From: Jason Woolf (jasonw@cadence.com)
Date: Mon Feb 09 2004 - 07:56:23 PST
Shalom,
> Steven,
>
> Looking more closely at the example in 12.8 of the proposal, we see
> that the "defparam m.n.p" is not affecting a parameter outside the
> generate, but at the same level as the defparam itself.
>
> Your formulation would not catch that case.
>
The defparam is not inside a generate, so the rule that we're discussing
doesn't apply to this example. Even if the defparam were inside a generate,
it should be okay to use a hierarchical name that starts at a scope outside the
generate but still refers to a parameter inside the generate.
-Jason
> Shalom
>
>
> On Mon, 2 Feb 2004, Steven Sharp wrote:
>
> > >Are you saying that it is too
> > >complicated, or that a code developer shouldn't have to understand
elaboration
> > >in order to write a model? I would have to disagree with you if that is
the
> > >case. True, it wasn't necessary before, but it's a different world with
> > >generates.
> >
> > I don't think that the developer needs to understand the process that
> > elaboration goes through, with the intermediate steps of resolving
> > parameters. The restriction can be stated in a way that just depends
> > on the resulting hierarchy. No defparams in the hierarchy under a
> > generate can affect a parameter outside that generate. It is a little
> > hard to express the concept in English, but the concept itself should
> > be easy to understand.
>
> --
> Shalom Bresticker Shalom.Bresticker@motorola.com
> Design, Verification & Reuse Methodology Tel: +972 9 9522268
> Motorola Semiconductor Israel, Ltd. Fax: +972 9 9522890
> POB 2208, Herzlia 46120, ISRAEL Cell: +972 50 441478
>
This archive was generated by hypermail 2.1.4
: Mon Feb 09 2004 - 07:42:48 PST
and
sponsored by Boyd Technology, Inc.