[Bug tree-optimization/83518] [8/9 Regression] Missing optimization: useless instructions should be dropped

[rguenther at suse dot de]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83518

--- Comment #13 from rguenther at suse dot de <rguenther at suse dot de> ---
On Tue, 9 Jul 2019, wilco at gcc dot gnu.org wrote:

> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83518
> 
> Wilco <wilco at gcc dot gnu.org> changed:
> 
>            What    |Removed                     |Added
> ----------------------------------------------------------------------------
>                  CC|                            |wilco at gcc dot gnu.org
> 
> --- Comment #12 from Wilco <wilco at gcc dot gnu.org> ---
> (In reply to Richard Biener from comment #11)
> > Fixed on trunk, not something for backporting.
> 
> The pr83518.C test fails on AArch64 and Arm (and presumably more targets). Are
> you missing options or relying on specific vector widths?

I'm not relying on vectorization (it passes for me with 
-fno-tree-vectorize).  But I'm relying on unrolling.  And also on
constant folding - maybe this is the usual

 arr[0] = 5;
 arr[1] = 4;
...

vs.

 arr = .LC0;

thing?