This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: virtual inline optimizations
- From: Richard Henderson <rth at redhat dot com>
- To: Reza Roboubi <reza at linisoft dot com>
- Cc: "gcc at gcc dot gnu dot org" <gcc at gcc dot gnu dot org>
- Date: Sat, 12 Oct 2002 10:03:20 -0700
- Subject: Re: virtual inline optimizations
- References: <3DA840A2.8DCE17F9@linisoft.com>
On Sat, Oct 12, 2002 at 08:32:50AM -0700, Reza Roboubi wrote:
> For instance, xx->foo() should have a compile time resolvable binding if
> xx points to a class at the _bottom_ of the class hierarchy(am I wrong
> about that??)
There is no "bottom" of the class hierarchy unless one is doing
whole-program analysis. If you're only looking at one source file,
someone somewhere else could create a further derivation.
What is more likely is that in some situations we'll be able to
deduce the true type of the object referred to by xx, and from there
call the proper method directly.
r~