Newsgroups: comp.parallel.pvm
From: andy@elroy.jpl.nasa.gov (Andy Boden)
Subject: Q: Anything flakey with pvm_catchout?
Organization: Jet Propulsion Laboratory -- California Institute of Technology
Date: 29 Jun 1995 17:05:59 GMT
Message-ID: <3sumho$kjb@elroy.jpl.nasa.gov>

Hi, I've got a quick question about pvm_catchout.  I hope this isn't
in the FAQ...

I'm using pvm3.3.7 and a master/slave code that has been percolating
along for some time.  I've recently tried to add the use of
pvm_catchout to catalog the output of the slaves.  When I first added
the call to pvm_catchout I noticed that the declaration statement was
commented out in the distributed pvm3.h, which of course made me
suspicious.  But I dutifully removed the comment from the header file
and compiled without further incident.

Now I'm occasionally seeing "Bus Error"s (Sparc-eese) and the
attendant crashes in the master code originating in pvm routines when
I try to receive output from slaves.

I have removed the usage of pvm_catchout from my master code, and it
is even as we speak humming along happily (as before).

So the obvious questions are:

1) What is the reason that pvm_catchout is commented out of the pvm3.h
header file in the default 3.3.7 distribution?  (I haven't kept the
older releases around, so I can't easily check when/if the routine
*wasn't* commented out.)

2) Presuming there is a known problem with pvm_catchout, is this the
likely source of the observed bus errors, or should I keep looking?

Thanks for your time and concern in advance.

							- A.B.

-- 


"...it doesn't matter how beautiful your theory is, it doesn't matter
how smart you are -- if it doesn't agree with experiment, it's wrong."

						- R.P. Feynman

