Juan 14/04/2022:MNH/gradient_m.f90/shuman_device.f90, Bypass Cray GPU  bug/opti: use present_cr + do concurrent + acc_nv loop where needed

Juan 19/05/2021: add loop independent collapse , for all do conccurent & BR_ routine , or the do loop seq !!!

Philippe 23/10/2019: OpenACC: rename _OPENACC key to MNH_OPENACC (_OPENACC is a standard OpenACC macro)

Philippe 03/10/2019: OpenACC: replace 'declare present' by 'data present' directives for assumed-size dummy arrays (to be standard compliant)

Philippe 02/10/2019: OpenACC: remove declare present directives from module declaration sections (to be standard compliant)

Philippe 19/01/2017: OpenACC: gradient + shuman: ported modifications to device versions of subroutines

Philippe 27/06/2016: 2 shuman versions: with or without linearized loops + use of shuman instead of macros for OpenACC in ppm.f90

Philippe 23/06/2016: OpenACC: split shuman.f90 into 2 files: 1 with and 1 without the *_DEVICE versions