This is the mail archive of the
java@gcc.gnu.org
mailing list for the Java project.
Re: does gcj do inlining when generating bytecodes?
- From: Per Bothner <per at bothner dot com>
- To: tromey at redhat dot com
- Cc: Adam Megacz <gcj at lists dot megacz dot com>, java at gcc dot gnu dot org
- Date: Mon, 17 Dec 2001 10:21:35 -0800
- Subject: Re: does gcj do inlining when generating bytecodes?
- References: <867krlsqba.fsf@megacz.com> <87vgf5vi87.fsf@creche.redhat.com>
Tom Tromey wrote:
>>>>>>"Adam" == Adam Megacz <gcj@lists.megacz.com> writes:
>>>>>>
>
> Adam> I asked earlier if .java -> .o would inline, and the answer was
> Adam> yes. Will GCJ inline small final/private methods within a single
> Adam> class when compiling .java -> .class?
>
> Nope.
>
> We don't really do much, if any, optimizing when generating bytecode.
> We could, but it has never been seen as very important.
Also, if a tree-based inliner is written (I believe that was one of the
intents, given that front-ends are now encouraged to represent
whole functions/methods as trees, but I don't know the status),
then it might be easier to do this.
--
--Per Bothner
per@bothner.com http://www.bothner.com/per/