This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Regression for trunk on i686-pc-linux-gnu
- From: Diego Novillo <dnovillo at redhat dot com>
- To: Richard Henderson <rth at redhat dot com>
- Cc: Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>, "gcc at gcc dot gnu dot org" <gcc at gcc dot gnu dot org>
- Date: Tue, 27 Jul 2004 19:14:16 -0400
- Subject: Re: Regression for trunk on i686-pc-linux-gnu
- Organization: Red Hat Canada
- References: <10407272150.AA25883@vlsi1.ultra.nyu.edu> <20040727230917.GA2086@redhat.com>
On Tue, 2004-07-27 at 19:09, Richard Henderson wrote:
> Having looked at the various uses of is_gimple_addr_expr_arg, I'm
> of the opinion that INDIRECT_REF should be added back to it. Every
> place uses build_fold_addr_expr, or the equivalent, which does fold
> away the INDIRECT_REF immediately.
>
Wait. Why accept it then? We don't accept INDIRECT_REF <ADDR_EXPR <>>,
why should we accept the reverse? Particularly, when we always use it
with the folder.
If we accept it and then forget to use the folder, we'll get into
trouble.
Diego.