AHA-2940UW problems with Linux driver

[Daniel M. Eischen]

> I've been through a similar debugging effort in the past although in that
> particular case, it was a SEEK RAID array and the OS was FreeBSD.  In the
> end, it turned out to be a SEEK problem where the Adaptec was queuing so
> much I/O, so quickly that the SEEK would get bogged down in handling it's
> internal request queue and either crash or return QUEUE FULL conditions in
> odd cases.  SEEK was unable to reproduce these problems with any other
> platform/controller pair and insisted that it was a device driver problem.
> Eventually SEEK admitted it was their fault when the SCSI analyzer traces
> exonerated the aic7xxx driver.
>
> I'll help out where I can in this effort, but as my development platform
> is FreeBSD, not Linux and the code that I run on a daily basis has diverged
> to a large extent from what Linux (and FreeBSD current for that matter) 
> run, I may not be able to offer exact fixes.

They're not running with your last released sequencer/driver.  Linux
2.0.30 was based on your driver from well over a year ago I believe.
I've pointed him at the latest driver which is in 2.0.31-pre-7 (don't
you love those names :-)).

Dan Eischen
deischen at iworks.InterWorks.org