[sv-bc] RE: [sv-ec] open-ended ranges

On the contrary, I think Arturo is looking at an old BNF. In 1800,

open_value_range ::= value_range

value_range ::=
    expression
  | [ expression : expression ]

Note that $ is a legal form of expression:

expression -> primary -> $

So, at least from the pure BNF, [$:$] would be legal.
There is no statement in the LRM which could be interpreted as
explicitly forbidding it, I think.

Shalom

________________________________________
From: Swapnajit Chakraborti [mailto:swapnaj@cadence.com] 
Sent: Wednesday, March 08, 2006 9:52 AM
To: Arturo Salz; Bresticker, Shalom; sv-bc@eda.org; sv-ec@eda.org;
sv-ac@eda.org
Subject: RE: [sv-ec] open-ended ranges 

Arturo,
 
Thanks for the clarification. Actually the LRM I am looking at
seems outdated as it doesn't have the new rule for
open_value_range.
 
Thx,
Swapnajit.

________________________________________
From: Arturo Salz [mailto:Arturo.Salz@synopsys.com] 
Sent: Wednesday, March 08, 2006 1:15 PM
To: Swapnajit Chakraborti; Bresticker, Shalom; sv-bc@eda.org;
sv-ec@eda.org; sv-ac@eda.org
Subject: RE: [sv-ec] open-ended ranges 
Swapnajit,

You found the right reference.

As for your questions:
bins b1[] = {[3:4], $, 9} is definitely not legal.
I believe [$:$] is also illegal, and this is supported by the BNF:

open_value_range ::= 
         expression
       | [ expression : expression ]
       | [ expression : $  ]
       | [$ : expression ]

Otherwise [$:$] becomes a more verbose form for the whole value range,
which is the same as specifying nothing. 

            Arturo

________________________________________
From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org] On Behalf Of
Swapnajit Chakraborti
Sent: Tuesday, March 07, 2006 11:10 PM
To: Bresticker, Shalom; sv-bc@eda.org; sv-ec@eda.org; sv-ac@eda.org
Subject: RE: [sv-ec] open-ended ranges 

When I tried a search I also found '$' references in many places of LRM.
For coverage purpose, I think, 8.19, 18.5 & A.10.20 are sufficient.
In 8.19, I found the following text which tells us how to calculate the
range bounds for $.
"A bound specified by $ shall represent the lowest or highest value for
the type of the expression on the left-hand side."
From A.10.20, it appears that '$' cannot be specified as single value
for bins. For example,
bins b1[] = {[3:4], $, 9} is not legal. Is it correct? Also should we
allow [$:$] for bin ranges?
I see no issue as it is just another way of specifying the complete
value range for a variable.
 
Thx,
Swapnajit.
 

________________________________________
From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org] On Behalf Of
Bresticker, Shalom
Sent: Wednesday, March 08, 2006 11:56 AM
To: sv-bc@eda.org; sv-ec@eda.org; sv-ac@eda.org
Subject: RE: [sv-ec] open-ended ranges 
I think the complete set of sections where $ is described in its
different uses is:

5.14

6.3.2.1

8.19

17.5
17.6
17.7.2

18.4
18.5

22.4

A.2.5
A.2.10
A.8.4
A.10.20
A.10.22

Shalom

________________________________________
From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org] On Behalf Of
Bresticker, Shalom
Sent: Wednesday, March 08, 2006 5:21 AM
To: Swapnajit Chakraborti; sv-ec@eda.org
Subject: RE: [sv-ec] open-ended ranges 

Also see Mantis 966 and 1350.

Shalom

________________________________________
From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org] On Behalf Of
Bresticker, Shalom
Sent: Tuesday, March 07, 2006 10:40 PM
To: Swapnajit Chakraborti; sv-ec@eda.org
Subject: RE: [sv-ec] open-ended ranges 

Well, first it is somewhat described just before 18.5.1.

But see 6.3.2.1 and particularly 8.19.

Shalom

________________________________________
From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org] On Behalf Of
Swapnajit Chakraborti
Sent: Tuesday, March 07, 2006 5:38 PM
To: sv-ec@eda.org
Subject: [sv-ec] open-ended ranges 

It is described in LRM that open-ended ranges can be 
specified for coverbins. But I coudn't find the exact 
semantic of this syntax in the LRM. 
 
For example,
[$:5] => what this means for a coverpoint of 3bit size ?
[5:$] => what this means for a coverpoint of 3bit size ?
 
Does it exist somewhere in the LRM? I have seen
it in queues chapter. But that's not exactly what I
am looking for.
 
Thx,
Swapnaji