Re: CVS version builds but doesn't run

[Matt Porter <mmporter@home.com>]

On Tue, Nov 02, 1999 at 09:17:51AM -0700, Erik Andersen wrote:
> On Tue Nov 02, 1999 at 01:28:28PM +0000, Craig H. Block wrote:
> > Erik Andersen wrote:
> > > 
> > > or we can leave busybox/init.c alone and we can
> > > create a /etc/init.d/rcS init script that does
> > > something like:
> > > 
> > >     #!/bin/sh
> > >     /bin/dbootstrap
> > > 
> > > I really don't care which option folks prefer...
> > > Let me know, and I will make it so.
> > > 
> > 
> > I like the script approach.  Its more standard.  C code gives me a
> > headache :)
> > 
> 
> Good, me too.  It is a lot more standard, and should help 
> folks avoid the temptation of making special-case versions
> of busybox for every application.

I've changed my tree so rcS spawns /sbin/dbootstrap, however it just 
hangs.  I believe it is segfaulting since I was seeing segfaults when I
built standalone versions of dbootstrap.  I've done test versions using the
release_notes_test rule and others which all segfault in strlen().
Normally I'd blame gcc/libc on PowerPC, but I have changed my build
environment since I built good images with the release of boot-floppies
2.2.1 which was pre new busybox.  My script echos so at least I know init
is spawning dbootstrap.

I will try this on a video console machine since I can't see a lot of error
messages on serial console.

--
Matt Porter
mmporter@home.com
This is Linux Country. On a quiet night, you can hear Windows reboot.