errata/209: Re: errata/209: hierarchical_identifier

From: Shalom.Bresticker@motorola.com
Date: Tue Oct 28 2003 - 02:30:00 PST

  • Next message: Jayaram Bhasker: "RE: negative interconnect delay"

    The following reply was made to PR errata/209; it has been noted by GNATS.

    From: Shalom.Bresticker@motorola.com
    To: etf-bugs@boyd.com
    Cc:
    Subject: Re: errata/209: hierarchical_identifier
    Date: Tue, 28 Oct 2003 12:28:21 +0200 (IST)

     Hi,
     
     With respect to the disable_statement form in A.6.5 as
     
     "disable" hierarchical_block_identifier ";"
     
     I wrote in the attached mail below that
     
     "in the disable statement, we would write hierarchical_block_identifier
     with an optional index (only one,not multi-dimensional, and with
     a constant value)."
     
     But now I think we do not need to allow indexing of the block identifier
     in the disable statement.
     
     My reasoning is that we cannot create an array of named blocks.
     The closest thing to that is an array of scopes created by a generate-loop.
     But such scopes are not named parallel or sequential blocks.
     Such scopes can themselves contain named blocks, but that would be
     a hierarchical level below that of the generated scope array.
     
     Comments?
     
     Thanks,
     Shalom
     
     
     On Sun, 2 Feb 2003, Shalom Bresticker wrote:
     
    > Date: Sun, 02 Feb 2003 14:51:59 +0200
    > From: Shalom Bresticker <Shalom.Bresticker@motorola.com>
    > To: Dennis Marsa <drm@xilinx.com>
    > Cc: etf@boyd.com
    > Subject: Re: hierarchical_identifier
    >
    > Dennis Marsa wrote:
    >
    > > > > > {identifier ["["unsigned_number"]"] .} identifier
    > >
    > > For the purposes of the BNF, I think this is OK.
    > >
    > > However, looking at this again, this definition requires that a
    > > hierarchical_identifier end with an identifier. That is, it cannot
    > > end with an index operation: a.b[1].
    > >
    > > The current set of rules for hierarchical_identifier seem to allow it
    > > to end with an index operation.
    >
    > You are correct about this.
    >
    > However, my analysis of the BNF shows that the only use of hierarchical_identifier as a scope
    > and not as a net/variable/event is for hierarchical_block_identifier, which itself is only
    > used in one place, in the disable statement.
    >
    > So what I want to do is define hierarchical_identifier like all other identifiers in the
    > BNF (after we approved #200),
    > that is, to end with an identifier name, without indexes. Then, where we need to add optional
    > indexes, we do so according to the need. So in the disable statement, we would write
    > hierarchical_block_identifier with an optional index (only one,not multi-dimensional, and with
    > a constant value).
    >
    > I think this makes the BNF much more consistent and unambiguous.
     
     --
     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
     

    This archive was generated by hypermail 2.1.4 : Tue Oct 28 2003 - 02:30:13 PST and
    sponsored by Boyd Technology, Inc.