Linux-Hardware Digest #432, Volume #13 Wed, 16 Aug 00 09:13:07 EDT
Contents:
Re: Dual processor board? ([EMAIL PROTECTED])
Re: Question ([EMAIL PROTECTED])
Re: PC133 ? (Kenneth R�rvik)
Re: Dual processor board? (Kenneth R�rvik)
Wrong title: DPT on restart ("Andreas Moroder")
DPR Problems on restart ("Andreas Moroder")
buying a new printer (Philipp Jeyapalan)
Promise FastTrak66 - only one channel is UDMA ("Jake")
cat /proc/tty/driver/serial (Florian Forster)
Re: Not all memory detected under linux (Peter Robin Hiesinger)
Re: 45 GB HD - How to get past BIOS limitation? (Alberto Gianoli)
Re: HP scanject 5p, I can't configure it (Robert Jones)
Re: Vibra 128PCI (Terence Chan)
Re: Not all memory detected under linux ("Peter T. Breuer")
Re: added hard drive and zip drive stopped working (Pierre Dupuy)
Re: cat /proc/tty/driver/serial (Pierre Dupuy)
Re: deleted partition table (Pierre Dupuy)
----------------------------------------------------------------------------
Subject: Re: Dual processor board?
From: [EMAIL PROTECTED]
Date: Wed, 16 Aug 2000 07:23:06 GMT
"D. Stimits" <[EMAIL PROTECTED]> writes:
> One sample of why, from the log:
> kernel: unexpected IRQ vector 217 on CPU#1!
I really need to get hold of the kernel source. It's slightly
frustrating looking at the MP Spec and trying to reverse-engineer how
Linux does it. I really don't even want to make any guesses at this
point, not without looking at the code.
> Apparently this new functionality requires adding irq handler code,
> but the code doesn't exist, so it sends it to limbo.
The MP Spec mentions that an APIC may generate a spurious interrupt
under certain circumstances ("for example, if an interrupt goes
inactive just after the first INTA cycle but before the second INTA
cycle"). It also says that the distributed APIC architecture is more
susceptible to this problem.
Intel says you should check the state of the device so that you know
if the interrupt was a real one or not. Does the PCI bus provide a
way to do this?
> It is good, however, to
> see progress on this subject. Apparently the IRQ assertion needs a
> handler. Chances are that the hardware that the assertion uses
> hard-codes or firm-codes 217 as a standardized value (or soon-to-be).
Or perhaps the 8th bit was reserved and not ignored in the kernel as
it should have been.
> Will anyone be able to use this new information to help stabilize the
> i840 boards?
I'll help, for sure. But I don't even have an 840 board; good ol'
440GX and 440FX for me. Anyone who wants to help may wish to look at
Intel's Multi-Processor Specification, v1.4, which is at:
http://developer.intel.com/design/intarch/manuals/242016.htm
As soon as I get my soon-to-be Linux computer operational, I will
certainly look at the source.
--
Eric McCoy ([EMAIL PROTECTED])
"The politics of failure have failed! We need to make them work again!"
- Kodos, The Simpsons
------------------------------
Subject: Re: Question
From: [EMAIL PROTECTED]
Date: Wed, 16 Aug 2000 07:25:03 GMT
erik m jonte <[EMAIL PROTECTED]> writes:
> What is the equivalent (if any) of Scandisk in Linux?
fsck, which will, very likely, just run e2fsck.
--
Eric McCoy ([EMAIL PROTECTED])
"corsair, n. A politician of the seas."
- Ambrose Bierce, The Devil's Dictionary
------------------------------
Subject: Re: PC133 ?
From: [EMAIL PROTECTED] (Kenneth R�rvik)
Date: Wed, 16 Aug 2000 07:25:32 GMT
[EMAIL PROTECTED] wrote in <[EMAIL PROTECTED]>:
>Sandhitsu R Das <[EMAIL PROTECTED]> writes:
>
>> I have set my DRAM:Freq ratio to 4:3 (I have K7V - soft-menu control) -
>> which is supposed to make the DRAM run at 133 MHz. And the system _is_
>> running - does that prove that I have PC133 SDRAM ?
>
>No. It proves that, whether you have PC133 or PC66 RAM, it's capable
>of running at 133MHz well enough not to produce any immediate errors.
>
>Viking and Kingston memory, I believe, have a sticker on each chip
>that says.
Also look for the labels on the chips on the DIMM - often there is a -6 or
similar. pc66 is often -10 (1/10ns -> 100MHz theoretical max, often no more
than 83MHz in practice). pc100 will often be -8 or better, so pc133 needs
to be even faster than that. PLease beware that not all chips conform to
this - there are 10ns chips that will happily run at 100MHz, but I don't
think there are that many 8ns that will run at 133(?). Anyway, since you
are running your memory at 133MHz, smile and be happy :)
--
Kenneth R�rvik 91841353/22718452
Steenstrupsgate 5 B [EMAIL PROTECTED]
0554 OSLO home.no.net/stasis
------------------------------
Subject: Re: Dual processor board?
From: [EMAIL PROTECTED] (Kenneth R�rvik)
Date: Wed, 16 Aug 2000 08:12:52 GMT
[EMAIL PROTECTED] wrote in <[EMAIL PROTECTED]>:
>As soon as I get my soon-to-be Linux computer operational, I will
>certainly look at the source.
Or you can go to this site: http://lxr.linux.no/source/
--
Kenneth R�rvik 91841353/22718452
Steenstrupsgate 5 B [EMAIL PROTECTED]
0554 OSLO, Norway home.no.net/stasis
------------------------------
From: "Andreas Moroder" <andreas[nospam][EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: Wrong title: DPT on restart
Date: Wed, 16 Aug 2000 09:59:55 +0200
pleas look at the previous message
------------------------------
From: "Andreas Moroder" <andreas[nospam][EMAIL PROTECTED]>
Crossposted-To: comp.os.linux.misc
Subject: DPR Problems on restart
Date: Wed, 16 Aug 2000 09:31:59 +0200
Hello,
on our Server we have ad DPT 3334 Controller with 6 disks ind 3 RAID-1
Pairs.
Whiel running all works well.
Last week we did shutdown the machine two times because we changed the
tapedrive.
Both time after restart we got a hd-failure with one of the disks of the
second pair with the "failed" led on.
( one time the upper disk, the second time the lower, not allways the same )
The first time we changed the failed disk.
The second time i reformatted the failed disk on another machine an
reinserted it. Now all seems to be ok, the is no "failed" led anymore.
Ca anyone tell me if this i a problem of the controller or do I need to send
a command to the controller before shutting down the system ( to let the
controller really flush all the data to the disks ).
I anyone had the same problems, please let me know
Andreas Moroder
P.S. please send me a e-mail amoroder@se-nord.[nospam]provinz.bz.it without
[nospam] for humans
------------------------------
From: Philipp Jeyapalan <[EMAIL PROTECTED]>
Subject: buying a new printer
Date: Wed, 16 Aug 2000 10:48:36 +0200
Hi all,
I'm thinking of getting a new printer. I'm intending to buy a color inkprinter
(so no laser) and have had a lot of trouble with my HP 710c as I try to get it
working under Linux. Does anybody have any adwise for me, which product to buy
or which one I should keep away from. I'll be glad for any tips and hints.
Thanks a lot.
Philipp
ps. I've already checked http://www.linuxprinting.org
------------------------------
From: "Jake" <[EMAIL PROTECTED]>
Subject: Promise FastTrak66 - only one channel is UDMA
Date: Wed, 16 Aug 2000 09:01:51 GMT
Hi All,
I have a Promise FastTrak66 RAID card (BIOS v1.21 build 13)
running under RedHat 6.2 (kernel upgraded to 2.2.16 with ide patch)
using a ASUS CUV4X mobo.
I have two 75 GB drives attached to the Promise FastTrak66 and
each 75 GB drive is seperate (not striped) as jumpered for master.
One disk (master) per channel. I'm really just trying to use the
RAID card as an UDMA66 controller at the moment. Eventually
I'll run RAID 0.
cat /proc/pci
RAID storage controller: Promise Technology IDE UltraDMA/66 (rev 1).
Medium devsel. IRQ 10. Master Capable. Latency=32.
I/O at 0xb400 [0xb401].
I/O at 0xb000 [0xb001].
I/O at 0xa800 [0xa801].
I/O at 0xa400 [0xa401].
I/O at 0xa000 [0xa001].
Non-prefetchable 32 bit memory at 0xde800000 [0xde800000].
Bus 1, device 0, function 0:
so I added this to/etc/lilo.conf
append="idebus=66 ide2=0xb400,0xb002 ide3=0xa800,0xa402"
then ran /sbin/lilo. Reboot.
Both drives show up when LILO loads (/dev/hde and /dev/hdg)
but only one is recognized as UDMA66 capable.
% dmesg
Uniform Multi-Platform E-IDE driver Revision: 6.30
ide: Assuming 66MHz system bus speed for PIO modes
VP_IDE: IDE controller on PCI bus 00 dev 21
VP_IDE: chipset revision 16
VP_IDE: not 100% native mode: will probe irqs later
PDC20262: IDE controller on PCI bus 00 dev 78
PDC20262: chipset revision 1
PDC20262: not 100% native mode: will probe irqs later
PDC20262: (U)DMA Burst Bit ENABLED Primary MASTER Mode Secondary MASTER
Mode.
ide2: BM-DMA at 0xa000-0xa007, BIOS settings: hde:pio, hdf:pio
hda: IBM-DTLA-307075, ATA DISK drive
hde: IBM-DTLA-307075, ATA DISK drive
hdg: IBM-DTLA-307075, ATA DISK drive
ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
ide2 at 0xb400-0xb407,0xb002 on irq 10
ide3 at 0xa800-0xa807,0xa402 on irq 10 (shared with ide2)
hda: IBM-DTLA-307075, 73308MB w/1916kB Cache, CHS=9345/255/63, UDMA(66)
hde: IBM-DTLA-307075, 73308MB w/1916kB Cache, CHS=148945/16/63, UDMA(66)
hdg: IBM-DTLA-307075, 73308MB w/1916kB Cache, CHS=148945/16/63
The problem is /dev/hdg, which is on the second channel of
the Promise board, isn't UDMA capable but the first channel on
the Promise card (/dev/hde) is udma capable. I have something setup
incorrectly I presume.
%hdparm -t /dev/hde
/dev/hde:
Timing buffered disk reads: 64 MB in 2.04 seconds = 31.37 MB/sec
%hdparm -t /dev/hdg
/dev/hdg:
Timing buffered disk reads: 64 MB in 11.51 seconds = 5.56 MB/sec
Any ideas why /dev/hdg isn't be recognized as UDMA capable
but /dev/hde is?
Thanks,
Jake
------------------------------
From: Florian Forster <[EMAIL PROTECTED]>
Subject: cat /proc/tty/driver/serial
Date: Wed, 16 Aug 2000 11:23:57 +0200
Hello
doing a cat on /proc/tty/driver/serial gives me lines like this
0: uart:16550A port:3F8 irq:4 baud:9600 tx:0 rx:0
tx should be transmitted as i guess and rx revieved.
But sometimes I also get outputs with
fe: 21
what does fe mean or where can I gain detailled infos about this output.
Already searched the I-Net and the Kernel Documentation but unfortunately
found nothing.
Thx in advance
Flo
------------------------------
From: Peter Robin Hiesinger <[EMAIL PROTECTED]>
Subject: Re: Not all memory detected under linux
Date: Wed, 16 Aug 2000 11:36:18 +0200
Hi!
I followed the discussion about RAM detection, because in my Linux
Laptop only 64 of 128M were detected. mem=128M lets it correctly see the
128M during startup, but /proc/meminfo then says there are only 111M.
Does anyone know where the missing 17M might be gone?
robin
------------------------------
From: Alberto Gianoli <[EMAIL PROTECTED]>
Subject: Re: 45 GB HD - How to get past BIOS limitation?
Date: Wed, 16 Aug 2000 12:51:00 +0200
sideband wrote:
> Gene Montgomery wrote:
>
> > Jake wrote:
> > >
> > > Gene Montgomery <[EMAIL PROTECTED]> wrote in message
> > > news:[EMAIL PROTECTED]...
> > > > A little added info. From the mini HOW-TO called Large-Disk, I
> > > > note that in addition to the BIOS limitation, there is a Linux
> > > > limitation in older kernels. Does anyone know where the break
> > > > comes in? That is, does 2.2.14 handle disks larger than the 33.8
> > > > GB limitation mentioned in the mini-HOWTO?
> > >
> > > I has able to attach/see/access two 75 GB drives using RedHat 6.2,
> > > which I beleive ships with kernel 2.2.14, without any modification
> > > on both the primary and secondary motherboard IDE channels.
> > >
> > > ASUS CUV4X motherboard
> > > IBM Deskstar 75 GB
> > > RedHat6.2
> > >
> > > I ened up upgrading the kernel to 2.4.0test5 to get it to
> > > work with a RAID controller card though.
> > >
> > > Jake
> >
> > Thanks, Jake. I will concentrate on the BIOS limitation,
> > under the assumption that RH6.2 (2.2.14) has solved the 33.8
> > GB limitation.
> >
> > Gene
>
> Gene:
>
> The what the BIOS detects, and what Linux can use are two wholly
> different things... Linux goes straight to the drive to determine its'
> size and attributes, rather than using the BIOS, so you should be able to
> do exactly what you want to do with that 45G drive regardless of what the
> BIOS says.
>
> HTH
>
> -SSB
If your BIOS is too old (and hangs when detecting the harddisk), and you are
not
booting from that disk (this was my situation), you can do the following:
just tell
the bios there is no harddisk! (but keep the ide interface active) Once
Linux will
find an active ide interface it will probe it independently, overruling
whatever the
bios says. Of course you have to check which version of fdisk (and
eventually
mount) you are using: RH6.2 was fine for me (IBM deskstar 75GB), while RH6.1
was not.
cheers,
alberto
--
---
Alberto Gianoli - PPE div. bat 32 2-A14, CERN - 1211 Geneve 23, Switzerland
=======================================================================
Phone : 00-41-22-767-9827 | Email : [EMAIL PROTECTED]
Fax : 00-41-22-767-8920 |
------------------------------
From: Robert Jones <[EMAIL PROTECTED]>
Subject: Re: HP scanject 5p, I can't configure it
Date: Wed, 16 Aug 2000 06:30:58 -0500
"Peter T. Breuer" wrote:
> Robert Jones <[EMAIL PROTECTED]> wrote:
> : Guy Maskall wrote:
> :> "Alastair (LiQUiDx) Tse" wrote:
> :> > danny-mio wrote:
> :> > > I have a scanner HP scanject 5p and I don't know how to use it on Linux,
> :> > > it's comunicating with my pc by the card included with the scanner (SCSI i
> :> > > suppose).
>
> If it's the same one I have, it's a perfectly normal scsi scanner. No
> problems. Of course you have to throw away the lobotomized half-a-scsi
> controller card that comes with it (it's most of an ncr5x...) and use
> a real controller. I use a cheap $20 symbios 8bit card.
>
> :> > > I'd appretiate any hint
> :> >
> :> > Try using SANE - http://www.mostang.com/sane/
> :> >
> :> > I've got my HP 5P to work, but only from 1.01. Doesn't seem to work
> :> > under 1.02 (?).
>
> I've never been able to comprehend the sane docs. Maybe it's time I
> looked at them again. But the hpscanbpm utility has always worked fine
> for me (good with tkscan!) and so I've never had the incentive.
>
> :> You should need a driver for the card. A lot of the cards bundled with scanners
> :> are barely scsi-standard compliant. If you can find a driver for it and it
> :> works then great, but you may be needing to buy a 'proper'scsi card.
>
> Indeed.
>
> : I'm running RH6.0 with a 2.2.16 kernel. A friend has offered me a HP-IIP scanner
> : that he no longer uses, along with the interface card that "came with it". I
> : already have an Adaptec 2940 installed, running a DAT tape drive. Will the
> : scanner work with the Adaptec SCSI card? What other software will I need to scan
>
> Yes. None if it's a scsi scaner (they support a minimal set of
> commands).
>
> : images? Is there any OCR software available for Linux that will work with this
> : combination?
>
> Oh! OCR. That's a different proprietary story. It's all in software, of
> course.
>
> Peter
Thanks for the answers and for the observations re sane. I have been warned! I find
myself reverting to a couple of scripts using tar for my backups for the very same
reason.
And yes, OCR >is< a whole 'nother matter. Back in the middle 80's, I had a couple of
customers who owned standalone OCR machines (~$35,000 each, thank you very much) that
were used as input devices on machines I serviced. Rip the covers off the one that
worked best, and discover the PDP-? (8, I believe) with a pair of 8" floppies. IF
every typewriter in the building used the same font and IF the typeballs were clean
and IF the sheets didn't suffer in their internal mail, it would deliver upwards of
95% accuracy!
I know things are much better than that on the Dark Side; hoping that something is
also available for Linux. If not, what the hell! It might give me an excuse to boot
Windows every now and then. ;-)
Regards
------------------------------
From: Terence Chan <[EMAIL PROTECTED]>
Subject: Re: Vibra 128PCI
Date: Wed, 16 Aug 2000 19:58:34 +0800
Oh! Thank a lot. I can listen music with RH now. But the sound volume is
mush low then the one I hear with xxx98. Why?
lobotomy wrote:
> Say yes after it plays/doesn't play the test sound, and let it set up
> the driver. Then try to play a sound. When I was using my Yamaha
> OPL3/SAx with RH 6.0, the sndconfig test never worked, but XMMS and
> other programs outputting sound worked fine. I would think you would
> have little problem the ES1371 (yes, that's what it really is -
> Creative bought out Ensoniq and renamed their line of cards) under
> linux, as it is PCI and should be configured pretty automatically.
>
> On Sun, 13 Aug 2000 15:24:53 +0800, Terence Chan <[EMAIL PROTECTED]>
> wrote:
>
> >I've installed RH 6.2 on my machine. I tried to set up my Vibra 128PCI
> >sound card with sndconfig. It detect the sound card as "Ensoniq | Es1371
> >Audio PCI-97]. But hear nothing when testing. What can I try now?
> >
> >
> >--
> >-(^)-(^)- ! Terence Chan
> > b ! mailto:[EMAIL PROTECTED]
> > --- !
> >
> >
> >
--
-(^)-(^)- ! Terence Chan
b ! mailto:[EMAIL PROTECTED]
--- !
------------------------------
From: "Peter T. Breuer" <[EMAIL PROTECTED]>
Subject: Re: Not all memory detected under linux
Date: 16 Aug 2000 12:08:47 GMT
Peter Robin Hiesinger <[EMAIL PROTECTED]> wrote:
: I followed the discussion about RAM detection, because in my Linux
: Laptop only 64 of 128M were detected. mem=128M lets it correctly see the
: 128M during startup, but /proc/meminfo then says there are only 111M.
It doesn't "say" anything. It's a load of figures. How are you
interpreting them so as tobelieve that only 111M of ram is available to
you! Try and understand that (a) it is YOU who is saying that, not the
machine, and (b) we have no interest in commenting on your beliefs
instead of the data, and (c) we need the data to comment on.
: Does anyone know where the missing 17M might be gone?
Nowhere of course. Show the data to us for comment, but don't expect
anything from a statement like 'memory is missing'. Here for example
is the output of /proc/meminfo for my 128MB machine:
cat /proc/meminfo
total: used: free: shared: buffers: cached:
Mem: 131395584 125083648 6311936 22073344 45395968 38891520
Swap: 264200192 43274240 220925952
MemTotal: 128316 kB
MemFree: 6164 kB
MemShared: 21556 kB
Buffers: 44332 kB
Cached: 37980 kB
SwapTotal: 258008 kB
SwapFree: 215748 kB
Now which one of those are you looking at? Are you maybe using a video
card that uses system ram? (this is a possibility). Are you looking at
"used" instead of "total". Or are you worried abut "free"?
Peter
------------------------------
From: Pierre Dupuy <[EMAIL PROTECTED]>
Subject: Re: added hard drive and zip drive stopped working
Date: Wed, 16 Aug 2000 14:54:08 +0200
did you check the master / slave jumpers ?
If the bios is not too old, you can try the auto IDE detection: does it
work ?
last, be careful that the disk letter (the x in /dev/hdx) can vary when
you adr or remove a disk.
Pierre
Cliff Bergman wrote:
>
> Hi. I'm having a problem.
>
> I have a new PC that came with an IDE hard drive and an IDE zip drive.
> Everything seemed to be working normally. I added a second IDE hard
> drive and suddenly the zip drive is misbehaving. It can't seem to read
> any of my preexisting zip disks. And furthermore, it can't read a brand
> new zip disk. If I look at such a disk with, say, the command
> fdisk -l /dev/hdd,
> it claims that the disk does not have a valid partition table.
>
> Does anybody know what could cause such a change in behaviour? Needless
> to say, neither Dell (the hardware manufacter) or RedHat can figure out
> the problem.
>
> Vital statistics:
> Dell Precision 420
> 20gb Maxtor Hard Drive (IDE)
> 250gb Zip drive (IDE)
> Redhat 6.2
>
> The machine also has an IDE CD-ROM drive. I added a 3gb IDE hard drive.
>
> Thanks for any help.
> cliff bergman
>
> real email address: cbergman 'at' iastate.edu
------------------------------
From: Pierre Dupuy <[EMAIL PROTECTED]>
Subject: Re: cat /proc/tty/driver/serial
Date: Wed, 16 Aug 2000 14:56:41 +0200
frame error ?
These are reception errors, this can happen when you connect on the fly
during a transmission, or when you misconfigure the speed, the number of
data bits or parity bit.
Florian Forster wrote:
>
> Hello
>
> doing a cat on /proc/tty/driver/serial gives me lines like this
> 0: uart:16550A port:3F8 irq:4 baud:9600 tx:0 rx:0
>
> tx should be transmitted as i guess and rx revieved.
> But sometimes I also get outputs with
>
> fe: 21
>
> what does fe mean or where can I gain detailled infos about this output.
------------------------------
From: Pierre Dupuy <[EMAIL PROTECTED]>
Subject: Re: deleted partition table
Date: Wed, 16 Aug 2000 14:50:41 +0200
As far as I remember, jaz drive come configured with a single partition
number 4 (eg /dev/sda4).
With fdisk, you can try to reinsteall such a partition (new, primary,
4), change the type to dos, and write the new partition table.
To be sure, just look at another jaz drive :-)
Cordialement,
Pierre
[EMAIL PROTECTED] wrote:
>
> Sean <[EMAIL PROTECTED]> wrote:
>
> > However, not being too bright, I managed to delete the partition table using
> > the linux fdisk on a 1G Jazz drive that had a single dos partition on it. No
> > changes have been made since.
------------------------------
** 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
******************************
