Re: Compilation question

Hi Arpad.. 
I agree with Kevin that this is not something that
must necessarily be defined in the LRM, but rather
would be a product of the LRM definition of what
things can depend on a parameter, and where in the
simulation cycle compilation happens.. 

Since I would assume you are talking about a "pre-run"
compilation - ie separate from the actual simulation
run. Its my understanding of the LRM that the
parameter should be modifiable at RUN time.. (ie with
a -defparam on the command line) so I would assume
that a requirement of the compiler would be NOT to
compile for one specific value of the parameter .. -
unless perhaps one used a parameter to define the
width of a signal (not something I got in the habit of
doing since my normal overall environment didn't have
the right kind of flexibility to make it a practical
proposition.. though I have seen it commonly used by
those with a more text-based overall design
environment than I was used to.)

Hope this kind of info helps you.. 
Jonathan




--- Kevin Cameron <kevin@sonicsinc.com> wrote:

> 
> The question should be aimed at your simulator
> supplier - rather than 
> the committee. Compilation is (or isn't) a
> requirement of the 
> implementation, so you won't find it discussed much
> in the LRM.
> 
> Kev.
> 
> Muranyi, Arpad wrote:
> 
> >Happy New Year to everyone!
> >
> >I have a question about the compilation of modules.
>  Let's
> >take an example which is a model for a resistor, in
> which
> >the resistance value is parameterized.
> >
> >1)  Does this module have to be re-compiled for
> every value
> >of the parameter value (in multiple instances in
> the same
> >simulation, or in the same instance of a sweep
> simulation)?
> >
> >Let's modify the problem, and consider a module for
> a PWL
> >source, for example, in which the data table is an
> array
> >parameter.
> >
> >2)  If the size of the array parameter is fixed
> inside the
> >module, will this module need to be recompiled for
> each
> >instance or iteration that causes the data to be
> different?
> >
> >Let's modify this PWL module so that the array size
> is
> >dynamic, i.e. it is determined by another
> parameter,
> >therefore the various instances are allowed to have
> >a different array length.
> >
> >3)  Will this module need to be recompiled for each
> instance
> >if the instances get different array parameter
> sizes?
> >
> >4)  Will this module need to be recompiled for each
> >iteration of a sweep simulation if the array size
> is
> >not the same between the iterations?  (I don't know
> >if this could realistically happen or not...)
> >
> >5)  Any related comments, cautions, suggestions?
> >
> >Thanks,
> >
> >Arpad
>
>=============================================================
> >
> >
> >  
> >
> 
>