From: Adam Krolnik (adamk@cyrix.com)
Date: Tue Jan 26 1999 - 08:55:03 PST
Regarding views and macros/arguments.
I think it is worthwhile to be able to associate a set
of arguments (to the compiler)/macro definitions with a view.
Justifications
A debug view enables debug statements with a macro definition.
A timing accurate model requires the compiler argument -Xkaq=0x1000
to enable the timing stuff.
A subset of a design expects a specific timescale.
It would also be required that if a view was invoked that had
a macro definition or compiler argument specified, it would
be implicitly setup without required user intervention - otherwise
the macro/arg is just a documentation line.
<p>But I do not agree with the specified functionality of collecting
modules into a view that have that macro defined/directive set.
Nor do I agree with switching views when a macro is defined on
a command line or other places.
Justifications
This is unreasonable approach to collect a set of modules
into a view. Instead, use `view to specify the view for
the modules. You already specified the argument - why not
the view.
It becomes confusing - action at a distance. I do this and it
does something to a module unrelated by name/reference. It
is difficult to determine what happens - resulting in searches
for something common.
The macro/arg definitions should not be used as another name
to wildcard modules together with. Macros and arguments have
a meaning - not for views.
Regarding config file options/statements:
I believe the current spec is trying to give allow
specification of the same material in many different places (source
files, config files, command line.) This expressiveness seems
more than is really needed.
E.g.
Is it really necessary to allow respecification of the actual
module instantiated in a config rather than at the instantiation?
Is it necessary to allow one to refer to an instance name to
specify alternative libraries/views?
Is it necessary to allow one to group modules based on compiler
directive settings?
Adam Krolnik
Verification Engineer
Cyrix - NSC.
Richardson TX. 75085
This archive was generated by hypermail 2.1.4
: Mon Jul 08 2002 - 12:53:25 PDT
and
sponsored by Boyd Technology, Inc.