[Bug tree-optimization/48271] ICE: SIGSEGV in tree_ssa_dominator_optimize (tree-ssa-dom.c:2964) with -O -ftree-vrp -fno-guess-branch-probability -fnon-call-exceptions

[rguenth at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48271



Richard Guenther <rguenth at gcc dot gnu.org> changed:



           What    |Removed                     |Added

----------------------------------------------------------------------------

             Status|UNCONFIRMED                 |ASSIGNED

   Last reconfirmed|                            |2011.03.24 12:34:35

         AssignedTo|unassigned at gcc dot       |rguenth at gcc dot gnu.org

                   |gnu.org                     |

     Ever Confirmed|0                           |1



--- Comment #1 from Richard Guenther <rguenth at gcc dot gnu.org> 2011-03-24 
12:34:35 UTC ---

Program received signal SIGSEGV, Segmentation fault.

0x0000000000d8ad0f in single_succ_p (bb=0x0)

    at /space/rguenther/src/svn/trunk/gcc/basic-block.h:545

545       return EDGE_COUNT (bb->succs) == 1;

(gdb) up

#1  0x0000000000d8a8c9 in tree_ssa_dominator_optimize ()

    at /space/rguenther/src/svn/trunk/gcc/tree-ssa-dom.c:737

737               if (single_succ_p (bb) == 1



so it seems the CFG is not in the same state as it was when setting that

bit.  Which seems sort-of obvious as we thread jumps before doing this

cleanup.  I suppose simply checking for the existance of the block is

enough.



I have such a patch.