kern/159663: sockets don't work though nullfs mounts

[Robert Millan]

The following reply was made to PR kern/159663; it has been noted by GNATS.

From: Robert Millan <rmh at freebsd.org>
To: FreeBSD-gnats-submit at freebsd.org, freebsd-bugs at freebsd.org
Cc: Kostik Belousov <kostikbel at gmail.com>, Adrian Chadd <adrian at freebsd.org>, 
	Josef Karthauser <joe at freebsd.org>, freebsd-fs at freebsd.org
Subject: Re: kern/159663: sockets don't work though nullfs mounts
Date: Sun, 25 Sep 2011 17:32:27 +0200

 2011/9/24 Robert Millan <rmh at freebsd.org>:
 > I found a thread from 2007 with further discussion about this problem:
 >
 > http://lists.freebsd.org/pipermail/freebsd-fs/2007-February/002669.html
 
 Hi,
 
 I've looked at the situation in a bit more detail, for now only with
 sockets in mind (not named pipes).  My understanding is (please
 correct me if I'm wrong):
 
 - nullfs holds reference counts for each vnode, but sockets have their
 own mechanism for reference counting (so_count / soref / sorele).
 vnode reference counting doesn't protect against socket being closed,
 which would leave a stale pointer in the upper nullfs layer.
 
 - Increasing the reference count of the socket itself can't be done in
 null_nodeget() because this function is merely a getter whose call
 doesn't indicate any meaningful event.
 
 - It's not clear to me that there's any event in time where the socket
 reference can be increased.  If mounting a nullfs were that event,
 then all existing sockets would be soref'ed but we wouldn't be
 soref'ing future sockets created in the lower layer after the mount.
 This doesn't seem correct.
 
 - Possible solution: null_nodeget() semantics are replaced with
 something that actually allows vnodes in the upper layer to be created
 and destroyed.
 
 - Possible solution: upper layer has a memory structure to keep track
 of which sockets in the lower layer have been soref'ed.