about non-compatible optimization (was: Re: patch to bring java vtables closer to g++ abi conformance)

Tuomas Leikola tuomas.leikola@digia.com
Tue Jan 29 04:16:00 GMT 2002


----- Original Message -----
From: "Tom Tromey" <tromey@redhat.com>
To: "Bryce McKinlay" <bryce@waitaki.otago.ac.nz>
Cc: "Tuomas Leikola" <tuomas.leikola@digia.com>; <java@gcc.gnu.org>
Sent: Tuesday, January 29, 2002 7:38 AM
Subject: Re: about non-compatible optimization (was: Re: patch to bring java
vtables closer to g++ abi conformance)


> >>>>> "Bryce" == Bryce McKinlay <bryce@waitaki.otago.ac.nz> writes:
>
> Bryce> In fact a virtual method call (in a loop) in GCJ is *faster*
> Bryce> than a static method call (in Java or C).
>
> This is very interesting, because it means that even if we can
> devirtualize a method call (via the eventual AST code), we might not
> want to in certain situations.
>
> Tom
>

unless devirtualization can assume the class is already initialized, and it
most of all enables the inliner to do it's stuff.



More information about the Java mailing list