3.3.1 was just released, FTR :)

@github-actions crossbow submit -g cpp

@ursabot please benchmark lang=C++

Benchmark runs are scheduled for commit dd3ce38. Watch https://buildkite.com/apache-arrow and https://conbench.arrow-dev.org for updates. A comment will be posted here when the runs are complete.

Revision: dd3ce38

Submitted crossbow builds: ursacomputing/crossbow @ actions-a1dcd2c5d1

Task	Status
example-cpp-minimal-build-static
example-cpp-minimal-build-static-system-dependency
example-cpp-tutorial
test-build-cpp-fuzz
test-conda-cpp
test-conda-cpp-valgrind
test-debian-13-cpp-amd64
test-debian-13-cpp-i386
test-debian-experimental-cpp-gcc-15
test-fedora-42-cpp
test-ubuntu-22.04-cpp
test-ubuntu-22.04-cpp-20
test-ubuntu-22.04-cpp-bundled
test-ubuntu-22.04-cpp-emscripten
test-ubuntu-22.04-cpp-no-threading
test-ubuntu-24.04-cpp
test-ubuntu-24.04-cpp-bundled-offline
test-ubuntu-24.04-cpp-gcc-13-bundled
test-ubuntu-24.04-cpp-gcc-14
test-ubuntu-24.04-cpp-minimal-with-formats
test-ubuntu-24.04-cpp-thread-sanitizer

Thanks for your patience. Conbench analyzed the 3 benchmarking runs that have been run so far on PR commit dd3ce38.

There were 13 benchmark results indicating a performance regression:

• Pull Request Run on amd64-m5-4xlarge-linux at 2026-04-27 09:08:52Z
  • TouchArea (C++) with params=size:16777216/real_time/threads:4, source=cpp-micro, suite=arrow-memory-pool-benchmark
  • TouchArea (C++) with params=size:1048576/real_time/threads:32, source=cpp-micro, suite=arrow-memory-pool-benchmark
• and 11 more (see the report linked below)

The full Conbench report has more details.

There seems to be a big regression on the memory pool benchmark on the TouchArea benchmark.
An example of one of the tests:
From mean: 110,760,324,636.8867 B/s
To mean: 84,367,944,824.52159 B/s
-23.83%
If I am interpreting the numbers correctly.

It's a very synthetic benchmark and doesn't measure any real-world operation, so I think we can ignore it.

It's a very synthetic benchmark and doesn't measure any real-world operation, so I think we can ignore it.

ok, then I don't think the rest of CI failures are related.