[Libre-soc-dev] daily kan-ban update 06sep2020

[Cesar Strauss]

Last week:

* Moved write_gtkw to nmutil.gtkw
It's now available for general use.
Usage example at:
https://git.libre-soc.org/?p=nmutil.git;a=blob;f=src/nmutil/test/example_gtkwave.py;h=3ea19c548d6eace10d4550d3af92308acfef91eb;hb=HEAD
But, maybe you might want to wait for the tutorial.
Also, the API will probably evolve, as I try it for real on more test cases.
Contributions and suggestions are welcome.

* Documented its build dependency, on nmutil's README.md and setup.py.
It didn't seem necessary to modify HDL_workflow, since "python setup.py
develop" already takes care of it.

* Started writing the GTKWave tutorial.
For now, it's a high level overview of what is to come.
https://libre-soc.org/Documentation/gtkwave_tutorial/
(work in progress)

* Tested the proposed fix for a cxxsim bug.
The race condition was fixed, but GTKWave still shows traces changing on
the falling clock edge.
But, only on the original, non-reduced test case.
Will try to narrow down the test case further (again). It's fun.


This week:

* Provide a reduced test case for the cxxsim bug.

* Fill out the GTKWave tutorial with actual content

Next (as time permits):

* Interested in writing a formal proof for
src/soc/experiment/alu_fsm.py, to acquire some experience.

* Add GTKWave documents to:
  - src/soc/experiment/alu_hier.py
  - src/soc/experiment/test/test_compalu_multi.py
  - src/soc/experiment/compldst_multi.py

* Write parallel unit tests for test_compalu_multi.py and
compldst_multi.py, from the experience already gained on alu_hier.py.

* Maybe, write a function to walk an nMigen top-level Module and output
a GTKWave document of its signals, colored by hierarchy.

Regards,

Cesar