This is a tiny proof of concept about function hooking under x86 and Unix.
See main.c for a basic code example.
A word on how it works
Let's say we have two functions original_func and new_func, and we want to redirect all the calls to the former to the latter.
In order to do that, we modify the beginning of the code of original_func to insert an instruction jumping to new_func.
This instruction will be a jmp, whose parameter is the distance between original_func and new_func, minus the size of the jump instruction (because the current address at the time of the jump is original_func + size_of_the_jmp_instruction)
Before actually writing that instruction, we also backup the bytes it will overwrite, so we can restore them if needed.
That way, we can easily remove a hook by copying the backed-up bytes back to their original location in original_func.