Re: [PATCH v2 4/4] mm/mglru: fix anon-only reclaim evicting file pages when swappiness=max

[Ridong Chen <ridong.chen@linux.dev>]

On 7/20/2026 10:48 AM, Qi Zheng wrote:
> 
> 
> On 7/18/26 9:43 PM, Barry Song wrote:
>> On Sat, Jul 18, 2026 at 5:54 PM Ridong Chen <ridong.chen@linux.dev> 
>> wrote:
>>>
>>> From: Ridong Chen <chenridong@xiaomi.com>
>>>
>>> The previous patch fixes this issue for the traditional LRU, which also
>>> exists in MGLRU [1]. Fix this by checking whether swappiness is
>>> SWAPPINESS_ANON_ONLY in get_swappiness() first, and returning 0 from
>>> get_nr_to_scan() when swappiness is SWAPPINESS_ANON_ONLY and anon pages
>>> cannot be reclaimed, to avoid useless work.
>>
>> Is this intended to avoid unnecessary work, or to prevent file folios
>> from being reclaimed incorrectly?
> 
> Based on the test results below, it seems primarily intended to prevent
> file pages from being mistakenly reclaimed. Also, bailing out early in
> the SWAPPINESS_ANON_ONLY + !can_reclaim_anon_pages case is to avoid
> unnecessary work.
> 
> Hi Ridong, perhaps the commit message could be clearer. ;)
> 

Thanks for the feedback. Let me clarify:

When SWAPPINESS_ANON_ONLY is set, get_swappiness() returns the 
swappiness value, and get_type_to_scan() can always return LRU_GEN_ANON. 
So if anonymous pages cannot be reclaimed, we bail out early to avoid 
unnecessary work. In that case, get_nr_to_scan() just returns 0.

If we do not return early in get_nr_to_scan() under the 
SWAPPINESS_ANON_ONLY + !can_reclaim_anon_pages condition, then in 
isolate_folios(), it may fall back to scanning file pages, but this only 
happens when scanned = 0, which is not the usual case.


```
shrink_one
try_to_shrink_lruvec
   get_swappiness // return SWAPPINESS_ANON_ONLY if swappiness=max
   get_nr_to_scan // return 0 if anon can't be reclaimed, skip
   evict_folios
     isolate_folios // !scanned falling back to anyther type if !scanned
```

I'll update the commit message to make this clearer in the next version.

>>
>>>
>>> The test result:
>>> Before fix:
>>>
>>>    # cat /sys/kernel/mm/lru_gen/enabled
>>>    0x0007
>>>    # cat memory.stat
>>>    anon 204800
>>>    file 67108864
>>>    ...
>>>    pgsteal_proactive 0
>>>    pgscan_proactive 0
>>>
>>>    # echo "64M swappiness=max" > memory.reclaim
>>>    # cat memory.stat
>>>    anon 208896
>>>    file 0
>>>    ...
>>>    pgsteal_proactive 16384
>>>    pgscan_proactive 16384
>>>
>>> After fix:
>>>
>>>    # cat memory.stat
>>>    anon 188416
>>>    file 67215360
>>>    kernel 1970176
>>>    ...
>>>    pgsteal_proactive 0
>>>    pgscan_proactive 0
>>>
>>>    # echo "64M swappiness=max" > memory.reclaim
>>>    -bash: echo: write error: Resource temporarily unavailable
>>>    # cat memory.stat
>>>    anon 204800
>>>    file 67215360
>>>    ...
>>>    pgsteal_proactive 0
>>>    pgscan_proactive 0
>>>
>>> [1] https://sashiko.dev/#/patchset/20260717113300.214717-1- 
>>> ridong.chen@linux.dev
>>>
>>> Fixes: 68a1436bde00 ("mm: add swappiness=max arg to memory.reclaim 
>>> for only anon reclaim")
>>> Signed-off-by: Ridong Chen <chenridong@xiaomi.com>
>>> ---
>>>   mm/vmscan.c | 14 +++++++++++++-
>>>   1 file changed, 13 insertions(+), 1 deletion(-)
>>>
> 
> Overall, this looks good to me.
> 
> Acked-by: Qi Zheng <qi.zheng@linux.dev>
> 
> Thanks,
> Qi
> 
> 

-- 
Best regards
Ridong