Re: [PATCH] mm, mlock, vmscan: no more skipping pagevecs

From: Shakeel Butt
Date: Tue Nov 21 2017 - 12:20:24 EST

On Tue, Nov 21, 2017 at 7:32 AM, Johannes Weiner <hannes@xxxxxxxxxxx> wrote:
> On Sat, Nov 04, 2017 at 03:43:12PM -0700, Shakeel Butt wrote:
>> When a thread mlocks an address space backed by file, a new
>> page is allocated (assuming file page is not in memory), added
>> to the local pagevec (lru_add_pvec), I/O is triggered and the
>> thread then sleeps on the page. On I/O completion, the thread
>> can wake on a different CPU, the mlock syscall will then sets
>> the PageMlocked() bit of the page but will not be able to put
>> that page in unevictable LRU as the page is on the pagevec of
>> a different CPU. Even on drain, that page will go to evictable
>> LRU because the PageMlocked() bit is not checked on pagevec
>> drain.
>> The page will eventually go to right LRU on reclaim but the
>> LRU stats will remain skewed for a long time.
>> However, this issue does not happen for anon pages on swap
>> because unlike file pages, anon pages are not added to pagevec
>> until they have been fully swapped in.
> How so? __read_swap_cache_async() is the core function that allocates
> the page, and that always puts the page on the pagevec before IO is
> initiated.
>> Also the fault handler uses vm_flags to set the PageMlocked() bit of
>> such anon pages even before returning to mlock() syscall and mlocked
>> pages will skip pagevecs and directly be put into unevictable LRU.
> Where does the swap fault path set PageMlocked()?
> I might just be missing something.

No, you are right. I got confused by
lru_cache_add_active_or_unevictable() in do_swap_page() but missed the
preceding comment that says "ksm created a completely new copy". I
will fix the the commit message as well.