Files
oldlinux-files/ftp-archives/tsx-11.mit.edu/1996-10-07/mail-archive/linux-devel/Volume2/digest228
2024-02-19 00:24:15 -05:00

665 lines
24 KiB
Plaintext

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, 26 Sep 94 08:13:07 EDT
Subject: Linux-Development Digest #228
Linux-Development Digest #228, Volume #2 Mon, 26 Sep 94 08:13:07 EDT
Contents:
Windows DLL-type linking possible...? (BATES, ROBERT P.)
HDD Controller Question (Cary B. Abend)
Re: Warnings with libc 4.5.26. (Bill C. Riemers)
Re: elf benchmarks (getting closer) (NightHawk)
Re: 900 MHz CB band??? (Richard Stone)
Re: Future of linux -- the sequel (Donald Becker)
Re: [STATUS] Linus Floppy Driver Development (jbarrett@onramp.net)
Linux on Pentium P90 PCI---which motherboard? (Vaughan R. Pratt)
Re: Don't use Linux?! (Andre Schoorl)
DMA problems on Dell XPS-90 (Steve Kann)
Re: Missing stdarg.h and (sort of) varargs.h (Eric V. Smith)
[?] DIP with auto-redial? (John Voth)
WANTED: Routers and Terminal Servers (Tony Schwartz)
Re: bind() bug, might have a fix (Neil Charley)
2nd Attempt: Menus for Dialin Users (Tony Schwartz)
Re: exabyte 8505 (HIGGINS@DELBOX.ZER.DE)
IRC Server for Linux?? (Bart Kindt)
Support for third IDE drive using 2nd IDE controller? (David Simmons)
Re: Linux on CD (Jeff Kesselman)
Re: CD-ROM gurus: PLEASE ADD PHOTOCD support to Mitsumi (Heiko Schlittermann)
Bug: telnet to Linux node, sometimes no "login:" prompth (Seppo Kallio)
----------------------------------------------------------------------------
From: r1b0804@rigel.tamu.edu (BATES, ROBERT P.)
Subject: Windows DLL-type linking possible...?
Date: 23 Sep 94 20:56:00 GMT
Howdy! I'm currently working on a Windows-based app, and want to try to port
it to Linux under XF86... However, one of the cornerstones of this project is
the ability to relink the function libs on the fly, without having to exit and
restart the application... Is there anything even remotely similar available on
the Linux or any other Un*x platform?
Thanks,
Rob
Please respoind to rbates@zgs.com
==============================================================================
Robert Bates Disclaimer:
"I make no claims..."
Texas A & M University r1b0804@sigma.tamu.edu
==============================================================================
------------------------------
From: cary@gwis2.circ.gwu.edu (Cary B. Abend)
Subject: HDD Controller Question
Date: 23 Sep 1994 19:45:30 GMT
I now have four hard drives. Two 212s one 80 and one 540. My current
system has one of the 212s and the 540, but it would be nice to mount the
other two drives as well. I just installed the Slackware 2.0
distribution [kernel 1.1.18] without a problem (all mishaps were my fault
and correctable.)
My question: I know that IDE cards that support 4 HDD exist and work
under DOS (presumably with some kind of BIOS bypass). Will (or which
of) these cards work with Linux?
--
Cary B. Abend "Why does my mouth always taste like old
cary@gwis2.circ.gwu.edu carpet in the morning?"
"Unknown. Checking medical database."
-- Susan Ivanova and the B5 Computer
"Signs and Portents"
------------------------------
From: bcr@k9.via.term.none (Bill C. Riemers)
Subject: Re: Warnings with libc 4.5.26.
Date: 23 Sep 1994 21:41:54 GMT
Reply-To: bcr@physics.purdue.edu
>>>>> "Michael" == Michael H Price <mhp1@Ra.MsState.Edu> writes:
Michael> Before I actually install libc-4.5.26, am I supposed to
Michael> get a million warnings while compiling it or should it
Michael> compile quietly?
I don't get a million, but I do get alot. It is scary to think that
the basic library running everything as so much questionable code.
The problem I always have, is if I compile libc myself, the gcc
'-g' option doesn't work. However, when I install the precompiled
version the '-g' option works fine. There is probably some key
instruction I'm missing, but what I do is:
make clean;make depend;make
This doesn't install anything (probably because I never compile as
root...) Also, there are no install commands listed in the README
file. So I just manually copy all the files with:
su -c 'cp `find . -name lib\*.so.\*` /lib;cp `find . -name lib\*.sa -or
-name lib\*.a` /usr/lib;ldconfig -v'
the *.sa *.a files to /usr/lib. There is probably some file critical
to debugging with a different extention that I'm not copying.
Bill
------------------------------
From: fsosi@j51.com (NightHawk)
Subject: Re: elf benchmarks (getting closer)
Date: 26 Sep 1994 00:02:41 -0400
Dan Connolly (connolly@hal.com) wrote:
: libc is a fairly important bunch of code to optimize. Perhaps we'll
: find some 1-liner optimization that will make the vast majority of
: linux applications run 5% faster -- or an optimization that will make
: one key application (like gcc) or one key subsystem (like TCP/IP) run
: 10% faster. Not likely, but worth the effort, if just for the
: knowledge gained.
I don't have the resources to futher fine tune libc and gcc. People
who have faster machines/big HDs should give it a try.
NH
------------------------------
From: rstone@infi.net (Richard Stone)
Subject: Re: 900 MHz CB band???
Date: 26 Sep 1994 04:36:36 GMT
Jay Ashworth (jra@zeus.IntNet.net) wrote:
: >The RatShack FM walkie talkies are 900mhz, as is the new spread spectrum
: >hand-held fone from Uniden... there are also some paging freqs up in that
: >area... but like the WaveLan.... nothing real powerful or long range..
: Forgive me, but _which_ Radio Shack HT's are those? I assume you mean the
: "business band" radios, made (I think) by Uniden. They live on
: "transient" frequencies in the 150MHz business radio band. There are no
: transient freq at 900, as far as I know.
Not CB, per se, but the 900MHz freq. is available for unlicensed consumer
use for short-range spread-spectrum tx/rx. The "latest" cordless phones are
one consumer application. More details? Write the FCC.
--
Richard S. Stone Network Engineer
The Engineering Design Group
"If it ain't broke, don't fix it!" "If it *is* broke, pay us to fix it!"
2-FOR-1 DEAL: "We'll break it for you and then fix it; for one low price!"
rstone@edgp.com rstone@infi.net
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
/* disclaimer.h */
printf("The opinions expressed above are my own, and do not necessarily
represent those of the Engineering Design Group or its affiliates.\n")
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
------------------------------
From: becker@cesdis.gsfc.nasa.gov (Donald Becker)
Subject: Re: Future of linux -- the sequel
Date: 21 Sep 1994 23:20:51 -0400
In article <35mrbo$a6@pandora.las-vegas.nv.us>,
Eric J. Schwertfeger <eric@pandora.Las-Vegas.NV.US> wrote:
>No, there's DD0-DD15, 16 bits. The reason IDE is slower is because without
>ATA-2 DMA Mode (preferably multiword mode 2), the CPU has to do the
>transfers itself, whereas most SCSI disks do busmastering. The ATA-2 draft
>spec (which is already being used) has transfer rates up to 13M/Sec, which
>puts it between Fast SCSI II and Fast/Wide SCSI II. I'm seriously
>considering getting a busmastering IDE controller just to write DMA drivers
>for linux on it :-)
The ATA-2 interface designs I've seen don't use anything nearly as
complicated as a busmaster interface. They basically add two gates to
the bus driver enables and connect the DREQ and DACK (DMA request and grant)
lines to the bus. The data transfer can then be done by the DMA controller
(which is hopefully faster on modern machines than the abysmal PC/AT
implementation).
Is anyone familiar with a different type of ATA-2 controller implementation?
This type of slave DMA transfer only really increases performance when
multiple sector mode is used. 512 byte sector transfers are below
break-even for setting up the DMA controller and taking an interrupt, but a
4K or 8K slave DMA transfer is a significant win. The tradeoff of the IDE
interface vs. SCSI then becomes the lower IDE setup cost vs. the ability to
chain SCSI commands and the the related ability to disconnect. And with
only two devices per low-cost controller, doing disconnects on IDE doesn't
make sense.
A final note: programmed-I/O is *not* evil. It primes the cache with
soon-to-be-used data and avoids coherency problems and interrupt overhead.
Nor is the ISA bus evil. It's very inexpensive, and has enough bandwidth for
most I/O devices. It's only when you combine the ISA bus and heavy-duty
programmed-I/O that a fast processor wastes most of its cycles waiting on
the bus.
--
Donald Becker becker@cesdis.gsfc.nasa.gov
USRA-CESDIS, Center of Excellence in Space Data and Information Sciences.
Code 930.5, Goddard Space Flight Center, Greenbelt, MD. 20771
301-286-0882 http://cesdis.gsfc.nasa.gov/pub/people/becker/whoiam.html
------------------------------
From: jbarrett@onramp.net
Subject: Re: [STATUS] Linus Floppy Driver Development
Date: Mon, 26 Sep 94 00:53:17 PDT
<niemidc@clark.net> writes:
>
> The basic problem is that the floppy drive in no way notifies the
> rest of the system when a disk has been inserted. This means that
> the floppy drive must be polled periodically so as to notice when
> a disk appears. This is not impossible, as a test already exists
>
> ------------------------------------------------------------------
> David C. Niemi (SLMA, Herndon, Virginia, USA) niemidc@clark.net
> Know the difference between the color of the wine and the color of
> the glass. (Jalaluddin Rumi)
> ------------------------------------------------------------------
>
Since when does the drive not notify the system of a disk change....
one of the big features of the 1.2 and 1.44 drives is the "DISK CHANGE LINE"
that tells you if a floppy has been changed since the drive was last selected!
About the only use of this line that I've ever seen is in FastBack for DOS...
so that FastBack can start immediatly writing to a disk as soon as it has been
inserted... I have no idea what port or bit this line is visible on... but it
is available somewhere out there is I/O land......
I think that an auto-mounter option would be a good idea for those of us who
have to build turnkey systems to be used by the unix-illiterate... It would
sure save some of the bogus code I've had to write to test if a floppy is
inserted, formatted, etc.
It might not be a bad idea if the automounter did not interfere with the usual
device nodes for the floppies... so you could still TAR/CPIO/Etc. to an auto-
mounted device??
Or you could (Deity of your choice forgive) go as far as MS Windows and pop up
a window asking if you want to format a floppy if the auto-mounter cannot
detect the filesystem type when you first attempt to access the device...
Naaaaa... then the automounter would have to be aware of which user/console
made the request to access the device...
Over the edge again
John Barrett
------------------------------
From: pratt@Sunburn.Stanford.EDU (Vaughan R. Pratt)
Crossposted-To: comp.os.linux.misc
Subject: Linux on Pentium P90 PCI---which motherboard?
Date: 23 Sep 1994 21:15:37 GMT
If Linux runs on your Pentium P90 PCI, or you know of a working such,
I'd appreciate knowing what motherboard did the trick.
--
Vaughan Pratt http://boole.stanford.edu/boole.html
------------------------------
From: aschoorl@uglz.UVic.CA (Andre Schoorl)
Subject: Re: Don't use Linux?!
Date: Fri, 23 Sep 94 21:25:42 GMT
David Holland <dholland@husc7.harvard.edu> wrote:
>Michael Schumacher (hightec@sbusol.rz.uni-sb.de) wrote:
>
> > : 3. The kernel versions change faster than the speed of light. If you
> > : ask for a "stable" version, you'll be teached that there are two
> > : versions: 1.0 (production) and 1.1 (hacker's paradise).
At least they ARE being developed, DOS, Windows, etc... sure doesn't
change much very fast. Plus Linux is free. :^)
>Updating the kernel takes maybe half an hour to unpack and configure,
>and maybe another half hour (on a decently fast machine) unattended to
>compile.
Yikes! Answering all the questions to make config and doing a make dep
takes maybe 10 min, and my machine takes 12 min to compile the kernel.
I only have a DX2 66. I wonder how the Pentium fairs.
(btw I'm running kernel 1.1.50)
>Furthermore, keeping 100 machines up to date is a lot easier with Unix
>than with DOS - there are tools for it.
Definitely. Normally PCs are horrible at networking compared to Unix
(in this case Linux). Granted, Unix is harder to use for the general
user, but still.
--
Andre Schoorl <aschoorl@engr.uvic.ca> PGP key available via finger/keyserver
C.Eng Student, U.Vic, Canada. -> LINUX! <- http://www-engr.uvic.ca/~aschoorl
------------------------------
From: stevek@panix.com (Steve Kann)
Crossposted-To: alt.sys.pc-clone.dell
Subject: DMA problems on Dell XPS-90
Date: 26 Sep 1994 02:56:19 -0400
I have been having problems with the DMA system on DELL XPS-90 machines.
I run Linux on them, 1 of them is a "server" of sorts, the other 20
serve as dual-purpose Xterminals/Dos workstations.
I am no DMA expert, but I have been getting errors from the two main DMA
devices on the systems: namely the floppy driver, and the lance ethernet
driver (I have AT-1500's). The errors are not very descriptive, (the
floppy driver says "Dma Error?", while the lance driver reports "Bus
Master Arbitration Problem". The errors seem to be a temporary
unavailability of the DMA controller. This doesn't seem to have a great
affect on the floppy, (since it goes so slow, I'd imagine, it has plenty
of time to retry the DMA transfer), but it often causes the lance driver
to get a FIFO error, because it cannot keep it's buffers filled while
performing a transmit.
I have not spoken to anyone at DELL yet, but I just wanted to see if
anyone else has had similar problems with these machines. They all pass
Dell's Diags, but I don't think that really means all that much.
Also, I have two other types of machine that are configured similarly
(same ethernet card, same kernel floppy/lance drivers), and they all do
not give these errors (Dell 486D50, Dell Omniplex P90).
Anyone else having these problems? Anyone have a similar setup, and NOT
having these problems?
-SteveK
--
- Steve
stevek@cooper.edu
stevek@midnite.roslyn.ny.us
------------------------------
From: esmith@access4.digex.net (Eric V. Smith)
Subject: Re: Missing stdarg.h and (sort of) varargs.h
Date: 25 Sep 1994 21:28:42 -0400
In article <Cwp389.M8t@news.cern.ch>, Dan Pop <danpop@cernapo.cern.ch> wrote:
>Mistery! /usr/include/stdarg.h does not exist, but gcc doesn't complain.
>Have I just discovered a gcc bug?
I know this is off topic and doesn't belong here, but...
As those who hang out in comp.std.c would tell you, this is acceptable
behavior. ANSI C doesn't require that standared include files be
in actual files. They could just tell the compiler "hey, you can
now use the stdarg stuff."
Eric.
------------------------------
From: jdv@ee.ualberta.ca (John Voth)
Subject: [?] DIP with auto-redial?
Date: 26 Sep 1994 06:01:30 GMT
Greetings Linux Community:
I am in search of a DIP or DIP-alike program that has auto redial
functions built into it.
I am in constant competition with others trying to connect to my
university and the DIP I have now just flops at a busy signal.
Any ideas?
John
--
________________________________________________________
| John D. Voth | Computer Engineering |
| jdv@ee.ualberta.ca | University of Alberta |
| voth@cs.ualberta.ca | Alberta, Canada |
------------------------------
From: tony@teleport.com (Tony Schwartz)
Crossposted-To: comp.os.linux.admin
Subject: WANTED: Routers and Terminal Servers
Date: Sun, 25 Sep 1994 20:56:29
I have cash and a need for some routers and Terminal Servers. If you have any
extras around that you want to part with, please email me with descriptions
and pricing.
Thanks in advance.
Tony Schwartz
Portland, OR
tony@teleport.com
(Not related to teleport.com, just a client)
------------------------------
From: N.Charley-CSSE93@cs.bham.ac.uk (Neil Charley)
Subject: Re: bind() bug, might have a fix
Date: 26 Sep 1994 03:39:01 GMT
(I originally posted this to comp.os.linux.admin yesterday, someone
suggested I post it here too, followups to both groups from THIS article)
Whilst testing out my talker code I happened upon a rather nasty bug
in Linux, I certainly hope it isn't meant to be a feature!
Basically for non-systems ports, ANY user can at ANY time bind() any
port number, if a process already 'owns' the port, even if the process
is owned by root! This means that if you run some sort of server,
someone else can come along and run a program to bind() the port, and
any new connections go to it. With the help of a friend I managed to
track down where the bind() for AF_INET sockets is actually done, and
worked out a fix. It seems to work fine, got it running on test atm.
The code now allows uid 0 to always grab a port as before, and any
other user to grab non-system (>= 1024) ports if and only if there
isn't a running process that is bound to that port. I'm using the
->dead data member to test if there is an 'alive' process attached to
the port, is this correct?
I have a diff patch against the sources for 1.1.51, but have been
unable to find who is the best person to send this to for it to be
checked and maybe made official, any offers?
-Neil
------------------------------
From: tony@teleport.com (Tony Schwartz)
Subject: 2nd Attempt: Menus for Dialin Users
Date: Wed, 21 Sep 1994 22:50:42
Please direct me to menus for Dialin users.
Tony Schwartz
------------------------------
From: HIGGINS@DELBOX.ZER.DE
Subject: Re: exabyte 8505
Date: Sat, 24 Sep 1994 10:01:00 +0200
dford@trc.amoco.com meinte am 21.09.94
zum Thema "exabyte 8505":
> Does linux support Exabyte 8505 tape drives, if not, is anybody working
> drivers for the 8505. Also, is there a Fortran compiler for linux yet or F2C
> available.
A general answer :
Linux supports all types of SCSI-streamers. This includes
- QIC 60/150/525/... [QIC-80/QIC-117 are floppy-streamers and not SCSI]
- Exabyte 8200/8500/...
- HP-DAT / Wang-DAT
SCSI-streamers are sequential-access-devices and the scsi-commands are
equal for all streamers. So .... no problem !
>
> David A. Ford
> dford@trc.amoco.com
Ciao
Higgins
--
You can escape the gates of hell, say DOG and WINDOG,
USE LINUX :-) !
## CrossPoint v3.0 ##
------------------------------
From: bart@dunedin.es.co.nz (Bart Kindt)
Subject: IRC Server for Linux??
Date: Tue, 27 Sep 1994 02:30:06 GMT
Hi. Is there a IRC (Internet Relay Chat) Server available for Linux?
Any replies by E-Mail please...
Thanks,
====================================================================================
Bart Kindt (ZL4FOX) System Operator, Efficient Software NZ LTD, Dunedin, New Zealand
====================================================================================
------------------------------
From: simmons@EE.MsState.Edu (David Simmons)
Subject: Support for third IDE drive using 2nd IDE controller?
Date: 26 Sep 1994 07:02:54 GMT
Reply-To: simmons@EE.MsState.Edu
Is there any possible way to support more than two IDE hard drives
in Linux? I want to use the second IDE controller in a Gateway
machine (it's ISA, the first one is a PCI) to put a third drive on.
The bios recognizes the third drive properly...
Thanks for any info.
David
--
David Simmons, System Administrator simmons@ee.msstate.edu
Mississippi State University Electrical and Computer Engineering
Visit my home page! http://www.ee.msstate.edu/~simmons
------------------------------
From: jeffpk@netcom.com (Jeff Kesselman)
Subject: Re: Linux on CD
Date: Thu, 22 Sep 1994 04:38:34 GMT
In article <CwDx2M.ww@cs.bsu.edu>, <fagarcia@cs.bsu.edu> wrote:
>I was having a chat with someone over the net and we came to this..
>What if Linux came fully implemented (X & all the disk sets) on a CD and
>all you would have to do is boot off the CD rom and have the
>settings/option files (ie ~/.seyon inittab & the rc scrips) in your HD.
>
>I mean, this would save a lot of diskspace ;)
>
>just a vague thought.
>
>-Frank
>
No offense, but you're either not up on the true speed of CD-ROM drives
or you don't realize how often UNIX hits the media.
You can do this with a number of commerical releases of Linux, including
the Yygdrasil release, but in practice running UNIX off of a 300kb/sec
device is painful and running X off of such a device is glacial.
Its alright as a way to look at UNIX, maybe, for a brief period, but
TOTALLY unstatisfactory for doing real work.
Sorry.
Jeff Kesselman
------------------------------
Crossposted-To: comp.os.linux.help
From: heiko@lotte.sax.de (Heiko Schlittermann)
Subject: Re: CD-ROM gurus: PLEASE ADD PHOTOCD support to Mitsumi
Date: Mon, 26 Sep 1994 06:59:04 GMT
In article <35a9cp$ig3@rutcor.rutgers.edu>,
Tamas Badics <badics@rutcor.rutgers.edu> wrote:
>Dear CD-ROM Gurus,
>
>Would some of you consider writing this minor improvement for the Mitsumi
>driver mcd.c?
I'm trying to write a new Mitsumi driver. But I've never handled
Photo-CD nor I have such a CD.
If you could be patient ... If it's time for me to think about
Photo-CD-Support I'll ask about some hints. At first the
functionality of the `old' (unifix) driver should be done.
-- heiko
------------------------------
From: kallio@tukki.jyu.fi (Seppo Kallio)
Crossposted-To: comp.os.linux.help
Subject: Bug: telnet to Linux node, sometimes no "login:" prompth
Date: 26 Sep 1994 08:25:09 GMT
Could someone help!
I have strange problems in my Linux box. It sometimes hangs all
new incomming telnet users. The Linux system is somehow responding to
telnet: it is opening the window on NCSA Telnet but does not write the
"login:" prompth.
In log files start to repeat these errors:
/usr/adm/syslog:
Sep 26 10:02:27 silmu inetd[39]: accept (for telnet): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for smtp): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for telnet): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for smtp): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for telnet): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for smtp): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for telnet): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for smtp): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for telnet): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for smtp): Try again
Sep 26 10:02:27 silmu inetd[39]: accept (for telnet): Try again
/usr/adm/messages:
Sep 26 10:02:25 silmu kernel: sock_socket: no more sockets
Sep 26 10:02:25 silmu kernel: NET: sock_accept: no more sockets
Sep 26 10:02:25 silmu last message repeated 13 times
Sep 26 10:02:25 silmu kernel: NET: sock_accept:
no more <6>Sep 26 10:02:25 kernel: NET: sock_accept: no more sockets
Sep 26 10:02:26 silmu kernel: NET: sock_accept: no more sockets
Sep 26 10:02:26 silmu last message repeated 14 times
Sep 26 10:02:28 silmu in.telnetd[9188]: connect from haigha.pccc.jyu.fi
Sep 26 10:02:26 silmu kernel: NET: sock_accept:
no <6>Sep 26 10:02:26 kernel: NET: sock_accept: no more sockets
Kernel version is 1.1.49
I can sometimes get this jam by running netstat -i -c on two terminals
at same time but not every time.
HW: 66MHz 486, one 2GB scsi aha-disk, 32M RAM, 2500 user accounts (!)
less than 10 users at this time.
Seppo Kallio
kallio@jyu.fi
------------------------------
** 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
******************************