akshaytayal opened a new issue, #13048:
URL: https://github.com/apache/gluten/issues/13048
## Summary
Apache Spark's `branch-4.2` introduces a large number of new unit tests
relative to `branch-4.1`. To support Spark 4.2 in Gluten (Velox backend) with
parity, these newly added tests should be ported/enabled and triaged
(supported, fallback, or intentionally disabled) as part of the Spark 4.2.x
effort.
Umbrella: #12569 · Related trackers: #11910, #11550
## Scope
- **3,826** high-confidence new test cases (test name not present anywhere
in `branch-4.1`)
- **580** unique upstream Spark PRs · **23** modules
- Languages: Scala **3,024**, Python **452**, Java **350**
- (An additional 69 lower-confidence entries — generic-name reuse or
cross-file moves — are excluded here and tracked separately.)
### New tests by module
| Module | New tests |
|---|---:|
| `sql/core` | 1907 |
| `sql/catalyst` | 854 |
| `python` | 452 |
| `core` | 156 |
| `sql/pipelines` | 150 |
| `sql/connect/server` | 96 |
| `resource-managers/kubernetes/core` | 47 |
| `sql/connect/client/jvm` | 47 |
| `udf/worker/core` | 31 |
| `common/unsafe` | 15 |
| `connector/avro` | 14 |
| `connector/protobuf` | 14 |
| `sql/hive` | 12 |
| `resource-managers/yarn` | 10 |
| `common/utils` | 4 |
| `common/utils-java` | 3 |
| `connector/kafka-0-10-sql` | 3 |
| `connector/kinesis-asl` | 3 |
| `sql/hive-thriftserver` | 3 |
| `connector/docker-integration-tests` | 2 |
| `mllib` | 1 |
| `resource-managers/kubernetes/core/volcano` | 1 |
| `sql/api` | 1 |
## Proposed work
- [ ] Triage each new test for the Velox backend (pass / fallback /
disable-with-reason)
- [ ] Enable the tests that pass under Gluten
- [ ] Record intentionally disabled suites in the Spark 4.x disabled-tests
tracker (#11550)
## Methodology
Computed by diffing `apache/spark` `branch-4.1..branch-4.2`. A "new test" is
a test case (Scala `test("...")`/`*Test(`, Java `@Test`, Python `def test_*`)
present in 4.2 but not in 4.1, with git rename detection and a cross-check
against the entire 4.1 test tree to exclude moved/renamed tests. Each test is
attributed to the first commit in range that introduced it. SQL
golden-file/generated tests are out of scope. The full per-test list (with
commit/PR links) is available on request.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
