fcntl bug

Mon Aug 24 15:17:00 GMT 2009

Christopher Faylor <cgf-use-the-mailinglist-please <at> cygwin.com> writes:
> >But POSIX does (and Linux at least obeys this part of POSIX, whether or
> >not its man page says so):
>> I checked in a fix for this a while ago. It's in the latest snapshot.

While we're at it, fcntl and dup2 both have another minor bug. POSIX states 
that fcntl(0,F_DUPFD,10000000) should fail with EINVAL (not EBADF) and the 
similar dup2(0,10000000) should fail with EBADF (not EMFILE). Gotta love the 
inconsistency of historical interfaces. (This assumes, of course, that 
OPEN_MAX is less than 10000000). dup2 should never fail with EMFILE, and fcntl
(F_DUPFD) should fail with EMFILE only if the target is within the range of the 
current OPEN_MAX (aka sysconf(_SC_OPEN_MAX)) but all fds from that point to the 
end of the table are currently open.
#include <unistd.h>
#include <fcntl.h>
#include <string.h>
#include <errno.h>
#include <stdio.h>
int main()
{
 int i = dup2 (0, 10000000);
 printf ("%d %d %s\n", i, errno, strerror (errno));
 i = fcntl (0, F_DUPFD, 10000000);
 printf ("%d %d %s\n", i, errno, strerror (errno));
 return 0;
}
Solaris:
% ./foo
-1 9 Bad file number
-1 22 Invalid argument
Cygwin 1.7:
$ ./foo
-1 24 Too many open files
-1 9 Bad file descriptor
-- 
Eric Blake
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple