Merge remote-tracking branch 'apache/main' into jvm-shuffle-refactor-3

Author: andygrove

.github/actions/java-test/action.yaml

@@ -32,16 +32,21 @@ inputs:
   scan_impl:
     description: 'The default Parquet scan implementation'
     required: false
-    default: 'native_comet'
+    default: 'auto'
   upload-test-reports:
     description: 'Whether to upload test results including coverage to GitHub'
     required: false
     default: 'false'
+  skip-native-build:
+    description: 'Skip native build (when using pre-built artifact)'
+    required: false
+    default: 'false'
 
 runs:
   using: "composite"
   steps:
     - name: Run Cargo release build
+      if: ${{ inputs.skip-native-build != 'true' }}
       shell: bash
       # it is important that we run the Scala tests against a release build rather than a debug build
       # to make sure that no tests are relying on overflow checks that are present only in debug builds
@@ -141,7 +146,3 @@ runs:
          path: "**/target/surefire-reports/*.txt"
          retention-days: 7 # 1 week for test reports
          overwrite: true
-
-    - name: Upload coverage results
-      if: ${{ inputs.upload-test-reports == 'true' }}
-      uses: codecov/codecov-action@v5

.github/actions/rust-test/action.yaml

@@ -21,11 +21,7 @@ description: "Run Rust tests"
 runs:
   using: "composite"
   steps:
-    - name: Check Cargo fmt
-      shell: bash
-      run: |
-        cd native
-        cargo fmt --all -- --check --color=never
+    # Note: cargo fmt check is now handled by the lint job that gates this workflow
 
     - name: Check Cargo clippy
       shell: bash

.github/actions/setup-spark-builder/action.yaml

@@ -22,8 +22,12 @@ inputs:
     description: 'The Apache Spark short version (e.g., 3.5) to build'
     required: true
   spark-version:
-    description: 'The Apache Spark version (e.g., 3.5.7) to build'
+    description: 'The Apache Spark version (e.g., 3.5.8) to build'
     required: true
+  skip-native-build:
+    description: 'Skip native build (when using pre-built artifact)'
+    required: false
+    default: 'false'
 runs:
   using: "composite"
   steps:
@@ -51,7 +55,15 @@ runs:
         restore-keys: |
           ${{ runner.os }}-spark-sql-
 
-    - name: Build Comet
+    - name: Build Comet (with native)
+      if: ${{ inputs.skip-native-build != 'true' }}
       shell: bash
       run: |
         PROFILES="-Pspark-${{inputs.spark-short-version}}" make release
+
+    - name: Build Comet (Maven only, skip native)
+      if: ${{ inputs.skip-native-build == 'true' }}
+      shell: bash
+      run: |
+        # Native library should already be in native/target/release/
+        ./mvnw install -Prelease -DskipTests -Pspark-${{inputs.spark-short-version}}

.github/workflows/benchmark-tpcds.yml

@@ -46,6 +46,13 @@ jobs:
         with:
           python-version: "3.10"
 
+      - name: Setup Java
+        uses: actions/setup-java@v5
+        with:
+          distribution: 'temurin'
+          java-version: '17'
+          cache: 'maven'
+
       - name: Install dependencies
         run: |
           set -x