Category : UNIX Files
Archive   : INTLX003.ZIP
Filename : INTLX003.TXT

 
Output of file : INTLX003.TXT contained in archive : INTLX003.ZIP


From: [email protected] (Howard Honig)

Subject: DOS and LINUX cohabitating on the same Disc?
Message-ID: <[email protected]>
Date: 12 Feb 92 23:51:02 GMT
Organization: Hewlett-Packard Co., Corvallis, OR, USA
Lines: 17


Hi there,

I'm sure this is a FAQ but it isn't covered in the Linux Information Sheet
so here goes anyway.

Can Linux and DOS coexist on the same IDE drive in different partitions?
If so, how do I boot up or run Linux?

Another question:

How many different partitions are recommended on a hard drive and what
is the recommended size of each of those partitions.

Thanks in advance.

Howard Honig


[next article]
From: [email protected] (Charles Hedrick)

Subject: Re: A better math library - mathlib-2.1.tar.Z
Message-ID:
Date: 13 Feb 92 05:50:49 GMT
References: <[email protected]> <[email protected] ntel.com>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 10
To: [email protected]

>By the way, anybody else looking at pax. I've got it compiled and working.
>It creates multi-volume tar files, but when I try to restore the files,
>I get a corrupted archive error at the end of the first disk. Ah well,
>it works just fine for single disk archives, tar and cpio.

If you're talking about floppy disks, there's a kernel bug. At the end
of the disk, it returns EIO instead of end of file. Could that be
your problem? I believe the fix has been posted. I'm at Rutgers
rather than at home, so I can't give you the fix, but it's an obvious
goof in the block I/O read routine.


[next article]
From: [email protected] (Philip Copeland)

Subject: Mylogin3
Message-ID: <[email protected]>
Date: 13 Feb 92 09:21:42 GMT
Sender: [email protected] (Usenet poster for nntp (tj - 17/12/91))
Reply-To: [email protected]
Organization: Bristol-Polytechnic
Lines: 29
Nntp-Posting-Host: carrot

I've compiled (and linked without fatal errors) the mylogin3 package. This
Works well except i have a minor problem,... it's not running as a gettty
has anyone been able to spawn 'login' from this package (on tsx-11.mit.edu)
onto other tty's.. currently i run it from the /etc/rc file (ok very bad idea)
but i cannot run /bin/doshell /dev/tty1 /bin/login & as this simply throws
login into the back of processing allowing the rc to complete and run a root
shell
On top of this, when i login, the system refuses to let me run another
shell on tty2 (i'm root?). I was going to use the qpl package but found that
it refused to compile (_free_s and _ttyname (yes i did put them in the include)
). Anyway, the grieving aside, has anyone (poe?) found an elegant way around
this?
Just another thing, linux kernel paniced last night.... disk i/o error
in 0302 (hur hur hur, mirth, snigger), I found that sector 6 offset form the
partition had become bad (after running norton's calibrate at 80 int. through
the disk before starting to port linux,... so much for the kernel panic, you
should have seen me at this stage!.. anyway to cut a long story short, i found
that writting sector 6 back from nortons 'de' in maint mode, cured the problem.
the thing is, that I was getting HDtimout errors for a while which manifested
itself -> panic over a 3 day period then the weird linux debug *P=null and
then the system failure (i've and ide, 386sx 16 mhz) luckily sector 6 was blank
in the first place but i doubt that this would be the case everytime.

===============================================================================
Philip Copeland - alias 'Bryce' (course: BScITIS)
JANET : [email protected]
Internet : p_copela%[email protected]
"... I can resist anything but temptation..."
-------------------------------------------------------------------------------


[next article]
From: [email protected] (Jim Winstead Jr.)

Subject: More stuff at TSX-11
Message-ID: <[email protected]>
Date: 21 Feb 92 03:06:08 GMT
Sender: [email protected] (Jim Winstead Jr.)
Organization: Harvey Mudd College, Claremont, CA 91711
Lines: 37

A while back I uploaded a new set of the GNU Text Utilties to TSX-11,
with the fixed sort, so if you've still got the bad sort from the
previous upload available, you will want to at least grab the seperate
sort that has been uploaded. If you're looking for all the neat GNU
text utilities (such as sort, cut, paste, join, head, tail, wc, cmp,
and most of the other basic text filters on unix systems), just grab

tsx-11.mit.edu:pub/linux/binaries/usr.bin/text-1.1.tar.Z

I compiled the GNU bc utility, which is "an arbitrary precision
calculator language," and is part of the POSIX draft standard. It
appears to work fine, but I'm open to criticisms.... Find it as:

tsx-11.mit.edu:pub/linux/binaries/usr.bin/bc.tar.Z

Next, I compiled the GNU fgrep for Linux. It's not part of POSIX
(since it duplicates (e)grep to a large degree, but it is a fairly
fast, fairly small grep for those cramped for disk space or just
love having all sorts of extra utilties around. (It seems like my
Linux partitions are empty!) You can find it as:

tsx-11.mit.edu:pub/linux/binaries/usr.bin/fgrep.tar.Z

Lastly, I've been playing around compiling games/other fun stuff on
Linux lately, and I've come up with rain (BSD), larn (pd), and best
of all: worms (BSD), which I've modified to have _color_ worms! So,
worms and rain, at least, will show up on TSX-11 in the near future.
(Once I apply the color diffs to my system so I can fix worms' colors,
and fix a small problem with worms (it crashes mightily if you try
and tell it how many worms to use. Sigh.).) <- I've been programming
too much!
---
--
Jim Winstead Jr. (CSci '95) | "Catch a fish!"
Harvey Mudd College | -Geddy Lee,
[email protected] | San Diego Sports Arena
Disclaimer: Mine, not theirs! | January 20, 1992


[next article]
From: [email protected] (Wayne Hayes)

Subject: Re: Sz was Re: Wanted: terminal program(not kermit)
Message-ID: <[email protected]>
Date: 23 Feb 92 03:29:38 GMT
References: <[email protected]> <[email protected]>
Lines: 49

I've gotten sz and rz to work fine under Linux: version 3.02 6-04-89,
by Chuck Forsberg. The system that worked for the make was sysvr3.

I use some of the tools that came with the Minix zmodem distribution,
namely bgzmod. ("background zmodem"). Since Linux has job control, I
don't even have to change screens. I type "rz" or "sz" on the remote
system, get back to the kermit prompt, do *not* exit kermit, but
suspend it with ^Z. Then just type "bgz sz " or "bgz rz".
When it's done, I "fg" to kermit, type "connect", and I'm back.

Here's the relevant bgz... files:

****** bgz (which should be in your path, and is the one you call) ***
#!/bin/sh
case "$1" in
r?) cd /usr/tmp ;; # put recieved things in /usr/tmp
*) ;;
esac
exec /usr/lib/zterm/bgzmod "[email protected]" /dev/ttyd1



***** bgzmod, which came with the Minix distribution of zterm ****
#!/bin/sh
# bgzmod - run zmodem transfers in the background
# called by zterm
ZDIR=/usr/lib/zterm
command="$1"
shift
case $command in
rz|rx) ;;
sz|sx) files=''
for i
do
files="$files $i"
done
;;
*) echo "Bad command - exiting" >/dev/tty
exit 1;
;;
esac
stty -even 8 $command $files
(${ZDIR}/bell;echo "Zmodem Transfers Completed";echo "Exit to shell and type 'zt erm' to enter terminal") >/dev/tty

--
Why don't sheep shrink when it rains?
- seen on A&E's Evening At The Improv
Wayne Hayes INTERNET: [email protected] CompuServe: 72401,3525


[next article]
From: [email protected] (Adam Goldberg)

Subject: fdisk / mkfs problems - ST251R & ST22
Summary: keyword
Message-ID:
Date: 23 Feb 92 03:21:37 GMT
Sender: [email protected] (USENET News System)
Organization: Iowa State University, Ames IA
Lines: 29
Originator: [email protected]

A friend of mine, who purchased a 386 motherboard for the express
purpose of running Linux (hooray) is having some problems getting
things up & going.

The boot process goes just fine, but fdisk reports NOTHING. It
dies without any sort of output whatsoever.

Furthermore, skipping the mkfs portion (and just guessing the #
of blocks) reports bad blocks before the first data block (or
something to that effect).

He has an 386/25 with Award BIOS, a ST251R (42 meg) RLL drive, with
a ST22 controller card. (and 4M memory...)

Does ANYONE have ANY idea what's wrong, or what to do, or what
to try, or ANYTHING AT ALL. HELP!

(Reply to either me ([email protected]) or him ([email protected]))

Thanks.

--
Adam Goldberg ! Ask me how IOWA STATE UNIVERSITY is
[email protected] ! censoring my usenet access!!
..!uunet!iastate.edu!goldberg ! >>I don't need a BIG BROTHER!!<<
--
Adam Goldberg ! Ask me how IOWA STATE UNIVERSITY is
[email protected] ! censoring my usenet access!!
..!uunet!iastate.edu!goldberg ! >>I don't need a BIG BROTHER!!<<


[next article]
From: [email protected] (Justin Twiss)

Subject: Call me stupid (problems with installing Gcc)
Summary: I'm having problems installing Gcc
Keywords: Gcc Installation Linux
Message-ID: <[email protected]>
Date: 23 Feb 92 03:31:55 GMT
Sender: [email protected] (netnews admin account)
Organization: University of Denver, Dept. of Math & Comp. Sci.
Lines: 14


Call me stupid (please, no mails addressed to stupid ๐Ÿ™‚ ) but I'm having
a
few problems installing the Gcc binaries under Linux... I can tar
them across, and have them sittig in my work directory, and can
use Gcc-Ar to unarchive the library';s etc, but just wondering WHERE DO THEY
GO?

Ie, is there a directory structure that is required? If not, can anyone
suggest something because I don';;t like having everything in one
directory, it looks messy ๐Ÿ™‚


[email protected]


[next article]
From: [email protected] (Boston)

Subject: Re: poe-IGL login.c not setting TERM
Message-ID: <[email protected]>
Date: 23 Feb 92 02:42:05 GMT
References: <[email protected]>
Sender: [email protected] (Usenet poster)
Organization: State University of New York at Stony Brook
Lines: 15
Nntp-Posting-Host: engws8.ic.sunysb.edu

In article <[email protected]> [email protected] wr ites:
>I see that agetty.c and login.c in version 1.1 of poe-IGL intentional
>don't set TERM. This is oaky if not using SVGA at boot time. User
>.profile can set TERM only if knows what terminal is set in tty1 which
>was set by bootimage/rootimage stuff. QUES: how to access tty1's TERM
>when tty5 login BTW: should tty1 ever come back (in a do forever) or
>should it be in the background so that tty1 is available YES: I put
>doshell for tty2 and tty3 in /etc/rc [helpful if there was a readme
>file]

Ouch you seem to be ahead of me. I just got it to compile, and I do not have a
tty5, so I removed it from inittab, when I boot up, it just sits there. So I
reload it, and make a symbolic link between tty5 and tty1, the problem is now
that when I do login, I'm on as a dumbterminal with funky key settings, any
idea how to get it to work right?


[next article]
From: [email protected] (Brian Mcbee)

Subject: Re: poe-IGL login.c not setting TERM
Message-ID: <[email protected]>
Date: 23 Feb 92 05:16:06 GMT
References: <[email protected]> <[email protected] sunysb.edu>
Sender: [email protected]
Organization: Oregon State University, CS Dept.
Lines: 10
Nntp-Posting-Host: jacobs.cs.orst.edu

I don't know if this is maybe fixed in the latest release of poe-IGL, but
I hacked my login so that it grabbed TERM from /etc/inittab. Seemed like
the logical thing to do.


--
Brian McBee [email protected] or [email protected]
"The strongest reason for the people to retain the right to keep and bear
arms is, as a last resort, to protect themselves against tyranny in
government." Thomas Jefferson


[next article]
From: [email protected] (Brian Mcbee)

Subject: A couple of problems...
Keywords: Gcc Installation Linux
Message-ID: <[email protected]>
Date: 23 Feb 92 05:40:24 GMT
References: <[email protected]>
Sender: [email protected]
Organization: Oregon State University, CS Dept.
Lines: 20
Nntp-Posting-Host: jacobs.cs.orst.edu

I have been hacking on linux now for a couple of weeks,
with almost uniformly good results. I have poe-IGL,
shoelace, mtools, and most other patches etc installed.
I do have two problems though:

1- I am unable to use 720k disks in my 1.44 meg 3.5 inch
drive. I have tried all the likely looking minor
device numbers. Anyone else get this to work?

2- How do people backup their systems? Trying to create
a large tar file either on floppies or to a disk file
almost inevitably locks up my system. A known
problem? Maybe fixed in 0.13? ๐Ÿ™‚


--
Brian McBee [email protected] or [email protected]
"The strongest reason for the people to retain the right to keep and bear
arms is, as a last resort, to protect themselves against tyranny in
government." Thomas Jefferson


[next article]
From: [email protected] (Boston)

Subject: Re: Call me stupid (problems with installing Gcc)
Keywords: Gcc Installation Linux
Message-ID: <[email protected]>
Date: 23 Feb 92 04:57:50 GMT
References: <[email protected]>
Sender: [email protected] (Usenet poster)
Organization: State University of New York at Stony Brook
Lines: 22
Nntp-Posting-Host: engws8.ic.sunysb.edu

In article <[email protected]> [email protected] (Ju stin Twiss) writes:
>
>Call me stupid (please, no mails addressed to stupid ๐Ÿ™‚ ) but I'm having
>a
>few problems installing the Gcc binaries under Linux... I can tar
>them across, and have them sittig in my work directory, and can
>use Gcc-Ar to unarchive the library';s etc, but just wondering WHERE DO THEY
>GO?
>
>Ie, is there a directory structure that is required? If not, can anyone
>suggest something because I don';;t like having everything in one
>directory, it looks messy ๐Ÿ™‚

Well, I too have all files in one dir /usr/gcc so I have a shell script in
/usr/bin that looks like this:
/usr/gcc/gcc -B/usr/gcc/gcc- -I/usr/gcc/include -L/usr/gcc $*
That works okay, but I also have some symbolic links for the programs such
as gcc-ar they are linked to:
gcc-ar
ar
gar
all in /usr/bin (THough the exe itself is in /usr/gcc)


[next article]
From: [email protected] (Charles Hedrick)

Subject: new libc copyright
Message-ID:
Date: 22 Feb 92 18:39:05 GMT
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 31

I'd like to suggest that people review the dj copying file carefully
before incorporating the new libraries into a release. I think it's a
bad idea for OS's to have libraries that impose royalties on software
written using them, whether the royalties are monetary or political.
(dj's conditions are a combination.) My suspicion is that if I showed
copying.dj to my lawyer friends, they'd just laugh, but it is still
dangerous. If it's valid, then technically it's a violation to
distribution Berkeley applications for Linux, since they are not
covered under the GPL. (They're covered under more liberal licensing.
DJ requires that requiring his code must carry the GPL, a restriction
stronger than the GPL itself.) We haven't gone through years of
de-ATT'ing systems just to pick up more restrictions. I've noticed
that there are pretty complete library sources in the Berkeley sources
area on ftp.uu.net. Indeed most of the supposed djcode is either
trivial or it's Berkeley code with the statement "This file may have
been modified by DJ Delorie...". I think the concept of saying "This
file may be copyright" is legally dubious. I would suggest redoing
the new libraries based on the original Berkeley code, unless we're
*really* sure that the dj notices can be ignored.

By the way, in general I support the Gnu Public License. I do not
support it or similar restrictions in libraries, because that makes it
difficult to write anything other than Gnu code for the system, both
truly restriction-free and commercial. Recent versions of the GPL
have apparently have improvements for libraries (I can't be sure,
because despite all the Gnu code in Linux, I can't find the GPL
anywhere -- at Rutgers we only seem to have the old version), but
still appear to require special precautions that I suspect many
vendors would not be willing to deal with. While there are probably
no vendors currently waiting eagerly to port code to Linux, I don't
see any reason to set things up to make that impossible.


[next article]
From: [email protected] (Charles Hedrick)

Subject: vXprintf doesn't work in new libraries
Message-ID:
Date: 22 Feb 92 19:29:02 GMT
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 8

The new libraries define vfprintf and friends as expecting multiple
arguments. Under the old libraries, SunOS, and the new BSD code, they
expect a vararg pointer, i.e. the address of a list of arguments.
This causes the error printer used by the Gnu fileutils (which passes
the address of a list) to print incorrect error messages. I'll
be including a fixed version of the v..printf functions in a
posting later today (which also includes fixed versions of
functions dealing with timezones -- I've taken the BSD code).


[next article]
From: [email protected] (Charles Hedrick)

Subject: losing serial input
Message-ID:
Date: 23 Feb 92 02:16:22 GMT
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 29

A number of people have reported problems with losing serial input. I
see it with both kermit and ka9q, even when nothing else is going on,
but more severely when anything is happening in another window. I've
looked into it a bit more, and have found at least two separate
problems. First, it appears that input interrupts are being missed.
We've talked about a solution: queuing the copy_to_cooked, rather than
doing it directly from interrupt level. This doesn't solve the
problem. In fact it looks like do_timer is simply running too long
with interrupts disabled. At least if I understand what's going on,
an interrupt gate disables interrupts, i.e. acts like cli. Thus all
of do_timer is running with interrupts off. By putting an sti() at
the beginning of do_timer, and putting cli/sti around things in it
that I thought might need it, I've made it so that I can ftp files
reliability, as long as nothing else is going on. However if anything
is going on in another fork, this loses big. Apparently the
scheduling mechanism is working OK, because I can interrupt the
process and get to the command level in ka9q. So the select is
responding immediately. But it seems like there's a very high
character drop rate. This is true whether the competing task is doing
I/O or not. I tried a simple loop in the shell, with
while :
do
:
done
That brings the file transfer to a screeching halt.

Linus indicated that he thought he had fixed this for 0.13. If he's
willing to send me the changes via email, I'd be interested to try
them.


[next article]
From: [email protected] (Charles Hedrick)

Subject: where's the rename patch?
Message-ID:
Date: 22 Feb 92 20:07:09 GMT
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 5

The new libc (i.e. the one posted with the new gcc and math lab)
requires the kernel rename system call. That's not in 0.12, but I
know it was made available as a patch. I don't see it on either
tsx-11 or banjo. I'm trying to avoid using the transatlantic links.
Could somebody tell me where it is?


[next article]
From: [email protected] (Jeffrey Comstock)

Subject: Re: just testing. ignore, please.
Message-ID: <[email protected]>
Date: 23 Feb 92 14:20:42 GMT
References: <[email protected]>
Organization: Sewer of Source Code
Lines: 10

Whoa - this one is kind of hard to ignore:

Organization: Kurchatov Institute of Atomic Energy, Moscow

Greetings from Minnesota in the USA Dima.

--
Jeffrey R. Comstock
INET [email protected]
CW -. .-. ----- -..


[next article]
From: [email protected] (Dylan Smith)

Subject: libtermlib.a
Message-ID: <[email protected]>
Date: 23 Feb 92 13:55:20 GMT
Organization: Bristol Polytechnic, England
Lines: 9

Hello!
I'm trying to port the game 'hack' to linux (the adventure game) and it needs
libtermlib.a
Where can I get this (or the sources for it?)
Thanks.

--
Email : JANET [email protected] | Everywhere else [email protected]
FidoNet Address 2:252/204 | Data (v22/v22bis) phone 0491 875104


[next article]
From: [email protected] (Mr GM Rehmet)

Subject: RE: making filesys - cannot open /dev/hdX
Message-ID:
Date: 23 Feb 92 11:38:48 GMT
Sender: [email protected] (Rhodes University NNTP server)
Organization: Rhodes University, Grahamstown, South Africa
Lines: 33

Hi,

I've just started playing around with Linux - looks very nice.

When I tried to install it on my second hard disk however I ran into
some problems.

My system configuration is :-
80364-DX33 with
Hard drive 0 - Conner CP30104 (120M IDE drive)
Hard drive 1 - Seagate ST157A (44M IDE drive)

I created a 20M partition on my 2nd drive (split it into 2 - using
edpart, and then later part.exe from pdisk). When I tried to do a mkfs
from linux, I got the error message
"mkfs : cannot open /dev/hd8" (or something like that)

(I have had no hassles creating a filesystem on a 386 in the department
- I have no idea what sort of hard drive it has !)

I know this might well fall into the category of FAQ's, but I have to
limit the transfers I make from tsx-11 (etc) because I'm hogging a lot
of bandwidth on our 9600 baud line (yes South Africa is still in the
dark ages a bit :-)).

Thanks,

Cheers, Geoff.
--
NOTE NOTE NOTE NOTE NOTE NOTE NOTE NOTE NOTE NOTE NOTE NOTE NOTE

This is a sample signature file - please edit your file '.signature' and
change it to suit your particulars. Something along these lines is suitable:


[next article]
From: [email protected]

Subject: Compiling
Message-ID: <[email protected]>
Date: 23 Feb 92 21:35:05 GMT
Organization: Memphis State University
Lines: 3

What is a __stderr in text segment??? I recieve that error message everytime
I attempt to compile any of the login source. What libraries do you need
to compiler ps.c and df.c???


[next article]
From: [email protected] (I.J. Wells)

Subject: Re: Call me stupid (problems with installing Gcc)
Keywords: Gcc Installation Linux
Message-ID: <[email protected]>
Date: 23 Feb 92 21:17:29 GMT
References: <[email protected]> <[email protected] sbcs.sunysb.edu>
Sender: [email protected] (The news facility)
Reply-To: [email protected] (I.J. Wells)
Organization: U of Cambridge Computer Lab, UK
Lines: 18


In article <[email protected]>,
[email protected] (Boston) writes:
|> Well, I too have all files in one dir /usr/gcc so I have a shell script in
|> /usr/bin that looks like this:
|> /usr/gcc/gcc -B/usr/gcc/gcc- -I/usr/gcc/include -L/usr/gcc $*

I think the files in the gcc distribution need to be in /usr/local/bin for gcc
to find them without being told. However, it's worth making links (symbolic
or otherwise) to anything you'd like to run without running the 'gcc'
command, partticularly gcc-ar (as ar), gcc-ld (as ld) and gcc-cpp (as
cpp), although
there may be others. Ther links should go somewhere in the path, I used
/usr/gnu/bin since it seemed the most locgical but suit yourself.

Ian Wells.
-------------------------------------------------------------------------------
[email protected] on JANET, [email protected] on Internet


[next article]
From: [email protected] (Hongjiu Lu -- Graduate Student)

Subject: Re: GCC for Linux (fwd)
Message-ID: <[email protected]>
Date: 23 Feb 92 22:35:11 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Hongjiu Lu -- Graduate Student)
Organization: The Internet
Lines: 58

Forwarded message:
| From [email protected] Sun Feb 23 14:12 PST 1992
| Date: Sun, 23 Feb 92 16:37:52 EST
| From: [email protected] (Humberto Ortiz-Zuazaga)
| Subject: Re: GCC for Linux
| In-Reply-To: <[email protected]>; from
| "Hongjiu Lu -- Graduate Student" at Feb 23, 92 12:15 am
| To: [email protected] (Hongjiu Lu -- Graduate Student)
| Cc: [email protected] (Marc Corsini)
| Reply-To: [email protected]
| Message-Id: <[email protected]>
| X-Envelope-To: [email protected]
| X-Mailer: ELM [version 2.3 PL11]
|
| Hongjiu Lu -- Graduate Student wrote:
| > > them to tsx-11? If you did, did you sent a note to [email protected]
| > > telling him about the upload? I have not found the files as of 12:30am
| >
| > Maybe I should. Perhaps, he is testing it now. Try ftp.eecs.umich.edu
| > under /incoming.
|
| Yes, you definitely should. If not, they will not be made available. I
| did get them from umich, as you said, but tsx-11 incoming directory is
| write only, no reads allowed.
|

This gcc 1.40 and libc.a is just a quick hack to get rid of estdio. I
really don't like them. They will be replaced by 2.0 soon. The stdio
for 2.0 will be based on iostream.

|
| However, the libsoft.a supplied with newgcc.tar.Z has not been ranlib'd,
| and will NOT work (undefined symbols). I rebuilt the whole thing from
| source, but I suspect I only needed to ranlib the original to get it to
| work.
|

Sorry about that. Just do a ranlib is enough.

| The top level makefile needs some work, especially to work with floating
| point emulation. The DIRS values need to change depending on the math
| lib used, and the entry with mlinux needs stdio added to the list.
|

Yes. Add stdio. Please try mlinux if you have a 387. I was told DJ's
copyright notice is funny. The only DJ's code in gcc 2.0 will be just
11 assembler functions in mlinux. They are not very hard to write if
you know some 387. I rewrote the part of mlinux. I hope someone will
finish it. Of courses, they should be at least as good as DJ's code.
Mathlib 2.1 may be a good start.


H.J.
--
School of EECS Internet: [email protected]
Washington State University BITNET: [email protected]
Pullman, WA 99164 Phone: (509) 335-6470 (O)
USA (509) 334-6315 (H)


[next article]
From: [email protected] (Jordan M Kossack)

Subject: uucp ???
Message-ID: <[email protected]>
Date: 23 Feb 92 23:13:37 GMT
Sender: [email protected] (Jordan Marc Kossack)
Distribution: usa
Organization: Rice University, Houston Texas
Lines: 19

Howdy

Is there UUCP available for LINUX?

I've booted LINUX off of the floppy and played around with it, but
I hesitate to install it on the hard disk w/o uucp funtionality.
I'm presently running XENIX on my machine, but between the problems
I've had with _its_ uucp and the lack of source for XENIX, I would
rather be running LINUX if I could do email and news.

Any advice? Other than porting it myself? ๐Ÿ™‚


--
Jordan M Kossack | OK ... so I'm mounting a filesystem when I get
[email protected] | email. It's some guy trying to sell me Berkeley
+1 713 270 9056 | UNIX. I said, "Forget it; It's just not AT&T."




[next article]
From: [email protected] (Toshi Morita)

Subject: GCC install problems
Keywords: gcc help install insanity
Message-ID: <[email protected]>
Date: 22 Feb 92 22:07:05 GMT
Lines: 13



I've gotten past the first major hurdle of "gcc can't find cpp" by
renaming gcc-cpp to cpp, gcc-as to as, etc, but now gcc can't find the
include file directory.

Will some kind soul kind tell me how to inform gcc of the whereabouts of
my include files before my sanity, already teetering on the brink of
the chasm of chaos, plunges downward, never to be retrieved by man nor
the minions of Darkness?

Toshi Morita
[email protected]


[next article]
From: [email protected] (Peter Hawkins)

Subject: SML on Linux?
Message-ID:
Date: 24 Feb 92 03:10:07 GMT
Sender: [email protected] (USENET News System)
Organization: University of Massachusetts, Amherst
Lines: 1

did anyone successfully compile SML-NJ on Linux on a 386?


[next article]
From: [email protected] (Drew Eckhardt)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Keywords: v86, video
Message-ID: <[email protected]>
Date: 24 Feb 92 05:17:50 GMT
References: <[email protected]> <[email protected] .edu> <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 50
Nntp-Posting-Host: juliet.cs.colorado.edu

In article <[email protected]> [email protected] (D.Bolla) writes:
>In article <[email protected]> [email protected] (Curtis Yarvi n) writes:
>>In article <[email protected]> [email protected] (Dre w Eckhardt) writes:
>>>We want X11 up on Linux. Since not everybody has an ATI or Tseng chipset, we
>>>also want to support the other cards out there - Trident 8800 / 8900,
>>>Genoa, etc.
>>
>>What's wrong with device drivers? Ioctls for bank switching and so forth,
>>and maybe a mappable frame buffer, and porting X should be clean enough.
> ^^^^^^^^^^^^^^^^^^^^^
> YES

If you read the post, I was suggesting v86 support of BIOS for
-INITIALIZATION- and that a syscall should be provided for
bank switching. These accomodated, we have -ONE- driver
for virtually all SVGA cards.


Also, I suggested that this was a stopgap meassure until real
support could be provided for everything.

>Agreed !
>I have a NON standard VGA card. It seems that I should be interested in
>using the BIOS, right ?
>NO.
>This is UNIX. Let's leave DOS OUT !
>This doesn't mean that we loose the card. It means that we use UNIX
>tools to achieve the same result.

>Eg: What about starting all the time using VGA and when the system is up

Maybe you have a VGA. Maybe I have a SVGA. Not everyone does.

>and running using ioctl to select various modes ?
>It IS possible to avoid BIOS. !
>
>Damiano

Yes, I can write a driver for the Trident card, for my ST-02 (being
debugged), etc. This is optimal. However, it takes time. Having
BIOS available as a fall back means it works now. For any bizarre
hardware out there with a BIOS interface (99% of what's outthere).

As far as using BIOS for kicking into SVGA modes - this is initialization.
The same code we're going to write for every SVGA chip out there.
Called once. Well, the x386 server lets you change resolution dynamically -
so maybe more than once. I honestly doubt that you would notice a 1 ms
delay on X startup.




[next article]
From: [email protected] (Curtis Yarvin)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Keywords: v86, video
Message-ID: <[email protected]>
Date: 24 Feb 92 05:44:44 GMT
References: <[email protected]> <[email protected]> <1992F [email protected]>
Sender: [email protected]
Organization: Brown University Department of Computer Science
Lines: 11

In article <[email protected]> [email protected] (Drew Eckh ardt) writes:
>
>Also, I suggested that this was a stopgap meassure until real
>support could be provided for everything.

I don't like stopgap measures; they have a way of becoming permanent. It
seems to me that stopgap measures are how Unix got as bloated as it is.

Anyone else share my prejudices?

c


[next article]
From: [email protected] (Drew Eckhardt)

Subject: poe-IGL sucks up cycles
Message-ID: <[email protected]>
Date: 24 Feb 92 05:36:17 GMT
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 3
Originator: [email protected]
Nntp-Posting-Host: juliet.cs.colorado.edu

The infinite loop in simpleinit eats up cycles. Under a load, you can
notice a difference in performance (386-33). A simple fix that works
is to put a sleep(1) at the end of the loop.


[next article]
From: [email protected] (Charles Hedrick)

Subject: new libc, file utils, date-related stuff
Message-ID:
Date: 23 Feb 92 07:37:08 GMT
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 106

I've just put up source and binary for libc, the Gnu file utilities,
and date-related programs on athos.rutgers.edu for anonymous FTP.
I'd appreciate it if the folks who are working on libc and the
kernel would take a look at this. The following description is
from the README, READ.tim-futil. (Sorry for the cryptic name --
14 characters, you know).

The changes to libc are:
imported BSD versions of the time-related stuff, setenv, getenv
fixed vprintf and friends
These are based on the new libc that came with GCC 1.40.

This file describes the following sets of patches:

futilbin.tar.Z futilsrc.tar.Z
libbin.tar.Z libsrc.tar.Z
kern-time.pat

These are patches to the kernel and the new libc (the one with gcc
1.40 and improved math), as well as new copies of the Gnu file
utilities compiled with the new libc, and a couple of new programs
related to dates.

futilbin contains new binaries of the Gnu file utilities
futilsrc is the source to them. This is fairly unchanged from
the Gnu source, but a couple of minor things had to be done.
libbin contains a new version of libc.a, as well as the binaries
for the time-related programs: date, zic, setzone, and the
source from which you can build /usr/share/zoneinfo and
/etc/localtime. I'm giving that as source because people
may need to change it. Once you've unloaded this stuff,
you'll want to look at /usr/src/zic/README. Basically,
you just type "make install" as root in /usr/src/zic,
and it will put the time zone stuff in /usr/share/zoneinfo.
You'll want to make sure the Makefile has the right
local time zone.
libsrc contains source to the libc modules I have changed -- not
all of libc -- and to date, zic, and setzone.
kern-time.pat contains diffs to implement the rename system
call (because the new libc uses it -- without this, the
new "mv" won't work), and to make settimeofday set the
CMOS clock.

The major changes should be apparent from this list. I started out to
implement the "date" program and time zones. I need time zone code
because I run my CMOS clock on GMT. (I do this so that I don't have
to adjust it every time daylight time happens. Unix handles this
automatically.) You can of course continue to run your clock on local
time. If you want to do things this way, then you don't need to do
anything with zic or setzone -- if the time zone descriptions are
missing, the library routines will treat the clock as local.

Once I did this, I decided to rebuild the file utilities, largely so
that "ls" would show the right creation times. But it also fixes du
to show Kbytes instead of 512 byte blocks. (I'm not sure why it was
using 512 blocks before -- this behavior is only suppose to happen if
you set the environment variable POSIX_ME_HARDER. The new mv then
pushed me into implemented rename in the kernel, and in order to do
"make install", I had to fix rename so you can rename on top of
existing files. The bug fixes for vXprintf also resulting from making
the file utilities work.

Time zone are a mess. Partly for historical reasons: the new zoneinfo
thing is used by libraries like ctime, localtime, etc. But the kernel
calls, gettimeofday and settimeofday, and the ftime library routine
use an old convention, where the system has a single time zone, and a
daylight saving code that's a small integer (0 for none, 1 for US).
SunOS has a program tzsetup that looks at /etc/localtime and tries to
figure out the old time zone and daylight savings code that's closest
to the information in the new format. BSD has removed ftime
completely, claiming that it's impossible to map from the new system
to the old one, and everybody should convert. I've taken the
intermediate strategy of letting you set a time zone using setzone.
Presumably this will be run in /etc/rc. It takes as arguments the
values that xxxtimeofday and ftime want: minutes west of GMT and
daylight savings code.

The other oddity is raised by the CMOS clock. Linux initializes the
time based on the CMOS clock. Unfortunately that clock is used mostly
by MS/DOS, which runs on local time rather than GMT. Thus when you
initialize this way, you end up with Linux running on local time, not
GMT. When you set the time zone using setzone (or the settimeofday
system call with the tz argument specified), Linux will adjust its
internal clock to GMT. If you keep your CMOS clock on GMT, you should
use setzone with the -u option. This prevents Linux from adjusting
the time (by sort of a hack -- it sets the time to the current time --
when both the time zone and time are set at the same time,
settimeofday does not adjust the clock to GMT). If you choose not to
use time zones at all, but simple let the clock stay on local time,
you can forget about setzone completely.

In summary, there are three reasonable time strategies:

1) ignore it. Don't set up zoneinfo and don't use setzone.
Your internal clock will be on local time, set from
the CMOS clock.

2) set up timezones, and run your CMOS clock on local time,
as in MS/DOS. To do this, set up zoneinfo as
indicated above, and run setzone in /etc/rc.
E.g. if you are on the east cost, do
/etc/setzone 300 1

3) as above, but your CMS clock is on GMT. Set up zoneinfo
and run setzone, but with the -u option
/etc/setzone -u 300 1


[next article]
From: [email protected] (Charles Hedrick)

Subject: Re: A couple of problems...
Keywords: Gcc Installation Linux
Message-ID:
Date: 23 Feb 92 17:44:55 GMT
References: <[email protected]> <[email protected] S.ORST.EDU>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 12

>2- How do people backup their systems? Trying to create
> large tar file either on floppies or to a disk file
>almost inevitably locks up my system. A known
>problem? Maybe fixed in 0.13? ๐Ÿ™‚

Linus has been hoping that these hangs (which occur to several people)
are due to running out of memory, or possibly to coming close to
running out (a situation which causes special code in the kernel to
run to free up memory). I saw a great improvement when I allocated a
partition for swapping. I have had only one hang since, and it was
late yesterday when I was experimenting with some dubious kernel
changes.


[next article]
From: [email protected] (Charles Hedrick)

Subject: Re: A couple of problems...
Keywords: Gcc Installation Linux
Message-ID:
Date: 23 Feb 92 17:52:48 GMT
References: <[email protected]> <[email protected] S.ORST.EDU>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 22

I just realized one thing that will cause I/O errors when doing
backups: if you have any bad blocks on the disk, "mkfs -c" will put
them into /.badblocks. If you do a full backup, tar will try to read
that file, and of course will cause I/O errors. You should use the -X
option to tar to exclude /.badblocks.

By the way, here's the way I do full backups. I haven't gotten around
to doing partials yet, but at least keep a timestamp so I do can a
find -newer. Note that what I'm doing is piping the backup into
mwrite, which writes a file under DOS. Obviously this requires a lot
of disk space, since I end up with two copies of Linux, one under DOS.
I have a cartridge tape drive that works only under DOS, so I then
bring up DOS and do a backup there. If you do that, don't forget to
keep a copy of mread somewhere under DOS, since you'd need that to
restore from backups.

#!/bin/sh
tar -c -b 200 -X /etc/dump.excludes -f - / | /usr/local/bin/mwrite - c:/unixback /dump.tar
touch /etc/dump0.stamp

/etc/dump.excludes is a file listing things not to backup. It
currently lists only /.badblocks.


[next article]
From: [email protected] (Charles Hedrick)

Subject: Re: no subject (file transmission)
Message-ID:
Date: 23 Feb 92 23:11:31 GMT
References: <[email protected]>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 10
To: hlu%[email protected]

I hadn't tried any floating point with the old compiler. But with the
new one, conversion from real to integer is wrong. It's rounding
instead of truncating. I don't know enough about the 387 to be able
to tell what the problem is, but it could certainly be the floating
point emulation in the kernel. I'm using -lsoft, and I don't have a
387. I'm much closer to having a releasable port of icon, but some of
the tests fail, and one even crashes. I suspect it's because of this.
(I get array boundary out of bounds on some tests involving random
number. I think because of rounding it's trying to access the next
element beyond the end.)


[next article]
From: [email protected] (Charles Hedrick)

Subject: new icon, math library fixes, libc fix
Message-ID:
Date: 24 Feb 92 07:09:55 GMT
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 35
Cc: [email protected], [email protected]

On athos.rutgers.edu, in pub/linux, I've just put a copy of icon, as
icon.tar.Z. It is designed to be used with a distribution of Icon,
from cs.arizona.edu, though if you don't need any of the libraries,
the binaries here may be enough for you. (This isn't quite like libc
-- the libraries are various optional modules. You can easily import
the ones you need and compile them individually.) This tar file puts
things in /usr/icon/v8/bin and /usr/icon/v8/config/unix/linux. The
first contains the execuatables icont and iconx (the translator and
runtime, respectively). The second contains the implementation
specific source files. The linux port did not require modifying any
of the main source file tree, so you should be able to install the
Linux configuration subdirectory and then following the normal
procedures for rebuilding icon. I've run all the tests, and
everything works. Note that building iconx uses -lm. If you're
using libsoft.a (as I did here), you should make sure a copy of
libsoft.a is available as /usr/lib/libm.a.

In order to get things to this point, I had to fix several things.
The floating point fixes are in math.fixes:
- fix crt0.s to initialize the 387 to truncate towards
zero. The coode produced by GCC seems to assume
this. Without it, assigning a floating point
number to an integer uses rounding instead of
truncation. This caused disasters to the icon
validation tests.
- fix sqrt to be able to take sqrt(0)
- add ecvt, which is needed by icon
By the way, libsoft.a apparently had not been ranlibbed. I
got a complaint about a missing routine that went away
after I ranlibbed it.

I also had to fix fwrite. If you tried to write a data structure that
had a -1 in just the wrong place, the write would blow up. I've put
this fix into libsrc.tar.Z and libbin.tar.Z, which I described in a
previous posting.


[next article]
From: [email protected] (Douglas E. Quale)

Subject: new libraries
Message-ID: <[email protected]>
Date: 24 Feb 92 07:11:32 GMT
Sender: [email protected] (The News)
Organization: Undergraduate Projects Lab, UW Madison
Lines: 107

Good stuff.

Do we have a timetable for the rest of the GNU C standard library?
After encouragement from Bruce Evans and John Kohl I found that Chris
Torek's new stdio implementation from the BSD Networking Release 2
compiles with relatively few changes. This stdio is designed to be
ANSI and BSD compatible and is much better than the old BSD code.
If the GNU libc.a is going to be a long wait, we might want to use
this stdio.

In fact, enough of the BSD routines compile so that when combined with
what we already have in Linux to the best of my knowledge all that is
missing for a POSIX compatible libc.a is fpathconf, mkfifo, pathconf,
siglongjmp, sigsetjmp, strtod, tcdrain, tcflush and tcsendbreak. In
particular, the BSD code contains all the POSIX time functions and
works with the TZ environment variable.

About the new library: sysconf.c needs to be fixed to return the correct
answers for job control and saved ids. This is utterly trivial, but we
don't want to overlook it, even though I haven't seen a single application
that uses it. (The GNU utilities would use sysconf if any of the values
could change at runtime. In Linux, all these values are constants and
can be snarfed from at compile time.)

Several folks have (quite reasonably) complained about the error messages
in sys_errlist. Following is the error list I use. (Add errlist.o to
the objects in gen/Makefile.)

# This is a shell archive. Save it in a file, remove anything before
# this line, and then unpack it by entering "sh file". Note, it may
# create directories; files and directories will be owned by you and
# have default permissions.
#
# This archive contains:
#
# errlist.c
# strerror.c
#
echo x - errlist.c
sed 's/^X//' >errlist.c << 'END-of-errlist.c'
Xchar * sys_errlist[] = {
X "Unknown error", /* 0 */
X "Operation not permitted", /* EPERM */
X "No such file or directory", /* ENOENT */
X "No such process", /* ESRCH */
X "Interrupted system call", /* EINTR */
X "I/O error", /* EIO */
X "No such device", /* ENXIO */
X "Argument list too long", /* E2BIG */
X "Exec format error", /* ENOEXEC */
X "Bad file descriptor", /* EBADF */
X "No child processes", /* ECHILD */
X "Try again", /* EAGAIN */
X "Out of memory", /* ENOMEM */
X "Permission denied", /* EACCES */
X "Invalid address", /* EFAULT */
X "Block device required", /* ENOTBLK */
X "Device or resource busy", /* EBUSY */
X "File exists", /* EEXIST */
X "Cross device link", /* EXDEV */
X "No such device", /* ENODEV */
X "Not a directory", /* ENOTDIR */
X "Is a directory", /* EISDIR */
X "Invalid argument", /* EINVAL */
X "File table overflow", /* ENFILE */
X "Too many open files", /* EMFILE */
X "Inappropriate ioctl", /* ENOTTY */
X "Text file busy", /* ETXTBSY */
X "File too large", /* EFBIG */
X "No space left on device", /* ENOSPC */
X "Invalid seek", /* ESPIPE */
X "Read-only file system", /* EROFS */
X "Too many links", /* EMLINK */
X "Broken pipe", /* EPIPE */
X "Math argument out of domain", /* EDOM */
X "Result too large", /* ERANGE */
X "Resource deadlock would occur",/* EDEADLK */
X "Filename too long", /* ENAMETOOLONG */
X "No locks available", /* ENOLCK */
X "Function not implemented", /* ENOSYS */
X "Directory not empty" /* ENOTEMPTY */
X };
X
X#define NR_ERRORS ((sizeof (sys_errlist))/(sizeof(char *))-1)
X
Xint sys_nerr = NR_ERRORS;
X
END-of-errlist.c
echo x - strerror.c
sed 's/^X//' >strerror.c << 'END-of-strerror.c'
X#include
X
Xextern int sys_nerr;
Xextern char *sys_errlist[];
X
Xchar * strerror(int n)
X{
X if (n<0 || n>sys_nerr) n=0;
X return sys_errlist[n];
X}
X
END-of-strerror.c
exit

--
Doug Quale
[email protected]


[next article]
From: [email protected] (Ivan Cheng)

Subject: REQUEST: swapon binary
Message-ID:
Date: 24 Feb 92 13:09:27 GMT
Sender: [email protected] (Operator)
Distribution: alt
Organization: Mathematics and Computer Science, Aalborg University
Lines: 8


Where can I find the binaries for swapon? I am unable to compile it
(only 2 Mb-RAM ๐Ÿ™

If it isn't available on ftp, could someone please mail an uuencoded
version of the binaries?

-Ivan


[next article]
From: [email protected] (KiYun Roe)

Subject: Re: A couple of problems...
Keywords: Gcc Installation Linux
Message-ID: <[email protected]>
Date: 24 Feb 92 12:22:25 GMT
References: <[email protected]> <[email protected] S.ORST.EDU>
Sender: [email protected] (Usenet poster)
Organization: State University of New York at Stony Brook
Lines: 37
Nntp-Posting-Host: sbstaff2

>>Trying to create
>> large tar file either on floppies or to a disk file
>>almost inevitably locks up my system.
>
>Linus has been hoping that these hangs (which occur to several people)
>are due to running out of memory, or possibly to coming close to
>running out (a situation which causes special code in the kernel to
>run to free up memory).

I have been experiencing the same sort of hangs, but I can't see how
they have anything to do with running out of memory, unless you mean
buffer memory. My machine will hang if I try to make a large tar file,
and the times I've checked there was plenty of memory left for the tar
program. A while ago I wrote a little program that did nothing but
write to a large file. The program itself required very little memory
but it would hang the system every time.

I've experienced fewer hangs since I implemented two changes: (1) I
modified the kernel to give me 512KB more buffer memory (on my 4MB
machine the default was really too small, and I have swapping enabled,
so I don't worry about running out of memory); and (2) I revised my
backup script to create separate tars for each directory under / and
/usr, so that none of my tar files is more than about 2.5MB.

I think this hanging is easily the most serious deficiency in Linux,
but I have made zero progress in identifying the cause. I hope Linus
figures it out before he releases 0.13.

KiYun Roe [email protected]
Department of Computer Science
SUNY at Stony Brook
Stony Brook, NY 11794-4400 (516) 632-7675
--
KiYun Roe [email protected]
Department of Computer Science
SUNY at Stony Brook
Stony Brook, NY 11794-4400 (516) 632-7675


[next article]
From: [email protected]

Subject: poe-IGL version 1.1 initial tty settings
Message-ID: <[email protected]>
Date: 24 Feb 92 15:36:08 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 5

SOrry about previous messge BTW: BS works okay if put stty sane in
.profile poe-IGL initializes ERASE to [email protected]

BTW: termcap without :am: did NOT help elvis with getting off by 1
line number (at least on small test samples).


[next article]
From: [email protected] (A. V. Le Blanc)

Subject: Re: Elvis and Linux
Message-ID: <[email protected]>
Date: 23 Feb 92 10:47:10 GMT
References: <[email protected]>
Organization: Computing Centre, University of Manchester
Lines: 25

In article <[email protected]> [email protected] (KiYun Roe) writes:
>I experienced a few minor problems using Elvis for a vi-replacement
>under Linux. First, screen refresh didn't work quite right. I didn't
>look at the problem closely, but it seemed that autowrap wasn't handled
>properly by Elvis or Linux. I took out the :am: setting from the
>termcap entry, and screen refresh now works properly.

I find that there are two irritating 'features' of elvis:

1) Long lines are not wrapped. On all systems on which I have
used elvis, this has meant that the display is unreliable
whenever long lines occur.

2) There is no buffer which saves the last :s command. On Unix,
in vi, I can type, for example, ':s/^........................//',
and then if I want to do this on another line or in another file,
I type ':s'. This does not work in elvis.

I seem to recall that my problem (1) was described by the designer
of elvis as a feature: he does not approve of the default behaviour
of vi. I would very much like an option which would change this.

A. V. Le Blanc
University of Manchester
[email protected]


[next article]
From: [email protected] (A. V. Le Blanc)

Subject: bash 1.12, textutils 1.2, mtools 2.0.5
Message-ID: <[email protected]>
Date: 23 Feb 92 16:48:37 GMT
Organization: Computing Centre, University of Manchester
Lines: 26

The following files are now available by anonymous ftp from
hpb.mcc.ac.uk [130.88.200.7]:

167596 Feb 23 16:25 pub/linux/mcc-bin/bash-1.12.tar.Z
271270 Feb 23 11:20 pub/linux/mcc-bin/textutils-1.2.tar.Z
1422 Feb 23 16:41 pub/linux/mcc-diffs/mtools-2.0.5.diffs.Z

The mtools diffs shows how to create a Linux entry in devices.c
for 3.5 and 5.25 inch drives, and for a C drive which has
12-bit fat's. Many thanks to all who helped with this.

bash-1.12.tar.Z contains a binary for this, plus the patches
I applied to create it. It has one known bug: the jobs command
does not print correctly the name of the command being executed
in each job.

textutils-1.2.tar.Z contains the patches applied to the GNU source,
plus binaries for the commands cat, cmp, comm, csplit, cut, expand,
fold, head, join, nl, paste, pr, sort, split, sum, tac, tail, tr,
unexpand, uniq, and wc.

Comments and suggestions about these files should be sent to
me at the address below:

A. V. Le Blanc
[email protected]


[next article]
From: [email protected] (D.Bolla)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Keywords: v86, video
Message-ID: <[email protected]>
Date: 24 Feb 92 15:04:19 GMT
References: <[email protected]> <[email protected]> <1992F [email protected]>
Reply-To: [email protected] (D.Bolla)
Organization: Computing Lab, University of Kent at Canterbury, UK.
Lines: 15

In article <[email protected]> [email protected] (Curtis Yarvin ) writes:
>In article <[email protected]> [email protected] (Drew Eck hardt) writes:
>>Also, I suggested that this was a stopgap meassure until real
>>support could be provided for everything.
>I don't like stopgap measures; they have a way of becoming permanent. It
>seems to me that stopgap measures are how Unix got as bloated as it is.
I agree. Stopgat measures are BAD.
It is more reasonable to design a clear system ( that may require a bit
more programming ) than a cludge.

Again.. Let's leave the BIOS out. !

BTW My vga is indeed a supervga.....

Damiano


[next article]
From: [email protected]

Subject: Re: poe-IGL login.c not setting TERM
Message-ID: <[email protected]>
Date: 24 Feb 92 16:33:09 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 7

Brian Mcbee's suggestion to grab it out of /etc/inittab not very good
idae if working with virtual consoles. The user has no choice, must use
SVGA mode selected at boot time. My Tseng VGA has modes 0 1 2
3 4 so that a real kludge is to append 0 thru 4 to end of login name
when login . Would be nice to capture boot time stuff and place in
file. tty1 has stuff as does doshell --- I guess I should get source
for doshell and fix .profile


[next article]
From: [email protected] (Hongjiu Lu -- Graduate Student)

Subject: Re: new icon, math library fixes, libc fix
Message-ID: <[email protected]>
Date: 24 Feb 92 17:06:45 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 20

| In order to get things to this point, I had to fix several things.
| The floating point fixes are in math.fixes:
| - fix crt0.s to initialize the 387 to truncate towards
| zero. The coode produced by GCC seems to assume
| this. Without it, assigning a floating point
| number to an integer uses rounding instead of
| truncation. This caused disasters to the icon
| validation tests.
|

Please DO NOT change crt0.s. This is a bug in gnulib. I will try to put
a correct one on tsx-11.mit.edu later today.


H.J.
--
School of EECS Internet: [email protected]
Washington State University BITNET: [email protected]
Pullman, WA 99164 Phone: (509) 335-6470 (O)
USA (509) 334-6315 (H)


[next article]
From: [email protected] (Peter Orbaek)

Subject: Re: poe-IGL login.c not setting TERM
Message-ID: <[email protected]>
Date: 24 Feb 92 15:26:50 GMT
References: <[email protected]> <[email protected] sunysb.edu>
Sender: [email protected] (Peter Orbaek)
Organization: DAIMI: Computer Science Department, Aarhus University, Denmark
Lines: 21

[email protected] (Boston) writes:

>Ouch you seem to be ahead of me. I just got it to compile, and I do not have a
>tty5, so I removed it from inittab, when I boot up, it just sits there. So I
>reload it, and make a symbolic link between tty5 and tty1, the problem is now
>that when I do login, I'm on as a dumbterminal with funky key settings, any
>idea how to get it to work right?

What do you mean? If you remove the tty5 line from inittab and reboot, there
should be no dependencies on tty5 at all.

For now you should set your TERM variable in your .profile. I'll try to
fix the TERM problem in the next release. It should be set to the value given
in /etc/inittab

- Peter.
--
Peter Orbaek ----------------- [email protected] | ///
Hasle Ringvej 122, DK-8200 Aarhus N, DENMARK | ///
| \\\///
"Strong typing is for people with weak memories" | \XX/


[next article]
From: [email protected] (Peter Orbaek)

Subject: Re: poe-IGL login.c not setting TERM
Message-ID: <[email protected]>
Date: 24 Feb 92 15:29:59 GMT
References: <[email protected]> <[email protected] sunysb.edu> <[email protected]>
Sender: [email protected] (Peter Orbaek)
Organization: DAIMI: Computer Science Department, Aarhus University, Denmark
Lines: 20

[email protected] (Brian Mcbee) writes:

>I don't know if this is maybe fixed in the latest release of poe-IGL, but
>I hacked my login so that it grabbed TERM from /etc/inittab. Seemed like
>the logical thing to do.

Not login.c ?!

If you run login from the commandline, it shouldn't take TERM from /etc/inittab

Login should simply enherit the TERM variable from it's parent. (It doesn't
now and that's a bug, admitted!)

- Peter.

--
Peter Orbaek ----------------- [email protected] | ///
Hasle Ringvej 122, DK-8200 Aarhus N, DENMARK | ///
| \\\///
"Strong typing is for people with weak memories" | \XX/


[next article]
From: [email protected] (Peter Orbaek)

Subject: Re: poe-IGL sucks up cycles
Message-ID: <[email protected]>
Date: 24 Feb 92 15:34:00 GMT
References: <[email protected]>
Sender: [email protected] (Peter Orbaek)
Organization: DAIMI: Computer Science Department, Aarhus University, Denmark
Lines: 18

[email protected] (Drew Eckhardt) writes:

>The infinite loop in simpleinit eats up cycles. Under a load, you can
>notice a difference in performance (386-33). A simple fix that works
>is to put a sleep(1) at the end of the loop.

init should sleep most of the time, as it sits in a wait() 99% of the time.

Besides, ps tells me that it IS sleeping almost always.

Maybe something's wrong with your inittab?

- Peter.
--
Peter Orbaek ----------------- [email protected] | ///
Hasle Ringvej 122, DK-8200 Aarhus N, DENMARK | ///
| \\\///
"Strong typing is for people with weak memories" | \XX/


[next article]
From: [email protected] (Tommy Thorn)

Subject: Re: new libc copyright
Message-ID: <[email protected]>
Date: 24 Feb 92 11:24:50 GMT
References:
Sender: [email protected] (Tommy Thorn)
Organization: DAIMI: Computer Science Department, Aarhus University, Denmark
Lines: 10

I couldn't agree more. While I prefer Berkley type copyright
Gnu Public Library Licence is acceptable too. Any thing stronger
would make Linux useless for lots of different purposes
including commercial usage.


--
/Tommy Thorn, another LPF member. Terminate software patents.
Join The League for Programming Freedom, mail [email protected]
Are *you* sure that *your* program isn't already covered by patents?


[next article]
From: [email protected] (Linus Benedict Torvalds)

Subject: Re: A couple of problems...
Keywords: Gcc Installation Linux
Message-ID: <[email protected]>
Date: 24 Feb 92 16:06:53 GMT
References: <[email protected]> <[email protected]>
Organization: University of Helsinki
Lines: 26

In article <[email protected]> [email protected] ( KiYun Roe) writes:
>
>I think this hanging is easily the most serious deficiency in Linux,
>but I have made zero progress in identifying the cause. I hope Linus
>figures it out before he releases 0.13.

Yes, I think I've got it licked: it was /both/ a memory problem and a
filesystem problem. No wonder people have had hangups. Due to the
bugs, 0.13 might be a bit late, but I still expect it to be out about
March 5th or so..

The filesystem problem gets much worse when writing out the buffer-cache
to disk (several requests in the request-queue), and is especially
noticeable on slow devices (floppies and old drives). Writing big files
out to floppy is one of the ways you /might/ hang linux currently, but
it depends on a lot of factors. 0.13 will definitely have these bugs
corrected, I just hope there aren't other ones... Other than that 0.13
will be a relatively simple release: not very much new features, mostly
bug-fixes and kernel enhancements.

Re: system call overlapping - yes some of the system calls overlap, and
could be removed, but it's a real pain to recompile everything: I did
that when I implemented sigactions, and removed the old signal() system
call, and it wasn't fun. Not for the faint-of-hearted.

Linus


[next article]
From: [email protected] (I Reid)

Subject: Bye bye symlinks?
Summary: Where have all those sysmlinks gone?
Message-ID: <[email protected]>
Date: 24 Feb 92 10:25:47 GMT
Distribution: alt
Organization: Edinburgh University
Lines: 5

Quick question..... I had symlinks but now ln says it can only manage hard
links. All my sources are latest releases and I don't think I've done anything
odd to the kernel..... any ideas anyone?

Iain


[next article]
From: [email protected] (Marc CORSINI)

Subject: gcc1-40, FAQ
Message-ID: <[email protected]>
Date: 24 Feb 92 00:42:28 GMT
Organization: GRECO Programmation du CNRS & LaBRI - Bordeaux,France
Lines: 41

Hi Linuxers,
The next release of FAQ will come out in one day or two, in advance this
is a short Q/A samples about gcc-1.4 of [email protected]
Any comment appreciated

==============
QUESTION: What are the contents of the files newggc.tar.Z newlib.tar.Z ?

ANSWER: newgcc.tar.Z contains all the binaries/libraries needed to use
gcc-1.40 whilst newlib.tar.Z contains the library sources.

QUESTION: Is the installation identical to the previous gcc stuff ?

ANSWER: No, after uncompressing the newgcc.tar.Z you have to untar it
below /usr cause the libraries /must/ go to /usr/lib (I've tried under
/usr/local and the compiler complains "ld: no file crt0.o").
Moreover the distrib contains cc1, cpp in the directory lib and
gcc ar as ld strip size ranlib under bin.
So you have to link cc1 and cpp under bin, and link
gar/ar gas/as gld/ld to prevent a lot off /stupid/ problems

QUESTION: What are the main differencies with the old release?

ANSWER: There were some bugs in the old port that have been corrected,
moreover this package contains 387 support (that I have not yet tested)
there is libm.a (for those with 387)
libsoft.a (for those without, I for example)
libtermcap.a (from tput 1.10)

The -mstring-insns option is no longuer needed nor supported ๐Ÿ™ [As
an example to recompile (successfully) linux you have to thraw
away this flag in all the Makefile]

BTW for those who want to recompile the lib sources you have to change all
the paths in Makefile(s) to fit with your organization. Notice also that
include files have changed (stdio.h which is no more ansi complient)

==========


Marc


[next article]
From: [email protected] (Manolis Tsangaris)

Subject: Re: Patches to have reboot and /dev/kmem working
Message-ID:
Date: 24 Feb 92 19:35:08 GMT
References: <[email protected]>
Sender: [email protected] (The News)
Organization: U of Wisconsin CS Dept
Lines: 27

I hate to point out that even though the problem of /dev/kmem is
interesting to many of us, it seems that the discussion shifts
from providing kernel data access, to operating system concepts on
``kernel-user interfaces''. Although it would be undemocratic to deny
discussions on issues like ``kernel-user interfaces'',
I think that alt.os.linux postings should be more oriented towards
implementation choices. There are many other groups to argue about concepts,
and after all, Linux is not a new operating system but a free implementation
of Unix on the 386/486 architectures.

Yes, the "/dev/kmem"/nlist interface has problems (security being one of them)
but nevertheless it has been proposed and used by other Unix implementations.
It is SIMPLE and consistent to the ``everything is a file'' philosophy.
Remember Unix started as simple and understandable operating system.

Many thanks to [email protected] (D.Bolla) for his patches.
Let us use them and settle the issue for now.
There certainly many more features to be added to Linux
to get a usable free Unix.

--mt
+-------------------------------+-------------------------------------------+
|Manolis M. Tsangaris |Email: [email protected], uunet!uwvax!mt |
|Computer Sciences Department |Phone: +608 262 6624,+608 231 1058 |
|University of Wisconsin-Madison|Fax: +608 262 9777 |
+-------------------------------+-------------------------------------------+
P.S. Flames to my account please ๐Ÿ™‚


[next article]
From: [email protected] (Mike Miller)

Subject: Re: Patches to have reboot and /dev/kmem working
Message-ID: <[email protected]>
Date: 24 Feb 92 19:40:14 GMT
References: <[email protected]>
Organization: Research and Evaluation Dept. Portland Public Schools, Portland Or egon.
Lines: 75

[email protected] (D.Bolla) writes:


(Stuff deleted)

>Anyway.... under Sun there are three programs that "look" into /dev/kmem
>ps, w, pstats. The amount of information you can get out of them is very
>big. The point that I am tryng to make here is:
>To obtain the same amount of information from syscalls you need a
>GREAT number of variations and parameters. Is this worth the work ?
>( Considering the small amount of programs that need that information )
>Again.....

Maybe there are so few programs because so few people have access to this
information...

>Negative points of kernel implementation.
>1) Extra (Unnecessary) code in the kernel -> Kernel bigger.
^^^^^^^^^
Depends on who you talk to.

>2) Possible introduction of ERRORS -> Kernel panic.

Thats what alpha and beta testing are for. If your messing with your own
system, then you should expect this from time to time. Its not that
big a deal ussually. You don't do production work on a OS hacking
machine anyway if you are sane.

>3) Increased number of syscall and paramentrs -> Complexity

Not if done properly. Its not the number of calls, its the way they
are implamented, and how well they are documented.

>4) Problems on returning lists of data in user space

>To me point 2 is already enough NOT to use this method. If you can live
>with possible kernel panics.......

I live with them on my Sun 3/75. I don't like it, but it happens. I sure
hope the serial port code in linux is better then suns.

>Positive points of a library implementation.
>1) NO extra code in the kernel. -> Save memory
>2) Impossibility to add ERRORS to the kernel code. -> NO panic

As somebody pointed out here, it is still possible to panic the kernal
if you read stuff from the wrong place. Not to mention writting.

>3) Easy to update, modify even by average user.

Really? Do you really want an average user messing around with the system
libraries? (Or am I misunderstanding you?)

And somebody has to write the code to give them access either way.

>4) No problems in allcating memory for the result.

>What it is important is to DEFINE what information we want to
>extract from /dev/kmem and who can use them. After that is defined it
>is possible to discuss about implementaion.

No argument from me on this part. I sort of like the /proc idea myself.
But I think its a bit limited. What would be ideal is to be able to
have unix like file protections of portions of a running program. Hmm,
you could /dev/kmem as /proc/1... or /dev/proc/1...

That would probably be a pain to do, but VERY powerfull since users could
get access to all sorts of interesting data on their program(s). Shared
libraries would kind of cause a problem though...

--
* Insert your favorite funny/witty saying here, but make it small. *
* [email protected] or [email protected] or [email protected] *
* Michael J. Miller Jr. Stand. disc. applies. *
* Please Reply if the content of this message implies that one is expected. *


[next article]
From: [email protected] (Lawrence C. Foard)

Subject: finally works ๐Ÿ™‚
Message-ID: <[email protected]>
Date: 24 Feb 92 21:58:47 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Lawrence C. Foard)
Organization: The Internet
Lines: 7


I finally had a free day to finish IPC. I've got tinymush working with it
so I'm pretty confident it will work for any demons. The only thing I'm not
sure about at this point is how to make it into a patch, I've got the old
virtual console patch, floppy patch and this all in the same kernel. If I do
a diff of the present kernel with the old (floppy+vc patch) kernel will this
work right?


[next article]
From: [email protected] (R. Ramesh)

Subject: Help with botting PROPERLY
Message-ID: <[email protected]>
Date: 24 Feb 92 22:45:01 GMT
Sender: [email protected]
Reply-To: [email protected]
Organization: The University of Texas at Dallas, ACC
Lines: 23
Nntp-Posting-Host: ramesh.utdallas.edu

Hi gurus:

I have Linux installed on hardrive and boot using shoelace. Since shoelace
requires a boot partition selection, I need to press a key. Apparently
shoelace does not clear some buffer (I don't know which one) and the key is
still available to linux as it boots. Consequently I cannot set the SVGA modes
as the key left in the buffer becomes the answer to the SVGA mode question.
Could some one tell me a way out of this. In particular, could some one give me
details as to how to clear this buffer containing the key left by shoelace
before the SVGA prompt.

BTW, I don't want to insert a floppy each time to boot. I prefer the shoelace
approach. Please answer, if you can, with this constraint.

Is there any one out there who owns a Gateway PC that has the cntrl-alt-del
(reboot) working. The current code does not seem to pull the
reset line in my PC, and thus simply hang the m/c. An attempt to get out of
protected mode and jump to BIOS reboot address craps my CMOS. Help!!!! I hate
pressing reset button (even though it is meant for that).

Thanks.

Ramesh


[next article]
From: [email protected] (Hongjiu Lu -- Graduate Student)

Subject: bug fix for the new gcc 1.40
Message-ID: <[email protected]>
Date: 24 Feb 92 22:59:36 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 44

Hi,

I just fixed a FP bug in gnulib, a bug in sqrt(0.0) of libsoft.a and put
a new newgcc.tar.Z on tsx-11.mit.edu.

The cc1 now will support dbx. Now gcc will look for /usr/local/lib and
also check /usr/local/gcc-include. Since gcc is the native compiler on
Linux, I don't see why we should use /usr/local for the standard stuffs
of gcc.

The binary utilities have the same names as they in Unix. There is no
prefix 'g'.

When you get the new newgcc.tar.Z, unpack it in /usr. Everything will
fall in the right place.

If you don't have a 387, use -lsoft instead of -lm when you use math
functions.

Some other bugs have been reported in libc.a come with the new gcc
1.40. I am working on gcc 2.0 and a new libc.a. Some of the bugs
in the old libc.a will disappear in the new libc.a. Most of the codes in
./unistd will stay in the new libc.a. But the other parts may be very
different from the old one.

The new libc.a will be ANSI (and BSD?) compatible. I will try to port
as much gnu libc.a as possible, as long as I don't have to change any
syscalls.

I hope someone will take over the old libc.a so that I can spend more
time on the new one. But please send me the bugs. I will make sure it
won't happen in the new one.

BTW, if there is another serious bug in gcc, I will be happy to try to
fix it.

Thanks.

H.J.
--
School of EECS Internet: [email protected]
Washington State University BITNET: [email protected]
Pullman, WA 99164 Phone: (509) 335-6470 (O)
USA (509) 334-6315 (H)


[next article]
From: [email protected] (Marc CORSINI)

Subject: FAQ cdiff
Message-ID: <[email protected]>
Date: 24 Feb 92 23:21:44 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 819



Hi Linux_land,
below is the hard-reading Cdiff of the FAQ. It's nearly half big as
the entire FAQ.

$$$$$$$$$$$$$$$$$$$$$

*** faq.1 Mon Feb 3 10:12:09 1992
--- faq.2 Mon Feb 24 23:46:01 1992
***************
*** 4,17 ****

Most credits of this work to Linus, Robert and Ted, the rest was
either on the list posted by many (real) activists, not me ;-), either
! in some other news groups, or else by direct posting to me (I haven't
! copyrighted them, so thanks to every one who participated even
! indirectly to this FAQ).

! [To find what has changed, view the `Cdiff-faq' posting. As this FAQ
! organization is not stable yet, Cdiff-faq is twice bigger as the whole
! stuff ๐Ÿ™‚ and will not be posted this time. The last-change-date of
! this posting is always "two minutes ago". :-)]

This is the introduction to a list of frequently asked questions (FAQ
for short) about Linux with answers (Yeap!). This article contains a
--- 4,15 ----

Most credits of this work to Linus, Robert and Ted, the rest was
either on the list posted by many (real) activists, not me ;-), either
! in some other news groups, or else by direct posting to me (thanks
! Humberto). I haven't copyrighted them, so thanks to every one who
! participated even indirectly to this FAQ.

! [To find what has changed, view the `Cdiff-faq' in my previous posting.
! The last-change-date of this posting is always "two minutes ago". :-)]

This is the introduction to a list of frequently asked questions (FAQ
for short) about Linux with answers (Yeap!). This article contains a
***************
*** 19,26 ****

This FAQ is supposed to reduce the noise level ๐Ÿ˜‰ in the alt.os.linux
newsgroup (and also the 'linux-activists' mailing list), and spare the
! time of many activists.

Please suggest any change, rephrasing, deletions, new questions,
answers ...
Please include "FAQ" in the subject of messages sent to me about FAQ.
--- 17,28 ----

This FAQ is supposed to reduce the noise level ๐Ÿ˜‰ in the alt.os.linux
newsgroup (and also the 'linux-activists' mailing list), and spare the
! time of many activists. I will post it twice a month, since there are
! more and more new incomers, and new features.

+ BTW I think this FAQ should be available at the main Linux sites in
+ the doc directory (have you read this Ary, Robert, Ted ?)
+
Please suggest any change, rephrasing, deletions, new questions,
answers ...
Please include "FAQ" in the subject of messages sent to me about FAQ.
***************
*** 31,45 ****
Thanks in advance,
Marc

! Remind the vote in Feb. 18th for comp.os.linux

Future Plan:
- - make available Cdiff of the FAQ last release.
- mark question/answer that have heavily changed recently
- verification/location/organization for files available
! via FTP
- cross posting this to news.answers as soon as comp.os.linux
! is created

================================8<=====8<==============================
CONTENTS
--- 33,49 ----
Thanks in advance,
Marc

! Remind the vote for comp.os.linux;
! post the YES vote to [email protected]
! The voting period will last through 00:00 UTC (7:00pm EST), 19 March
! 1992.

Future Plan:
- mark question/answer that have heavily changed recently
- verification/location/organization for files available
! via FTP (I've tried what a mess!!)
- cross posting this to news.answers as soon as comp.os.linux
! is created (may be earlier, on march)

================================8<=====8<==============================
CONTENTS
***************
*** 50,55 ****
--- 54,60 ----
V. SOME CLASSICAL PROBLEMS
VI. INSTALLATION HINTS
VII. FEATURES
+ VIII. MORE HINTS

I. LINUX GENERAL INFORMATION
=============================
***************
*** 60,66 ****
advantage of being free. It is still in beta-testing, but is slowly
getting useful even for somewhat real developement. The current
version is 0.12, date: Jan. 14th 1992. The next version will come out
! in the end of Feb.

LINUX 0.12 is a freely distributable UNIX clone. It implements a
subset of System V and POSIX functionality. LINUX has been written
--- 65,71 ----
advantage of being free. It is still in beta-testing, but is slowly
getting useful even for somewhat real developement. The current
version is 0.12, date: Jan. 14th 1992. The next version will come out
! in the end of February, beginning of March as the Linux v0.13/0.95

LINUX 0.12 is a freely distributable UNIX clone. It implements a
subset of System V and POSIX functionality. LINUX has been written
***************
*** 73,78 ****
--- 78,89 ----
and task primitives.


+ QUESTION: What is the current state of Linux?
+
+ ANSWER: do "finger [email protected]", or read the
+ alt.os.linux newsgroup.
+
+
QUESTION: I've just heard about linux, what should I do to get it?

ANSWER: First read all this FAQ, and the INFO-SHEET monthly post, then
***************
*** 144,151 ****
QUESTION: What's about the copywright of linux.

ANSWER: In the previous release, Linux was copyrighted by Linus
! Torvalds but free. The copyright will change (v0.13 ?) to follows the
! same set of rules as the GNU copyleft.


QUESTION: Should I be a UNIX and/or a DOS wizard to install/use Linux?
--- 155,162 ----
QUESTION: What's about the copywright of linux.

ANSWER: In the previous release, Linux was copyrighted by Linus
! Torvalds but free. The copyright will change (v0.13 /0.95 ?) to
! follows the same set of rules as the GNU copyleft.


QUESTION: Should I be a UNIX and/or a DOS wizard to install/use Linux?
***************
*** 154,160 ****
easier for you if you know things about Unix. Right now Linux is used
by more than 400 persons, very few of them enhance the kernel, some
adds/ports new soft and most of us are only (but USEFUL) beta testers.
! So join us and choose your "cast"


QUESTION: What are the differences, pros and cons compared to Minix ?
--- 165,171 ----
easier for you if you know things about Unix. Right now Linux is used
by more than 400 persons, very few of them enhance the kernel, some
adds/ports new soft and most of us are only (but USEFUL) beta testers.
! So join us and choose your "caste"


QUESTION: What are the differences, pros and cons compared to Minix ?
***************
*** 163,169 ****
Cons:
- Linux is not as mature as Minix, there is less working software right now.
- Linux v0.12 doesn't have real login yet, you are root all the time.
! this will change pretty soon probably in the next version (end Feb.)
- Linux only works on 386 and 486 processors.
- Linux needs 2M of memory just to run, 4M to be useful.
- Linux is a more traditional unix kernel, it doesn't use message passing.
--- 174,181 ----
Cons:
- Linux is not as mature as Minix, there is less working software right now.
- Linux v0.12 doesn't have real login yet, you are root all the time.
! this will change pretty soon in the next version (end Feb.). Right
! now, you can use the Peter Orbaek init/login package (see VIII.)
- Linux only works on 386 and 486 processors.
- Linux needs 2M of memory just to run, 4M to be useful.
- Linux is a more traditional unix kernel, it doesn't use message passing.
***************
*** 188,194 ****
QUESTION: If my PC runs under Linux, is it possible to ftp, rlogin,
rsh etc.. to other Unix boxes?

! ANSWER: Not yet, but kermit has been ported to Linux.


QUESTION: Does linux do paging? Can I have virtual memory on my small
--- 200,206 ----
QUESTION: If my PC runs under Linux, is it possible to ftp, rlogin,
rsh etc.. to other Unix boxes?

! ANSWER: Not yet, but kermit has been ported to Linux, and the ka9q too.


QUESTION: Does linux do paging? Can I have virtual memory on my small
***************
*** 209,216 ****

QUESTION: Does the bigger program sizes mean I can run X?

! ANSWER: X is not ported to linux, and though I hope it will be some
! day, I cannot guarantee it. It's big, and wants a lot from the system.


II. LINUX USEFUL ADRESSES
--- 221,229 ----

QUESTION: Does the bigger program sizes mean I can run X?

! ANSWER: X is not ported to linux, and I hope it will be some day,
! people are working hard on it but it's big, and wants a lot from
! the system.


II. LINUX USEFUL ADRESSES
***************
*** 229,234 ****
--- 242,249 ----
directory linux
puffin.doc.ic.ac.uk (146.169.3.7):
directory /pub/os/Linux
+ ustsun.s.u-tokyo.ac.jp (133.11.11.11):
+ directory misc/linux

You might want to check out which of these is the most up-to-date.

***************
*** 266,281 ****
sent to [email protected] are posted to
alt.os.linux

! DO NOT mail "I want to [un]subscribe" to the mailing-list/news, use
the request-address. IF not your mail-box will be over-crowded by
activists.

! Questions and bug-reports can be sent either to the mailing-list or to
"[email protected]", depending on which you find more
appropriate.

BTW People are working on the organization of Linux, this is done on
! an apart mailing-list.

linux-standards: Discussion of distribution and directory standards
for the Linux operating system, including directory structure, file
--- 281,296 ----
sent to [email protected] are posted to
alt.os.linux

! DO NOT mail "I want to [un]subscribe" to the newsgroup, use
the request-address. IF not your mail-box will be over-crowded by
activists.

! Questions and bug-reports can be sent either to the newsgroup or to
"[email protected]", depending on which you find more
appropriate.

BTW People are working on the organization of Linux, this is done on
! a separate mailing-list.

linux-standards: Discussion of distribution and directory standards
for the Linux operating system, including directory structure, file
***************
*** 284,291 ****
Requests to be added to this list must be sent to:
[email protected]

! QUESTION: Does there exist a place where the traffic of the
! mailing-list and newsgroup is kept?

ANSWER: Yes, on nic and tsx-11 (see the ftp adresses above)

--- 299,306 ----
Requests to be added to this list must be sent to:
[email protected]

! QUESTION: Does there exist a place where the traffic of the newsgroup
! is kept?

ANSWER: Yes, on nic and tsx-11 (see the ftp adresses above)

***************
*** 306,313 ****
although I double-checked with fdisk, and it's correct.

ANSWER: Be sure you give the size in BLOCKS, ie 1024 bytes, not
! sectors. The mkfs doesn't work for very big partition. Also,
! make doubly certain that you have the correct partition.

There are a few rules about this: /dev/hd0 and /dev/hd5 are always the
same under linux and minix. DO NOT USE THEM, they are the whole raw
--- 321,328 ----
although I double-checked with fdisk, and it's correct.

ANSWER: Be sure you give the size in BLOCKS, ie 1024 bytes, not
! sectors. The mkfs doesn't work for very big partition (over than 64
! Megs). Also, make doubly certain that you have the correct partition.

There are a few rules about this: /dev/hd0 and /dev/hd5 are always the
same under linux and minix. DO NOT USE THEM, they are the whole raw
***************
*** 318,324 ****
programs sort this information on the screen only, some will write the
sorted information back to the hard disk. Linux assumes that the
first entry is hd1, and so if some utility starts sorting/reordering
! the table these things can change.

REMARK Minix does some reordering.

--- 333,340 ----
programs sort this information on the screen only, some will write the
sorted information back to the hard disk. Linux assumes that the
first entry is hd1, and so if some utility starts sorting/reordering
! the table these things can change. Moreover, do not use extended
! partition (this is in the installation notes).

REMARK Minix does some reordering.

***************
*** 446,455 ****
extensive explanatory note can be found in the mailing-list archives
(25th Jan. 92).


QUESTION: What must I do to mkfs a floppy?

! ANSWER: blocks are of size 1K so 1.44 floppy is 1440 blocks.


QUESTION: When I run kermit under Linux, I get "Warning, Read access
--- 462,478 ----
extensive explanatory note can be found in the mailing-list archives
(25th Jan. 92).

+ BTW It might be useful to set three (3) separated partitions for
+ Linux, one for the root, another one for the usr and a third one for
+ swap, as an illustration, my root partition is 10Meg, the usr is 40Meg
+ and the swap partition is 8Meg (twice the size of RAM on my box). As
+ an experience I have used MS-DOS fdisk to partition my two hd and got
+ no peculiar difficulties.

QUESTION: What must I do to mkfs a floppy?

! ANSWER: blocks are of size 1K so 1.44 floppy is 1440 blocks. The
! floppy has to be formatted before this will work.


QUESTION: When I run kermit under Linux, I get "Warning, Read access
***************
*** 466,472 ****


QUESTION: du works just fine on directories, except on / and /dev,
! moreover "ls -l" returns big number on /dev. Why?

ANSWER: This is a "feature" added in Linux 0.12; it was originally
present in Minix; more specifically, when you stat a device file
--- 489,495 ----


QUESTION: du works just fine on directories, except on / and /dev,
! moreover "ls -l" returns either big or negative number on /dev. Why?

ANSWER: This is a "feature" added in Linux 0.12; it was originally
present in Minix; more specifically, when you stat a device file
***************
*** 487,499 ****
VI. INSTALLATION HINTS
======================

QUESTION: I've got all the things on site ??? but I don't know what
goes where.

! ANSWER: include.tar.Z goes to /usr/include; ggcbin.tar.Z goes in
! /usr/local/lib except gcc which goes in /usr/local/bin. Moreover each
! gcc-xxx of /usr/local/lib should be linked with gxxx and xxx in
! /usr/local/bin. system.tar.Z contains the latest sources of the
system files (mkswap, mkfs, fsck and fdisk). In version 0.12
utilbin.tar.Z has been replaced by fileutil.tar.Z and utils.tar.Z
which contains a new tar to handle the symbolic links, make, uemacs
--- 510,526 ----
VI. INSTALLATION HINTS
======================

+ QUESTION: Where can I find the basic starting help?
+
+ ANSWER: You have to download the INSTALL notes, and more specially
+ the 0.11 and 0.12
+
+
QUESTION: I've got all the things on site ??? but I don't know what
goes where.

! ANSWER: include.tar.Z goes to /usr/include; system.tar.Z contains
! the latest sources of the
system files (mkswap, mkfs, fsck and fdisk). In version 0.12
utilbin.tar.Z has been replaced by fileutil.tar.Z and utils.tar.Z
which contains a new tar to handle the symbolic links, make, uemacs
***************
*** 501,506 ****
--- 528,550 ----
separately.


+ QUESTION: I don't know how to install gcc stuff, is there special places?
+
+ ANSWER: It depends on the release of gcc you are using. Right now
+ there are two packages : the original one gccbin.tar.Z contains all
+ the gcc-1.37 binary distrib; and quite recently there is a
+ newgcc.tar.Z which contains the gcc-1.40 stuff. Choose yours
+
+ A) gccbin.tar.Z goes in /usr/local/lib except gcc which goes in
+ /usr/local/bin. Moreover each gcc-xxx of /usr/local/lib should be
+ linked with gxxx and xxx in /usr/local/bin.
+
+ B) newgcc.tar.Z goes in /usr, then uncompress and untar it. Files are
+ directed to /usr/lib, /usr/include and /usr/bin. You have to link ar,
+ as, ld with gar, gas and gld, this will prevent some error while using
+ make (especially whilst re-compiling Linux kernel). More information can
+ be found in section VIII.
+
QUESTION: When I use the images, and i type "tar xvf ..." I got
"command not found". What did I wrong?

***************
*** 592,608 ****
QUESTION: gcc complains about not finding crt0.o and the system
include files What am I doing wrong ?

! ANSWER: The include files normal place is in /usr/include. libc.a and
! *.a should be in /usr/lib or /usr/local/lib


! QUESTION: While compiling some GNU packages gcc chocs on regex.c with
an insn code, what can I do?

! ANSWER: There is a little bug in the port of gcc, this will be
corrected on the port of v2.0 (with g++). Right now throw away the -O
flag (to compile regex) and every thing will be alright.


QUESTION: I tried to port a /new/ version of gnu stuff. But in the
linking phase, gcc complains about the missing libg.a.
--- 636,655 ----
QUESTION: gcc complains about not finding crt0.o and the system
include files What am I doing wrong ?

! ANSWER: The include files normal place is in /usr/include. lib*.a and
! *.o should be in /usr/lib or /usr/local/lib


! QUESTION: While compiling some GNU packages gcc chokes on regex.c with
an insn code, what can I do?

! ANSWER: There is a little bug in the port of gcc-1.37, this will be
corrected on the port of v2.0 (with g++). Right now throw away the -O
flag (to compile regex) and every thing will be alright.

+ BTW there are some minor bugs with this release of gcc because it was
+ compiled under linux-0.10, whith odd libraries. These problems have
+ disappeared with the new gcc-1.40 package.

QUESTION: I tried to port a /new/ version of gnu stuff. But in the
linking phase, gcc complains about the missing libg.a.
***************
*** 618,624 ****
a couple of problems with it: it isn't as straightforward as you'd
imagine. Race conditions, inclusion checking etc..

! It will befinitely be in 0.13 or 14. It's not /that/ difficult, just
needs a little thought. In the meantime, you might fake it with "cp
+recursive" and "rm -rf". Be careful with that though..
You can also write a bash script, which simulates the mvdir (really a
--- 665,671 ----
a couple of problems with it: it isn't as straightforward as you'd
imagine. Race conditions, inclusion checking etc..

! It will befinitely be in 0.13/0.95. It's not /that/ difficult, just
needs a little thought. In the meantime, you might fake it with "cp
+recursive" and "rm -rf". Be careful with that though..
You can also write a bash script, which simulates the mvdir (really a
***************
*** 714,719 ****
--- 761,767 ----


VII. FEATURES
+
=============

QUESTION: I've read that linux has virtual consoles, what must I do to
***************
*** 734,741 ****

QUESTION: What kind of shell is /bin/sh ?

! ANSWER: It's the Bourne Again Shell, If you are interested in the
! version just type sh -version and you get the response


QUESTION: Does there exist a man page for **** ?
--- 782,789 ----

QUESTION: What kind of shell is /bin/sh ?

! ANSWER: It's the Bourne Again Shell, bash-1.11 and compilation was
! straightforward (Linus dixit), just "make" that's all or nearly.


QUESTION: Does there exist a man page for **** ?
***************
*** 746,768 ****
database provided. The files in the cat1 dir are pre-formatted man
pages that the man program can use.

! BTW there is no roff,troff nor nroff for Linux.


! QUESTION: What are the editors available in linux

! ANSWER: Right now there are uemacs and elvis-1.4, some one (R. Blum)
is working on some other vi clone. The port of emacs 18.57 has been
done by John T Kohl, files can be found at the different sites
at nic it's in the directory xtra
at tsx-11 it's in the directory ports/emacs-18.57.

-
QUESTION: Does there exist a printer package for Linux?

ANSWER: There are lp patches for linux.0.12, which implement a
parallel printer interface and feature a greatly improved driver
! design. the patches are in lp.12.tar.Z


QUESTION: Does there exist a ps for Linux?
--- 794,820 ----
database provided. The files in the cat1 dir are pre-formatted man
pages that the man program can use.

! BTW there is no roff,troff nor nroff for Linux. Cawf 2.0 works just
! fine for simple man pages, and a partial ms support too.


! QUESTION: What are the editors available in linux?

! ANSWER: Right now there are uemacs, elvis-1.4, some one (R. Blum)
is working on some other vi clone. The port of emacs 18.57 has been
done by John T Kohl, files can be found at the different sites
at nic it's in the directory xtra
at tsx-11 it's in the directory ports/emacs-18.57.
+ Also the port of mg (micro gnu) has been done and can be found at
+ least at athos.rutgers.edu (128.6.4.4) in pub/linux, mg is the binary
+ and mg.tar.Z is the sources file.

QUESTION: Does there exist a printer package for Linux?

ANSWER: There are lp patches for linux.0.12, which implement a
parallel printer interface and feature a greatly improved driver
! design. the patches are in lp.12.tar.Z As I have no printer (yet), I
! don't know how good it is.


QUESTION: Does there exist a ps for Linux?
***************
*** 813,823 ****

QUESTION: Is there only the %$#@ keyboard ?

! ANSWER: There are Finnish, French, German, UK, and US keyboards. Set
! it in linux/kernel/chr_drv/keyboard.S, then compile the kernel again.

BTW a Dannish keyboard patch (Tommy Thorn) is now available look the
! digest19 vol#1 file.


QUESTION: Does there exist shared libs ?
--- 865,877 ----

QUESTION: Is there only the %$#@ keyboard ?

! ANSWER: There are Finnish, French, German,and US keyboards. Set it in
! linux/kernel/chr_drv/keyboard.S, then compile the kernel again.

BTW a Dannish keyboard patch (Tommy Thorn) is now available look the
! digest19 vol#1 file. There is an Uk keyboard too (Jan 25. by Leblanc),
! I have not seen it in the newsgroup so ask directly to
! [email protected]


QUESTION: Does there exist shared libs ?
***************
*** 859,864 ****
--- 913,1118 ----
patched version of the files changed before /any/ CO either by you or

make

+ VIII. MORE HINTS
+ ================
+ This part is quite recent and try to keep track of the different
+ information that appeared in alt.os.linux and on the list since
+ beginning of February. I have not checked all of these information.
+ But the one I know correct are marked by a (+). Fill free to contact
+ me and infirm/confirm what is written below. Moreover take care to use
+ the correct library and include stuff for linux-0.12 and the ad-hoc
+ gcc you use !!!
+
+
+ QUESTION: How can I backup my Hd under Linux ?
+
+ ANSWER: I know at least two ways. One possibility is tar and mtools,
+ another possibility is the diskbackup/diskrestore of Diamano Bolla
+ (digest44 vol. #1) which saves big hd to floppies using the
+ stdin/stdout.
+
+
+ QUESTION: How to use setterm: for the novice?
+
+ (+)ANSWER:The setterm utility provides access to most of Virtual Consoles
+ (VCs) functionality. You can set your screen up to blank at 10
+ minutes using:
+ setterm -blank 10
+
+ You can set colors, and clear the screen. For a full list of commands,
+ just type "setterm" with no arguments.
+
+ There are a few tricks with the screen dumper can really make VCs go a
+ long way. Here are a few of the common ones that I use:
+
+ setterm dump
+
+ Dumps the contents of the current VC to screen.dump (in the current dir).
+
+ setterm dump 4
+
+ Dumps the contents of VC 4 to screen.dump
+
+ setterm -file mydumpfile -dump 4
+
+ Dump the contents of VC 4 to the file mydumpfile
+
+ setterm -file /dev/tty0 -dump 4
+
+ Dumps the contents of VC 4 to the current VC.
+
+ setterm -file /dev/tty4 -dump
+
+ Dumps the contents of the current VC to VC 4.
+
+ setterm -file /dev/tty64 -dump
+
+ Dumps the contents of the current VC to the serial port.
+ Handy if you are logged on and want to paste a screen full without
+ having to resort to doing a file transfer.
+
+ setterm -file mydumpfile -append 4
+
+ Appends to instead of overwriting the dump file. Useful if you
+ have several screens you wish to concatenate.
+
+
+ QUESTION: I know there are VC, but where is the setterm stuff?
+
+ (+)ANSWER: It is in virtcons.tar.Z (at least on nic). It provides a
+ number of features that are not available with the standard release of
+ linux-0.12, among them blanking/dumping screen.
+
+
+ QUESTION: While using emacs in 80x25 mode, the mode line is constantly
+ moving around, why?
+
+ ANSWER: This appear to be a bug in the scroll region handling of the
+ console. Some one has suggested to simply remove the last entry in the
+ vt100 termcap that refers to the scroll region escape. Namely remove
+ the cs entry.
+
+
+ QUESTION: I have read that there are patches for an init/login in
+ Linux-0.12, where can I find them.
+
+ (+)ANSWER: There are many, one is the qpl-init.tar.Z package available
+ at the 3 main sites, but it is not that good (it's a quick hack).
+ There is also the Peter Orbaek init/login stuff that can be down-load
+ from ftp.daimi.aau.dk (130.225.16.1); the last version is in the file
+ poeigl-1.1.tar.Z and it is in pub/Linux-source. Among other things it
+ contains a patch to main.c of Linux kernel v0.12 (directory init). You
+ have to recompile the kernel (after patching it) to use /etc/init
+ rather than /etc/rc at boot time.
+
+
+ QUESTION: How can I change user password in login?
+
+ (+)ANSWER: There is no passwd program yet, you should do it by hand
+ with the encrypt stuff provided.
+
+
+ QUESTION: When I use make as non root, it doesn't work, why?
+
+ (+)ANSWER: ?????, the message is either (null) setuid ..., or (null)
+ setgid... I don't know how to fix it.
+ BTW This problem does not exist with the pmake (make for BSD 4.3 Reno
+ and BSD 4.4) package.
+
+
+ QUESTION: How can I get Linux to boot directly from the harddisk?
+
+ ANSWER: Right now, this can be done via the shoelace stuff or the
+ bootany package.
+
+
+ QUESTION: Sometimes, when I want to remove a directory, I get an error
+ message, is it a (known) bug?
+
+ (+)ANSWER: No, There is no bug at all, you probaly have another shell
+ on another VC whose working directory is either the one you try to
+ remove, either a subdirectory of it.
+
+
+ QUESTION: can anyone give me a sample /etc/inittab entry for login
+ from a pc attached to serial line /dev/tty65?
+
+ ANSWER: "Humberto speaking :)"
+ I've got a Hayes Smartmodem 1200 hooked up to /dev/tty65, and log
+ in daily from work. My inittab entry is
+
+ tty65:console:/bin/agetty -m -t 30 1200 tty65
+
+ but it works without the -t OK too. The real problem I suspect you're
+ having is initializing the modem properly. Before setting up logins, you
+ have to disable echo and enable auto-answer on the modem. I do this here
+ by firing up kermit at 1200, connecting, then sending 'ate0s0=5'
+ followed by enter to the modem.
+
+
+ QUESTION: When compiling some code, cc1 complains about some insn
+ code, what's that?
+
+ (+)ANSWER: An insn is an internal representation that gcc uses when
+ compiling. The main part of gcc is to take ordinary c (or c++) code,
+ and compile it, while ding optimizations in insn part, which is
+ soft/hard independant. Then another part which is hard/Os dependant
+ takes the insns and translate it in assembly language. The fix is only
+ to turn off the optimization flag (-O) or download the new gcc release
+ (1.4) which has been made available at tsx-11 (newgcc.tar.Z and the
+ ad-hoc libraries newlibc.tar.Z).
+
+ BTW I've checked tsx-11 on Sat. 22nd Feb. and nothing is available,
+ but you can find one in /incoming at ftp.eecs.umich.edu.
+
+ QUESTION: While compiling some stuff, I'm getting the following
+ error message:
+ Undefined symbol ___addsf3 referenced from text segment
+ as well as ___mulsf3 and __cmpsf2.
+ These symbols are not in the program or in it's header files.
+
+ ANSWER: These are math helper functions, and you can usually compile
+ these programs to use the kernel floating point routines by adding
+ '-m80387' to the compiler switches. If the program does any wierd
+ fp math (exp(), sin()) it'll die when you run it though.
+
+
+ QUESTION: What are the enhancement of the newgcc.tar.Z ?
+
+ (+)ANSWER: There were some bugs in the old port that have been corrected,
+ moreover this package contains 387 support
+ there is libm.a (for those with 387)
+ libsoft.a (for those without, I for example)
+ libtermcap.a (from tput 1.10)
+
+ The -mstring-insns option is no longer needed nor supported ๐Ÿ™ [As
+ an example to recompile (successfully) linux you have to throw away
+ this flag in all the Makefile]
+
+ BTW Notice also that include files have changed (stdio.h which is no
+ more ansi compliant). Right now (Feb. 24th), there are some little
+ bugs which will be fixed in a short while.
+ The port of gcc 2.0 has begun ...
+
+
+ QUESTION: I've been trying to get Linux to run on my [3/4]86 box. It
+ can't even boot. Any suggestions?
+
+ (+)ANSWER: The most common error/problem is writing the bootimage to a
+ low density disk. It fits, but the bootstrap code will only recognize
+ high density disk. So try to format explicitely disk as high density:
+ - for 3.5", 'format a: /n:18 /t:80 '
+ - for 5.25", 'format a: /n:15 /t:80 '
+
+ QUESTION: Does there exist games, languages (other than C), and
+ anything which make the system more friendly?
+
+ ANSWER: Yes, among other things there are rogue and yatzee; TeX;
+ Prolog.. but in general, if you want some extra tool port it to Linux
+ this is also a good beta-testing exercice.
+
===================8<==========>8================
+
+


~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# LaBRI | #
# 351 cours de la Liberation | e-mail: [email protected] #
# 33405 Talence Cedex | Marc-Michel Corsini #
# | #
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


[next article]
From: [email protected] (Drew Eckhardt)

Subject: Re: Problems booting off login kernel
Message-ID: <[email protected]>
Date: 25 Feb 92 00:51:22 GMT
References:
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 26
Nntp-Posting-Host: romeo.cs.colorado.edu

In article [email protected] (Sean C. Cox) writes:
>
> Yesterday, I tried making the "login" kernel for Linux. The compile

>went ok, and I dd'ed the Image to a new boot floppy. I hit my RESET
>button and tried to boot the new system. Here's where the problem hit..
> Things started up ok, I got the "Loading......" prompt, but before it
>finished the "usual" amount of dots, I got a whole bunch of
>
> @X: 0400
> AX: 020D
> E600
> CX: 0401
> DX: 0000.
>
>Where the "." after the DX: line seems (I assume) to be from the "Loading.."
>string. I can still boot off of the floppy I made from the distribution
>bootimage, but not this new one. My guess is that I was using the wrong
>bs= parameter for the dd, but I tried several values (8128, 4096, & then
>not using a bs=) and none seemd to work.
> Anyone have any ideas?

Bad floppy disk. Sector 1, track 4, head 0 not found.

Solution :

Reformat the floppy disk.


[next article]
From: [email protected] (Theodore Ts'o)

Subject: Re: bug fix for the new gcc 1.40
Message-ID: <[email protected]>
Date: 25 Feb 92 04:20:24 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 37


From: Hongjiu Lu -- Graduate Student
Date: Mon, 24 Feb 92 14:54:32 PST

I just fixed a FP bug in gnulib, a bug in sqrt(0.0) of libsoft.a and put
a new newgcc.tar.Z on tsx-11.mit.edu.

It's since been installed in ~ftp/pub/linux/binaries/usr.bin/newgcc.tar.Z.

I've also taken the time to process all of the uploaded files that have
hit ~ftp/incoming in the past few days. To make things easier on me
(I've been fairly time-crunched at work lately), I'd appreciate it if:

* People at least their name, and preferably their email
addresses as well in README files. A date is a good thing,
too.

* README files --- please! Ideally there should be a README
file bundled in the tar file, and ideal it has been bundled up
by doing "tar cvf ../foo.tar *", since that will make sure that
the README file is one of the first files in the tar file,
which makes things easier for me when I try to figure out what
the heck it is you've uploaded.

* Please send me tar files, or compressed tar files. While I
can deal with other formats, it's less convenient for me,
and I can't be sure other people can deal with them.

* If you want something processed fast, please send me mail
describing what you've uploaded and what names you've used.
If you don't send me mail, it may take a couple of days
before I find time to scan the incoming directory.

Thanks! The easier it is for me to process something, the faster it
will make it out there....

- Ted


[next article]
From: [email protected] (Drew Eckhardt)

Subject: Hackers' technotes to Nondestructive Repartitioning
Summary: Non-destructive repartitioning made easy
Keywords: partitioning, nondestructive
Message-ID: <[email protected]>
Date: 25 Feb 92 06:39:53 GMT
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 460
Nntp-Posting-Host: caesar.cs.colorado.edu

Reformatting your hard disk, and reinstalling all your DOS programs from
scratch can be a real nuissance, and is unecessary.

Breathe a sigh of relief : it is possible to non-destructively repartition
hard disks.

Notes : Before continuing, make sure you have a RECENT BACKUP.

I assume that you understand hex arithmatic, and are not
afraid of a little assembler or DEBUG.

Actually, a decent partition and disk editor will get you around
this - NU works gret.

Of course, the modern programmer doesn't use anything but a
source level debugger - so I've included some helpful hints
and the command syntaxes. However, there is no room for
handholding here : if you screw up, you might have to use
that backup. Don't do it unless you are confident in your
abilities.

Also, this procedure only works with NON-EXTENDED DOS partitions,
< 64K logical sectors, (DOS 4 large partitions add additional BPB
fields that I am unsure of - roughly the same procedure applies there
though. According to Townsend' Advanced MS DOS : Expert Techniques
for programmers

offset 26h will have the signature byte 29h if this is
the case, 20h a dword containing the number of sectors if
volume size > 64K sectors

I still use MS-Loss 3.3, with an ~82M partition under disk mangler
and fall into the tested category)

Large partitions, handled by a third party partition manager
and handled so that there are < 64K logical sectors
work - this was the case with my SCSI disk.

I will lay down what general procedures you need to know (
required to read / write the raw disk), as well as the
data structures we are dealing with. Then I will proceed
with the entire procedure, which applies the general procedures
in reading and modifying the data structures. If it looks like
a tech manual - it is. If you don't grok non-destructive
repartitioning, you shouldn't be doing it.

DEBUG has a Hexaritmatic command, h which will add and
subtract the two operands. You may find this useful.

IE : I have loaded sector 0 into memory at 0200, and wish to know
the address of the partition table at 1be.

-h 200 1be
03BE 0042

Where 03BE is the sum, and 042 the difference.

DEBUG prints a segment before the offset : note that your segments
will probably not match. The offset is what's important.

The 80x86 family is LITTLE ENDIAN. This means least significant
byte first - ie the internal representation of 0x12345678 would be
78 56 34 12. When dealing with multi-byte quantities, keep this
in mind.

When I say word, I mean word as in the Intel documentation :
16 bits. dword is 32 bits.

DISK BIOS addresses cylinder, head as zero based, sector
as 1 based. Same thing for the partition table.

DOS addresses sectors as 0 based, from the start of the
logical partition, and as logical sectors which may
consist of multiple physical sectors.

Unless otherwise noted, all numbers are hex.

You're better off using Norton Utilities - but Debug works fine
too.

This document is sort of a quick hack 8^)

Tools :

Required : DEBUG, and a disk defragmenter

Optional : partition editors (NOTE!!!! Make sure these DO NOT perform
any formatting, and allow you to edit partitions in the
REAL order they appear on the disk.), the Linux FDISK program,
utilities that save an image of the boot sector FATs, and
directory (IE Norton's Format Recover), a raw disk editor
(Norton Utilities NU)..

Procedures :
Editing memory with debug :

d ADDRESS l LENGTH

will dump memory

f ADDRESS l LENGTH values

will fill memory.


Reading and writing DOS logical sectors (using debug):

Reading is accomplished using the debug l command.

l ADDRESS DRIVE SECTOR COUNT

Where ADDRESS is the hex address of where to put the data,

DRIVE is a 0 based drive number (IE A:=0, B:=1, C:=2, etc. If there
is only one floppy drive, it is considered both A: and B: as it is
in DOS)

SECTOR is a zero based sector number. Of interest to us are
sector 0, the boot sector, and the sectors immediately following it -
the FAT's.


COUNT is the number of sectors to read.


So, to read in the first 10 sectors of the FAT on my E: partition (the
FAT starts at sector 1), storing them at 0200, I would enter

l 0200 4 0 A

Writing is done with the W command, which takes the same
parameters. Assuming I had edited the boot sector of E: at 0200 in memory,
and wanted to write it back to disk, I would type in

w 0200 4 0 1


Using Norton Utilities NU:

xplore disk, hoose Item, ector, dit/display

Unfortunately, absolute sector 0 falls outside of all partitions (this
is where the partition table is), and we need to use a different
procedure for it.

Reading / Writing absolute sector 0:
The following debug / assembler interaction shows how to read absolute disk
sector 0 , replace xx with 80 for hard disk 0, 81 for hard disk 1:

-a 0100
1984:0100 mov ax, 0201
1984:0103 mov bx, 0200
1984:0106 mov cx, 0001
1984:0109 mov dx, 00xx
1984:010C int 13
1984:010E int 20
1984:0110 ^C

-


-g = 0100

This will read sector 0 into DS:0200. To write it back,

-a 0100
1984:0100 mov ax, 0301
1984:0103 ^C


-g = 0100.


Using Norton Utilities NU :
Under xplore disk, choose hoose item, bsolute sector, dit / display



Structures :

1. The Partition table

The partition table resides at absolute sector 0 (ie
cylinder 0, head 0, sector 1) on all harddisks. It is accessed
by a short bootstrap loader on that sector, which reads the partition
table and then picks a partition from which to load the boot sector for
the operating system.


The partition table itself resides at offset 1be. It is 64 bytes (decimal)
in length, plus the two byte signature 55 AA. When dealing with the
partition table, make sure byte 40 (offset 1fe of the sector) is 55 and
byte 41 (offset 1ff of the sector) is
of the sector




The partition table is subdivided into FOUR 16 byte entries, fielded
as follows :
offset length field
0 byte bootable 80h = bootable, 0 = not
1 byte starting head number
2 word starting cylinder (and sector - sector is 1 based
high byte is low byte of cylinder, low byte low 6 bits is
sector, high 2 bits of low byte high 2 bits of cylinder)


typically, sector = 0.

4 byte system 1 = primary DOS, 12 bit, 4 = primary DOS 16 bit,
5 = extended DOS, 8 = NON-DOS (might be usable)
5 byte ending head number


6 word ending cylinder / sector

8 dword starting sector (relative to begining of disk - THIS IS ZERO
based)

C dword number of sectors



My partition table on drive 0 looks like :

-d 3be
1984:03BE 80 01 ..
1984:03C0 01 00 01 05 22 08 22 00-00 00 0A 07 00 00 00 00 ....".".........
1984:03D0 01 09 51 05 E2 2B 2C 07-00 00 E4 7F 02 00 00 00 ..Q.b+,...d.....
1984:03E0 00 00 00 00 00 00 00 00-00 00 00 00 00 00 00 00 ................
1984:03F0 00 00 00 00 00 00 00 00-00 00 00 00 00 00 55 AA ..............U*

Interpreting this, we can see that I have two partitions in use -
partition 1 and 2, the rest being blank.


My first partition is bootable (80)
FAT size is 12 bits (type = 01)
It starts at head 1, cylinder 0, sector 1, which is sector 22
(34 decimal) relative to the start of the disk.
It is 0000070a sectors in length (1802 decimal), and ends on
head 5, cylinder 8, sector 22 (34 decimal).


My second partition is non-bootable (00).
It is type 51 (Disk Mangler) meaning I need to
find out fat size some other way.

It starts at Head 0, cylinder 9, sector 1, which is sector 72c
(1836 decimal) relative to the start of the disk.
It is 27FE4 (163812 decimal) sectors in length, and ends on
head 5, cylinder 32b (811 decimal), sector 22 (34 decimal)..

See how I got ending cylinder? In Hex, each digit is a nibble. You can
easily convert to binary a nibble at a time. IE E2 becomes 1110 0010


The low 6 bits (sector) are 10 0010 = 22
The high 2 bits are 11 = 3

So high byte cylinder is 03, low byte is 2B so cylinder = 032b

Got it?



2. The Boot sector

The second important piece of data is the bootsector. There are a number
of fields we are interested in. I have ommitted the DOS 5 extended fields
(Can't give you an answer I'm 100% sure on), as well as fields unecessary
to our procedure.

Fields we are interested in :
Offset Size Field
b word bytes per LOGICAL sector - divide by 512 to get physical
to logical mapping

d byte sectors per cluster. Multiply by logical sector size and
divide by 1024 (decimal) to get K / cluster

13 word total number of LOGICAL sectors. This is one field
extended by DOS5.

18 word sectors per track

1a word heads

Clusters = sectors / sectors per cluster.

Since my first partition on drive 1 is fairly boring, we'll look at
partition 2 - E:

-l 0200 4 0 1
-d 0200
1984:0200 EB FE 90 4E 4F 53 59 53-54 45 4D 00 08 04 01 00 k~.NOSYSTEM.....
1984:0210 02 00 02 F9 9F F8 0A 00-22 00 06 00 2C 07 80 00 ...y.x.."...,...

The first three bytes are either a near jump over the actual BPB or
a short jump and a NOP. There follows an eight byte system ID, and then
the Good Stuff.

I have 800 (2048 decimal) bytes per logical sector, meaning 4 physical
sectors = 1 logical sector.
.
There are 4 logical sectors per cluster, meaning there is 8K in
each cluster, 16 logical sectors per cluster.

There are 9ff9 (40953 decimal) logical sectors in my partition, meaning
163812 decimal physical sectors (Notice how this number matches the
number in the partition table.)

There are 22 (34 decimal) sectors per track, and 6 heads.

3. The FAT

The final interesting data structure is the FAT. If you have some other
reliable means of insuring all data is at the begining of the partition
you can ignore this.

If there are > 4096 clusters, you definately have a 16 bit FAT.

The FAT will begin with an FFF8 signature - 16 bit or (FF8) 12 bit.
What follows
is one entry for each cluster of the disk - with FFFF (FFF) marking
end of file, FFF7 (FF7) marking a bad sector, 0000 (000) empty, basically
anything else a next cluster pointer.

Basically, all that were interested in is that there are roughly enough
0 entries all together, with only BAD marks and no end of file or next
cluster pointers.

12 bit entries are packed two per three bytes - low byte = low byte
first cluster, middle byte low nibble is high nibble for the first cluster,
middle byte high nibble is low nibble for the second cluster, high byte
is high byte for the second cluster.


Step 1:
Make a FULL backup.

1a:
If you have a program to save off an image of the boot sector,
do it now. Dumping the boot sector / partition sectors to
floppy disk (do an rcx:200 in debug, then a w FILENAME to
save to a file) would also accomplish the same thing.

This will save you the trouble of having to restore an entire
backup if you screw up either of these sectors.

Step 2:

Defragment. All data must be contiguous.


Step 3:
Decide where the partition goes, and how big it is.

The new partition must fullfil these requirements :

1. It resides at the end of the original partition.
2. It is situated such that the original partition has a boundary
that ends on a FULL logical sector (In my example, this is four
physical sectors).

In my case, I am simply starting testing on my already packed SCSI
disk, and want a ~1 megabyte partition, or 2048 sectors.

This partition ends at :

cylinder 32b (812 decimal), head 5, sector 22 (34 decimal)

If I divide by 34 sectors per track (available in the boot record),
I get 60 tracks. If I divide by 6 heads, I get 10 cylinders.

So, I will start my partition at

cylinder 802 decimal (322 hex)
sector 0
head 0
relative sector 802 * 34 * 6 decimal= 163608


I will adjust the size of my E: drive down by 510 logical (2040) (decimal)
physical sectors, and make this my new partition.

Note that this falls on the required logical sector boundary.

Likewise, I will reduce the size of my second partition by 2040 (decimal)
sectors, to 161772 (decimal) sectors and reduce the ending cylinder to
801 (decimal).

Step 4:

Verify that all data is contiguous, and we have enough free space
for the partition we are about to create.

Use a reliable disk space map utility, if none is available you will
have to do so manually. Again, NU can fill this role.


To do this manually, first determine FAT size

check - partition table
- number of clusters (roughly - (sectors - 1) / sectors per cluster)
if > 4096, definately 16 bit

After you've decided on 12 bits or 16 bits,

Find the first 0 cluster (The disk has been defragmented?)

Count consecutive 0 or BAD FAT entries (1 per 2 bytes on 16 bits,
3 per 2 bytes on 12 bits), multiply by K per cluster (as determined
from the boot sector - see data structures above), and be happy.

If there are intervening NEXT CLUSTER pointers, or END OF FILE (FFFF),
try a different defragmenter.

Also, not that some defragmenters will not move hidden files. You will
have to unhide these, defragment, and rehide these.



5. Adjust boot record of old DOS partition. Change # sectors field,
offset 13 hex (word), be sure to adjust in number of
LOGICAL SECTORS, not physical sectors (consult sector
size field to determine physical / logical sector mapping)

After image for my example :
-l 0200 4 0 1
-d 0200
1563:0200 EB FE 90 4E 4F 53 59 53-54 45 4D 00 08 04 01 00 k~.NOSYSTEM.....
1563:0210 02 00 02 FB 9D F8 0A 00-22 00 06 00 2C 07 80 00 ...{.x.."...,...

6. Adjust partition table. Change ending head, cylinder, sector
field of old partition, total sectors of old partition,
add new partition.

Avoid overlap. Note that new partition table becomes active
on reboot.

After image for our example:
-d 03be
1563:03BE 80 01 ..
1563:03C0 01 00 01 05 22 08 22 00-00 00 0A 07 00 00 00 00 ....".".........
1563:03D0 01 09 51 05 E2 21 2C 07-00 00 EC 77 02 00 00 00 ..Q.b!,...lw....
1563:03E0 C1 22 FF 05 E2 2B 18 7F-02 00 F8 07 00 00 00 00 A"..b+....x.....
1563:03F0 00 00 00 00 00 00 00 00-00 00 00 00 00 00 55 AA ..............U*

Yes, I know 0xff is a bizarre identifier for the linux partition. That's
how I did my first one, that's how I did this one.

References :
Norton, Peter Norton's Programmer's Guide to the IBM PC
Townsend, Advanced MS-DOS Expert Techniques for Programmers

Hopefully this has been enlightening, and possibly even useful to some.

Mail comments to [email protected]


[next article]
From: [email protected] (Marc CORSINI)

Subject: FAQ (sorry for the delay)
Message-ID: <[email protected]>
Date: 25 Feb 92 06:48:26 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Marc CORSINI)
Organization: The Internet
Lines: 1117


Hi, sorry for the delay but my system breakdown after the posting of CDIFF

Hi Linuxers!

The originall FAQ 1st version was posted in Dec. 19, by Robert Blum,

Most credits of this work to Linus, Robert and Ted, the rest was
either on the list posted by many (real) activists, not me ;-), either
in some other news groups, or else by direct posting to me (thanks
Humberto). I haven't copyrighted them, so thanks to every one who
participated even indirectly to this FAQ.

[To find what has changed, view the `Cdiff-faq' in my previous posting.
The last-change-date of this posting is always "two minutes ago". :-)]

This is the introduction to a list of frequently asked questions (FAQ
for short) about Linux with answers (Yeap!). This article contains a
listing of the sections, followed by the question/answer part.

This FAQ is supposed to reduce the noise level ๐Ÿ˜‰ in the alt.os.linux
newsgroup (and also the 'linux-activists' mailing list), and spare the
time of many activists. I will post it twice a month, since there are
more and more new incomers, and new features.

BTW I think this FAQ should be available at the main Linux sites in
the doc directory (have you read this Ary, Robert, Ted ?)

Please suggest any change, rephrasing, deletions, new questions,
answers ...
Please include "FAQ" in the subject of messages sent to me about FAQ.
Please use [email protected] whatever will be the From part
of this message.


Thanks in advance,
Marc

Remind the vote for comp.os.linux;
post the YES vote to [email protected]
The voting period will last through 00:00 UTC (7:00pm EST), 19 March
1992.

Future Plan:
- mark question/answer that have heavily changed recently
- verification/location/organization for files available
via FTP (I've tried what a mess!!)
- cross posting this to news.answers as soon as comp.os.linux
is created (may be earlier, on march)

================================8<=====8<==============================
CONTENTS
I. LINUX GENERAL INFORMATION
II. LINUX USEFUL ADRESSES
III. INSTALLATION and SECURITY
IV. LINUX and DOS
V. SOME CLASSICAL PROBLEMS
VI. INSTALLATION HINTS
VII. FEATURES
VIII. MORE HINTS

I. LINUX GENERAL INFORMATION
=============================

QUESTION: What is linux?

ANSWER: Linux is a small unix for 386-AT computers, that has the added
advantage of being free. It is still in beta-testing, but is slowly
getting useful even for somewhat real developement. The current
version is 0.12, date: Jan. 14th 1992. The next version will come out
in the end of February, beginning of March as the Linux v0.13/0.95

LINUX 0.12 is a freely distributable UNIX clone. It implements a
subset of System V and POSIX functionality. LINUX has been written
from scratch, and therefore does not contain any AT&T or MINIX
code--not in the kernel, the compiler, the utilities, or the
libraries. For this reason it can be made available with the complete
source code via anonymous FTP. LINUX runs only on 386/486 AT-bus
machines; porting to non-Intel architectures is likely to be
difficult, as the kernel makes extensive use of 386 memory management
and task primitives.


QUESTION: What is the current state of Linux?

ANSWER: do "finger [email protected]", or read the
alt.os.linux newsgroup.


QUESTION: I've just heard about linux, what should I do to get it?

ANSWER: First read all this FAQ, and the INFO-SHEET monthly post, then
go to the nearest ftp site (see below), download the Images there are
two a rootimage and a bootimage (in general in the images directory),
download the INSTALL and RELNOTES files. Find the rawrite utility
(for example at tsx-11 it's in /pub/linux/INSTALL), then rawrite the
images on high density floppies (5.25 or 3.5), finally boot on the
root diskette and that's it.


QUESTION: Does it run on my computer?

ANSWER: Linux has been written on a clone-386, with IDE drives and a
VGA screen. It should work on most similar setups. The harddisk should
be AT-standard, and the system must be ISA. A high density floppy
drive -- either 5".25 or 3".5

Otherwise the requirements seem relatively small: a 386 (SX, DX or any
486). Any video card of the following: Hercules, CGA, EGA, (S)VGA.

It needs at least 2M to run, and 4M is definitely a plus. It can
happily use up to 16M (and more if you change some things). People
are working on a SCSI-driver.

BTW it works for some ESDI drive (Joincom controller with Magtron
drive after you have commented out the "unexpected hd interrupt"-message
from hd.c). And quite recently for some SCSI adapter (adaptech-1542)
by Tommy Thorn in digest19 vol#1 of alt.os.linux.


QUESTION: How would this operate in an OS/2 environment?

ANSWER: Fine.


QUESTION: Will linux run on a PC or 286-AT? If not, why?

ANSWER: Linux uses the 386 chip protected mode functions extensively,
and is a true 32-bit operating system. Thus x86 chips, x<3, will
simply not run it.


QUESTION: How big is the 'complete' Linux package?

ANSWER: Well, the boot and root image diskettes are about 750k
compressed. The kernel sources are about 200k compressed, and the
libc sources are another 170k compressed. The GNU C compiler is 670k,
and the other miscellaneous unix utilities are probably a bit over a
megabyte.

Now add sources to whatever you want to port and compile yourself.
The sources to GNU emacs are about 3 megabytes, compressed. Groff (a
troff replacement) is just over 1 megabyte.

If you think this is big, remember that the OS/2 2.0 Limited
Availability release is 20 1.44 megabyte diskettes.


QUESTION: Since how long does linux exist?

ANSWER (partial): Few months, v0.10 went out in Nov. 91, v0.11 in Dec.
and the current version 0.12 is available since Jan. 14th 92. But even
it is pretty recent it is quite reliable. There are very few and small
bugs and in its current state it is mostly useful for people who are
willing to port code and write new code.


QUESTION: What's about the copywright of linux.

ANSWER: In the previous release, Linux was copyrighted by Linus
Torvalds but free. The copyright will change (v0.13 /0.95 ?) to
follows the same set of rules as the GNU copyleft.


QUESTION: Should I be a UNIX and/or a DOS wizard to install/use Linux?

ANSWER: Not at all, just follow the install rules, of course it will be
easier for you if you know things about Unix. Right now Linux is used
by more than 400 persons, very few of them enhance the kernel, some
adds/ports new soft and most of us are only (but USEFUL) beta testers.
So join us and choose your "caste"


QUESTION: What are the differences, pros and cons compared to Minix ?

ANSWER (partial):
Cons:
- Linux is not as mature as Minix, there is less working software right now.
- Linux v0.12 doesn't have real login yet, you are root all the time.
this will change pretty soon in the next version (end Feb.). Right
now, you can use the Peter Orbaek init/login package (see VIII.)
- Linux only works on 386 and 486 processors.
- Linux needs 2M of memory just to run, 4M to be useful.
- Linux is a more traditional unix kernel, it doesn't use message passing.

Pros:
- Linux is free, and freely distributable.
- Linux has some advanced features such as:
- Memory paging with copy-on-write
- Demand loading of executables
- Page sharing of executables
- Multi-threaded file system
- job control and virtual memory, virtual consoles and pseudo-ttys.
- Linux is a more traditional unix kernel, it doesn't use message
passing.


QUESTION: Does Linux use TSS segments to provide multitasking?

ANSWER: Yes!


QUESTION: If my PC runs under Linux, is it possible to ftp, rlogin,
rsh etc.. to other Unix boxes?

ANSWER: Not yet, but kermit has been ported to Linux, and the ka9q too.


QUESTION: Does linux do paging? Can I have virtual memory on my small
machine?

ANSWER: Linux0.12 does do paging, but it's pretty rudimentary. It's
even more beta than the rest of linux. How it selects which page to be
thrown out of memory is particularly primitive. It is there though and
it does appear to work


QUESTION: Can I have tasks spanning the full 4GB of addressable 386
memory? No more 64kB limits like in coherent or standard minix?

ANSWER: Linux does limit the task-size, but at a much more reasonable
64MB (MEGA-byte, not kilos), so bigger programs are no problem.


QUESTION: Does the bigger program sizes mean I can run X?

ANSWER: X is not ported to linux, and I hope it will be some day,
people are working hard on it but it's big, and wants a lot from
the system.


II. LINUX USEFUL ADRESSES
=========================

QUESTION: Where can I get linux?

ANSWER: Linux can be gotten by anonymous ftp from
nic.funet.fi (128.214.6.100):
directory /pub/OS/Linux
Tupac-Amaru.Informatik.RWTH-Aachen.DE (137.226.112.31):
directory /pub/msdos/replace
tsx-11.mit.edu (18.172.1.2):
directory /pub/linux
ftp.eecs.umich.edu (141.212.99.7):
directory linux
puffin.doc.ic.ac.uk (146.169.3.7):
directory /pub/os/Linux
ustsun.s.u-tokyo.ac.jp (133.11.11.11):
directory misc/linux

You might want to check out which of these is the most up-to-date.

If you don't have ftp-capability, you are in trouble. See next Q/A. If
you have no uncompress utility, there are a lot even for DOS, have a
look on SIMTEL, or else use facilities provided by some sites to
uncompress for you. Don't do that if you can, because it's lengthy,
expensive and causes troubles to other users on ftp sites.


QUESTION: I do not have FTP access, what can I do to get linux?

ANSWER: Try to contact a friend on the net with those access, or try
mailserver/ftpmail server otherwise contact [email protected] You
might try mailing "[email protected]" with "help" in the body of
the mail. If you choose ftpmail server (example: [email protected]),
with "help" in the body, the server will send back instructions and
command list. As an exemple to get the list of files available at tsx-11
in /pub/linux send:

mail [email protected]
subject: anything
reply
connect tsx-11.mit.edu
chdir /pub/linux
dir -R
quit

QUESTION: Is there a newsgroup or mailing-list about linux? Where can
I get my questions answered? How about bug-reports?

ANSWER: alt.os.linux is formed, and comp.os.linux is on the way, for
those who can't access to the news you can ask for digest to:
[email protected] On the other hand, mail
sent to [email protected] are posted to
alt.os.linux

DO NOT mail "I want to [un]subscribe" to the newsgroup, use
the request-address. IF not your mail-box will be over-crowded by
activists.

Questions and bug-reports can be sent either to the newsgroup or to
"[email protected]", depending on which you find more
appropriate.

BTW People are working on the organization of Linux, this is done on
a separate mailing-list.

linux-standards: Discussion of distribution and directory standards
for the Linux operating system, including directory structure, file
location, and release disk format.

Requests to be added to this list must be sent to:
[email protected]

QUESTION: Does there exist a place where the traffic of the newsgroup
is kept?

ANSWER: Yes, on nic and tsx-11 (see the ftp adresses above)

III. INSTALLATION and SECURITY
==============================

QUESTION: How can I be sure I won't be writing over anything
important? I have to use DOS on my machine, and I don't want to
lose any files.

ANSWER: Back up everything. Just in case. Then, write some easily
recognizable pattern to the partition you have reserved for linux,
using some DOS tool. You can then use "cat /dev/hdX" under linux to
examine which of the partitions you used.


QUESTION: Linux mkfs doesn't accept the size I give the device,
although I double-checked with fdisk, and it's correct.

ANSWER: Be sure you give the size in BLOCKS, ie 1024 bytes, not
sectors. The mkfs doesn't work for very big partition (over than 64
Megs). Also, make doubly certain that you have the correct partition.

There are a few rules about this: /dev/hd0 and /dev/hd5 are always the
same under linux and minix. DO NOT USE THEM, they are the whole raw
disk, not partitions. Also if a partition is on drive 1 under minix
(ie /dev/hd1-4), it is drive 1 under linux as well. Moreover, there
is no real consensus on whether partition #1 is the first partition on
the disk, or is the first entry in the partition table. Some parition
programs sort this information on the screen only, some will write the
sorted information back to the hard disk. Linux assumes that the
first entry is hd1, and so if some utility starts sorting/reordering
the table these things can change. Moreover, do not use extended
partition (this is in the installation notes).

REMARK Minix does some reordering.

A useful hack is to make each of your partitions a different size.
Then after any editting or possible change to the partition table you
can boot a floppy system and run fdisk (linux's, not DOS) to see if
the assignments still hold.


QUESTION: I have a one partitionned 40Mb disk. If I run mkfs, what
happens?

ANSWER: If you do that, you will have an empty 40Mb Linux file system
(currently the same fs as Minix). You should, at least, make on your
hard disk, one partition per operating system you want to use.


QUESTION: I mounted the linux filesystem, and copied the files from
the root-disk to the harddisk. Now I cannot find them any more, and
somethimes linux dies with a "panic: trying to free unused inode".

ANSWER: You have probably forgot to sync before rebooting. Linux, like
all unices, use a "buffer cache" to speed up reads and writes to disk.
On a machine that has enough memory, this buffer-cache is 1.5MB, and
if you forget to sync before exiting, it may not be fully written out
to disk. Re-mkfs and re-install (or try to use the preliminary fsck,
but remember that although fsck tries to correct the faults it finds,
it may fail.)


IV. LINUX and DOS
=================

QUESTION: Is it possible to access to DOS world from Linux

ANSWER: Yes, there is the mtools package (with patches for devices.c)
The original sources of mtools can be found at any places not only at
nic, tupac and tsx-11, and the patches for Linux (with fix for big DOS
partitions are in the directory patches or ports). Moreover you should
download the file patch.Z to apply patches ๐Ÿ™‚

QUESTION: the mtools package won't work. I get an ENOENT error message
for all devices.

ANSWER: mtools needs to be told which device to look for. Use 'ln' or
'mknod' to create a special file called "/dev/dosX", where X is A, B,
C, X or Y. A and B are for floppies (12 bits), C is for hard disk and
X, Y for any. This file should point to the device you want to read.
About the minor/major pair have a look in section INSTALLATION HINTS.


QUESTION: What is as86.tar.Z ?

ANSWER: It's the port of Bruce Evans' minix assembler, you need it to
be able to recompile Linux at your convenience.


QUESTION: Turbo (Microsoft) Assembler won't compile the Linux boot
code. In fact, some of the opcodes in these files look completely
unfamiliar. Why?

ANSWER: The Linux boot codes are written in Bruce Evans' minix
assembler, which has the same opcodes as the original minix assembler
ported to linux get as86.tar.Z Anyway there are a few differences
between these and normal DOS assemblers:

- No segments - everything is in the same segment (at least in the
bootsectors and setup, as they don't use the .data segments)

- mov[b|w|l] are shorter versions of mov ax,[byte|word|long] ptr
[XXX].
This is how unix assemblers normally give the size (byte, word or
long).
Gas has similar constructs.

- There is no "jmp short", the opcodes are "j" for a short jump and
"jmp" for a long one.

- "jmpi" is a jump with a segment:offset pair. I don't know how this
is
written in DOS assembly.


V. SOME CLASSICAL PROBLEMS
==========================

QUESTION: While running du I get "Kernel panic: free_inode: bit
already cleared". Also, du produces a ENOENT error for all the files
in certain of my directories. What's going on?

ANSWER: These are both consistent with a bad file-system. That's
relatively easy to produce by not syncing before rebooting, as linux
usually has 1.5MB of buffer space held in memory (unless you have <=4M
RAM, in which case the buffers are only about 0.5MB). Also linux
doesn't do anything special about the bit-map blocks, and as they are
used often, those are the thing most likely to be in memory. If you
reboot, and they haven't been written to disk ...

Just do an fsck on the device, the -a flag might repair it otherwise,
the only thing to do is to reinstall the filesystem from the Images.

A sync is done only every 30 seconds normally (standard unix
practice), so do one by hand (some people think you should do 3 syncs
after each other, but that's superstition), or by logging out from the
startup-shell, which automatically syncs the system. Unmounting a
filesystem also syncs it (but of course you can never unmount root).

Another (sad) possibility is that you have bad blocks on your disk.
Not very probable, as they would have to be in the inode-tables, just
a couple of blocks in size. Again there aren't programs available to
read a disk for bad sectors and put them in some kind of
"bad-sector-file". On IDE drives this is no problem (bad sectors are
automatically mapped away).


QUESTION: How can I partition my hard-drive to use Linux?

ANSWER: There are (at least) two ways to answer this. The easy way is
probably to use a program which will do it for you, such as the MS-DOS
fdisk, Minix fdisk, Xenix/Unix fdisk, or programs such as edpart.exe
or part.exe.

On the other hand, you can use a disk editor and modify the contents
of the partition table directly. This has been already done, and an
extensive explanatory note can be found in the mailing-list archives
(25th Jan. 92).

BTW It might be useful to set three (3) separated partitions for
Linux, one for the root, another one for the usr and a third one for
swap, as an illustration, my root partition is 10Meg, the usr is 40Meg
and the swap partition is 8Meg (twice the size of RAM on my box). As
an experience I have used MS-DOS fdisk to partition my two hd and got
no peculiar difficulties.

QUESTION: What must I do to mkfs a floppy?

ANSWER: blocks are of size 1K so 1.44 floppy is 1440 blocks. The
floppy has to be formatted before this will work.


QUESTION: When I run kermit under Linux, I get "Warning, Read access
to lock directory denied". What am I doing wrong?

ANSWER: Nothing, you just need to create /usr/spool/uucp, which is
where kermit like to lock files.


QUESTION: du seems buggy when i used it the number of disk occupation
is wrong.

ANSWER: Take care, if you want size in kbytes use the -k flags.


QUESTION: du works just fine on directories, except on / and /dev,
moreover "ls -l" returns either big or negative number on /dev. Why?

ANSWER: This is a "feature" added in Linux 0.12; it was originally
present in Minix; more specifically, when you stat a device file
belong to a block device, it will return the maximum size of the block
device in the st_size member of the stat structure. If you don't like
it, it's very simple to patch it out. Look in the fs/inode.c, in the
subroutine read_inode().


QUESTION: When I try to (un)compress many files in one command, the
command partially fails?

ANSWER: This is a bug, many partial fixes are floating around but ..
You can solve it by a bash command "for i in whateverfiles;do
compress $i; done"


VI. INSTALLATION HINTS
======================

QUESTION: Where can I find the basic starting help?

ANSWER: You have to download the INSTALL notes, and more specially
the 0.11 and 0.12


QUESTION: I've got all the things on site ??? but I don't know what
goes where.

ANSWER: include.tar.Z goes to /usr/include; system.tar.Z contains
the latest sources of the
system files (mkswap, mkfs, fsck and fdisk). In version 0.12
utilbin.tar.Z has been replaced by fileutil.tar.Z and utils.tar.Z
which contains a new tar to handle the symbolic links, make, uemacs
kermit and minor programs (sed,...). Other utilities have been ported
separately.


QUESTION: I don't know how to install gcc stuff, is there special places?

ANSWER: It depends on the release of gcc you are using. Right now
there are two packages : the original one gccbin.tar.Z contains all
the gcc-1.37 binary distrib; and quite recently there is a
newgcc.tar.Z which contains the gcc-1.40 stuff. Choose yours

A) gccbin.tar.Z goes in /usr/local/lib except gcc which goes in
/usr/local/bin. Moreover each gcc-xxx of /usr/local/lib should be
linked with gxxx and xxx in /usr/local/bin.

B) newgcc.tar.Z goes in /usr, then uncompress and untar it. Files are
directed to /usr/lib, /usr/include and /usr/bin. You have to link ar,
as, ld with gar, gas and gld, this will prevent some error while using
make (especially whilst re-compiling Linux kernel). More information can
be found in section VIII.

QUESTION: When I use the images, and i type "tar xvf ..." I got
"command not found". What did I wrong?

ANSWER: Nothing, in the distribution the tar is in compressed form
(lack of place). You have first to copy tar.Z on another disk/diskette
and uncompress it, this command is available on your diskettes.


QUESTION: What is the "em" binary?

ANSWER: Em is micro-EMacs (probably version 3.10).


QUESTION: It seems that $#@! ported on linux don't run correctly

ANSWER: Possible, but check first if the size of your file corresponds
to the one on the ftp sites, if it is then do a complete report of the
error, try to correct it.


QUESTION: Does anyone port this to linux?, if not i'll compile it

ANSWER: First check on the sites, have a look to the info-sheet
monthly post and also available on sites. Have a look in the "old"
digest files and mail-archives of linux-activists, these are kept at
least at tsx-11 and nic possibly at tupac. Look also at the GNU(*)
utilities to see if someone has already written a freely distribuable
version. Ask then on the list/news.


(*) GNU stands for GNU's Not Unix, which (besides being a recursive
acronym) is a project started by the Free Software Foundation (the FSF)
to write a freely distributable version of Unix. The GNU kernel is
named HURD, and is based on Mach. It is currently being written, and is
not yet done. Many of the GNU utilities, however, are completed and are
much more functional than the original Unix utilities. Since they are
freely available, Linux is using them as well.


QUESTION: I've ported *** to Linux, what should i do to add it in the
standard distribution?

ANSWER: Read first the previous Q/A, then to make something available to
others you have to post it (with cdiffs of the source, a short README if
needed) in the incoming directory of one of nic,tupac,tsx-11, then drop
a short note to the list/group and to the site advisor.
On nic it's [email protected] (Ari Lemmke)
On tupac it's [email protected] (Robert Blum)
On tsx-11 it's [email protected] (Ted Ts'o)


QUESTION: I want to port *** to Linux, what are the flags?

ANSWER: Recall that Linux implements subset of SYSV and POSIX, so
-DUSG and -DPOSIX work in general. Moreover throw away most of the ld
flags such as -ltermcap, -lg, since the libg.a and libtermcap.a are
missing.


QUESTION: Linux lacks on ****/ Linux has a bug in ***, what are the
rules to enhance/correct the kernel?

ANSWER: Before anything check if some one else is working on that
subject, contact those people. Test your improvment (it should work is
NOT enough), then send the patches in cdiffs form to Linus and/or the
list, moreover the localization must be clear. This does NOT mean that
bug-reports and patches are not accepted.


QUESTION: I seem to be unable to compile anything with gcc. Why?

ANSWER: If you have only 2 MB RAM, gcc will die silently without
compiling anything. You must have at least 4 MB to do compilations

BTW Since swapping is possible, I have heard that compilation is
possible with only 2Meg and a lot disk traffic ๐Ÿ™‚ Isn't it great?

QUESTION: I'm using a program that uses signal handlers which are
installed using sigaction() with the SA_NOMASK, and they get a general
protection error right after the signal handler tries to return.
What's going wrong?

ANSWER: You are using a libc.a that has an out-of-date signal.o and
sig_restore.o file, and they don't know how to deal with SA_NOMASK.
(The one in gccbin.tar.Z is out-of-date). Get the new libc.a and put
it in /usr/lib or /usr/local/lib


QUESTION: gcc complains about not finding crt0.o and the system
include files What am I doing wrong ?

ANSWER: The include files normal place is in /usr/include. lib*.a and
*.o should be in /usr/lib or /usr/local/lib


QUESTION: While compiling some GNU packages gcc chokes on regex.c with
an insn code, what can I do?

ANSWER: There is a little bug in the port of gcc-1.37, this will be
corrected on the port of v2.0 (with g++). Right now throw away the -O
flag (to compile regex) and every thing will be alright.

BTW there are some minor bugs with this release of gcc because it was
compiled under linux-0.10, whith odd libraries. These problems have
disappeared with the new gcc-1.40 package.

QUESTION: I tried to port a /new/ version of gnu stuff. But in the
linking phase, gcc complains about the missing libg.a.

ANSWER: Yes this is well known, throw away the flag -g that's all,
anyway libg.a is /only/ for debugging purpose.


QUESTION: I can't move directories under Linux, why ?, what can I do?

ANSWER: mvdir needs kernel resources which don't exist yet. The
reason the rename system call isn't implemented yet is that there are
a couple of problems with it: it isn't as straightforward as you'd
imagine. Race conditions, inclusion checking etc..

It will befinitely be in 0.13/0.95. It's not /that/ difficult, just
needs a little thought. In the meantime, you might fake it with "cp
+recursive" and "rm -rf". Be careful with that though..
You can also write a bash script, which simulates the mvdir (really a
good exercice ๐Ÿ™‚ )

This also appears to be the reason why mv's of sym links do not work
properly.

BTW Linus has already done it for linux v0.12 it's a patch available
in the digest10 vol #1 of alt.os.linux.

QUESTION: What are the device minor/major numbers?

ANSWER: (early Linus mail Nov. 6th 91, last update Jan. 19th 92)
Memory devices: Major = 1 (characted devices) minor
0 /dev/ram
1 /dev/mem
2 /dev/kmem - not implemented (easy, but I haven't done it)
3 /dev/null
4 /dev/port (implemented, but untested - don't play with it)

example: "mknod /dev/null c 1 3"


Floppy disks: Major = 2 (block devices)

minor = drive + 4*type, drive = 0,1,2,3 for A,B,C or D-diskette

type 1: 360kB floppy in 360kB drive (5.25")
2: 1.2M floppy in 1.2M drive (5.25")
3: 360kB floppy in 720kB/1.44Mb drive (3.5")
4: 720kB floppy in 720kB/1.44Mb drive (3.5")
5: 360kB floppy in 1.2M drive (5.25")
6: 720kB floppy in 1.2M drive (5.25")
7: 1.44M floppy in 1.44M drive (3.5")

Thus minor nr for a 1.44Mb floppy in B is: 1 + 4*7 = 29, and to read
an old 360kB floppy in a 1.2M A-drive you need to use minor= 0 + 4*5
= 20.

Example: "mknod /dev/PS0 b 2 28" (b for block: 2 for floppy, 28 for
1.44 in A)


Hard disks: Major = 3 (block devices) minor
0 /dev/hd0 - The whole hd0, including partition table sectors
etc.
1 /dev/hd1 - first partition on hd0
...
4 /dev/hd4 - fourth partition on hd0
5 /dev/hd5 - The whole hd1, again including partition table info
6 /dev/hd6 - first partition on hd1
...
9 /dev/hd9 - fourth partition on hd1

NOTE! Be /very/ careful with /dev/hd0 and /dev/hd5 - you seldom need
them, and if you write to them you can destroy the partition tables:
something you probably don't want. The only things that use /dev/hd0
are things like "fdisk" etc.

NOTE 2!! The names for hd's are the same as under minix, but I think
minix orders the partitions in some way (so that the partition numbers
will be in the same order as the partitions are physically on the
disk). Linux doesn't order anything: it has the partitions in the
same order as in the partition table (ie /dev/hd1 might be physically
after /dev/hd2).

Tty's: Major = 4 (character devices) minor
0 /dev/tty0 - general console 1 -
63 - reserved for virtual console
64-127 - reserved for serial io
128- - reserved for pty's

And more particularly we have:
64 /dev/tty64 - com1
65 /dev/tty65 - com2


QUESTION: How to start Linux from drive B?

ANSWER: There is a DOS utility called boot_b.exe (look at DOS ftp).
Another simple way is to open the box and invert the cables.

QUESTION: The program boot_b works fine /but/ once the first disk is
read the system go back to the first drive, any hints?

ANSWER: Yes, change the bootimage in just the same way that you change
it to boot on the hard drive, execept that the major/minor pair is
different. All these information are in the file INSTALL-0.10.
Remember that if you use a sun or other endian machine, you will need
to reverse the byte order when you run the filter program (also in the
same file).


VII. FEATURES

=============

QUESTION: I've read that linux has virtual consoles, what must I do to
get them?

ANSWER: Yes there are, you can access them with the left -key
together with -key. But first you have to activate them with doshell
/dev/ttyn /bin/sh & where n is the number of the Virtual console.
With the Linux 0.12 Images distribution, two consoles are available,
but if you want to use the second do first:
"doshell /dev/tty2 /bin/sh&"
Moreover if you want to activate automatically the Virtual console,
you should add this command in the /etc/rc file.

BTW: the serial ports are now /dev/tty64 and /dev/tty65. tty0 is the
general console. tty128- are reserved to pty's


QUESTION: What kind of shell is /bin/sh ?

ANSWER: It's the Bourne Again Shell, bash-1.11 and compilation was
straightforward (Linus dixit), just "make" that's all or nearly.


QUESTION: Does there exist a man page for **** ?

ANSWER: Download man.tar.Z from your favorite linux ftp site, there is
most of the fileutils man page -- either **** or g****, example there
is nothing on ld, but there is for gld ๐Ÿ™‚ --, check the whatis
database provided. The files in the cat1 dir are pre-formatted man
pages that the man program can use.

BTW there is no roff,troff nor nroff for Linux. Cawf 2.0 works just
fine for simple man pages, and a partial ms support too.


QUESTION: What are the editors available in linux?

ANSWER: Right now there are uemacs, elvis-1.4, some one (R. Blum)
is working on some other vi clone. The port of emacs 18.57 has been
done by John T Kohl, files can be found at the different sites
at nic it's in the directory xtra
at tsx-11 it's in the directory ports/emacs-18.57.
Also the port of mg (micro gnu) has been done and can be found at
least at athos.rutgers.edu (128.6.4.4) in pub/linux, mg is the binary
and mg.tar.Z is the sources file.

QUESTION: Does there exist a printer package for Linux?

ANSWER: There are lp patches for linux.0.12, which implement a
parallel printer interface and feature a greatly improved driver
design. the patches are in lp.12.tar.Z As I have no printer (yet), I
don't know how good it is.


QUESTION: Does there exist a ps for Linux?

ANSWER: Yes in the vc2.tar.Z file, however the command is very kludgy.

QUESTION: It's nice to have the df utility, but it would be nicer if
it would give statistics of the root file system. Would it be
difficult to do?

ANSWER: surely not, in your file /etc/rc, instead of the line
> etc/mtab
put the following
echo "/dev/hdX (root)" > /etc/mtab
where the X is the hard drive you use as root partition.


QUESTION: How do I make swapping work?, when I boot I get the
following message: "Unable to get size of swap device"

ANSWER: There are two ways (as far as I know). Recompile the kernel
and makes the swap device available (have a look in the main
Makefile). Or else you can modify the bootimage, in that case you have
to put a full device number in the word reserved for the swap device:
the logic is the same as for the root device. Thus, if you want
/dev/hd2 to be your swap-device, you'll have to write a 2 (minor nr)
into byte 506, and a 3 (= harddisk) into byte 507. You can use
virtually the same program as for the root-device (look in INSTALL
files).


QUESTION: When I boot I get one of the following messages:
"Unable to find swap signature" or "Bad swap-space bitmap"

ANSWER: You probably forgot to make your swap-device, use the mkswap
command.


QUESTION: How do I know if it is swapping?

ANSWER: You will notice it :)) First of all, Linux tells you at boot
time, "Adding swap: XXX pages of swap space", and if you start running
out of memory, you will notice that the disk will work overtime, and
things slow down. Generally a 2Meg RAM will make the system swap
constantly while running gcc, 4 Meg will swap occasionnaly when
optimizing big files (and having other things active, such as make).


QUESTION: Is there only the %$#@ keyboard ?

ANSWER: There are Finnish, French, German,and US keyboards. Set it in
linux/kernel/chr_drv/keyboard.S, then compile the kernel again.

BTW a Dannish keyboard patch (Tommy Thorn) is now available look the
digest19 vol#1 file. There is an Uk keyboard too (Jan 25. by Leblanc),
I have not seen it in the newsgroup so ask directly to
[email protected]


QUESTION: Does there exist shared libs ?

ANSWER: They seem to work. The kernel features are in Linux 0.12
already.


QUESTION: Does Linux permit/support bitmapped graphics on vga/svga
cards?

ANSWER: No, there is no interface for graphics operations on Linux
(yet).

QUESTION: There are a lot of patches available (fd patch, lp patch
login patch ...) can I be fairly confident the subsequent patches will
work?

ANSWER: No you can't, patching is a real beta tester art :)). People
are not working on the same patched release, so you have to check if
the patches you already applied works on the same kernel part, if not,
/great/, just apply them. If yes, check if there is an order, patch
creator knows that, and (should) try to warn patch user (in other
words: beta tester) otherwise you should edit the patch files (and
possibly make a brief note to others on this list/newsgroup or even
a cdiff) before applying them, another solution is to keep cool and
wait for the next version of Linux where, in general, the
modifications have been done but this behavior is /not/ Linux
helpful.


QUESTION: I got the patches on some ftp sites, and applied them to the
kernel and tried to compile. It didn't !!. Are the patches buggy?

ANSWER: Before remake, just do a make clean in the directories
involved by the patches. This will force a rebuild of the .o and .a
files.
If you have a RCS running on your source tree, did you checked a
patched version of the files changed before /any/ CO either by you or
make

VIII. MORE HINTS
================
This part is quite recent and try to keep track of the different
information that appeared in alt.os.linux and on the list since
beginning of February. I have not checked all of these information.
But the one I know correct are marked by a (+). Fill free to contact
me and infirm/confirm what is written below. Moreover take care to use
the correct library and include stuff for linux-0.12 and the ad-hoc
gcc you use !!!


QUESTION: How can I backup my Hd under Linux ?

ANSWER: I know at least two ways. One possibility is tar and mtools,
another possibility is the diskbackup/diskrestore of Diamano Bolla
(digest44 vol. #1) which saves big hd to floppies using the
stdin/stdout.


QUESTION: How to use setterm: for the novice?

(+)ANSWER:The setterm utility provides access to most of Virtual Consoles
(VCs) functionality. You can set your screen up to blank at 10
minutes using:
setterm -blank 10

You can set colors, and clear the screen. For a full list of commands,
just type "setterm" with no arguments.

There are a few tricks with the screen dumper can really make VCs go a
long way. Here are a few of the common ones that I use:

setterm dump

Dumps the contents of the current VC to screen.dump (in the current dir).

setterm dump 4

Dumps the contents of VC 4 to screen.dump

setterm -file mydumpfile -dump 4

Dump the contents of VC 4 to the file mydumpfile

setterm -file /dev/tty0 -dump 4

Dumps the contents of VC 4 to the current VC.

setterm -file /dev/tty4 -dump

Dumps the contents of the current VC to VC 4.

setterm -file /dev/tty64 -dump

Dumps the contents of the current VC to the serial port.
Handy if you are logged on and want to paste a screen full without
having to resort to doing a file transfer.

setterm -file mydumpfile -append 4

Appends to instead of overwriting the dump file. Useful if you
have several screens you wish to concatenate.


QUESTION: I know there are VC, but where is the setterm stuff?

(+)ANSWER: It is in virtcons.tar.Z (at least on nic). It provides a
number of features that are not available with the standard release of
linux-0.12, among them blanking/dumping screen.


QUESTION: While using emacs in 80x25 mode, the mode line is constantly
moving around, why?

ANSWER: This appear to be a bug in the scroll region handling of the
console. Some one has suggested to simply remove the last entry in the
vt100 termcap that refers to the scroll region escape. Namely remove
the cs entry.


QUESTION: I have read that there are patches for an init/login in
Linux-0.12, where can I find them.

(+)ANSWER: There are many, one is the qpl-init.tar.Z package available
at the 3 main sites, but it is not that good (it's a quick hack).
There is also the Peter Orbaek init/login stuff that can be down-load
from ftp.daimi.aau.dk (130.225.16.1); the last version is in the file
poeigl-1.1.tar.Z and it is in pub/Linux-source. Among other things it
contains a patch to main.c of Linux kernel v0.12 (directory init). You
have to recompile the kernel (after patching it) to use /etc/init
rather than /etc/rc at boot time.


QUESTION: How can I change user password in login?

(+)ANSWER: There is no passwd program yet, you should do it by hand
with the encrypt stuff provided.


QUESTION: When I use make as non root, it doesn't work, why?

(+)ANSWER: ?????, the message is either (null) setuid ..., or (null)
setgid... I don't know how to fix it.
BTW This problem does not exist with the pmake (make for BSD 4.3 Reno
and BSD 4.4) package.


QUESTION: How can I get Linux to boot directly from the harddisk?

ANSWER: Right now, this can be done via the shoelace stuff or the
bootany package.


QUESTION: Sometimes, when I want to remove a directory, I get an error
message, is it a (known) bug?

(+)ANSWER: No, There is no bug at all, you probaly have another shell
on another VC whose working directory is either the one you try to
remove, either a subdirectory of it.


QUESTION: can anyone give me a sample /etc/inittab entry for login
from a pc attached to serial line /dev/tty65?

ANSWER: "Humberto speaking :)"
I've got a Hayes Smartmodem 1200 hooked up to /dev/tty65, and log
in daily from work. My inittab entry is

tty65:console:/bin/agetty -m -t 30 1200 tty65

but it works without the -t OK too. The real problem I suspect you're
having is initializing the modem properly. Before setting up logins, you
have to disable echo and enable auto-answer on the modem. I do this here
by firing up kermit at 1200, connecting, then sending 'ate0s0=5'
followed by enter to the modem.


QUESTION: When compiling some code, cc1 complains about some insn
code, what's that?

(+)ANSWER: An insn is an internal representation that gcc uses when
compiling. The main part of gcc is to take ordinary c (or c++) code,
and compile it, while ding optimizations in insn part, which is
soft/hard independant. Then another part which is hard/Os dependant
takes the insns and translate it in assembly language. The fix is only
to turn off the optimization flag (-O) or download the new gcc release
(1.4) which has been made available at tsx-11 (newgcc.tar.Z and the
ad-hoc libraries newlibc.tar.Z).

BTW I've checked tsx-11 on Sat. 22nd Feb. and nothing is available,
but you can find one in /incoming at ftp.eecs.umich.edu.

QUESTION: While compiling some stuff, I'm getting the following
error message:
Undefined symbol ___addsf3 referenced from text segment
as well as ___mulsf3 and __cmpsf2.
These symbols are not in the program or in it's header files.

ANSWER: These are math helper functions, and you can usually compile
these programs to use the kernel floating point routines by adding
'-m80387' to the compiler switches. If the program does any wierd
fp math (exp(), sin()) it'll die when you run it though.


QUESTION: What are the enhancement of the newgcc.tar.Z ?

(+)ANSWER: There were some bugs in the old port that have been corrected,
moreover this package contains 387 support
there is libm.a (for those with 387)
libsoft.a (for those without, I for example)
libtermcap.a (from tput 1.10)

The -mstring-insns option is no longer needed nor supported ๐Ÿ™ [As
an example to recompile (successfully) linux you have to throw away
this flag in all the Makefile]

BTW Notice also that include files have changed (stdio.h which is no
more ansi compliant). Right now (Feb. 24th), there are some little
bugs which will be fixed in a short while.
The port of gcc 2.0 has begun ...


QUESTION: I've been trying to get Linux to run on my [3/4]86 box. It
can't even boot. Any suggestions?

(+)ANSWER: The most common error/problem is writing the bootimage to a
low density disk. It fits, but the bootstrap code will only recognize
high density disk. So try to format explicitely disk as high density:
- for 3.5", 'format a: /n:18 /t:80 '
- for 5.25", 'format a: /n:15 /t:80 '

QUESTION: Does there exist games, languages (other than C), and
anything which make the system more friendly?

ANSWER: Yes, among other things there are rogue and yatzee; TeX;
Prolog.. but in general, if you want some extra tool port it to Linux
this is also a good beta-testing exercice.

===================8<==========>8================


[next article]
From: [email protected]

Subject: keyboard.S patch for control-space
Message-ID: <[email protected]>
Date: 25 Feb 92 07:11:51 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 40

Well, I'm hooked! My time spent in DOS is decreasing rapidly. The port of
GNU Emacs was the final straw. I can't live without my Emacs!

But... to make using Emacs easier for those of us used to DEC keyboards,
here is a patch to keyboard.S which maps control-space to [email protected] This is
far easier to type and almost second nature to many Emacs users with DEC
keyboards.

Line numbers may vary depending on which other patches you & I have applied.
Patch is to version 0.12, of course.

*** keyboard.S~ Tue Feb 25 02:02:05 1992
--- keyboard.S Mon Feb 24 23:18:13 1992
***************
*** 525,531 ****
subb $32,%al
2: testb $0x0c,_kmode /* ctrl */
je 3f
! cmpb $64,%al
jb 3f
cmpb $64+32,%al
jae 3f
--- 525,535 ----
subb $32,%al
2: testb $0x0c,_kmode /* ctrl */
je 3f
! cmpb $32,%al /* ctrl-space -> [email protected] */
! jne 5f
! movw $0,%al
! jmp 4f
! 5: cmpb $64,%al
jb 3f
cmpb $64+32,%al
jae 3f

----
Dave Gentzel
[email protected]
Proud member of the "League for Programming Freedom"
Keep Your Lawyers Off My Computer!


[next article]
From: [email protected]uddy.uucp (sweh)

Subject: Re: A couple of problems...
Keywords: Gcc Installation Linux
Message-ID: <[email protected]>
Date: 24 Feb 92 20:08:45 GMT
References: <[email protected]> <[email protected] S.ORST.EDU>
Organization: Spuddy's Public Usenet Domain
Lines: 20

In article [email protected] edu (Charles Hedrick) writes:
:>2- How do people backup their systems? Trying to create
:> large tar file either on floppies or to a disk file
:>almost inevitably locks up my system. A known
:>problem? Maybe fixed in 0.13? ๐Ÿ™‚

I get this when tar'ing to floppy, but not to a large disk file.

:Linus has been hoping that these hangs (which occur to several people)
:are due to running out of memory, or possibly to coming close to
:running out (a situation which causes special code in the kernel to
:run to free up memory). I saw a great


[next article]
From: [email protected] (Robert Blum)

Subject: Re: floating point
Keywords: coprocessor
Message-ID:
Date: 25 Feb 92 08:17:40 GMT
References: <[email protected]> <[email protected]>
Sender: [email protected] (Newsfiles Owner)
Organization: RBI - RWTH Aachen
Lines: 19
Nntp-Posting-Host: kaa


Hi Linuxers!
[email protected] (I.J. Wells) writes:

>One thing I would like, if it's easy(ish) to do, is to be able to remove
>the FP from a kernel. My machine has an FPU, so I'd like to be able to
>reclaim the memory FP emulation takes
I am working on a loadable FPU emulator, but even then, you can't reclaim
the memory for further use. Everything below 512 K is kernel space,
no matter if it contains code or not. This will probably become useful,
if we start to have dynamic memory management in the kernel

C U l8r,
Robert Blum


--
UNIX should not be able to be crashed from user space.
Crashes belong in the kernel! ([email protected])


[next article]
From: [email protected] (Michael K. Johnson)

Subject: lists...
Message-ID: <[email protected]>
Date: 25 Feb 92 01:38:36 GMT
Sender: [email protected]
Organization: St. Olaf College; Northfield, MN USA
Lines: 32



[email protected] (Craig Metz) writes:
(in linux-activists)

RE: patches

Well, we could do this. mjohnson, could you handle both the bug list
and patch list and get them going, and when summer nears, I'll handle them fo r
you? When fall comes, who knows. But it wouldn't be a bad idea to keep the
two lists together, since they have many things in common.

-Craig

Well, I havn't been keeping up _at_all_ on the patches, but I could
keep a list of what they mean. I am not an ftp admin, so if I am
being asked to keep a patches directory on an ftp site, some ftp admin
would have to let me on their site with extra privs, which many
(reasonably) wouldn't like. However, I am _NOT_ going through 2 Megs
of stuff _again_ to pull out patches, so if you want your patches in
my list, _mail_me_. I take no responsibility for remembering other
people's patches or finding them.

As far as the buglist goes, I am ready to post the 40K file.
-----REPEAT-----
40K file. If anyone has strenuous objections to such a file being
posted, please mail me soon. I intend to send it both to
[email protected] and alt.os.linux.

michaelkjohnson
[email protected]
I don't do .sig's.


[next article]
From: [email protected] (Greg Lee)

Subject: Re: Hackers' technotes to Nondestructive Repartitioning
Keywords: partitioning, nondestructive
Message-ID: <[email protected]>
Date: 25 Feb 92 12:13:38 GMT
References: <[email protected]>
Sender: [email protected] (News Service)
Organization: University of Hawaii
Lines: 18
Nntp-Posting-Host: uhunix.uhcc.hawaii.edu

In article <[email protected]> [email protected] (Drew Eckha rdt) writes:
}Reformatting your hard disk, and reinstalling all your DOS programs from
}scratch can be a real nuissance, and is unecessary.
}...
}Hopefully this has been enlightening, and possibly even useful to some.

It was very helpful -- thanks a lot. I just went through the process of
making a new partition using Norton Utilities (5.0), and it turned out
to be pretty easy. The procedure I followed was to:
(1) delete "image.idx", the hidden file left by the Norton Image program
(2) ask Speed Disk to do a full optimization to collect files at the
beginning of the disk
(3) use the partition editor of Disk Editor to add the new partition to
the partition table (I used type XENIX)
(4) reboot and use Disk Doctor to check all was ok

--
Greg Lee


[next article]
From: [email protected] (Joel M. Hoffman)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Keywords: v86, video
Message-ID: <[email protected]>
Date: 25 Feb 92 13:58:38 GMT
References: <[email protected]>
Sender: [email protected] (USENET News system)
Organization: University of Maryland, College Park
Lines: 36
Nntp-Posting-Host: next.wam.umd.edu

In article <[email protected]> [email protected] (Drew Eckhardt) writes:
>We want X11 up on Linux. Since not everybody has an ATI or Tseng chipset, we
>also want to support the other cards out there - Trident 8800 / 8900,
>Genoa, etc.
>
>Now, the only real differences between these chips is in setup (mode
>selection), and bank switching (to accomodate the 128K we have for video
>buffer and 1M of modern SVGA cards). Setup is the difficult part, with over
>20 registers that need to be set for each mode,
>but this is easily accomplished by BIOS.
>
>The problem is that we don't have BIOS in protected mode. Or do we?
>The i386 gives us a nifty little feature for running real mode programs
>as virtual 80x86 tasks. Unfortunately, these tasks must reside
>at absolute address 0. Our current mm code requires that all processes
>share the same page tables - meaning only one can live at
>absolute address 0. However, if all we want is one (the BIOS routines ) this
>is not a problem.
>
>Specifically what needs to be done : [...]

I think there are *many* good reasons to rewrite mm. First of all, we
(at least I) certainly want a DOS emulator down the road. The easiest
way to do this is to run the BIOS, and then DOS, in V86 mode, I think.
Also, I think a more general page table scheme would be nice on its
own merits. (I can see having more than 64 tasks, e.g.)

One place to start might be the code from DJGPP. DJGPP is a port of
GCC to DOS (!). It contains a lot of code for switching back and
forth between 32- and 16- bit modes.

The other option, of course, is to copy the BIOS into a process's
memory (in protected mode), and try to make use of it. I don't know
how difficult this would be.

-Joel


[next article]
From: [email protected] (Humberto Ortiz-Zuazaga)

Subject: Re: Problems booting off login kernel
Summary: Magic numbers?
Keywords: Register dump, floppies, magic numbers.
Message-ID: <[email protected]>
Date: 25 Feb 92 15:17:59 GMT
References: <[email protected]>
Organization: Univ. of Cincinnati
Lines: 20

In article <[email protected]> [email protected] (Drew Eckha rdt) writes:
>In article [email protected] (Sean C. Cox) writes:
>> Things started up ok, I got the "Loading......" prompt, but before it
>>finished the "usual" amount of dots, I got a whole bunch of
>>
>> @X: 0400
>> AX: 020D
>> E600
>> CX: 0401
>> DX: 0000.
>
>Bad floppy disk. Sector 1, track 4, head 0 not found.

Drew, I've seen you answer this kind of question several times before.
Do you have a list of the codes produced in this kind of error? Would
you like to post them? Maybe we can put them in the next FAQ.
--
Humberto Ortiz-Zuazaga INTERNET: [email protected]
Dept. of Physiology & Biophysics BITNET: [email protected]
University of Cincinnati CIS: 72301,2303


[next article]
From: [email protected] (Adam Goldberg)

Subject: Re: fdisk / mkfs problems - ST251R & ST22
Message-ID:
Date: 25 Feb 92 15:59:16 GMT
Sender: [email protected]iastate.edu (USENET News System)
Organization: Iowa State University, Ames IA
Lines: 27
Originator: [email protected]

A friend of mine, who purchased a 386 motherboard for the express
purpose of running Linux (hooray) is having some problems getting
things up & going.

The boot process goes just fine, but fdisk reports NOTHING. It
dies without any sort of output whatsoever.

Furthermore, skipping the mkfs portion (and just guessing the #
of blocks) reports bad blocks before the first data block (or
something to that effect).

He has an 386/25 with Award BIOS, a ST251R (42 meg) RLL drive, with
a ST22 controller card. (and 4M memory...)

Does ANYONE have ANY idea what's wrong, or what to do, or what
to try, or ANYTHING AT ALL. HELP!

(Reply to either me ([email protected]) or him ([email protected]))

Thanks.

--
Adam Goldberg ! Ask me how IOWA STATE UNIVERSITY is
[email protected] ! censoring my usenet access!!
..!uunet!iastate.edu!goldberg ! >>I don't need a BIG BROTHER!!<<

--


[next article]
From: hlu%[email protected]

Subject: Re: Gcc 2.0, new libc.a and alpha test
Message-ID: <[email protected]>
Date: 25 Feb 92 17:41:37 GMT
Sender: [email protected] (Mr Background)
Reply-To: hlu%[email protected]
Organization: The Internet
Lines: 18

|
| Note that Gcc 2.0 as released compiles objective C into code that
| calls functions in an objective C runtime library. That library has
| not been written. (NeXT contributed the compiler front end for
| objective C, but not the library.)
|
| Carl Witty
| [email protected]
|

Oops. I screwed up this one. Ok. No objective C for Linux.

H.J.
--
School of EECS Internet: [email protected]
Washington State University BITNET: [email protected]
Pullman, WA 99164 Phone: (509) 335-6470 (O)
USA (509) 334-6315 (H)


[next article]
From: [email protected] (Al Clark)

Subject: Can't run off of hard disk.
Keywords: hard disk, boot
Message-ID:
Date: 25 Feb 92 20:29:45 GMT
Organization: Netcom - Online Communication Services (408 241-9760 guest)
Lines: 44

I am using 0.12 boot and root images.

Status: Generated 1.2 Mb boot and root disks.
Booted successfully.
Back to DOS, used 'edpart' to make linux partitions
Partition Sectors Use
1 1 - 251 (Already dos)
2 251 - 410 hd2 - root 20400 blocks
3 411 - 570 hd3 - swap 20400 blocks
4 571 - 1000 hd4 - user 54697 blocks

Back to linux; got sizes above from fdisk;
mkfs -c /dev/hd2 20400 and mkfs -c /dev/hd4 54697
mount /dev/hd2 /user
copied all of bin dev etc usr tmp to /user
cd /user/usr/bin - uncompress tar.Z
Loaded some binaries from 1.4 MB floppy B.
tar -xvf /dev/PS1 elvis.tar.Z
tar -Zxvf elvis.tar.Z
Back to DOS.
Used Linus' edit program to modify boot.12 -> boothd2.12
comp gives:

C:ROOT.12 and C:ROOTHD2.12

Compare error at OFFSET 1FC (508)
File 1 = 0
File 2 = 2
Compare error at OFFSET 1FD (509)
File 1 = 0
File 2 = 3
EOF mark not found

Rawrite boothd2.12 to a:
Boot linux; still uses floppy. Can mount /dev/hd2 as /user,
still get data on hd2. After unmounting, tried
mount /dev/hd2 /
got error 16?

Now what??
Am I missing something obvious???
--
Al - [email protected] - My opinions are my own.
*** Commit acts of random kindness and senseless beauty! ***


[next article]
From: [email protected] (Al Clark)

Subject: swapon binary
Keywords: swapon binary
Message-ID: <[email protected]>
Date: 25 Feb 92 21:40:59 GMT
Organization: Netcom - Online Communication Services (408 241-9760 guest)
Lines: 7

In checking tsx-11.mit.edu, all I can find is swapon.tar.Z; I find it
contains only the source. Since I have only 4Mb RAM, I need swap to
compile; so ... I need a binary of swapon. Any pointers???

--
Al - [email protected] - My opinions are my own.
*** Commit acts of random kindness and senseless beauty! ***


[next article]
From: [email protected] (Andrew Stevens)

Subject: Two Controllers
Message-ID:
Date: 25 Feb 92 15:24:01 GMT
Sender: [email protected]
Distribution: alt
Organization: /users/as/.organization
Lines: 22


Dear all,

I am planning on installing Linux on a new 486 box equipped with
only a 40M hard-drive. Given the lack of space, I would be keen
to press into serivce a 60M MFM drive that I posses.

The tricky questions are thus: (1) What flavour of MFM controller
can LINUX drive (2) are there likely to be any complications
in driving both an MFM controller and an IDE drive from LINUX /
DOS? I have visions of much kernel tweakery and endless
fiddling with interrupt jumpers etc

I trust this question isn't hopelessly naive...


Andrew
--
Andrew Stevens
Programmming Research Group JANET: [email protected]
11 Keble Road, Oxford, England UUCP: ...!uunet!mcvax!ukc!ox-prg!as
OX1 3QD +44 0865 272563


[next article]
From: [email protected] (Ben Lippolt)

Subject: pfdisk with 2 HDs
Message-ID:
Date: 25 Feb 92 08:34:13 GMT
Sender: [email protected] (USEnet News)
Reply-To: [email protected]
Organization: PTT Research
Lines: 22
Nntp-Posting-Host: freyr

Hi,

I just got pfdisk and installed it (scary to mess around with boot-blocks!).
Now I got a question: how do I use it with 2 hard disks?
My configuration is as follows:
partition 1 2 3 4
disk 1 (106 Mb) unused unused DOS(active) DOS(extension)
disk 2 (106 Mb) Linux(boot) Linux(swap) Linux Linux

I installed "bootactv.bin" on /dev/hd0. When I boot it finds the
active partition on disk 1 and if I don't press a key it boots DOS.
When I press a key I can choose 1-4, but only on the first disk.
How do I get it to boot partition 1 on disk 2? I assumed I had to install
a boot-block and some other programs on my Linux root-partition, so I
looked at shoelace and installed the programs as I should. I installed
the Linux bootimage-0.12 as /vmlinux and edited /etc/config accordingly.
I ran laceup and finally I ran pfdisk on /dev/hd5 again, but it still
only boots from the first disk.

So, how do I do it?

Ben.


[next article]
From: [email protected] (Ben Lippolt)

Subject: Re: bug fix for the new gcc 1.40
Message-ID:
Date: 25 Feb 92 09:30:53 GMT
References: <[email protected]>
Sender: [email protected] (USEnet News)
Reply-To: [email protected]
Organization: PTT Research
Lines: 10
Nntp-Posting-Host: freyr

[email protected] (Hongjiu Lu -- Graduate Student) writes:
>I just fixed a FP bug in gnulib, a bug in sqrt(0.0) of libsoft.a and put
>a new newgcc.tar.Z on tsx-11.mit.edu.

I've been looking on tsx-11 about once a day since the first announcement
of gcc-1.40, but I'm unable to find it. "/incoming" is unreadable and I
can't find it anywhere else. What's the status of it?


Ben.


[next article]
From: [email protected] (Marc CORSINI)

Subject: Booting Linux
Message-ID: <[email protected]>
Date: 24 Feb 92 19:02:26 GMT
References:
Sender: [email protected]
Organization: Greco Programmation du CNRS
Lines: 25

Hi there,
Quite often I can read messages like the following in the newsgroup:

" I boot from the bootimage disk and get the following message:

CX:xxxx
DX:yyyy
@X:zzzz
AX:wwww
mmmm
"

Could some of the debugger guru send ME a brief note about the analysis
I think this could be included in the FAQ

Marc


--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
# LaBRI | #
# 351 cours de la Liberation | e-mail: [email protected] #
# 33405 Talence Cedex | #
# | #
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~


[next article]
From: [email protected] (Paul Richards)

Subject: Standards docs.
Message-ID: <[email protected]>
Date: 25 Feb 92 15:46:22 GMT
Sender: [email protected] (Network News System)
Organization: University of Wales College at Cardiff
Lines: 20
X-Mailer: Cardiff Computing Maths PP Mail Open News Gateway

Where can I get hold of standards documents e.g. ANSI C, POSIX and
anything else we're using for linux. In particular, are there any
on-line sources.


Has anyone started work on ethernet drivers. I was thinking of looking
at the clarkson packet drivers. If anyone could give me any pointers,
like which device driver should I look at to give me some idea of
structure etc, I might start to get somewhere.

I've got an isolan 4110, so I'll be working on that. If someone who
knows about the clarkson drivers and writing linux device drivers wants
to work on this with me then let me know. In particular, someone with a
different ethernet card would be handy.
--
Paul Richards at Cardiff university, UK.

[email protected] Internet: spedpr%[email protected]
UUCP: [email protected] or ...!uunet!mcsun!ukc!cf!thor!spedpr
+++


[next article]
From: [email protected] (cs87ejh)

Subject: Re: A couple of problems...
Message-ID: <[email protected]>
Date: 25 Feb 92 13:58:21 GMT
References: <[email protected]>
Organization: Brunel University, Uxbridge, UK
Lines: 36

[email protected] (sweh) writes:
: In article [email protected] s.edu (Charles Hedrick) writes:
: :>2- How do people backup their systems? Trying to create
: :> large tar file either on floppies or to a disk file
: :>almost inevitably locks up my system. A known
: :>problem? Maybe fixed in 0.13? ๐Ÿ™‚
:
: I get this when tar'ing to floppy, but not to a large disk file.
:
: :Linus has been hoping that these hangs (which occur to several people)
: :are due to running out of memory, or possibly to coming close to
: :running out (a situation which causes special code in the kernel to
: :run to free up memory). I saw a great

I found a good solution to the backup and restore problem. There is a
disksplit.tar file on tsx11, that helps. I have used it and have had no hangs
whatsoever.

An example usage is:

tar cvf - bin dev usr etc .. | compress | diskbackup

To restore your files:

diskrestore | uncompress | tar xvf -

It never hung the machine (no swap 3mb extended), and is very efficient
(because of compression). Of course if a disk fails ...

Roger
--
+=============================================================================+
| [email protected] Roger Binns Brunel University - UK |
|:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::|
| Humans were created by water to move it uphill |
+=============================================================================+


[next article]
From: [email protected] (Brian Syme)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Keywords: v86, video
Message-ID: <[email protected]>
Date: 25 Feb 92 17:39:04 GMT
References: <[email protected]> <[email protected]> <1992F [email protected]> <[email protected]>
Organization: Glasgow University Computing Service
Lines: 26

In <[email protected]> [email protected] (D.Bolla) writes:
>[thread about video mode changes.]

I don't think the actual mode changes are much of the problem with regards to
SVGA cards, basically all you do is pick up a table of values and stuff them
into the registers. This is straightforward from the Linux kernel (about 10
lines of C - in fact have a look at the X386 code, that's where I got it from.)
It certainly isn't worth the horrors of setting the machine up to do a BIOS
call.. though, of course if we want DOS emulation.. actually I think I'd be
happiest if I could just mount a DOS filesystem under linux..

The real problem with SVGA's is the bank switching and pallette control while
in SVGA mode and actually writing to the thing. Unless the card supports a
linear memory map mode, you are forced to do this, and you need to perform
port I/O to do it (ioctl's are too slow - you have to do lots of them at times)
currently Linix seems to blackhole port I/O from user space ๐Ÿ™

Anyway - apart from the graphic support, what's missing from 0.13 that we still
need to get X up and running? (Not that I like X, I'm just lumping it..)

Brian
--
------------------------------------------------------------------------------
Brian Syme, Glasgow University Library. | Janet: [email protected]
| Proper: [email protected]
------------------------------------------------------------------------------


[next article]
From: [email protected] (cs87ejh)

Subject: interim
Message-ID: <[email protected]>
Date: 25 Feb 92 18:18:54 GMT
Organization: Brunel University, Uxbridge, UK
Lines: 20

Is anybody prepared to make an interim distibution of Linux (ie something
between .12 and .13|.95.

I am getting snowed under different libraries, floating point, compilers
and patches. It would be nice to have a complete set of patched kernel,
libraries etc at a decent version.

I am also trying to port tcsh and am having considerable problems with
signals, and control is never returned to the tty when an external command
is executed in the foreground. I think both of these would be fixed if I
have applied the correct patches.

Roger

--
+=============================================================================+
| [email protected] Roger Binns Brunel University - UK |
|:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::|
| Humans were created by water to move it uphill |
+=============================================================================+


[next article]
From: [email protected] (Philip Copeland)

Subject: catman
Message-ID: <[email protected]>
Date: 25 Feb 92 21:59:26 GMT
Sender: [email protected] (Usenet poster for nntp (tj - 17/12/91))
Reply-To: [email protected]
Organization: Bristol-Polytechnic
Lines: 16
Nntp-Posting-Host: eggs


ok so i'm subscribed to alt.pulp... anyway, i have been looking around
and cannot find the sources for catman to generate manual page entrys
with.

I have a good idea that it would want nroff and troff but i'm prepared to
smeg the code and get it working as i'm not fond of the present method
of editing the whereis file.

alternively, has someone else done this? if not i volunteer (subconciously
not a good idea) to crack the code on this (what have i said ?!??).

comments?

Phil
=--=


[next article]
From: [email protected] (`Grave' Dave Gymer)

Subject: Backing up Linux partitions (was Re: A couple of problems...)
Keywords: Gcc Installation Linux
Message-ID: <[email protected]>
Date: 25 Feb 92 09:32:46 GMT
References: <1992Feb23.033155.248[email protected]> <[email protected] S.ORST.EDU>
Organization: Nottingham University
Lines: 15

In article <[email protected]> [email protected] (Brian Mcbee) writes:
>2- How do people backup their systems? Trying to create
>a large tar file either on floppies or to a disk file
>almost inevitably locks up my system. A known
>problem? Maybe fixed in 0.13? ๐Ÿ™‚

I'm still working on getting a Linux system up and running, but you could try
afio which was posted to comp.sources.*something* (unix? misc?) a while back;
I've used it on my Atari ST to back up Minix paritions to floppies. Once I have
a working Linux I'll port it myself (if nobody else has done it).
--
`Grave' Dave Gymer | ## ## ##### ##### ##### | [email protected]
42 St Mary's Park, | | \_/ | | ____ ____| ___/ | These opinions are
Louth, Lincs, | | | | | / | mine, so get your
LN11 0EF, ENGLAND | # # # ##### ##### | hands off 'em!


[next article]
From: [email protected] (Benjamin Cline)

Subject: Re: More questions about Linux ๐Ÿ™‚
Message-ID: <[email protected]>
Date: 26 Feb 92 00:03:07 GMT
References: <[email protected]> <[email protected]>
Sender: [email protected] (Usenet News Administration)
Organization: MathSoft, Inc.
Lines: 28
Nntp-Posting-Host: quine

In article <[email protected]> [email protected] (Tommy Thorn ) writes:
>
>Would everybody please stop calling Linux *a small system*. It's small
>yes, but already very complete, and I hope see several more things yet, like:
>
(unimportant stuff deleted)

> o IP/TCP, NFS
>
>I think most of it will come eventually, just because people need/miss it.
>
>/Tommy
>
>--
>/Tommy Thorn, another LPF member. Terminate software patents.
>Join The League for Programming Freedom, mail [email protected]
>Are *you* sure that *your* program isn't already covered by patents?

Who exactly "owns" NFS, I was under the impression that it was licensed from
sun.

Benjamin

--
-------------------------------------------------------------------------------
Benjamin Cline | Never trust a computer you can't pick up!
Systems Engineering |----------------------------------------------------------
MathSoft, Inc. | #include


[next article]
From: [email protected] (Drew Eckhardt)

Subject: Re: catman
Message-ID: <[email protected]>
Date: 26 Feb 92 04:27:47 GMT
References: <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 37
Nntp-Posting-Host: romeo.cs.colorado.edu

In article <[email protected]> [email protected] .ac.uk writes:
>
>ok so i'm subscribed to alt.pulp... anyway, i have been looking around
>and cannot find the sources for catman to generate manual page entrys
>with.
>
>I have a good idea that it would want nroff and troff but i'm prepared to
>smeg the code and get it working as i'm not fond of the present method
>of editing the whereis file.
>

Cawf is a fairly reasonable nroff substitute, and comes with man and ms
macro packages. It prints anoying warning messages, but these are
easily deactivated -

>alternively, has someone else done this? if not i volunteer (subconciously
>not a good idea) to crack the code on this (what have i said ?!??).
>
>comments?
>
>Phil
>=--=

Editing whatis :

When I added some things to whatis, I decided that writing a short
shell script would be simpler. It's marred by my imperfect use of
sed, and dies when it encounters a non-standard man page, but I have a
makewhatis script if people are interested.

A better solution would be Tom Christensen's makewhatis and man
perl scripts - they make use of dbm databases, and will maintain compressed
cat pages, etc. dbm and col are required.

Cat pages :

Cawf is not too slow,


[next article]
From: [email protected] (Dan Stromberg)

Subject: tsx-11, new upload (was Re: bug fix for the new gcc 1.40)
Message-ID: <[email protected]>
Date: 26 Feb 92 04:16:13 GMT
References: <[email protected]>
Organization: College of Engineering, University of Cincinnati
Lines: 41

In article <[email protected]> [email protected] writes:

>I've also taken the time to process all of the uploaded files that have
>hit ~ftp/incoming in the past few days. To make things easier on me
>(I've been fairly time-crunched at work lately), I'd appreciate it if:

(Good suggestions deleted)

Three of those things I uploaded, without attribution:
- sdbm, an ndbm clone
- m4, a preprocessor
- John Haugh's shadow password suite.

I used the first two, and a number of pieces of poe's work, in building
the third.

The third contains:

chage chfn chpasswd chsh dialup.h dpasswd faillog gpasswd groupadd
groupdel groupmod groups id libshadow.a login login.defs logoutd
mkpasswd newgrp newusers passwd pwconv pwd.h pwunconv shadow.h
su sulogin useradd userdel usermod

If anything, it's too much stuff. My 3M root partition is nearly
full. Shared libraries for this would be kind of nice.

On tsx-11, the shadow password stuff is split into two files:
shadow-src.tar.Z and shadow-bin.tar.Z. You *can* use the binaries
without grabbing the source, but there's a small kernel patch in
the src file, which makes it look nicer... It adds ECHONL support,
so when you hit enter after typing a password, the enter is echoed.

shadow-src contains a README file. shadow-bin contains a script
"quick.install", taken from a "make -n install", to get it
running on your system with minimal effort.

> - Ted

Thanks, Ted.

- Dan


[next article]
From: [email protected] (Lawrence C. Foard)

Subject: tubes on tsx-11
Message-ID: <[email protected]>
Date: 26 Feb 92 09:55:59 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Lawrence C. Foard)
Organization: The Internet
Lines: 11


I've uploaded my kernel+some demo programs to tsx-11.mit.edu as tubes.tar.Z.
I tried to patch a fresh kernel and then make a diff from that but ran into
some weird problems which might also explain the problems people had with
the floppy patch. The fresh kernel seems to have some things mine doesn't.

The kernel I uploaded is the one I've been using, floppy+vc patchs.
All the changes in fs/*.c (except select) are surrounded with #ifdefs so
they could be cut and pasted in to another kernel. I'm hoping that once
Linux has VFS it will be possible to make a simple patch instead of
modifying a bunch of files.


[next article]
From: [email protected] (Michael Bethune)

Subject: Re: More questions about Linux ๐Ÿ™‚
Message-ID: <[email protected]>
Date: 26 Feb 92 12:59:32 GMT
References: <1992Feb19.074831.6031 @daimi.aau.dk> <[email protected]>
Organization: Pyramid Technology Corp PL, Melb., Australia
Lines: 36

In article <[email protected]> [email protected] (Benjamin Cl ine) writes:
>In article <[email protected]> [email protected] (Tommy Thor n) writes:
>>
>>Would everybody please stop calling Linux *a small system*. It's small
>>yes, but already very complete, and I hope see several more things yet, like:
>>
>> o IP/TCP, NFS
>>
>>I think most of it will come eventually, just because people need/miss it.
>>
>Who exactly "owns" NFS, I was under the impression that it was licensed from
>sun.
>

Sun have placed the specification of all three components of NFS into the
public domain, RPC (Remote Procedure Calls, and in this case they have also
placed an implementation into the public domain), XDR ( External Data
Representation) and the NFS protocol itself which sits on top of the above two.

They did this to encourage the wide spread adoption of NFS and were
successful in this.

i.e. OSF provide an implementation of NFS in OSF/1 which originates from Guelph
University (where ever that is) which doesn't include any Sun Microsystem's
Code and which consequently isn't covered by any Sun Copyrights.

This style of enlightened marketing has been one of the great things about
Sun, hopefully ( in spite of recent signs to the contrary :-() sun will
continue this open approach.

Cheers.
--
Michael Bethune Disclaimer: I have no commercial
Independent Unix Consultant. relationship with Pyramid and
Phone: +61 3 018 538103 do not represent them in any way.
Melbourne, Australia.


[next article]
From: [email protected] (R. Ramesh)

Subject: May be a small bug in poe-IGL
Message-ID: <[email protected]>
Date: 26 Feb 92 16:12:17 GMT
Sender: [email protected]
Reply-To: [email protected]
Organization: The University of Texas at Dallas, ACC
Lines: 23
Nntp-Posting-Host: ramesh.utdallas.edu

Hi everyone:

I saw an article a couple of days ago complaining that init takes too much
cpu cycles. I got into the same problem yesterday when I installed the poeigl
stuff for the first time. Like Peter (author, poeigl) points out init should
in fact be sleeing most of the time. A look at the code and the README file
reveals that all error messages are displayed on /dev/console and I didn't
have console. SO I made /dev/tty1 as console and then I found out that init was
forking like crazy with exec failing. A look at inittab didn't reveal anything
at first sight. But after a close look I found a blank line in it. A quick
look at init.c revealed that blank lines in inittab are not ignored. So, my
init was happily forking and execing a null string as file to execute.
Naturally it was failing and restarting etc... Deleting the null line and
sending a SIGHUP to init solved the problem.

All those who had similar problems may want to check their inittab for null
lines and spelling mistakes. Also don't forget the /dev/console.

Peter, if you plan to fix this can you also make login and agetty to write
error messages to /dev/console so that by linking appropriate /dev/tty? we can
move error messages as per our liking.

Ramesh.


[next article]
From: [email protected] (Mark Saltzman)

Subject: trouble dialing in with poe-IGL
Message-ID: <[email protected]>
Date: 26 Feb 92 18:34:39 GMT
Sender: [email protected]
Reply-To: [email protected] (Mark Saltzman)
Organization: Case Western Reserve University, Cleveland, OH (USA)
Lines: 13
Nntp-Posting-Host: cwns2.ins.cwru.edu


When I try to dial in to poe-IGL using a hayes compatible 2400 baud
modem i get stuff like this:
(none) login: password: Login incorrect
repeated a few times, without giving me the chance to respond to the
login: prompt. What options do I need to give agetty in inittab? What
options should my terminal emulator use (XON/XOFF, CTS/RTS, etc...).
Also, I have not yet dared to modify my kernel in order to have it run
init. When I want to use it, I just type init & from the shell. Is
this ok?
thanks for any info,
-mark


[next article]
From: [email protected] (Paul Richards)

Subject: linux-standards
Message-ID: <[email protected]>
Date: 26 Feb 92 12:48:16 GMT
Sender: [email protected] (Network News System)
Organization: University of Wales College at Cardiff
Lines: 13
X-Mailer: Cardiff Computing Maths PP Mail Open News Gateway

There seem to be problems getting mail to banjo.concert.net from the uk
(maybe elsewhere?). I can't therefore subscribe to linux-standards. Does
anyone know what the problem is. Could an alternative route be
established?

In the meantime could I be subscribed, there is no problem with mail the
other way.
--
Paul Richards at Cardiff university, UK.

[email protected] Internet: spedpr%[email protected]
UUCP: [email protected] or ...!uunet!mcsun!ukc!cf!thor!spedpr
+++


[next article]
From: [email protected] (Paul Richards)

Subject: startup shell?
Message-ID: <[email protected]>
Date: 26 Feb 92 13:27:38 GMT
Sender: [email protected] (Network News System)
Organization: University of Wales College at Cardiff
Lines: 15
X-Mailer: Cardiff Computing Maths PP Mail Open News Gateway

I've created a /home/root directory and modified /etc/passwd accordingly
but it doesn't work. It's just occured to me that /etc/passwd isn't even
used since I don't have the login stuff installed yet.

What happens when I start up. The home directory is always set to
/usr/root and a set of environment variables appear from somewhere.

What is setting things up this way and will I have to get the login
stuff working to change it.
--
Paul Richards at Cardiff university, UK.

[email protected] Internet: spedpr%[email protected]
UUCP: [email protected] or ...!uunet!mcsun!ukc!cf!thor!spedpr
+++


[next article]
From: [email protected] (Vinter Brian)

Subject: HELP (or am I an idiot?)
Message-ID: <[email protected]>
Date: 26 Feb 92 08:07:54 GMT
Organization: Mathematics and Computer Science, Aalborg University
Lines: 50

Help please!!!

It's not that I want to be stupid but I'm getting a little paranoid....
The problem is (yes again) instalation to HD.
I know that there is a partion.faq and I read it a million times, give or
take a few, but it still dont work.

Now the setup is as follows...
My first harddisk is a 43 MB drive, I have split it into two partitions
at 23 MB and 20 MB, I got DOS running on the first partition and I want to
install Linux on the second partition, just like everybody else...

I boot Linux from floppy, fdisk tells me that I have two drives and that
the second drive holds 20046 blocks, nice!

mkfs -c /dev/hd2 20046 works allright two, no sweat.

Mounting /dev/hd2 /user goes well two.

The recursive copy function works, and I can cd /user and actually play around
the harddisk, this stuff is just okay, but I sync (yes 3 times don't wanna
upset the UNIX god where ever he is:))

Reset and boot the boot disk which is changed to 508=3 509=2 comes up with
the usual Press enter to see SVGA modes....
.
.
.
Insert rootsystem into floppy and press enter (I cant remember the exact words
but something like that you know it...)

BUT NOW WHEN I PRESS ENTER I JUST GET

Reset floppy called.
Reset floppy called.
Reset floppy called.
Reset floppy called.
Reset floppy called.
Reset floppy called.

You get the picture ?

I have tried the above drill about 15 times, please what do I do wrong?


Otherwise Linux seems just like what I needed, and I lokking forward to get
it running.


vinter


[next article]
From: [email protected] (Philip Copeland)

Subject: emacs...shell problem? or just me?
Message-ID: <[email protected]>
Date: 26 Feb 92 19:32:04 GMT
Sender: [email protected] (Usenet poster for nntp (tj - 17/12/91))
Reply-To: [email protected]
Organization: Bristol-Polytechnic
Lines: 42
Nntp-Posting-Host: beans

Well I'm knackered. I just finished showing off LINUX 0.12 to all the main
computing staff (CSM admissions tutor, SUN transputer manager, main unix
administrators and other asorted staff) (i've never been so terified in my
life!)

I 'modified' 6 machines (opel, emerald, diamond.. etc) and created an almost
exact duplicate of our main sun system's directory hirarchy and passwd files
.... extended thanks go to linus torvolds, [email protected], [email protected],
and all the other people i've pestered over the last few days 8)

verdict?... they all want one and bug**r the sun solaris effort that they had
all been reading about!... tom (sys manager of csd) wants xwindows, will (sys
of the workstation network) wants a 386.. badly, andy (transputer boss & sys
admin) will possibly donate a couple of tranputer 486's.... and wants a copy
for the one he has at home,... (sigh)...

QUOTE of the day: (Andy) ... look tom,.. it's unix,.. it's real unix!...

all went exceedingly well and except for the occasional bsd sysv conflict,
they all seemed quite at home. (I think it was the food that i provided that
set it all off actually).


The thing is, these people are seasoned bsd unix administrators and although
almost everything worked, there was ONE small/big? problem

EMACS 18.57 !!!!!

tom tried to load up a shell in emacs and ... it hung,... the processor
died and i felt 2" high... (gurr) i tried it out and sure enough
emacs hangs trying to start a new shell

has anyone else come across this? or is it something that i've not done?
the other features are all there but the shell seems broke,... i can't
recover from it either 8( not even in a second virt con

the machine is a 386 16mhz sx 40 meg hd, 9 meg swap space(overkill) 4 meg
memory...

any takers?

Phil
=--=


[next article]
From: [email protected] (Paul C. Janzen)

Subject: Floating point performance
Message-ID: <[email protected]>
Date: 26 Feb 92 19:16:46 GMT
Sender: [email protected]
Organization: The University of Texas at Austin, Austin TX
Lines: 25


I just got GCC-1.40 and the libsoft.a. ๐Ÿ™‚ ๐Ÿ™‚ ๐Ÿ™‚

So I took a typical differential equation solution program (alright,
it's braindead, but it's interesting) and compiled it using

gcc -O -Wall -o vibr vibr.c -lsoft

and timed the results. It took *9 TIMES* as long as the same program
compiled under MessedupDos using BC++ 3.0 to run. (only a few doshells
running, besides the program itself.)

All the algorithm does is take a sin(), a sqrt(), and some
multiplications and divisions on each step.

Is libsoft.a just slow, or is it the overhead of Linux, or what?
Would I really see a >10x performance gain by getting a coprocessor?

Thanks!

------------------------------------------------------------------------
Paul C. Janzen Practical people would be more practical
[email protected] if they would take a little more time for
dreaming. -- J. P. McEvoy
------------------------------------------------------------------------


[next article]
From: [email protected] (Drew Eckhardt)

Subject: Making whatis, roff
Summary: bash script to make whatis
Keywords: whatis
Message-ID: <[email protected]>
Date: 27 Feb 92 00:47:35 GMT
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 38
Nntp-Posting-Host: ophelia.cs.colorado.edu

Ok, as promised here it is. It only makes whatis from roff manpages, and will
puke if the man pages are not standard format
(IE : .SH NAME
name1[,name2, name3 ...] - description
is not there)

Note that by changing it to look for the resulting "straight ASCII"
of NAME , and by extracting section from the file name,
you can work on CAT pages instead.

Output is to stdout.

Problem : I don't have roff!
Solution : I have uploaded cawf binaries to nic.funet.fi under
incomming. The only changes to the distribution sources
were to ignore non-fatal error messages.

Simply install it, make a symlink to nroff from cawf,
and be happy.
--makewhatis--
#!/bin/sh
cd /usr/man/
(
for DIRECTORY in `ls -d man* `; do
cd /usr/man/${DIRECTORY}
FILES=`ls`
if test "$FILES"; then
for FILE in $FILES; do
WHAT=`echo $FILE | awk -F. '{print $1}'`
THSECTION=`grep .TH $FILE | awk '{print $3}' | head -1`
WHERE=$[`grep -n ".SH NAME" $FILE | awk -F: '{print $1}' ` + 1]
head -$WHERE $FILE | tail -1 | tr '\\' ' ' | \
awk -F- '{printf("%s('${THSECTION}') \t- %s\n"), $1,$2 }'
done
fi
cd ..
done
) | sort


[next article]
From: [email protected] (Mark Saltzman)

Subject: some bugs in gcc 1.40
Message-ID: <[email protected]>
Date: 27 Feb 92 02:03:25 GMT
Sender: [email protected]
Reply-To: [email protected] (Mark Saltzman)
Organization: Case Western Reserve University, Cleveland, OH (USA)
Lines: 18
Nntp-Posting-Host: cwns2.ins.cwru.edu


It seems that the new gcc has some problems with converting floating
point numbers to ints. Try 'printf( "%d\n", (int) 0.51 )'. It appears
to be rounding instead of truncating.

I have also noticed some problems with v*printf. v*printf should accept
a va_list as a formal parameter which is then passed on to _doprnt.
However in the new libc.a, v*printf acts just like *printf, with a
variable formal parameter list. It then attempts to create the va_list
itself which it then incorrectly passes on to _doprnt.

I have also
noticed dramatic reductions in floating point performance in comparison
with the old libc.a. Oh yeah, the floor() and ceil() functions (or
maybe just one of them) cause a segmentation fault.

mark


[next article]
From: [email protected] (Quillen Edward O)

Subject: BASH trifle
Message-ID: <[email protected]>
Date: 27 Feb 92 00:25:05 GMT
Sender: [email protected]
Distribution: usa
Organization: Dept. of Computer Science, Michigan State University
Lines: 21
Originator: [email protected]


This is probably a really silly question, but a necessary one for me
to be productive with Linux:
How do I imbed the current working directory in the prompt?
In csh I use:
alias cd 'cd \!*; set prompt="<`dirs`>"'

I can't get this to work with Bash.( using PS1 etc...)

+++++++++++++++++++++++++++++++++++++++++++++++++
+++ ED (Edward Quillen) 6-1293 +++
+++ Vet Teaching Hosp. +++
+++ Email: [email protected] +++
+++ [email protected] +++
+++ [email protected] +++
+++++++++++++++++++++++++++++++++++++++++++++++++
--
+++++++++++++++++++++++++++++++++++++++++++++++++
+++ ED (Edward Quillen) 6-1293 +++
+++ Vet Teaching Hosp. SYSTEM MANAGER +++
+++ Email: [email protected] +++


[next article]
From: [email protected] (Al Clark)

Subject: Re: Can't run off of hard disk.
Message-ID:
Date: 26 Feb 92 23:43:18 GMT
References:
Organization: Netcom - Online Communication Services (408 241-9760 guest)
Lines: 13

[email protected] (Al Clark) writes:
: I am using 0.12 boot and root images.

- report on modification of boot image omitted --

: Am I missing something obvious???

Thanks to the sharp eyes of [email protected], who noted that I was
modifying the "root" rather than "boot" floppy. Which was missing
something obvious.
--
Al - [email protected] - My opinions are my own.
*** Commit acts of random kindness and senseless beauty! ***


[next article]
From: [email protected] (Paul M Schwartz)

Subject: WD ESDI works, but only 1 partition...
Message-ID: <[email protected]>
Date: 27 Feb 92 03:58:00 GMT
Sender: [email protected]
Organization: University at Buffalo
Lines: 13
News-Software: VAX/VMS VNEWS 1.41
Nntp-Posting-Host: ubvmsb.cc.buffalo.edu


My WD xx07-2 controller with sector translation seems to work just fine
under Linux; however, I can only mkfs a single partition. Anyone have any
ideas? All the partitions are of the same type. They are hd7-9, 10, 32,and
8 megs. The 10 meg works just fine, but mkfs refuses to work on the other
two. When I make it one big partition, mkfs works just fine. Help please.
I'm gonna try to make the 3rd partition a swap drive tonight.


PauL M SchwartZ | Follow men's eyes as they look to the skys
PUNK!, geez, anyways | the shifting shafts of shining
[email protected] | weave the fabric of their dreams.
[email protected] | - RUSH -


[next article]
From: [email protected] (Adam Thompson)

Subject: Problem with UNCOMPRESS
Message-ID: <[email protected]>
Date: 27 Feb 92 03:25:47 GMT
Organization: University of Manitoba, Winnipeg, Canada
Lines: 37

I've installed Linux (v.0.12, no patches yet - read on to see why)
on my second HD (40Meg, Connors IDE somethingorother), and leave my first
40M HD as DOS. I boot linux off floppy. (I can diable speed switching, so
it's acceptably fast)

I d/l (using kermit - when will serial I/O work properly?) all the compiler
binaries, as86, bison, flex, gcc, and stuck them into /tmp.
Then I try 'uncompress as86.tar.Z' ... it works for a little while, then
hangs the machine. I can type for about ~15-25 seconds after I execute
(fg or bg - doesn't matter except in timing) and then the whole machine just
sits there ... a nice-looking paperweight.

I have 4M RAM, no swap space. No error messages. It'll sync until it dies,
so it seems there's a specific point at which uncompress stops...

the same thing happens no matter which file I try to uncompress.
I didn't have this problem with the tar.Z that came on the root image.
I don't have this problem when I 'compress' something and then 'uncompress'
again.

Any ideas, people?

BTW: when will 0.13 be released? 2) what is the secondary v#(0.95) that I keep
seeing ?

-Adam Thompson

--
= Adam Thompson ---- Computer Engineering ---- University of Manitoba =
= [email protected] = "When you have eliminated the improbable, =
= ...!uunet!decwrl!alberta!\ = whatever is left, however impossible, =
= ccu.UManitoba.CA!umthom61 = must be the answer." =
--
= Adam Thompson ---- Computer Engineering ---- University of Manitoba =
= [email protected] = "When you have eliminated the improbable, =
= ...!uunet!decwrl!alberta!\ = whatever is left, however impossible, =
= ccu.UManitoba.CA!umthom61 = must be the answer." =


[next article]
From: [email protected] (kevin dahlhausen)

Subject: emacs question
Message-ID: <[email protected]>
Date: 27 Feb 92 04:53:35 GMT
Sender: [email protected] (kevin dahlhausen)
Organization: The Ohio State University Department of Computer and Information S cience
Lines: 12



I've seen this once before, but don't remember ever seeing
the answer. The status-line in Emacs scrolls up the screen
when you do a page up or down. ^L clears the mess.
Is there a soloution to this problem?

Thanks

=======================================================================
Kevin Dahlhausen "Never hire a programmer that
[email protected] can communicate with people."


[next article]
From: [email protected] (Charles Hedrick)

Subject: Re: catman
Message-ID:
Date: 26 Feb 92 22:27:15 GMT
References: <[email protected]> <[email protected] colorado.edu>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 4
To: [email protected]

I agree that cawf is OK for normal man pages (though I had to add a
definition for the SS macro). However Berkeley is now using a
different package, mandoc. Does anyone know of a macro definition
file for mandoc that works with cawf?


[next article]
From: [email protected] (Risto Kankkunen)

Subject: Re: where to get ALL of linux?
Message-ID: <[email protected]>
Date: 25 Feb 92 10:35:57 GMT
References: <1992Feb21.122 [email protected]>
Organization: University of Helsinki, Department of Computer Science
Lines: 30

I.J. Wells writes:
> At nic, there appears to be both archives of source and source file by
> file, which would be OK except there are several tar files with no idea
> as to what's in them. A few hints as to what's worth fetching in the
> README's wouldn't hurt.

I think it would be best to have all the stuff (at least sources) file
by file. Then you could browse around in the directories and get what
you need. If you want whole packages, you can use the special features
of the ftp server in nic. Quote from the /README file:

> NON-STANDARD FTP FEATURES:
>
> This FTPSERVER has some special features which may help on grabbing
> files from here:
>
> .Z Compress designated file while
> fetching from this system.
> .tar Make a TAR (using GNU tar-1.07) of dir.
> .tar Make a TAR (- " -) of file.
> .tar.Z Make a compressed tar of dir.
> .tar.Z Make a compressed tar of file.
> (Why to tar a single file? To get its date & time
> information too...)
> +".Z" How to say... There exists file: .Z,
> you ask it to be UNCOMPRESSED for FTP transfer...
> (For all of you who can't uncompress a'la BSD UNIX.)

--
no sig today


[next article]
From: [email protected] (D.Bolla)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Keywords: v86, video
Message-ID: <[email protected]>
Date: 27 Feb 92 08:50:09 GMT
References: <[email protected]> <[email protected] du>
Reply-To: [email protected] (D.Bolla)
Organization: Computing Lab, University of Kent at Canterbury, UK.
Lines: 48

In article <[email protected]> [email protected] (Joel M. Hoffman ) writes:
>In article <[email protected]> [email protected] (Drew Eckhardt) writes:
>I think there are *many* good reasons to rewrite mm. First of all, we
>(at least I) certainly want a DOS emulator down the road. The easiest
>way to do this is to run the BIOS, and then DOS, in V86 mode, I think.
Let's face it. A V86 mode is almost useless.
What kind of programs (decent programs) do you see that use an 8086 ?
( That's a V86 mode ).....
If you are not using the V86 mode then you have to start to catch
interrupts, look for memory access amnd all dirty stuff.
It seems to me that people underestimate the complexity of building
an almost buf free OS.

If you and other people like to live with a DOS type OS ( That also
crashes quite often ) well..... I see a sorrow end for Linux.

>Also, I think a more general page table scheme would be nice on its
>own merits. (I can see having more than 64 tasks, e.g.)
Yes. I agree with this BUT there are other things FIRST.
Linux is not "standard" yet. In the sense that the math emulation in the
kernel is not finished ( And math emulation relly has to be in kernel )

The OS in general has very little problems in compatibility that make you
do some changes in the source code of PD programs to make them compile and
run.

This has to be fixed. MM and the rest can wait.


>One place to start might be the code from DJGPP. DJGPP is a port of
>GCC to DOS (!). It contains a lot of code for switching back and
>forth between 32- and 16- bit modes.

Again, kernel bigger ( probably 4mb ram will not be enought anymore )
Errors around.... For what ? To play with DOS ?
Remembar that DOS is not so clean. Not only you have do deal with DOS
but with all the rubbish of Extended Expanded EMM .....
If your emulator is not EQUAL to DOS then you DOS program will hang.

Do you want DOS to hang in the middle of you wordprocessor ?
I don't

>The other option, of course, is to copy the BIOS into a process's
>memory (in protected mode), and try to make use of it. I don't know
>how difficult this would be.
Please forget it.

Damiano


[next article]
From: [email protected] (Joel M. Hoffman)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Keywords: v86, video
Message-ID: <[email protected]>
Date: 27 Feb 92 11:49:32 GMT
References: <[email protected]> <[email protected] du> <[email protected]>
Sender: [email protected] (USENET News system)
Organization: University of Maryland, College Park
Lines: 26
Nntp-Posting-Host: next.wam.umd.edu

In article <[email protected]> [email protected] (D.Bolla) writes:
>In article <[email protected]> [email protected] (Joel M. Hoffma n) writes:
>>In article <[email protected]> [email protected] (Dre w Eckhardt) writes:
>>I think there are *many* good reasons to rewrite mm. First of all, we
>>(at least I) certainly want a DOS emulator down the road. The easiest
>>way to do this is to run the BIOS, and then DOS, in V86 mode, I think.
>Let's face it. A V86 mode is almost useless.
>What kind of programs (decent programs) do you see that use an 8086 ?
>( That's a V86 mode ).....

The fact of the matter is that there are many good DOS programs:
WordPerfect and Quicken come immediately to mind. While I don't
actually use WP and more, I know many people that do, and I'd like to be
able to use Linux without giving up such nice DOS programs. Most
programs still run under a simple 8086. I also have several games for
DOS that I simply cannot port to Linux. (Scrabble, Chess, etc.)

>
>Do you want DOS to hang in the middle of you wordprocessor ?
>I don't
>

If the DOS emulator dies it's not nearly to critical as if the OS
dies. Under Linux, everything else will keep running. Not SO bad....

-Joel


[next article]
From: [email protected] (Bruce Varney)

Subject: Re: BASH trifle
Message-ID: <[email protected]>
Date: 27 Feb 92 13:30:16 GMT
References: <[email protected]>
Sender: [email protected]
Distribution: usa
Organization: Purdue University Computing Center
Lines: 29

In article <[email protected]> [email protected] u (Quillen Edward O) writes:
}
}This is probably a really silly question, but a necessary one for me
}to be productive with Linux:
} How do I imbed the current working directory in the prompt?
}In csh I use:
}alias cd 'cd \!*; set prompt="<`dirs`>"'

1) The \!* construct does not work in aliases under bash as it does in csh.
2) There are already special prompt thingies set up for bash.
\t the time
\d the date
\n CRLF
\s the name of the shell, the basename of $0 (the
portion following the final slash)
\w the current working directory
\W the basename of the current working directory
\u the username of the current user
\h the hostname
\# the command number of this command
\! the history number of this command
\$ if the effective UID is 0, a #, otherwise a $
\nnn character code in octal
\\ a backslash

Thus, if you want your prompt to be "hostname:directory -> ", then do
PS1="\h:\w -> "

Bruce


[next article]
From: [email protected] (Linus Benedict Torvalds)

Subject: Re: Floating point performance
Message-ID: <[email protected]>
Date: 27 Feb 92 13:42:13 GMT
References: <[email protected]>
Organization: University of Helsinki
Lines: 31

In article <[email protected]> [email protected] (Paul C. Janzen) write s:
>
>[math program with soft-float]. It took *9 TIMES* as long as the same program
>compiled under MessedupDos using BC++ 3.0 to run. (only a few doshells
>running, besides the program itself.)
>
>All the algorithm does is take a sin(), a sqrt(), and some
>multiplications and divisions on each step.
>
>Is libsoft.a just slow, or is it the overhead of Linux, or what?
>Would I really see a >10x performance gain by getting a coprocessor?

As I have moved toward software emulation in the kernel, things have
generally gotten slower: I never even tried to make my soft-float
routines fast (did you get the right results?) - my feeling was that
anybody who wanted math-capabilities beyond a few adds etc would have a
387 or 486. The kernel emulation doesn't make things faster: it just
makes it /easier/, as you never need care about the existence of the
387.

In fact the current math-emulation is probably pessimal when it comes to
the speed of things like 'sin()' etc: I'd guess sin and other more
complicated functions use the emulator add/mul to calculate the results,
so when calculating sin you get /a lot/ of the math-not-present
exceptions. Things will probably get /a bit/ better when all the
commands are inside the kernel (not trapping all the time), but even
then I'd suggest you get a math-coprocessor if you intend to calculate a
lot: I wouldn't be the least surprised to see much more than >10x
performance gains if you use sin etc extensively.

Linus


[next article]
From: [email protected] (Humberto Ortiz-Zuazaga)

Subject: Re: Floating point performance
Summary: 80387 is MUCH faster.
Keywords: libsoft.a, math, 80387.
Message-ID: <[email protected]>
Date: 27 Feb 92 15:48:15 GMT
References: <[email protected]> <[email protected]>
Organization: Univ. of Cincinnati
Lines: 30

In article <[email protected]> [email protected] FI (Linus Benedict Torvalds) writes:
>In article <[email protected]> [email protected] (Paul C. Janzen) writ es:
>>
>>[math program with soft-float]. It took *9 TIMES* as long as the same program

>>Is libsoft.a just slow, or is it the overhead of Linux, or what?
>>Would I really see a >10x performance gain by getting a coprocessor?
>
>As I have moved toward software emulation in the kernel, things have
>generally gotten slower: I never even tried to make my soft-float
>routines fast (did you get the right results?) - my feeling was that

I think he was using H.J.'s newlibc, not the kernel stuff.

>lot: I wouldn't be the least surprised to see much more than >10x
>performance gains if you use sin etc extensively.

Actually, I have a 20MHz SX at home without a coprocessor, and a 16MHz
SX with a coprocessor at work. I compiled a program with libsoft and
libm, and tested the libsoft compiled program at home. 50 iterations in
3 minutes (it does many ldexp's). At work, off of a floppy disk, the
same program, 50 iterations in less than a second.

Interestingly, I can't tell the difference between the libsoft and libm
linked programs on the machine with the coprocessor. They both run just
as fast. Do libsoft linked programs use the coprocessor when present?
--
Humberto Ortiz-Zuazaga INTERNET: [email protected]
Dept. of Physiology & Biophysics BITNET: [email protected]
University of Cincinnati CIS: 72301,2303


[next article]
From: [email protected] (Adam Goldberg)

Subject: Re: emacs question
Message-ID:
Date: 27 Feb 92 16:12:21 GMT
References: <[email protected]>
Sender: [email protected] (USENET News System)
Organization: Iowa State University, Ames IA
Lines: 21
Originator: [email protected]

In <[email protected]> [email protected] du (kevin dahlhausen) writes:

>I've seen this once before, but don't remember ever seeing
>the answer. The status-line in Emacs scrolls up the screen
>when you do a page up or down. ^L clears the mess.
>Is there a soloution to this problem?

All you need to do is:

TERM=vt100

then run emacs and everything'll be just dandy.

--
Adam Goldberg ! Ask me how IOWA STATE UNIVERSITY is
[email protected] ! censoring my usenet access!!
..!uunet!iastate.edu!goldberg ! >>I don't need a BIG BROTHER!!<<
--
Adam Goldberg ! Ask me how IOWA STATE UNIVERSITY is
[email protected] ! censoring my usenet access!!
..!uunet!iastate.edu!goldberg ! >>I don't need a BIG BROTHER!!<<


[next article]
From: [email protected] (Wayne Davison)

Subject: v86 mode is [not] useless
Keywords: v86
Message-ID: <[email protected]>
Date: 27 Feb 92 16:59:55 GMT
References: <[email protected]> <[email protected] du> <[email protected]>
Organization: Borland International
Lines: 18

D.Bolla ([email protected]) wrote:
> Let's face it. A V86 mode is almost useless.
> What kind of programs (decent programs) do you see that use an 8086 ?

You are seriously underestimating the useability of V86 mode. The DOS
memory manager QEMM uses V86 mode to provide all sorts of loadhigh features
and memory management services. This broke programs that tried to switch
into protected mode and do their own memory management, but it was so
popular that most programs were re-written to include VCPI support (which
allows them to work with or without QEMM being present). These days you
also have to support DPMI so you can run Windows 3.0 without resorting to
real mode. Look at OS/2 2.0 -- they are providing an OS that uses V86
dos boxes complete with DPMI (and I believe VCPI) support, so it CAN be
done. Don't think that it's going to be easy, though.
--
\ /| / /|\/ /| /(_) Wayne Davison
(_)/ |/ /\|/ / |/ \ [email protected]
(W A Y N e)


[next article]
From: [email protected] (Hayes John H)

Subject: Re: gcc 1.04 libraries and utilities
Keywords: gcc libraries include
Message-ID: <[email protected]>
Date: 27 Feb 92 16:17:56 GMT
Sender: [email protected]
Reply-To: [email protected] (Hayes John H)
Organization: Dept. of Computer Science, Michigan State University
Lines: 9
Originator: [email protected]

I recently ftp's newgcc.tar and installed it on my machine ( I never got
gccbin.tar to work properly ) and tried to compile a c program ( man.c ).
I received many comlaints about missing items ( the header files were found
after some editing ). The question is, what library and include package
should I be using with gcc 1.04 and where should I get and ho should it be
installed. Also, are the standard library functions available such as
printf, scanf, fopen etc?

Thanks, John Hayes


[next article]
From: [email protected] (I Reid)

Subject: Hard disk boot (full), non shoelace questions
Message-ID: <[email protected]>
Date: 27 Feb 92 14:58:27 GMT
Distribution: alt
Organization: Edinburgh University
Lines: 22

I had a look at shoelace before but didn't like the look of it but would like
to boot completely from my hard disk (disabling floppy seek makes a big
difference to my boot up speed). I played around with bootany (a boot record
replacement which looks as though it will do the job) but, on experimentation
it successfully booted up into dos but couldn't boot linux. This wan't really
surprising because it had no image to tell it what to do. In the makefile for
linux (kernel) there is a disk: entry which I use to build an image for floppy
boot up. How do I get this into the appropriate place in my partition from
within the standard makefile? With a full backup I tried changing the dd etc
/dev/PS0 to dd etc /dev/hd3 and successfully attacked the kernel (in a fatal
and unpleasant way) so I would guess that this isn't the way to do things :-).

So...... the question is.....

In the disk entry in the main makefile a) what do I write and b) where do i
write it (i.e. which tools with which parameters?)

Many thanks,
Iain

p.s. How many blocks on a 1.44 HD disk? I thought I remembered it being 1440
but tar lost data with this (tar -L 1440 -Mzpf /dev/PS0 as a backup strategy)


[next article]
From: [email protected] (Adam Thompson)

Subject: Yikes! Thanks!
Message-ID: <[email protected]>
Date: 27 Feb 92 18:50:03 GMT
Organization: University of Manitoba, Winnipeg, Canada
Lines: 20

I'll make this short...
thanks to everbody who reminded me to 'set type binary' in kermit before
downloading!

My only question is, why does uncompress HANG on a not-quite-valid .Z
file ?? Shouldn't it be able to recognize that 7-bit data isn't quite
sufficient for a .Z file?

-Adam Thompson

--
= Adam Thompson ---- Computer Engineering ---- University of Manitoba =
= [email protected] = "When you have eliminated the improbable, =
= ...!uunet!decwrl!alberta!\ = whatever is left, however impossible, =
= ccu.UManitoba.CA!umthom61 = must be the answer." =
--
= Adam Thompson ---- Computer Engineering ---- University of Manitoba =
= [email protected] = "When you have eliminated the improbable, =
= ...!uunet!decwrl!alberta!\ = whatever is left, however impossible, =
= ccu.UManitoba.CA!umthom61 = must be the answer." =


[next article]
From: [email protected] (R. Ramesh)

Subject: A question on standards
Message-ID: <[email protected]>
Date: 27 Feb 92 20:09:38 GMT
Sender: [email protected]
Reply-To: [email protected]
Organization: The University of Texas at Dallas, ACC
Lines: 10
Nntp-Posting-Host: ramesh.utdallas.edu

I got hold of poeigl package a couple of days ago and installed it. I found
that it uses inittab/securetty instead of ttytab and ttys. I do not know the
standards. So, I ask those who know: Which is the standard in init/getty/login
setup ttytab or inittab? If poeigl is nonstandard I am thinking of
standardizing it with some stuff borrowed from qpl-init stuff. Of course, I
don't want to do this if some one is already working on it. Could relevant
folks reply to these please. If you prefer, you can e-mail to me at
[email protected]

Ramesh


[next article]
From: [email protected] (Bruce Harlick)

Subject: Re: v86 mode is [not] useless
Keywords: v86
Message-ID: <[email protected]>
Date: 27 Feb 92 20:06:19 GMT
References: <[email protected]> <[email protected]> <1992Feb [email protected]>
Sender: [email protected] (News Service)
Organization: Stanford University
Lines: 19

In article <[email protected]> [email protected] (Wayne Daviso n) writes:
>

>Look at OS/2 2.0 -- they are providing an OS that uses V86
>dos boxes complete with DPMI (and I believe VCPI) support, so it CAN be
>done. Don't think that it's going to be easy, though.

Sadly, VCPI will NOT run with OS/2 v2.0. You are quite right about
DPMI, however. I have heard that Phar Lap (DOS extender people) will
rewrite their extender to use DPMI, so new apps using that extender
will run under OS/2. I have found that while noticably slower than
under DOS, Lotus 123 v3.1 runs quite nicely under the DOS box of OS/2.
V86 machines *can* produce a very usable DOS environment.

Will
[email protected]
------------------------
SAIC does not share my opinions.
[email protected] is a friends account, but you can send me mail here, too.


[next article]
From: [email protected] (cs87ejh)

Subject: Re: startup shell?
Message-ID: <[email protected]>
Date: 27 Feb 92 14:39:25 GMT
References: <[email protected]>
Organization: Brunel University, Uxbridge, UK
Lines: 20

[email protected] (Paul Richards) writes:
: I've created a /home/root directory and modified /etc/passwd accordingly
: but it doesn't work. It's just occured to me that /etc/passwd isn't even
: used since I don't have the login stuff installed yet.
:
: What happens when I start up. The home directory is always set to
: /usr/root and a set of environment variables appear from somewhere.
:
: What is setting things up this way and will I have to get the login
: stuff working to change it.

/usr/root is hard coded into linux/init/main.c - have a look in there.

Roger
--
+=============================================================================+
| [email protected] Roger Binns Brunel University - UK |
|:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::|
| Humans were created by water to move it uphill |
+=============================================================================+


[next article]
From: [email protected] (Steven Harvey)

Subject: thnks for kermit & com4 help
Keywords: kermit com4
Message-ID: <[email protected]>
Date: 27 Feb 92 16:16:48 GMT
Organization: Northern Arizona University, Flagstaff AZ
Lines: 1

Thanks everyone for the kermit & com4 advice. It all works fine now.


[next article]
From: [email protected] (Roger Christie - CONTRACTOR Writer)

Subject: Installation problems
Message-ID: <[email protected]>
Date: 27 Feb 92 20:47:27 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Roger Christie - CONTRACTOR Writer)
Organization: The Internet
Lines: 25



Hi,

I've successfully got the floppy system up and gathered up the disk part itioning
programs, but there are two things I have questions about:

1. A viable way to patch the bootimage. Can this be done via debug on the DOS si de. Can
someone give a reasonably simple description of the procedure.

2. Assuming I get the root filesystem up on hd, I'm somewhat confused as to how to
go about bring over the other tools. Mtools is not part of the rootimage. Here's what
I'm doing:
a. Download to a Sun SPARCstation.
b. Write the file to a DOS floppy on the Sun
c. Take it home and copy it on to the DOS side of my PC.

Now here is where I'm confused. It sounds as though (reading the Install + Relea se notes)
that I can use rawrite to write the files back to floppy and then get at them fr om the Linux side.
This sounds supicious to me. If true, is the floppy a DOS floppy, or a floppy wi th a Linux filesystem
on it. I assume there is SOME way to do this or else mtools would be on the root image.

Thanks for any and all help

Roger Christie


[next article]
From: [email protected] (Drew Eckhardt)

Subject: Re: Problem with UNCOMPRESS
Message-ID: <[email protected]>
Date: 27 Feb 92 22:46:16 GMT
References: <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 12
Nntp-Posting-Host: juliet.cs.colorado.edu

In article <[email protected]> [email protected] (A dam Thompson) writes:
>I d/l (using kermit - when will serial I/O work properly?) all the compiler
>binaries, as86, bison, flex, gcc, and stuck them into /tmp.
>Then I try 'uncompress as86.tar.Z' ... it works for a little while, then
>hangs the machine. I can type for about ~15-25 seconds after I execute
>(fg or bg - doesn't matter except in timing) and then the whole machine just
>sits there ... a nice-looking paperweight.
>

If you have corrupt .Z files, this is what happens. Your choices are
1. Re download.
2. Get tars on floppy from some one else.


[next article]
From: [email protected] (Greg Lee)

Subject: Re: catman
Message-ID: <[email protected]>
Date: 27 Feb 92 22:58:03 GMT
References: <[email protected]> <[email protected] colorado.edu>
Sender: [email protected] (News Service)
Organization: University of Hawaii
Lines: 12
Nntp-Posting-Host: uhunix.uhcc.hawaii.edu

I have a program "fmt" that interprets many MAN commands
directly, and a very few nroff commands, so it is fast.
You can link it to the name "nroff" and "nroff -man" gives
a rough approximation of the real thing. (You can also
link it to the name "tex" and get an even smaller subset
of real TeX.)

It's available by ftp from ftp.Hawaii.edu
(= uhunix2.uhcc.Hawaii.edu) as /pub/linguist/fmt.tar.Z.

--
Greg Lee


[next article]
From: [email protected] (Jim Lick)
Newsgroups: comp.unix.bsd,comp.unix.sysv386,alt.os.linux

Subject: Re: Networking 2 release, BSDI, etc
Message-ID:
Date: 27 Feb 92 23:42:38 GMT
References: <[email protected]>
Organization: University of California, Santa Barbara
Lines: 61

[crossposted reply to a question in the new comp.unix.bsd group]

Hopefully this post will clear up some confusion about all the BSD and
BSD-like systems running around out there:

BSDI is producing a product called BSD386. It is supposed to be shipping
in beta form 'soon'. It will contain complete source code for $995. It
will include device support for things such as SCSI boards (Adaptek 1542B)
which are not in other BSD releases. I think they were planning on a
production release this summer, but seeing how the beta ship date has
slipped back, it's anyone's guess. But in any case, this is your best
bet in the near term for BSD on 386/486 machines. It will be ready to
run, be supported, offer some basic device support, and will still have
all source code available. (The source written by BSDI will not be
freely distributable.)

Next there is the BSD Networking Release II Tape. This is the latest
release of the free portions of BSD 4.3. There are still missing pieces
though, so as of right now you can't just download the things and run.
They also come only in source code form, no binaries. However, upon
looking closer, it was discovered that there were really only a few
pieces missing. Most of the 386 support was already there, and only
a few kernel routines were omitted because they contained AT&T code.
There is currently an effort by some netters to fill in the missing
pieces in order to complete the system. There is a mailing list to
coordinate efforts here. I can't remember the address offhand, but
someone just posted some info on the list here. The bad news is that
only basic device support is included, so it may be a while before a
free (for example) SCSI driver will be available.

There is also a system called Linux. This is a unix-like system
written from scratch by one guy over the last year or so, with help
from other netters over the last few months. It is not a complete
system yet, but it does boot, and has a compiler (gcc) and shell
(bash) working on it, among other things. There is a new group
for the system, alt.os.linux. With much input from enthusiastic
netters, progress is being made at a fierce rate.

In the far future there are two developments which may help. First,
BSD4.4 is rumored to be free of AT&T code, include many new features,
and support for many systems. I have heard that work is in progress
or pretty much completed for 386, Sun 4, and Sun 3 systems. However,
from what I have heard, release is still a long ways away. I have
heard someone say that 'next fall' will be the earliest it will
come out. And like BSD NetII it will probably include only minimal
device support for the 386.

The other future development is the Gnu OS based on their 'Hurd'
kernel. No telling when this will be out though.

So although there are some exciting things coming in the future,
right now the options are still very limited.


Jim Lick
Work: University of California | Play: 6657 El Colegio #24
Santa Barbara | Isla Vista, CA 93117-4280
Dept. of Mechanical Engr. | (805) 968-0189 voice/msg
2311 Engr II Building | "Young enough to wonder/Old enough
(805) 893-4113 | to wander/Fool enough to think he
[email protected] | knows the way" - Lilac Time


[next article]
From: [email protected] (Marc CORSINI)

Subject: uemacs kills symbolic link !!
Message-ID: <[email protected]>
Date: 28 Feb 92 00:11:09 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 25


Hi,

ln does not affect the right access (lrwxr-xr-x instead of lrwxrwxrwx)
this is well-known but I don't know if it as something to do with the
links disappeared.

IDENTIFICATION

pwd : /
action: ln -s /usr/root/.profile
verification: ok except the access
action: uemacs .profile (* slight modification *)
action: save and exit
action: ls -ail
observation: that's it, no more symbolic link

I've tried with gnu-emacs the link is kept /but/ a new file .profile~
is created.
I've tried with elvis-1.4 the link is kept, and no new file is created

BTW: uemacs is 3.11a

Hope this helps
Marc


[next article]
From: [email protected] (Curtis Yarvin)

Subject: Re: Problem with UNCOMPRESS
Message-ID: <[email protected]>
Date: 28 Feb 92 00:30:28 GMT
References: <[email protected]> <[email protected] orado.edu>
Sender: [email protected]
Organization: Brown University Department of Computer Science
Lines: 18

In article <[email protected]> [email protected] (Drew Eckh ardt) writes:
|In article <[email protected]> [email protected] ( Adam Thompson) writes:
|>I d/l (using kermit - when will serial I/O work properly?) all the compiler
|>binaries, as86, bison, flex, gcc, and stuck them into /tmp.
|>Then I try 'uncompress as86.tar.Z' ... it works for a little while, then
|>hangs the machine. I can type for about ~15-25 seconds after I execute
|>(fg or bg - doesn't matter except in timing) and then the whole machine just
|>sits there ... a nice-looking paperweight.
|>
|
|If you have corrupt .Z files, this is what happens. Your choices are
|1. Re download.
|2. Get tars on floppy from some one else.

Even if you do have corrupt .Z files, this shouldn't happen. If compress
can crash a machine this way, there is a kernel bug involved.

c


[next article]
From: [email protected] (Hai Pham)

Subject: setting video modes & emacs keys..
Message-ID: <[email protected]>
Date: 28 Feb 92 04:09:53 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Hai Pham)
Organization: The Internet
Lines: 20

I've been going thru the FAQ list and back issues of the Linux-Activists
Digests, but I haven't been able to figure out how on earth does one
change the default VGA 80x25 text mode into something more interesting
like 80x50 or 80x60. I've tried using setterm and TERM=con80x60
to no avail. HELP??

I've also been trying to figure out how to make emacs use the
cursor keys and editing keys (like PGUP/PGDOWN, etc). I know it
can be done because I've been using the cursor keys in emacs on
an Alliant fx/2800.. Anybody out there who can point me in the
right direction? (I'm pretty new to the wonders of emacs,
and know next to nothing of lisp, so you'll have to take it
slow.. ๐Ÿ™‚

Thanks,
Hai

ps. I think my first question really ought to be on the FAQ list..

email: [email protected]


[next article]
From: [email protected] (Charles Hedrick)

Subject: Re: Floating point performance
Keywords: libsoft.a, math, 80387.
Message-ID:
Date: 28 Feb 92 03:39:40 GMT
References: <[email protected]> <[email protected]> <199 [email protected]>
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 22

>Interestingly, I can't tell the difference between the libsoft and libm
>linked programs on the machine with the coprocessor. They both run just
>as fast. Do libsoft linked programs use the coprocessor when present?

Maybe. libsoft is really for things like the trig functions. For
normal multiplication and addition you don't use libsoft at all.
Unless you tell gcc to compile for software emulation, you get 387
instructions in the code. They are done in the hardware if you have
387, or in the kernel if you don't.

If you're doing lots of trig functions, then you'll see the
performance of libsoft. But even then you're not necessarily going to
slow down a machine with a 387. All libsoft does is replace the sin,
sqrt, etc. instructions that are built into the 387 (which Linux does
not emulate) with algorithms using only basic operations like addition
and multiplication (which Linux does emulate). If libsoft is compiled
with default gcc settings, the basic operations will still be done by
the 387 hardware if you have it. Complex algorithms aren't
necessarily any faster when implemented in hardware than software. In
fact sometimes software can be faster. It can normally be more
complex, so it can take test for more special cases where shortcuts
are possible.


[next article]
From: [email protected] (Rick)

Subject: Here's a thought...
Message-ID: <[email protected]>
Date: 28 Feb 92 05:56:38 GMT
Organization: Telecom Australia, TNE Computer Support Services
Lines: 7

Multi-processing with 486s - ie 2 or more 486s on the bus running the
one OS!

Linus?

Rick.
(Add emoticons as neccessary)


[next article]
From: [email protected] (Al Clark)

Subject: Re: Problem with UNCOMPRESS
Message-ID: <[email protected]>
Date: 28 Feb 92 05:53:54 GMT
References: <[email protected]> <[email protected] orado.edu> <[email protected]>
Organization: Self
Lines: 29

In article <[email protected]> [email protected] (Curtis Yarvin ) writes:
>In article <[email protected]> [email protected] (Drew Eck hardt) writes:
>|In article <[email protected]> [email protected] (Adam Thompson) writes:
>|>I d/l (using kermit - when will serial I/O work properly?) all the compiler
>|>binaries, as86, bison, flex, gcc, and stuck them into /tmp.
>|>Then I try 'uncompress as86.tar.Z' ... it works for a little while, then
>|>hangs the machine. I can type for about ~15-25 seconds after I execute
>|>(fg or bg - doesn't matter except in timing) and then the whole machine just
>|>sits there ... a nice-looking paperweight.
>|>
>|
>|If you have corrupt .Z files, this is what happens. Your choices are
>|1. Re download.
>|2. Get tars on floppy from some one else.
>
>Even if you do have corrupt .Z files, this shouldn't happen. If compress
>can crash a machine this way, there is a kernel bug involved.
>

By an amazing coincidence, I did this, (kermit text mode on .Z file)
right after mailing Adam Thompson the probable cause. I found it hung
the particular shell; the other shells available via virtual console
still worked, and when I killed the uncompress from one of them, the
original shell also worked. So I guess just the uncompress process
hangs. (At least sometimes)

--
Al - [email protected] - My opinions are my own.
*** Practice random kindness and senseless acts of beauty! ***


[next article]
From: [email protected] (Paul C. Janzen)

Subject: Xon / xoff bug?
Summary: It works once.
Message-ID: <[email protected]>
Date: 28 Feb 92 05:37:08 GMT
Sender: [email protected]
Organization: The University of Texas at Austin, Austin TX
Lines: 51

I am used to using ^S / ^Q for "flow control" on a terminal; that is,
if something is scrolling by, and I want to see it, my fingers hit
the ^S key to stop it. Well, this didn't work at first in Linux, so
I got stty and did a
stty ixon
which should have done the trick, I thought.

So I typed
cat huge_file
and about two pages into it I hit ^S. Great! It stopped! ^Q -restarted
it. Ohmighod, I thought I had it made.

BZZZZT. Thank you for playing.

Yes, cat restarted the file after I pressed ^Q. And about half a page
of huge_file scrolled up the screen. And then...NOTHING!!! Not a peep.
I could interrupt the cat job fine (^C) and then everything was back
to normal.

Well, I figured, I'm a hacker, I'll dive headfirst into the kernel
and see what's happening. (BZZZT! Alright, my first choice of a place
to put a printk was poor. But now I can say that I've actually used
a kernel that printed a "HEY!" before every single character. :-))

As far as I can tell, this is what happens when I type ^S:
the routine in kernel/chr_drv/tty_io.c that goes:
if (I_IXON(tty) ... /* tests for c==stop_char, etc.*/
gets called,

followed by (some random number between 1 and 4) calls to
console.c's con_write(), where it just goes down to the
line:
if (tty->stopped)
and breaks.

and when I press ^Q it screams through the tty_io.c line
that goes if c==START_CHAR() and does all its tty->writes().

Well, that all seems right to me, so I can't figure out what
is going on.

That's about all I can tell you. Something somewhere isn't getting
reset, but I can't figure out what it is. Anyone?


------------------------------------------------------------------------
Paul C. Janzen Documentation is the castor oil of
[email protected] programming. Managers know it must be
good because the programmers hate it so
much.
------------------------------------------------------------------------


[next article]
From: [email protected] (Jeffrey Comstock)

Subject: Re: A question on standards
Message-ID: <[email protected]>
Date: 28 Feb 92 03:06:12 GMT
References: <[email protected]>
Organization: Sewer of Source Code
Lines: 13

In article <[email protected]> [email protected] writes:
>I got hold of poeigl package a couple of days ago and installed it. I found
>that it uses inittab/securetty instead of ttytab and ttys. I do not know the
>standards. So, I ask those who know: Which is the standard in init/getty/login
>setup ttytab or inittab? If poeigl is nonstandard I am thinking of

ttytab is standard in bsd.
inittab is standard in sysv.

--
Jeffrey R. Comstock
INET [email protected]
CW -. .-. ----- -..


[next article]
From: [email protected] (Michael Bethune)

Subject: Re: Few comments about Linux
Message-ID: <[email protected]>
Date: 28 Feb 92 12:17:26 GMT
References: <[email protected]> <[email protected] use.co.uk> <[email protected]>
Organization: Pyramid Technology Corp PL, Melb., Australia
Lines: 70

In article <[email protected]> [email protected] (Mark William Hopkins) writes:
>In article <[email protected]> [email protected] (D.J.Wal ker-Morgan) writes:
>>
>>|> A 486-33 outperforms a low end SPARC, and my 386-33 is easily twice
>>|> as fast as my HP Bobcats.
>>
>In computer science, you learn that if you want speed, you write better
>algorithms and use parallelism instead of beating a faster processor against a
>brick wall.

Thank you for reminding me about what my degree taught me. ๐Ÿ˜Ž

[nonsense deleted]

>In any case, it's not speed that counts, it's functionality. The bottom line
>is: what microprocessor has the best low-level hardware support for operating
>systems?
>
>I won't answer the question, because I don't know what a SPARC supports
>(semaphores, memory segments/protection, access levels, caching, or whatever).
>I do know that the 386 and above have ideal OS support, though it tends to
>suffer from having a long heritage dating back to the 8080 (when people were
>still living in caves).

Just what do mean by 'the bottom line is: what micoprocessor has the best
low-level hardware support for operating systems'?

This 'bottom line' makes no sense. What operating systems? This is crucial
to understanding what kind of support is useful. There is a whole lot of
difference between what good in supporting a small address space operating
system like MS-DOS without virtual memory and what is useful to support
say UNIX.

Afterall an 8088 provides all the 'low-level support' MS-DOS needs.

The point being there are various operating systems that make various demands
upon a CPU. More to the point, Operating systems like any program make
demands upon a whole range of sub-systems, CPU, memory, bus traffic, I/O
sub-systems etc etc. Looking at CPU speed/capabilities in isolation is
a favourite past time of the MIP merchants and marketoids, it says very
little about ultimate machine performance for a given application.

This is very relevant to 486s versus Sparc.

To make the point more strongly, what is useful about a
12 SPEC mark CPU, if its sitting idle in wait states waiting for
slow memory sub-systems, or slow I/O sub-systems.

A high performance desktop machine requires balance between its different
components.

In general Intel based PCs tend to be built for a low price point where
performance versus price trade offs are inevitably made.

Consider the ISA bus, it has a through put of not much more than 5 Megabytes
per second. Even the EISA bus is crippled by poor bandwidth to retain
compatability with the ISA bus.

Sun Sparc workstations in contrast tend to balance a capable processor
with good bus performance, good memory sub-systems, mass-storage
peripherals and some of the best ethernet performance in the business.

On the other hand Intel based PC's provide peerless single user performance,
with non-virtual memory operating systems like MS-DOS.

--
Michael Bethune Disclaimer: I have no commercial
Independent Unix Consultant. relationship with Pyramid and
Phone: +61 3 018 538103 do not represent them in any way.
Melbourne, Australia.


[next article]
From: [email protected] (D.Bolla)

Subject: Re: v86 mode is [not] useless
Keywords: v86
Message-ID: <[email protected]>
Date: 28 Feb 92 13:51:14 GMT
References: <[email protected]> <[email protected]> <1992Feb [email protected]>
Reply-To: [email protected] (D.Bolla)
Organization: Computing Lab, University of Kent at Canterbury, UK.
Lines: 45

In article <[email protected]> [email protected] (Wayne Daviso n) writes:
>D.Bolla ([email protected]) wrote:
>> Let's face it. A V86 mode is almost useless.
>> What kind of programs (decent programs) do you see that use an 8086 ?

>You are seriously underestimating the useability of V86 mode. The DOS
>memory manager QEMM uses V86 mode to provide all sorts of loadhigh features
>and memory management services. This broke programs that tried to switch
>into protected mode and do their own memory management, but it was so

Yes the point is that linux is not stable as it is now ( and it is simple
now ). Loading a lot of stuff will make it so unstable to be useless.

Can I remind to you that:
A program that produce the WRONG result fast is useless...

For an OS there is no clear line between a good and bad result but the
same principle apply. If I use linux in a serious way I want it to be
stable !!!!
No point to have a lot of features that panic !

>popular that most programs were re-written to include VCPI support (which
>allows them to work with or without QEMM being present). These days you
>also have to support DPMI so you can run Windows 3.0 without resorting to
>real mode. Look at OS/2 2.0 -- they are providing an OS that uses V86
>dos boxes complete with DPMI (and I believe VCPI) support, so it CAN be
>done. Don't think that it's going to be easy, though.

Is it worth it ? Is it worth it to argue that it can be done ?
Of course it can be done !

BUT

The list of priorities has MANY more things in front of it.
Let's have first:

A STABLE KERNEL (The actual version still hang too often)
A FAST math emulation in the kernel.
A STABLE IPC
A STABLE TCP-IP
A STABLE X11

And then we may think about DOS and other things....

Damiano


[next article]
From: hlu%[email protected]

Subject: FAQ: the new gcc 1.40
Message-ID: <[email protected]>
Date: 28 Feb 92 17:25:33 GMT
Sender: [email protected] (Mr Background)
Reply-To: hlu%[email protected]
Organization: The Internet
Lines: 88

=============================================================
QUESTION: Where is the new gcc 1.40?

ANSWER: It's on tsx-1.mit.edu under /pub/linux/binaries/compilers. It's
called newgcc.tar.Z

QUESTION: What is the content of newgcc.tar.Z?

ANSWER: newgcc.tar.Z contains all the binaries, header files and
libraries needed to use gcc-1.40.

QUESTION: How do I install it?

ANSWER: Just do

cd /usr
tar xvofvz xxxx/newgcc.tar.Z

Note: Some header files will be overwritten by the ones in
newgcc.tar.Z. If you rely on some changed header files. Then do

cd /usr/local
tar xvofvz xxxx/newgcc.tar.Z

I believe gcc will look for /usr/local first for anything. You may want
to either change your Makefiles or link gar/ar gas/as gld/ld to prevent
a lot off /stupid/ problems. But since gcc is the default compiler for
Linux, the prefix 'g' should be dropped.

QUESTION: What are the main differences with the old release?

ANSWER: Those bugs related to estdio in the old port have been
corrected, like FP output and "compress *." And this package contains
387 support. There are
libm.a (for those with 387)
libsoft.a (for those without)

libtermcap.a (from tput 1.10) is separated from libc.a. The
-mstring-insns option is no longer needed nor supported.

QUESTION: I find FP to integer is not right. What is wrong?

ANSWER: There is a bug in gnulib. That should have been fixed. Get a
new one from tsx-11.mit.edu. If it still doesn't work right, drop me a
note.

QUESTION: Is stdio ANSI compatible?

ANSWER: No. Read the stdio.h in newgcc.tar.Z. It will be fixed in gcc
2.0. The stdio will be replaced by an ANSI compatible one based on BSD
4.4. You should read the stdio.h in newgcc.tar.Z to avoid the trouble
for now.

QUESTION: When will gcc 2.0 be released?

ANSWER: Very soon.

QUESTION: Why does -O fail on some files?

ANSWER: The -O option may fail when the INLINE functions (string.h) are
passed as parameters. There is nothing wrong with compiler(?). They
will be fixed in gcc 2.0. You can change the source to avoid that.

QUESTION: Where is the source code of the new libc.a?

ANSWER: Drop me a note. The only difference is stdio. Since it will be
replaced in gcc 2.0, I suggest you wait for 2.0.

QUESTION: Why cannot gcc exec cpp and cc1?

ANSWER: Someone suggests me to put cc1 and cpp in /usr/local/lib. I
changed the driver. But I forgot to move cc1 and cpp. It will be solved
in gcc 2.0. For now, just do
ln -s /usr/lib/cpp /usr/local/lib
ln -s /usr/lib/cc1 /usr/local/lib

QUESTION: Where is limits.h?

ANSWER: You may need to get the lastest header files, aka the ones from
the kernel sources.
=============================================================

H.J.
--
School of EECS Internet: [email protected]
Washington State University BITNET: [email protected]
Pullman, WA 99164 Phone: (509) 335-6470 (O)
USA (509) 334-6315 (H)


[next article]
From: [email protected] (Jiansheng Zhao)

Subject: gcc 2.0
Message-ID: <[email protected]>
Date: 28 Feb 92 17:14:33 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 82


Xref: unixg.ubc.ca ubc.unix:68 bc.unix:23 bc.general:292 ubc.general:198
Newsgroups: ubc.unix,bc.unix,bc.general,ubc.general
Path: unixg.ubc.ca!ubc-cs!acton
From: [email protected] (Donald Acton)
Subject: gcc-2.0 available for anonymous ftp from cs.ubc.ca
Message-ID: <[email protected]>
Sender: [email protected] (Usenet News)
Organization: Computer Science, University of B.C., Vancouver, B.C., Canada
Distribution: na
Date: Sun, 23 Feb 92 21:11:28 GMT

gcc-2.0 is available for anonymous ftp from cs.ubc.ca and can be found
in /pub/archive/gnu/gcc-2.0.tar.Z. Postscript versions of the gcc and
cpp manuals built from the texinfo files shipped with this release are
available as /pub/archive/gnu/manuals_ps/{gcc,cpp}-2.0.Z

Note that gcc seems to be the do everything C compiler as it supports
ANSI C, Objective C and C++ although it sounds like the Objective C
support might be incomplete. Poking around in the distribution I
noticed a makefile for MS-DOS.

Donald Acton


------The release announcement follows-------------------------------
~From: [email protected] (Richard Stallman)
~Newsgroups: gnu.gcc.help,gnu.gcc.announce,gnu.g++.announce
~Subject: GCC 2.0 released
~Date: 22 Feb 92 08:54:57 GMT

People have been waiting for GCC 2.0 for a year. Now it's
available for anonymous ftp from /pub/gnu/gcc-2.0.tar.Z
on prep.ai.mit.edu. There are no diffs from version 1--such
diffs would be too large to be useful.


Version 2 of GCC can generate code for the IBM PC/RT, the IBM RS/6000,
the Motorola 88000, the Acorn RISC machine (not fully tested), the AMD
29000 and the HP-PA (700 or 800), in addition to several machines
supported by version 1 (Motorola 68000, Vax, Sparc, National
Semiconductor 32000, Intel 386 and 860, and Mips). Ports for the IBM
370, the Intel 960, the Clipper, the Tron (a Japanese standard
computer architecture) and the NCUBE are on their way, but there is no
estimate of when they will be available. Note that using GCC to
compile for the HP-PA requires as yet unreleased versions of other GNU
software such as the assembler and linker.

Version 2 can generate output files in a.out, COFF, ECOFF, ELF, XCOFF,
VAX-VMS and OSF-Rose formats when used with a suitable assembler. It
can produce debugging information in several formats: BSD stabs, COFF,
ECOFF, ECOFF with stabs symbols, VAX-VMS and DWARF. (We may support
XCOFF for the RS/6000 in the future.)

Version 2 can be easily configured as a cross-compiler, running on one
platform while generating code for another.

Version 2 supports compatible calling conventions for function calling
and return values on the Sparc (unlike version 1) as well as the other
machine types.

Early testing of GCC Version 2.0 indicates that it produces faster
code for SPARC computers than Sun's latest released compilers (both
bundled and unbundled). It is also the fastest known compiler for
the Motorola 88k.


In addition to ANSI C, GCC Version 2.0 includes support for the C++
and Objective C languages. Objective C is an object-oriented language
which adds to C features similar to Smalltalk. The front end for the
Objective C language was donated by NeXT Computers, Inc., which uses
GCC as the basis for their NeXTstep operating system. (Run-time
support for the Objective C language is still under development.)

GCC extends the C language to support nested functions, non-local
gotos, taking the address of program labels, and unnamed structures as
function arguments (among other things). There are also many new
warnings for frequent programming mistakes.

GCC Version 2 can produce position-independent code for several types
of CPU: 68000, 88000, 80386, Sparc, and RS/6000. Supporting PIC on
additional suitable CPU types is not too difficult a task.


[next article]
From: hlu%[email protected]

Subject: Re: newgcc installation error (fwd)
Message-ID: <[email protected]>
Date: 28 Feb 92 17:25:33 GMT
Sender: [email protected] (Mr Background)
Reply-To: hlu%[email protected]
Organization: The Internet
Lines: 46

Forwarded message:
| From @nic.funet.fi:[email protected] Fri Feb 28 09:12 PST 1992
| X-Delivery-Notice: SMTP MAIL FROM does not correspond to sender.
| Date: Fri, 28 Feb 1992 19:02:09 +0200
| From: tim
| Subject: Re: newgcc installation error
| Sender: [email protected]
| To: [email protected]
| Cc: [email protected]
| Message-Id: <[email protected]>
| In-Reply-To: <[email protected]>; from "Marc CORSINI" a t
| Feb 28, 92 5:01 pm
| X-Mailer: ELM [version 2.3 PL11]
| X-Envelope-To: [email protected]
|
| >
| >
| >
| > May be the cc1 and cpp could be link in /usr/bin (I think this is the
| > only thing I've done which was not in the readme)

NO. NO. Someone suggests me to put cc1 and cpp in /usr/local/lib. I
changed the driver. But I forgot to move cc1 and cpp. It will be solved
in gcc 2.0.

| >
| yes! I made sym-links from /usr/lib/cc1 and /usr/lib/cpp to
| /usr/bin/cc1 and /usr/bin/cpp and the problem was solved.
|


| there doesn't seem to be an limits.h in /usr/include...i commented the
| line #include out of stdio.h and was able to compile some
| stuff successfully. i'll worry about getting a limits.h later.
|
|

You may need to get the lastest header files, aka the ones from the
kernel sources.

H.J.
--
School of EECS Internet: [email protected]
Washington State University BITNET: [email protected]
Pullman, WA 99164 Phone: (509) 335-6470 (O)
USA (509) 334-6315 (H)


[next article]
From: [email protected] (cm445a17)

Subject: Re: Few comments about Linux
Message-ID: <[email protected]>
Date: 28 Feb 92 21:11:09 GMT
References: <[email protected]> <[email protected] use.co.uk> <[email protected]> <[email protected]>
Organization: University of Maryland University College
Lines: 76

[email protected] (Michael Bethune) writes:

>In article <[email protected]> [email protected] (Mark Willia m Hopkins) writes:
>>In article <[email protected]> [email protected] (D.J.Wa lker-Morgan) writes:
>>>
>>>|> A 486-33 outperforms a low end SPARC, and my 386-33 is easily twice
>>>|> as fast as my HP Bobcats.
>>>
>>In computer science, you learn that if you want speed, you write better
>>algorithms and use parallelism instead of beating a faster processor against a
>>brick wall.

>Thank you for reminding me about what my degree taught me. ๐Ÿ˜Ž

>[nonsense deleted]

>>In any case, it's not speed that counts, it's functionality. The bottom line
>>is: what microprocessor has the best low-level hardware support for operating
>>systems?
>>
>>I won't answer the question, because I don't know what a SPARC supports
>>(semaphores, memory segments/protection, access levels, caching, or whatever).
>>I do know that the 386 and above have ideal OS support, though it tends to
>>suffer from having a long heritage dating back to the 8080 (when people were
>>still living in caves).

>Just what do mean by 'the bottom line is: what micoprocessor has the best
>low-level hardware support for operating systems'?

>This 'bottom line' makes no sense. What operating systems? This is crucial
>to understanding what kind of support is useful. There is a whole lot of
>difference between what good in supporting a small address space operating
>system like MS-DOS without virtual memory and what is useful to support
>say UNIX.

>Afterall an 8088 provides all the 'low-level support' MS-DOS needs.

>The point being there are various operating systems that make various demands
>upon a CPU. More to the point, Operating systems like any program make
>demands upon a whole range of sub-systems, CPU, memory, bus traffic, I/O
>sub-systems etc etc. Looking at CPU speed/capabilities in isolation is
>a favourite past time of the MIP merchants and marketoids, it says very
>little about ultimate machine performance for a given application.

>This is very relevant to 486s versus Sparc.

>To make the point more strongly, what is useful about a
>12 SPEC mark CPU, if its sitting idle in wait states waiting for
>slow memory sub-systems, or slow I/O sub-systems.

>A high performance desktop machine requires balance between its different
>components.

>In general Intel based PCs tend to be built for a low price point where
>performance versus price trade offs are inevitably made.

>Consider the ISA bus, it has a through put of not much more than 5 Megabytes
>per second. Even the EISA bus is crippled by poor bandwidth to retain
>compatability with the ISA bus.


>Sun Sparc workstations in contrast tend to balance a capable processor
>with good bus performance, good memory sub-systems, mass-storage
>peripherals and some of the best ethernet performance in the business.

>On the other hand Intel based PC's provide peerless single user performance,
>with non-virtual memory operating systems like MS-DOS.

>--
>Michael Bethune Disclaimer: I have no commercial
>Independent Unix Consultant. relationship with Pyramid and
>Phone: +61 3 018 538103 do not represent them in any way.
>Melbourne, Australia.

Please, this is a news group about Linux. It is NOT a platforms nor
an operating systems discussion group. Take your topics to were they
belong (i.e. another newsgroup). Thank you!


[next article]
From: [email protected] (Wayne Davison)

Subject: Re: v86 mode is [not] useless
Keywords: v86, ramblings
Message-ID: <[email protected]>
Date: 28 Feb 92 21:51:44 GMT
References: <[email protected]> <[email protected]> <[email protected] lcon.ukc.ac.uk>
Organization: Borland International
Lines: 45

D.Bolla ([email protected]) wrote:
> Let's face it. A V86 mode is almost useless.

I said:
> You are seriously underestimating the useability of V86 mode.

D.Bolla responds:
> Yes the point is that linux is not stable as it is now ( and it is simple
> now ). Loading a lot of stuff will make it so unstable to be useless.

If you notice I said nothing about when or if a V86 mode should be done. Only
that an 8086 mode running under a [34]86 OS is much more powerful than simply
an 8086 chip. _My_ point was to correct an inaccuracy, and I left the other
points of your article untouched. But as long as I'm here, let me address
_your_ point.

I agree that the stability of Linux should be the highest goal, along with
providing a large enough set of the basic kernel services to make it useful
(things like the IPC and VFS efforts). However, I disagree that people
talking about and maybe even beginning to develop other new-fangled features
is detrimental to the product. Keep in mind that the public community of
Linux hackers is quite large and diverse. There are already people focusing
on riding the kernel of bugs and providing the enhancements necessary to
run things like X windows. If this effort is under-staffed, then by all
means let this be known so that someone else can help out.

Just don't think that throwing more people at a task will necessarily speed
it up. The rule is to choose a few good people (or just one) to target a
specific area that needs improvement and allow that group (person) to focus
on it. Having someone working on an improved memory manager that would
support a V86 mode is a great idea. When and if it gets included into the
mainstream distribution is then an integration decision that can happen after
it's done. I argue that having these side-line development efforts is not
detracting from the central goal of a version 1.0 Linux in any way as long
as we have a core group of people who are already working on the basics.

Remember that Linus has a very good track record of not throwing in just any
new feature that comes along. And if Linus passes the baton on to another
to succeed him in this job, he or she will have to be equally conservative
in only adding the things to the mainstream distribution that are well thought
out and well implemented.
--
\ /| / /|\/ /| /(_) Wayne Davison
(_)/ |/ /\|/ / |/ \ [email protected]
(W A Y N e)


[next article]
From: [email protected] (Ian Wells)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Message-ID:
Date: 28 Feb 92 15:59:01 GMT
References: <[email protected]>
<[email protected]> <[email protected]>
<[email protected]>
Sender: [email protected] (The news facility)
Organization: CompScis Anonymous, Cambridge branch
Lines: 29
In-reply-to: [email protected]'s message of 27 Feb 92 11:49:32 GMT

In article <[email protected]> [email protected] (Joel M. Hoffma n) writes:

[ ... comments about nice DOS programs ... ]

WordPerfect and Quicken come immediately to mind. While I don't
actually use WP and more, I know many people that do, and I'd like to be
able to use Linux without giving up such nice DOS programs. Most
programs still run under a simple 8086. I also have several games for
DOS that I simply cannot port to Linux. (Scrabble, Chess, etc.)

Yes, so use DOS for them. It would make Linux unnecessariy complicated to
add DOS. It's developing as a nice operating system in its own right, forcing
support for another OS onto it is taking a step backward. If you want a good
WP, game, etc. write your own for Linux or keep DOS on your disk.

>
>Do you want DOS to hang in the middle of you wordprocessor ?
>I don't
>

If the DOS emulator dies it's not nearly to critical as if the OS
dies. Under Linux, everything else will keep running. Not SO bad....

-Joel

One point about this is that if DOS trickery has to be coded into the MM then
it's a lot more likely to hang Linux at the same time. It's also a lot more
likely that Linux will hang given the amount of bugs adding such code would
produce.


[next article]
From: [email protected] (Peter Orbaek)

Subject: Re: startup shell?
Message-ID: <[email protected]>
Date: 28 Feb 92 16:57:20 GMT
References: <[email protected]>
Sender: [email protected] (Peter Orbaek)
Organization: DAIMI: Computer Science Department, Aarhus University, Denmark
Lines: 24

[email protected] (Paul Richards) writes:

>I've created a /home/root directory and modified /etc/passwd accordingly
>but it doesn't work. It's just occured to me that /etc/passwd isn't even
>used since I don't have the login stuff installed yet.

>What happens when I start up. The home directory is always set to
>/usr/root and a set of environment variables appear from somewhere.

The init/main.c code in the kernel sets up that home directory, and sets the
TERM and HOME environments too. The other environmentvariables are set up
by the shell.

>What is setting things up this way and will I have to get the login
>stuff working to change it.

Yes. You will have to use the login stuff to change root's homedir.

- Peter.
--
Peter Orbaek ----------------- [email protected] | ///
Hasle Ringvej 122, DK-8200 Aarhus N, DENMARK | ///
| \\\///
"Strong typing is for people with weak memories" | \XX/


[next article]
From: [email protected] (Peter Orbaek)

Subject: Re: trouble dialing in with poe-IGL
Message-ID: <[email protected]>
Date: 28 Feb 92 16:53:18 GMT
References: <[email protected]>
Sender: [email protected] (Peter Orbaek)
Organization: DAIMI: Computer Science Department, Aarhus University, Denmark
Lines: 32

[email protected] (Mark Saltzman) writes:

>When I try to dial in to poe-IGL using a hayes compatible 2400 baud
>modem i get stuff like this:
>(none) login: password: Login incorrect
>repeated a few times, without giving me the chance to respond to the
>login: prompt. What options do I need to give agetty in inittab? What
>options should my terminal emulator use (XON/XOFF, CTS/RTS, etc...).

1. Be sure that you run version 1.1 and not 1.0. 1.0 had some problems with
the ttybuffers being full of garbage.

2. Did you turn the modem into silent mode with an "ATE" command? If
not then getty is confused by the "CONNECT" message.

3. Did you use the -m flag to getty? This is recommended.

Set your terminal-emulator to use xon/xoff. Getty intelligently adjusts for
parity and such.

>Also, I have not yet dared to modify my kernel in order to have it run
>init. When I want to use it, I just type init & from the shell. Is
>this ok?

Yes. It's not ideal but it should work.

- Peter.
--
Peter Orbaek ----------------- [email protected] | ///
Hasle Ringvej 122, DK-8200 Aarhus N, DENMARK | ///
| \\\///
"Strong typing is for people with weak memories" | \XX/


[next article]
From: [email protected] (D.Bolla)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Keywords: v86, video
Message-ID: <[email protected]>
Date: 28 Feb 92 18:50:44 GMT
References: <[email protected]> <[email protected] du>
Reply-To: [email protected] (D.Bolla)
Organization: Computing Lab, University of Kent at Canterbury, UK.
Lines: 40

In article <[email protected]> [email protected] (Joel M. Hoffman ) writes:
>In article <[email protected]> [email protected] (Drew Eckhardt) writes:
>>We want X11 up on Linux. Since not everybody has an ATI or Tseng chipset, we
>>also want to support the other cards out there - Trident 8800 / 8900,
>>Genoa, etc.
>>Now, the only real differences between these chips is in setup (mode
>>selection), and bank switching (to accomodate the 128K we have for video
>>buffer and 1M of modern SVGA cards). Setup is the difficult part, with over

As somebody else wrote the switching can be easly done without touching
the BIOS or the DOS.

>I think there are *many* good reasons to rewrite mm. First of all, we
>(at least I) certainly want a DOS emulator down the road. The easiest

I think that everything is easy when you look using abstract concepts.
Things get very BAD when you try to implement them.
The idea is that things are NOT so simple as they appear !
Let's leave DOS out... Please !
BTW If you want DOS, WINDOWS, MINIX and all the rest buy OS/2 .
It is supposed to be the answer to this kind of problems.....

>way to do this is to run the BIOS, and then DOS, in V86 mode, I think.
>Also, I think a more general page table scheme would be nice on its
>own merits. (I can see having more than 64 tasks, e.g.)
Yes of course it is nice...
It will be nice to have a CRAY instead !
The 386 is NOT a cray. 64 tasks with 64 mb each is a very reasonable limit
If you need much more then you probably need another machine too.
I am not sayng that linux should not be improved I am sayng that
there are others thing FIRST !

>The other option, of course, is to copy the BIOS into a process's
>memory (in protected mode), and try to make use of it. I don't know
>how difficult this would be.
Why don't you try and have fun ?
As long as things are NOT in the kernel the worse that should happen
is that your process dies.. ๐Ÿ™‚

Damiano


[next article]
From: [email protected] (Linus Benedict Torvalds)

Subject: Re: Floating point performance
Keywords: libsoft.a, math, 80387.
Message-ID: <[email protected]>
Date: 28 Feb 92 09:20:37 GMT
References: <[email protected]> <[email protected]> <199 [email protected]>
Organization: University of Helsinki
Lines: 42

In article <[email protected]> [email protected] (Humberto Ortiz-Zuazaga) writes:
>
>I think he was using H.J.'s newlibc, not the kernel stuff.

Yes, but even that one uses the kernel for addition and multiplications,
I think (unless it uses some all-integer cordic algorithm, which I
doubt). Thus the math-emulator in the kernel still gets called (a lot)
when doing things like sin etc: this is /extremely/ slow, and is the
reason floating point performance was better in the older versions. The
math-emulator has to calculate effective addresses etc for every add or
mul: much slower than just a simple function call.

Some people may wonder why use an emulator at all, if it's slower than
having function calls. There are reasons: never fear. One of the reasons
is that gcc for the 386 assumes at least minimum float-emulation, unless
you want to patch every new version of gcc that comes out. An equally
pressing reason is that with an emulator, the /same/ binary will work on
both a 387-machine and on one without any coprocessor: no need for the
soft-float libraries (well, right now it's still needed for sin/sqrt
etc, but that will change eventually (not in 0.13)).

>Actually, I have a 20MHz SX at home without a coprocessor, and a 16MHz
>SX with a coprocessor at work. I compiled a program with libsoft and
>libm, and tested the libsoft compiled program at home. 50 iterations in
>3 minutes (it does many ldexp's). At work, off of a floppy disk, the
>same program, 50 iterations in less than a second.

Probably an extreme example, but yes, it pays to have a coprocessor.
Most (well, many anyway) people probably do, and those that haven't got
one probably never wanted to calculate a lot in the first place.

>Interestingly, I can't tell the difference between the libsoft and libm
>linked programs on the machine with the coprocessor. They both run just
>as fast. Do libsoft linked programs use the coprocessor when present?

libsoft uses the coprocessor when present for most things: all the
things that are present in the kernel (additions, loads etc). Libsoft
is used only for the more complex functions, which don't have kernel
support yet (and even they get /much/ faster with a coprocessor: they
are built up from adds/muls).

Linus


[next article]
From: [email protected]

Subject: Undefined Symbol: __stderr in text segment
Message-ID: <[email protected]>
Date: 28 Feb 92 23:49:44 GMT
Organization: Memphis State University
Lines: 6

What in the world is a Undefined symbol: __stderr in text segment???? No
matter how many different versions of GCC and no matter how many different
ways I go about installing it, I can *NEVER* get a damned thing compiled
with out error coming up somewhere. It usually happens when the compiler
is using something link ld /usr/lib/crt.o /usr/lib/gnulib. Anyone care
to mention my simply mistake???


[next article]
From: [email protected] (Harry G. Varnis)

Subject: mkswap sources
Message-ID: <[email protected]>
Date: 28 Feb 92 17:36:31 GMT
Sender: [email protected]
Organization: Network Systems Corporation
Lines: 9
Originator: [email protected]
Nntp-Posting-Host: herring

Would someone post source for mkswap (or a pointer if it's archived),
please?

I'm trying to understand the size constraints that apparently
exist with _swapon_ and friends. Any enlightenment appreciated.

Regards,
--
Harry Varnis (612) 493-1042


[next article]
From: [email protected] (Robert Mudry)

Subject: Linux for the Amiga
Keywords: amiga
Message-ID:
Date: 29 Feb 92 04:26:02 GMT
Organization: Netcom - Online Communication Services (408 241-9760 guest)
Lines: 11

Just out of curiousity, is there any hope of linux being ported over to,
let us say, an Amiga? ๐Ÿ˜‰ Any info would be great! Lie to me if you must.. ๐Ÿ˜‰

--
[email protected]

"In general, it is safe and legal to kill your children and their children"
-POSIX Prg Gt, by Donald Lewine, O'Reilly & Associates, 1991, p.110 (On
process termination)

glurp.


[next article]
From: [email protected] (Kevin Cummings)

Subject: mount can't open lock file?
Message-ID: <[email protected]>
Date: 28 Feb 92 15:56:12 GMT
Sender: [email protected] (Kevin Cummings)
Organization: Prime Computer R&D
Lines: 33

Well, last night I bit the bullet, and repartitioned my 80MB drive 1
from 3 26MB SSTOR partitions, into 2 32MBs, an 8MB, and a 4MB.

I rebooted LINUX (which is on my 70MB drive 0), and tried to mount
the new 8MB partition under /files. I ran fdisk to get the block
sizes (under LINUX), and ran mkfs on the 8 MB partition (/dev/hd7),
and mkswap on the 4 MB partition (/dev/hd6). when I try and mount
/dev/hd7 /files, mount complains about not being able to open a
lock-file. Of course it doesn't say WHAT lock file, or WHERE,
or WHY it can't open it! As a control test, I tried to mount
a floppy filesystem (empty) that I had made when I installed
LINUX on /dev/PS1, and got the same error message. fsck can find
no errors on the devices I'm trying to mount. Can anyone explain
what is wrong? Do I need to boot from floppy again, and fsck my
root partition?

Also, anyone know where I can find a swapon binary so I can
use the 4MB of swap space that mkswap says I created?

Thanx.

=================================================================
Kevin J. Cummings Prime Computer Inc.
20 Briarwood Road 500 Old Connecticut Path
Framingham, Mass. Framingham, Mass.

InterNet: [email protected]
UUCP: uunet!primerd.Prime.COM!cummings

Std. Disclaimer: "Mr. McKittrick, after careful consideration,
I've come to the conclusion that your new
defense system SUCKS..." -- War Games
=================================================================


[next article]
From: [email protected] (Heien-Kun Chiang)

Subject: Used copy of PC Minix wanted.
Message-ID: <[email protected]>
Date: 29 Feb 92 06:40:29 GMT
Sender: [email protected] (News)
Organization: University of Illinois at Urbana
Lines: 16

Hi there,

I am a newcomer in OS and just begins to learn how does it work. I have
followed both comp.os.minix and alt.os.linux for a while, and have installed
linux.012 successfully in my 386 machine and quite happy with it.

Since I am a beginner, linux might be a little bit difficult to me. I
think PC Minix might be more suitable for me at this moment.
If anyone of you no longer needs (or uses) PC Minix and am willing to sell it,
please send me an e-mail.

[email protected]
or
[email protected]

Thanks a lot.


[next article]
From: [email protected] (Linus Benedict Torvalds)

Subject: Re: Xon / xoff bug?
Message-ID: <[email protected]>
Date: 29 Feb 92 09:42:09 GMT
References: <[email protected]emx.uucp>
Organization: University of Helsinki
Lines: 27

In article <[email protected]> [email protected] (Paul C. Janzen) write s:
>I am used to using ^S / ^Q for "flow control" on a terminal; that is,
>if something is scrolling by, and I want to see it, my fingers hit
>the ^S key to stop it. Well, this didn't work at first in Linux, so
>I got stty and did a
> stty ixon
>which should have done the trick, I thought.
>
>So I typed
> cat huge_file
>and about two pages into it I hit ^S. Great! It stopped! ^Q -restarted
>it. Ohmighod, I thought I had it made.
>
>BZZZZT. Thank you for playing.
> [ deleted ]

Yes, XON/XOFF isn't quite ready in 0.12 - this in one more of the fixes
for 0.13. The problem is simply that ^Q doesn't correctly wake up the
writing process, so only the buffer already written will be output, and
then the process just sleeps away... Nothing bad happens, it's just not
noticing that it can start writing again.

In my version (and thus 0.13) this is corrected, and ^S/^Q work all
right. There are some other rewrites in there too: tty-io has been
cleaned up a bit.

Linus


[next article]
From: [email protected] (Charles Carlson)

Subject: pcomm: has anyone ported it?
Message-ID: <[email protected]>
Date: 29 Feb 92 03:48:26 GMT
Sender: [email protected]
Organization: University of Manitoba, Winnipeg, Canada
Lines: 4
Nntp-Posting-Host: gibson.cc.umanitoba.ca

I'm wanting to get Pcomm working with Linux..
Has anyone managed it?

Charles


[next article]
From: [email protected]

Subject: Announcement: poeigl 1.2 (init/getty/login)
Keywords: init login getty write getlogin wall users mesg
Message-ID: <[email protected]>
Date: 29 Feb 92 13:16:26 GMT
Sender: [email protected]
Organization: Acadia University
Lines: 47
To: [email protected]

I am happy to announce version 1.2 of the poe-IGL package. Once again it should
fix all the reported bugs and defiencies of version 1.1.

Most notably:

* The TERM environment variable is now carried through from the kernel all the
way to the shell. The termcap field in /etc/inittab is thereby overridden
by what you choose at boot time. The setting in inittab is still in effect
for non-console tty's such as modem ports.

* Login now correctly handles passwords longer than 8 characters (ie. it does
not send the superflous chars to the shell)

* The ERASE char is now set properly by getty.

* Init now handles the HUP and TSTP signals properly.

* Two new commands have been added: users and mesg.
(Both are small shell scripts, and "users" requires that you have "awk").

The package still includes: init, getty, login, wall, write, UFC-crypt, etc.

In order to get the TERM setting from the boot-time selection an other patch
to init/main.c is necessary. A context-diff against the ORIGINAL 0.12 kernel
is included.

I'm afraid that this release wont make it into 0.13 | 0.95. Oh well.

You can ftp the package from

ftp.daimi.aau.dk (130.225.16.27)

in the file

pub/Linux-source/poeigl-1.2.tar.Z

I also uploaded it to

nic.funet.fi

in the

pub/OS/Linux/incoming

directory.

- Peter ([email protected])


[next article]
From: [email protected] (Mark Saltzman)

Subject: Fix for make problem with setuid?
Message-ID: <[email protected]>
Date: 29 Feb 92 18:19:41 GMT
Sender: [email protected]
Reply-To: [email protected] (Mark Saltzman)
Organization: Case Western Reserve University, Cleveland, OH (USA)
Lines: 5
Nntp-Posting-Host: cwns1.ins.cwru.edu


Has anyone found a fix for problem with using make under a login other
than root?
-mark


[next article]
From: [email protected] (Mark Lanett)
Newsgroups: comp.unix.bsd,comp.unix.sysv386,alt.os.linux

Subject: Re: Networking 2 release, BSDI, etc
Message-ID: <[email protected]>
Date: 29 Feb 92 20:42:46 GMT
References: <[email protected]>
Sender: [email protected]
Organization: University of Illinois at Urbana-Champaign
Lines: 10

[email protected] (Jim Lick) writes:

>The other future development is the Gnu OS based on their 'Hurd'
>kernel. No telling when this will be out though.

Actually, the "GNU OS" will be a gnu filesystem (HURD) running on Mach
(the unix successor being developed by CMU).

The hurd's mailing list is [email protected]; the Main Man is
[email protected] (Michael I Bushnell).


[next article]
From: [email protected] (Charles Hedrick)

Subject: problem with termcap entry
Message-ID:
Date: 29 Feb 92 20:33:09 GMT
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 4

Now and then I've been getting the cursor in the wrong place. I
believe this is because "do" is defined as linefeed, and linefeed also
does a CR. I suggest changing the "do" property from linefeed to
"\E[B".


[next article]
From: [email protected] (Charles Hedrick)

Subject: initial tty speed
Message-ID:
Date: 29 Feb 92 20:53:12 GMT
Organization: Rutgers Univ., New Brunswick, N.J.
Lines: 3

Linus: in 0.13, I suggest making the default speed of the console
tty's be 38400, rather than 0. With a speed of 0, emacs thinks you're
on a very slow modem, and goes into slow search mode.


[next article]
From: [email protected] (Adam Goldberg)

Subject: Re: v86 mode is [not] useless
Keywords: v86
Message-ID:
Date: 1 Mar 92 00:35:58 GMT
References: <[email protected]> <[email protected] du> <[email protected]> <[email protected]>
Sender: [email protected] (USENET News System)
Organization: Iowa State University, Ames IA
Lines: 24
Originator: [email protected]

In <[email protected]> [email protected] (Wayne Davison) write s:

>real mode. Look at OS/2 2.0 -- they are providing an OS that uses V86
>dos boxes complete with DPMI (and I believe VCPI) support, so it CAN be
>done. Don't think that it's going to be easy, though.

Not that it makes any difference (to alt.os.linux people, anyway), but
OS/2 2.0 will provide DPMI, but _not_ VCPI. DPMI does not allow the
client program to determine which protection ring to run in, VCPI on
the other hand, DOES allow the client to determine which ring to
run in.

Windows, on the other hand, is going to be a little nasty to deal with.
It'll probably be possible to run it in 'real' mode, but not 'standard'
or '386'. Not that it matters, anyway...X is better.

--
Adam Goldberg ! Ask me how IOWA STATE UNIVERSITY is
[email protected] ! censoring my usenet access!!
..!uunet!iastate.edu!goldberg ! >>I don't need a BIG BROTHER!!<<
--
Adam Goldberg ! Ask me how IOWA STATE UNIVERSITY is
[email protected] ! censoring my usenet access!!
..!uunet!iastate.edu!goldberg ! >>I don't need a BIG BROTHER!!<<


[next article]
From: [email protected]

Subject: poe-IGL troubles
Message-ID: <[email protected]>
Date: 1 Mar 92 00:47:26 GMT
Organization: Memphis State University
Lines: 3

How do you get agetty in poe-IGL to answer a tty hooked up to a modem????
is /dev/tty65 com2??? Also, the WHO program doesn't seem to like to look
up users in the /etc/utmp file and just to display a NIL listing.


[next article]
From: [email protected] (Michael Campbell)

Subject: Re: Linux-Activists Digest #74
Message-ID: <[email protected]>
Date: 1 Mar 92 01:15:14 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected] (Michael Campbell)
Organization: The Internet
Lines: 1

Is anyone working on SCSI drivers?


[next article]
From: [email protected] (Bill Henning)

Subject: Re: Easy video mode changes, nonstandard disk support, and v86 mode
Keywords: v86, video
Message-ID: <[email protected]>
Date: 26 Feb 92 04:18:14 GMT
References: <[email protected]> <[email protected]> <1992F [email protected]>
Organization: BC News and Mail
Lines: 7

Actually we should start discussing the level of kernal support for graphics.
I'd prefer a loadable library or device driver approach, with a standardized
call interface. Incidentally, as far as I know 16 bit SVGA's based on TSENG
4000's are capable of being mapped to a 1Mb linear memory segment in the
first 16Mb of the address space, as well as 16 bit access to their video
memory. Now if I could just find the information for doing so, and the code
to initialize various graphics modes (I am particularly interested in 800x600
at 256 colors) I would start writing pieces of a graphics library...


[next article]
From: [email protected] (Drew Eckhardt)

Subject: SCSI support
Summary: It works!!!!!
Message-ID: <[email protected]>
Date: 1 Mar 92 03:31:58 GMT
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 7
Nntp-Posting-Host: caesar.cs.colorado.edu

Good news! I've got SCSI support (my low level driver is for the Seagate
ST01 / ST02, johnsonm is working on a DTC low level driver, and tthorn on
an adaptec) up and running.

I will upload the code after it's been cleaned up and commented a little
bit better - those hardy souls wanting something sooner (or a Seagate ST0x
bootable image) please send E-mail.


[next article]



[next article]
From: [email protected] (Brian Mcbee)

Subject: Re: SCSI support
Message-ID: <[email protected]>
Date: 1 Mar 92 07:57:48 GMT
References: <[email protected]>
Sender: [email protected]
Organization: Oregon State University, CS Dept.
Lines: 17
Nntp-Posting-Host: jacobs.cs.orst.edu

In article <[email protected]> [email protected] (Drew Eckha rdt) writes:
>Good news! I've got SCSI support (my low level driver is for the Seagate
>ST01 / ST02, johnsonm is working on a DTC low level driver, and tthorn on
>an adaptec) up and running.
>
>I will upload the code after it's been cleaned up and commented a little
>bit better - those hardy souls wanting something sooner (or a Seagate ST0x
>bootable image) please send E-mail.

I take it that this is a replacement for the hd driver currently in the kernel.
How difficult will it be to hack it so I can have both controllers?

--
Brian McBee [email protected] or [email protected]
"The strongest reason for the people to retain the right to keep and bear
arms is, as a last resort, to protect themselves against tyranny in
government." Thomas Jefferson


[next article]
From: [email protected]

Subject: swapon hint
Message-ID: <[email protected]>
Date: 1 Mar 92 16:06:08 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 8

In order to get swapon not to report 99 error, must make linux with
SWAP_DEV=NONE (NONE in all capitals). /etc/rc then should contain line:
swapon /dev/hdx where x = [12346789] BTW: yatzee didnot hang with
swapon (Haven't try it enough to know if swapon effective --- didn't
expect a simple curses game to swap on 8MB system).

BTW: tar -xv -b 1 -f filename.tar WORKS even when uncompress junk
appended to file. (default for GNU tar is N=20 blocks of 512). John


[next article]
From: [email protected] (Marc CORSINI)

Subject: Various problems and some solutions
Message-ID: <[email protected]>
Date: 1 Mar 92 22:48:26 GMT
Sender: [email protected] (Mr Background)
Reply-To: [email protected]
Organization: The Internet
Lines: 39



hi,
Today I've applyed different patches on the kernel and features which arise in
various problem
1) my env is:
386Sx 4Mo VC, gcc-1.4, gnuemacs, uemacs, swapon and a 2Megs swap-file
2) the patches and features I try to add
mmap, dev/kmem, Peter init-1.2

The problems: (except the classical patches applying)
a) after the mmap patches I try to run the ega it works for the strips
but hangs during the dots test
Moreover the blank feature of VC won't work anymore (funny remark is
that when working on dev/tty2 , the blank put the screen to "swap" to
dev/tty1; and when working on dev/tty1 *only* half screen blanks)
b) after dev/kmem of Diamano
gnuemacs cannot anymore save files the error message is
IO error writing : bad file nr
[the only change i made was the -NR-reboot value]
c) for the Peter init-1.2, the package won't compile with the new
include (stdio.h is no longer STDC) and the link fails for stderr
to solve it I use the old include files and link with the library of
gcc-1.37 and all works as for the previous release

I think that Peter should try to adapt his package / or some kind soul (and
C guru) could make include and library more uniform (I can't do that since I
don't even know why the port of gcc-1.4 has not been made 1.37 compatible)

I will dig on the IO error in gnuemacs (which do not exist in uemacs)
BTW I don't understand why uemacs is XON/XOFF sensitive in init/login
and not otherwise. Nor gnuemacs is not XON/XOFF sensitive

Finally I've seen in a recent news someone try to mkfs a swap
partition (or have I misread). A mkswap will cure the "cannot find
swap signature"

Marc
Any clue wellcome


[next article]
From: [email protected] (Drew Eckhardt)

Subject: Re: SCSI support
Message-ID: <[email protected]>
Date: 1 Mar 92 14:58:06 GMT
References: <[email protected]> <[email protected] .EDU>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 26
Nntp-Posting-Host: caesar.cs.colorado.edu

In article <[email protected]> [email protected] (Brian Mcbee) writes:
>In article <[email protected]> [email protected] (Drew Eckh ardt) writes:
>>Good news! I've got SCSI support (my low level driver is for the Seagate
>>ST01 / ST02, johnsonm is working on a DTC low level driver, and tthorn on
>>an adaptec) up and running.
>>
>>I will upload the code after it's been cleaned up and commented a little
>>bit better - those hardy souls wanting something sooner (or a Seagate ST0x
>>bootable image) please send E-mail.
>
>I take it that this is a replacement for the hd driver currently in the kernel.

Guess again.

>How difficult will it be to hack it so I can have both controllers?
>


No work. The SCSI drivers are entirely separate from the normal
hard disk driver. In fact, the SCSI drivers configure themselves
at boot time, detecting which host adapters are present, where
those host adapters are in the address space, and what scsi
devices are on the scsi bus(es).

This allows the same distribution kernel to be used for normal, Seagate
SCSI, and soon other SCSI hosts as well.


[next article]
From: [email protected] (Drew Eckhardt)

Subject: SCSI drivers are uploaded
Summary: uploaded and everything!
Message-ID: <[email protected]>
Date: 1 Mar 92 15:47:13 GMT
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 46
Nntp-Posting-Host: caesar.cs.colorado.edu

Ok, here's the current state of things :

I was up all night fixing bugs, changing the time-consuming
parts of the seagate driver to inline assembler (less than 1/3 the
size of the 'C' it replaced, over 4 times as fast) and documenting
things a little better.


And not only does it still work, but the seagate driver is about 4 times
faster than the straight 'C' version without arbitration, and an order of
magnitude faster than the version with arbitration.

I have uploaded the following files to nic.funet.fi and tsx-11.mit.edu
under incomming :

bootimage-0.12.seagate.Z bootable image for seagate st01/03
host adapters at C8000. Heavily
modified.

fdisk.scsi.tar.Z fdisk that recognizes hard disk 0, 1,
and the first scsi disk.

scsi.shar.Z source code to this mess.

scsi.user.doc.Z user documentation that says
these are the major minor numbers,
and what the rest of the files are.

The current flaws :

High level drivers :
sd.c needs some cleaning up still, and I'd like to move
most of the error handling and request sense down to the
mid level drivers.

st.c :
it's a stub.

seagate low level :
The uploaded bootimage is hardcoded to
address c8000 (The detection routines are there, but I only
have a signature that detects a BIOS revision 2.00 - have to
scrounge my mail or pull the chip from my dead ST-02 and
get the other signatures ), and currently the routine is not interrupt driven.
(Not sure how much we'll gain from this.)

Enjoy.


[next article]
From: [email protected] (Linus Benedict Torvalds)

Subject: Re: Small filesystem problem
Message-ID: <[email protected]>
Date: 1 Mar 92 15:20:13 GMT
References: <[email protected]>
Organization: University of Helsinki
Lines: 33

Ok, this problem has come up so many times I'd better answer with a
follow-up.

In article <[email protected]> [email protected] (D.Bolla) writes:
> [ deleted ]
># mkfs -c /dev/hd9 6247
>mkfs: unable to open /dev/hd9
>
>I suppose it is because is over the 64Mb limit ...
>Is this the reason ?

No, the reason is a very stupid mistake I did when making the root
floppy: I just copied my /dev-directory over to the floppy. No problem,
but for the fact that I don't have a /dev/hd7-9, so the special files
simply aren't on the floppy and thus they won't get installed on your
harddisk either...

The problem is happily very easy to correct: you just make the necessary
special files with mknod:

# mknod /dev/hd8 b 3 8
# mknod /dev/hd9 b 3 9

and everything should work again.

>Will the next release have a better filesystem ?

No. 0.13 still uses the minix fs, but the routines have been much better
localized, and with some additional (minor) work, it will at least be
possible to write a better fs. The 0.12 filesystem was very much
hardcoded for the minix layout.

Linus


[next article]
From: [email protected] (Joel M. Hoffman)

Subject: EGA/VGA screen fonts in Linux
Message-ID: <[email protected]>
Date: 1 Mar 92 15:51:40 GMT
Sender: [email protected] (USENET News system)
Organization: University of Maryland at College Park
Lines: 28
Nntp-Posting-Host: wam.umd.edu

I'd like to have access to soft fonts in Linux (primarily because I do
a lot of work in Hebrew). First of all, has anyone started working on
this? I don't think it will be be very difficult. But I'm still
trying to decide what the best interface will be. Reprogramming the
video card, of course, can affect not just the process running but
also other virtual terminals, so the font change belongs in the
kernal. But what should the system call be?

One possiblity is to allow ordinary users to download any font they
see fit, with perhaps a system call to take a block of memory and use
it for the font table. Ordinary users would them be able to download
nonsense into the font table, and mess up all the displays. Perhaps
only processes owning a virtual terminal should have access to this
feature.

Another possibility is to make a list of fonts in a non-writable
directory, and only load fonts from that list, so ordinary users can't
screw things up too much.

Ideally, the font-change should only affect the virtual terminal from
which it was requested, but I'm not sure I know how to do that. I
suppose it's a matter of keeping track of the video RAM and
storing/restoring it, but I haven't looked at the VC code in the
kernal yet.

Comments? Thoughts? Suggestions?

-Joel


[next article]
From: [email protected] (Drew Eckhardt)

Subject: Killing init, shutdown, reaping zombies, anoying kernel messages
Message-ID: <[email protected]>
Date: 1 Mar 92 16:14:07 GMT
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 24
Nntp-Posting-Host: caesar.cs.colorado.edu

On normal Unix machines, killing init will take you down into
single user mode. We should do this, making init run from an
exec - fork instead of an exec.

Shutdown :
If you accidentally leave elvis running on a file, in another window,
you end up with a permanantly busy file if you reset. A shutdown
command / syscall which killed everything else off would be nice.

Reaping zombies :
Every so often, my workstation goes around, makes init (pid 1) the parent
of all zombies, and then kills them properly. With the 64 proc process table
that fills up with crond zombies in just a few days, linux definately
needs this.

Anoying kernel messages :
instead of going to the fg console, printk() and panic() calls should print
to a user specified device - virtual console, tty, or lp. Come to
think of it, a syslog would be nice.

This would send all kernel messages one place, where they would not
interfere with other activities on the system, and let developers see
debugging output from their kernel hacks one place, while still
working unimpaired on another terminal (virtual or real).


[next article]
From: [email protected] (Drew Eckhardt)

Subject: Re: EGA/VGA screen fonts in Linux
Message-ID: <[email protected]>
Date: 1 Mar 92 16:26:50 GMT
References: <[email protected]>
Sender: [email protected] (The Daily Planet)
Organization: University of Colorado at Boulder
Lines: 12
Nntp-Posting-Host: caesar.cs.colorado.edu

In article <[email protected]> [email protected] (Joel M. Hoffman ) writes:
>I'd like to have access to soft fonts in Linux (primarily because I do
>a lot of work in Hebrew). First of all, has anyone started working on
>this? I don't think it will be be very difficult. But I'm still
>trying to decide what the best interface will be. Reprogramming the
>video card, of course, can affect not just the process running but
>also other virtual terminals, so the font change belongs in the
>kernal. But what should the system call be?
>

Well, X is imminient. X supports soft fonts in a wide variety of shapes
and sizes.


[next article]
From: [email protected] (Lael)

Subject: Slip?
Message-ID:
Date: 1 Mar 92 16:34:50 GMT
Organization: University of New Mexico, Albuquerque
Lines: 8


Has anyone ported slip to linux yet? Or even better
the compressed slip?


Lael
--
email: [email protected] bitnet: [email protected]


[next article]
From: [email protected] (Joel M. Hoffman)

Subject: Re: EGA/VGA screen fonts in Linux
Message-ID: <[email protected]>
Date: 1 Mar 92 17:19:54 GMT
References: <[email protected]> <[email protected] edu>
Sender: [email protected] (USENET News system)
Organization: University of Maryland at College Park
Lines: 20
Nntp-Posting-Host: wam.umd.edu

In article <[email protected]> [email protected] (Drew Eckha rdt) writes:
>In article <[email protected]> [email protected] (Joel M. Hoffma n) writes:
>>I'd like to have access to soft fonts in Linux (primarily because I do
>>a lot of work in Hebrew). First of all, has anyone started working on
>>this? I don't think it will be be very difficult. But I'm still
>>trying to decide what the best interface will be. Reprogramming the
>>video card, of course, can affect not just the process running but
>>also other virtual terminals, so the font change belongs in the
>>kernal. But what should the system call be?
>>
>
>Well, X is imminient. X supports soft fonts in a wide variety of shapes
>and sizes.

But I don't want to clobber my system with X just so I can get Hebrew.
Also, I already have everything set up to work in text-mode with
Hebrew (originally on modified VT-100 terminals in Israel).

-Joel
([email protected])


[next article]
From: [email protected] (Derron Simon)

Subject: Linux 0.13?
Message-ID: <[email protected]>
Date: 1 Mar 92 19:17:54 GMT
Sender: [email protected]
Organization: University of Pennsylvania
Lines: 10
Nntp-Posting-Host: eniac.seas.upenn.edu

When is version 0.13 scheduled to be released? I don't want to hassle with
all the patches and updates to 0.12 so I'd like to wait until 0.13 if it is
not going to be long from now. Thanks,
Derron

--
/* Derron Simon (SEAS 1994) * Tipsy fuddled boozy groggy elevated / */
/* [email protected] * Prime did edit her / Hellborn elfchild */
/* * roadhog mountain fortune hunter / Man */
/* University of Pennsylvania * beheader her - Fluff's Travels (Phish) */


[next article]
From: [email protected] (Adam Thompson)

Subject: Re: initial tty speed
Message-ID: <[email protected]>
Date: 1 Mar 92 18:41:23 GMT
References:
Organization: University of Manitoba, Winnipeg, Canada
Lines: 27

In [email protected] (Char les Hedrick) writes:

>Linus: in 0.13, I suggest making the default speed of the console
>tty's be 38400, rather than 0. With a speed of 0, emacs thinks you're
>on a very slow modem, and goes into slow search mode.

Wouldn't it be more to the point to make the speed (-1) ? That's usually how
'maximum speed' is shown. Not being a UNIX internals wizard, I don't know
if this is peculiar to some driver we don't have...

BTW: so when will 0.13 be out? I'm waiting... 'cause I just got my new
9600/v22/v32/v32b/v42/v42b/MNP5 modem, and can't use it at anything faster
then **4800** baud !!
(serial port: loses characters...)

-Adam Thompson

--
= Adam Thompson ---- Computer Engineering ---- University of Manitoba =
= [email protected] = "When you have eliminated the improbable, =
= ...!uunet!decwrl!alberta!\ = whatever is left, however impossible, =
= ccu.UManitoba.CA!umthom61 = must be the answer." =
--