Assembly Divisions: Algorithms & #s Not Fitting Registers

[mtanti]

First of all i'de like to know if there is an efficiant algorithm which assembly uses to divide two numbers.
Second I'de like to know how do you divide numbers which don't fit into a register. I'm talking about an array of bytes in RAM to be divided.

10x!

 

[saltydog]

First of all i'de like to know if there is an efficiant algorithm which assembly uses to divide two numbers.
Second I'de like to know how do you divide numbers which don't fit into a register. I'm talking about an array of bytes in RAM to be divided.

10x!

Well, you could use multi-precision arithmetic if it's really large numbers or very high precision. I tell you what though, that would be quite a neat challenge in assembly. What about just C++? You know that's pretty close to assembly speed. Anyway, just search the web for "multi-precision arithmetic". The algorithm is straight-forward and intuitive once you work with it for a while.

As far as what assembly does, 2's compliments come to mind but I'd really have to review on that one.

 

[dduardo]

In ASM you can do integer division by using DIV. If you want to do floating point division use FDIV.

Here is a C library for doing arbitrary precision arithmetic:

http://www.swox.com/gmp/index.orig.html

There are more specialized libraries depending on how many bits of precision you really need. This is just a general case library.