21d18c459f
was previously called "subversion" as "build", remove old cruft from newvers.sh XXX very lighly tested build on NetBSD/hp300 1.4.1, though could not do full XXX build due to toolchain differences to -current |
||
---|---|---|
.. | ||
boot.c | ||
conf.c | ||
devopen.c | ||
dmareg.h | ||
en.c | ||
enreg.h | ||
installboot.sh | ||
machdep.c | ||
Makefile | ||
newvers.sh | ||
README | ||
rtc.c | ||
scsi.c | ||
scsireg.h | ||
scsivar.h | ||
sd.c | ||
srt0.s | ||
version |
$NetBSD: README,v 1.1.1.1 1998/06/09 07:53:06 dbj Exp $ NeXT standalone bootblocks. Rolf Grossmann, Dec 1994 Started work based on files from hp300/stand. boot.c was from post-1.0 sparc/stand/boot.c, modified to work for the needs of the NeXT PROM, i.e. it wants to call the kernel, so the bootblock has to return the entry point. The code does not try multiple names for te kernel, as I've seen it in some other architectures' boot code. (The copied code simply didn't do that ;)) It also doesn't prompt if the argument to boot ends with a questionmark '?', like the NeXT bootblock does. Do we need this? (Why should the bootblock as again when you can specify everything on the boot command line?) Most files have nothing to do with their original version anymore. The whole code is a mixture of my own ideas, various other netbsd code I've looked at (like the sparc scsi code, the independent scsi code, and the needs of the standalone library). In contrast to NeXT's bootblocks, mine keep the PROM's idea of what the boot parameters are, i.e. logical disk number (the number the disk would get as sd*), the lun and the partition. TODO Make some additional improvements