> No. PID 1 is still busy after you kill init (ignoring the fact that you
> just can't kill -9 1, the kernel catches that) since init becomes a zombie
> (nobody waits for it).
>
couldn't you do (after having crashed init)
kill_everybody_else(TM); /* so they don't come and grab our free
slot */
go_to_last_pid();
waitpid(1,NULL,WNOHANG); /* get zombie init's status to get rid of
it*/
fork()....
'cos you're root so you should be able to wait for pid 1? Then you have
pid1 and root so you can change the securelevel as you wish.
Or am I confused?
Other low pids may not be security sensitive resources, but as long as
PID 1 has special powers, it is.. Which leads to the question why dosen't
the kernel look after PID 1 more carefully?
<http://www.tardis.ed.ac.uk/~mikedlr/biography.html>
Scottish Climbing Archive: <http://www.tardis.ed.ac.uk/~mikedlr/climbing/>
Linux/Unix clone@ftp://src.doc.ic.ac.uk/packages/linux/sunsite.unc-mirror/docs/