Re: [PATCH v10 2/5] mm: page_isolation: check specified range for unmovable pages

[Zi Yan <ziy@nvidia.com>]

[-- Attachment #1: Type: text/plain, Size: 4678 bytes --]

On 12 Apr 2022, at 9:10, David Hildenbrand wrote:

> On 06.04.22 17:18, Zi Yan wrote:
>> From: Zi Yan <ziy@nvidia.com>
>>
>> Enable set_migratetype_isolate() to check specified sub-range for
>> unmovable pages during isolation. Page isolation is done
>> at MAX_ORDER_NR_PAEGS granularity, but not all pages within that
>> granularity are intended to be isolated. For example,
>> alloc_contig_range(), which uses page isolation, allows ranges without
>> alignment. This commit makes unmovable page check only look for
>> interesting pages, so that page isolation can succeed for any
>> non-overlapping ranges.
>>
>> Signed-off-by: Zi Yan <ziy@nvidia.com>
>> ---
>
> [...]
>
>>  /*
>> - * This function checks whether pageblock includes unmovable pages or not.
>> + * This function checks whether the range [start_pfn, end_pfn) includes
>> + * unmovable pages or not. The range must fall into a single pageblock and
>> + * consequently belong to a single zone.
>>   *
>>   * PageLRU check without isolation or lru_lock could race so that
>>   * MIGRATE_MOVABLE block might include unmovable pages. And __PageMovable
>> @@ -28,12 +30,14 @@
>>   * cannot get removed (e.g., via memory unplug) concurrently.
>>   *
>>   */
>> -static struct page *has_unmovable_pages(struct zone *zone, struct page *page,
>> -				 int migratetype, int flags)
>> +static struct page *has_unmovable_pages(unsigned long start_pfn, unsigned long end_pfn,
>> +				int migratetype, int flags)
>>  {
>> -	unsigned long iter = 0;
>> -	unsigned long pfn = page_to_pfn(page);
>> -	unsigned long offset = pfn % pageblock_nr_pages;
>> +	unsigned long pfn = start_pfn;
>> +	struct page *page = pfn_to_page(pfn);
>
>
> Just do
>
> struct page *page = pfn_to_page(start_pfn);
> struct zone *zone = page_zone(page);
>
> here. No need to lookup the zone again in the loop because, as you
> document "must ... belong to a single zone.".
>
> Then, there is also no need to initialize "pfn" here. In the loop header
> is sufficient.
>

Sure.

>> +
>> +	VM_BUG_ON(ALIGN_DOWN(start_pfn, pageblock_nr_pages) !=
>> +		  ALIGN_DOWN(end_pfn - 1, pageblock_nr_pages));
>>
>>  	if (is_migrate_cma_page(page)) {
>>  		/*
>> @@ -47,8 +51,11 @@ static struct page *has_unmovable_pages(struct zone *zone, struct page *page,
>>  		return page;
>>  	}
>>
>> -	for (; iter < pageblock_nr_pages - offset; iter++) {
>> -		page = pfn_to_page(pfn + iter);
>> +	for (pfn = start_pfn; pfn < end_pfn; pfn++) {
>> +		struct zone *zone;
>> +
>> +		page = pfn_to_page(pfn);
>> +		zone = page_zone(page);
>>
>>  		/*
>>  		 * Both, bootmem allocations and memory holes are marked
>> @@ -85,7 +92,7 @@ static struct page *has_unmovable_pages(struct zone *zone, struct page *page,
>>  			}
>>
>>  			skip_pages = compound_nr(head) - (page - head);
>> -			iter += skip_pages - 1;
>> +			pfn += skip_pages - 1;
>>  			continue;
>>  		}
>>
>> @@ -97,7 +104,7 @@ static struct page *has_unmovable_pages(struct zone *zone, struct page *page,
>>  		 */
>>  		if (!page_ref_count(page)) {
>>  			if (PageBuddy(page))
>> -				iter += (1 << buddy_order(page)) - 1;
>> +				pfn += (1 << buddy_order(page)) - 1;
>>  			continue;
>>  		}
>>
>> @@ -134,11 +141,18 @@ static struct page *has_unmovable_pages(struct zone *zone, struct page *page,
>>  	return NULL;
>>  }
>>
>> -static int set_migratetype_isolate(struct page *page, int migratetype, int isol_flags)
>> +/*
>> + * This function set pageblock migratetype to isolate if no unmovable page is
>> + * present in [start_pfn, end_pfn). The pageblock must intersect with
>> + * [start_pfn, end_pfn).
>> + */
>> +static int set_migratetype_isolate(struct page *page, int migratetype, int isol_flags,
>> +			unsigned long start_pfn, unsigned long end_pfn)
>
> I think we might be able do better, eventually not passing start_pfn at
> all. Hmm.

IMHO, having start_pfn and end_pfn in the parameter list would make the
interface easier to understand. Otherwise if we remove start_pfn,
the caller needs to adjust @page to be within the range of [start_pfn,
end_pfn)

>
> I think we want to pull out the
> start_isolate_page_range()/undo_isolate_page_range() interface change
> into a separate patch.

You mean a patch just adding

unsigned long isolate_start = pfn_max_align_down(start_pfn);
unsigned long isolate_end = pfn_max_align_up(end_pfn);

in start_isolate_page_range()/undo_isolate_page_range()?

Yes I can do that.

>
> Let me try to give it a shot, I'll try hacking something up real quick
> to see if we can do better.

Sure. Thanks.

--
Best Regards,
Yan, Zi

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