Modifications to correct inversed gaussian grids.
(in work) Jun 5 2015 (ok for Mark)
Modifications from Valin to replace wordint with F2CL
(in work_valin) Sept 10 2015
Correct bug in ezwdint to allocate latlon vectors Jan 2016
Replace fstinl with fstinf and fstsui in ez_lireEnrPositionnels Sept 2016
Correct bug in gdxywdval-memory overlap, gdllfxy-memory leak (Vanh S. ) Jan 2017

IBM (AIX):
. ssmuse-sh -d hpcs/201402/02/base -d hpcs/ext/xlf_13.1.0.10


intel13sp1u2:
. ssmuse-sh -d hpcs/201402/02/base -d hpcs/201402/02/intel13sp1u2

using Compiler rules from: 
/users/dor/armn/lib/userlibs/Linux_x86-64/intel13sp1u2/Compiler_rules
and put link in 
$HOME/userlibs/Linux_x86-64/intel13sp1u2/Compiler_rules
to make it the same rules as in librmn.a

For GPSC
make obj (OPTIMIZ=2)

For HARE (broadwell)
make obj OPTIMIZ=3 CFLAGS="-fast-transcendentals -no-prec-div -ip -no-prec-sqrt"

make obj
