>>> esp0: error:csr=b2930a13<INT,ERR,DRAINING=0,IEN,ENDMA,DSBL_SCSI_DRN,BURST=0,TCI
>>> esp0: DMA error; resetting
>>> esp0: SCSI bus reset

I just got this too (though in my case there was a space before "csr";
I trust that was a copying mistake in the original post).

>> ...how's the drive cabled to the fas?  internal SCA?

Yes, in my case - two internal SCA drives, nothing connected to the
external connector.  Appended after my signature is an edited-down
serial console capture.  I booted diskless, as you can see from the
bootpath: line.  The "big data transfer command" was copying from the
network to sd1, using two executables that lived on the NFS root area;
the machine on the other end of the connection (which was also the root
area's NFS server) was at only 10Mbps, so the U1's hme wasn't working
hard.  The filesystem on sd1 was on the whole disk and was created
under Solaris; I ran NetBSD fsck_ffs on it and it fixed the clean bit
but found nothing else to criticize.

I found the lack of a closing > on the "error:" line somewhat
disturbing.  After investigating, I ascribe that, plus the missing
DMAON bit and the missing second half of TCIDIS, to the buffer being
only 64 characters long; however, that does not explain what happened
to the DSBL_CSR_DRN bit (0x00010000), which appears in DDMACSR_BITS and
is set in the printed hex value but is not listed in the unpacking -
perhaps the source I'm looking at (lsi64854reg.h 1.5) does not match
the source Matt used to build that kernel?

If anyone has any suggestions for useful experiments for me to try, or
useful information I left out here, I'm certainly listening.

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B

NetBSD 1.6A (GENERIC) #0: Mon May 27 22:43:53 PDT 2002
    mthomas@matt02.allegronetworks.com:/u1/obj/sys/arch/sparc64/compile/GENERIC
total memory = 128 MB
avail memory = 109 MB
using 832 buffers containing 6656 KB of memory
bootpath: /sbus@1f,0/SUNW,hme@e,8c00000
mainbus0 (root): SUNW,Ultra-1
cpu0 at mainbus0: SUNW,UltraSPARC @ 167 MHz, version 0 FPU
cpu0: physical 32K instruction (32 b/l), 16K data (32 b/l), 512K external (64 b/l)
...
sbus0 at mainbus0 addr 0xfffcc000: clock = 25 MHz
DVMA map: ff800000 to ffffe000
IOTSB: 984000 to 986000
...
esp0 at sbus0 slot 14 offset 0x8800000 vector 20 ipl 3: dma rev fas
: FAS366/HME, 40MHz, SCSI ID 7
scsibus0 at esp0: 8 targets, 8 luns per target
...
scsibus0: waiting 2 seconds for devices to settle...
sd0 at scsibus0 target 0 lun 0: <SEAGATE, ST32550W SUN2.1G, 0414> SCSI2 0/direct fixed
sd0: 2048 MB, 3511 cyl, 11 head, 108 sec, 512 bytes/sect x 4194995 sectors
esp0: wide mode 1
sd0: sync (100.0ns offset 15), 16-bit (20.000MB/s) transfers, tagged queueing
sd1 at scsibus0 target 1 lun 0: <SEAGATE, ST32550W SUN2.1G, 0416> SCSI2 0/direct fixed
sd1: 2048 MB, 3511 cyl, 11 head, 108 sec, 512 bytes/sect x 4194995 sectors
esp0: wide mode 1
sd1: sync (100.0ns offset 15), 16-bit (20.000MB/s) transfers, tagged queueing
...
# ...big data transfer command...
esp0: error: csr=b2930a13<INT,ERR,DRAINING=0,IEN,ENDMA,DSBL_SCSI_DRN,BURST=0,TCI
esp0: DMA error; resetting
esp0: SCSI bus reset