Re: MVS under VM\ESA

--- In [email protected], Jay Maynard <jmaynard@...> wrote:
>
> On Fri, Aug 13, 2010 at 02:37:20PM -0400, Dave McGuire wrote:
> > 
> >   Does such a modified 3.8j exist?  I'm very much interested in doing
> > this.  I guess I was just assuming all along that it'd work out-of-the-box.
> 
> This is the same problem as running 3.8J on a P/390. I do that by IPLing in
> single step mode, manually altering CR7 appropriately, the setting it to run
> and going on. It should be easy to patch IEANIP00 to do this. MVS never
> touches CR7 after IPL.

I assume by CR7 you mean CR0.7, as CR7 is to do
with a secondary address space.

CR0.7 is described here:

http://publib.boulder.ibm.com/cgi-bin/bookmgr/BOOKS/DZ9AR006/3.4.1.1?SHELF=&DT=19990630131355&CASE=

but doesn't seem to match the description given
by Ivan either.

How many instructions do you need to single-step
through before altering the appropriate register?

What is the instruction that you look for before
stepping over it?

I'm not sure what to look for in the source code.

Note that the ieavnipm and ieavnpm3 source files
have been brought up-to-date so far.

http://mvs380.cvs.sourceforge.net/viewvc/mvs380/mvssrc/asm/ieavnipm.asm?view=markup

Also, I assume that if this is an otherwise unused
bit, MVS 3.8j will continue to run on standard
Hercules S/370?

Also, why is this necessary at all, from VM or
P390's perspective?

And also, how does VM/ESA or P/390 manage to cope
with MVS 3.8j doing old-style I/O, or 370 DAT?
Do they support S/370 mode?  If so, why the
requirement for a change in CR0 or whatever?

Thanks.  Paul.