This is the primary patch for cc0 removal on the H8 port.  It doesn't
have any of the optimization work and many patterns are simply disabled
at this time.  It's working well enough to not regress the testsuite.

The H8 is similar to the m68k and other ports in that the vast majority
of instructions clobber the condition codes, including most of the
arithmetic insns that reload needs to use.  While there's a special
adds/subs that does not modify the condition codes, they only accept
constant addends of 1, 2 and 4.  With that in mind, this port does not
expose cc0 until after reload.  So most patterns are defined using
define_insn_and_split.  The splitter adds the appropriate clobbers.  A
subsequent patch will add patterns that set the condition codes and
returns the port to generating roughly equivalent code to before
removing cc0 support.

One might argue that this is an effective rewrite since nearly every
pattern changes and as such I've been a bit more liberal than usual on
the ChangeLog entry.  For the highly mechanical changes, I've just
called out the change for the .md file as a whole rather than every
pattern that got changed.

This is a combination of my son's work as well as my own.

I'll be pushing this to the trunk shortly.  Subsequent patches will be
restoring code quality by using the compare-elim pass to eliminate
unnecessary test/compare insns.

Jeff