[cfe-dev] [RFC] Approach for C++1y N3639 (runtime-sized arrays with automatic storage duration)

[Ronan Keryell]

>>>>> On Thu, 9 May 2013 14:48:11 -0700, Richard Smith <richard at metafoo.co.uk> said:

    Ronan>     How to deal with C function definitions such as:
           
    Ronan>  void foo(int n, int m, float array[n][m]) {
    
    Ronan>     }
    
    Ronan>     where do you put the test for n and m?

    Richard> I don't think a test is necessary here, because this
    Richard> declaration doesn't allocate storage. Instead, any relevant
    Richard> test should have been performed when the argument passed as
    Richard> 'array' was created. I definitely see value in verifying
    Richard> that the 'array' argument can be used to access at least n
    Richard> * m floats, but I think that belongs in a separate check
    Richard> (maybe ASan could be taught to verify this).

I was thinking to this kind of creative programming using the example
above:
{
  float a[100][10];
  foo(-5000, -365, a);
}

But you are right, it is to be tested by some more or less static
analyzers instead, since it is not related to the allocation itself.
-- 
  Ronan KERYELL                            |\/  Phone:  +1 650 386 6482
  SILKAN Wild Systems                      |/)
  4962 El Camino Real #201                 K    Ronan.Keryell at silkan.com
  Los Altos, CA 94022                      |\   skype:keryell
  USA                                      | \  http://silkan.com