On Mon, 08 Sep 2003 02:31, Herbert Poetzl wrote; > The main reason for removing it was that this > capability had #30, and there are currently 32 > bits available for capabilities, so I didn't > want to use them up lightly ...
I'm sure an extra 4 bytes per process table entry won't hurt :-) In fact on 64 bit architectures it probably won't even take that. Sounds good, if a lot of work, but it needs to be done. I wonder whether the `magicness' of security context 1 shouldn't be controlled by a capability instead? It would mean that the root context has full access to see all processes, but that was the case anyway - it could just use chcontext. -- Sam Vilain, [EMAIL PROTECTED] If we make peaceful revolution impossible, we make violent revolution inevitiable. JOHN F KENNEDY
