>Laine, if it is any comfort, there are similar problems with the 2940UW on
>FreeBSD (same person is responsible for the code).

Not with running Ultra drives.  The only pending bug report I have is from
you, Darren, and it has to do with a tape drive.

It wouldn't surprise me if the NetBSD driver didn't handle Ultra as well
as the FreeBSD driver since it is a fairly old version of my work.  Now
that things have started to stabilize again in the FreeBSD driver, I plan
to renew my efforts of porting it back to NetBSD.

>I'm forming the opinion that the 2940UW is just another crap piece of PC
>hardware.

It's either that or your tape drive "is just another crap piece of PC
hardware". 8-)

>One of the FreeBSD people who had problems went to a 3940UW and his problems
>were solved.

This is exactly the oposite of what was reported.  He went from a 3940 to
two 2940s and his problems went away because his motherboard wasn't 
enabling memory mapped I/O on the DEC bridge chip and he turned on the 
option to use memory mapped I/O.  This was a configuration issue (and 
perhaps also a problem with the FreeBSD pci code in that it doesn't have
a mechanism to detect this problem similar to what Chris just added to
NetBSD).

>Darren

--
Justin T. Gibbs
===========================================
  FreeBSD: Turning PCs into workstations
===========================================