Merge branch 'develop' into feature/logging-updates

.github/workflows/check-spelling.yml

@@ -0,0 +1,44 @@
+name: Checks - spelling
+
+on:
+  pull_request:
+  workflow_dispatch:
+
+jobs:
+  spelling_check:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Install python3-pip
+        run: |
+          sudo apt update
+          sudo apt install -y --no-install-recommends python3-pip
+
+      - name: Install codespell
+        run: pip install codespell
+
+      - name: Print codespell version
+        run: codespell --version
+
+      - name: Check out repository code
+        uses: actions/checkout@v4
+        with:
+          submodules: true
+
+      - name: Run codespell
+        run: |
+          ./scripts/spelling.sh
+
+      - name: Run git diff to see if anything changed
+        run: /usr/bin/git diff --name-only --exit-code
+
+      - name: Run git diff if we failed
+        if: failure()
+        run: /usr/bin/git diff > spelling.patch
+
+      - name: Archive diff as a patch if we failed
+        uses: actions/upload-artifact@v3
+        if: failure()
+        with:
+          name: spelling.patch
+          path: |
+            ${{ github.workspace }}/spelling.patch

CHANGELOG.md

@@ -10,6 +10,9 @@
 
 Fixed the loading of ARKStep's default first order explicit method.
 
+Fixed a CMake bug regarding usage of missing "print_warning" macro
+that was only triggered when the deprecated `CUDA_ARCH` option was used.
+
 ### Deprecation Notices
 
 ## Changes to SUNDIALS in release 7.1.1
@@ -1789,7 +1792,7 @@ update their code to set the corresponding `ops` structure member,
 `matvecsetup`, to `NULL`.
 
 The generic SUNMatrix API now defines error codes to be returned by SUNMatrix
-operations. Operations which return an integer flag indiciating success/failure
+operations. Operations which return an integer flag indicating success/failure
 may return different values than previously.
 
 A new SUNMatrix (and SUNLinearSolver) implementation was added to facilitate
@@ -1901,7 +1904,7 @@ function signatures have been changed including MRIStepCreate which now
 takes an ARKStep memory structure for the fast integration as an input.
 
 The reinitialization functions `ERKStepReInit`, `ARKStepReInit`, and
-`MRIStepReInit` have been updated to retain the minimum and maxiumum step
+`MRIStepReInit` have been updated to retain the minimum and maximum step
 size values from before reinitialization rather than resetting them to the
 default values.
 
@@ -1926,7 +1929,7 @@ being built.
 
 Fixed a memory leak in the PETSc `N_Vector` clone function.
 
-Fixed a memeory leak in the ARKODE, CVODE, and IDA F77 interfaces when not using
+Fixed a memory leak in the ARKODE, CVODE, and IDA F77 interfaces when not using
 the default nonlinear solver.
 
 Fixed a bug in the ARKStep time-stepping module in ARKODE that would result in
@@ -2638,7 +2641,7 @@ with sparse direct solvers.
 
 #### KINSOL
 
-The Picard iteration return was chanegd to always return the newest iterate upon
+The Picard iteration return was changed to always return the newest iterate upon
 success.
 
 A minor bug in the line search was fixed to prevent an infinite loop when the

CMakeLists.txt

@@ -187,7 +187,7 @@ include(SundialsSetupCompilers)
 include(SundialsSetupTPLs)
 
 # ===============================================================
-# Build options to be proccessed last
+# Build options to be processed last
 # ===============================================================
 
 include(SundialsBuildOptionsPost)

CODEOWNERS

@@ -3,6 +3,6 @@
 
 # These owners will be the default owners for everything in
 # the repo. Unless a later match takes precedence,
-# globabl owners will be requested for review when someone
+# global owners will be requested for review when someone
 # opens a pull request.
 * @balos1 @gardner48 @cswoodward @drreynolds

CONTRIBUTING.md

@@ -1,6 +1,6 @@
 # Contributing to SUNDIALS
 
-There are two primary ways of contributing to SUNDIALS. The first way is by particpating
+There are two primary ways of contributing to SUNDIALS. The first way is by participating
 in the development of SUNDIALS directly through contributions of code to the primary
 [SUNDIALS repository](https://github.com/LLNL/sundials). This is the best way to contribute
 bug fixes and minor improvements. At this time, the SUNDIALS team does not have the resources

README.md

@@ -15,50 +15,62 @@ equation (ODE) systems, differential-algebraic equation (DAE) systems, and
 nonlinear algebraic systems:
 
 * ARKODE - for integrating stiff, nonstiff, and multirate ODEs of the form
+
   $$M(t) \\, y' = f_1(t,y) + f_2(t,y), \quad y(t_0) = y_0$$
 
 * CVODE - for integrating stiff and nonstiff ODEs of the form
+
   $$y' = f(t,y), \quad y(t_0) = y_0$$
 
 * CVODES - for integrating and sensitivity analysis (forward and adjoint) of
   ODEs of the form
+
   $$y' = f(t,y,p), \quad y(t_0) = y_0(p)$$
 
 * IDA - for integrating DAEs of the form
+
   $$F(t,y,y') = 0, \quad y(t_0) = y_0, \quad y'(t_0) = y_0'$$
 
 * IDAS - for integrating and sensitivity analysis (forward and adjoint) of DAEs
   of the form
+
   $$F(t,y,y',p) = 0, \quad y(t_0) = y_0(p), \quad y'(t_0) = y_0'(p)$$
 
 * KINSOL - for solving nonlinear algebraic systems of the form
+
   $$F(u) = 0 \quad \text{or} \quad G(u) = u$$
 
 ## Installation ##
 
-For installation directions see the [online install guide](https://sundials.readthedocs.io/en/latest/Install_link.html),
-the installation chapter in any of the package user guides, or INSTALL_GUIDE.pdf.
+For installation directions, see the [getting started](https://sundials.readthedocs.io/en/latest/sundials/index.html#getting-started)
+section in the online documentation. In the [released tarballs](https://github.com/LLNL/sundials/releases),
+installation directions are also available in `INSTALL_GUIDE.pdf` and the
+installation chapter of the user guides in the `doc` directory.
 
 Warning to users who receive more than one of the individual packages at
-different times: Mixing old and new versions of SUNDIALS may fail. To avoid
-such failures, obtain all desired package at the same time.
+different times: Mixing old and new versions of SUNDIALS may fail. To avoid such
+failures, obtain all desired package at the same time.
 
 ## Support ##
 
-Full user guides for all of the SUNDIALS packages are available [online](https://sundials.readthedocs.io)
-and in the [doc](./doc) directory. Additionally, the [doc](./doc) directory
-contains documentation for the package example programs.
+Full user guides for all of the SUNDIALS packages are available [online](https://sundials.readthedocs.io).
+In the [released tarballs](https://github.com/LLNL/sundials/releases), the `doc`
+directory includes PDFs of the user guides and documentation for the example
+programs. The example program documentation PDFs are also available on the
+[releases page](https://github.com/LLNL/sundials/releases).
 
 For information on recent changes to SUNDIALS see the [CHANGELOG](./CHANGELOG.md)
 or the introduction chapter of any package user guide.
 
 A list of Frequently Asked Questions on build and installation procedures as
-well as common usage issues is available on the SUNDIALS [FAQ](https://computing.llnl.gov/projects/sundials/faq).
-For dealing with systems with unphysical solutions or discontinuities see the
-SUNDIALS [usage notes](https://computing.llnl.gov/projects/sundials/usage-notes).
+well as common usage issues is available on the SUNDIALS
+[FAQ](https://computing.llnl.gov/projects/sundials/faq). For dealing with
+systems with nonphysical solutions or discontinuities see the SUNDIALS
+[usage notes](https://computing.llnl.gov/projects/sundials/usage-notes).
 
-If you have a question not covered in the FAQ or usage notes, please submit
-your question to the SUNDIALS [mailing list](https://computing.llnl.gov/projects/sundials/mailing-list).
+If you have a question not covered in the FAQ or usage notes, please submit your
+question as a [GitHub issue](https://github.com/LLNL/sundials/issues) or to the
+SUNDIALS [mailing list](https://computing.llnl.gov/projects/sundials/mailing-list).
 
 ## Contributing ##
 
@@ -76,14 +88,15 @@ any publications reporting work done using SUNDIALS packages.
 
 The SUNDIALS library has been developed over many years by a number of
 contributors. The current SUNDIALS team consists of Cody J. Balos,
-David J. Gardner, Alan C. Hindmarsh, Daniel R. Reynolds, and Carol S. Woodward.
-We thank Radu Serban for significant and critical past contributions.
+David J. Gardner, Alan C. Hindmarsh, Daniel R. Reynolds, Steven B. Roberts, and
+Carol S. Woodward. We thank Radu Serban for significant and critical past
+contributions.
 
 Other contributors to SUNDIALS include: James Almgren-Bell, Lawrence E. Banks,
 Peter N. Brown, George Byrne, Rujeko Chinomona, Scott D. Cohen, Aaron Collier,
 Keith E. Grant, Steven L. Lee, Shelby L. Lockhart, John Loffeld, Daniel McGreer,
-Yu Pan, Slaven Peles, Cosmin Petra, Steven B. Roberts, H. Hunter Schwartz,
-Jean M. Sexton, Dan Shumaker, Steve G. Smith, Shahbaj Sohal, Allan G. Taylor,
+Yu Pan, Slaven Peles, Cosmin Petra, H. Hunter Schwartz, Jean M. Sexton,
+Dan Shumaker, Steve G. Smith, Shahbaj Sohal, Allan G. Taylor,
 Hilari C. Tiedeman, Chris White, Ting Yan, and Ulrike M. Yang.
 
 ## License ##
@@ -93,7 +106,7 @@ and [NOTICE](./NOTICE) files for details. All new contributions must be made
 under the BSD 3-clause license.
 
 **Please Note** If you are using SUNDIALS with any third party libraries linked
-in (e.g., LAPACK, KLU, SuperLU_MT, PETSc, or *hypre*), be sure to review the
+in (e.g., LAPACK, KLU, SuperLU_MT, PETSc, *hypre*, etc.), be sure to review the
 respective license of the package as that license may have more restrictive
 terms than the SUNDIALS license.
 

benchmarks/advection_reaction_3D/kokkos/ParallelGrid.hpp

@@ -72,7 +72,7 @@ class ParallelGrid
 public:
   // Constructor that creates a new ParallelGrid object.
   // [in] - the memory helper to use for allocating the MPI buffers
-  // [in,out] comm - on input, the overal MPI communicator, on output, the cartesian communicator
+  // [in,out] comm - on input, the overall MPI communicator, on output, the cartesian communicator
   // [in] a[] - an array of length 3 which defines the domain [a,b]
   // [in] b[] - an array of length 3 which defines the domain [a,b]
   // [in] npts[] - an array of length 3 which defines the number of mesh points in each dimension
@@ -563,7 +563,7 @@ class ParallelGrid
 
   GLOBALINT nx, ny, nz;   /* number of intervals globally       */
   int nxl, nyl, nzl;      /* number of intervals locally        */
-  int npx, npy, npz;      /* numner of processes                */
+  int npx, npy, npz;      /* number of processes                */
   sunrealtype dx, dy, dz; /* mesh spacing                       */
   sunrealtype ax, ay, az; /* domain in [a, b]                   */
   sunrealtype bx, by, bz;

benchmarks/advection_reaction_3D/kokkos/arkode_driver.cpp

@@ -33,7 +33,7 @@ typedef struct
   SUNNonlinearSolver local_nls;
 }* TaskLocalNewton_Content;
 
-/* Content accessor macors */
+/* Content accessor macros */
 #define GET_NLS_CONTENT(NLS) ((TaskLocalNewton_Content)(NLS->content))
 #define LOCAL_NLS(NLS)       (GET_NLS_CONTENT(NLS)->local_nls)
 

benchmarks/advection_reaction_3D/kokkos/check_retval.h

@@ -23,7 +23,7 @@
  * opt == 0  means the function allocates memory and returns a
  *           pointer so check if a NULL pointer was returned
  * opt == 1  means the function returns an integer where a
- *           value < 0 indicates an error occured
+ *           value < 0 indicates an error occurred
  * --------------------------------------------------------------*/
 static int check_retval(void* returnvalue, const char* funcname, int opt, int myid)
 {

benchmarks/advection_reaction_3D/kokkos/rhs3D.hpp

@@ -23,7 +23,7 @@
  * --------------------------------------------------------------*/
 
 /* Compute the advection term f(t,y) = -c (grad * y). This is done using
-   upwind 1st order finite differences.  At present, only periodic boudary
+   upwind 1st order finite differences.  At present, only periodic boundary
    conditions are supported, which are handled via MPI's Cartesian
    communicator (even for serial runs). */
 static int Advection(sunrealtype t, N_Vector y, N_Vector ydot, void* user_data)

benchmarks/advection_reaction_3D/raja/ParallelGrid.hpp

@@ -45,7 +45,7 @@ class ParallelGrid
 public:
   // Constructor that creates a new ParallelGrid object.
   // [in] - the memory helper to use for allocating the MPI buffers
-  // [in,out] comm - on input, the overal MPI communicator, on output, the cartesian communicator
+  // [in,out] comm - on input, the overall MPI communicator, on output, the cartesian communicator
   // [in] a[] - an array of length 3 which defines the domain [a,b]
   // [in] b[] - an array of length 3 which defines the domain [a,b]
   // [in] npts[] - an array of length 3 which defines the number of mesh points in each dimension
@@ -508,7 +508,7 @@ class ParallelGrid
 
   GLOBALINT nx, ny, nz; /* number of intervals globally       */
   int nxl, nyl, nzl;    /* number of intervals locally        */
-  int npx, npy, npz;    /* numner of processes                */
+  int npx, npy, npz;    /* number of processes                */
   REAL dx, dy, dz;      /* mesh spacing                       */
   REAL ax, ay, az;      /* domain in [a, b]                   */
   REAL bx, by, bz;

benchmarks/advection_reaction_3D/raja/advection_reaction_3D.hpp

@@ -89,7 +89,7 @@ struct UserData
   N_Vector vmask;
   N_Vector wmask;
 
-  /* problem paramaters */
+  /* problem parameters */
   sunrealtype xmax; /* maximum x value              */
   sunrealtype A;    /* concentration of species A   */
   sunrealtype B;    /* w source rate                */

benchmarks/advection_reaction_3D/raja/arkode_driver.cpp

@@ -33,7 +33,7 @@ typedef struct
   SUNNonlinearSolver local_nls;
 }* TaskLocalNewton_Content;
 
-/* Content accessor macors */
+/* Content accessor macros */
 #define GET_NLS_CONTENT(NLS) ((TaskLocalNewton_Content)(NLS->content))
 #define LOCAL_NLS(NLS)       (GET_NLS_CONTENT(NLS)->local_nls)
 

benchmarks/advection_reaction_3D/raja/check_retval.h

@@ -23,7 +23,7 @@
  * opt == 0  means the function allocates memory and returns a
  *           pointer so check if a NULL pointer was returned
  * opt == 1  means the function returns an integer where a
- *           value < 0 indicates an error occured
+ *           value < 0 indicates an error occurred
  * --------------------------------------------------------------*/
 static int check_retval(void* returnvalue, const char* funcname, int opt, int myid)
 {

benchmarks/diffusion_2D/diffusion_2D.cpp

@@ -11,7 +11,7 @@
  * SPDX-License-Identifier: BSD-3-Clause
  * SUNDIALS Copyright End
  * -----------------------------------------------------------------------------
- * Shared implementaiton file for 2D diffusion benchmark problem
+ * Shared implementation file for 2D diffusion benchmark problem
  * ---------------------------------------------------------------------------*/
 
 #include "diffusion_2D.hpp"

benchmarks/diffusion_2D/diffusion_2D.hpp

@@ -159,7 +159,7 @@ struct UserData
   sunrealtype* Ssend = NULL;
   sunrealtype* Nsend = NULL;
 
-  // Send requests for neighor exchange
+  // Send requests for neighbor exchange
   MPI_Request reqSW;
   MPI_Request reqSE;
   MPI_Request reqSS;
@@ -192,7 +192,7 @@ struct UserData
 
 struct UserOutput
 {
-  // Ouput variables
+  // Output variables
   int output     = 1;    // 0 = no output, 1 = stats output, 2 = output to disk
   int nout       = 20;   // number of output times
   N_Vector error = NULL; // error vector

benchmarks/diffusion_2D/main_arkode.cpp

@@ -329,7 +329,7 @@ int main(int argc, char* argv[])
     sunrealtype dTout = udata.tf / uout.nout;
     sunrealtype tout  = dTout;
 
-    // Inital output
+    // Initial output
     flag = uout.open(&udata);
     if (check_flag(&flag, "UserOutput::open", 1)) { return 1; }
 
@@ -524,7 +524,7 @@ void UserOptions::help()
   cout << endl;
   cout << "Integrator command line options:" << endl;
   cout << "  --rtol <rtol>           : relative tolerance" << endl;
-  cout << "  --atol <atol>           : absoltue tolerance" << endl;
+  cout << "  --atol <atol>           : absolute tolerance" << endl;
   cout << "  --nonlinear             : disable linearly implicit flag" << endl;
   cout << "  --order <ord>           : method order" << endl;
   cout << "  --fixedstep <step>      : used fixed step size" << endl;

benchmarks/diffusion_2D/main_cvode.cpp

@@ -310,7 +310,7 @@ int main(int argc, char* argv[])
     sunrealtype dTout = udata.tf / uout.nout;
     sunrealtype tout  = dTout;
 
-    // Inital output
+    // Initial output
     flag = uout.open(&udata);
     if (check_flag(&flag, "UserOutput::open", 1)) { return 1; }
 
@@ -479,7 +479,7 @@ void UserOptions::help()
   cout << endl;
   cout << "Integrator command line options:" << endl;
   cout << "  --rtol <rtol>           : relative tolerance" << endl;
-  cout << "  --atol <atol>           : absoltue tolerance" << endl;
+  cout << "  --atol <atol>           : absolute tolerance" << endl;
   cout << "  --ls <cg|gmres|sludist> : linear solver" << endl;
   cout << "  --lsinfo                : output residual history" << endl;
   cout << "  --liniters <iters>      : max number of iterations" << endl;

benchmarks/diffusion_2D/main_ida.cpp

@@ -305,7 +305,7 @@ int main(int argc, char* argv[])
     sunrealtype dTout = udata.tf / uout.nout;
     sunrealtype tout  = dTout;
 
-    // Inital output
+    // Initial output
     flag = uout.open(&udata);
     if (check_flag(&flag, "UserOutput::open", 1)) { return 1; }
 
@@ -460,7 +460,7 @@ void UserOptions::help()
   cout << endl;
   cout << "Integrator command line options:" << endl;
   cout << "  --rtol <rtol>           : relative tolerance" << endl;
-  cout << "  --atol <atol>           : absoltue tolerance" << endl;
+  cout << "  --atol <atol>           : absolute tolerance" << endl;
   cout << "  --ls <cg|gmres|sludist> : linear solver" << endl;
   cout << "  --liniters <iters>      : max number of iterations" << endl;
   cout << "  --epslin <factor>       : linear tolerance factor" << endl;

benchmarks/nvector/plot_nvector_performance_results.py

@@ -15,7 +15,7 @@
 # This script plots the output from test_nector_performance_* and assumes:
 #   1. vector lengths are powers of two starting from 0, and
 #   2. output files are named: output_nelem_nvec_nsum_ntest_timing.txt
-# where nelem is the number of elements in the vector, nvec is the nuber of
+# where nelem is the number of elements in the vector, nvec is the number of
 # vectors, nsum is the number of sums, ntest is the number of tests, and timing
 # indicates if timing was enabled.
 # -----------------------------------------------------------------------------