Subject: sparc boot blocks
To: None <port-sparc@NetBSD.ORG>
From: Paul Kranenburg <email@example.com>
Date: 12/02/1996 14:50:07
I've changed installboot(8) to munge on the (first-stage) bootblock as it
is written to disk, to have it work correctly on both `oldmon' (sun4)
and OBP machines. I used the trick to make the first few words in the
image look both like a valid header to OBP machines and like instructions
that branch past the first 8 words.
Please let me know if the bootblocks start to fail on you.
Thanks to Metthew Jacob and Mike Tuciarone for ensuring that such a trick
would indeed work.
Mike's other suggestions about writing the bootcode in Fcode and/or coding
an UFS package in < 4k, so one can do away with the current multi-stage
approach, look very interesting too..
BTW. A snapshot of -current including a miniroot that has the new bootblocks
is at ftp.netbsd.org.