Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 10 Aug 1998 22:47:37 -0400
From:      Matthew Patton <patton@sysnet.net>
To:        freebsd-smp@FreeBSD.ORG
Subject:   hanging second processor
Message-ID:  <l03110719b1f55f1e7876@[192.168.1.10]>

next in thread | raw e-mail | index | archive | help
I'm running -current as of 2 days ago. The box is a Dual PPRO 240 (180's
overclocked). I don't know where to lay the blame on this one yet but here
goes.

There is no user load but I'm running Distributed.net's rc5des crackers on
the machine. It may take 5 minutes or 3 days but I eventually "loose" the
second CPU. The rc5des process that was apparently running on it hangs, and
I can't kill it no matter what I do. A shutdown (clean) and a hardware
reset is needed to unwedge the processor. The main processor runs just
peachy 100% of the time.

Is the problem buggy rc5des code? Or some sort of race condition or bug in
the FreeBSD scheduler? Apparently the program doesn't fork or thread itself
to take advantage of 2 processors otherwise I would have specified
numcpu=2. If I run 1 instance, it gives me 690kkeys/sec (the expected rate
for a single processor). If I run 2 instances I get 690kkyes/sec/processor
which leads me to believe each CPU is getting an instance. Anyone have an
idea as to the problem? Are there some SMP aware programs that I can run
for days that will hammer the CPU's?

--------
"You need only reflect that one of the best ways to get yourself a
 reputation as a dangerous citizen these days is to go around repeating
 the very phrases which our founding fathers used in their struggle for
 independence,"  - Charles A. Beard (American historian)



To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-smp" in the body of the message



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?l03110719b1f55f1e7876>