GH-48921: [C++] Bump mimalloc to 3.2.7

[pitrou]

Rationale for this change

Bump mimalloc to the latest version.

On Linux platforms, this seems to make some Parquet reading benchmarks significantly faster.

Are these changes tested?

By existing CI builds.

Are there any user-facing changes?

Hopefully no regressions, but memory allocation behavior may change.

• GitHub Issue: [C++] Bump mimalloc to 3.2.7 #48921

[github-actions]

Thanks for opening a pull request!

If this is not a minor PR. Could you open an issue for this pull request on GitHub? https://github.com/apache/arrow/issues/new/choose

Opening GitHub issues ahead of time contributes to the Openness of the Apache Arrow project.

Then could you also rename the pull request title in the following format?

GH-${GITHUB_ISSUE_ID}: [${COMPONENT}] ${SUMMARY}

or

MINOR: [${COMPONENT}] ${SUMMARY}

See also:

• Other pull requests
• Contribution Guidelines - Contributing Overview

[pitrou]

@github-actions crossbow submit -g cpp

[pitrou]

@github-actions crossbow submit wheelcp314*

[pitrou]

@ursabot please benchmark

[pitrou]

Two CI failures are relevant, I've opened upstream issues:

• assertion failures on Debian i386
• crash on macOS if forking from non-main thread

[pitrou]

@ursabot please benchmark

[pitrou]

@github-actions crossbow submit -g cpp

[pitrou]

@github-actions crossbow submit wheelcp314*

[pitrou]

I've diagnosed the test freezes on Debian i386 and reported the issue.

[pitrou]

@github-actions crossbow submit -g r

[pitrou]

@github-actions crossbow submit -g cpp

[pitrou]

@github-actions crossbow submit -g packaging

[pitrou]

@github-actions crossbow submit -g r -g cpp

[github-actions]

Revision: cfa3114

Submitted crossbow builds: ursacomputing/crossbow @ actions-0fbdfc8620

Task	Status
example-cpp-minimal-build-static
example-cpp-minimal-build-static-system-dependency
example-cpp-tutorial
r-binary-packages
r-recheck-most
test-build-cpp-fuzz
test-conda-cpp
test-conda-cpp-valgrind
test-debian-12-cpp-amd64
test-debian-12-cpp-i386
test-debian-experimental-cpp-gcc-15
test-fedora-42-cpp
test-r-alpine-linux-cran
test-r-arrow-backwards-compatibility
test-r-depsource-system
test-r-dev-duckdb
test-r-devdocs
test-r-extra-packages
test-r-gcc-11
test-r-gcc-12
test-r-install-local
test-r-install-local-minsizerel
test-r-linux-as-cran
test-r-linux-rchk
test-r-linux-sanitizers
test-r-linux-valgrind
test-r-m1-san
test-r-macos-as-cran
test-r-offline-maximal
test-r-ubuntu-22.04
test-r-versions
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

[pitrou]

@ursabot please benchmark lang=C++

[rok]

Benchmark runs are scheduled for commit 275a07b. 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.

[pitrou]

@rok trying to benchmark this one

[pitrou]

The C++ benchmarks seem to have run correctly. The R benchmarks can fail because of #48912

[rok]

The C++ benchmarks seem to have run correctly. The R benchmarks can fail because of #48912

Ah interesting, I was about to bring this one up.

[zanmato1984]

+1

[conbench-apache-arrow]

Thanks for your patience. Conbench analyzed the 3 benchmarking runs that have been run so far on PR commit 275a07b.

There were 3 benchmark results indicating a performance regression:

• Pull Request Run on amd64-c6a-4xlarge-linux at 2026-01-21 13:06:17Z
  • CopyBitmapWithoutOffset (C++) with params=8192, source=cpp-micro, suite=arrow-bit-util-benchmark
  • MemoryBandwidth (C++) with params=/33554432/real_time/threads:1, source=cpp-micro, suite=arrow-io-memory-benchmark
• and 1 more (see the report linked below)

The full Conbench report has more details.