Working with 128 bit numbers in C

[Wiretron]

Hello, how do you go about working with 128 bit numbers in C?

[laserlight]

You would use or write a fixed precision arithmetic library that supports 128 bit numbers, or an arbitrary precision arithmetic library like GMP.

[Wiretron]

I need to be able to write the code myself.

[laserlight]

So, you basically want to write your own bignum library?

[MacGyver]

So you should probably write some library to do it. Could be a number of ways of going about it.

[Wiretron]

For instance, how could i store the 128 bit number?

[laserlight]

For instance, how could i store the 128 bit number?

In an array. The question is an array of what, and that depends on your approach. For example, you could make it a string, and this would make it pretty trivial to read and write, but operations would be relatively slow.

[Wiretron]

Would that involve representing the number as a binary number in order to store it in an array?

[matsp]

It's not particularly hard to add/subtract big numbers [a 128 bit number using long long -> 64 bit numbers should be two add or sub operations, plus a quick check for overflow].

Division is a bit more tedious, as it would involve [at least] 128 shift operations - by far not impossible, but a bit lengthy.

Multiply ends up on the "inbetween" scale.

--
Mats

[Wiretron]

So do I use binary?

[matsp]

How should I know, but I think it makes sense.

--
Mats