Re: does gfortran do double double ??

On Mon, Jan 08, 2007 at 02:44:38PM -0700, Reg Clemens wrote:
Most Fortran Compliers I have seen on 32 bit machines have a command
line option that converts double precision from 64 -> 128 bits.

I see some options in gfortran to control variable lenght, but none that
do the above.

Check out:

-fdefault-double-8
Set the "DOUBLE PRECISION" type to an 8 byte wide.

I saw that, and find it rather strange.
Single precision in normal 32bit, hence 4bytes.
That makes double 8bytes by default, so why does this statement exist?
I would want something that said -fdefault-double-16


--
Reg.Clemens
reg@xxxxxxx


--
fedora-list mailing list
fedora-list@xxxxxxxxxx
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list

Relevant Pages

  • Re: Possible F77 Code Improvement ??
    ... The problem is that gfortran should know it's going to execute ... REAL*10 computations but doesn't set precision control accordingly. ... 32-bit compilers include the option /Qpc80 to set 64-bit precision mode, ...
    (comp.lang.fortran)
  • Re: Possible F77 Code Improvement ??
    ... Can you be more specific in what you think the bug is? ... The problem is that gfortran should know it's going to execute ... REAL*10 computations but doesn't set precision control accordingly. ... It appears that GCC will use the FPU control ...
    (comp.lang.fortran)
  • Re: glibc detected
    ... gfortran for compiling.) ... the version of gfortran that you are using. ... is a single precision constant; ... have a portability problem) ...
    (comp.lang.fortran)
  • Re: another question about precision
    ... because this will "truncate" them to their natural precision. ... I don't know about gfortran in this regard. ... Few compilers optimize well for them. ... precision mode. ...
    (comp.lang.fortran)
  • Re: does gfortran do double double ??
    ... I see some options in gfortran to control variable lenght, ... Set the "DOUBLE PRECISION" type to an 8 byte wide. ... Dave Ihnat ...
    (Fedora)