Update score_baselibs commit hash to latest version - #355
Merged
Merged
Conversation
4og
requested review from
AlexanderLanin,
FScholPer,
MaximilianSoerenPollak,
PiotrKorkus,
antonkri,
dcalavrezo-qorix,
lurtz,
nradakovic,
opajonk and
pawelrutkaq
as code owners
September 22, 2026 09:33
Member
Author
|
The build issue in communication is addressed in eclipse-score/communication#1129 |
4og
force-pushed
the
4og_update_baselibs
branch
from
September 22, 2026 13:25
3b3faa5 to
2dc5a5e
Compare
Member
Author
|
The other issues related to deprecated rust targets are addressed in eclipse-score/lifecycle#681 and eclipse-score/persistency#431 |
4og
force-pushed
the
4og_update_baselibs
branch
from
September 22, 2026 16:58
b5afb2d to
85c8aa3
Compare
…azel configuration
4og
force-pushed
the
4og_update_baselibs
branch
from
September 22, 2026 17:18
85c8aa3 to
ff36a99
Compare
…n-report patch Baselibs eclipse-score/baselibs#618 fixed the module_verification_report post_template to include all components and switched baselibs' own docs/verification_report/module_verification_report.rst to generate its report via the shared module_verification_report post_template, keeping the id doc__baselibs_verification_report (the template derives the module id by stripping doc__/_verification_report from the need's own id, so this id is now load-bearing). patches/baselibs/005-rename-manual-verification-report-id.patch is no longer compatible with that: reapplying it would rename the id to doc__baselibs_manual_verification_report, which the template would then misparse as module id baselibs_manual, producing an empty report. - Bump score_baselibs known_good pin to 9216215f3a8468f3487884b472e1f325ef36f92b (main after #618). - Drop patch 005 from known_good.json and regenerate bazel_common/score_modules_target_sw.MODULE.bazel. - Simplify docs/verification_report/modules/baselibs/baselibs_verification_report.rst: it no longer redeclares a document with the (now colliding) id, it just links to the report baselibs itself contributes via its docs bundle mount. - Refresh MODULE.bazel.lock via 'bazel mod deps --lockfile_mode=update'.
… wrapper Remove docs/verification_report/modules/baselibs/, which only redeclared a redundant document need. Link modules/index.rst directly to /modules/score_baselibs/verification_report/module_verification_report, the report baselibs itself contributes via its docs bundle mount, so clicking "Baselibs" jumps straight to it instead of an intermediate wrapper page.
…el.lock bazel mod tidy in CI removes the obsolete registryFileHashes entries for buildifier_prebuilt/8.2.0.2, which are no longer part of the resolved dependency graph. Sync the lockfile to fix the bzlmod-lock-check.
antonkri
added a commit
that referenced
this pull request
Sep 24, 2026
Skipping the extraction removed the module from the coverage table entirely, which is silent in the same way the zero rows were. Emit a row with status 'skipped' instead, and suppress the coverage annotation for those, since the unit test annotation for the same module already names the cause. The run of #355 shows why the table cannot simply be trusted: the coverage published for the two modules whose tests never ran was character-identical to the module tested just before them. score_config_management_cpp lines 90.8% functions 84.4% branches 67.1% score_time_cpp lines 90.8% functions 84.4% branches 67.1% score_logging_cpp lines 77.6% functions 81.8% branches 51.2% score_lifecycle_cpp lines 77.6% functions 81.8% branches 51.2%
Contributor
|
Fix for Lifecycle: eclipse-score/logging#271 |
baselibs 0bde8113d ("memory: remove deprecated string manipulation
shims") dropped the forwarding headers that still lived under
score/memory; string_literal, string_comparison_adaptor and
split_string_view have been in //score/string_manipulation since
980a5510d. config_management never followed, so against the baselibs
commit this branch pins, config_provider_impl.h fails to compile with
fatal error: score/memory/string_comparison_adaptor.h: No such file
Point the include at the new path and add the Bazel dependency
explicitly. It was only reachable transitively via
score/json/internal/model before, which is not something a header
should rely on.
Upstream has nothing to wait for: neither main nor any open PR or
branch in eclipse-score/config_management touches this header, so the
patch stays until the module migrates itself.
While here, drop the duplicate "bazel_patches" key in the module's
known_good.json entry - the later one silently won.
…selibs Two independent baselibs restructurings caught up with score_logging at the same time: 980a5510d string_literal moved to //score/string_manipulation 9c2a4481b Rust components moved from //src/log to //score/log_rust 0bde8113d the deprecated score/memory shims were finally removed Against the baselibs commit this branch pins, that leaves four dead labels: score/datarouter/error -> score/memory:string_literal score_log_bridge (2x) -> src/log/score_log score_log_bridge_cpp_init -> src/log/score_log The first one breaks the x86_64 Linux and QNX8 image builds, the Rust ones abort analysis of the whole module, which is why score_logging reported zero executed tests. The same three files are already rewritten this way in eclipse-score/logging#271, but that is a syslog-backend feature PR tied to baselibs#502, not a migration, and it is behind main. Nothing else upstream addresses this, so carry the change here until the module migrates on its own.
The merge with main brought in #367, which moves score_logging from c8e85fd8 to b4f272b0. The patch still applied there, but with the two Rust hunks drifting 10 and 11 lines. Regenerate it against b4f272b0 so it sits on its context exactly. No content change: the same four labels are rewritten.
|
The created documentation from the pull request is available at: docu-html |
antonkri
approved these changes
Sep 25, 2026
antonkri
added a commit
that referenced
this pull request
Sep 28, 2026
* ci: name the modules whose tests failed quality_runners.py already records an exit code per module, but the job summary it renders uses the columns module/passed/failed/skipped/total. A module whose bazel invocation aborted during analysis therefore shows up as all zeroes - indistinguishable from one that simply has no tests, and `failed` even claims zero failures. In #355 that hid two broken modules behind these rows: | score_logging | 0 | 0 | 0 | 0 | | score_config_management | 0 | 0 | 0 | 0 | Add a status column, emit ::error annotations so GitHub names the module above the step list of the run, and close the log with an explicit block naming the failures instead of a pprint dump. Mirror the summary into a single comment on the pull request, rewritten on every run. The job summary is only reachable behind the "Details" link, so the conversation view showed nothing beyond "test_and_docs - Failing after 45m". Also skip coverage extraction when the test run failed. genhtml reads the .dat file from a fixed location, which after a failed run still holds the previous module's data: the coverage report published for score_config_management in #355 in fact contained score_time's numbers. * ci: list skipped coverage runs instead of dropping them Skipping the extraction removed the module from the coverage table entirely, which is silent in the same way the zero rows were. Emit a row with status 'skipped' instead, and suppress the coverage annotation for those, since the unit test annotation for the same module already names the cause. The run of #355 shows why the table cannot simply be trusted: the coverage published for the two modules whose tests never ran was character-identical to the module tested just before them. score_config_management_cpp lines 90.8% functions 84.4% branches 67.1% score_time_cpp lines 90.8% functions 84.4% branches 67.1% score_logging_cpp lines 77.6% functions 81.8% branches 51.2% score_lifecycle_cpp lines 77.6% functions 81.8% branches 51.2% * ci: colour the status column with an emoji prefix Markdown has no way to colour a table cell that survives both a GitHub comment and the Sphinx build of the same file, so carry the colour in an emoji and keep the word next to it for readers without emoji support. * ci: write the job summary from the script instead of commenting on the PR Review feedback: the pull-requests:write scope should not gain another user, and the summary is better assembled where the data already exists. The script now appends the two tables to GITHUB_STEP_SUMMARY directly, so the shell step that cat'ed the files and the github-script step that mirrored them into a pull request comment both go away. Nothing reads the markdown files back any more - they are written purely for the documentation build - which also removes the need to delete them up front. The failing modules stay visible without opening the log: the ::error annotations render above the step list of the run.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
This pull request updates the dependency on the
score_baselibsmodule to a newer commit. The change ensures that both the Bazel module file and theknown_good.jsonreference the same, latest commit for consistency across builds.Dependency update:
score_baselibsdependency to commitf2af2ddb0688abf3bf7516f87bd740afbf1965f8in bothbazel_common/score_modules_target_sw.MODULE.bazelandknown_good.json.