Skip to content

why: tell whether a CVE is covered, and if not, why - #212

Merged
igorpecovnik merged 1 commit into
mainfrom
tooling/why
Nov 3, 2026
Merged

igorpecovnik merged 1 commit into
mainfrom
tooling/why

Conversation

@igorpecovnik

Copy link
Copy Markdown
Member

Goal: one command answers "do we cover CVE-X, and if not, why".

$ ./fasttrack.sh why CVE-2025-37834 CVE-2026-9923
CVE-2025-37834: mm/vmscan: don't try to reclaim hwpoison folio
  6.18 (base 6.18.60): not affected: fixed in mainline 6.15, before this series
  6.12 (base 6.12.115): fixed in stable 6.12.26
  6.1 (base 6.1.191): included: 1b0449544c64 mm/vmscan: don't try to reclaim hwpoison folio
      patches/6.1/0112-1b0449544c64-mm-vmscan-don-t-try-to-reclaim-hwpoison-folio.patch

CVE-2026-9923: not covered: not a Linux kernel CVE (Chrome: Google Chrome). armbian/cve tracks kernel CVEs only.
  • Each run writes cves.json: status per CVE and series, with reason and patch file. The workflow commits it.
  • why needs no sync. Without a local feed, it reads the record from CVE.org.
  • report.md and cves.json use one classification. report.md output is unchanged (byte-identical on a 6.1 + 6.18 run).

Tests: unit tests pass; a new test covers why.

Each run writes cves.json: the status of each CVE per series.
`fasttrack why <CVE>` reads it with the CVE record. A CVE from
another CNA gets "not a Linux kernel CVE". report.md and cves.json
use one classification, so they agree.

Signed-off-by: Igor Pecovnik <igor@armbian.com>
@igorpecovnik
igorpecovnik merged commit d357a79 into main Nov 3, 2026
1 check passed
@igorpecovnik
igorpecovnik deleted the tooling/why branch November 3, 2026 05:04
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.

1 participant