This is the mail archive of the
mailing list for the GCC project.
Re: [tree-ssa] Tail recursion improvement
- From: law at redhat dot com
- To: Diego Novillo <dnovillo at redhat dot com>
- Cc: Zdenek Dvorak <rakdver at atrey dot karlin dot mff dot cuni dot cz>, "gcc-patches at gcc dot gnu dot org" <gcc-patches at gcc dot gnu dot org>, Jan Hubicka <jh at suse dot cz>, Richard Henderson <rth at redhat dot com>
- Date: Tue, 02 Mar 2004 12:04:46 -0700
- Subject: Re: [tree-ssa] Tail recursion improvement
- Reply-to: law at redhat dot com
In message <firstname.lastname@example.org>, Diego Novillo writ
>On Tue, 2004-03-02 at 13:37, email@example.com wrote:
>> You've moved tail recursion to a point before we've got aliasing informatio
>> which means that we haven't computed virtual uses/defs. This ultimately le
>> to transforming tail recursive calls that ought to be left alone.
>Perhaps the operand scanner may offer help here by marking CALL_EXPRs
>with has_volatile_ops if aliases haven't been computed yet.
But it would seem to me that would, in effect, prevent any tail recursion
optimization from happening at all. Really the tail recursion stuff needs
to run after we've computed aliasing information.
Your patch may still be a wise idea, but I don't think it's really the
solution to the problem.