[Bug tree-optimization/27039] Unable to determine # of iterations for a simple loop

[rguenth at gcc dot gnu dot org]

------- Comment #1 from rguenth at gcc dot gnu dot org  2006-04-05 09:57 -------
Confirmed.  That gives us a testcase at least.

Now, back to the folding problem of

  PTR +- CST  CMP  PTR +- CST

where all of PTR / CST are of pointer type naturally and unsigned usually.
The problem was that the frontends/middle-end introduce pointer overflow via
presenting us with PTR + (unsigned)-CST.  Now, we may argue that if (signed)CST
is positive, that this didn't happen, and we can do the comparison in either
signed or unsigned mode.  Of course this won't help for p - 4 < p + 4, because
this is exactly where the above condition would trigger.

Does this sound reasonable?  Can anyone punch a hole in this argument?


-- 

rguenth at gcc dot gnu dot org changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
     Ever Confirmed|0                           |1
   Last reconfirmed|0000-00-00 00:00:00         |2006-04-05 09:57:15
               date|                            |


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27039