gmp implementation documentation

[Thomas Woldert]

Hi,

i'm looking for a detailed documentation, how the floating-point type mpf_t
is implemented. The documentation on the gmp-website just gives information
about how to work with the library. Of course I could analyze the source
code, but I'm not very good in C, so it would take to much time.

Also i'm interested in any papers about how to implement floating-point
datatypes working with 'contiguous mantissa'.

Thanx a lot,
Heinz