Subject: re: Modification To CD-ROM Device Driver (To Read Partition
To: matthew green ,Rob Ginn <>
From: Rob Ginn <>
List: tech-kern
Date: 04/19/1999 21:36:51

No, "boot cdrom" isn't broken, but the default root partition
is based on the default boot partition.  So, if you are on a
Sun 670 (for example), the boot cdrom command makes the system
boot the (sun) partition label entry cd0d.  When NetBSD comes up,
it defaults to mounting root at /dev/cd0d (which doesn't exist
in the fake cd label we have now).  Thus you have to specify boot
options.  Not a disaster, but certainly less convenient.

As for the throw away issue, to make a bootable CD now, one
makes a CD that looks like this:

[iso image with root filesystem][install image with (unused) root filesystem]

The install image is _only_ used to bootstrap the system, the rest
of the process loads the kernel from the iso image at the start
of the disc.  The root filesystem that is mounted is also there.
Thus, the root filesystem on the end of the disc is a "throw away".

As to why it's faster, I'm not really sure, but my guess is
that it is the speed of the filesystem.  And yes, the speed
difference is amazing.

- Rob

At 01:21 PM 4/19/99 +1000, matthew green wrote:
>hi rob.
>i have some patches floating around to me scsipi/cd.c to make it use
>readdisklabel() to get the real disklabel on the disk... the problem
>was, that then i wasn't able to mount the real partitions there as
>now the types didn't match, or something... it was a while ago.
>    Install CD's for Sparc run much faster (at _least_ 5x faster) and can be
>       run with a simple "boot cdrom" command at the BIOS.  It is even
>       comfortable now to run the whole thing off the CD as an emergency
>"boot cdrom" isn't broken.  at least, it wasn't last i tried it when i
>was doing the bootable sparc cdrom stuff.  do you know why is it faster?
>that is interesting...
>    The boot image on the end of the disc is no longer a "throw-away".  My 
>       install CD's now have an ISO image with the tarballs followed by a
>       copy of the miniroot (FFS) which is actually executed (instead of a
>       copy of the miniroot data on the ISO partition being used).  I did
>       have to re-write the mkbootcd program, but that didn't take long.
>i don't really understand this...
>   If there is no heartburn, I'll submit the change request with the diffs
>   (or email it if that's simpler).
>send-pr is the best way.