Re: [PATCH] mm, compaction: Skip all pinned pages during scan

[Khalid Aziz <khalid.aziz@oracle.com>]

On 5/11/23 21:45, Andrew Morton wrote:
> On Thu, 11 May 2023 10:55:16 -0600 Khalid Aziz <khalid.aziz@oracle.com> wrote:
> 
>> Pinned pages can not be migrated. Currently as
>> isolate_migratepages_block() scans pages for compaction, it skips
>> any pinned anonymous pages. All pinned pages should be skipped and
>> not just the anonymous pinned pages. This patch adds a check for
>> pinned page by comparing its refcount with mapcount and accounts for
>> possible extra refcounts. This was seen as a real issue on a
>> customer workload where a large number of pages were pinned by vfio
>> on the host and any attempts to allocate hugepages resulted in
>> significant amount of cpu time spent in either direct compaction or
>> in kcompatd scanning vfio pinned pages over and over again that can
>> not be migrated.
>>
>> Signed-off-by: Khalid Aziz <khalid.aziz@oracle.com>
>> Suggested-by: Steve Sistare <steven.sistare@oracle.com>
>> ---
>>   mm/compaction.c | 33 +++++++++++++++++++++++++++++----
>>   1 file changed, 29 insertions(+), 4 deletions(-)
>>
>> diff --git a/mm/compaction.c b/mm/compaction.c
>> index 5a9501e0ae01..d1371fd75391 100644
>> --- a/mm/compaction.c
>> +++ b/mm/compaction.c
>> @@ -764,6 +764,32 @@ static bool too_many_isolated(pg_data_t *pgdat)
>>   	return too_many;
>>   }
>>   
>> +/*
>> + * Check if this base page should be skipped from isolation because
>> + * it is pinned. This function is called for regular pages only, and not
>> + * for THP or hugetlbfs pages. This code is inspired by similar code
>> + * in migrate_vma_check_page(), can_split_folio() and
>> + * folio_migrate_mapping()
>> + */
>> +static inline bool is_pinned_page(struct page *page)
>> +{
>> +	unsigned long extra_refs;
>> +
>> +	/* anonymous page can have extra ref from page cache */
> 
> "from swapcache"?
> 
>> +	if (page_mapping(page))
>> +		extra_refs = 1 + page_has_private(page);
>> +	else
>> +		extra_refs = PageSwapCache(page) ? 1 : 0;
>> +
>> +	/*
>> +	 * This is an admittedly racy check but good enough to determine
>> +	 * if a page should be isolated
> 
> "cannot be migrated"?
> 

Thanks, Andrew! I will fix these and send out v2 patch.

--
Khalid