[When] Add support for unflattened tuples

[rsetaluri]

Allows compiling when blocks to MLIR with flatten_all_tuples=False.

[rsetaluri]

@leonardt this is basically #1267 rebased on #1322.

[codecov-commenter]

Codecov Report

Patch coverage: 97.29% and project coverage change: -0.03% ⚠️

Comparison is base (4475c36) 86.12% compared to head (726e71e) 86.09%.

❗ Current head 726e71e differs from pull request most recent head 7c86be9. Consider uploading reports for the commit 7c86be9 to get more accurate results

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1324      +/-   ##
==========================================
- Coverage   86.12%   86.09%   -0.03%     
==========================================
  Files         171      171              
  Lines       17869    17897      +28     
==========================================
+ Hits        15389    15409      +20     
- Misses       2480     2488       +8     

Files Changed	Coverage Δ
magma/backend/mlir/translation_unit.py	92.70% <50.00%> (-0.64%)	⬇️
magma/backend/mlir/hw.py	98.56% <83.33%> (-0.46%)	⬇️
magma/backend/mlir/compile_to_mlir_opts.py	100.00% <100.00%> (ø)
magma/backend/mlir/hardware_module.py	96.84% <100.00%> (+0.01%)	⬆️
magma/backend/mlir/magma_ops.py	100.00% <100.00%> (ø)
magma/backend/mlir/mlir_compiler.py	95.45% <100.00%> (ø)
magma/backend/mlir/when_utils.py	99.41% <100.00%> (+0.05%)	⬆️
magma/backend/mlir/xmr_utils.py	91.93% <100.00%> (+0.13%)	⬆️
magma/inline_verilog.py	98.11% <100.00%> (+0.89%)	⬆️
magma/passes/when.py	100.00% <100.00%> (ø)
... and 2 more

... and 1 file with indirect coverage changes

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.