[Bug other/59924] [4.9 Regression] uninit memory hog

["jakub at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=59924

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

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2014-01-23
                 CC|                            |davidxl at gcc dot gnu.org,
                   |                            |jakub at gcc dot gnu.org
   Target Milestone|---                         |4.9.0
            Summary|[4.9 Regression] memory hog |[4.9 Regression] uninit
                   |/ build failure with        |memory hog
                   |-Wformat                    |
                   |-Werror=format-security     |
     Ever confirmed|0                           |1

--- Comment #1 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
Started with r206309.  WIth r206308 -O1 -Wall it builds just fine even with
ulimit -v 100000, with r206309 -O1 -Wall it doesn't build even with ulimit -v
2000000.  With that limit, xmalloc_failed is hit on trying to grow
norm_preds vector with 15354379 elements (at least the bunch of elements of
that vector I've looked at were all vectors with a single element).
So, either there is some bug, or we need to cap the size of the normalization
and give up above the limit.