[Bug other/47167] Performance regression in numerical code

["martin at mpa-garching dot mpg dot de" <gcc-bugzilla at gcc dot gnu dot org>]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47167

--- Comment #9 from Martin Reinecke <martin@mpa-garching.mpg.de> 2011-01-19 17:26:31 UTC ---
(In reply to comment #8)
> Can you check if the following patch solves your problem?

Yes, this patch gets performance back to normal on the 4.5 branch and on trunk.
Great!

> The differences in GIMPLE of the patch do not explain the code-differences
> though, so it might be just bad luck that the patch regressed things for
> you.  I can see other unwanted differences though.

I would of course be happy if the code generation would be less "erratic", and
if the nice performance I'm seeing does not depend on my luck ;)
So if I can do anything to help optimizing this kind of code more consistently,
please let me know! Of course, I'm more into numerics than into compiler
writing ...