[Simh] DEQNA boot

[Bob Supnik]

It would depend on how you got there.

If the system was wired to go directly to the boot ROM on powerup (mode 
2, I believe), the microcode would set PS = 340. However, if you entered 
microODT and then jumped to the boot ROM via 173000G, the microcode 
would set PS = 0. Note that the SimH PDP11 doesn't simulate the boot 
ROMs; it uses workalikes.

Nonetheless, the clock shouldn't interrupt... provided that the system 
had a real clock card, with an interrupt disable. If it was an original 
11/23 with the half height card, then the line clock (BEVENT L) had no 
interrupt disable, just like on the LSI11/2. So a clock interrupt was 
certainly possible if the bootstrap was started manually.

Tim Litt pointed out that I could add a switch for starting boot code 
with PS = 0, but I can't see a use case that makes sense.

/Bob

On 10/21/2013 10:49 AM, Alan Frisbie wrote:
>> 2. boot routines should jam PS = 340, which is what is expected today.
>>
>> #2 is based on an examination of the official boot ROM listings. While
>> there may be paths through the boot ROMs that don't set PS = 340, the
>> primary path does that, and it is known to work with all the existing
>> boot ROMs in the simulator.
> My memory is far from perfect, but I recall a problem with
> the DEQNA boot ROM (internal to the DEQNA, not external ROM)
> in which it did not set PS=340.   As I recall, it loaded and
> then executed code in low memory (the vector area).   If a
> clock tick occured before the boot code finished, it would
> dispatch to somewhere out in the weeds.
>
> This was from a project I did in 1982, so my memory of the
> details may be faulty.   As I recall, we were using an LSI-11/23.
> If necessary, I can search for my notes on that project.
>
> Alan Frisbie
>
> --
> This e-mail is a natural hand made product.  The slight
> variations in spelling and grammar enhance its individual
> character and beauty and in no way are to be considered
> flaws or defects.
>
> --  Alan E. Frisbie               Frisbie at Flying-Disk.Com
> --  Flying Disk Systems, Inc.
>