You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Closes the E10 enforcement-mode flip and tidies the audit-phase artefacts that
were always slated for removal at the close-out. With S10.12–S10.19 done and
the latest main carrying 0 cppcheck-misra findings and 0 clang-tidy naming
warnings, the gates can flip without code work.
Bundle
cppcheck-misra → gating. .github/workflows/ci.ymlanalyze-cppcheck job — flip --error-exitcode=0 → 1 on both the "Run cppcheck-misra" step and the
"Generate cppcheck-misra XML report" step.
Add Platform/MbedTls/Source/ to cppcheck-misra invocation.
Append -IPlatform/MbedTls/Interface and Platform/MbedTls/Source/ to
both cppcheck-misra invocations. Verification step only — the adapter was
developed against the local warning-mode check (S08.07).
Widen the non-MISRA cppcheck XML pass.
The diagnostic XML pass at line 511–522 is scoped to Core/Source/ only;
widen to Core/Source/ + Platform/*/Source/ to match the cppcheck-misra
scope. (Build-time cppcheck already gates Platform/ via CMAKE_C_CPPCHECK
with --error-exitcode=1, but the artefact upload should report the wider
surface.)
clang-tidy naming → gating.
Root .clang-tidy — change WarningsAsErrors: '*,-readability-identifier-naming'
to WarningsAsErrors: '*' and remove the soft-freeze comment block (lines
34–37 reference the wrong story — S10.18 vs. S10.20 — so the easiest fix
is to delete the now-stale paragraph entirely).
Delete docs/misra-conformance.md.
Audit working doc, frozen since S10.12. Live truth lives in docs/misra-deviations.md + misra_suppressions.txt. No remaining
cross-references in code or CI — only DEVLOG history references it.
Widen analyze-format to Platform/.
Currently find Core/Interface Core/Source Tests Bdd/Targets …. Add Platform. (The epic lists this as "optional" — including it because
E10's close-out is the natural moment, and format drift on Platform/
has nothing to push it back.)
Mop-up nit from S10.22 PR refactor: S10.22 enum naming convention restoration (SCREAMING_SNAKE) #391. Tests/DatagramFake.h:18-19 comment says "calls beyond the configured
slots return SOLIDSYSLOG_DATAGRAM_SEND_RESULT_SENT" — wording is
ambiguous and reads wrong against the _FAILED fallback at DatagramFake.c:126. Re-word for clarity. (The other deferred nit was
in misra-conformance.md, which gets deleted in item 5 — no fix needed.)
Acceptance
All status checks green on the PR (cppcheck-misra and naming gates now
exercised in error mode against a clean tree).
git grep misra-conformance.md returns no matches outside DEVLOG.md.
analyze-format step exercises a Platform/ file.
Re-proving the gates actually fail on a deliberate violation was considered
and skipped — the gates fired dozens of times during S10.07–S10.19, so
re-proving in this PR adds churn without information.
Out of scope
No production code changes. No new deviations or suppression edits.
The build-time cppcheck pass (CMAKE_C_CPPCHECK, --error-exitcode=1
since the cppcheck preset landed) is unchanged.
Closes the E10 enforcement-mode flip and tidies the audit-phase artefacts that
were always slated for removal at the close-out. With S10.12–S10.19 done and
the latest main carrying 0 cppcheck-misra findings and 0 clang-tidy naming
warnings, the gates can flip without code work.
Bundle
cppcheck-misra → gating.
.github/workflows/ci.ymlanalyze-cppcheckjob — flip--error-exitcode=0→1on both the "Run cppcheck-misra" step and the"Generate cppcheck-misra XML report" step.
Add
Platform/MbedTls/Source/to cppcheck-misra invocation.Append
-IPlatform/MbedTls/InterfaceandPlatform/MbedTls/Source/toboth cppcheck-misra invocations. Verification step only — the adapter was
developed against the local warning-mode check (S08.07).
Widen the non-MISRA cppcheck XML pass.
The diagnostic XML pass at line 511–522 is scoped to
Core/Source/only;widen to
Core/Source/ + Platform/*/Source/to match the cppcheck-misrascope. (Build-time cppcheck already gates Platform/ via
CMAKE_C_CPPCHECKwith
--error-exitcode=1, but the artefact upload should report the widersurface.)
clang-tidy naming → gating.
Root
.clang-tidy— changeWarningsAsErrors: '*,-readability-identifier-naming'to
WarningsAsErrors: '*'and remove the soft-freeze comment block (lines34–37 reference the wrong story — S10.18 vs. S10.20 — so the easiest fix
is to delete the now-stale paragraph entirely).
Delete
docs/misra-conformance.md.Audit working doc, frozen since S10.12. Live truth lives in
docs/misra-deviations.md+misra_suppressions.txt. No remainingcross-references in code or CI — only DEVLOG history references it.
Widen
analyze-formatto Platform/.Currently
find Core/Interface Core/Source Tests Bdd/Targets …. AddPlatform. (The epic lists this as "optional" — including it becauseE10's close-out is the natural moment, and format drift on Platform/
has nothing to push it back.)
Mop-up nit from S10.22 PR refactor: S10.22 enum naming convention restoration (SCREAMING_SNAKE) #391.
Tests/DatagramFake.h:18-19comment says "calls beyond the configuredslots return SOLIDSYSLOG_DATAGRAM_SEND_RESULT_SENT" — wording is
ambiguous and reads wrong against the
_FAILEDfallback atDatagramFake.c:126. Re-word for clarity. (The other deferred nit wasin misra-conformance.md, which gets deleted in item 5 — no fix needed.)
Acceptance
exercised in error mode against a clean tree).
git grep misra-conformance.mdreturns no matches outside DEVLOG.md.analyze-formatstep exercises a Platform/ file.Re-proving the gates actually fail on a deliberate violation was considered
and skipped — the gates fired dozens of times during S10.07–S10.19, so
re-proving in this PR adds churn without information.
Out of scope
CMAKE_C_CPPCHECK,--error-exitcode=1since the cppcheck preset landed) is unchanged.
Parent epic: #12