[Bug c/51015] [4.7 Regression] ice in vect_determine_vectorization_factor with -O3

[dominiq at lps dot ens.fr]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=51015

Dominique d'Humieres <dominiq at lps dot ens.fr> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2011-11-07
                 CC|                            |ira at gcc dot gnu.org,
                   |                            |jakub at redhat dot com
            Summary|ice in                      |[4.7 Regression] ice in
                   |vect_determine_vectorizatio |vect_determine_vectorizatio
                   |n_factor with -O3           |n_factor with -O3
     Ever Confirmed|0                           |1

--- Comment #1 from Dominique d'Humieres <dominiq at lps dot ens.fr> 2011-11-07 
23:48:01 UTC ---
Reduced test:

typedef unsigned long long __u64;
static __u64 ext2_max_sizes[16 -
       10 + 1];

void e2fsck_pass1()
{
 int i;
 __u64 max_sizes;

 for (i = 10; i <= 16; i++) {
  max_sizes = 12 + (1ULL << ((i) - 2));
  max_sizes = max_sizes + (1ULL << ((i) - 2)) * (1ULL << ((i) - 2));
  max_sizes = max_sizes + (1ULL << ((i) - 2)) * (1ULL << ((i) - 2)) * (1ULL <<
((i) - 2));
  ext2_max_sizes[i - 10] = max_sizes;
 }

}

r179960 is OK
r180683 gives the ICE