Test R package with Alpine Linux

.github/workflows/r_tests.yml

@@ -2,116 +2,140 @@ name: XGBoost-R-Tests
 
 on: [push, pull_request]
 
-env:
-  GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
-
 permissions:
   contents: read # to fetch code (actions/checkout)
 
 concurrency:
   group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref }}
   cancel-in-progress: true
 
+env:
+  BRANCH_NAME: >-
+    ${{ github.event.pull_request.number && 'PR-' }}${{ github.event.pull_request.number || github.ref_name }}
+  USE_DOCKER_CACHE: 1
+  GITHUB_PAT: ${{ secrets.GITHUB_TOKEN }}
+
 jobs:
-  test-Rpkg:
-    runs-on: ${{ matrix.os }}
-    name: Test R on OS ${{ matrix.os }}, R ${{ matrix.r }}, Compiler ${{ matrix.compiler }}, Build ${{ matrix.build }}
+  build-containers:
+    name: Build CI containers (${{ matrix.container_id }})
+    runs-on:
+      - runs-on
+      - runner=linux-amd64-cpu
+      - run-id=${{ github.run_id }}
+      - tag=r-tests-build-containers-${{ matrix.container_id }}
+    strategy:
+      fail-fast: false
+      matrix:
+        container_id:
+          - xgb-ci.gpu_build_r_rockylinux8
+          - xgb-ci.alpine
+    steps:
+      # Restart Docker daemon so that it recognizes the ephemeral disks
+      - run: sudo systemctl restart docker
+      - uses: actions/checkout@v4
+        with:
+          submodules: "true"
+      - name: Build ${{ matrix.container_id }}
+        run: bash ops/docker_build.sh ${{ matrix.container_id }}
+  test-rpkg:
+    name: Test R on ${{ matrix.name }}
+    needs: build-containers
+    runs-on:
+      - runs-on=${{ github.run_id }}
+      - runner=linux-amd64-cpu
+      - tag=r-tests-test-rpkg-${{ matrix.name }}
     strategy:
       fail-fast: false
       matrix:
         include:
-          - os: windows-latest
-            r: release
-            compiler: mingw
-            build: autotools
-          - os: ubuntu-latest
-            r: release
-            compiler: none
-            build: cmake
+          - name: rockylinux-cmake
+            container_id: xgb-ci.gpu_build_r_rockylinux8
+            build_tool: cmake
+          - name: rockylinux-autotools
+            container_id: xgb-ci.gpu_build_r_rockylinux8
+            build_tool: autotools
+          - name: alpine-autotools
+            container_id: xgb-ci.alpine
+            build_tool: autotools
     env:
       R_REMOTES_NO_ERRORS_FROM_WARNINGS: true
-
     steps:
-      - name: Install system dependencies
-        run: |
-          sudo apt update
-          sudo apt install libcurl4-openssl-dev libssl-dev libssh2-1-dev libgit2-dev libglpk-dev libxml2-dev libharfbuzz-dev libfribidi-dev
-        if: matrix.os == 'ubuntu-latest'
+      # Restart Docker daemon so that it recognizes the ephemeral disks
+      - run: sudo systemctl restart docker
       - uses: actions/checkout@v4
         with:
           submodules: 'true'
+      - name: Fetch container from cache
+        run: bash ops/docker_build.sh ${{ matrix.container_id }}
+      - name: Test R
+        run: |
+          bash ops/pipeline/test-rpkg.sh ${{ matrix.container_id }} ${{ matrix.build_tool }}
 
+  test-rpkg-windows:
+    runs-on: windows-latest
+    name: Test R on Windows
+    env:
+      R_REMOTES_NO_ERRORS_FROM_WARNINGS: true
+    steps:
+      - uses: actions/checkout@v4
+        with:
+          submodules: 'true'
       - uses: r-lib/actions/setup-r@v2
         with:
-          r-version: ${{ matrix.r }}
-
+          r-version: release
       - name: Cache R packages
         uses: actions/cache@v4
         with:
           path: ${{ env.R_LIBS_USER }}
-          key: ${{ runner.os }}-r-${{ matrix.r }}-7-${{ hashFiles('R-package/DESCRIPTION') }}
-          restore-keys: ${{ runner.os }}-r-${{ matrix.r }}-7-${{ hashFiles('R-package/DESCRIPTION') }}
-
+          key: ${{ runner.os }}-r-release-7-${{ hashFiles('R-package/DESCRIPTION') }}
+          restore-keys: ${{ runner.os }}-r-release-7-${{ hashFiles('R-package/DESCRIPTION') }}
       - uses: actions/setup-python@v5
         with:
           python-version: "3.10"
           architecture: 'x64'
-
       - uses: r-lib/actions/setup-tinytex@v2
-
       - name: Install dependencies
         shell: Rscript {0}
         run: |
           source("./R-package/tests/helper_scripts/install_deps.R")
-
       - name: Test R
+        shell: powershell
         run: |
-          python ops/script/test_r_package.py --compiler='${{ matrix.compiler }}' --build-tool="${{ matrix.build }}" --task=check
-        if: matrix.compiler != 'none'
-
-      - name: Test R
-        run: |
-          python ops/script/test_r_package.py --build-tool="${{ matrix.build }}" --task=check
-        if: matrix.compiler == 'none'
+          python ops/script/test_r_package.py --compiler='mingw' `
+            --build-tool='autotools' --task=check
 
   test-R-on-Debian:
     name: Test R package on Debian
     runs-on: ubuntu-latest
     container:
       image: rhub/debian-gcc-release
-
     steps:
       - name: Install system dependencies
         run: |
           # Must run before checkout to have the latest git installed.
           # No need to add pandoc, the container has it figured out.
-          apt update && apt install libcurl4-openssl-dev libssl-dev libssh2-1-dev libgit2-dev libglpk-dev libxml2-dev libharfbuzz-dev libfribidi-dev git -y
-
+          apt update && apt install libcurl4-openssl-dev libssl-dev libssh2-1-dev libgit2-dev \
+            libglpk-dev libxml2-dev libharfbuzz-dev libfribidi-dev git -y
       - name: Trust git cloning project sources
         run: |
           git config --global --add safe.directory "${GITHUB_WORKSPACE}"
-
       - uses: actions/checkout@v4
         with:
           submodules: 'true'
-
       - name: Install dependencies
         shell: bash -l {0}
         run: |
           Rscript -e "source('./R-package/tests/helper_scripts/install_deps.R')"
-
       - name: Test R
         shell: bash -l {0}
         run: |
           python3 ops/script/test_r_package.py --r=/usr/bin/R --build-tool=autotools --task=check
-
       - uses: dorny/paths-filter@v3
         id: changes
         with:
           filters: |
             r_package:
               - 'R-package/**'
-
       - name: Run document check
         if: steps.changes.outputs.r_package == 'true'
         run: |

R-package/configure

@@ -646,6 +646,7 @@ ac_includes_default="\
 ac_header_cxx_list=
 ac_subst_vars='LTLIBOBJS
 LIBOBJS
+DMLC_DEFS
 BACKTRACE_LIB
 ENDIAN_FLAG
 OPENMP_LIB
@@ -1504,6 +1505,68 @@ fi
 
 } # ac_fn_cxx_try_link
 
+# ac_fn_cxx_check_func LINENO FUNC VAR
+# ------------------------------------
+# Tests whether FUNC exists, setting the cache variable VAR accordingly
+ac_fn_cxx_check_func ()
+{
+  as_lineno=${as_lineno-"$1"} as_lineno_stack=as_lineno_stack=$as_lineno_stack
+  { printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking for $2" >&5
+printf %s "checking for $2... " >&6; }
+if eval test \${$3+y}
+then :
+  printf %s "(cached) " >&6
+else $as_nop
+  cat confdefs.h - <<_ACEOF >conftest.$ac_ext
+/* end confdefs.h.  */
+/* Define $2 to an innocuous variant, in case <limits.h> declares $2.
+   For example, HP-UX 11i <limits.h> declares gettimeofday.  */
+#define $2 innocuous_$2
+
+/* System header to define __stub macros and hopefully few prototypes,
+   which can conflict with char $2 (); below.  */
+
+#include <limits.h>
+#undef $2
+
+/* Override any GCC internal prototype to avoid an error.
+   Use char because int might match the return type of a GCC
+   builtin and then its argument prototype would still apply.  */
+#ifdef __cplusplus
+extern "C"
+#endif
+char $2 ();
+/* The GNU C library defines this for functions which it implements
+    to always fail with ENOSYS.  Some functions are actually named
+    something starting with __ and the normal name is an alias.  */
+#if defined __stub_$2 || defined __stub___$2
+choke me
+#endif
+
+int
+main (void)
+{
+return $2 ();
+  ;
+  return 0;
+}
+_ACEOF
+if ac_fn_cxx_try_link "$LINENO"
+then :
+  eval "$3=yes"
+else $as_nop
+  eval "$3=no"
+fi
+rm -f core conftest.err conftest.$ac_objext conftest.beam \
+    conftest$ac_exeext conftest.$ac_ext
+fi
+eval ac_res=\$$3
+	       { printf "%s\n" "$as_me:${as_lineno-$LINENO}: result: $ac_res" >&5
+printf "%s\n" "$ac_res" >&6; }
+  eval $as_lineno_stack; ${as_lineno_stack:+:} unset as_lineno
+
+} # ac_fn_cxx_check_func
+
 # ac_fn_cxx_try_run LINENO
 # ------------------------
 # Try to run conftest.$ac_ext, and return whether this succeeded. Assumes that
@@ -2192,6 +2255,8 @@ ac_link='$CXX -o conftest$ac_exeext $CXXFLAGS $CPPFLAGS $LDFLAGS conftest.$ac_ex
 ac_compiler_gnu=$ac_cv_cxx_compiler_gnu
 
 
+DMLC_DEFS=""
+
 ### Check whether backtrace() is part of libc or the external lib libexecinfo
 { printf "%s\n" "$as_me:${as_lineno-$LINENO}: checking Backtrace lib" >&5
 printf %s "checking Backtrace lib... " >&6; }
@@ -2866,7 +2931,17 @@ if test "x$ac_cv_lib_execinfo_backtrace" = xyes
 then :
   BACKTRACE_LIB=-lexecinfo
 else $as_nop
-  BACKTRACE_LIB=''
+  BACKTRACE_LIB=''; DMLC_DEFS="$DMLC_DEFS -DDMLC_LOG_STACK_TRACE=0"
+fi
+
+
+### Check whether fopen64 is available
+ac_fn_cxx_check_func "$LINENO" "fopen64" "ac_cv_func_fopen64"
+if test "x$ac_cv_func_fopen64" = xyes
+then :
+
+else $as_nop
+  DMLC_DEFS="$DMLC_DEFS -DDMLC_USE_FOPEN64=0"
 fi
 
 
@@ -3205,6 +3280,7 @@ fi
 
 
 
+
 ac_config_files="$ac_config_files src/Makevars"
 
 ac_config_headers="$ac_config_headers config.h"

R-package/configure.ac

@@ -22,10 +22,17 @@ CPPFLAGS=`"${R_HOME}/bin/R" CMD config CPPFLAGS`
 LDFLAGS=`"${R_HOME}/bin/R" CMD config LDFLAGS`
 AC_LANG(C++)
 
+DMLC_DEFS=""
+
 ### Check whether backtrace() is part of libc or the external lib libexecinfo
 AC_MSG_CHECKING([Backtrace lib])
 AC_MSG_RESULT([])
-AC_CHECK_LIB([execinfo], [backtrace], [BACKTRACE_LIB=-lexecinfo], [BACKTRACE_LIB=''])
+AC_CHECK_LIB([execinfo], [backtrace],
+             [BACKTRACE_LIB=-lexecinfo],
+             [BACKTRACE_LIB=''; DMLC_DEFS="$DMLC_DEFS -DDMLC_LOG_STACK_TRACE=0"])
+
+### Check whether fopen64 is available
+AC_CHECK_FUNC(fopen64, [], [DMLC_DEFS="$DMLC_DEFS -DDMLC_USE_FOPEN64=0"])
 
 ### Endian detection
 AC_ARG_VAR(USE_LITTLE_ENDIAN, "Whether to build with little endian (checks at compile time if unset)")
@@ -83,6 +90,7 @@ AC_SUBST(OPENMP_CXXFLAGS)
 AC_SUBST(OPENMP_LIB)
 AC_SUBST(ENDIAN_FLAG)
 AC_SUBST(BACKTRACE_LIB)
+AC_SUBST(DMLC_DEFS)
 AC_CONFIG_FILES([src/Makevars])
 AC_CONFIG_HEADERS([config.h])
 AC_OUTPUT

R-package/src/Makevars.in

@@ -22,6 +22,7 @@ PKG_CPPFLAGS = \
     -I$(PKGROOT)/include \
     -I$(PKGROOT)/dmlc-core/include \
     -I$(PKGROOT) \
+    @DMLC_DEFS@ \
     $(XGB_RFLAGS)
 
 PKG_CXXFLAGS = \

ops/docker/ci_container.yml

@@ -27,6 +27,9 @@ xgb-ci.gpu_build_r_rockylinux8:
     CUDA_VERSION_ARG: "12.4.1"
     R_VERSION_ARG: "4.3.2"
 
+xgb-ci.alpine:
+  container_def: alpine
+
 xgb-ci.gpu:
   container_def: gpu
   build_args:

ops/docker/dockerfile/Dockerfile.alpine

@@ -0,0 +1,20 @@
+FROM alpine:3.20
+ENV TZ=UTC
+RUN apk add --no-cache bash
+SHELL ["/bin/bash", "-c"]
+RUN apk add --no-cache R R-dev R-doc g++ shadow
+
+ENV GOSU_VERSION=1.10
+
+# Install lightweight sudo (not bound to TTY)
+RUN set -ex; \
+    wget -nv -O /usr/local/bin/gosu "https://github.com/tianon/gosu/releases/download/$GOSU_VERSION/gosu-amd64" && \
+    chmod +x /usr/local/bin/gosu && \
+    gosu nobody true
+
+# Default entry-point to use if running locally
+# It will preserve attributes of created files
+COPY docker/entrypoint.sh /scripts/
+
+WORKDIR /workspace
+ENTRYPOINT ["/scripts/entrypoint.sh"]

ops/docker/entrypoint.sh

@@ -27,7 +27,7 @@ then
         "${CI_BUILD_USER}" || true
     export HOME="/home/${CI_BUILD_USER}"
     shopt -s dotglob
-    cp -r /root/* "$HOME/"
+    cp -r /root/* "$HOME/" || true
     chown -R "${CI_BUILD_UID}:${CI_BUILD_GID}" "$HOME"
 
     # Allows project-specific customization

ops/pipeline/test-rpkg-impl.sh

@@ -0,0 +1,17 @@
+#!/bin/bash
+## Test R package on Linux
+
+set -euo pipefail
+
+if [[ "$#" -lt 1 ]]
+then
+  echo "Usage: $0 {autotools, cmake}"
+  exit 1
+fi
+
+build_tool="$1"
+
+set -x
+
+gosu root Rscript ./R-package/tests/helper_scripts/install_deps.R
+python3 ops/script/test_r_package.py --build-tool="${build_tool}"