Subject: Linux-Development Digest #721
From: Digestifier <Linux-Development-Request@senator-bedfellow.MIT.EDU>
To: Linux-Development@senator-bedfellow.MIT.EDU
Reply-To: Linux-Development@senator-bedfellow.MIT.EDU
Date:     Sun, 15 May 94 18:13:07 EDT

Linux-Development Digest #721, Volume #1         Sun, 15 May 94 18:13:07 EDT

Contents:
  1.0.7 kernel NULL ptr dereference (John Aycock)
  Re: Special plea to all developers and ftp sites... (Charles E Meier)
  Re: LOGISCAN V0.0.1 released (CHRISTOPHER M MAY)
  Re: NeXT like voice mail ? (Bill Chin)
  Driver for smart card (Bob Perkins)
  SIGHUP - Deep Kernal Guts question! (Sean Puckett)
  LOGISCAN V0.0.1 released (Henning Holtschneider)
  Re: Zeos Pantera OK ? (neal kettler)
  Adaptec VL SCSI (John M. Tippitt)
  Re: CD-ROM's do all read and write now? (Russel Mar)
  Stupid Question (maybe) PCI? (Steve Kann)
  workaround to dip 3.3.7 found (Joerg Weule)
  Re: Terminator-Power Question (Claus-Dieter Bredl)
  Q: How do I find the PCMCIA TCIC base address on my notebook? (Rob Whitener)

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

From: aycock@cpsc.ucalgary.ca (John Aycock)
Subject: 1.0.7 kernel NULL ptr dereference
Date: Sat, 14 May 1994 16:39:08 GMT

I'm running a 1.0.7 kernel with IDE performance patches and PPP.  It
seemed quite stable, but last night I got this:

Unable to handle kernel NULL pointer dereference at address 00000412
Oops: 0000
EIP: 0010:0011e9ff
EFLAGS: 00010246
eax: 0400 ebx: 0fde ecx: 0 edx: 0ba8
esi: 0 edi: 01 ebp: bfff0000 esp: 0763f88
ds: 018 es: 018 fs: 02b gs: 02b ss: 018
Process update (pid: 9, process nr: 5, stackpage=763000)
Stack: 0 01e bfffd8b bffff30 0
Code: 80 78 12 00 74 22 83 7c 24 20 00 74 04 85 f6 75 08 bf 01 00

The EIP value falls between _sync_buffers and _sync_dev in my kernel.  It
didn't kill my machine, but I thought I'd mention it anyway.
:ja


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

From: cemeier@magnus.acs.ohio-state.edu (Charles E Meier)
Subject: Re: Special plea to all developers and ftp sites...
Date: 15 May 1994 14:44:14 GMT

In article <2r54bm$epo@klaava.Helsinki.FI>,
Lars Wirzenius <wirzeniu@cc.Helsinki.FI> wrote:
>jon@gtex02.us.es writes:
>> Why do you need to use mess-dos 1440K discs? if you are using a sun (and
>> I happen to know there are 60 odd in the ugrad lounge up there),
>
>It's faster to hoard a computer room with a dozen PCs running DOS and 
>transfer to a dozen disks in parallel than to use a Sun.  (Using the Sun
>lets you read quite a lot of news at the same time, of course, but sometimes
>even that isn't good enough.)
>
>Of course, that isn't a reason for the ftp sites to split things up.
>Those using floppies should do it themselves, there are dozens of programs
>to do it, both for Unix and DOS (I assume Windows has some too).  
>
>BTW, why should ftp sites split things in 1440 kB pieces, and not 1200 kB
>or 720 kB or 360 kB pieces?
>
>--
>Lars.Wirzenius@helsinki.fi  (finger wirzeniu@klaava.helsinki.fi)
>My name is Ozymandias, king of kings/Look on my works, ye Mighty, and despair!

While we are on the subject of ftp sites files, let me add a new plea to 
uploaders.  Please, if you are going to tar a set of files (foo.1 foo.2 foo.3 
...) in directory bar, cd to the next directory up (cd ..) and run "tar -cf 
bar.tar ./bar".  

If you just stay in bar when you create the tar file, people who ftp the file 
to, say, /var/tmp and then run "tar -xf bar.tar" can get quite confused as to
what was actually in bar.tar if there is a lot of other stuff in /var/tmp.

I've learned to run "tar -tf bar.tar" on new files and "mkdir bar; mv bar.tar 
./bar; cd bar; tar -xf bar.tar" if it wasn't packaged properly, but still
it is a minor hassle that could be avoided.  The network utilities are 
particularly bad in this respect.

Thanks for listening

Charles Meier

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

From: cmay@titan.ucs.umass.edu (CHRISTOPHER M MAY)
Crossposted-To: comp.os.linux.misc
Subject: Re: LOGISCAN V0.0.1 released
Date: 15 May 1994 15:40:15 GMT

Henning Holtschneider (hh@hhdo.ping.de) wrote:
: Hi you, out there!

: There recently was a discussion about a Logitech scanner driver for Linux in
: col.help and now I found this message in a German Fidonet newsgroup:

: [for those of you who don't understand enough: the author has written a
: driver, based on the mustek driver, for the Scanman32 and Scanman+. He has
: uploaded the file to tsx-11.mit.edu as logiscan.tar.gz]

Hi, I have a scanman32 and it has just been collecting dust.  I was
wondering if you knew a more exact location for this file, as I could 
not find it on tsx-11, and 'archie logiscan.tar.gz' found no matches.

Thanks for the info, though, it's good to know that at least there's
something out there for my logitech handheld scanner.
 
-Chris May, Computer Science, University of MA, Amherst
-       Technical Assistant, P.C. Maintenance Lab


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

From: bchin@is-next.umd.edu (Bill Chin)
Crossposted-To: comp.os.linux.admin,comp.os.linux.misc
Subject: Re: NeXT like voice mail ?
Date: 15 May 1994 16:00:04 GMT

Shahid Ikram Butt (sib1@Ra.MsState.Edu) wrote:

: I have not installed Linux on my machine yet, I am looking to do that 
: real soon. A quick question. Is anybody working on NeXT-like voice 
: mail system for Linux ? I like the ability to send/receive voice mail on my
: computer.  Just about everybody has sounblaster or compatible card and
: it shouldn't be difficult to stnadardize something like that. 

: If we could get voice messages and gifs in the sig file of a voice message
: we'd very close to video-conferencing without wasting all that bandwidth.
: The data could also be automatically compressed/decompressed 
: uudecoded/uuencoded etc ? 

You are talking about MIME, an Internet standard for sending "richer"
messages than straight ascii.  You can send graphics, binaries, sounds
and others using MIME compatible mailers on many platforms including
Linux. 

: How easy or difficult something like this should be to write ? Does something 
: like this already exist for Linux ? I haven't noticed in any Linux users' sigs
: as I do in NeXT owners' sigs. Imagine putting "Linux mail accepted here" next
: to your email address. 

NeXTmail was developed way before there was an Internet standard to exchange
"rich" messages.  In some ways, NeXTmail is still superior, but not
in one of the most important - there are only a few people that can send/
receive NeXTmail.  Even NeXT is building in MIME support into their
bundled mail app, and there are several commercial and freeware solutions
to send/receive MIME under NEXTSTEP.  Also, if one wanted to be able
to handle NeXTmail, it isn't *that* hard.  NeXTmail is nothing more
than an uuencoded Rich Text Format document with simple extensions to handle
attachments. However, the demand probably just isn't there and one still
has to deal with funky things like character set mapping and font
compatibility (usually X11 <-> Display Postscript).  There is little reason
to create yet another non-standard e-mail message protocol.

.Bill Chin
bchin@andi.org


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

From: news@edus.oau.org (Bob Perkins)
Subject: Driver for smart card
Date: Sun, 15 May 1994 13:46:47 GMT

I would like to have a driver for a smart card for my linux box.
If someone will send me a roadmap for a device driver for the
serial ports in the linux kernel, I will do what I can.

I have all the board ports, registers, commands, etc, but am
looking for a map to tie it all together. If this was assembler,
I could have the driver working in a couple of hours, but just
learning sanscrit ( some times known as 'C' ). :-)

e-mail, please: rcp@edus.oau.org
---
rcp@edus.oau.org

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

From: nate@loreli.ftl.fl.us (Sean Puckett)
Subject: SIGHUP - Deep Kernal Guts question!
Date: Fri, 13 May 1994 04:25:16 GMT

We're developing a BBS under Linux 1.0, and we've found that 
many applications, including bash, ksh and perl, do not pass
on the HUP signal from do_tty_hangup (tty_io.c)'s call to 
kill_sl (exit.c).
 
The kernal apparently wants to send just one HUP/CONT pair
only to the session leader, and none to any other processes
on the lost tty.  The session leader (shell) is sucessfully
killed on hangup, every time, but its children are not.
 
We need all active processes attached to a lost tty to get
a proper HUP signal.  We're considering patching tty_io.c
to do what we want, but first we need to know if the kernal 
was designed to act 'broken' on purpose -- that is, if there 
is something else we should be doing besides patching the 
kernal to effect the change we need.
 
If this problem has been addressed already in versions after
1.0, just point us to an ftp site, and we'll solve our own
worries.
 
TIA

--
                       ..:: nate@loreli.ftl.fl.us ::..
             ..:: Sean Puckett - Albino Frog Software, Inc. ::..
          ..:: The Right Reverend Aural Hardly, MSK, BoC, FCoC ::..

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

Crossposted-To: comp.os.linux.misc
From: hh@hhdo.ping.de (Henning Holtschneider)
Subject: LOGISCAN V0.0.1 released
Date: Thu, 12 May 1994 12:57:53 GMT

Hi you, out there!

There recently was a discussion about a Logitech scanner driver for Linux in
col.help and now I found this message in a German Fidonet newsgroup:

[for those of you who don't understand enough: the author has written a
driver, based on the mustek driver, for the Scanman32 and Scanman+. He has
uploaded the file to tsx-11.mit.edu as logiscan.tar.gz]

[ Article crossposted from fido.ger.linux ]
[ Author was Andy Beck ]
[ Posted on Sat, 07 May 1994 12:06:08 +0200 ]

Hallo All!
==========

Ich hab' mir in den letzten Tagen die Arbeit gemacht,einen Scannertreiber
fuer den Logitech SCANMAN+ (Scanman32) zu schreiben ...

Das Ding sollte auch mit dem aelteren SCANMAN und ggf. (teilweise) einem 
neueren Typ,den die Logi-Software als "NCRSCAN" identifiziert,laufen.

Es handelt sich um ein Kernel-Loadable-Module,es werden also die modutils
benoetigt.Ein Demo-Programm [fuer svgalib] ist auch dabei.

Ich hab' das Teil auf tsx.mit.edu hochgeladen,wo es demnaechst wahrscheinlich
unter ALPHA/scanner/logiscan.tar.gz o.ae. zur Verfuegung stehen wird.

Das Ding basiert zu grossen Teilen (Geruest und ioctl-Schnittstelle) auf dem 
M105-Treiber von Thomas Faehnle (Thomas.Faehnle@uni-ulm.de). Software,die 
fuer diesen Treiber geschrieben wurde,und keine speziellen Faehigkeiten des
Mustek benutzt und die Zeilenbreite korrekt via IOCTL holt,sollte ohne
weitere Veraenderungen mit dem Logi-Treiber laufen.

Leider ist es mir unmoeglich, Thomas zu erreichen (der Mailserver seiner Uni 
sagt 'Addresse unknown'),um mich mit ihm ueber die weitere Entwicklung der 
Scanner-Ioctl-Schnittstelle,etc. zu unterhalten.
Sollte jemand von euch in der Lage sein,ihn zu erreichen oder ihn gar kennen,
so moege er ihn bitten,mich 'mal unter <becka@hp.rz.uni-duesseldorf.de> 
oder via Fido anzumailen.
Ich waere auch schon gluecklich ueber eine funktionierende Mail-Adresse.

CU,Andy

... Line noise provided by German Telekom !
-- 
      Henning Holtschneider * Bauernkamp 41 * 44339 Dortmund * Germany
          Member of OASE - European Shareware Authors Organization
  Internet: hh@hhdo.ping.de  Fidonet: 2:2444/1099  Telefax: +49 231 7285296

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

Crossposted-To: comp.os.linux.admin,comp.os.linux.misc
From: kettler@CS.ColoState.EDU (neal kettler)
Subject: Re: Zeos Pantera OK ?
Date: Sun, 15 May 1994 19:17:29 GMT

In article <sib1.768979718@isis.msstate.edu>,
Shahid Ikram Butt <sib1@Ra.MsState.Edu> wrote:
>
>I am considering buying a Zeos Pantera machine with 486DX2-66, 16MB RAM,
>3PCI 5 ISA slots, on board Adaptec 6360 SCSI controller chip and a Diamond
>Speedstar 64 PCI (1meg) video card.
>
>I am wondering if someone out there has the same or a similiar machine 
>running Linux ?  I would hate to buy the machine and find out I made a 
>mistake. Any help will be greatly appreciated. 
>
>Thanks a million.
>
>
>Shahid
>   ______________________sib1@Ra.Msstate.Edu____________________________
>                  Cruising Information Super Highway
>                  WWW : http://www.msstate.edu/~sib1
        Just one problem with that setup... The Diamond Speedstar. Diamond
doesn't like to make the programming specs freely available, so that
card may not work with X. The diamond stealth won't work, but I'm not
sure about the Speedstar 64. My reccomendation is to play it safe
and get something based on the ATI Mach 32 chip. Good luck.

-- 
========== Neal Kettler < kettler@beethoven.cs.colostate.edu > ==========
The following sentence is false.
The preceding sentence is true.
=========================================================================

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

From: jmt5@Ra.MsState.Edu (John M. Tippitt)
Subject: Adaptec VL SCSI
Date: 13 May 1994 01:46:50 -0500


Does anyone know of any drivers for an Adaptec AHA-2842VL SCSI adapters?
Or plans to write one?
Better yet, will it work with the existing Adaptec drivers?

thanks,
john m. tippitt

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

Crossposted-To: comp.os.os2.programmer.misc,comp.os.minix,comp.os.mach
From: bbs.ruscal@tsoft.net (Russel Mar)
Subject: Re: CD-ROM's do all read and write now?
Date: Wed, 11 May 1994 22:00:34 GMT

Gess Shankar <gess@knex.mind.org> writes:

> rme81977@zach.fit.edu (Brian Fritz/ ADVISOR Russell (MAE)) writes:
> 
> > Can someone tell me if all CD drives read and write now?  Which ones
> > are the "top of the line"?
> > Thanks
> > -BRian
> 
> All CD-ROM drives only READ. Hence the ROM moniker. CD-Recordables
> are available, however. You use a recordable media (same form factor
> as a CD-ROM) and premastered images can then be written (say in
> ISO-9660 format) to the CD-R media. This can be read by all CD-ROM
> drives.
> 
> While prices for CD-R recorders are coming down (Quad speed Yamaha
> around $4000 e.g.), it is not a piece of equipment that you use for
> reading and writing as in a WORM or hard disk. While they are good
> for writing, you need to prestage everything and write a whole track
> at-once. And the READ performance is abominable. Archiving and producing
> CD-ROM masters/small production runs are primary applications.
> 
> I have not seen any support from any premastering software vendors for
> OS/2. I am working on one.
++++++++++++++
Be aware that CERT has posted a warning that a virus has been posted to 
various bbs's that advertises to be a CHINON CD-ROM driver that does the 
ol'scramble-the-hard-drive trick.  The driver sucks the user in by making 
the claim that it will make CHINON CD-ROM drives read/writable.

Posted to alt.security or comp.sucurity.anoounce a few days ago.


======================================================================
Primary addr:   bbs.ruscal@tsoft.net (UUCP apple.com!tsoft!bbs.ruscal)
                Tsoft Public Access Unix, Mt. View, CA 415/428-0190
Secondary addr: ruscal@west.darkside.com (Home of Waffle BBS Software)

"The last run of Twinkies was 1978; If you want'm fresh, wait'l 1996."

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

From: stevek@panix.com (Steve Kann)
Subject: Stupid Question (maybe) PCI?
Date: 13 May 1994 02:55:31 -0400

I know this is probably a FAQ by now, but is there any support by the
kernel for PCI busses?

I am not sure if support is even necessary, as I don't really know about
low-level bussing protocols, but we (at school) are planning on testing
out a couple of machines to run linux on, and don't want to waste money
on PCI machines just to test find out it won't work.

Thanks for any information, or pointers to any information!

- Steve
Cooper Union, NY Ny


--
- Steve

stevek@panix.com
stevek@magnum.cooper.edu



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

From: weule@hp817s.rz.uni-duesseldorf.de (Joerg Weule)
Subject: workaround to dip 3.3.7 found
Date: 15 May 1994 22:14:03 +0200

Hi,

when I want to use dip a second time, dip blocks while opening the line.
I had to reboot my computer to call dip a second time.

Now I have found a work around this problem: (dip-3.3.7)

    1. open the line with O_NDELAY
    2. change the flow with fcntl to blocking mode

Now I can call dip a second time and it works fine. Can I understand
this trick ???

Thanks in advance and have fun with dip

J"org

===============================================================================
J"org Weule                  Mathematisches Institut, Abteilung f"ur Informatik
Raum: 25.13.00.36                                   Heinrich-Heine-Universit"at
Tel.: +49-211-311-3716, priv:+49-211-751409              Universit"atsstra"se 1
weule@cs.uni-duesseldorf.de                                 D-40225 D"usseldorf
-- 
J"org Weule                  Mathematisches Institut, Abteilung f"ur Informatik
Raum: 25.13.00.36                                   Heinrich-Heine-Universit"at
Tel.: +49-211-311-3716, priv:+49-211-751409              Universit"atsstra"se 1
weule@cs.uni-duesseldorf.de                                 D-40225 D"usseldorf

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

From: cdb@tph116.fkp.physik.th-darmstadt.de (Claus-Dieter Bredl)
Subject: Re: Terminator-Power Question
Date: 13 May 1994 07:16:22 GMT

Rob Janssen (rob@pe1chl.ampr.org) wrote:
: In <1994May12.031528.958@prism.demon.co.uk> david@prism.demon.co.uk (David Metcalfe) writes:

: >Drew Eckhardt (drew@kinglear.cs.colorado.edu) wrote:
: >: If the device(s) driving TERMPWR can't source enough current, Bad Things(tm)
: >: happen, and if all devices providing TERMPWR get turned off (I've seen it 
: >: happen), Worse Things(tm) happen, like devices see an asserted RST signal, 
: >: etc.

: >: So, I allways jumper all devices to provide termination power.

: >It s interesting you say that because I recently added a new DEC SCSI
: >disk to my system.  The manual recommended enabling TERMPWR so I did.
: >However, it interfered with my WANGTEK 5250 external tape drive that
: >was also sitting on the bus, so I had to disable TERMPWR on the DEC
: >drive.

: Interfered in what way?
: Did the tape drive short-circuit the terminator power?

: (if it is not that, I really cannot imagine how supplying terminator
: power can interfere with anything)

Terminator power is a distinct line on the SCSI bus, fed with +5V from
an arbitrary SCSI device. It must be provided from at least one source,
of course. BUT BEWARE: If there are no precautions, that +5V may end up
reverse-feeding SCSI devices having their local power supplies turned off...
So, it depends on your individual SCSI devices whether you might face
such a situation as a result of more than one device jumpered to provide
terminator power... (My installation is setup to provide that power
via its ADAPTEC 1542B controller only, since I regularly remove power from
an external DAT (gathering dust the other way): no problems so far, the
internal SCSI disk is as happy as before.

(my opinion, not to be taken as recommendation)

CDB

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

From: whitener@dfw.paranet.com (Rob Whitener)
Subject: Q: How do I find the PCMCIA TCIC base address on my notebook?
Date: Fri, 13 May 1994 16:06:41 GMT

I have a CompuAdd 425TX notebook with a PCMCIA 2.0 slot.  By looking at the
DOS PCMCIA drivers load up, I deduct that I have a Databook TCIC-2/N 
contoller.  I am attempting to use Donald Becker's 3c589 code, which has
a socket service routine that expects the TCIC to be at address 0x240. 
Mine is not at this address.

Any one have an idea of how I can find the address?

I called CompuAdd support, but could not find anyone knowledgeable enough to
answer the question.

Please reply by email. Thanks in advance.

Rob Whitener
whitener@dfw.paranet.com

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


** FOR YOUR REFERENCE **

The service address, to which questions about the list itself and requests
to be added to or deleted from it should be directed, is:

    Internet: Linux-Development-Request@NEWS-DIGESTS.MIT.EDU

You can send mail to the entire list (and comp.os.linux.development) via:

    Internet: Linux-Development@NEWS-DIGESTS.MIT.EDU

Linux may be obtained via one of these FTP sites:
    nic.funet.fi				pub/OS/Linux
    tsx-11.mit.edu				pub/linux
    sunsite.unc.edu				pub/Linux

End of Linux-Development Digest
******************************
