fix[codegen]: remove silent fallback for bool kwargs in venom pipeline

[banteg]

What I did

Fixed Venom builtin lowering for boolean keyword arguments that semantic analysis already accepted as folded constants.

This covers abi_decode(..., unwrap_tuple=CONST_FALSE) and print(..., hardhat_compat=CONST_TRUE).

How I did it

Updated the Venom bool kwarg extraction helpers in the ABI and debug-print builtins to use kw_node.reduced() before checking for boolean or integer literals.

If a provided bool kwarg still cannot be recognized after reduction, the helper raises a compiler panic instead of silently falling back to the default. That keeps the absent-kwarg default path, but avoids masking unexpected accepted input.

How to verify it

• uv run --python 3.12 pytest tests/unit/compiler/venom/test_builtin_kwargs.py tests/functional/builtins/codegen/test_abi_decode.py::test_abi_decode_folded_unwrap_tuple_kwarg tests/functional/syntax/test_print.py::test_print_folded_hardhat_compat_kwarg -q
• uv run --python 3.12 ruff check vyper/codegen_venom/builtins/abi.py vyper/codegen_venom/builtins/misc.py tests/unit/compiler/venom/test_builtin_kwargs.py tests/functional/builtins/codegen/test_abi_decode.py tests/functional/syntax/test_print.py

Commit message

_get_bool_kwarg in abi.py and misc.py silently returned the default
value when the kwarg node was not a NameConstant or Int literal. this
masked two bugs at once: folded constants like UNWRAP=False (a Name, not
NameConstant) were ignored, and any unexpected node type was swallowed
without error.

call kw_node.reduced() before the type checks so folded constants
resolve to their actual literal form. replace the silent fallback
with a CompilerPanic to make future issues immediately visible.

Description for the changelog

Fix Venom builtin lowering for folded constant boolean keyword arguments.

[charles-cooper]

i think it's worse that we are falling back to something rather than not unfolding. we should remove the return default fallback line in these builtins (and actually all builtins as applicable)

[banteg]

Updated. The helpers still return the default when the kwarg is absent, but a provided bool kwarg that cannot be interpreted after folding now raises CompilerPanic instead of falling back to the default.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 92.81%. Comparing base (e57e6ac) to head (6e25d45).
⚠️ Report is 1 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #4984      +/-   ##
==========================================
+ Coverage   92.71%   92.81%   +0.09%     
==========================================
  Files         187      187              
  Lines       27626    27629       +3     
  Branches     4776     4776              
==========================================
+ Hits        25614    25644      +30     
+ Misses       1363     1336      -27     
  Partials      649      649              

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[banteg]

Updated to use kw_node.reduced() in both bool-kwarg helpers. I also added direct unit coverage for both the folded path and the unreduced/error path, so the Codecov patch complaint should clear on the new run.

[charles-cooper]

Suggested change

	raise CompilerPanic(f"unfoldable boolean kwarg: {kwarg_name}")
	raise CompilerPanic(f"unfoldable boolean kwarg: {kwarg_name}", kw_node)

[charles-cooper]

Suggested change

	raise CompilerPanic(f"unfoldable boolean kwarg: {kwarg_name}")
	raise CompilerPanic(f"unfoldable boolean kwarg: {kwarg_name}", kw_node)

[banteg]

Applied the suggestion in both bool-kwarg helpers: the CompilerPanic now carries the offending kwarg node. Validated with the focused bool-kwarg, abi_decode, print tests, and ruff check.

[banteg]

Correction: #4984 was merged while this follow-up was being pushed, so the branch update is not part of the merged PR.

I opened the audited follow-up as #5001: #5001

Summary of the follow-up:

• removed every Venom _get_kwarg_value copy across abi, create, system, and misc
• added shared reduced-kwarg helpers so builtin callers receive kw.value.reduced()
• made required literal kwargs fail closed with CompilerPanic instead of silently falling back to defaults
• kept runtime fallback only for create_from_blueprint(code_offset=...), where runtime values are allowed
• found and fixed an adjacent ABIEncode.fetch_call_return raw ensure_tuple read that ICEd on ensure_tuple=CONST before codegen

Validated with focused Venom helper, abi_encode, abi_decode, print, constants, create, raw_call, and send tests; ruff check is clean on all touched files.