USER VISIBLE CHANGES IN GNU MP IN VERSION 2

* The mpn functions is now documented.  These functions are intended for
  very time critical applications, or applications that need full control
  over memory allocation.
* There are several new mpn functions: mpn_sqrt, mpn_gcd, mpn_set_str,
  mpn_get_str, mpn_add_1, mpn_sub_1, mpn_add_n, mpn_sub_n, mpn_mul_1,
  mpn_addmul_1, mpn_submul_1, and mpn_mul_n.
* New functions for arbitrary precision floating point arithmetic, beginning
  with mpf.  Associated type mpf_t.
* New contributed mpz functions, including a faster GCD.
* New variable types (mpz_t and mpq_t) are available that makes syntax of
  mpz and mpq calls nicer.  The MP_INT and MP_RAT types are still available
  for compatibility.
* More portable.
* Uses Cygnus configure.  This makes it possible to choose target
  architecture and CPU variant, to compile into a separate object directory,
  and to cross-compile.
* Carefully optimized assembler for important inner loops.  Support for DEC
  Alpha, Amd 29000, HPPA 1.0 and 1.1, Intel x86, and i960, Motorola MC68000,
  MC68020, MC88100, and MC88110, National NS32000, IBM POWER, MIPS R3000,
  R4000, SPARCv7, SPARCv8, and DEC VAX.  Some support also for Advanced RICS
  machines (ARM), Clipper, IBM ROMP (RT), Intel i860, Motorola/IBM PowerPC,
  Pyramid AP/XP
* Faster, thanks to the assembler code, new algorithms, and general tuning.
  In particular, the speed on machines without GCC is improved.
* Sources are reorganized.
