It is possible to call your assembly language code from Python. In fact, it can be insanely easy. It can also be difficult. What differentiates the two are the arguments you need to pass.
Python stores data in a very different way than C and therefore assembly language. All Python data types are objects, which for sets and dicts can make things difficult and outside the scope of this book.
If you want to explore the subject deeply, we suggest this link
Take this trivial C function:
int square(int x) {
return x * x;
}
It could be written in this way using our Apple / Linux convergence macros:
#include "apple-linux-convergence.S"
/*
gcc -fPIC -shared -o my_square.so function.S
*/
.p2align 2
.text
GLABEL square
#if defined(__APPLE__)
_square:
#else
square:
#endif
mul x0, x0, x0
ret
.end
This function is a leaf and requires no interaction with the stack.
Notice there is no main()
.
Using the following command line, turn your assembly language module into a "shared object" or .so file:
gcc -fPIC -shared -o my_square.so function.S
-fPIC
means generate position independent code. This allows the library
to be relocated wherever the program loading the library wants it.
Then, in Python:
from ctypes import *
so_file = "./my_square.so"
my_funcs = CDLL(so_file)
print(my_funcs.square(10))
It just works.