RE: [sv-ec] Program Block / Reactive Region Use Cases & Requirements

From: Stuart Siutherland <stuart_at_.....>
Date: Thu Dec 14 2006 - 17:16:59 PST
While maintaining backwards compatibility is our "prime directive", there
are many times over the years that we have compromised on backward
compatibility because it was the right thing to do for USERS of the
standard.  IMHO, this is one of those times.  This change to have anything
called from a program block thread execute in program block regions is
intuitive to the user, and it greatly simplifies writing test programs.  The
user won't need to worry about when to use blocking/nonblocking assignments,
what package a task/function is imported from.  Code brought over from a
Verilog-style testbench will run correctly without have to manually figure
out which assignments have to be converted to nonblocking, ...

Let's make the standard better for the end user of the standard and SV
tools, make this change, add a note to the LRM that the 1800-2005 standard
used a different event scheduling rule, and forget about dragging this
discussion on and on and on!

My $0.02 (or less, after adjusting for inflation)...

Stu
~~~~~~~~~~~~~~~~~~~~~~~~~
Stuart Sutherland
Sutherland HDL, Inc.
stuart@sutherland-hdl.com
503-692-0898
 

> -----Original Message-----
> From: owner-sv-ec@server.eda.org 
> [mailto:owner-sv-ec@server.eda.org] On Behalf Of Steven Sharp
> Sent: Thursday, December 14, 2006 3:42 PM
> To: sv-ec@server.eda.org; Neil.Korpusik@Sun.com
> Subject: Re: [sv-ec] Program Block / Reactive Region Use 
> Cases & Requirements
> 
> 
> >From: Michael Burns <michael.burns@freescale.com>
> >
> >Second, it has come to my attention (from a discussion with 
> Rob Slater) that 
> the
> >new program block semantics proposal as it is now may have a 
> backwards
> >compatibility issue as well.
> 
> 
> The new proposal definitely breaks backwards compatibility.  
> For example,
> it changes the behavior of a delay in a module task called 
> from a program.
> By the 2005 standard, the task would wake up in the active region.  By
> the new proposal it would wake up in the reactive region.  This is not
> backwards compatible.
> 
> Steven Sharp
> sharp@cadence.com
> 
> 
> 
Received on Thu Dec 14 17:17:04 2006

This archive was generated by hypermail 2.1.8 : Thu Dec 14 2006 - 17:17:28 PST