Subject: Linux-Development Digest #926
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:     Mon, 18 Jul 94 08:13:05 EDT

Linux-Development Digest #926, Volume #1         Mon, 18 Jul 94 08:13:05 EDT

Contents:
  Re: How to know the patch lev (Robert Adams)
  Re: obsolete routing request (Eckard Kopatzki)
  SCSI speed-ups? (Roberts Simon J)
  Fix for Versa XFree... but... (KromeKing)
  Re: FreePort functional equivalent for Linux? (Phil Hughes)
  Re: Fix for Versa XFree... but... (Marc La France)
  changes between 1.1.5 and 1.1.10 wanted (Dragon Slayer)
  Reschedule of the SVNET meeting for 386BSD, Release 1.0 (Jesus Monroy Jr)
  Re: Returning free'd memory (Hugh Emberson)
  SUP for Linux? (Russell Street)
  Re: obsolete routing request (Mitchum DSouza)
  link() system call (A N Burton)
  Re: Speaker Driver Uploaded to TSX-11 (Michael Beck)

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

From: robert@oblique.ksmith.com (Robert Adams)
Subject: Re: How to know the patch lev
Date: Sun, 17 Jul 1994 04:18:18 GMT

David Dyer-Bennet (ddb@terrabit.mn.org) wrote:
: riku.saikkonen@compart.fi (Riku Saikkonen) writes:

: >>I want to know what patch level of 1.0 I am running. I tried "uname"
: >>it only tels me it is "1.0".
: >It's 1.0.0 then...
: When I type "uname" on my system, it just says "Linux".

uname -a                  gives all info
uname -m                  gives hardware type     <i386 i486>
uname -n                  network hostname
uname -r                  release version number  <linux kernel version>
uname -s                  operating system name   <same as no args>
uname -v                  operating system version

This is the option list.  type uname -r  to get the version

Robert
-- 
Robert J. Adams                                         Oblique Enterprises 
robert@oblique.ksmith.com                      who wants dos? windoze? os/2 
adamsr@hamlet.uncg.edu                                   when one has UNIX? 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

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

From: root@stevie.isar.muc.de (Eckard Kopatzki)
Subject: Re: obsolete routing request
Date: 17 Jul 1994 15:37:58 GMT

In article <30abmj$ask@nyx10.cs.du.edu>, kkoller@nyx10.cs.du.edu (captain sarcastic) writes:
> 
> Oops, it's happening.
> 
> I knew that if I kept upgrading the kernel, eventually I'd have to 
> upgrade something else.  syslogd (i think) is breaking with a message 
> "obsolete routing request" when I first boot.
> 
> Is it syslogd?
> 
> How do I fix?
> 
> What's the best way to keep the newest versions of system software, in 
> addition to the kernel?
> 

Kernel routing tables have changed somewhere around 1.1.20, so your ifconfig
and route commands are outdated. Get the new net-0.32d-NET3 (or so ...). It's
somewhere in the sunacm directory tree on sunsite.

-- 
Eckard Kopatzki           Internet eko@isar.muc.de
Therese-Giehse-Allee 53     CompuServe 100024,2175
D-81739 Muenchen, Germany     Voice +49-89-6378103

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

From: robesj3@csc.canterbury.ac.nz (Roberts Simon J)
Subject: SCSI speed-ups?
Date: 18 Jul 1994 00:16:49 GMT

Just installed Slackware 1.2 - and now I hear rumours of massive speed-ups
for SCSI (1542B in particular) under kernel 1.1.xx.  Is this true?  What sort
of improvements?  Does 1.0.9 (slackware 2.0) have these improvements?

Simon



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

From: raunn@stat.tamu.edu (KromeKing)
Subject: Fix for Versa XFree... but...
Date: 18 Jul 1994 02:41:29 GMT

I hassled the NEC people into giving me the assembly code that
switches video-modes on their Versa laptops(Fn-F3), effectively
providing a hard-reset to the chips, which is necessary for restoring
the correct video mode on the wd90c24 when exiting XFree.

Unfortunately, I am absolutely clueless when it comes to assembly,
especially under linux. This is the code:

switch between LCD                      mov ax,7F7Fh
and simultaneous mode                   mov bx,4104h
                                        int 10h

switch to CRT-only mode                 mov ax,7F7Fh
                                        mov bx,4105h
                                        int 10h

switch to LCD-only mode                 mov ax,7F7Fh
                                        mov bx,4106h
                                        int 10h

Can someone help me implement these instructions in Linux? This will
fix the Versa series laptop's problem with XFree.

Replies by email welcomed.
--
||"I slept with Faith, and found a corpse  ||   -=/> Raun D. Nohavitza </=-
|| in my arms on awaking; I drank and      ||        irc:   KromeKing
|| danced all night with Doubt, and found  ||      email: raunn@tamu.edu
|| her a virgin in the morning."-A. Crowley||       G.C.v1.0.1 follows:

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

Crossposted-To: comp.os.linux.admin
From: fyl@eskimo.com (Phil Hughes)
Subject: Re: FreePort functional equivalent for Linux?
Date: Sun, 17 Jul 1994 16:24:12 GMT

Richard Whittaker (rwhittak@orion.docwhitehorse.doc.ca) wrote:
: Has anyone out there got something functionally equivalent to FreePort? I'd
: like to setup a freenet, and have all of the necesarry resources the I need
: (except cold hard cash).. I'd like to try setting such a free-net beast up
: on a Linux box, and would like to know if someone has a less expensive option
: out there, but one that retains the functionality (NetNews, E-Mail, Account
: Generation, etc...) of the Freeport package...

I have this idea on the back burner.  I was involved in SCN, a project to
create a freenet in Seattle.  Too many decisions happened that I disagreed
with and I have dropped out.  But I know a lot about FreePort and would
like to see a real "Free"net created.  Here is my take on the situation.

The current FreePort software has three serious defects:
  - It isn't free
  - It is a total performance dog
  - The user interface isn't configurable to any reasonable extent
    (What I mean here is that there is no well-defined interface layer
    where you could build a graphical connection or one for vision
    inpaired or whatever.)

The fact that isn't freely redistributable is enough to scratch it from my
list.  The whole idea of Freenets just begs for free software.  But it
also means that development isn't in the open and you end up with the
other two problems.

The performance dog part of it is because it is a bunch of scripts and
programs that talk to each other and to Unix utilities to accomplish
tasks.  A perfectly good way to prototype a system but certainly not what
you want for a maximum performance/minimum cost system that you want to
have replicated around the world.  It also means that the interface
changes when you do various things For example, help is displayed using
more but the way you get out of more (and therefore help) is not the same
as the way you return up a menu level all the rest of the time.  About 8
years ago I wrote a simple BBS using C-shell scripts and Unix commands.
It was done to offer a particular set of info on line, it was a dog but it
worked.  And I write it in a weekend.  The current FreePort software
reminds me of that BBS.

There is an alternative.  A system called TSO (Tri-State Online) is
implemented from scratch using the same user interface model as FreePort.
It is fast.  Numbers like 100 users on a Sun IPX (IPC?) were what they
were talking.  Ten users on FreePort on the same hardware was more like
it.  TSO also offers a user interface with a well-defined layer so you can
expand it.  It looks the same as FreePort expect the pager is integrated
into the other software so it always works the same and there are no
occasional Unix error messages.  But it costs a lot of money.

Once I saw that SCN was falling for the get bad software and pump a lot of
cpu cycles at it I saw it as a dead end for me.  The idea of a Freenet in
Seattle is nice but, to me, the goal was to make a Freenet system that
anyone could put together for a minimum cost anywhere.  (Guess that's
think globally, act locally.)

I have a lot of info (like the TSO manual) some design ideas, some test
code (if I can still find it and a desire to see this sort of project
happen.  But I don't have the time.  I went off and started Linux Journal
because I thought it had more potential to make things happen.  And, based
on the response, I think I am right.  But, I still want a read "FREE"net
to happen and I think Linux is where it should happen.

So, the offer.  I am willing to use Linux Journal to help promote the
idea.  Clearly much of the design discussion would be on the Internet but
adding coverage in Linux Journal could get people not yet on the Internet
to decide to get involved, get feedback from those who would probably just
be a user as to design decisions and generally get more exposure for the
whole idea.  I would love to put time into the project but can't really
see that as possible in the short term.  But, if something starts to
happen or someone wants to coordinate an effort to get the ball rolling,
please keep me informed.  I think it is important and will support it in
any way I can.

To contact me about this (or any Linux Journal related stuff that I need
to hear about) send e-mail to phil@fylz.com.  (General Linux Journal
questions should be sent to linux@ssc.com, subscription questions to
subs@ssc.com)
--
Phil Hughes, Publisher, Linux Journal (206) 524-8338
usually phil@fylz.com, sometimes fyl@eskimo.com

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

From: tsi@gpu.srv.ualberta.ca (Marc La France)
Subject: Re: Fix for Versa XFree... but...
Date: 18 Jul 1994 04:06:57 GMT

In article <30cq4p$3bt@news.tamu.edu>,
 KromeKing (raunn@stat.tamu.edu) wrote:
> I hassled the NEC people into giving me the assembly code that
> switches video-modes on their Versa laptops(Fn-F3), effectively
> providing a hard-reset to the chips, which is necessary for restoring
> the correct video mode on the wd90c24 when exiting XFree.

> Unfortunately, I am absolutely clueless when it comes to assembly,
> especially under linux. This is the code:

> switch between LCD                    mov ax,7F7Fh
> and simultaneous mode                 mov bx,4104h
>                                       int 10h

> switch to CRT-only mode                       mov ax,7F7Fh
>                                       mov bx,4105h
>                                       int 10h

> switch to LCD-only mode                       mov ax,7F7Fh
>                                       mov bx,4106h
>                                       int 10h

> Can someone help me implement these instructions in Linux? This will
> fix the Versa series laptop's problem with XFree.

GOOD Luck!  Those are BIOS calls...

+----------------------------------+-----------------------------------+
|  Marc Aurele La France           |  work:   1-403-492-9310           |
|  Computing and Network Services  |  fax:    1-403-492-1729           |
|  352 General Services Building   |  email:  tsi@gpu.srv.ualberta.ca  |
|  University of Alberta           +-----------------------------------+
|  Edmonton, Alberta               |                                   |
|  T6G 2H1                         |     Standard disclaimers apply    |
|  CANADA                          |                                   |
+----------------------------------+-----------------------------------+

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

From: ds2@dragonslair.caltech.edu (Dragon Slayer)
Subject: changes between 1.1.5 and 1.1.10 wanted
Date: 18 Jul 1994 05:43:55 GMT

Can someone please send me (or post) a summary of the changes that occurred
between versions 1.1.5 and 1.1.10? I am trying to find out why the sound 
driver stopped working in the newer version. I have tried everything from 
pulling all non-essential card to reconfiguring the kernel. My problem is 
whenever I cat a .au file (i.e. cat file.au > /dev/audio), all I get is 
a buzz; and whenever I try to record, I still get a buzz even w/o an
input. I am guessing it is some sort of DMA/IRQ thing, but knowing what 
changed will help. If anyone wants to offer suggestions, I will be glad 
to hear them (and provide needed info).

Thanks in advance.

-- 
Dragon's Lair, USA -- Dragon Slayer -- +1.818.395.1406
<a href="http://dragonslair.caltech.edu/~ds2/">ds2@dragonslair.caltech.edu</a>

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

Crossposted-To: comp.os.os2.programmer.misc,comp.os.minix,comp.os.mach,comp.unix.bsd,comp.unix.pc-clone.32bit,comp.os.386bsd.development
From: jmonroy@netcom.com (Jesus Monroy Jr)
Subject: Reschedule of the SVNET meeting for 386BSD, Release 1.0
Date: Mon, 18 Jul 1994 07:29:47 GMT

 
 
     /===========================================================\
     |                                                           |
     |  Reschedule of the SVNET meeting for 386BSD, Release 1.0  |
     |                                                           |
     \===========================================================/
 
 
        The meeting scheduled by SVNET for Wenesday - July 20, 1994
        has been tenatively rescheduled for August 17, 1994.
        The Reason the reschedule is due to the delayed release
        of 386bsd.
 
 
        For those of you in need of you BSD fix here is a quote from
        a letter to those waiting for the CD-ROM from DR. Dobbs:
 
 
        Dear John:
 
        Thank you for your recent order for the 386BSD Refernce
        CD-ROM!     We'd like to keep you abreast of the status of
        your order.
 
        The 386BSD Refernce CD-ROM is still in final development at
        this time.   The authors, William and Lynne Jolitz, are taking
        some additional development time in order to include even more
        annotations to the kernel source.  Plus, they are also
        including selected topics from their forthcoming book "386BSD
        From The Inside Out" - certainly a bonus feature that is not in
        the original product specs!  So, while we have held off on
        putting the putting the product to bed, we hope you'll
        understand that we are taking the time to make what is already
        an outstanding product even more useful and valuable for you.
 
        We anticipate shipping the 386BSD Refernce CD-ROM in early
        September, 1994.  Be assured that your credit card or check
        will not be chared or cashed until we ship your CD-ROM.
 
        If you have any question, please call me... 
 
 
                Thank you for your patience!
 
                Sincerely,
                Stan Barnes
                Manager-special Projects
 

-- 
Jesus Monroy Jr                                          jmonroy@netcom.com
Zebra Research
/386BSD/device-drivers /fd /qic /clock /documentation
___________________________________________________________________________

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

From: hugh@hugh.cosc.canterbury.ac.nz (Hugh Emberson)
Subject: Re: Returning free'd memory
Date: 18 Jul 1994 09:22:09 GMT


In article <1994Jul15.025533.8154@unlv.edu> ftlofaro@unlv.edu (Frank Lofaro) writes:
   [.. munch ..]

           How hard would it be to allow such a kernel syscall:
   unmap_from_process_memory_space(begin_addr, end_addr)

   or something similar. What about mprotect() while we are at it?

           I've been thinking of hacking on both, but I don't know _THAT_
   much about Linux MM yet (any one got some info?)

No kernel hacking is necessary.  I believe that GNU malloc (gmalloc)
can be hacked to return memory to the system.  Actually gmalloc does
try to return blocks to the system now, but only if the blocks are up
against the heap limit (i.e. end_of_block == sbrk(0)).

My hack would go something like this:

   Instead of sbrk, use mmap("/dev/zero") to allocate blocks (gmalloc
   uses page size blocks for its internal allocation and splits these
   into smaller fragments as necessary, things larger than 1/2 page
   size get 1 or more blocks to themselves.

   When free determines that a block can be returned to the system
   (disregarding the end_of_block == sbrk(0) condition), unmap it.

Of course it is a bit harder than this.  Gmalloc has a table for
storing per block information in called _heapinfo which is indexed by
block number (the number of blocks from the start of the heap),
in order to keep this table small it would be a good idea to make the
mmap allocate blocks contiguously (across calls to mmap) and to force
mmap to "fill in the gaps" when blocks have been unmapped i.e. if the
heap looks like this:

        #########....#######                    # = mapped block
                 ^                              . = unmapped block
                 1          

then it would be best to force mmap to map (using MAP_FIXED) in pages
at 1 rather than at the end of the heap if possible.  So it would be a
good idea to maintain a list of unmapped pages, I think this could be
added into the _heapinfo table.

I'd like to do this, I'm pretty sure it's possible, but I have a
thesis to finish and several other things to do.  If anyone else would
like to do it, that would be neat.

Actually, what we really need is a compacting garbage collector :-)

Cheers,
        Hugh











        
-- 
hugh@cosc.canterbury.ac.nz
hugh@hugh.cosc.canterbury.ac.nz

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

From: russells@ccu1.auckland.ac.nz (Russell Street)
Subject: SUP for Linux?
Date: 18 Jul 1994 08:23:33 GMT
Reply-To: r.street@auckland.ac.nz (Russell Street)

Has anyone ported `SUP' (the Software Update Protocol, used to
distribute FreeBSD et al) to Linux?

If so, where so?

TIA
Russell

==============================================================
Russell Street (r.street@auckland.ac.nz)
        "If there is anything more important than my ego around
        here I want it caught and shot -- now!"


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

From: Mitchum DSouza <m.dsouza@mrc-apu.cam.ac.uk>
Subject: Re: obsolete routing request
Date: 18 Jul 1994 06:36:11 -0400
Reply-To: m.dsouza@mrc-apu.cam.ac.uk


kkoller@nyx10.cs.du.edu (captain sarcastic) says:
| Oops, it's happening.
| 
| I knew that if I kept upgrading the kernel, eventually I'd have to 
| upgrade something else.  syslogd (i think) is breaking with a message 
| "obsolete routing request" when I first boot.
| 
| Is it syslogd?
| 
| How do I fix?
| 
| What's the best way to keep the newest versions of system software, in 
| addition to the kernel?
| 

get the new net tools from

     sunacm.swan.ac.uk:
        /pub/misc/Linux/Networking/PROGRAMS/NetTools/net-tools-1.1.29.tar.gz

Mitch

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

From: anb@doc.ic.ac.uk (A N Burton)
Subject: link() system call
Date: 18 Jul 1994 11:49:14 GMT

Would anyone know if any of the file systems available under linux allow
'root' to create _HARD_ links to directories?

Is there any documentation describing the working of the various file systems?
(E.g., minix, ext, ext2 etc..)


Replies by  e-mail,

Ariel BURTON
anb@doc.ic.ac.uk

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

From: beck@informatik.hu-berlin.de (Michael Beck)
Subject: Re: Speaker Driver Uploaded to TSX-11
Date: 18 Jul 1994 13:51:42 +0200

In article <Ct3uxD.Kyr@sed.csc.com>, jruschme@sed.csc.com (John Ruschmeyer) writes:
|> Since several people have asked about this...
|> 
|> This weekend I installed the PC speaker driver from tsx-11 into my
|> 1.1.29 system.  Because, as several people have mentioned, the driver
|> was written for 0.99.12, I needed to install the patches by hand.  So,
|> to make it easier for everyone else, I have done the following:
|> 
|> 1) Reorganized the sources to fit the current kernel sources hierarchy.
|> 
|> 2) Made them user-selectable by Configure.
|> 
|> 3) Rolled a new set of patches against 1.1.29.  (Probably will work with
|>    1.0.9, but I haven't tested.)
|> 
|> 4) Uploaded the whole thing to the /Incoming directory on tsx-11.mit.edu.
|> 
|> I've also notified the admin, so it should be moved to the correct place
|> shortly.
|> 
|> <<<John>>>

Sorry, but it seems you have upgraded a very old version. The actual
version is 0.6 with all the features above and some more. It can be found
on sunsite (I think) and at ftp.informatik.hu-berlin.de:pub/os/linux/hu-sound
Version 0.6 works on 1.0.x and 1.1.x kernels. The next version 0.7 will
be released soon. I really suggest you to install the current version,
because versions <0.6 may crash a slow machine.





-- 

  Michael Beck               beck@informatik.hu-berlin.de

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


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