Started by an SCM change5943ebd14c5ba84fb2caf021b7f96b1d5b7ddba93f93ebd14c5ba84fb2caf021b7f96b1d5b7ddba93f9refs/remotes/origin/master3ebd14c5ba84fb2caf021b7f96b1d5b7ddba93f93ebd14c5ba84fb2caf021b7f96b1d5b7ddba93f9refs/remotes/origin/master3ebd14c5ba84fb2caf021b7f96b1d5b7ddba93f93ebd14c5ba84fb2caf021b7f96b1d5b7ddba93f9refs/remotes/origin/masterhttps://github.com/larson-group/clubb.githttp://carson.math.uwm.edu/jenkins/blue/organizations/jenkins/clubb_BFB_gfortran_test/detail/clubb_BFB_gfortran_test/594/artifactshttp://carson.math.uwm.edu/jenkins/blue/organizations/jenkins/clubb_BFB_gfortran_test/detail/clubb_BFB_gfortran_test/594/changeshttp://carson.math.uwm.edu/jenkins/blue/organizations/jenkins/clubb_BFB_gfortran_test/detail/clubb_BFB_gfortran_test/594/http://carson.math.uwm.edu/jenkins/blue/organizations/jenkins/clubb_BFB_gfortran_test/detail/clubb_BFB_gfortran_test/594/testsfalsefalse#594853596847420clubb_BFB_gfortran_test #594594false5942126SUCCESS1773736149070http://carson.math.uwm.edu/jenkins/job/clubb_BFB_gfortran_test/594/run_scripts/run_scm.pysrc/clubb_driver.F902108b26241ca7695b3358c5f905385e79189cdc21773679567000http://carson.math.uwm.edu/jenkins/user/hueblerGunther Hueblerhuebler@uwm.eduForgot to add the changes to make concurrency safe - which caused random failures when using the new parallel run_scm_all.py with options like -out_dir or -ngrdcol 2026-03-16 11:46:07 -05002108b26241ca7695b3358c5f905385e79189cdc2Forgot to add the changes to make concurrency safe - which caused random failures when using the new parallel run_scm_all.py with options like -out_dir or -ngrdcoleditrun_scripts/run_scm.pyeditsrc/clubb_driver.F90cmake/toolchains/linux_x86_64_gfortran.cmakecmake/toolchains/linux_x86_64_gcc.cmakecmake/toolchains/linux_x86_64_nvfortran.cmakecmake/toolchains/linux_x86_64_ifx.cmakecmake/toolchains/linux_x86_64_nvhpc.cmakecmake/toolchains/linux_x86_64_intel.cmakecompile.pycmake/toolchains/linux_x86_64_cce.cmake6b49cb739e4c6970da64b808e791b85a227b44b31773718458000http://carson.math.uwm.edu/jenkins/user/noreplynoreplynoreply@github.comSlightly more general compile setup (#1277) * Making compiler lookup a little more general, this allows us to compile on Derecho too. Now it uses LMOD_FAMILY_COMPILER by default, then tries FC as a backup, and has a map of names so that things like intel/intel-oneapi map to the same things. Also adding a toolchain for the cray compiler (cce). * Removing old toolchains 2026-03-16 22:34:18 -05006b49cb739e4c6970da64b808e791b85a227b44b3Slightly more general compile setup (#1277)deletecmake/toolchains/linux_x86_64_gfortran.cmakedeletecmake/toolchains/linux_x86_64_ifx.cmakeaddcmake/toolchains/linux_x86_64_nvhpc.cmakeaddcmake/toolchains/linux_x86_64_gcc.cmakeeditcompile.pyaddcmake/toolchains/linux_x86_64_cce.cmakedeletecmake/toolchains/linux_x86_64_nvfortran.cmakeaddcmake/toolchains/linux_x86_64_intel.cmakecmake/deps/GPTLDependency.cmake3ebd14c5ba84fb2caf021b7f96b1d5b7ddba93f91773718471000http://carson.math.uwm.edu/jenkins/user/noreplynoreplynoreply@github.comGPTL cmake update (#1279) * Making compiling with gptl more robust. Before it could only compile when we used openmp too, but that's been worked around, and now it can compile on nvhpc too thanks to the static and disabled-shared flags that were added. Also works with gcc now. * Adding comment * Small cleanup and a better comment 2026-03-16 22:34:31 -05003ebd14c5ba84fb2caf021b7f96b1d5b7ddba93f9GPTL cmake update (#1279)editcmake/deps/GPTLDependency.cmakegithttp://carson.math.uwm.edu/jenkins/user/noreplynoreplynoreplyhttp://carson.math.uwm.edu/jenkins/user/hueblerGunther HueblerhueblerfalseStarted by an SCM change595bdee6e6dbac76a996cdf73a88522c47affdd4d90bdee6e6dbac76a996cdf73a88522c47affdd4d90refs/remotes/origin/masterbdee6e6dbac76a996cdf73a88522c47affdd4d90bdee6e6dbac76a996cdf73a88522c47affdd4d90refs/remotes/origin/masterbdee6e6dbac76a996cdf73a88522c47affdd4d90bdee6e6dbac76a996cdf73a88522c47affdd4d90refs/remotes/origin/masterhttps://github.com/larson-group/clubb.gitfalse#5952606814847420clubb_BFB_gfortran_test #595595false5952218SUCCESS1773822548482http://carson.math.uwm.edu/jenkins/job/clubb_BFB_gfortran_test/595/jenkins_tests/clubb_nvhpc_gpu_vs_cpu_diffness/Jenkinsfilejenkins_tests/clubb_nvhpc_gpu_column_mirror_test/Jenkinsfile02123793194ca95295bbd90c35129faf395b95251773771251000http://carson.math.uwm.edu/jenkins/user/hueblerGunther Hueblerhuebler@uwm.eduToolchain names were changed, so the references to them should've changed in the jenkins tests. 2026-03-17 13:14:11 -050002123793194ca95295bbd90c35129faf395b9525Toolchain names were changed, so the references to them should've changed in the jenkins tests.editjenkins_tests/clubb_nvhpc_gpu_vs_cpu_diffness/Jenkinsfileeditjenkins_tests/clubb_nvhpc_gpu_column_mirror_test/Jenkinsfilerun_scripts/test_monoflux_limiter_GPU.py36ca61e3203b201832c5ca74ce764fc7e97786a81773771352000http://carson.math.uwm.edu/jenkins/user/hueblerGunther Hueblerhuebler@uwm.eduToolchain names were changed, so the references to them should've changed in this test too. 2026-03-17 13:15:52 -050036ca61e3203b201832c5ca74ce764fc7e97786a8Toolchain names were changed, so the references to them should've changed in this test too.editrun_scripts/test_monoflux_limiter_GPU.pysrc/G_unit_test_types/pdf_parameter_tests.F90src/CLUBB_core/adg1_adg2_3d_luhar_pdf.F90src/G_unit_tests.F90src/G_unit_test_types/spurious_source_test.F90c15b1d3c005fdffa76902e976acdb5db5dbde0121773804486000http://carson.math.uwm.edu/jenkins/user/noreplynoreplynoreply@github.comADG1_w_closure consistency fix (#1282) * G_unit_test / ADG1_w_closure fix. A recent commit (bff619bdb2836f83ac658be8b8c258d41aaa6e14) introduced a change to ADG1_ADG2_responder_params that caused an inconsistency in the pdf calculations that the G_unit tests picked up on. This corrects it, but is also BIT_CHANGING, since it changes the calculation slightly when wp2 is exactly tolerance. * Adding test to ensure err_info is correctly set by adg1 routines when calling with bad inputs - also moving wp2 check to ADG1_w_closure because it makes more sense there 2026-03-17 22:28:06 -0500c15b1d3c005fdffa76902e976acdb5db5dbde012ADG1_w_closure consistency fix (#1282)editsrc/G_unit_test_types/pdf_parameter_tests.F90editsrc/CLUBB_core/adg1_adg2_3d_luhar_pdf.F90editsrc/G_unit_tests.F90editsrc/G_unit_test_types/spurious_source_test.F90src/CLUBB_core/matrix_solver_wrapper.F90jenkins_tests/clubb_error_handling_gfortran_test/Jenkinsfilerun_scripts/test_fatal_error_handling.pysrc/CLUBB_core/lapack_wrap.F90bdee6e6dbac76a996cdf73a88522c47affdd4d901773805495000http://carson.math.uwm.edu/jenkins/user/noreplynoreplynoreply@github.comFixing clubb_error_handling_gfortran_test (#1281) * Fixing error test by making the code check for NaNs in all solver cases - before it was only checking in Lapack. Also making a python version of the test and updating jenkins script to use it. * A little cleanup * Adding more comments and making a little data structure for the tests 2026-03-17 22:44:55 -0500bdee6e6dbac76a996cdf73a88522c47affdd4d90Fixing clubb_error_handling_gfortran_test (#1281)editsrc/CLUBB_core/lapack_wrap.F90editsrc/CLUBB_core/matrix_solver_wrapper.F90editjenkins_tests/clubb_error_handling_gfortran_test/Jenkinsfileaddrun_scripts/test_fatal_error_handling.pygithttp://carson.math.uwm.edu/jenkins/user/noreplynoreplyhttp://carson.math.uwm.edu/jenkins/user/hueblerGunther Hueblerfalse596http://carson.math.uwm.edu/jenkins/job/clubb_BFB_gfortran_test/596/594http://carson.math.uwm.edu/jenkins/job/clubb_BFB_gfortran_test/594/Started by an SCM change593bff619bdb2836f83ac658be8b8c258d41aaa6e14bff619bdb2836f83ac658be8b8c258d41aaa6e14refs/remotes/origin/masterbff619bdb2836f83ac658be8b8c258d41aaa6e14bff619bdb2836f83ac658be8b8c258d41aaa6e14refs/remotes/origin/masterbff619bdb2836f83ac658be8b8c258d41aaa6e14bff619bdb2836f83ac658be8b8c258d41aaa6e14refs/remotes/origin/masterhttps://github.com/larson-group/clubb.gitfalse#5932281239847420clubb_BFB_gfortran_test #593593false5931966SUCCESS1773649748168http://carson.math.uwm.edu/jenkins/job/clubb_BFB_gfortran_test/593/jenkins_tests/clubb_plot_nvhpc_gpu_vs_cpu/Jenkinsfilesrc/CLUBB_core/pdf_closure_module.F90src/G_unit_test_types/pdf_parameter_tests.F90src/CLUBB_core/adg1_adg2_3d_luhar_pdf.F90run_scripts/run_scm_all.pyjenkins_tests/clubb_nvhpc_gpu_column_mirror_test/Jenkinsfilejenkins_tests/clubb_driver_test/Jenkinsfilebff619bdb2836f83ac658be8b8c258d41aaa6e141773630552000http://carson.math.uwm.edu/jenkins/user/noreplynoreplynoreply@github.comStability fix (#1276) * Adding multiprocesses to run_scm_all.py, making the GPU jenkins tests only use 1 thread though, otherwise the different processes will fight for GPU usage * Making the if statement in ADG1_ADG2_responder_params use a strict inequality - this was causing numerical issues when fields were close to tolerance. This will be slightly BIT_CHANGING for most cases. * Adding comments to run_scm_all.py and removing the if-case in ADG1_ADG2_responder_params in favor of an explicit error check * Changing nprocs to nproc and updating jenkins tests 2026-03-15 22:09:12 -0500bff619bdb2836f83ac658be8b8c258d41aaa6e14Stability fix (#1276)editjenkins_tests/clubb_driver_test/Jenkinsfileeditsrc/G_unit_test_types/pdf_parameter_tests.F90editsrc/CLUBB_core/pdf_closure_module.F90editjenkins_tests/clubb_nvhpc_gpu_column_mirror_test/Jenkinsfileeditjenkins_tests/clubb_plot_nvhpc_gpu_vs_cpu/Jenkinsfileeditsrc/CLUBB_core/adg1_adg2_3d_luhar_pdf.F90editrun_scripts/run_scm_all.pygithttp://carson.math.uwm.edu/jenkins/user/noreplynoreplyfalse594http://carson.math.uwm.edu/jenkins/job/clubb_BFB_gfortran_test/594/592http://carson.math.uwm.edu/jenkins/job/clubb_BFB_gfortran_test/592/