Skip to content

Fix dangling aiter.ops.flydsl.utils import breaking test collection - #5225

Closed
coderfeli wants to merge 1 commit into
mainfrom
fix/flydsl-utils-import-test
Closed

coderfeli wants to merge 1 commit into
mainfrom
fix/flydsl-utils-import-test

Conversation

@coderfeli

@coderfeli coderfeli commented Sep 3, 2026 •

Copy link
Copy Markdown
Collaborator

Summary

op_tests/test_fused_moe_s1_fuse_fp8q.py imports a module that no longer exists, aborting collection of the whole file on any shard that picks it up:

File "/workspace/op_tests/test_fused_moe_s1_fuse_fp8q.py", line 24, in <module>
    from aiter.ops.flydsl.utils import is_flydsl_available
ModuleNotFoundError: No module named 'aiter.ops.flydsl.utils'

…lection

#5116 removed aiter/ops/flydsl/utils.py and swept every caller that
existed in main at the time. op_tests/test_fused_moe_s1_fuse_fp8q.py
landed afterwards via #4994, whose CI base predated the removal, so it
still imports the deleted helper. The two PRs touch disjoint paths, so
git merged them without conflict and neither CI run could see the break.

The import is module scope and feeds pytestmark, so collection aborts:

  File "op_tests/test_fused_moe_s1_fuse_fp8q.py", line 24
    from aiter.ops.flydsl.utils import is_flydsl_available
  ModuleNotFoundError: No module named 'aiter.ops.flydsl.utils'

This fails the whole file on any shard that picks it up (seen on
MI300X shard 5/8). Main has had no completed Aiter Test run since
#4994 landed, so it went unnoticed.

Drop the import and the availability skipif, matching how #5116 handled
the same guard in op_tests/flydsl_tests/test_flydsl_moe_a8w4.py. FlyDSL
is a pinned hard dependency (flydsl==0.3.2 in requirements.txt), which
is why #5116 retired the probe; the remaining gfx950 gate is unchanged.

Test plan:
- pytest op_tests/test_fused_moe_s1_fuse_fp8q.py -> 2 passed on gfx950
  (previously ModuleNotFoundError at collection)
- No references to aiter.ops.flydsl.utils or is_flydsl_available remain
- black --check, compileall clean

Co-Authored-By: Claude <noreply@anthropic.com>
@coderfeli
coderfeli requested a review from a team September 3, 2026 04:07
@coderfeli coderfeli changed the title [FlyDSL] Fix dangling aiter.ops.flydsl.utils import breaking test collection [CI] hot-fix dangling aiter.ops.flydsl.utils import breaking test collection Sep 3, 2026
@github-actions

github-actions Bot commented Sep 3, 2026

Copy link
Copy Markdown
Contributor

🏷️ CI Guide

Runs automatically on every PR:

  • ✅ Pre-checks (submodule verification, code formatting)
  • ✅ Aiter op tests (gfx942 + gfx950)
  • ✅ Triton tests on MI35X (only when aiter/ops/triton/** or related paths are changed)

Extended tests (opt-in via labels):

Label Tests
ci:gfx1250-ffm-triton Run the five-shard gfx1250 FFM Triton test suite
ci:triton-300x Run an additional Triton test job on MI300X in PRs; main branch always runs both MI35X and MI300X
multigpu Aiter multi-GPU tests on the 8-GPU runner
ci:sglang SGLang integration tests: DeepSeek-R1-MXFP4 accuracy, Qwen 3.5 accuracy
ci:atom ATOM benchmark: DeepSeek-R1-0528, GPT-OSS-120B
ci:atom_full ATOM accuracy suite for PR and main models from ATOM models_accuracy.json
ci:vllm vLLM benchmark: GPT-OSS-120B, DeepSeek-R1-0528, Kimi-K2.5
ci:all All standard extended tests (excludes ci:atom_full)

Only add ci:atom_full for FlyDSL or Triton upgrades.
Add labels via the sidebar or gh pr edit 5225 --add-label <label>

PR title tags & labels:
Component tags ([Triton/Gluon], [HIP], [CK], [ASM], ...) are added to the PR title and as PR labels automatically from the changed files and re-synced on every push — change-type tags like [fix]/[Perf], op tags like [MLA], and human labels (ci:*) are left untouched. Add the no-auto-title label to opt this PR out.

@github-actions github-actions Bot changed the title [CI] hot-fix dangling aiter.ops.flydsl.utils import breaking test collection Fix dangling aiter.ops.flydsl.utils import breaking test collection Sep 3, 2026
@valarLip

valarLip commented Sep 3, 2026

Copy link
Copy Markdown
Collaborator

reverted the bad one

@valarLip valarLip closed this Sep 3, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants