ALF meeting, July 15 2002 @ Mentor Graphics, Wilsonville, Oregon Attendees --------- Dennis Brophy - Mentor Graphics Gregory Hackney - Mentor Graphics Peter Christiansen - Magma (phone) Joe Daniels - at large Bob Digracia - ASC (phone) Alex Zamfirescu - ASC Kevin Grotjohn - LSI Logic Srinivas Madaboosi - Fujitsu Wolfgang Roethig - NEC Balchandra Thatte - Sequence Parimal Zaveri - Sequence Next meeting ------------ Tentatively scheduled for Mon Sept. 16 @ NEC Santa Clara, 9AM - 5PM One hour timeslot will be allocated for ALF tutorial, based on popular demand. Action item review ================== Old Action Items ---------------- A.I. Design and test magazine - Alex Create one page, take picture from ALF presentation Update: Alex to contact Peter Ashenden OPEN A.I. Call Paul Menchini for ballot constituency - Wolfgang OPEN A.I. Sign up country experts for ALF support in IEC meeting - Alex update: IEC meeting happened in April, ALF is proposed as "new work item" ONGOING A.I. Announce VirtualEDACafe tradeshow on ALF and OLA reflector - Wolfgang DONE A.I. round up IEEE members for participation in ballot (at least 3 per company) - all Update: specific assignment: Identify 3 IEEE members from Sequence - Parimal Identify 3 IEEE members from Magma - Peter Identify 3 IEEE members from Fujitsu - Srinivas Madaboosi - Srinivas Identify 3 IEEE members from LSI Logic - Kevin Grotjohn Identify 3 IEEE members from Mentor - Greg Identify 3 more IEEE members from Mentor - Dennis Issue wwith ballot is not negative vote, but absence of response. Therefore we need a critical mass of people who will respond. ONGOING A.I. Explain pitfalls, if INCLUDE statement is not properly used - Alex ASC making assessment through ALF compiler development ONGOING A.I. schedule conf call for restrict_class - Wolfgang Desired participants: Kevin, Bhalachandra, Alex, Tim Ehrler Need to prepare a table comparing the feature sets of the different proposals Update: Resolving the issue in today's face-to-face meeting VOID A.I. find character naming terminology used in unicode set or ISO??? - Alex Update: Reviewing today DONE A.I. Propose a smart way to deal with smart quotes - Alex Update: not an issue, if standard is based on ASCII character set only CLOSED A.I. send proposal for SI-units now for circulation to Paul - Wolfgang "SI" here stands for "Systeme international" (French), not signal integrity. OPEN New Action items ---------------- A.I. find IEEE doc expert - Joe / Alex A.I. fill out the IEC form by August 21 (IEC meeting @ Mentor in San Jose) - Wolfgang A.I. send pointer for IEEE membership enrollment to ALF reflector - Wolfgang Doc review ========== Section 6 --------- 6.1, Table 14: ^ = caret put ASCII code as additional column into table 14 remove commas between (), <> etc. 6.6: change unit-symbol into quantity-symbol Table 21: put x-ref to SI system change "lexical value" into something better (SI-prefix?) 6.12: Keyword identifier: give a reasoning why more restrictive than general identifier Reasoning: encourages use of english (or other natural language) words as keywords rather than kryptic keywords involving numbers, symbols etc. Section 7 --------- 7.3: change "string" into "string value" 7.15: last paragraph still not marked as a comment. Section 8 --------- 8.3: Ned rule for dealing with inheritance of compiler-detectable conflicting information. Alex proposes: hide conflicting information, don't build data structure Kevin proposes: build datastructure, leave resolution up to the application Discussion to be continued. Section 9 --------- 9.5.2: last paragraph: remove items in parenthesis Explain that swap_class and restrict_class are not the only mechanism for cell usage. Default situation: no cell has a restrict_class or swap_class, application figures out everything on its own. 9.5.3: "Design transformation" should be qualified by using cells of a library. Restrict_class is only relevant if the design transformation actually uses a cell in a library. Table 28: Synthesis: "provides" functional equivalence, instead of "preserve" Layout: "create artwork" may be too restrictive, "use for artwork" is better Restrict_class: proposal a) and b) accepted make clear that b) is narrowing down a) Work item review ================ Proposed new feature for std header file: SEMANTICS declaration --------------------------------------------------------------- This feature is used in appendix B, to explain a rule formally instead of using english text. Advantage: configurable parser/rule checker instead of hardcoding a rule based on a document. Alex argues that the SEMENATICS keywword is misused and the rules are actually SYNTAX rules. Joe disagrees with Alex, based on distinction between "syntax" and "semantics" within other standards. Restrict_class and Swap_class ----------------------------- Proposed section 9.5.2 and 9.5.3 accepted, with minor wording changes as noted above. Macro definition for vector expression (new) -------------------------------------------- Proposed on the whiteboard by Wolfgang. Proposal: support the creation of a special identifier as a symbol for a vector expression. Use the # character as 1st character for distinction against regular identifiers. Purpose: compact representation of complex vector expressions, eventually better interporability with high-level event spec such as "Sugar" etc. Kevin: Make this a "define" clause, similar to Verilog. Alex: Stay within the scope of ALF language, but use regular identifier. Discussion to be continued. Other work items ---------------- - New feature for lib xref: ASSOCIATE statement - Amended PURPOSE and USAGE annotation, new annotations - Units for arithmetic models - Alex - Simplification of SHIFT statement (new) - Electrical components in context of CELL (new) Not reviewed explicitely today, but available in work doc. USAGE annotation was reviewed implicitely as it appears in context of restrict_class definition (see Semantics 4).