Repository navigation
tests: sample the lower-feed shape bound robustly - #644
luisleo526 wants to merge 1 commit into
Conversation
…rline one once test_native_intrabar_lower_lookup compares the time of a 4x lower feed with a 6.0 bound. On a shared CI runner each leg is the best of three samples with independent scheduler noise, so a true ratio near 4 occasionally read 6.08 and failed main's post-merge run once (4000 bars 0.1140 s, 16000 bars 0.6934 s), then passed on a re-run. Take the best of five user-CPU samples per leg, and give a ratio in [6, 12) exactly one fresh measured pair, keeping the smaller ratio. A ratio of 12 or more (a quadratic lookup measures about 16) is judged as measured with no retry, and the bound stays 6.0 with CHECK(ratio < kShapeBound) as the only verdict, so a persistently high ratio still fails. The decision is a small header-only helper in tests/ratio_timing.hpp, exercised inside the same test binary on fixed sample sets: the observed noisy pair with a quiet re-measure passes, the same pair with a high re-measure still fails, a quadratic pair is not retried, and a clean linear pair is not retried. Co-Authored-By: Claude Sonnet 5.5 <[email protected]>
![graphify-labs[bot]](https://avatars.githubusercontent.com/in/4356270?s=60&v=4){kind=small}
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 1 advisory finding(s) below merit a look before merge.
Not checked: tests were not run; no formal proof of the changed code.
Graphify review — findings
Tightens the lower-feed linearity timing test against shared-runner noise: each leg now takes the best of five user-CPU samples instead of three, and ratio_with_remeasurement gives a ratio in [6, 12) one fresh measured pair and keeps the smaller of the two ratios. A ratio at or above 12 fails without a retry, and a passing ratio stands as measured, so a real quadratic regression is still caught. The retry path takes the lower ratio and treats a zero small-leg time as ratio 0, which is the spot where a bad retry measurement could slip under the bound.
Worth a look
- Degenerate retry (small==0) yields ratio 0 and silently passes the bound —
tests/ratio_timing.hpp:47· Escalate · medium- NOT verified (no proof, no reproducing execution) — needs human review
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 46 functions depend on the 46 functions this change touches.
Health — grade A; 1 existing hotspot(s) in the area this change touches (pre-existing, not introduced here):
main()— 0 callers, 6 callees (medium)
Verification — 46 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 46 function(s) in the blast radius were not formally verified this run
|
Superseded by #647 (same change on current main, folded with its sibling). Branch kept. |
Makes the lower-feed shape-bound check in
test_native_intrabar_lower_lookuprobust to a noisy shared runner without weakening it.The test compares a 4x-bars run against a base run and asserts the ratio stays under the fixed shape bound (6.0, unchanged). One measured run on a shared runner exceeded it by 1.3% while the same test passed elsewhere in the same CI run. The check now takes the best of N samples per arm, plus a bounded re-measure (
tests/ratio_timing.hpp), so a single noisy sample cannot fail it. A real super-linear cost still fails: fixed-sample rows include a persistent-high pair that must still fail and a quadratic pair that must not be retried into a pass.Test sources only.
🤖 Generated with Claude Code