Linux-Development-Sys Digest #392, Volume #6 Wed, 10 Feb 99 13:13:53 EST
Contents:
Re: Lightnux, the slim Linux (Sami Tikka)
Re: K6-2 and Linux, Are there any Bug? ([EMAIL PROTECTED])
Re: Modest next goal for Linux (Peter Samuelson)
Re: Create Bootable RH 5.2 CD? (Mark Bergsma)
UNIX/C++/Emb. eng. wanted (Garynlang)
/dev/gfx project: Modular graphics card driver. (Bjorn De Meyer)
Re: K6-400 "kernel paging request" errors ("David R. Bergstein")
Re: Printer problem with 2.2.1 kernel (Christian Ordig)
smbmount and Kernel 2.2.1 (Christian Ordig)
SUMMARY: How to detect CPU clock speed? ([EMAIL PROTECTED])
Re: SUMMARY: How to detect CPU clock speed? (Stefan Monnier)
2.2.1, Genius NetMouse and gpm 1.13/1.16 (Bruce Smith)
Re: Terminals (Cray)
Re: K6-400 "kernel paging request" errors (Jeff McWilliams)
Re: Shared library programming (Markus Schutz)
Re: SUMMARY: How to detect CPU clock speed? ([EMAIL PROTECTED])
----------------------------------------------------------------------------
From: Sami Tikka <[EMAIL PROTECTED]>
Subject: Re: Lightnux, the slim Linux
Date: 09 Feb 1999 21:16:55 +0200
Douglas Kilpatrick <[EMAIL PROTECTED]> writes:
> Check out the "Diskless" mini-howto. Its a little out of date; I don't
> think the kernel and ramdisk will fit on a single floppy anymore. The
> basic idea should still be sound.
Just a couple of months ago I built a 1.4M disk with 2.0.36 + TCP/IP +
ramdisk for some utilities. Granted, it was very tuned to the job,
nothing extra and just the driver for the network card on the PC I was
going to use it. And there wasn't too much space on the ramdisk
either. You couldn't fix an X server there.
But you could fit in there something that would mount some disks from
somewhere and run X server and stuff over the net.
--
Sami Tikka, [EMAIL PROTECTED], http://www.iki.fi/sti/
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: K6-2 and Linux, Are there any Bug?
Crossposted-To:
alt.os.linux.slackware,comp.os.linux.hardware,comp.os.linux.misc,comp.os.linux.setup
Date: Tue, 09 Feb 1999 17:40:36 GMT
In comp.os.linux.misc Timothy J. Lee <[EMAIL PROTECTED]> wrote:
: Does an "i430TX" motherboard support a 300 MHz K6-2 processor and
: 128 MB of RAM?
Well, it won't cache the 2nd 64MB of RAM, since it doesn't have enough tagram,
but if it has the proper voltage settings, it should do the K6-2 (my TX board
doesn't have the proper voltage settings).
--
patowic jurai net
Cabal Obsidian Order # i. Cabal #e. TINC.
"Do you expect me to talk?"
"No, Mr. Bond, I expect you to _die_."
------------------------------
From: [EMAIL PROTECTED] (Peter Samuelson)
Crossposted-To: comp.os.linux.advocacy
Subject: Re: Modest next goal for Linux
Date: 28 Jan 1999 23:13:27 -0600
Reply-To: Peter Samuelson <[EMAIL PROTECTED]>
[John De Hoog <[EMAIL PROTECTED]>]
> "Faster file access. Version 2.2 can store filenames in a high-speed
> cache in memory, meaning that users won't have to wait for the
> computer to retrieve the information off relatively slow hard disks."
Right, the dcache.
> OK, maybe CNet news isn't the right place to be finding out this
> stuff.<g>
CNet probably has no more clue what it means than you did, but whoever
they are parroting got it right.
Linux has always had block caching, so that for a block device such as
a hard disk partition, the OS caches all the blocks it knows about
(previously read or written) and the cache grows to use available
memory. (Unused memory is wasted memory.) The cache grows and shrinks
per memory demand. So far so good; any decent OS will do this much.
But in the 2.1 series they introduce a directory entry cache, or
`dentry cache' for short, or `dcache' for shorter. This is simply a
cache of `dentries' (file pathnames, hashed for speed) holding such
useful information as inode numbers. When a user process either reads
a directory or asks for a specific filename, Linux 2.2 looks up the
dentry in the dcache and, if not found, starts searching the usual way.
In the case of a dcache hit this is *much* faster than just searching
through the inodes and directory contents, even if *those* are already
in the block cache. AFAIK most Unices do not have a dcache. Someone
said NT does; I don't know.
--
Peter Samuelson
<sampo.creighton.edu!psamuels>
------------------------------
From: Mark Bergsma <[EMAIL PROTECTED]>
Subject: Re: Create Bootable RH 5.2 CD?
Date: Tue, 09 Feb 1999 14:54:53 +0100
Dan O'Reilly wrote:
>
> How does one go about creating a bootable LINUX RedHat 5.2 CD? I have
> an HP 7200 IDE CD burner on my PC.
There's a program for doing just that. I don't know it's name anymore,
but I've seen it somewhere in the Fresmeat application-index.
(www.freshmeat.net)
--
Mark
[EMAIL PROTECTED]
- Linux 2.2.1 on an i686 (266.24 BogoMIPS) -
------------------------------
From: [EMAIL PROTECTED] (Garynlang)
Subject: UNIX/C++/Emb. eng. wanted
Date: 9 Feb 1999 19:26:04 GMT
We have an excellent permanent position in
Irvine (CA) for an embedded engineer with
UNIX/C++ knowledge. US Citizenship req.
Pls. contact us for futher details.
Gary
Gary N. Lang
Vice President of ACD,Inc.
E-mail: [EMAIL PROTECTED]
http://www.acdcon.com/
------------------------------
From: Bjorn De Meyer <[EMAIL PROTECTED]>
Subject: /dev/gfx project: Modular graphics card driver.
Date: Wed, 10 Feb 1999 13:14:46 +0100
Hi.
I've started programming a modular device driver for graphic cards.
I need more information about low-level programming of the VGA and
SVGA chipsets, and about how to use stacked modules. Can anyone point me
to online documentation about these subjects?
I know that there is at least one similar project going on, so don't flame
me for that. I want to make my device driver a bit different: simple,
fast, free even for commercial use, and hardware-oriented.
Thanks in advance.
Bjorn De Meyer.
IA #i^4
Homepage: http://studwww.rug.ac.be/~bgdmeyer
E-mail: [EMAIL PROTECTED]
Licentiate Chemistry.
Coordinator of the "Tales of Artah: First Sword" CRPG project.
"I've shaven. I'm happy. I have no religion. Don't pray for me."
Illusion of Gaia 2
------------------------------
From: "David R. Bergstein" <[EMAIL PROTECTED]>
Crossposted-To: linux.dev.kernel,comp.os.linux.hardware,comp.os.linux.misc
Subject: Re: K6-400 "kernel paging request" errors
Date: Wed, 10 Feb 1999 07:33:53 -0500
Thanks - the tip seems to have worked and my system is now rock stable.
Prior to cleaning out the CPU grease & dust I had upgraded to Linux
Kernel 2.2.1 and another similar system crash occurred - this was
motivation enough for me to open the case and try your suggrestion.
Cheers,
- David
"David R. Bergstein" wrote:
>
> José Ureña wrote:
> >
> > When using Heat Tranfer compound (grease), remember to remove any
> > excess from the sides and walls of the heat sink. (the Radiator)
> > The compound tends to collect dust and turns into a kind of insulator
> > that prevents the heat from excaping the heat sink.
> >
> That's a great tip - I will check for this next time I service the unit.
>
> Thanks again,
> -
> David R. Bergstein
> Systems Engineer and Blues Musician
> Rockville, MD
> ---------------------------------------------------------------------------
> [EMAIL PROTECTED] [EMAIL PROTECTED]
> SE & Blues Musician Home Page Heart of Blue - Playin' the Blues for
> You!
> http://www.erols.com/dbergst http://heartofblue.com
> ---------------------------------------------------------------------------
--
David R. Bergstein
Systems Engineer and Blues Musician - http://www.erols.com/dbergst
Heart of Blue - http://heartofblue.com
------------------------------
From: [EMAIL PROTECTED] (Christian Ordig)
Subject: Re: Printer problem with 2.2.1 kernel
Date: Wed, 10 Feb 1999 13:37:23 +0100
Reply-To: "Christian Ordig" <[EMAIL PROTECTED]>
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Piniek aka Piotr Ingling) writes:
> Dnia Sun, 07 Feb 1999 05:54:15 -0800, P Wong <[EMAIL PROTECTED]>
> napisa�(a):
>
>>I just compiled the 2.2.1 kernel and I can't use the printer. I'm new to
>>this kernel thing so I problably forgot to mark something necessary for
>>using the printer. It worked on the 2.0.34 kernel.The config file is
>>attached
>
> Have you read the Documentation/parport.txt? It describes changes that were
> made to parallel port driver. The main change affecting most users is the way
> ports are numbered. The first port that is found, no matter under what
> address, is /dev/lp0 - so if your port is on 378 (/dev/lp1 in the old days) it
> will become /dev/lp0.
>
>
> Piotr Ingling
>
> e-mail: [EMAIL PROTECTED]
That's the first problem and the second occured when I compiled parport as
module and didn't give any module parameters. After compiling parport into the
Kernel 2.2.1 everything works fine. (I mean printing ;-)
--
Christian Ordig | Homepage: http://thor.prohosting.com/~chrordig/
Germany | eMail: [EMAIL PROTECTED]
__ _ |
/ / (_)__ __ ____ __ | Why Linux? Because it is free, stable, and
/ /__/ / _ \/ // /\ \/ / | bugs can be fixed in source opposed to waiting
/____/_/_//_/\_,_/ /_/\_\ | for a stable WinTendo from Micro$oft.
------------------------------
From: [EMAIL PROTECTED] (Christian Ordig)
Subject: smbmount and Kernel 2.2.1
Date: Wed, 10 Feb 1999 13:59:38 +0100
Reply-To: "Christian Ordig" <[EMAIL PROTECTED]>
Hi,
I just upgraded my SuSE Linux 6.0 to kernel 2.2.1. They claim it to be "Kernel
2.2.x ready", but I just had to upgrade some things... nevermind. Everything
wotks fine again, without smbmount. I downloaded Samba 2.0, as adviced in the
Kernel documentation and compiled it. The compilation of smbfs 2.0.2 stops with
many error messages. Smbfs 2.1.51 compiles fine, but when I try to mount a share
I get a "file not found" error. I read something about a new command line syntax
of smbmount, but the man page isn't up-to-date, yet. "smbmount -h" tells me
"smbmount <mount-point> [options]" as syntax - but who tells smbmount which
share to mount???
Thank you for answering.
--
Christian Ordig | Homepage: http://thor.prohosting.com/~chrordig/
Germany | eMail: [EMAIL PROTECTED]
__ _ |
/ / (_)__ __ ____ __ | Why Linux? Because it is free, stable, and
/ /__/ / _ \/ // /\ \/ / | bugs can be fixed in source opposed to waiting
/____/_/_//_/\_,_/ /_/\_\ | for a stable WinTendo from Micro$oft.
------------------------------
From: [EMAIL PROTECTED]
Subject: SUMMARY: How to detect CPU clock speed?
Date: Tue, 09 Feb 1999 19:11:06 GMT
Thanks to all who replied.
Summarising, there is a CPU clock field in newer (2.1.x) kernels, but not in
production ones, neither there is a ready API call for that. Also, the format
of cpuinfo will depend on the type of CPU involved.
To answer the questions from posts and mails:
1. This is used in LSF (Load Sharing Facility) by Platform Computing Corp.
If interested, see http://www.platform.com for more info. 2. I have answers
for most platforms I'm currently interested in (Solaris, HP-UX, Irix, even
NT, ...), but not for Linux. 3. Because of the specifics of the project I
cannot deploy any tight looping algorithm and rather stick with [possibly
incorrect because of over/under clocking] information. 4. No, x86 is not
assumed. Our current ports support x86 and Alpha, and may support others.
(That's why I'd rather use API call than 'cat /proc/cpuinfo')
Please, if you disagree with my summary and and/or can offer a way to find out
this value, let me know.
In article <79nuqk$tk7$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] wrote:
> Hi,
>
> I need to find out the CPU clock speed. cat'ing files in /proc didn't help,
> neither did apropos search on a box. It seems (at least to me) that this value
> must be readily available and I'm simply overlooking something. API call is
> preferred (nlist would do just fine if I knew the symbol to search for).
>
> And no, I cannot use bogomips because of certain restrictions on the
> uniformity of the values, imposed by the project.
>
--
Igor Shpigelman "Stand aside in the coming battle and though
Yet Another UNIX Hacker thou fightest be not thou the warrior." M.C.
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
From: Stefan Monnier
<[EMAIL PROTECTED]>
Subject: Re: SUMMARY: How to detect CPU clock speed?
Date: 09 Feb 1999 16:34:33 -0500
>>>>> "Karmadon" == Karmadon <[EMAIL PROTECTED]> writes:
> Summarising, there is a CPU clock field in newer (2.1.x) kernels, but not in
> production ones, neither there is a ready API call for that. Also, the format
2.2 is considered `production' and does have the MHz info.
> 1. This is used in LSF (Load Sharing Facility) by Platform Computing Corp.
Load balancing cares about relative performance of the computers (which may or
may not be related to processor frequency).
So if your machines are homogeneous, MHz and BogoMIPS will be just as useful as
any other measure. If they are heterogenous, then MHz are about as useless as
BogoMIPS.
> for most platforms I'm currently interested in (Solaris, HP-UX, Irix, even
> NT, ...), but not for Linux.
The closest I can think of is BogoMIPS. It's bogus as the name implies, but
it's not really more bogus than MHz.
> 3. Because of the specifics of the project I
> cannot deploy any tight looping algorithm and rather stick with [possibly
> incorrect because of over/under clocking] information.
Does that mean that the time to get the MHz rating is relevant ?
Timing a somewhat random little loop would be more portable across OSes,
would give a less meaningless performance estimate and might not take that
much longer than querying the OS.
> 4. No, x86 is not
> assumed. Our current ports support x86 and Alpha, and may support others.
> (That's why I'd rather use API call than 'cat /proc/cpuinfo')
I despise text interfaces, so I completely agree. I don't know of any other
way to get that data in Linux, tho. Maybe it's available via `sysctl' ?
Stefan
------------------------------
From: [EMAIL PROTECTED] (Bruce Smith)
Subject: 2.2.1, Genius NetMouse and gpm 1.13/1.16
Date: Wed, 10 Feb 1999 20:32:23 GMT
Help
Since upgrading to 2.2.1, gpm has been doing strange things. Button
assignments change without warning and when I switch back from X
to console, whatever is in the cut buffer, gets pasted to the command line.
I am running what remains of a Slackware 3.4 distribution, upgraded over the
years, Linux 2.2.1 and XFree86 3.3.3. Any ideas?
Bruce Smith
<[EMAIL PROTECTED]>
------------------------------
From: [EMAIL PROTECTED] (Cray)
Subject: Re: Terminals
Date: Tue, 09 Feb 1999 22:49:23 +0100
Timothy K Canfield wrote:
>
> Are terminals, like ttyp0, managed by the kernel? Is there a common
> device driver which programs like xterm and telnetd use or do they all
> have their own?
>
> Thanks
the key you're looking for is the command line interpreter. all
terminals under unix are based on /bin/bash . to answer your question:
the terminals neither do have their own interface nor it's a kernel
function. it's simply a little program called "bash", which is started
in the terminal window.
bye the way - Cray
------------------------------
From: [EMAIL PROTECTED] (Jeff McWilliams)
Crossposted-To: linux.dev.kernel,comp.os.linux.hardware,comp.os.linux.misc
Subject: Re: K6-400 "kernel paging request" errors
Date: 10 Feb 1999 01:08:44 GMT
In article <[EMAIL PROTECTED]>, Christian Aasland wrote:
>
>Stefan Lucke wrote:
>
>> Did you try other FSB clock rates?
>> With my K6-2 400 it is impossible to compile the kernel with FSB 100MHz
Might be a bad CPU, I'm running 100MHz FSB on a K6-2 300 using an Asus P5-A
and have compiled the 2.2.0-pre5, 6, 7 kernels a couple times without
trouble.
Jeff
--
Jeff McWilliams
"The mystery boy has gone to war, in the fields of death you'll find him."
------------------------------
Date: Tue, 09 Feb 1999 23:45:28 +0100
From: Markus Schutz <[EMAIL PROTECTED]>
Reply-To: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.development.apps
Subject: Re: Shared library programming
XuYifeng wrote:
> I am relate new to linux programming. currently I
> want to build a shared library, problem is the library
> need automatic init and finish function like DllMain
> function in WindozeNT which will be invoked at library
> loaded or freed ?
>
> Is there a shared library programming guide avail?
Hi,
Take a look at libtool (from your favourite FTP GNU mirror site). It
will help you with the Makefile to build your shared lib.
I'm not sure you will be able to mimic DllMain.
Markus
--
A designer knows he has achieved perfection not when there is
nothing left to add, but when there is nothing left to take away.
Antoine de Saint-Exup�ry
=================================================================
Markus SCH�TZ
8, Ch. des Aub�pines Phone: ++41 21 646 9362
CH-1004 Lausanne mailto:[EMAIL PROTECTED]
Switzerland
http://www.geocities.com/SiliconValley/Foothills/9297
=================================================================
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: SUMMARY: How to detect CPU clock speed?
Date: Wed, 10 Feb 1999 16:25:08 GMT
Well, all I want is a way to map a specific piece of hadrware to a specific
performance metric. Performance metrics can be obtained in miriads of
different ways and will depend on the applications involved. At the same
time, the processor type and its speed will be the same as long as you don't
upgrade :-)
In article <79qcff$sla$[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (Dave Platt) wrote:
> >Load balancing cares about relative performance of the computers (which may
or
> >may not be related to processor frequency).
> >So if your machines are homogeneous, MHz and BogoMIPS will be just as useful
as
> >any other measure. If they are heterogenous, then MHz are about as useless
as
> >BogoMIPS.
>
> Right. Even if they're of the same processor family and clock speed,
> the actual performance may vary significantly due to factors other
> than these.
>
> E.g. compare a 100 MHz Pentium system with a 256k L2 writeback cache,
> with an otherwise-identical system whose L2 cache stick has been
> pulled off of the motherboard. They'll have the same clock rate, and
> (I believe) the same BogoMIPS number, and the same amount of memory...
> and their actual performance when running real-world applications will
> be _grossly_ different. Same issue if you run a first-generation
> Intel Celeron processor (no L2 cache) against an identical-clock-
> rate Pentium II.
>
> If you really want a metric for load-balancing which is going to be
> useful for a given project, I don't know of any reliable way other
> than to actually run a few of the actual programs involved on each
> machine, measure how fast they run in practice, and set the
> load-balancing metrics accordingly.
>
> --
> Dave Platt [EMAIL PROTECTED]
> Visit the Jade Warrior home page: http://www.radagast.org/jade-warrior/
> I do _not_ wish to receive unsolicited commercial email, and I will
> boycott any company which has the gall to send me such ads!
>
--
Igor Shpigelman "Stand aside in the coming battle and though
Yet Another UNIX Hacker thou fightest be not thou the warrior." M.C.
============= Posted via Deja News, The Discussion Network ============
http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
------------------------------
** 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: [EMAIL PROTECTED]
You can send mail to the entire list (and comp.os.linux.development.system) via:
Internet: [EMAIL PROTECTED]
Linux may be obtained via one of these FTP sites:
ftp.funet.fi pub/Linux
tsx-11.mit.edu pub/linux
sunsite.unc.edu pub/Linux
End of Linux-Development-System Digest
******************************
