For kicks, I frobbed with David's new 5380 driver, and I'm pleased to 
announce that I have DVMA working on my 4/260.  With the new driver, I've 
checked and mounted my filesystems, created a few files, removed them, 
etc.  It's quite a bit more zippy (it certainly feels an order of 
magnitude faster, and the drive light has gone from a slow flicker to a 
near-solid "on" when doing disk operations..)

The "sw" stuff should work too, but in pio-mode.  I basically 
short-circuit to non-dma in that case, since I don't know much more 
about that controller than where the registers are.

Anyhow, this is before the changes Gordon made to that driver today, so 
I'll merge those diffs in with what I have ...

There's a few iffies I still have ... Mostly, I remember having a 
converstaion with Chuck and Paul about the interaction bewteen the 
write-back cache on the 4/200-series and DVMA, but I don't remember what 
we decided the solution was (this came up in a conversation about the 
4/200-series and the Xylogics driver...).  In any case, I can't really 
risk data corruption on this box, so I'm going to be cautious with it for 
a while, or at least until I a) can find the e-mail outlining the 
solution, or b) hear back from Chuck and/or Paul on the subject :-)

Anyhow, thought you'd be interested.

Oh, BTW ... I must say "kudos" to David once again.  The driver is 
extremely well designed .. the modifications (besides the autoconfig 
stuff, which I pulled directly from the old si driver) to make it fit in 
with the sparc world (well, VME world, anyhow) were quite trivial.  In 
fact, a fair number of them were to work around #define evil I had to do 
in ncr_sireg.h to get the register layout of the "sw" right...

Ciao.

------------------------------------------------------------------------------
Jason R. Thorpe                                         thorpej@Xenotropic.COM

           Just me and my collection of obsolete computer gear(s).