SUSv2: Can we have MAXFLOAT in <math.h> ?

[J. Johnston]

"H. Peter Anvin" wrote:
> 
> J. Johnston wrote:
> >
> > Done.
> >
> 
> Dumb question, perhaps: does newlib support non-IEEE fp?
> 

Newlib assumes IEEE floating-point format.  For example, see 
libm/common/sf_logb.c.   Some variations are supported;
for example, there is a flag _FLT_LARGEST_EXPONENT_IS_NORMAL
to specify that there are no NaNs or infinity.

-- Jeff J.