This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug tree-optimization/69720] [6 Regression] wrong code at -O3 on x86_64-linux-gnu
- From: "rguenth at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Mon, 08 Feb 2016 13:11:44 +0000
- Subject: [Bug tree-optimization/69720] [6 Regression] wrong code at -O3 on x86_64-linux-gnu
- Auto-submitted: auto-generated
- References: <bug-69720-4 at http dot gcc dot gnu dot org/bugzilla/>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=69720
--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
So we end up vectorizing the outer loop in
for (; a<6; a++)
{
b ^= e;
tem = d[a*6];
for (c = 0; c < 2; c++)
tem |= 9;
d[a*6] = tem;
}
where the store-motioned d[] load wasn't vectorizable before the change.
We generate totally stupid vectorized b ^= e and then fuck up trying to
generate totally stupid vectorized d :/