Linux-Development-Sys Digest #964, Volume #7 Wed, 21 Jun 00 09:13:13 EDT
Contents:
Disk Thrashing (Octalman - Wilbur Killebrew)
help regarding battery utilities (piyush mishra)
Re: Packaging for Release (Octalman - Wilbur Killebrew)
Re: Why so many collisioins ? (Michael Powe)
Re: OT: Re: CodeWarrior for Linux. Comments ? (David Minor)
Thread Specific Memory (David Minor)
Re: How to mount floppey drive ? ([EMAIL PROTECTED])
kernel books (Philippe Peyre)
64bit division (Daniel Krippner)
Re: kernel books (Arnaud Westenberg)
UDI (Gundu)
top-half to bottom-half ("Mikko Jaakkola")
Howto debug kernel deadlock ? (Ulrich Leodolter)
Re: installing procinfo 17 (termcap.h missing) ([EMAIL PROTECTED])
Re: top-half to bottom-half (Alan Donovan)
Re: data acquisition boards ("Casper Helenius")
Re: TCP question (ecivasa)
Re: 64bit division (Aki M Laukkanen)
----------------------------------------------------------------------------
From: Octalman - Wilbur Killebrew
Subject: Disk Thrashing
Date: Wed, 21 Jun 2000 05:22:13 GMT
Reply-To: [EMAIL PROTECTED]
I have run kernel versions 2.2.9 (stock SuSE 6.2), 2.2.14 and 2.2.16, with varying
experince with disk thrashing.
StarOffice 5.1 Mail, when running with kernel versions 2.2.9 and 2.2.16, thrashes the
disk drive when deleting messages. With kernel Version 2.2.9, SO Mail sometimes takes
up to 1 minute 30 seconds to delete a message, accompanied by intense, thumping disk
thrashing, but with kernel Version 2.2.14, never more than about ten seconds, with
only light disk activity. When I upgraded to 2.2.16, thrashing returned, but isn't as
bad as with 2.2.9; max SO Mail delete is around 45 seconds, but seldom less than 20
seconds.
This is with an AMD K6-III/400, 128MB RAM and 128MB swap. Drive is a WD IDE (not
EIDE) hard disk. I compiled both 2.2.14 and 2.2.16 with SuSE's (stock SuSE 6.2) gcc
2.7.2 compiler.
There is a problem with kernel 2.2.9 in that it can enter a race condition (running
the swap, I understand) and eventually will crash. That happened to me a lot with
2.2.9, but never with 2.2.14 or 2.2.16.
Any comments or suggestions?
------------------------------
From: piyush mishra <[EMAIL PROTECTED]>
Subject: help regarding battery utilities
Date: Wed, 21 Jun 2000 05:30:06 GMT
hi friends,
i am a garduate student and am relatively new to linux. i have installed
the redhat version 6.2 on my laptop (ibm thinkpad 600e series). i am
looking for battery utilities/tools that can be used for studying the
power consumption characteristics of application running on linux. any
help/guidance will be much appreciated. thanking you for your time,
regards,
piyush
--
Posted via CNET Help.com
http://www.help.com/
------------------------------
From: Octalman - Wilbur Killebrew
Subject: Re: Packaging for Release
Date: Wed, 21 Jun 2000 05:58:12 GMT
Reply-To: [EMAIL PROTECTED]
[Posted and mailed]
In article <8ikr1m$9em$[EMAIL PROTECTED]>,
Avi Shmidman <[EMAIL PROTECTED]> writes:
> Hello,
> We have compeleted development of a Linux device driver for our PCI
> hardware (an avionics simulator), as well as a user mode API for
> interfacing with the card.
> The question now is, how should we distribute this? Should we use the
> redhat RPM scheme, or will that help only redhat users? Should we
> provide the source of the device driver, or just of the user mode API?
> Should we provide executable examples programs, or just source?
> Thank You,
> Avi Shmidman
> Excalibur Systems
SuSE and several other distributions use RPM, as well as Red Hat. For those
distributions that use RPM, it is an easy, convenient way to maintain. Debian
uses their own installer/maintainer. For the rest, .tar.gz files have to do.
If you want to make an RPM, it would help many (probably more than 50% of the
"market") but you would want to make the .tar.gz's available too, for those
that don't use RPM. If you don't want to build an RPM, even RPM-based
distributions can install and use .tar.gz's, they just aren't maintained in
the RPM database.
------------------------------
From: Michael Powe <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.networking
Subject: Re: Why so many collisioins ?
Date: 20 Jun 2000 23:10:00 -0700
=====BEGIN PGP SIGNED MESSAGE=====
Hash: SHA1
>>>>> "Chuck" == Chuck Swiger <[EMAIL PROTECTED]> writes:
Chuck> In comp.os.linux.networking Cedric Blancher
Chuck> <[EMAIL PROTECTED]> wrote:
>> "Boris Pran" <[EMAIL PROTECTED]> a �crit dans le message news:
>>> It works but, an enormous amount of collisions are detected.
>>> Collision led on the hub is on almost all the time. It
>>> significantly slowes down the communication.
Chuck> You've probably got a cabling problem or else your 10/100
Chuck> card is not autonegotiating to 10Mbs/half-duplex properly.
>> If your HUB is a 10/100Mbps one, D-Link card is ten time faster
>> than the other. It has to slow down for the other card not to
>> get flooded.
Chuck> A hub either runs at 10 or 100; a device which can handle
Chuck> both speeds is a switch, not a hub (by definition). The
Chuck> 10/100 card should simply run at 10Mbs; congestion control
Chuck> is not handled by the card per se, but by the TCP sliding
Chuck> window.
A switch, by definition, is a "smart" hub that will figure out which
machine is to receive a packet and send the packet to only that port,
rather than, in typical ethernet fashion, spraying the packet to every
port. Switches are most often used to connect different segments of a
network.
I have a 10/100 hub and some machines on it are plugging in at 100,
some at 10, depending on the card in the particular machine.
mp
- --
BOYCOTT AMAZON http://www.gnu.org/philosophy/amazon.html BOYCOTT AMAZON
"Public opinion's always in advance of the Law." -- Galsworthy
Michael Powe Portland, Oregon USA
=====BEGIN PGP SIGNATURE=====
Version: GnuPG v1.0.1 (GNU/Linux)
Comment: Mailcrypt 3.5.5/GnuPG v1.0.1 http://www.gnupg.org
iD8DBQE5UFwi755rgEMD+T8RAgRbAKCYoMR2q0BIQ2m1o6/29i0n1Vz1LQCfcP4E
5f013YYnQ/9U1hS4oq83v6k=
=FtVG
=====END PGP SIGNATURE=====
------------------------------
From: David Minor <[EMAIL PROTECTED]>
Subject: Re: OT: Re: CodeWarrior for Linux. Comments ?
Date: Wed, 21 Jun 2000 09:08:18 +0300
That's for sure.
CW is a great development environment, I used it for years on the Mac.
But it has one fatal flaw under Linux. It doesn't use standard Make!
Standard Make is what a mouse is to a Mac! Think of how successful CW
would have been on a Mac if it only had a command line interface. Get
with it guys! I can't even consider buying CW for my group without make
support.
David Minor
Bryan Hackney wrote:
> MWRon wrote:
> >
> [...]
> > Ron
> >
> > --
> > When it came time to show off Mac OS X applications
> > Steve Jobs chose those built with CodeWarrior
> >
> > Metrowerks, a Motorola Company - Ron Liechty
> > "Software Starts Here" - [EMAIL PROTECTED]
> [...]
>
> You won't win any friends or arguments, or impress anyone around here
> by invoking the name Steve Jobs.
>
> --
> Bryan Hackney / BHC / [EMAIL PROTECTED]
> http://www.FreeClassAds.com/
> http://bhconsult.com/
> http://bhconsult.com/bh/pgp.key
------------------------------
From: David Minor <[EMAIL PROTECTED]>
Subject: Thread Specific Memory
Date: Wed, 21 Jun 2000 09:14:46 +0300
Does thread specific memory provide protection for the blocks pointed to
by keyed variables, or are only the keyed variables themselves safe? I
suspect the latter but it's not mentioned in any of the documentations.
David Minor
Orbotech
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: How to mount floppey drive ?
Crossposted-To: linux.help,comp.os.linux.development.apps,comp.os.linux.setup
Date: 21 Jun 2000 03:39:55 -0400
In linux.help Robert <[EMAIL PROTECTED]> wrote:
> I am getting erro as mount: you must specify the file type when i am using
> the commad : mount /dev/fd0 /mnt
> i am also getting error as mount: /dev/fd0 is not a valid block when i am
> using the command
> mount -t ext2 /dev/fdo /mnt
1: Is the floppy formatted?
2: How? Is it an MSDOS disk?
When I want to mount an msdos or windows disk (I use vfat so it can handle
long file names) I use:
mound -t vfat /dev/fd0 /mnt/floppy
(or wherever I want to mount the directory)
(Oh - you have to have vfat support for that, don't you? With so many
different packages for Linux one never knows what is included. I use
TurboLinux 6 which installs vfat support if you ask for it during the
install.)
Getting the "not a valid block" indicates that it is trying to mount the disk,
but the file system is NOT "ext2" (or maybe there is no file system - if the
disk is not formatted - but heck, any disk you purchase now is probably
formatted for dos or windows).
so .. the third question is:
3: Have you got support for the disk's file format installed (e.g. vfat)?
------------------------------
From: Philippe Peyre <[EMAIL PROTECTED]>
Subject: kernel books
Date: 21 Jun 2000 11:49:21 +0200
I have found two books references on the Linux kernel.
Have you any information to help me to chose :
- "Linux Kernel Internals" M. Beck
- "the Linux Kernel book" R. Card, E. Dumas and F. Mevel
------------------------------
Date: Wed, 21 Jun 2000 09:48:18 +0200
From: Daniel Krippner <[EMAIL PROTECTED]>
Subject: 64bit division
Hi everyone!
I don't know if this is the right place for my problem... sorry if it
ain't.
I need to do a division of 64bit values in a module for x86 Linux. Is
there a function somewhere in the kernel that does this, or can anyone
give me a pointer to some assembly source?
cu
Daniel
--
[EMAIL PROTECTED]
Student of Communication Engineering/Computer Networking
Furtwangen University of Applied Sciences
------------------------------
Date: Wed, 21 Jun 2000 08:33:51 +0200
From: Arnaud Westenberg <[EMAIL PROTECTED]>
Subject: Re: kernel books
Philippe Peyre wrote:
>
> I have found two books references on the Linux kernel.
> Have you any information to help me to chose :
>
> - "Linux Kernel Internals" M. Beck
>
> - "the Linux Kernel book" R. Card, E. Dumas and F. Mevel
I have the latter, it's clear and understandable but perhaps a bit
boring. It deals with kernel 2.0 but this doesn't really matter because
you should read the source anyway.
Every chapter starts with the basic concepts and becomes more detailed
towards the end of the chapter discussing detailed description of
implementation.
I don't really have an opinion about it but it definately will help you
understand the kernel.
If you like I can send you the table of contents.
Greetings Arnaud
------------------------------
From: Gundu <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking
Subject: UDI
Date: Wed, 21 Jun 2000 14:03:23 +1000
Wanted to know if linux has a implemented Uniform driver Interface, If
so,
how can i use UDI for a IBM token ring NIC.
Thanks
------------------------------
From: "Mikko Jaakkola" <[EMAIL PROTECTED]>
Subject: top-half to bottom-half
Date: Wed, 21 Jun 2000 09:14:47 GMT
Hello everyone!
Has anyone measured how long it takes to transfer execution from top-half to
bottom-half. I know it is subject to entries on the immediate-queue and to
the processor-speed but I am sure someone has measured it.
Best Regards, Mikko
------------------------------
From: Ulrich Leodolter <[EMAIL PROTECTED]>
Subject: Howto debug kernel deadlock ?
Date: 21 Jun 2000 11:15:39 +0200
Problem:
Deadlock about once a month on a old Pentium 166
SCSI System, always under heavy network load
during ADSM backup (network card old NE2000)
There are absolute NO error messages in /var/log/messages
console is dead and SysRQ keys dont work.
SystemInfo:
RedHat 6.1
linux-2.2.14 (+ raid & nfs patches)
Any hints/suggestions how to debug/fix this problem?
Thanks
--
o------------------------------------------------o
Ulrich Leodolter
University of Vienna, Institute of Psychology
Email: [EMAIL PROTECTED]
Tel.: +43 1 4277 47828
o------------------------------------------------o
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: installing procinfo 17 (termcap.h missing)
Date: 21 Jun 2000 05:57:56 -0400
> From: <[EMAIL PROTECTED]>
> Reply to: [EMAIL PROTECTED]
> Date: Wed, 21 Jun 2000 04:30:07 GMT
> Organization: Posted via Supernews, http://www.supernews.com
> Newsgroups:
> comp.os.linux.development.system
> Followup to: newsgroup
>hi,
>i'm running RH 5.1. I downloaded procinfo17 and trying to install it.
>i made the changes to the Makefile, however the source files requires
>termcap.h. I don't seem to have it on my system.
>There is a comment in the Makefile that one may need to use ncurses. what
>is that? should i use the curses.h instead of termcap.h in the source file?
>thanks,
>aner
ncurses is a higher-level library that provides (among other things) a
termcap-style interface. There's a termcap.h to correspond - GNU termcap
also has a termcap.h file, so you really want to use the one that matches
the library you are using. In either case, you need the corresponding
development header-files.
The current version of ncurses is 5.0 (19991023)
There's an faq at
http://dickey.his.com/ncurses/ncurses.faq.html
------------------------------
From: Alan Donovan <[EMAIL PROTECTED]>
Subject: Re: top-half to bottom-half
Date: Wed, 21 Jun 2000 11:40:19 +0100
Mikko Jaakkola wrote:
> Has anyone measured how long it takes to transfer execution from top-half to
> bottom-half. I know it is subject to entries on the immediate-queue and to
> the processor-speed but I am sure someone has measured it.
It depends (of course). As you know the top-half _is_ the hardware
interrupt, running truly asynchronously to the system, and the bottom
half is one of a list of functions to be executed "when the swapper gets
around to it".
I found (with a logic analyser) that my tophalf, which simply resets the
HW interrupt and then marks the BH, consistently reset the HW interrupt
within about 5us (about 8us if sharing the IRQ with another device).
Looking at my notes, the BH was run typically about 400us after the
interrupt. However I'm not sure whether the various other numbers I have
relate to start times or durations of my BH, so I won't confuse you by
adding them. But certainly, 400us is a resonable figure for start time
although I can't qualify that with variance data etc.
These rather ad-hoc figures come from a 300MHz K6.
alan
--
========================================================================
Alan Donovan [EMAIL PROTECTED] http://www.imerge.co.uk
Imerge Ltd. +44 1223 875265
------------------------------
From: "Casper Helenius" <[EMAIL PROTECTED]>
Subject: Re: data acquisition boards
Date: Wed, 21 Jun 2000 12:11:35 GMT
Also, take a look at
http://www.tmworld.com/articles/2000/05_linux.htm
It's a two part article on developing drivers for DAQ-boards.
Combine it with the Linux Kernel Module Programming HOW-TO and you're well
on your way.
(I know, I was.. :-) )
Best regards,
Casper Helenius,
Denmark
goulmik skrev i meddelelsen <[EMAIL PROTECTED]>...
>Hi,
>
>I'm searching for any kind of informations about Linux and Data
>Acquisition boards
>known supported hardware
>configuration
>visualisation of plots collected
>Related websites
>books
>
>so if you can help me, feel free to contact me...
>thank
>
------------------------------
From: ecivasa <[EMAIL PROTECTED]>
Subject: Re: TCP question
Crossposted-To: comp.unix.programmer,comp.unix.programmer,comp.unix.solaris
Date: Tue, 20 Jun 2000 06:02:22 -0700
Hi Sonyea,
I am Srinivas, working in ERICSSON ( R & D ) Center in INDIA.
Yor Question is interesting and ofcourse a little bit confusing.
Here I am trying to answer your Questions to extent of my known
knowledge in TCP/IP. If you feel wrong, sorry for it and give me reply
to " [EMAIL PROTECTED]" by correcting.
1. What happens when a client tries to connect to a server, whose
quequ is already full. I understand TCP being a reliable
protocol will resend the SYN request again. Does this mean that
whatever the delay is between accept() on teh server
side, the client will ultimately ensure success even on a single
connect() by resending the SYN s for an infinite time. There
is a mention about 75sec on Richard Stevens. But in this case is
connect() going to return an ETIMEDOUT or what.
Answer:
when the client tries to connect to the server and the at the server
if the Que is full, the server discards the incomming messages. But as
you know, it is the responsibility of the client to retransmit the
packets once again. But the client not at all retransmit the same
packet infinite number of times. The client retransmit the packets by
using the time interval " Exponnentially back off ". There is a limit
to this "Exponential back off". Once that limit reached means there is
no furhter doubling in the Time interval to wait for receiving the "
SYN +ACK " segment from the server.Once Max Back off reached, here
onwards it sends the segments with this interval. But there is limit on
number of retries also. Hence the retransmission will be stopped as
soon as number of retries over.
2. The queues( ESTABLISHED & SYN_RECEVD ) are allocated on a system
wide basis or on a per socket basis. I m almost certain that it has to
be on a per socket basis since a system wide queue can result in a bad
process killing off thecomplete system. But I would like you guys to
confirm this.
Answer:
As you know the TCP/IP servers are concurrent servers,
that is, the main server creates a child process as soon as the
client's request is succeeded. The system may provide more than one
service. Hence the requests come to system for different services and
hence on different sockets.
The queues are allocated on a per socket basis, but there is
limitation on system as well.
* Sent from RemarQ http://www.remarq.com The Internet's Discussion Network *
The fastest and easiest way to search and participate in Usenet - Free!
------------------------------
From: [EMAIL PROTECTED] (Aki M Laukkanen)
Subject: Re: 64bit division
Date: 21 Jun 2000 12:57:09 GMT
In article <[EMAIL PROTECTED]>, Daniel Krippner wrote:
>I need to do a division of 64bit values in a module for x86 Linux. Is
>there a function somewhere in the kernel that does this, or can anyone
>give me a pointer to some assembly source?
No, sorry there is not. I've myself added lib/math.c and div64() there.
There was some discussion on lkml about this topic but I can not remember
if explicitly called div64() would be a candidate for inclusion in the
kernel. Kernel does not link with libgcc because if implicitly called
these functions could take a considerable time without the programmer
realizing it.
For example on a Celeron (Mendocino) the div64 operation takes about
550 cycles.
Just take gcc/libgcc2.c and extract __divdi3 from there or for
experiments link with libgcc.
--
D.
------------------------------
** 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
******************************
