Subject: Can't find swap partition?
To: None <port-pmax@netbsd.org>
From: Ted Spradley <tsprad@metronet.com>
List: port-pmax
Date: 12/16/1994 20:19:58
I've had some measure of success building a new kernel, "self-hosted".
Executive Summary: a few little glitches, it boots and runs, but
doesn't swap.
I got the Dec 3 tar-balls from Iowa State and set out to compile a new
kernel, on a system running Jonathan Stone's binaries. Mostly
successful, but a few minor difficulties:
1) sys.tar.gz was corrupted and wouldn't gunzip (Michael or Michael?
probably long since fixed by now), so I went to sun-lamp to get a better
copy.
2) 'make depend' wouldn't go because /usr/bin/mkdep tried to invoke cpp
with a whole long list of source files, and cpp wants only one input
file and one output file. I just changed 'cpp' to 'gcc' in mkdep and
that took care of that.
3) vnode_if.h had the extra commas in the function prototypes. I
figured, Aha! I have awk sources from research.att.com. I bet that'll
work right. Nope. Got compile errors, probably yacc incompatibilities,
I guess. I punted that problem and edited vnode_if.h by hand.
4) everything compiled but the linker came up with undefined
references to '_end' and '_edata'. That was the original ld binary
that Jonathan put up for ftp back in September. I also had the as
and ld he put on gregorio.stanford.edu about a week ago. I switched
to that ld (not as) and it went fine.
The reason I didn't use that one in the first place is that with the
new as and ld I can't "gcc hello.c", it gets the converse error,
undefined references to 'end'. This business with the underscores has
me confused.
5) When I tried to boot it it got a panic in swapinit. It was
configured "swap generic", so I changed to "root on rz4a swap on rz4b dumps on
rz4b" and tried again. This one doesn't panic, but does complain that
it can't find rz4b to swap to. I checked and disklabel shows there
really is a 30 Megabyte b partition. What am I missing here?
--
Ted Spradley tsprad@metronet.com Usual disclaimers apply.
Information tends to drive out knowledge. [...] many people cannot
tell the difference between information and knowledge, not to mention
wisdom, which even knowledge tends sometimes to drive out. -Heinz Pagels