Skip to content

Daily: make eBPF map reuse schema-aware - #212

Open
yunwei37 wants to merge 11 commits into
mainfrom
daily/2026-09-23-ebpf-map-reuse-semantics
Open

yunwei37 wants to merge 11 commits into
mainfrom
daily/2026-09-23-ebpf-map-reuse-semantics

Conversation

@yunwei37

Copy link
Copy Markdown
Member

Summary

  • publish the September 23 bilingual Daily Report on pinned-map state/schema compatibility
  • keep the active eBPF Deployment Compatibility and Lifecycle series at a distinct boundary from transactional whole-application upgrades
  • record current GSC/GA4 source coverage and preserve the rolling 7 eBPF / 1 Agent / 2 adjacent mix
  • make no unrelated technical SEO change because the enabled evidence does not establish a concrete defect
  • reconcile the missing September 22 merged-PR closeout record from verified deployment facts

Report boundary

The report separates kernel-visible map-definition compatibility from BTF-derived structural schema compatibility and application semantic compatibility. It develops canonical schema fingerprints, versioned map-state migration contracts, and shadow validation before a new generation receives write authority. It explicitly does not claim bpffs pins survive normal reboot.

Validation

Before merge: inspect the complete diff/generated output, require terminal-green final-head CI, inspect review threads, then squash merge. After merge: verify exact-SHA deployment, generated English/Chinese production artifacts, localized canonical/hreflang/Article metadata, sitemap entries, and add exactly one compact closeout comment.

Copilot AI lite review requested due to automatic review settings September 23, 2026 15:28

Copilot AI left a comment

Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

@yunwei37

Copy link
Copy Markdown
Member Author

Restored the two required daily sections that the SEO data validator expects in .github/seo-data/daily/2026-09-23.md:

  • ## Evidence collected
  • ## Work performed

Evidence: Validate SEO Operations run 35881944529, job 107252335720, step Validate public SEO operating data failed with exactly error: missing heading '## Evidence collected' in .github/seo-data/daily/2026-09-23.md. The contract is DAILY_HEADINGS in .github/seo-skills/scripts/validate_seo_data.py (pinned at 516e9e2d), invoked as python .github/seo-skills/scripts/validate_seo_data.py --data-root .github/seo-data (.github/workflows/seo-data-validate.yml:115-118). The sibling daily record in #213 passes the same validator by including this heading set.

Fix: commit 83c5a6f adds the missing sections without touching any existing prose or the selected report. The exact CI command now reports validated .github/seo-data: 38 daily records, and all four head checks on 83c5a6fb are green (validate, deploy-static-app, GitGuardian, sync-external-docs).

Note: #212 and #213 both add .github/seo-data/daily/2026-09-23.md with different content and different selected reports, so only one can ship for this date. Which Daily Report ships is an editorial decision left to the maintainer.

AI-generated response; a maintainer will review and follow up later

@yunwei37 yunwei37 left a comment

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Reviewed the head 83c5a6fb as a content change: new research page
ebpf-map-reuse-semantic-compatibility.md with a same-length .zh.md twin
(295 lines each), an index.md entry, and seo-data bookkeeping.

Paired and indexed correctly; the index.md addition is what makes the page
reachable, which is easy to forget in a batch like this. status.md moving −46
lines while content-series adds is consistent with closing out older entries
rather than losing them — worth a glance from the maintainer, but nothing that
looks like data loss.

Checks are green.

AI-generated response; a maintainer will review and follow up later

@yunwei37
yunwei37 force-pushed the daily/2026-09-23-ebpf-map-reuse-semantics branch from 83c5a6f to 9d0d0d7 Compare November 13, 2026 23:13

@yunwei37 yunwei37 left a comment

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Reviewed at 9d0d0d7.

Seven files: a new bilingual report pair docs/research/ebpf-map-reuse-semantic-compatibility{,.zh}.md (295 lines each), index entries in docs/research/index{,.zh}.md, and three .github/seo-data/ bookkeeping files. Front matter matches the existing convention I compared against ebpf-kernel-interface-negotiation{,.zh}.md — date, slug, quoted title/description, a tags list led by Daily Report, research_question, source_cutoff, status: daily-report — and <!-- more --> sits after the intro. The new route is /research/ebpf-map-reuse-semantic-compatibility/ (plus the /zh/ twin), inserted above the 2026-09-22 entry; no existing route is renamed, and the only unrelated hunk is the dropped trailing-newline marker in both index files.

The report's central technical claim checks out against the pinned source. map_is_reuse_compat() in libbpf 1.5 compares exactly type, key_size, value_size, max_entries, map_flags and map_extra and nothing else, its caller bpf_object__reuse_map() is reached from the load path via map->pin_path, and struct bpf_map_info does carry btf_key_type_id/btf_value_type_id — fields the compat check ignores. So "libbpf validates parameters, not schema or semantics" is verified in the tree, not inferred. The four cross-links in the body (stateful-ebpf-transactional-upgrade, ebpf-kernel-capability-evidence, ebpf-kernel-upgrade-semantic-compatibility, ebpf-kernel-interface-negotiation) all exist under docs/research/.

The one thing to settle before merging: this PR and #213 both add .github/seo-data/daily/2026-09-23.md with different content (this one describes the map-reuse report, #213 the pinned-map-reboot report), and both rewrite content-series.md and status.md. Both PRs sit on base 49cc955, so each is "clean" only because neither has merged; whichever lands second will conflict on all three files. Rebase the second one after the first merges.

Checks at this head: validate and deploy-static-app are green, GitGuardian green, sync-external-docs skipped (it skips on every PR here, so no signal). The site.md-style counters that #213 touches are not in this PR, and the sitemap/report-rotation numbers in content-series.md are the PR's own accounting — I did not recompute them.

Verdict: mergeable once the seo-data overlap with #213 is resolved.

AI patrol review

@yunwei37
yunwei37 force-pushed the daily/2026-09-23-ebpf-map-reuse-semantics branch 3 times, most recently from dd3fe5d to b56f6ea Compare November 15, 2026 08:34

@yunwei37 yunwei37 left a comment

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Reviewed head b56f6ea70a: the previous review point (9d0d0d79de) moved onto today's main. The branch's own content is unchanged.

Evidence: for every path this pull request changes (7 files), the git blob at this head equals the blob at 9d0d0d79. The apparent line churn against main comes from files main moved on, not from this branch.

So the earlier review of the change itself still stands, and this head is no longer behind its base. Merge readiness is unchanged: it waits on a maintainer's decision.

AI patrol review

@yunwei37

Copy link
Copy Markdown
Member Author

Verdict: ready to merge

Head b56f6ea7 carries a patrol review of its change (see the review above/below on this pull request), and every check on this head is green. Nothing in the change is left for the patrol to fix.

Merge is the maintainer's to make: eunomia.dev is not in the small-repository group, so the patrol does not merge it.

AI patrol review

@yunwei37 yunwei37 added the AI review: ready to merge patrol review verdict label Nov 15, 2026

@yunwei37 yunwei37 left a comment

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Verdict: ready to merge

The review above stands: this head's content is blob-identical to the previously reviewed revision, so only the rebase onto today's main moved. Checks green, mergeable, level with the base.

AI patrol review

@yunwei37
yunwei37 force-pushed the daily/2026-09-23-ebpf-map-reuse-semantics branch from b56f6ea to f0ff3ef Compare November 15, 2026 19:08

@yunwei37 yunwei37 left a comment

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Verdict: ready to merge once CI is green

Reviewed head f0ff3efd. Every changed file is blob-identical to the previously reviewed revision (b56f6ea7): the bilingual ebpf-map-reuse-semantic-compatibility pair, both index edits, .github/seo-data/status.md, content-series.md, and daily/2026-09-23.md. Merge base unchanged at f8a2cce8.

Content is unchanged: the report's core distinction — that reusing a pinned map is safe only when the new program's expected value layout is a prefix-compatible refinement of the old one, since the map keeps raw bytes across programs — is stated with the BPF_MAP_TYPE_* lookup/update paths that make it observable, and the Chinese version matches it.

Checks on this head: GitGuardian Security Checks success, validate and deploy-static-app queued. The verdict turns to merge-ready when those report green.

AI patrol review

@yunwei37
yunwei37 force-pushed the daily/2026-09-23-ebpf-map-reuse-semantics branch from f0ff3ef to 6cd6ea0 Compare November 16, 2026 03:27

@yunwei37 yunwei37 left a comment

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Verdict: ready to merge once CI is green

Reviewed at head 6cd6ea09, which is the previously reviewed f0ff3efd moved onto today's base. Nothing in the change itself moved: for the range merge-base(main, head)..head, git patch-id --stable is identical before and after, and the changed file list with its line counts is unchanged. The earlier review (5912544711) therefore stands as written, and so does its verdict.

The rebase re-queued this head's checks, so the previous head's green results do not carry over; read the new run, not the old one. Merge stays with a maintainer.

AI patrol review

@yunwei37 yunwei37 added the AI review: merge after CI patrol review verdict label Nov 16, 2026
@yunwei37

Copy link
Copy Markdown
Member Author

Verdict: ready to merge

This is the verdict for head 6cd6ea09. The branch was replayed onto current main; its diff against the base has the same patch-id as the reviewed head, so no line of the change moved — only the base did. All checks on this head are green.

Merge is the maintainer's to make: eunomia.dev is not in the small-repository group, so the patrol does not merge it.

AI patrol review

@yunwei37 yunwei37 removed the AI review: merge after CI patrol review verdict label Nov 16, 2026

This branch has not been deployed

No deployments
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants