Re: Are f.p. manipulation functions only used in initialization?

In article <fv2h8l$2p4t$3@xxxxxxxxxx>,
"FX" <coudert@xxxxxxxxxxxxx> writes:
Hi James,

Testing with gfortran I found that
RRSPACING, SCALE, SET_EXPONENT, and, SPACING were broken like this,
while EXPONENT, FRACTION, and NEAREST seemed to work

Thanks for your example, I will try to reduce it and understand what's
happening; if it happens to not be a Fortran front-end issue, I'll alert
our great Win64 developers to it.


It's an OS issue. Look at -fdump-tree-original.

--
Steve
http://troutmask.apl.washington.edu/~kargl/
.

Relevant Pages

  • Re: Are f.p. manipulation functions only used in initialization?
    ... RRSPACING, SCALE, SET_EXPONENT, and, SPACING were broken like this, ... EXPONENT returned garbage for REAL*10 and FRACTION ...
    (comp.lang.fortran)
  • Re: Are f.p. manipulation functions only used in initialization?
    ... RRSPACING, SCALE, SET_EXPONENT, and, SPACING were broken like this, ... while EXPONENT, FRACTION, and NEAREST seemed to work ...
    (comp.lang.fortran)
  • Re: Im a bit rusty
    ... | floating-point capabilty to my asm printf function to verify results ... | exponent, ... But conversion of the binary mantissa by the remaining fraction of the ...
    (comp.lang.asm.x86)
  • Re: Optimizing exponentiation
    ... > The C library includes frexp() and ldexpto separate the exponent ... See the fraction, exponent, and scale intrinsics in F90 or later. ...
    (comp.lang.fortran)
  • Re: computing t = pow(-11.5, .333)
    ... and the exponent is a fraction. ... In standard C this would be something like t = pow, but with this combination of arguments there is a domain error and the result is a percolating NaN. ...
    (comp.lang.c)