Linux-Hardware Digest #321, Volume #10 Tue, 25 May 99 13:13:30 EDT
Contents:
some problem with my hd (�O�G�ۤv�D�R)
Re: All the current OSes are idiotic (was Re: Is Windows for idiots?) (westprog)
Modem ID ([EMAIL PROTECTED])
Re: i'm killed 3 cd-rom drives!!! ([EMAIL PROTECTED])
Re: 3c509b: How to turn off PnP? (David Fox)
OMNIS Studio RAD Tool available Soon...... ([EMAIL PROTECTED])
IBM 22 GByte hard disk (Nick Birkett)
Re: 3c509b: How to turn off PnP? ("jamal")
Re: Modem ID (Rob Clark)
Re: IDE faster than SCSI UW? (Andy Longton)
Re: Linux supporting UPS? (Peter)
Re: ASUS P2BD vs. SuperMicro P6DBE (bryan)
Re: 3c509b: How to turn off PnP? ("jamal")
Data Acquisition Card + Counter ("Patrick Riphagen")
Re: MSI motherboard compatibility with Linux 2.2? (Marc Mutz)
DLT 7000 performance problem under linux (hvergonet)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (�O�G�ۤv�D�R)
Subject: some problem with my hd
Date: 24 May 1999 16:50:51 GMT
I install linux (RedHat 6.0 and kernel 2.2.8) in hdb
and I divide into two partition(linux native and swap)
and oneday I remove my hdb from my computer(for learning)
and re-plug in it.
However, when I start linux, some problem happens in the progress
of setup and loading
error message:
Warn:unable to open an initial console
kernel panic:No init found. Try passing init=option to kernel
please help with my linux and my hd
I can't stand reinstalling RedHat for I have important data
in hdb
so...
--
[1;33;41m�� Origin: �F�j�̯�����BBS�� 140.119.185.146 �� From: Full.Dorm9.nccu.edu.tw
------------------------------
From: westprog <[EMAIL PROTECTED]>
Crossposted-To:
comp.lang.java.advocacy,comp.os.linux.advocacy,comp.os.ms-windows.advocacy
Subject: Re: All the current OSes are idiotic (was Re: Is Windows for idiots?)
Date: Tue, 25 May 1999 15:36:53 GMT
In article <7icb8q$rh9$[EMAIL PROTECTED]>,
"Osvaldo Pinali Doederlein" <[EMAIL PROTECTED]> wrote:
> westprog <[EMAIL PROTECTED]> wrote in message
> news:7ic2lb$9i4$[EMAIL PROTECTED]...
> > "Osvaldo Pinali Doederlein" <[EMAIL PROTECTED]> wrote:
> > I think that the best that can be done is to restrain programs as to
> > what they can do. To specify exactly what you want the program to
> > do,
> > and disallowing it from doing anything else, is affectively the same
> > thing as writing the program. It is unfortunate but true that once
> > we
> > allow programs to do something useful, we allow them to do something
> > harmful.
> <science-fiction>I think in the future we can have some kind of formal
> specification of everything that our programs do, and operating
> systems/runtimes/VMs/whatever will be able to verify than some binary
> abides
> to that spec. For example, UML and CORBA are both moving to make some
> constraints part of software components, then we would need lots of
> intelligence in the runtimes to check constraints on binaries...
> bytecode
> would be probably the only path for software distribution. Users very
> concerned with security, say, some militar facility, would have
> administrators capable to read the formal specs of each and every
> software
> they get, and tell if it looks safe.</science-fiction> But there's a
> long way to go here.
There seem to be two approaches to get there - new systems like Java
that sit on top of existing systems, and develop new ideas, and
completely new systems built from the ground up. I think that both are
required, as people won't accept completely new systems initially.
I believe that the security found in Java bytecode can be replicated in
native code, but only with
> > > This is being done by any respectable OS for a long time. It
> > > seems
> > > the idea of Spin is making application code safe even running it
> > > on _privileged_ mode.
> > Safer, perhaps. Of course, it always depends what level of
> > > privilege.
> > I envisage systems where programs have much finer-grained levels of
> > privilege, and that privilege would even vary within programs - a
> > database program could use a component and decide how much to trust
> > it,
> > possibly denying it same privileges as the rest of the program, even
> > while accessing it in-process. I don't know if that is available
> > with Javabeans yet.
> In Java you can define your own permissions and security managers.
> Generally, if you have any central coordinator for some kind of object
> e.g. by using factories -- you can easily insert all sorts of checks.
> There
> are tricks like controlling inheritance on class loading time, so you
> can
> prevent attacks like defining a derived class that override the secure
> bits of its parent with unsecure code. The Java core libs do this a
> lot, you
> can't override freely some classes that are not 'final'. And each
> bean can protect itself from bad collaborators, e.g. using
> PropertyVetoException
> which is usually thrown in the "invalid value for this attribute"
> situation, but can also be smart enough to throw in the
> "unauthorized sender" situation.
One of the things I would love to see in the new wonder operating system
are system APIs which propagate exceptions rather than returning status
codes. I would also like to see the APIs in an object format rather than
a
> > > I guess it
> > > would be a good idea to run them on kernel mode on protected
> > > operating
> > > systems, bypassing lots of overhead that's useless because the JVM
> > > takes care of the safety belts.
> > Depends how trusting you are, and how thoroughly the code has been
> > tested. There is always the possibility of some unforeseen problem
> > arising. I would prefer to have nothing running in kernel mode but a
> > simple scheduler.
>
> Well it seems the microkernel religion is dying, since the major OS to
> claim
> to do it (WinNT) has some megabytes of code running in kernel mode,
> and Linus decided that the whole idea sucks and it's not even worth
> paying lip service. ;-)
I think that the designer of a new OS could do worse than find all the
points of commonality between NT and Linux (and there are many) and
eliminate them from the new system.
> I don't think the inclusion of an entire JVM would make
> most
> OSen significantly less robust. And we don't need to put in kernel
> mode
> things like jitters; it suffices to have the performance-critical
> parts
> (interpreter, native libs, GC). Native libs which are not really
> performance critical could be transformed in user-mode services, e.g.
> JDBC drivers.
>
> And we can probably get big advantage on integrating the garbage
> collector tightly with the OS's virtual memory manager. This is an
> old dream, there
> is a lot of research on paging-friendly, OS-friendly, cache-friendly
> GC; but
> not a lot of implementation as never before Java a GC system was
> popular
> enough to justify such surgery on any major (commercial) OS. I would
> expect
> Sun to try that first, or maybe IBM... if the idea is good anyway; I
> only know enough to say that the idea is cool. ;-)
There are a lot of system services that could be provided in this way.
Back in the bad old days of Dos, Wordperfect and 1-2-3 and all the other
programs had to implement their own printer drivers. Then Windows 3.0
came along, and you just needed one driver per printer. There are dozens
of wheels out there being reinvented every day. Garbage collection is
just one example.
--
J.
--== Sent via Deja.com http://www.deja.com/ ==--
---Share what you know. Learn what you don't.---
------------------------------
From: [EMAIL PROTECTED] ([EMAIL PROTECTED])
Subject: Modem ID
Date: Tue, 25 May 1999 15:54:03 GMT
Reply-To: [EMAIL PROTECTED]
I will soon be making the switch to Linux and am taking inventory of
my hardware to see if all is compatible. I am currently running
Windows NT 4.0 and think everything will be compatible except my
modem. I want to find out if it is, but am not sure how. It is not
the original that shipped with my Micron PC, it is a replacement
USR/3COM internal fax/modem I bought from a local custom 'puter shop.
NT 4 identifies it as a U.S. Robotics 56K FAX INT modem. However, I
cannot find out any more information. The disks that came with it
(they have the .ini driver files) have the following id on the labels:
AKITA (1787-81)/PYTHON(1749-81)
P/N: 1.018.1500-00
Anyone know if this is a winmodem or used this modem before?
Thanks in Advance,
Pearce
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
Reply To: pearce AT email DOT com
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
LEGAL NOTICE: Pursuant to U.S. Code, Title 47,
Chapter 5, Subchapter II, Section 227, any
and all unsolicited commercial email sent
to this address is subject to a fee
of $500.00 U.S. Emailing denotes
acceptance of these terms.
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux,comp.os.linux.admin,comp.os.linux.help
Subject: Re: i'm killed 3 cd-rom drives!!!
Date: Tue, 25 May 1999 15:14:32 GMT
By now I`m at CD-rom drive 4, 5 and 6 on my 3 computers, 3 of the first
drives died within the first few months I had them, two while playing
games under one of the MS-pseudo-OSes and the third while listening to
an Audio-CD under linux.
The lesson ? CD-rom drives are fragile beasts but fortunately any
reputable vendor will replace them at no charge if they are not too
old...
\Gandalf
--== Sent via Deja.com http://www.deja.com/ ==--
---Share what you know. Learn what you don't.---
------------------------------
From: d s f o x @ c o g s c i . u c s d . e d u (David Fox)
Subject: Re: 3c509b: How to turn off PnP?
Date: 25 May 1999 08:25:19 -0700
[EMAIL PROTECTED] writes:
> In article <[EMAIL PROTECTED]>,
> Michael Gibson <[EMAIL PROTECTED]> wrote:
> >You should have gotten a disk with the card. On it is a DOS program called
> >3c5x9cfg. If you run this it will allow you to configure/test the card. Under
> >the configure options is the ability to turn off PnP and set static resources.
>
>
> What about for purists? If you're going to require DOS before you can
> setup Linux, you've required MS into the equation. Ugly, ugly. (Not
> 3com's or Linux' fault really...)
http://rpmfind.net/linux/RPM/contrib/libc6/i386/3c5x9utils-1.0-3.i386.html
--
David Fox http://hci.ucsd.edu/dsf xoF divaD
UCSD HCI Lab baL ICH DSCU
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.m68k,comp.os.linux.misc
Subject: OMNIS Studio RAD Tool available Soon......
Date: Tue, 25 May 1999 16:06:58 GMT
****************************************************************
* Posted by Newsgroup AutoPoster! It's NOT registered yet! *
****************************************************************
Please see the following link to read our latest press release regarding the
forthcoming launch of OMNIS Studio on the LINUX operating system platform.
http://www.omnis-software.com/whatsnew/press/linux.html
The beta version will be available from July 1999, with full release in
September this year.
We believe we have a real first here, as OMNIS Studio represents a true
4GL Rapid application development system that is binary compatible with
both Windows and Macintosh machines, and soon LINUX.
This means that developers are totally free to choose their preferred
development platform, and then deploy applications without alteration in
all of the above environments.
If you require any more information about OMNIS Studio, or would like to
know about some of the many commercial applications written in OMNIS
that will soon be available for the LINUX environment please don't
hesitate to contact me.
Kind Regards,
Richard Darsa
OMNIS Software Ltd.
mailto:[EMAIL PROTECTED]
------------------------------
From: Nick Birkett <[EMAIL PROTECTED]>
Subject: IBM 22 GByte hard disk
Date: Tue, 25 May 1999 16:48:24 +0100
We recently bought a 22Gbyte hard disk for use under Linux.
I can partition this and format it fine. It is used as scartch storage
hence only one huge parition (we have some enourmous files).
The disk : IBM 22Gbyte, 7200 rpm, 2Mbyte cache UDMA.
Problem - seems slow but this may be due to the large paritition.
hdparm -t /dev/hda1
/dev/hda1:
Timing buffered disk reads: 32 MB in 10.40 seconds = 3.08 MB/sec
This is also the same figure I get timing a 600Mbyte file transfer.
I was expecting something like 10-12MB/s.
Any tips for tuning gratefully received.
Thanks,
Nick
Please remove NOSPAM from email reply address.
===========================================================
Other info:
Boot messages
kernel: hda: IBM-DJNA-372200, ATA DISK drive
kernel: hdc: Hewlett-Packard CD-Writer Plus 8100, ATAPI CDROM drive
kernel: hdd: IOMEGA ZIP 100 ATAPI, ATAPI FLOPPY drive
kernel: ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
kernel: ide1 at 0x170-0x177,0x376 on irq 15
kernel: hda: IBM-DJNA-372200, 21557MB w/1966kB Cache, CHS=43800/16/63,
(U)DMA
kernel 2.2.x
------------------------------
From: "jamal" <[EMAIL PROTECTED]>
Subject: Re: 3c509b: How to turn off PnP?
Date: Wed, 26 May 1999 00:17:16 +0800
i finally got the file at http://sam.via.ecp.fr/linux/formation/fips/
haven't try the program yet
jamal
jamal <[EMAIL PROTECTED]> wrote in message
news:7iei0c$e9u$[EMAIL PROTECTED]...
> surprisingly guys, i bought two of them.
> no diskette at all.....
> i went to 3com homepage, can't locate the 3c5x9cfg file
> if it is in DOS, that's mean, i had to install MSDOS first?
>
> jamal
------------------------------
Subject: Re: Modem ID
From: [EMAIL PROTECTED] (Rob Clark)
Date: Tue, 25 May 1999 16:24:57 GMT
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
>I will soon be making the switch to Linux and am taking inventory of
>my hardware to see if all is compatible. I am currently running
>Windows NT 4.0 and think everything will be compatible except my
>modem. I want to find out if it is, but am not sure how. It is not
>the original that shipped with my Micron PC, it is a replacement
>USR/3COM internal fax/modem I bought from a local custom 'puter shop.
>NT 4 identifies it as a U.S. Robotics 56K FAX INT modem. However, I
>cannot find out any more information. The disks that came with it
>(they have the .ini driver files) have the following id on the labels:
>
>AKITA (1787-81)/PYTHON(1749-81)
>P/N: 1.018.1500-00
Both the Python and Akita modems are real modems with jumpers. Whichever
one you have, it should be fine with Linux. (Be sure to disable PnP on
the modems and use the jumpers-- it's just easier).
Rob Clark, [EMAIL PROTECTED]
http://www.o2.net/~gromitkc/winmodem.html <--Linux/modem compat. list
------------------------------
From: Andy Longton <[EMAIL PROTECTED]>
Subject: Re: IDE faster than SCSI UW?
Reply-To: [EMAIL PROTECTED]
Date: 25 May 99 12:58:49 GMT
Patrick Colbeck <[EMAIL PROTECTED]> wrote:
> A single SCSI drive will not be much of an improvement if at all over
> a single EIDE drive. For a single user system I would normaly
> recommend and EIDE drive as its cheaper as fast as SCSI. Where SCSI
> really scores is when you have multiple drives chained of a SCSI
> controller.
Was I reading the results from Bonnie properly?
In sum: The amount of data transfered -- not the type of bus -- is the
most important factor when determing performance for a single drive. (?)
> The point about SCSI is that the drives themselves are intelligent as
> compared to EIDE drives which are dumb (thats one reason why they are
> cheaper - less circuitry). If you have several SCSI drives on a chain
> the controller can just say to the first drive get me these blocks of
> data then disconnect from the drive and move on to the next and ask it
> fro the data, and so on. It can then reconnect to the first drive and
> the blocks it wanted will be ready. With an EIDE drive the controller
> has to handhold the drive and tell t exactly what to do and so has to
> wait till the drive gets it the data before moving on to the next
> drive.
Well, I have two SCSI-UW systems and have noticed a benifit on the
slower machine (AMD 486-166). On the faster one (Pentium II 450),
with busmastering IDE, the spindle speed tied with the CPU speed
seem to have the largest impact.
Since there is only so much CPU available, and the busmastering
available under IDE seems to have about the same demands as SCSI, why
not use two UDMA drives in an IDE system ... and only move to SCSI
when both available chanels are used up. This would require some
thought and planning when setting up the hardware and partitions but
I have to do this anyway to properly configure a SCSI system.
Of course, I could have my Adaptec 2940 misconfigured! Any tuning
tips or URLs would be appreciated! (Esp. RAID 0 optimal cluster size
for the IBM and Viking drives ... and advice on if it's even worth it
to use these two disimmilar drives in a RAID 0 configuration.)
> Also SCSI can do neat tricks like device to device copying without
> going through the system bus. For example two SCSI drives or a SCSI
> drive and a tape streamer can copy data between themselves directly
> without any hit on the host system CPU (well except for a bit to kick
> the copy into life in the first place etc).
I'm getting an occasional "hickup" when accessing the CD-ROM. I've got
the system loaded; UMAX 1220s scanner, Toshiba 32x CD, Yamaha 4260t
CDRW, and the IBM and Viking 4.5g drives. I can get the actual settings
for the devices if necessary (Guesstimate: all devices have disconnect
enabled though the bus speeds differ (40mbs/drive, 10mbs/CD,
??mbs/scanner. Except for ID 0 (Viking), the other devices are placed
in on IDs 1-6 in no specific order.).
> This is of course a terrible over-simplification and there are other
> pros anc cons but a good rule of thumb is that SCSI drives will
> outperform EIDE ones in multiuser systems or where there is more than
> one drive.
------------------------------
From: Peter <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup
Subject: Re: Linux supporting UPS?
Date: Tue, 25 May 1999 12:07:33 GMT
> Anyone know if there is a built-in support for an UPS unit, in Linux?
> I want to have Linux making a nice halt if receiving a signal from the
UPS,
> connected preferrably
> to a serial port.
> I use SuSE 6.1 with kernel 2.2.1.
I've been using apcupsd daemon (can't locate the readme files, so unable
to provide the URL) with APC SmartUPS via serial link without a hitch
for about half a year now (on a RedHat 5.3 box). It is supposed to
support 'dumb' UPS as well, and can shutdown other networked
workstations in case of power failure. Try some search engine for
'apcupsd' and it should come up.
BTW, it's free.
Peter
--== Sent via Deja.com http://www.deja.com/ ==--
---Share what you know. Learn what you don't.---
------------------------------
From: bryan <[EMAIL PROTECTED]>
Subject: Re: ASUS P2BD vs. SuperMicro P6DBE
Date: Tue, 25 May 1999 13:08:22 GMT
supermicro shipped me a board with such an old bios (and there were no
newer ones on their site). stupid stuff like not seeing drives >
8gig. annoying.
asus never does that - their bios is always current and accurate. its
worth the money.
Maxim Bazhenov <[EMAIL PROTECTED]> wrote:
: I am planing to buy a dual P-III system to use with Linux (RedHat) and
: I am deciding now
: between two motherboards: ASUS P2BD and SuperMicro P6DBE. It looks like
: they have pretty
: similar features, however P6DBE is less expensive (by about $70). Are
: these two motherboards
: both compatible with Linux? Does anybody has any good/bad experience
: with any of them
: while running Linux? Thanks a lot for any advice.
: maxim
: P.S. Please, replay also to my e-mail address: [EMAIL PROTECTED]
--
Bryan
------------------------------
From: "jamal" <[EMAIL PROTECTED]>
Subject: Re: 3c509b: How to turn off PnP?
Date: Wed, 26 May 1999 00:04:37 +0800
surprisingly guys, i bought two of them.
no diskette at all.....
i went to 3com homepage, can't locate the 3c5x9cfg file
if it is in DOS, that's mean, i had to install MSDOS first?
jamal
Don Baccus <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> In article <7ibr2o$3uc$[EMAIL PROTECTED]>, jamal <[EMAIL PROTECTED]> wrote:
> >I am using 2 3c509b. I can't get the network working.
> >I've tried adding append in lilo.conf & alias in conf.modules.
> >Still not working.......... Can somebody tell me how to off
> >the PnP on the card?
>
> Use the configuration program on the floppies that came
> with the card. It's named 3c5x9cfg or something similar.
> --
>
> - Don Baccus, Portland OR <[EMAIL PROTECTED]>
> Nature photos, on-line guides, at http://donb.photo.net
------------------------------
From: "Patrick Riphagen" <[EMAIL PROTECTED]>
Subject: Data Acquisition Card + Counter
Date: Tue, 25 May 1999 14:16:16 +0200
Hello,
Can someone tell me if a counter on a DIO card needs a driver?
I have a linux driver for the Computer Boards CIO-DIO 24.
The same board also exists with a counter, but I don't know if a seperate
driver is needed to use this.
------------------------------
Date: Tue, 25 May 1999 14:58:57 +0200
From: Marc Mutz <[EMAIL PROTECTED]>
Subject: Re: MSI motherboard compatibility with Linux 2.2?
Carl Kumaradas wrote:
>
> I'm purchasing a system that comes with an MSI MS-5169 motherboard,
> which uses an Aladdin-5 M1531/M1543 chipset. It will have a K6-2/400
> CPU.
>
> One of the changes in Linux 2.3.1 listed at LinuxHQ is "IDE ALI M15x3
> chipset support added" which worries me a bit since it implies that
> this cipset is not supported in previous kernel versions (which I will
> use).
>
> Does anyone know if Linux 2.2 supports this chipset? Am I worried for
> no reason?
>
Linux does not need to be configured for a specific motherboard to run
correctly. However, if you want to use the full power of your *IDE*
drives (SCSI is another matter), then Linux needs to have drivers for
the onboard-IDE-channels, so's it can put them from PIO (=slow) to
DMA-mode (=fast). But that's not an issue if you are running a home
computer. It becomes an issue if you're running a server or a
workstation, where processor time is critical and cannot be thrown away
to load data from to HD. But then you should use SCSI...
Marc
------------------------------
From: hvergonet <[EMAIL PROTECTED]>
Subject: DLT 7000 performance problem under linux
Date: Tue, 25 May 1999 14:17:45 +0200
Dear Linux users,
A:
# cat files | cpio -ocav | dd bs=32k of=/dev/st0
24478883 blocks # cpio (output) 512 byte blocks
0+3336377 records in # dd (output)
0+3336377 records out
results in a 12.5GB (1G = 10^9) backup and uses 7h30, or 28 Meg per
minute. The DLT 7000 specs. state that the drive is capable of a
sustained data transfer of 600 MB per minute (with compression). During
backup the drive continuously spins up and down, maybe we are unable to
feed the drive with enough data.
We have tried to tune the parameters in /etc/stinit.def but didn't
notice any speed improvements.
/etc/stinit.def
# The common definitions that can usually be used
{buffer-writes read-ahead async-writes }
# Tandberg DLT 7000
manufacturer=QUANTUM model = "DLT7000" {
buffering can-bsr can-partitions auto-lock
mode1 blocksize=0 density=0x85 # 85937 bpi, 52 quad tracks, serial
cart tape 70 Gb, compression
mode2 blocksize=0 density=0x84 # 85937 bpi, 52 quad tracks, serial
cart tape 35 Gb, no compression
}
Please is there someone who could you help us in determining the
problem, is it a Linux problem ? Is there a way to boost the backup
speed ? Does anyone have experience with DLT 7000 drives ?
Regards,
Henk
--- system configuration ---
- cat /proc/version
Linux version 2.0.34 ([EMAIL PROTECTED]) (gcc version 2.7.2.3) #1
Fri May 8 16:05:57 EDT 1998
- mt -f /dev/nst0 status
SCSI 2 tape drive:
File number=0, block number=0, partition=0.
Tape block size 0 bytes. Density code 0x1b (unknown).
Soft error count since last status=0
General status bits on (41010000):
BOT ONLINE IM_REP_EN
- cat /proc/scsi/scsi
Host: scsi0 Channel: 00 Id: 01 Lun: 00
Vendor: QUANTUM Model: DLT7000 Rev: 2151
Type: Sequential-Access ANSI SCSI revision: 02
Henk Vergonet.
--
REDWOOD Software and services B.V.
phone:+31.30.6354573
fax :+31.30.6354550
http://www.redwood.nl
------------------------------
** 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.hardware) 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-Hardware Digest
******************************
