Unix Technical Digest V1 #32

[Ron Heiby (The Moderator)]

Unix Technical Digest       Tue, 26 Mar 85       Volume  1 : Issue  32

Today's Topics:
                4.2BSD boot from slave 1 disk (3 msgs)
                    Disk partition sizes (2 msgs)
----------------------------------------------------------------------

Date: 19 Mar 85 03:02:35 GMT
From: steve at milo.UUCP (Steven Kahn)
Subject: 4.2BSD boot from slave 1 disk

I have a 4.2BSD  system I wish to boot from an ra81,  which
happens to be slave 1 (not 0) on the uda50. I can certainly
do:

	>>> B ANY
	:ra(1,0)vmunix

but I'd rather  not type that each time.  Is there a way to
modify /sys/floppy/rasboo.cmd so that

	>>> B RAS

will boot from slave 1 instead of 0?

	Any suggestions would be greatly appreciated.
-- 
steve kahn
johns hopkins applied physics lab, laurel, md
...!seismo!umcp-cs!aplvax!milo!steve

------------------------------

Date: 19 Mar 85 06:36:49 GMT
From: chris at umcp-cs.UUCP (Chris Torek)
Subject: 4.2BSD boot from slave 1 disk

To boot from ra(1,0) by default (or by B RAS/B RAM from the 780
console), you have to hack up the boot program (/sys/stand/boot.c, BOOT
on the floppy) a bit.  What we did was use r9 to specify the device and
partition (though using the high word of r10 might be better, since
it's already set to 0).  Basically it's a matter of replacing the
string at the top with the right numbers (it already replaces the "xx"
based on r10, a.k.a. devtype).

Then all you need to do is edit RASBOO.CMD, RAMBOO.CMD, and DEFBOO.CMD,
to include the device and partition information.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 4251)
UUCP:	{seismo,allegra,brl-bmd}!umcp-cs!chris
CSNet:	chris at umcp-cs		ARPA:	chris at maryland

------------------------------

Date: 22 Mar 85 04:48:40 GMT
From: jsdy at hadron.UUCP (Joseph S. D. Yao)
Subject: 4.2BSD boot from slave 1 disk

> I have a 4.2BSD  system I wish to boot from an ra81,  ...

>       ... Looking at /sys/mdec/raboot.s, it looks like
> either register R3 or R9 or both are involved in unit selection.
> However, I tried Depositing each with 1 during booting,
> and neither approach worked.  ...

The mdec raboot goes in the boot block.  It does not get read in
unless something else pulls it in.  In particular, on milo (a VAX
11/780), your floppy-disc program BOOT always gets read in first,
and then pulls whatever it wants off your hard disc.  On machines
like the 11/750, where you always ROM-boot off the first block,
you can change the unit number in R3 and get different units.  But
I don't believe that the 4.2 mdec/raboot will work for an 11/780.
In any case, you would have to go back to using the DEC VMB.EXE
loader with boot flags (R5) set to block boot (8).

Or, as Chris Torek points out, you can hack up /sys/stand/boot.c
to do the same thing.

	Joe Yao		hadron!jsdy at seismo.{ARPA,UUCP}

------------------------------

Date: 17 Mar 85 07:01:21 GMT
From: notes at isucs1.UUCP
Subject: Disk partition sizes

	A question on file system sizes.  Our Vax 730 running BSD4.2 is 
partitioned using the information out of the /etc/disktab file.  We have an
rb80 disk drive partitioned as folows:
/dev/rb1a	sec used 15884  sec avail 16058  cyl 0 - 36  
/dev/rb1b	sec used 33440  sec avail 33852  cyl 37 - 114
/dev/rb1g	sec used 82080  sec avail 82460  cyl 115 - 304
/dev/rb1h	sec used 110143 sec avail 110236 cyl 305 - 558

	When I run diskpart 8 it recommends the following differences:
/dev/rb1d	15884 sec  cylinders 115 - 151
/dev/rb1e	55936 sec  cylinders 152 - 280
/dev/rb1f	120466 sec cylinders 281 - 558
/dev/rb1g	192510 sec cylinders 115 - 558
/dev/rb1h	unused

	We would like to partition our disk with the g partition described
in the second list.  This would give us the standard a & b partitions with
the remainder of the disk in partition g.  Has any body ever done this and
does it work?  

	My second question is why does diskpart 8 only recomend 192510 sectors 
when there are 192539 sectors available according to the way the diskpart manual
page describes how to compute it.  Compute as follows: (192696 the total 
possible number of sectors ) - (1 track,31 sectors on an rb80; for the 
replicated copies of the bad144 table) - (126 sectors to which bad sectors can 
be mapped)= 192539.  Where do the numbers that diskpart 8 is recommending come 
from and why wouldn't I be able to use the number calculated above.  Also if 
these calculations are correct our h partition configured the berkeley way is 
stomping on the bad144 info as it only leaves 94 sectors at the end of the disk 
rather then the recommended 157 sectors.  
						-Jeff Rule

UUCP:	{umn-cs,csu-cs,okstate}!isucs1!exnet
[Beat em or burn em they all go up pretty easy.] - 'Night of the Living Dead'

------------------------------

Date: 23 Mar 85 02:52:41 GMT
From: chris at umcp-cs.UUCP (Chris Torek)
Subject: Disk partition sizes

I don't have any specifics on rb80 disks, but one general thing to keep
in mind when computing 4.2BSD partition sizes: Berkeley probably isn't
using that drive, probably doesn't have any information about it, and
probably is just using whatever was in the software someone else gave
them.  If it seems to be wrong, there's a good chance it's wrong.
-- 
In-Real-Life: Chris Torek, Univ of MD Comp Sci Dept (+1 301 454 4251)
UUCP:	{seismo,allegra,brl-bmd}!umcp-cs!chris
CSNet:	chris at umcp-cs		ARPA:	chris at maryland

------------------------------

End of Unix Technical Digest
******************************
-- 
Ronald W. Heiby / ihnp4!{wnuxa!heiby|wnuxb!netnews}
AT&T Information Systems, Inc.
Lisle, IL  (CU-D21)