[PATCH 5/5] Account for the slub objects

[Pavel Emelyanov]

[snip]

>> +int slub_alloc_notify(struct kmem_cache *s, void *obj, gfp_t gfp)
>> +{
>> +	struct page *pg;
>> +	struct kmem_container *cnt;
>> +	struct kmem_container **obj_container;
>> +
>> +	pg = virt_to_head_page(obj);
>> +	obj_container = pg->cgroups;
>> +	if (unlikely(obj_container == NULL)) {
>> +		/*
>> +		 * turned on after some objects were allocated
>> +		 */
>> +		if (slub_newpage_notify(s, pg, GFP_ATOMIC) < 0)
>> +			goto err;
>> +
>> +		obj_container = pg->cgroups;
>> +	}
>> +
>> +	rcu_read_lock();
>> +	cnt = task_kmem_container(current);
>> +	if (res_counter_charge(&cnt->res, s->size))
> 
> Is s->size measure in pages or bytes? I suspect it is bytes.

In bytes, of course :) the struct anon_vma (for example) is difficult
to account in pages :P

>> +		goto err_locked;
>> +
>> +	css_get(&cnt->css);
>> +	rcu_read_unlock();
>> +	obj_container[slab_index(obj, s, page_address(pg))] = cnt;
>> +	return 0;
>> +
>> +err_locked:
>> +	rcu_read_unlock();
>> +err:
>> +	return -ENOMEM;
>> +}
>> +
>> +void slub_free_notify(struct kmem_cache *s, void *obj)
>> +{
>> +	struct page *pg;
>> +	struct kmem_container *cnt;
>> +	struct kmem_container **obj_container;
>> +
>> +	pg = virt_to_head_page(obj);
>> +	obj_container = pg->cgroups;
>> +	if (obj_container == NULL)
>> +		return;
>> +
>> +	obj_container += slab_index(obj, s, page_address(pg));
>> +	cnt = *obj_container;
>> +	if (cnt == NULL)
>> +		return;
>> +
>> +	res_counter_uncharge(&cnt->res, s->size);
>> +	*obj_container = NULL;
>> +	css_put(&cnt->css);
>> +}
>> +
> 
> Quick check, slub_free_notify() and slab_alloc_notify() are called
> from serialized contexts, right?

Yup.

>> +int slub_on_notify(struct kmem_cache *cachep)
>> +{
>> +	return (is_kmalloc_cache(cachep) ? -EINVAL : 0);
>> +}
>> +
> 
> I know you've mentioned several times in the comments that
> kmalloc slab's cannot be accounted for, could you please
> add a big comment here as well.

This is useless comment - I will implement the kmalloc accounting
soon, so it will just go away.

> Also, looks like if I turn on notification for kmalloc

You cannot, but let's go on with this assumption :)

> slab's, the listener fails, which will cause all
> allocations to fail?

Nope, just this cache will not generate events and that's it.