Looking through the MV II sources, it appears that the watch chip is
simply not implemented. The watch chip should have registers at
200B8000. The MV II emulator has "128 bytes" of NVR (non-volatile RAM)
starting at that address instead.
The NVR implementation isn't right either. Although the watch chip
occupies addresses 200B8000-200B807E, the bytes of the watch chip are
accessed on word (16b) boundaries. So there are really only 64 bytes
available - 14 bytes of registers and 50 bytes of battery-backed up RAM.
Byte access is undefined.
The byte corresponding to CSR D defaults to 0. The VRT (valid time) bit
is supposed to be set by hardware when the register is read. Because
this doesn't happen, VRT always remains 0, and VMS will always ask for
the time and date.
See chapter 11 of the KA630 processor spec for all the gory details.
So yes, feel free to file a bug report.
/Bob Supnik
On 4/11/2016 10:16 AM, [email protected] wrote:
Message: 1
Date: Mon, 11 Apr 2016 14:33:21 +0200
From: Wilm Boerhout<[email protected]>
To: SIMH Mailing List<[email protected]>
Subject: [Simh] MicroVAX II time across boots
Message-ID:<[email protected]>
Content-Type: text/plain; charset=utf-8; format=flowed
How does a real MicroVAX II preserve its time across boots?
How is this implemented in simh MicroVAX II? i.e. what virtual device
should I attach to a file so that the time is preserved in VMS?
My configuration (below) does not preserve the time (VMS V7.3 fully patched)
Thanks,
/Wilm
_______________________________________________
Simh mailing list
[email protected]
http://mailman.trailing-edge.com/mailman/listinfo/simh
