Subject: Linux-Development Digest #996
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:     Sat, 6 Aug 94 22:13:05 EDT

Linux-Development Digest #996, Volume #1          Sat, 6 Aug 94 22:13:05 EDT

Contents:
  Re: How to do inb from a user program? (Grant Edwards)
  Some serious problems with 1.1.39 (floppy and CD mounting) (Steve DuChene)
  Re: PLIP problems solved! (mostly...) (Johannes Stille)
  Re: S3 support at the console (Tim)
  Re: Adaptec AHA 2940 & S3964 (Rob Janssen)
  Re: How to do inb from a user program? (Matthias Urlichs)
  Re: Enhanced IDE?? (Mark Lord)
  Adaptec vs. 1.1.37+ (Clayton Haapala)
  Re: Linux-1.1.38 broken for 2 Adaptor system [FIX] (Eric Youngdale)
  Backspace behavior (Sam Oscar Lantinga)
  Linux, X, DOSemu versions... (Bibek Sahu)
  Re: Adaptec vs. 1.1.37+ (R. Schalk)
  Re: Interesting idea for lilo developers (Matthias Urlichs)
  Re: SLIP, CSLIP, PPP and modems (Dominik Kubla)
  Re: -= good programmer's editor for X? (Dominik Kubla)
  Re: SLIP, CSLIP, PPP and modems (Andrew R. Tefft)
  Re: File Locking (David Ferovick)

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

From: grante@reddwarf.rosemount.com (Grant Edwards)
Subject: Re: How to do inb from a user program?
Date: Thu, 4 Aug 1994 22:06:54 GMT

Richard Roth (rothr@iia.org) wrote:
: Thomas E Zerucha (zerucha@shell.portal.com) wrote:

: Check out the ioperm(port, count, ???) function call. I forgot to doc the 
: two params but (port, 1, 1) will allow inb()/outb() to that port.

: The headers you need are: 
:       fcntl.h (maybe)
:       asm/io.h
:       linux/hdreg.h   (maybe)
:       linux/fs.h (maybe)

: Sorry for the maybe's but this is from a code segment.

You've also got to remember to turn on optimization when compiling,
otherwise it won't work.  The functions inb and outb are inline
funcitons defined in asm/io.h.  Unless optimization is on the inline
definitions will be treated as extern declarations, and you'll get
complaints about inb and outb being unresolved.

--
Grant Edwards                                 |Yow!  INSIDE, I have the same
Rosemount Inc.                                |personality disorder as LUCY
                                              |RICARDO!!
grante@rosemount.com                          |

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

From: s0017210@cc.ysu.edu (Steve DuChene)
Subject: Some serious problems with 1.1.39 (floppy and CD mounting)
Date: 6 Aug 1994 05:23:40 GMT

        I have the following system:    Gateway 2000 486DX4-100
                                        w/ PCI bus
                                        540Mb Western Digital hard drive
                                        Sony CDU-33A CD-ROM drive
                                        ATI Graphics Ultra Pro video card

        I have been trying since 1.1.24 to get a kernel that will let me mount
        a iso9660 CD in my CD-ROM drive. I tried 1.1.24 first, then tried using 
        the new ALPHA CDU-33A driver as a loadable module. No luck, so I upgraded
        to 1.1.34 and tried the Sony CDU driver there, same result. I just got done
        compiling 1.1.39 and I am including the errors as the were output after I
        tried twice to mount one of the CD's that I have (these work fine with a 
        1.0 kernel):

        jaguar:~# mount -t iso9660 -o ro /dev/cdrom /mnt/cdrom/
        mount: /dev/cdrom: can't read superblock
        jaguar:~# mount -t iso9660 -o ro /dev/cdrom /mnt/cdrom/
        general protection: 0000
        EIP:    0010:00174e2c
        EFLAGS: 00010212
        eax: 00c3ed7c   ebx: 001a326c   ecx: 00000000   edx: 00000000
        esi: 00190118   edi: 00000000   ebp: 00000002   esp: 00c32e20
        ds: 0018   es: 0018   fs: 002b   gs: 002b   ss: 0018
        Process mount (pid: 74, process nr: 18, stackpage=00c32000)
        Stack: 70000000 00000000 00000000 00007000 00000000
        Code: c3 90 90 90 9d 98 0a 19 00 00 75 4a
        Segmentation fault

-- 
| Steven A. DuChene   sduchene@cis.ysu.edu  or  s0017210@cc.ysu.edu      
| Youngstown State University  | Computer Science / Math / Mech. Eng.
|They all laughed at Albert Einstein. They all laughed at Columbus. 
|Unfortunately, they also all laughed at Bozo the Clown. 

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

Crossposted-To: comp.os.linux.help
From: johannes@titan.westfalen.de (Johannes Stille)
Subject: Re: PLIP problems solved! (mostly...)
Date: Sat, 6 Aug 1994 18:17:35 GMT

In article <CtuwxL.D5t@utnetw.utoledo.edu> ddelsig@uoft02.utoledo.edu writes:
>Hi all,
>
>     Good news!  I finally got PLIP working on my machine.  I'm posting this
>mostly for the sake of the few other guys who futilely posted for help with
>me on this subject.  Here are my experiences:
[...]
>I had a great deal of confusion about what the exact pin configuration of the
>cable was supposed to be.  Some people said to connect 16, some said not to. 
>everybody seemed to say that the diagram that they had was the standard
>laplink cable.  Here's what I've gotten to work under 1.0.9.  I found it
>using veronica, in the file plip.doc.  I'm including the whole document for
>the sake of others who want to be further confused.
>
>\begin{plip.doc}
>
>  This is the wiring diagram for the cable necessary to link two
>  machines together by their parallel ports.  This is the same wiring
>  diagram used by a Turbo LapLink cable, available for sale from many
>  PC parts outlets.
>
>  1 - 1
>  2 - 15
>  3 - 13
>  4 - 12
>  5 - 10
>  6 - 11
>  7 nc
>  8 nc
>  9 nc
>  10 - 5
>  11 - 6
>  12 - 4
>  13 - 3
>  14 - 14
>  15 - 2
>  16 - 16
>  17 - 17
>  18 nc
>  19 nc
>  20 nc
>  21 nc
>  22 nc
>  23 nc
>  24 nc
>  25 - 25 (ground)
>
>\end{plip.doc}

This might be a standard LapLink cable.

But the connections 1-1, 14-14, 16-16, and 17-17 are not needed for
PLIP. They connect output lines on both sides, BTW.

[...]
>
>My last problem was with my laptop (a Compaq Concerto).  The way it was 
>configured was that it had one parallel port on IRQ 7 at address 3BC.  Linux
>recognized the fact that it was on 3BC, but assigned IRQ 5 to it, which is 
>probably the standard IRQ for that address.  Luckily, I was able to fiddle 
[...]

No, the standard IRQ for I/O address 0x3BC is IRQ 7. Complain to Alan
Cox about the bad default configuration in Linux. The Linux
configuration can be changed in the drivers/net/Space.c file.

IRQ 7 is the hardware default for both the 0x3BC and the 0x378 I/O
address. This means that if you have printer ports at both addresses,
you have to check and possibly to reconfigure your hardware. The default
IRQ for I/O address 0x278 is IRQ 5.

>with the configurations of my laptop via config utilities that came with
>the thing, and was able to set the IRQ to 7 and the address to 378, which
>is the standard place for plip1, lp1, and LPT1.  I determined that this was

Please don't use the DOS/BIOS device names LPT* in this context. The
BIOS looks for printer ports at 0x3BC, 0x378, and 0x278 in this order
and assigns numbers to the ports found. Linux has a fixed assignment of
lp0/plip0 = 0x3BC, lp1/plip1 = 0x378, lp2/plip2 = 0x278.
This means there is no 1:1 relation between Linux and DOS printer port
names.

        Johannes

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

Subject: Re: S3 support at the console
From: tims@kcbbs.gen.nz (Tim)
Date: 6 Aug 94 11:50:46 GMT

Christopher M. May (cmay@titan.ucs.umass.edu) wrote:
: I'm also interested in finding register/programming info for
: the S3 805.  Should I contact the manufacturer of my card or
: S3?  Or should I just look at the XFree Accel server code?

Get vgadoc3.zip - this has text files on registers for practically all SVGA
cards including S3. Archie should turn it up somewhere.


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

From: rob@pe1chl.ampr.org (Rob Janssen)
Subject: Re: Adaptec AHA 2940 & S3964
Reply-To: pe1chl@rabo.nl
Date: Sat, 6 Aug 1994 21:23:43 GMT

In <3209op$9ov@news.cs.tu-berlin.de> theboss@cs.tu-berlin.de (Alexander Nareyek) writes:

>Hi,

>I'm looking for some drivers:

>- Adaptec AHA 2940  (SCSI-PCI)
>- S3-964  (miroCrystal 20SV PCI)

>Are there any (Alpha, Beta...), if, where?
>If not, are they under construction? When
>will they be available?

Check the Projects-FAQ.
It is available in comp.os.linux.announce

Rob
-- 
=========================================================================
| Rob Janssen                | AMPRnet:   rob@pe1chl.ampr.org           |
| e-mail: pe1chl@rabo.nl     | AX.25 BBS: PE1CHL@PI8UTR.#UTR.NLD.EU     |
=========================================================================

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

From: urlichs@smurf.noris.de (Matthias Urlichs)
Subject: Re: How to do inb from a user program?
Date: 6 Aug 1994 21:58:20 +0200

In comp.os.linux.development, article <1994Aug4.220654.7481@rosevax.rosemount.com>,
  grante@reddwarf.rosemount.com (Grant Edwards) writes:
> 
> You've also got to remember to turn on optimization when compiling,
> otherwise it won't work.  The functions inb and outb are inline
> funcitons defined in asm/io.h.  Unless optimization is on the inline
> definitions will be treated as extern declarations, and you'll get
> complaints about inb and outb being unresolved.

Which is a mystery to me -- why aren't they declared "static inline"
instead of "extern inline"? The net effect with optimization should be
exactly the same, but without optimization, the "static inline" version
compiles and links correctly.

-- 
FRANK'S BUREAUCRATIC REWRUTE:
Old MacDonald had an agricultural real estate tax abatement.
-- 
Matthias Urlichs        \ XLink-POP N|rnberg  | EMail: urlichs@smurf.noris.de
Schleiermacherstra_e 12  \  Unix+Linux+Mac    | Phone: ...please use email.
90491 N|rnberg (Germany)  \   Consulting+Networking+Programming+etc'ing     42

Click <A HREF="http://smurf.noris.de/~urlichs/finger">here</A>.

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

From: mlord@bnr.ca (Mark Lord)
Crossposted-To: comp.os.linux.admin
Subject: Re: Enhanced IDE??
Date: 4 Aug 1994 22:33:08 GMT

In article <Ctz9sM.48M@haapi.mn.org> clay@haapi.mn.org writes:
>
>Just for an example, how would you handle this Seagate drive of mine?
>
>The physical drive params are: 1057 cylinder, 16 heads, 63 sectors/track.
>The BIOS, of course, would like to limit that to 1024 cyls.

If your BIOS will allow you to pretend the drive is 528/32/63
instead of 1057/16/63, then that might be your best bet.  Linux 1.1.37+
copes with this as automatically as many modern BIOSs, and you'll never
have to fuss about the 1024cyl limitation in the partition table, LILO,
or DOS.

>Because it physically doesn't have more than 16 heads, I would like to see
>it work like this:
>
>       Set BIOS to 1024 cyls/16 heads/63 sectors
>
>       When Linux boots (with kernel 1.1.37 or higher), the actual drive
>       parms are queried and stored, and then used in programs such as
>       linux fdisk.

Sure, but trickier.  Use LILO command-line parameters to the kernel,
to pass the true geometry.  Just ignore the linux fdisk messages, so
long as your last partition starts below 1024.  Also, ensure that your
linux boot partition is entirely below 1024 (for LILO).
-- 
mlord@bnr.ca    Mark Lord       BNR Ottawa,Canada       613-763-7482

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

From: clay@haapi.mn.org (Clayton Haapala)
Subject: Adaptec vs. 1.1.37+
Date: Fri, 5 Aug 1994 18:59:27 GMT

I have a very similar setup (Adaptec 1542B set to defaults) and run with
no problems up to and including 1.1.39.  BUT -- I have only a single
SCSI device.  Could that be a difference?
-- 
Clay Haapala                    "Well, there was the process of sitting around
clay@haapi.mn.org                and wishing I had more computer stuff."
                                        -- Dilbert

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

From: ericy@cais.cais.com (Eric Youngdale)
Subject: Re: Linux-1.1.38 broken for 2 Adaptor system [FIX]
Date: 5 Aug 1994 22:03:56 GMT

In article <Cu04tL.6EK@park.uvsc.edu>, Matt Day <mday@park.uvsc.edu> wrote:
>Here's a message I just sent to the Linux KERNEL mailing list which
>contains a fix for the problem you're seeing:
>
>1.1.38 crashes with "scsi_devices corrupt (sd)" on my system.  I have an
>Adaptec 1542C, a couple SCSI hard disks and a SCSI CD-ROM drive.  I
>tracked down the problem to the `rscsi_disks' memory not being
>initialized to 0, which causes sd_attach() to think that some or all of

        Thanks Matt!  I missed that one.  For everyone else, sorry for the
inconvenience.  As many of you know, I am working on changing things so
that low-level scsi drivers can be loaded as loadable modules.  The worst
is over now, I think, at least as far as wholesale changes like this.  

        On my system at home, I am now able to load the in2000 and
scsi_debug drivers as loadable modules, but all it does not is call the
detect routine (they do not yet call scan_scsis yet).  In other words, it
is just a curiosity at the moment, and it will be a while before you will
be able to do anything useful.  Unforunately I had to modify the
Scsi_Host_Template structure once again - I forgot to add an (int*) to be
used for the reference count for the loadable modules. 

-Eric


-- 
"The woods are lovely, dark and deep.  But I have promises to keep,
And lines to code before I sleep, And lines to code before I sleep."

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

From: slouken@cs.ucdavis.edu (Sam Oscar Lantinga)
Subject: Backspace behavior
Date: Sat, 6 Aug 1994 05:31:50 GMT


        Why is it that backspace on the virtual console
echoes backslash and then the character erased?  Is there any
reason not to use "\b \b" to erase the character from the 
screen?  

        It becomes very frustrating when you are backspacing
over a pathname and you have to calculate the number of backspace
keys you need to press, and then read your text backwards to
verify that you have backed over the proper characters.

My .02 cents. :)

Happy birthday to me, happy birthday to me!  *grin*


        -Sam            (Born August 7th, 1974)



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

Date: 06 Aug 1994 13:07:52 -0000
From: leviathan@commonlink.com (Bibek Sahu)
Reply-To: leviathan@commonlink.com
Subject: Linux, X, DOSemu versions...

        I realize that the Linux versions change almost every day, and I saw a post
in the announce area that prompted me to ask about X, & since I have a lot of
DOS stuff, I'm asking about DOSemu...
        But anyway...
        What is the latest version of the Linux kernel (as of your reply)?  What is
the latest [relatively] STABLE version?
        Has X11R6 been ported to Linux yet?
        What is the latest version of DOSemu?

        Thank you for any and all help.  I feel kind of stupid asking all this, but
I haven't used Linux in a while due to problems.  I ordered the 2 CD set from
InfoMagic, and will do a complete reinstall when they arrive.  Again, Thanks!

                                        - Bibek


"The road to truth is long, and lined the entire way with annoying bastards"


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

Date: Sat, 6 Aug 1994 22:20:50 +0200
From: U001295@vm.uci.kun.nl (R. Schalk)
Subject: Re: Adaptec vs. 1.1.37+

In article <Cu2sr4.1BF@haapi.mn.org>
clay@haapi.mn.org (Clayton Haapala) writes:

>
>I have a very similar setup (Adaptec 1542B set to defaults) and run with
>no problems up to and including 1.1.39.  BUT -- I have only a single
>SCSI device.  Could that be a difference?

Well I run 1.1.39 on an adaptec 1542b with:
2 SCSI-II disks
1 SCSI-I disk
1 SCSI-I Tape
1 SCSI-I cdrom
1 SCSI-II cdrom
and don't have a single problem. Well, maybe there are slightly different
controllers around? Or maybe it's the devices hanging on it?
Ronald

********************************************************************
* ing. Ronald Schalk, afdeling CS, sectie COOS                     *
* Universitair Centrum Informatievoorziening (UCI)                 *
* University of Nijmegen (KUN)    snailmail: Geert Grooteplein 41  *
* e-mail : R.Schalk@uci.kun.nl               6525 GA Nijmegen      *
********************************************************************

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

From: urlichs@smurf.noris.de (Matthias Urlichs)
Subject: Re: Interesting idea for lilo developers
Date: 6 Aug 1994 22:17:59 +0200

In comp.os.linux.development, article <31t0lo$jq6@news.u.washington.edu>,
  tzs@u.washington.edu (Tim Smith) writes:
> 
> If "loadlin" does what it sounds like it does, what is the need
> for boot managers, anyway?  Note that MS-DOS 6.xx allows you to
> have multiple configurations, which you can easily select from
> a menu when config.sys is processed.

The big problem with this is that MS-DOG isn't free. 

-- 
All experience is an arch wherethrough gleams that untravelled world.
                        -Tennyson
-- 
Matthias Urlichs        \ XLink-POP N|rnberg  | EMail: urlichs@smurf.noris.de
Schleiermacherstra_e 12  \  Unix+Linux+Mac    | Phone: ...please use email.
90491 N|rnberg (Germany)  \   Consulting+Networking+Programming+etc'ing     42

Click <A HREF="http://smurf.noris.de/~urlichs/finger">here</A>.

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

From: kubla@goofy.zdv.Uni-Mainz.DE (Dominik Kubla)
Subject: Re: SLIP, CSLIP, PPP and modems
Date: 06 Aug 1994 20:51:27 GMT


I have some questions regarding modems too:

1. How do i create a SLIP router, that is a Linux-PC that handles incoming
   SLIP connections only as ROUTER to the ethernet. We enabled proxy-arp
   for the relevant IP numbers but the Linux machine was not routing ...
   Am i missing something?

2. How to hangup the phone line on an incoming call without closing the
   device? I need this for a callback solution to prevent mgetty from
   claiming the line again and eventually allowing an incoming call on it.

Any suggestions?

Thanks in advance,
  Dominik Kubla
--
+-------------------------------------------------------------------------+
| eMail: Dominik.Kubla@Uni-Mainz.DE                                       |
| sMail: Dominik Kubla, Lannerstrasse 53, 55270 Ober-Olm, F.R. of Germany |
+-------------------------------------------------------------------------+

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

From: kubla@goofy.zdv.Uni-Mainz.DE (Dominik Kubla)
Subject: Re: -= good programmer's editor for X?
Date: 06 Aug 1994 20:53:25 GMT


Try xcoral, it has a nice user-interface; shell, c and c++-modes and uses
ANSI-C as internal script language (for all of you who hate elisp :-) )

Cheers,
  Dominik
--
+-------------------------------------------------------------------------+
| eMail: Dominik.Kubla@Uni-Mainz.DE                                       |
| sMail: Dominik Kubla, Lannerstrasse 53, 55270 Ober-Olm, F.R. of Germany |
+-------------------------------------------------------------------------+

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

From: teffta@csp195.erie.ge.com (Andrew R. Tefft)
Subject: Re: SLIP, CSLIP, PPP and modems
Date: Sat, 6 Aug 1994 20:24:58 GMT

In article <1994Aug5.210700.15292@brtph560.bnr.ca>,
Jay Denebeim P025 <denebeim@bnr.ca> wrote:
>
>Unlike xmodem or zmodem or whatever, SLIP/PPP are 'real' networking
>protocols. That means that you can be doing several things on the wire
>at the same time, FTPing, Telnetting, etc.  There's only so much
>bandwidth though, so the more things you do the slower they go.

I think the real answer to this whole question is related to that.
If you watch while you're doing an ftp transfer over ppp, a lot of
acks go from the receiver to the sender. If you watch an interactive
session while you are also ftp'ing, you will see the delays caused
by the queued up packets of the ftp transfer. Put those together
and you might guess that the queued up packets from a transfer
in one direction will delay the acks from the other direction.

It's then a question of timing, not bandwidth. A dedicated bidirectional
protocol can either a) be streaming, where timing of acks is not
important, or b) properly time its acks so that they are not delayed
by the data going the other direction.
-- 

Andy Tefft               - new, expanded .sig -     teffta@erie.ge.com

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

From: davidf@austin.ibm.com (David Ferovick)
Subject: Re: File Locking
Date: Sat, 6 Aug 1994 22:21:27 GMT

In article <31uq6i$65g@urchin.umiacs.umd.edu> kashmir@umiacs.umd.edu (Mike Grupenhoff) writes:
>dillon@apollo.west.oic.com (Matthew Dillon) wrote:
>>
>>    Why not just use flock() ?
>>
>>    I dunno about it working over NFS tho, don't think it does.
>>
>>                                      -Matt
>
>The point of using lockf/fcntl is that it does work over nfs, although 
>it suffers from many race conditions.  I think the main reason linux lacks
>lockf is because there is no port of statd/lockd which handle the lock 
>requests.
>

This port will be in progress starting the 19th of this month... I have all
the information collected thanks to several people who pointed out sources
for various pieces of information on NFS and NFS file locking.  

Dave Ferovick
(dave@paris.utsa.edu)

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


** 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
******************************
