Verification of the CMake-based build system

From AtmosWiki
Jump to: navigation, search

This page contains the details regarding my (Liam Bindle) verification of the CMake-based build system.

Verification for GEOS-Chem Classic

  • Compiles each run directory with Make and CMake BSs.
    • Make's build uses make mpbuild (are all of the specialty simulations (e.g. tagHg) meant to be compiled with mpbuild?). If not, I'll need to rerun those tests.
  • Build logs are scraped for preprocessor definitions, compiler options, and compiled files.
  • The scraped definitions, flags, and files are sorted, duplicates are removed, and then diffed.
    • This means unique values for each log file are being diffed (i.e. not on a file-by-file basis).
    • Preprocessor definitions and compiler options are defined globally in Makefile_header.mk, so not diffing file-by-file seems good enough (and the implementation of the tests is a lot easier).
  • Test takes about 7 hours to run.
  • Numbers in cells are the number of differences that were found.

dev/12.4.0

Verification report for dev/12.4.0-gd597567 (2019-06-06)
Run directory Compiled? Same KPP? Same definitions? Same flags? Same files?
geosfp_025x03125_CH4_na KPP/Standard 6 14 6
geosfp_025x03125_tropchem_as KPP/Tropchem 6 14 6
geosfp_025x03125_tropchem_na KPP/Tropchem 6 14 6
geosfp_2x25_aciduptake KPP/Standard 6 14 6
geosfp_2x25_aerosol KPP/Standard 6 14 6
geosfp_2x25_CH4 KPP/Standard 6 14 6
geosfp_2x25_CO2 KPP/Standard 6 14 6
geosfp_2x25_complexSOA KPP/Tropchem 6 14 6
geosfp_2x25_complexSOA_SVPOA KPP/SOA_SVPOA 6 14 6
geosfp_2x25_Hg KPP/Standard 6 14 6
geosfp_2x25_marinePOA KPP/Standard 6 14 6
geosfp_2x25_masscons KPP/Standard 7 14 6
geosfp_2x25_POPs KPP/Standard 6 14 6
geosfp_2x25_RRTMG KPP/Tropchem 7 15 5
geosfp_2x25_standard KPP/Standard 6 14 6
geosfp_2x25_tagCH4 KPP/Standard 6 14 6
geosfp_2x25_tagCO KPP/Standard 6 14 6
geosfp_2x25_tagO3 KPP/Standard 6 14 6
geosfp_2x25_TransportTracers KPP/Standard 6 14 6
geosfp_2x25_tropchem KPP/Tropchem 6 14 6
geosfp_4x5_aciduptake KPP/Standard 6 14 6
geosfp_4x5_aerosol KPP/Standard 6 14 6
geosfp_4x5_benchmark KPP/Standard 7 14 6
geosfp_4x5_CH4 KPP/Standard 6 14 6
geosfp_4x5_complexSOA KPP/Tropchem 6 14 6
geosfp_4x5_complexSOA_SVPOA KPP/SOA_SVPOA 6 14 6
geosfp_4x5_gc_timing KPP/Standard 6 14 6
geosfp_4x5_HEMCO KPP/Standard 6 1 29 OK. Extra flag is -r8 in ISORROPIA and tpcore (doesn't affect hemco_standalone)
geosfp_4x5_Hg KPP/Standard 6 14 6
geosfp_4x5_marinePOA KPP/Standard 6 14 6
geosfp_4x5_POPs KPP/Standard 6 14 6
geosfp_4x5_RRTMG KPP/Tropchem 7 15 5
geosfp_4x5_standard KPP/Standard 6 14 6
geosfp_4x5_tagCH4 KPP/Standard 6 14 6
geosfp_4x5_tagCO KPP/Standard 6 14 6
geosfp_4x5_tagHg KPP/Standard 6 14 6
geosfp_4x5_tagO3 KPP/Standard 6 14 6
geosfp_4x5_TOMAS15 KPP/Tropchem 8 14 4
geosfp_4x5_TOMAS40 KPP/Tropchem 8 14 4
geosfp_4x5_TransportTracers KPP/Standard 6 14 6
geosfp_4x5_tropchem KPP/Tropchem 6 14 6
merra2_05x0625_CH4_na KPP/Standard 6 14 6
merra2_05x0625_tropchem_as KPP/Tropchem 6 14 6
merra2_05x0625_tropchem_na KPP/Tropchem 6 14 6
merra2_2x25_aerosol KPP/Standard 6 14 6
merra2_2x25_CH4 KPP/Standard 6 14 6
merra2_2x25_CO2 KPP/Standard 6 14 6
merra2_2x25_complexSOA KPP/Tropchem 6 14 6
merra2_2x25_complexSOA_SVPOA KPP/SOA_SVPOA 6 14 6
merra2_2x25_Hg KPP/Standard 6 14 6
merra2_2x25_masscons KPP/Standard 7 14 6
merra2_2x25_POPs KPP/Standard 6 14 6
merra2_2x25_RRTMG KPP/Tropchem 7 15 5
merra2_2x25_standard KPP/Standard 6 14 6
merra2_2x25_tagCH4 KPP/Standard 6 14 6
merra2_2x25_tagCO KPP/Standard 6 14 6
merra2_2x25_tagO3 KPP/Standard 6 14 6
merra2_2x25_TransportTracers KPP/Standard 6 14 6
merra2_2x25_tropchem KPP/Tropchem 6 14 6
merra2_4x5_aerosol KPP/Standard 6 14 6
merra2_4x5_CH4 KPP/Standard 6 14 6
merra2_4x5_complexSOA KPP/Tropchem 6 14 6
merra2_4x5_complexSOA_SVPOA KPP/SOA_SVPOA 6 14 6
merra2_4x5_Hg KPP/Standard 6 14 6
merra2_4x5_POPs KPP/Standard 6 14 6
merra2_4x5_RRTMG KPP/Tropchem 7 15 5
merra2_4x5_standard KPP/Standard 6 14 6
merra2_4x5_tagCH4 KPP/Standard 6 14 6
merra2_4x5_tagCO KPP/Standard 6 14 6
merra2_4x5_tagHg KPP/Standard 6 14 6
merra2_4x5_tagO3 KPP/Standard 6 14 6
merra2_4x5_TransportTracers KPP/Standard 6 14 6
merra2_4x5_tropchem KPP/Tropchem 6 14 6

Version 12.3.2

Quick 1-day benchmark comparison

I set up four copies of the geosfp_4x5_benchmark run directory with the following modifications.

Modification Value
Simulation start 2016-07-01
Simulation end 2016-07-02
Tran/conv timestep 1800 seconds
Chem/emis timestep 3600 seconds

These run directories were for the four combinations of Intel and GNU compilers, and CMake and Make (ref) build systems. I ran the resulting geos.sp executables. The MD5 hashes of the restart files (output) for 2016-07-02 are shown below.

$  md5sum **/GEOSChem.Restart.20160702_0000z.nc4
5041454f99cee13c01a58c6e95f4e383  gnu/cmake/geosfp_4x5_benchmark/GEOSChem.Restart.20160702_0000z.nc4 
5041454f99cee13c01a58c6e95f4e383  gnu/ref/geosfp_4x5_benchmark/GEOSChem.Restart.20160702_0000z.nc4 
5ecc65c3ddfe932d755fef9071f34a0a  intel/cmake/geosfp_4x5_benchmark/GEOSChem.Restart.20160702_0000z.nc4 
5ecc65c3ddfe932d755fef9071f34a0a  intel/ref/geosfp_4x5_benchmark/GEOSChem.Restart.20160702_0000z.nc4

These MD5 hashes show that geos.sp built with CMake and Make (ref) build systems output bit-wise identical restart files for Intel and GNU compilers. The last commit before these tests were ran was 9d27642.

Scraped build log diffs

Verification report for 12.3.2-92-g9d27642 with Intel 17.0.5 (2019-05-29)
Run directory Compiled? Same KPP? Same definitions? Same flags? Same files?
geosfp_025x03125_CH4_na 6
geosfp_025x03125_tropchem_ch 6
geosfp_025x03125_tropchem_na 6
geosfp_2x25_aciduptake 6
geosfp_2x25_aerosol 6
geosfp_2x25_CH4 6
geosfp_2x25_CO2 6
geosfp_2x25_complexSOA 6
geosfp_2x25_complexSOA_SVPOA 6
geosfp_2x25_Hg 6
geosfp_2x25_marinePOA 6
geosfp_2x25_masscons 6
geosfp_2x25_POPs 6
geosfp_2x25_RRTMG 5
geosfp_2x25_standard 6
geosfp_2x25_tagCH4 6
geosfp_2x25_tagCO 6
geosfp_2x25_tagO3 6
geosfp_2x25_TransportTracers 6
geosfp_2x25_tropchem 6
geosfp_4x5_aciduptake 6
geosfp_4x5_aerosol 6
geosfp_4x5_benchmark 6
geosfp_4x5_CH4 6
geosfp_4x5_complexSOA 6
geosfp_4x5_complexSOA_SVPOA 6
geosfp_4x5_gc_timing 6
geosfp_4x5_HEMCO Should be fixed in 477046c.
geosfp_4x5_Hg 6
geosfp_4x5_marinePOA 6
geosfp_4x5_POPs 6
geosfp_4x5_RRTMG 5
geosfp_4x5_standard 6
geosfp_4x5_tagCH4 6
geosfp_4x5_tagCO 6
geosfp_4x5_tagHg 6
geosfp_4x5_tagO3 6
geosfp_4x5_TOMAS15 See geoschem/geos-chem issue #38. Previous tests @12.1.1 passed.
geosfp_4x5_TOMAS40
geosfp_4x5_TransportTracers 6
geosfp_4x5_tropchem 6
merra2_05x0625_CH4_na 6
merra2_05x0625_tropchem_as 6
merra2_05x0625_tropchem_na 6
merra2_2x25_aerosol 6
merra2_2x25_CH4 6
merra2_2x25_CO2 6
merra2_2x25_complexSOA 6
merra2_2x25_complexSOA_SVPOA 6
merra2_2x25_Hg 6
merra2_2x25_masscons 6
merra2_2x25_POPs 6
merra2_2x25_RRTMG 5
merra2_2x25_standard 6
merra2_2x25_tagCH4 6
merra2_2x25_tagCO 6
merra2_2x25_tagO3 6
merra2_2x25_TransportTracers 6
merra2_2x25_tropchem 6
merra2_4x5_aerosol 6
merra2_4x5_CH4 6
merra2_4x5_complexSOA 6
merra2_4x5_complexSOA_SVPOA 6
merra2_4x5_Hg 6
merra2_4x5_POPs 6
merra2_4x5_RRTMG 5
merra2_4x5_standard 6
merra2_4x5_tagCH4 6
merra2_4x5_tagCO 6
merra2_4x5_tagHg 6
merra2_4x5_tagO3 6
merra2_4x5_TransportTracers 6
merra2_4x5_tropchem 6
Verification report for 12.3.2-92-g9d27642 with GNU 5.4 (2019-05-29)
Run directory Compiled? Same KPP? Same definitions? Same flags? Same files?
geosfp_025x03125_CH4_na 6
geosfp_025x03125_tropchem_ch 6
geosfp_025x03125_tropchem_na 6
geosfp_2x25_aciduptake 6
geosfp_2x25_aerosol 6
geosfp_2x25_CH4 6
geosfp_2x25_CO2 6
geosfp_2x25_complexSOA 6
geosfp_2x25_complexSOA_SVPOA 6
geosfp_2x25_Hg 6
geosfp_2x25_marinePOA 6
geosfp_2x25_masscons 6
geosfp_2x25_POPs 6
geosfp_2x25_RRTMG RRTMG doesn't compile with GNU compilers. See here.
geosfp_2x25_standard 6
geosfp_2x25_tagCH4 6
geosfp_2x25_tagCO 6
geosfp_2x25_tagO3 6
geosfp_2x25_TransportTracers 6
geosfp_2x25_tropchem 6
geosfp_4x5_aciduptake 6
geosfp_4x5_aerosol 6
geosfp_4x5_benchmark 6
geosfp_4x5_CH4 6
geosfp_4x5_complexSOA 6
geosfp_4x5_complexSOA_SVPOA 6
geosfp_4x5_gc_timing 6
geosfp_4x5_HEMCO Should be fixed in 477046c.
geosfp_4x5_Hg 6
geosfp_4x5_marinePOA 6
geosfp_4x5_POPs 6
geosfp_4x5_RRTMG RRTMG doesn't compile with GNU compilers. See here.
geosfp_4x5_standard 6
geosfp_4x5_tagCH4 6
geosfp_4x5_tagCO 6
geosfp_4x5_tagHg 6
geosfp_4x5_tagO3 6
geosfp_4x5_TOMAS15 See geoschem/geos-chem issue #38. Previous tests @12.1.1 passed.
geosfp_4x5_TOMAS40
geosfp_4x5_TransportTracers 6
geosfp_4x5_tropchem 6
merra2_05x0625_CH4_na 6
merra2_05x0625_tropchem_as 6
merra2_05x0625_tropchem_na 6
merra2_2x25_aerosol 6
merra2_2x25_CH4 6
merra2_2x25_CO2 6
merra2_2x25_complexSOA 6
merra2_2x25_complexSOA_SVPOA 6
merra2_2x25_Hg 6
merra2_2x25_masscons 6
merra2_2x25_POPs 6
merra2_2x25_RRTMG RRTMG doesn't compile with GNU compilers. See here.
merra2_2x25_standard 6
merra2_2x25_tagCH4 6
merra2_2x25_tagCO 6
merra2_2x25_tagO3 6
merra2_2x25_TransportTracers 6
merra2_2x25_tropchem 6
merra2_4x5_aerosol 6
merra2_4x5_CH4 6
merra2_4x5_complexSOA 6
merra2_4x5_complexSOA_SVPOA 6
merra2_4x5_Hg 6
merra2_4x5_POPs 6
merra2_4x5_RRTMG RRTMG doesn't compile with GNU compilers. See here.
merra2_4x5_standard 6
merra2_4x5_tagCH4 6
merra2_4x5_tagCO 6
merra2_4x5_tagHg 6
merra2_4x5_tagO3 6
merra2_4x5_TransportTracers 6
merra2_4x5_tropchem 6



Histroy

Version 12.3.2

Build comparison

Flags matching the following regexes were omitted:

  • -I -L: link and include directories are handled differently between the build systems and problems would result in compiler/linker errors so these can safely be ignored.
  • -l: CMake links with absolute paths so linker options look different. Any problems would result in linker errors or immediate runtime errors to these can safely be ignored.
  • -q*openmp: Make-based BS uses -openmp whereas CMake is using -qopenmp (based on FindOpenMP.cmake). These can be verified manually.
  • -Wl,: Again, igoring linker commands.
  • -free-form: gfortran automatically uses free form for F90 files and all GC's free form files are .F90. Errors would result in compiler errors.

Intel 17

Verification report for 12.3.2-90-gdb36590.dirty (2019-05-28)
Run directory Compiled? Same KPP? Same definitions? Same flags? Same files?
geosfp_025x03125_CH4_na 6
geosfp_025x03125_tropchem_ch 6
geosfp_025x03125_tropchem_na 6
geosfp_2x25_aciduptake 6
geosfp_2x25_aerosol 6
geosfp_2x25_CH4 6
geosfp_2x25_CO2 6
geosfp_2x25_complexSOA 6
geosfp_2x25_complexSOA_SVPOA 6
geosfp_2x25_Hg 6
geosfp_2x25_marinePOA 6
geosfp_2x25_masscons 6
geosfp_2x25_POPs 6
geosfp_2x25_RRTMG 5
geosfp_2x25_standard 6
geosfp_2x25_tagCH4 6
geosfp_2x25_tagCO 6
geosfp_2x25_tagO3 6
geosfp_2x25_TransportTracers 6
geosfp_2x25_tropchem 6
geosfp_4x5_aciduptake 6
geosfp_4x5_aerosol 6
geosfp_4x5_benchmark 6
geosfp_4x5_CH4 6
geosfp_4x5_complexSOA 6
geosfp_4x5_complexSOA_SVPOA 6
geosfp_4x5_gc_timing 6
geosfp_4x5_HEMCO
geosfp_4x5_Hg 6
geosfp_4x5_marinePOA 6
geosfp_4x5_POPs 6
geosfp_4x5_RRTMG 5
geosfp_4x5_standard 6
geosfp_4x5_tagCH4 6
geosfp_4x5_tagCO 6
geosfp_4x5_tagHg 6
geosfp_4x5_tagO3 6
geosfp_4x5_TOMAS15
geosfp_4x5_TOMAS40
geosfp_4x5_TransportTracers 6
geosfp_4x5_tropchem 6
merra2_05x0625_CH4_na 6
merra2_05x0625_tropchem_as 6
merra2_05x0625_tropchem_na 6
merra2_2x25_aerosol 6
merra2_2x25_CH4 6
merra2_2x25_CO2 6
merra2_2x25_complexSOA 6
merra2_2x25_complexSOA_SVPOA 6
merra2_2x25_Hg 6
merra2_2x25_masscons 6
merra2_2x25_POPs 6
merra2_2x25_RRTMG 5
merra2_2x25_standard 6
merra2_2x25_tagCH4 6
merra2_2x25_tagCO 6
merra2_2x25_tagO3 6
merra2_2x25_TransportTracers 6
merra2_2x25_tropchem 6
merra2_4x5_aerosol 6
merra2_4x5_CH4 6
merra2_4x5_complexSOA 6
merra2_4x5_complexSOA_SVPOA 6
merra2_4x5_Hg 6
merra2_4x5_POPs 6
merra2_4x5_RRTMG 5
merra2_4x5_standard 6
merra2_4x5_tagCH4 6
merra2_4x5_tagCO 6
merra2_4x5_tagHg 6
merra2_4x5_tagO3 6
merra2_4x5_TransportTracers 6
merra2_4x5_tropchem 6


  • The dirty tag is because I removed the NC_DIAG definition. Included in next commit which is 5e1b981


GNU 5.4

Verification report for 12.3.2-90-gdb36590.dirty (2019-05-28)
Run directory Compiled? Same KPP? Same definitions? Same flags? Same files?
geosfp_025x03125_CH4_na 6
geosfp_025x03125_tropchem_ch 6
geosfp_025x03125_tropchem_na 6
geosfp_2x25_aciduptake 6
geosfp_2x25_aerosol 6
geosfp_2x25_CH4 6
geosfp_2x25_CO2 6
geosfp_2x25_complexSOA 6
geosfp_2x25_complexSOA_SVPOA 6
geosfp_2x25_Hg 6
geosfp_2x25_marinePOA 6
geosfp_2x25_masscons 6
geosfp_2x25_POPs 6
geosfp_2x25_RRTMG
geosfp_2x25_standard 6
geosfp_2x25_tagCH4 6
geosfp_2x25_tagCO 6
geosfp_2x25_tagO3 6
geosfp_2x25_TransportTracers 6
geosfp_2x25_tropchem 6
geosfp_4x5_aciduptake 6
geosfp_4x5_aerosol 6
geosfp_4x5_benchmark 6
geosfp_4x5_CH4 6
geosfp_4x5_complexSOA 6
geosfp_4x5_complexSOA_SVPOA 6
geosfp_4x5_gc_timing 6
geosfp_4x5_HEMCO
geosfp_4x5_Hg 6
geosfp_4x5_marinePOA 6
geosfp_4x5_POPs 6
geosfp_4x5_RRTMG
geosfp_4x5_standard 6
geosfp_4x5_tagCH4 6
geosfp_4x5_tagCO 6
geosfp_4x5_tagHg 6
geosfp_4x5_tagO3 6
geosfp_4x5_TOMAS15
geosfp_4x5_TOMAS40
geosfp_4x5_TransportTracers 6
geosfp_4x5_tropchem 6
merra2_05x0625_CH4_na 6
merra2_05x0625_tropchem_as 6
merra2_05x0625_tropchem_na 6
merra2_2x25_aerosol 6
merra2_2x25_CH4 6
merra2_2x25_CO2 6
merra2_2x25_complexSOA 6
merra2_2x25_complexSOA_SVPOA 6
merra2_2x25_Hg 6
merra2_2x25_masscons 6
merra2_2x25_POPs 6
merra2_2x25_RRTMG
merra2_2x25_standard 6
merra2_2x25_tagCH4 6
merra2_2x25_tagCO 6
merra2_2x25_tagO3 6
merra2_2x25_TransportTracers 6
merra2_2x25_tropchem 6
merra2_4x5_aerosol 6
merra2_4x5_CH4 6
merra2_4x5_complexSOA 6
merra2_4x5_complexSOA_SVPOA 6
merra2_4x5_Hg 6
merra2_4x5_POPs 6
merra2_4x5_RRTMG
merra2_4x5_standard 6
merra2_4x5_tagCH4 6
merra2_4x5_tagCO 6
merra2_4x5_tagHg 6
merra2_4x5_tagO3 6
merra2_4x5_TransportTracers 6
merra2_4x5_tropchem 6
  • The dirty tag is because I removed the NC_DIAG definition. Included in next commit which is 5e1b981

Version 12.1

Intel 17

Verification report for 20fbf09 (2019-05-16)
Run directory Compiled? Same KPP? Same definitions? Same flags? Same files?
geosfp_025x03125_CH4_na 5
geosfp_025x03125_tropchem_ch 5
geosfp_025x03125_tropchem_na 5
geosfp_2x25_aciduptake 5
geosfp_2x25_aerosol 5
geosfp_2x25_CH4 5
geosfp_2x25_CO2 5
geosfp_2x25_complexSOA 5
geosfp_2x25_complexSOA_SVPOA 5
geosfp_2x25_Hg 5
geosfp_2x25_marinePOA 5
geosfp_2x25_masscons 5
geosfp_2x25_POPs 5
geosfp_2x25_RRTMG 4
geosfp_2x25_standard 5
geosfp_2x25_tagCH4 5
geosfp_2x25_tagCO 5
geosfp_2x25_tagO3 5
geosfp_2x25_TransportTracers 5
geosfp_2x25_tropchem 5
geosfp_4x5_aciduptake 5
geosfp_4x5_aerosol 5
geosfp_4x5_benchmark 5
geosfp_4x5_CH4 5
geosfp_4x5_complexSOA 5
geosfp_4x5_complexSOA_SVPOA 5
geosfp_4x5_gc_timing 5
geosfp_4x5_HEMCO Todo: Add hemco_standalone target.
geosfp_4x5_Hg 5
geosfp_4x5_marinePOA 5
geosfp_4x5_POPs 5
geosfp_4x5_RRTMG 4
geosfp_4x5_standard 5
geosfp_4x5_tagCH4 5
geosfp_4x5_tagCO 5
geosfp_4x5_tagHg 5
geosfp_4x5_tagO3 5
geosfp_4x5_TOMAS15 1 3 I believe this definitions difference is expected. Make's build is defining NC_HAS_COMPRESSION despite NC_DIAG=n.
geosfp_4x5_TOMAS40 1 3
geosfp_4x5_TransportTracers 5
geosfp_4x5_tropchem 5
merra2_05x0625_CH4_na 5
merra2_05x0625_tropchem_as 5
merra2_05x0625_tropchem_na 5
merra2_2x25_aerosol 5
merra2_2x25_CH4 5
merra2_2x25_CO2 5
merra2_2x25_complexSOA 5
merra2_2x25_complexSOA_SVPOA 5
merra2_2x25_Hg 5
merra2_2x25_masscons 5
merra2_2x25_POPs 5
merra2_2x25_RRTMG 4
merra2_2x25_standard 5
merra2_2x25_tagCH4 5
merra2_2x25_tagCO 5
merra2_2x25_tagO3 5
merra2_2x25_TransportTracers 5
merra2_2x25_tropchem 5
merra2_4x5_aerosol 5
merra2_4x5_CH4 5
merra2_4x5_complexSOA 5
merra2_4x5_complexSOA_SVPOA 5
merra2_4x5_Hg 5
merra2_4x5_POPs 5
merra2_4x5_RRTMG 4
merra2_4x5_standard 5
merra2_4x5_tagCH4 5
merra2_4x5_tagCO 5
merra2_4x5_tagHg 5
merra2_4x5_tagO3 5
merra2_4x5_TransportTracers 5
merra2_4x5_tropchem 5


Verification report for 9c40efe (2019-05-15)
Run directory Compiled? Same KPP? Same definitions? Same flags? Same files?
geosfp_025x03125_CH4_na 5
geosfp_025x03125_tropchem_ch 5
geosfp_025x03125_tropchem_na 5
geosfp_2x25_aciduptake 5
geosfp_2x25_aerosol 5
geosfp_2x25_CH4 5
geosfp_2x25_CO2 5
geosfp_2x25_complexSOA 1 5 definitions fixed in 6445d22
geosfp_2x25_complexSOA_SVPOA didn't run because accidental regex match with "complexSOA" run directories; fixed for tonights test
geosfp_2x25_Hg 1 19 definitions fixed in 20fbf09: turned off GTMM coupling in CMake build by default
geosfp_2x25_marinePOA 5
geosfp_2x25_masscons 5
geosfp_2x25_POPs 5
geosfp_2x25_RRTMG 4
geosfp_2x25_standard 5
geosfp_2x25_tagCH4 5
geosfp_2x25_tagCO 5
geosfp_2x25_tagO3 5
geosfp_2x25_TransportTracers 5
geosfp_2x25_tropchem 5
geosfp_4x5_aciduptake 5
geosfp_4x5_aerosol 5
geosfp_4x5_benchmark 5
geosfp_4x5_CH4 5
geosfp_4x5_complexSOA 1 5 definitions fixed in 6445d22
geosfp_4x5_complexSOA_SVPOA didn't run because accidental regex match with "complexSOA" run directories; fixed for tonights test
geosfp_4x5_gc_timing 5
geosfp_4x5_HEMCO
geosfp_4x5_Hg 1 19 definitions fixed in 20fbf09: turned off GTMM coupling in CMake build by default
geosfp_4x5_marinePOA 5
geosfp_4x5_POPs 5
geosfp_4x5_RRTMG 4
geosfp_4x5_standard 5
geosfp_4x5_tagCH4 5
geosfp_4x5_tagCO 5
geosfp_4x5_tagHg fixed in 6ca7376
geosfp_4x5_tagO3 5
geosfp_4x5_TOMAS15 2 3 definitions fixed in 6445d22; Make's build is defining NC_HAS_COMPRESSION even though NC_DIAG=n, so 1 difference is expected; had to run Make's build manually to pass TOMASXX=yes NC_DIAG=n
geosfp_4x5_TOMAS40 2 3
geosfp_4x5_TransportTracers 5
geosfp_4x5_tropchem 5
merra2_05x0625_CH4_na 5
merra2_05x0625_tropchem_as 5
merra2_05x0625_tropchem_na 5
merra2_2x25_aerosol 5
merra2_2x25_CH4 5
merra2_2x25_CO2 5
merra2_2x25_complexSOA 1 5 definitions fixed in 6445d22
merra2_2x25_complexSOA_SVPOA didn't run because accidental regex match with "complexSOA" run directories; fixed for tonights test
merra2_2x25_Hg 1 19 definitions fixed in 20fbf09: turned off GTMM coupling in CMake build by default
merra2_2x25_masscons 5
merra2_2x25_POPs 5
merra2_2x25_RRTMG 4
merra2_2x25_standard 5
merra2_2x25_tagCH4 5
merra2_2x25_tagCO 5
merra2_2x25_tagO3 5
merra2_2x25_TransportTracers 5
merra2_2x25_tropchem 5
merra2_4x5_aerosol 5
merra2_4x5_CH4 5
merra2_4x5_complexSOA 1 5 definitions fixed in 6445d22
merra2_4x5_complexSOA_SVPOA didn't run because accidental regex match with "complexSOA" run directories; fixed for tonights test
merra2_4x5_Hg 1 19 definitions fixed in 20fbf09: turned off GTMM coupling in CMake build by default
merra2_4x5_POPs 5
merra2_4x5_RRTMG 4
merra2_4x5_standard 5
merra2_4x5_tagCH4 5
merra2_4x5_tagCO 5
merra2_4x5_tagHg fixed in 6ca7376
merra2_4x5_tagO3 5
merra2_4x5_TransportTracers 5
merra2_4x5_tropchem 5
  • Hg, complexSOA, and TOMAS simulations have a wrong definition
    • Hg: The Make BS doesn't build GTMM with make mpbuild, so GTMM_Hg doesn't get defined in the Make BS whereas it is in the CMake BS.
    • complexSOA: should be on a reduced grid, but CMake isn't defining GRIDREDUCED
    • TOMAS: CMake BS is missing GRIDREDUCED and NC_HAS_COMPRESSION (even though NC_DIAG isn't defined)
  • TOMASs with Make BS has to be built mangually because NC_DIAG=n isn't set in the test script.


GTMM_Hg isn't defined in the Make BS. This could just


Verification report for cdc0ef0 (2019-05-14)
Run directory Compiled? Same KPP? Same definitions? Same flags? Same files?
geosfp_025x03125_CH4_na 1 2 5
geosfp_025x03125_tropchem_ch 2 5
geosfp_025x03125_tropchem_na 2 5
geosfp_2x25_aciduptake 2 5
geosfp_2x25_aerosol 1 2 5
geosfp_2x25_CH4 1 2 5
geosfp_2x25_CO2 1 2 5
geosfp_2x25_complexSOA 2 5
geosfp_2x25_complexSOA_SVPOA
geosfp_2x25_Hg 1 2 19
geosfp_2x25_marinePOA 2 5
geosfp_2x25_masscons 1 2 5
geosfp_2x25_POPs 1 2 5
geosfp_2x25_RRTMG 2 4
geosfp_2x25_standard 2 5
geosfp_2x25_tagCH4 1 2 5
geosfp_2x25_tagCO 1 2 5
geosfp_2x25_tagO3 1 2 5
geosfp_2x25_TransportTracers 2 5
geosfp_2x25_tropchem 2 5
geosfp_4x5_aciduptake 2 5
geosfp_4x5_aerosol 1 2 5
geosfp_4x5_benchmark 2 5
geosfp_4x5_CH4 1 2 5
geosfp_4x5_complexSOA 2 5
geosfp_4x5_complexSOA_SVPOA
geosfp_4x5_gc_timing 2 5
geosfp_4x5_HEMCO
geosfp_4x5_Hg 1 2 19
geosfp_4x5_marinePOA 2 5
geosfp_4x5_POPs 1 2 5
geosfp_4x5_RRTMG 2 4
geosfp_4x5_standard 2 5
geosfp_4x5_tagCH4 1 2 5
geosfp_4x5_tagCO 1 2 5
geosfp_4x5_tagHg
geosfp_4x5_tagO3 1 2 5
geosfp_4x5_TOMAS15 1 1 1
geosfp_4x5_TOMAS40 1 1 1
geosfp_4x5_TransportTracers 2 5
geosfp_4x5_tropchem 2 5
merra2_05x0625_CH4_na 1 2 5
merra2_05x0625_tropchem_as 2 5
merra2_05x0625_tropchem_na 2 5
merra2_2x25_aerosol 1 2 5
merra2_2x25_CH4 1 2 5
merra2_2x25_CO2 1 2 5
merra2_2x25_complexSOA 2 5
merra2_2x25_complexSOA_SVPOA
merra2_2x25_Hg 1 2 19
merra2_2x25_masscons 1 2 5
merra2_2x25_POPs 1 2 5
merra2_2x25_RRTMG 2 4
merra2_2x25_standard 2 5
merra2_2x25_tagCH4 1 2 5
merra2_2x25_tagCO 1 2 5
merra2_2x25_tagO3 1 2 5
merra2_2x25_TransportTracers 2 5
merra2_2x25_tropchem 2 5
merra2_4x5_aerosol 1 2 5
merra2_4x5_CH4 1 2 5
merra2_4x5_complexSOA 2 5
merra2_4x5_complexSOA_SVPOA
merra2_4x5_Hg 1 2 19
merra2_4x5_POPs 1 2 5
merra2_4x5_RRTMG 2 4
merra2_4x5_standard 2 5
merra2_4x5_tagCH4 1 2 5
merra2_4x5_tagCO 1 2 5
merra2_4x5_tagHg
merra2_4x5_tagO3 1 2 5
merra2_4x5_TransportTracers 2 5
merra2_4x5_tropchem 2 5

Notes for cdc0ef0:

  • Mismatched definitions appear to be GRIDREDUCED not getting defined. Addressed in 4eaf7c7
  • Mismatched flags appear to be -fPIC -O3 getting set when they shouldn't be. Addressed in 757c746
  • TOMASs didn't build with Make BS because NC_DIAG=n isn't set in the test script.

GNU 5.4

Verification report for 369547e (2019-05-17)
Run directory Compiled? Same KPP? Same definitions? Same flags? Same files?
geosfp_025x03125_CH4_na 1 5
geosfp_025x03125_tropchem_ch 1 5
geosfp_025x03125_tropchem_na 1 5
geosfp_2x25_aciduptake 1 5
geosfp_2x25_aerosol 1 5
geosfp_2x25_CH4 1 5
geosfp_2x25_CO2 1 5
geosfp_2x25_complexSOA 1 5
geosfp_2x25_complexSOA_SVPOA 1 5
geosfp_2x25_Hg 1 5
geosfp_2x25_marinePOA 1 5
geosfp_2x25_masscons 1 5
geosfp_2x25_POPs 1 5
geosfp_2x25_RRTMG
geosfp_2x25_standard 1 5
geosfp_2x25_tagCH4 1 5
geosfp_2x25_tagCO 1 5
geosfp_2x25_tagO3 1 5
geosfp_2x25_TransportTracers 1 5
geosfp_2x25_tropchem 1 5
geosfp_4x5_aciduptake 1 5
geosfp_4x5_aerosol 1 5
geosfp_4x5_benchmark 1 5
geosfp_4x5_CH4 1 5
geosfp_4x5_complexSOA 1 5
geosfp_4x5_complexSOA_SVPOA 1 5
geosfp_4x5_gc_timing 1 5
geosfp_4x5_HEMCO
geosfp_4x5_Hg 1 5
geosfp_4x5_marinePOA 1 5
geosfp_4x5_POPs 1 5
geosfp_4x5_RRTMG
geosfp_4x5_standard 1 5
geosfp_4x5_tagCH4 1 5
geosfp_4x5_tagCO 1 5
geosfp_4x5_tagHg 1 5
geosfp_4x5_tagO3 1 5
geosfp_4x5_TOMAS15
geosfp_4x5_TOMAS40
geosfp_4x5_TransportTracers 1 5
geosfp_4x5_tropchem 1 5
merra2_05x0625_CH4_na 1 5
merra2_05x0625_tropchem_as 1 5
merra2_05x0625_tropchem_na 1 5
merra2_2x25_aerosol 1 5
merra2_2x25_CH4 1 5
merra2_2x25_CO2 1 5
merra2_2x25_complexSOA 1 5
merra2_2x25_complexSOA_SVPOA 1 5
merra2_2x25_Hg 1 5
merra2_2x25_masscons 1 5
merra2_2x25_POPs 1 5
merra2_2x25_RRTMG
merra2_2x25_standard 1 5
merra2_2x25_tagCH4 1 5
merra2_2x25_tagCO 1 5
merra2_2x25_tagO3 1 5
merra2_2x25_TransportTracers 1 5
merra2_2x25_tropchem 1 5
merra2_4x5_aerosol 1 5
merra2_4x5_CH4 1 5
merra2_4x5_complexSOA 1 5
merra2_4x5_complexSOA_SVPOA 1 5
merra2_4x5_Hg 1 5
merra2_4x5_POPs 1 5
merra2_4x5_RRTMG
merra2_4x5_standard 1 5
merra2_4x5_tagCH4 1 5
merra2_4x5_tagCO 1 5
merra2_4x5_tagHg 1 5
merra2_4x5_tagO3 1 5
merra2_4x5_TransportTracers 1 5
merra2_4x5_tropchem 1 5
  • Compiler option difference is Make's BS is defining -ffree-form, whereas CMake's BS is relying on gfortran automatically using this option for ".F90" source files.

Compiler error while compiling RRTMG mod in GeosCore:

/misc/data10/bindle/test4/GC/GeosCore/rrtmg_rad_transfer_mod.F:803:11:

        IF (INTROP(I,J,L).EQ..TRUE.) THEN
           1
Error: Logicals at (1) must be compared with .eqv. instead of .eq.

Compiler error while compiling HCOX:

hcox_tomas_jeagle_mod.F90:142:14:

     IF ( .NOT. ExtState%TOMAS_Jeagle ) RETURN
              1
Error: Operand of .not. operator at (1) is INTEGER(4)

Start the tests:

nohup ./build_loop.sh &> results/job.log & 
disown