This is an automated email from the ASF dual-hosted git repository.
marin-ma pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/gluten.git
The following commit(s) were added to refs/heads/main by this push:
new 5964ac0abc [GLUTEN-6887][VL] Daily Update Velox Version (2026_08_05)
(#12692)
5964ac0abc is described below
commit 5964ac0abc07fa40ff5fc34d82acb4fc59cc83d2
Author: Gluten Performance Bot
<[email protected]>
AuthorDate: Sun Aug 9 16:33:38 2026 +0100
[GLUTEN-6887][VL] Daily Update Velox Version (2026_08_05) (#12692)
* [GLUTEN-6887][VL] Daily Update Velox Version (dft-2026_08_05)
Upstream Velox's New Commits:
6dab64802 by zhichenxu-meta, fix(rpc): Size RPCOperator output from
preferredOutputBatchRows (#18298)
fe7761f5a by Bikramjeet Singh Vig, fix(tz): Regenerate TimeZoneNames.cpp to
cover Europe/Kyiv
1692d7fbc by Krishna Pai, fix(ci): Move stalled ubuntu-latest jobs to
4-core-ubuntu (#18392)
334e4ec92 by oerling, feat(torchwave): in-place buffer reuse and
clone-elision infra
bda33beec by RindsSchei225e, refactor(writer): Move the writer dependencies
to dwio/nimble/writer (#18362)
549c16c87 by RindsSchei225e, test(iceberg): Guard IcebergNimbleInsertTest
behind VELOX_ENABLE_NIMBLE
a821a1a6d by Shrinidhi Joshi, docs: Add hashtable caching blog
872a94d9a by Eric Fu, feat(hive): Add database name to FileScanBatchEvent
(#18249)
3106ba073 by Patrick Wilson, feat(cudf): Add GPU TopNRowNumber operator
with tests
87f381979 by zhichenxu-meta, refactor: Fold RPCNode call arguments into a
CallTypedExpr (#18267)
1d9c83a94 by Bikramjeet Singh Vig, refactor(fuzzer): Share the Presto
expression fuzzer skip list
6737a72da by Pedro Pedreira, perf(vector): Return OPAQUE values by const
ref from DecodedVector::valueAt
b60289eb9 by Eric Fu, fix(iceberg): Default-initialize
FieldIdColumnSpec::requiredSubfields
690569ed2 by kevinwilfong, fix(SortBuffer): Release accumulated rows once
all output is produced (#18357)
31ab7fc05 by Pedro Pedreira, refactor(type): Consolidate is_shared_ptr
trait into Type.h (#18365)
3adc0c5b9 by abhinavmuk04, Fix lock-order-inversion in RPCState waiter
wakeup (#18310)
58e5c678e by dependabot[bot], build(ci): Bump actions/upload-artifact from
4.6.2 to 7.0.1
12cee0b55 by dependabot[bot], build(ci): Bump actions/download-artifact
from 4.3.0 to 8.0.1
a3dd57068 by Pramod Satya, feat(cuDF): Refactor expression subsystem to
remove CPU expression dependency
ef1caf839 by Natasha Sehgal, feat: Add TrackedExecutor to velox/common/base
(#18359)
0cfbfe451 by Maria Basmanova, fix(type): Keep upper 64 bits of HUGEINT in
Variant serde (#18351)
be5f084ad by Krishna Pai, fix(build): Suppress clang-tidy
nullability-completeness via --extra-arg
47d5b9dff by Ping Liu, feat(parquet): Read columns by field ID
5331684b0 by Mohammad Linjawi, fix(HashAggregation): Do not drop input
batches under downstream back-pressure
a9390325b by Shruti Shivakumar, fix(exec): Handle custom join bridges in
mixed grouped execution
249d7c8a8 by dependabot[bot], build(ci): Bump actions/setup-python from
6.0.0 to 7.0.0
a72b46638 by dependabot[bot], build(ci): Bump actions/checkout from 5.0.0
to 7.0.1
f5aa02e15 by Dong Wang, misc: Fix typos in Task.cpp
fd022bd17 by Patrick Wilson, fix(cudf): Fix build-side visibility race in
CudfNestedLoopJoinProbe
4fc985cee by LingBin, feat(memory): Include allocator failure message in
allocateAligned error
98d131acf by LingBin, fix: Correct addLeafPool threadSafe comment in Memory
f4de7283f by Shruti Shivakumar, fix(cudf): Deduplicate probe rows in
leftSemiFilterJoin with multiple build batches
01c46469d by LingBin, fix(docs): Render '/*argName=*/' literally in
CODING_STYLE.md
f99b97da4 by Bikramjeet Singh Vig, fix(exec): Compile OutputTransportEntry
under Clang 15 (#18307)
1626c647a by RindsSchei225e, feat(dwrf): Optionally verify compressed pages
round-trip on write
c2927fc94 by Krishna Pai, fix: Restore Clang-15 nightly (scheduled Fuzzer
Jobs) build (#18081)
e698feccd by Raymond Lin, fix: PushDictionaryToRowVectorLeaves loses struct
nulls under a null-free dictionary
f73dce58b by kevinwilfong, fix(dwrf): Fix flat map struct-key UAF on
stringKeys_ realloc (#18319)
8978b89f4 by Bikramjeet Singh Vig, fix(ci): Fix pyvelox wheels workflow
(#18340)
d5ac4db8a by Ke, feat(fuzzer): Cross-config dictionary read path
verification in TableEvolutionFuzzer (#18295)
b22b52d3f by JiaKe, fix(docs): Fix configs.rst Parquet options table
rendering in GitHub preview
d47f2125d by Bikramjeet Singh Vig, test(parquet): Disable compression
round-trip test (#18330)
3edd5c7a2 by Muhammad Haseeb, feat(cudf): Support sub-splits and filters in
cudf Iceberg
dbde3f27f by PRASHANT GOLASH, refactor(nimble): Rename leftover chunkIndex
identifiers to chunkStats (#18328)
5d3e95af3 by Bikramjeet Singh Vig, fix(parquet): Preserve explicit writer
options
f1480d35c by meta-codesync[bot], perf(exec): Avoid redundant serializer
reinit in PartitionedOutput clear
b00ee2dd0 by Daniel Bauer, feat(cudf): Push down kNegatedBigintRange
subfield filter
24949d1b3 by Scottie Hanff, fix(s3): Improve error messages and retryable
detection
456fd34c2 by Karthikeyan, fix(compression): Fix LZO decompressor on ARM64
dd7d12175 by Wechar Yu, fix: Support dictionary-wrapped approx_percentile
intermediate arrays
5fd40273e by Shruti Shivakumar, fix(cudf): Use operator IDs for hash join
peer probe lookup
285e9abf5 by Bikramjeet Singh Vig, docs(blog): Compare jemalloc and mmap
allocators (#18324)
63497f6d7 by Apurva Kumar, feat: Thread per-query session credential into
Iceberg read and write paths
15b846c8c by Bikramjeet Singh Vig, fix(build): Disable
nullability-completeness in clang-tidy
3b2665a0d by Shruti Shivakumar, fix(cudf): Fix NLJ mismatch emission when
no probe/build columns in output
aa849f182 by Zhenyuan Zhao, fix: Forward an optional post-processor through
IcebergColumnHandle (#18301)
578504b72 by Karthikeyan, fix(hdfs): Fail cleanly when libhdfs cannot be
loaded
548306bab by Ping Liu, refactor: Move orc format specific configs out of
dwio::common
116e9bf06 by Apurva Kumar, fix: Correct Iceberg deletion-vector-v1 CRC-32
to match spec (#18296)
Signed-off-by: glutenperfbot <[email protected]>
---------
Signed-off-by: glutenperfbot <[email protected]>
Co-authored-by: glutenperfbot <[email protected]>
Co-authored-by: root <[email protected]>
Co-authored-by: Rui Mo <[email protected]>
Co-authored-by: BInwei Yang <[email protected]>
Co-authored-by: Rong Ma <[email protected]>
---
cpp/velox/CMakeLists.txt | 5 +++++
cpp/velox/utils/ConfigExtractor.cc | 10 ++++++++--
ep/build-velox/src/get-velox.sh | 4 ++--
3 files changed, 15 insertions(+), 4 deletions(-)
diff --git a/cpp/velox/CMakeLists.txt b/cpp/velox/CMakeLists.txt
index 78baafef8f..e347381887 100644
--- a/cpp/velox/CMakeLists.txt
+++ b/cpp/velox/CMakeLists.txt
@@ -555,6 +555,10 @@ if(ENABLE_GPU)
facebook::velox::velox_cudf_expression
${VELOX_BUILD_PATH}/velox/experimental/cudf/expression/libvelox_cudf_expression.a
)
+ import_library(
+ facebook::velox::velox_cudf_expression_registry
+
${VELOX_BUILD_PATH}/velox/experimental/cudf/expression/libvelox_cudf_expression_registry.a
+ )
import_library(
facebook::velox::velox_cudf_exec
${VELOX_BUILD_PATH}/velox/experimental/cudf/exec/libvelox_cudf_exec.a)
@@ -585,6 +589,7 @@ if(ENABLE_GPU)
facebook::velox::velox_cudf_iceberg_connector
facebook::velox::velox_cudf_hive_connector
facebook::velox::velox_cudf_expression
+ facebook::velox::velox_cudf_expression_registry
cudf::cudf
velox_curl)
endif()
diff --git a/cpp/velox/utils/ConfigExtractor.cc
b/cpp/velox/utils/ConfigExtractor.cc
index 109cf7b807..d38745416b 100644
--- a/cpp/velox/utils/ConfigExtractor.cc
+++ b/cpp/velox/utils/ConfigExtractor.cc
@@ -26,6 +26,7 @@
#include "velox/connectors/hive/HiveConfig.h"
#include "velox/connectors/hive/storage_adapters/s3fs/S3Config.h"
#include "velox/dwio/common/Options.h"
+#include "velox/dwio/dwrf/common/Config.h"
#include "velox/dwio/parquet/common/ParquetConfig.h"
namespace gluten {
@@ -244,6 +245,11 @@ std::string parquetSessionProperty(std::string_view key) {
std::string(key);
}
+std::string orcSessionProperty(std::string_view key) {
+ return
facebook::velox::dwio::common::formatConfigPrefix(facebook::velox::dwio::common::FileFormat::ORC,
"_") +
+ std::string(key);
+}
+
} // namespace
std::shared_ptr<facebook::velox::config::ConfigBase>
createHiveConnectorSessionConfig(
@@ -263,7 +269,7 @@ std::shared_ptr<facebook::velox::config::ConfigBase>
createHiveConnectorSessionC
conf->get<std::string>(kParquetMaxTargetFileSize, "0B"); // 0 means no
limit on target file size
configs[facebook::velox::connector::hive::HiveConfig::kIgnoreMissingFilesSession]
=
conf->get<bool>(kIgnoreMissingFiles, false) ? "true" : "false";
-
configs[facebook::velox::connector::hive::HiveConfig::kParquetUseColumnNamesSession]
=
+
configs[parquetSessionProperty(facebook::velox::parquet::ParquetConfig::kUseColumnNamesSession)]
=
conf->get<bool>(kParquetUseColumnNames, true) ? "true" : "false";
configs[facebook::velox::connector::hive::HiveConfig::kAllowInt32NarrowingSession]
=
conf->get<bool>(kAllowInt32Narrowing, true) ? "true" : "false";
@@ -273,7 +279,7 @@ std::shared_ptr<facebook::velox::config::ConfigBase>
createHiveConnectorSessionC
// native reader. When Spark's orc.force.positional.evolution is set, force
// position-based mapping for the whole scan by disabling name-based mapping
// (ColumnMappingMode::kPosition), matching OrcUtils.requestedColumnIds.
-
configs[facebook::velox::connector::hive::HiveConfig::kOrcUseColumnNamesSession]
=
+
configs[orcSessionProperty(facebook::velox::dwrf::Config::kOrcUseColumnNamesSession)]
=
conf->get<bool>(kOrcForcePositionalEvolution, false) ? "false" : "true";
configs[parquetSessionProperty(facebook::velox::parquet::ParquetConfig::kWriterPageSizeSession)]
=
conf->get<std::string>(kWriteParquetPageSizeBytes, "1MB");
diff --git a/ep/build-velox/src/get-velox.sh b/ep/build-velox/src/get-velox.sh
index 491d5413ca..5df0b0c28a 100755
--- 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_07_28
-VELOX_ENHANCED_BRANCH=ibm-2026_07_28
+VELOX_BRANCH=dft-2026_08_05
+VELOX_ENHANCED_BRANCH=ibm-2026_08_05
VELOX_HOME=""
RUN_SETUP_SCRIPT=ON
ENABLE_ENHANCED_FEATURES=OFF
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
