Macros using __NO_LONG_DOUBLE_MATH

[Joseph Myers]

On Mon, 8 Jun 2015, Wilco Dijkstra wrote:

> Making GLIBC support sNaN correctly is likely a lot of work - many functions
> start like this: if (isinf(x) || x == 0.0)... And there seems to be no test 

But for the bulk of functions that should raise "invalid" on sNaN, 
starting like that is no problem at all (it's fine for == 0.0 to raise 
"invalid" in that case, no need to move to iszero).

-- 
Joseph S. Myers
joseph@codesourcery.com