Started by an SCM change49ed9ae72a5eb28dbc036dd3e14b9312eea840feaced9ae72a5eb28dbc036dd3e14b9312eea840feacrefs/remotes/origin/mastered9ae72a5eb28dbc036dd3e14b9312eea840feaced9ae72a5eb28dbc036dd3e14b9312eea840feacrefs/remotes/origin/mastered9ae72a5eb28dbc036dd3e14b9312eea840feaced9ae72a5eb28dbc036dd3e14b9312eea840feacrefs/remotes/origin/masterhttps://github.com/larson-group/clubb.gitfalse#492997781240811clubb_nvhpc_gpu_column_mirror_test #4949false498411FAILURE1743148986099http://carson.math.uwm.edu/jenkins/job/clubb_nvhpc_gpu_column_mirror_test/49/src/CLUBB_core/advance_clubb_core_module.F90src/clubb_driver.F90"src/print *, \"thlm_init = \", thlm_init.f90"src/CLUBB_core/stats_clubb_utilities.F90ed9ae72a5eb28dbc036dd3e14b9312eea840feac1743109355000http://carson.math.uwm.edu/jenkins/user/noreplynoreplynoreply@github.comClubb driver upgrades (#1225) * Moving stuff around in preparation for preparation for GPUization * Small modifications, adding explicit loops * Small bug fix and explicit loop additions * Removing ngrdcol dimension from initialize_clubb, essentially reverting changes. * Removing array slicing by using local variables of nzm/nzt dim * GPUization of initialization variables * Removing need for single_col versions of the pdf_params, this (weirdly) saves a huge amount of runtime * Adding comment * Adding 2D version of various routines, including initialize_clubb, but only calling initialize_clubb with 1 column. * Fixes to make work with latest clubb changes * Merging with master 2025-03-27 16:02:35 -0500ed9ae72a5eb28dbc036dd3e14b9312eea840feacClubb driver upgrades (#1225)editsrc/CLUBB_core/advance_clubb_core_module.F90editsrc/clubb_driver.F90editsrc/CLUBB_core/stats_clubb_utilities.F90add"src/print *, \"thlm_init = \", thlm_init.f90"githttp://carson.math.uwm.edu/jenkins/user/noreplynoreplyfalse50http://carson.math.uwm.edu/jenkins/job/clubb_nvhpc_gpu_column_mirror_test/50/48http://carson.math.uwm.edu/jenkins/job/clubb_nvhpc_gpu_column_mirror_test/48/