SV-EC Meeting Minutes 25 November 2002 Active (At one or more of last 4 meetings) (-----aa-aaaa) v Brad Pierce (Synopsys) (a-aaaa-aaa-a) v Cliff Cummings (Sunburst) (aaaaa-aaaa-a) v David Smith (Synopsys) (-aaa-a-aaaa-) v Dennis Brophy (ModelTech) (---aaa-a-aaa) v Francoise Martinolle (Cadence) (aaaaaaaaaaaa) v Karen Pieper (Synopsys) (aaa-aaaaaaaa) v Kevin Cameron (National) (aaaapaaaaaaa) v Mehdi Mohtashemi (Synopsys) (-aaaaaaaa-aa) v Neil Korpusik (Sun) (-aaaaaaa-aaa) v Stefen Boyd (Boyd Technology, Inc.) (--------aa-a) v Stu Sutherland (-----a----aa) n Arturo Salz (Synopsys) (---------aa-) n Don Mills (LCDM Engineering) (-----a-a-a-a) n Tim Corcoran (WHDL) (-aaaaa--a---) n Tom Fitzpatrick (Co-Design) Inactive (Missed last 4 meetings) (---a-aa-----) n Kurt Takara (0-in) (--a--a-a----) n Michael McNamara (Verisity) (--a---------) n Adam Krolnik (LSI Logic) (---a-a------) n Alec Stanculescu (Fintronic) (-aaa--------) n Dave Kelf (Co-Design) (aa-aa-------) n Heath Chambers (HMC) (aaa---------) n Paul Graham (Cadence) (aapaa-------) n Peter Flake (Co-Design) (a-----------) n Roy Armoni (Intel) (aapa-a------) n Simon Davidmann (Co-Design) (aa---a------) n Steven Sharp (Cadence) (-----a------) n Stephen Meier (Synopsys) (-----a------) n Zeev Kirshenbaum (Verisity) a => Attended p => Attended by proxy - => Missed v => Eligible to vote at next meeting (based on attendance only) n => Not eligible to vote at next meeting (based on attendance only) Action Items: 1) David: Put latest Testbench Document on web set. 2) Kevin: propose where alias should be inserted into LRM (including discussion from today). 3) David: send status on issues being moved over to EC from BC 4) ALL: reveiw all proposals on the reflector, channel,references, etc. to be prepared. Stephen Boyd will take notes. 1. Review minutes from the 11 November meeting and approve David proposed, Mehdi second, no discussion, Cliff & Stu abstain (attendance), passed 2. Review Action items (see minutes from 11 November meeting) 3. Mehdi to see when merged can be put on web site At http://www.eda.org/sv-ec/SVTestbench.pdf All action items completed 3. Establish meeting schedule Noted that 1-20 is MLK Day, and 2-17 is also US school holiday Dates and milestones were presented to Accellera board Francoise - Please make face to face on adjacent days Full day possibly on 1-21 or 1-13 (if BC will reconsider face to face meeting) Milestones: 1. December 15th for complete integration into 3.1 LRM (Stu trying for 4th) 2. December 31st integration of all extensions in 3.1 LRM 3. January 15th, 2003 complete review of 3.1 LRM (continue work on integration of other sections into the 3.1 LRM from SV-CC, SV-AC, SV-BC) 4. February 24th, 2003 complete 3.1 LRM (complete at DEVCON) 5. May 1, 2003 SV 3.1 sent to Accellera board 6. June 1, 2003 SV 3.1 standard complete Meetings: 12-9 LRM Review 12-16 LRM Review 1-6 LRM review, Target for Alias and Random Constraints merged in LRM 1-21 Full day person to person meeting 2-3 LRM review 2-17 LRM review complete and available for DEVCON on 24 Feb 3-3 3-17 3-31 Technology freeze - only clarification and editing 4-14 4-28 LRM complete and 3.1 finished at committee - pass to board 4. Review Enhancement status and proposals. a) Alias Proposal Francoise - What about issues with net type (need to define net type resolution). Kevin - Think it's done. Cliff - I don't think jumpered ports are well defined in 1364-2001. Proposal - Net types used in alias must be equivalent. Cliff moves, Stefen second, Kevin abstain, passed b) Random Constraints Mehdi gave overview Stefen - Saw questions about syntax matching SV and AC committee requirements Arturo - Classes and expressions are compatible with SV, added three new operators. Kevin - AMS uses 'from' operator Arturo - Proposal has all been implemented already. Francoise - Only place to specify constraints is in class? Arturo - no, can be passed as argument... but only place for random variables is in class David - Creating noise operator for mixed signal might be possible... but this is specific for AC requirements of stimulus Stefen - What about operator issue in Adam's 11/21 email? Arturo - No problem yet, AC is still working through proposals. We are consistent. Stefen - What about Adam's questions regarding rand vs randc? Arturo - His questions were mostly on difference Stefen - I think the concern was on creating confusingly similar keywords. Issue of keywords is important for getting past 1364. How about his proposed 'rand unique'? Arturo - No precedence for two word types? Stefen - reg signed Tim - What about standardizing the random generation sequence? Arturo - Random number generation isn't hard, but a constraint solver has many ways to implement... and is much harder to standardize. Cliff - How about simple solver that's standard with open door to adding enhanced solvers that are used for all tools. Similar to VCD... users depend on vendor specific VCD, but there still is a standard format. Many saw the random generation and constraints as important to standardize, users want to repeat same sequences no matter what tool. David - Not so hard to standardize, but area of active work for companies. Tim - Likes Cliff's suggestion of standard algorithms with opportunity for vendors to offer enhanced. 5. Review Open Issues from testbench review a) REV-24 - Keyword vs $task Discussion about starting to do context sensitive keywords. David - In some ways classes give us new scope. Stefen - Is it fair to say that using keyword instead of $task is not preferred? David - Interpreted it as reverse... what about new keyword operators 'dist' and 'inside'? Arturo - Well, you could do that with most keywords... you are just moving the problem. David - Will capture discussion and close issue. b) REV-25- new operators instead of new system tasks David - Same issue. David - some issues are still open... will resolve when we review LRM. Will put copy up on site as soon as available from Stu. Kevin - Wants to add issue related to backannotation on interfaces. Needs to be done as enhancement. Karen - It's not any different than any other port connections. With the exception of static tasks and functions, everything is inlined. David - Looks like the problem needs to be clarified and the issue doesn't really look like it belongs on EC (since it's already under discussion in BC). 6. Meeting closed