Re: [PATCH v5 3/4] mm: thp: use folio_batch to handle THP splitting in deferred_split_scan()

[Qi Zheng <qi.zheng@linux.dev>]

On 11/6/25 10:52 PM, Wei Yang wrote:
> On Wed, Oct 15, 2025 at 02:35:32PM +0800, Qi Zheng wrote:
>> From: Muchun Song <songmuchun@bytedance.com>
>>
>> The maintenance of the folio->_deferred_list is intricate because it's
>> reused in a local list.
>>
>> Here are some peculiarities:
>>
>>    1) When a folio is removed from its split queue and added to a local
>>       on-stack list in deferred_split_scan(), the ->split_queue_len isn't
>>       updated, leading to an inconsistency between it and the actual
>>       number of folios in the split queue.
>>
>>    2) When the folio is split via split_folio() later, it's removed from
>>       the local list while holding the split queue lock. At this time,
>>       the lock is not needed as it is not protecting anything.
>>
>>    3) To handle the race condition with a third-party freeing or migrating
>>       the preceding folio, we must ensure there's always one safe (with
>>       raised refcount) folio before by delaying its folio_put(). More
>>       details can be found in commit e66f3185fa04 ("mm/thp: fix deferred
>>       split queue not partially_mapped"). It's rather tricky.
>>
>> We can use the folio_batch infrastructure to handle this clearly. In this
>> case, ->split_queue_len will be consistent with the real number of folios
>> in the split queue. If list_empty(&folio->_deferred_list) returns false,
>> it's clear the folio must be in its split queue (not in a local list
>> anymore).
>>
>> In the future, we will reparent LRU folios during memcg offline to
>> eliminate dying memory cgroups, which requires reparenting the split queue
>> to its parent first. So this patch prepares for using
>> folio_split_queue_lock_irqsave() as the memcg may change then.
>>
>> Signed-off-by: Muchun Song <songmuchun@bytedance.com>
>> Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
>> Reviewed-by: Zi Yan <ziy@nvidia.com>
>> Acked-by: David Hildenbrand <david@redhat.com>
>> Acked-by: Shakeel Butt <shakeel.butt@linux.dev>
>> ---
>> mm/huge_memory.c | 87 +++++++++++++++++++++++-------------------------
>> 1 file changed, 41 insertions(+), 46 deletions(-)
>>
>> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
>> index a68f26547cd99..e850bc10da3e2 100644
>> --- a/mm/huge_memory.c
>> +++ b/mm/huge_memory.c
>> @@ -3782,21 +3782,22 @@ static int __folio_split(struct folio *folio, unsigned int new_order,
>> 		struct lruvec *lruvec;
>> 		int expected_refs;
>>
>> -		if (folio_order(folio) > 1 &&
>> -		    !list_empty(&folio->_deferred_list)) {
>> -			ds_queue->split_queue_len--;
>> +		if (folio_order(folio) > 1) {
>> +			if (!list_empty(&folio->_deferred_list)) {
>> +				ds_queue->split_queue_len--;
>> +				/*
>> +				 * Reinitialize page_deferred_list after removing the
>> +				 * page from the split_queue, otherwise a subsequent
>> +				 * split will see list corruption when checking the
>> +				 * page_deferred_list.
>> +				 */
>> +				list_del_init(&folio->_deferred_list);
>> +			}
>> 			if (folio_test_partially_mapped(folio)) {
>> 				folio_clear_partially_mapped(folio);
>> 				mod_mthp_stat(folio_order(folio),
>> 					      MTHP_STAT_NR_ANON_PARTIALLY_MAPPED, -1);
>> 			}
>> -			/*
>> -			 * Reinitialize page_deferred_list after removing the
>> -			 * page from the split_queue, otherwise a subsequent
>> -			 * split will see list corruption when checking the
>> -			 * page_deferred_list.
>> -			 */
>> -			list_del_init(&folio->_deferred_list);
> 
> @Andrew
> 
> Current mm-new looks not merge the code correctly?
> 
> The above removed code is still there.
> 
> @Qi
> 
> After rescan this, I am confused about this code change.
> 
> The difference here is originally it would check/clear partially_mapped if
> folio is on a list. But now we would do this even folio is not on a list.
> 
> If my understanding is correct, after this change, !list_empty() means folio
> is on its ds_queue. And there are total three places to remove it from
> ds_queue.
> 
>    1) __folio_unqueue_deferred_split()
>    2) deferred_split_scan()
>    3) __folio_split()
> 
> In 1) and 2) we all clear partially_mapped bit before removing folio from
> ds_queue, this means if the folio is not on ds_queue in __folio_split(), it is
> not necessary to check/clear partially_mapped bit.

In deferred_split_scan(), if folio_try_get() succeeds, then only the
folio will be removed from ds_queue, but not clear partially_mapped.

> 
> Maybe I missed something, would you mind correct me on this?
>