[Bug tree-optimization/58984] [4.8/4.9 Regression] wrong code at -Os and above on x86_64-linux-gnu in 64-bit mode

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=58984

Jakub Jelinek <jakub at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |ASSIGNED
           Assignee|unassigned at gcc dot gnu.org      |jakub at gcc dot gnu.org

--- Comment #6 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Created attachment 31152
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=31152&action=edit
gcc49-pr58984.patch

Untested fix.  As the access size isn't preserved, I'm checking TYPE_SIZE
(TREE_TYPE (v->value)), I believe it should be the same, otherwise we'd have an
invalid store in the caller.  I'll try to bootstrap/regtest an instrumented
version of this to log in the cases where this punted.