difference among sun4u, sun4m, sun4d
Mathon
What's the difference among sun4u, sun4m, sun4d etc.
uname manual says(indicates) that this is an obsolete
machine hardware name(class). I also remember that
I read somewhere that ISV is encouraged to use uname -p
instead to make new developed program more portable.
Anyway, I am still curious about what the original meaning
of sun4u, sun4m, sun4d etc.
TIA
---
Mathon
Neil W Rickert
>What's the difference among sun4u, sun4m, sun4d etc.
>uname manual says(indicates) that this is an obsolete
>machine hardware name(class). I also remember that
They are different versions of sparc hardware. The sun4u is the
ultra sparc, the most recent of those you list. I think solaris 8 is
the last that will install on sun4d, and solaris 9 is the last that
will install on sun4m (or do I have that backwards).
Richard L. Hamilton
Originally, I think the difference between sun4, sun4c, sun4d, sun4e,
sun4m, and sun4u corresponded to the bus structure (VME on sun4, Sbus on
sun4c, Sbus+Mbus on sun4m (sun4d similar, but had something else too);
don't know enough about the sun4e to know what it used. There tended to
be some CPU differences implicit as well, but I'm not sure whether that's
coincidence or not. But at sun4u, it just seems to mean an UltraSPARC,
whether with Sbus or PCI bus for peripherals, UPA,
UPA+Gigaplane/GigaplaneXB, or whatever the newer stuff uses.
I think for awhile there was sun4u1 which referred to the E10k with
its UPA+GigaplaneXB; I haven't looked at an E10K running Solaris 8
or later recently, so I don't know if sun4u1 is still used in that way.
There are also sun, sun4, sun4c, sun4m, ..., etc. (but no sun4u,
because that came into existence after the machid(1) stuff was already
declared obsolete) commands, which are all linked together and simply
return truth values (0 return code=true, anything else=false) depending on
the system they're running on (determined by sysinfo with SI_ARCHITECTURE,
SI_MACHINE, or as a last resort, SI_HW_PROVIDER options, plus some special
cases), for use in shell scripts. Those certainly aren't favored anymore;
machid(1) says
The machid family of commands is obsolete. Use uname -p and
uname -m instead.
In fact it exists strictly for compatibility with the SunOS 4.x days.
--
mailto:rlh...@smart.net http://www.smart.net/~rlhamil
Lasik/PRK theme music:
"In the Hall of the Mountain King", from "Peer Gynt"
Tony Walton
That's correct (apart from the SS600 which was sun4m but wouldn't run
anythung later than 2.5.1). sun4c stops at Solaris 7.
IIRC the 'c' means "Campus" - the SS1 was known as the "Campus" - while
the 'm' in sun4m means "multiprocessor". sun4c was exclusively single
processor while sun4m would take more than one. I'm not sure what the
'd' stood for in sun4d (if anything), and (again, IIRC), the 'e' in
sun4e is just "the letter after d".
--
Tony
Richard L. Hamilton
Tony Walton <tony.walton@s_u_n.com> writes:
> Neil W Rickert wrote:
> > Mathon <mat...@yeah.net> writes:
> >
> >
> >> What's the difference among sun4u, sun4m, sun4d etc. uname manual
> >> says(indicates) that this is an obsolete machine hardware
> >> name(class). I also remember that
> >
> >
> > They are different versions of sparc hardware. The sun4u is the
> > ultra sparc, the most recent of those you list. I think solaris 8 is
> > the last that will install on sun4d, and solaris 9 is the last that
> > will install on sun4m (or do I have that backwards).
>
>
> That's correct (apart from the SS600 which was sun4m but wouldn't run
> anythung later than 2.5.1). sun4c stops at Solaris 7.
Wouldn't _officially_. And the SPARCstation Voyager, although sun4m,
stopped at Solaris 7. But it's not hard to hack 'em to bypass the
check (google a bit to find it if curious), although the result will
be limited - power management no longer works on the Voyager, and anything
across the VME bus wouldn't work on the SS6xx.
> IIRC the 'c' means "Campus" - the SS1 was known as the "Campus" - while
> the 'm' in sun4m means "multiprocessor". sun4c was exclusively single
> processor while sun4m would take more than one. I'm not sure what the
> 'd' stood for in sun4d (if anything), and (again, IIRC), the 'e' in
> sun4e is just "the letter after d".
>
That's probably the story behind the particular letters (like hme ==
"Happy Meal Ethernet"), but IIRC the point of the distinction was bus
structure (as well as CPU perhaps, esp. in the case of the sun4u). It may
also correlate with the old way of encoding architecture info into the
high-order byte of the hostid (the specifics of which I'd have to dig
through 2.5.1 include files to find, although the rule was that if the
high-order bit was on, those ways no longer applied).
Tony Walton
> That's probably the story behind the particular letters (like hme ==
> "Happy Meal Ethernet"), but IIRC the point of the distinction was bus
> structure (as well as CPU perhaps, esp. in the case of the sun4u).
Indeed, but I didn't see the point in my repeating information which
you'd already given in such full and helpful detail.
--
Tony
Karl Schendel
Tony Walton <tony.walton@s_u_n.com> wrote:
> Neil W Rickert wrote:
> > Mathon <mat...@yeah.net> writes:
> >
> >
> >> What's the difference among sun4u, sun4m, sun4d etc. uname manual
> >> says(indicates) that this is an obsolete machine hardware
> >> name(class). I also remember that
> ...
> IIRC the 'c' means "Campus" - the SS1 was known as the "Campus" - while
> the 'm' in sun4m means "multiprocessor". sun4c was exclusively single
> processor while sun4m would take more than one. I'm not sure what the
> 'd' stood for in sun4d (if anything), and (again, IIRC), the 'e' in
> sun4e is just "the letter after d".
The sun4d was the SS-2000 and later SS-1000, the D was for Dragon
I believe. Sun4d and sun4m both were sbus+mbus supersparc based,
but the sun4d had an xdbus and the sun4m didn't.
Karl
Thomas Tornblom
The sun4d:s had an XDBus. The Supersparc processors can be run
with either an MBus or XDBus. AFAIK this makes the Hypersparcs
incompatible with sun4d sustems, as they are MBus only.
>
> --
> Tony
>
Thomas (uses up way too many valuable brain cells with useless
knowledge )
Richard L. Hamilton
Richard L. Hamilton
Thanks - I'd forgotten xdbus (although not that there was an
extra layer there). Now ISTR xdbus was almost "packet-switched",
cool if not as fast as the crossbars that came later.
Richard L. Hamilton
Was that also why the extra meg of cache on the SM-82 only was useable on
the sun4d? (I had a 20 with two SM-82's in it that still kicked pretty
good even using only half the cache, although of course one had to watch
the temp, esp. w. two fast-for-the-time drives in there, which was probably
why the SM-82 in the 20 wasn't supported...although maybe not why the
thing finally got cranky a few years later).
Peter Eriksson
>Neil W Rickert wrote:
> > Mathon <mat...@yeah.net> writes:
> >
> >
> >> What's the difference among sun4u, sun4m, sun4d etc. uname manual
> >> says(indicates) that this is an obsolete machine hardware
> >> name(class). I also remember that
> >
> >
> > They are different versions of sparc hardware. The sun4u is the
> > ultra sparc, the most recent of those you list. I think solaris 8 is
> > the last that will install on sun4d, and solaris 9 is the last that
> > will install on sun4m (or do I have that backwards).
>That's correct (apart from the SS600 which was sun4m but wouldn't run
>anythung later than 2.5.1). sun4c stops at Solaris 7.
Actually the Sun4/600 series machines *does* run later versions
of Solaris just fine - it's just that the VME board drivers are missing,
but if you don't need those then it's quite simple to boot a
later Solaris version. Just follow the instructions here:
http://www.ifm.liu.se/~peter/sun -> 2.8-on-600.txt
(That file contains instructions for booting Solaris 8 on a Sun4/600).
Also see the file "patch_ss600_5.8.sh").
- Peter
--
--
Peter Eriksson <pe...@ifm.liu.se> Phone: +46 13 28 2786
Computer Systems Manager/BOFH Cell/GSM: +46 705 18 2786
Physics Department, Linköping University Room: Building F, F203
Luca
messaggio news:10ip8to...@corp.supernews.com...
> In article <cgh5gn$ns8$1...@usenet.cso.niu.edu>,
> Neil W Rickert <ricke...@cs.niu.edu> writes:
>
> I think for awhile there was sun4u1 which referred to the E10k with
> its UPA+GigaplaneXB; I haven't looked at an E10K running Solaris 8
> or later recently, so I don't know if sun4u1 is still used in that way.
A Standard E10K with 5.8 running reports sun4u
Luca
Thomas Tornblom
It was called sun4u1 when it first arrived with Solaris 2.5.1. I don't
remember when it was merged into sun4u.
>
> Luca
>
>
Thomas
Casper H.S. Dik
>"Luca" <em...@byrequest.com> writes:
For a while the code for E10K lived in a different part of the
company (the FPS/Cray/SGI bit we bought) and the changes for it
hadn't been merged in the common code base. That was a bug.
Casper
--
Expressed in this posting are my opinions. They are in no way related
to opinions held by my employer, Sun Microsystems.
Statements on Sun products included here are not gospel and may
be fiction rather than truth.