Subject: [sv-cc] sv-cc Meeting minutes 02/04/2003
From: Joao Geada (Joao.Geada@synopsys.com)
Date: Tue Feb 04 2003 - 10:01:01 PST
sv-cc minutes, Feb 04, 2003
- Joao proposed, Bassan seconded acceptance of last week's minutes
- Joao & Bassam representing sv-cc at the scheduling working group
Both confirmed acceptance
- Joe Daniels has started the LRM writing process
- Andrzej has sent out ascii text for the C side of DirectC LRM
Merges together all the C side pieces of DirectC
Discussion on some corrections in the C interface related to handling
of packed/unpacked arrays, dealing with "unused" bits, dealing with
"endianess" and "signedness" of types across SV<->foreign language boundary
Discussion proceeded related to C representation of SV data, specifically
when SV values do not fit neatly into C types (eg 10 bit value)
Suggestion that DirectC interface extended with convinience function(s)
to remove "unused" bits and correctly extending "sign" bits as necessary.
Discussion got to establishing that this is specifically on how to interpret
bit patterns BUT that the bit pattern will not change across the boundary.
Further details to be discussed on mailing list
Next discussion on bit ordering representation. Andrzej proposes that
C layout always have high order bit in same place, regardless of how
those bits are numbered on the SV side
ie regardless whether SV is a[7:0] or a[0:7], on C side the MSB is always
bit 7 and LSB is bit 0
Francoise and Andrzej explored the issues with various representation methods
Discussion to continue to mailing list
Next overview of section 3.4, describing how inout and output values modified
by DirectC are "noticed" by the simulator. Kevin continues to have issues
with this approachl; Andrzej and Francoise agreed on approach
Section 3.5: Semantics of "context" calls. Without context, using
other APIs is specified as "undefined behavior". Michael would
prefer stronger wording, stating that you *must not* use other APIs
or that such usage is an *error*
Kevin continues to have major concerns over this behavior and will raise
more detailed issues on the mailing list.
Section 3.6; pure functions
Michael et all raised the point that it is still not explicitly stated that
pure functions may not have inout and output arguments.
Minor clarification to explicitly note that pure functions can have return
values.
Section 3.7; memory ownership rules
short discussion, not issues
- Joe Daniels made points it is better to explicitly clarify and even have
redudant text in several places to ensure that users understand the
standard and that implementations are compatible
- Swapnajit asked Joao on status of VPI entity relation diagrams. Asked if
these can be given to Joe this week. Joao confirmed.
Joao asked wether to have one set of diagrams merging assertion&coverage
or to have separate sets for each. All decided that it would be preferable
to have separate sets of diagrams
- discussion re string requirements/proposal with sv-ec
Joao stated that sv-ec proposal is fine
Swapnajit to contact Joao offline to finalize this
- Francoise asked if sv-ac has confirmed that assertion API access is OK
Swapnajit stated that the API was given to sv-ac and that no comments
have been received.
Francoise requested that we get explicit confirmation on this API
- Francoise asked if the issue list on the web is updated
Swapnajit stated that this is *not* updated and needs to be corrected.
- Kevin asked if scheduling working group has a reflector.
Swapnajit stated that this is still in nascent state and that he has no
further information. Group being led by David Smith (sv-ec chair)
Reason for this group is that scheduling affects all committees
(bc, ac, ec, cc); Francoise asked if this group will get updates on
progress. Yes.
- discussion on what will be the agenda for tomorrow's meeting
Attendance:
============
Swapnajit Mittra
Joao Geada
Ghassan Khoory
Andrzej Litwiniuk
Francoise Martinole
Kevin Cameron
Michael Rohleder
Bassam Tabbara
Joe Daniels
John Stickely
John Ameroux
Doug Warmke
Stuart Swan
==============================================================================
Joao Geada, PhD Principal Engineer Verif Tech Group
Synopsys, Inc TEL: (508) 263-8083
344 Simarano Drive, Suite 300, FAX: (508) 263-8069
Marlboro, MA 01752, USA
==============================================================================
This archive was generated by hypermail 2b28 : Tue Feb 04 2003 - 10:02:27 PST