Sun-Spots-Request@RICE.EDU (William LeFebvre) (05/25/88)
SUN-SPOTS DIGEST Monday, 23 May 1988 Volume 6 : Issue 97
Today's Topics:
Re: Interrupt button under SunView (2)
Re: Consult/Gateway
Re: silo overflow
Ciprico Rimfire 3220
Oops: my termcap entry
More on iechkcca panics
fig2tex
Wanted - board to do group 4 compression FAST
Any info on big, quiet disks
Need TEKTRONIX emulator for PC via Ethernet Network
SunOS 4.0 Device Drivers?
tape file systems?
LaserWriter II running on Suns?
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 available via anonymous FTP from "titan.rice.edu".
For volume X, issue Y, "get sun-spots/vXnY". They are also accessible
through the archive server: mail the request "send sun-spots vXnY" to
"archive-server@rice.edu" or mail the word "help" to the same address
for more information.
----------------------------------------------------------------------
Date: Fri, 20 May 88 08:17:04 EDT
From: Chuck Musciano <chuck@trantor.harris-atd.com>
Subject: Re: Interrupt button under SunView (1)
If your application is running under the window system, all you need to do
is to attach an event proc to the base frame. This event proc will see
every event that is passed to your window, regardless of whether the
number crunching is going on or not. You can then look for a certain
keystroke (^C would be good) and act on it accordingly. A sample proc
would be
static Notify_value event_proc(frame, event, arg, type)
Frame frame;
Event *event;
Notify_arg arg;
Notify_event_type type;
{
if (event_id(event) == '\003' && event_is_down(event)) {
/* set some flag telling the integration to stop */
return(NOTIFY_DONE);
}
else
return(notify_next_event_func(frame, event, arg, type));
}
After you have created your base frame, before you call
window_main_loop(), insert the event proc with:
notify_interpose_event_func(frame, event_proc, NOTIFY_SAFE);
Your event proc will see every event passed to the base frame, including
keystrokes. You can also monitor the state of the window, to track
whether it is opened or closed. Examples of this, and more information,
can be found in the chapter on the Notifer, in the SunView Programmer's
Guide.
Chuck Musciano
Advanced Technology Department
Harris Corporation
(407) 727-6131
ARPA: chuck@trantor.harris-atd.com
------------------------------
Date: Fri, 20 May 88 13:32:58 PDT
From: gandalf@csli.stanford.edu (Juergen Wagner)
Subject: Re: Interrupt button under SunView (2)
Two alternatives:
[1] Use the notifier to setup a signal handler for, say SIGUSR2. Then fork a
process which opens a small window containing that magic interrupt button.
When the button is depressed, a signal SIGUSR2 is sent to the parent.
The parent can then do a longjmp to the position where it can restart
(be sure to trash any malloc'ed space).
[2] Use polling, either by setting a notifier on an input fd, or by explicitly
writing a loop calling the poll function.
In my view, the first alternative is preferrable because you don't get any
computational overhead until the user really hits the button. The tool can
also be used in a very general way (i.e. "signaltool <sig> <pid>
<message>" popping up a window with a button labelled <message>, sending a
signal <sig> to process <pid> when hit).
I've used that to monitor certain background processes where I'd like to
avoid to have to type in a "ps" and a "kill".
Juergen "Gandalf" Wagner, gandalf@csli.stanford.edu
Center for the Study of Language and Information (CSLI), Stanford CA
------------------------------
Date: Fri, 20 May 1988 07:41-EDT
From: Ralph.Hyre@ius3.ius.cs.cmu.edu
Subject: Re: Consult/Gateway
In article <1311@uokmax.UUCP> rob@uokmax.UUCP (Robert K. Shull) writes:
>Has anyone used a product called (I think) Consult/Gateway for the Sun 3?
>It is supposed to allow using up to 4 ethernets under SunOS 3.x.
[Could this be true? Are third-party vendors selling SunOS bug-fixes?]
Doesn't the operating system (ab)normally support this? We had problems
with most of the network code in SunOS 3.4, but upgrading to 3.5 fixed
most of these. Most people just have one or two ethernet boards, but I
didn't think there was any inherent limitation on the number of interfaces
in the Berkeley networking software. Routed and friends should handle it
just fine.
------------------------------
Date: Fri, 20 May 1988 14:37-EDT
From: Michael.Mills@b.gp.cs.cmu.edu
Subject: Re: silo overflow
there are several possible causes. the most common in our domain is from
lance 10M ethernet controllers which are used in 3/50 suns; the error
looks something like:
le0: silo overflow
or occasionally something like:
le0: ... extra byte(s) bug [silo overflow]
the first message indicates that the character input silo (input buffer)
overflowed before it could be reduced (serviced); the second refers to a
lance chip bug i think.
there are a couple other silos used in suns; the zilog serial driver (zs*)
and the systech interface (mti0) come to mind. both have similar silo
error messages when their capacity is exceeded, but since all of these
messages are preceded with a mnemonic it isn't difficult to identify the
culprit.
the lance error is usually handled gracefully, the dropped packets are
requeried and retransmitted. if the problem becomes increasingly
prevalent then it is usually a sign of handling problems in the kernel
(not the controller), but for the most part there is no need to fix
anything because the cause is often just a confluence of events that
retards performance temporarily.
md mills
carnegie-mellon university
computer science dept.
mills@b.gp.cs.cmu.edu
[[ Sorry to hear about your shift key! Hope it's feeling better
soon. :-) --wnl ]]
------------------------------
Date: Tue, 17 May 88 14:07:59 EDT
From: ames!srs!matt@sally.utexas.edu (Matt Goheen)
Subject: Ciprico Rimfire 3220
We have finally succeeded in getting one of our 3220s into operation.
Ciprico was valiant in their efforts to help us, but in the end, we don't
really know what the problem was (or is).
We were trying to install the 3220 (which is the Sun size version of the
3200 SMD controller, it also has 4 B cable connectors rather than just 2)
in a rack mount 3/180. We were using the last slot in the card cage (slot
#12 in our case). For some reason, the controller just didn't work. It
would acknowledge itself at boot time, but doing just about anything with
"rfutil" would crash the kernel (in the driver routine putpb(), I think).
Specifically, doing a `g' command to set the disk geometry would ALWAYS
crash the system.
Well, it turns out the problem was one of two things:
1) Our slot #12 is bad
2) Rimfire 3220s don't work in the last slot
I prefer to believe #1 (in which case everyone can ignore this entire
message), but #2 is a possibility. The only physical difference between
the last slot and every other slot (other than the P2 connections) is that
the last slot has no jumper for P1_IACKIN to P1_IACKOUT. This jumper is
supposed to be removed when using the Rimfire, but for the last slot, it
shouldn't make any difference (the signals that the jumper controls are
for determining priority for boards that interrupt at the same level --
closest to the CPU wins, you USE the jumper when the board doesn't use
this feature).
Anyway, this is just a warning to anyone trying to install a Rimfire in
the last slot. If it isn't working and you don't know why, try moving to
a different slot (ours is now in #10 with the second one to be installed
in #11).
If you have a Rimfire in the last slot and it works fine, I would love to
hear from you.
uucp: {rutgers,ames}!rochester!srs!matt Matt Goheen
maybe-nets: matt@srs.uucp OR matt%srs.uucp@harvard.harvard.edu
------------------------------
Date: Fri, 20 May 88 13:19:50 edt
From: mlijews@nswc-wo.arpa
Subject: Oops: my termcap entry
Oops,
Seems like I was thinking of two things at once when I submitted the ho
termcap entry. The correct entry is ho=\E[H { \E is the symbol for ESC in
termcap}.
Please be wary when and if you install this this entry. Seems there is
some as yet unexplained subtlety in Suntools which causes vi and emacs to
choke when this entry is present, though they perform correctly under the
console. So please be careful to thouroughly check out this entry if you
decide to include it in your termcap.
An easy procedure to test out new termcap entries is to write the entry to
a file, say /usr/home/termcap. Then:
unsetenv TERMCAP
setenv TERMCAP /usr/home/termcap
[[ The "unsetenv" is not necessary. --wnl ]]
Then you can simply fiddle around with /usr/home/termcap w/o having to
keep reinitializing you environment.
Sorry for any problems I may have caused.
Mike Lijewski (mlijews@nswc-wo.arpa)
Applied Math Branch
NSWC
Silver Spring, MD 20903
------------------------------
Date: 20 May 88 12:49:04 GMT
From: roy%phri@uunet.uu.net (Roy Smith)
Subject: More on iechkcca panics
Looks like the cancer is spreading. Remember those iechkcca panics I told
you about yesterday? Well, we got another one about 5 hours after the
last one, on a different system (this time a 3/160C). Getting pretty
spooky. [insert twilight zone music] Possibly notable point; the 3/160
is on a new vampire tap which I installed just about a week ago, violating
one of the rules. I reinstalled a tap in the same place where there was
one before (but I did drill a new tap hole on the other side of the
cable). It passed the extended external loopback test, for what that's
worth.
Roy Smith, System Administrator
Public Health Research Institute
455 First Avenue, New York, NY 10016
{allegra,philabs,cmcl2,rutgers}!phri!roy -or- phri!roy@uunet.uu.net
------------------------------
Date: 20 May 88 19:29:09 GMT
From: mds@bu-cs.bu.edu (Michael Siegel)
Subject: fig2tex
Well I'm sitting at my Sun trying to take a figure from fig and place it
in a Latex document. I have no problem in the translation from Fig to Tex.
However when I try to run LateX I run out of memory when all I am doing is
trying to draw a circle.
I got the Pictex files from a.cs.uiuc.edu but was not able to make the
Pictex file (returns with pictex.tex has too many command lines) but I
doubt this is the problem sin the LateX file calls all the Pictex files
explicitly.
To some it up an example of a LateX file that takes the results from
fig2tex and produces a figure would be greatly appreciated.
Thanks for the help
---Michael
------------------------------
Date: Wed, 18 May 88 16:44:05 edt
From: kevin <mnetor!syntron!jtsv16!kevin@uunet.uu.net>
Subject: Wanted - board to do group 4 compression FAST
I am looking for a board which can do CCITT group 4 compression fast. The
board must be compatible with the PC bus or VME bus on Sun 3 or 386i
systems. By fast, I mean sub second times on a 300 dpi 8.5 x 11 image. I
realize this is asking a lot, and I may have to settle for something
slower.
If anyone has heard of such a device, could you please respond via e-mail.
Thanks in advance,
Kevin Brighton
JTS Computer Systems Ltd. UUCP: {geac|utzoo|utgpu}!syntron!jtsv16!kevin
Toronto +1 416 665-8910 or: suncan!jtsv16!kevin
------------------------------
Date: 18 May 88 15:57:00 MST
From: diegert@sandia-2.arpa
Subject: Any info on big, quiet disks
We are looking for quiet alternatives to the pair of Fuji Eagles
(M2361A's) in a Sun 3/280. The Fuji manual spec's the M2361 at 60 dB A
(each). Is the new, larger SMD Sun is shipping also quieter? Thanks!
------------------------------
Date: Sat, 21 May 88 01:09 N
From: <BAGLIONI%ASTRFI.INFNET%IBOINFN.BITNET@icnucevm.cnuce.cnr.it>
Subject: Need TEKTRONIX emulator for PC via Ethernet Network
Is there anybody using the PC as GRAPHIC terminal over an Ethernet network
along with SUN and/or VAX machines?
We' re looking for a TEKTRONIX or a VT125 emulator (although TEK
compatibility is preferred) for PC to replace the TELNET.EXE program (a
pure VT100 emulator) provided by SUN with the PC-NFS package.
We have not heard of anybody with our same concern, so we have posted this
request. Does anybody know if SUN is devoloping such a product for the
next future, or if it's at least possible to buy the source code for
writing our own version of TELNET with graphic capabilities? We also
believe that any package like FANSI-Console with added graphic
capabilities would do the job.
Any hint will be appreciated,
Thanks a lot
Roberto Baglioni
BAGLIONI%ASTRFI.INFNET@IBOINFN.BITNET
------------------------------
Date: 21 May 88 03:34:14 GMT
From: tim%vsi1@uunet.uu.net (Tim Richardson )
Subject: SunOS 4.0 Device Drivers?
Vicom makes a real-time image processing system. Since Unix and SunOS are
not real-time we have imbeded an additional microcomputer in our system
which handles all the tasks which have real-time requirements. This
computer, along with our other hardware has required several 'hacks' to
the memory allocation handlers and the building of our own device drivers.
Has anyone yet gotten anything which was running under 3.xx Sun0s like the
above, to work in SunOS 4.0???
If there is anyone out there who has done drivers or the like, please let
us know.
Tim Richardson VP Engineering
VICOM Systems Inc. sun!pyramid----\
2520 Junction Ave. uunet!ubvax-----!vsi1!tim
San Jose, CA 95134 ucbvax!tolerant/
+1-408-432-8660
------------------------------
Date: Tue, 17 May 88 14:39:04 EDT
From: ames!srs!matt@sally.utexas.edu (Matt Goheen)
Subject: tape file systems?
Just for fun, we were trying to put a file system on the block tape device
to see how slow it was (i.e. would we want to actually use this feature?).
After a small struggle, I was able to get a file system onto the tape
drive via:
dd if=/dev/rxy1d of=/dev/mt8 bs=4k
The "rxy1d" is a small file system (about 10Mb). I was then able to run
"fsck" on "/dev/mt8" (probably "/dev/rmt8" would work fine as well). It
took about 1/2 hour but it worked ("fsck" was even smart enough to do the
check in read-only mode).
The next step was to try to mount the file system on a directory. The
following SHOULD work:
mount /dev/mt8 /mnt
This would invariably return EBUSY. Making a fresh directory and
attempting to mount on that gave the same result. I even went so far as
to write the following little program:
#include <sys/mount.h>
extern int errno;
main()
{
int val;
struct ufs_args myfs;
myfs.fspec = "/dev/mt8";
val = mount(MOUNT_UFS, "/mnt", M_RDONLY, &myfs);
printf("return value: %d, errno: %d\n", val, errno);
}
This would also return EBUSY. Does anyone know what's going on? Is this
a bug or am I doing something wrong?
BTW, this is a Sun 3/180 running 3.2. The tape drive is a GCR CacheTape
hooked to a Xylogics 472 (xt), in case any of that matters...
uucp: {rutgers,ames}!rochester!srs!matt Matt Goheen
maybe-nets: matt@srs.uucp OR matt%srs.uucp@harvard.harvard.edu
------------------------------
Date: Sat, 21 May 88 08:15:41 edt
From: mlijews@nswc-wo.arpa
Subject: LaserWriter II running on Suns?
Does anyone have an Apple Laser Writer II up and running? With the
LaserWriter being hard to come by, I bought two II's w/ the expectation
that they would work under the Transcript , which runs my other
LaserWriters. But this has turned out to not be the case. A call to Sun
has revealed that the II is still in Beta testing. To run I'm told it
will require buying a newer version of Transcript! They did say that you
should be able to get it to work as a straight line printer by removing
the filters in printcap. I have done this, but this doesn't work either.
If anyone has gotten a II to print, I would appreciate a copy of your
printcap, as well as any other useful info/flames. Thanks.
Mike Lijewski (mlijews@nswc-wo.arpa)
Applied Math Branch
NSWC
Silver Spring, MD 20903
------------------------------
End of SUN-Spots Digest
***********************