On Fri, Oct 28, 2011 at 05:49:59AM -0600, Eric Blake wrote:
> On 10/28/2011 02:28 AM, Daniel P. Berrange wrote:
> >>>+ if ((fd = dup(msg->fds[slot]))< 0) {
> >>>+ virReportSystemError(errno,
> >>>+ _("Unable to duplicate FD %d"),
> >>>+ msg->fds[slot]);
> >>>+ return -1;
> >>>+ }
> >>
> >>Do we want to be using gnulib's dup_cloexec ("cloexec.h") or
> >>fcntl(fd, F_DUPFD_CLOEXEC, 0) here, so that our dup doesn't leak
> >>into a third-party child if we are linked into some larger
> >>multithreaded app?
> >
> >dup3() is the new glibc API for this ? Does gnulib support
> >that yet ?
>
> Gnulib supports dup3(). But dup3() is not the same as dup(); it is
> the same as dup2(). That is:
>
> dup(fd) maps to fcntl(fd, F_DUPFD_CLOEXEC,0) == dup_cloexec(fd)
> dup2(fd1, fd2) maps to dup3(fd1, fd2, O_CLOEXEC)
Ah ok, I see what you mean.
Daniel
--
|: http://berrange.com -o- http://www.flickr.com/photos/dberrange/ :|
|: http://libvirt.org -o- http://virt-manager.org :|
|: http://autobuild.org -o- http://search.cpan.org/~danberr/ :|
|: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :|
--
libvir-list mailing list
libvir-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/libvir-list