`n` functions, each a cascade of cross-type conversions: scalar widening (int/uint -> float), splats, mixed initializer lists (per-element coercion), vector compose/truncate, and explicit narrowing. Stresses the coercion / conversion-cost engine in semantic checking rather than operator overloading. Scaling null: n scales conversion sites against a FIXED type set; ideal cost is O(n).
bucket: typecheck · mode: module · flags: (none)
compileInner split into phase buckets (named leaves + (self) residuals) stacked across the sweep sizes — the top edge is compileInner, so you can see which phase drives the scaling.
floor-subtracted power-law fit (t − floor) = a·Nk; floor = the minimal workload (fixed per-compile cost), k the global exponent, top-2× the local high-end doubling ratio.
| N range | floor (ms) | k (work) | fit R² | t(Nmin) | t(Nmax) | top-2× |
|---|---|---|---|---|---|---|
| 300–2400 | 11 | 1.00 | 1.000 | 61 | 418 | 1.93× |
compileInner grows by 357 ms across the sweep; the mutually-exclusive phase buckets below partition that growth exactly (no nested-timer double counting). × lin is the same metric as the top-level panels, per bucket: the end point vs a linear expectation anchored to the bucket's share of the minimal floor and fitted on the low-N half — 1.0 = grew exactly linearly, >1 bends up. The super-linearity lives where × lin (and k) are red.
| bucket | t@N=300 | t@N=2400 | Δ ms | share | × lin | ∝Nk |
|---|---|---|---|---|---|---|
| SemanticChecking | 35 | 225 | +190 | 53% | 0.96× | 0.96 |
| generateIR | 19 | 143 | +124 | 35% | 1.00× | 0.99 |
| parseTranslationUnit | 3 | 26 | +23 | 6% | 1.07× | 1.05 |
| generateOutput (self) | 4 | 24 | +20 | 6% | 1.23× | 1.22 |
Near-constant (≤2% of growth each): compileInner (self) (0→0 ms), frontEndExecute (self) (0→0 ms).
| N | compileInner | SemanticChecking | frontEndExecute |
|---|---|---|---|
| 300 | 61 | 35 | 57 |
| 600 | 112 | 62 | 106 |
| 1200 | 216 | 117 | 204 |
| 2400 | 418 | 226 | 394 |