Affects:
* store_dlsch_buffer()
* assign_rbs_required()

When the RLC layer returns error codes (like dropped packets), this is still
counted as PDCP layer traffic. In order to reflect dropped packets or
malfunctioning, this patch makes that upon an RLC error code, the function is
exited immediately. In this case, it is not counted as additional PDCP traffic.

An example is too much UDP traffic. If the RLC buffer is saturated and the
packet therefore refused, incoming traffic won't be counted anymore. With this
patch, the PDCP layer statistics won't count such traffic anymore.

the integration of the nfapi-L2-emulator-for-develop (f254107b) added more
types in the RCconfig_flexran(), but they were partially wrong. This commit
fixes those types after the rebase of this branch on top of 2018.w19

* give the correct UE_id instead an index of the protobuf message (they can
  diverge from each other)
* in the PDCP part of the RAN API, the RAN API has to make the conversion UE_id
  (MAC) to UID (PDCP)
* a follow up commit could clean up the interfaces:
  - the MAC functions take the UE_id
  - the RRC functions take the RNTI
  - the PDCP functions take the UID
  this would mean the caller has to find those ID but would decrease
  computations since the RAN API functions do not have to do that themselves

If the buffer in the RLC is full, testing the buffer state over and over again
slows the DL traffic down. In order to circumvent this, the PDCP will drop any
data during a configurable time (compile-time) before it delivers data to RLC
again.

To change this, see the constant TM_SKIP_FULL_BUF_MS in pdcp.h.

- there can be a race between the FlexRAN agent reading the RC.mac and its
  allocation in main.c
- In order to circumvent this, change the allocation by allocating
  everything "into" a local variable
- finally "make it visible" by storing  the pointer in RC.mac

don't fail in library utility function, return with error code instead and let
application decide what is best on how to proceed

* check whether a library has been loaded before; in this case, don't allocate
  new memory but load from old library again
* vital parameters, previous exit_fun are checked by AssertFatal()
* structure has changed a bit to make it easier to follow (instead of nested
  ifs includes one goto to the end of the function)
* formatting has been improved

* put mutex lock around instance_prach
* destroy attributes

* instead of the lengthy YAML parsing, can now handle protobuf message
* after setting the parameters, send a message to ENB_APP to perform it
* old YAML way still works

* join pthread_FH (ru_thread) once all condition variables have been set
* join RU FHTX only if it has been started (check as for pthread_create())
* join PRACH thread in the monolithic case
* send broadcast on condition variable cond_eNBs, as there can be multiple
  waits on it
* integrate stop_ru(*ru) into kill_RU_proc(*ru)
* Correct memory freeing

* init_td_thread()/init_te_thread() are only initialized depending on condition
  get_nprocs() > 2 && codingw
* apply the same the same for the functions kill_td_thread()/kill_te_thread()

* properly allocate memory for scheduler name/sorting
* free mem of sc_update, include todo for slice_config
* set has_x to one, add fct whether sorting update necessary
* set n_ul/n_dl to correct value after every loop iteration
* support for delete of multiple slices
* allocate memory for scheduler name in slice_config
* check that we really set the scheduler

- in order to facilitate the steps as verification and parameter setting,
  slices are "filled up" with default values:
  * if it is an update, with the values of the concerned slice
  * if it is a new one, with the parameters of slice 0
- because of this, it can be assumed that all slices have all parameters

- whether a scheduler callback has to be changed is now handled by the agent