Re: [PATCH v2] mm/memcg: use exist interface to get css from memcg

[Michal Hocko <mhocko-AlSwsSmVLrQ@public.gmane.org>]

On Wed 18-07-12 19:08:54, Wanpeng Li wrote:
> use exist interface mem_cgroup_css instead of &mem->css.

This interface has been added to enable mem->css outside of
mm/memcontrol.c (where we define struct mem_cgroup). There is one user
left (hwpoison_filter_task) after recent clean ups.

I think we shouldn't spread the usage inside the mm/memcontrol.c. The
compiler inlines the function for all callers added by this patch but I
wouldn't rely on it. It is also unfortunate that we cannot convert all
dereferences (e.g. const mem_cgroup).
Moreover it doesn't add any additional type safety. So I would vote for
not taking the patch but if others like it I will not block it.

> V2->V1:
> * restore the mistake modify in mem_cgroup_css
> 
> Signed-off-by: Wanpeng Li <liwanp-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
> ---
>  mm/memcontrol.c |   89 ++++++++++++++++++++++++++++--------------------------
>  1 files changed, 46 insertions(+), 43 deletions(-)
> 
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index f72b5e5..f27084c 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -645,7 +645,7 @@ retry:
>  	 */
>  	__mem_cgroup_remove_exceeded(mz->memcg, mz, mctz);
>  	if (!res_counter_soft_limit_excess(&mz->memcg->res) ||
> -		!css_tryget(&mz->memcg->css))
> +		!css_tryget(mem_cgroup_css(mz->memcg)))
>  		goto retry;
>  done:
>  	return mz;
> @@ -899,7 +899,7 @@ struct mem_cgroup *try_get_mem_cgroup_from_mm(struct mm_struct *mm)
>  		memcg = mem_cgroup_from_task(rcu_dereference(mm->owner));
>  		if (unlikely(!memcg))
>  			break;
> -	} while (!css_tryget(&memcg->css));
> +	} while (!css_tryget(mem_cgroup_css(memcg)));
>  	rcu_read_unlock();
>  	return memcg;
>  }
> @@ -935,10 +935,10 @@ struct mem_cgroup *mem_cgroup_iter(struct mem_cgroup *root,
>  		root = root_mem_cgroup;
>  
>  	if (prev && !reclaim)
> -		id = css_id(&prev->css);
> +		id = css_id(mem_cgroup_css(prev));
>  
>  	if (prev && prev != root)
> -		css_put(&prev->css);
> +		css_put(mem_cgroup_css(prev));
>  
>  	if (!root->use_hierarchy && root != root_mem_cgroup) {
>  		if (prev)
> @@ -963,9 +963,10 @@ struct mem_cgroup *mem_cgroup_iter(struct mem_cgroup *root,
>  		}
>  
>  		rcu_read_lock();
> -		css = css_get_next(&mem_cgroup_subsys, id + 1, &root->css, &id);
> +		css = css_get_next(&mem_cgroup_subsys, id + 1,
> +					mem_cgroup_css(root), &id);
>  		if (css) {
> -			if (css == &root->css || css_tryget(css))
> +			if (css == mem_cgroup_css(root) || css_tryget(css))
>  				memcg = container_of(css,
>  						     struct mem_cgroup, css);
>  		} else
> @@ -997,7 +998,7 @@ void mem_cgroup_iter_break(struct mem_cgroup *root,
>  	if (!root)
>  		root = root_mem_cgroup;
>  	if (prev && prev != root)
> -		css_put(&prev->css);
> +		css_put(mem_cgroup_css(prev));
>  }
>  
>  /*
> @@ -1150,7 +1151,8 @@ bool __mem_cgroup_same_or_subtree(const struct mem_cgroup *root_memcg,
>  		return true;
>  	if (!root_memcg->use_hierarchy || !memcg)
>  		return false;
> -	return css_is_ancestor(&memcg->css, &root_memcg->css);
> +	return css_is_ancestor(mem_cgroup_css(memcg),
> +				&root_memcg->css);
>  }
>  
>  static bool mem_cgroup_same_or_subtree(const struct mem_cgroup *root_memcg,
> @@ -1183,7 +1185,7 @@ int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *memcg)
>  		task_lock(task);
>  		curr = mem_cgroup_from_task(task);
>  		if (curr)
> -			css_get(&curr->css);
> +			css_get(mem_cgroup_css(curr));
>  		task_unlock(task);
>  	}
>  	if (!curr)
> @@ -1195,7 +1197,7 @@ int task_in_mem_cgroup(struct task_struct *task, const struct mem_cgroup *memcg)
>  	 * hierarchy(even if use_hierarchy is disabled in "memcg").
>  	 */
>  	ret = mem_cgroup_same_or_subtree(memcg, curr);
> -	css_put(&curr->css);
> +	css_put(mem_cgroup_css(curr));
>  	return ret;
>  }
>  
> @@ -1251,7 +1253,7 @@ static unsigned long mem_cgroup_margin(struct mem_cgroup *memcg)
>  
>  int mem_cgroup_swappiness(struct mem_cgroup *memcg)
>  {
> -	struct cgroup *cgrp = memcg->css.cgroup;
> +	struct cgroup *cgrp = mem_cgroup_css(memcg)->cgroup;
>  
>  	/* root ? */
>  	if (cgrp->parent == NULL)
> @@ -1396,7 +1398,7 @@ void mem_cgroup_print_oom_info(struct mem_cgroup *memcg, struct task_struct *p)
>  
>  	rcu_read_lock();
>  
> -	mem_cgrp = memcg->css.cgroup;
> +	mem_cgrp = mem_cgroup_css(memcg)->cgroup;
>  	task_cgrp = task_cgroup(p, mem_cgroup_subsys_id);
>  
>  	ret = cgroup_path(task_cgrp, memcg_name, PATH_MAX);
> @@ -2275,12 +2277,12 @@ static int __mem_cgroup_try_charge(struct mm_struct *mm,
>  again:
>  	if (*ptr) { /* css should be a valid one */
>  		memcg = *ptr;
> -		VM_BUG_ON(css_is_removed(&memcg->css));
> +		VM_BUG_ON(css_is_removed(mem_cgroup_css(memcg)));
>  		if (mem_cgroup_is_root(memcg))
>  			goto done;
>  		if (nr_pages == 1 && consume_stock(memcg))
>  			goto done;
> -		css_get(&memcg->css);
> +		css_get(mem_cgroup_css(memcg));
>  	} else {
>  		struct task_struct *p;
>  
> @@ -2316,7 +2318,7 @@ again:
>  			goto done;
>  		}
>  		/* after here, we may be blocked. we need to get refcnt */
> -		if (!css_tryget(&memcg->css)) {
> +		if (!css_tryget(mem_cgroup_css(memcg))) {
>  			rcu_read_unlock();
>  			goto again;
>  		}
> @@ -2328,7 +2330,7 @@ again:
>  
>  		/* If killed, bypass charge */
>  		if (fatal_signal_pending(current)) {
> -			css_put(&memcg->css);
> +			css_put(mem_cgroup_css(memcg));
>  			goto bypass;
>  		}
>  
> @@ -2344,29 +2346,29 @@ again:
>  			break;
>  		case CHARGE_RETRY: /* not in OOM situation but retry */
>  			batch = nr_pages;
> -			css_put(&memcg->css);
> +			css_put(mem_cgroup_css(memcg));
>  			memcg = NULL;
>  			goto again;
>  		case CHARGE_WOULDBLOCK: /* !__GFP_WAIT */
> -			css_put(&memcg->css);
> +			css_put(mem_cgroup_css(memcg));
>  			goto nomem;
>  		case CHARGE_NOMEM: /* OOM routine works */
>  			if (!oom) {
> -				css_put(&memcg->css);
> +				css_put(mem_cgroup_css(memcg));
>  				goto nomem;
>  			}
>  			/* If oom, we never return -ENOMEM */
>  			nr_oom_retries--;
>  			break;
>  		case CHARGE_OOM_DIE: /* Killed by OOM Killer */
> -			css_put(&memcg->css);
> +			css_put(mem_cgroup_css(memcg));
>  			goto bypass;
>  		}
>  	} while (ret != CHARGE_OK);
>  
>  	if (batch > nr_pages)
>  		refill_stock(memcg, batch - nr_pages);
> -	css_put(&memcg->css);
> +	css_put(mem_cgroup_css(memcg));
>  done:
>  	*ptr = memcg;
>  	return 0;
> @@ -2445,14 +2447,14 @@ struct mem_cgroup *try_get_mem_cgroup_from_page(struct page *page)
>  	lock_page_cgroup(pc);
>  	if (PageCgroupUsed(pc)) {
>  		memcg = pc->mem_cgroup;
> -		if (memcg && !css_tryget(&memcg->css))
> +		if (memcg && !css_tryget(mem_cgroup_css(memcg)))
>  			memcg = NULL;
>  	} else if (PageSwapCache(page)) {
>  		ent.val = page_private(page);
>  		id = lookup_swap_cgroup_id(ent);
>  		rcu_read_lock();
>  		memcg = mem_cgroup_lookup(id);
> -		if (memcg && !css_tryget(&memcg->css))
> +		if (memcg && !css_tryget(mem_cgroup_css(memcg)))
>  			memcg = NULL;
>  		rcu_read_unlock();
>  	}
> @@ -2795,7 +2797,7 @@ int mem_cgroup_try_charge_swapin(struct mm_struct *mm,
>  		goto charge_cur_mm;
>  	*memcgp = memcg;
>  	ret = __mem_cgroup_try_charge(NULL, mask, 1, memcgp, true);
> -	css_put(&memcg->css);
> +	css_put(mem_cgroup_css(memcg));
>  	if (ret == -EINTR)
>  		ret = 0;
>  	return ret;
> @@ -2816,7 +2818,7 @@ __mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *memcg,
>  		return;
>  	if (!memcg)
>  		return;
> -	cgroup_exclude_rmdir(&memcg->css);
> +	cgroup_exclude_rmdir(mem_cgroup_css(memcg));
>  
>  	__mem_cgroup_commit_charge(memcg, page, 1, ctype, true);
>  	/*
> @@ -2835,7 +2837,7 @@ __mem_cgroup_commit_charge_swapin(struct page *page, struct mem_cgroup *memcg,
>  	 * So, rmdir()->pre_destroy() can be called while we do this charge.
>  	 * In that case, we need to call pre_destroy() again. check it here.
>  	 */
> -	cgroup_release_and_wakeup_rmdir(&memcg->css);
> +	cgroup_release_and_wakeup_rmdir(mem_cgroup_css(memcg));
>  }
>  
>  void mem_cgroup_commit_charge_swapin(struct page *page,
> @@ -3083,7 +3085,7 @@ mem_cgroup_uncharge_swapcache(struct page *page, swp_entry_t ent, bool swapout)
>  	 * mem_cgroup_get() was called in uncharge().
>  	 */
>  	if (do_swap_account && swapout && memcg)
> -		swap_cgroup_record(ent, css_id(&memcg->css));
> +		swap_cgroup_record(ent, css_id(mem_cgroup_css(memcg)));
>  }
>  #endif
>  
> @@ -3135,8 +3137,8 @@ static int mem_cgroup_move_swap_account(swp_entry_t entry,
>  {
>  	unsigned short old_id, new_id;
>  
> -	old_id = css_id(&from->css);
> -	new_id = css_id(&to->css);
> +	old_id = css_id(mem_cgroup_css(from));
> +	new_id = css_id(mem_cgroup_css(to));
>  
>  	if (swap_cgroup_cmpxchg(entry, old_id, new_id) == old_id) {
>  		mem_cgroup_swap_statistics(from, false);
> @@ -3184,7 +3186,7 @@ int mem_cgroup_prepare_migration(struct page *page,
>  	lock_page_cgroup(pc);
>  	if (PageCgroupUsed(pc)) {
>  		memcg = pc->mem_cgroup;
> -		css_get(&memcg->css);
> +		css_get(mem_cgroup_css(memcg));
>  		/*
>  		 * At migrating an anonymous page, its mapcount goes down
>  		 * to 0 and uncharge() will be called. But, even if it's fully
> @@ -3227,7 +3229,7 @@ int mem_cgroup_prepare_migration(struct page *page,
>  
>  	*memcgp = memcg;
>  	ret = __mem_cgroup_try_charge(NULL, gfp_mask, 1, memcgp, false);
> -	css_put(&memcg->css);/* drop extra refcnt */
> +	css_put(mem_cgroup_css(memcg));/* drop extra refcnt */
>  	if (ret) {
>  		if (PageAnon(page)) {
>  			lock_page_cgroup(pc);
> @@ -3268,7 +3270,7 @@ void mem_cgroup_end_migration(struct mem_cgroup *memcg,
>  	if (!memcg)
>  		return;
>  	/* blocks rmdir() */
> -	cgroup_exclude_rmdir(&memcg->css);
> +	cgroup_exclude_rmdir(mem_cgroup_css(memcg));
>  	if (!migration_ok) {
>  		used = oldpage;
>  		unused = newpage;
> @@ -3306,7 +3308,7 @@ void mem_cgroup_end_migration(struct mem_cgroup *memcg,
>  	 * So, rmdir()->pre_destroy() can be called while we do this charge.
>  	 * In that case, we need to call pre_destroy() again. check it here.
>  	 */
> -	cgroup_release_and_wakeup_rmdir(&memcg->css);
> +	cgroup_release_and_wakeup_rmdir(mem_cgroup_css(memcg));
>  }
>  
>  /*
> @@ -3575,7 +3577,7 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
>  				next_mz =
>  				__mem_cgroup_largest_soft_limit_node(mctz);
>  				if (next_mz == mz)
> -					css_put(&next_mz->memcg->css);
> +					css_put(mem_cgroup_css(next_mz->memcg));
>  				else /* next_mz == NULL or other memcg */
>  					break;
>  			} while (1);
> @@ -3593,7 +3595,7 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
>  		/* If excess == 0, no tree ops */
>  		__mem_cgroup_insert_exceeded(mz->memcg, mz, mctz, excess);
>  		spin_unlock(&mctz->lock);
> -		css_put(&mz->memcg->css);
> +		css_put(mem_cgroup_css(mz->memcg));
>  		loop++;
>  		/*
>  		 * Could not reclaim anything and there are no more
> @@ -3606,7 +3608,7 @@ unsigned long mem_cgroup_soft_limit_reclaim(struct zone *zone, int order,
>  			break;
>  	} while (!nr_reclaimed);
>  	if (next_mz)
> -		css_put(&next_mz->memcg->css);
> +		css_put(mem_cgroup_css(next_mz->memcg));
>  	return nr_reclaimed;
>  }
>  
> @@ -3679,9 +3681,9 @@ static int mem_cgroup_force_empty(struct mem_cgroup *memcg, bool free_all)
>  	int ret;
>  	int node, zid, shrink;
>  	int nr_retries = MEM_CGROUP_RECLAIM_RETRIES;
> -	struct cgroup *cgrp = memcg->css.cgroup;
> +	struct cgroup *cgrp = mem_cgroup_css(memcg)->cgroup;
>  
> -	css_get(&memcg->css);
> +	css_get(mem_cgroup_css(memcg));
>  
>  	shrink = 0;
>  	/* should free all ? */
> @@ -3722,7 +3724,7 @@ move_account:
>  	/* "ret" should also be checked to ensure all lists are empty. */
>  	} while (res_counter_read_u64(&memcg->res, RES_USAGE) > 0 || ret);
>  out:
> -	css_put(&memcg->css);
> +	css_put(mem_cgroup_css(memcg));
>  	return ret;
>  
>  try_to_free:
> @@ -3933,7 +3935,7 @@ static void memcg_get_hierarchical_limit(struct mem_cgroup *memcg,
>  
>  	min_limit = res_counter_read_u64(&memcg->res, RES_LIMIT);
>  	min_memsw_limit = res_counter_read_u64(&memcg->memsw, RES_LIMIT);
> -	cgroup = memcg->css.cgroup;
> +	cgroup = mem_cgroup_css(memcg)->cgroup;
>  	if (!memcg->use_hierarchy)
>  		goto out;
>  
> @@ -4770,7 +4772,7 @@ static void __mem_cgroup_free(struct mem_cgroup *memcg)
>  	int node;
>  
>  	mem_cgroup_remove_from_trees(memcg);
> -	free_css_id(&mem_cgroup_subsys, &memcg->css);
> +	free_css_id(&mem_cgroup_subsys, mem_cgroup_css(memcg));
>  
>  	for_each_node(node)
>  		free_mem_cgroup_per_zone_info(memcg, node);
> @@ -4926,7 +4928,7 @@ mem_cgroup_create(struct cgroup *cont)
>  		mem_cgroup_put(memcg);
>  		return ERR_PTR(error);
>  	}
> -	return &memcg->css;
> +	return mem_cgroup_css(memcg);
>  free_out:
>  	__mem_cgroup_free(memcg);
>  	return ERR_PTR(error);
> @@ -5145,7 +5147,8 @@ static enum mc_target_type get_mctgt_type(struct vm_area_struct *vma,
>  	}
>  	/* There is a swap entry and a page doesn't exist or isn't charged */
>  	if (ent.val && !ret &&
> -			css_id(&mc.from->css) == lookup_swap_cgroup_id(ent)) {
> +			css_id(mem_cgroup_css(mc.from)) ==
> +					lookup_swap_cgroup_id(ent)) {
>  		ret = MC_TARGET_SWAP;
>  		if (target)
>  			target->ent = ent;
> -- 
> 1.7.7.6
> 

-- 
Michal Hocko
SUSE Labs
SUSE LINUX s.r.o.
Lihovarska 1060/12
190 00 Praha 9    
Czech Republic