[Bug tree-optimization/79224] [7/8 Regression] Large C-Ray slowdown

[aldyh at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79224

--- Comment #19 from Aldy Hernandez <aldyh at gcc dot gnu.org> ---
In my sandbox I see no discernible difference between GCC 6 and current
mainline (r256265).

I am testing with:

./xgcc -B./ -O3 -ffast-math -funroll-loops -march=core-avx2 -lm c-ray-f.c
cat ~/scene |  ./a.out -s 5000x4000 >/dev/null

The numbers for GCC 6 and mainline are pretty close.  I ran each version 4
times on a Linux box under runlevel 3 (no X).  The average is:

Mainline: 5048 milliseconds
GCC 6:    5059 milliseconds

I am testing on: Intel(R) Core(TM) i7-6600U CPU @ 2.60GHz

If y'all agree with my tests, should we close this PR?