Sun-Spots-Request@RICE.EDU (William LeFebvre) (03/29/88)
SUN-SPOTS DIGEST Thursday, 24 March 1988 Volume 6 : Issue 38
Today's Topics:
Re: Incompatbities between Sun diag programs?
Re: Multi-process debugging using Dbxtool
Re: More than one dump on 1/4 inch tape
Re: TCP packet size bug in 3.4 AND 3.5 (2)
Re: Problems rewinding a cartridge tape on a SUN-3
Cleaning up after SunOS 3.5 Tools
SoftPC - a software coprocessor from Insignia
Wanted, FIG to Postscript filter (f2ps)
Fast Clock problem
Reading/copying arbitrary 3/50 tapes?
Can backups be done with system in multi-user mode?
Mixing local and nd swap?
program to convert Rasterfile to ln03 (sixels)?
Sun SCSI CD-ROM device driver?
Adding a cgtwo to a 3/75M?
The acronym "SPARC"
Send contributions to: sun-spots@rice.edu
Send subscription add/delete requests to: sun-spots-request@rice.edu
Bitnet readers can subscribe directly with the CMS command:
TELL LISTSERV AT RICE SUBSCRIBE SUNSPOTS My Full Name
Recent backissues are stored on "titan.rice.edu". For volume X, issue Y,
"get sun-spots/vXnY". They are also accessible through the archive
server: mail the word "help" to "archive-server@rice.edu".
----------------------------------------------------------------------
Date: Wed, 16 Mar 88 10:29:48 EST
From: lokkur.UUCP!scs@citi.umich.edu (Steve Simmons)
Subject: Re: Incompatbities between Sun diag programs?
Reference: v6n27
William P. Malloy (wpm@gateway.mitre.org) writes:
>...I tried to run the new diag (which shows up as version 1.7) that comes
>with 3.5 to read the disk labels. The ``verify'' command could not find
>the disk labels ALTHOUGH SunOS obviously could as it could boot. I also
>cannot re-label the disks, the write's fail.
We have had exactly the same problem with some extremely old CDC-9771s.
It turns out that if you don't give *exactly* the same answers every time
when diag asks you questions, you can't find the labels. The key field we
stumble over is "type". 0, 1 and 2 are valid answers, and a Sun engineer
long ago told us "pick one -- it doesn't matter". It turns out that once
you pick one *you have to stick to it*. If when using diag on a "type 1"
disk I specify "type 2", it can't read the label but it *can* boot.
>I don't know if Sun did something extremely stupid or I did . . .
More likely it's just a matter of diag being "upgraded" and those of us
with nonsupported drives being caught out. It's not reasonable to expect
Sun to test our configurations, after all. We found differences in the
prompts (more questions, like "interleave?") between 3.0 and 3.2. This
made it difficult to give the same answers since we hadn't been asked the
question before! We have 3.5, but haven't tried it on our CDCs yet.
>Maybe it's just Sun's way of telling us to buy our disks from them. :-)
It's just the price we pay for going our own way. But I'm sure Sun
wouldn't mind if we decided to buy the disks from them because of it! :-)
Steve Simmons
Schlumberger CAD/CAM, Ann Arbor, MI.
scs@lokkur.uucp (home)
simmons@applga.uucp (work)
------------------------------
Date: Wed, 16 Mar 88 09:36:50 PST
From: weiser.pa@xerox.com
Subject: Re: Multi-process debugging using Dbxtool
Reference: v6n28
Slight correction to the message from Danny Backx, who says "(more
general: no file can be debugged if more than one process is runnung it)".
Actually, as I stated earlier, a file can be debugged when several
processes are running it if the file has OMAGIC (obtained by the -N flag
to cc or ld.)
-mark
------------------------------
Date: Wed, 16 Mar 88 23:09:34 EST
From: ted@braggvax.arpa
Subject: Re: More than one dump on 1/4 inch tape
Reference: v6n29
We regularly get more than one dump on 1/4 inch tape with no problems.
The trick is to use /dev/nrst[08] so that the tape does not rewind between
dumps. On our sun2s we are easily able to get a dump of /, /pub and a
client partition all on one rst0 tape.
Ted Nolan
ted@braggvax.arpa
------------------------------
Date: Thu, 17 Mar 88 08:33:55 PST
From: suncan!sunvcr!kirk!spock!terry@sun.com (arden Terry)
Subject: Re: TCP packet size bug in 3.4 AND 3.5 (1)
Reference: v6n28
To permanently fix that annoying TCP/IP packet size bug in SUN OS 3.[45]
edit the following file, re-generate a kernel, and reboot.
cd /sys/OBJ
adb -w tcp_input.o
tcp_mss+0xac?w 400
(was at 200)
tcp_mss+0xbc?w 400
$q
Good-luck.
Terry Arden
MacMillan Bloedel Research
------------------------------
Date: Thu, 10 Mar 88 23:39:50 EST
From: Mike Muuss <mike@BRL.ARPA>
Subject: Re: TCP packet size bug in 3.4 AND 3.5 (2)
Reference: v6n28
[[ This was relayed to me via "stpeters%dawn.tcpip@ge-crd.arpa". I should
have put it in a digest sooner, but "oh well"!. --wnl ]]
...Phil here at BRL experienced the same problems when using 3 different
TCPs here, which definitely points the finger at the TCP on poor <expo>.
I'm the original proponent of the tcp_mss code, and moving it to 1024 will
actually cause you significantly more trouble than the 512 setting, as it
will case the packets to be fragmented as they enter the IMP-based
networks (ARPANET, MILNET). If you are trying to squeeze a few more data
bytes per packet from the network, be certain that this relationship
holds:
tcp_mss <= Network_MTU - Max_Header_Size
In the case of IMPs, the MTU is 1006. [[ Actually, it's 1007. --wnl ]]
The max header (TCP+IP) isn't huge, perhaps 42 bytes :-). Therefore, your
tcp_mss should not exceed 964 bytes. If you want to be really
conservative, a number like 920 or 930 would allow lots of room for IP
Options (which are rare, but not unheard of). Even at 920 that would give
you a 1.8x performance advantage. HOWEVER, note that no system is
REQUIRED to accept a packet with more than 512 data bytes, so by
implementing this change you may reduce your interoperability. No UNIX
system is likely to mind, but some non-UNIX systems (esp. some MULTICS
machines) are unhappy with larger MSS settings.
If you induce IP fragmentation through the core network system, you will
do very much worse than with the current setting. The problem is that if
one fragment of a multi-fragment IP datagram is dropped, then the whole
transmission (all fragments) are not useful, and have to be re-sent. If
each TCP transmission fits in a single IP datagram, then TCP has the
opportunity to be more sensible with it's retransmission strategy.
At BRL, we run SUNOS 3.4, with the "Van" TCP code replacing the Sun code.
I believe he offers the fixes in both source and Sun-binary-patch form.
It's a real winner, consult the archives of the <TCP-IP> list for details.
Best,
-Mike
------------------------------
Date: Fri, 18 Mar 88 09:06:58 CST
From: Jim Knutson <knutson@mcc.com>
Subject: Re: Problems rewinding a cartridge tape on a SUN-3
Reference: v6n30
On most Suns that use the SCSI bus for both the cartridge tape and disk,
the controller/cartridge drive will hold the bus during ANY tape operation
(including rewind or retension). You will still be able to operate so
long as you don't need to swap or page. Some of the newer
controller/drive combinations (like those found on a 3/140) will
relinquish the bus if running SunOS 3.3 (or was it 3.4) or higher during a
tape operation which doesn't require the bus.
Jim Knutson
knutson@mcc.com
im4u!milano!knutson
------------------------------
Date: Thu, 17 Mar 88 10:20 EST
From: Roy Richter <RICHTER@gmr.com>
Subject: Cleaning up after SunOS 3.5 Tools
Here's a script that will clean up most of the debris from the new tool
versions in SunOS 3.5. The part new to this mailing list is in the second
section. The new versions of CommandTool and MailTool don't clean up
their scratch files if you exit suntools while they are running.
#
# cleanup_suntools, a script to cleanup after suntools
# invoke along with suntools, e.g.,
# % suntools; cleanup_suntools
#
# Roy Richter GM Research Labs 3/17/88
#
# get rid of the pseudo-ttys that are hanging around
# use clear_ttys from Sun-Spots archive server,
# described in Sun-Spots Digest, v6n9
#
echo -n "clearing ttys..."
$HOME/bin/clear_ttys.x
echo "done."
#
# clear out the scratch files left in SunOS 3.5
# by MailTool, CommandTool, ...
# You could also add ttyselection, winselection, and the shelf
# if you don't want anybody peeking at them after you leave
#
# one typically has /tmp linked to /usr/tmp; if so, leave alone.
# if not, change the following line to where you have /tmp linked.
#
set tmplink=/usr/tmp
#
echo -n "clearing $tmplink ..."
find $tmplink -user `whoami` \
\( -name MT\* -o -name Text\* -o -name tty.txt\* \) \
-exec rm {} \; -prune
echo "done."
#
# end of cleanup_suntools
------------------------------
Date: Thu, 17 Mar 88 15:55:21 GMT
From: Zdravko Podolski <mcvax!insignia!zdrav@uunet.uu.net>
Subject: SoftPC - a software coprocessor from Insignia
[[ Some may view this message as an advertisement, but I have reviewed the
article and have given it my "seal of approval". It is a factual
description of a product that, based on previous discussion, will likely
be of interest to some readers. And, yes, one of the facts stated is the
price. (Just trying to calm the flames before they are sent.) --wnl ]]
SoftPC (tm) - IBM PC emulation in software
There have been several queries about SUNs and PCs coexisting, cooperating
and sharing files.
Insignia Solutions have developed a software solution to the problem,
called SoftPC. This is an emulation of an IBM PC/XT at the hardware level.
The software runs PC applications at about XT speeds on a 3/160 and AT
speeds on a 3/260.
SoftPC supports ill behaved programs too, such as MS Flight. The emulation
includes a 'hard disk' which is a large UNIX file and also file sharing
with UNIX. The data format is completely compatible and sharable with PC's
on for example PC/NFS.
Copy protected software can be installed on the 'hard disk' either using
an external floppy disk drive, or by slaving a real PC to your SUN.
Display adapters supported are currently MDA and CGA, with colour on a
colour SUN, and emulated colour (different patterns) on b/w. Other display
emulations are in preparation.
SoftPC also allows the emulated serial ports (2 on an XT) to be hooked up
to files on the host SUN, to the real serial ports or to pipes. Thus
SoftPC can drive modems or printers directly, or talk to a spooler.
The one off price is 595 US Dollars and SoftPC should be available from
your dealer now. If you cannot find it the distributor in the US is
Softsel of Inglewood, CA.
SoftPC is also available for the MACII (AUX and MACOS), Tektronix 43XX,
Silicon Graphics IRIS, MIPS and the Intergraph Clipper based workstations.
Development work includes both porting to new CPU's (we showed a VAX
version at DECUS in Dec 87) and further performance improvements.
Zdravko Podolski
UK: Insignia Solutions Ltd, Victoria House, 28-38 Desborough St,
High Wycombe, Bucks, HP11 2NF, England
Email: ...!ukc!insignia!zdrav or zdrav%insignia.uucp@ukc.ac.uk
Tel: +44 494 459426, Fax: +44 494 459540, Telex: 83389 BUSENT G
US: Insignia Solutions Inc, 1255 Post St, Suite 625, San Francisco, CA 94109
tel 415 885 4455
(TM) SoftPC is a trademark of Insignia Solutions
UNIX is a trademark of ATT Bell Labs
SUN is probably a trademark of SUN Microsystems
PC/XT and PC/AT are trademarks of IBM
Other things are probably other people's trademarks and I apologise
if I missed any
------------------------------
Date: 16 Mar 88 14:10:42 GMT
From: mcvax!dnlunx!lippolt@uunet.uu.net (Ben Lippolt)
Subject: Wanted, FIG to Postscript filter (f2ps)
Does anybody has, or know about, a conversion program which can convert
the output from FIG (a MacDraw lookalike for SUNs) to Postscript. A
conversion program from pic to postscript would also be sufficient,
because we already have a fig to pic converter.
Thanks.
Ben Lippolt Telephone: +31 70 435439
PTT-Dr. Neher Laboratories Telefax: +31 70 436477
P.O.Box 421 Telex: 31236 dnl nl
2260 AK Leidschendam UUCP: ..!mcvax!dnlunx!lippolt
Netherlands EARN: lippolt@hlsdnl5
------------------------------
Date: Thu, 17 Mar 88 13:09:55 PST
From: chaos%gojira.Berkeley.EDU@jade.berkeley.edu
Subject: Fast Clock problem
Problem: The clock on a 3/50 now gains a half hour a day.
Context: Sun 3.2. The 3/50 was turned off for about 10 minutes. On power
up, it complained that it had lost 30 hours. The clock has worked fine
for the last year + and it had never gotten so confused after a power
down.
Question: Is this a dead battery? If so, how to replace it and what with?
If not, any suggestions?
Tangential Issue: After many months of procrastinating I am considering
upgrading to 3.5. The manuals tell what has been fixed, but experience
shows that some things will be broken as well. Anything major that I
should know about?
Thanks,
Jim Crutchfield
Physics, UCB
Berkeley, CA
(415) 642-1287
[[ All important question: have you booted the machine since the first of
the year? Have you installed the patch for the "leap year clock drift"
bug? If the answers are "no", then that is your problem. Retrieve volume
6 issue 1 of the Sun-Spots digests. It contains the necessary patch to
fix this bug. (Are there really still people out there who haven't
encountered this problem yet?) --wnl ]]
------------------------------
Date: 14 Mar 88 15:49:59 GMT
From: Olaf Pors <op@boole.acc.virginia.edu>
Subject: Reading/copying arbitrary 3/50 tapes?
Does anyone know how to read/copy a Sun 3/50 1/4-inch cartridge (st0 or
st8) tape? Once you know the block sizes and number of files, it's easy
to use dd, but I'm not aware of any way to determine the block sizes given
an "unknown" tape. It seems that the st driver reads as many physical
tape blocks as needed to satisfy the dd "bs" parameter.
My purpose is to be able to easily produce backup copies of important
tapes.
Olaf Pors
Academic Computing Center
Gilmer Hall
University of Virginia
Charlottesville, Va. 22903
804-924-0633
op@virginia.edu (Bitnet)
or op@bigfoot.acc.virginia.edu
------------------------------
Date: Wed, 16 Mar 88 11:55:23 EST
From: uunet!helios!hydra!lalonde@ut-sally.UUCP (Terry Lalonde)
Subject: Can backups be done with system in multi-user mode?
I wondering if there is a way to backup Sun systems without having to goto
single-user mode, as this causes background processes (long cpu intensive
jobs) to die.
Does anyone have a solution to this or recommendations ?
Thanks in advance
Terry
------------------------------
Date: Thu, 17 Mar 88 10:04:40 CST
From: Jim Knutson <knutson%SW.MCC.COM@mcc.com>
Subject: Mixing local and nd swap?
We would like to be able to add some additional swap space to a Sun with
local disk (and swap) by using an ND partition on a server. Has anyone
had any luck with this?
Jim Knutson
knutson@mcc.com
im4u!milano!knutson
------------------------------
Date: Thu, 17 Mar 88 11:11:13 EST
From: ajb%cornea.mitre.org@gateway.mitre.org
Subject: program to convert Rasterfile to ln03 (sixels)?
I'm interested in printing sun rasterfiles on our LN03-PLUS (with full
page bitmap memory). Has anyone written a rasterfile to LN03 conversion
program ?
Thanks
Alan J Broder
The MITRE Corporation
ajb@mitre.arpa
------------------------------
Date: 17 Mar 88 22:32:59 GMT
From: harry@rainy.atmos.washington.edu (Harry Edmon)
Subject: Sun SCSI CD-ROM device driver?
Does anyone have a device driver for a SCSI CD-ROM? I am especially
interested in one that reads "High Sierra" format disks.
Harry Edmon UUCP: harry@rainy.atmos.washington.edu or
(206) 543-0547 uw-beaver!geops!rainy!harry
Department of Atmospheric Sciences
University of Washington BITNET: HARRY@UWARITA
------------------------------
Date: Fri, 18 Mar 88 10:52:37 EST
From: jas@monk.proteon.com (John A. Shriver)
Subject: Adding a cgtwo to a 3/75M?
While the Sun-3/75 is low on power, it is even lower on cooling. This is
probably why you don't want to put a cgtwo board in it. Anything oomphier
than a SCSI board will probably die of heat prostration, or cuase the CPU
card to do so. One of the goals of the Sun-3/110 chassis was to solve the
power and cooling restrictions of the Sun-[23]/75 chassis.
------------------------------
Date: 16 Mar 88 18:03:25 GMT
From: govett@avsd.uucp (David Govett)
Subject: The acronym "SPARC"
According to the Feb. 8 edition of Nikkei Electronics (Tokyo),
SPARC stands for Scarable Processor ARChitecture.
Do the Japanese know something we should?
------------------------------
End of SUN-Spots Digest
***********************