standards/60772: _Bool and bool should be unsigned
Jason Evans
jasone at canonware.com
Wed Dec 31 16:50:16 PST 2003
>Number: 60772
>Category: standards
>Synopsis: _Bool and bool should be unsigned
>Confidential: no
>Severity: serious
>Priority: high
>Responsible: freebsd-standards
>State: open
>Quarter:
>Keywords:
>Date-Required:
>Class: sw-bug
>Submitter-Id: current-users
>Arrival-Date: Wed Dec 31 16:50:14 PST 2003
>Closed-Date:
>Last-Modified:
>Originator: Jason Evans
>Release: FreeBSD 4.9-RC i386
>Organization:
>Environment:
System: FreeBSD canonware.com 4.9-RC FreeBSD 4.9-RC #12: Fri Oct 17 22:15:36 PDT 2003 jasone at canonware.com:/usr/src/sys/compile/CANONWARE i386
>Description:
_Bool and bool must be unsigned, so that structure bit fields work as
expected.
>How-To-Repeat:
#include <stdio.h>
#include <stdbool.h>
/* _Bool and bool must be unsigned. Otherwise, stucture bit fields do not work
* correctly, due to how sign extension works. */
struct foo {
bool white:1; /* Note :1 (bit field). */
};
struct bar {
bool white:2;
};
int
main(void)
{
struct foo foo;
struct bar bar;
bool white;
foo.white = false;
fprintf(stderr, "%u (expect 0)\n", foo.white);
fprintf(stderr, "%u (expect 1)\n", !foo.white);
foo.white = !foo.white; /* Breaks. */
fprintf(stderr, "%u (expect 1)\n", foo.white);
fprintf(stderr, "%u (expect 0)\n", !foo.white);
bar.white = false;
fprintf(stderr, "%u (expect 0)\n", bar.white);
fprintf(stderr, "%u (expect 1)\n", !bar.white);
bar.white = !bar.white;
fprintf(stderr, "%u (expect 1)\n", bar.white);
fprintf(stderr, "%u (expect 0)\n", !bar.white);
white = false;
fprintf(stderr, "%u (expect 0)\n", white);
fprintf(stderr, "%u (expect 1)\n", !white);
white = !white;
fprintf(stderr, "%u (expect 1)\n", white);
fprintf(stderr, "%u (expect 0)\n", !white);
return 0;
}
>Fix:
Using bit fields with more than one bit works around the problem, but is
not a satisfactory workaround. This bug can cause obscure problems, since
something like:
if (foo.white == true) ...;
breaks.
>Release-Note:
>Audit-Trail:
>Unformatted:
More information about the freebsd-standards
mailing list