MINOR: [Dev][C++] Allow ubuntu-cpp-thread-sanitizer Docker build with Ubuntu 24.04

[pitrou]

Install the clang-rt libraries that are necessary to link Thread Sanitizer-enabled binaries. Also fix use of deprecated BufferReader constructor in some tests, so that compilation with CLang 18 succeeds.

Note that the C++ test suite still fails on Flight tests, as tracked in GH-36552.

[pitrou]

cc @kou

[pitrou]

@github-actions crossbow submit ubuntucpp*

[github-actions]

Revision: 0019b58

Submitted crossbow builds: ursacomputing/crossbow @ actions-d48e76f193

Task	Status
test-ubuntu-20.04-cpp
test-ubuntu-20.04-cpp-bundled
test-ubuntu-20.04-cpp-minimal-with-formats
test-ubuntu-20.04-cpp-thread-sanitizer
test-ubuntu-22.04-cpp
test-ubuntu-22.04-cpp-20
test-ubuntu-22.04-cpp-emscripten
test-ubuntu-22.04-cpp-no-threading
test-ubuntu-24.04-cpp
test-ubuntu-24.04-cpp-gcc-13-bundled
test-ubuntu-24.04-cpp-gcc-14

[kou]

Do we need this too?

diff --git a/dev/tasks/tasks.yml b/dev/tasks/tasks.yml
index 6e1f7609a9..50ef83905d 100644
--- a/dev/tasks/tasks.yml
+++ b/dev/tasks/tasks.yml
@@ -1168,16 +1168,16 @@ tasks:
       flags: "-e ARROW_ENABLE_THREADING=OFF"
       image: ubuntu-cpp
 
-  test-ubuntu-20.04-cpp-thread-sanitizer:
+  test-ubuntu-24.04-cpp-thread-sanitizer:
     ci: github
     template: docker-tests/github.linux.yml
     params:
       env:
         # clang-tools and llvm version need to be synchronized so as
         # to have the right llvm-symbolizer version
-        CLANG_TOOLS: 11
-        LLVM: 11
-        UBUNTU: 20.04
+        CLANG_TOOLS: 18
+        LLVM: 18
+        UBUNTU: 24.04
       image: ubuntu-cpp-thread-sanitizer
 
   test-ubuntu-20.04-cpp-minimal-with-formats:

[pitrou]

We can't because of #36552

[pitrou]

Wait, let me try with those LLVM/CLang versions

[pitrou]

Indeed, the Flight tests still fail with those versions. However, CLang 18 also surfaces some minor C++ issues that I can fix here.

[pitrou]

@kou Do these changes look ok?

[kou]

+1

I haven't tested this on local but it will work.

[conbench-apache-arrow]

After merging your PR, Conbench analyzed the 4 benchmarking runs that have been run so far on merge-commit 7c8909a.

There were no benchmark performance regressions. 🎉

The full Conbench report has more details. It also includes information about 1 possible false positive for unstable benchmarks that are known to sometimes produce them.