Subject: Linux-Development Digest #936
From: Digestifier <Linux-Development-Request@senator-bedfellow.MIT.EDU>
To: Linux-Development@senator-bedfellow.MIT.EDU
Reply-To: Linux-Development@senator-bedfellow.MIT.EDU
Date:     Fri, 22 Jul 94 16:13:14 EDT

Linux-Development Digest #936, Volume #1         Fri, 22 Jul 94 16:13:14 EDT

Contents:
  Re: ASUS486SP3G Linux users? (Michael Will)
  Re: ASUS486SP3G Linux users? (Michael Will)
  Re: Replacing part of shared libc (Mitchum DSouza)
  Re: What topics should be discussed here? (sravet@bangate.compaq.com)
  Re: link() system call (Mitchum DSouza)
  Re: Hacking odd-ball Floppy drive device drivers (Rob Janssen)
  Re: Linux Performance Enhance ? (Jim Balter)
  QMAGIC is nice, but impure ZMAGIC is smaller (Wolfram Gloger)
  German Linux Kernel Book (Stephen Collyer)
  Dosemu, xterm & 8bit-chars (Kai Arstila)
  Use of network 'select' command (...In Bed)
  Re: 1.1.32 kernel broken? (Linus Torvalds)
  Re: Anyone else lust for threads? (Cameron N. Heide)
  Re: Linux Performance Enhance ? (Jim Balter)
  Re: gcc-2.6.0: help w/asm (Richard Earnshaw)
  TLI library for linux? (Diane Tang)
  Re: GOTO haters ..Re: Linux Performance Enhance ? (Jeff Dege)
  Re: loadable modules: undefined interuptible_sleep_on (Bjorn Ekwall)
  Re: Starting New DIP (Bill Hay)

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

From: michaelw@desaster.student.uni-tuebingen.de (Michael Will)
Subject: Re: ASUS486SP3G Linux users?
Date: Fri, 22 Jul 1994 09:18:41 GMT

broadley@neurocog.lrdc.pitt.edu (Bill Broadley) writes:
>I'd like to talk to any ASUS PCI/I-486SP3G linux users, especially if
>they use the on board scsi.
>Anything bad to say about the board?  I want to buy one solely for the
>purpose of running linux.  Thanks.
It is a fine board, has has the saturn-chipset rev. 4 which seems to have
got rid of the problems the old 486SP3-Board had with saturn chipset rev. 2.

It should perform excellent.

To use the onboard SCSI-Chip ncr53c810 you need Drews NCR53c810-Driver. It is
not yet in the standard-kernel (why not?) but works almost perfect. The only
problem you might get is with disk-geometry-detection. While the syscall
is implemented in rev. 3 of the code, it does not work for me. So I have
to tell fdisk the geometry by hand (dparam.com running on DOS can tell you
the parameters. it comes with LILO and is included in the SCSI-Howto.) 
Lilo does not work for me - sigh...

...but apart from that the driver works flawless - never lost a byte.

I am running the old buggy ASUS PCI/I-486SP3 with NCR53c810 and ATI-GUP-PCI
without any problems. Any performance-switch I could enable in the BIOS
is turned on, never had a crash. The ASUS PCI/I-486SP3G you mention should
be even faster. I would say it is an excellent choice.

Cheers, Michael Will
-- 
. .         Michael Will <michaelw@desaster.student.uni-tuebingen.de> 
 .      cs-student in Tuebingen, Germany, Europe, Earth, Solar-System, [...]

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

From: michaelw@desaster.student.uni-tuebingen.de (Michael Will)
Subject: Re: ASUS486SP3G Linux users?
Date: Fri, 22 Jul 1994 09:51:54 GMT

I forgot to mention the place you get the NCR-Driver from:
tsx-11.mit.edu:/pub/linux/ALPHA/scsi/ncr or something alike.

In the Slackware-2.0-distribution are said to be two ncr-disks now too.

Cheers, Michael Will
-- 
. .         Michael Will <michaelw@desaster.student.uni-tuebingen.de> 
 .      cs-student in Tuebingen, Germany, Europe, Earth, Solar-System, [...]

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

From: Mitchum.DSouza@mrc-apu.cam.ac.uk (Mitchum DSouza)
Subject: Re: Replacing part of shared libc
Date: 22 Jul 1994 10:24:57 GMT

In article <30mdpl$s7c@charon.citicorp.com>, philip@charon.citicorp.com (Philip
Gladstone) writes:
|> 
|> Hi,
|>      I want to replace the resolver functions in the shared
|>      version of libc. I've looked at the DLL tools stuff, but it
|>      seems that I need to rebuild the entire libc from scratch.
|> 
|>      Is this really the case, or is there some short cut that I
|>      can take?

I haven't had time to try it yet, but your should just be able to take the
"res" subdirectory from the bind distribution and substitute it into the
the libc-linux source tree by renaming it "inet".

Failing that you can just wait until the next libc is out which may have the
new bind code in it.

Mitch

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

From: sravet@bangate.compaq.com
Subject: Re: What topics should be discussed here?
Date: Wed, 20 Jul 1994 17:49:15 GMT

In article <1994Jul20.160240.15427@cc.ntnu.edu.tw> s21008@cc.ntnu.edu.tw (DL83-08) writes:
>I asked for help about whether or not a WangDAT could work on Linux. I
>received several very useful replies. The problem is, when I finally got a
>chance to get at the machine (486DX33) and installed Linux Slackware 1.2,
>compiled a kernel with ATA1542 SCSI Tape support, the SCSI was not detected
>during boot-up. Just said (SCSI 0 drives 0 tapes). I was under the
>impression that it was as easy as compile and run. What could posibly be the
>problem? Any help?
>
>Thanks in advance. This is rather important, and urgent!

This post is not flame bait.

I see lots of posts like this that really have nothing to do
with development.  I have considered *politely* emailing the
poster to say that questions such as this are better posed and
answered in c.o.l.help.  What is the general feeling about these
sorts of posts?  I'm not net.cop, but this group does have a lot
of off-topic traffic.  Is there an official charter?  Should there
be a weekly posting that says what is appropriate to post here?

again, I don't want to start a war here, but......

--steve

Steve Ravet     sravet@bangate.compaq.com
"Baby you're a genius when it comes to cooking up some chili sauce...."

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

From: Mitchum.DSouza@mrc-apu.cam.ac.uk (Mitchum DSouza)
Subject: Re: link() system call
Date: 22 Jul 1994 13:42:43 GMT

In article <CtBv66.672@seneca.ix.de>, hm@seneca.ix.de (Harald Milz) writes:
|> A N Burton (anb@doc.ic.ac.uk) wrote:
|> : > Would anyone know if any of the file systems available under linux allow
|> : > 'root' to create _HARD_ links to directories?
|> 
|> I assume you mean "loopback mounts" (that's at least what they are called 
|> under SunOS and AIX). No, I've been trying to convince several people
|> to implement this (me not being a kernel hacker), with no success yet. 
|> 

loopback devices (with encryption) are already available with Ted Tyso's
loop driver.

Mitch

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

From: rob@pe1chl.ampr.org (Rob Janssen)
Subject: Re: Hacking odd-ball Floppy drive device drivers
Reply-To: pe1chl@rabo.nl
Date: Fri, 22 Jul 1994 07:20:49 GMT

In <30mi4f$663@scapa.cs.ualberta.ca> gordon@cs.ualberta.ca (Gordon Atwood) writes:

>Can someone with the technical background comment on the advisibility/
>practicality of writing a device driver to read odd-ball formats?

>Or better yet, point me at someone/thing that already does it?

>In particular, the Apple ][e Prodos formatted floppy disk (5.25").

There are difficult technical problems when reading these disks on a
PC.  Maybe not completely unsolvable, but at least it will require a lot
of hacking.

I think a better bet is to read them on the original system, and setup
some method to transfer them to the Linux machine using a serial line
or so.

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

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

From: jqb@netcom.com (Jim Balter)
Subject: Re: Linux Performance Enhance ?
Date: Fri, 22 Jul 1994 14:24:08 GMT

In article <DAVIS.94Jul16095716@pacific.mps.ohio-state.edu>,
John E. Davis <davis@amy.tch.harvard.edu> wrote:
>In article <jqbCt0tpM.MsK@netcom.com> jqb@netcom.com (Jim Balter) writes:
>   Setting static what to zero will not fail, in the sense that Rob's
>   original code failed, by not doing all three tests.  Having what be 0
>   instead of 2 after failing all 5+ iterations of the loop is not much
>   of a failure; it doesn't much matter what "what" is upon the next entry in
>   that case.
>
>In order to make this statement, you have made use of knowledge of the
>calling routine.

Of course I did.

>Without that knowledge, setting what to zero is not right.

Of course it isn't.

Now, did you have a point?
-- 
<J Q B>

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

From: u7y22ab@sun4.lrz-muenchen.de (Wolfram Gloger)
Subject: QMAGIC is nice, but impure ZMAGIC is smaller
Date: 22 Jul 1994 17:46:20 +0200

Hi,

In a recent note on the kernel channel, Eric Youngdale has announced
that a future version of binutils will only support QMAGIC binaries.
While I am very happy with this type of format for demand paged
executables, so far I haven't been able to generate an impure QMAGIC
binary, i.e. combine the -qmagic and -N linker switches.  I always
seemed to end up with a ZMAGIC impure executable.  These are really
small and IMHO useful e.g. for root disks (where every byte counts)
and for programs like sync, where I am pretty sure they run for such
a short period of time that they won't be swapped out (in which case
the disadvantage of impurity would show up, as far as I understand
it).

Am I missing something ?

Thanks,

Wolfram.

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

From: stephen@dogmatix.inmos.co.uk (Stephen Collyer)
Subject: German Linux Kernel Book
Date: Wed, 20 Jul 1994 13:19:27 GMT

I've heard that there is a published book in German on the Linux kernel.
Could some kind soul let me know the publisher, ISBN, price etc. I'm 
interested in getting a copy.

I'd appreciate email replies - I won't see any posted response due to a
flaky newsfeed :-(

Danke im Voraus

Steve Collyer.

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

From: arstila@kl-linux.helsinki.fi (Kai Arstila)
Crossposted-To: comp.os.linux.help
Subject: Dosemu, xterm & 8bit-chars
Date: 20 Jul 1994 12:03:40 GMT

What is the reason that I can't use 8-bit characters when running Dosemu
in xterm? I know that my xterm can show the whole IBM-characters set
when I eg. type a file consisting those characters in dosemu ( outlook
of the characters of course depends on the font used in xterm), however when
I try to cut and paste 8-bit characters or press a national character
button in my finnish keyboard I hear a single beep from dosemu (I can
change the sound using speaker -option in .dosrc -file). My xterm and
tty should be 8-bit clean because I can use 8-bit characters with tcsh
and editor (uemacs) without problems.

I found the following quote from DOSEMU-HOWTO:

> 6.2.  Is it possible to run dosemu in an xterm?
>
>  Evmorfopoulos Dimitris (devmorfo@mtu.edu) reported (12/10/93) that
>
>  It is possible to run dosemu under an xterm, but without any graphics,
>  and with no more than 128 characters.
>
>  H. Peter Anvin (hpa@hook.eecs.nwu.edu) adds (2/6/94)
>
>  no more than 128 characters... unless you use a special font.  Someone
>  has an IBM PC (CP 437) X font for use with ANSI-graphics on BBS's, as
>  well as a patch for Xterm that *might* include character translation
>  (for letters like \305, \304, \326 for example.
>

However, this does not say anything about the reason for the problem.

*******************************************************************************
Kai Arstila                                 phone: +358-0-708 4009
Accelerator Laboratory, Univ. of Helsinki   fax:   +358-0-708 4042
P.O. Box 43                                 email: Kai.Arstila@Helsinki.FI
FI-00014 Helsinki, Finland                  finger:arstila@kl-linux.helsinki.fi 
*******************************************************************************

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

From: cavenewt@netcom.com (...In Bed)
Subject: Use of network 'select' command
Date: Fri, 22 Jul 1994 14:27:49 GMT

I'm using 'select' to tell me when socket decriptors become ready for reading
so I don't block on a read.  (As I'm reading from 2 decriptors).  These
descriptors come from TCP connections.  However, select never becomes
ready.  I've made a tight little loop, and printed IS_SET(sock,sockfdlist)
after the select (with a timeout set of 1 second and 1 usec), and none of
the sockets claim to be ready after receiving data.  I know they have received
data, because I've typed it, and a netstat shows stuff in the queue waiting
to be read.


Is there something strange with the Linux select command, or will I have to post
a code sample to get a more acuurate response?

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

From: torvalds@cc.Helsinki.FI (Linus Torvalds)
Subject: Re: 1.1.32 kernel broken?
Date: 21 Jul 1994 09:08:36 +0300

In article <30kjjn$gnq@watnews1.watson.ibm.com>,
Uri Blumenthal <uri@watson.ibm.com> wrote:
>Hi,
>       After installation of 1.1.32 I got lots of
>       these:
>
>/var/log/notice:
>Jul 20 18:54:39 angmar linux: del_timer() called with timer not initialized
>Jul 20 18:55:25 angmar linux: del_timer() called with timer not initialized
>Jul 20 18:57:26 angmar last message repeated 1635 times
>Jul 20 19:07:27 angmar last message repeated 1888 times
>
>       Comments? Fixes? 

I had a temporary brain disorder yesterday, and the above is due to a
one-liner horrible bug. linux/kernel/sched.c, del_timer():

-       if (p->next || p->prev)
+       if (timer->next || timer->prev)
                printk("del_timer() ...");

ie change the 'p's into 'timer's on that particular line that tests
before the printk().. 

I also missed a 'rss++' in mm/memory.c. Ugh. I'll make a 1.1.33 today
that fixes those two at least.

                Linus

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

From: heide@ee.ualberta.ca (Cameron N. Heide)
Subject: Re: Anyone else lust for threads?
Date: 22 Jul 1994 15:30:28 GMT

ddt@idcube.idsoftware.com (David Taylor) writes:

>I may have missed something in the FAQ's, but assuming that there
>is no thread support in Linux, are there plans to add or document
>it?  I'm saddened by the fact that our next generation sound code
>can be developed under NEXTSTEP, Irix, Solaris, etc., but apparently
>not Linux because of a lack of threads.  Am I wrong?  *hope, hope*

I've actually been hacking around with the BSD libpthread package
with moderate success (the library compiles, the application links
and runs, but it won't actually switch threads yet.  Grrr....).

I thought I saw a message some time ago about someone having ported
pthreads successfully, but the message has long since expired (if
it exists).  I'd also appreciate hearing about any success with this
package.

--
Cameron Heide (heide@ee.ualberta.ca)  "I'll be socially unpopular!  More so!"


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

Crossposted-To: comp.lang.c
From: jqb@netcom.com (Jim Balter)
Subject: Re: Linux Performance Enhance ?
Date: Fri, 22 Jul 1994 14:30:26 GMT

In article <774702537snz@genesis.demon.co.uk>,
Lawrence Kirby <fred@genesis.demon.co.uk> wrote:
>In article <Ct6HJw.GDI@and.nl> jos@and.nl "Jos Horsmeier" writes:
>
>>Not that it matter much actually, but I beg to differ here. Duff's
>>device is an application of a very sneaky loophole in the syntax
>>of the C language with very interesting semantical consequences ...
>>I think I like that thingie ...
>
>I agree that C is one of the languages where it can be implemented but it
>isn't the only one (clearly it can be implemented in machine code so
>it isn't inherently a C language issue).

Huh?  You can nest a do loop inside a switch in machine language?


Duff's Device is not merely a branching into the first iteration of a loop.
It is syntactically specific way of doing this.  It can only be implemented
in a language where switch cases fall through and where switch labels
can occur inside an inner block.  I can think of two languages other than C
in which this can be done: C++ and Objective C.  Hmm, wonder what these have
in common?
-- 
<J Q B>

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

Crossposted-To: gnu.gcc.help
From: rwe@cs.utwente.nl (Richard Earnshaw)
Subject: Re: gcc-2.6.0: help w/asm
Date: Fri, 22 Jul 1994 14:22:32 GMT

In article <30mkpi$8np@archive.ny.jpmorgan.com>, cae@aww.ny.jpmorgan.com (Caleb Epstein) writes:
|> static inline unsigned char xchgb(unsigned char reg,
|>                                volatile unsigned char *mem)
|> {
|>   asm("xchgb %0,%1" :
|>       "=r" (reg), "=m" (*(unsigned char *)mem) :
|>       "0" (reg), "1" (*(unsigned char *)mem));               /* line 309 */
|>   return reg;
|> }
|> 

I'm not a 386 expert, but try

static inline unsigned char xchgb(unsigned char reg,
                                  volatile unsigned char *mem)
{
   asm ("xchgb %0, %1" :
        "+r" (reg), "+m" (*(unsigned char *)mem));
   return reg;
}

'+' in a constraint says that the operand is read as well as written.

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

From: dtang@virtual12.harvard.edu (Diane Tang)
Subject: TLI library for linux?
Date: Fri, 22 Jul 1994 15:42:40 GMT

I'm looking for a TLI library for linux (libtli.a or something like
that), and I was wondering:
a. does it exist
b. if so, where can I get it from?

I plan on developing some communications software using TLI, and
I would desperately like to avoid having to do testing on DOS/Windows and 
Novell NETWARE, so if this exists for linux, I would be most pleased.

If you could e-mail your responses to dtang@das.harvard.edu, I would
be most appreciative.

Thanks!
--Diane

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

From: jdege@winternet.com (Jeff Dege)
Subject: Re: GOTO haters ..Re: Linux Performance Enhance ?
Date: 22 Jul 1994 13:36:51 GMT

Bill C. Riemers (bcr@k9.via.term.none) wrote:

: Not that I hate goto's, but in general I would consider both of these
: examples bad programming?  Why?  Because a goto loop, or a while()
: loop should only be used when you are willing to face the possability
: of an infinite loop.  (Very rarely the case.)  Otherwise you should
: use a for() loop (or the equivlent of it) so you have a counter to 
: insure even if something in your code is broken, eventually the
: program will exit the loop.

:                               Bill

   Jon Bentley's book ``Programming Pearls'' (or was it the sequel, ``More
Programming Pearls''?) has a discussion of loop-invariants that is must
reading for anyone who has these sort of vague concerns about whether his
loops will terminate.  Personally, I was more likely to have a loop iterate
one too many or one too few times, but it's all the same sort of thing.
There are methods for determining that a loop is written such that it
will do what you want it to do, and Bentley's explanation is one of the
clearest.

===========================
.sig under construction

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

From: bj0rn@blox.se (Bjorn Ekwall)
Subject: Re: loadable modules: undefined interuptible_sleep_on
Date: 22 Jul 94 15:54:18 GMT

RAINER SCHIELE INFORMATIK (81264@novell1.rz.fht-mannheim.de) wrote:
 > Hello

 > Yesterday i played with the loadable modules and write my first pseudo-
 > device driver. When i used interruptible_sleep_on i my driver and trying to 
 > load the module, i get an undefined interruptible_sleep_on. The 
 > wake_up_interruptible symbol is found. 

 > Then i going to kernel/ksyms.c and add the bad symbols. Now my Question is, 
 > can i do it that way (It works, for now :->) or must i do anything else and 
 > why aren't that symbols not defined !

Yup, that's all there is to it!
Since noone had had any use for the one you were missing, it wasn't
exported.  Although not all kernel symbols should be exported, really...

Someday there might be a concencus on what symbols should be exported
and what symbols should not, but until then...

 > Thanks Rainer

Bjorn Ekwall == bj0rn@blox.se

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

From: wish@dumain.demon.co.uk (Bill Hay)
Subject: Re: Starting New DIP
Date: Mon, 18 Jul 1994 22:41:55 +0000

Paul Cadach (paul@paul.east.alma-ata.su) wrote:

> I did make some changes in dip-3.3.7 for auto-redialing when carrier is
> dropped. It very simple: I checking for zero return of 'read' function in
> tty_getc procedure. The kernel changing tty-function's table when modem
> switching CD from high to low (drop carrier), and the 'read' function in
> 'hangupped' table will always return 'zero'.

> If it interested for somebody I can put sources of dip-3.3.7 with my
> changes on mailserv@east.alma-ata.su.

I'd be very interested in your changes.


--
Bill Hay

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


** FOR YOUR REFERENCE **

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

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

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

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

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

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