From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <gcc-return-55940-listarch-gcc=gcc.gnu.org@gcc.gnu.org>
Received: (qmail 23875 invoked by alias); 1 Jul 2002 18:58:59 -0000
Mailing-List: contact gcc-help@gcc.gnu.org; run by ezmlm
Precedence: bulk
List-Archive: <http://gcc.gnu.org/ml/gcc/>
List-Post: <mailto:gcc@gcc.gnu.org>
List-Help: <http://gcc.gnu.org/ml/>
Sender: gcc-owner@gcc.gnu.org
Received: (qmail 23839 invoked from network); 1 Jul 2002 18:58:56 -0000
Received: from unknown (HELO executor.cambridge.redhat.com) (195.224.55.237)
  by sources.redhat.com with SMTP; 1 Jul 2002 18:58:56 -0000
Received: from cuddles.cambridge.redhat.com (vpn50-1.rdu.redhat.com [172.16.50.1])
	by executor.cambridge.redhat.com (Postfix) with ESMTP
	id 24646ABAFC; Mon,  1 Jul 2002 19:58:52 +0100 (BST)
Received: (from aph@localhost)
	by cuddles.cambridge.redhat.com (8.11.6/8.11.0) id g61Iwvg16057;
	Mon, 1 Jul 2002 19:58:57 +0100
X-Authentication-Warning: cuddles.cambridge.redhat.com: aph set sender to aph@cambridge.redhat.com using -f
From: Andrew Haley <aph@cambridge.redhat.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit
Message-ID: <15648.42609.739440.682415@cuddles.cambridge.redhat.com>
Date: Mon, 01 Jul 2002 11:58:00 -0000
To: Daniel Berlin <dberlin@dberlin.org>
Cc: Mark Mitchell <mark@codesourcery.com>,
	Per Bothner <per@bothner.com>, Fergus Henderson <fjh@cs.mu.OZ.AU>,
	"gcc@gcc.gnu.org" <gcc@gcc.gnu.org>,
	"java@gcc.gnu.org" <java@gcc.gnu.org>
Subject: Re: RFC: Java inliner
In-Reply-To: <Pine.LNX.4.44.0207011455550.20102-100000@dberlin.org>
References: <15648.40350.493795.623862@cuddles.cambridge.redhat.com>
	<Pine.LNX.4.44.0207011455550.20102-100000@dberlin.org>
X-SW-Source: 2002-07/txt/msg00027.txt.bz2

Daniel Berlin writes:
 > On Mon, 1 Jul 2002, Andrew Haley wrote:
 > 
 > > Mark Mitchell writes:
 > >  > 
 > >  > The current inliner already has mechanisms for language-specific
 > >  > extensions.  If those can be used, or it can be easily extended so that
 > >  > they can be used, great.
 > > 
 > > Well I tried, I really did.  But the inliner we have at the moment
 > > uses C-private tree nodes to such an extent that I think it isn't
 > > possible without totally messing up the code.  I could replace every
 > > occurrence of these nodes with macros or function pointers, but the
 > > result would not be pretty.  The real answer in the long term is, of
 > > course, SIMPLE.
 > 
 > Is the java inliner usable in languages other than java (IE could we 
 > replace the current c-inliner with your java inliner)?

No, because C needs the special C tree nodes.

 > Or did you do the same thing, and create java specific nodes?

We use only the standard nodes.  However, the way we interpret a
couple of these nodes, such as BLOCK, is Java specific.

Andrew.