RE: [sv-ac] reminder to vote on mantis 1550

Hi John,

I agree that this is what should happen, but as you say yourself, it is
not stated in the algorithm. Should we add that there? Or in 1550 just
for the sampled value fncts? I'd still like to add that that result of
initializationa nd evaluation sets values that can be thought of as
being valid from -oo till 0-.

ed

  

> -----Original Message-----
> From: John Havlicek [mailto:john.havlicek@freescale.com] 
> Sent: Friday, January 26, 2007 10:08 AM
> To: Eduard.Cerny@synopsys.COM
> Cc: john.havlicek@freescale.com; Eduard.Cerny@synopsys.COM; 
> sv-ac@eda-stds.org
> Subject: Re: [sv-ac] reminder to vote on mantis 1550
> 
> Hi Ed:
> 
> I never said anything got scheduled into the preponed region.
> 
> Look at the scheduling algorithm.  The step
> 
>   initialize the values of all nets and variables
> 
> executes before the execution of the time 0 slot begins.  My
> interpretation is that this step includes execution of the 
> declaration assignments.
> 
> The value of an expression should be readable in the 
> preponed region.  What I am saying is that if an expression
> is read (or sampled) in the preponed region of the time 0 slot,
> then it should see the effects of the 
> 
>   initialize the values of all nets and variables
> 
> In other words, the declaration assignments are reflected in 
> the preponed values in the time 0 slot.
> 
> Do you disagree?
> 
> J.H.
> 
> 
> > X-MimeOLE: Produced By Microsoft Exchange V6.5
> > Content-class: urn:content-classes:message
> > Date: Fri, 26 Jan 2007 06:56:31 -0800
> > Thread-Topic: [sv-ac] reminder to vote on mantis 1550
> > Thread-Index: AcdBWOUQu1Vg9ZovRt+2v6wWbB8FGAAAQpnQ
> > From: "Eduard Cerny" <Eduard.Cerny@synopsys.com>
> > Cc: <sv-ac@eda-stds.org>
> > X-OriginalArrivalTime: 26 Jan 2007 14:56:32.0702 (UTC) 
> FILETIME=[2AF6ADE0:01C7415A]
> > 
> > John, you cannot schedule anything into a preponed region, 
> or can you?=20
> > 
> > From page 123:
> > The Preponed region provides for a PLI callback control point that
> > allows PLI application routines to access
> > data at the current time slot before any net or variable has changed
> > state. Within this region, it is illegal to
> > write values to any net or variable or to schedule an event 
> in any other
> > region within the current time slot.
> > 
> > How does that give values in preponed region?
> > 
> > ed
> > 
> > 
> > > -----Original Message-----
> > > From: John Havlicek [mailto:john.havlicek@freescale.com]=20
> > > Sent: Friday, January 26, 2007 9:47 AM
> > > To: Eduard.Cerny@synopsys.COM
> > > Cc: john.havlicek@freescale.com; Eduard.Cerny@synopsys.COM;=20
> > > sv-ac@eda-stds.org
> > > Subject: Re: [sv-ac] reminder to vote on mantis 1550
> > >=20
> > > Hi Ed:
> > >=20
> > > > I have been looking at Clause 9 of the LRM and could not=20
> > > find anything
> > > > specific with regards to  the value of expressions in 
> the preponed
> > > > region at time 0. Perhaps I am looking at the wrong place.
> > >=20
> > > Here is the argument I gave in a previous email:
> > >=20
> > >    First, based on 9.3.1, I think that the declaration 
> assignments for
> > >    static variables execute as part of=20
> > >   =20
> > >       initialize the values of all nets and variables
> > >   =20
> > >    Then the initialization events are scheduled into the 
> time 0 slot.
> > >    This does not mean that any assignments from an 
> initial block are
> > >    performed yet. =20
> > >   =20
> > >    Then, assuming that there is some event, we execute the=20
> > > time 0 slot.
> > >   =20
> > >    Then the preponed region of the time 0 slot is executed.
> > >   =20
> > >    Thus, the declaration assignments for static variables=20
> > > will influence
> > >    the preponed values in the time 0 slot.
> > >=20
> > > I agree that we will have to think about how $past, etc. are=20
> > > defined and what this means for time 0 slot evaluation when we
> > > resolve 1698.
> > >=20
> > > Best regards,
> > >=20
> > > John H.
> > >=20
> > >=20
> > > > X-MimeOLE: Produced By Microsoft Exchange V6.5
> > > > Content-class: urn:content-classes:message
> > > > Date: Fri, 26 Jan 2007 06:19:12 -0800
> > > > Thread-Topic: [sv-ac] reminder to vote on mantis 1550
> > > > Thread-Index: AcdA50wluAaOiDR2QrCpyFXIT1AqLAAbLhmA
> > > > From: "Eduard Cerny" <Eduard.Cerny@synopsys.com>
> > > > Cc: <sv-ac@eda-stds.org>
> > > > X-OriginalArrivalTime: 26 Jan 2007 14:19:13.0540 (UTC)=20
> > > FILETIME=3D[F451AC40:01C74154]
> > > >=20
> > > > Hi John,
> > > >=20
> > > > I have been looking at Clause 9 of the LRM and could not=20
> > > find anything
> > > > specific with regards to  the value of expressions in 
> the preponed
> > > > region at time 0. Perhaps I am looking at the wrong place.
> > > >=20
> > > > Also, it is not just sampled value at time 0, but also=20
> > > $past (and the
> > > > other sampled value functions) that depend on values at=20
> > > clock ticks in
> > > > the past (<0). This is why I think that there should be=20
> > > something said
> > > > about these values, along the lines I mentioned earlier -=20
> > > the default
> > > > and values at declarations are assumed to exist from time=20
> > > -oo. Perhaps
> > > > we should add that expressions dependent on these <0 
> values are also
> > > > evaluated to get their values ate time <0.  I agree with=20
> > > your thought
> > > > process.
> > > >=20
> > > > Best regards,
> > > > edd
> > > >=20
> > > > =3D20
> > > >=20
> > >=20
> 

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.