[sv-ec] RE: Wildcard Bin Expansion - clarification needed

Hi Shalom,

Thanks for taking a look into this. I would request all experts to review this
requirement and let us know their views. We need to close this issue soon
so as to avoid confusions.

Regds,
Swapnajit

________________________________
From: Bresticker, Shalom [mailto:shalom.bresticker@intel.com]
Sent: Thursday, December 23, 2010 3:06 PM
To: Vaibhav Bhutani; 'sv-ec@eda.org'
Cc: Swapnajit Chakraborti
Subject: RE: Wildcard Bin Expansion - clarification needed

I looked at this a little.

It seems to be that the LRM is almost completely silent on how bins are named (deliberately?), even on more basic points.

About the only clue about intention is this paragraph:

"Each automatically created bin will have a name of the form of auto[value] where value is either a single coverage point value or the range of coverage point values included in the bin-in the form low:high. For enumerated types, value is the named constant associated with a particular enumerated value."

So for example, in what format is "value"? Is it necessarily in unsized unbased integer format?

How are transition bins named?

Etc.

I think the entire subject needs to be handled, almost from the very beginning.

And as you point out, the LRM should clarify how many bins are created in wildcard bins.

Regards,
Shalom

From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org] On Behalf Of Vaibhav Bhutani
Sent: Thursday, December 09, 2010 2:18 PM
To: 'sv-ec@eda.org'
Cc: Vaibhav Bhutani; Swapnajit Chakraborti
Subject: [sv-ec] Wildcard Bin Expansion - clarification needed

Hi Shalom and All Coverage Experts,

This is a query regarding the Mantis discussion http://www.eda.org/svdb/view.php?id=1474 on wildcard bins.
Here it is concluded that wildcard vector bins should not expand the wildcard expressions to create separate bins for each expanded value.
Only the wildcard values should be used with bin-names. Hence in the following example,
wildcard bins b1[] = {4'bx110, 4'1x00};
the two bins will be named as b1[4'bx110] and b1[4'1x00].

But in the LRM there is no example/clarification on this for wildcard vector bins. I have quoted following excerpt from LRM,
which, I believe, can make user interpret that the expanded 2-state values of the wildcard vector bins should be used as bin names.
Hence for the above example, there will be four bins which are b1[4'0110], b1[1110], b1[1000], b1[1100].

LRM-2009 19.5.3, Pg 502
"A wildcard bin definition only considers 2-state values; sampled values containing X or Z are excluded"

The above statement clarifies that the sampled value should not have X or Z .
In this case, there is no point in keeping the wildcard bin definition unexpanded because the sampled value is a 2-state value (0 or 1) and not "don't care value".
Hence showing the exact 2-state value with bin name of vector bins makes sense because user knows which exact value matched. The only intention to provide wildcard entries in case
of wildcard bins is to match the wildcards with 0 or 1.

Now, for non-wildcard vector bins, for which === operator is used, it makes more sense to show the wildcard value as bin name. For example,
for "bins b1[] = {4'bx110, 4'1x00};", we can show 2 bins like b1[4'bx110], b1[4'1x00].

Please let us know your comments.

Vaibhav Bhutani | SMTS, Incisive Comprehensive Coverage, Functional Verification | Cadence

P: 91.120.3984000 M: 91.9899811123 F: 91.120.3984203 www.cadence.com<http://www.cadence.com>

--
This message has been scanned for viruses and
dangerous content by MailScanner<http://www.mailscanner.info/>, and is
believed to be clean.
---------------------------------------------------------------------
Intel Israel (74) Limited
This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.