mindspore-ai / mindspore-ai/hyper-parallel
[Bug]: MegaMoe Level1 在无 Git 元数据的 CI 目录中写入证据失败
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 53
- Forks
- 63
- Avg merge
- 23h 45m
- Merged PRs (30d)
- 63
Description
🐛 Describe the bug
MegaMoe Level1 在 CI 的归档测试目录中完成计算和校验后,写入测试证据时执行 git rev-parse HEAD。该目录不含 Git 元数据,命令返回 128,导致 rank 0 退出失败。
受影响的两个场景为共享资源验证和原默认接口验证;其余 rank 已通过。公共证据采集位于 tests/torch/multicore/_mega_moe_utils.py。
最小复现:
import subprocess
import tempfile
with tempfile.TemporaryDirectory() as root:
subprocess.check_output(["git", "rev-parse", "HEAD"], cwd=root, text=True)
错误包含 fatal: not a git repository,并抛出 subprocess.CalledProcessError,退出码为 128。在 Git worktree 内执行时能够获取 SHA,因此通常不会触发该问题。
Expected behavior
源码 SHA 应作为可降级的辅助证据。没有 Git 仓库或 Git 不可用时,记录 source_sha: null 和失败原因,继续保留版本、文件哈希及各 rank 的验证结果。正常仓库仍应记录实际 SHA,其他证据采集错误和测试断言应正常报告。
Additional context
修复 PR:https://gitcode.com/mindspore/hyper-parallel/pull/1368
已使用真实 Git 命令复现原错误,并在模拟设备信息及 2/4 rank 汇总的证据写入回归中验证降级行为;这项回归不代表重新执行了 NPU ST。
Environment info
- CI:Ascend910B,测试从不含 Git 元数据的归档目录运行。
- 本地最小复现:cann920 环境中的 Python 3.10.19。
schema_version: 1
source: gitcode
gitcode_repo: mindspore/hyper-parallel
gitcode_issue: 375
source_url: https://gitcode.com/mindspore/hyper-parallel/issues/375
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start in tests/torch/multicore/_mega_moe_utils.py and review the evidence-collection path that runs git rev-parse HEAD. Reproduce it from a Git-free temporary directory, then run the evidence-writing regression covering simulated device information and 2/4-rank aggregation; done means missing Git records source_sha as null with the failure reason while preserving other evidence and normal SHA capture.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- git, python
- Domain
- testing-qa
- Issue type
- Bug
- Difficulty
- 2/5
- Estimated time
- 1-3 hours
- Activity status
- Stale
- Clarity
- Clearly specified
- Newbie friendliness
- 25/100