[BusyBox] My plans for an init.c patch to make ids & reloading different.

Tom Oehser tom at toms.net
Sat Apr 27 07:21:03 UTC 2002


>It's also been requested that runlevels be added (which is really
>where we're heading with /dev/initctl, HUP support, etc.) and that
>the use of serial console doesn't (or optionally doesn't) kill the
>/dev/ttyx'en as well.
>
>Putting all this together, sounds to me like near-rewrite. If you're
>taking the lead on this, I'll be glad to pitch in.

Whoa!  I have sysvinit-2.60 compiled to only 17217 bytes, and the
busybox init.o is already 13K or something.  Overall I think the
busybox init has too many bells and whistles already, not too few,
and if I was to do a scratch-rewrite, I would just aim for a 2K
nasm init.  I'm just trying to find the shortest path, quick-and-
dirty, to fixing a bug introduced in tomsrtbt by replacing sysvinit
with busybox init.  It turns out that the shortest path unfortunately
includes staring at linked lists of actions, but the likely result of
me having to do that will be a reduced and more limited init for me,
and a patch that others (or the project if it wants to) can use all
or part of, as a side effect.

-Tom




More information about the busybox mailing list