ISSUE 117

Number 117
Category errata
Synopsis 19.1,19.6,19.9: `unconnected_drive and `celldefine
State lrmdraft
Class errata-discuss
Arrival-DateSep 16 2002
Originator Shalom.Bresticker@motorola.com
Release 2001b: 19.1, 19.6, 19.9
Environment
Description
---------- Forwarded message ----------
Date: Mon, 16 Sep 2002 08:27:41 -0700
From: Brad Pierce <Brad.Pierce@synopsys.com>

1) In section 19.9, "`endcelldefine" should be "`nounconnected_drive".

2) Section 19.1, 19.6, and 19.9 lack syntax boxes. This is especially
important for section 19.9 .

3) In section 19.1 is it an error when `celldefine/`endcelldefine
directives do not come in pairs? For `celldefine it is only "advisable"
to bracket with an `endcelldefine. In contrast, 19.9 says of the
`unconnected_drive/`nounconnected_drive directives that "These
directives shall be specified in pairs." Is it an error if an
`endcelldefine directive is issued before a `celldefine directive?


Fix
No syntax boxes will be added.

In section 19.1, REPLACE

"It is advisable to pair each `celldefine with an `endcelldefine."

WITH

"It is advisable to pair each `celldefine with an `endcelldefine,
but it is not required. The latest occurrence of either directive
in the source controls whether modules are tagged as cell modules."

In section 19.9, REPLACE

"These directives shall be specified in pairs, and outside of the
module declarations."

WITH

"It is advisable to pair each `unconnected_drive with a
`nounconnected drive, but it is not required. The latest occurrence
of either directive in the source controls what happens to unconnected
ports. These directives shall be specified outside of the module
declarations."

In section 19.9, REPLACE

"The `resetall directive includes the effect of a `endcelldefine
directive."

WITH

"The `resetall directive includes the effect of a `nounconnected_drive
directive."
Audit-Trail

From: Steven Sharp <sharp@cadence.com>
To: etf-bugs@boyd.com
Cc:
Subject: errata/117: comments
Date: Mon, 10 Mar 2003 22:44:27 -0500 (EST)

I did some investigation of the questions in part 3. Verilog-XL doesn't
care about pairing of the `celldefine/`endcelldefine directives. There
can be multiple `celldefine directives in a row, and multiple `endcelldefines
in a row or before any `celldefines. NC-Verilog is similarly forgiving, and
I checked what it does with the directives. The behavior appears to be that
`celldefine turns on the tagging (whether it was on before or not), and
`endcelldefine turns off the tagging (whether it was on before or not).

Steven Sharp
sharp@cadence.com


From: Steven Sharp <sharp@cadence.com>
To: etf-bugs@boyd.com
Cc:
Subject: errata/117: Partial Proposal
Date: Wed, 25 Jun 2003 18:12:16 -0400 (EDT)

Here are proposals for the descriptions of `celldefine/`endcelldefine and
`unconnected_drive/`nounconnected_drive that match the behavior of Verilog-XL.
I borrowed the wording partly from `default_nettype, for consistency. It
is not a full proposal because it doesn't cover the syntax box issue.

In section 19.1, REPLACE

"It is advisable to pair each `celldefine with an `endcelldefine."

WITH

"It is advisable to pair each `celldefine with an `endcelldefine,
but it is not required. The latest occurrence of either directive
in the source controls whether modules are tagged as cell modules."

In section 19.9, REPLACE

"These directives shall be specified in pairs, and outside of the
module declarations."

WITH

"It is advisable to pair each `unconnected_drive with a
`nounconnected drive, but it is not required. The latest occurrence
of either directive in the source controls what happens to unconnected
ports. These directives shall be specified outside of the module
declarations."

In section 19.9, REPLACE

"The `resetall directive includes the effect of a `endcelldefine
directive."

WITH

"The `resetall directive includes the effect of a `nounconnected_drive
directive."


Steven Sharp
sharp@cadence.com

------------- End Forwarded Message -------------


Steven Sharp
sharp@cadence.com

Unformatted



Hosted by Boyd Technology