Re: owner field in `struct fs'

[Philipp Rumpf]

On Sun, Jun 25, 2000 at 08:51:12PM +1000, Andrew Morton wrote:
> Alan Cox wrote:
> > 
> > > Here it is (it compiles both for UP and SMP, I can't test it on SMP though).
> > >
> > > --- linux/kernel/module.c     Fri Jun 23 20:17:18 2000
> > > +++ linux-prumpf/kernel/module.c      Sat Jun 24 11:53:05 2000
> > 
> > CPU1                            CPU2
> >                                 spinlock(&card->lock)
> > rmmod
> >                 -FREEZE->
> >                                 spin on freeze
> > cleanup_module
> > spinlock(&card->lock)
> > 
> > And the box is dead.
> 
> Except CPU2 shouldn't be diddling with device-specific locks when the
> module refcount is zero.

But it can.  There's no point in disallowing it, is there ?

> Al, the problem we've hit with the netdevice drivers is that they are
> registered and appear in the namespace _when their refcount is zero_. 
> That is, register_netdevice() is called from within the context of the
> module constructor.  (Well, that's just one of the problems..)
> 
> So netdevice methods such as open(), do_ioctl(), set_multicast_list(),
> etc can be and are called when the module refcount is zero.

With the grab-all-CPUs patch, you can arbitrarily call functions with
refcount == 0, as long as they don't go to sleep and 

Not a problem, as long as they increase the module use count before going
to sleep (and are called out of process context).

> I think the 'grab all other CPUs' trick solves a lot of the problems. 
> It seems to be a sensible thing to do when you're unloading kernel text
> anyway.  But there remains the possibility that someone has done a
> shedule() from within the context of a zero-refcount module :(

Yes.  schedule() out of a zero-refcount module is a bug.  In fact, it's
quite easy to make it a BUG() as well:

struct module *lookup_module(unsigned long pc)
{
	struct module *mod;

	for(mod = &kernel_module; mod; mod = mod->next) {
		if((pc >= (unsigned long)mod) && (pc < (unsigned long)mod+mod->size))
			return mod;
	}

	return NULL;
}

in schedule():

if((mod = lookup_module(__builtin_return_address(0))) && (atomic_read(&mod->uc.usecount) == 0))
	BUG();

	Phiilpp