Port-vax archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: POLYx addressing-mode questions



> [...], plus if you run VMS on SIMH emulating a model that doesn't
> have POLY instructions, won't VMS emulate them properly, and
> therefore a test can tell you?

If you have VMS, yes.  I'd even be willing to run it for this purpose,
depending on the costs - is there a no-money version of VMS available
with at least an assembler and linker?  I could turn POLYx into
emulated-instruction traps in my code and then run VMS, and even dust
off those decades-old memories of VMS while I'm at it.  If, that is,
VMS wouldn't pitch a fit at seeing what otherwise looks like a KA630
taking emulated-instruction traps for instructions that are, according
to the VARM I have, implemented by the KA630.

> Which models of VAX have POLY instructions?

Assuming the list in appendix B of the VARM I have is (a) accurate and
(b) comprehensive, every VAX includes POLY support for every
floating-point type it supports.  I suspect it is accurate; I don't
know how comprehensive it is.  Here's what it says:

The MicroVAX-I has POLY for the floating-point types it implements,
which can be F and D or can be F and G, depending on which model of
uV-I you're talking about.

The MicroVAX-II also has POLY for the floating-point types it
implements, which are F, D, and G.

The 730 (and 725, which "is a repackaged version of the [730]")
"includes all the instructions" and also is "the first to include
G_floating and H_floating as standard".  I assume F and D floats are
present as well, though this is not explicitly stated.

The 750 "includes all the instructions (G_floating and H_floating are
available as an option)".

The 780 also "includes all the instructions (G_floating and H_floating
are available as an option)".  "The floating point architecture was
revised in January 1979 to add two new data types (G_floating and
H_floating) and to take faults instead of traps on floating exceptions
for the original data types (F_floating and D_floating) as well as the
two new data types.  All VAX-11/780 processors have been modified to
take faults instead of traps."

The 782 "is a dual processor [780] with shared memory".

The 785 "is identical to the [780] from the point of view of software,
except that the [785] has increased performance and has a bit set in
the SID".

The 8200 "includes all the instructions and architecturally defined
processor registers, but does not include compatibility mode".

The 8300 is a dual-processor 8200.

The 8500 is a lower-performance lower-cost 8700.

The 8600 "includes all the instructions, architecturally defined
processor registers, and compatibility mode".

The 8650 "is...a field upgrade of the [8600]".  I assume this means
it's got the same support, but this is not explicitly stated.

The 8700 "is a single-processor version of the [8800].  It can be
field-upgraded to a VAX 8800 by addition of a section processor".  I
suspect "section" is a typo for "second".

The 8800 "includes all the instructions and architecturally defined
processor registers, but does not include PDP-11 compatibility mode".

/~\ The ASCII				  Mouse
\ / Ribbon Campaign
 X  Against HTML		mouse%rodents-montreal.org@localhost
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B


Home | Main Index | Thread Index | Old Index