[RFC] Full float128, third iteration

[Dominique Dhumieres]

FX,

Preliminary tests on x86_64-apple-darwin10.4.0:

(1) I need to compile with -lquad (even if I use the patch in
http://gcc.gnu.org/ml/fortran/2010-09/msg00250.html ,
although I may need to bootstrap to use it?),

(2) the following test

program a
integer, parameter :: k1 = selected_real_kind (precision (0.0_8) + 1) ! Set to IBM Extended kind to see problem.
integer, parameter :: knd = selected_real_kind (precision (0.0_k1) + 1) ! Set to IBM Extended kind to see problem.
   integer b, e, p
   real(knd) x,y
   b = radix(x)
   e = maxexponent(x)
   p = digits(x)
   x = huge(x)
   print*, x
   print*, (1._knd - real(b,knd)**(-p)) * real(b,knd)**(e-1) * b
   print*, x-(1._knd - real(b,knd)**(-p)) * real(b,knd)**(e-1) * b
end program a

fails at compile time with

Undefined symbols:
  "__gfortran_pow_r16_i4", referenced from:

(3) gfortran.dg/promotion.f90 is failing and contrary to Steve I think
this should be fixed.

Otherwise the few other tests behave as expected.

Thanks for tha work and stay tuned!-)

Dominique