SunOS 4.x boot disk layout

[Doc Shipley]

Jules Richardson wrote:

> On Sun, 2005-03-27 at 10:05 -0600, Doc Shipley wrote:
> 
>>   Given the output of fdisk, I'm going to guess that your Linux kernel 
>>doesn't support the SunOS 4 filesystem type.  I've got an email out to a 
>>friend who just installed on a 3/60 yesterday, so mayhap he can shedsome 
>>light on this.  Meantime, see if your kernel has "sysv" filesystem 
>>support.  I have a glimmer of a memory that that's what you'll need.  ;)
> 
> 
> Ahh, that's interesting. I expected it to be an ffs-derived format - in
> other words using a type of 'ufs' under linux and supplied with a
> "ufstype=sun" flag to the mount prog.

   Actually, that's my bad.  A little googling reveals that v4.x is 
indeed ufs.

> (Incidentally I noticed that linux fdisk defaults to reporting the
> blocks count in 1024-byte blocks, rather than the 512-byte blocks that
> the disk itself is using)

   That confuses the students in my Linux classes as well, since fdisk 
reports in 1024-byte blocks even on ext partitions that have 
4096-byte-block filesystems.  :)

>>From what Patrick says, it should just work assuming I have the
> filesystem type right; the root partition is the first on the disk and
> incorporates the partition table, so has an offset of 0. Maybe support
> under Linux is a little broken (I know already that ufs/sun makes
> assumptions about the endian-ness of the filesystem in some places)
> 
> Hmmm...

   It looks like the endian issue may be your problem.  Don't have 
another SPARC to hang the disk from?


	Doc