Linux-Hardware Digest #850, Volume #10 Mon, 26 Jul 99 09:14:08 EDT
Contents:
Re: turn a computer into a network hub? (James Knott)
Re: Occasional timeouts DMA'ing to IDE drive (Rintala Matti)
Re: spindle count and RAID performance [was: RAID controler-RAID 5 (Helge Hafting)
Re: Asus AMD K6-2 450Mhz 128K system ("Peter Christy")
Re: Epson Stylus Colour 440 (Duncan Grisby)
GDI Printer support (Ralf Wienzek)
Re: What is LVD SCSI? (Helge Hafting)
ECC-SDRAM: Supported? (Thomas Stieler)
Re: #@$%&%( WINMODEMS are a pain in the A#$^& (Helge Hafting)
Re: Ensoniq AudioPCI and microphone weirdness ? (Petter Reinholdtsen)
Re: Mic working on Ensoniq AudioPCI es1370? (Petter Reinholdtsen)
Re: Poem about a former roommate (John Thompson)
Re: Need support for Asus V3800 Riva TNT2. ("Global Refund Portugal")
Re: spin down HDD (Helge Hafting)
Re: kernel panic: unable to mount root fs on 03:42 ([EMAIL PROTECTED])
Re: building efficient fileserver on Linux - hardware questions (Warner Bruns)
can't compile SMP driver with -D__SMP__ in 2.2.X ("robert_c")
Re: ECC-SDRAM: Supported? (Abdullah Ramazanoglu)
----------------------------------------------------------------------------
From: [EMAIL PROTECTED] (James Knott)
Crossposted-To: comp.os.linux.networking,
Subject: Re: turn a computer into a network hub?
Date: Sun, 25 Jul 1999 11:48:32 -0400
Reply-To: James Knott <[EMAIL PROTECTED]>
In article <[EMAIL PROTECTED]>,
[EMAIL PROTECTED] (John Doe) wrote:
>On Tue, 20 Jul 1999 15:45:32 +1000, dkwok <[EMAIL PROTECTED]> wrote:
>paranoids think that anyone who does that is doing a commericial although
>it is true that i get about $100 comission for each person from this
>newsgroup visiting my sponsor. also just to be completely ethical i
>should say that i have a very sophisticated way to figure out the
>effectiveness of my commercial. basically it involves a spy satellite
>watching every computer user on earth.
So, your real name is ***BILL GATES***!!! Now we know what all those
satellites are really for. ;-)
--
E-mail [EMAIL PROTECTED]
_________________________________________________________________________
The above opinions are my own and not those of ISM Corp., a subsidiary of
IBM Canada Ltd.
------------------------------
From: [EMAIL PROTECTED] (Rintala Matti)
Subject: Re: Occasional timeouts DMA'ing to IDE drive
Date: 26 Jul 1999 11:32:26 +0300
In article <[EMAIL PROTECTED]> "Gerson G. H. Cavalheiro"
<[EMAIL PROTECTED]> writes:
> Jerome problem:
> Jul 24 15:12:54 jbuhler kernel: hda: timeout waiting for DMA
> Jul 24 15:12:54 jbuhler kernel: hda: irq timeout: status=0xd0 { Busy }
> Jul 24 15:12:55 jbuhler kernel: ide0: reset: success
I've had the same problem with my Gigabyte PentiumII motherboard (with
the Intel 440BX chipset, I think) and my Seagate ST36530A drive. It
has been happening occasionally with all kernels I've tried up to (and
including) 2.2.10. Last weekend I even patched the kernel with latest
UDMA patches, but the problem remained.
Sorry I cannot be more specific, I'm away from my machine now.
============ Matti Rintala =========== [EMAIL PROTECTED] ============
"Your reality, sir, is lies and balderdash and I'm delighted to say
that I have no grasp of it whatsoever."
(Baron von Münchausen in "The Adventures of Baron von Münchausen")
------------------------------
From: Helge Hafting <[EMAIL PROTECTED]>
Subject: Re: spindle count and RAID performance [was: RAID controler-RAID 5
Date: Mon, 26 Jul 1999 11:14:28 +0200
You missed an optimization when doing small writes. With 20 drives you
don't have to re-read from every drive order to recalculate parity.
Just read the stripe you are about to overwrite, and the parity stripe.
Then xor the parity stripe (in memory) with the stripe you are going
to overwrite. Now the old stuff is effectively removed by calculation,
and you
have the same parity stripe you would get if you xored the 18 other
drive stripes - except you didn't have to read them. Now
xor in the new stripe, and write the new stripe and the parity stripe
back.
You get no more than two reads and two writes for a single small write -
no
matter how many drives there is in your RAID-5 array.
The time loss may be even smaller, as the two reads is done in parallel,
same for the two writes.
Scott Marlowe wrote:
[...]
> This isn't how I've learned RAID 5 theory. What I learned goes like this:
>
> Each write, you create a data stripe for all but one drive, create a parity
> stripe by XORing all the data stripes, and place the parity stripe and data
> stripes on the drives. The parity stripe round robins across all the drives,
> so that it looks like this for each write:
>
> A B C D
> D D D P
> P D D D
> D P D D
> D D P D
[...]
> Another thing that can kill you on a large drive number raid array is small
> writes. If you have a database that does changes in the range of hundreds of
> bytes at a time, and you've got say, 20 drives with a stripe size of 4k, you've
> got to retreive the 20x4k of data, change 1 or 100 bytes, recalculate the
> parity strip, and write out the 20x4k. Reads are fast, but all writes require
> this type of operation. If you're building a FTP server where the average file
> size is 100k and above, then the 20x4k is no big deal, and you would likely see
> good performance there.
>
------------------------------
From: "Peter Christy" <[EMAIL PROTECTED]>
Subject: Re: Asus AMD K6-2 450Mhz 128K system
Date: Mon, 26 Jul 1999 08:46:03 +0100
Hmmm! Not sure if this will help or not, as my board is a P5A-B and it
sounds as if it may be slightly different. I'm running a K6-2 350
overclocked to 400 without problems.
According to my instructions, there are two sets of voltage jumpers VI00 &
VI01 (near the SIMM slots) set the volts for the SIMMs, 3.5 volt by default,
which is a jumper across pins 1&2 of VI01 only.
The processor volts are set by FOUR sets if pins, VID0, VID1, VID2 & VID3
which are located close to the audio and fan connectors. According to my
book (which is a little out of date, as it only covers up to the 333!) all
the AMD .25 micron processors use 2.2v dual plane, and the jumpers should be
VID0 1-2, VID1 2-3, VID2 1-2, VID3 1-2.
I seem to remember looking up more up to date info on the Asus website when
I fitted my K6-2. I don't recall that there were any differences, but I just
used the recommended settings and it all works fine. As I said, my board is
a -B variant and may be slightly different. Hope this helps.
Pete
[EMAIL PROTECTED]
------------------------------
From: Duncan Grisby <[EMAIL PROTECTED]>
Subject: Re: Epson Stylus Colour 440
Date: 26 Jul 1999 10:15:01 +0100
In article <[EMAIL PROTECTED]>,
Doug Kelly <[EMAIL PROTECTED]> wrote:
>Can I infer from the above that there is a way to put the 740 into a
>mono mode? I just recently successfully managed to print via Samba to a
>740 on a Win95 machine, but anything other than plain text is printing
>using the color jets. Is there a way to force the use of black ink
>only? I did a bit of looking at the upp file and tried printing to file
>from windows and trying to identify a 'page start string' but to no
>avail (not that I spent that much time on it....) . Any pointers
>greatly appreciated, both from a quality and $$$-spent-on-ink
>perspective.
Well, you can successfully print using black only on a 440 by putting
-dupColorModel=/DeviceGray
-dupRendering=/ErrorDiffusion
-dupWhiteTransfer="{ 1.0 0.0 }"
in the .upp file. For some reason greyscale printing is white-based
rather than black-based, so without the transfer function, everything
comes out white on black.
However, I don't think this is the same as a mono mode, since I
haven't been able to convince it to use 64 jets, only the 21 you can
get in colour.
Cheers,
Duncan
--
-- Duncan Grisby \ Research Engineer --
-- AT&T Laboratories Cambridge --
-- http://www.uk.research.att.com/~dpg1 --
------------------------------
From: Ralf Wienzek <[EMAIL PROTECTED]>
Subject: GDI Printer support
Date: Mon, 26 Jul 1999 10:59:55 +0200
Hi,
anyone heard about a support for GDI Printer
(especially for NEC Superscript 610)?
Thanks in advance
Ralf
------------------------------
From: Helge Hafting <[EMAIL PROTECTED]>
Subject: Re: What is LVD SCSI?
Date: Mon, 26 Jul 1999 11:27:16 +0200
John Jacques wrote:
>
> Hello, I've been looking over SCSI controller cards and the newer models
> have LVD support. What does this mean? Thanks!
LVD = Low Voltage differential.
The low voltage part means lover voltage than older scsi equipment.
The differential part means 1 and 0 are signalled as a voltage
difference
between two wires instead of using a common ground for all the lines.
This allows up to 12 meters of cabling while going 80MB/s, which is
good.
The non-LVD setups have a problem that higher speed require shorter
total
cable length, while maintaining a large minimum distance between drives.
This means that you can't even connect 7 fast scsi-2 drives (10MB/s
max!)
because the minimum distance between drives forces a longer than maximum
cable. This gets worse with higher speeds, resulting in "mysterious"
trouble
when the os actually is capable of using all the devices simultaneously.
LVD solves this with much longer maximum cabling, allowing more devices
as well as being convenient in big cases or with the separate drive
cabinets
you need if you actually install 15 drives.
Helge Hafting
------------------------------
From: Thomas Stieler <[EMAIL PROTECTED]>
Subject: ECC-SDRAM: Supported?
Date: 26 Jul 1999 10:09:41 GMT
Hello,
I have some problems with defect memory-modules.
Know I want to know, in which way linux is able to support
ECC-memory-modules.
>From what I know these modules can correct one-bit-errors and can
recongnize two-bit errors.
Know my question: how is the behavior of the linux-kernel?
I think, that an two bit error should trigger an NMI, and the system
should be halted immediatly.
Thats good, or better than let the system working until the random
errors crash the system...
But of course it would be the best, if the system will tell you about
one-bit-errors, so you can see, that the memory is unstable!
I have searched for some information about this, but I couldn't find
infos about that!
Can anybody help me and give me some information??
regards,
Thomas
------------------------------
From: Helge Hafting <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,at.linux
Subject: Re: #@$%&%( WINMODEMS are a pain in the A#$^&
Date: Mon, 26 Jul 1999 11:41:24 +0200
Scott Marlowe wrote:
>
> First, an apology to recross posting this across groups, but Little Fish, you
> need to take some responsibility here too. If I went down to my Local Chevy
> Dealer and ordered a 10 Ton moving truck and told him, yeah, but I'm gonna be
> using it to haul "Bananas in Brazil", I'm not sure he'd know the exact way to
> setup that truck for that, and would likely tell me so. I'd have to explain to
Exactly. He would *tell* that he don't know the banana business.
> him all the intricacies of hauling "Bananas in Brazil" (A musical group, by the
> way) he might have a better understand of what I need.
>
> It's your job to know which modems are and are not winmodems right now. Why?
> Because no sales people know, they don't care, and you and I are still a tiny
> blip on the Radar screen of their sales pie chart.
It is still the salespeoples responsibility to say something like
"huh - winmodems? linux? We don't understand that part of your
requirement,
please check it yourself/be clearer" instead of just handing out a
modem,
claiming it is ok for the *specified* use, hoping there will be no fuss
afterwards.
If they don't understand about linux, try asking for a modem that will
work equally
well with NT, with DOS, and with a macintosh. The salespople should
know those,
and I don't think anybody has software drivers for all of them.
Helge Hafting
------------------------------
From: [EMAIL PROTECTED] (Petter Reinholdtsen)
Subject: Re: Ensoniq AudioPCI and microphone weirdness ?
Date: 26 Jul 1999 09:46:32 GMT
[Gabriel L. Somlo]
> I recently got an Ensoniq AudioPCI ES1371. When I tried to record stuff
> with a regular microphone from BestBuy, it didn't work.
I just discovered that I had to make the microphone a recording source
to get it to work. I used kmix, the KDE mixer, to make sure the only device
marked red (RecSource) was the microphone and it started working. :-)
--
##> Petter Reinholdtsen <## | [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] (Petter Reinholdtsen)
Subject: Re: Mic working on Ensoniq AudioPCI es1370?
Date: 26 Jul 1999 09:47:55 GMT
[Petter Reinholdtsen]
> I'm trying to get the microphone working on my Ensoniq AudioPCI card.
I just discovered how: Make the microphone a 'RecSource' using kmix,
the KDE mixer.
Stupid default setting!
--
##> Petter Reinholdtsen <## | [EMAIL PROTECTED]
------------------------------
From: John Thompson <[EMAIL PROTECTED]>
Subject: Re: Poem about a former roommate
Date: Sun, 25 Jul 1999 21:55:25 -0600
Matt Garman wrote:
>
> This is my first post to this NG. I wrote this about a former
> roommate. Go nuts with critiques, comments, concerns, flames, etc.
[clip]
Well, you asked for it:
too long, and no "on-topic" content.
--
-John ([EMAIL PROTECTED])
------------------------------
From: "Global Refund Portugal" <[EMAIL PROTECTED]>
Subject: Re: Need support for Asus V3800 Riva TNT2.
Date: Mon, 26 Jul 1999 12:12:38 +0100
www.riva.com has the XFree86 drivers for all riva chipset cards...
Ian Willmore <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Hi all
>
> I've got the above mentioned display card and want to eventually run
Redhat
> 6
> on my new machine but am concerned about support for my display card.
>
> Will appreciate any help.
>
> Thanks
> Ian Willmore
> Cape Town
>
>
>
>
>
------------------------------
From: Helge Hafting <[EMAIL PROTECTED]>
Crossposted-To:
comp.os.linux.misc,alt.os.linux,alt.os.linux.slackware,comp.os.linux.setup
Subject: Re: spin down HDD
Date: Mon, 26 Jul 1999 13:58:08 +0200
Lindoze 2000 wrote:
>
> I use hdparm to spin down my HDD after 1/2 hr or so.
> the problem is, it spins up again after 1/2 hr. then it spins back down.
> has anyone had that problem?
> why wont it stay asleep?
> the system seems idle. no hdd activity detected.
This could be syslog. It will write to /ver/log/syslog
now and then, even if nothing happens. (So you may know
the last time the machine was alive - if it ever dies.)
Check if /var/log/syslog contains "-- MARK --" now and then.
This is configurable, you can turn off the MARKing by
adding the parameter "-m 0" at the place syslogd starts.
This is probably in /etc/init.d somewhere.
Helge Hafting
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: comp.os.linux.misc,comp.os.linux.setup
Subject: Re: kernel panic: unable to mount root fs on 03:42
Date: 26 Jul 1999 09:20:49 GMT
Scorpio <[EMAIL PROTECTED]> wrote:
> Hi all.
> I've been runnuing 2.2.10 kernel for a while, and had to replace a
> motherboard in my machine. After that, the kernel would not boot and
> produces teh message:
>
> Partition check:
> hda: hda1
^^^^^^^^^
> hdb: hdb1 hdb2 hdb3 <> hdb4
> hdc: hdc1 hdc2 <hdc5>
> [...]
> Kernel panic: VFS: unable to mount root fs on 03:42
^^^^^
The rootfs is set incorrectly: You don't have any logic partition on hda but
your kernel tries to mount the 38th logic partition (03:42 are the
major:minor device numbers of /dev/hda42). Edit your lilo.conf (if you boot
from HDD) or run rdev on the floppy (if you use such a thing for booting)
to set the right root partition.
Peter
--
Peter Gritsch <[EMAIL PROTECTED]>
`... so I'd rather you didn't try any last-minute stuff.'
I *AM* LAST-MINUTE STUFF, said Death, standing up.
[Terry Pratchett, Hogfather]
------------------------------
From: [EMAIL PROTECTED] (Warner Bruns)
Subject: Re: building efficient fileserver on Linux - hardware questions
Date: 26 Jul 1999 11:58:57 GMT
Reply-To: [EMAIL PROTECTED] (Warner Bruns)
Since the file-server will serve its files via a network,
doesn't the speed of the disk and of the cpu way outperform
the speed of the network?
Just a thought.
Warner
------------------------------
From: "robert_c" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.development.apps,comp.os.linux.development.system
Subject: can't compile SMP driver with -D__SMP__ in 2.2.X
Date: 26 Jul 1999 12:07:06 GMT
Hello:
Could someone tell me the answer about compiling error in SMP case(and I
make two experiments for this problem).
[Experiment1:]
Question 1:
when I compile a program(it works fine in kernel 2.0.X, ie. in UP)
with -D__SMP__, gcc told me some errors like following.
[root@robert MyDriver]# gcc -D__SMP__ -O -w -c MyDriver.c
/usr/include/asm/smp.h: In function hard_smp_processor_id':
In file included from /usr/include/linux/smp.h:11,
from /usr/include/linux/sched.h:20,
from MyDriver.h:23,------------------------->ie. #include
<linux/sched.h>
/usr/include/asm/smp.h:209: APIC_BASE' undeclared (first use in this
function)
/usr/include/asm/smp.h:209: (Each undeclared identifier is reported only
once
/usr/include/asm/smp.h:209: for each function it appears in.)
/usr/include/asm/smp.h:209: APIC_ID' undeclared (first use in this function)
[root@robert MyDriver]#
============================================================================
=========================================
MyDriver.h like following.
#ifndef __KERNEL__
# define __KERNEL__
#endif
#ifndef MODULE
# define MODULE
#endif
#include <linux/modules.h>
#include <linux/sched.h>
#include <linux/kernel.h>
............
[Experiment 2:]
But, when I make a change from -D__SMP__ (in compile time) to write the
definition inside MYDriver.h like following and try again, compiling is OK,
but insmod MYDriver.o has some error.
MyDriver.h:
#include <linux/config.h>
#ifdef
CONFIG_SMP<-----------------------------------------------------------------
============================================================================
=======
#undef __SMP__ <----------------------
#define __SMP__1<-----------------------
#endif <----------------------
#ifndef __KERNEL__
# define __KERNEL__
#endif
#ifndef MODULE
# define MODULE
#endif
....
+++++++++++++++++++++++++++++++++++++++++++++++++
Error like this:
[root@robert MyDriver]# gcc -O -w -c MyDriver.c
<--------Compile OK.
[root@robert MyDriver]# insmod -f Mydriver
Warning: kernel-module version mismatch
Question 2:
./MyDriver.o was compiled for kernel version 2.2.5-15<------Why UP,
not SMP?(because I have define __SMP__ like above MyDriver.h
while this kernel is version 2.2.5-15smp
./MyDriver.o: unresolved symbol schedule_timeout
./MyDriver.o: unresolved symbol request_region
./MyDriver.o: unresolved symbol kmalloc
./MyDriver.o: unresolved symbol unregister_chrdev
./MyDriver.o: unresolved symbol register_chrdev
./MyDriver.o: unresolved symbol release_region
./MyDriver.o: unresolved symbol boot_cpu_data
./MyDriver.o: unresolved symbol kfree
./MyDriver.o: unresolved symbol __verify_write
./MyDriver.o: unresolved symbol memcpy_fromfs
./MyDriver.o: unresolved symbol check_region
./MyDriver.o: unresolved symbol jiffies
./MyDriver.o: unresolved symbol printk
./MyDriver.o: unresolved symbol add_timer
Question 3:
And Why has many unresolved symbols like above? (I can't make sure it is
relavant with -DMODVERSIONS or not, because I didn't use -DMODVERSIONS
option in compile time). I still make third experiment,ie. copy MYDriver.o
to /lib/modules/2.2.5-15smp/misc directory, and issues comand like:
insmod -f MyDriver.o, but still fail.
So, Could you give any advives.
------------------------------
From: Abdullah Ramazanoglu <[EMAIL PROTECTED]>
Subject: Re: ECC-SDRAM: Supported?
Date: Mon, 26 Jul 1999 15:08:31 +0300
Thomas Stieler wrote:
> Know my question: how is the behavior of the linux-kernel?
> I think, that an two bit error should trigger an NMI, and the system
> should be halted immediatly.
> Thats good, or better than let the system working until the random
> errors crash the system...
>
> But of course it would be the best, if the system will tell you about
> one-bit-errors, so you can see, that the memory is unstable!
>
ECC logic is handled completely by hardware. So there should be a BIOS
function to collect ECC status from hardware. Then OS (or user) can
collect it from BIOS.
Another question : Is there a standard way defined in PC architecture,
in terms of ECC status communication between BIOS and OS ? If not (which
is what I think), then a more likely solution would be BIOS recording
ECC events to somewhere (CMOS or dedicated disk area), so that user or
OS can retrieve it later. Since there is no standard way of doing this,
mo/bo manufacturer should write an OS patch, so that the OS would be
able to communicate ECC events with BIOS.
--
Abdullah Ramazanoglu ( aramazanoglu AT demirbank DOT com DOT tr )
------------------------------
** 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
******************************
