[ wxwindows-Bugs-1572314 ] arrarys with sizeof(T)<sizeof(base::base_type) on big-endian

[SourceForge.net]

Bugs item #1572314, was opened at 2006-10-06 20:29
Message generated for change (Tracker Item Submitted) made by Item Submitter
You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=109863&aid=1572314&group_id=9863

Please note that this message will contain a full copy of the comment thread,
including the initial issue submission, for this request,
not just the latest update.
Category: Common
Group: None
Status: Open
Resolution: None
Priority: 5
Submitted By: Oskar Linde (oskar_linde)
Assigned to: Nobody/Anonymous (nobody)
Summary: arrarys with sizeof(T)<sizeof(base::base_type) on big-endian

Initial Comment:
A bug in dynarray.h when making arrays storing types
smaller than base_type.

Example:

WX_DEFINE_ARRAY_SHORT(unsigned char, MyArr);

...

MyArr arr;

arr.Add(17);
assert(arr.Last() == 17); // Fails on big endian systems

while:

arr.SetCount(1); 
arr[0] = 17;
assert(arr.Last() == 17); // Succeeds

The problem lies in Item() and operator[] referencing
the underlying array in a different way from what Add()
and Insert() does.

As far as I can see, the bug is there both with and
without wxUSE_STL, but I havn't verified.

A snippet from dynarray.h that shows the problem:

Here, T == char, and base_type == short

  T& operator[](size_t uiIndex) const                 
               
    { return (T&)(base::operator[](uiIndex)); }       
              

  void Add(T lItem, size_t nInsert = 1)               
               
    { base::Add((base_type)lItem, nInsert); }         
               

In the first case, a (short &) is cast into a (char &)
(!)... In the second case, lItem argument is cast into
a (short) before being inserted in the underlying
short* array.

Those two behaviors give different results on big
endian systems. It is analogous to the following variants:

short *internal_data;

(1) *((char *)&internal_data[x]) = 17;
(2) internal_data[x] = (short) 17;

----------------------------------------------------------------------

You can respond by visiting: 
https://sourceforge.net/tracker/?func=detail&atid=109863&aid=1572314&group_id=9863