++Newbie Antigloss's Avatar
    Join Date
    May 2005
    Posts
    212

    a weird problem

    hi guys and gals, here i've got a problem and ask for your kindly help.

    Code:
    #include <cstdlib>
    #include <cmath>
    #include <iosfwd>
    #include <iostream>
    #include <iterator>
    #include <vector>
    #include <cstring>
    #include <iostream>
    #include <stdexcept>
    #include <string>
    #include <sstream>
    using namespace std;
    
    class t1 {
    public:
        t1() { cout << "t1: " << this << endl; }
        virtual ~t1() { cout << "~t1: " << this << endl; }
        int a, b, c;
        char d, f, g;
    };
    
    class t2 : public t1 {
    public:
        t2() { cout << "t2: " << this << endl; }
        ~t2() { cout << "~t2: " << this << endl; }
        int i1, i2, i3;
    };
    
    int main()
    {
        t1* p11 = new t1;
        delete p11;  // OK
    
        t1* p12 = new t1[2];
        delete [] p12;  // OK
    
        t2* p21 = new t2[2];
        delete [] p21;  // OK
    
        t1* p22 = new t2;
        delete p22;  // OK
    
        t1* p23 = new t2[2];
        delete [] p23;  // Core Dump, why?
    }
    why can't we delete an array of type t2 by deleting a 'pointer of type t1(p23)' that is actually pointing to 'new t2[2]', whereas we do can delete an object of type t2 by deleting a 'pointer of type t1(p22)'

  2. #2
    Cat without Hat CornedBee's Avatar
    Join Date
    Apr 2003
    Posts
    8,893
    Forget deleting. You should simply never have a pointer to base pointing to an array of derived. Pointer arithmetic is messed up for that.

    You can have a pointer to base to individual elements of an array of derived. This is not a syntactic difference (C++ doesn't have a separate notion of 'this pointer points at the start of an array'), but an important consideration nonetheless. Of course, since the pointer isn't considered to point at the array, you can't delete the array through it.

    As for delete itself, the C++ standard has the following to say in [expr.delete]p3:
    In the second alternative (delete array), if the dynamic type of the object to be deleted differs from its static type, the behavior is undefined.
    In other words, deleting an array through a pointer to base is a bad idea.
    The concrete reason for this is
    a) that there is no use case for this, since as I explained above, having such a pointer in the first place is bad, and
    b) that it simplifies implementation for compiler writers if they don't have to consider this case.
    All the buzzt!
    CornedBee

    "There is not now, nor has there ever been, nor will there ever be, any programming language in which it is the least bit difficult to write bad code."
    - Flon's Law

  3. #3
    Registered User
    Join Date
    Jun 2005
    Posts
    6,205
    Quote Originally Posted by CornedBee View Post
    b) that it simplifies implementation for compiler writers if they don't have to consider this case.
    In Antigloss's code, the compiler could detect the difference.

    The problems come in if the creation and deletion happen in functions within separate compilation units - the compiler would have no way of detecting the mismatch (short of changing compilation model to something other than separate compilation of source files). The whole point of the separate compilation model is that behaviour of a program won't vary if a function is moved between source files.
    Right 98% of the time, and don't care about the other 3%.

  4. #4
    Cat without Hat CornedBee's Avatar
    Join Date
    Apr 2003
    Posts
    8,893
    In Antigloss's code, the compiler could detect the difference.
    Difference between what?
    A static analyzer could easily detect the UB here, but it would require actual flow analysis, which gets very hard very quickly in more complex examples.
    All the buzzt!
    CornedBee

    "There is not now, nor has there ever been, nor will there ever be, any programming language in which it is the least bit difficult to write bad code."
    - Flon's Law

  5. #5
    Registered User
    Join Date
    Jun 2005
    Posts
    6,205
    Quote Originally Posted by CornedBee View Post
    Difference between what?
    Difference between defined and undefined behaviours, or (if you prefer) the difference between the cases within the example.
    Right 98% of the time, and don't care about the other 3%.

Popular pages Recent additions subscribe to a feed

Similar Threads

  1. Weird Problem With Pointer
    By DarkDot in forum C++ Programming
    Replies: 3
    Last Post: 05-07-2007, 07:50 PM
  2. Weird problem on '02 3.4L V6 auto
    By VirtualAce in forum A Brief History of Cprogramming.com
    Replies: 8
    Last Post: 01-11-2006, 11:05 PM
  3. Really Weird itoa Problem
    By Grantyt3 in forum C++ Programming
    Replies: 8
    Last Post: 12-19-2005, 11:44 PM
  4. Replies: 6
    Last Post: 05-12-2005, 03:39 AM
  5. Weird class problem!
    By aker_y3k in forum C++ Programming
    Replies: 2
    Last Post: 09-25-2002, 06:12 AM

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21