GH-51254: [C++][CI] Move ubuntu-cpp-bundled-offline to C++ Extra

.github/workflows/cpp_extra.yml

@@ -113,6 +113,11 @@ jobs:
               -e BUILD_WARNING_LEVEL=PRODUCTION
             runs-on: "runs-on=${{ github.run_id }}/family=x8i.2xlarge/volume=80gb/spot=capacity-optimized"
             title: AMD64 Ubuntu Large Memory Tests
+          - image: ubuntu-cpp-bundled-offline
+            runs-on: ubuntu-latest
+            envs:
+              - UBUNTU=24.04
+            title: AMD64 Ubuntu Bundled Offline
           - image: conda-cpp
             run-options: >-
               -e ARROW_USE_MESON=ON

ci/docker/alpine-linux-3.22-cpp.dockerfile

@@ -104,7 +104,6 @@ ENV ARROW_ACERO=ON \
     AWSSDK_SOURCE=BUNDLED \
     google_cloud_cpp_storage_SOURCE=BUNDLED \
     MUSL_LOCPATH=/usr/share/i18n/locales/musl \
-    PATH=/usr/lib/ccache/bin:$PATH \
     # We can remove this once
     # https://gitlab.alpinelinux.org/alpine/aports/-/work_items/18353
     # is fixed.

ci/docker/debian-13-cpp.dockerfile

@@ -137,7 +137,6 @@ ENV ARROW_ACERO=ON \
     Azure_SOURCE=BUNDLED \
     google_cloud_cpp_storage_SOURCE=BUNDLED \
     ORC_SOURCE=BUNDLED \
-    PATH=/usr/lib/ccache/:$PATH \
     PYTHON=python3 \
     simdjson_SOURCE=BUNDLED \
     xsimd_SOURCE=BUNDLED

ci/docker/debian-experimental-cpp.dockerfile

@@ -143,6 +143,5 @@ ENV ARROW_ACERO=ON \
     CXX=g++${gcc:+-${gcc}} \
     google_cloud_cpp_storage_SOURCE=BUNDLED \
     ORC_SOURCE=BUNDLED \
-    PATH=/usr/lib/ccache/:$PATH \
     PYTHON=python3 \
     xsimd_SOURCE=BUNDLED

ci/docker/fedora-42-cpp.dockerfile

@@ -109,7 +109,6 @@ ENV ARROW_ACERO=ON \
     opentelemetry_cpp_SOURCE=BUNDLED \
     PARQUET_BUILD_EXAMPLES=ON \
     PARQUET_BUILD_EXECUTABLES=ON \
-    PATH=/usr/lib/ccache/:$PATH \
     PYARROW_TEST_GANDIVA=OFF \
     simdjson_SOURCE=BUNDLED \
     xsimd_SOURCE=BUNDLED

ci/docker/ubuntu-22.04-cpp-minimal.dockerfile

@@ -108,5 +108,4 @@ ENV ARROW_ACERO=ON \
     CMAKE_GENERATOR="Unix Makefiles" \
     PARQUET_BUILD_EXAMPLES=ON \
     PARQUET_BUILD_EXECUTABLES=ON \
-    PATH=/usr/lib/ccache/:$PATH \
     PYTHON=python3

ci/docker/ubuntu-22.04-cpp.dockerfile

@@ -225,7 +225,6 @@ ENV absl_SOURCE=BUNDLED \
     ORC_SOURCE=BUNDLED \
     PARQUET_BUILD_EXAMPLES=ON \
     PARQUET_BUILD_EXECUTABLES=ON \
-    PATH=/usr/lib/ccache/:$PATH \
     PYTHON=python3 \
     simdjson_SOURCE=BUNDLED \
     xsimd_SOURCE=BUNDLED

ci/docker/ubuntu-24.04-cpp-minimal.dockerfile

@@ -105,5 +105,4 @@ ENV ARROW_ACERO=ON \
     CMAKE_GENERATOR="Unix Makefiles" \
     PARQUET_BUILD_EXAMPLES=ON \
     PARQUET_BUILD_EXECUTABLES=ON \
-    PATH=/usr/lib/ccache/:$PATH \
     PYTHON=python3

ci/docker/ubuntu-24.04-cpp.dockerfile

@@ -219,7 +219,6 @@ ENV absl_SOURCE=BUNDLED \
     ORC_SOURCE=BUNDLED \
     PARQUET_BUILD_EXAMPLES=ON \
     PARQUET_BUILD_EXECUTABLES=ON \
-    PATH=/usr/lib/ccache/:$PATH \
     PYTHON=python3 \
     simdjson_SOURCE=BUNDLED \
     xsimd_SOURCE=BUNDLED

compose.yaml

@@ -514,7 +514,6 @@ services:
       ARROW_OFFLINE: ON
       # Apache ORC always uses external orc-format.
       ARROW_ORC: OFF
-      CMAKE_GENERATOR: "Unix Makefiles"
     volumes: *ubuntu-volumes
     command: *cpp-command
 

cpp/CMakeLists.txt

@@ -263,6 +263,9 @@ if(ARROW_USE_CCACHE
    AND NOT CMAKE_C_COMPILER_LAUNCHER
    AND NOT CMAKE_CXX_COMPILER_LAUNCHER)
 
+  # XXX This does not detect if ccache is enabled externally, for example
+  # by adding `/usr/lib/ccache` to PATH in which case `cc` resolves to
+  # `/usr/lib/ccache/cc`. Such double-wrapping of ccache is suboptimal.
   find_program(CCACHE_FOUND ccache)
 
   if(CCACHE_FOUND)

cpp/thirdparty/download_dependencies.sh

@@ -38,6 +38,8 @@ download_dependency() {
   local url=$1
   local out=$2
 
+  echo "Downloading ${url}" 1>&2
+
   wget --quiet --continue --output-document="${out}" "${url}" || \
     (echo "Failed downloading ${url}" 1>&2; exit 1)
 }

dev/tasks/tasks.yml

@@ -382,15 +382,6 @@ tasks:
         UBUNTU: 22.04
       image: ubuntu-cpp-bundled
 
-  test-ubuntu-24.04-cpp-bundled-offline:
-    ci: github
-    template: docker-tests/github.linux.yml
-    params:
-      env:
-        UBUNTU: 24.04
-      image: ubuntu-cpp-bundled-offline
-      timeout: 120
-
   test-ubuntu-24.04-cpp-gcc-13-bundled:
     ci: github
     template: docker-tests/github.linux.yml