[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <af188933-9677-c733-00f4-1d042425522f@huawei.com>
Date: Mon, 12 Jul 2021 19:10:03 +0800
From: Miaohe Lin <linmiaohe@...wei.com>
To: Michal Hocko <mhocko@...e.com>
CC: <akpm@...ux-foundation.org>, <hannes@...xchg.org>,
<vbabka@...e.cz>, <axboe@...nel.dk>, <iamjoonsoo.kim@....com>,
<alexs@...nel.org>, <apopple@...dia.com>, <willy@...radead.org>,
<minchan@...nel.org>, <david@...hat.com>, <shli@...com>,
<hillf.zj@...baba-inc.com>, <linux-mm@...ck.org>,
<linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 2/5] mm/vmscan: remove misleading setting to sc->priority
On 2021/7/12 15:24, Michal Hocko wrote:
> On Sat 10-07-21 18:03:26, Miaohe Lin wrote:
>> The priority field of sc is used to control how many pages we should scan
>> at once while we always traverse the list to shrink the pages in these
>> functions. So these settings are unneeded and misleading.
>
> I dunno. I agree that priority is not really used as these operate on
> page lists but I am not sure this is worth touching.
When I investigated the vmscan code, I thought the order here would control the
proportion of the pages in the list to shrink. So I prefer to remove these.
Thanks a lot for review and reply!
>
>> Signed-off-by: Miaohe Lin <linmiaohe@...wei.com>
>> ---
>> mm/vmscan.c | 2 --
>> 1 file changed, 2 deletions(-)
>>
>> diff --git a/mm/vmscan.c b/mm/vmscan.c
>> index 6483fe0e2065..fbe53e60b248 100644
>> --- a/mm/vmscan.c
>> +++ b/mm/vmscan.c
>> @@ -1702,7 +1702,6 @@ unsigned int reclaim_clean_pages_from_list(struct zone *zone,
>> {
>> struct scan_control sc = {
>> .gfp_mask = GFP_KERNEL,
>> - .priority = DEF_PRIORITY,
>> .may_unmap = 1,
>> };
>> struct reclaim_stat stat;
>> @@ -2327,7 +2326,6 @@ unsigned long reclaim_pages(struct list_head *page_list)
>> unsigned int noreclaim_flag;
>> struct scan_control sc = {
>> .gfp_mask = GFP_KERNEL,
>> - .priority = DEF_PRIORITY,
>> .may_writepage = 1,
>> .may_unmap = 1,
>> .may_swap = 1,
>> --
>> 2.23.0
>
Powered by blists - more mailing lists