Adding a hard drive

[D. Hugh Redelmeier]

| From: Paul King <sciguy-Ja3L+HSX0kI at public.gmane.org>
| 
| I have a 320G hard drive recently purchased, and after resolving some IRQ
| conflicts with my network card (this HD runs off a PROMISE card, an IDE card
| like my Ethernet card), I later try to format the HD, and it fails to format
| under both XP and Linux. Both systems can see the HD. The HD is a Seagate 320
| GB IDE with a 16 MB cache.

I've not used one of these cards.  Sometimes they have a BIOS that
creates a geometry that Linux does not see.  In fact, some of that
geometry stuff was actually removed from Linux in recent years.

Mind your, the error messages don't look like geometry errors to me.  The 
problems look to be at a lower level.

| The HD appears as /dev/hdg, and fdisk showed that it had no format to begin
| with (unusual in my recent experience). I gave it a partition under fdisk, and
| selected "W95 32 Bit" as I wanted to see the HD under both systems (I have
| dual boot). Then I attempted to format it using mkfs.vfat, and the
| 
| Buffer I/O error on device hdg1, logical block 30
| end_request: I/O error, dev hdg, sector 187
| Buffer I/O error on device hdg1, logical block 31
| Warning: could not read block 0: Attempt to read block from filesystem
| resulted
| in short read
| end_request: I/O error, dev hdg, sector 63
| Warning: could not erase sector 0: Attempt to write block from filesystem
| resulted in short write

Those look bad to me (but I'm not an expert).  Does dmesg tell you more?

| The motherboard is an ASUS A7V running an Athlon K7 processor and 1 GB of RAM.

I vaguely recollect that VIA chipsets of that era could have problems
with some PCI cards.  I'm not saying that this is what you are
hitting.

| The HD is on the same cable as a DVD-ROM (one which is only lightly used). The
| HD is intended to be used as storage, with no executables. The HD is hooked up
| to a 2-port PROMISE IDE card.

Are you sure that the promise card supports anything other than hard
drives?  Some RAID cards of that did not support ATAPI.  I never
knew whether this was a hardware of driver implementation.  (I had an
ABIT BP6, with a RAID chip onboard.  So did the main IDE driver
maintainer (since replaced).  He said to only use hard drives on it,
not cdroms, and he should know and was in a position to fix it if it
were fixable.)

Are you using an IDE cable suitable for ultradma?  You know, the ones
with twice as many conductors?

Certainly the first thing I would try would be to take everything else
off the channel.

If your controller is really old, you might be experiencing the ~137G
limit of the old IDE protocol.  Unlikely, I think.

Consider testing the drive on your main IDE interface, without your
original drive.  You haven't got much to lose since anything on the
new drive is toast.

Don't forget to set the jumpers correctly.  I've heard that the new
IDE cables (the ones with twice the conductors) are "cable select".

I've not read it, but a glance at this suggests it might be worth
reading: http://en.wikipedia.org/wiki/AT_Attachment

| Any help people can give me regarding my hard drive problem will be helpful.

All these were shots in the dark.  Good luck!
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists