Repository navigation
fix: replace the unmaintained YAML parser - #2279
Merged
Merged
Conversation
Replace serde_yaml and unsafe-libyaml with deserialize-only serde-saphyr to clear RUSTSEC-2024-0449 without suppressing advisories. Preserve language ordering and duplicate-language diagnostics, strict schema fields, YAML 1.2 booleans, and surgery operation validation for non-finite values. Add regressions for duplicate fields, numeric spellings, and all six shipped surgery examples. Refs #2278.
inureyes
added
status:review
Cover non-finite add coefficients and interpolation ratios, including the deliberately accepted quoted canonical special values. Record the parser migration and completed validation in bilingual reports; the added test execution and Clippy gate remain pending at report preparation. Refs #2278.
6 tasks done
inureyes
added a commit
that referenced
this pull request
Oct 12, 2026
Merge main at fc45284 so PR #2277 uses the maintained safe-Rust YAML parser and no longer inherits the RUSTSEC-2024-0449 cargo-deny failure. The split-mtp source remains unchanged. Validation: configured version, kernel, Llama, ROCm, Python, and WebUI contracts pass; cargo deny check, formatting, and diff checks pass. Refs #1784, #2279
5 tasks done
inureyes
added a commit
that referenced
this pull request
Oct 23, 2026
The Activity performance gate judged five alternating pairs by a bare median against the 2% budget and also failed any mode whose baseline CV exceeded 5%. On `lablup-dgxspark21` the per-pair noise (SD 3.6 points, kurtosis 12 to 25) makes both rules fire on unchanged code. Each mode now samples until an interval decides, and a mode that cannot be decided fails as noise, not as a slowdown. ## The rule - Per pair: `(off - on) / off * 100`, alternating order as before. Point estimate: the median. Interval: the distribution-free sign-test interval `[x(k), x(n+1-k)]` at 1% per tail (98% two-sided). It needs no variance estimate and no random bootstrap, so JavaScript and Python agree bit for bit. - `investigate` (slowdown): the interval lies wholly above the 2% budget. - `within-target`: it does not, and its upper bound is at most 4%. - Otherwise add 2 pairs, from 12 up to 40; still undecided at 40 is `inconclusive`, which fails with a message naming measurement noise. - Baseline CV and the paired range stay in the summary as diagnostics and decide nothing. The 2% budget is unchanged. - The 4% bound is the precision a pass must reach, not a looser budget: a 4% bound gives 0.15% null inconclusive per mode, while 3% gives 2.4%, which breaks the 1% false-failure target (table below). - All constants live in `webui/scripts/activity-performance-policy.json`. The verifier re-derives every mode from the raw samples with a Python port (`scripts/webui/activity_performance_verdict.py`) and derives the sample and preflight counts from `paired_runs`. The summarizer has no samples, so it checks each summary's internal consistency under the same policy. Both languages are tested against `webui/scripts/activity-performance-fixtures.json`. ## Resampling estimate Noise model: all 636 complete diagnostics artifacts on this runner (2026-09-15 to 2026-10-22, 6,360 pairs), centred on a true 0% change, iid pairs drawn from that pool plus a per-run shift N(0, 0.83^2). 0.83 is the between-run SD estimated from the post-#2119 runs; one-visible and two-visible means in the same run are uncorrelated (r = -0.07). 20,000 simulated runs per row, sequential looks included: | True degradation | investigate | inconclusive | within-target | mean pairs | |---|---|---|---|---| | 0% | 0.06% | 0.15% | 99.79% | 13.5 | | 2% | 11.4% | 7.0% | 81.6% | 20.4 | | 4% | 78.6% | 8.5% | 12.9% | 21.6 | | 5% | 96.0% | 2.3% | 1.6% | 17.1 | | 6% | 99.67% | 0.27% | 0.07% | 14.1 | | 8% | 100% | 0% | 0% | 12.4 | - False failure on a true 0% change: 0.21% per mode, about 0.4% per CI job (two modes), about 0.6% in `full` mode. With iid pairs and no run shift it is below 0.01%. - Detection: at least 95% for true effects of 5% and above, 99.7% at 6%. At 4% it is only 79%, so the injected slowdown for acceptance is sized at about 6%. - Rejected alternatives: a 3% upper bound (2.4% null inconclusive); passing only when the interval lies wholly below 2% (far more pairs); mean with a t-interval and Hodges-Lehmann (similar power, but the mean is hurt by the heavy tails and HL assumes a symmetry the data lacks, skew -0.7). - Assumption: the guarantee holds while run-level shifts stay near the measured 0.8 points. A shift SD of 1.5 gives 2.4% false `investigate`; keeping such shifts out is the job of the #1949 host-quiet precheck. - Warmup: there is no first-pair bias (pair 0 mean -0.50 vs -0.48 overall, and no order effect), so no leading pairs are discarded. The existing warmup request runs 1.6% slower than the first baseline, and that request is already excluded. ## Wall time Measured per pair on 153 recent runs: one-visible 7.2 s median (7.7 s max), two-visible 8.4 s (9.2 s max), plus about 34 s before the first pair. In CI (visible-only-headed), the worst case at 40 pairs per mode is about 12 minutes, and the expected null run is about 4 minutes (about 2 today). `full` mode adds a hidden block, so its worst case is about 20 minutes, inside the 1800 s activity timeout. `ci.yml` timeouts are unchanged. ## Fixture provenance The fixture copies the per-sample logs of the `webui-activity-performance-diagnostics-Linux` artifacts, none of which had expired: 35423171315 a1, 35425297407 a1/a2/a3, 35425050365 a1, 35421166203 a1, 35567777700 a1 (#1944), 35588958676 a1 (#1946) and 38265665905 a1/a2 (#2279). At their recorded five pairs, every one of them returns `continue` (no verdict), never `investigate`. Their concatenated pairs decide `within-target` at 12 pairs in both modes. The same pairs with the Activity arm 6% slower decide `investigate`, and spread-amplified or drifting (CV 12%) variants exercise `inconclusive` and CV-only cases. ## Test plan - [x] `pnpm --dir webui run unit` and `pnpm --dir webui run lint` - [x] `make verify-webui-helper-tests`, including the new `activity_performance_verdict_tests.py` - [x] `python3 scripts/webui/verify_activity_performance_tests.py` and `summarize_activity_evidence_tests.py` - [x] `make verify-versions verify-env-var-docs verify-python-tooling` - [x] The new gate inside the real `WebUI installed artifact` job on this PR, run [39853154267](https://github.com/lablup/mlxcel/actions/runs/39853154267): one-visible within-target (median -0.15%, 98% interval [-1.83%, 1.25%]) and two-visible within-target (median -0.76%, [-2.64%, 0.47%]), both at 12 pairs, with the verifier and summarizer accepting the evidence; the verifier took 3 min 42 s. ## Acceptance runs on `lablup-dgxspark21` - Five consecutive green `WebUI installed artifact` runs of this tree (head a7cc03f): run [39853154267](https://github.com/lablup/mlxcel/actions/runs/39853154267) attempts 1 to 5 (jobs 120084449655, 120097378392, 120110916929, 120116442394, 120121248222). - Injected slowdown, red with `investigate` in both modes: run [39870659579](https://github.com/lablup/mlxcel/actions/runs/39870659579) on throwaway draft PR #2554 (closed, never merged). one-visible median 4.58%, 98% interval [3.89%, 5.68%]; two-visible median 8.05%, interval [7.11%, 9.62%]; 12 pairs each. The injection sleeps 400 us per decode tick for each open `/ui-api/v1/events` stream, which the Activity page holds open exactly while visible. - The true effect was not measured beforehand on a separate harness (a local build on the shared CI host was not permitted); it is characterised only by the gate's own median and interval above, both inside the 4 to 8% target for one-visible. - The issue's suggested injection does not work on GB10. A per-poll CPU burst of 80 spinning threads (run 39862888632) raised the runnable tasks to 46 to 87 in the visible arms, but decode stayed at 152 tok/s in both arms. A delay keyed on Activity polls (runs 39866915190 and 39868894929) also measured 0%: the instrumented run showed that no operations poll ever landed inside a decode window, because each poll was answered the moment a request finished. Closes #1925
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Replace the CLI and surgery YAML parser with deserialize-only
serde-saphyr 1.3.0, removingserde_yamlandunsafe-libyamlfrom the lockfile and resolving RUSTSEC-2024-0449 without an advisory exception.Preserve ordered language priorities,
DuplicateLanguageCode, duplicate schema-field rejection, YAML 1.2 booleans, and ordinary merge-key handling. Recover canonical non-finite values from Serde's tagged/untagged buffering so existing operation-specific finite-value validation remains effective. Quoted canonical.nan/.inf/-.infnow follow the same handling as their unquoted forms; malformed-input diagnostic formatting comes from the new parser.Validation: all 34 focused language-bias tests and all 28 surgery configuration tests pass, including all six shipped surgery examples.
cargo deny checkpasses all four checks; formatting and diff checks pass. Repository contract checks also pass (123 Python tooling tests and 49 WebUI fixtures). The alpha/ratio and quoted-special regressions pass. CI Clippy, CPU-only link, and cargo-deny pass on dd36abb; the duplicate local CUDA Clippy rebuild was stopped after CI succeeded.Closes #2278