[Bug libstdc++/50724] isnan broken by -ffinite-math-only in g++

[paolo.carlini at oracle dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50724

--- Comment #5 from Paolo Carlini <paolo.carlini at oracle dot com> 2011-10-14 14:56:27 UTC ---
I haven't really followed this discussion, but I remember a very similar one
some time ago, and I suspect that part of the confusion stems from the meaning
of "finite-math": maybe some users don't consider calling std::is_nan part of a
*mathematical computation* proper, they consider it something preliminary to
it, they would like to see *only* arithmetic expressions, transcendental
functions calls, etc, optimized per finite-math. I don't know if I have been
able to explain myself well enough. Anyway, mine is just a side observation,
maybe worth keeping in mind only when writing the documentation.