cfg merge part 23 - simple loop analysis code

[Mark Mitchell]

> Question: does your loop unrolling code fix the regression we have
> from 3.0 to 3.1 in e.g. opt/6405?
>
> How about unsharing the loop header for
>   http://gcc.gnu.org/ml/gcc-patches/2002-06/msg01392.html
>
> If the answer for both of these is positive (if with a bit extra
> code), I think we ought to consider merging this anyway.
>
> Mark, what do you think?

I think if we can slide this in a day or two late, that's fine.  There's
nothing magic about a particular day.

On the other hand, if we're looking at weeks (which is what Jan indicated),
I think we need to wait for 3.3.

(My advice to people with stuff on branches is to merge frequently from
the mainline and keep testing; if it's in good shape when we hit stage
1 for GCC 3.3 you'll be able to get your stuff in very quickly at that
point.)

-- 
Mark Mitchell                mark@codesourcery.com
CodeSourcery, LLC            http://www.codesourcery.com