Fix non-openmp build. (#5566)

* Add test to Jenkins.
* Fix threading utils tests.
* Require thread library.

CMakeLists.txt

@@ -95,6 +95,8 @@ if (USE_CUDA)
   message(STATUS "CUDA GEN_CODE: ${GEN_CODE}")
 endif (USE_CUDA)
 
+find_package(Threads REQUIRED)
+
 if (USE_OPENMP)
   if (APPLE)
     # Require CMake 3.16+ on Mac OSX, as previous versions of CMake had trouble locating
@@ -128,7 +130,7 @@ foreach(lib rabit rabit_base rabit_empty rabit_mock rabit_mock_static)
   # Explicitly link dmlc to rabit, so that configured header (build_config.h)
   # from dmlc is correctly applied to rabit.
   if (TARGET ${lib})
-    target_link_libraries(${lib} dmlc)
+    target_link_libraries(${lib} dmlc ${CMAKE_THREAD_LIBS_INIT})
   endif (TARGET ${lib})
 endforeach()
 
@@ -138,6 +140,7 @@ if (R_LIB)
 endif (R_LIB)
 
 # core xgboost
+list(APPEND LINKED_LIBRARIES_PRIVATE Threads::Threads ${CMAKE_THREAD_LIBS_INIT})
 add_subdirectory(${xgboost_SOURCE_DIR}/plugin)
 add_subdirectory(${xgboost_SOURCE_DIR}/src)
 target_link_libraries(objxgboost PUBLIC dmlc)

Jenkinsfile

@@ -63,6 +63,7 @@ pipeline {
           parallel ([
             'build-cpu': { BuildCPU() },
             'build-cpu-rabit-mock': { BuildCPUMock() },
+            'build-cpu-non-omp': { BuildCPUNonOmp() },
             'build-gpu-cuda9.0': { BuildCUDA(cuda_version: '9.0') },
             'build-gpu-cuda10.0': { BuildCUDA(cuda_version: '10.0') },
             'build-gpu-cuda10.1': { BuildCUDA(cuda_version: '10.1') },
@@ -221,6 +222,22 @@ def BuildCPUMock() {
   }
 }
 
+def BuildCPUNonOmp() {
+  node('linux && cpu') {
+    unstash name: 'srcs'
+    echo "Build CPU without OpenMP"
+    def container_type = "cpu"
+    def docker_binary = "docker"
+    sh """
+    ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/build_via_cmake.sh -DUSE_OPENMP=OFF
+    """
+    echo "Running Non-OpenMP C++ test..."
+    sh """
+    ${dockerRun} ${container_type} ${docker_binary} build/testxgboost
+    """
+    deleteDir()
+  }
+}
 
 def BuildCUDA(args) {
   node('linux && cpu') {
@@ -259,7 +276,7 @@ def BuildJVMPackages(args) {
     ${docker_extra_params} ${dockerRun} ${container_type} ${docker_binary} tests/ci_build/build_jvm_packages.sh ${args.spark_version}
     """
     echo 'Stashing XGBoost4J JAR...'
-    stash name: 'xgboost4j_jar', includes: 'jvm-packages/xgboost4j/target/*.jar,jvm-packages/xgboost4j-spark/target/*.jar,jvm-packages/xgboost4j-example/target/*.jar'
+    stash name: 'xgboost4j_jar', includes: "jvm-packages/xgboost4j/target/*.jar,jvm-packages/xgboost4j-spark/target/*.jar,jvm-packages/xgboost4j-example/target/*.jar"
     deleteDir()
   }
 }

src/CMakeLists.txt

@@ -88,12 +88,15 @@ if (XGBOOST_BUILTIN_PREFETCH_PRESENT)
     -DXGBOOST_BUILTIN_PREFETCH_PRESENT=1)
 endif (XGBOOST_BUILTIN_PREFETCH_PRESENT)
 
+find_package(Threads REQUIRED)
+list(APPEND SRC_LIBS Threads::Threads ${CMAKE_THREAD_LIBS_INIT})
+
 if (USE_OPENMP OR USE_CUDA)  # CUDA requires OpenMP
   find_package(OpenMP REQUIRED)
   list(APPEND SRC_LIBS OpenMP::OpenMP_CXX)
-  set(LINKED_LIBRARIES_PRIVATE "${LINKED_LIBRARIES_PRIVATE};${SRC_LIBS}" PARENT_SCOPE)
   target_link_libraries(objxgboost PRIVATE OpenMP::OpenMP_CXX)
 endif (USE_OPENMP OR USE_CUDA)
+set(LINKED_LIBRARIES_PRIVATE "${LINKED_LIBRARIES_PRIVATE};${SRC_LIBS}" PARENT_SCOPE)
 
 # For MSVC: Call msvc_use_static_runtime() once again to completely
 # replace /MD with /MT. See https://github.com/dmlc/xgboost/issues/4462

src/common/column_matrix.h

@@ -227,8 +227,7 @@ class ColumnMatrix {
     /* missing values make sense only for column with type kDenseColumn,
        and if no missing values were observed it could be handled much faster. */
     if (noMissingValues) {
-      const int32_t nthread = omp_get_max_threads();  // NOLINT
-      #pragma omp parallel for num_threads(nthread)
+#pragma omp parallel for num_threads(omp_get_max_threads())
       for (omp_ulong rid = 0; rid < nrow; ++rid) {
         const size_t ibegin = rid*nfeature;
         const size_t iend = (rid+1)*nfeature;

src/common/threading_utils.h

@@ -110,10 +110,12 @@ class BlockedSpace2d {
 
 // Wrapper to implement nested parallelism with simple omp parallel for
 template<typename Func>
-void ParallelFor2d(const BlockedSpace2d& space, const int nthreads, Func func) {
+void ParallelFor2d(const BlockedSpace2d& space, int nthreads, Func func) {
   const size_t num_blocks_in_space = space.Size();
+  nthreads = std::min(nthreads, omp_get_max_threads());
+  nthreads = std::max(nthreads, 1);
 
-  #pragma omp parallel num_threads(nthreads)
+#pragma omp parallel num_threads(nthreads)
   {
     size_t tid = omp_get_thread_num();
     size_t chunck_size = num_blocks_in_space / nthreads + !!(num_blocks_in_space % nthreads);

src/metric/rank_metric.cc

@@ -189,12 +189,7 @@ struct EvalAuc : public Metric {
           rec[j - gptr[group_id]] = {h_preds[j], j};
         }
 
-        if (omp_in_parallel()) {
-          std::stable_sort(rec.begin(), rec.end(), common::CmpFirst);
-        } else {
         XGBOOST_PARALLEL_SORT(rec.begin(), rec.end(), common::CmpFirst);
-        }
-
         // calculate AUC
         double sum_pospair = 0.0;
         double sum_npos = 0.0, sum_nneg = 0.0, buf_pos = 0.0, buf_neg = 0.0;
@@ -557,11 +552,7 @@ struct EvalAucPR : public Metric {
           continue;
         }
 
-        if (omp_in_parallel()) {
-          std::stable_sort(rec.begin(), rec.end(), common::CmpFirst);
-        } else {
         XGBOOST_PARALLEL_SORT(rec.begin(), rec.end(), common::CmpFirst);
-        }
 
         // calculate AUC
         double tp = 0.0, prevtp = 0.0, fp = 0.0, prevfp = 0.0, h = 0.0, a = 0.0, b = 0.0;

tests/cpp/CMakeLists.txt

@@ -69,8 +69,8 @@ set_target_properties(
 target_link_libraries(testxgboost
   PRIVATE
   ${GTEST_LIBRARIES}
-  ${LINKED_LIBRARIES_PRIVATE}
-  OpenMP::OpenMP_CXX)
+  ${LINKED_LIBRARIES_PRIVATE})
+
 target_compile_definitions(testxgboost PRIVATE ${XGBOOST_DEFINITIONS})
 set_output_directory(testxgboost ${xgboost_BINARY_DIR})
 

tests/cpp/common/test_threading_utils.cc

@@ -38,7 +38,10 @@ TEST(ParallelFor2d, Test) {
       return kDim2;
   }, kGrainSize);
 
-  ParallelFor2d(space, 4, [&](size_t i, Range1d r) {
+  auto old = omp_get_max_threads();
+  omp_set_num_threads(4);
+
+  ParallelFor2d(space, omp_get_max_threads(), [&](size_t i, Range1d r) {
     for (auto j = r.begin(); j < r.end(); ++j) {
       matrix[i*kDim2 + j] += 1;
     }
@@ -47,12 +50,17 @@ TEST(ParallelFor2d, Test) {
   for (size_t i = 0; i < kDim1 * kDim2; i++) {
     ASSERT_EQ(matrix[i], 1);
   }
+
+  omp_set_num_threads(old);
 }
 
 TEST(ParallelFor2dNonUniform, Test) {
   constexpr size_t kDim1 = 5;
   constexpr size_t kGrainSize = 256;
 
+  auto old = omp_get_max_threads();
+  omp_set_num_threads(4);
+
   // here are quite non-uniform distribution in space
   // but ParallelFor2d should split them by blocks with max size = kGrainSize
   // and process in balanced manner (optimal performance)
@@ -66,7 +74,7 @@ TEST(ParallelFor2dNonUniform, Test) {
     working_space[i].resize(dim2[i], 0);
   }
 
-  ParallelFor2d(space, 4, [&](size_t i, Range1d r) {
+  ParallelFor2d(space, omp_get_max_threads(), [&](size_t i, Range1d r) {
     for (auto j = r.begin(); j < r.end(); ++j) {
       working_space[i][j] += 1;
     }
@@ -77,6 +85,8 @@ TEST(ParallelFor2dNonUniform, Test) {
       ASSERT_EQ(working_space[i][j], 1);
     }
   }
+
+  omp_set_num_threads(old);
 }
 
 }  // namespace common

tests/cpp/helpers.cc

@@ -301,9 +301,12 @@ std::unique_ptr<DMatrix> CreateSparsePageDMatrix(
     batch_count++;
     row_count += batch.Size();
   }
+#if defined(_OPENMP)
   EXPECT_GE(batch_count, 2);
   EXPECT_EQ(row_count, dmat->Info().num_row_);
-
+#else
+#warning "External memory doesn't work with Non-OpenMP build "
+#endif  // defined(_OPENMP)
   return dmat;
 }