Gurus only: Linux finds a non-existant IDE HD and reports weird kernel messages

Gurus only: Linux finds a non-existant IDE HD and reports weird kernel messages

Post by <dra.. » Fri, 01 May 1998 04:00:00



I have an interesting one here:

Pentium133Mhz (HX-chipset), blah blah blah running Slackware v3.2 and
a stock 2.0.33 kernel.  The machine has been very stable but we
recently started getting a lot of these kernel errors:

        03:05: rw=0, want=2001762903, limit=717664
        attempt to access beyond end of device

What is "03:05"?  It isn't the date or time.  I beleive it's
a code for a specific device.  /dev/hdb possibly?  Read on..

Anyway, we also noticed that the /dev/hdb partition table seemed
to be corrupted:

Disk /dev/hdb: 64 heads, 63 sectors, 525 cylinders

Nr AF  Hd Sec  Cyl  Hd Sec  Cyl   Start    Size ID
 1 00   1   1    0  15  63  203      63  205569 83
Partition 1 has different physical/logical endings:
     phys=(203, 15, 63) logical=(50, 63, 63)
Partition 1 does not end on cylinder boundary:
     phys=(203, 15, 63) should be (203, 63, 63)
 2 00   0   1  204  15  63  313  205632  110880 82
Partition 2 has different physical/logical beginnings (non-Linux?):
     phys=(204, 0, 1) logical=(51, 0, 1)
Partition 2 has different physical/logical endings:
     phys=(313, 15, 63) logical=(78, 31, 63)
Partition 2 does not end on cylinder boundary:
     phys=(313, 15, 63) should be (313, 63, 63)
 3 00   0   1  314  15  63  618  316512  307440 83
Partition 3 has different physical/logical beginnings (non-Linux?):
     phys=(314, 0, 1) logical=(78, 32, 1)

Etc... Anyway, to make a long story short, we moved the drive
from /dev/hdb to /dev/hdd and the partition table as FINE now.  
Fixed?  Wrong.  Now when you would boot the machine, dmesg would
report:

        hdb: non-IDE drive, CHS=1023/64/63
        hdb: INVALID GEOMETRY: 64 PHYSICAL HEADS?

But there wasn't any drive attached as /dev/hdb.  Fine, so we
disabled the SLAVE drive on the primary IDE controller.  SAME
THING!  Reset the BIOS to default settings, reconfigured it,
rebooted.. SAME THING!

I can't say I've ever seen anything like this before in all
my 7+ linux years.  Any ideas?

Please send responses via email to dranch at
"trinnet _-dot-_ net" and I will post solutions once it is
figured it out.

PS.  Sorry for the Anti-SPAM email address in the header and in
        the footer but what can I do?

--David

 
 
 

Gurus only: Linux finds a non-existant IDE HD and reports weird kernel messages

Post by Ulrich Teiche » Fri, 01 May 1998 04:00:00



>I have an interesting one here:
>Pentium133Mhz (HX-chipset), blah blah blah running Slackware v3.2 and

[deleted very strange story]

With a buggy RZ-100 or CMD640 IDE controler? Then you need the workarounds
for these shitty thingys compiled in your kernel.

HTH,
Uli
--

Stormweg 24               |listening to: Seasons In The Sun (Me First & The
24539 Neumuenster, Germany|Gimme Gimmies), It's the Sickening Sound (Doktor X)

 
 
 

Gurus only: Linux finds a non-existant IDE HD and reports weird kernel messages

Post by Simon McCab » Sat, 02 May 1998 04:00:00



> I have an interesting one here:

<big snip>

I have an almost identical problem that has ocurred twice on the slave
drive 1st IDE controller.
My WD21200 is no longer recognised by the BIOS (Gateway 2000 Pentium Pro
200 FX chipset)
after I patitioned it and made the file sys on the first partition.
This is a replacement drive from
WD (however the first one failed on my previous computer a 486DX2 66 it
was also connected
as a slave drive and both systems had 3 drives...)

The current symptoms are
(1) BIOS does not register a drive
(2) Linux (2.0.29 Slackware 3.2) sees incorrect information 64 heads as
in previous post and also
returns the identifier incorrectly.
(3) The Gateway diagnostic program tells me all is well with the drive
but cannot read the drive
identifier replacing it with <!   !    !   !>

As a previous post has suggested this may be due to a buggy
IDE controller but why does it only
die when Linux touches the drive (does Linux bypass INT13 ?) and what
has the controller done
to the drive to render it currently unrecognisable.  Western Digital
were not very forthcoming on
a technical explanation suggesting that I return the drive for another
replacement.  That is not a
solution IMO.

Patches I can find an explanation as to why the drive is no longer
recognised I cannot...

Your thoughts appreciated (and I will check the IDE controller chipset
right away)

Simon

------------------------------------------------------------------
Dr. Simon McCabe, Computational Physics Group,  Physics Department
University of York,  YO1 5DD           Tel: +44 1904 43(2276/2251)

 
 
 

1. kernel-api docbook fix for now non-existant PCI hotplug file

Here is a trivial documentation fix for some of recent PCI hotplug changes
that enables 'make htmldocs' to complete.

--- Documentation/DocBook/kernel-api.tmpl.orig  2003-03-11 16:14:25.000000000 -0800

      </sect1>
      <sect1><title>PCI Hotplug Support Library</title>
 !Edrivers/hotplug/pci_hotplug_core.c
-!Edrivers/hotplug/pci_hotplug_util.c
      </sect1>
      <sect1><title>MCA Architecture</title>
        <sect2><title>MCA Device Functions</title>

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in

More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

2. Good Linux CD-ROMs?

3. arch/i386/kernel/setup.c references non-existant website

4. Reducing the size of the primary paging space

5. Linux Threat: non-existant

6. ftape how do you use it

7. Patch?: linux-2.5.59/sound/soundcore.c referenced non-existant errno variable

8. help:sblive driver installation

9. regular hits for non-existant file...?

10. Non-existant IP-addresses

11. Conf.modules non-existant?

12. KDE2.0 compilation problems (make trying to enter non-existant directory)

13. HELP: Root shell non-existant!