[Bug tree-optimization/70729] Loop marked with omp simd pragma is not vectorized

[jakub at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70729

--- Comment #30 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Another thing to think of, e.g.
void
baz (int *p, int *q)
{
  #pragma omp simd safelen(2)
  for (int i = 0; i < 1024; i++)
    p[4 * i] += q[0];
}
for aliasing p[4 * 1022] I think still applies that if (&q[0] == &p[4 * 1022])
then p[4 * 1022] == 0 upon entry to the loop, because while the penultimate
iteration could be executed in the SIMD chunk, it could as well be executed
before the last iteration and thus if it changes its value, it would mean
undefined behavior.