This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: TREE_READONLY in lookup_tmp_var
- From: Richard Henderson <rth at redhat dot com>
- To: Richard Kenner <kenner at vlsi1 dot ultra dot nyu dot edu>
- Cc: gcc at gcc dot gnu dot org
- Date: Fri, 20 Aug 2004 13:20:18 -0700
- Subject: Re: TREE_READONLY in lookup_tmp_var
- References: <10408201303.AA02453@vlsi1.ultra.nyu.edu>
On Fri, Aug 20, 2004 at 09:03:25AM -0400, Richard Kenner wrote:
> You removed the setting of TREE_READONLY there.
It was wrong. These temporaries *are* set more than once.
> That caused 14 ACATS regressions!
Really.
Well, no one else can tell because no one else can build the ada front
end at all. This is your fault for not committing anything to fix the
things that you know are wrong. In particular, the alloca fix is a
complete show-stopper.
I should remind you that batching changes is a bad habit.
> In an elaboration routine, we often set a field of a global variable to
> the address of a CONSTRUCTOR that just has integer constants.
Why are you ever taking the address of a CONSTRUCTOR? A constructor
is not an lvalue. Use a real variable. *Especially* if you are
relying on it being treated a particular way, namely, by setting
TREE_STATIC+TREE_READONLY.
r~