Skip to content

ci: make clang-format non-blocking in lint workflow - #51

Merged
balbasty merged 1 commit into
mainfrom
claude/lint-nonblocking
Aug 2, 2026
Merged

ci: make clang-format non-blocking in lint workflow#51
balbasty merged 1 commit into
mainfrom
claude/lint-nonblocking

Conversation

@balbasty

Copy link
Copy Markdown
Collaborator

Summary

  • Add continue-on-error: true at the clang-format job level in .github/workflows/lint.yaml so the job keeps running and reporting on every PR but no longer fails the overall required check.
  • codespell is untouched and remains blocking.
  • The diff-scoping logic (merge-base diff, git-clang-format-18, extension list) and the codespell job are untouched.
  • Appended one line to the top-of-file comment block noting that clang-format is now report-only via continue-on-error, without altering the existing explanation of why it's diff-scoped.

Test plan

  • Confirm the Lint workflow runs on this PR.
  • Confirm codespell still runs and blocks as before.
  • Confirm clang-format runs and, if it finds violations, is shown as a non-blocking failure (does not fail the required check) rather than a hard failure.

Workstream: claude-jitfields-to-fastfields

🤖 Generated with Claude Code


Generated by Claude Code

@balbasty
balbasty merged commit 93d46e3 into mainAug 2, 2026
3 checks passed
@balbasty
balbasty deleted the claude/lint-nonblocking branch August 2, 2026 07:52
@balbastyClaude

Copy link
Copy Markdown
CollaboratorAuthor

Review summary (triage, jitfields-to-fastfields): merged as squash. Same mechanical change as the other 5 repos in this batch (continue-on-error: true on the clang-format job only, codespell untouched) — see fastfields-kernels#65 for the full review note.

Workstream: claude-jitfields-to-fastfields


Generated by Claude Code

balbasty added a commit that referenced this pull request Aug 5, 2026
Picks up fastfields-cpu-lib#82, which bumps fastfields-cpu-impl to #69's
fix for #51 (relax_bending_rls_/relax_bending_jrls_ used a 2*niter
colour-loop bound instead of pow<ndim>(3)*niter, leaving most of the
field unrelaxed for bending-order RLS/JRLS relaxation) and adds a
deterministic regression test for it.
part of #33
Co-authored-by: Claude <noreply@anthropic.com>
Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant

@balbasty