moepgf finite field arithmetic library ====================================== The moepgf finite field arithmetic library offers high-performance operations over GF(2), GF(4), GF(16), and GF(256) using SSE/AVX/NEON SIMD extensions on x86/ARM platforms. Installation ------------ To install moepgf you can use the following commands: ./configure make make install See ./configure --help for details on the available parameters. For example you can use --prefix=DIR to specify the installation prefix. To check the results before installation you can issue: make check If you use a git checkout you must run the following commands first: apt-get install automake autoconf autoconf-archive autotools-dev libtool autoreconf -fi Compilers and Options --------------------- The library requires a version of the GNU C compiler supporting AVX2 instrinsics (x86) or NEON instrinsics (ARM). Otherwise, compilation will fail. Documentation ------------- All functions users should interface with are defined and documented in includedir/moepgf/moepgf.h. Knwon bugs ---------- - The function moepgf_init() currently accepts only GF_SELFTEST and GF_ALGORITHM_BEST as algorithms, i.e., either the most trivial implementation or the best implementation available according to the supported feature set is returned. In order to get, for instance, the SSE2 implementation on a processor supporting AVX2, you have to use gf_get_algorithms() which returns a list with all algorithms. This should not be of concern in any practical cases since the all suboptimal implementations are only useful for benchmarks. Where to get the latest ----------------------- moepgf is available at moep80211.net.