Merge Pull Request #13363 from trilinos/Trilinos/master_merge_2024081…

…6_175818

Automatically Merged using Trilinos Master Merge AutoTester
PR Title: b'Trilinos Master Merge PR Generator: Auto PR created to promote from master_merge_20240816_175818 branch to master'
PR Author: trilinos-autotester

.github/workflows/clang_format.yml

@@ -12,7 +12,7 @@ jobs:
 
     steps:
     - uses: actions/checkout@692973e3d937129bcbf40652eb9f2f61becf3332 # v4.1.7
-    - uses: DoozyX/clang-format-lint-action@11b773b1598aa4ae3b32f023701bca5201c3817d # v0.17
+    - uses: DoozyX/clang-format-lint-action@d7f6a5bada32b7ea520b5918416e92997678e3fd # v0.18
       with:
         source: './packages/muelu ./packages/tempus ./packages/teko ./packages/xpetra'
         exclude: './packages/tempus/examples'
@@ -22,7 +22,7 @@ jobs:
     - run: git diff HEAD > format_patch.txt
     - run: if [ "$(cat format_patch.txt)" == "" ] ; then rm format_patch.txt ; else cat format_patch.txt; fi
 
-    - uses: actions/upload-artifact@0b2256b8c012f0828dc542b3febcab082c67f72b # v4.3.4
+    - uses: actions/upload-artifact@834a144ee995460fba8ed112a2fc961b36a5ec5a # v4.3.6
       id: upload-artf
       if: ${{ hashFiles('format_patch.txt') != '' }}
       with:

.github/workflows/codeql.yml

@@ -62,7 +62,7 @@ jobs:
 
     # Initializes the CodeQL tools for scanning.
     - name: Initialize CodeQL
-      uses: github/codeql-action/init@afb54ba388a7dca6ecae48f608c4ff05ff4cc77a # v3.25.15
+      uses: github/codeql-action/init@eb055d739abdc2e8de2e5f4ba1a8b246daa779aa # v3.26.0
       with:
         languages: ${{ matrix.language }}
         build-mode: ${{ matrix.build-mode }}
@@ -85,6 +85,6 @@ jobs:
           make -j 2
       
     - name: Perform CodeQL Analysis
-      uses: github/codeql-action/analyze@afb54ba388a7dca6ecae48f608c4ff05ff4cc77a # v3.25.15
+      uses: github/codeql-action/analyze@eb055d739abdc2e8de2e5f4ba1a8b246daa779aa # v3.26.0
       with:
         category: "/language:${{matrix.language}}"

.github/workflows/dependency-review.yml

@@ -17,7 +17,7 @@ jobs:
     runs-on: ubuntu-latest
     steps:
       - name: Harden Runner
-        uses: step-security/harden-runner@0d381219ddf674d61a7572ddd19d7941e271515c # v2.9.0
+        uses: step-security/harden-runner@5c7944e73c4c2a096b17a9cb74d65b6c2bbafbde # v2.9.1
         with:
           egress-policy: audit
 

.github/workflows/scorecards.yml

@@ -58,14 +58,14 @@ jobs:
       # Upload the results as artifacts (optional). Commenting out will disable uploads of run results in SARIF
       # format to the repository Actions tab.
       - name: "Upload artifact"
-        uses: actions/upload-artifact@0b2256b8c012f0828dc542b3febcab082c67f72b # v4.3.4
+        uses: actions/upload-artifact@834a144ee995460fba8ed112a2fc961b36a5ec5a # v4.3.6
         with:
           name: SARIF file
           path: results.sarif
           retention-days: 5
 
       # Upload the results to GitHub's code scanning dashboard.
       - name: "Upload to code-scanning"
-        uses: github/codeql-action/upload-sarif@afb54ba388a7dca6ecae48f608c4ff05ff4cc77a # v3.25.15
+        uses: github/codeql-action/upload-sarif@eb055d739abdc2e8de2e5f4ba1a8b246daa779aa # v3.26.0
         with:
           sarif_file: results.sarif

packages/amesos2/example/SimpleSolve_File.cpp

@@ -53,15 +53,21 @@ int main(int argc, char *argv[]) {
 
   bool printMatrix   = false;
   bool printSolution = false;
+  bool checkSolution = false;
   bool printTiming   = false;
   bool allprint      = false;
   bool verbose = (myRank==0);
-  std::string filename("arc130.mtx");
+  std::string mat_filename("arc130.mtx");
+  std::string rhs_filename("");
+  std::string solvername("Superlu");
   Teuchos::CommandLineProcessor cmdp(false,true);
   cmdp.setOption("verbose","quiet",&verbose,"Print messages and results.");
-  cmdp.setOption("filename",&filename,"Filename for Matrix-Market test matrix.");
+  cmdp.setOption("filename",&mat_filename,"Filename for Matrix-Market test matrix.");
+  cmdp.setOption("rhs_filename",&rhs_filename,"Filename for Matrix-Market right-hand-side.");
+  cmdp.setOption("solvername",&solvername,"Name of solver.");
   cmdp.setOption("print-matrix","no-print-matrix",&printMatrix,"Print the full matrix after reading it.");
   cmdp.setOption("print-solution","no-print-solution",&printSolution,"Print solution vector after solve.");
+  cmdp.setOption("check-solution","no-check-solution",&checkSolution,"Check solution vector after solve.");
   cmdp.setOption("print-timing","no-print-timing",&printTiming,"Print solver timing statistics");
   cmdp.setOption("all-print","root-print",&allprint,"All processors print to out");
   if (cmdp.parse(argc,argv) != Teuchos::CommandLineProcessor::PARSE_SUCCESSFUL) {
@@ -76,7 +82,7 @@ int main(int argc, char *argv[]) {
 
   const size_t numVectors = 1;
 
-  RCP<MAT> A = Tpetra::MatrixMarket::Reader<MAT>::readSparseFile(filename, comm);
+  RCP<MAT> A = Tpetra::MatrixMarket::Reader<MAT>::readSparseFile(mat_filename, comm);
   if( printMatrix ){
     A->describe(*fos, Teuchos::VERB_EXTREME);
   }
@@ -98,29 +104,32 @@ int main(int argc, char *argv[]) {
   RCP<MV> X = rcp(new MV(dmnmap,numVectors));
   X->randomize();
 
-  /* Create B
-   *
-   * Use RHS:
-   *
-   *  [[10]
-   *   [10]
-   *   [10]
-   *   [10]
-   *   [10]
-   *   [10]]
-   */
+  // Create B
   RCP<MV> B = rcp(new MV(rngmap,numVectors));
-  B->putScalar(10);
-
+  if (rhs_filename == "") {
+    /*
+     * Use RHS:
+     *
+     *  [[10]
+     *   [10]
+     *   [10]
+     *   [10]
+     *   [10]
+     *   [10]]
+     */
+    B->putScalar(10);
+  } else {
+    B = Tpetra::MatrixMarket::Reader<MAT>::readDenseFile (rhs_filename, comm, rngmap);
+  }
 
   // Constructor from Factory
   RCP<Amesos2::Solver<MAT,MV> > solver;
-  if( !Amesos2::query("Superlu") ){
-    *fos << "SuperLU solver not enabled.  Exiting..." << std::endl;
+  if( !Amesos2::query(solvername) ){
+    *fos << solvername << " solver not enabled.  Exiting..." << std::endl;
     return EXIT_SUCCESS;
   }
 
-  solver = Amesos2::create<MAT,MV>("Superlu", A, X, B);
+  solver = Amesos2::create<MAT,MV>(solvername, A, X, B);
 
   solver->symbolicFactorization().numericFactorization().solve();
 
@@ -140,6 +149,24 @@ int main(int argc, char *argv[]) {
     }
   }
 
+  if( checkSolution ){
+    const Scalar one = Teuchos::ScalarTraits<Scalar>::one ();
+    RCP<MV> R = rcp(new MV(rngmap,numVectors));
+    A->apply(*X, *R);
+    R->update(one, *B, -one);
+    for (size_t j = 0; j < numVectors; ++j) {
+      auto Rj = R->getVector(j);
+      auto Bj = B->getVector(j);
+      auto r_norm = Rj->norm2();
+      auto b_norm = Bj->norm2();
+      if (myRank == 0) {
+        *fos << "Relative Residual norm = " << r_norm << " / " << b_norm << " = "
+             << r_norm / b_norm << std::endl;
+      }
+    }
+    if (myRank == 0) *fos << std::endl;
+  }
+
   if( printTiming ){
     // Print some timing statistics
     solver->printTiming(*fos);

packages/framework/get_dependencies.sh

@@ -19,6 +19,11 @@ function tril_genconfig_assert_pwd_is_git_repo() {
   fi
 }
 
+function retry_command() {
+  cmd=$1
+  ${cmd} || { echo "Retrying after 1m..." ; sleep 60 ; ${cmd} ; } || { echo "Retrying after 5m..." ; sleep 300 ; ${cmd} ; }
+}
+
 function tril_genconfig_clone_or_update_repo() {
   git_url=$1
   sub_dir=$2
@@ -33,10 +38,12 @@ function tril_genconfig_clone_or_update_repo() {
     echo "STATUS: ${sub_dir}: Fetching remote repo"
     cd ${sub_dir}
     tril_genconfig_assert_pwd_is_git_repo
-    git fetch
+    cmd="git fetch"
+    retry_command "${cmd}"
   else
     echo "STATUS: ${sub_dir}: Cloning from '${git_url}'"
-    git clone ${git_url} ${sub_dir}
+    cmd="git clone ${git_url} ${sub_dir}"
+    retry_command "${cmd}"
     cd ${sub_dir}
   fi
 
@@ -52,7 +59,8 @@ function tril_genconfig_clone_or_update_repo() {
   if [[ "${has_submodules}" == "has-submodules" ]] ; then
     echo
     echo "STATUS: ${sub_dir}: Update submodules"
-    git submodule update --force --init --recursive
+    cmd="git submodule update --force --init --recursive"
+    retry_command "${cmd}"
     cd - > /dev/null
   elif [[ "${has_submodules}" != "" ]] ; then
     echo "ERROR: argument '${has_submodules}' not allowed!  Only 'has-submodules' or ''!"

packages/ifpack2/cmake/Dependencies.cmake

@@ -1,7 +1,7 @@
 SET(LIB_REQUIRED_DEP_PACKAGES Belos Teuchos Tpetra KokkosKernels)
 SET(LIB_OPTIONAL_DEP_PACKAGES Xpetra Zoltan2Core ThyraTpetraAdapters Amesos2 ShyLU_NodeBasker ShyLU_NodeHTS ShyLU_NodeFastILU)
 SET(TEST_REQUIRED_DEP_PACKAGES Belos Galeri)
-SET(TEST_OPTIONAL_DEP_PACKAGES Amesos2 ShyLU_NodeHTS ML AztecOO Epetra)
+SET(TEST_OPTIONAL_DEP_PACKAGES Amesos2 ShyLU_NodeHTS ML AztecOO Epetra Zoltan2Core)
 SET(LIB_REQUIRED_DEP_TPLS)
 SET(LIB_OPTIONAL_DEP_TPLS HYPRE Cholmod Lemon METIS MPI)
 SET(TEST_REQUIRED_DEP_TPLS)

packages/ifpack2/src/Ifpack2_BlockComputeResidualVector.hpp

@@ -790,7 +790,7 @@ namespace Ifpack2 {
                const MultiVectorLocalViewTypeB &b_,
                const MultiVectorLocalViewTypeX &x_) {
         IFPACK2_BLOCKHELPER_PROFILER_REGION_BEGIN;
-        IFPACK2_BLOCKHELPER_TIMER("BlockTriDi::ComputeResidual::<SeqTag>");
+        IFPACK2_BLOCKHELPER_TIMER_WITH_FENCE("BlockTriDi::ComputeResidual::<SeqTag>", ComputeResidual0, execution_space);
 
         y = y_; b = b_; x = x_;
         if constexpr (is_device<execution_space>::value) {
@@ -818,7 +818,7 @@ namespace Ifpack2 {
                const MultiVectorLocalViewTypeX &x_,
                const MultiVectorLocalViewTypeX_Remote &x_remote_) {
         IFPACK2_BLOCKHELPER_PROFILER_REGION_BEGIN;
-        IFPACK2_BLOCKHELPER_TIMER("BlockTriDi::ComputeResidual::<AsyncTag>");
+        IFPACK2_BLOCKHELPER_TIMER_WITH_FENCE("BlockTriDi::ComputeResidual::<AsyncTag>", ComputeResidual0, execution_space);
 
         b = b_; x = x_; x_remote = x_remote_;
         if constexpr (is_device<execution_space>::value) {
@@ -892,7 +892,7 @@ namespace Ifpack2 {
                const MultiVectorLocalViewTypeX_Remote &x_remote_,
                const bool compute_owned) {
         IFPACK2_BLOCKHELPER_PROFILER_REGION_BEGIN;
-        IFPACK2_BLOCKHELPER_TIMER("BlockTriDi::ComputeResidual::<OverlapTag>");
+        IFPACK2_BLOCKHELPER_TIMER_WITH_FENCE("BlockTriDi::ComputeResidual::<OverlapTag>", ComputeResidual0, execution_space);
 
         b = b_; x = x_; x_remote = x_remote_;
         if constexpr (is_device<execution_space>::value) {

packages/ifpack2/src/Ifpack2_BlockHelper.hpp

@@ -10,6 +10,7 @@
 #ifndef IFPACK2_BLOCKHELPER_IMPL_HPP
 #define IFPACK2_BLOCKHELPER_IMPL_HPP
 
+#include "Ifpack2_BlockHelper_Timers.hpp"
 
 namespace Ifpack2 {
 
@@ -155,15 +156,6 @@ namespace Ifpack2 {
     };
 #endif
 
-
-#if defined(HAVE_IFPACK2_BLOCKTRIDICONTAINER_TIMERS)
-#define IFPACK2_BLOCKHELPER_TIMER(label) TEUCHOS_FUNC_TIME_MONITOR(label);
-#define IFPACK2_BLOCKHELPER_TIMER_FENCE(execution_space) execution_space().fence();
-#else
-#define IFPACK2_BLOCKHELPER_TIMER(label)
-#define IFPACK2_BLOCKHELPER_TIMER_FENCE(execution_space)
-#endif
-
 #if defined(KOKKOS_ENABLE_CUDA) && defined(IFPACK2_BLOCKHELPER_ENABLE_PROFILE)
 #define IFPACK2_BLOCKHELPER_PROFILER_REGION_BEGIN \
     KOKKOS_IMPL_CUDA_SAFE_CALL(cudaProfilerStart());
@@ -404,7 +396,7 @@ namespace Ifpack2 {
       void ireduce(const int sweep, const bool force = false) {
         if ( ! force && sweep % sweep_step_) return;
 
-        IFPACK2_BLOCKHELPER_TIMER("BlockTriDi::NormManager::Ireduce");
+        IFPACK2_BLOCKHELPER_TIMER("BlockTriDi::NormManager::Ireduce", Ireduce);
 
         work_[1] = work_[0];
 #ifdef HAVE_IFPACK2_MPI
@@ -433,7 +425,7 @@ namespace Ifpack2 {
         // early return
         if (sweep <= 0) return false;
 
-        IFPACK2_BLOCKHELPER_TIMER("BlockTriDi::NormManager::CheckDone");
+        IFPACK2_BLOCKHELPER_TIMER("BlockTriDi::NormManager::CheckDone", CheckDone);
 
         TEUCHOS_ASSERT(sweep >= 1);
         if ( ! force && (sweep - 1) % sweep_step_) return false;
@@ -481,7 +473,7 @@ namespace Ifpack2 {
     void reduceVector(const ConstUnmanaged<typename BlockHelperDetails::ImplType<MatrixType>::impl_scalar_type_1d_view> zz,
                       /* */ typename BlockHelperDetails::ImplType<MatrixType>::magnitude_type *vals) {
       IFPACK2_BLOCKHELPER_PROFILER_REGION_BEGIN;
-      IFPACK2_BLOCKHELPER_TIMER("BlockTriDi::ReduceVector");
+      IFPACK2_BLOCKHELPER_TIMER("BlockTriDi::ReduceVector", ReduceVector);
 
       using impl_type = BlockHelperDetails::ImplType<MatrixType>;
       using local_ordinal_type = typename impl_type::local_ordinal_type;

packages/ifpack2/src/Ifpack2_BlockHelper_Timers.hpp

@@ -0,0 +1,36 @@
+// @HEADER
+// *****************************************************************************
+//       Ifpack2: Templated Object-Oriented Algebraic Preconditioner Package
+//
+// Copyright 2009 NTESS and the Ifpack2 contributors.
+// SPDX-License-Identifier: BSD-3-Clause
+// *****************************************************************************
+// @HEADER
+
+#ifndef IFPACK2_BLOCKHELPER_TIMERS_HPP
+#define IFPACK2_BLOCKHELPER_TIMERS_HPP
+
+
+namespace Ifpack2 {
+
+  namespace BlockHelperDetails {
+
+#if defined(HAVE_IFPACK2_BLOCKTRIDICONTAINER_TIMERS)
+#define IFPACK2_BLOCKHELPER_TIMER(label, varname) TEUCHOS_FUNC_TIME_MONITOR_DIFF(label, varname);
+#define IFPACK2_BLOCKHELPER_TIMER_FENCE(execution_space) execution_space().fence();
+#define IFPACK2_BLOCKHELPER_TIMER_DEFAULT_FENCE() Kokkos::DefaultExecutionSpace().fence();
+#else
+#define IFPACK2_BLOCKHELPER_TIMER(label, varname)
+#define IFPACK2_BLOCKHELPER_TIMER_FENCE(execution_space)
+#define IFPACK2_BLOCKHELPER_TIMER_DEFAULT_FENCE()
+#endif
+
+#define IFPACK2_BLOCKHELPER_TIMER_WITH_FENCE(label, varname, execution_space) \
+  IFPACK2_BLOCKHELPER_TIMER_FENCE(execution_space) \
+  IFPACK2_BLOCKHELPER_TIMER(label, varname)
+
+  } // namespace BlockHelperDetails
+
+} // namespace Ifpack2
+
+#endif