Subject: AW: Brainstorming VHDL 200x
From: Wolfgang.Ecker@infineon.com
Date: Wed Apr 19 2000 - 11:11:55 PDT
Yust as an add on. The synthesis tool, that I use gives
warnings also, however does not exactly analyse if only
slices of composite signals are used.
-----Ursprüngliche Nachricht-----
Von: Evan Lavelle [mailto:eml@riverside-machines.com]
Gesendet am: Mittwoch, 19. April 2000 16:30
An: stds-vasg@ieee.org
Betreff: Re: Brainstorming VHDL 200x
Stephen Bailey wrote:
> > Good idea. I suppose the only downside is that simulation of clocked
> > processes will be slower if this is used indiscriminately. I suspect
> > that there will also be combinatorial cases in which pre- and post-synth
> > sims will differ (can't think of one right now) but this will be a case
> > of buyer bewares.
>
> You are missing the point. What Lance has proposed here is for
> *combinatorial* processes and not *sequential* processes (in usage).
That's why I said "if used indiscriminately". If the feature exists,
then there's nothing to stop it being used for clocked processes. The
sim's not going to complain, but will run slower.
> Synthesis style guidelines would still apply. Therefore synthesis tools
> would expect to only see the "clk" and "asynchronous control" (if any)
> signals in the sensitivity list for a sequential process.
The synth tools I use ignore the sensitivity list, but may issue a
warning if it's "wrong". I don't believe that I've seen warnings issued
if there are too many signals in the list. Certainly, a F/F template
will produce a F/F, irrespective of the contents of the sensitivity
list.
> Synthesis
> guidelines expect all signals read in a combinatorial process to be in the
> static sensitivity list. If the sensitivity list is incomplete, synthesis
> completes it. This can result in differences between simulation behavior
> and synthesis results.
Conversely, if the list is complete when it should have been incomplete,
then there's a potential for problems. It's easy to think of simulation
cases to demonstrate this, but more difficult to come up with something
synthesisable - that's why I said I couldn't immediately think of a
case.
I'm not disagreeing with Lance - as I said, I think it's a good idea,
but the user has to beware. It shouldn't be used indiscriminately.
Evan
This archive was generated by hypermail 2b28 : Wed Apr 19 2000 - 11:39:16 PDT