[sv-cc] SV-CC Meeting minutes for 06/07/2006

Minutes of 06/07/2006 SV-CC Meeting.

ATTENDEES
000000000000000
666666666665555
000000000001111
655443332212110
021212101010302
740629515157096
xxxxxxxxxxxxxxx Charles Dawson
x-x-xxxxxxx-xxx Ankur Kuchlous
xxxxxxxxxxxxxxx Ralph Duncan
x-xxxxxxxxxxxxx Jim Vellenga
xx-xxxxxxxxxxxx Steven Dovich
xx-xx-xxxxxx-xx Andrzej Litwiniuk
xxxxxxxxxxxx-xx Abigail Moorhouse
-xxxxxxxxx-xxxx Michael Rohleder
xxxxxxxxxxx---- Ghassan Khoory
-xxxxxx-------- Chuck Berking
xxxxxxx-------- Bassam Tabbara
xxxxxx--xxx---x Francoise Martinolle
-xxx----------- Amit Kohli
-------------x- Nasim Hussain
--------------x Sachchidananda Patel
x--x--x-------- Stu Sutherland


1.  Reviewed Patent information.

   - Chas reviewed the patent information.


2.  Reviewed minutes of the 05/24/2006 Meeting.

   - Ralph pointed out that section describing "b. Scope manipulating and
     export function execution from within a non context import function.
     This is undefined behavior as per the LRM (as per proposal in Mantis
     item 1456). " is actually not correct.  Will discuss further during
     debate on 1456.  JimV pointed out that the 1456 does not say that.
     Ralph verified that he is correct.  Needs a friendly amendment.
   - Steve/Ralph.  ACCEPTED (unanimous) as modified


3.  Liaisons

   - Francoise to report on PASSED SV-BC items
     SV-BC discussed the PAR around the interactions with other standards such
     as AMS, VHDL, and system-C.  Could these be done through something like an
     API?  SV-BC thinks the PAR can mean a lot of work.  Stu pointed out that
     it is vaguely worded so that we can address things if we decide we want to
     do so.  SV-BC is going to ask for clarification and prioritization.

   - No one reported on other meetings


4.  New business

   - ITC Issue (Item 1456)

     Friendly amendment to F.8 from:
       Similarly, the behavior of exported tasks or functions is undefined
       when they are invoked by a DPI call chain that lacks the context
       characteristic.
     to
       Similarly, the behavior of exported tasks or functions is erroneous
       when they are invoked by a DPI call chain that lacks the context
       characteristic.

     Andrzej pointed out that there is some cost to checking this.  Ralph
     thought it was minimal.  Francoise considered the possibility of
     leaving it as undefined allows us the possibility of supporting it in
     the future.  Ralph agreed.  If we leave it undefined, we could skip the
     check.  Long discussion on the ethics of having something undefined
     versus an error.

     Ralph withdraws his friendly amendment.

     Ralph/JimV.  PASSED (unanimous)

   - Assignment patterns issues

     Postponed until next meeting.

   - JimV brought up a question on size of strings.

     Does anyone have a preference as to whether strings is in bits or
     characters.  Should it be the same as dynamic arrays?  Steve thinks
     that vpiSize should be clear, and may be causing confusion now.  Chas
     disagrees.  It always means bits unless it is on an object that contains
     other objects; then it means the number of elements.

   - Item 1482

     Ralph noted that there is a proposal for Item 1482 which we can put on
     the agenda for next meeting.  Unfortunately, we should wait until 1481
     is resolved in SV-BC.

   - No other issues.

5.  Reviewed of items with proposals.

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

7.  Old Business

   - All to review Item 1488.
     Ralph commented that he made some minor changes to help with issues around
     setjmp and longjump.

Motion to adjourn.  1:02pm.  Steve/JimV.  PASSED (unanimous)

8.  Action items

   - JimV to look into other things which can have size changes as discussed in 0985.
   - Chas to send everyone the frame docs for the diagrams.
   - Chuck to write a document that describes the possible solutions for the
     compatibility issues, and they're pros and cons.
   - Francoise and Bassam to continue work on assignment patterns.

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

   - Item 1488 Address user uncertainty on DPI context behavior
   - Item 1482 State how tag values are obtained in DPI packed union arguments
   - Item 0964 27.31 Assertion -- vpiAssertType should be vpiType

10. Next meeting
   The next SV-CC meeting will be on 06/21/2006.
   The next P1800 meeting will be on 06/19/2006.

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