Subject: [sv-ec] Minutes of 21 January 2004 meeting
From: David W. Smith (david.smith@synopsys.com)
Date: Wed Jan 21 2004 - 17:29:56 PST
Here are the meeting minutes from today's meeting.
Regards
David
SV-EC Meeting Minutes
21 January 2004 10:00 am. Wednesday
(rrrrrrrrrrxrxrxrrrr)
Voting Members (3/4 or > 75%)
(aaaaaaaaaaaaaaaaaaa) Arturo Salz (Synopsys)
(-aaaaaaaaaaaa-aaaa-) Brad Pierce (Synopsys)
(--aaaa-aaa---a-aaa-) Cliff Cummings (IEEE 1364)
(aaaaa-aaaa-aaaaaaa-) Dave Rich (Synopsys)
(aaaaaaaaaaaaaaaaaaa) David Smith (Synopsys)
(-aaa-aaa-a-aap-p-aa) Dennis Brophy (ModelTech)
(aaaaapaaaaaa-aaaaa-) Jay Lawrence (Cadence)
(aaa-aaaaaaaaaaaaaaa) Michael Burns (Motorola)
(-aaaaaaaaaaaaaaaaaa) Mehdi Mohtashemi (Synopsys)
(aa-aaaaaaaaaaaaaaaa) Neil Korpusik (Sun)
(--aaaaaaaaaaaaa--aa) Ray Ryan (ModelTech)
|||||||||||||||||||_ 21 January
||||||||||||||||||__ 5 January
|||||||||||||||||___ 15 December
||||||||||||||||____ 8 December
|||||||||||||||_____ 1 December
||||||||||||||______ 24 November
|||||||||||||_______ 17 November
||||||||||||________ 11 November
|||||||||||_________ 3 November
||||||||||__________ 27 October
|||||||||___________ 20 October
||||||||____________ 13 October
|||||||_____________ 29 September
||||||______________ 15 September
|||||_______________ 2 September
||||________________ 18 Aug
|||_________________ 4 Aug
||__________________ 21 July
|___________________ 7 July
Non-Voting Members (attendance based)
(------a------------) Chris Spear (Synopsys)
(-------------s-s---) Doug Warmke (ModelTech)
(-----s-------------) Francoise Martinolle (Cadence)
(--a-aaa-a----------) Jeff Freedman (ModelTech)
(-----------a-------) Peter Flake
(---------------a---) Ron Goodstein (First Shot Logic Simulation and Design)
(---a-----------aa--) Stefen Boyd (IEEE 1364)
(-a---a----------aa-) Stu Sutherland (IEEE 1364)
Guests (non-voting)
(--a-a-a----------a-) Don Mills (LCDM Engineering)
(-----a-------------) James Young (HP)
(-a-----------------) Kevin Cameron (National)
r => Regular meeting
x => Extra meeting (Presence counts for attendance, absence does not)
a => Attended
p => Attended by proxy
s => Attended as proxy
- => Missed
Action Items:
[identified with AI (#) in this text, # refers to AI number]
Added this week (please see the site for existing action items):
Minutes 1/21/04 taken by Mehdi Mohtashemi
1. Review of the meeting minutes
http://www.eda.org/sv-ec/Minutes/SV-EC-Minutes-2004-January-5.txt
Motion: Accept Minutes of 5 January
Moved: Mehdi
Second: Arturo
Abstain: Neil (did not read)
Opposed: None
Passed
2. Review of open Action Items
All closed
David: Any other discussions on reviewing assume proposal.
Michael: No other points with the proposal.
3. Review of Inter-committee dependencies
SV-BC passed Issue 117 to us - ERR-70
Chairs passed bounded queues to us - ERR-71
4. Review Errata list
Proposals:
ERR-70 (Arturo Salz): Default arguments in tasks
Arturo: Disallows output type ports having defaults. If default value
expression not visible from caller side, not an error. Default
values only with ansi-styles.
Michael: It means functions, tasks and methods.
Arturo: Yes.
Ray: Function call in default expression.
Arturo: Yes, we can change it to identifier. It can have anything, it
is not limited to constants.
David: It should be any combination of expression.
Ray: Also note that the expression is evaluated at the call time.
Neil: We should add that.
Ray: It does make it better that it is not a constant., each time it is
called, includes any combination of
Default values and expression evaluated each time the function is
David: What are expressions composed of: Literals, variables and
functions?
Ray: Maybe say that the expression can include variable.
Arturo: Two points, it does not have to be constant and the scope with
respect to caller. It is evaluated at the call time but evalauated
in the scope of the caller subroutine.
Neil: It should be divided into two pieces. First you mention what it
is then you can define the evaluation.
Arturo: It has to be visible in the declaration and the caller.
Ray: So if the default value is declared X, if the X in defined in
two places, which X does it pickup.
Arturo: The caller.
David: Suggest a wording; Evaluated in the scope of the caller and
is visible in both caller and subroutine.
Arturo: Evaluated in the scope of the caller each time it is called.
David: It is clear, how about the visibility.
Arturo: The problem is non-synthesizability of the constructs.
David: Three sentences. Also the section number is 10.4.3.
Two changes suggested here.
Ray: Final edition of that: Expression is not evaluated if it is
not used.
Arturo: That is already there,
Ray: Also important to indicate that it is not evaluated.
David: Expression is not evaluated if not used and need not be visible
to the caller.
Neil: Statement from c++ book.
David: any other discussions.
Final suggested wording.
The default_value is an expression. It is evaluated
in the scope of the caller each time the subroutine is called.
The elements of the expression must be visible at the scope
of both the caller and the subroutine. The expression is not
evaluated if it is not used and need not be visible at the scope
of the caller.
Motion: Accept ERR-70 with changes above
Moved: Ray
Second: Arturo
Abstain: None
Opposed: None
Passed
ERR-71 (Dave Rich): Bounded queues
Arturo: The only change is to allow users to specify a maximum size of
the queue. I did the proposal of the dynamic queues.
You can specify a : and a number after the $ sign.
Michael: Why do this?
Arturo: This makes the queue synthesizable.
Neil: Does this say positive integer. Add:
constant_expression must evaluate to a positive
integer value.
David: We were using the phrase, constant expression to be positive
expression, not including 0.
Arturo: I think it was non-negative vs positive.
Ray: 0 would make it useless.
Arturo: Would it be an error to use it? Should we define the word
positive.
David: It is used in too many places. Constant expression must be
evaluated to a positive integer value should be added.
Michael: Specifying a non-zero in the BNF would be ok?
Arturo: We should not do this.
David: Brad sent out email, commentary for bnf changes,
Neil: Question, the example shows 255, but it should be 256 enteries,
0 to 255.
Ray: It is stated in the next paragraph.
Neil: Specify the maximum size, i.e, by index,
David: Max size is limited by the max of index.
Neil: It looks ok,
Michael: The example makes it quite clear.
Motion: Accept ERR-71 with change above
Moved: Arturo
Second: Michael
Abstain: None
Opposed: None
Passed
ERR-72 (Brad Pierce): Parameter as localparam
David: Brad's suggestion for adding one sentence for the localparam.
Michael: Since it is compilation unit, there is no difference between
param and localparam.
Motion: Accept ERR-72
Moved: Arturo
Second: Michael
Abstain: None
Opposed: None
Passed
ERR-73 (David Smith): Tagged examples need renamed
David: Tagged union work in BC generated new tagged keyword, EC
examples had tagged in them, so we need to change them, I chose
tagged_st.
Neil: st stands for struct. I am seeing some question mark show up
in my netscape email.
David: It must be the browser
Motion: Accept ERR-73
Moved: Neil
Second: Arturo
Abstain: None
Opposed: None
Passed
5. LRM Review
David: I have received feeback from Brad bot noone else.
Neil: I have looked at it, written up some, will complete it and
send it.
David: Most of these changes are non-controversial, I will go ahead and
submit them with draft4 changes.
Ray: Not yet any chance to look at them.
David: how about on consistency checks.
Michael: Started but not done yet.
David: It should have all the changes from AC in it.
Michael: There is still one thing, adding subroutine calls to the
sequences, they are still working on it. Method calls, it is still
being worked on.
Arturo: They had allowed pure function call inside, no side-affects.
They
are considering those changes. Inside assertions, (called local
variables), assertion variables, need to get the information out of
assertions to the rest of testbench. They have added capability to
call a task as opposed to function, at the end of sequence.
David: There are some errata in AC, but limited.
Arturo: They thought about putting print statement, but they need more.
David: CC has one more meeting on Friday, so draft4 should be complete,
to be available at end of next week, but more likely first week of
February.
Summary and Assignments:
Glossary
Dave
Verification of all cross references
Sections 1-7: Ray
Sections 8-14: Stu
Sections 15-21: Neil
Sections 21-29: Brad
Check all changes for consistency and correctness
Assertions - Michael
C API -
Classes/Randomization - Mehdi
Constraints - Arturo
Types - Dave
6. Review 3.1a Extensions and discussion
All closed
7. Meeting Logistics
Next meeting scheduled for 2 February 2004 from 11:00am until 1:00pm
Meeting on 16 February falls on President's day - reschedule?
Move to 18 February from 10 till 12.
Focus on editorial review and any open errata.
8. Next Meeting
Monday February 2, 2004, 11:00am-1:00 pm PST
9. Meeting adjourned at: 10:55 am.
This archive was generated by hypermail 2b28 : Wed Jan 21 2004 - 17:38:40 PST