SV-AC_Minutes_2011_07_12 < P1800

P1800 Web>SystemVerilogAssertionCommittee>SVACMeetingMinutes>SV-AC_Minutes_2011_07_12 (2011-07-19, ErikSeligman) (raw view)EditAttach
Minutes of SV-AC meeting

Date: 2011-07-12

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=9628489]]

Agenda:

-------

- Reminder of IEEE patent policy.

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

- Minutes approval

- Champions Feedback

Mantis 3385 Possible ambiguity when deferred assertion action block

           (Erik)

           Opposed: Language: First person sentences, etc.

Mantis 2476 Need clarification about system functions $onehot, etc

           (Erik)

           Opposed: Consider using a queue of logic rather than

                                                fixed 4-bit argument.

Mantis 3113 Add port_identifier to constant_primary BNF for sequences,

           properties and checkers

           (Laurence)

           Opposed: constant_range_expression too restrictive Mantis 3135 Verbal explanation of nexttime and always is misleading for

           multiple clocks.

           (Dana)

           Friendly Amendment: Font Problem

- New issues

Mantis 3015       Examples of $fatal have bad arguments

- Issue resolution/discussion

Mantis 3033       Allow procedural control statements is checkers

- Enhancement progress update

- Opens

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:

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

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

  v[xxxxxxxxx-xxxxxxxxx-xxxxxxxxxxxx-xxxxxxxxxxxxx-] Eduard Cerny

(Synopsys)

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

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

(Synopsys)

  v[-xxxxxx........................................] Shaun Feng (Freescale)

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

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

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

  v[xxx-xxxxx-xxxx-xxxxxxxxxxxxxxx--xxxxxx-xxxxxxxx] Scott Little

(Freescale)

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

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

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

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

(Synopsys)

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

  n[-----------xx---xx-------x.....................] Srini Venkataramanan (CVC Pvt Ltd)

  n[-x.............................................] Vibarajan Viswanathan (Marvell)

    |- attendance on 2011-07-12

  |--- voting eligibility on 2011-07-12

Minutes

-------

- Reminder of IEEE patent policy.

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

- Minutes Approval

Erik: Move to approve minutes

Scott:     Second

     Vote results: 10y, 0n, 0a

- Champions Feedback

Mantis 3385 Possible ambiguity when deferred assertion action block

           (Erik)

           Opposed: Language: First person sentences, etc.

Erik: Has modified the proposal to resolve Champions objections.

     Change was to change the sentence to third-person.

Tom: Move to approve

Erik: Second

     Vote results 10y, 0n, 0a

Mantis 2476 Need clarification about system functions $onehot, etc

           (Erik)

           Opposed: Consider using a queue of logic rather than

                 fixed 4-bit argument.

Erik: Dave Rich says that a fixed-size queue should be synthesizable

     Some examples were too obvious to Champions.

     Should be ready for an e-mail vote.

Jacob:     Not sure any synthesis tool will support it.

     What about an unpacked array of fixed size.

     A queue is a special case of an unpacked array of fixed size.

Ed:  Only other option would be to allow only single bit arguments,

    but a variable number of arguments.

     E.g. dumpvars allows variable number of arguments

Erik: Like the idea of variable number of arguments.

     Will re-write the proposal using this construct.

(Manisha joined)

Mantis 3113 Add port_identifier to constant_primary BNF for sequences,

     properties and checkers

     (Laurence)

     Opposed: constant_range_expression too restrictive

Lawrence: Has made changes for champions comments

     Recommend an E-mail vote.

Dmitry:    Will call for an e-mail vote.

- New issues

Mantis 3015       Examples of $fatal have bad arguments

Dmitry:          Entered by Shalom.

           Will call for an e-mail vote.

- Issue resolution/discussion

Mantis 3033       Allow procedural control statements is checkers

Dmitry:    

Anupam:          What is reason for keeping everything in Reactive.

     Only thing I can think of is Free variables.

Dmitry:          Can't assign from seq.triggered(). i.e. in always_ff

Dmitry:    Should we keep always inside checker?

     The always construct inside a checker is limited to be identical

     to always_ff

     Tom in favor of keeping: backward compatibility.

Ben: What about local variables in always.

     In module could define a variable local to the always, and

     can have blocking assignment to that local variable.

Tom: Current proposal: no blocking assigns allowed at all.

Ben: Okay with deprecating always, but . .

Dmitry     Deprecate or not?

     Jacob: not sure

     Scott: yes

     Tom: no

     Ed:  yes

     Lawrence:  yes

     Anupam     don't care

     Erik yes

     Samik yes

     Manish     don't care but need to be sure we will never use again

Dmitry: Can't we just "un-deprecate" it if we need it in the future?

Dmitry:    Static prefix restriction

     Removed it.

     Was put in for singale assignment rule. Now that we have always_ff,

     the always_ff imposes restriction that no other process can assign

     to the variable.

Dmitry: Initial assignment

     Restriction on initial assignments to variables is needed because

     of non-determinism because of procedure ordering at time 0.

Dmitry: Non-determinism example

Tom: Don't need this example. Need checker sampling to make always_ff

     act like a sequential element.

Anupam:    What is the use for deferred assertions in always_ff blocks?

Dmitry:    Immediate assertions in checkers. Currently illegal.

     Now don't know why it's necessary to make them illegal

     Any objections to removing this restriction?

Anupam:    Why were they illegal in the first place?

Dmitry: We didn't allow if statements in always blocks, and immediate

     assertions act as an if.

Dmitry:    Will modify proposal in response to comments

Meeting adjourned.
Topic revision: r1 - 2011-07-19 - 21:47:47 - ErikSeligman
P1800