[Bug tree-optimization/45427] Number of iteration analysis bogus

[rguenth at gcc dot gnu dot org]

------- Comment #6 from rguenth at gcc dot gnu dot org  2010-08-27 17:26 -------
You can see that analysis happening with the C testcase on unpatched trunk
when looking at the cunrolli dump at -O2 for example.

Analyzing # of iterations of loop 1
  exit condition [p_4(D), + , 1](no_overflow) != 0B
  bounds on difference of bases: -18446744073709551615 ... 0
  result:
    # of iterations -(long unsigned int) p_4(D), bounded by 0


-- 


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=45427