ELPA 2021.05.001 release

See merge request !73

modify max_stored_rows

See merge request !70

Configure test

See merge request !69

It turned out that users ignored the warnings of the test programs
if their MPI library did not provide "MPI_THREAD_SERIALIED" or
"MPI_THREAD_MULTIPLE".

Thus for safty reasons, ELPA does from now on during the call to
"elpa_setup" check the provided threading level of the MPI library.
If the provided level is too low, ELPA will do the following

- limit the number of OpenMP threads **internal** to ELPA to 1
- print a warning about this
- ignore the settings of the user via
-- the OMP_NUM_THREADS variable
-- calling the set-method set("omp_threads", value)

These settings will **not** affect the run-time of a threaded
blas and lapack library, **if** the number of threads for these
libraries can be controlled independently of the OMP_NUM_THREADS
variable (for example for Intel's MKL one can use MKL_NUM_THREADS)

In case of MPI calculations, ELPA does _not_ anymore rely on the user
to check that his provided BLACSGRID setup is reasonable.
Now ELPA does check this (again) during ELPA_SETUP and returns with
an error if this check fails

Change buffer size for the pure blocked configuration

See merge request !68

in pure blocked config the buffer size should be slightly increased to accomodate message sizes

Correct the path to the USERS_GUIDE.md

See merge request !67

USERS_GUIDE.md is now under ./documentation