586
Subj : Re: Math error ??? doen't make sense!
To   : borland.public.cpp.borlandcpp
From : waking
Date : Sat Jul 03 2004 02:07 am

On Fri, 2 Jul 2004 20:31:26 -0400, "Stan DeGroff" <sdegroff@bright.net> wrote:

>There ain't no overflow. as shown in the z result above.

As I already explained, the overflow and truncation occur as
the result of the product exceeding the capacity of a 16-bit int.
Since both x and y are 16-bit ints, the *intermediate* result will
be stored in a 16-bit int. The typecast I showed forces promotion
to 32-bit long for the calculation. The multiplication occurs *before*
the assignment to a 32-bit long, hence its size is not relevant to
the capacity used for the product.

>Multiply in Borland product must have both inputs to a multiply
>and the output all the same size variables.

It depends on the values being stored. Also, promoting one of 
either the multiplier or multiplicand results in automatic promotion
of the other and of the intermediate temporary variable used for
the result. This is the way it is in the C and C++ languages. It is not
unique to Borland. See the C FAQ at 
http://www.eskimo.com/~scs/C-faq/faq.html
especially 3.14: 

"Why doesn't the code 
int a = 1000, b = 1000;
long int c = a * b;
work? "

http://www.eskimo.com/~scs/C-faq/q3.14.html


--
Wayne A. King 
(waking@idirect.com, Wayne_A_King@compuserve.com)

.

0