PROM fields are at octal (not decimal) offsets + there are more fields of interest

[David Rosenberg]

The Multics Differences document specified the locations of PROM fields at octal offsets, but Multics interpretted those offsets as if they were decimal. We since found out that ISOLTS interprets the offsets as being octal (as the Multics Differences document specifies). The simulated PROM has been adjusted to place fields at the octal offsets. Multics has to be corrected to expect them there.

Issue #112 documents this and also documents new fields that the DPS8M simulator project added to the simulated PROM so that Multics can know which version of the simulator it is running and can issue a warning if that version of Multics knows that it would be adversely affected by the (older) version of the simulator on which it is running. Code should be added to Multics to::

1. Read the additional information from the PROM
   
2. (Possibly) store it somewhere appropriate
   
3. Write it to a log file
   
4. Issue a warning message if the currently running version of Multics knows that it could be adversely affected by a known problem in the running simulator that has been corrected in a later version of the simulator

[David Rosenberg]

This ticket contains three separate, but closely related, tasks. One is a defect and two are enhancements:

1. (Defect): Read CPU Model number, CPU Serial Number, and Ship Date fields at offsets interpreted as octal, rather than decimal, numbers.
2. (Enhancement): Read and report new information added to the PROM (e.g., Release type, release version number, target architecture, and target operating system, etc.)
3. (Enhancement): If this version of the simulator is not a general release (i.e., the Release type character is not an "R"), issue a warning. Compare this version of the simulator to the latest released version known at the time of this Multics release. If this version of the simulator is earlier, issue a warning. In particular, if this earlier version of the simulator is known to have a defect that could adversely affect the operation of Multics, include that in the warning

[Eric Swenson]

​https://s3.amazonaws.com/eswenson-multics/public/mcrs/MCR10120.pdf

[Eric Swenson]

Fixed with MR12.8-1033 install.

[Eric Swenson]

Actually, fixed with MR12.8-1037: MR12.8h install.