Subject: Linux-Development Digest #42
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:     Fri, 19 Aug 94 03:13:09 EDT

Linux-Development Digest #42, Volume #2          Fri, 19 Aug 94 03:13:09 EDT

Contents:
  ISO Development Environment for Linux? (David T. Margrave)
  Re: Howto write Device Drivers. (Jonathan Magid)
  Re: Howto write Device Drivers. (Randy Hootman)
  1.1.45: cannot login (Joerg Wedeck)
  Re: Socket Timeouts in Linux 1.1 (Rob Janssen)
  Thoughts on /proc/irq, etc... (Archie Cobbs)
  Re: Ejecting of CD on umount (Daniel Quinlan)
  Re: Looking for color ls source. (Orhan Unal)
  Re: too many kernel patches? (Daniel Quinlan)
  Re: setting textmode from program (Frank Lofaro)
  Re: Suggest:SCSI Tape File System (H. Peter Anvin)
  Re: Future of Linux (H. Peter Anvin)
  Report on SVNET: Bill Jolitz's Talk; My brain hurts, too. (Jordan K. Hubbard)

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

From: dmargrav@clark.net (David T. Margrave)
Subject: ISO Development Environment for Linux?
Date: 19 Aug 1994 03:09:02 GMT

Are there any plans to port the ISO Development Environment (X.500) to
Linux?

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

From: jem@bittyblue.oit.unc.edu (Jonathan Magid)
Subject: Re: Howto write Device Drivers.
Date: 18 Aug 1994 17:14:03 GMT

In article <32v7k1$kth@pheidippides.axion.bt.co.uk>,
Mark Baker <mbaker@phal.chilli> wrote:
>I have a frame grabber card and suspect that the simplest way of getting a device
>driver for it is to write one myself.
>
>Problem:- Is there an authoritive text on how to write Linux/Unix device
>drivers. I have designed sbc's before so it should be a managable task. Its
>just that I do not know where to start...

You can read the Linux kernel hackers guide- its on sunsite.unc.edu
in /pub/Linux/docs/LDP/kernel-hackers-guide.

You might also want to correspond with  Steffen W. Schilke (sws@tora.robin.de),
as the Linux Projects-FAQ says that he is working on a driver for
the Metec VDG-8 Framegrabber.

cheers,
jem.
--
jem@sunsite.unc.edu\/sunSITE admin



-- 

<a href="http://sunsite.unc.edu/jem">A sig.</a>

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

From: rph@netcom.com (Randy Hootman)
Subject: Re: Howto write Device Drivers.
Date: Thu, 18 Aug 1994 19:09:16 GMT

Try looking at the HOW-TO's? There is a scsi and a device driver how-to.

Hope this helps.

Randy


Mark Baker (mbaker@phal.chilli) wrote:
: I have a frame grabber card and suspect that the simplest way of getting a device
: driver for it is to write one myself.

: Problem:- Is there an authoritive text on how to write Linux/Unix device
: drivers. I have designed sbc's before so it should be a managable task. Its
: just that I do not know where to start...


: Thanks in advance.

-- 

///////////////////////////////////////////////////////////////////////
     "In recognizing the humanity of our fellow beings,
      we pay ourselves the highest tribute." - Thurgood Marshall
//////////////////////////////////////////////////////////////////////
Randy Hootman                Randysoft Software             (408) 229-0119


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

From: jw@peanuts.informatik.uni-tuebingen.de (Joerg Wedeck)
Subject: 1.1.45: cannot login
Date: 18 Aug 1994 22:04:33 GMT

Hi all,

when I try to let run linux 1.1.45 i run into the problem that I cannot 
login any more, not with getty_ps nor with agetty. getty_ps says something
like: sh: cannot exec binary file /bin/login.
agetty shows up with an exec format error for /bin/login. when booting again
with an older kernel, the file system tells me that the ext2 root files system
of the kernel previously run is ruined, especially many of the tty devices
are then eliminated by e2fsck ...
I have observed this behavior concerning crashing the filesystem since pl 41
(pl 40 seemed to work ok, but i am not absolutely shure), the problems
with login occured sinc pl 42.

My system has a PCI motherboard of Intel, AMI Bios, 486 DX-2/66, IDE
onboard controller, PCI video card spea v7 mercury light, NO PCI scis
controller, but an ISA adaptec 1542 CF SCSI controller.
There is a 545 MB Conner Disk connected
to the IDE adapter, which is also used for starting linux via lilo,
linux itself is completely installed on the IBM scsi disk, which
is connected to the adaptec scsi controller, also a wangtek scsi tape
is connected to that controller.


Joerg

--
==================================+======================================
Joerg Wedeck                      | E-Mail: jw@peanuts.informatik\
Universitaet Tuebingen            |             .uni-tuebingen.de
Lehrstuhl fuer Techn. Informatik  |
Sand 13                           |
72076 Tuebingen                   |
Germany                           |
==================================+======================================
       Deswegen bin ich ja Optio, und Du nur Legionaer :-)
=========================================================================

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

From: rob@pe1chl.ampr.org (Rob Janssen)
Subject: Re: Socket Timeouts in Linux 1.1
Reply-To: pe1chl@rabo.nl
Date: Thu, 18 Aug 1994 21:56:55 GMT

In <3301ed$elu@senator-bedfellow.MIT.EDU> thomasc@ATHENA.MIT.EDU (Thomas Wallace Colthurst) writes:

>I know of two ways of making a socket timeout while 
>attempting to read data, neither of which seems to work
>with Linux.

>Method #1:  setsockopt( skt, SOL_SOCKET, SO_RCVTIMEO,
>                        &tv, sizeof(tv) );
>Problem:  SO_RCVTIMEO isn't defined in sys/socket.h or linux/socket.h

>METHOD #2:  fd_set readfds;  FD_ZERO(readfds);  FD_SET( skt, &readfds );
>            select( 1, &readfds, NULL, NULL, NULL, &tv );
>Problem:  select always waits till the timeout, and then returns 0.


>Am I using either of these methods incorrectly?  Is there another
>method I should be using?

You use select() incorrectly.  The first argument must not be 1, but skt+1.

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

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

From: archie@qin.CS.Berkeley.EDU (Archie Cobbs)
Subject: Thoughts on /proc/irq, etc...
Date: 18 Aug 94 19:59:32 GMT

So now there's /proc/irq, in 1.1.45?

Here's an idea, tell me what you think... have a directory
called /proc/hardware. In it would be any junk that is
hardware specific, for instance,

        /proc/hardware/irq
        /proc/hardware/dma
        /proc/hardware/mmu
        /proc/hardware/cpu-temperature

...or whatever. That way we keep this stuff isolated for the
sake of future portings of linux to other platforms.

But now, is "/proc" really the right name for this directory?
Maybe we should have "/info" and then "/info/proc", "/info/hardware",
and so on... OK, I guess now I'm actually suggesting this. For
backwards compatibility, just "ln -s /info/proc /proc"...

This reminds me of the MIB system used by SNMP to control
network devices, where you have a huge hierarchy of information
and statistics down to very specific detail (even manufacturer
specific statistics). Similar issues involved...

-Archie


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

From: quinlan@freya.yggdrasil.com (Daniel Quinlan)
Crossposted-To: comp.os.linux.help
Subject: Re: Ejecting of CD on umount
Date: 16 Aug 1994 20:19:10 GMT
Reply-To: quinlan@yggdrasil.com


Warwick Ward-Cox <wwar@lostlink.alt.za> writes:

> After looking through the source for the
> /usr/src/linux/drivers/block/sbpcd.c cdrom driver I noticed a define
> for EJECT the line looks like this :
>
> #define EJECT 1 /*.....*/
>
> Now can I just comment this line out to stop the CD auto ejecting on
> umount'ing, as I dont want the system automatically ejecting and
> destrying my CD tray which sits in side a closed case.

Don't comment it out.  I think the proper change is to change the "1"
to a "0".

> How about in future having a question about ejecting the CD in the
> make config process..?

It would make more sense to have an option passable to umount(1) and
a related ioctl in the kernel.

-- 
Daniel Quinlan                         Yggdrasil Computing Inc. (408) 261-6630
(quinlan@yggdrasil.com)                "Free software for the rest of us"

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

From: unal@uwnuc1.physics.wisc.edu (Orhan Unal)
Subject: Re: Looking for color ls source.
Date: 19 Aug 1994 04:09:42 GMT

     Paul Walker wrote in article <CupDp8.G1M@suncad.camosun.bc.ca> :
>
>
>Slackware has a color version 'ls' from the GNU fileutils 3.9.
>
>Does anyone know where the source is?  I wish to port it to a Sun.
>
>
>
>--
>  |)
> /|_    P A U L   W A L K E R
>((|,)   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
> ~|~    paul@malcolm.camosun.bc.ca    COMPTEC92100@CAMOSUN.BC.CA
>  |     
>
>

I don't know if there is anything special about slackware version of color ls,
but Using fileutils-3.9.tar.gz from prep.ai.mit.edu and color-ls-3.9.0.2.patch.gz
from sunsite.unc.edu I was able to compile a color version of ls for Sun Sparc.

----
********************************************************
*  Orhan Unal  *  Email: unal@uwnuc1.physics.wisc.edu  *
********************************************************

-- 
----
********************************************************
*  Orhan Unal  *  Email: unal@uwnuc1.physics.wisc.edu  *
********************************************************

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

From: quinlan@freya.yggdrasil.com (Daniel Quinlan)
Subject: Re: too many kernel patches?
Date: 17 Aug 1994 21:02:08 GMT
Reply-To: quinlan@yggdrasil.com


Johannes Rest <rest@IKP.Uni-Koeln.DE> writes:

> every time i look on the sites with kernel sources i find a new
> patch, but is this really necessary? It seems to me that every
> kernel developer who has a new idea to bring into the kernel
> generates a patch and we're at the next kernel level. After this i
> find several mail complaining about malfunctions in the new kernel,
> which will be obviously patched in the next kernel.  for me this
> doesn't make sense. Any opinions?

Kernel release 1.0 is intended for people who want a stable release.
1.1 carries no guarantees and is intended to be a developers testing
release.

If you can't take the heat, stay out of the kitchen.

Dan

-- 
Daniel Quinlan                         Yggdrasil Computing Inc. (408) 261-6630
(quinlan@yggdrasil.com)                "Free software for the rest of us"

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

From: ftlofaro@unlv.edu (Frank Lofaro)
Subject: Re: setting textmode from program
Date: Mon, 15 Aug 94 22:12:32 GMT

In article <1994Aug15.111231.4126@tudedv.et.tudelft.nl> boekhold@morra.et.tudelft.nl (Maarten Boekhold (Who'd you expect??)) writes:
>Hi,
>
>I was wondering if it is possible to set the textmode of the active VC 
>from a user-program (if necessary setuid root). I guess it must be 
>possible, since lilo (or is it the kernel?) can do it on startup (and 
>under dos you can use 'mode con: lines=' or something like that).
>
>So what I want is to enable the user of the program to change the 
>textmode of the VC from a menu-item or config-file from ('standard') 
>80x25 up to 132x43 or whatever is possible with the video-card.
>

SVGALIB has a utility to restore textmodes from a file. You use the 
kernel to get the console into any desired mode, and use the SVGALIB to 
save the register values in a file.

But we _DO_ need some kernel support. We need a way to tell the console driver 
that the screen size has changed. stty only tells tthe tty driver.

We can leave the actual mode changes out of the kernel itself though.


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

Crossposted-To: comp.os.linux,comp.os.linux.misc
From: hpa@ahab.eecs.nwu.edu (H. Peter Anvin)
Subject: Re: Suggest:SCSI Tape File System
Reply-To: hpa@nwu.edu (H. Peter Anvin)
Date: Tue, 16 Aug 1994 03:06:26 GMT

Followup to:  <32jv0g$e0f@prism.mindware.brisnet.org.au>
By author:    clb@prism.mindware.brisnet.org.au (Chris Burke)
In newsgroup: comp.os.linux
>
> I have been pondering a bit and have worked out a reasonable way to allow
> a mountable file system for high speed tapes. 
>

I think the best way to do this would probably be to do it through a
userfs daemon; userfs permits a user-mode program to run a filesystem
while at the same time have access to all the standard amenities of
user mode (like access to other filesystems, etc).

        /hpa
 
>  BUT   -  Has it been done for LINUX ?
> 
> If not (and I have looked) - this is my proposal : that a modified ext2 
> file system be made with the following attributes :
> (Note Primary is SMALL+HIGH SPEED, Secondary is LARGE+SLOW SPEED)
> 
>       1. Allows the directory entry portion to be stored on a primary 
>            block device to the actual data. Of course there could only
>            be 1 for each. Only entries of a normal data type would be
>            redirected to tape - or alternately and in a totally non-standard
>            fashion implement a new 'offline' bit for files stored on the
>            secondary device. The aim is to ensure directories, devices and
>            named pipes are not redirected to the secondary device.
> 
>       2. Upon initially mounting the device - the entire directory
>            structure is loaded onto the primary block device from the
>            secondary block device.
> 
>         3. Upon unmounting the device - the entire directory structure is
>            stored from the primary block device onto the secondary block
>            device.
> 
>       4. During idle time directory changes on the primary device
>            are reflected to the secondary device.
> 
> (*)Also a block scsi tape device needs to be created to allow "random"
>  access to the tape device. This device is stbn standing for SCSI Tape
>  Block number 'n'.
> 
> The aim of this is as follows :
> 
>    Create for example a ext2new partition as /dev/hda4(primary 4Meg)
>    /dev/stb0(secondary 2Gig) allowing a large file system accessible
>    drive to be created for archival purposes.
> 
> My immediate problems are :
> 1. I only started on Linux 6 weeks ago.
> 2. I only have one drive subsystem (IDE) so testing is impossible.
> 3. I NEED HELP.
> 
> If anyone is interested give me an email, if not I'll still be working
> on it just for the fun of it.
> 
> PS : This idea is modelled on the idea of similar products for DOS which
> store the directory information as a DOS file and the data on either indexed
> floppies or tape.
> 
> --
>  Christopher Burke           \ "Be positive,     \             CUSTOM COMPUTER
>  MindWare, PO Box 1247         \    become         \     SOFTWARE(Dos+Windows)
>  Carindale Q. 4152 AUSTRALIA     \    a proton"      \    and SYSTEMS ANALYSIS
>  clb@prism.mindware.brisnet.org.au \ Eric Lectron('88) \ Phone:+61-(7)-3984000


-- 
INTERNET: hpa@nwu.edu             --- Allah'u'ahba ---
IBM MAIL: I0050052 at IBMMAIL     HAM RADIO:   N9ITP or SM4TKN
FIDONET:  1:115/511 or 1:115/512  STORMNET:    181:294/1 or 181:294/101
Microsoft: The Second Evil IBMpire!

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

From: hpa@ahab.eecs.nwu.edu (H. Peter Anvin)
Subject: Re: Future of Linux
Reply-To: hpa@nwu.edu (H. Peter Anvin)
Date: Tue, 16 Aug 1994 05:23:08 GMT

Followup to:  <cjsCuLI6p.6As@netcom.com>
By author:    cjs@netcom.com (cjs)
In newsgroup: comp.os.linux.development
> 
> But one thing you really have to admit, printing under Unix (any unix) really
> blows. Its even worse then the early days of DOS. But I especually miss
> being able to kick in a printer driver and dump lines, fonts, and pictures 
> to it with C commands.
> 
> But at least with Linux, all you have to do is wrestle lprd.. I hear that SCO
> makes you do triple-summer-salts while hand-cuffed and blind-folded in order
> to get the printer support working.
> 

#define printer_driver  Ghostscript

Now you can dump lines, fonts and pictures with commands (not in C,
but in PostScript).

Seriously.  Using a magic filter and Ghostscript printing on UNIX is
really nice.  I recently wrote a magic filter for in-house use... I'll
probably release it for general consumption when I have gotten around
to writing some docs for it.  :)

        /hpa

-- 
INTERNET: hpa@nwu.edu             --- Allah'u'ahba ---
IBM MAIL: I0050052 at IBMMAIL     HAM RADIO:   N9ITP or SM4TKN
FIDONET:  1:115/511 or 1:115/512  STORMNET:    181:294/1 or 181:294/101
"Life is complex.  It has real and imaginary parts." -- urlichs@smurf.sub.org

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

From: jkh@violet.berkeley.edu (Jordan K. Hubbard)
Crossposted-To: comp.unix.bsd,comp.os.386bsd.development,comp.os.386bsd.bugs,comp.os.386bsd.apps,comp.os.386bsd.questions,comp.os.386bsd.misc,comp.os.os2.programmer.misc,comp.os.minix,comp.os.mach,comp.unix.pc-clone.32bit
Subject: Report on SVNET: Bill Jolitz's Talk; My brain hurts, too.
Date: 19 Aug 1994 04:40:40 GMT

Well, I went to Bill's little talk too, though my views are perhaps not quite
so rose-colored as the esteemed Jesus Monroy's.

After driving the 95 minutes from Concord, CA to Mountain View, I arrived
somewhat stiff but hopeful (I'd never heard Bill talk before).  At last,
the much vaunted and definitely long awaited 386BSD 1.0 was to be unveiled!
What amazing secrets from the labyrinthian depths of Bill's mind might
be revealed?  Had he really rewritten the kernel in Prolog, thus to
guess the user's every desire before even logging in?  Only time would
tell!

Well, in retrospect, I can say that much time passed but very little was
told, and if anything truly rests in the labyrinthian depths of Bill's
mind then he certainly didn't share many of the TECHNICAL DETAILS with us.

I knew we were all in trouble when Bill spent 25 minutes on the first slide,
and it was with a great sinking feeling that I realized there were at least
6 more to go.  The first slide could be deftly summarized as:

        "There are many problems in operating systems.  Solving them the
         right way is a good thing.  386BSD 1.0 solves them the right way.
         386BSD is good."

Exactly HOW this has been accomplished we must sadly remain in the dark
about, but all credit to Bill - if he ever tried to sell used cars like he
tries to sell his operating system (or his book), he'd be a bleedin'
millionaire!  Clearly someone who's missed his calling.

I mean, I really don't mean to be catty, but Bill's ENTIRE multi-hour talk
could be summarized by the following paragraph:

        "Hello!  Blah blah blah blah blah blah blah hype hype hype hype
         hype hype hype hype blah blah blah blah blah god I'm amazingly
         gifted, aren't I? hype hype hype more hype blah blah hyperbole
         marketspeak blah blah merged buffer cache blah blah even more
         hype than you thought possible 417 sentences ago more blah blah
         kernel threads!  heap good! blah blah 386BSD fine stuff! blah
         blah blah stupid joke [polite laughter] hype hype gratuitous
         change hype blah blah blah even more gratuitous changes blah
         blah um um um um uh uh uh blah blah pause for hype from Lynn,
         who's sitting in the the front row and staring in rapt admiration
         at her brain-boffin hubby, blah blah and finally some more hype
         followed by a quick blast at all the interlopers who have shown
         such incredibly bad taste in not following Bill like proper little
         soldiers and have instead done irrepairable damage to the BSD
         community by ignoring the stunning brilliance of his torch."

As my friend said to me in a whispered aside: "Bill never uses 10 words
where 200 will do the job just as well."  I wouldn't know, of course,
having already gone partially deaf by that point.

Anyway, Bill didn't really say anything that actually LOOKING AT THE DAMN
CODE wouldn't have said about 40 times better, and with infinitely greater
accurately but, unfortunately, that same code is not available for comment.
Ever fearful that we might criticise his code if we ever got the chance to
actually see it, Bill's still keeping it tightly under wraps, no matter what
Jesus might say to the contrary.  Me, I'll believe in the emperor's clothes
when they're actually sitting on an ftp site somewhere!  So far, all I've
seen have been some exceedingly grandiose slides and a machine in the corner
that proclaimed to run 1.0 but could have been Linux with a different set of
prompts for all I know! :-)  Judging by its appearance, we may at least rest
assured that Bill has the capability to VIEW DOCUMENTS FROM WINDOWS well in
hand!  [No, I'm not kidding - that was the demo]

Speaking of our friend Jesus, he was there, faithfully filming his idol
and mentor for future posterity.  I now know where Jesus gets his marketing
talents - from Bill!  They're so alike it's scary - I think someone should
go for a genetic sample off those two, just to see how closely they match.
Or maybe not, I guess I'd really rather not know.  In any case, I'd almost
be willing to buy the freakin' video tape just to be able to FAST FORWARD it
to the one or two bits where I actually brought my head up off my chest to
listen to something that Bill said that sounded vaguely interesting.  Not to
say that those features actually exist, since all we have are Bill's
increasingly dubious words to go by, but they might make a possible blueprint
for some more open and public effort someday.  I'm sure Jesus himself will
visit us in the near future to recap the evening and all of these hypothetical
features, though I'm sure I don't need to tell you all to take his words with
a couple of dump trucks full of salt.

Anyway, I came away hoping for some meat and what I came away with was lots
and lots of soggy bread.  I guess I should not have been so surprised,
given Bill's miserable release and vaporvision track record, but I'd sort
of thought that EVENTUALLY, just through sheer weight of peer pressure and
continental drift if nothing else, Bill would finally release his operating
system and discuss the details of its many and varied wonders.  No such
luck, just 2 years worth of accumulated marketspeak and a Real Soon Now
date on the software.  God, I have to sit down - the feelings of Deja' Vu are
simply overwhelming!  Ok, so I'm not being entirely constructive here
and probably shouldn't go so much for Bill's throat.  He can't help who
he is, and were it not for certain...  Environmental factors...  He might
even be a reasonable guy to work with; there's no question that he's an
innovative thinker, perhaps even a brilliant one, but his interpersonal
skills are completely and totally lacking (I won't even mention his
wife - *shudder*!) and he's in love with the sound of his own voice.
The Emperor remains on his throne, but he's buck-naked.

I shall be giving his next talk a wide miss, probably to the great
enjoyment of us both.

                                                Jordan

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


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