Jean-Claude,

you can expect your compiler to ignore your "inline" declaration. Or do something else with it. It is just a hint. [1]

A sufficiently smart compiler might use compile-time type inference to find a method matching some set of argument types and then replace the function invocation by the body of that method. To me, it appears to be questionable whether such a smart compiler would depend on inline declarations.

I am just a user, though, and maybe one of the compiler writers on this list want to add some of the fine points surrounding this :)

-Hans

[1] http://clhs.lisp.se/Body/d_inline.htm

2014-07-10 0:00 GMT+02:00 Jean-Claude Beaudoin <jean.claude.beaudoin@gmail.com>:

Hello CL pros,

I am trying to figure out the meaning of the inline declaration when the so qualified function happens to be a generic function. As in:

(declaim (inline foo))
(defgeneric foo (a b))

What can be expected of a compiler when it then faces a call to foo while the above is in effect?

Is this an obvious no-op or did some find something useful to do in such a context?

Thanks for your help,

JCB

_______________________________________________
pro mailing list
pro@common-lisp.net
http://common-lisp.net/cgi-bin/mailman/listinfo/pro