Use of C99 extra long double math functions after r236148

David Chisnall theraven at
Tue May 28 11:55:44 UTC 2013

On 28 May 2013, at 12:12, Bruce Evans <brde at> wrote:

> Inlining the big function do_hard_work() helps for gcc on
> amd64 (about 5% faster), but makes no significant difference for clang.
> The previous testing was mostly with gcc.

How are you inlining?  With the C99 inline keyword, which changes the linkage type but only provides and advisory hint to the compiler with regard to inlining (which, in a modern compiler, is largely ignored), or with the always_inline attribute, which forces the compiler to inline the function?


More information about the freebsd-numerics mailing list