Subject: SystemVerilog 22nd Committee Meeting 4/8/02
From: Dave Kelf (davek@co-design.com)
Date: Mon Apr 08 2002 - 11:48:13 PDT
SystemVerilog 22nd Committee Meeting Minutes
April 8th, 2002
Teleconference
Lead Vassilios, Minutes Dave Kelf
Attendance
(aaaaaaaaaaaa-aaaaaaaaaa) Vassilios Gerousis *
(aa-aaaaraaa-aaarar-aaaa) Dave Kelf *
(-------------aaa--aaaa=) John Sanguinetti
(aaaaaaaaaaaaaa--a-aa---) Dennis Brophy *
(aaaaa-aaaaaaaaaaaaaaaaa) Stu Sutherland *
(--------a--------aaaaa=) David Knapp
(aaaaaaa--aaaaaraaar-aaa) Tom Fitzpatrick *
(------a-arraa-aaaaaaa=) Phil Moorby
(aaaa--aaa-aaaa-aaaaaaaa) Anders Nordstrom *
(aaa---aaaaaa--a-aaaaaaa) Cliff Cummings *
(a--aaaaaaaaaraaa-aaaaaa) Simon Davidmann *
(aaaaaaaaaaaaaaaaaaa====) Peter Flake *
(aaa-aa--aaaa-aaaaa-aaaa) Stefen Boyd *
(aa----aaaaaaaaaaaaaaaaa) David Smith *
(a-aa--aaa-aaa--a-a--aa=) Mike McNamara *
(aaaaaaaaaaaaaaaa=======) Kevin Cameron *
(aa-aaaaaa-a-aa-a=======) Andy Tsay *
(aaaaaaaa-aa============) Alec Stanculescu *
(a-aa-----a=============) Adam Krolnik *
(aaaa--aaa==============) Paul Graham *
(--------a==============) David Seifert
(---aa==================) Alex Zamfirescu
Please note - if you join the meeting late - please identify yourself so we
can keep this list up to date and the voting rights accurate
Agenda:
BNF Discussion and Stefen's notes
Packed Arrays
Deprecation
.* / .name proposal review
FSM and State keyword
1. Stefen BNF Queries
Stefen had some questions contained in his note on the BNF. All resolved.
- ACTION: Peter to write a sentence on a minor change from Verilog2K1 BNF
on the fact that tasks and functions can now be empty.
Question on whether jumper ports should be allowed with the newer ANSI
syntax. Consensus was to remove from 3.0 - possibility of revisit in 3.1.
2. Paul's Packed Array Proposal
Discussed contents of proposal contained in email from Paul entitled
Proposal for Array Types, sent on 4/3/02
Vote:
Proposed by Paul, Seconded by Peter
Kevin Against
Rest For
Issue of whether we should include packed arrays.
Mac: If we accept this proposal for general arrays, do we still need packed
arrays.
Peter - yes - relates to data storage, can be treated as vector.
Vote:
Proposal by Stu to replace multi-dimensional packed arrays and retain only
single dimensional (i.e. what we have now). Theory is that all
functionality can be performed with packed structures
Seconded Kevin
Roll Call Vote:
Vassilios Gerousis Casting Vote (Against)
Dave Kelf Against
Stu Sutherland For
Tom Fitzpatrick Against
Anders Nordstrom Abstain
Cliff Cummings Against
Peter Flake Against
David Smith Against
Mike McNamara For
Kevin Cameron For
Andy Tsay For
Alec Stanculescu Abstain
Adam Krolnik Abstain
Paul Graham For
6 Against
5 For
3 Abstain
Packed Arrays included, as they are in current draft.
Minor issues brought up with proposal:
Stu: some confusion of the use of the keywords integer and bit vector.
Suggest use of vector
"Slice" of a vector preferred to other possible terms like "part select".
Stu: Definition should be added
ACTION: Paul to send Stu example on how to handle array concatenation
$length instead of $size suggested - sounds OK to everyone
Question of the value to be returned when an error input is provided to a
function like $left. Should return X. Generally decided after discussion.
$increment sounds better than low, high
3. FSM Vote last week
Result of email vote was to defer FSM discussion until 3.1
State keyword elimination vote
Abstain: Tom, Paul, Kevin, DaveK
For: All others
The "State" keyword will not be included in the reserved keyword list.
4 .name / .*
Proposal now submitted by Cliff capturing previous discussions.
Port connection rules. .name should follow the general
Verilog/SystemVerilog rules for port connections. Remove redundant 3rd
bullet in that section.
Vote:
Propose unconnected ports do not have to be explicitly listed, as with
current port connection rules
Second Cliff
Against: Adam
For: Everyone else
Note that everyone thought that .* did need to have unconnected ports
explicitly named.
Cliff will redo proposal and we can have an email vote on this.
ACTION: Cliff to email proposal for vote
5. Deprecating Items
Cliff provided proposed wording this morning after last weeks discussion.
We need to have an email vote on this. ACTION: DaveK will organize.
Note from Vassilios: Only document review issues from now on, no more major
changes.
Next Meeting
Monday April 15th, 9:00am PST
Call In Number 405 244 5555 access code 3715
Stu will release next draft of spec including BNF for Friday. We will
review next meeting.
______________________________
Dave Kelf
VP Marketing
Co-Design Automation, Inc.
Tel: 1 877 6 CODESIGN ext 404
Mobile: 1 617 571 9883
Fax: 1 781 662 2281
Email: davek@co-design.com
Web: www.co-design.com
www.superlog.org
Latest News:
http://www.co-design.com/news/index.htm
"Faster, Smarter Verilog"
______________________________
This archive was generated by hypermail 2b28 : Mon Apr 08 2002 - 11:50:04 PDT