Add CTest test cases from existing tests (run_tests.sh)

[siggmo]

This PR should not be merged before PR #19!

Adressing #9

CMake comes with a testing tool called CTest. This PR adds the existing test cases from test/run_tests.sh as CTest test cases. Instead of maintaining a dedicated test shell script, CTest allows to define the tests in the existing CMake build system and offers useful functions for testing like measuring execution time, collecting outputs, checking for assertions etc.

You can run the tests by calling ctest in the build directory. To see the full output, add the -VV flag.

Features

• Add the four test cases LinearThermalIsotropic, LinearElasticIsotropic, PseudoPlasticLinearHardening and VonMisesPlasticLinearIsotropicHardening.
• Use maximum possible number of MPI processes (more than 8 caused this error: "[ FANS3D_Grid ] ERROR: Please decrease the number of processes or increase the grid size to ensure that each process has at least 4 boxels in the x direction.").
• HDF5 output files go into <build-dir>/test/, such that running the tests only creates/alters files in the build tree

Open points

• No result checking! Should we add that?
• Do we stay with max. number of MPI processes favoring execution time, or rather focus on reproducability and fix the number?

[siggmo]

Apparently MPI isn't able to obtain the required "slots" when running in the GitHub Action. Locally on my machine it worked... I have to investigate that further.

Here's the error:

Test project /FANS/build
    Start 1: LinearThermalIsotropic
1/4 Test #1: LinearThermalIsotropic ....................***Failed    0.04 sec
--------------------------------------------------------------------------
There are not enough slots available in the system to satisfy the 4
slots that were requested by the application:

  /FANS/build/FANS

Either request fewer slots for your application, or make more slots
available for use.

A "slot" is the Open MPI term for an allocatable unit where we can
launch a process.  The number of slots available are defined by the
environment in which Open MPI processes are run:

  1. Hostfile, via "slots=N" clauses (N defaults to number of
     processor cores if not provided)
  2. The --host command line parameter, via a ":N" suffix on the
     hostname (N defaults to 1 if not provided)
  3. Resource manager (e.g., SLURM, PBS/Torque, LSF, etc.)
  4. If none of a hostfile, the --host command line parameter, or an
     RM is present, Open MPI defaults to the number of processor cores

In all the above cases, if you want Open MPI to default to the number
of hardware threads instead of the number of processor cores, use the
--use-hwthread-cpus option.

Alternatively, you can use the --oversubscribe option to ignore the
number of available slots when deciding the number of processes to
launch.

Do you have any ideas what could be going wrong?

[sanathkeshav]

"For example, GitHub Actions only have 2 cores available for Windows and Linux, yet 3 for macOS."

mentioned in: vercel/turborepo#761

can't request 4 cores for building or for testing that's all.

[siggmo]

"For example, GitHub Actions only have 2 cores available for Windows and Linux, yet 3 for macOS."

mentioned in: vercel/turborepo#761

can't request 4 cores for building or for testing that's all.

Ahh okay makes sense! I used the CMake function ProcessorCount to determine the available cores, but I guess then sth. is not working correctly with this function.

[sanathkeshav]

hehe