test(verilog): terminate the r_ca_2 fixture's call statement

[gHashTag]

Closes #2399

What

r_ca_2_synthetic_no_comment_only_call_argument never reaches its subject. Its fixture
fails to parse:

Error: Compile error: parse error in fn 'driver' near line 8:
       unexpected token after expression statement: KwReturn

The test's name is wrong about the cause

Isolated with four probes:

fixture	result
consume([1,2,3,4]) then return	parse error
consume([1,2,3,4]); then return	OK
consume(7) then return	parse error
consume([1,2,3,4]) as last statement	OK

The array literal is irrelevant. The condition is an expression statement without a
terminating ;, followed by another statement. The fixture omits the semicolon.

Fix

One character. With consume([1, 2, 3, 4]); both tests in the file pass, and the test
finally measures what it names: the emitter writes consume(0 /* TODO: array literal [1,2,3,4] not yet lowered to Verilog */) — a real argument plus a comment, not a
comment standing in for one.

Note what this does not fix. Array literals in call position are genuinely not
lowered; the emitter says so in its own output and emits 0. R-CA-2 is the narrower rule
that the argument must not be comment-only, and that rule holds.

Related measurement, offered as context and not as a claim of breakage

Sweeping t27c parse over all 650 .t27 files under specs/: 496 parse, 154 do not.
The 496 matches the repository's own recorded baseline, so this is the known state, not a
regression.

The single largest failure kind is the same mechanism as above — 28 of the 154 report
unexpected token after expression statement. For example specs/server/session.t27:187:

assert(SessionState::Pending == 0)
assert(SessionState::Active == 1)

Whether the parser should accept newline-terminated expression statements is a language
decision, not a defect I am asserting. 717 spec files use ; line endings, so the
semicolon form is clearly the norm; the question is only whether the other form should be
rejected. Raising it rather than answering it.

I also nearly reported a much larger and much wronger number: t27c gen-verilog fails on
649 of 650, because most specs are not hardware specs and that backend legitimately
declines them. t27c corpus's own help text warns about exactly this class of measurement
— "a parser error count moves three orders of magnitude from one character and RISES when
a real defect is fixed (T119)".

Why it stayed red

Target 65 of 73; cargo test -p t27c --tests stops at the first failing target (#2382).
This is 1 of the 7 in #2386, leaving 1.

Related: #2386, #2382.

[github-actions]

📓 NotebookLM Notebook linked to this PR

• Notebook ID: b83263109fb055dc
• URL: https://notebooklm.google.com/notebook/b83263109fb055dc

This notebook contains session context, decisions, and artifacts for this work.

[github-actions]

PR Dashboard

Generated at: 2026-08-22 10:48:03 UTC

• test(verilog): terminate the r_ca_2 fixture's call statement #2400 test(verilog): terminate the r_ca_2 fixture's call statement | Created: 2026-08-22 | Updated: 2026-08-22 | 17 failing / 26 total
• corpus: freeze the damage, patch one class at a time, and say where one rule stops #2161 corpus: freeze the damage, patch one class at a time, and say where one rule stops | Created: 2026-08-14 | Updated: 2026-08-21 | 0 failing / 9 total
• Unbreak FPGA CI and correct two unbacked hardware claims #2081 Unbreak FPGA CI and correct two unbacked hardware claims | Created: 2026-08-11 | Updated: 2026-08-20 | 5 failing / 27 total

Summary

Status	Count
Total Open PRs	3
PRs with Failing Checks	1
PRs with All Checks Green	2
READY	1
FAILING	1
PENDING	0

Seal Status

• ⚠️ STALE -- sha256(compiler.rs)=fba627661c36 != manifest seal=87e5cbd3ad94.
  The committed NMSE numbers were certified against an older compiler.rs.
  Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).

[github-actions]

📓 NotebookLM Notebook linked to this PR

• Notebook ID: b83263109fb055dc
• URL: https://notebooklm.google.com/notebook/b83263109fb055dc

This notebook contains session context, decisions, and artifacts for this work.

[github-actions]

PR Dashboard

Generated at: 2026-08-22 11:04:38 UTC

• feat(ci): conformance vectors execute against generated RTL #2403 feat(ci): conformance vectors execute against generated RTL | Created: 2026-08-22 | Updated: 2026-08-22 | 1 failing / 1 total
• test(verilog): terminate the r_ca_2 fixture's call statement #2400 test(verilog): terminate the r_ca_2 fixture's call statement | Created: 2026-08-22 | Updated: 2026-08-22 | 16 failing / 26 total
• corpus: freeze the damage, patch one class at a time, and say where one rule stops #2161 corpus: freeze the damage, patch one class at a time, and say where one rule stops | Created: 2026-08-14 | Updated: 2026-08-21 | 0 failing / 9 total
• Unbreak FPGA CI and correct two unbacked hardware claims #2081 Unbreak FPGA CI and correct two unbacked hardware claims | Created: 2026-08-11 | Updated: 2026-08-20 | 5 failing / 27 total

Summary

Status	Count
Total Open PRs	4
PRs with Failing Checks	1
PRs with All Checks Green	3
READY	1
FAILING	1
PENDING	0

Seal Status

• ⚠️ STALE -- sha256(compiler.rs)=ee73b41dba8f != manifest seal=87e5cbd3ad94.
  The committed NMSE numbers were certified against an older compiler.rs.
  Run scripts/reseal-check.sh locally for the two-step reseal command (advisory; not a merge gate).