[Bug math/19869] floor and fmod not consistent with each other

[fweimer at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=19869

--- Comment #6 from Florian Weimer <fweimer at redhat dot com> ---
(In reply to Robert Dodier from comment #5)
> (In reply to Florian Weimer from comment #4)
> > I think this is just one of the cases where the specification of fmod does
> > not uniquely determine the value to be returned.  Both 0 and
> > 1.5999999999999996 are valid because 8.0 - 4 * 1.6 = 1.5999999999999996 and
> > 8.0 - 5 * 1.6 = 0.
> 
> Florian, thanks for taking the time to comment. It seems to follow from what
> you're saying that the integer part of 8/1.6 might be either 4 or 5, and
> these are equally valid. Doesn't it have to be one or the other?

Not according to the POSIX specification.

> Incidentally 8.0 - 5*1.6 is actually not 0.0 but rather something like
> -4e-16.

Under which rounding mode?  I can't reproduce this.

-- 
You are receiving this mail because:
You are on the CC list for the bug.