[WIP] Add Display statement.

[ghost]

Intent: fixes #432.

• Rebase of dc6a805, which adds the Display statement with support in simulation.
• Minimally working RTLIL backend support.
  • Confirmed this works in CXXRTL.

TODO:

• Synchronous Display statements
• Lots of cleaning up and poking-around testing.
• Unit tests.
• Documentation.

[codecov]

Codecov Report

Merging #871 (90b6583) into main (c7da6c1) will decrease coverage by 0.74%.
The diff coverage is 44.02%.

@@            Coverage Diff             @@
##             main     #871      +/-   ##
==========================================
- Coverage   83.73%   82.99%   -0.74%     
==========================================
  Files          54       54              
  Lines        7681     7788     +107     
  Branches     1868     1896      +28     
==========================================
+ Hits         6432     6464      +32     
- Misses       1048     1123      +75     
  Partials      201      201              

Files	Coverage Δ
amaranth/__init__.py	85.71% <ø> (ø)
amaranth/hdl/__init__.py	100.00% <100.00%> (ø)
amaranth/hdl/dsl.py	99.73% <100.00%> (ø)
amaranth/sim/_pyrtl.py	96.07% <88.23%> (-0.53%)	⬇️
amaranth/back/rtlil.py	76.97% <25.00%> (-0.34%)	⬇️
amaranth/hdl/xfrm.py	95.37% <79.16%> (-0.85%)	⬇️
amaranth/hdl/ast.py	84.74% <25.28%> (-5.25%)	⬇️

... and 1 file with indirect coverage changes

📣 We’re building smart automated test selection to slash your CI/CD build times. Learn more

[whitequark]

Thank you for this contribution. Project policy requires that substantial changes (such as this one) go through the RFC process first. As this feature (along with its sister feature #427, which will use the same kind of format strings) is a high-priority one on the roadmap, I will go ahead and write the RFC for it @wanda-phi ended up writing the RFC. As always, your feedback in the RFC process will be appreciated.

[whitequark]

Thank you for submitting this pull request. We have iterated on the proposed design; as such, support for this feature has been accepted via RFC 50 and implementation is tracked in issue #1186. Since it significantly differs from this PR I'll go ahead and close it, noting that we may still choose to reuse some of the implementation.