Newsgroups: comp.parallel.pvm
From: papadopo@cs.utk.edu (Philip Papadopoulos)
Subject: Re: using pvm_pking for float
Organization: CS Department, University of Tennessee, Knoxville
Date: 15 Mar 1995 08:46:38 -0500

In article <3k604j$6n8@magus.cs.utah.edu> hsong@asylum.cs.utah.edu (Hwal Song) writes:
>would this work?
>
>  float a[10], b;
>
>  ........
>  pvm_pkint(a, 10*sizeof(float)/sizeof(int), 1)
>  pvm_pkint(&b, sizeof(float)/sizeof(int), 1);
>
>
>do you any difficulties in unpacking?
>what would cause any confilicts with implementational definition?
Maybe. Maybe not.  You can do it and PVM won't choke, but your
data may not be what you expect. (Well it won't choke as long
as (sizeof(float) % sizeof(int))  == 0)

PVM works in a heterogeneous setting -- this means
that data types are translated from one machine representation to
another without user intervention.  For integers , one worries about
sign and zero extension and byte order. By picking out the partial bit
pattern of a float and calling it an int, you may get some bizarre 
behaviour.  If you are just trying to move bits (i.e., you don't
want PVM to do translation, you are on a HOMOGENOUS machine) then
use raw data encoding. 
>
>thank you.


-Phil Papadopoulos

