GH-50471: [C++][Python] Bump bundled mimalloc to 3.4.1

[raulcd]

Rationale for this change

A new version of mimalloc was released. The existing version of mimalloc can cause a segmentation fault as seen on the referenced issue.

What changes are included in this PR?

Bumping the mimalloc version.

Are these changes tested?

Yes, via CI and archery.
I've also tested the minimal reproducer from the issue:

import threading

def imp():
    import pyarrow  # dlopens libarrow -> bundled mimalloc initializes on this thread

def alloc():
    import pyarrow as pa
    pa.array(["x"] * 5000)

a = threading.Thread(target=imp); a.start(); a.join()
b = threading.Thread(target=alloc); b.start(); b.join()

The above cause a segfault on every execution with pyarrow 25.0.0 on a local Python 3.14. With the new built wheel with the new mimalloc version, I haven't been able to reproduce the segfault after several attempts.

Are there any user-facing changes?

No

• GitHub Issue: [C++][Python] SIGSEGV in bundled mimalloc mi_thread_init when libarrow is first loaded on a non-main thread that exits (mimalloc 3.3.x, pyarrow 25.0.0) #50471

[github-actions]

⚠️ GitHub issue #50471 has been automatically assigned in GitHub to PR creator.

[Copilot]

Pull request overview

Updates the bundled mimalloc version used by Arrow’s C++ (and thus Python wheel) builds to address a known multi-threaded initialization crash scenario.

Changes:

• Bump ARROW_MIMALLOC_BUILD_VERSION from v3.3.1 to v3.4.1.
• Update the corresponding SHA256 checksum for the mimalloc source archive.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated no new comments.

[raulcd]

Unfortunately the two failures are suspicious and look related. Those seeming to affect only macOS arm64 doesn't make it easy to debug/diagnose.

[Copilot]

Pull request overview

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

[raulcd]

I'll close this PR as there's a second PR from @pitrou that is covering similar problems. Both the initial fix on that PR and the bump seem required to be done together:

• GH-50428: [C++] Better mimalloc configuration on macOS #50549