apache · infvg · Mar 11, 2026 · Mar 11, 2026 · Mar 11, 2026 · Mar 11, 2026
diff --git a/.github/workflows/velox_nightly.yml b/.github/workflows/velox_nightly.yml
@@ -171,7 +171,8 @@ jobs:
         run: |
           cd $GITHUB_WORKSPACE/ && \
           ./build/mvn clean install -Pspark-4.0 -Pscala-2.13 -Pjava-21 -Pbackends-velox -Pceleborn -Puniffle -Piceberg -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
-          ./build/mvn clean install -Pspark-4.1 -Pscala-2.13 -Pjava-21 -Pbackends-velox -Pceleborn -Puniffle -Piceberg -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
+          # iceberg/hudi support for spark-4.1 is not avaiable, skip it first
+          ./build/mvn clean install -Pspark-4.1 -Pscala-2.13 -Pjava-21 -Pbackends-velox -Pceleborn -Puniffle -Pdelta -DskipTests -Dmaven.source.skip
       - name: Upload bundle package
         uses: actions/upload-artifact@v4
         with:
@@ -287,8 +288,8 @@ jobs:
           ./build/mvn clean install -Pspark-3.4 -Pjava-17 -Pbackends-velox -Pceleborn -Puniffle -Piceberg -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
           ./build/mvn clean install -Pspark-3.5 -Pjava-17 -Pbackends-velox -Pceleborn -Puniffle -Piceberg -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
           ./build/mvn clean install -Pspark-4.0 -Pscala-2.13 -Pjava-17 -Pbackends-velox -Pceleborn -Puniffle -Piceberg -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
-          # iceberg support for spark-4.1 with java-17 is not avaiable, skip it first
-          ./build/mvn clean install -Pspark-4.1 -Pscala-2.13 -Pjava-21 -Pbackends-velox -Pceleborn -Puniffle -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
+          # iceberg/hudi support for spark-4.1 is not avaiable, skip it first
+          ./build/mvn clean install -Pspark-4.1 -Pscala-2.13 -Pjava-17 -Pbackends-velox -Pceleborn -Puniffle -Pdelta -DskipTests -Dmaven.source.skip
       - name: Upload bundle package
         uses: actions/upload-artifact@v4
         with:
@@ -323,8 +324,8 @@ jobs:
         run: |
           cd $GITHUB_WORKSPACE/ && \
           ./build/mvn clean install -Pspark-4.0 -Pscala-2.13 -Pjava-21 -Pbackends-velox -Pceleborn -Puniffle -Piceberg -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
-          # iceberg support for spark-4.1 with java-21 is not avaiable, skip it first
-          ./build/mvn clean install -Pspark-4.1 -Pscala-2.13 -Pjava-21 -Pbackends-velox -Pceleborn -Puniffle -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
+          # iceberg/hudi support for spark-4.1 is not avaiable, skip it first
+          ./build/mvn clean install -Pspark-4.1 -Pscala-2.13 -Pjava-21 -Pbackends-velox -Pceleborn -Puniffle -Pdelta -DskipTests -Dmaven.source.skip
       - name: Upload bundle package
         uses: actions/upload-artifact@v4
         with:
@@ -486,8 +487,8 @@ jobs:
           ./build/mvn clean install -Pspark-3.4 -Pjava-17 -Pbackends-velox -Pceleborn -Puniffle -Piceberg -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
           ./build/mvn clean install -Pspark-3.5 -Pjava-17 -Pbackends-velox -Pceleborn -Puniffle -Piceberg -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
           ./build/mvn clean install -Pspark-4.0 -Pscala-2.13 -Pjava-17 -Pbackends-velox -Pceleborn -Puniffle -Piceberg -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
-          # iceberg support for spark-4.1 with java-17 is not avaiable, skip it first
-          ./build/mvn clean install -Pspark-4.1 -Pscala-2.13 -Pjava-17 -Pbackends-velox -Pceleborn -Puniffle -Phudi -Pdelta -Ppaimon -DskipTests -Dmaven.source.skip
+          # iceberg/hudi support for spark-4.1 with java-17 is not avaiable, skip it first
+          ./build/mvn clean install -Pspark-4.1 -Pscala-2.13 -Pjava-17 -Pbackends-velox -Pceleborn -Puniffle -Pdelta -DskipTests -Dmaven.source.skip
       - name: Upload bundle package
         uses: actions/upload-artifact@v4
         with:

diff --git a/...velox/src-iceberg/test/scala/org/apache/gluten/execution/enhanced/VeloxIcebergSuite.scala b/...velox/src-iceberg/test/scala/org/apache/gluten/execution/enhanced/VeloxIcebergSuite.scala
@@ -16,6 +16,7 @@
  */
 package org.apache.gluten.execution.enhanced
 
+import org.apache.gluten.config.VeloxConfig
 import org.apache.gluten.execution._
 import org.apache.gluten.tags.EnhancedFeaturesTest
 
@@ -383,4 +384,139 @@ class VeloxIcebergSuite extends IcebergSuite {
       }
     }
   }
+  test("iceberg max aggregate with nulls enhanced true") {
+    withSQLConf(VeloxConfig.ENABLE_ENHANCED_FEATURES.key -> "true") {
+      withTable("store_sales_test_nulls_enhanced_true") {
+        spark.sql("""
+                    |create table store_sales_test_nulls_enhanced_true (
+                    |  ss_item_sk int,
+                    |  ss_sales_price decimal(7,2)
+                    |) using iceberg
+                    |""".stripMargin)
+
+        spark.sql("""
+                    |insert into store_sales_test_nulls_enhanced_true values
+                    |(1, 200.00),
+                    |(2, 200.00),
+                    |(3, null),
+                    |(4, 199.98),
+                    |(5, 199.96)
+                    |""".stripMargin)
+
+        val result = spark
+          .sql("select max(ss_sales_price) from store_sales_test_nulls_enhanced_true")
+          .collect()
+
+        assert(result.length == 1, "Should return 1 row")
+        assert(result(0).get(0) != null, "MAX should not return NULL")
+        assert(
+          result(0).getDecimal(0).doubleValue() == 200.00,
+          s"MAX should return 200.00, but got ${result(0).get(0)}"
+        )
+      }
+    }
+  }
+
+  test("iceberg max aggregate without nulls enhanced true") {
+    withSQLConf(VeloxConfig.ENABLE_ENHANCED_FEATURES.key -> "true") {
+      withTable("store_sales_test_no_nulls_enhanced_true") {
+        spark.sql("""
+                    |create table store_sales_test_no_nulls_enhanced_true (
+                    |  ss_item_sk int,
+                    |  ss_sales_price decimal(7,2)
+                    |) using iceberg
+                    |""".stripMargin)
+
+        spark.sql("""
+                    |insert into store_sales_test_no_nulls_enhanced_true values
+                    |(1, 200.00),
+                    |(2, 200.00),
+                    |(3, 200.00),
+                    |(4, 199.98),
+                    |(5, 199.96),
+                    |(6, 199.96),
+                    |(7, 199.92),
+                    |(8, 199.92),
+                    |(9, 199.92),
+                    |(10, 199.90),
+                    |(11, null)
+                    |""".stripMargin)
+
+        val result = spark
+          .sql("select max(ss_sales_price) from store_sales_test_no_nulls_enhanced_true where ss_sales_price is not null")
+          .collect()
+
+        assert(result.length == 1, "Should return 1 row")
+        assert(result(0).get(0) != null, "MAX should not return NULL")
+        assert(
+          result(0).getDecimal(0).doubleValue() == 200.00,
+          s"MAX should return 200.00, but got ${result(0).get(0)}"
+        )
+      }
+    }
+  }
+
+  test("iceberg max aggregate with nulls enhanced false") {
+    withSQLConf(VeloxConfig.ENABLE_ENHANCED_FEATURES.key -> "false") {
+      withTable("store_sales_test_nulls_enhanced_false") {
+        spark.sql("""
+                    |create table store_sales_test_nulls_enhanced_false (
+                    |  ss_item_sk int,
+                    |  ss_sales_price decimal(7,2)
+                    |) using iceberg
+                    |""".stripMargin)
+
+        spark.sql("""
+                    |insert into store_sales_test_nulls_enhanced_false values
+                    |(1, 200.00),
+                    |(2, 199.98),
+                    |(3, null),
+                    |(4, 199.96)
+                    |""".stripMargin)
+
+        val result = spark
+          .sql("select max(ss_sales_price) from store_sales_test_nulls_enhanced_false")
+          .collect()
+
+        assert(result.length == 1, "Should return 1 row")
+        assert(result(0).get(0) != null, "MAX should not return NULL")
+        assert(
+          result(0).getDecimal(0).doubleValue() == 200.00,
+          s"MAX should return 200.00, but got ${result(0).get(0)}"
+        )
+      }
+    }
+  }
+
+  test("iceberg max aggregate without nulls enhanced false") {
+    withSQLConf(VeloxConfig.ENABLE_ENHANCED_FEATURES.key -> "false") {
+      withTable("store_sales_test_no_nulls_enhanced_false") {
+        spark.sql("""
+                    |create table store_sales_test_no_nulls_enhanced_false (
+                    |  ss_item_sk int,
+                    |  ss_sales_price decimal(7,2)
+                    |) using iceberg
+                    |""".stripMargin)
+
+        spark.sql("""
+                    |insert into store_sales_test_no_nulls_enhanced_false values
+                    |(1, 200.00),
+                    |(2, 199.98),
+                    |(3, 199.96),
+                    |(4, null)
+                    |""".stripMargin)
+
+        val result = spark
+          .sql("select max(ss_sales_price) from store_sales_test_no_nulls_enhanced_false where ss_sales_price is not null")
+          .collect()
+
+        assert(result.length == 1, "Should return 1 row")
+        assert(result(0).get(0) != null, "MAX should not return NULL")
+        assert(
+          result(0).getDecimal(0).doubleValue() == 200.00,
+          s"MAX should return 200.00, but got ${result(0).get(0)}"
+        )
+      }
+    }
+  }
 }
diff --git a/dev/HardwareConcurrency.h b/dev/HardwareConcurrency.h
@@ -0,0 +1,27 @@
+/*
+ * Copyright (c) Meta Platforms, Inc. and affiliates.
+ *
+ * Licensed under the Apache License, Version 2.0 (the "License");
+ * you may not use this file except in compliance with the License.
+ * You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+
+#pragma once
+
+namespace folly {
+
+unsigned int hardware_concurrency() noexcept;
+
+unsigned int available_concurrency() noexcept {
+  return hardware_concurrency();
+}
+
+} // namespace folly
diff --git a/dev/ci-velox-buildshared-centos-8.sh b/dev/ci-velox-buildshared-centos-8.sh
@@ -18,5 +18,7 @@
 set -e
 
 source /opt/rh/gcc-toolset-11/enable
+# Fixme: this is a hack to use new header, should upgrade folly instead.
+cp dev/HardwareConcurrency.h /usr/local/include/folly/system/HardwareConcurrency.h
 ./dev/builddeps-veloxbe.sh --run_setup_script=OFF --build_arrow=OFF --build_tests=ON \
     --build_examples=ON --build_benchmarks=ON
diff --git a/dev/vcpkg/ports/folly/adding-api.patch b/dev/vcpkg/ports/folly/adding-api.patch
@@ -0,0 +1,34 @@
+diff --git a/folly/system/HardwareConcurrency.cpp b/folly/system/HardwareConcurrency.cpp
+index 71213c80a..cd1852034 100644
+--- a/folly/system/HardwareConcurrency.cpp
++++ b/folly/system/HardwareConcurrency.cpp
+@@ -36,4 +36,18 @@ unsigned int hardware_concurrency() noexcept {
+   return std::thread::hardware_concurrency();
+ }
+
++unsigned int available_concurrency() noexcept {
++#if defined(__linux__) && !defined(__ANDROID__)
++  cpu_set_t cpuset;
++  if (!sched_getaffinity(0, sizeof(cpuset), &cpuset)) {
++    auto count = CPU_COUNT(&cpuset);
++    if (count != 0) {
++      return count;
++    }
++  }
++#endif
++
++  return std::thread::hardware_concurrency();
++}
++
+ } // namespace folly
+diff --git a/folly/system/HardwareConcurrency.h b/folly/system/HardwareConcurrency.h
+index 7c497634c..71dfb5db1 100644
+--- a/folly/system/HardwareConcurrency.h
++++ b/folly/system/HardwareConcurrency.h
+@@ -20,4 +20,6 @@ namespace folly {
+
+ unsigned int hardware_concurrency() noexcept;
+
++unsigned int available_concurrency() noexcept;
++
+ } // namespace folly
diff --git a/dev/vcpkg/ports/folly/portfile.cmake b/dev/vcpkg/ports/folly/portfile.cmake
@@ -13,6 +13,7 @@ vcpkg_from_github(
         disable-uninitialized-resize-on-new-stl.patch
         fix-unistd-include.patch
         fix-absolute-dir.patch
+        adding-api.patch
 )
 file(REMOVE "${SOURCE_PATH}/CMake/FindFastFloat.cmake")
 file(REMOVE "${SOURCE_PATH}/CMake/FindFmt.cmake")

diff --git a/ep/build-velox/src/get-velox.sh b/ep/build-velox/src/get-velox.sh
@@ -18,8 +18,8 @@ set -exu
 
 CURRENT_DIR=$(cd "$(dirname "$BASH_SOURCE")"; pwd)
 VELOX_REPO=https://github.com/IBM/velox.git
-VELOX_BRANCH=dft-2026_03_10-iceberg
-VELOX_ENHANCED_BRANCH=ibm-2026_03_10
+VELOX_BRANCH=dft-2026_03_11-iceberg
+VELOX_ENHANCED_BRANCH=temp-fix
 VELOX_HOME=""
 RUN_SETUP_SCRIPT=ON
 ENABLE_ENHANCED_FEATURES=OFF

diff --git a/ep/build-velox/src/setup-centos8.sh b/ep/build-velox/src/setup-centos8.sh
@@ -43,8 +43,9 @@ export CC=/opt/rh/gcc-toolset-11/root/bin/gcc
 export CXX=/opt/rh/gcc-toolset-11/root/bin/g++
 DEPENDENCY_DIR=${DEPENDENCY_DIR:-$(pwd)/deps-download}
 
-FB_OS_VERSION="v2024.07.01.00"
-FMT_VERSION="10.1.1"
+FB_OS_VERSION="v2026.01.05.00"
+FMT_VERSION="11.2.0"
+FAST_FLOAT_VERSION="v8.0.2"
 BOOST_VERSION="boost-1.84.0"
 GEOS_VERSION="3.10.7"
 
@@ -58,7 +59,7 @@ function install_build_prerequisites {
   dnf_install epel-release dnf-plugins-core # For ccache, ninja
   dnf config-manager --set-enabled powertools
   dnf update -y
-  dnf_install ninja-build curl ccache gcc-toolset-11 git wget which
+  dnf_install ninja-build curl ccache gcc-toolset-11 git wget which expat-devel gettext-devel
   dnf_install yasm
   dnf_install autoconf automake python39 python39-devel python39-pip libtool
   pip3.9 install cmake==3.28.3
@@ -78,6 +79,24 @@ function install_conda {
   dnf_install conda
 }
 
+function install_git {
+  # Remove an older version if present.
+  dnf remove -y git
+  wget_and_untar https://github.com/git/git/archive/v2.52.0.tar.gz git
+  (
+    cd ${DEPENDENCY_DIR}/git
+    make prefix=/usr/local all -j$(nproc)
+    make prefix=/usr/local install
+  )
+}
+
+function install_xxhash {
+  wget_and_untar https://github.com/Cyan4973/xxHash/archive/refs/tags/v0.8.1.tar.gz xxhash
+  (
+    cd ${DEPENDENCY_DIR}/xxhash
+    make && make install
+  )
+}
 
 function install_gflags {
   # Remove an older version if present.
@@ -136,6 +155,11 @@ function install_fizz {
   cmake_install_dir fizz/fizz -DBUILD_TESTS=OFF
 }
 
+function install_fast_float {
+  wget_and_untar https://github.com/fastfloat/fast_float/archive/refs/tags/"${FAST_FLOAT_VERSION}".tar.gz fast_float
+  cmake_install_dir fast_float -DBUILD_TESTS=OFF
+}
+
 function install_folly {
   wget_and_untar https://github.com/facebook/folly/archive/refs/tags/${FB_OS_VERSION}.tar.gz folly
   cmake_install_dir folly -DFOLLY_HAVE_INT128_T=ON -DFOLLY_NO_EXCEPTION_TRACER=ON
@@ -160,7 +184,7 @@ function install_duckdb {
   if $BUILD_DUCKDB ; then
     echo 'Building DuckDB'
     wget_and_untar https://github.com/duckdb/duckdb/archive/refs/tags/v0.8.1.tar.gz duckdb
-    cmake_install_dir duckdb -DBUILD_UNITTESTS=OFF -DENABLE_SANITIZER=OFF -DENABLE_UBSAN=OFF -DBUILD_SHELL=OFF -DEXPORT_DLL_SYMBOLS=OFF -DCMAKE_BUILD_TYPE=Release
+    cmake_install_dir duckdb -DGIT_COMMIT_HASH="6536a77" -DBUILD_UNITTESTS=OFF -DENABLE_SANITIZER=OFF -DENABLE_UBSAN=OFF -DBUILD_SHELL=OFF -DEXPORT_DLL_SYMBOLS=OFF -DCMAKE_BUILD_TYPE=Release
   fi
 }
 
@@ -179,6 +203,7 @@ function install_geos {
 
 function install_velox_deps {
   run_and_time install_velox_deps_from_dnf
+  run_and_time install_git
   run_and_time install_conda
   run_and_time install_gflags
   run_and_time install_glog
@@ -187,6 +212,8 @@ function install_velox_deps {
   run_and_time install_boost
   run_and_time install_protobuf
   run_and_time install_fmt
+  run_and_time install_fast_float
+  run_and_time install_xxhash
   run_and_time install_folly
   run_and_time install_fizz
   run_and_time install_wangle

diff --git a/gluten-substrait/src/main/scala/org/apache/gluten/execution/JoinExecTransformer.scala b/gluten-substrait/src/main/scala/org/apache/gluten/execution/JoinExecTransformer.scala
@@ -118,7 +118,6 @@ trait HashJoinLikeExecTransformer extends BaseJoinExec with TransformSupport {
         fromFields.length == toFields.length &&
         fromFields.zip(toFields).forall {
           case (l, r) =>
-            l.name.equalsIgnoreCase(r.name) &&
             sameType(l.dataType, r.dataType)
         }
 

diff --git a/gluten-ut/spark41/src/test/scala/org/apache/gluten/utils/velox/VeloxTestSettings.scala b/gluten-ut/spark41/src/test/scala/org/apache/gluten/utils/velox/VeloxTestSettings.scala
@@ -780,7 +780,7 @@ class VeloxTestSettings extends BackendTestSettings {
   // Generated suites for org.apache.spark.sql
   enableSuite[GlutenCacheManagerSuite]
   enableSuite[GlutenDataFrameShowSuite]
-  // TODO: 4.x enableSuite[GlutenDataFrameSubquerySuite]  // 1 failure
+  enableSuite[GlutenDataFrameSubquerySuite]
   enableSuite[GlutenDataFrameTableValuedFunctionsSuite]
   enableSuite[GlutenDataFrameTransposeSuite]
   enableSuite[GlutenDeprecatedDatasetAggregatorSuite]