On Fri, 2007-03-16 at 12:49 -0800, Andrew Morton wrote:
> > Ok, I've just done a few tests with the attached patch. It basically
> > creates a freeze_machine_run function that is equivalent in interface to
> > stop_machine_run, but uses freeze_processes / thaw_processes to stop the
> > machine.
I've never been convinced that the freezer was a good idea.
stop_machine is a damn big hammer, but it works.