It now takes paths into account correctly

cc @HanatoK I fixed the problem. Namely, plumed was linking by mistake
another boost version (homebrew).

Notice that this breaks previous "check" target, that I think
was anyway useless.

check: does the check with the non-installed plumed
installedcheck: does the check with the installed plumed

In case MPIEXEC was passed at configure time, it is used.
In case not, env var PLUMED_MPIRUN is checked (for backward compatibility).
Otherwise, `mpirun` is used.

There was some "return 1". Now that exit code is checked, this would
have triggered an error

Report a failure in the following two cases:
- unknown test type
- test program exiting with non-zero code

hopefully the last time..

appliable -> applicable

failures are moved to the end of the list

(suggested by @valsson)

The parameter order in DRRAxis::setPeriodicity is fixed, so regenerate the
regtest in drr/rt-extract2dpbc.

The parameter order in DRRAxis::setPeriodicity is fixed, so regenerate the
regtest in drr/rt-extract2dpbc.

It doesn't run on travis-ci, but on my mac it runs successfully.

Can be used to skip a regtest based on some complex test.
E.g., to skip a test on mac add this to the config file:

function plumed_custom_skip(){
  if [ "$(uname)" == "Darwin" ] ; then
    return 0
  fi
}