[RFC] Full float128, third iteration

[Jerry DeLisle]

On 09/09/2010 01:58 PM, FX wrote:
>
> Things that need to be tested:
>    * I/O, again and again and again :)

The following program now works correctly.  I have also tested some basic 
namelist I/O with real 16 variables and all looks OK.  I am seeing some possible 
issues with the F03 rounding modes.  Still testing, may just be from having 
extra precision.

program huge_real16_formatted
   integer, parameter :: k = 16
   real(kind=k) :: a,b(2), c
   character(len=180) :: tmp
   ! Test real(k) scalar and array formatted IO with big value
   b(:) = huge (1.0_k)/2
   print *, 'real(16) big value:      ', b(1)
   write (tmp, *) b
   read (tmp, *) a, c
   print *, 'same value read again:   ', a
   print *, 'difference: looks OK now ', a-b(1)
   ! Test with really big value
   b(:) = huge (1.0_k)
   print *, 'huge value: ', b(1)
   write (tmp, *) b
   read (tmp, *) a, c
   print *, "tmp=", tmp
   print *, "These all work now!"
   print *, a
   print *, b(1)
   print *, b(2)
   print *, c
   if (a /= b(1)) call abort ()
   if (c /= b(2)) call abort ()
end program huge_real16_formatted

$ ./a.out
  real(16) big value:        5.94865747678615882542879663314003508E+4931
  same value read again:     5.94865747678615882542879663314003508E+4931
  difference: looks OK now    0.0000000000000000000000000000000000
  huge value:   1.18973149535723176508575932662800702E+4932
  tmp=  1.18973149535723176508575932662800702E+4932 
1.18973149535723176508575932662800702E+4932 

  These all work now!
   1.18973149535723176508575932662800702E+4932
   1.18973149535723176508575932662800702E+4932
   1.18973149535723176508575932662800702E+4932
   1.18973149535723176508575932662800702E+4932
jerry@eeepc:~/gcc/prs/quad$ gfc quad2.f90
jerry@eeepc:~/gcc/prs/quad$ ./a.out
  real(16) big value:        5.94865747678615882542879663314003508E+4931
  same value read again:     5.94865747678615882542879663314003508E+4931
  difference: looks OK now    0.0000000000000000000000000000000000
  huge value:   1.18973149535723176508575932662800702E+4932
  tmp=  1.18973149535723176508575932662800702E+4932 
1.18973149535723176508575932662800702E+4932 

  These all work now!
   1.18973149535723176508575932662800702E+4932
   1.18973149535723176508575932662800702E+4932
   1.18973149535723176508575932662800702E+4932
   1.18973149535723176508575932662800702E+4932