Compiling GCHP

From Geos-chem
Revision as of 18:38, 8 August 2018 by Lizzie Lundgren (Talk | contribs)

Jump to: navigation, search

Previous | Next | Getting Started with GCHP

  1. Hardware and Software Requirements
  2. Downloading Source Code
  3. Obtaining a Run Directory
  4. Setting Up the GCHP Environment
  5. Compiling
  6. Basic Example Run
  7. Run Configuration Files
  8. Advanced Run Examples
  9. Output Data
  10. Developing GCHP


Overview

You are ready to compile GCHP after you have obtained the GCHP source code, have acquired and set up your run directory, and have loaded your environment. First time compilation can take a while (30 minutes or more) because it requires building large modules. GCHP is made up of several different modules, all communicating via NASA's Modeling, Analysis and Prediciton Layer (MAPL). MAPL is built on the Earth System Modeling Framework (ESMF), and both systems are highly flexible but highly complex, with significant compilation times. Fortunately, MAPL, ESMF, and a third external component for advection (FV3 dycore), do not need to be recompiled unless parts of their code have been modified. Subsequent compile times will therefore be significantly shorter, on the same order of magnitude as compiling GEOS-Chem Classic.

This section provides instructions on how to perform a full ("clean") compilation, including MAPL, ESMF, and FV3 dycore, as well as a partial ("standard") compilation when you want to keep your existing MAPL, ESMF, and advection builds as is. The "clean" compilation MUST be done the first time you use GCHP since it cleans and compiles all modules. The "standard" compilation may be used for subsequent recompilation after updating GEOS-Chem code. Be careful not to wipe out the entire GCHP build when cleaning unless you intentionally want to do so!

Compiling GCHP produces a GCHP executable called geos that is stored in the bin/ sub-directory of your GEOS-Chem source code directory. When you compile from the run directory using the Makefile (as described on this page) a copy will also be stored in your run directory for easy use. Beware that once you have an executable you must not rename your run directory or your source code directory or the executable will no longer work. However, you can copy the executable elsewhere for use. If you do this then it is also a good idea to copy lastbuild and compile.log to your new run directory to keep track of where your executable came from.

Initial Compilation

To compile GCHP from scratch, type the following within the run directory:

make compile_clean

This command will execute ./build.sh compile_clean (described in the advanced topics section below) and send all log output to both the terminal window and to a log file called compile.log. Git information for both source code repositories at compile time are recorded in a separate file called lastbuild for future reference. Upon successful compilation, the following empty files are created in your GCHP source code directories:

  • GCHP/ESMF/esmf.install
  • GCHP/Shared/mapl.install
  • GCHP/FVdycoreCubed_GridComp/fvdycore.install

These files indicate successful compilation per module and their presence prompts the GCHP source code Makefile to skip compiling each component during future builds so as to save time. The *.install files are deleted when you clean MAPL, ESMF, and FVdycore source code such as when running make compile_clean or make superclean. If you are ever uncertain on whether your ESMF, MAPL, and FV3 dycore builds have been wiped out, check if these files exist.

Partial Compilation

For subsequent GCHP compilations where you only need to rebuild the GEOS-Chem code, you can therefore type the following within the run directory for a quick compilation:

make compile_standard

If you updated MAPL code only, you can still do a partial compilation to avoid recompiling ESMF. Do this with:

make compile_mapl

Additional Makefile Options

The GCHP run directory Makefile contains several options for cleaning and compiling GCHP using make, as well as options for other functionality. Options are printed out by typing the following at the command line:

make help

This will output something like this to your screen.

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%%%    GCHP Run Directory Makefile Options    %%%
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Usage: make TARGET [OPTIONAL-FLAGS]

Default threads for compiling is hard-coded as 6

----------------------------------------------------------
TARGET may be one of the following:
----------------------------------------------------------

To print settings to screen:
  printbuildinfo   Print out git info for GC and GCHP repos
  help             If you got this far, you know this one

To clean source code:
  clean_all        Clean all source code using build.sh
  clean_gc         Clean GEOS-Chem classic only using build.sh
  clean_mapl       Clean MAPL and FVdyCore using build.sh
  clean_nuclear    Clean everything using build.sh
  superclean       Remove exe, logs, output, and clean ALL code

To compile:
  compile_clean     Include everything
  compile_standard  Skip MAPL, FvDyCore, and ESMF; debug off
  compile_debug     Skip MAPL, FvDyCore, and ESMF; debug on
  compile_mapl      Skip ESMF

To remove run directory files:
  cleanup_output   Remove output data, log files, and executable
  cleanup_data     Remove output data
  cleanup_logs     Remove log files
  cleanup_exe      Remove executable

To run interactively:
  run_interactive

To clean, build, and run interactively:
  gchp_clean       Include everything
  gchp_standard    Skip MAPL, FvDyCore, and ESMF; debug off
  gchp_debug       Skip MAPL, FvDyCore, and ESMF; debug on
  gchp_mapl        Skip ESMF

Some very useful options include the following:

  • make printbuildinfo: prints git information such as branch and last commit for the GEOS-Chem and GCHP repositories your CodeDir symbolic link points to.
  • make cleanup_output: cleans your run directory to remove all traces of your last run, including log and output files, but will not remove your executable or last build information.

Checking for Successful Compilation

If GCHP compiled without issues, you should see:

###################################
### GCHP compiled successfully! ###
###################################

As described above, compiler output, including warnings and error messages, will be stored in a log file called compile.log. If you inspect your log file or watch the compile output real-time, you will notice that there are many warnings and error messages. Many of these errors are due to leftover components in MAPL which are no longer used but which still try to compile. As long as the compilation continues, these errors can almost always be ignored and they will have no effect on the operation of GCHP.

You may copy your executable to other run directories but please be aware that you can only successfully run it if you maintain your original directory where you compiled. This is because the GCHP executable will include paths to ESMF libraries that incorporate the CodeDir symbolic link in the run directory path. If you encounter an error while running GCHP that an ESMF shared library is missing, check that the exact path in the error message is still valid.

If you run into problems and contact us for help, please be sure to send us your log file.

Advanced Compilation Topics

The Makefile in the run directory invokes a bash shell script called build.sh that includes changing into the source code directory and invoking source code Makefiles. This section details the capabilities of build.sh and source code make options for advanced users who wish to better understand how GCHP works. You may skip this section and return to it at a later date when you want to dive deeper into how GCHP is compiled.

build.sh

Whenever build.sh is executed it checks if you have set up your environment by querying if the environment variable MPI_ROOT is defined. If it is not, you will be prompted to choose a .bashrc, source it, and export the BASHRC variable. Instructions on how to do this are printed to the terminal and are also described on the Setting up the GCHP environment wiki page. Note that the script gives you the option to choose your .bashrc file from a list of files it finds in your run directory. If none of them are appropriate, for example if you are not working on the Harvard Odyssey or Compute Canada Glooscap compute clusters, then you will need to create a .bashrc file using these sample files as an example.

If you have multiple terminal windows and cannot remember if you have loaded your environment, a quick check is execute the help option of build.sh.

./build.sh help

Clean and compile options of build.sh will also print for your reference. For example, when working on the Harvard Odyssey compute cluster you output will look as follows:

MPI_ROOT is to /n/sw/fasrcsw/apps/Comp/intel/15.0.0-fasrc01/mvapich2/2.2-fasrc01
Script name:
  build.sh
Arguments:
  Accepts single argument indicating clean and/or compile settings.
  Currently implemented arguments include:
     clean_gc         - classic only
     clean_nuclear    - GCHP, ESMF, MAPL, FVdycore (be careful!)
     clean_all        - classic, GCHP, ESMF, MAPL, FVdycore (be careful!)
     clean_mapl       - mapl and fvdycore only
     compile_debug    - turns on debug flags, no cleaning
     compile_standard - no cleaning
     compile_mapl     - includes fvdycore
     compile_clean    - cleans and compiles everything (be careful!)
Example usage:
  ./build.sh compile_standard

The default compile commands for compile_clean, compile_standard, and compile_mapl are as follows:

   make -j${SLURM_NTASKS} NC_DIAG=y   CHEM=tropchem EXTERNAL_GRID=y   \
                          DEBUG=y     DEVEL=y       TRACEBACK=y       \
                          MET=geos-fp GRID=4x5      NO_REDUCED=y      \
                          UCX=n       hpc

The compile command for compile_debug is the same as above with BOUNDS=yes and FPEX=yes added for debugging purposes.

If you have only changed GEOS-Chem code and have not touched the GCHP structural files (e.g. MAPL), then you can recompile with ./build.sh compile_standard to save time. Running this will recompile only changed files (excluding MAPL, FVdycore, and ESMF). To run it and also output stderr and stdout to log file compile.log, run the following make command:

make compile_standard

If you find that you need to fully recompile GEOS-Chem, however, you can still avoid recompiling the MAPL framework code (as well as components such as the FV3 dycore). You can do this simply by calling make realclean within your top-level source code directory prior to running make compile_standard. This will perform a clean build of the core GEOS-Chem components without necessitating a full recompilation of the more static components of GCHP.

Source Code Makefiles

At its core, building GCHP simply requires the inclusion of an additional option, HPC=yes, in the source code directory make command beyond what is typically included in GEOS-Chem compilation. Appending HPC=yes causes the make procedure to look for the contents of GCHP, and bypasses much of the serial GEOS-Chem code. For example, base GEOS-Chem classic source code is cleaned using the following command:

 make HPC=yes realclean 

After the process completes, the GCHP/ subdirectory is cleaned by changing into that directory and invoking:

 make EXTERNAL_GRID=yes DEVEL=yes the_nuclear_option 

Calling make with the_nuclear_option defined in GCHP/Makefile removes all compiled code in the directory. As implied by the name, this is not to be used lightly! The nuclear option is an alias, calling three commands in turn: wipeout_esmf, wipeout_mapl, and wipeout_fvdycore. Each of these components can take a long time (30 minutes or more) to compile, and all are structural components that are not affected by small changes in GEOS-Chem.

If you are interested in seeing the core make commands used for all options provided in the run directory Makefile, see the commands defined in build.sh. Note that calling make always invokes the Makefile in the directory you are in and changing into the GEOS-Chem and GCHP source code directories occurs within build.sh. Also note that the GRID=4x5 argument has no effect on the resolution at which the model runs since GCHP functions on the cubed-sphere grid. However, passing GRID is required so that components of GEOS-Chem Classic are functional.


Previous | Next | GCHP Home