Re: [sv-bc] uwire & wire -vs- reg

One of the areas that should likely be explicitly considered
here is the impact of forces on a "collapsed var port".  The
LRM suggests/states that net collapsing is not required but is
really only an optimization.  In the presence of forces, I don't
believe that is really true since the implied continuous assignment
semantics that one otherwise uses would not propagate strength.
This can lead to different simulation results.  If we want to
define/allow some form of collapsing for var ports, we should
probably define how forces are supposed to behave (particularly
since a force is considered neither procedural nor continuous in
terms of the writer rules).

Gord.

Steven Sharp wrote:
>> From: "Rich, Dave" <Dave_Rich@mentor.com>
> 
>> I don't see how this logic optimization is any different than what one
>> would use to perform port net collapsing. Any global optimization that
>> normally looks for the absence of forces, hierarchical references, or
>> PLI write access could perform the same port collapsing on logic types
>> in those cases because there is no difference between a wire and a var
>> kind.
> 
> I agree that you may be able to recover some of the performance loss
> by collapsing var ports in some cases also.  As you mention, this would
> require analysis to determine the cases where this optimization was
> valid, i.e. that it produces behavior that would be valid for the
> design without port collapsing (I don't say equivalent, since it might
> change the original behavior to some other valid behavior, perhaps with
> different event ordering).
> 
> For nets, port collapsing is explicitly allowed, whether it changes
> the visible behavior or not.  So net ports can legally be collapsed in
> cases where var ports cannot.  This means that nets still have a
> performance advantage, though I don't know how much.
> 
> 
> Steven Sharp
> sharp@cadence.com
> 
> 

-- 
--------------------------------------------------------------------
Gordon Vreugdenhil                                503-685-0808
Model Technology (Mentor Graphics)                gordonv@model.com


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