Subject: Linux-Development Digest #983
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:     Thu, 4 Aug 94 06:13:04 EDT

Linux-Development Digest #983, Volume #1          Thu, 4 Aug 94 06:13:04 EDT

Contents:
  Re: SLIP, CSLIP, PPP and modems (Starcon SysAdmin)
  Kernel change summary 1.1.21 -> 1.1.22 (Russell Nelson)
  Kernel change summary 1.1.24 -> 1.1.25 (Russell Nelson)
  Kernel change summary 1.1.25 -> 1.1.26 (Russell Nelson)
  Kernel change summary 1.1.26 -> 1.1.27 (Russell Nelson)
  SCSI scanner support ? (Mark K Vallevand)
  Token Ring Driver Status? (Jason Zarin)
  RPC: where is authunix_create_default() ? (Pat Niemeyer)
  Re: Linux backup of MSDOS? (Salvador Pinto Abreu)
  auto irq problems after 1.1.16 (Ted Deppner)
  Re: 1.1.38 just hangs on my P60/PCI (James W. Smith)
  Re: gcc and Linux (compiling)... (Sunny Yum)
  Re: 1.1.38 just hangs on my P60/PCI (Laurent Chemla)
  Re: Using rs232 control lines from program? (Lynx man)
  gcc 2.6.0 produces many errors (Felix v.Leitner)
  Help needed compiling SLIPLOGIN (Bart Kindt)
  gcc 2.6.0 intern. comp. error (Klaus ZLOEBL)
  Re: -= good programmer's editor for X? (David Fox)

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

From: yuriev@astro.ocis.temple.edu (Starcon SysAdmin)
Subject: Re: SLIP, CSLIP, PPP and modems
Date: 3 Aug 1994 21:53:07 GMT

Jay Denebeim P025 (denebeim@bnr.ca) wrote:
: In article <31jcqs$9dc@cronkite.ocis.temple.edu> yuriev@astro.ocis.temple.edu (Starcon SysAdmin) writes:
: >In this case can you please explain results of the following file transfers:
: >
: >Linux <--> Linux using Hydra on 1.7Mb & 1.4Mb ZIP files: total time 17 min
: >34 sec 
: >
: >Linux <--> Linux via SLIP (ftp transfer). Total time: 28 min 48 sec. 
: Alex, you're comparing apples and oranges.  A network is not the same
: thing as a file transfer. 

But how could it increase time by factor 2? 

: SLIP is used to send arbitrary packets over
: the phone line onto a network.  Its not limited to sending files.
: Because of this, there is quite a bit more information that is
: required for each packet then a file transfer takes.  Because of this,
: it is definately going to be slower.
: It probably shouldn't be a factor of two though.  Are you sure you
: wern't using the link for something else at the same time?

I'm positive. I took 2 Linux machines, disconnected them both from
Internet, recompiled kernels 1.0.8, 1.0.8, 1.1.33 without everything but
SLIP support and tried it. Results were approximately the same for all
kernels i tried. Both machines are running on i486DX-33 with 8 Mb RAM and
were tested with

a) US Robotics Sportsters 14400
b) Zoom 14400
c) Calpak 2400
d) DataSystemXE 14400

All results in this group are aprox. the same: slip is twice slower.

e) US Robotics HST DS (HST connect)

SLIP is aprox 5-7% slower than filetransfer. Which is OK, considering time
HST needs to turn around.

Best wishes,
Alex Yuriev

yuriev@astro.ocis.temple.edu  &  root@starcon.main.temple.edu

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

From: nelson@crynwr.crynwr.com (Russell Nelson)
Subject: Kernel change summary 1.1.21 -> 1.1.22
Date: 03 Aug 1994 06:19:20 GMT

Kernel configuration for network drivers greatly revamped.
Back off the "Keep track of the number of hard drives correctly."
        change from 1.1.21 because it doesn't work for everybody, but
        helps IBM ValuePoints.
Add new flags to the VT-100 emulator, one to display control chars,
        and another to toggle the high bit.
Select no control and no high bits on CSI m 10
Select control and no high bits on CSI m 11
Select control and high bits on CSI m 12
Select underscore on CSI m 38
Select no underscore on CSI m 39
Added 3c505 driver.
Added Ansel Communications EISA ethernet driver.
Added support for the EISA DEPCA, DE422.
Added Cabletron E2100 ethernet driver.
Added Zenith Z-Note ethernet driver.
Label for {"MAXTOR","XT-8760S","B6B"} added.
Added TCP Window clamping.
In the earlier routine added to verify that an area is in memory,
        don't verify if the user space has been mapped to kernel space.
Add de620 to NET_MODULES list.
Set the socket state to CONNECTED even if we've only sent a SYN.
Don't print the "...forwarding disabled!" error.
Autobind added to ipx code.
 *              Alan Cox        :       TCP fast path debugging
 *              Alan Cox        :       Window clamping
 *              Michael Riepe   :       Bug in tcp_check()

Process the last segment of data if the packet contained a FIN (which
        it used to do at 1.1.13, but somewhere between 1.1.13 and
        1.1.21, someone botched it).
--
-russ <nelson@crynwr.com>    http://www.crynwr.com/crynwr/nelson.html
Crynwr Software   | Crynwr Software sells packet driver support | ask4 PGP key
11 Grant St.      | +1 315 268 1925 (9201 FAX)  | What is thee doing about it?
Potsdam, NY 13676 | LPF member - ask me about the harm software patents do.

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

From: nelson@crynwr.crynwr.com (Russell Nelson)
Subject: Kernel change summary 1.1.24 -> 1.1.25
Date: 03 Aug 1994 15:45:12 GMT

Put bootable kernel into /boot instead of /.
Added support for Sound FX as host for Sony CD-ROM.
Apparently, there is a bug in the floppy FIFO that requires you to
        force the FIFO to accept a byte even though it insists it
        isn't ready.
Even though the dates are old, these changes are apparently new to 1.1.25:
        October 1993 Bernd Huebner and Ruediger Helsch, Unifix Software GmbH,
        Braunschweig, Germany: Total rework to speed up data read operation.
        Also enabled definition of irq and address from bootstrap, using the
        environment. linux/init/main.c must be patched to export the env.
        November 93 added code for FX001 S,D (single & double speed).
        February 94 added code for broken M 5/6 series of 16-bit single speed.
ramdisk load code moved into its own subroutine, so that we can
        gram(sic) the floppy IRQ/DMA.
3c505 driver changes: renamed to elplus, hard reset needed on a warm
        reset from DOS after running some drivers, autodetect/autoirq
        added
HP LANcard+ drivers added to autoprobe list.
ac32000.c renamed to ac3200.c.
Added Schneider & Koch G16 Ethernet driver.
SLIP and CSLIP fixes for memory leaks.
znote.c renamed to znet.c.

SCSI change notes from Eric Youngdale
* sd.c, sr.c: Use new check_media_change and revalidate
        file_operations fields.
* st.c, st.h: Add changes from Kai Makisara, dated Jun 22.
* hosts.h: Change SG_ALL back to 0xff.  Apparently soft error in
        /dev/brain resulted in having this bumped up.  Change first
        parameter in bios_param function to be Disk * instead of index
        into rscsi_disks.
* sd_ioctl.c: Pass pointer to rscsi_disks element instead of index to
        array.
* sd.h: Add struct name "scsi_disk" to typedef for Scsi_Disk.
* scsi.c: Remove redundant Maxtor XT8760S from blacklist.  In
        scsi_reset, add printk when DEBUG defined.
* All low level drivers: Modify definitions of bios_param in
        appropriate way.

When flushing filesystem buffers, reset b_req also.
Finish adding media_change calls.
Allow locking on non-regular files.
If they were trying to unlock, and it worked, don't return an error.
An NFS xid mismatch no longer causes the request to be trashed.
When a socket dies, set the dead flag.
When returning an ICMP error, return the IP address corresponding to
        the interface.
Respond to broadcast ICMP_ECHO (ping) requests.
--
-russ <nelson@crynwr.com>    http://www.crynwr.com/crynwr/nelson.html
Crynwr Software   | Crynwr Software sells packet driver support | ask4 PGP key
11 Grant St.      | +1 315 268 1925 (9201 FAX)  | What is thee doing about it?
Potsdam, NY 13676 | LPF member - ask me about the harm software patents do.

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

From: nelson@crynwr.crynwr.com (Russell Nelson)
Subject: Kernel change summary 1.1.25 -> 1.1.26
Date: 03 Aug 1994 16:02:55 GMT

In config.in, don't include AT1500 support by default, and do include
        NE2000 support by default.
Switch to release 2.2 of SoundBlaster Pro CD-ROM driver.  Include
        better support for multiple Matsushita controllers.  Include
        media_change changes.
Check for real tty in set_termios and open line discipline calls.
If the serial driver couldn't get a transmit buffer from the kernel,
        it returned the positive error number instead of the correct
        negative error.
No need to busy the tty if we're delaying the close.
Existance test for serial ports improved.
Now can add serial ports at run-time, to support PCMCIA modems.
Make sure that line discipline changes are run past set_ldisc().
Make sure ptys have the usage count on the other tty set correctly.
Make sure the shared memory stuff is known by the general VM manager.
--
-russ <nelson@crynwr.com>    http://www.crynwr.com/crynwr/nelson.html
Crynwr Software   | Crynwr Software sells packet driver support | ask4 PGP key
11 Grant St.      | +1 315 268 1925 (9201 FAX)  | What is thee doing about it?
Potsdam, NY 13676 | LPF member - ask me about the harm software patents do.

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

From: nelson@crynwr.crynwr.com (Russell Nelson)
Subject: Kernel change summary 1.1.26 -> 1.1.27
Date: 03 Aug 1994 16:08:35 GMT

3c509 driver starts to support multiple cards.
If there wasn't a check_media_change for a fs driver, you'd get spurious
        disk change messages from the kernel.
NFS file IO buffer size quadrupled.
Fixed the ICMP error status of net/host unreachable
Fixed broadcast ping properly
--
-russ <nelson@crynwr.com>    http://www.crynwr.com/crynwr/nelson.html
Crynwr Software   | Crynwr Software sells packet driver support | ask4 PGP key
11 Grant St.      | +1 315 268 1925 (9201 FAX)  | What is thee doing about it?
Potsdam, NY 13676 | LPF member - ask me about the harm software patents do.

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

From: root@markus.rsvl.unisys.com (Mark K Vallevand)
Subject: SCSI scanner support ?
Date: Wed, 3 Aug 1994 03:21:29 GMT

Is there any support for SCSI scanners?  Drivers, apps?  Inquiring minds, etc.

-- 
Regards,                           Posted from home: mkv@ernie.rsvl.unisys.com
Mark K Vallevand.                              work: markus@rsvl.unisys.com
"God not only plays dice, He somethimes throws them where they can't be seen."
  Stephan W Hawking
--
Regards,                           Posted from home: mkv@ernie.rsvl.unisys.com
Mark K Vallevand.                              work: markus@rsvl.unisys.com
"God not only plays dice, He somethimes throws them where they can't be seen."
  Stephan W Hawking

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

From: jzarin@nyx10.cs.du.edu (Jason Zarin)
Subject: Token Ring Driver Status?
Date: 3 Aug 1994 16:03:12 -0600


My office uses token ring, and I really want to run linux here on my
machine.  Is there any active development going on to support token
ring, and if so, what is its status?

Thanks.

-- 
:::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::::
::      Jason Zarin     :: zarin@econ.sscnet.ucla.edu                      ::
:: Grad Student at UCLA :: "To an economist, real life is a special case." ::
::::::::::::::::finger jzarin@nyx.cs.du.edu for PGP public key:::::::::::::::

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

From: pat@sbctri.sbc.com (Pat Niemeyer)
Subject: RPC: where is authunix_create_default() ?
Date: 2 Aug 94 22:03:52 GMT
Reply-To: pat@sbctri.sbc.com


I can't seem to find authunix_create_default()...

It doesn't seem to be anywhere in the rpc-0.9 stuff or with the kernel
stuff...



Thanks,
Pat

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

From: spa@fct.unl.pt (Salvador Pinto Abreu)
Subject: Re: Linux backup of MSDOS?
Date: Tue, 2 Aug 1994 21:50:28 GMT

In article <31l7u7$ve@sun.cais.com> toehser@cais.cais.com (Tom Oehser) writes:

> >Tom Oehser <toehser@cais2.cais.com> wrote:
>
> >>I am trying "dd if=/dev/hda | gzip --fast | dd of=/dev/nrmt0"
>
> >What is the advantage of this over gzip </dev/hda >/dev/nrmt0 ?
>
> Is there one?  I am just ignorant and *nix is my newest operating system.
> I knew that some programs didn't like the raw devices, so I used dd for them.

if your filesystems are not completely full; you may want to try this
before compressing (either form):

        for FS in $FILESYSTEMS; do
          cp /dev/zero $FS/.zero.
          rm $FS/.zero.
        done

the idea is to fill the unused space with zeros so that gzip will do a
better job.

../salvador
--

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

From: tdeppne@eis.calstate.edu (Ted Deppner)
Subject: auto irq problems after 1.1.16
Date: 3 Aug 1994 09:41:05 -0700

Greetings all.  I've not seen any mention of this problem in the last 
month, so I'll post an article about it.  If it's already been hashed, be 
merciful with me.

I've a 486dx2-66 AMD machine, Trident 8900c, 420 and 345meg IDE, 16meg 
memory, SoundBlaster 16 irq 5 dma 1,6 port 220, Ne2000 clone port 300, 
irq 15, and 3.5", 5-1/4" floppies.

with kernel 1.1.16 all the above is autodetected and all irqs are 
correct.  I recompiled the kernel of course with support for only the 
above SB16 and Ne2000 cards.

the next kernel I used was 1.1.30, then 1.1.35, then 1.1.37.  Something 
about these kernels doesn't do correct IRQ detection.  the SB16 comes up 
okay, but the 30,35,37 kernels think my NE2000 is on IRQ 6, the floppy 
interrupt.  Hence, Ne2000 doens't work, and the floppies fail in a polite, 
yet complete fashion.

I didn't see any change in the autoirq procedures in drivers/net/misc.c, 
and there are no significant changes in drivers/net/ne2.c that I could 
find regarding IRQs, so I'm at a loss as to why they fail.

My machine configuration has not changed at all, and I can still boot 
1.1.37 and have it fail, then next boot bring up 1.1.16 (alah lilo) and 
have everything work correctly.  Same kernel config options.

I've been messing with kernels since .96 so I don't think I'm missing 
something obvious, but I suppose I could be.  Of course I have hard coded 
the correct interupt in ne2.c, but I'd be happy if I didn't have to.

Any ideas?  I'd be happy to test this out further if needed.


---
Ted Deppner
    tdeppne@eis.calstate.edu

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

From: zjws08@trc.amoco.com (James W. Smith)
Subject: Re: 1.1.38 just hangs on my P60/PCI
Reply-To: zjws08@trc.amoco.com
Date: Wed, 3 Aug 1994 17:35:39 GMT

In article 7tg@sheckley.cnam.fr, chemla@cnam.cnam.fr (Laurent Chemla) writes:
>
>
>--
>The title says it all..
>I just get the message saying 'Now booting the kernel' and nothing
>else happens. It's the first time since 0.96 Linux just doesn't work
>at all on one of my machines :-)
>
>Configuration:
>Pentium P60/PCI (Intel motherboard),
>Adaptec 1542B
>Sound Blaster CDROM
>
>Any hints ?

me too.
I thought I was having a problem with the ncr53c810 alpha scsi driver. Perhaps its
 the pentium.  The floppy drive light stays on too, implying a hang. I tried generating
 a new boot disk in case it was a bad sector - no luck.

Configuration:
Pentium P60/PCI (Intel)
Nexstor ncr53c810 scsi with 1 gig seagate and plexor CDROM
Matrox video.






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

From: n5298@cray.com (Sunny Yum)
Subject: Re: gcc and Linux (compiling)...
Date: Wed, 3 Aug 1994 21:20:38 GMT

In article <31o0mr$qao@news.u.washington.edu>,
Tim Smith <tzs@u.washington.edu> wrote:
>Urs Thuermann <thuerman@ibr.cs.tu-bs.de> wrote:
>>No, if it does, it is broken.  Functions must always have a name, and
>>variable declarations always need at least one type specifier, storage
>>class specifier or type modifier.
>
>Variable declarations don't always need one of those, at least explicitly.
>For example:
>
>       a;
>       main()
>       {
>               a = 1;
>       }
>
>is OK.  The first line declares "a" as an int.


Now that is _gross_.  Anyhow, in that case, I wouldn't tell you that
the compiler was broken, but that the _programmer_ was!

-- 
Sunny D. Yum (n5298@marvin.cray.com) -- My Opinions Are MINE!
GCS d-- H-- s g+(-) p? au->+++ a- w+ v+(*) C++++(++)$ UL++++ P->++++ L++(+++)
3- E>++ N++ K- W(---) M-- V-- -po+ Y+ t(-) 5++ j- R- G? tv(+) b++ D++(+++)
B(-) e+>++ u h!>++ f+ r n-(----) y?

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

From: chemla@cnam.cnam.fr (Laurent Chemla)
Subject: Re: 1.1.38 just hangs on my P60/PCI
Date: 3 Aug 1994 16:55:54 GMT

Mark Lord (mlord@bnr.ca) wrote:
: Sounds like what I see when I forget to run lilo after installing a new kernel.
: -- 
: mlord@bnr.ca  Mark Lord       BNR Ottawa,Canada       613-763-7482

--
Would be a great idea if I was using lilo, but I use bootlin.com..
Thanks anyway. Nice try.
--
Laurent Chemla : chemla@cnam.cnam.fr or laurent@brasil.frmug.fr.net
Brasil BBS  - +33 1 44 67 08 44 -  Atari France developpers support

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

From: ez008579@dale.ucdavis.edu (Lynx man)
Subject: Re: Using rs232 control lines from program?
Date: Wed, 3 Aug 1994 22:32:25 GMT

dlacroix@guilder.bevd.blacksburg.va.us (David La Croix) writes:

>I'd like to know how to program for the serial ports under LINUX.
>Specifically I'd like to be able to use the flow control lines to 
>send/detect a boolean case.

>any RTFMs are welcomed, but I'd like to know which Ms.

I have a similar (the same?) problem.  I want to migrate from QNX 2.x to
Linux.  Under QNX there's this convenient function:

    unsigned tty_ctrl(fp,mask,control)
    FILE *fp;
    unsigned mask;
    unsigned control;

Example usage:  force DTR and RTS lines off:

    new_state = tty_ctrl(fp, 0x03, 0x00);

These and similar functions are of course not Posix or System V or anything.
I've combed through the kernel source code, grep'ed the /usr/include/*/*.h files,
and haven't come up with any promising leads, but some clues in sys/serial.h and
linux/serial_reg.h.

I'd prefer not to purchase QNX 4.0 but I have a copy of its stty manual page.
It describes the +DTR, -DTR, +RTS, and -RTS options to the stty command.
It claims that its stty is Posix-compliant; no other Posix machine I've looked
at supported these options.

--
--dudley                                <dudley@water.ca.gov>

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

From: leitner@cs.tu-berlin.de (Felix v.Leitner)
Subject: gcc 2.6.0 produces many errors
Date: 3 Aug 1994 17:04:24 GMT

Hi !

I was wondering whether it's gcc's fault or the new kernel's. I have installed
1.1.29 and the old gcc (2.5.8) worked reliably.

Now that I have gcc 2.6.0 installed it fails with an internal error on
groff 1.09, it can compile xhdvi 0.5b, but xhdvi dumps core under certain
circumstances.

Do other people have difficulties with gcc 2.6.0, too ?

BTW: I compiled it myself, so maybe I did something wrong there.

Felix

PS: Yes, I know the FAQ, gcc errors belong to gnu.somewhere, but if I am
the only one I don't want to bother them and new gcc-2.5.x came every two
days and 2.6.1 is still not there.



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

Crossposted-To: comp.os.linux.help
From: bart@dunedin.es.co.nz (Bart Kindt)
Subject: Help needed compiling SLIPLOGIN
Date: Fri, 5 Aug 1994 05:07:06 GMT

Hi,

I am using Sliplogin for a multi-line dial-in server. I currently use the 
pre-complied binairy which I downlaoded from 
sunsite.unc.edu/pub/Linux/system/Serial

The program often (but not always) does not terminate after receiving a SIGHUP 
when a dialer disconnect, with the result the dial-in line is hanging forever.

I tried to compile the source, but run into trouble.

First, some #include files where in the wrong place; I found them and moved 
them in the directory Make was looking in.

Then the compiler gave the error:

.. -lshadow: no such file or directory   (or something like that). This was a 
compiler directive. I removed it, and I also removed the directive LIBS = 
-lbsd because this is not a BSD system... I replaced it with -DLINUX.  

Another compile; no errors anymore, but a Warning:

Warning: implicit declaration of function 'system'.

This refers to the 'system()' call which executes the Login and Logout script 
files, which set the ifconfig and route.

I had a look at the source of the *DIP* program, and added a handfull of 
#include files I found in that source.  This fixed the above Warning about the 
system call.  (:-)

However, when I tried the newly compiled sliplogin, the system() call was not 
working; e.g the route was not set.  I put some syslog() commands arround it, 
to check on the status before and after the call, and according to the syslog, 
the system() call is executed without error... But it is not.

I am a Pascal programmer, and don't realy know what I am doing here, although 
I can gues a bit.

I only want one thing: A Sliplogin which is *correctly* compiled for Linux 
Kernel 1.1.29, and which does not HANG all the time!  At the moment, I cannot 
leave the computer, because after one of these 'hangers' my dial-in server is 
down, until I manually kill the Sliplogin process.

Please help!

Bart.

=================================================
Bart Kindt (ZL4FOX/PA2FOX), Dunedin, New Zealand.
=================================================

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

From: zloebl@piis10.joanneum.ac.at (Klaus ZLOEBL)
Subject: gcc 2.6.0 intern. comp. error
Date: 4 Aug 1994 09:47:54 GMT

When i was compiling the 1.1.36 kernel
with the flag -O3  (new in gcc 2.6.0)
the compiler got sig 6 (IOT) compiling linux/fs/dcache.c

So what is the IOT signal or instruction?

is it a compiler error ready to be postet to gnu.g++.bug?


--

Klaus Zloebl          | E-Mail: zloebl@piis10.joanneum.ac.at
Joanneum Research     | PSI   : PSI%2631102911::ZLOEBL
Steyrergasse 17       |
A-8020 Graz           | Phone: ++43/316/876/243
AUSTRIA               |

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

From: fox@graphics.cs.nyu.edu (David Fox)
Subject: Re: -= good programmer's editor for X?
Date: 04 Aug 1994 04:38:29 GMT

In article <31pkcd$dhv@wombat.cssc-syd.tansu.com.au> chrisb@cssc-syd.tansu.com.au (Chris Bitmead) writes:

] > Not to mention I'm not a believer in having everything AND the
] > kitchen sink installed in my editor.
] 
] So remove the lisp files that don't interest you.

Boy, if people just spent half the time learning emacs that
they spend looking for an editor that doesn't do as much...
--
David Fox                                               xoF divaD
NYU Media Research Lab                     baL hcraeseR aideM UYN

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


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