Subject: RE: [sv-ec] Motorola vote on sending the standard to the board
From: Kevin Cameron x3251 (Kevin.Cameron@nsc.com)
Date: Thu Apr 24 2003 - 15:26:40 PDT
> From: "David W. Smith" <david.smith@synopsys.com>
> To: sv-ec@eda.org
>
> Greetings,
> Just to set the record straight. There was no better support fine grain
> control of individual threads in 3.0. There were two proposals that would
> have met these requirements (when with 3.0 syntax and one building from 3.1
> fork...join syntax) that were made during 3.1. They were not processed for
> the following reasons:
>
> 1. It was not a primary objective in 3.1 to have support this for this.
> 2. There is a way using global variables ot provide some support within 3.1
> (awkward though it is).
> 3. There was insufficient time to handle this issue with the body of other
> issues that we were obligated to handle.
I.e. it was left half-baked as it was in 3.0, or rather one deficient
implementation was replaced by another with more idiosyncratic syntax.
> 4. There was no proposal provided that had been implemented in any form.
I seem to remember citing the Posix Threads implementation as a model for
process control. All it needed was a mechanism to get a thread(/group) id
- an extension the 3.0 syntax supported better than the 3.1 does.
> Having said this it clearly is an issue is of importance to many people. It
> will definitely go on the list of Post-3.1 issues and I hope that we can
> have a solution (as a committee) and resolve it quickly.
Are there any outstanding issues from 3.0 also being transferred?
Item #3 on the Post-3.1 list needs to be considered in conjunction with
back-annotation and user-defined type driver resolution, can you add
those two items to the list too please.
Regards,
Kev.
> Regards
> David
>
> David W. Smith
> Synopsys Scientist
>
> Synopsys, Inc.
> Synopsys Technology Park
> 2025 NW Cornelius Pass Road
> Hillsboro, OR 97124
>
> Voice: 503.547.6467
> Main: 503.547.6000
> FAX: 503.547.6906
> Email: david.smith@synopsys.com
> http://www.synopsys.com
>
>
>
> -----Original Message-----
> From: owner-sv-ec@eda.org [mailto:owner-sv-ec@eda.org] On Behalf Of Kevin
> Cameron x3251
> Sent: Thursday, April 24, 2003 1:31 PM
> To: sv-ec@eda.org; Michael.Burns@motorola.com
> Subject: Re: [sv-ec] Motorola vote on sending the standard to the board
>
>
>
> > From: "Michael Burns" <Michael.Burns@motorola.com>
> >
> > Hi,
> >
> > Motorola votes "yes" on sending the latest System Verilog 3.1 to the
> > Accellera board for approval, with the following comment:
> >
> > -- We're unhappy that the syntax for spawning child threads appears to
> > be making it difficult to define a syntax extension that will allow
> > control (i.e. killing) of individual threads, recognizing that the
> > existing syntax was inherited from the earlier standard. The lack
> > of an individual thread control mechanism may limit our adoption of
> > SV3.1.
> >
> > Mike Burns
> > Motorola
>
> There was better syntax from SuperLog in the 3.0 version of SV that got
> removed. That fact that it got removed tends to indicate to me that that LRM
> was premature (I voted against it), and that this one is equally
> premature.
>
> Kev.
>
>
>
>
This archive was generated by hypermail 2b28 : Thu Apr 24 2003 - 15:46:24 PDT