[0467] daemon@ATHENA.MIT.EDU (Digestifier)  Linux Digest Archive  06/14/92 14:30 (623 lines)
Subject: Linux-Activists Digest #289
From: Digestifier <Linux-Activists-Request@BLOOM-PICAYUNE.MIT.EDU>
To: Linux-Activists@BLOOM-PICAYUNE.MIT.EDU
Reply-To: Linux-Activists@BLOOM-PICAYUNE.MIT.EDU
Date:     Sun, 14 Jun 92 14:30:08 EDT

Linux-Activists Digest #289, Volume #2           Sun, 14 Jun 92 14:30:08 EDT

Contents:
  Re: Problem with inb and outb.... Segmentation Fault?!?? (Michael O'Reilly)
  Re: Linux & ESDI? (Ed June)
  Does X on Linux support MS bus mouse? (Delman Lee)
  Re: minicom (was: minix lp daemon) (Miquel van Smoorenburg)
  Compiling the Kernel with GCC2.11c (Sakari Aaltonen)
  Help compiling Samil 3.1.25 please!  :-( (Martin Mueller)
  Mtools (Gene Choi)
  0.96a+ SCSI problems (Brian E. Gallew)
  Re: It Is Multiuser, Is It Not? (jgifford@attmail.com)
  Re: GNU tar dies on multi-volume archive.  Help. (Al Clark)
  Re: Minicom V1.3 posted in comp.os.minix (Joachim Riedel)
  GNU tar dies on multi-volume archive. Help. (Mike Davies)
  Re: Help compiling Samil 3.1.25 please! :-( (Bill Jones@cambria.columbus.oh.us)
  Re: Problems galore!!.. (Eric Hazen)
  Linux issue of UNIX USER and BBS system (Steve Hanson)

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

From: oreillym@tartarus.uwa.edu.au (Michael O'Reilly)
Subject: Re: Problem with inb and outb.... Segmentation Fault?!??
Date: 14 Jun 92 10:26:04 GMT

techs@pioneer.unm.edu (Erik Fichtner) writes:
: Well, Here I am working on the dos emulator....  more things
: work, but trying to get the inb and outb functions to work
: causes a segmentation fault... 
: 
: Here's what i did...
: 
: #include <asm/io.h>
: 
: whenever there's a call to either inb or outb, which are assembled
: in the .h file, it crashes the program with a segmentation fault.
: 
: Why doesn't this code work?  it's distributed with the compiler, I'd 
: expect it to work.   /usr/src/linux/include/asm/io.h doesn't work 
: and it should since the KERNEL uses these functions without any 
: problems.  
: What's wrong here??!

User programs CAN'T do arbitrary I/O to ports. You have to go via the
ioperm sys call to explicity allow it. In general you shouldn\'t need to
go to the ports as linux can get very confused if you do. Allowing the
dos emulator to do all the port I/O it wants is almost guarenteed to
crash linux.. ;)

A better way to go would be to emulate the port in question. I.e. if it
is trying to directly access the serial port, translate this to read()
and write() and ioctl()'s on a serial device.

: Totally confused,
: Techs 
: [techs@pioneer.unm.edu][Space and Planatary Image Facility]

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

From: june@f308.n133.z1.FIDONET.ORG (Ed June)
Subject: Re: Linux & ESDI?
Date: 13 Jun 92 13:18:57 GMT

On <Jun 12 14:04>, Uucp of 1:369/11 wrote to June:

 U> From: cybernet!mips.mitek.com!tk (Tom Kimball)
 U> To:   f308.n133.z1.fidonet.org!june
 U> Date: Fri, 12 Jun 92 07:24:26 -0700

 U> In article <34.2A383DF2@gisatl.fidonet.org> you write:
>Has anyone gotten Linux to run on an ESDI controller/HD?
>I have an UltraStor 12C with 1MB, so I'm most interested in that, 
>but would like to hear from all ESDIr's...

 U> I have a UltraStor /ESDI and have installed Linix.  Works great.
 U> -- 
 U> Tom Kimball             tk@oc.com
 U> OpenConnect Systems     (214)308-0497
 U> Carrollton, TX  75006

Thanks, good to know, now to send those disks...


--  
Ed June

Fidonet:   1:133/308, Atlanta's OS/2 Users Group BBS, 404-471-1549
Internet:  ed_june@over.mind.org
Internet:  june@gisatl.fidonet.org
UUCP:      ...!emory!uumind!overmind!ed_june
UUCP :     ...!gatech!warlord!gisatl!308!june

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

From: delman.lee@engineering.oxford.ac.uk (Delman Lee)
Subject: Does X on Linux support MS bus mouse?
Date: 14 Jun 92 12:35:59 GMT

The subject says it all. Thanks in advance!
______________________________________________________________________________

  Delman Lee
  Dept. of Engineering Science,
  Oxford University,                         JANET: delman.lee@uk.ac.ox.eng
  Parks Road,              Internet, BITNET & UUCP: delman.lee@eng.ox.ac.uk
  Oxford, OX1 3PJ, U.K..                    ( UUCP: ...!uknet!oxuniv!delman )
______________________________________________________________________________

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

From: miquels@htsa.aha.nl (Miquel van Smoorenburg)
Subject: Re: minicom (was: minix lp daemon)
Date: Sun, 14 Jun 1992 10:53:36 GMT

In article <1992Jun14.071213.29150@cbnewse.cb.att.com> sph1@cbnewse.cb.att.com (stephen.p.hill) writes:
[ stuff about minix lp package deleted, since I haven't tested it ]
>Stephen Hill
>sph1@cbnewse.att.com
>
>P.S.  I also tried the "minicom" program posted to the minix 
>group a few days ago.  It compiles and runs, but it seems 
>to have a little trouble with the screen.  The status bar 
>eventually gets messed up and keeps flashing on and off, 
>and I have also had it hang up on calls unexpectedly.

Err, well, I am the author of minicom.

I *did* compile it under Linux, but I don't have a 386 box at home
(only a 286) so I couldn't really test it. We have a Linux box at
school, so we can play with it, but it doesn't have anything attached
to the com ports apart from a mouse.

It *does* run perfectly under HP/UX 8.0 which is also a Sysv compatible OS,
and it compiled fine under Linux. It also did show the garbage from
the mouse perfectly, so I assumed it worked..

If someone does get the bugs out, or knows what's going on, could you
mail me about it? I could then post a 'official' patch.

Thanks - Mike.



-- 
%     Miquel van Smoorenburg, Baljuwstraat 20, 2461SL Langeraar, Holland.     %
% miquels@maestro.htsa.aha.nl,miquels@drinkel.nl.mugnet.org (home, in 2 weeks)%
%                   Calm down - it's just ones and zeros                      %

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

From: sakaria@vipunen.hut.fi (Sakari Aaltonen)
Subject: Compiling the Kernel with GCC2.11c
Date: 14 Jun 92 12:52:18 GMT
Reply-To: sakaria@vipunen.hut.fi (Sakari Aaltonen)

The following might help if you try to compile the 0.96a (or 0.95c+) kernel
with GCC2.11c.

GCC2.11c (for Linux) is organized such that gcc is used for everything, 
i.e., trying to run cpp or cc results in "command not found".

On the other hand, the kernel Makefile's (10 or 11 of them) all include
CPP     =cpp -nostdinc ..........

Additionally, there is a line in kernel/blk_drv/scsi/Makefile that says
cc .............

If you replace the cpp's with "gcc -E"'s and cc with "gcc", it helps.
-- 
=============================================================================
Sakari Aaltonen                             Helsinki University of Technology
Email: sakaria@vipunen.hut.fi
---  You can't keep a Finn down without him getting all red in the face  ----

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

From: mundm@zelator.in-berlin.de (Martin Mueller)
Subject: Help compiling Samil 3.1.25 please!  :-(
Date: 14 Jun 92 12:45:13 GMT

Hi!

I've got problems compiling Samil 3.1.25.
make -k depend fails with the following messages:

rm -f directors/Makefile routers/Makefile transports/Makefile
sh mkdrivtab.sh ../conf/driver/unix-generic
Build drivertab.c, backup in .drivertab.c
make[1]: *** [directors/Makefile] Error 1
sh mkdrivtab.sh ../conf/driver/unix-generic
Build drivertab.c, backup in .drivertab.c
make[1]: *** [routers/Makefile] Error 1
sh mkdrivtab.sh ../conf/driver/unix-generic
Build drivertab.c, backup in .drivertab.c
make[1]: *** [transports/Makefile] Error 1
chmod -w Makefile
make[1]: Target `depend' not remade because of errors.

I'm using the config-files from tsx-11! I think the problem is,
that there is no Makefile in the subdirs directors, routers and transports!

The "real" make fails with the same errors, though everything exept
alloc.c and drivetab.c compiles fine.
When executing a "make -k" I get the following errors:
[stuff deleted]
Build targets under pd/pathalias ...
make[2]: *** No way to make target `y.tab.c'.
make[2]: Target `all' not remade because of errors.
[stuff deleted]
Build targets under src ...
sh mkdrivtab.sh ../conf/driver/unix-generic
Build drivertab.c, backup in .drivertab.c
make[1]: *** [directors/Makefile] Error 1
sh mkdrivtab.sh ../conf/driver/unix-generic
Build drivertab.c, backup in .drivertab.c
make[1]: *** [routers/Makefile] Error 1
sh mkdrivtab.sh ../conf/driver/unix-generic
Build drivertab.c, backup in .drivertab.c
make[1]: *** [transports/Makefile] Error 1
. ./defs.sh; $CC $CFLAGS $INCLUDES -c alloc.c
alloc.c: In function `xrealloc':
alloc.c:98: void value not ignored as it ought to be
alloc.c: In function `xfree':
alloc.c:121: void value not ignored as it ought to be
make[1]: *** [alloc.o] Error 1
. ./defs.sh; $CC $CFLAGS $INCLUDES -c drivertab.c
drivertab.c:22: unterminated string or character constant
drivertab.c:22: possible real start of unterminated constant
make[1]: *** [drivertab.o] Error 1
make[1]: Target `all' not remade because of errors.

Any help appreciated!

                        M&M

-- 
========================================================================
========================================================================

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

From: genie@netcom.com (Gene Choi)
Subject: Mtools
Date: Sun, 14 Jun 92 13:57:55 GMT


Hi everyone.  Thanks for all your help.  It's been very help,
and have started to figure out where I went wrong, etc..

I still can't seem to get a working version of mtools yet.
I'm trying to compile it under cc 2.1.  Does it need
2.11 (latest version?)?  Anyways, I'd greatly appreciate it
if someone could mail me a uuencoded file of the program, if
it is possible.

Thanks in advance.

Gene

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

From: bg11+@andrew.cmu.edu (Brian E. Gallew)
Subject: 0.96a+ SCSI problems
Date: 14 Jun 92 15:31:40 GMT

I seem to be having problems with SCSI now.  I have been using the
pre-96 kernel forever, and having no problem with it.  I just
downloaded the 0.96a+ bootimage from headrest.woz.colorado.edu, and
tried to boot it.  I get the following important messages :-)

Linus ver 0.96a - 16 May 26
[blah]
... SCSI ... v1.8...
[blah]
sd371916 : block size assumed to be 512 bytes, disk size 1GB
SCSI host 0 timed out - aborting command

and then the kernel hangs!  needless to say, I don't have a scsi
device #371916.  Part of the previous blah is the normal stuff for sd0
which I do have.  Any ideas?

                                  -Brian

You drop the bomb -more-
It goes off... -more-
=========================================================================
|  My cluelessness is more real than apparent...                        |
=========================================================================
|  DEC measures their benchmarks with a calendar!!                      |
=========================================================================

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

From: jgifford@attmail.com
Subject: Re: It Is Multiuser, Is It Not?
Reply-To: jgifford@attmail.com
Date: Sun, 14 Jun 1992 16:09:06 GMT

You might try the following inittab line(worked for me)

ttys1:vt100:/bin/getty 9600 ttys1

The -m switch tells getty that it is a modem connecting(I think) .
I hope this helps you, and if not, just let me know, and I will try something else.  :)
good luck,
Jim Gifford
jgifford@attmail.com

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

From: aclark@netcom.com (Al Clark)
Subject: Re: GNU tar dies on multi-volume archive.  Help.
Date: Sun, 14 Jun 92 15:37:50 GMT

In article <1992Jun12.192948.7047@wam.umd.edu> joel@wam.umd.edu (Joel M. Hoffman) writes:
>I'm trying to backup my /acc partition, using the following:
>
>       tar +listed-incremental /etc/backup/full.acc +verbose +create
>        +multi-volume +tape-length 1400000 +compress +one-file-system
>        +file /dev/a: /acc/
>
>(it's all on one line, of course).
>       
>but it dies before the end of the first disk with:
>
>       tar (child): only wrote 0 of 10240 bytes to /dev/a:
>       acc/joel/letters/personal/mike525.tex
>       acc/joel/letters/personal/mike525.dvi
>
>What's going on?  I purposely used tape-length 1400000, which is a
>little less than 1440000, because I know there have been some problems
>reading the last bit of floppies.  Any ideas?
>
>``Tar +version'' tells me GNU tar version 1.10.  This is /not/ the
>buggy tar distributed with 0.95; it untars gcc without a problem.  I'm
>running 0.96 with the first patch applied.
>
>-Joel
>(joel@wam.umd.edu)Newsgroups: comp.os.linux
>
>

I haven't used the option, but I was doing some research for a tar.1 manual
I'm working on, and it appears that the +tape-length option specifies "tape" 
size in 1K blocks, thus you should use:

         +tape-length 1400

to get the effect you want.  I have NOT tested this.  Good luck.

Al

-- 
Al - aclark@netcom.com - My opinions are my own.
      *** Practice random kindness and senseless acts of beauty! ***

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

From: joachim@jr.sub.org (Joachim Riedel)
Crossposted-To: comp.os.coherent,alt.sources
Subject: Re: Minicom V1.3 posted in comp.os.minix
Date: 14 Jun 92 09:31:12 GMT

miquels@htsa.aha.nl (Miquel van Smoorenburg) writes:
>       I have just osted the newest version (1.3) of my Minicom
> comms package in comp.os.minix. This maybe of interest to a lot of
> people, since it now runs on:
> 
>   Minix, BSD 4.3, Linux, Sysv, HP/UX, SVR2, Coherent and probably more.

Europe/Germany:

If someone did not get comp.os.minix or missed Miquel's articles then
he may poll minicom from ECSBBS (see signature for details).
It's file # 46 in area downloads.

Cheers,

Joachim

PS: there is a login gast/guest which now again allows to download files
(it was for a short time disabled because someone tried to fool me.....)
Choose <f> for file transfers, <g> to group downloads, <d> download and
choose file 46 and afterwards XMODEM,KERMIT,ZMODEM,YMODEM.

Nevertheless is using ECSBBS's BBS-shell free of charge so no reason to
not login with real-name and serial-no (which allows to have also the latest
bug fix updates (uucico.p40.Z for example))

+----------------------------+--------------------+-------------------------+
| Joachim Riedel             | joachim@jr.sub.org | jr@connie.de.convex.com |
| Geschwister-Scholl-Str. 48 +----------------------------------------------|
| D-6050 Offenbach am Main   | European Coherent Support BBS: +49 69 858711 |
| Tel. +49 69 85 62 25       | V22bis, V32, V32bis (T3000 and USR Dual Std. |
+---------------------------------------------------------------------------+

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

From: mike@hectic.demon.co.uk (Mike Davies)
Subject: GNU tar dies on multi-volume archive. Help.
Date: 14 Jun 92 16:39:38 GMT


In article <82flksq.aclark@netcom.com> aclark@netcom.com writes:

> I haven't used the option, but I was doing some research for a tar.1 manual
> I'm working on, and it appears that the +tape-length option specifies "tape" 
> size in 1K blocks, thus you should use:
> 
>          +tape-length 1400
> 
> to get the effect you want.  I have NOT tested this.  Good luck.
> 

I've just done a backup (well I *hope* I did) without using the 
tape-length parameter. It seemed to work okay and came out at about
the right number of disks. Can tar pick up the device size from the
driver ?

Anyway ... I had done some tests before that and found that the 
+compress option caused a similar error to that original quoted.
I didn't use +compress, surprise, surprise !


Mike Davies   mike@hectic.demon.co.uk
              mikedav@cix.compulink.co.uk
              +44 635 36746

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

From: Bill Jones@cambria.columbus.oh.us
Subject: Re: Help compiling Samil 3.1.25 please! :-(
Date: 14 Jun 92 18:27:05 GMT

> I've got problems compiling Samil 3.1.25.
> make -k depend fails with the following messages:
>
> rm -f directors/Makefile routers/Makefile transports/Makefile
> sh mkdrivtab.sh ../conf/driver/unix-generic
> Build drivertab.c, backup in .drivertab.c
> make[1]: *** [directors/Makefile] Error 1
> sh mkdrivtab.sh ../conf/driver/unix-generic
> Build drivertab.c, backup in .drivertab.c
> make[1]: *** [routers/Makefile] Error 1
> sh mkdrivtab.sh ../conf/driver/unix-generic
> Build drivertab.c, backup in .drivertab.c
> make[1]: *** [transports/Makefile] Error 1
> chmod -w Makefile
> make[1]: Target `depend' not remade because of errors.

> I'm using the config-files from tsx-11! I think the problem is,
> that there is no Makefile in the subdirs directors, routers and transports!

I didn't use these config files (I didn't know there was a working smail
at tsx-11).  I used the source from ftp.uu.net.

> The "real" make fails with the same errors, though everything exept
> alloc.c and drivetab.c compiles fine.
> When executing a "make -k" I get the following errors:
> [stuff deleted]
> Build targets under pd/pathalias ...
> make[2]: *** No way to make target `y.tab.c'.
> make[2]: Target `all' not remade because of errors.
> [stuff deleted]
> Build targets under src ...
> sh mkdrivtab.sh ../conf/driver/unix-generic
> Build drivertab.c, backup in .drivertab.c
> make[1]: *** [directors/Makefile] Error 1
> sh mkdrivtab.sh ../conf/driver/unix-generic
> Build drivertab.c, backup in .drivertab.c
> make[1]: *** [routers/Makefile] Error 1
> sh mkdrivtab.sh ../conf/driver/unix-generic
> Build drivertab.c, backup in .drivertab.c
> make[1]: *** [transports/Makefile] Error 1


[additional errors deleted]


(This is a copy of my mail message regarding the problem.  This same
question seems to have arisen more than once.  Perhaps it's time to
put it into the FAQ or fix bash or whatever it is that is causing the
problem.)


I ran into the same problems.  There seems to be some sort of a bug
in bash which is keeping that portion of the shell script from
running.

My solution (which is a kludge which some people cannot do because
they don't have access to another machine) was to put smail on
a DECStation 3100, run the initial make there, tar everything up
and port it to my linux box.  Then, copy drivertab.c and the Makefiles
from the directors, routers, and transports directories somewhere safe,
run the initial make again under linux (necessary to get the
dependencies correct), copy drivertab.c back to where it belongs
as well as all the makefiles (Note that the intial make MUST be a
make -k otherwise you'll never get anything done), edit the makefiles
in the directors, routers, and transports directories to reflect
the correct dependencies.

Now, cd the the directors directory, run make there.  Do the same
for routers and transports.  Go back to the conf (I may be mixing
packages in me head here but I think it is the conf directory) directory
and run make normally.  If you make any additional changes, do not
allow make to rebuild the dependencies without having a copy of
these files somewhere.

It's a long process, but with a little practice, you can get the
entire smail package built in about an hour on a 386SX machine (not
counting porting time to and from another machine).  It also has
the benefit of working, using elm2.3pl11 as the user interface and
procmail as the local mail agent.

Now if someone could just help me get C-News working :).

Bill

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

From: hazen@budoe.bu.edu (Eric Hazen)
Subject: Re: Problems galore!!..
Date: 14 Jun 92 17:53:30 GMT

In article <#sflp+d.genie@netcom.com> genie@netcom.com (Gene Choi) writes:

   Another solution to this problem might be the fact that 
   kermit is not transferring the files correctly.  It seems that
   every file transferred over kermit is slightly (a few bytes)
   smaller than on the remote host original file size.

Last night I installed Kermit5A (ALPHA) from tsx-11 and had the same
problems.  You have to use BINARY file mode to transfer anything but
text files.  I found that I had to say "set file type binary" on BOTH
ends (local and remote).  Also, the transfer speed is much faster if you
set the packet length to something longer than the default (assuming
your remote kermit supports it).
I put these commands in my ".kermrc" file in my home directory, and
everthing works well now:

   set file type binary                 ; use binary file transfer
   set file name literal                ; don't mess with file names
   set receive packet-length 1000       ; receive long packets
   set send packet-length 1000          ; send long packets
   set prompt "Linux-Kermit> "          ; so I know where I am :)

I should note that I have only been able to dial my modem by hand, using
ATDT command.  My config:  0.96a (no patches), 0.95a root, Kermit5A


--
    Eric Hazen

        Boston U Physics / 617-353-4117
        hazen@buphy.bu.edu / hazen@buphyc.bitnet / BUPHYC::HAZEN

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

From: sjh@unixuser (Steve Hanson)
Subject: Linux issue of UNIX USER and BBS system
Date: 13 Jun 92 16:12:54 GMT

This post serves multiple purposes.  I just got the rest of waffle
up and running and need to do a post to see if I actually
am getting out to  the real world.

More importantly - for those of you who have either wanted to
get your hands on a copy of Linux or wanted to subscribe
to my newsletter UNIX USER, It is now easier to do
either.  We're now accepting Visa and Mastercard for
subscriptions.  There's been some interest in this
group, so I thought I'd bring it up.  A 1 year subscription
to UNIX USER is $35 in the US.  Just the Linux issue 
(wwhich includes the MCC Interim release and notes, FAQ,
etc.) is $15.  The Linux issue ships out toward the end of
the week, for those of you who have been waiting.

Also, our BBS is now somehwat more stable and interesting.
We are starting up a news feed and email service.  We will
have to charge something for news and email REAL soon
now, but we haven't decided on what to charge.  For now
its free.  You can call it at (708)879-8633, which is also
our voice mail number.  If you dial in to the BBS, please
make sure you are set for a fairly long (up to 60 seconds)
wait for a connect tone, as you have to wait for the voice
mail intro to finish.  2400 baud, 8N1.  You can subscribe
to UNIX USER or order the Linux disks by leaving a message
for the SYSOP with the feedback command.  We'll need
your mail address, MC or VISA number, and expiration 
date.  There will be a subscription door soon, but we
need to get this issue out the door first.

sjh@unixuser (Steve Hanson)
UNIX USER bbs -- (708)879-8633

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


** FOR YOUR REFERENCE **

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

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

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

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

Linux may be obtained via one of these FTP sites:
    nic.funet.fi				pub/OS/Linux
    tsx-11.mit.edu				pub/linux
    tupac-amaru.informatik.rwth-aachen.de	pub/msdos/replace

The current version of Linux is 0.96a released on May 22, 1992

End of Linux-Activists Digest
******************************
--[0467]--

