c++/4341: g++ built-in type double/float looses precision

Tue Sep 18 02:37:00 GMT 2001

On 18 Sep 2001 nathan@gcc.gnu.org wrote:

>     not a bug. Please read http://www.acm.org/pubs/citations/journals/surveys/1991-23-1/p5-goldberg/
>     and other information about why floating point numbers
>     are not real numbers

Though, entirely incidentally to this PR, C99 recommends that we should
warn when a hexadecimal floating constant can't be exactly represented in
its type, and we don't do that either.

-- 
Joseph S. Myers
jsm28@cam.ac.uk