P4 on sun3/sun3x (Re: NetBSD/sun3 10.0_RC2 problems on sun 3/60)

Le lun. 15 janv. 2024 =C3=A0 01:00, foo bar <[email protected]> a =C3=A9cr=
it :
> On Sun, Jan 14, 2024 at 3:35=E2=80=AFAM Izumi Tsutsui <[email protected].=
ne.jp> wrote:
> > Oh, actually 501-1248 works on 3/80? (Mine has 501-1443)
> > "THE SUN HARDWARE REFERENCE" part 1 says P4 on 3/80 is not same
> > as P4 on 3/60, but maybe it's about physical shape..
> >  https://mail-index.netbsd.org/port-sun3/2023/05/07/msg000199.html
> I think that's talking about the physical shape of the cards. As far
> as I can tell the 3/80 is the only machine with a P4 slot that does
> not have holes for BNC connectors so the cards won't physically fit in
> the machine. I had to take the motherboard out, remove the back plate,
> and put the motherboard in a sparcstation clone case to use this
> cgfour.

Well, thank you, that seems to confirm that the P4 is logically/electricall=
y
the same across sun and sun3x, or at least they are 'backward compatible'
(they could have added /STERM on an otherwise not conflicting pin to
open the option of synchronous cycle to the P4 device, though by that
time SPARC was the future and sun3/sun3x were not likely to be
updated).
The partial pinout that was shared with me on this list strongly suggested
so, and my own continuity checking confirmed a few more signals.
Has a 3/60 framebuffer works in a 3/80, that means at least all the signals
for asynchronous cycle (of the '020/'030) are in the same spot, and
so do most/all of the power signals, and no conflicts.

Unfortunately my 3/80 motherboard blew a cap when I tried to power it
up with my homemade ATX adapter (https://github.com/rdolbeau/SS1_ATX).
The pinouts of the adapter seem OK to me and I haven't figured out why,
I presume the cap died of old age...
Anyway the mobo didn't give any sign of life.on the serai port when i
tried again, so for now my idea of a P4FPGA to follow up on the
SBusFPGA is stalled.

Cordially,

--=20
Romain Dolbeau