RE: [sv-ac] reminder to vote on mantis 1550

Hi John,

I agree with you that we don't have to specify which value will $sampled
function return at time 0 in this proposal: it should return the value
from the preponed region. If there are problems with the variable
initialization definition they should be handled in the clause
describing the simulation semantics. As Dave mentioned, the definitions
should be ideally handled in the same place. We can add a reference to
this clause here.

Thanks,
Dmitry

-----Original Message-----
From: owner-sv-ac@server.eda.org [mailto:owner-sv-ac@server.eda.org] On
Behalf Of John Havlicek
Sent: Friday, January 26, 2007 3:14 AM
To: Eduard.Cerny@synopsys.com
Cc: john.havlicek@freescale.com; Eduard.Cerny@synopsys.com;
sv-ac@server.eda-stds.org
Subject: Re: [sv-ac] reminder to vote on mantis 1550

Hi Ed:

I don't think we are complicating anything.

I think that the preponed value of an expression in the time 0 slot 
is already defined by the scheduling algorithm.

So you should be able to delete this text entirely and just say
that for _every_ time slot, a call to $sampled in that time slot
returns the preponed value of its argument in that time slot.

If you say that in the time 0 slot $sampled follows a different rule,
then that is complicating things.

Try this thought experiment.  Suppose that there is an expression e 
that combines some static variables with declaration assignments.  
Suppose that no assignment in the time 0 slot changes these
variables.  Should a call to $sampled(e) in the time 0 slot yield
a different result than a call to $sampled(e) in the time 1 slot?

My opinion is "no".

J.H.

> X-MimeOLE: Produced By Microsoft Exchange V6.5
> Content-class: urn:content-classes:message
> Date: Thu, 25 Jan 2007 13:43:00 -0800
> Thread-Topic: [sv-ac] reminder to vote on mantis 1550
> Thread-Index: AcdAxdG186prUgy2SBOcNKMORR/kCAAA4F8A
> From: "Eduard Cerny" <Eduard.Cerny@synopsys.com>
> Cc: <sv-ac@eda-stds.org>
> X-OriginalArrivalTime: 25 Jan 2007 21:43:01.0498 (UTC)
FILETIME=[C967DDA0:01C740C9]
> 
> Hi John,
> 
> This is exactly what I wanted to avoid, having to evaluate the
> expression based on the initial assigned and default values. I.e., in
> the case of an expression, just use the default value of its type. But
> if the majority thinks that it is more appropriate to do wht you
> suggest, fine. But we are here dealing with a corner case, because in
> most applications, a disable iff  or $assertoff takes care of time 0.
Is
> it worth complicating it?
> 
> ed
> 

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.