- use (starpu_cpu_func_t) 1 trick, same as cuda_func
    - cpu funtions are not defined anymore avoiding the dependency to coreblas
    - add #if !defined(CHAMELEON_SIMULATION) where it is needed
    - remove dependency to the coreblas library (become useless)
    - remove useless simucblas, simulapacke libraries
    - remove CHAMELEON_SIMULATION_MAGMA cmake variable and definition
      - keep using CHAMELEON_USE_CUDA and CHAMELEON_USE_MAGMA to consider CUDA kernels
      - this avoid to introduce useless new variables
    - work on messages

to change the default values for tag_width(=31) and tag_sep(=24).
Today, it serves only with starPU.

To use your own progress indicator,
- Define a function with prototype "void my_update_progress(int currentValue, int maximumValue)"
- Pass it to chameleon with "ierr=MORSE_Set_update_progress_callback(my_update_progress)"
- Activate progress indicator with MORSE_Enable(MORSE_PROGRESS)

morse_desc_init_user(): If one of the function get_* is NULL, we switch back to the default, like in morse_desc_init()

This allows to change only 1 or 2 of the 3 functions, and keep the other
unchanged. For example, in OOC, only the 1st is modified (to always return NULL).

Remove large blocks of duplicated code by calling morse_desc_init_user() from morse_desc_init() and morse_desc_init_diag()

to avoid reverse dependency libcoreblas->libchameleon
set_coreblas_gemm3m_enabled()/get_coreblas_gemm3m_enabled() allow to set/get
this variable.

This routine, available in MKL, does a product in 6n^3 ops instead of 8n^3
but is interesting only for "large enough" matrices (to be tested...)
Potentially, we gain 25 % in all complex computations.
It could be interesting to look for it / implement it in cuda.

!!! Note that the flop counters are not updated         !!!
!!! In C/Z accuracy, most flops counter should be x0.75 !!!

IT is OFF by default
It is activated with MORSE_Enable(MORSE_GEMM3M)
In the timing routines, it is activated with --gemm3m

IT is OFF by default
It is activated with MORSE_Enable(MORSE_PROGRESS)
In the timing routines, it is activated with --progress
No progress is printed for tasks faster than 10 seconds

The new codelet is added for all 3 runtimes but only tested with starpu ;-)
An example test7 is added : it is a copy of test6 that uses the new function to build the matrices

Fix a large bunch of warnings, and there are still mistakes that should be fixed before the SC release

instead of introducing RUNTIME_distributed_barrier

add cudablas library to make calls to cuda kernels (magma here, cublas will follow), no more calls to magma in runtime/starpu codelets

MORSE_Distributed_size, MORSE_Distributed_rank so that applications do not
hardcode the use of MPI.

Introduce RUNTIME_distributed_rank, RUNTIME_distributed_size,
RUNTIME_distributed_barrier, so that MORSE does not hardcode the use of MPI
either.

This allows to use simgrid-mpi.

add MORSE_Pause/Resume function to avoid CPU consumption when no tasks have to be executed - call starpu_pause/resume, do nothing for quark (no need)

in MORSE_Desc_Create_User we can use the matrix in mat field without allocating it --> avoid to free it in this case, users could need it and will manage deallocation

change the way we include our own header files --> relative to the root - when plasma is in the same env, chameleon can take some headers not belonging to it (ex: #include descriptor.h, this file states in plasma install dir also) which make compilation errors