Faulty hard disk hanging machine?

[Metod Kozelj]

Hello,

On Sat, 30 Jun 2001, I wrote:

> On Fri, 29 Jun 2001, I wrote:
> 
> > I have a presumably faulty hard disk, which can completely hang the
> > machine.
> 
> Well, not really as it show out. Today I tested the same hardware, but now
> running kernel 2.2.19 and aic7xxx 6.1.11 and everything works just fine,
> without any problems.

A while ago I was complaining about the new driver (6.1.x) not behaving
well while driving DEC RZ28D dec through Adaptec AIC-7881U controller. The
symptoms were complete lockup or, in best case, weird kernel oopses. The
cure was to limit the depth of TCQ. At that time my conclusion was that
there was some weirdness in disk's firmware.

Now I have just the same problem with IBM DDYS-T18350N and again, the
problems are cured by limiting the depth of TCQ to the exactly same
number: 32.

Does this mean anything?
Please note, that I've never had any problems with the Seagate ST34572N on
narrow channel. It happily limits itself to 64 tagged commands and runs
solid. Another note: the same setup runs rock solid running linux 2.2.19
with aic7xxx 6.1.11 without any queue depth limits.

Setup:

Linux 2.4.5 on Alpha SX164
aic7xxx 6.1.13
adaptec 2940UW
disks DEC RZ28D and IBM DDYS-T18350N on wide channel
disk Seagate ST34572N and CD-ROM NEC CD-ROM DRIVE:465 on narrow channel

Peace!
  Mkx

---- perl -e 'print $i=pack(c5,(41*2),sqrt(7056),(unpack(c,H)-2),oct(115),10);'


To Unsubscribe: send mail to majordomo at FreeBSD.org
with "unsubscribe aic7xxx" in the body of the message