Q46735: Loop Optimization May Cause Improper Type Cast of int

Article: Q46735
Product(s): See article
Version(s): 5.10   | 5.10
Operating System(s): MS-DOS | OS/2
Keyword(s): ENDUSER | | mspl13_c
Last Modified: 25-JUL-1989

In certain situations, compiling with loop optimization may cause a
type cast of an int to a long or unsigned long to produce an incorrect
value if the cast precedes a loop. When an int is type cast to a long,
the correct conversion sign extends. Thus, a negative integer remains
negative. On the other hand, when casting an int to an unsigned long,
the correct conversion is to sign extend to long, THEN convert the
long to unsigned long. This means a negative integer will be cast to a
really huge positive number when cast to an unsigned long. This
expected behavior is documented on page 141 of the "Microsoft C 5.1
Optimizing Compiler Language Reference." Loop optimization may cause
errors in both of these type casts. Disabling loop optimization or
rearranging the code corrects the problem.

The following program, when compiled with the /Oal or /Ox option,
displays this erroneous type casting. In the third assignment
statement, -31536 is cast to a long and assigned to lNum. The result
should still be -31536, but lNum ends up being 34000. The fourth
assignment statement casts this same number to an unsigned long and
assigns it to ulNum. The expected result, as described above, is the
huge positive number 4294935760, but the value assigned to ulNum is
again 34000.

Program Sample
--------------

#include <stdio.h>

void main(void)
{
   int iNum;
   unsigned uNum;
   long lNum;
   unsigned long ulNum;

        lNum = 34000;
        printf ("lNum = %ld\n", lNum);
        iNum = (int) lNum;
        printf ("iNum = (int) lNum = %d \n", iNum);
        lNum = (long) iNum;
        printf ("lNum = (long) iNum = %ld \n", lNum);
        ulNum = (unsigned long) iNum;
        printf ("ulNum = (unsigned long) iNum = %lu \n", ulNum);

        for (lNum = 0L; lNum < 17 ; lNum += 1)
               ;
        printf ("%s", "done");
}

The following is the program output:

lNum = 34000                                <== Correct
iNum = (int) lNum = -31536                  <== Correct
lNum = (long) iNum = 34000                  <== WRONG! (should be -31536)
ulNum = (unsigned long) iNum = 34000        <== WRONG! (should be 4294935760)
done

THE INFORMATION PROVIDED IN THE MICROSOFT KNOWLEDGE BASE IS PROVIDED "AS IS" WITHOUT WARRANTY OF ANY KIND. MICROSOFT DISCLAIMS ALL WARRANTIES, EITHER EXPRESS OR IMPLIED, INCLUDING THE WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT SHALL MICROSOFT CORPORATION OR ITS SUPPLIERS BE LIABLE FOR ANY DAMAGES WHATSOEVER INCLUDING DIRECT, INDIRECT, INCIDENTAL, CONSEQUENTIAL, LOSS OF BUSINESS PROFITS OR SPECIAL DAMAGES, EVEN IF MICROSOFT CORPORATION OR ITS SUPPLIERS HAVE BEEN ADVISED OF THE POSSIBILITY OF SUCH DAMAGES. SOME STATES DO NOT ALLOW THE EXCLUSION OR LIMITATION OF LIABILITY FOR CONSEQUENTIAL OR INCIDENTAL DAMAGES SO THE FOREGOING LIMITATION MAY NOT APPLY.

Copyright Microsoft Corporation 1986-2002.