This is the mail archive of the
mailing list for the GCC project.
Re: Various reworkings of PLACEHOLDER stuff
- From: Diego Novillo <dnovillo at redhat dot com>
- To: Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>
- Cc: Steven Bosscher <s dot bosscher at student dot tudelft dot nl>, "gcc-patches at gcc dot gnu dot org" <gcc-patches at gcc dot gnu dot org>
- Date: 29 May 2003 14:23:00 -0400
- Subject: Re: Various reworkings of PLACEHOLDER stuff
- Organization: Red Hat Canada
- References: <10305281109.AA10642@vlsi1.ultra.nyu.edu>
On Wed, 2003-05-28 at 07:09, Richard Kenner wrote:
> And PLACEHOLDER_EXPR is not part of GIMPLE...
> Since it's a defined tree node, it most certainly should be (I assume
> WITH_RECORD_EXPR isn't either), but not much has to be done with it
> since it's used only when converting trees to RTL.
Nope. GIMPLE is only a subset of GENERIC which, in turn, should be
language independent to a large extent. The idea is for every language
to lower their representations into something that can be expressed with
In some cases, the constructs have very unique control or data flow
properties, and we may need to extend GENERIC. For instance, we did
this for try/catch. But in principle, if you can express it in RTL,
then you can express it in GENERIC/GIMPLE.