[openPMD] Communication of upstream beam particle number

[SeverinDiederichs]

In the parallel beam I/O, the total number of beam particles of the upstream ranks is required. This number is passed via blocking MPI communication down the pipeline.

It was tested that it gives the correct behavior on GPUs:

[1,4]<stdout>:num particles upstream before exchange 0
[1,4]<stdout>:num particles upstream after exchange 7
[1,3]<stdout>:num particles upstream before exchange 0
[1,3]<stdout>:num particles upstream after exchange 2901
[1,2]<stdout>:num particles upstream before exchange 0
[1,2]<stdout>:num particles upstream after exchange 49869
[1,1]<stdout>:num particles upstream before exchange 0
[1,1]<stdout>:num particles upstream after exchange 97008
[1,0]<stdout>:num particles upstream before exchange 0
[1,0]<stdout>:num particles upstream after exchange 99994

• Small enough (< few 100s of lines), otherwise it should probably be split into smaller PRs
• Tested (describe the tests in the PR description)
• Runs on GPU (basic: the code compiles and run well with the new module)
• Contains an automated test (checksum and/or comparison with theory)
• Documented: all elements (classes and their members, functions, namespaces, etc.) are documented
• Constified (All that can be const is const)
• Code is clean (no unwanted comments, )
• Style and code conventions are respected at the bottom of https://github.com/Hi-PACE/hipace
• Proper label and GitHub project, if applicable

[MaxThevenet]

Looks great, thanks! Two cosmetic points:

• Names RecvNumParticlesUpstreamRanks and PassNumParticlesUpstreamRanks are not fully consistent (the opposite or Recv would be Send, and the second function is better described with e.g. PassNumParticlesDownstreamRanks). In fact, we already have names for sending data along the pipeline, this is called Wait and Notify. Could you name these functions accordingly, e.g., WaitNumParticles and NotifyNumParticles?
• Could you move the calls to these functions to Wait and Notify? I think it's cleaner and more consistent (let's have all the pipelined data handled there.)

[MaxThevenet]

Suggested change

	/**< Loop over beams and pass total number of beam particles on upstream ranks */
	void PassNumParticlesUpstreamRanks (MPI_Comm a_comm_z);
	/** Loop over beams and pass total number of beam particles on upstream ranks */
	void PassNumParticlesUpstreamRanks (MPI_Comm a_comm_z);

Doxygen syntax is

int something /**< description */

or

/** description */
int something