Subject: Re: Proposed requirements for SV assertion API
From: Alain Raynaud (alain@tensilica.com)
Date: Tue Aug 20 2002 - 15:55:37 PDT
I'd like to follow-up on an issue raised by Francoise during today's
conference call.
The question was whether the API should apply to a simulation model only
or should we support formal tools also.
My basic take on this is that if it's free, let's try to address the
biggest set possible: that would make the life of customers so much
easier if inter-operability is available at the API level.
In particular, are there any reasons why the API for accessing the
assertions should be different during simulation and through a formal
tool? Clearly, there will be some API that would be formal-specific, and
we probably don't want to provide anything in that area, but I feel we
could spend some effort to identify a "core" API that will apply to a
broad set of tools.
My guess is that 80% of what we'll define can have a broad usage,
therefore there is no reason to restrict it to interfacing with a
simulation kernel. Requirements #1, #2, #3, #4, #7 seem broad in nature.
Requirements #6 and #9 seem simulation-specific.
[I have a similar argument for merging tests for the coverage API - see
separate e-mail thread]
Alain Raynaud
Tensilica, Inc.
This archive was generated by hypermail 2b28 : Tue Aug 20 2002 - 15:57:59 PDT