This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] Remove GMP in favor of MPFR
On Sun, Jul 04, 2004 at 05:17:32PM +0100, Paul Brook wrote:
>
> I think these numbers come from gcc/real.c. However reading a bit further
> uncovers the following comment:
> /* Recall that IEEE numbers are interpreted as 1.F x 2**exp,
> whereas the intermediate representation is 0.F x 2**exp.
> Which means we're off by one. */
>
> There's also Note 13.4 from the standard, which looks a lot like ieee single
> precision format.
>
In learning what MPFR does with underflow, I wrote a simple
program that repeated divided 1 by 10. Here's a small
sampling of MPFR's binary and decimal output:
1.10011001100110011001101e-4 1.000000e-1
1.01000111101011100001010e-7 1.000000e-2
1.00000110001001001101110e-10 9.999999e-4
1.10100011011011100010110e-14 9.999999e-5
It appears MPFR will give us 1.F x 2**exp, which is the behavior
specified by the Fortran standard.
--
Steve