I must to write a floating point multiprecision library by myself (it's a practice for my studies). Anyone knows where I can find the explanation of these algorithms (and pseudo-code) with some tricks for speed up? Yes, I know that there is many opensource libraries but these libraries are too big and complex for me... :confused:
Posted on 2003-08-10 22:22:06 by wildark
Posted on 2003-08-10 22:31:54 by bitRAKE
GMP is really perfect, and I recommend that you install Cygwin to build it.
The problem is that it is optimized for the processor where it is compiled, because the library supports almost all x86 processors !

I had to extract the ASM routines for my own purpose, and it was quite long to do, since I had to compile, disassemble the .O files, and take the routines I needed (the .M4 files are in f***ing GAS format).
Posted on 2003-08-11 03:19:55 by MCoder