[Bug middle-end/54563] [4.7/4.8 Regression] ICE in redirect_eh_edge_1, at tree-eh.c:2215

[jakub at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54563

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

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

--- Comment #8 from Jakub Jelinek <jakub at gcc dot gnu.org> 2012-09-14 
19:09:51 UTC ---
Created attachment 28192
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=28192
gcc48-pr54563.patch

Actually it doesn't seem to be too hard to handle it in execute_cse_sincos, the
gsi_replace calls already pass true as last argument, so no need to call
maybe_cleanup_eh_stmt, just gimple_purge_dead_eh_edges needs to be called if
we've changed the last stmt in a bb.  Untested yet.