After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit a625340.

Benchmarks with performance regressions

There were 0 possible performance regressions, according to the lookback z-score method.

All benchmark runs analyzed:

• Commit Run on arm64-t4g-2xlarge-linux at 2025-10-22 17:13:38Z
• Commit Run on amd64-m5-4xlarge-linux at 2025-10-22 17:30:23Z
• Commit Run on amd64-c6a-4xlarge-linux at 2025-10-22 16:52:24Z
• Commit Run on test-mac-arm at 2025-10-22 21:45:13Z

Unstable benchmarks with performance regressions

The following benchmark results indicate a possible performance regression, but are known to sometimes produce false positives when applying the lookback z-score analysis.

• Commit Run on amd64-m5-4xlarge-linux at 2025-10-22 17:30:23Z

  • BM_UnpackUint32 (C++) with params=Avx512Unaligned/8/64, source=cpp-micro, suite=arrow-bpacking-benchmark
  • BM_UnpackUint32 (C++) with params=Avx2Unaligned/8/64, source=cpp-micro, suite=arrow-bpacking-benchmark
  • FloatParsing (C++) with params=, source=cpp-micro, suite=arrow-value-parsing-benchmark
  • AllocateDeallocate (C++) with params=/size:4096/real_time/threads:1, source=cpp-micro, suite=arrow-memory-pool-benchmark

• Commit Run on amd64-c6a-4xlarge-linux at 2025-10-22 16:52:24Z

  • TakeStringMonotonicIndices (C++) with params=524288/1, source=cpp-micro, suite=arrow-compute-vector-selection-benchmark
  • ArrayArrayKernel (C++) with params=<DivideChecked, Int32Type>/size:524288/inverse_null_proportion:0, source=cpp-micro, suite=arrow-compute-scalar-arithmetic-benchmark
  • BM_UnpackUint64 (C++) with params=Avx2Unaligned/1/1024, source=cpp-micro, suite=arrow-bpacking-benchmark
  • ArrayArrayKernel (C++) with params=<Divide, Int32Type>/size:524288/inverse_null_proportion:0, source=cpp-micro, suite=arrow-compute-scalar-arithmetic-benchmark
  • TakeFSLInt64MonotonicIndices (C++) with params=524288/1, source=cpp-micro, suite=arrow-compute-vector-selection-benchmark
  • CaseWhenBenchStringContiguous (C++) with params=65536/99, source=cpp-micro, suite=arrow-compute-scalar-if-else-benchmark
  • BM_UnpackUint64 (C++) with params=Avx2Unaligned/2/64, source=cpp-micro, suite=arrow-bpacking-benchmark
  • ArrayArrayKernel (C++) with params=<DivideChecked, Int32Type>/size:524288/inverse_null_proportion:100, source=cpp-micro, suite=arrow-compute-scalar-arithmetic-benchmark
  • CaseWhenBenchStringContiguous (C++) with params=65536/0, source=cpp-micro, suite=arrow-compute-scalar-if-else-benchmark
  • TakeFSLInt64RandomIndicesWithNulls (C++) with params=524288/1, source=cpp-micro, suite=arrow-compute-vector-selection-benchmark
  • TakeFSLInt64RandomIndicesNoNulls (C++) with params=524288/1, source=cpp-micro, suite=arrow-compute-vector-selection-benchmark