Re: [Gm2] More about LONGINT fixes

[Gaius Mulley]

Izo <address@hidden> writes:

> Gaius Mulley wrote:
> 
> >Hi,
> >
> >I believe I've fixed LONGINT constant for an opteron based system..
> >It is slightly more tricky for 32 bit systems due to the data type
> >correspondence between C and Modula-2 which is outlined below:
> >
> >GNU Modula-2              GNU C
> >======================================
> >...
> >CARDINAL                  unsigned int
> >LONGCARD                  long unsigned int
> >SHORTCARD                 short unsigned int
> >BOOLEAN                   int
> >.....
> >
> 
> I am not quite sure that this is true. I have to check, but I think
> that the C's unsigned int only takes the positive part of the signed
> int, which means that the GNU M2's CARDINAL would have the range
> 0..7FFFFFFF instead of 0..FFFFFFFF. The similar would then apply to
> the LONGCARD and SHORTCARD.

Practically in GCC I believe it does use the full range for an
unsigned int. It does seem as if C supports CARDINAL via this
construct. The reference manual p196 in "Programming in C (ANSI
Ed)", 1990 says:

"Unsigned integers, declared using the keyword unsigned, obey the laws
of arithmetic modulo 2^n where n is the number of bits in a
representation, and thus arithmetic on unsigned quantities can never
overflow. The set of non-negative values that can be stored in a
signed object is a subset of the values that can be stored in the
corresponding unsigned object, and the representation for the
overlapping values is the same."

> And for the BOOLEAN. It is true, indeed that the size of BOOLEAN is
> not defined by standard but left to be defined by design. Yet, my
> opinion would be that for the sake of the language intercompatibility
> would be the best to define the BOOLEAN as char. At least through
> special compiler option ?

yes this is probably the right thing to do.

regards,
Gaius