Subject: [sv-cc] RE: [sv-ac] Re: assertion/coverage API questions
From: Bassam Tabbara (bassam@novas.com)
Date: Tue Mar 04 2003 - 10:22:12 PST
Hi Adam,
> Hi Joao;
>
> Thanks for the answers...
Yes thanks Joao for the timely response!
> >1. we could not agree on a canonical definition for
> expression coverage.
>
> What about a method to create additional coverage and place
> it into the database? On the market are many tools that deal
> with coverage and collecting information. Yet few have the
> ability to incorporate additional information so as to
> produce an integrated reporting interface. Writing a standard
> for existing things is useful but of more use is something
> that goes further.
>
> Would it be possible to define additional coverage elements?
> Such as ...
>
> >4. No. Currently the assertion coverage "covers" only
> whether a particular
>
> Unfortunate. One can write more information into assertions
> than coverage statements. Maybe 3.2 can address possibilities
> within a covered sequence, values chosen, timing between
> subsequences, etc.
The (assertion/expression coverage) API as is DOES have room (i.e.
placeholder) for additional granularity. In particular in the next rev
of the doc (very soon), we'll clarify the wording that a "sequence" can
be accessed (just like an "assert"). Also the "step" mentioned in the
documentation is the placeholder for adding timing granularity (aka
completion notification) for subsequences within a sequence. For now, we
only list the finest step i.e. clockstep. All other granularities can be
put here using this mechanism (3.2).
As for your example of "value chosen in a set", it is also not too much
of a stretch to think of it in terms of this "stepping" mechanism in
(sub)sequences (where the listener gets info on the current value of the
sequence as it evolves).
Bottom line, the answer is "yes": I think the placeholders are there, at
least Joao and I had some discussion on this, and revamped a bit with
your and Surrendra's feedback. Of course elaborating how this "stepping"
happens (concatenation, repetition in regexp and so on), needs to be
discussed in SV-AC in 3.2 to bind the (stepping) mechanism to the
semantics.
Thx.
-Bassam.
This archive was generated by hypermail 2b28 : Tue Mar 04 2003 - 10:23:34 PST