This is a simple header file targeting evaluation of computation overhead between C code snippets.
It also contains some helper macros for implementing a benchmark easier, such as the CPU cache line alignment.
- AMD64 (x86-64); must be compiled with GCC/CLANG.
- i386(x86) and MSVC supports are in the plan.
Reference 'example_avx256rotl.c' source file.
It is an example code for the performance comparison of an AVX2 based 'bit-rotate-to-left' implementation to the traditional 'bit-rotate-to-left'.
- Be sure the bug was not already reported on Issues.
- If the issue seems not duplicated, open a new one from Here.
- Attaching additional information such as core-dump makes maintainers happier.
- Open a new Pull Request.
- When we meet someday, I'll buy you a pint of chilling beer.