Re: [PATCH v3 09/15] mm/shmem, swap: remove redundant error handling for replacing folio

[Baolin Wang <baolin.wang@linux.alibaba.com>]

On 2025/9/12 20:36, Kairui Song wrote:
> On Fri, Sep 12, 2025 at 4:22 PM Baolin Wang
> <baolin.wang@linux.alibaba.com> wrote:
>> On 2025/9/11 00:08, Kairui Song wrote:
>>> From: Kairui Song <kasong@tencent.com>
>>>
>>> Shmem may replace a folio in the swap cache if the cached one doesn't
>>> fit the swapin's GFP zone. When doing so, shmem has already double
>>> checked that the swap cache folio is locked, still has the swap cache
>>> flag set, and contains the wanted swap entry. So it is impossible to
>>> fail due to an XArray mismatch. There is even a comment for that.
>>>
>>> Delete the defensive error handling path, and add a WARN_ON instead:
>>> if that happened, something has broken the basic principle of how the
>>> swap cache works, we should catch and fix that.
>>>
>>> Signed-off-by: Kairui Song <kasong@tencent.com>
>>> Reviewed-by: David Hildenbrand <david@redhat.com>
>>> ---
>>>    mm/shmem.c | 42 ++++++++++++------------------------------
>>>    1 file changed, 12 insertions(+), 30 deletions(-)
>>>
>>> diff --git a/mm/shmem.c b/mm/shmem.c
>>> index 410f27bc4752..5f395fab489c 100644
>>> --- a/mm/shmem.c
>>> +++ b/mm/shmem.c
>>> @@ -1661,13 +1661,13 @@ int shmem_writeout(struct folio *folio, struct swap_iocb **plug,
>>>                }
>>>
>>>                /*
>>> -              * The delete_from_swap_cache() below could be left for
>>> +              * The swap_cache_del_folio() below could be left for
>>>                 * shrink_folio_list()'s folio_free_swap() to dispose of;
>>>                 * but I'm a little nervous about letting this folio out of
>>>                 * shmem_writeout() in a hybrid half-tmpfs-half-swap state
>>>                 * e.g. folio_mapping(folio) might give an unexpected answer.
>>>                 */
>>> -             delete_from_swap_cache(folio);
>>> +             swap_cache_del_folio(folio);
>>>                goto redirty;
>>>        }
>>>        if (nr_pages > 1)
>>> @@ -2045,7 +2045,7 @@ static struct folio *shmem_swap_alloc_folio(struct inode *inode,
>>>        new->swap = entry;
>>>
>>>        memcg1_swapin(entry, nr_pages);
>>> -     shadow = get_shadow_from_swap_cache(entry);
>>> +     shadow = swap_cache_get_shadow(entry);
>>
>> Again, there are still some issues with the patch split. The swapcache
>> related APIs replacement should be placed in Patch 8, otherwise there
>> will be buidling errors after applying Patch 8.
>>
>> With this issue fixed:
>> Reviewed-by: Baolin Wang <baolin.wang@linux.alibaba.com>
>>
> 
> Hi Baolin
> 
> Yeah you are right, I need to move these few changes to patch 8.
> 
> BTW I also found that the WARN_ON and irq unlock needs following fix,
> the stats update need to be done with irq disabled:

Yes, indeed, I overlooked this.

> diff --git a/mm/shmem.c b/mm/shmem.c
> index 957e40caba6e..c4d491c93506 100644
> --- a/mm/shmem.c
> +++ b/mm/shmem.c
> @@ -2121,14 +2121,14 @@ static int shmem_replace_folio(struct folio
> **foliop, gfp_t gfp,
>          /* Swap cache still stores N entries instead of a high-order entry */
>          xa_lock_irq(&swap_mapping->i_pages);
>          for (i = 0; i < nr_pages; i++) {
> -               WARN_ON_ONCE(xas_store(&xas, new));
> +               WARN_ON_ONCE(xas_store(&xas, new) != old);
>                  xas_next(&xas);
>          }
> -       xa_unlock_irq(&swap_mapping->i_pages);
> 
>          mem_cgroup_replace_folio(old, new);
>          shmem_update_stats(new, nr_pages);
>          shmem_update_stats(old, -nr_pages);
> +       xa_unlock_irq(&swap_mapping->i_pages);
> 
>          folio_add_lru(new);
>          *foliop = new;