[Bug target/70232] [6 regression] excessive stack usage with -O2

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=70232

--- Comment #8 from Richard Biener <rguenth at gcc dot gnu.org> ---
But not sure how this became a regression - does GCC 5 really do better here?

With -fno-tree-dominator-opts the warning no longer triggers and we get

        sub     sp, sp, #428

so it may be another jump-threading related regression.  With additional
-fno-tree-vrp we get to

        sub     sp, sp, #140

> grep '<< 56' t.c.211t.optimized
  _68 = _38 << 56;
  _98 = _97 << 56;

for both and with VRP and DOM

> grep '<< 56' t.c.211t.optimized
  _68 = _38 << 56;
  _98 = _97 << 56;
  _137 = _138 << 56;
  _375 = _376 << 56;
  _342 = _343 << 56;
  _170 = _169 << 56;
  _603 = _602 << 56;
  _636 = _635 << 56;
  _209 = _208 << 56;
  _683 = _682 << 56;
  _716 = _715 << 56;
  _239 = _238 << 56;
  _523 = _522 << 56;
  _556 = _555 << 56;
  _300 = _301 << 56;
  _37 = _76 << 56;
  _443 = _442 << 56;
  _476 = _475 << 56;

so we get nine(!) copies of it.