Linux-Hardware Digest #459, Volume #9 Thu, 18 Feb 99 23:13:30 EST
Contents:
HP 8140 Linux ??? ("Timothy J. Suhr")
Re: Same Disk RAID and Mirroring (Peter da Silva)
Using 1.4Mb floppies with an Imation LS-120. (Carl Beaudry)
Re: ATI 3D Rage IIc - Getting on my tits! (Daren Scot Wilson)
Re: Matrox Millenium II AGP ("David A. Frantz")
Re: Wheel mouse ("DigitalLiquid")
Re: Mastrox Millenium G200 AGP and XFree ("David A. Frantz")
Re: Same Disk RAID and Mirroring (Andy Glew)
RE: Diamond Viper 550 Riva TNT (Anantha)
Re: 3c509 + 3c59x = Barfff ("Oliver Cronk")
Re: 3151 Terminal. (Derek Kwan)
Re: Windows & Linux File Transfer ? (Colin Day)
Re: ATI Mach 64 clockchip settings (Colin Day)
floppy driver instability in 2.1.x/2.2.x?
Re: Lexmark 1100 Color Jetprinter & Redhat 5.1 ([EMAIL PROTECTED])
Re: C Programming for ISA Card ([EMAIL PROTECTED])
External Modems (Linux Newbie)
Re: ADSL and Linux (Miguel Cruz)
Re: PCI modems in linux? ("Eugene")
----------------------------------------------------------------------------
From: "Timothy J. Suhr" <[EMAIL PROTECTED]>
Subject: HP 8140 Linux ???
Date: Thu, 18 Feb 1999 16:54:21 -0800
Greetings.
I have an HP Pavilion 8140 computer and I would appreciate if somebody would
look through my list of parts and check if any won't work in Linux. Okay,
here's the specs:
Intel P 200 MHz CPU
Asus Mainboard TX97-XV
96 MB EDO RAM (may upgrade to 128)
Intel 82371AB/EB PCI Bus Master IDE Controller
Standard Floppy Disk Controller
Seagate 4.0 GB harddrive
ATI 264-VT3 (DirectX)
Analog Devices AD1816 Sound Controller
GoldStar CRD-8240B CD-ROM
IOMEGA ZIP 100
GENERIC NEC FLOPPY DISK
Intel 82371AB/EB PCI to USB Universal Host Controller
HP M70 Multi-Media Montor
Hewlett-Packard DeskJet 1000C
Thanks,
T.J. Suhr
------------------------------
From: [EMAIL PROTECTED] (Peter da Silva)
Crossposted-To: comp.arch,comp.arch.storage,alt.os.linux,comp.periphs
Subject: Re: Same Disk RAID and Mirroring
Date: 17 Feb 1999 22:35:14 GMT
In article <[EMAIL PROTECTED]>, Andy Glew <[EMAIL PROTECTED]> wrote:
>A recent issue of IEEE Computer (I think) has a paper that proposes
>allocating a separate disk, or a portion of the local disk, as a disk
>cache - absorbing writes when DRAM based disk caches overflow
>during a write burst, and then emptying this "disk based write buffer"
>to the ordinary disk blocks once the write burst is over.
Sort of like a journalled file system?
And then the next step is to not bother moving the blocks, and you're looking
at logfs/Advfs/WAFL.
No?
--
In hoc signo hack, Peter da Silva <[EMAIL PROTECTED]>
`-_-' "Heb jij vandaag je wolf al geaaid?"
'U`
"Tell init(8) to lock-n-load, we're goin' zombie slaying!"
------------------------------
From: Carl Beaudry <[EMAIL PROTECTED]>
Subject: Using 1.4Mb floppies with an Imation LS-120.
Date: Wed, 17 Feb 1999 23:42:17 -0500
Reply-To: [EMAIL PROTECTED]
Since 120Mb Imation LS-120 disks work as removable IDE drives under
Linux and can be used to boot Linux from either a 120Mb or 1.4Mb
floppiee, the question becomes: why have a regular floppy disk drive at
all?
The only remaining use for a 1.4M floppy drive is reading and writing
1.4Mb floppies which the LS-120 is capable of doing at the hardware
level.
Has anyone figured out how to create the special files in the /dev/
directory needed to treat an IDE Imation LS-120 disk installed at, say,
/dev/hdc as *either* a regular 1.4Mb floppy drive *or* a 120Mb drive
(depending on the media of course)?
It's got to be possible, but I don't know enough about the block
structure of the drives to do it and didn't want to dig around the major
& minor numbering scheme if there's any easy answer floating around out
there.
Has anyone done this yet?
--Carl
PS.Please copy responses via e-mail as my news connection is flakey
------------------------------
From: Daren Scot Wilson <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: ATI 3D Rage IIc - Getting on my tits!
Date: Thu, 18 Feb 1999 20:24:12 +0000
I run a Compaq Presario 5140, ATI Rage etc. I don't remember exactly what
version, all those funny letters and numbers. But it's working great for me.
My X config file doesn't even use chip ID, as the one respondent said you'd
have to use.
I had much trouble with the Rage, but made discoveries by searching the
internet. Results are on my web page,
http://www.newcolor.com/darenw/computer/presario.html.
But even better, the latest XFree86, version 3.3.3.1, uses the Rage with no
special fussing.
Still, it was hard to get the right combo of modes in the config file to get
1024x768 in 24 bit color, but eventually managed to do this too.
My X Config file is appended to this. It's short 'cause i knocked out all
comments and modes i don't really need. It only runs 24 bit color, though,
since we artists have no use for less ;-)
--
Daren Scot Wilson
[EMAIL PROTECTED]
www.newcolor.com
----
"A ship in a harbor is safe, but that is not what ships are built for"
-- William Shedd
Section "Files"
RgbPath "/usr/X11R6/lib/X11/rgb"
FontPath "/usr/X11/lib/X11/fonts/75dpi/"
FontPath "/usr/X11/lib/X11/fonts/100dpi/"
FontPath "/usr/X11/lib/X11/fonts/misc/"
FontPath "/usr/X11/lib/X11/fonts/Type1/"
FontPath "/usr/X11/lib/X11/fonts/Speedo/"
FontPath "/usr/X11/lib/X11/fonts/freefont/"
FontPath "/usr/X11/lib/X11/fonts/sharefont/"
EndSection
Section "ServerFlags"
# Disable Ctrl-Alt-Backspace; use
# DontZap
EndSection
Section "Keyboard"
Protocol "Standard"
AutoRepeat 300 5
# To set the LeftAlt to Meta, RightAlt key to ModeShift,
# RightCtl key to Compose, and ScrollLock key to ModeLock:
# LeftAlt Meta
# RightAlt ModeShift
# RightCtl Compose
# ScrollLock ModeLock
LeftAlt Meta
RightAlt Meta
ScrollLock Compose
RightCtl Control
XkbRules "xfree86"
XkbModel "microsoft"
XkbLayout "us"
EndSection
Section "Pointer"
Protocol "PS/2"
Device "/dev/mouse"
Emulate3Buttons
# Emulate3Timeout 50
EndSection
Section "Monitor"
Identifier "Nokia-plain"
VendorName "Nokia"
ModelName "447W"
HorizSync 30-64
VertRefresh 48-100
Modeline "640" 26.1 640 680 776 806 480 535 537 597
Modeline "800" 36 800 832 892 980 600 620 638 653
Modeline "1024" 60 1024 1072 1232 1288 768 788 796 833
EndSection
Section "Device"
Identifier "Presario"
VendorName "ATI"
Chipset "ATI"
VideoRam 4096
EndSection
Section "Screen"
Driver "accel"
Device "Presario"
Monitor "Nokia-plain"
Subsection "Display"
Depth 32
Modes "1024" "800" "640"
ViewPort 0 0
Virtual 1024 768
EndSubsection
Subsection "Display"
Depth 8
Modes "1024" "800 640"
ViewPort 0 0
Virtual 1024 768
EndSubsection
EndSection
------------------------------
From: "David A. Frantz" <[EMAIL PROTECTED]>
Subject: Re: Matrox Millenium II AGP
Date: Thu, 18 Feb 1999 20:54:49 -0500
Once you have 3.3.3.1 downloaded you will be well on your way to superior X
performance. cofiguration is the same as any other Xfree server, pretty
much works great
dave
John E. Hagensieker wrote in message <7aghhf$[EMAIL PROTECTED]>...
>I have the Matrox Millenium II AGP 8 meg card on an Asus P5-A-B.
>Running Red Hat 5.2.
>
>My card is listed but I cannot run X windows at any resolution or monitor
>type.
>
>Downloading the Xfree86 3.3.3.1-1 as we speak.
>
>Has anybody else been unable or better yet able to configure this card with
>Linux?
>
>
>
------------------------------
From: "DigitalLiquid" <[EMAIL PROTECTED]>
Subject: Re: Wheel mouse
Date: Fri, 19 Feb 1999 02:39:32 +0100
G�nther schrieb in Nachricht <[EMAIL PROTECTED]>...
>Brian Rutledge wrote:
>But I can scroll with the wheel yet. The three buttons work fine.
take a look at this page:
http://www.inria.fr/koala/colas/mouse-wheel-scroll/
...works fine!!!
------------------------------
From: "David A. Frantz" <[EMAIL PROTECTED]>
Subject: Re: Mastrox Millenium G200 AGP and XFree
Date: Thu, 18 Feb 1999 20:57:06 -0500
Christian;
The AGP version works great, infact its fantastic! Jus make sure you have
the latest version of Xfree availiable, that would be 3.3.3.1.
dave
Christian Tagtachian wrote in message <7ah21p$d9e$[EMAIL PROTECTED]>...
>I would like to know how well this card is supported...
>I read that XFree supports it, but it doesn't state if it is the AGP or PCI
>version... I am about to get
>an AGP one, and wanted to be sure it works ok before I do.
>By the way, it is an 8 megs one.
>Thanks
>Christian
>
>
>
>
------------------------------
From: Andy Glew <[EMAIL PROTECTED]>
Crossposted-To: comp.arch,comp.arch.storage,alt.os.linux,comp.periphs
Subject: Re: Same Disk RAID and Mirroring
Date: Thu, 18 Feb 1999 20:07:56 -0600
> You are looking at the disk from an abstract, system viewpoint, where
> sectors discrete entities, separated by gaps. Within each sector, each bit
> is stored as a discrete element, just like the old 1200 baud (and slower)
> modems transmitted each bit as a separate tone.
>
> In reality, these days a disk is a rather different animal.
Your presumption of my ignorance is excessive.
I.e. I know all this.
> So, let's consider a radial scratch...
>
> HOWEVER... there is absolutely no reason to presume that our radial scratch
> will be kind enough to occur within the data area of the sector. Suppose it
> occurs towards the end of the ECC segment. Even though the user data is
> fine, by corrupting the ECC the drive *has* to determine that the sector is
> unreadable... IN ADDITION, the scratch may have resulted in the drive being
> unable to read the preamble for the next sector, which means that the
> adjacent sector is also unreadable.
>
> And suddenly you have a problem that your RAID solution cannot fix.
First, as someone else has commented, the ECC codes I am familiar with
are not special - an error anywhere, in the data or code, is detectable.
The ECC part is not special.
Second, if the error results in the next preamble being jiggered,
then we have an error than spans 2 disk blocks. Such an error is correctable
by assigning physically adjacent disk blocks to different parity groups
- in my previous posts I called this "interleaving" the parity groups,
which is similar to, but not necessarily exactly the same as,
disk block interleaving for performance.
(If I were in a tit-for-tat mood, I would now go on and explain disk interleaving
in detail, since that is the sort of condescension that your explanation of flux
reversal codings comes across as. I assume, however, that you know this
already, and I will try not to be so condescending.)
If you have errors that span 3 disk blocks, then 3 way interleaving of the parity
groups
can correct it. 4 disk blocks, 4 way interleaving, and so on. Potentially,
errors that span up to 1/5 of the blocks on a track can be corrected in this manner.
However, obviously - or at least I thought obviously - parity group interleaving would
be bad for write performance. Basically, single and small block writes
are more expensive than multiple block (over)writes, and parity group interleaving
would move further out the point at which full multiple block (over)writes
of an entire parity group would be done. If your entire workload is large
writes, e.g. if it is a purely a log structured filesystem, with a bit of NVRAM
you would not care, but if there is a random access short block burst write
component you would care.
So, again, my question is this:
What fraction of disk surface errors span 1 block,
2 blocks, 3 blocks ... an entire track?
- i.e. what degree of parity group interleaving
would be required for an acceptable level
of protection?
Compare this to
What fraction of errors are correctable by block ECC?
And similarly
What fraction of errors span two or more adjacent tracks?
(although I note that parity groups within the same track
can correct such errors, as long as they are confined
to less than 1/5th of the full circle.
What other patterns of error locality are there?
E.g. does it ever happen that an entire cylinder goes,
but not other cylinders?
===
I will admit that this conversation has annoyed me,
but it has had the effect of making me more and more
confident that simple block parity groups, in the same track
(or just sequential, which is mainly in the same track)
would be effective on regular disks
(a) if the workload is read-only
(b) given the techniques for handling the write overheads
we have discussed.
The write overhead issues are real, but probably doable
with a challenge.
The read overhead issues are minor.
I.e. I think that it is doable. And no argument to the contrary,
that I find credible, has been presented.
===
I frequently find that the people who argue most vociferously against
a new idea are mapping it into something else that they are thinking
of. Often a better idea.
Hence I ask, Malcolm: do you have any interesting ideas for increasing
single disk reliability?
====
Finally, the term "Same disk RAID" is a poor one. Better to call it
"same disk block parity groups".
------------------------------
From: [EMAIL PROTECTED] (Anantha)
Subject: RE: Diamond Viper 550 Riva TNT
Date: 18 Feb 1999 05:19:51 GMT
Try getting the updates for XFree86 which has support for the latest cards
Updates can be got from
ftp://ftp.cc.gatech.edu/pub/Linux/distributions/redhat/updates/5.2/i386/
http://www.redhat.com/support/docs/rhl/rh52-errata-general.html
Please check the following page also about the server info:
http://www.redhat.com/support/docs/rhl/XFree86-upgrade.html
------------------------------
From: "Oliver Cronk" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking,comp.os.linux.setup,comp.os.linux.install
Subject: Re: 3c509 + 3c59x = Barfff
Date: Fri, 19 Feb 1999 01:32:53 -0000
I think you mean RJ-45 not RJ-41!
J. Ken Zur wrote in message ...
>Question....
>How does one tell the driver to use the 3c509 AIU, or the RJ41 interface?
>The driver picked my BNC and I am using the RJ41. Help!!!!!! I'd really
love
>to get this interface up and running.
>
>
>Thanks in advance
>
>
>aasland wrote in message <[EMAIL PROTECTED]>...
>>Good news! I got a response from Donald Becker, here's what he wrote:
>>
>>> I compiled them into the kernel, so that counts as built-in, right?
>>
>>Yup.
>>
>>> However, they do show up in kerneld.
>>
>>Hmmm, what do you mean by that?
>>
>>Anyway, you should put the following line in /etc/lilo.conf:
>> append = "ether=0,0,eth1 ether=0,0,eth2 ether=0,0,eth3"
>>
>>Note: some old versions of LILO were picky about spaces, so put in two (as
>>above).
>>
>>Alternately, you can use the 3c509 driver as a module.
>>
>>
>
>
>
------------------------------
From: Derek Kwan <[EMAIL PROTECTED]>
Crossposted-To: comp.unix.aix
Subject: Re: 3151 Terminal.
Date: Thu, 18 Feb 1999 23:49:53 GMT
Jason Williams wrote:
> Your question is a little weird but I think the answer to your question is
> Yes you can use other types of terminals with an RS6k.
>
> As for a regular dumb terminal???? What type is regular.
>
> I can't recall if a 3151 has two input ports but it does have an output :)
OOPS, I mean 3153, not 3151. Just do a search on 3153 at www.ibm.com.....
>
>
> JW
>
> Derek Kwan <[EMAIL PROTECTED]> wrote in article
> <[EMAIL PROTECTED]>...
> >
> > Hello World,
> >
> > I have just order a IBM 3151 terminal, and I was wondering if it is
> > possible to use as a regular dumb terminal. For example hook up with my
> > Linux box (BTW, isn't 3151 have 2 comm ports?) so I can switch between
> > terminals?
> >
> > Derek
> >
> >
------------------------------
From: Colin Day <[EMAIL PROTECTED]>
Crossposted-To: comp.linux,comp.os.linux.questions,linux.redhat
Subject: Re: Windows & Linux File Transfer ?
Date: Fri, 19 Feb 1999 02:25:13 +0000
"W. Christopher Everhart" wrote:
> Tom Emerson wrote:
>
> [snip]
>
> > these partitions are on the same drive, just different formats. Linux
> > supports MANY file systems other than ext2fs -- FAT, FAT32, VFAT (?), and
>
> VFAT is just MSDOS (FAT) with support for long filenames. It's great for
> sharing with a Win95 system that isn't using FAT32.
>
> [snip]
VFAT does support FAT32 partitions. Ever since 2.0.34.
Colin Day [EMAIL PROTECTED]
------------------------------
From: Colin Day <[EMAIL PROTECTED]>
Subject: Re: ATI Mach 64 clockchip settings
Date: Thu, 18 Feb 1999 05:51:58 +0000
James wrote:
> I'm trying to use use RH 5.1 with my Ati Mach 64. During install, when I
> select Autoprobe for card detection, I gett a message stating that Linux
> can't detect my video Ram.
> I get a list of chipsets to select from, but I don't know which one to
> use.
> When i try to run X, I get a message "cannot open display". Any help
> with this would be greatly appreciated.
> Thanx in advance
Is there an Mach64 VT option? Also, you might not want to do the clockchip
settings.
Colin Day [EMAIL PROTECTED]
------------------------------
From: [EMAIL PROTECTED] ()
Subject: floppy driver instability in 2.1.x/2.2.x?
Date: 19 Feb 1999 03:06:47 GMT
hello,
i think i've noticed a bit of instability in the floppy driver used in the
2.1.x and 2.2.x kernels. i unfortunately have no core dumps or kernel panic
information to submit, but just my hunch that after using linux for 3-4
years without a single crash, something is very wrong.
the first time i noticed this was on a new machine at work which i had just
upgraded to a 2.1.x kernel. i was writing something to a floppy to
sneakernet (with tar) and the whole machine locked up *hard*. i attributed
this to a bug in the kernel or shaky hardware, and moved on (although a bit
perturbed, since previously i could brag to NT users that the only time i've
seen a linux machine lock up is with a buggy X server.)
however, it has happened 2-3 times total on that machine, still running
2.1.131.
just now, this floppy-induced lockup happened on my *home* machine, which
has been absolutely rock solid for the past 3 years running 2.0.x kernels.
i upgraded to the 2.2.x kernel back when it was released and haven't touched
the floppy drive until now, when i was doing some stuff with cpio. my
machine locked up while playing an mp3 and cpio'ing.
any ideas!? has anyone else seen this kind of behaviour?! this is definitely
not typical linux stability...
--
::: John Randolph <jrandATfastDOTnet> ::::: linux 2.0.xx
::: PGP key: http://www.users.fast.net/~jrand/pgpkey.txt
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To: linux.redhat.install,comp.os.linux.setup
Subject: Re: Lexmark 1100 Color Jetprinter & Redhat 5.1
Date: Fri, 19 Feb 1999 02:34:39 GMT
On Wed, 17 Feb 1999 12:46:30 GMT, [EMAIL PROTECTED] wrote:
>Hi All:
>
> I've installed Redhat 5.1 on my PC. I recently purchased the Lexmark 1100
>Color Jetprinter. Judging from several postings in this and other newsgroups,
>Lexmark does not provide drivers for Linux (Fat chance!!), and Linux (RedHat,
>Debian, Suse, etc.??) does not support Lexmark printers. However, the
>printer's manual does say that it is compatible with HP Deskjet 500C drivers.
>This is listed in Redhat's printer compatibility list. So, I installed the HP
>Deskjet 500C filter. But, I am still unable to print anything. Either a blank
>page gets printed or the printer head(??) prints everything on the same spot
>(leaving a big messy blot on the paper). Is there anybody out there with more
>experience with this type of problem? Should I set something in the CMOS
>setup (Bidirectional printing??), or maybe disable PostScript printing, or
>install it as a plain, old text-only printer?
>
>Any hints would be appreciated.
>
>Thanx in advance,
>Swaroop ([EMAIL PROTECTED])
>
>-----------== Posted via Deja News, The Discussion Network ==----------
>http://www.dejanews.com/ Search, Read, Discuss, or Start Your Own
the lexmark compatibility with the 500c is only at the lower res. try
300x300. worked for the 1020.
------------------------------
From: [EMAIL PROTECTED]
Crossposted-To:
alt.os.linux,comp.os.linux.development.system,comp.os.linux.misc,linux.dev.c-programming,linux.dev.serial
Subject: Re: C Programming for ISA Card
Date: 17 Feb 1999 16:20:41 GMT
In comp.os.linux.development.system Rick Wheeler <[EMAIL PROTECTED]> wrote:
> I have an ISA Industrial Digital I/O card, caple of driving up to 64
> Can anyone provide programming examples, documentation or other advise that
> may be of benefit to me?
Hello,
take a look at my program, IO256Control. Its a network server
that i've programmed, fully configurable...
You can find it on any sunsite.unc.edu/pub/Linux/apps/circuits
I programmed it about an year ago tough...It seems that the server
to be compiled with the new glibc must have -lcrypt, but you can get a patch
by Ken for this on my homepage http://student.dee.uc.pt/~slug/
Bye,
Nuno Sucena
------------------------------
From: [EMAIL PROTECTED] (Linux Newbie)
Subject: External Modems
Date: Thu, 18 Feb 1999 06:05:17 GMT
Is it safe to say that most external modems will work with
Linux? I'm getting a used Cardinal 28.8, will that work well in
linux?
--
"Linux: The best things in life are free"
------------------------------
From: [EMAIL PROTECTED] (Miguel Cruz)
Subject: Re: ADSL and Linux
Date: 18 Feb 1999 06:02:02 GMT
Nat <[EMAIL PROTECTED]> wrote:
> Has anyone ever installed ADSL modem on linux box? ... I get a good deal
> with ADSL service with a frem ADSL Modem. I have no idea how that modem is
> going to work with my box.
I have Bell Atlantic ADSL service with a Westell external modem. It took all
of 5 minutes to set up, ethernet card and all. God bless that clever kernel.
It recognized the card automatically, all I had to do was stick a few
'ifconfig' and 'route' commands in the inet rc file.
miguel
------------------------------
From: "Eugene" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.networking,comp.os.linux.setup
Subject: Re: PCI modems in linux?
Date: Fri, 19 Feb 1999 03:43:30 GMT
AFAIK all PCI modems are winmodems
Doug wrote in message <[EMAIL PROTECTED]>...
>Can a PCI modem be used in linux? If so how? Something called a modem
>enumerator is installed in windows along with the modem itself and i
>dont know what that is. Its creative modmeblaster DI5630 v.90. Its
>being used as PnP right now but there are jumpers on it I dont have the
>manual so im trying to find out if com and irq can be hard set..
>Thanks for any help and please email me a response if possible at
>ratchet at tir dot com
>Doug
>
------------------------------
** 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
******************************
