RE: [sv-ec] Embedded covergroups - mantis #1239

Swapnajit,
 
I agree that with the modified syntax, the special case syntax for
assignment
to a 'covergroup_variable_identifier' is no longer needed. The more
general
rule for a 'variable_identifier' is sufficient.
 
I have updated the proposal to delete the special syntax rule for
'covergroup_variable'.
 
Ray


________________________________

	From: Swapnajit Chakraborti [mailto:swapnaj@cadence.com] 
	Sent: Sunday, February 12, 2006 11:58 PM
	To: Ryan, Ray; sv-ec@eda.org
	Subject: RE: [sv-ec] Embedded covergroups - mantis #1239
	
	

	Ray, 

	I have a question regarding the following: 

	>Finally, there is a syntax correction such that the call to 
	>'new' can be ommitted in the declaration of a covergroup 
	>variable for a non-embedded covergroup. I.E. a regular 
	>covergroup variable can be declared without instantiating 
	>(constructing) the covergroup in the variable declaration. 

	The above syntax is already supported by the first option on RHS
of the 
	following rule, unless we say that for declaring covergroup
variables one has to choose 
	the  rule with covergroup prefix only. 

	variable_decl_assignment ::= 
	variable_identifier { variable_dimension } [ = expression ]
<---- This will also enable "cg cg_inst;" kind of decl 
	| .. 
	| [ covergroup_variable_identifier ] = new [ ( list_of_arguments
) ]15  


	I agree what you recommended will rectify the last option of the
above rule 
	to handle "cg cg_inst" kind of scenario. My only point was this
is already 
	handled by an existing option in the rule and we are allowing
redundancy in the BNF. 

	Thx, 
	Swapnajit. 

	>-----Original Message----- 
	>From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org
<mailto:owner-sv-ec@eda.org> ] On 
	>Behalf Of Ryan, Ray 
	>Sent: Saturday, February 11, 2006 4:28 AM 
	>To: sv-ec@eda.org 
	>Subject: [sv-ec] Embedded covergroups - mantis #1239 
	> 
	>Attached is a proposal for Mantis #1239. 
	> 
	>The proposal 'clarifies' that a covergroup declared within a 
	>class is an 'embedded' covergroup. It makes the embedded 
	>covergroup TYPE anonymous. 
	>The embedded covergroup declaration also implicitly creates a 
	>covergroup variable (of the anonymous type). This replaces a 
	>number of special rules regarding implicit variable 
	>declaration and limitations on the usage of an 'embedded' 
	>covergroup type (the restriction are not needed since the type 
	>is not visable). 
	> 
	>There also was a special syntax (and note) that allows the 
	>variable name to be omitted (implicit) in a variable 
	>declaration of an embedded covergroup type. This optional 
	>syntax can be removed since variables of embedded covergroups 
	>are now always implicitly declared - and cannot be explicitly
declared. 
	> 
	>Finally, there is a syntax correction such that the call to 
	>'new' can be ommitted in the declaration of a covergroup 
	>variable for a non-embedded covergroup. I.E. a regular 
	>covergroup variable can be declared without instantiating 
	>(constructing) the covergroup in the variable declaration. 
	> 
	>- Ray 
	>