Linux-Hardware Digest #400, Volume #13 Fri, 11 Aug 00 05:13:13 EDT
Contents:
ATA-100 on ATA-66 moboard -- hdparms... (Alessandro Giachino)
Re: [Fwd: Ultra 66 Support] (Alessandro Giachino)
Re: Help!!! IDE Tape drive setup. (David Vaught-Alexander)
Re: How can i install redhat5.2 with scsi ultra160 hard disk(aic7892)?
([EMAIL PROTECTED])
Re: Advice for celeron vs pentium under Linux ("John Mazza")
Re: Advice for celeron vs pentium under Linux ("John Mazza")
Re: Touchpad & pc_keyb.c (Christopher Michael Collins ())
Lexmark Optra 40? (John Hong)
Re: 20gb (Kenneth R�rvik)
Re: getting dsl, what h/w should i get? ([EMAIL PROTECTED])
Re: 20.4GB HD, bios limitation, EZ-drive or any other? (Stuart Fotheringham)
YAMAHA sound card (Stuart Fotheringham)
Re: ISDN 128K modem (sideband)
2 bttv cards... /dev/video1 does not give inputsignal (Arne Van Theemsche)
Re: 20gb (Stuart Fotheringham)
Re: ATA-100 on ATA-66 moboard -- hdparms... (Bartek Kostrzewa)
----------------------------------------------------------------------------
From: Alessandro Giachino <[EMAIL PROTECTED]>
Subject: ATA-100 on ATA-66 moboard -- hdparms...
Date: Fri, 11 Aug 2000 04:36:56 GMT
Hello,
my ATA100 drive is sloooow I think because moboard IDE
controller is ATA66.
Can anybody suggest how to use hdparms in a stable way
to tweak the drive to run in ATA66 mode?
'hdparms -d1 -c1 /dev/hda' seems to give better transfer
rates, when checking with 'hdparms -Tt' options.
(I think then I would have to use -k option to keep settings
and put line in rc.local....)
Anything else beyond this which may help?
Thank you,
AG
--
Alessandro Giachino -- Software and Sales Engineer
EKK Inc. -- http://www.ekkinc.com
2065 West Maple Ste. C309, Walled Lake MI 48390
(+1)248.624.9957 tel / (+1)248.624.7158 fax
------------------------------
From: Alessandro Giachino <[EMAIL PROTECTED]>
Subject: Re: [Fwd: Ultra 66 Support]
Date: Fri, 11 Aug 2000 04:43:26 GMT
VIA is what I have on my motherboard too.
That is why I was concerned...
AG
James Richard Tyrer wrote:
>
> EKK wrote:
>
> >
> >
> > ------------------------------------------------------------------------
> >
> > Subject: Re: Ultra 66 Support
> > Date: Mon, 07 Aug 2000 01:08:17 -0700
> > From: EKK <[EMAIL PROTECTED]>
> > Organization: EKK Inc
> > Newsgroups: comp.os.linux.hardware
> > References: <N6_h5.2822$[EMAIL PROTECTED]>
><8ma89k$b4o$[EMAIL PROTECTED]> <BOpi5.4000$[EMAIL PROTECTED]>
><[EMAIL PROTECTED]> <hW%i5.4670$[EMAIL PROTECTED]>
><[EMAIL PROTECTED]>
> >
> > This is a README on the U66 patch directory of kernel FTP site:
> >
> > ftp://ftp.kernel.org/pub/linux/kernel/people/hedrick/
> >
> > WARNING!!!!!!!!!!
> >
> > VIA code require you pass idebus=XX
> >
> > Where XX must == 25, 33, 37, or 41
> >
> > This reflects IDE-PCI BUS CLOCK!!
> >
> > You must know your SYSTEM........
> >
> > Failure to do so can have damage that is not recoverable.
> >
> > 0519:
> >
> > This should fix the VIA problems.
> >
> > Does anyone know what that means? I.e. how do I find out what
> > my idebus is?
> >
> > AND
> >
> > Does this patch only affect IDE PCI cards, or also onboard IDE
> > controllers?
> >
> > Thanks,
> >
> > AG
> > --
> >
> > Alessandro Giachino, Software Engineer
> >
> > EKK Inc.
> > 2065 West Maple C309 tel. 248-624-9957
> > Walled Lake MI 48390 fax. 248-624-7158
> > _____________________________________________
> > http://www.ekkinc.com
>
> VIA is a brand of chip. If you don't have that brand, don't worry. Promise uses
>their own chips, but I have VIA on my Mother Board.
>
> Also, there is a new patch posted:
>
> ide.2.2.16.all.20000805.patch.bz2
>
> This one is compressed.
>
> JRT
--
Alessandro Giachino -- Software and Sales Engineer
EKK Inc. -- http://www.ekkinc.com
2065 West Maple Ste. C309, Walled Lake MI 48390
(+1)248.624.9957 tel / (+1)248.624.7158 fax
------------------------------
From: David Vaught-Alexander <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.install,linux.redhat.misc
Subject: Re: Help!!! IDE Tape drive setup.
Date: Tue, 08 Aug 2000 19:35:42 -0700
Hi!
Have you tried just writing to the tape? Here's what I do (there is no mount
procedure or configuration file entry).
# tar cpv filename > /dev/ht0
Anyway, when you say mount/use the drive, what do you mean? I've only tried
archiving. Did you try the above tar command for your drive?
p.s. My IDE tape unit stop working the other day. Maybe it went bad. But when I
check to be sure it is a real block device it isn't ......... but used to be!.
(from BRU manual)
# ls -l /dev/tape
# brw-rw-rw 6 bin bin 2, 52 May 06 /dev/tape
I don't get this anymore...which makes sense. 99% sure my drive failed ;-(
> I have a HP Colorado 20GB IDE tape backup unit, and I am running RedHat 6.2,
> when I boot it sees the drive as HDD. But I cannot mount/use the drive.
> Please help with detailed instructions or a link. I am not toooo new to
> Linux, but have never used a tapedrive with it.
>
> --Shawn
------------------------------
Subject: Re: How can i install redhat5.2 with scsi ultra160 hard disk(aic7892)?
From: [EMAIL PROTECTED]
Date: Fri, 11 Aug 2000 05:56:50 GMT
[EMAIL PROTECTED] writes:
> When i install redhat5.2,kernel couldn't find my hard disk?
> redhat 5.2 didn't support scsi ultra160 aic7892,only support
> aic788x.How can i install redhat5.2?
Modern Linux kernels should support all AIC-78xx chipsets. Use a more
current RedHat.
--
Eric McCoy ([EMAIL PROTECTED])
"misfortune, n. The kind of fortune that never misses."
- Ambrose Bierce, The Devil's Dictionary
------------------------------
From: "John Mazza" <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: Re: Advice for celeron vs pentium under Linux
Date: Fri, 11 Aug 2000 06:16:30 GMT
Actually, they did advise them that the machine should be a single processor
unit, but they insisted on the dual processor Xeon, which BTW we ate the
cost of when it didn't work out for them (we now have in the office a nearly
$20k machine that can bang out a SETI block in under 4 hours). Sometimes it
isn't vendor competence that is the issue, but client familiarity with their
own requirements that is. Not many companies will work with someone the way
they did, even taking a $20,000 charge without a restocking fee on a totally
custom machine when the client's specified configuration didn't work out for
them.
Oh - BTW - the company involved is very familiar with multiprocessor setups,
and is currently the only Intel-authorized OEM for 8 and 16 processor
configurations (Compaq applied, but didn't like Intel's terms, so they
backed out of the deal).
I agree that the application is key to whether or not an SMP solution is
appropriate in any particular situation, however often the clients do not
know their apps well enough to give the appropriate information to make the
determination.
Now, as to the raw performance numbers, the K7 machine still outpaced
anything the client had ever seen. The dual Xeon does have about a 15% edge
IF the apps can take advantage of SMP, otherwise it is at a severe
disadvantage due to its lower clock speed on the individual processors.
For all PRACTICAL purposes, I have seen that the Intel and AMD processors
are very close in performance, with the AMD seeming to have a bit of an edge
price-wise. I would be very interested in trying a K7-900 with a big block
of RDRAM or DDR memory as well to see just how much of a boost the 200 MHz
front-end bus gives the unit. Right now, I am inclined to believe that the
K7s are being somewhat restrained by the PC-100 and PC-133 memory being used
not allowing them to take full advantage of their archetecture and faster
FSB, but I will wait until I can actually test this theory before making a
final statement on the matter.
I trust this provides some clarification of my position.
smp root <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> what a shame that the company that sold the hardware didn't
> advise the client whether or not an smp based board would be
> suitable. if the athlon did beat it by being more than twice as
> fast i'd say shame on the company that sold them the kit. and i
> wouldn't be to proud about having worked for them. still it's
> better than working for them, i suppose:-) it's always worth
> checking with a *competent* company before investing in hardware,
> and is the speed gain was so dramatic on the athlon, that's bad
> advice not poor hardware. check before going down the smp road
> that it will work better for you than a monoprocessor setup.
> intel or amd. still waiting for amd smp...
>
>
> -----------------------------------------------------------
>
> Got questions? Get answers over the phone at Keen.com.
> Up to 100 minutes free!
> http://www.keen.com
>
------------------------------
From: "John Mazza" <[EMAIL PROTECTED]>
Crossposted-To: linux.redhat.misc
Subject: Re: Advice for celeron vs pentium under Linux
Date: Fri, 11 Aug 2000 06:20:23 GMT
I agree. In the "real world" many clients run older commercial software
that is poorly written, but they still need to obtain usable performance
from. Retraining their staff or getting the software vendor to rewrite the
code may both be impractical, so therefore sometimes we have to take what we
are given and try to help as best we can.
In any event, I am still very impressed by the K7 machine. It is a real
screamer.
Jes�s M. NAVARRO <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]...
> Hi John:
>
> John Mazza wrote:
> >
> > I've been running Linux on an AMD K6-2 350 clocked to 366 MHz with no
> > problems whatsoever. AMD is making some absolutely killer processors
> > nowadays, and beating the pants off Intel price-wise at the same time.
I
> > have also worked with the Athlon processors, and they are great too.
> > Haven't had a chance to check out the "Duron" yet, but will as soon as
> > possible.
> >
> > A company I do some work with just delivered an Athlon 900 to a client
> > equipped with 1 GB of PC 133 SDRAM, Ultra/160 SCSI drives, and other
> > goodies. This client originally specified a $18,000 dual Xeon 550 (2 MB
> > Cache version) with 1 GB of 800 MHz RDRAM for this machine, and it was
> > disappoiningly slow for their application (we are talking SEVERE
engineering
> > number crunching here). The Athlon blew away everything they had by a
> > factor of at least 2, was the fastest they had ever seen, and put them
> > seriously into the AMD camp for high-end workstation applications.
BTW -
> > the benchmarks blew away an Alpha 600 MHz machine by a VERY comfortable
> > margin! (Guess there really is something to be said for a 200 MHz
front-end
> > bus after all.)
> >
>
> Well, yours seems to me a hardly believable afirmation...
> First, you stay that the machine will be a numbercruncher, then probably
> using a single app most of (computing) time, so I believe any faster
> processor will do better than a dual machine at the same equivalent
> speed (roughly half and a bit more per processor)... but this is not
> because the AMD vs Intel stuff but because most probably your app is not
> able to take advantage of the dual architecture (it might probe that
> your client's app behalves from bad -using one CPU up to 100% and
> letting the other unused, to horribly -loosing most of the time jumping
> processes between both processors, on these circumstances).
> Second, you stay that the Athlon blew away *everything* by a factor of
> 2, which is the very first notice I had this way (including AMD
> assertions, which one can expect to be a bit biassed): even AMD stays
> that its FPU is still not as good as that from the P6 (although by only
> a sligth margin).
> Third, probably my English is not as good as it could be, but I don't
> understand your last assertion about Alphas. Does it mean that the
> Alpha was blown away by the benchmarks (that is, it did it poorly) or
> that Alpha blew away any rival regarding its benchmarks? I do believe
> the second it's the true... again, not because its faster (and wider)
> frontal bus, but because it has better MIPS rank (well, that's the
> advantage of RISC, isn't it?
>
> Anyway, if the crunchingnumber app is well programmed (not miracolously
> well programmed, just well) it will use extend use of CPU preload, which
> roughly means that when moving big bunches of numbers, the bigger cache
> of the Xeon CPU should beat by far the Athlon when running at the same
> speed... So, I would suggest (being the crunchnumbering really the
> bottelneck) that a XXXMHz PIII Xeon should be from fairly to much faster
> than an Athlon at the same speed on these circumstances. My opinion is
> that, in these circumstances, the *real* problem were that your client
> didn't design good enough what they needed, nothing to do with a AMD vs
> Intel question.
>
> --
> SALUD,
> Jes�s
> ***
> [EMAIL PROTECTED]
> ***
------------------------------
From: [EMAIL PROTECTED] (Christopher Michael Collins ())
Subject: Re: Touchpad & pc_keyb.c
Date: 11 Aug 2000 07:04:44 GMT
In article <8ms2c8$r5s$[EMAIL PROTECTED]>,
Christopher Michael Collins () <[EMAIL PROTECTED]> wrote:
>
>Hello,
>
>I have a Compaq Presario 1685 with a synaptics touchpad.
>
>It has a kernel hack as it jumps and false clicks and
>basically doesn't work straight outta the box.
Would some kernel hacker please explain the fundamental
problem with the touchpad mouse and compaq 16xx laptops.
Is there a kernel and/or patch that will allow
this to work flawlessly(without gpm). For use
in X.
--Chris
------------------------------
From: [EMAIL PROTECTED] (John Hong)
Subject: Lexmark Optra 40?
Date: 11 Aug 2000 07:10:22 GMT
Can anyone still find this printer for sale?
------------------------------
Subject: Re: 20gb
From: [EMAIL PROTECTED] (Kenneth R�rvik)
Date: Fri, 11 Aug 2000 07:21:49 GMT
[EMAIL PROTECTED] (Harsh) wrote in <8n003u$pk7$[EMAIL PROTECTED]>:
> i have a 20gb hdd.
>
> wanna install redhat6.x . . .
>
> partitions are 10gb + 5gb + 5gb
>
> redhat shows 10gb+10gb. . .
>
>please guide
> can i install on secondary drive
Some more info would be OK, what kind of partitions are they? (DOS, FAT32,
ext2, primary, extended, logical etc...???)
--
Kenneth R�rvik 91841353/22718452
Steenstrupsgate 5 B [EMAIL PROTECTED]
0554 OSLO home.no.net/stasis
------------------------------
From: [EMAIL PROTECTED]
Subject: Re: getting dsl, what h/w should i get?
Date: Fri, 11 Aug 2000 07:37:11 GMT
Ray Tayek <[EMAIL PROTECTED]> wrote:
> In article <[EMAIL PROTECTED]>, [EMAIL PROTECTED] says...
>> No No No you don't want any kind of internal modem, you need an external
>> ethernet modem,
> ok, isa for regular dial-p modem and an external modem for dsl?
ISA for nothing. Stick to external.
------------------------------
Date: Fri, 11 Aug 2000 17:39:46 +0930
From: Stuart Fotheringham <[EMAIL PROTECTED]>
Subject: Re: 20.4GB HD, bios limitation, EZ-drive or any other?
After boot, LILO uses the BIOS to load the kernel.
The problem is that a lot of BIOSs can only access 1024 cylinders.
So each bootable partitions' boot code needs to be within 1024 cylinders.
Typically this is:
1024 cylinders * 63 sectors / cylinder * 255 heads * 512 bytes / sector
=> 8.4Gbytes.
Once the kernel gets going, it can access all of the disk.
Maybe you should reduce the first two partitions + linux /boot partition
to just less than 8.4Gbyte.
[EMAIL PROTECTED] wrote:
> James Richard Tyrer wrote:
> >
> > [EMAIL PROTECTED] wrote:
> >
> > > Hi,
> > >
> > > I have a 4 GB hda for windoze, 4 GB hdb for Debian and just installed a
> > > 20.4 GB maxtor HD and also running partition commender's boot manager. I
> > > would like to use this 20.4 gb hdd for debian only.
> > >
> > > My bios does not support bigger than 8.4 GB. I got the latest bios
> > > upgrade from gateway2000, but that didn't do anything. I tried setting
> > > the hard disk settings in the bios manually (instead of automatic
> > > configuration) with 39683 cylinder (manufacturer specs says this is the
> > > max cylinder), did not help. I included the 'append = "hdd =
> > > 39683,16,63"' in lilo.conf (as suggested by linuxdoc-howto web site),
> > > still can not get more than 8.4 gb. This howto says that linux can
> > > bypass the bios settings, but it didn't or I missed something.
> > >
> > > Is there any other way to get the 20.4 gb from this harddrive?
> > > Should I install EZ-Drive software that comes from maxtor? Will this
> > > screw up the things?
> > >
> > > Thanks.
> >
> > If you are using the whole disk for Linux, you don't have to use
> > EZ-Drive. The BIOS only has to be able to find the /boot directory, not
> > the whole disk. Make a small (10 Mbyte) partition for /boot at the
> > beginning of the disk and all will be fine.
> >
> > JRT
>
> Yes, I only want to use it for linux. So how to do it? Thanks.
------------------------------
Date: Fri, 11 Aug 2000 17:43:27 +0930
From: Stuart Fotheringham <[EMAIL PROTECTED]>
Subject: YAMAHA sound card
Does anyone know how to get the Yamaha sound card working under RH6.2
------------------------------
From: sideband <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: ISDN 128K modem
Date: Fri, 11 Aug 2000 04:23:32 -0400
"David .." wrote:
> Looking for recommendations for a good stable 128K ISDN modem internal
> or external to use with linux. Will be used to give internet access to
> 24 users systems.
>
> Thanks.
> --
> Confucius say: He who play in root, eventually kill tree.
> Registered with the Linux Counter. http://counter.li.org
> ID # 123538
I've had good luck with the NETGEAR XM128 and a high speed serial card.
The modem's serial port is capable of 460.8Kbps, so get a serial card
capable of at least that. The cost is nice too... $169 for the modem, and
$30 for a decent serial card that will handle the thruput.
HTH.
-SSB
------------------------------
From: Arne Van Theemsche <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux
Subject: 2 bttv cards... /dev/video1 does not give inputsignal
Date: Fri, 11 Aug 2000 10:21:55 +0200
Hi
I have two bttv cards (bt878 chipset hauppage new).. The two work fine,
when I use them apart, both work. I can capture images from composite1
input.
No the problem raises....when i put the two cards together in a pc, i
can capture images from composite1@video0, but when i plug in into the
other bttv-card, en change the device to video1,all i get is
blue-image... The cards are both detected (syslog), get an own
interrupt, and the /dev/video1 is recognized (by the realproducer, and
by xawtv/webcam since i get a blue signal).
Is there a pitfall somewhere to use two bttv cards?
thx in advance
Arne
------------------------------
Date: Fri, 11 Aug 2000 17:49:09 +0930
From: Stuart Fotheringham <[EMAIL PROTECTED]>
Subject: Re: 20gb
Probably the 5Gb + 5Gb are logical partitions on the second 10Gb
partition.
Harsh wrote:
> i have a 20gb hdd.
>
> wanna install redhat6.x . . .
>
> partitions are 10gb + 5gb + 5gb
>
> redhat shows 10gb+10gb. . .
>
> please guide
> can i install on secondary drive
------------------------------
Date: Fri, 11 Aug 2000 11:06:12 +0200
From: Bartek Kostrzewa <[EMAIL PROTECTED]>
Subject: Re: ATA-100 on ATA-66 moboard -- hdparms...
Alessandro Giachino wrote:
>
> Hello,
>
> my ATA100 drive is sloooow I think because moboard IDE
> controller is ATA66.
>
Are you using a shielded cable?
--
Bartek kostrzewa - [EMAIL PROTECTED]
<<< http://technoage.web.lu >>>
------------------------------
** 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
******************************
