If that helps, I have a:
ahb0 targ 1 lun 0: <FUJITSU, M2624S-512, 0202> SCSI2 0/direct fixed
sd1 at scsibus0: 496MB, 1429 cyl, 11 head, 64 sec, 512 bytes/sec

running a -current i386 kernel.

->If so, that would be strong evidence that it's the pmax scsi driver
->that's at fault, rather than the drive itself not handling
->write-clustering.

What's so magic on the SCSI-Side of write-clustering?

Gruesse
       Oskar

 ==============================================================================
 					Tragbar ist, was nicht herunterfaellt.
						    oskar@zappa.unna.ping.de
 ==============================================================================