This is the mail archive of the
mailing list for the GCC project.
Re: loop optimizer naivety
- From: Eric Botcazou <ebotcazou at libertysurf dot fr>
- To: Glen Nakamura <glen at imodulo dot com>
- Cc: gcc at gcc dot gnu dot org
- Date: Tue, 23 Sep 2003 10:38:05 +0200
- Subject: Re: loop optimizer naivety
- References: <20030923075625.GA3047@modulo.internal.net>
> FYI, I've been using the following patch by Jan Hubicka in my local tree:
Thanks for the pointer!
> The patch was intended for PR opt/10185 and possibly PR opt/10087,
> but was never reviewed. As far as I can tell, the testcase in PR
> opt/12340 doesn't segfault when using this patch... Since the patch fixes
> symptoms similar to what you described, I thought it you may be
That's indeed exactly the same problem and Jan's patch cures it by
supplementing delete_trivially_dead_insns, which is really dumb.
I see you fixed PR opt/10087 by tweaking the loop optimizer. However, it
still thinks that the biv is incremented twice and that the total increment
is 2, which badly fools the unroller.