[Bug tree-optimization/28029] [4.1 Regression] wrong optimization with -ftree-vectorize

[pinskia at gcc dot gnu dot org]

------- Comment #2 from pinskia at gcc dot gnu dot org  2006-06-14 14:05 -------
The code is basicially the same as:
  void multiply(float *data_, const float *op, unsigned int size_)
    {
      for (unsigned int i=0; i<size_; ++i)
        data_[i] *= op[i];
    }

And what happens is op is data_ + 3 and size_ is 6, we will get the wrong
answer as there will be no feedback in the loop.

Anyways this is a 4.1 bug fixed already in 4.2.0


-- 

pinskia at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |pinskia at gcc dot gnu dot
                   |                            |org
  GCC build triplet|i686-pc-linux-gnu           |
   GCC host triplet|i686-pc-linux-gnu           |
 GCC target triplet|i686-pc-linux-gnu           |
           Keywords|                            |wrong-code
            Summary|[4.2 regression] missed     |[4.1 Regression] wrong
                   |optimization with -ftree-   |optimization with -ftree-
                   |vectorize                   |vectorize
   Target Milestone|---                         |4.1.2
            Version|4.2.0                       |4.1.0


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28029