Re: [sv-bc] E-mail Ballot: Respond by 8AM PST, Mon, Dec 10, 2007

From: Gordon Vreugdenhil <gordonv_at_.....>
Date: Tue Dec 04 2007 - 16:30:05 PST
Steven Sharp wrote:
> 
>>> SVDB 2225 ___Yes   _X_No  
>>> http://www.eda.org/svdb/view.php?id=2225
> 
> There are a number of other problems in this section that should be
> fixed, but I will confine my comments to the text being changed.
> 
> In the first text change, the name of a parent module declaration is
> allowed instead of its instance name.  The same would presumably apply
> to a program or interface name.  However, this was not mentioned
> in this part of the text before, so maybe it is a separate issue.
> 
> In the second text change, the revised text would read
> 
> "If not found and the current scope is not the module scope, look for
> the name in the enclosing scope, repeating as necessary until the name
> is found, or the compilation unit scope is reached."
> 
> Notice the "and the current scope is not the module scope".  That would
> imply that we don't look in the compilation unit scope if the current
> scope is the module scope.  If it is assumed to apply to each repetition,
> it would prevent us from ever looking in the compilation unit scope.


Well, as is that is true.  With the 1809 rules it does not.  The
compilation unit searching for task/func names in an upwards
resolution is explicitly discussed there in terms of how such
names interact with the basic algorithm.

> This was basically a combination of a bottom-testing loop with an
> if-statement, to make it act like a while loop.  That has a problem
> if you change the condition on one without the other.
> 
> The simplest fix is to strike the "and the current scope is not the module 
> scope".  We don't need to prevent a look upward starting from the module
> scope any more, since we will always be going up at least once before we
> reach the compilation unit scope.

I didn't do that specifically due to the 1809 interaction.

> In the final text added to the end of item b), it should not say
> "design unit scope is reached".  It should say "module scope is reached".
> I think we have agreed that upward hierarchical searches do not look in
> the compilation unit scope of parent instances, only in the compilation
> unit scope of the reference.

"design unit" here covers interfaces, etc. as well as modules.

The "compilation unit" is not a design unit so it is specifically
excluded by that term.

Gord.

-- 
--------------------------------------------------------------------
Gordon Vreugdenhil                                503-685-0808
Model Technology (Mentor Graphics)                gordonv@model.com


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.
Received on Tue Dec 4 16:30:22 2007

This archive was generated by hypermail 2.1.8 : Tue Dec 04 2007 - 16:30:36 PST