[GLUTEN-6887][VL] Daily Update Velox Version (2026_06_06)

[rui-mo]

Upstream Velox's New Commits:

85f2cd5d0 by Rui Mo, feat: Enable TIMESTAMP_UTC in CompactRow and UnsafeRowFast (17643)
b6ca0a902 by sungwoo-XCENA, fix: Add back CustomMemoryResource with bug fix (17698)
64d7b5e00 by Shaojie Li, perf(parquet): Batched DELTA decoding with SIMD bit-unpack (17728)
ebe134939 by Muhammad Haseeb, feat(cudf): Enable ABFS via BufferedInput in Velox-cudf (17562)
5e7b483de by Bradley Dice, fix: Remove Fedora tzdata pin (17739)
6edea8f8a by Zac Wen, fix: Return done instead of nullopt from HiveIndexSource EmptyIterator (17740)
1e1738388 by Henry Edwin Dikeman, fix(expr): Push OR of varchar ranges down as subfield filter (17736)
9b8ea5d85 by Huameng (Michael) Jiang, refactor(writer): Clean up writer performance stats (17719)
7e7342162 by Kevin Wilfong, fix: Initialize null children in SelectiveMapAsStructColumnReader getValues (17720)
fd8f0cf07 by dependabot[bot], build(ci): Bump docker/setup-qemu-action from 3.6.0 to 4.1.0 (17692)
bfaf4056e by lifulong, refactor(parquet): Refactor parquet reader test (17560)
196cdc238 by dependabot[bot], build(ci): Bump docker/login-action from 3.6.0 to 4.2.0 (17691)
2bac72430 by Mariam-Almesfer, feat: Support Spark hour(timestamp_utc) function (17557)
d20ef330a by Minni Mittal, feat(spark): Add 2-arg decimal ceil/floor special forms for Spark (17664)
72ac75b61 by Muhammad Haseeb, fix(cudf): Run Velox-cudf iceberg tests in CI (17704)
ae29a679d by Reema Alzaid, feat(cudf): Add GPU array access support for element_at, subscript, and get (17243)
f03e832cc by brijrajk, feat: Register Spark transform_values function (17669)
9092f52da by Rui Mo, fix: Skip the fuzzer test for Spark regexp_instr function (17709)
8092d7ea2 by Rui Mo, feat: Allow subfield rename and deletion for ORC format (17551)
2eabd14b2 by Zac Wen, fix: Handle NULL partition values in makePartitionKey (17723)
0dbf1731f by Rui Mo, feat: Support partition value of TIMESTAMP_UTC type (17631)
42b0ac8c4 by Apurva Kumar, feat: Populate recordCount on DWRF/ORC IcebergDataSink commits (17654)
058b1a314 by Daniel Bauer, build: Exclude ucx-exchange sources from clang-tidy (17715)
e3a3fc318 by Daniel Bauer, feat: Add per-node transport type annotations to PlanFragment (17567)
3c55d6639 by Pratik Pugalia, fix(spark): Handle timezone gaps in DATE to TIMESTAMP cast (17660)

velox_branch: https://github.com/IBM/velox/commits/dft-2026_06_06

Related issue: #6887

[Copilot]

Pull request overview

Updates Gluten’s Velox integration to track the next daily Velox branches and adjusts a Hive connector default to match updated Parquet behavior expectations for missing struct fields.

Changes:

• Bump the default Velox and “enhanced” Velox branches in the Velox fetch script to 2026_06_06.
• Enable HiveConfig::kParquetNullStructForMissingFields by default in the Hive connector config map.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated no comments.

File	Description
ep/build-velox/src/get-velox.sh	Updates the default Velox branch pins to the 2026_06_06 daily branches.
cpp/velox/utils/ConfigExtractor.cc	Sets a new default Hive connector config for Parquet handling of missing struct fields.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[prestodb-ci]

Rebase job