Re: [RFC PATCH 2/4] mm/compaction: optimize >0 order folio compaction with free page split.

[Zi Yan <ziy@nvidia.com>]

[-- Attachment #1: Type: text/plain, Size: 3334 bytes --]

On 18 Sep 2023, at 3:34, Baolin Wang wrote:

> On 9/13/2023 12:28 AM, Zi Yan wrote:
>> From: Zi Yan <ziy@nvidia.com>
>>
>> During migration in a memory compaction, free pages are placed in an array
>> of page lists based on their order. But the desired free page order (i.e.,
>> the order of a source page) might not be always present, thus leading to
>> migration failures. Split a high order free pages when source migration
>> page has a lower order to increase migration successful rate.
>>
>> Note: merging free pages when a migration fails and a lower order free
>> page is returned via compaction_free() is possible, but there is too much
>> work. Since the free pages are not buddy pages, it is hard to identify
>> these free pages using existing PFN-based page merging algorithm.
>>
>> Signed-off-by: Zi Yan <ziy@nvidia.com>
>> ---
>>   mm/compaction.c | 40 +++++++++++++++++++++++++++++++++++++++-
>>   1 file changed, 39 insertions(+), 1 deletion(-)
>>
>> diff --git a/mm/compaction.c b/mm/compaction.c
>> index 868e92e55d27..45747ab5f380 100644
>> --- a/mm/compaction.c
>> +++ b/mm/compaction.c
>> @@ -1801,9 +1801,46 @@ static struct folio *compaction_alloc(struct folio *src, unsigned long data)
>>   	struct compact_control *cc = (struct compact_control *)data;
>>   	struct folio *dst;
>>   	int order = folio_order(src);
>> +	bool has_isolated_pages = false;
>>  +again:
>>   	if (!cc->freepages[order].nr_free) {
>> -		isolate_freepages(cc);
>> +		int i;
>> +
>> +		for (i = order + 1; i <= MAX_ORDER; i++) {
>> +			if (cc->freepages[i].nr_free) {
>> +				struct page *freepage =
>> +					list_first_entry(&cc->freepages[i].pages,
>> +							 struct page, lru);
>> +
>> +				int start_order = i;
>> +				unsigned long size = 1 << start_order;
>> +
>> +				list_del(&freepage->lru);
>> +				cc->freepages[i].nr_free--;
>> +
>> +				while (start_order > order) {
>> +					start_order--;
>> +					size >>= 1;
>> +
>> +					list_add(&freepage[size].lru,
>> +						&cc->freepages[start_order].pages);
>> +					cc->freepages[start_order].nr_free++;
>> +					set_page_private(&freepage[size], start_order);
>
> IIUC, these split pages should also call functions to initialize? e.g. prep_compound_page()?

Not at this place. It is done right below and above "done" label. When free pages
are on cc->freepages, we want to keep them without being post_alloc_hook() or
prep_compound_page() processed for a possible future split. A free page is
only initialized when it is returned by compaction_alloc().

>
>> +				}
>> +				post_alloc_hook(freepage, order, __GFP_MOVABLE);
>> +				if (order)
>> +					prep_compound_page(freepage, order);
>> +				dst = page_folio(freepage);
>> +				goto done;
>> +			}
>> +		}
>> +		if (!has_isolated_pages) {
>> +			isolate_freepages(cc);
>> +			has_isolated_pages = true;
>> +			goto again;
>> +		}
>> +
>>   		if (!cc->freepages[order].nr_free)
>>   			return NULL;
>>   	}
>> @@ -1814,6 +1851,7 @@ static struct folio *compaction_alloc(struct folio *src, unsigned long data)
>>   	post_alloc_hook(&dst->page, order, __GFP_MOVABLE);
>>   	if (order)
>>   		prep_compound_page(&dst->page, order);
>> +done:
>>   	cc->nr_freepages -= 1 << order;
>>   	return dst;
>>   }


--
Best Regards,
Yan, Zi

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