Linux-Hardware Digest #854, Volume #14            Fri, 1 Jun 01 10:13:05 EDT

Contents:
  Re: WinTV BTTV driver with NO SOUND (John McCabe)
  Re: PCI Expansion card for USB Client(device) required ("Anders Larsen")
  video cards ("Didier Link")
  Re: Two tulip ethernet cards in a single RH 6.2 machine? ([EMAIL PROTECTED])
  Re: Maxtor ATA/100 Card: Kernel 2.4.2 not setting correctly? ("Wayne Osborn")
  Re: any recommendation for a dual monitor card? ("Wayne Osborn")
  Re: Linux Clustering Hardware ?s ("Peet Grobler")
  Maximum number of disk units to be connected on a Linux server ? (A. GUILLEVIC)
  Re: advansys SMP problems (Juergen Sauer)
  usblp0: device node registration failed ("Martin Schmidauer")
  Re: Linux, CPU HLT instruction et software cooling ! (Georges Goncalves)
  Re: Linux, CPU HLT instruction et software cooling ! (Georges Goncalves)
  Re: Linux, CPU HLT instruction et software cooling ! (H Dziardziel)
  Hercules Prohphet 4500 (Kyro 2) ("Dan Watt")

----------------------------------------------------------------------------

From: John McCabe <[EMAIL PROTECTED]>
Crossposted-To: alt.os.linux,alt.os.linux.mandrake
Subject: Re: WinTV BTTV driver with NO SOUND
Date: Fri, 01 Jun 2001 10:20:32 +0100

I've a similar weird problem with the WinTV PVR.... When I leave the machine off 
overnight and then boot up into Mandrake 8.0 there is
no sound, there isn't even a volume control in either Zapping or Xawtv..

The only way to get it back is to shutdown the system, boot into window, run the TV 
there, then reboot back into Mandrake and the sound
now works..
 Bloody weird.

/John


------------------------------

From: "Anders Larsen" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.embedded
Subject: Re: PCI Expansion card for USB Client(device) required
Date: Fri, 01 Jun 2001 11:18:15 +0200

On Thu, 31 May 2001 17:38:58 +0200, "Girish G" <[EMAIL PROTECTED]>
wrote:

> Don't know if this mail befits this newsgroup. However, could not find
> a better place to post to. So here goes.
> 
> We need an PCI board for USB CLIENT(b'coz our hardware does not have a
> USB port, either client or host and we need a USB client sitting on
> PCI). Any recommendations. We have searched the web but did not come
> up with anything worthwhile
> 
> Please note, the board must be PCI

Hi,

I don't know where you're located, but here in Germany a USB interface
card (2 ports, PCI) is available from the supplier Secomp
(http://www.secomp.de/ - order no. 15.06.2152).
The catalogue doesn't state any manufacturer, so I guess it comes
from somwhere in the far east (the low price tag, EUR 23,-, comfirmes
this); on the picture you can see that it's based on a VIA chip.

HTH
cheers
  Anders
-- 
"In theory there is no difference between theory and practice.
 In practice there is."  - Yogi Berra

------------------------------

From: "Didier Link" <[EMAIL PROTECTED]>
Subject: video cards
Date: Fri, 01 Jun 2001 11:45:17 +0200

Hello everybody,
I will upgrade my voodoo3 3000 AGP for a new cards.
I have the choice beetween a geforce2 MX or a voodoo5 5000 AGP.
What is the better under Linux (XFree 4.0.3, T-Bird 1GHz) ?
 Thank you for your interest.

Did

------------------------------

From: [EMAIL PROTECTED]
Subject: Re: Two tulip ethernet cards in a single RH 6.2 machine?
Date: Fri, 01 Jun 2001 10:08:49 GMT

Peter T. Breuer <[EMAIL PROTECTED]> wrote:
: Richard Cobbe <[EMAIL PROTECTED]> wrote:
:> Ga Mu <[EMAIL PROTECTED]> writes:
:> The only trick, of course, is that I have the driver compiled into the
:> kernel, not built as a module.  I get the impression from the Ethernet
:> HOWTO that this should work---doesn't it?

: It could, but why bother? It's always better to have modules. That way
: you can play with the driver parameters easily, which is precisely what
: you want to do! What kernel parameters did you pass to the driver at
: bootup?

: I don't know if the tulip driver supports more than one card, btw. Check.
: If not, you'll have to rename one of the modules and be specific about
: which card it controls. It might be wise to spearate the IRQs of the
: cards too, as I don't know if they can share for real or not, or if the
: driver supports a shared interrupt ..

: Peter

I was only able to get my two tulip cards to work compiled as a module
with separate IRQ's on RedHat 6.2 and two straight DEC DE435's, although 
I also tried an off brand which uses the tulip.

-- 

------------------------------

From: "Wayne Osborn" <[EMAIL PROTECTED]>
Subject: Re: Maxtor ATA/100 Card: Kernel 2.4.2 not setting correctly?
Date: Fri, 01 Jun 2001 18:07:55 +0800

In article <nMvR6.18330$[EMAIL PROTECTED]>, "Michael
Barton" <[EMAIL PROTECTED]> wrote:

> I've got a Maxtor ATA/100 PCI card installed, with a Maxtor ATA/100 80GB
> drive attached.  Kernel Version 2.4.2 boots up fine with it.
> 
> If I use hdparm to check the device, it sats:
> 
> /dev/hde:
> 
>  Model=Maxtor 98196H8, FwRev=ZAH814Y0, SerialNo=V80732FC Config={ Fixed
>  }
>  RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=57
>  BuffType=DualPortCache, BuffSize=2048kB, MaxMultSect=16, MultSect=16
>  CurCHS=16383/16/63, CurSects=-66060037, LBA=yes, LBAsects=160086528
>  IORDY=on/off, tPIO={min:120,w/IORDY:120}, tDMA={min:120,rec:120} PIO
>  modes: pio0 pio1 pio2 pio3 pio4
>  DMA modes: mdma0 mdma1 mdma2 udma0 udma1 udma2 udma3 udma4 *udma5
> 
> Note that udma5 is checked.
> 

yes, but that does not mean it is running udma5.... Its the max support
ed in the drive.

> 
> But if Icheck the status of the card using 'cat /proc/ide/pdc202xx', I
> get:
> 
>                                 PDC20267 Chipset.
> ------------------------------- General Status
> --------------------------------- Burst Mode                           :
> enabled Host Mode                            : Normal Bus Clocking      
>                   : 66 External IO pad select                        :
> 10 mA Status Polling Period                : 0 Interrupt Check Status
> Polling Delay : 0 --------------- Primary Channel ----------------
> Secondary Channel -------------
>                 enabled                          enabled
> 66 Clocking     enabled                          disabled
>            Mode PCI                         Mode PCI
>                 FIFO Empty                       FIFO Empty
> --------------- drive0 --------- drive1 -------- drive0 ----------
> drive1 ------
> DMA enabled:    yes              no              no                no
> DMA Mode:       UDMA 4           NOTSET          NOTSET NOTSET
> PIO Mode:       PIO 4            NOTSET           NOTSET NOTSET
> 
> Note thate drive0 on the Primary Channel is in UDMA 4 mode, and the
> Clocking is set to 66 (consistent with UDMA 4) - not 100, which it
> should be for UDMA 5.
> 
> Any thoughts?
> 

I'm no expert here, just got my 1st ATA100 drive myself. I understand
that udma5 is not supported yet?

> ---
> 
> Michael Barton
> Engineering Software Solutions, Inc.
> 303-948-6199
> [EMAIL PROTECTED]
> 
> 
> 
> 


-- 
  Wayne A. Osborn, SCADA Engineer.[dnar AT iinet DOT net DOT au]
  Registered Linux User #212818.  [2.2.16-22] [i686]
  6:00pm  up 21:18,  3 users,  load average: 1.00, 1.00, 1.00
  ...BASIC is the Computer Science equivalent of `Scientific Creationism'.

------------------------------

From: "Wayne Osborn" <[EMAIL PROTECTED]>
Subject: Re: any recommendation for a dual monitor card?
Date: Fri, 01 Jun 2001 18:09:27 +0800

In article <9f65di$bur$[EMAIL PROTECTED]>, "Peter White"
<[EMAIL PROTECTED]> wrote:

> Steven <[EMAIL PROTECTED]> wrote in message
> news:GewR6.27141$[EMAIL PROTECTED]...
>> Thanks for the rec... Are you using their PowerDesk package?  I am
>> leaning towards them just because of the idea that they are actively
>> trying to
> make
>> a Linux dual monitor solution work.
>>
>> Thanks
>>
>>
>>
>>
> No - I didn't know they did a powerdesk package for Linux! Maybe I'll
> check it out, but I'm quite happy with how its working in KDE and Gnome,
> so not too bothered about additional s/w. I did try their software under
> NT (maybe that was powerdesk?), before I moved my box to linux. From
> what I remember it gave me "virtual desktops" that I could switch
> between. Well, you get this anyway in KDE, so with the default 4
> desktops for each head, I have 8 - more than enough. I can use
> "Xinerama" (comes with KDE) which makes a single virtual desktop, but I
> prefer being able to use the heads independently. Incidently, I followed
> all the many threads on the Matrox site about configuring
> XFree86Config-4 (I think), and the best results came from just using the
> Mandrake Control Center - Display setup (I did have to fix the fonts in
> that file though.
> 
> 

Just got the G450 myself, how did you configure Gnome for dual-head???


-- 
  Wayne A. Osborn, SCADA Engineer.[dnar AT iinet DOT net DOT au]
  Registered Linux User #212818.  [2.2.16-22] [i686]
  6:00pm  up 21:18,  3 users,  load average: 1.00, 1.00, 1.00
  ...BASIC is the Computer Science equivalent of `Scientific Creationism'.

------------------------------

From: "Peet Grobler" <peetgr at absa.co.za>
Subject: Re: Linux Clustering Hardware ?s
Date: Fri, 1 Jun 2001 12:32:32 +0200

If you don't mind me asking...

What software do you use for the clustering? E.g. what are you running on
top of Linux, is it a lot of work to install/tune, etc. I'm looking into
doing it as well. I've read through a great deal of sites, and understand
what it's about, but how did you implement it?

Regards,
Peet
David Konerding wrote in message ...
>On 31 May 2001 08:40:38 -0700, Dan Mesimer <[EMAIL PROTECTED]> wrote:
>> Greetings all,
>>
>>     I have been scouring through the Newsgroups and a multitude of
>> websites for an answer to my questions, but cannot seem to find it.
>> So I will ask you all.
>>
>> I would like to setup a small cluster at home for educational value
>> using MOSIX.  I have a AMD 600 and an old Pentium 133 that I would
>> like to use.  My question is:  I already have a NIC in the AMD and
>> have it hooked up to my Cable modem and all (actually through a switch
>> which has other computers hooked up to it so as to share the cable
>> modem which is fine).  I want to know if it is possible to basically
>> add a NIC to the AMD and the old pentium and have them use these new
>> nics as the network for the 2 node Cluster?  And if I did this would
>> it cause problem with my internet access on my main (AMD)?  Would I
>> have to set them up as routers?  Or am I making this way to hard?
>
>Sounds perfectly reasonable to me.  I have a similar set up: a
>"master" node which is a dual-homed host (two NIC's on two networks).
>eth0 (the first NIC) is a "true" internet address, reachable from
>anywhere (like the NIC attached to your cable modem) and eth1
>(the second NIC) uses one of the so-called "private IP's" such as 10.*.*.*
>and 192.168.*.*.  I made sure to configure the master node carefully such
>that the default route was to the internet (using my exterior gateway
address).
>For convenience I set up a DNS server and NFS server on the master node,
>as well as a firewall on the master node which disallowed any access to the
>DNS and NFS from outside.  I gave the master node's eth1 the address
10.0.0.1
>and all the cluster nodes are 10.0.0.2 to 10.0.0.whatever.
>
>I didn't actually make the master node a "router" per se since I didn't
>need to access the internet from the cluster nodes.  Therefore I simply
>didn't enable ip forwarding in the kernel...
>
>Dave



------------------------------

From: [EMAIL PROTECTED] (A. GUILLEVIC)
Subject: Maximum number of disk units to be connected on a Linux server ?
Date: 1 Jun 2001 03:52:44 -0700

Hello,

To sweep away disk space limit we had planned to plug our Domino Server to a 
HP XP256 storage bay. Our server runs Red Hat 6.2 with kernel 2.2.18. We chose
a Qlogic Fiber Channel card for the server. The first step was to allocate around
140 Gbyte of storage space shared in 8 Gbytes LUNs (that is about 17 LUNs).
Unfortunately, only a part of all these LUNs were finally detected when 
Fiber Channel link was activated. If I add all the SCSI disks I already have
and the new units, no more than 15 units are detected. IS there a limit to the
number of disks kernel can see ? Neither storage bay nor Qlogic cards can gather
units themselves.

                     Thanks is advance for your suggestions.

------------------------------

From: Juergen Sauer <[EMAIL PROTECTED]>
Subject: Re: advansys SMP problems
Date: Fri, 1 Jun 2001 10:40:14 +0000 (UTC)

Anthony Ewell <[EMAIL PROTECTED]> schrieb
am Thu, 31 May 2001 22:44:49 -0700 in comp.os.linux.hardware:
> Hi All,
>    I have a dual PIII-600 (MSI MS-6120, V1.9 bios)
> with a Megaraid and an Advansys (ConnectCom) 3940u2w
> card.

>    I can boot perfectly under "linux.up" (single processor
> kernel).  But, when I boot up with the dual kernel ("linux"),
> boot will slow to a crawl as the Advansys drive crashes it was
> through two time outs per bus address.  Needless to say,
> anything on the 3940u2w does not work after it FINALLY boots up.

>    Does anyone have any suggestions?
This task you should really ask the advansys Support for it.
The Advansys Supporters are supporting Linux directly from
manufacturer.
They should give a solution for you.
mfG
        Jojo


-- 
J�rgen Sauer - AutomatiX GmbH, +49-4209-4699, [EMAIL PROTECTED] **
** Das Linux Systemhaus - Service - Support - Server - L�sungen **
http://www.automatix.de to Mail me: remove: -not-for-spawm-     **

------------------------------

From: "Martin Schmidauer" <[EMAIL PROTECTED]>
Subject: usblp0: device node registration failed
Date: Fri, 1 Jun 2001 14:41:28 +0200

Hi,

I've got a problem with my USB printer. I use kernel 2.4.5.  If I send a
file to /dev/usb/lp0 notjing happens.
My kernel messages:

Jun  1 01:24:17 odin kernel: usb.c: deregistering driver usblp
Jun  1 01:24:45 odin kernel: usb.c: registered new driver usblp
Jun  1 01:24:45 odin kernel: printer.c: usblp0: device node registration
failed
Jun  1 01:24:45 odin kernel: printer.c: usblp0: USB Bidirectional printer
dev 2 if 0 alt 0
Jun  1 01:24:45 odin kernel: usb.c: usblp driver claimed interface cfa17d40
Jun  1 01:24:45 odin kernel: printer.c: v0.8 Michael Gee, Pavel Machek,
Vojtech Pavlik, Randy Dunlap
Jun  1 01:24:45 odin kernel: printer.c: USB Printer Device Class driver

tia,
-Martin.





------------------------------

From: Georges Goncalves <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Linux, CPU HLT instruction et software cooling !
Date: Fri, 01 Jun 2001 15:11:15 +0200



faeychyld wrote:

> Just an observation here, Linux uses sys logging and cron
> to perform various housekeeping tasks constantly.
> 
> I am not sure if the cpu would be halted and started each time these
> functions were accessed. Windows approach is quite different,it will
> sit idle for hours with the drive(s) shut down, so why not the CPU?.

I know that but the machines are Idle between 97.3 and 99.8% of their
time
and especially when I go through my tests :)

-- 
Georges 'Melkor' Goncalves <[EMAIL PROTECTED]>

------------------------------

From: Georges Goncalves <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.setup
Subject: Re: Linux, CPU HLT instruction et software cooling !
Date: Fri, 01 Jun 2001 15:22:11 +0200



Serban-Mihai Popescu wrote:

> Are you sure you're comparing apples to apples? Maybe lm_sensors is not
> properly tuned and could show wrong numbers.

Both HMonitor and LMSensors are giving the GOOD temperatures. Every temp
sample
I took was (and is) confirmed by the BIOS display report (1� max of
variation)

> A better way would be to let the system settle down and, in both Windoze
> and Linux, to quickly reboot and go see what the bios says about the
> temperature. 15*C difference would not go away that fast.

Both Windows and Linux run at an average of 55�C as HMonitor and
LMSensors report 
(confirmed by a quick reboot and a look at the BIOS report). If I
activate the
"CPU Low Power Mode" on HMonitor (win), within 1 minute, the 55�C
temperature
drops to 37-39�C (immediately rebooting the machine and checking the
BIOS also
reports this NEW temperature). I'd just like to know what to do on Linux
to obtain
the same result.

-- 
Georges 'Melkor' Goncalves <[EMAIL PROTECTED]>

------------------------------

From: [EMAIL PROTECTED] (H Dziardziel)
Subject: Re: Linux, CPU HLT instruction et software cooling !
Date: Fri, 01 Jun 2001 13:46:52 GMT

On Thu, 31 May 2001 11:30:10 +0200, Georges Goncalves
<[EMAIL PROTECTED]> wrote:

>
>
>Andr� David wrote:
>
>> > What puzzles me is that all modern OSes (WinNT/2K, Linux and others) are
>> > ALREADY meant to use the CPU HLT command when they're Idle, but, the
>> > temperature remains constant (does not drop) HMonitor claims using the
>> > HLT command and the temperature really drops.
>> >
>> > Who's right, who's wrong ?
>> >
>> 
>> Are you really sure that in Linux you don't have some deamon running in
>> the background?
>
>I have lots of them (Apache, ProFTPd, SSH, LDAP, Bind, X, PostgreSQL,
>Sensord,
>Postfix, SolidPOP3d and Samba) but most of the time, my machine is
>between
>98.3 and 99.6% Idle with a load of 0.00 :-)
>
>BTW, my Windows machine also have at least one daemon running (SMB for
>instance)
>but the temperature drops.
>
>-- 
>Georges 'Melkor' Goncalves <[EMAIL PROTECTED]>

Hello, the Linux hlt does work or it would not be part of the kernel?
My tests using external cpu heatsink thermocouple sensors and current
draw metering have consistently shown Linux idle power consumption
(and cpu temperaure) is about the same as hlt enabled windows except
in (mobile cpu) laptops where it seems to be a bit less idle.   I
believe maybe that is due to laptop bios being tailored for windows
and that particular laptop power management and perhaps Linux software
clock and similar design functions intrinsic to basic running.  But it
could also merely be that laptop current measurements can be easier
and more accurate than line powered as I have no intention of cutting
into internal wiring for this.  In any case the difference is
negligible, a few percent.  They were made with all daemons that I
could find and control, off.    

In your case with all those daemons, irrespective of the idle/load
figures which I am not at all aware of how calculated (includes
daemons etc?), comparing the two o/s should be done with daemons put
to bed during the test.  Especially all the network stuff.

Also, perhaps the AMD implementation of hlt differs?  Regards  

------------------------------

From: "Dan Watt" <[EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.help
Subject: Hercules Prohphet 4500 (Kyro 2)
Date: Fri, 1 Jun 2001 09:06:41 -0700

What support is there for the Prophet 4500 (Kyro 2 chipset) in Linux? Right
now Win2k is still my primary platform, but I am slowly migrating to Linux,
and due to recent video card problems, I am considering purchasing a Prophet
4500. I have Mandrake 8 installed already and have used it with my TNT2, and
Im not too concearned about 3d acceleration in linux (at least not for a
while).

Basically, will the card work at say 1280x1024 with generic or hardware
specific drivers (if available)?

Dan Watt
[EMAIL PROTECTED]



------------------------------


** 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 by posting to comp.os.linux.hardware.

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
******************************

Reply via email to