[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [MiNT] MiNT termination

> > I wouldn't opt for changing MiNT's exit behaviour, as it may be useful
> > for some people.
> I agree.  
> And I wouldn't like the idea that terminating the AES would
> unconditionally kill MiNT because I would like the idea
> of being able to launch at kill the AES whenever I want
> without having to reboot.  

You got me wrong. Exiting AES would cause a reboot only (ONLY!!!) when AES
is init! When AES is init, you have no process to leave from the AES to,
so a reboot (or whatever) follows. If you started AES from a text shell
(i.e. MiNT starts init.prg, init.prg starts (finally) login, login starts
shell, shell starts AES), exiting AES whouldn't cause a reboot, but just a
return to the shell.

Generally the idea was: LEAVING INIT (pid 1) causes a reboot, like it
causes a return to the TOS now. Then, IF and ONLY IF AES is pid 1 (=
init), leaving AES causes same stuff.

The current behaviour (leaving pid 1 causes an *attempt* to go back to
TOS) may be useful - i can admit - but also can be annoying if you
accidentally kill pid 1 from remote: (repeating myself) there's no way to
get the control back on the machine, because it sits there in TOS.

So what about making it configurable. I.e. some sort of a(nother) MINT.CNF
keyword like EXIT=n, where n = 0 is default (as now), n = 1 is reboot, n =
2 is halt or something similar.

> Alternative proposal:  If init goes to runlevel 0 the system is
> halted.  Runlevel -1 (or any unused) would cause the kernel to
> try (only try!) to exit gracefully and leave a plain TOS 
> environment.



Konrad M.Kokoszkiewicz
|mail: draco@mi.com.pl                  | Atari Falcon030/TT030/65XE |

** Ea natura multitudinis est,
** aut servit humiliter, aut superbe dominatur (Liv. XXIV,25)
** U pospolstwa normalne jest, ze albo sluzy ono unizenie,
** albo bezczelnie sie panoszy.