feat: [WIP] Add advection-2d testcase as benchmark

[sbstndb]

Description

-	Add advection-2 (the balanced version) as benchmark
-	Support of MPI through Google benchmark
-	TODO :
	-	resolve benchmark_set
	-	Add other testcases as benchmark
	-	use different stencil size and type

Related issue

How has this been tested?

Tested on my personal machine

• with mpirun -np 1
• with mpirun -np 4 which is faster

Code of Conduct

By submitting this PR, you agree to follow our Code of Conduct

• I agree to follow this project's Code of Conduct

[sbstndb]

I have to manage the WITH_MPI=OFF case before merging.

[sbstndb]

At this point, the new benchmark output is :

----------------------------------------------------------------------------
Benchmark                                  Time             CPU   Iterations
----------------------------------------------------------------------------
ADVECTION_2D<5, 8>/iterations:1         <result>            1
ADVECTION_2D<5, 10>/iterations:1        <result>            1
ADVECTION_2D<5, 12>/iterations:1        <result>            1
ADVECTION_2D<5, 14>/iterations:1        <result>            1
ADVECTION_2D<6, 8>/iterations:1         <result>            1
ADVECTION_2D<8, 10>/iterations:1         <result>            1
ADVECTION_2D<10, 12>/iterations:1       <result>           1
ADVECTION_2D<12, 14>/iterations:1      <result>            1
ADVECTION_2D<6, 6>/iterations:1        <result>            1
ADVECTION_2D<8, 8>/iterations:1         <result>            1
ADVECTION_2D<10, 10>/iterations:1       <result>            1
ADVECTION_2D<12, 12>/iterations:1        <result>            1
ADVECTION_2D<14, 14>/iterations:1       <result>           1

Hence we have :

• Uniform
• MRA with max-level - min-level = 2
• MRA where min_level = 5

I think this could be merged.