[sv-ac] RE: semicolon problem in 2173

 
Hi John,

I definitely see the problem lucky you caught this.

Another solution to this problem would be to leave property_spec as is
and remove the semicolon after property_spec inside property_declaration
(without property_statement_spec), this would allow however writing
without semicolon at the end of current property_expr inside
property_declaration, what do you think ?


I'm attaching a version which implements your solution (and an earlier
comment by Shalom).

Yaniv


-----Original Message-----
From: John Havlicek [mailto:john.havlicek@freescale.com] 
Sent: Wednesday, March 12, 2008 13:58
To: Fais Yaniv
Cc: sv-ac@eda.org
Subject: semicolon problem in 2173

Hi Yaniv:

I am reviewing 2173 for Champions.

   There are some semicolon problems that need to be fixed.  

   1. When property_spec is used in a concurrent assertion statement,
      no semicolon is required.  This needs to be preserved for
      backward compatibility.  Under this proposal, a semicolon is
      required for the property_statement form "property_expr ;".

   2. In a property declaration, there is an explicit ";" after the
      property_spec.  This proposal will result in two semicolons when
      the property_statement has the form "property_expr ;".

   A possible solution is to have both property_spec and
property_statement_spec,
   where the former is unchanged, i.e.

      property_spec ::= [clocking_event] [disable iff
(expression_or_dist)] property_expr

   and the latter requires a property_statement in place of the
property_expr.
   Then use property_spec in the assertion statements (unchanged) and
use 
   property_statement_spec in the property declaration, getting rid of
the 
   explicit semicolon there.

J.H.


-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.