Re: [bug report] mm: huge_memory: enable debugfs to split huge pages to any order

[Zi Yan <ziy@nvidia.com>]

[-- Attachment #1: Type: text/plain, Size: 1262 bytes --]

On 7 Mar 2024, at 9:31, Dan Carpenter wrote:

> On Thu, Mar 07, 2024 at 09:20:09AM -0500, Zi Yan wrote:
>>
>> Can you check if the patch below fixes the issue? I checked the inputs from
>> debugfs and also inside split_huge_page_to_list_to_order().
>>
>> diff --git a/mm/huge_memory.c b/mm/huge_memory.c
>> index a81a09236c16..4d21e57a7d07 100644
>> --- a/mm/huge_memory.c
>> +++ b/mm/huge_memory.c
>> @@ -3052,6 +3052,9 @@ int split_huge_page_to_list_to_order(struct page *page, struct list_head *list,
>>         VM_BUG_ON_FOLIO(!folio_test_locked(folio), folio);
>>         VM_BUG_ON_FOLIO(!folio_test_large(folio), folio);
>>
>> +       if (new_order >= folio_order(folio))
>> +               return -EINVAL;
>> +
>>         /* Cannot split anonymous THP to order-1 */
>>         if (new_order == 1 && folio_test_anon(folio)) {
>>                 VM_WARN_ONCE(1, "Cannot split to order-1 folio");
>
> From a static analysis perspective I think just this is sufficient?  But
> I don't really know the code and don't object to the other parts of
> this path either.

You are right. The other two changes are in debugfs to avoid unnecessary
folio locking operations. They are more like an optimization.

--
Best Regards,
Yan, Zi

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 854 bytes --]