From: Alex Shi <alex.shi@linux.alibaba.com>
To: Alexander Duyck <alexander.duyck@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
Mel Gorman <mgorman@techsingularity.net>,
Tejun Heo <tj@kernel.org>, Hugh Dickins <hughd@google.com>,
Konstantin Khlebnikov <khlebnikov@yandex-team.ru>,
Daniel Jordan <daniel.m.jordan@oracle.com>,
Yang Shi <yang.shi@linux.alibaba.com>,
Matthew Wilcox <willy@infradead.org>,
Johannes Weiner <hannes@cmpxchg.org>,
kbuild test robot <lkp@intel.com>, linux-mm <linux-mm@kvack.org>,
LKML <linux-kernel@vger.kernel.org>,
cgroups@vger.kernel.org, Shakeel Butt <shakeelb@google.com>,
Joonsoo Kim <iamjoonsoo.kim@lge.com>,
Wei Yang <richard.weiyang@gmail.com>,
"Kirill A. Shutemov" <kirill@shutemov.name>,
Michal Hocko <mhocko@kernel.org>,
Vladimir Davydov <vdavydov.dev@gmail.com>,
Rong Chen <rong.a.chen@intel.com>
Subject: Re: [PATCH v16 18/22] mm/lru: replace pgdat lru_lock with lruvec lock
Date: Sat, 18 Jul 2020 22:15:02 +0800 [thread overview]
Message-ID: <62dfd262-a7ac-d18e-216a-2988c690b256@linux.alibaba.com> (raw)
In-Reply-To: <CAKgT0Ud+Dj-Q8Sxv8eDQhjM3fFHwnU_ZFEVG54debBennUmxAg@mail.gmail.com>
在 2020/7/18 上午5:38, Alexander Duyck 写道:
>> + return locked_lruvec;
>> +
>> + if (locked_lruvec)
>> + unlock_page_lruvec_irqrestore(locked_lruvec, *flags);
>> +
>> + return lock_page_lruvec_irqsave(page, flags);
>> +}
>> +
> These relock functions have no users in this patch. It might make
> sense and push this code to patch 19 in your series since that is
> where they are first used. In addition they don't seem very efficient
> as you already had to call mem_cgroup_page_lruvec once, why do it
> again when you could just store the value and lock the new lruvec if
> needed?
Right, it's better to move for late patch.
As to call the func again, mainly it's for code neat.
Thanks!
>
>> #ifdef CONFIG_CGROUP_WRITEBACK
>>
>> struct wb_domain *mem_cgroup_wb_domain(struct bdi_writeback *wb);
>> diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
>> index 14c668b7e793..36c1680efd90 100644
>> --- a/include/linux/mmzone.h
>> +++ b/include/linux/mmzone.h
>> @@ -261,6 +261,8 @@ struct lruvec {
>> atomic_long_t nonresident_age;
>> /* Refaults at the time of last reclaim cycle */
>> unsigned long refaults;
>> + /* per lruvec lru_lock for memcg */
>> + spinlock_t lru_lock;
>> /* Various lruvec state flags (enum lruvec_flags) */
>> unsigned long flags;
> Any reason for placing this here instead of at the end of the
> structure? From what I can tell it looks like lruvec is already 128B
> long so placing the lock on the end would put it into the next
> cacheline which may provide some performance benefit since it is
> likely to be bounced quite a bit.
Rong Chen(Cced) once reported a performance regression when the lock at
the end of struct, and move here could remove it.
Although I can't not reproduce. But I trust his report.
...
>> putback:
>> - spin_unlock_irq(&zone->zone_pgdat->lru_lock);
>> pagevec_add(&pvec_putback, pvec->pages[i]);
>> pvec->pages[i] = NULL;
>> }
>> - /* tempary disable irq, will remove later */
>> - local_irq_disable();
>> __mod_zone_page_state(zone, NR_MLOCK, delta_munlocked);
>> - local_irq_enable();
>> + if (lruvec)
>> + unlock_page_lruvec_irq(lruvec);
> So I am not a fan of this change. You went to all the trouble of
> reducing the lock scope just to bring it back out here again. In
> addition it implies there is a path where you might try to update the
> page state without disabling interrupts.
Right. but any idea to avoid this except a extra local_irq_disable?
...
>> if (PageLRU(page)) {
>> - struct pglist_data *pgdat = page_pgdat(page);
>> + struct lruvec *new_lruvec;
>>
>> - if (pgdat != locked_pgdat) {
>> - if (locked_pgdat)
>> - spin_unlock_irqrestore(&locked_pgdat->lru_lock,
>> + new_lruvec = mem_cgroup_page_lruvec(page,
>> + page_pgdat(page));
>> + if (new_lruvec != lruvec) {
>> + if (lruvec)
>> + unlock_page_lruvec_irqrestore(lruvec,
>> flags);
>> lock_batch = 0;
>> - locked_pgdat = pgdat;
>> - spin_lock_irqsave(&locked_pgdat->lru_lock, flags);
>> + lruvec = lock_page_lruvec_irqsave(page, &flags);
>> }
> This just kind of seems ugly to me. I am not a fan of having to fetch
> the lruvec twice when you already have it in new_lruvec. I suppose it
> is fine though since you are just going to be replacing it later
> anyway.
>
yes, it will be reproduce later.
Thanks
Alex
next prev parent reply other threads:[~2020-07-18 14:15 UTC|newest]
Thread overview: 66+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-07-11 0:58 [PATCH v16 00/22] per memcg lru_lock Alex Shi
2020-07-11 0:58 ` [PATCH v16 01/22] mm/vmscan: remove unnecessary lruvec adding Alex Shi
2020-07-11 0:58 ` [PATCH v16 02/22] mm/page_idle: no unlikely double check for idle page counting Alex Shi
2020-07-11 0:58 ` [PATCH v16 03/22] mm/compaction: correct the comments of compact_defer_shift Alex Shi
2020-07-11 0:58 ` [PATCH v16 04/22] mm/compaction: rename compact_deferred as compact_should_defer Alex Shi
2020-07-11 0:58 ` [PATCH v16 05/22] mm/thp: move lru_add_page_tail func to huge_memory.c Alex Shi
2020-07-16 8:59 ` Alex Shi
2020-07-16 13:17 ` Kirill A. Shutemov
2020-07-17 5:13 ` Alex Shi
2020-07-20 8:37 ` Kirill A. Shutemov
2020-07-11 0:58 ` [PATCH v16 06/22] mm/thp: clean up lru_add_page_tail Alex Shi
2020-07-20 8:43 ` Kirill A. Shutemov
2020-07-11 0:58 ` [PATCH v16 07/22] mm/thp: remove code path which never got into Alex Shi
2020-07-20 8:43 ` Kirill A. Shutemov
2020-07-11 0:58 ` [PATCH v16 08/22] mm/thp: narrow lru locking Alex Shi
2020-07-11 0:58 ` [PATCH v16 09/22] mm/memcg: add debug checking in lock_page_memcg Alex Shi
2020-07-11 0:58 ` [PATCH v16 10/22] mm/swap: fold vm event PGROTATED into pagevec_move_tail_fn Alex Shi
2020-07-11 0:58 ` [PATCH v16 11/22] mm/lru: move lru_lock holding in func lru_note_cost_page Alex Shi
2020-07-11 0:58 ` [PATCH v16 12/22] mm/lru: move lock into lru_note_cost Alex Shi
2020-07-11 0:58 ` [PATCH v16 13/22] mm/lru: introduce TestClearPageLRU Alex Shi
2020-07-16 9:06 ` Alex Shi
2020-07-16 21:12 ` Alexander Duyck
2020-07-17 7:45 ` Alex Shi
2020-07-17 18:26 ` Alexander Duyck
2020-07-19 4:45 ` Alex Shi
2020-07-19 11:24 ` Alex Shi
2020-07-11 0:58 ` [PATCH v16 14/22] mm/thp: add tail pages into lru anyway in split_huge_page() Alex Shi
2020-07-17 9:30 ` Alex Shi
2020-07-20 8:49 ` Kirill A. Shutemov
2020-07-20 9:04 ` Alex Shi
2020-07-11 0:58 ` [PATCH v16 15/22] mm/compaction: do page isolation first in compaction Alex Shi
2020-07-16 21:32 ` Alexander Duyck
2020-07-17 5:09 ` Alex Shi
2020-07-17 16:09 ` Alexander Duyck
2020-07-19 3:59 ` Alex Shi
2020-07-11 0:58 ` [PATCH v16 16/22] mm/mlock: reorder isolation sequence during munlock Alex Shi
2020-07-17 20:30 ` Alexander Duyck
2020-07-19 3:55 ` Alex Shi
2020-07-20 18:51 ` Alexander Duyck
2020-07-21 9:26 ` Alex Shi
2020-07-21 13:51 ` Alex Shi
2020-07-11 0:58 ` [PATCH v16 17/22] mm/swap: serialize memcg changes during pagevec_lru_move_fn Alex Shi
2020-07-11 0:58 ` [PATCH v16 18/22] mm/lru: replace pgdat lru_lock with lruvec lock Alex Shi
2020-07-17 21:38 ` Alexander Duyck
2020-07-18 14:15 ` Alex Shi [this message]
2020-07-19 9:12 ` Alex Shi
2020-07-19 15:14 ` Alexander Duyck
2020-07-20 5:47 ` Alex Shi
2020-07-11 0:58 ` [PATCH v16 19/22] mm/lru: introduce the relock_page_lruvec function Alex Shi
2020-07-17 22:03 ` Alexander Duyck
2020-07-18 14:01 ` Alex Shi
2020-07-11 0:58 ` [PATCH v16 20/22] mm/vmscan: use relock for move_pages_to_lru Alex Shi
2020-07-17 21:44 ` Alexander Duyck
2020-07-18 14:15 ` Alex Shi
2020-07-11 0:58 ` [PATCH v16 21/22] mm/pgdat: remove pgdat lru_lock Alex Shi
2020-07-17 21:09 ` Alexander Duyck
2020-07-18 14:17 ` Alex Shi
2020-07-11 0:58 ` [PATCH v16 22/22] mm/lru: revise the comments of lru_lock Alex Shi
2020-07-11 1:02 ` [PATCH v16 00/22] per memcg lru_lock Alex Shi
2020-07-16 8:49 ` Alex Shi
2020-07-16 14:11 ` Alexander Duyck
2020-07-17 5:24 ` Alex Shi
2020-07-19 15:23 ` Hugh Dickins
2020-07-20 3:01 ` Alex Shi
2020-07-20 4:47 ` Hugh Dickins
2020-07-20 7:30 ` Alex Shi
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=62dfd262-a7ac-d18e-216a-2988c690b256@linux.alibaba.com \
--to=alex.shi@linux.alibaba.com \
--cc=akpm@linux-foundation.org \
--cc=alexander.duyck@gmail.com \
--cc=cgroups@vger.kernel.org \
--cc=daniel.m.jordan@oracle.com \
--cc=hannes@cmpxchg.org \
--cc=hughd@google.com \
--cc=iamjoonsoo.kim@lge.com \
--cc=khlebnikov@yandex-team.ru \
--cc=kirill@shutemov.name \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=lkp@intel.com \
--cc=mgorman@techsingularity.net \
--cc=mhocko@kernel.org \
--cc=richard.weiyang@gmail.com \
--cc=rong.a.chen@intel.com \
--cc=shakeelb@google.com \
--cc=tj@kernel.org \
--cc=vdavydov.dev@gmail.com \
--cc=willy@infradead.org \
--cc=yang.shi@linux.alibaba.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).