In the previsous commit the physics modifications between cy48t1 and cy48t3
have been introduced in PHYEX but still using the cy48t1 pack.
This commit introduces new modifications (in arome/ext) to use the cy48t3 pack.
Results are different due to modifications introduced betwwen cy48t1 and cy48t3
in parts other than the physics.
So, a new reference is needed (modifications in check_commit_ial.sh, INSTALL_pack_ial.md).

Quentin 19/08/2022: bugfix at merge (mode_prandtl AROME version was used instead of MNH for RED variables)

- bugfix with ADAP
- FLOAT() to REAL() in lima_adjust*
- bugfix budget in rain_ice_elec
- others = minors or comments

The budget subroutines have been cleaned
Unused subroutines in the arome directory have been removed
Minor update of the check_commit_* scripts

Quentin 07/07/2022: turb, LDIAG_IN_RUN and XCURRENT_TKE_DISS moved outside of turb (passing as args)

Quentin 20/06/2022: bugfix, missing add arguments in call turb of optionnal PTR and PDISS for mesonh

Move some codes from arome specific directory to the common one
Hack FCM to build a shared lib
Compilation of fiat through FCM

For GPU, ice4_nucleation cannot be put in an include file
as included subroutine must not contain local variables.

mesonh/micro/rain_ice.f90 -> mesonh/micro/rain_ice_old.f90
mesonh/micro/rain_ice_red.f90 -> mesonh/micro/rain_ice.f90
bugfix for the new condensation interface
interface harmonisation betwwen mesonh/micro/rain_ice_red.f90 and arome/micro/rain_ice.F90

Original code contained a bug (autoconversion ri->rs was not set
to 0 if ldsoft when ri content gets lower than the threshold).
Introducing REPRO48 in ice4_slow allows to suppress the one
introduced in compute_pdf to compensate.

Simpler array declaration
ITPL initialisation
Cp computed using rain

REPRO48 key added in aro_adjust
PA/PB size declaration changed in ice4_tendencies (to allow bound checking)

Small bugfixes to enable the compilation of AROME from scratch

Quentin 18/03/2022: modd_turbn different in arome/mesonh (TYPE for grid-nesting in mesonh). Move common->arome

CMICRO variable passed to turb
REPRO48/REPRO55 used (instead of #ifndef REK) in sedimentation_stat

Hail and graupel (with ICE4 option) budgets was wrong in the initial commit and different,
but still wrong, in the previsous commit.
Bug correction in fast_rh protected by a REPRO48 macro

Quentin 02/03/2022: Merge MNH->COMMON mode_prandtl.F90 SMOOTH_FUNCTION and correction from mesonh (in REPRO48 for AROME)