[Patch] Moving FD passing into the header

[Lennart Poettering]

On Wed, 23.06.10 08:00, Colin Walters (walters at verbum.org) wrote:

> 
> On Tue, Jun 22, 2010 at 8:49 PM, Lennart Poettering <mzqohf at 0pointer.de> wrote:
> >
> > And again, and I said it a gazillion of times: nothing stops you from
> > serializing/deserializing the unix fd stuff not as unix fd, but as index
> > into the fd table. Then, things would behave almost identical to what
> > your patches offer, except that methods that carry unix fds are properly
> > typechecked and have complete introspection data. You could support that
> > just fine in GVariant with minimal work.
> 
> This makes me think of a third option here - make the 'h' type be
> DBUS_TYPE_OOB (out of band).  It's an integer index into some
> (possibly platform specific) data to be found in the header.  In your
> service documentation you'd mention that the 'h' in a particular
> method is a file descriptor, say.

Such an approach would suggest itself if we had more than one type of
OOB data to pass around. But I don't really see that. Looking through
the man pages on Linux only three kinds of data can be sent out-of-band
really so far:

SCM_CREDENTIALS (which we already cover a different way)
SO_TIMESTAMP (which is kinda useless here I think)
SCM_RIGHTS

And because things are the way they are I'd just support what we know
about and matters, and ignore the rest, that we don't know about.

Lennart

-- 
Lennart Poettering                        Red Hat, Inc.
lennart [at] poettering [dot] net
http://0pointer.net/lennart/           GnuPG 0x1A015CC4