-- This message has been scanned for viruses and dangerous content by MailScanner, and is believed to be clean. Champions April 10, 2008 Conference call Thursday 8-10AM PDT Attendees: ---------- 1. * Stu Sutherland 2. * Surrendra Dudani 3. * Brad Pierce 4. - Francoise Martinolle 5. * Shalom Bresticker (late) 6. * John Havlicek 7. * Dave Rich 8. * Neil Korpusik 9. * Karen Pieper 1. Review IEEE patent policy http://standards.ieee.org/board/pat/pat-slideset.ppt Move: Dave - assume that the patent policy was read Second: Brad Passed unanimously 2. List of Mantis items that are ready for review: ----------------------------------------------- 2.1 2336 SV-AC VPI part of 1757 - fixed - 2008-03-25: Approved by voice vote 8y/0n/0a. Brad - bug note from svcc saying contingent on 2100 #defines Neil - was under review by svcc when added to agenda Move: Brad - approve the proposal for Mantis item 2336 Second: John Passed unanimously 2.2 2335 SV-AC Minor fix for 1641 - fixed - 2008-03-20: Voice vote 5y/0n/0a Move: Brad - approve the proposal for Mantis item 2335 Second: John Passed unanimously 2.3 2334 SV-BC Champions feedback on 0002106 - no change required - On March 25, 2008 the SV-BC unanimously voted to resolve this issue with no action required. Move: brad - approve the resolution of "no change required" for Mantis item 2334 Second: Dave Passed unanimously 2.4 2304 SV-BC Champions feedback on Mantis item 2106 - fixed - On March 25, 2008 the SV-BC unanimously approved the attached proposal. Move: Brad - approve the proposal for Mantis item 2304 Second: Stu Passed unanimously 2.5 2269 SV-BC 6.24.1: sign/width cast unclear - fixed - On March 25, 2008 the SV-BC unanimously approved the attached proposal. Stu dissented via previous e-mail vote on this proposal but was not in attendance at the meeting of March 25. The SV-BC committee felt the issues raised by Stu are well addressed by the proposal but acknowledge that such specificity may make it more challenging to understand in some cases. - Email from Stu (April 01) - as pointed out by Shalom "I accept the arguments given by Shalom and Steven, and withdraw my objections to the proposal. What I consider most important is the LRM specify a non-ambiguous behavior, which I feel this proposal does." Brad - It may break backward compatibility. There was ambiguity in the LRM - The expression to be cast should be evaluated in self-determined fashion. - Now there is consistency with $signed, $unsigned. Move: Brad - approve the proposal for Mantis item 2269 Second: Stu Passed unanimously 2.6 2262 SV-CC Missing definition in vpi_user.h for vpiIsProtected - fixed - The SV-CC PASSED this item on 03/19/2008 (unanimous). Move: Stu - approve the proposal for Mantis item 2262 Second: Brad Passed unanimously 2.7 2251 SV-BC `include with "filename" - fixed - On March 25, 2008 the SV-BC unanimously approved the attached proposal. Brad - opened another mantis asking for more details. 2339 This proposal just clarifies the language. Stu - the last example - why was the file name changed? Brad - List.vh is in annex H of the lrm. It is capitalized there Dave - things are left up to the tools today. Stu - had a note from merge that some text was to be deleted. - This proposal doesn't really address the issue. Dave - when going to fix the problem it was realized there was a larger problem to be addressed. Brad - what is the original issue? Dave - the LRM usually doesn't get into this level of tool details John - are we expecting any contradictions? Dave - we are defining it to be undefined. Move: Dave - approve the proposal for Mantis item 2251 Second: Brad Opposed: stu - feels that it should be resolved in conjunction with 2339. Wants both to be resolved by svbc Passed with one opposed. 2.8 2235 SV-BC Clarifications needed for "ref" mode actuals - fixed - On March 25, 2008 the SV-BC approved the attached proposal. Two members abstained from the vote: Shalom (the term indexed select is misleading) Francoise (Unsure if this is an improvement) Dave - the whole definition of what is a variable is being brought into question. John - why does Shalom have an issue? Stu - there is no definition of "indexed select" Shalom - intent was to say select of a single item as opposed to a slice (a group of items). It isn't consistent with LRM terminology. Indexed part-select and non-indexed part-select. indexed part-select syntax using +: or -: as opposed to old syntax : and use indexed to refer to the new v2k syntax. In 2235 the author didn't make this distinction. The alternative would be to use a wording that was more accurate but significantly longer. The author agreed that the current wording isn't completely consistent with the rest of the LRM. It was felt important enough to get this passed in time to get it into the lrm. John - are these the only pieces of aggregates that can be bound? Stu - From: a non-sliced indexed select of an unpacked array. To: a "single element of an unpacked array". Shalom - an array is an entire variable. Move: Brad - send the proposal for Mantis item 2235 back to svbc for rewording. Second: Stu Passed unanimously 2.9 2226 SV-CC VPI Handle behavior for HDL objects of dynamic lifetime is not defined - fixed - The SV-CC PASSED this on 03/26/2008 with the following vote: Michael ABSTAINED (did not get to fully review the proposal) Abi, Chuck, Jim, John, Anil, Bassam, Ghassan, Francoise IN FAVOR There were many friendly amendments (see minutes). John to implement and delete old proposals. - Neil sent a request to the TC chair for an explanation in the bug notes. There was a new proposal uploaded after the latest vote in the svcc. - Chas says it IS ready for the Champions. Stu - some of this is confusing to the Editor, there is a high risk of the wrong set of changes being made. Clause 36 looks ok. - The intent was for Stu to get the Frame files back from sv-cc - Stu will most likely miss the fact that there are 4 documents The following set of issues were flagged: - Change the proposal file names, or merge the 4 pdf's. Otherwise the Editor will not recognize there are 4 files to process. - Make it clear the changes are with respect to draft 5 (not draft 4). - 36.2.2 last sentence of first paragraph - mentions deprecated text, This should be a note (not normative text). - Bottom of page 2 of the changes to clause 36, "NOTE-A VPI program that attempts to release an invalid handle is also erroneous" - this appears to be normative text. - The last sentence in the box on page 2 of 37.6, the text should be blue, the text should be removed from 37.6 and added to the deprecation annex c. - Remove the note at the bottom of page 2, section 37.6 - Review the entire set of proposals to ensure notes are being used properly. - vpi_free_obj is used a lot in the LRM. All of those places need to be updated. Is it used in any of the other proposals? They need to be checked. - clause 37 - page 2 adds new text to vpi_free_object section, but it is suppose to be deprecated. Move: John - send the proposal for Mantis item 2226 back to svcc Second: Stu Passed unanimously 2.10 2218 SV-CC Clarify meaning of vpi_compare_objects() - fixed - The SV-CC PASSED this Item on 03/12/2008 (unanimous). Move: Stu - approve the proposal for Mantis item 2218 Second: John Passed unanimously 2.11 2216 SV-CC Details missing from VPI Generates object diagram - fixed - The SV-CC PASSED this item on 03/12/2008 (unanimous). Stu - does an unnamed generate create a scope? - has a concern that this is changing generates. Move: Dave - approve the proposal for Mantis item 2216 Second: Stu Passed unanimously 2.12 2143 SV-CC How should VPI handle index expressions with side effects? - fixed - The SV-CC PASSED the "conservative" proposal on 03/26/2008 (unanimous) Jim Vellenga to remove other proposals. Stu - which proposal? Neil - see the bug note. (conservative) AI/Neil - send email to Chas to delete the other proposal. - needs to be done before next WG meeting. Move: Dave - approve the conservative proposal for Mantis item 2143 Second: Stu Passed unanimously 2.13 2094 SV-CC Need to extend VPI to handle class specializations - fixed - The SV-CC PASSED this on 03/26/2008 (unanimous) Brad - page 5, last sentence - with the body --> within the body Friendly amendment - to change this text (with --> within) AI/Chas - to fix before next WG meeting. Move: Dave - approve the proposal for Mantis item 2094 with a friendly amendment Second: Brad Passed unanimously 2.14 2054 SV-CC Draft 3a, Section 40 - Deprecate Data Read API - fixed - The SV-CC PASSED this on 03/26/2008 Vote as follows: Michael, Francoise, Ghassan OPPOSED Bassam ABSTAINED Abi, John, Chuck, Jim, Steve, Anil IN FAVOR Stu - need to update annex C - clause 40 and Annex N now deprecated. Friendly amendment on annex c change. Move: Dave - approve the proposal for Mantis item 2054 with friendly amendment. Second: Stu Abstain: Brad - doesn't know enough about it, appears to be a political decision. Passed with one Abstain 2.15 2008 SV-BC Glitch problem in unique/priority if/case - fixed - The latest proposal was approved by the SV-BC on March 17, 2008. The vote was not unanimous: Opposed: Stu (ambiguous severity level will lead to implementation differences that will be problematic for users) Abstain: Heath (agrees with Stu but no enough to oppose) Alex (joined the conversation late) Stu - Has an objection - In his training classes - talke about unique and priority. Users want these to be errors and not warnings. This proposal leaves it up to the tool to decide. It could lead to divergent behavior from different implementations. Agrees with the underlying content but against it for this reason. Move: Brad - approve the proposal for Mantis item 2008 Second: Shalom Opposed: Stu - for the reasons noted above Passed with one Opposed. Stu - There is a note about other proposals being merged. Shalom - Those others are assumed to be implemented first. This then goes on top. AI/??? - Update the proposal with a better note. 2.16 1932 SV-AC Introduce LTL and other temporal operators - fixed - was sent back by the champions (email vote of Feb 23) - The TC updated the proposal to address the Champions feedback - 2008-03-20: Voice vote 5y/0n/0a - Still has issues with keywords??? (changed "next" to "nexttime") John - There was a concern raised about the keyword next, so it was changed - The sv-ac has been through the keyword issue a couple of times. Alternatives seem to make the keywords too long. - 'until' may have also been pointed out as an issue. The svac had a strong opinion of keeping this keyword. Stu - As a user he is also concerned about some of the other new keywords. The issue was raised a couple of months ago. One vendor scanned for them. If no other vendors were interested we should go with what is in the proposal. John - See Table 16.25 p7, and, or, iff, within They should all begin with lower-case letters. AI/Neil - add a note to Editor on these capitalization problems. John - There are 2 documents - formal and 1932 Stu - The special symbols in Annex F are problematic. Shalom - The symbol font or the equation tools in Frame may work for those. Stu - Was unable to get either to do it for him. Today he is using little graphics for those special symbols. John - Send some source for that annex to John for him to look at it. Dave - Has anyone looked at the grammar in any level of detail? - Someone on the svac has the expertise? - For accellera we had someone run it through a parser to confirm. Brad - There is a lot of ambiguity here. This doesn't make it worse. John - Did not go through an automatic parser. Was reviewed by people to ensure that the syntax was natural. Stu - Is there anything in here that might be affected by checkers? John - No, this can be added even if checkers aren't added. - Set of operators aligned with temporal logic used for a lot of years. It is well-aligned with PSL. - There is one backward compatibility issue. Interpretation of a sequential property e (weak ---> strong). The svac was unanimous on this point. This would only affect formal tools. Now added syntax required to make something strong. Dave - Where does it say what a simulator will do with eventually? John - 16.12.13 describes it Dave - It appears that there is a need for some end of simulation computation. John - A similar problem can occur in the current LRM with sequences promoted to a property. Action blocks - similar problem. Neil - Should this be an issue for the svsc? Dave - Thinks enough time wasn't spent on this. Brad - The svsc work is independent of this issue. John - Users haven't been worried about the action blocks at the end. Tools have added abilities to deal with this. Stu - Has seen some of this in some tools. John - There are 2 cases - it is test dependent a. long enough for all pending to complete - should fail b. if stopped can be a lot of pending - tool can be told not to report on those. Stu - iff areas - inconsistent use of commas. Thinks more commas are needed. AI/Dave - merge the two pdf files. 080317 (two pdf files) Friendly amendments Shalom - LTL Page 6, extra comma on page 6 (blue text) Shalom - There are about 10 instances of const_expression. They should probably all be constant_expression. See example on page 11, ltl doc, nexttime Editor issue - fuscia on page 16 - no note. fuscia is blue in 1757 Move: Brad - Approve the proposal for Mantis item 1932, change const_expression to constant_expression. - add to bug note Second: John Opposed: Dave - for the reasons given above - needs more thorough review by implementors Passed with one Opposed <we ran out of time and did not review the following> 2.17 1757 SV-AC Property resets: accepton(b) P, rejecton(b) P - fixed - Was sent back to the svcc by the champions (December 20) - Extracted the vpi portion out into mantis 2336 - 2008-03-25: approved by voice vote 8y/0n/0a. - I did not explicitly check to ensure that the Champions feedback was addressed (Neil). 2.18 1688 SV-CC Performance of VPI access to memories and MDAs is inadequate - fixed - The latest proposal was not approved by the sv-cc. I sent a note to the Chair, requesting input. - The latest proposal will be reviewed by svcc in Wed 4/9 meeting (Chas) 2.19 1465 SV-BC 19.8: port declarations without directions - clarification - fixed - On March 25, 2008 the SV-BC unanimously approved the attached proposal. 3. Next meeting April 24 - Thursday Champions May 01 - Thursday Working Group meetingReceived on Mon Apr 21 18:56:33 2008
This archive was generated by hypermail 2.1.8 : Mon Apr 21 2008 - 18:56:36 PDT