Re: A problem with comctl32

[Dimitrie O. Paun]

On November 4, 2002 05:49 am, Rolf Kalbermatter wrote:
> > const HIMAGELIST himl = pimldp->himl;
>                                  ^^^^^^^^^^^^
>
> According to the same lines pimldp could also be NULL so this might as
> well cause a NULL pointer dereference and should be moved to after the
> check for "if (!pimldp || !(himl = pimldp->himl)) return FALSE;

Right, this needs to be moved below the if as well. The reasons they
are this way it's that we used to have internal helper functions that
were guaranteed these are not NULL. It's a copy & paste bug.
> Of course the question remains why would you call the function at all
> with a NULL pointer.

We can not control that, the function is directly callable by the app.