GlutenPerfBot opened a new pull request, #12928:
URL: https://github.com/apache/gluten/pull/12928

   Upstream Velox's New Commits:
   
   ```
   aa6376820 by Xiaoxuan, perf(nimble): Optimize boolean stream slicing (18748)
   f9f77bd8d by Ke, feat(nimble): Enable null generation in Nimble table 
evolution fuzzer (18715)
   4861a363b by Ke, feat(nimble fuzzer): Vary random encoding by chunk data 
(18729)
   aa0b8f884 by oerling, feat: TorchWave whole-graph AOT load/run API 
(AOTInductor alternative) (18728)
   78cc0d4f9 by Krishna Pai, build: Update FBOS to v2026.07.13.00 (18194) 
(18194)
   31626dd95 by oerling, fix(wave): Make the kernel cache safe to share between 
processes (18727)
   7a6fd80e9 by Zac, perf: Add all-null fast path to nullable materialization 
(18722)
   611009c7c by Xiaoxuan, feat(nimble): Add string shared dictionary encoding
   d6fd0ddf7 by shanhao-203, test(nimble): Exclude Huffman from writer fuzzing
   831406785 by PRASHANT GOLASH, perf(nimble): Answer integral constancy from 
min/max instead of unique counts (18676)
   3c877e616 by Rui Mo, fix(nimble): Update NullColumnReaderTest to enable 
runtime filter
   d1fbb180c by Hongze Zhang, feat: Add runtime stat 
`toIntermediateFastPathCalls` in HashAggregation
   c79c183e6 by Jiaan Geng, fix(hdfs): Retry transient read failures in 
HdfsReadFile with backoff + reopen
   24600ec80 by Xiaoxuan, perf(nimble): Preallocate index projector slice 
output buffer (18712)
   3221d03fc by Karthikeyan, fix(nimble): Satisfy gersemi and header-ownership 
pre-commit hooks
   b7be48aaa by Xiaoxuan, perf(nimble): Reduce index projector slice overhead 
(18708)
   195d3e366 by Karthikeyan, refactor(type): Drop Type.h from the 
simple-function header path so nvcc can parse them
   595a99244 by RindsSchei225e, fix(nimble): Re-size the merged flat map key 
buffer when the entry count repeats (18706)
   512d2409a by Ismaël Mejía, perf(parquet): Optimize DeltaBpDecoder skip and 
decode
   a797ecd7b by RindsSchei225e, refactor(reader): Move the reader factory out 
of fb/ (18677)
   88f949710 by Andrii Rosa, fix: Fix CacheShard::shutdown race and leak
   90167ecf7 by Ke, fix(nimble): Reject enableChunkIndex without enableChunking 
(18664)
   d6b8abf29 by Jimmy Lu, fix: Restore Huffman 12-bit guard, fix NaN stats 
compare
   97f32a018 by dependabot[bot], build(ci): Bump pypa/cibuildwheel from 4.0.0 
to 4.2.0
   b3afad61a by kevinwilfong, test(nimble): BlockBitPacking encoding does not 
support empty rows
   d142979db by Rui Mo, fix: Add split-level column mapping mode for file scans
   777dce4cf by meta-codesync[bot], Daily `arc lint --take CLANGFORMAT` (18696)
   9a3ac1535 by Ke, feat(nimble): Verify schema roundtrip and stripe group 
consistency in writer fuzzer (18671)
   437c80c2a by David-C-L, feat(nimble): SubIntSplit and Reordering 
microbenchmark drivers
   61e41f0fa by zhichenxu-meta, fix(rpc): Recover the congestion window after 
an overload backs it off (18603)
   eb846b0fd by Naveen Mahadevuni, fix(iceberg): Field id based matching for 
equality delete columns
   6449fafa0 by RindsSchei225e, fix(nimble): Fix null population when 
abandoning the dictionary (18628)
   4900660b1 by Kuo Zhao, fix(sparksql): Add high-precision DECIMAL to 
REAL/DOUBLE cast aligned with Spark
   062958646 by Ke, feat(nimble): Verify column statistics in writer fuzzer 
(18666)
   3bf3db72a by XiaoDu, feat(nimble): Make FOR read-only while keeping PFOR 
writable (18657)
   2b5cbb2eb by oerling, feat(torchwave): Allocation groups for shared 
lifetimes and fused concats (18668)
   ```
   
   velox_branch: https://github.com/IBM/velox/commits/dft-2026_08_30
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]


---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to