SV-AC_Minutes_2011_06_21 < P1800

P1800 Web>SystemVerilogAssertionCommittee>SVACMeetingMinutes>SV-AC_Minutes_2011_06_21 (2011-07-08, ErikSeligman) (raw view)EditAttach
Minutes of SV-AC Meeting

Date: 2011-06-21

Time: 16:00 UTC (9:00 PDT)

Duration: 1.5 hours

Dial-in information:

--------------------

Meeting ID:            38198

Phone Number(s):

1-888-813-5316         Toll Free within North America

Live Meeting:

[[https://webjoin.intel.com/?passcode=4797656]]

Agenda:

-------

- Reminder of IEEE patent policy.

See: http://standards.ieee.org/board/pat/pat-slideset.ppt

- Minutes approval

- Email ballot results

Mantis 3069 Passed: 6y, 0n, 0a

Mantis 3033 Failed: 0y, 5n, 1a

- New issues

- Issue resolution/discussion

3033: Allow procedural control statements is checkers 3552:Sequence methods // .triggered need further clarification

- Enhancement progress update

- Mantis status

Attendance Record:

------------------

         Legend:

                 x = attended

                 - = missed

                 r = represented

                 . = not yet a member

                 v = valid voter (2 out of last 3 or 3/4 overall)

                 n = not a valid voter

                 t = chair eligible to vote only to make or break a tie

Attendance re-initialized on 2010-07-06:

  v[xxx---xxxx-xxx-xxx...........................] Ashok Bhatt (Cadence)

  v[-xxxxxx-xxxxxxxxxx-xxx-xxxxxxxxx-x-xxxxx--xxx] Laurence Bisht (Intel)

  v[xxxxxxx-xxxxxxxxx-xxxxxxxxxxxx-xxxxxxxxxxxxx-] Eduard Cerny (Synopsys)

  v[xxxxx--------xx---xxx--x-xxxxxxx-xxxxx-xxxxxx] Ben Cohen (Accellera)

  n[--------------------------xx-x-xxx-x--xxxxxxx] Surrendra Dudani

(Synopsys)

  v[xxxxx........................................] Shaun Feng (Freescale)

  n[-x--x-x-xxxx-x-x----x-x-x--xx---xxxx---x-xxxx] Dana Fisman (Synopsys)

  n[------------------------xxxxx-xxxx-x-xxxxxxxx] John Havlicek

(Freescale)

  v[xxx-xx-xxxx-xxxxxxxxxxxxxxxx-xxx-xxxxxxxxxxxx] Tapan Kapoor (Cadence)

  v[-xxxxx-xxxx-x-x..............................] Jacob Katz (Intel)

  t[-xxx-xxxxxxxxxxxxxxxxxxxxxxx--xxxxxxxxxxxxxxx] Dmitry Korchemny (Intel ¿ Chair)

  v[x-xxxxx-xxxx-xxxxxxxxxxxxxxx--xxxxxx-xxxxxxxx] Scott Little (Freescale)

  v[xxxxxxxxx-xxxxxxxxxxxxxxxxx-xxxxxxxxx-xxxxxxx] Manisha Kulshrestha (Mentor Graphics)

  v[xxxxxxxxxxxxxxxxxxxxxxxxx-xxxxxxxxxxxxxxxxxxx] Anupam Prabhakar (Mentor Graphics)

  n[x--xxx-xxxx-xxx-xxx--x-xx-xxx-xx--xxxxxxx-xxx] Erik Seligman (Intel)

  v[x-xxxxxxxxxx-x-xxx-xxxx-xxxx--xxxxxx-xxxxxxx.] Samik Sengupta

(Synopsys)

  v[xxxxxxxxx-xxxxxxxxxxxxxxxxx-xxxxxxxxxxxxx-xxx] Tom Thatcher (Oracle ¿ Co-Chair)

  n[---------xx---xx-------x.....................] Srini Venkataramanan

 (CVC Pvt Ltd)

    |- attendance on 2011-06-21

  |--- voting eligibility on 2011-06-21

Minutes:

--------

- Reminder of IEEE patent policy.

See: http://standards.ieee.org/board/pat/pat-slideset.ppt

Participants were reminded of the IEEE patent policy.

- Minutes approval

Ben: Move to approve minutes

Samik:     Second

     Vote results: 9y, 0n, 1a

3033 Allow procedural control statements is checkers

Ben: Proposal gave lots of rationale for why things are done a certain

     way.

     Rest of LRM doesn't do this.

Erik: Yes, some of this should be moved to preamble of the proposal.

Ben: Blocking assignments used as a variable.

     A good idea to allow blocking assignments. But only as a local

     variable.

Tom: You can declare an automatic variable inside a for loop.

Ben  The idea would be to allow use as temporary variables.

     You can do this otherwise using a function call.

Tom: My feedback comments regarding the always_latch were incorrect.

     No change to the proposal needed for the example with always_latch.

Erik: Schedule discussion for when Dmitry comes back.

- Issue resolution/discussion

3552 Was moved to SV-BC

3195 Local Variable flow out.

Ben: Would like to revive this proposal.

Erik: Will review this proposal

Scott:     Will review this proposal

3478 Make local values accessible.

Ben: This seems to be a hole. Would like to fix it.

Scott, Erik: This would be a new issue that we would need

     Sampled values in classes & constraints

2476: Generalized count-bits function

Scott:     Champions feedback was different this time. Neil only updated

     The Mantis items, He didn't send out e-mail

Tom  Will ask Neil about Champions feedback

Erik: Dave rich opposed use of a bit vector argument (variable bit width)

     because vectors are usually extended when there is a mismatch.

     This function would be an exception: If the user specifies 2 bits,

     the value would not be extended to 4 bits, but used as is.

    He proposed a Logic queue. Or make it a constant 4-bit argument.

Tom: If it is a constant 4-bit argument, then if a user specifed a 2-bit

     argument, it would be extended, and would likely contain a 0, which

     would give unexpected results.

Scott:     Of the given options, prefer the queue.

Erik: Will update proposal with the queue idea.

     Probably will be ready by next meeting.

2412: Allow clock inference in sequences.

     Champions feedback:

Anupam:    Shalom's feedback: modules don't have arguments, they have ports.

Anupam:    Made an editorial change

Manisha: What about checkers?

Erik: BNF says that checkers have "ports"

Scott:     Is it proper to say "function instantiation?"

Tom: Yes, because Verilog's original functions are static.

Anupam:    VPI section mentions function instnace

Anupam:    Maybe should change text to "function call".

Manisha:   Function call should be a more correct usage.

3552: Sequence methods .triggered() needs clarification

     Dave Rich posted a comment

Manisha:   One thing not clear

     Is Dave saying we should not define it?

     Or is he saying that this should be illegal?

Ed:  Right now, it's left open.

Ed:  Suggest in proposal that the "bad usage" comments be changed to

     "behavior not defined"

Ed:  Does example code occur in a module or an interface?

     What happens when this is an interface passed to a program?

     In a program, the "if" sthatement would be ok

     If in an interface passed to a module, then the "if" would

     probably not work.

Ben: Could we test the code on a simulator?

Ed:  That's one approach. But we should study LRM to get an answer

     For this proposal, make the example into a module, then we don't

     have to worry about the Program case.

2412: Allow clock inference in sequences.

Anupam: Has uploaded modified proposal.

Ben: Prefer to e-mail vote it.

Tom: Will call for an e-mail vote.

Meeting adjouned.
Topic revision: r2 - 2011-07-08 - 20:36:37 - ErikSeligman
P1800