RE: [sv-bc] fork...join_any with no statements

I agree with Shalom.

I would also suggest that when the number of processes spawned by a fork
join_any statement is less than or equal to one, its semantics be
identical to
to that of a fork/join statement. Regarding this, 1364-2001 says:

	For parallel blocks, the start time is the same for all the
statements, 
	and the finish time is when the last time-ordered statement has
been 
	executed.

A literal reading of the above statement is similarly ambiguous to the
language
describing fork/join_any, where it is understood that the process
continues 
immediately.

	Arturo

-----Original Message-----
From: owner-sv-bc@eda.org [mailto:owner-sv-bc@eda.org] On Behalf Of
Bresticker, Shalom
Sent: Wednesday, March 01, 2006 7:31 PM
To: Steven Sharp; sv-bc@eda.org
Subject: RE: [sv-bc] fork...join_any with no statements

Intuitively, the second alternative is intended.

Shalom

> -----Original Message-----
> From: owner-sv-bc@eda.org [mailto:owner-sv-bc@eda.org] On
> Behalf Of Steven Sharp
> Sent: Thursday, March 02, 2006 5:17 AM
> To: sv-bc@eda.org
> Subject: [sv-bc] fork...join_any with no statements
> 
> What is supposed to happen if you have a fork...join_any that
> contains
> no statements?  A literal reading would indicate that it should
> hang
> forever, since it is never true that "any one of the processes
> spawned
> by this fork completes."
> 
> On the other hand, all of the processes have completed.  Since
> join_any
> is generally supposed to continue before or at the same time
> that join
> would have continued, it seems like it should continue
> immediately.
> 
> Steven Sharp
> sharp@cadence.com