[myhdl-list] About multiple simulations without leaving the interpreter

[Jan D. <ja...@ja...>]

Hello:

There is some new functionality available since 0.6dev6 that I still
have to document.

It's relatively advanced, and depending on your usage patterns you
may never need it. So if the following doesn't ring a bell, you can
safely ignore the info.

In a recent project I defined the internal signals "globally" in
a separate module. This way, I could define lookup functions to
make it easier to define top-level netlists and test benches (and
keep them up to date.)

Also, running my test plan meant that I did multiple simulations
without leaving the interpreter.

This usage pattern implies that after the first simulation, the
*same* (in terms of object identity) Signals are reused in the second
one.

This led to problems. For example, at the end of a simulation Signals
were not "reset" to their initial value, leading to "unexpected"
values at the start of the next simulation. In general, the Simulation
and Signal data structures were not cleaned up systematically at
the end of a simulation.

This is now solved: at the end of a simulation, things are cleaned up
so that a subsequent simulation in the *same* interpreter gives the
expected results: the same behavior as if the new simulation would
run in a *new* interpreter.

Best regards,

Jan

-- 
Jan Decaluwe - Resources bvba - http://www.jandecaluwe.com
Kaboutermansstraat 97, B-3000 Leuven, Belgium
     From Python to silicon:
     http://myhdl.jandecaluwe.com