On Sat, Oct 12, 2013 at 12:53:36AM -0700, Paul E. McKenney wrote:
> On Sat, Oct 12, 2013 at 04:25:08AM +0200, Hannes Frederic Sowa wrote:
> > I saw your patch regarding making rcu_assign_pointer volatile and wonder if
> > we
> > can still make it a bit more safe to use if we force the evaluation of the
> > to-be-assigned pointer before the write barrier. This is what I have in
> > mind:
> >
> > diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h
> > index f1f1bc3..79eccc3 100644
> > --- a/include/linux/rcupdate.h
> > +++ b/include/linux/rcupdate.h
> > @@ -550,8 +550,9 @@ static inline void rcu_preempt_sleep_check(void)
> > })
> > #define __rcu_assign_pointer(p, v, space) \
> > do { \
> > + typeof(v) ___v = (v); \
> > smp_wmb(); \
> > - (p) = (typeof(*v) __force space *)(v); \
> > + (p) = (typeof(*___v) __force space *)(___v); \
> > } while (0)
> >
> >
> > I don't think ___v must be volatile for this case because the memory barrier
> > will force the evaluation of v first.
> >
> > This would guard against cases where rcu_assign_pointer is used like:
> >
> > rcu_assign_pointer(ptr, compute_ptr_with_side_effects());
>
> I am sorry, but I am not seeing how this would be particularly useful.
Oh, I do not think it is useful. It should help enforcing the right memory
orderings if someone uses rcu_assign_pointer in such a nasty way.
It was meant by me as a "defensive rearragement" to guard rcu_assign_pointer
to accidentaly slip memory mutations across the barrier without hurting
compiler optimizations too bad (or at all).
> The point of rcu_assign_pointer() is to order the initialization of
> a data structure against publishing a pointer to that data structure.
> An example may be found in cgroup_create():
>
> name = cgroup_alloc_name(dentry);
> if (!name)
> goto err_free_cgrp;
> rcu_assign_pointer(cgrp->name, name);
>
> Here, cgroup_alloc_name() allocates memory for the name and fills in
> the name:
>
> static struct cgroup_name *cgroup_alloc_name(struct dentry *dentry)
> {
> struct cgroup_name *name;
>
> name = kmalloc(sizeof(*name) + dentry->d_name.len + 1, GFP_KERNEL);
> if (!name)
> return NULL;
> strcpy(name->name, dentry->d_name.name);
> return name;
> }
>
> So the point of the smp_wmb() in __rcu_assign_pointer() is to order the
> strcpy() in cgroup_alloc_name() to happen before the assignment of the
> name pointer to cgrp->name.
>
> To make this example fit your pattern, we could change the code in
> cgroup_create() to look as follows (and to be buggy):
>
> /* BAD CODE! Do not do this! */
> rcu_assign_pointer(cgrp->name, cgroup_alloc_name(dentry));
> if (!cgrp->name)
> goto err_free_cgrp;
>
> The reason that this is bad practice is that it is hiding the fact that
> the allocation and initialization in cgroup_alloc_name() needs to be
> ordered before the assignment to cgrp->name.
>
> Make sense?
Absolutely! But e.g. the pointer could be preallocated and no length
checks are needed so there is no need for an error path, I guess someone
could think it is safe to put the get_ptr_with_side_effects (and if the
side effect is only a bit flip in the strucutre pointed to by the value)
in the v argument of rcu_assign_pointer.
I also tought about adding a (*(&(v))) statement to enfore that only
lvalues where allowed as the value in rcu_assign_pointer, but that would
already cause compilation errors (e.g. rcu_assign_pointer(ptr, ptr|MARK)).
The nice thing with ACCESS_ONCE(p) is, that there is now no way to put
a non-lvalue expression into the first argument of rcu_assign_pointer. :)
Regarding the volatile access, I hope that the C11 memory model
and enhancements to the compiler will some day provide a better
way to express the semantics of what is tried to express here
(__atomic_store_n/__atomic_load_n with the accompanied memory model,
which could be even weaker to what a volatile access would enfore
now and could guarantee atomic stores/loads).
Greeting,
Hannes
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
