From: Gordon Vreugdenhil (gvreugde@synopsys.com)
Date: Wed Jan 15 2003 - 08:25:58 PST
Precedence: bulk
Shalom Bresticker wrote:
>
> Precedence: bulk
>
> Question:
> With the addition in 1364-2001 of multi-dimensional nets and array
> element
> bit-selects and part-selects, what was the intention to allow on the
> left-hand side of continuous assignments?
>
> The BNF of the LHS is net_lvalue, which allows an arbitrary array
> element constant bit-select or part-select.
>
> Was that the intention?
I can't really address the original intent, but I would certainly
consider a multi-dimensional net with all constant indices to be a
structural net expression. If not, then you couldn't use them in
a port connect either which is "clearly" against the expected use
with generate. Since the LHS of a continuous assign just needs to
be a structural net expression, I consider LHS constant-indexed
multidimensional net selects to be valid. BTW, this also includes
fully constant indexed part selects (I lump "indexed part select"
as a form of "part select" but didn't know if you were assuming that
as well in your comments above).
Gord.
> While we discussed this with respect to procedural assignments
> (e.g., http://www.boydtechinc.com/btf/archive/btf_2002/1815.html),
> I don't know what was the intent with respect to continuous assignments.
>
> Anyone remember?
>
> --
> Shalom Bresticker Shalom.Bresticker@motorola.com
> Design & Reuse Methodology Tel: +972 9 9522268
> Motorola Semiconductor Israel, Ltd. Fax: +972 9 9522890
> POB 2208, Herzlia 46120, ISRAEL Cell: +972 50 441478
>
> "The devil is in the details."
-- ---------------------------------------------------------------------- Gord Vreugdenhil gvreugde@synopsys.com Staff Engineer, VCS (Verification Tech. Group) (503) 547-6054 Synopsys Inc., Beaverton OR
This archive was generated by hypermail 2.1.4
: Wed Jan 15 2003 - 08:26:42 PST
and
sponsored by Boyd Technology, Inc.