[HECnet] decnet.org

[Johnny Billquist]

On 2012-07-11 17:16, Bob Armstrong wrote:
You can't take the default on *all* questions, since some don't have ny
defaults.
The CSR and vector for the terminal lines, for example... :-)

    You're right - I did have to enter the address for that one.

:-)

    Simh is rather simple minded about the terminal muxes - it considers the
DZ11/DZV11/DZQ11/CYX08/CXY16 all to be the same.   Although the hardware is
different as far as I know they all have the same programmer model so this
is not unreasonable. The only thing that differs is the number of lines per
board, and simh allows you to set that.

I thought the CXY muxes had a different interface than the DZ, but if I remember that one wrong, then I guess you should be ok.

    And the only floating vector in this configuration, AFAIK, is the DZ/CXY
and 300 is the first one, right?

Yes. But vectors should be less of an issue. However, the floating CSR address is the more important one.

    FWIW I did a VMS SYSGEN (which, among other things allows you to enter a
list of devices and will calculate for you the correct CSR/vector
assignments) and it agrees with simh in that regard.   It particular, it says
160100 and vector 300 for the DZ11.   As far as I know the PDP-11 and VAX
rules for address and vector assignments are the same, but it's possible RSX
does something different.   It's suspicious that the default answer for the
first DZ11 in the RSX SYSGEN is not the same as the default floating address
assignment.

No. For autoconfigure, the architecture does not matter. It's a set of rules for the bus, not the CPU.
The RSX SYSGEN is not clever enough to calculate a reasonable default value for CSR addresses of devices. So it just gives a fixed value as the default, if it gives a default at all.

    The other devices (MSCP disk, MSCP tape, NI) are all their default
addresses for the first such device.

Yes. Actually, the first device of those types have a fixed address.

You really need to make sure your hardware is correctly configured before
you can expect ACF to work.

    Sure, but that's the whole question.   I think this is the correct
configuration unless you can see what's wrong.

I thought the CXY looked different from a DZ, which is why I thought things were wrong. Since it would appear I was wrong on that, I can't see any problems. And in the next post, you even note yourself that it works fine, if you just select another CPU, so it would definitely appear that your config here is ok.

    The other possibility that there's some hardware in there that this
version of ACF doesn't know about.

Don't appear to be, and I would have been surprised if there had been.

    I guess I could do another SYSGEN and just take out the DZ/CXY from the
simh configuration.   That's easy enough to do.

I guess at this point, we're beyond that.
However, the fact that ACD fails if you select an 11/53 in simh, but works if you select an 11/73 makes me wonder if there is something funny in the 11/53 CPU type in simh.

	Johnny

-- 
Johnny Billquist                                   || "I'm on a bus
                                                                  ||   on a psychedelic trip
email: bqt at softjar.se                         ||   Reading murder books
pdp is alive!                                         ||   tryin' to stay hip" - B. Idol