SV-XC Committee meeting Monday June 11, 2012 Attendees xx 1. Dave Rich Mentor xx 2. Neil Korpusik Oracle xx 3. Dmitry Korchemny Intel xx 4. Shalom Bresticker Intel xx 5. Matt Maidment Intel -x 6. Stu Sutherland Editor-IEEE xx 7. Francoise Martinolle Cadence xx 8. Surrendra Dudani Synopsys -x 9. Dana Fisman Synopsys -x 10. Ghassan Khoory Synopsys -x 11. Yatin Trivedi Synopsys xx 12. Brad Pierce Synopsys -x 13. Arturo Salz Synopsys xx 14. Ed Cerny Synopsys x- 15. Brandon Tipp Intel x- 16. Steven Sharp Cadence x- 17. Mark Hartoog Synopsys x- 18. Jim Vellenga Cadence x- 19. Chuck Berking Cadence x- 20. Jonathan Bromley Accellera 1. Review IEEE patent policy ------------------------------------------------------ http://standards.ieee.org/board/pat/pat-slideset.ppt The chair brought everyone's attention to the patent policy. Steven, Mark - assume read 2. Approve minutes of May 21st meeting http://www.eda-twiki.org/sv-xc/hm/0193.html Shalom, Brad - Approve the minutes. unanimously approved 3. Review the Agenda and ask if there is any other business. None were added. 4. Editor update, if any. Are they any items that the editor is not able to handle directly with committee approval? Stu wasn't online today. 5. Review the list of Mantis items with no assigned owners. 1. 4131 -> Shalom 2. 4141 -> Dave Shalom - he just filed the item. It wasn't his. - The proposal is straight-forward 3. 4132 -> Jim Vellenga An sv-cc item. Francoise - the sv-cc haven't met since their first meeting. Chuck - someone from Frescale had signed up for 4132 in the sv-cc. Francoise - Michael Rohleder was waiting for the c file, it would be compiled to check for syntax errors. Charles needs to send the file to Michael. Matt - is the only issue to add void to all the required spots? Shalom - it should be simple to locate any other cases Chuck - it is trivial, it is just a matter of getting the .c file. Jim - Michael had the ability to run it through several tools. - Follow up with Charles, Stu, Michael 4. 4130 -> Chuck 5. 3659 -> Swapnajit An issue with wildcard bins Shalom - agrees that there is an issue and it should be dealt with. Francoise - it was filed by someone from Cadence (Swapnajit). - She will contact Swapnajit. 6. 3394 -> Shalom - There is a proposal, but there was never agreement on it - There is a question about legality of an example - This mantis is a duplicate of 2390 (filed May 2008) Steven - Dynamic arrays don't have new listed as a method. Neil - 7.5.1 describes new as being a constructor. Shalom - There is a proposal, the question is whether there is consensus - The example seems to contradict other text in the LRM. - This doesn't appear in the list of "assignment like syntax" Steven - if we make it legal, there are probably other problems that need to be dealt with. Shalom - if we just remove it, we don't need to make other changes. Steven - views it as an incorrect example. - examples help clarify when something is fuzzy. Mark - It sounds like it is an assignment like context. Steven - There is a strict definition of assignment like contexts. Francoise - Was this situation just missed? Matt - it may have been missed. Steven - it doesn't satisfy all of the requirements of an assignment like context. Jonathan - you can sometimes re-create a dynamic array to be larger than it was. Steven - if the LHS is smaller than the RHS you don't throw anything away. - This isn't a normal assignment like context. For a dynamic array, there can be a resizing occurring. - the only way to quickly deal with it is to remove an example. Shalom - the proposal doesn't declare it illegal, but proposes to remove it. Dave - thinks people prefer it to be legal. Mark - why do people think this syntax is unclear? Dave - Steven provide a list of what needs to be addressed if this is made legal? Francoise - allow it, but define it to be equivalent to a series of statements? Jonathan - the sequence of statements suggested by Francoise would conflict with initializing from another dynamic array. - take what is in the parens and copy to an imaginary dynamic array then assign that to the first dynamic array. - there are some other issues that should also be addressed which have nothing to do with new. - It isn't clear how to handle all possible types of assignment patterns. Some situations are currently ambiguous. Steven - could get a larger one as a temporary. Then apply that one using the other existing rules. Dave - it doesn't have to be implemented that way, as long as the end results behaves that way. Steven - a parameter could be involved. that size should be used. Dave - this example might be supported by existing implementations. AI/Jonathan - add a comment to 3394 with his feedback. AI/All - check for existing implementations of this construct. 7. 2182 - 2012 Ballot comment 56: Elaborate VPI diagrams for checkers Dmitry - it should be assigned to someone in the sv-cc Francoise - this will most likely not be able to be done for this PAR. There are other changes that would need to be carefully reviewed. The SV-CC didn't have time to look into it. Chuck - there were concerns about the original proposal. Other changes from 2012 need to be considered along with it. It is a significant amount of work. It will take a comprehensive review cycle. He expects it would take sever months (at least several meetings). They were dealing with broken stuff from previous versions of the standard, so this one didn't get done. Francoise - it can't be done in a couple of weeks. Dave - he would like to see the sv-cc continue to meet. 6. Review and vote on the list of Mantis items with proposals available. Dave - Should we have an email ballot? Neil - That sounds like a good idea. Steven - agrees Shalom - prefers that we have a few ballots instead of one large one. Dave - how about 2 email votes? one due June 18 one due June 25 - the date of our next conference call Dave - anyone who attended either meeting is eligible. - the 2 of 3 rule will kick in after the next meeting. - there are 18 items ready to be voted on. 7. General review of outstanding Mantis items needing further discussion. Dmitry - 3879 has two proposals. - he uploaded a new proposal yesterday that has some additional changes. Dave - Editorially you typically don't repeat a definition. Shalom - this is an exception of that rule. A returned value is usually a 1-bit self-determined value. When discussing a condition which is evaluated to determine what do to, it can be multi-bit. Because of this you can have true with different meanings. Dave - for a logical and 'true' isn't even defined. Shalom - in the sv-bc there was a sub-clause (12.4) where the value of true was defined. Dave - the two proposals need to be merged before we vote on it. Shalom - he will do the merge. Neil - it should be ok to make the same change in more places than what was flagged in the ballot comment. These are also very close to being editorial. 8. Next meeting June 25 (9am PDT) Adjourned 10:16am