Hi! On Mon, Dec 07, 1998 at 09:29:37PM +0100, Konrad Kokoszkiewicz wrote: > Additionally, I would be for moving the function of killing all processes > from MiNT to init. I.e. init, when hit by SIGTERM or similar signal, would > kill everyone (its children), then terminate. So if init terminates, the > kernel could assume that there are no processes left, i.e. just do some > Sync(), then straight reboot. > > On systems without init.prg, ie. where AES is used, would be the same. > I.e. the AES would only care on killing children at shutdown time, then > it would exit, what would be a signal for the kernel that it should reboot > the machine. > > I believe this also would solve the problem with assert from pipefs.c > sometimes occuring while shutting the system down. > > Any comments? Well, I think that KGMD's init.prg is very limited, and should be replaced with a more sophisticated one. I'm thinking of runlevels and other nifty stuff. A port (if not already done by someone) could then of course use MiNT's Shutdown() function on receiving SIGHUP (reboot?) and SIGTERM (halt?) I wouldn't opt for changing MiNT's exit behaviour, as it may be useful for some people. Ciao Thomas -- Thomas Binder (Gryf @ IRCNet) gryf@hrzpub.tu-darmstadt.de PGP-key available on request! binder@rbg.informatik.tu-darmstadt.de
Attachment:
pgpl5qjVylFDS.pgp
Description: PGP signature