Newsgroups: comp.parallel.pvm
From: hzhou@cs.utk.edu (Honbo Zhou)
Subject: Re: Is PVM Thread Safe?
Organization: CS Department, University of Tennessee, Knoxville
Date: 4 Jul 1995 12:43:24 -0400
Message-ID: <3tbr3cINN9a8@duncan.cs.utk.edu>

In article <3tbl9h$fb@NNTP.MsState.Edu> tony@aurora.cs.msstate.edu (Tony Skjellum) writes:
>The PVM API has unsafe aspects to it.  
>
>In pseudo-calls:
>	Initialize a buffer
>	Pack
>	Pack
>	...
>	Pack
>	Send or Mcast
>One needs to know which message a given pack applies to; in the
>thread world, this is not well-defined, unless we restrict ourselves
>to the PVM 2.x model of one message at a time, which is unlikely
>to be your choice.

The change of PVM user interface seems inevitable in the thread world.
The change is to have pvm_initsend() to return a send buffer id, lots
of stuff could be packed into this buffer just like the current PVM does.
 
>How have you modified the API to accomodate thread safety? Are you
>using a thread ID in calls?
>-Tony Skjellum

Yes, check ths LPVM home page at:

http://www.epm.ornl.gov/~zhou/ltpvm/ltpvm.html

There is a sample LPVM program there showing the changes.

A paper describe LPVM will also be found there in a couple of days.

Honbo Zhou 


