Re: [PATCH 5/5] mm/vmscan: fix misleading comment in isolate_lru_pages()

[Miaohe Lin <linmiaohe@huawei.com>]

On 2021/7/12 15:28, Michal Hocko wrote:
> On Sat 10-07-21 18:03:29, Miaohe Lin wrote:
>> We couldn't know whether the page is being freed elsewhere until we failed
>> to increase the page count.
> 
> This is moving a hard to understand comment from one place to another.

If get_page_unless_zero failed, the page could have been freed elsewhere. I think
this looks straightforward but doesn't help a lot. Are you preferring to just
remove this comment ?

Thank you.

> If anything this would benefit from what that elsewhere might be
> typically or simply drop the comment altogether.
> 
>>
>> Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
>> ---
>>  mm/vmscan.c | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/mm/vmscan.c b/mm/vmscan.c
>> index a74760c48bd8..6e26b3c93242 100644
>> --- a/mm/vmscan.c
>> +++ b/mm/vmscan.c
>> @@ -1891,7 +1891,6 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
>>  		 */
>>  		scan += nr_pages;
>>  		if (!__isolate_lru_page_prepare(page, mode)) {
>> -			/* It is being freed elsewhere */
>>  			list_move(&page->lru, src);
>>  			continue;
>>  		}
>> @@ -1901,6 +1900,7 @@ static unsigned long isolate_lru_pages(unsigned long nr_to_scan,
>>  		 * page release code relies on it.
>>  		 */
>>  		if (unlikely(!get_page_unless_zero(page))) {
>> +			/* It is being freed elsewhere. */
>>  			list_move(&page->lru, src);
>>  			continue;
>>  		}
>> -- 
>> 2.23.0
>