Problem with Maxtor ATA card

[Keith Lofstrom]

Atrus <atrus at bellsouth.net> writes:
>
> I have 5 ide devices. hda, hdb, hdc, and hdd are on the motherboard's
> IDE controler. The fifth device, hde, was a Maxtor drive hung off of an
> Adaptec 1200A IDE card. Since there is only on disk connected to this
> card, there is no raid involved. This is a dual boot system where hda is
> XXXindows, hdb is Fedora Core, hdc and hdd are CD-ROMS, and hde is a
> shared drive between the 2 OS's.
>
> Now everything was just peachy until the Adaptec card Bombed on me. I
> had a spare Maxtor Ultra 133 card that I replaced it with. If the maxtor
> card is installed, regardless of whether I have a disk attatched or not.
> Fedora fails to boot. It gives a Kernel Panic: no init found.

I have seen the same problem with the Maxtor Ultra 133 (rebranded Promise
TX2 Ultra-133 ) on one of my systems - and not on the other!  It seems to
be sensitive to the motherboard (!).   The non-working motherboard is an
ASUS P4T-E, while the working motherboard is an ABIT TH7II-RAID .  I have
the Maxtor/Promise incompatable motherboard working with a Siig dual
ATA-133 controller ( rebranded ACARD ULTRA DMA133, CN2487 ), so the
issue seems to be a three-way incompatability between Linux, the Maxtor
card, and the ASUS motherboard.  

One thing to watch out for is the /etc/fstab file;  I've noticed that
this is often configured to look for a labeled drive rather than something
like /dev/hde1, and this can get a little screwy when you mix together
multiple drives with boot areas on them.  But this does not sound like
your problem, and you seem to be fairly aware of how all this works.

Perhaps a reinstall from CD onto a spare drive would be an interesting
science project - there might be some hardware-sensitive magic that
grub-install invokes when it builds the MBR for the drive.  Or perhaps
there is a flash upgrade for your motherboard.

No other suggestions, except that when you do find a workaround, please
share it with us so it helps the next person with the problem.  If you
can isolate problem better than I can, please put something up on 
bugzilla - I imagine Alan Cox (who did a lot of work on the drivers
for these LBA48 boards) is eager to debug this :-)

Keith

--
Keith Lofstrom           keithl at ieee.org         Voice (503)-520-1993
KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon"
Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs