[sv-cc] SV-CC Meeting minutes for 02/01/2006

Minutes of 02/01/2006 SV-CC Meeting.

ATTENDEES
000000
665555
001111
212110
010302
157096
xxxxxx Charles Dawson
xx-xxx Ankur Kuchlous
xxxxxx Ralph Duncan
xxxxxx Jim Vellenga
xxxxxx Steven Dovich
xxx-xx Andrzej Litwiniuk
xxx-xx Abigail Moorhouse
x-xxxx Michael Rohleder
xx---x Francoise Martinolle
xx---- Ghassan Khoory
----x- Nasim Hussain
-----x Sachchidananda Patel

1.  Reviewed Patent information.

   - Charles Dawson reviewed the patent information.


2.  Reviewed minutes of the 01/18/2006 Meeting.

   - Steve/Ralph  ACCEPTED (unanimous)


3.  Liaisons

   - Chas talked about what happened at P1800 meeting.
     Chas requested the Frame documents for SV-CC sections.


4.  New business

   - Merging the two specifications.

     Discussed issue at length.

     Some questioned the schedule that Stu gave.  Editing part is probably
     okay, but review by committees will be the real work.

     Users and vendors are having to merge the standards mentally anyhow.
     What is worse, vendors may be making incompatible decisions about how
     to do it.

     The end of the balloting period is always hectic, and merging
     then would be infeasible.  Better to do it now.

     When we were making changes to the VPI object models, coordinating
     the changes in the two standards took a lot of extra work.  It would
     have taken us less work if the standards had already been merged.

     We found and fixed many of the inconsistencies between the two specs
     during the review period last time.  Might not be that difficult for
     the SV-CC.  Other committees may have more trouble.

     In general, we expect that merging now will speed fixes and other
     changes later.

     Completing the changes in two months (wall clock) is infeasible,
     although a raw draft is feasible in such a time.

     As a team, we expect that the draft merge will highlight inconsistencies
     and gaps that exist anyhow.  It will be better to merge at the beginning
     than at the end of the cycle so that have time to work through the
     resulting inconsistencies.  But many, if not all, of us do not consider a
     cycle of less than two years to be feasible.

     We believe that the effort is important enough that given a two year
     window, it is worthwhile and should start now.

     Wall clock time is different from effort.  How much other stuff
     will come up from efforts of other committees?

     Motion to direct the chair to report that the position of the
     SV-CC to P1800 is to merge the standards immediately, as long
     as the duration of the applicable PAR is not less than 2 years.
     Abigail/Steve.  PASSED (unanimous)

   - Dynamic Arrays
     Sized vs unsized packed dimensions.  Francoise pointed to
     a definitive section of the LRM that makes it clear that that
     is allowed.  So, do we have to make a change to DPI to deal with
     that?  No, all set.  Question on the reflector generated this debate.
     Andrzej has already responded to it.  This issue is resolved.

   - Comments as pragmas
     Francoise has an update from SV-BC.  We should not have comments
     in the LRM that has semantic meaning.  They will file a mantis
     item asking us to go through coverage section and resolve.  Pragmas
     should probably be replaced with Attributes.

   - DPI and Unions
     Ralph filed mantis Item 1322 on this.  There are substantial
     problems with implementing this.  The practical implications
     were missed when this feature was originally specified.
     Committee should try to look at it for next time.  Ralph thinks we
     should eliminate DPI support for Unions in most cases.  Francoise
     thinks we should poll users.  Ralph thinks no one is trying to use
     them because he would have heard about it by now.

Meeting ended at 1:10pm EST.


5.  Reviewed of items with proposals.

6.  Review SV-CC items with proposals (Straw poll only):

7.  Old Business

8.  Action items

   - JimV to look into other things which can have size changes as discussed in 0985.
   - Steve to help Nasim file an clarification request.
   - Andrzej to file a mantis item that there is no vpiChandleVar object.
   - Chas to send Abigail the frame docs for the diagrams.
   - Doug to put a new file into Mantis with the friendly amendment to 0924.

9.  Items for consideration at the next meeting (they already have proposals):

   - Item 1307 An open array is not like a multidimensional dynamic array
   - Item 1269 Open array output mode semantics
   - Item 0754 vpiConstraintItem relation missing from header file
   - Item 0804 27.55 Generates has bad assertion and typespec enclosures
   - Item 0791 27.43 There is no 'vpiAssignmentOp' in the include file
   - Item 0790 27.40 continue and break should be bold
   - Item 0755 vpiValueRange target should be range rather than distribution

10. Next meeting
   The next SV-CC meeting will be on 02/15/2006 at the regular time.
   The next P1800 meeting is on 02/21/2006.

--
Charles Dawson
Senior Engineering Manager
NC-Verilog Team
Cadence Design Systems, Inc.
270 Billerica Road
Chelmsford, MA  01824
(978) 262 - 6273
chas@cadence.com