implement in-kernel keys & keyring management [try #6]

Hi Linus, Andrew,

> I've modified my patch to avoid a locking error that Al Viro pointed out. If
> you have two keyrings A and B, if one process tries to link A to B, whilst
> in parallel another process tries to link B to A, it could end up creating a
> cycle in the graph. I've added code to serialise link calls with respect to
> one another to obviate that problem.

I've fixed another looking bug that that introduced.

Following a reply from Chris Wright, I've also sorted out session keyring
management to what is probably a more sane approach. Processes now start off
without session keyrings being assigned. A process is automatically subscribed
to the user's default session if it doesn't have a session keyring when it
tries to access it. Processes inherit their parents session keyring or lack
thereof at fork time. Processes can still manually join a session.

The patch can be found at:

	http://people.redhat.com/~dhowells/keys/keys-268rc2-6.dif...

David.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/