[PATCH v1 07/26] mm: memcontrol: prevent memory cgroup release in get_mem_cgroup_from_folio()

Qi Zheng posted 26 patches 3 months, 2 weeks ago
There is a newer version of this series
[PATCH v1 07/26] mm: memcontrol: prevent memory cgroup release in get_mem_cgroup_from_folio()
Posted by Qi Zheng 3 months, 2 weeks ago
From: Muchun Song <songmuchun@bytedance.com>

In the near future, a folio will no longer pin its corresponding
memory cgroup. To ensure safety, it will only be appropriate to
hold the rcu read lock or acquire a reference to the memory cgroup
returned by folio_memcg(), thereby preventing it from being released.

In the current patch, the rcu read lock is employed to safeguard
against the release of the memory cgroup in get_mem_cgroup_from_folio().

This serves as a preparatory measure for the reparenting of the
LRU pages.

Signed-off-by: Muchun Song <songmuchun@bytedance.com>
Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
---
 mm/memcontrol.c | 11 ++++++++---
 1 file changed, 8 insertions(+), 3 deletions(-)

diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index d484b632c790f..1da3ad77054d3 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -973,14 +973,19 @@ struct mem_cgroup *get_mem_cgroup_from_current(void)
  */
 struct mem_cgroup *get_mem_cgroup_from_folio(struct folio *folio)
 {
-	struct mem_cgroup *memcg = folio_memcg(folio);
+	struct mem_cgroup *memcg;
 
 	if (mem_cgroup_disabled())
 		return NULL;
 
+	if (!folio_memcg_charged(folio))
+		return root_mem_cgroup;
+
 	rcu_read_lock();
-	if (!memcg || WARN_ON_ONCE(!css_tryget(&memcg->css)))
-		memcg = root_mem_cgroup;
+retry:
+	memcg = folio_memcg(folio);
+	if (unlikely(!css_tryget(&memcg->css)))
+		goto retry;
 	rcu_read_unlock();
 	return memcg;
 }
-- 
2.20.1
Re: [PATCH v1 07/26] mm: memcontrol: prevent memory cgroup release in get_mem_cgroup_from_folio()
Posted by Harry Yoo 2 months, 3 weeks ago
On Tue, Oct 28, 2025 at 09:58:20PM +0800, Qi Zheng wrote:
> From: Muchun Song <songmuchun@bytedance.com>
> 
> In the near future, a folio will no longer pin its corresponding
> memory cgroup. To ensure safety, it will only be appropriate to
> hold the rcu read lock or acquire a reference to the memory cgroup
> returned by folio_memcg(), thereby preventing it from being released.
> 
> In the current patch, the rcu read lock is employed to safeguard
> against the release of the memory cgroup in get_mem_cgroup_from_folio().
> 
> This serves as a preparatory measure for the reparenting of the
> LRU pages.
> 
> Signed-off-by: Muchun Song <songmuchun@bytedance.com>
> Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
> ---

This patch looks correct to me, so:
Reviewed-by: Harry Yoo <harry.yoo@oracle.com>

with a few side question on current implementation rather than
this change...

>  mm/memcontrol.c | 11 ++++++++---
>  1 file changed, 8 insertions(+), 3 deletions(-)
> 
> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
> index d484b632c790f..1da3ad77054d3 100644
> --- a/mm/memcontrol.c
> +++ b/mm/memcontrol.c
> @@ -973,14 +973,19 @@ struct mem_cgroup *get_mem_cgroup_from_current(void)
>   */
>  struct mem_cgroup *get_mem_cgroup_from_folio(struct folio *folio)
>  {
> -	struct mem_cgroup *memcg = folio_memcg(folio);
> +	struct mem_cgroup *memcg;
>  
>  	if (mem_cgroup_disabled())
>  		return NULL;
>  
> +	if (!folio_memcg_charged(folio))
> +		return root_mem_cgroup;
> +
>  	rcu_read_lock();
> -	if (!memcg || WARN_ON_ONCE(!css_tryget(&memcg->css)))

I have no idea why we're already calling rcu_read_lock() when a folio pins
the memcg :)

> -		memcg = root_mem_cgroup;
> +retry:
> +	memcg = folio_memcg(folio);
> +	if (unlikely(!css_tryget(&memcg->css)))
> +		goto retry;
>  	rcu_read_unlock();
>  	return memcg;
>  }
> -- 
> 2.20.1

-- 
Cheers,
Harry / Hyeonggon
Re: [PATCH v1 07/26] mm: memcontrol: prevent memory cgroup release in get_mem_cgroup_from_folio()
Posted by Qi Zheng 2 months, 3 weeks ago

On 11/19/25 4:06 PM, Harry Yoo wrote:
> On Tue, Oct 28, 2025 at 09:58:20PM +0800, Qi Zheng wrote:
>> From: Muchun Song <songmuchun@bytedance.com>
>>
>> In the near future, a folio will no longer pin its corresponding
>> memory cgroup. To ensure safety, it will only be appropriate to
>> hold the rcu read lock or acquire a reference to the memory cgroup
>> returned by folio_memcg(), thereby preventing it from being released.
>>
>> In the current patch, the rcu read lock is employed to safeguard
>> against the release of the memory cgroup in get_mem_cgroup_from_folio().
>>
>> This serves as a preparatory measure for the reparenting of the
>> LRU pages.
>>
>> Signed-off-by: Muchun Song <songmuchun@bytedance.com>
>> Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
>> ---
> 
> This patch looks correct to me, so:
> Reviewed-by: Harry Yoo <harry.yoo@oracle.com>
> 
> with a few side question on current implementation rather than
> this change...
> 
>>   mm/memcontrol.c | 11 ++++++++---
>>   1 file changed, 8 insertions(+), 3 deletions(-)
>>
>> diff --git a/mm/memcontrol.c b/mm/memcontrol.c
>> index d484b632c790f..1da3ad77054d3 100644
>> --- a/mm/memcontrol.c
>> +++ b/mm/memcontrol.c
>> @@ -973,14 +973,19 @@ struct mem_cgroup *get_mem_cgroup_from_current(void)
>>    */
>>   struct mem_cgroup *get_mem_cgroup_from_folio(struct folio *folio)
>>   {
>> -	struct mem_cgroup *memcg = folio_memcg(folio);
>> +	struct mem_cgroup *memcg;
>>   
>>   	if (mem_cgroup_disabled())
>>   		return NULL;
>>   
>> +	if (!folio_memcg_charged(folio))
>> +		return root_mem_cgroup;
>> +
>>   	rcu_read_lock();
>> -	if (!memcg || WARN_ON_ONCE(!css_tryget(&memcg->css)))
> 
> I have no idea why we're already calling rcu_read_lock() when a folio pins
> the memcg :)

+1, this was introduced by 
https://github.com/torvalds/linux/commit/f77f0c7514789577125c1b2df145703161736359,
may be some defensive programming.

> 
>> -		memcg = root_mem_cgroup;
>> +retry:
>> +	memcg = folio_memcg(folio);
>> +	if (unlikely(!css_tryget(&memcg->css)))
>> +		goto retry;
>>   	rcu_read_unlock();
>>   	return memcg;
>>   }
>> -- 
>> 2.20.1
>