This repository contains the benchmark code for the paper Legendre PRF (Multiple) Key Attacks and the Power of Preprocessin.
cmake
3.20- c++ 20 compiler, something like
clang++-14
gmp
,gmpxx
To build the project simply run:
mkdir cmake-build-release
cd cmake-build-release/
cmake -DCMAKE_BUILD_TYPE=Release ..
make -j1
To run a benchmark, simply execute
python gen.py -r 33 --iterations 10 --bench --executable code --include main.h
or
python gen.py -r 33 --iterations 10 --bench --executable code_noprep --include main.h
to generate data for all prime up the bit length 32
, where each data point is
averaged over 10 measurements. Note that you have the set the m
value in the
python script gen.py
in line 94
depending on which setting (single-key or
multiple-key) you want to benchmark.
The resulting *.log
file can be converted into an tikz
readable format with
python script plot.py
.