Repository for refactoring Parla to a C++ runtime to avoid GIL during scheduling
Note: Ensure that the Cupy version and your CUDA Toolkit are compatible with each other and the architecture of your machine. On TACC systems we recommend installing CuPy (and if needed numba) via pip against the system CUDA version instead of pulling in binaries from conda-forge's cudatoolkit.
Can be built with:
git submodule update --init --recursive
pip install .
You may need to set C compilers explicily on Frontera.
module load gcc/12
make clean; CC=gcc CXX=g++ make
This will populate a build directory and an install into your active Python environment's site-packages folder.
If you are a developer, it may sometimes be necessary to remove the build directory manually to apply changes and remove the cached version.
Changes to most files in the src tree will trigger a rebuild of the relevant files without needing to clear the build directory.
Additionally, scikit-build-core supports editable installations that recompile changes on each run, please see this documentation.
You may need to set C compilers explicily on Frontera.
module load gcc/12
make clean; CC=gcc CXX=g++ make
Build options are set through the pyproject.toml: PARLA_ENABLE_NVTX=ON/OFF : Enables NVTX hooks in the C++ backend. PARLA_ENABLE_LOGGING=ON/OFF : Enables Binlog hooks in the Python & C++ backend for logging. PARLA_ENABLE_CUDA=ON/OFF : Enables CUDA C++ backend for runahead scheduling PARLA_ENABLE_HIP=ON/OFF : Enables HIP C++ backend for runahead scheduling
Logging is done through the high-performance binlog library. We wrap the logging function to dump to the same logs from Python, Cython, & C++. Logging from Python & Cython is more expensive than in C++ due to the cost of calling a c-extension, be aware of performance penalties from logging many things.
Logs are dumped when the Parla runtime exits. The name of the log file defaults to "parla.blog" but can be set with the PARLA_LOGFILE enviornment variable or as a parameter in the with Parla(logfile=filename) context.
If a Parla program is interrupted with a KeyboardInterrupt Exception (aka Ctrl+C) it should dump the log as the runtime exits gracefully. The same should hopefully now be true of exceptions raised within tasks or workers. But please raise an issue if you found one that isn't handled.
If a Parla program is interuppted with an error that causes a crash and core dump, the logfile can be recovered from the coredump directly with brecovery. See the binlog documentation for more information.
To read a logfile run:
bread <logfile>
The number of workers in the Parla runtime is controlled by the PARLA_NUM_THREADS enviornment variable. If it is not set it defaults to the number of cores on the host machine.