Re: [RFC PATCH 3/3] mm: unclutter THP migration

["Zi Yan" <zi.yan@cs.rutgers.edu>]

[-- Attachment #1: Type: text/plain, Size: 1858 bytes --]

On 31 Dec 2017, at 4:07, Michal Hocko wrote:

> On Fri 29-12-17 10:45:46, Zi Yan wrote:
>> On 29 Dec 2017, at 6:36, Michal Hocko wrote:
>>
>>> On Tue 26-12-17 21:19:35, Zi Yan wrote:
> [...]
>>>> And it seems a little bit strange to only re-migrate the head page, then come back to all tail
>>>> pages after migrating the rest of pages in the list “from”. Is it better to split the THP into
>>>> a list other than “from” and insert the list after “page”, then retry from the split “page”?
>>>> Thus, we attempt to migrate all sub pages of the THP after it is split.
>>>
>>> Why does this matter?
>>
>> Functionally, it does not matter.
>>
>> This behavior is just less intuitive and a little different from current one,
>> which implicitly preserves its original order of the not-migrated pages
>> in the “from” list, although no one relies on this implicit behavior now.
>>
>>
>> Adding one line comment about this difference would be good for code maintenance. :)
>
> OK, I will not argue. I still do not see _why_ we need it but I've added
> the following.
>
> diff --git a/mm/migrate.c b/mm/migrate.c
> index 21b3381a2871..0ac5185d3949 100644
> --- a/mm/migrate.c
> +++ b/mm/migrate.c
> @@ -1395,6 +1395,11 @@ int migrate_pages(struct list_head *from, new_page_t get_new_page,
>  				 * allocation could've failed so we should
>  				 * retry on the same page with the THP split
>  				 * to base pages.
> +				 *
> +				 * Head page is retried immediatelly and tail
> +				 * pages are added to the tail of the list so
> +				 * we encounter them after the rest of the list
> +				 * is processed.
>  				 */
>  				if (PageTransHuge(page)) {
>  					lock_page(page);
>
> Does that this reflect what you mean?

s/immediatelly/immediately

Yes. Thanks. :)

—
Best Regards,
Yan Zi

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 557 bytes --]