> -----Original Message-----
> From: geoffrey.coram@analog.com [mailto:geoffrey.coram@analog.com]
> Sent: Friday, May 14, 2004 10:06 AM
> To: Kevin Cameron
> Cc: VerilogA Device Modeling Reflector
> Subject: Re: Paramset Proposal
>
> If the corner matches but l&w don't, that's not a match.
Yes, but if I have a both set and the paramset only requires
one: that will be a match regardless of a potentially better match.
> All of the specifications have to match.
>
> You do have to be careful in the example on page 19 of paramsets-v4:
> the paramset with only l,w as parameters has to come before the one
> with l,w,ad,as,pd,ps, because an instance with only l,w specified
> could match either. You want it to match the one that computes
> ad,as,pd,ps, but I don't see that we need any rules that say it
> should pick that one even if it's listed later.
Seems error-prone as an approach. I think it will just cause too
many problems in large system designs.
> What we certainly *don't* want is for the simulator to have to
> search through every library file ever mentioned to make sure it
> has the "best" match for each instance.
>
> -Geoffrey
That's why I'm suggesting the procedural approach: no searching
required and predictable results.
If there are still multiple applicable paramsets the search mechanism
should start with the most restricted and work down, e.g. if there is
a paramset with "l&w" and "corner" restrictions it should be matched
before one that is just "l&w" or "corner", and if the choice remains
between "l&w" or "corner" then the user should indicate a preference.
I.e. the simulator should be able to sort out the proper search order
regardless of declaration order.
NB: Stuff entered through GUIs tends not to have much ordering to it.
Kev.
Kevin Cameron, CPU Technology, CA 94588, Tel.: (925) 225 4862
>
>
>
>
> Kevin Cameron wrote:
> > The implication of just picking the first match would be that if you
> > have a paramset that matches for "corner" and a different one
matching
> > for "l & w" and another which matches both (the best match), then
which
> > one you get depends on the order of declaration? I would prefer to
always
> > get the best match.
Received on Fri May 14 10:50:47 2004
This archive was generated by hypermail 2.1.8 : Fri May 14 2004 - 10:50:56 PDT