| Newsgroups |
gmane.comp.emulators.hercules390.mvs |
| Message-ID |
<[email protected]> |
> -----Original Message-----
> From: [email protected] <[email protected]>
> Sent: 21 September 2019 16:53
> To: [email protected]
> Subject: RE: [H390-MVS] Re: Wait State 00050064
>
> Robert Schreiber wrote:
>
> > S/370 has no idea what an LPAR is. Thus, the fact that it's running
> > in one should be shielded so that it cannot be confused.
> > @Tony_Harminc correctly pointed out that the image expects the machine
> > to look/act like GA22-7000-10.
> > The problem that I caused myself was because STIDP did not adhere to
> > the S/370 spec with the configuration that I
> > (wrongly) created.
>
> As I said to Dave, I would agree. I don't like the current behavior myself for the
> exact reason you and Tony offered.
>
> But as I also said to Dave, I was not (and still AM not!) in any position to
> question it, not having any experience with such beasts.
>
> I'm thinking your "FORCE370" option is best we can hope for right now.
>
> UNLESS of course, someone can step up and offer solid, clearly convincing
> evidence that the current behavior is absolutely, 100%, without any question,
> WRONG. I mean, perhaps such systems (as the ES/9000 that Dave mentioned)
> only supported 370 operating systems that were "LPAR-aware"? <shrug>
>
> (I don't like this issue any more than the rest of you!)
Fish,
I would say, as usual its "unclear" So looking at the ES/9000 Functional Characteristics
Page 1-12 says that
Control Programs for Basic Modes
===========================
The following control programs operate in basic mode:
• MVS/SP Version 3
• VM/ESA System Product Releases 1.0 and 1.1
• MVS/ESA System Product Release 4.1
• MVS/ESA System Product Release 4.2.
Control Programs for Logically Partitioned Operation
===========================================
The following System/370 control programs operate in LPAR mode:
• MVS/SP Version 2 Release 1.3
• MVS/SP Version 1 Release 3.5
• VM/SP Release 5 and subsequent releases
• VM/SP High Performance Option Release 4.2 and subsequent releases
• VSE/ Advanced Function Version 2 and Version 4
• VSE/SP Version, Version 3, and Version 4
Which seems odd as it means older versions of operating systems work in a 370 mode LPAR but not in a basic mode ES9000...
Page 1-8 says:-
You select the operating mode during system initialization.
This occurs for a single-image configuration or for each side of a partitioned configuration. The operating mode can be basic mode or LPAR mode.
When you select LPAR mode, you must specify the Architectural mode (ESA/390 or System/370) of each partition.
Page 3-3 says:-
A doubleword designated by the second operand address of the Store CPU ID (STIDP) instruction provides information about the ES/9000 Processor Complex..
More information can be found in IBM Enterprise Systems Architecture/390 Principles of Operation. A summary of the contents of the doubleword follows:
System/370 or ESA/390 Mode:
VV A SSSSS 9821 0000
LPAR Mode:
VV L PSSSS 9821 8888
• VV is the version code (two hexadecimal digits):
39 (Model 330)
49 (Model 340)
28 (Model 500)
38 (Model 580)
45 (Model 620)
65 (Model 720).
• ASSSSS or LPSSSS constitutes the central processor (CPU) identification number (six hexadecimal digits):
The first digit (A in the basic modes, L in LPAR mode) is the CPU address, as stored by the Store CPU Address (STAP) instruction (see
"CPU Address Identification" on page 7-15):
Digit Model
1 330
1 348
1 or 2 508
0 or 1 or 2 580
1 or 2 628 - (side 0)
3 or 4 620 - (side 1)
0 or 1 or 2 720 - (side 0)
3 or 4 or 5 720 - (side 1)
The next five digits (SSSSS) are, for the basic modes, selected from the serial number of the ES/9000 Processor Unit. In LPAR mode, the logical
partition identifier replaces the first of the five digits (PSSSS) The next four digits are 9021 (for the processor unit machine type).
The last four hexadecimal digits are 0000.
To me that implies that you get the second format when you are in an LPAR and in S/370 mode,
Dave Wade
> --
> "Fish" (David B. Trout)
> Software Development Laboratories
> http://www.softdevlabs.com
> mail: [email protected]
>
>