This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: 252.EON dropped 25% on between 8-31 and 9-02
- From: Josh Conner <jconner at apple dot com>
- To: Daniel Berlin <dberlin at dberlin dot org>
- Cc: gcc-patches <gcc-patches at gcc dot gnu dot org>
- Date: Thu, 21 Sep 2006 22:38:55 -0700
- Subject: Re: 252.EON dropped 25% on between 8-31 and 9-02
- References: <4aca3dc20609211518k6cc2a195o6f00235654ce05f5@mail.gmail.com>
Daniel Berlin wrote:
> Our SPEC2000 performance on 252.eon dropped by a pretty large amount
> between those days.
>
> The only patch I see that could have affected this was:
>
> +2006-09-01 Josh Conner <jconner@apple.com>
> +
> + PR c++/25505
> + * tree-nrv.c (dest_safe_for_nrv_p): New function.
> + (execute_return_slot_opt): Use it.
>
>
>
> Josh, I assume you access to SPEC, so could you take a look at this?
>
> (Sorry if i have pegged the wrong patch, i'm just going by the spec
> tester diffs).
Well, I don't see any difference on my i686-pc-linux-gnu box (actually,
I see a 1.5% speedup). I guess it's still possible my patch does
something horrific on the automated-test machine, but I'm thinking I
might not be the culprit. I'll try to find some cycles tomorrow to help
narrow this down further, since it is such a serious regression.
- Josh