On Sat, 4 Mar 2000, der Mouse wrote:

[ SS1+, ELC fail in the same mysterious way with X and/or a mud ]

# This is noteworthy largely because this behavior was very familiar.
# For a long time - probably over a year, though I'm not sure - I'd been
# plagued by exactly that symptom.  Seeing it here reminded me that it
# had been a while since I saw it, and in fact I cannot recall ever
# seeing it happen on an IPX.
# 
# Thus, I suspect that whatever was killing the mud is also what has been
# killing my X sessions for "a long time"...and it appears it happens on
# at least two and I think three SS1+s, at least three ELCs, and does not
# happen on at least two IPXes.
# 
# Now, of course, the $65,536 question: any guesses at the bug?

It almost seems to be SS1/SS1+-class specific.  I will hazard you will
have this problem on an IPC that someone might be able to lend you
(if i could, I'd do it -- don't have one), and I will bet equal money
that it does NOT happen on the SPARCstation 2.

This to me sounds like a memory-management or caching problem, because
the 2-class had (I believe) different revisions of MMU and/or on-chip
or external cache.

Subject: Hardware diff IPX/SS2?
Oh, and I'd like to ask a question here:  What's the difference between
a SPARCstation 2 and a SPARCstation IPX?  I seem to remember some of my
favourite buzzwords ("cache" and "contexts") being bandied about in
a comparison between the two.  It can't be the on-chip cache, since
a PoweruP goes into either an IPX or a SS2, so that would leave the
number of contexts it can handle (which is done...where?), or the
amount of L2 cache (which is on that chip to the left of the CPU on
an IPX).  This leads me to ask the question, if it's the latter,
is it possible to put more L2 cache into these things?

# 
# 					der Mouse
# 
# 			       mouse@rodents.montreal.qc.ca
# 		     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B
# 


				--*greywolf;
--
BSD: If you look through Windows