• Currently, the installation only works on linux machines.

  • CUMODP needs CUDA Toolkit 7.0 or later, on devices of compute capability 2.0 or above.

  • Maple and NTL are required for some of the verification and benchmarking.

  • Installation:
    • Download latest.tar.gz
    • $ tar -xzf latest.tar.gz
    • $ cd cumodp-2.0
    • $ make install
  • The following environmental variables will be automatically set by install_cumodp.sh:

    • CUMODP_HOME : The home directory of the library which is "CUMODP_HOME". By default, this variable should point to the directory where you have extracted the source files.

    • MODPN_HOME : the home directory of the MODPN library.

    • "$CUMODP_HOME/src" and "$MODPN_HOME/src" are added to "LD_LIBRARY_PATH".

  • Recompilation:

    Use the Makefile in the "$CUMODP_HOME" directory and simply execute "make". This command will compile and create the CUMODP library in $CUMODP_HOME/src.

  • Testing:

    There are test files written in C in the "new_tests" directory for all of the functionalities. There is Makefiles inside of each subdirectories.

If you have questions or bug reports, please contact: cumodp sdl [dot] csd [dot] uwo [dot] ca

© 2017, ORRCA

Ontario Research Center for Computer Algebra (ORRCA)
Department of Computer Science
The University of Western Ontario
London, Ontario, Canada N6A 5B7