This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug c++/69564] [5/6 Regression] lto and/or C++ make scimark2 LU slower
- From: "jason at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Tue, 23 Feb 2016 20:57:29 +0000
- Subject: [Bug c++/69564] [5/6 Regression] lto and/or C++ make scimark2 LU slower
- Auto-submitted: auto-generated
- References: <bug-69564-4 at http dot gcc dot gnu dot org/bugzilla/>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69564
--- Comment #7 from Jason Merrill <jason at gcc dot gnu.org> ---
(In reply to Richard Biener from comment #4)
> Yeah, didn't try to figure out whether the C vs. C++ thing is a
> regression. But I suspect the change to the C++ loop lowering.
Yes, the relatively small difference between C and C++ without LTO seems to be
due to the loop inversion that the C front end still does, and the C++ front
end doesn't. But if I add loop inversion back into the C++ front end so that
the .optimized output is indistinguishable, that resolves the difference
without LTO, but LTO still makes the C++ output much slower.