[GLUTEN-6887][VL] Daily Update Velox Version (2026_08_21)

[GlutenPerfBot]

Upstream Velox's New Commits:

522ee471d7 by zhichenxu-meta, fix(rpc): Release retained input vectors in RPCOperator::close() (#18610)
4b1d8eaf1e by oerling, feat(torchwave): CUDA-event stream ordering and GPU-idle timing
82187f392e by Maria Basmanova, fix: Coerce non-boolean AND and OR arguments to boolean (#18605)
c3413b270e by Han-Yin Chang, fix: Reject trailing date-time separator in Spark CAST(string as timestamp)
7ea2280899 by Maria Basmanova, fix: Coerce a non-boolean SWITCH condition to boolean
0d4e581aff by Karthikeyan, build(hdfs): Download aarch64 Hadoop native libs on ARM64
9019e37a17 by LingBin, refactor: Replace raw glog DCHECK with VELOX_DCHECK in velox/vector
d22adf1242 by Wechar Yu, refactor: Minor normalize gtest names for core
13acea01ab by Wechar Yu, refactor: Minor normalize gtest names for experimental dir
9be92d7a49 by Krishna Pai, feat(nimble): Add PFOR encoding benchmark (#18543)
5071055c26 by Minni Mittal, feat: Make Spark CAST(string as REAL/DOUBLE) ANSI-compliant
ffd016a0fa by Raymond Lin, feat(hive): Report per-operation latency and retry counts
7c2b38147b by XiaoDu, feat(nimble): Keep deprecated encodings read-only for PFOR encoding (#18547)
62b178b8fb by mohsaka, fix(parquet): Null-fill missing columns correctly in kParquetFieldId mode
11dc0d3498 by Hongze Zhang, feat: Support toIntermediate in AggregateCompanionAdapter::PartialFunction
f2f78eabbf by Mohammad Linjawi, fix(expr): Preserve ordinal field references in simplified evaluation
2ca5030f39 by Krishna Pai, build: Bump the OpenZL pin to the revision fbcode builds against (#18576)
ab9856565f by Natasha Sehgal, refactor: Drop the prefix from TrackedExecutor::reportTo
9de6a1ff37 by Kostas Xirogiannopoulos, refactor(velox): Extract shared HiveHash primitives into functions/lib/HiveHash.h (#18193)
1f971d3721 by Felipe Pessoto, fix(dwio): Correct row-index size when a scan has a filter and no child readers
d7b3de73c1 by Mohammad Linjawi, feat: Add Spark CAST(string as TIME) with ANSI support
73ac8855a1 by Rui Mo, docs: Update Spark function documents
ca800d0a95 by Yotam Perlitz, feat(cudf): Support masked (FILTER) aggregation for sum/count/min/max
3c3fb1605e by Daniel Bauer, refactor(exec): Rename ExchangeClient to InMemoryExchangeClient
b0896f95cd by Philo He, fix(spark): Enable ANSI-compliant cast from BOOLEAN to DECIMAL
38d78c605b by RindsSchei225e, fix(fuzzer): Emit the number of aggregates the fuzzer decides on (#18564)
c1d8445d3c by RindsSchei225e, feat(fuzzer): Run TableEvolutionFuzzer query shapes against an existing file
bbebba0085 by LingBin, refactor(memory): Use VELOX_MEM_LOG in Memory and MemoryArbitrator
f72fb2f247 by Zhu, Lipeng, perf(HashTable): K-way chain walk with per-slot buffered flush
cc87a86b50 by LingBin, fix(memory): Make the global arbitration thread name fully visible
8612db55bb by Patrick Ribbsaeter, fix(exec): Emit TopNRowNumber in-memory partitions in ascending rank order (#18494)
4ab29d9dc6 by dependabot[bot], build(ci): Bump actions/cache/save from 5.0.5 to 6.1.0
e60330d7c0 by Wechar Yu, fix(ci): Run clang-tidy for Breeze separately
02d85be583 by Ping Liu, fix: UBSan misaligned-load failures
a985eb8931 by Ismaël Mejía, feat(parquet): Add BYTE_STREAM_SPLIT decoder and INT64 decimal bulk path
0db6714253 by Karthikeyan, fix(fuzzer): Exclude UNKNOWN base type from HLL input generator
15ee98e650 by dependabot[bot], build(deps): Bump brace-expansion from 1.1.14 to 1.1.18 in /website
76b8c194c5 by Pradeep Vaka, fix(hive): Round-trip all FileProperties in split serde (#18481)
309858b751 by Rui Mo, feat: Add HUGEINT support to VectorHasher's distinct value-id path
a71143c57 by Hongze Zhang, build: Add support for LTO
8a6663d58 by Xiaoxuan, feat(nimble): Optimize result adapter batch decoding (18548)
9aa400146 by XiaoDu, fix(nimble): Apply encoding layouts to predefined FlatMap keys (18549)
44329572a by Mohammad Linjawi, build(nimble): Track benchmark data headers in CMake (18538)
de642fd73 by Rui Mo, fix: Filter on constant ScanSpec fields
375eeccf8 by Minni Mittal, feat(spark): Add ANSI mode support for CAST(NumericType as integral)
c4299b70b by Ismaël Mejía, fix(parquet): Skip fixed-length byte arrays without a length prefix
d242cac62 by Karthikeyan, fix(test): Load lazy vectors before DuckDB ingestion in test harness
54fea71cc by Rohit Jain, feat(nimble): Move Nimble into Velox (18468)
b83147180 by Joe Abraham, refactor(test): Rename outputType to targetColumns
1e5bfb8c6 by Devavret Makkar, fix(cuDF): Fix cuDF local exchange memory accounting

velox_branch: https://github.com/IBM/velox/commits/dft-2026_08_21

Related issue: #6887

[felipepessoto]

It doesn't contain the recent commits from Velox?

[zhouyuan]

https://github.com/apache/gluten/actions/runs/32488348268/job/96817237534?pr=12846

- Decimal type *** FAILED ***
  org.apache.spark.SparkException: Job aborted due to stage failure: Task 0 in stage 370.0 failed 1 times, most recent failure: Lost task 0.0 in stage 370.0 (TID 301) (8b2b40bc1b7f executor driver): org.apache.gluten.exception.GlutenException: org.apache.gluten.exception.GlutenException: Exception: VeloxUserError
Error Source: USER
Error Code: UNSUPPORTED
Reason: Filter(HugeintValuesUsingHashTable, deterministic, no nulls): mergeWith() is not supported.
Retriable: False
Function: mergeWith
File: /work/ep/build-velox/build/velox_ep/./velox/type/Filter.h
Line: 281
Stack trace:

the failure seems related with facebookincubator/velox#18159
Cc: @rui-mo

[github-actions]

Run Gluten Clickhouse CI on x86

[felipepessoto]

@ zhouyuan the PR description is updated with new commit, but the branch https://github.com/IBM/velox/commits/dft-2026_08_21
isn’t.

Could you confirm if the list in the PR description is the correct one? Asking because I have a dependency on 1f971d3721

update: nvm, the branch is updated now (not sure if I looked the wrong place yesterday)

[rui-mo]

Thanks @zhouyuan. Working on a fix in facebookincubator/velox#18626.

[prestodb-ci]

Rebase job