Re: gEDA: silly question on simulating a synthesized design with Icarus

[Larry Doolittle <ldoolitt@recycle.lbl.gov>]

On Fri, Aug 15, 2003 at 12:21:45PM -0700, I wrote:
> [When] I blithely turn on the "-S" switch [for a simulation]
> I get a ton of
> :0: internal error: NetProc::nex_output not implemented
> messages, that I can easily track back to $display()s,
> by adding
>       dump(cerr,4);
> to the end of void NetProc::nex_output(NexusSet&out).

I can eliminate the problem by putting (* ivl_synthesis_off *)
in front of the always blocks that contain $display() calls.
I don't think I'm supposed to need this, synthesis is supposed
to ignore procedure calls:
  Std 1364.1-2002 7.15
    All system tasks shall be ignored.
  Std 1364-1-2002 7.18
    PLI task calls shall be ignored.
and even if my construct _shouldn't_ be allowed (given the odd
combination of synthesis and simulation that Icarus implements),
I should get some more useful message than "internal error".

My next hurdle is reported as PR#786, "vvp parse error with synthesis".
  http://www.icarus.com/cgi-bin/ivl-bugs?findid=786

I'm still curious to hear from other people who may have explored
this dark corner of Icarus.

      - Larry