[Bug c/16403] New: Floating-point assignment to int is inconsistent

["gcc-bugzilla at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>]

	The program shown below assigns different values to i and j.
	The difference is that the result of '(int)5 * (double)1.2' is
	stored in an intermediate double-precision variable 'g' before it is
	assigned to j. This causes different rounding to take place. So the
	same double-precision calculation is rounded, or truncated, in
	two different ways when assigned to an int.

Environment:
System: Linux goya 2.4.18-3custom #1 SMP Wed Oct 16 18:51:01 IST 2002 i686 unknown
Architecture: i686

	
host: i686-pc-linux-gnu
build: i686-pc-linux-gnu
target: i686-pc-linux-gnu
configured with: ../gcc-3.4.1/configure --prefix=/usr/local/gnu/gcc/3.4.1-fixed --enable-languages=c,c++,f77 --with-gcc-version-trigger=/vobs/tools/xde/gcc-3.4.1/gcc/version.c --no-create --no-recursion

How-To-Repeat:
	Compile the following program with no options ('gcc f.c') and
	run a.out.  The output will be 'i = 5, j = 6'.
/***/
extern int printf (char*, ...);

int main (void)
{
	int j, i = 5;
	double g = i * 1.2;
	j = g;
	i *= 1.2;
	printf("i = %d, j = %d\n",i,j);
	return 0;
}
/***/

-- 
           Summary: Floating-point assignment to int is inconsistent
           Product: gcc
           Version: 3.4.1
            Status: UNCONFIRMED
          Severity: critical
          Priority: P2
         Component: c
        AssignedTo: unassigned at gcc dot gnu dot org
        ReportedBy: yuvalk at mainsoft dot com
                CC: gcc-bugs at gcc dot gnu dot org
 GCC build triplet: i686-pc-linux-gnu
  GCC host triplet: i686-pc-linux-gnu
GCC target triplet: i686-pc-linux-gnu


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=16403