@@ -30,16 +30,17 @@ is a `workflow_call` reusable invoked from the umbrella.
3030 | heavy job) |
3131 +-----------+-----------+
3232 |
33- +-----------+-----------+-----------+-----------+-----------+-----------+
34- | | | | | | |
35- v v v v v v v
36- pr_build_ pr_build_ pr_benchmark_ docs spark_3_5 spark_4_0 iceberg_1_10
37- linux macos check (push) (PR+push) (PR+push) (PR+push)
38- (PR+push) (PR+push) (PR+push)
39- | | |
40- v v v
41- spark_3_4 / spark_4_1 iceberg_1_8 / 1_9
42- (push or PR + label) (push only)
33+ +-----------------------------------+-----------------------------------+
34+ | | |
35+ v v v
36+ every PR + push push to main only PR with label, or push
37+ --------------- ----------------- ----------------------
38+ pr_build_linux docs spark_3_4 run-spark-3.4-tests
39+ pr_build_macos spark_4_0 run-spark-4.0-tests
40+ pr_benchmark_check iceberg_1_8 run-iceberg-tests
41+ spark_3_5 iceberg_1_9 run-iceberg-tests
42+ spark_4_1 iceberg_1_10 run-iceberg-tests
43+ iceberg_1_11
4344
4445 reusable workflows invoked via `uses:`:
4546 pr_build_linux.yml spark_sql_test_reusable.yml
@@ -50,25 +51,50 @@ is a `workflow_call` reusable invoked from the umbrella.
5051
5152## What runs when
5253
53- | Job in ` ci.yml ` | Triggered by | Path filter source |
54- | -------------------- | ------------------------------------------------ | ----------------------------------- |
55- | ` preflight ` | every PR / push to main / dispatch | none (always runs) |
56- | ` changes ` | every PR / push to main / dispatch | runs ` dev/ci/compute-changes.py ` |
57- | ` pr_build_linux ` | PR or push, paths matched | ` dev/ci/compute-changes.py ` |
58- | ` pr_build_macos ` | PR or push, paths matched | ` dev/ci/compute-changes.py ` |
59- | ` pr_benchmark_check ` | PR or push, paths matched | benchmark sources only |
60- | ` docs ` | push to main, paths matched | ` .asf.yaml ` , ` docs/** ` , ` docs.yaml ` |
61- | ` spark_3_5 ` | PR or push, paths matched | Spark 3.5 sources |
62- | ` spark_4_0 ` | PR or push, paths matched | Spark 4.0 sources |
63- | ` spark_3_4 ` | push, ** or** PR with ` run-spark-3.4-tests ` label | Spark 3.4 sources |
64- | ` spark_4_1 ` | push, ** or** PR with ` run-spark-4.1-tests ` label | Spark 4.1 sources |
65- | ` iceberg_1_10 ` | PR or push, paths matched | Iceberg sources |
66- | ` iceberg_1_8 ` | push only | Iceberg sources |
67- | ` iceberg_1_9 ` | push only | Iceberg sources |
54+ | Job in ` ci.yml ` | Triggered by | Path filter source |
55+ | -------------------- | --------------------------------------------------- | ----------------------------------- |
56+ | ` preflight ` | every PR / push to main / dispatch / PR label added | none (always runs) |
57+ | ` changes ` | every PR / push to main / dispatch / PR label added | runs ` dev/ci/compute-changes.py ` |
58+ | ` pr_build_linux ` | PR or push, paths matched | ` dev/ci/compute-changes.py ` |
59+ | ` pr_build_macos ` | PR or push, paths matched | ` dev/ci/compute-changes.py ` |
60+ | ` pr_benchmark_check ` | PR or push, paths matched | benchmark sources only |
61+ | ` docs ` | push to main, paths matched | ` .asf.yaml ` , ` docs/** ` , ` docs.yaml ` |
62+ | ` spark_3_5 ` | PR or push, paths matched | Spark 3.5 sources |
63+ | ` spark_4_1 ` | PR or push, paths matched | Spark 4.1 sources |
64+ | ` spark_3_4 ` | push, ** or** PR with ` run-spark-3.4-tests ` label | Spark 3.4 sources |
65+ | ` spark_4_0 ` | push, ** or** PR with ` run-spark-4.0-tests ` label | Spark 4.0 sources |
66+ | ` iceberg_1_11 ` | PR or push, paths matched | Iceberg sources |
67+ | ` iceberg_1_8 ` | push, ** or** PR with ` run-iceberg-tests ` label | Iceberg sources |
68+ | ` iceberg_1_9 ` | push, ** or** PR with ` run-iceberg-tests ` label | Iceberg sources |
69+ | ` iceberg_1_10 ` | push, ** or** PR with ` run-iceberg-tests ` label | Iceberg sources |
6870
6971A heavy job appears in the PR's checks list as a ` skipped ` entry whenever
7072its path filter or event criteria don't match. Skipped checks count as
71- passing for branch protection.
73+ passing for branch protection, so a name that can report ` skipped ` is not
74+ safe to make a required check.
75+
76+ ### Label events
77+
78+ ` ci.yml ` also fires on ` pull_request.types: [labeled] ` , so applying
79+ ` run-spark-3.4-tests ` , ` run-spark-4.0-tests ` or ` run-iceberg-tests ` starts the
80+ job that label gates without needing a new push. GitHub cannot filter a
81+ ` pull_request ` trigger by label name, so ** every** label added to a PR starts a
82+ run, including labels that gate nothing.
83+
84+ Two rules keep those runs from corrupting the PR's status:
85+
86+ - ` preflight ` and ` changes ` carry no event guard and run every time. A job held
87+ back by ` if: ` still publishes a check run under its own name with conclusion
88+ ` skipped ` , and the newest check run for a name is what the merge box,
89+ ` gh pr checks ` and required-status-check evaluation read. Guarding
90+ ` preflight ` on the label name used to let any unrelated label overwrite the
91+ commit run's real ` Preflight ` verdict with ` skipped ` , see
92+ [ #5007 ] ( https://github.com/apache/datafusion-comet/issues/5007 ) .
93+ - Every heavy job excludes ` labeled ` events unless the label just added is the
94+ one that gates it. Without that, applying a single label re-ran the entire
95+ heavy pipeline at a commit that had already been tested.
96+
97+ ` run-spark-4.1-tests ` gates nothing: ` spark_4_1 ` already runs on every PR.
7298
7399## Standalone workflows (not under the umbrella)
74100
@@ -83,17 +109,18 @@ umbrella doesn't watch, or operate independently of the rest of CI:
83109| ` stale.yml ` | Daily stale-PR closer. |
84110| ` take.yml ` | Issue-comment trigger for ` take ` / ` untake ` . |
85111| ` label_new_issues.yml ` | Issue trigger to apply ` requires-triage ` . |
112+ | ` label_prs.yml ` | Runs on ` pull_request_target ` so it can label pull requests opened from forks. |
86113
87114## Reusable workflows (called by ` ci.yml ` )
88115
89- | File | Called from ` ci.yml ` job(s) |
90- | --------------------------------- | -------------------------------------------------- |
91- | ` pr_build_linux.yml ` | ` pr_build_linux ` |
92- | ` pr_build_macos.yml ` | ` pr_build_macos ` |
93- | ` pr_benchmark_check.yml ` | ` pr_benchmark_check ` |
94- | ` docs.yaml ` | ` docs ` |
95- | ` spark_sql_test_reusable.yml ` | ` spark_3_4 ` , ` spark_3_5 ` , ` spark_4_0 ` , ` spark_4_1 ` |
96- | ` iceberg_spark_test_reusable.yml ` | ` iceberg_1_8 ` , ` iceberg_1_9 ` , ` iceberg_1_10 ` |
116+ | File | Called from ` ci.yml ` job(s) |
117+ | --------------------------------- | ------------------------------------------------------------ |
118+ | ` pr_build_linux.yml ` | ` pr_build_linux ` |
119+ | ` pr_build_macos.yml ` | ` pr_build_macos ` |
120+ | ` pr_benchmark_check.yml ` | ` pr_benchmark_check ` |
121+ | ` docs.yaml ` | ` docs ` |
122+ | ` spark_sql_test_reusable.yml ` | ` spark_3_4 ` , ` spark_3_5 ` , ` spark_4_0 ` , ` spark_4_1 ` |
123+ | ` iceberg_spark_test_reusable.yml ` | ` iceberg_1_8 ` , ` iceberg_1_9 ` , ` iceberg_1_10 ` , ` iceberg_1_11 ` |
97124
98125## Modifying path filters
99126
@@ -171,3 +198,11 @@ umbrella exposes per-job names like `CI / pr_build_linux / Lint`,
171198protection rules to point at the new names; the old standalone workflow
172199names (` Spark SQL Tests (Spark 3.5) ` , ` PR Build (Linux) ` , ...) no longer
173200exist as top-level workflows.
201+
202+ ` .asf.yaml ` currently declares no ` required_status_checks ` for ` main ` , only
203+ ` required_approving_review_count: 1 ` . Anything added there must be a name that
204+ never legitimately reports ` skipped ` , because GitHub counts a skipped check as
205+ passing. The bare caller-job names (` PR Build (Linux) ` , `Spark SQL Tests
206+ (Spark 3.5)`, ...) are not such names: a reusable workflow that actually runs
207+ publishes only its child jobs, so the bare name shows up solely when the caller
208+ was skipped.
0 commit comments