Repository navigation
golden: name and pin the x86_64 Linux divergences (#69) - #90
Merged
Merged
Conversation
DOCWATCH runs compare.py only for its counts, so the 50-54 golden cases that fail on the Linux runner had no names to pin. An always() step reruns compare.py against the staged artefact and prints its FAIL and SUMMARY lines; golden.log joins the uploaded gate log. Measurement only; it never fails the job.
Contributor
Gate on CI (issue #24)Full gate: 123 suites = 115 run on at least one CI runner + 8 local-only. The local gate (
Local-only: verified by no CI runner (8)
Suites that verified nothing, or failed, on a runner
Every suite, every runner
A SKIP is unverified on that runner, never passed. UNVERIFIED means the suite ran and every one of its tests skipped by name. |
The names come from PR #90's CI run. On x86_64 Linux only, a FAILING case whose exact name is pinned prints X86KNOWN with its issue and is counted in a new SUMMARY field x86_known, not in failed; a pinned case that passes is a plain pass, an unpinned failure stays a FAIL, and off x86_64 Linux the table is never read. DOCWATCH's golden_cases_failed keeps reading failed; golden_x86_known is a new count source. The self-test does not consult the table. Groups: nll_loss_forward 24, matmul 12, kaiser_window 8, gelu 4 (x86 upstream gelu(+inf) is NaN), sort/argsort ties 3 (#81), sdpa flash 2, add bf16 alpha 1. GOLDEN.md section 9 records the 47-54 run-to-run instability. test_x86pins.py (5): every pin names a built case, no whole-op pins, classification on and off x86. Nullified four ways (dead pin, dropped platform guard, classify always FAIL, whole-op match): each turns it red. Here: SUMMARY 12103/12103, 0 failed, x86_known=0.
b-re-w
changed the title
3 tasks
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
1 participant
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.
Refs #69. Adds an always() step to the Linux gate that runs
tests/golden/compare.pyvisibly and uploadsgolden.log, so the x86_64 divergences (the #40 classes) can be pinned by name in the next step. Measurement only: the step exits 0.test_cigate.pyandtest_publish.pypass locally.