[ale] hard drive won't mount

[Danny Cox]

All,

On Thu, 2002-06-20 at 14:08, keith frost wrote:
> the maxtors may have problems that only show up with
> the 2.4 kernel, but (and i mean no disrespect to the
> kernel people) isn't this where opensource is supposed
> to prevail.  you would think that somewhere there
> would be a module option that would allow the drive to
> think it was talking to a 2.2 kernel?

	Having spent the last year (almost) attempting to learn the ins and out
of the 2.4 IDE subsystem, I may have a little insight.

	One origin of CRC errors is attempting to run on a too-high DMA mode. 
UDMA modes above 2 require an 80 conductor cable.  While I've been able
to push a drive to a faster mode without such a cable, it's not
recommended (and the code prevents it [unless one modifies it]).

	The 2.4 IDE driver also has code which specifically starts winding back
the DMA mode when CRC errors occur.  I don't see where this state info
is displayed (in /proc), but it's in there.

	I do also remember the Maxtor drive.  If, after the first N I/Os, it
received no CRC/checksum failures, it *must* be working okay, right? 
So, the drive turned that check off, to be faster.  Sheesh!  BTW, the
very largest Maxtors only run at 5400 RPM.

	In the interest of full disclaimer, I work for Quantum.  On the other
hand, we no longer make hard drives, and we're using Western Digitals
now.  The IBMs we had kept on having spotty problems.  Most were fine,
but sometimes a whole run (whatever a "run" is) would exhibit problems. 
I don't know any more than that.  Fortunatly, the folks at
www.hardocp.com agree.  They don't like the IBM Deskstars either.

	Enough rambling.

-- 
kernel, n.: A part of an operating system that preserves the
medieval traditions of sorcery and black art.

Danny


---
This message has been sent through the ALE general discussion list.
See http://www.ale.org/mailing-lists.shtml for more info. Problems should be 
sent to listmaster at ale dot org.