[Bug ipa/60306] [4.9 Regression] Incorrect devirtualization "pure virtual method called"

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=60306

--- Comment #7 from Richard Biener <rguenth at gcc dot gnu.org> ---
Of course the bug seems to be

static bool
stmt_may_be_vtbl_ptr_store (gimple stmt)
{
  if (is_gimple_call (stmt))
    return false;
^^^

this.  I remember being very curious when seeing this ;)

Please at this point only fix that bug, not introduce more fancy stuff (we're
long past a stage where that is appropriate).