[Bug c/14384] Invalid use of extra precision floating-point with -O0 optimization

[wilson at specifixinc dot com]

------- Additional Comments From wilson at specifixinc dot com  2004-03-03 18:57 -------
Subject: Re:  New: Invalid use of extra precision floating-point
 with -O0 optimization

anz at obs-nice dot fr wrote:
> Arithmetic involving a float and a "lesser" type should be done in float in C89,
> at least when optimization is turned off.

This is false.  K&R C required float operations to be performed in 
double.  C89 allows float operations to be performed as float, but does 
not require it.

In my ANSI C 89 manual, this is section 3.2.1.5 Usual Arithmetic 
Conversions, last sentence
"The values of floating operands and of the results of floating 
expressions may be represented in greater precision and range than that 
required by the type; the types are not changed thereby."

This gives the compiler the option of using float/double/long double for 
operations on float types.  This is unfortunately necessary, because 
some hardware, the classic x86 FPU in particular, does not have a full 
set of operations.  It only has one, which for linux, is long double by 
default.


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=14384