After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit eaeb64b.

Benchmarks with performance regressions

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

All benchmark runs analyzed:

• Commit Run on amd64-c6a-4xlarge-linux at 2025-04-03 09:04:28Z
• Commit Run on arm64-t4g-2xlarge-linux at 2025-04-03 10:36:59Z
• Commit Run on amd64-m5-4xlarge-linux at 2025-04-03 09:47:43Z
• Commit Run on test-mac-arm at 2025-04-03 10:28:42Z

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-c6a-4xlarge-linux at 2025-04-03 09:04:28Z

  • IntegerParsing (C++) with params=, source=cpp-micro, suite=arrow-value-parsing-benchmark
  • ReferenceWorkloadCost (C++) with params=task_cost:1000/real_time, source=cpp-micro, suite=arrow-thread-pool-benchmark

• Commit Run on arm64-t4g-2xlarge-linux at 2025-04-03 10:36:59Z

  • CopyBitmapWithOffsetBoth (C++) with params=8192, source=cpp-micro, suite=arrow-bit-util-benchmark
  • BenchmarkBitmapAnd (C++) with params=131072/2, source=cpp-micro, suite=arrow-bit-util-benchmark
  • BenchmarkBitmapVisitBitsetAnd (C++) with params=32768/0, source=cpp-micro, suite=arrow-bit-util-benchmark
  • BenchmarkBitmapVisitBitsetAnd (C++) with params=131072/0, source=cpp-micro, suite=arrow-bit-util-benchmark
  • BenchmarkBitmapAnd (C++) with params=32768/1, source=cpp-micro, suite=arrow-bit-util-benchmark

• Commit Run on amd64-m5-4xlarge-linux at 2025-04-03 09:47:43Z

  • ExecuteScalarExpressionBaseline (C++) with params=/rows_per_batch:100000/real_time/threads:1, source=cpp-micro, suite=arrow-acero-expression-benchmark
  • BM_ByteStreamSplitEncode_Float_Sse2 (C++) with params=1024, source=cpp-micro, suite=parquet-encoding-benchmark
  • ExecuteScalarExpressionBaseline (C++) with params=/rows_per_batch:1000000/real_time/threads:1, source=cpp-micro, suite=arrow-acero-expression-benchmark
  • ExecuteScalarExpressionBaseline (C++) with params=/rows_per_batch:10000/real_time/threads:1, source=cpp-micro, suite=arrow-acero-expression-benchmark
  • ExecuteScalarExpressionBaseline (C++) with params=/rows_per_batch:1000/real_time/threads:1, source=cpp-micro, suite=arrow-acero-expression-benchmark

• Commit Run on test-mac-arm at 2025-04-03 10:28:42Z

  • FilterOverhead (C++) with params=simple_expression/batch_size:10000/null_prob:0/bool_true_prob:50/real_time, source=cpp-micro, suite=arrow-acero-filter-benchmark
  • FilterOverhead (C++) with params=complex_expression/batch_size:1000/null_prob:0/bool_true_prob:50/real_time, source=cpp-micro, suite=arrow-acero-filter-benchmark
  • FilterOverhead (C++) with params=selectivity_benchmark/batch_size:1000/null_prob:75/bool_true_prob:50/real_time, source=cpp-micro, suite=arrow-acero-filter-benchmark
  • ListSliceInt64ListViewWithStepAndStop (C++) with params=65536/10, source=cpp-micro, suite=arrow-compute-scalar-list-benchmark
  • FilterOverhead (C++) with params=ref_only_expression/batch_size:100000/null_prob:0/bool_true_prob:50/real_time, source=cpp-micro, suite=arrow-acero-filter-benchmark