Re: Fun with C

["George P." <george.priv@gmail.com>]

On Apr 17, 3:17 am, Georg Bauhaus <rm.dash-bauh...@futureapps.de>
wrote:
> On 4/16/11 11:12 PM, Ludovic Brenta wrote:
>
> > "Nasser M. Abbasi"<n...@12000.org>  writes:
> >> $ gcc -Wall  t.c
> >> $ ./a.out
> >> R = 33554416
>
> > So R = 2**25 - 16, the correct answer is -16, so the actual result is
> > the correct one except for bit 24, which is 1 when it should be zero.
>
> > Out of curiosity, is this actual result a predictable consequence of the
> > language definition, or is it undefined behavior?
>
> Why, it is the same in Ada.
>
> with Ada.Text_IO; use Ada.Text_IO;
> with Ada.Unchecked_Conversion;
>
> procedure Conv is
>
>      type UInt is mod 2**32;
>      type Int is range -2**31 .. 2**31 -1;
>
>      function To_UInt is new Ada.Unchecked_Conversion(Int, UInt);
>      N: UInt := 128;
>      I: Int := -2048;
>      R : Int;
> begin
>      R := Int(To_UInt(I) / N);
>      Put_Line ("R =" & Int'Image(R));
> end;

Not exactly, here you have at least freedom and understanding what is
being cast to what.  As I mentioned I have no problem C# giving
compile error, screeming - tell me exactly what to do here.  BTW, it
defaults result to a long int, complaining about assigning it to int.

Not to mentioned that negative integer number may be larger then
unsigned.  I am now looking at the rest the code to see where else
could be the same type of cr..p.

George