Re: [RFC PATCH 3/4] mm/shmem: optimize file read with folio batching

[Chi Zhiling <chizhiling@163.com>]

On 5/18/26 14:57, Baolin Wang wrote:
> On 5/15/26 5:47 PM, Chi Zhiling wrote:
>> From: Chi Zhiling <chizhiling@kylinos.cn>
>>
>> Optimize shmem file read by implementing folio batching in the read
>> iteration path. Only uptodate folios are added to the batch, ensuring
>> all folios in the batch are valid and ready for use without additional
>> checking.
>>
>> Signed-off-by: Chi Zhiling <chizhiling@kylinos.cn>
>> ---
>>   mm/shmem.c | 102 ++++++++++++++++++++++++++++++++++++++++++++++++-----
>>   1 file changed, 93 insertions(+), 9 deletions(-)
>>
>> diff --git a/mm/shmem.c b/mm/shmem.c
>> index 767610f78d0d..4bc4e463ca97 100644
>> --- a/mm/shmem.c
>> +++ b/mm/shmem.c
>> @@ -3348,16 +3348,100 @@ shmem_write_end(const struct kiocb *iocb, 
>> struct address_space *mapping,
>>       return copied;
>>   }
>> +static pgoff_t shmem_get_read_batch(struct address_space *mapping,
>> +        pgoff_t index, pgoff_t max, struct folio_batch *fbatch)
>> +{
>> +    XA_STATE(xas, &mapping->i_pages, index);
>> +    struct folio *folio;
>> +    pgoff_t end = max;
>> +
>> +    rcu_read_lock();
>> +    xas_for_each(&xas, folio, max) {
>> +        if (xas_retry(&xas, folio))
>> +            continue;
>> +        if (xa_is_value(folio)) {
>> +            end = xas.xa_index;
>> +            break;
>> +        }
>> +        if (!folio_try_get(folio))
>> +            goto retry;
>> +
>> +        if (unlikely(folio != xas_reload(&xas)))
>> +            goto put_folio;
>> +
>> +        end = folio_next_index(folio);
>> +
>> +        if (!folio_test_uptodate(folio)) {
>> +            xas_advance(&xas, end - 1);
>> +            folio_put(folio);
>> +            continue;
>> +        }
>> +        if (!folio_batch_add(fbatch, folio))
>> +            break;
>> +        xas_advance(&xas, end - 1);
>> +        continue;
>> +put_folio:
>> +        folio_put(folio);
>> +retry:
>> +        xas_reset(&xas);
>> +    }
>> +    rcu_read_unlock();
>> +
>> +    return end;
>> +}
> 
> I haven't looked at this patch in detail yet, but I feel this part 
> should belong in mm/filemap.c, or perhaps mm/filemap.c already has a 
> similar function implemented? Let's wait for Matthew's comments on this.

Yes, it's better to put it in filemap.c, but seems no other caller 
except shmem. :)


Thanks,