agrier@poofygoof.com asked;

> I used to get full sync
> transfers on all drives in my system, but now I get:
>
> sd0(asc0:0:0:0): asc0: timed out [ecb 0x80f1b000 (flags 0x1, dleft 2c, stat 0)], <state 5, nexus
x80f1b000, phase(l 1, c 1, p 1), resid 2c, msg(q 0,o 80) DMA active>
> sd0(asc0:0:0:0): asc0: timed out [ecb 0x80f1b000 (flags 0x41, dleft 2c, stat 0)], <state 5, nexus
x80f1b000, phase(l 1, c 1, p 1), resid 2c, msg(q 20,o 80) DMA active> AGAIN
> sd1: async, 8-bit transfers
> sd2: async, 8-bit transfers
> cd0: async, 8-bit transfers
>
> sd0 used to go 3.5MB/s sync, and as you can see, after first disk access
> all the disks get renegotiated as async.
>
> maybe it's something as simple as a timeout that needs to be extended?

I'm afraid not.  I can not recall whether NetBSD/pmax 1.5.4 was equiped with
its own asc driver.  NetBSD/pmax 1.6 features MI NCR driver which
is common across other ports, and it does not necessary mean achieving best
performance from particular hardware combinations.  I suggest you to make sure;
- Check you SCSI cabling.  Poor cable suffers machines badly.
- Remove or upgrade RZ56 with better one.  Mixing rusting-old-drive and
shiny-modern is a _bad_ practice.
- the driver complains about the 2nd (sd1) drive. It might be another trouble
source.

If you're serious to keep your DEC5000/200 useful, it's a good idea to try hard
to obtain PMAZB or PMAZC TC option cards.  They are DMA-accelerated,
and known far better than PIO PMAZ-A built inside DEC5000/200.

Toru Nishimura/ALKYL Technology