This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: More Q's regarding gimplification
- From: Steven Bosscher <stevenb at suse dot de>
- To: neroden at fastmail dot fm (Nathanael Nerode), gcc at gcc dot gnu dot org
- Date: Mon, 16 Aug 2004 14:05:27 +0200
- Subject: Re: More Q's regarding gimplification
- Organization: SUSE Labs
- References: <20040816112729.GA2904@fastmail.fm>
On Monday 16 August 2004 13:27, Nathanael Nerode wrote:
> gimple_predicate
> rhs_predicate_for (tree lhs)
> {
> if (is_gimple_tmp_var (lhs))
> return is_gimple_tmp_rhs;
> else if (is_gimple_reg (lhs))
> return is_gimple_reg_rhs;
> else
> return is_gimple_mem_rhs;
> }
>
> The question is this:
> Is it supposed to be possible to have
> is_gimple_tmp_var(lhs) && !is_gimple_reg(lhs)
> or is
> is_gimple_tmp_var
> supposed to imply
> is_gimple_reg?
The former.
Gr.
Steven