On Sun, Aug 30, 2009 at 06:32:59PM -0600, Eric Blake wrote:
>According to Eric Blake on 8/24/2009 9:15 AM:
>>While we're at it, fcntl and dup2 both have another minor bug. POSIX
>>states that fcntl(0,F_DUPFD,1000) should fail with EINVAL (not
>>EBADF) and the similar dup2(0,1000) should fa
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to Eric Blake on 8/24/2009 9:15 AM:
> While we're at it, fcntl and dup2 both have another minor bug. POSIX states
> that fcntl(0,F_DUPFD,1000) should fail with EINVAL (not EBADF) and the
> similar dup2(0,1000) should fail with EBAD
Christopher Faylor 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
On Fri, Aug 21, 2009 at 04:04:12PM -0600, Eric Blake wrote:
>According to cygwin at sipxx.com on 8/21/2009 3:56 PM:
>>It does reject the argument if the upper bound is reached. On the
>>lower end, the linux man page does not specify that negative numbers
>>are not allowed as arguments.
>
>But POSI
I see, I would concur based on the definition of the error code you
quoted. I consulted only my memory just now, since I ran the same
general test last week.
Eric Blake wrote:
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to cygwin at sipxx.com on 8/21/2009 3:56 PM:
It does reje
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
According to cygwin at sipxx.com on 8/21/2009 3:56 PM:
> It does reject the argument if the upper bound is reached. On the lower
> end, the linux man page does not specify that negative numbers are not
> allowed as arguments.
But POSIX does (and Linux
It does reject the argument if the upper bound is reached. On the lower
end, the linux man page does not specify that negative numbers are not
allowed as arguments. Note, that fcntl is not supposed to be the same as
dup2(), i.e., it does NOT duplicate the given fd into the new one
specified, bu
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1
fcntl is supposed to reject attempts to duplicate to an out-of-range fd.
On Linux, this correctly fails with EINVAL.
#include
#include
#include
#include
int main()
{
int i = fcntl (0, F_DUPFD, -1);
printf ("%d %d %s\n", i, errno, strerror (err
Hello,
Cygwin doesn't support advisory file locking. It implements file locking
via the Win32 api LockFile*. Find out more here:
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/fileio/base/lockfile.asp
And more specifically this part:
[...]
Locking a portion of a file for shared
Hello,
I try to comiple an run the distcc package - but I get errors while running.
I figured out, tha the problems are the fcntl() calls - perhaps
ther is a bug in cygwin?
The following code runs differently in cygwin and linux.
If I want ro lock the whole file twice, but still the same process,
10 matches
Mail list logo
