Re: [PATCH V2] mm/gup: Clear the LRU flag of a page before adding to LRU batch

[Ge Yang <yangge1116@126.com>]

在 2024/7/29 22:04, David Hildenbrand 写道:
> On 22.06.24 08:48, yangge1116@126.com wrote:
>> From: yangge <yangge1116@126.com>
>>
>> If a large number of CMA memory are configured in system (for example, 
>> the
>> CMA memory accounts for 50% of the system memory), starting a virtual
>> virtual machine, it will call pin_user_pages_remote(..., FOLL_LONGTERM,
>> ...) to pin memory.  Normally if a page is present and in CMA area,
>> pin_user_pages_remote() will migrate the page from CMA area to non-CMA
>> area because of FOLL_LONGTERM flag. But the current code will cause the
>> migration failure due to unexpected page refcounts, and eventually cause
>> the virtual machine fail to start.
>>
>> If a page is added in LRU batch, its refcount increases one, remove the
>> page from LRU batch decreases one. Page migration requires the page is 
>> not
>> referenced by others except page mapping. Before migrating a page, we
>> should try to drain the page from LRU batch in case the page is in it,
>> however, folio_test_lru() is not sufficient to tell whether the page is
>> in LRU batch or not, if the page is in LRU batch, the migration will 
>> fail.
>>
>> To solve the problem above, we modify the logic of adding to LRU batch.
>> Before adding a page to LRU batch, we clear the LRU flag of the page so
>> that we can check whether the page is in LRU batch by 
>> folio_test_lru(page).
>> Seems making the LRU flag of the page invisible a long time is no 
>> problem,
>> because a new page is allocated from buddy and added to the lru batch,
>> its LRU flag is also not visible for a long time.
>>
>> Cc: <stable@vger.kernel.org>
>> Signed-off-by: yangge <yangge1116@126.com>
>> ---
>>   mm/swap.c | 43 +++++++++++++++++++++++++++++++------------
>>   1 file changed, 31 insertions(+), 12 deletions(-)
>>
>> diff --git a/mm/swap.c b/mm/swap.c
>> index dc205bd..9caf6b0 100644
>> --- a/mm/swap.c
>> +++ b/mm/swap.c
>> @@ -211,10 +211,6 @@ static void folio_batch_move_lru(struct 
>> folio_batch *fbatch, move_fn_t move_fn)
>>       for (i = 0; i < folio_batch_count(fbatch); i++) {
>>           struct folio *folio = fbatch->folios[i];
>> -        /* block memcg migration while the folio moves between lru */
>> -        if (move_fn != lru_add_fn && !folio_test_clear_lru(folio))
>> -            continue;
>> -
>>           folio_lruvec_relock_irqsave(folio, &lruvec, &flags);
>>           move_fn(lruvec, folio);
>> @@ -255,11 +251,16 @@ static void lru_move_tail_fn(struct lruvec 
>> *lruvec, struct folio *folio)
>>   void folio_rotate_reclaimable(struct folio *folio)
>>   {
>>       if (!folio_test_locked(folio) && !folio_test_dirty(folio) &&
>> -        !folio_test_unevictable(folio) && folio_test_lru(folio)) {
>> +        !folio_test_unevictable(folio)) {
>>           struct folio_batch *fbatch;
>>           unsigned long flags;
>>           folio_get(folio);
>> +        if (!folio_test_clear_lru(folio)) {
>> +            folio_put(folio);
>> +            return;
>> +        }
>> +
>>           local_lock_irqsave(&lru_rotate.lock, flags);
>>           fbatch = this_cpu_ptr(&lru_rotate.fbatch);
>>           folio_batch_add_and_move(fbatch, folio, lru_move_tail_fn);
>> @@ -352,11 +353,15 @@ static void folio_activate_drain(int cpu)
>>   void folio_activate(struct folio *folio)
>>   {
>> -    if (folio_test_lru(folio) && !folio_test_active(folio) &&
>> -        !folio_test_unevictable(folio)) {
>> +    if (!folio_test_active(folio) && !folio_test_unevictable(folio)) {
>>           struct folio_batch *fbatch;
>>           folio_get(folio);
>> +        if (!folio_test_clear_lru(folio)) {
>> +            folio_put(folio);
>> +            return;
>> +        }
>> +
>>           local_lock(&cpu_fbatches.lock);
>>           fbatch = this_cpu_ptr(&cpu_fbatches.activate);
>>           folio_batch_add_and_move(fbatch, folio, folio_activate_fn);
>> @@ -700,6 +705,11 @@ void deactivate_file_folio(struct folio *folio)
>>           return;
>>       folio_get(folio);
>> +    if (!folio_test_clear_lru(folio)) {
>> +        folio_put(folio);
>> +        return;
>> +    }
>> +
>>       local_lock(&cpu_fbatches.lock);
>>       fbatch = this_cpu_ptr(&cpu_fbatches.lru_deactivate_file);
>>       folio_batch_add_and_move(fbatch, folio, lru_deactivate_file_fn);
>> @@ -716,11 +726,16 @@ void deactivate_file_folio(struct folio *folio)
>>    */
>>   void folio_deactivate(struct folio *folio)
>>   {
>> -    if (folio_test_lru(folio) && !folio_test_unevictable(folio) &&
>> -        (folio_test_active(folio) || lru_gen_enabled())) {
>> +    if (!folio_test_unevictable(folio) && (folio_test_active(folio) ||
>> +        lru_gen_enabled())) {
>>           struct folio_batch *fbatch;
>>           folio_get(folio);
>> +        if (!folio_test_clear_lru(folio)) {
>> +            folio_put(folio);
>> +            return;
>> +        }
>> +
>>           local_lock(&cpu_fbatches.lock);
>>           fbatch = this_cpu_ptr(&cpu_fbatches.lru_deactivate);
>>           folio_batch_add_and_move(fbatch, folio, lru_deactivate_fn);
>> @@ -737,12 +752,16 @@ void folio_deactivate(struct folio *folio)
>>    */
>>   void folio_mark_lazyfree(struct folio *folio)
>>   {
>> -    if (folio_test_lru(folio) && folio_test_anon(folio) &&
>> -        folio_test_swapbacked(folio) && !folio_test_swapcache(folio) &&
>> -        !folio_test_unevictable(folio)) {
>> +    if (folio_test_anon(folio) && folio_test_swapbacked(folio) &&
>> +        !folio_test_swapcache(folio) && 
>> !folio_test_unevictable(folio)) {
>>           struct folio_batch *fbatch;
>>           folio_get(folio);
>> +        if (!folio_test_clear_lru(folio)) {
>> +            folio_put(folio);
>> +            return;
>> +        }
> 
> Looking at this in more detail, I wonder if we can turn that to
> 
> if (!folio_test_clear_lru(folio))
>      return;
> folio_get(folio);
> 
> In all cases? The caller must hold a reference, so this should be fine.
> 

Seems the caller madvise_free_pte_range(...), calling 
folio_mark_lazyfree(...), doesn't hold a reference on folio.