Self-host: gen-c nests C comments in a nested repeat initializer and zeroes a repeated row #7353

Profile (read first)

You are working as Gamma (Compiler Core) on epic #5980: rewrite t27c in t27 itself.
The self-hosted core (specs/compiler/core/t27core.t27) must write exactly the bytes
t27c gen-c writes, so every gen-c defect is a defect the core has to refuse instead of learn.
Rules for this task:

• Change gen-c only inside the Boundary below. gen/ is generated: never hand-edit it.
• Build with cargo build --release -p t27c and use ./target/release/t27c from the repository
  root, not a t27c on PATH.
• Write the regression test first and watch it fail on the unfixed compiler. Then fix, and
  revert the fix once to watch the test go red again. Say in the PR that you did both.
• Any spec that stops compiling must be listed in the PR by path with its fix. A silent corpus
  change is a failed task.
• The PR body says Closes # plus this issue's number and Part of #5980. ASCII only in source files.

Context

gen-c writes a repeat initializer as a C comment followed by {0}. For a nested repeat the
comment contains another comment:

var g: [2][3]u8 = [_][3]u8{ [_]u8{0} ** 3 } ** 2;
=> static uint8_t g[2][3] = /* repeat: { /* repeat: { 0 } ** 3 */ {0} } ** 2 */ {0};

The inner */ closes the outer comment, so the C does not compile (expected ',' or ';'). The
form is the one the corpus uses for 2-D globals ([16][16]u8, 6 uses). Two more losses on the same path:

• a repeated non-zero row, [_][2]u8{ [_]u8{1, 2} } ** 2, becomes {0} and compiles: every element reads 0;
• a non-zero inner repeat, [_][K]i64{ [_]i64{7} ** K } ** R, puts the GNU range
  { [0 ... (K) - 1] = 7 } at the outer level ("array index range in initializer exceeds array bounds").

An explicit row list, [_][2]u8{ [_]u8{1, 2}, [_]u8{3, 4} }, is lowered correctly to
{ { 1, 2 }, { 3, 4 } }. The core (P1 slice 3) accepts that form and refuses an outer ** on a nested array with E_ARRAY.

Reproduce

cargo build --release -p t27c
printf 'module m;\nvar g: [2][2]u8 = [_][2]u8{ [_]u8{1, 2} } ** 2;\ntest t { assert(g[1][1] == 2); }\n' > /tmp/n.t27
./target/release/t27c gen-c /tmp/n.t27 > /tmp/n.c; echo "gen-c exit $?"
cc -std=c99 -DT27_TEST_MAIN /tmp/n.c -o /tmp/n && /tmp/n; echo "run exit $?"

Today: gen-c exit 0, then the run aborts on assert(g[1][1] == 2). With [_][3]u8{ [_]u8{0} ** 3 } ** 2 cc fails instead.

Acceptance criteria

• 1. The Reproduce run prints All 1 tests passed. and exits 0
• 2. [_][3]u8{ [_]u8{0} ** 3 } ** 2 and [_][K]i64{ [_]i64{7} ** K } ** R compile with cc -std=c99 and hold the right values
• 3. cargo test --release -p t27c --test genc_nested_repeat passes

Boundary

• bootstrap/src/compiler.rs
• bootstrap/tests/genc_nested_repeat.rs

User Scenarios & Testing

• Given a global var g: [16][16]u8 = [_][16]u8{ [_]u8{0} ** 16 } ** 16;,
  when ./target/release/t27c gen-c lowers it,
  then the C compiles and every element is 0; a repeated non-zero row keeps its values.

Requirements

• FR-001: gen-c MUST NOT nest a C comment inside a C comment.
• FR-002: a repeated row MUST keep its element values; a value MUST NOT become {0} unless it is zero.
• FR-900: cargo test --release -p t27c --test genc_nested_repeat MUST pass, and MUST fail with the fix reverted.
• FR-901: cargo test --release -p t27c --test core_selfhost and the corpus suite MUST still pass.

Part of #5980