RFA: redirect_jump_1: fix REG_EQUAL notes

[amylaar@spamcop.net]

> And I still disapprove of this idiom:
>
> >  {
> >!   redirect_exp_1 (&XEXP (note, 0), olabel, nlabel, jump);
> >!   apply_change_group ();
> >  }

P.S.: If the unnecessary calls to insn_invalid_p are what you are worried about,
my patch for PR rtl-optimization/20070:
http://gcc.gnu.org/ml/gcc-patches/2005-02/msg01062.html
splits apply_change_group basically into two functions, verify_changes and
confirm_change_group.  With this infrastructure, the call to apply_change_group
above could be replaced with a call to confirm_change_group.