>>>> "Hrvoje" == Hrvoje Niksic
<hniksic(a)iskon.hr> writes:
Hrvoje> Andy Piper <andy(a)xemacs.org> writes:
> At 09:19 AM 2/24/00 +0100, Hrvoje Niksic wrote:
> >Notice how in almost all files INLINE is #defined to _extern_ inline,
> >not to static or static inline as one would assume. This is so that
> >the inline functions defined in headers work right.
>
> RIght, inline IMHO should almost always be used in headers. If you
> define inline functions in a source file (as per redisplay-msw.c)
> any decent compiler should inline them declared as static
> anyway.
Hrvoje> Oh, I don't know about that. I'd still like to be able to just
Hrvoje> specify INLINE, meaning `static inline' or whatever is appropriate.
Hrvoje> You can't do that at all with the current setup.
Yes, you can - `static inline'.
Hrvoje> For instance, `gcc -O2' (or, for that matter, `gcc -O') doesn't
inline
Hrvoje> any functions, but it does inline the ones specifically requested.
Hrvoje> `gcc -O3' inlines all it can, but was reported as buggy under some
Hrvoje> architectures.
Whatever you do, `inline' will never be more than a hint to the compiler.