From: Hugh Dickins <hughd@google.com>
To: Michal Hocko <mhocko@kernel.org>
Cc: Hugh Dickins <hughd@google.com>, Baoquan He <bhe@redhat.com>,
Vlastimil Babka <vbabka@suse.cz>,
David Hildenbrand <david@redhat.com>,
linux-mm@kvack.org, pifang@redhat.com,
linux-kernel@vger.kernel.org, akpm@linux-foundation.org,
aarcange@redhat.com, Mel Gorman <mgorman@suse.de>
Subject: Re: Memory hotplug softlock issue
Date: Wed, 21 Nov 2018 17:53:33 -0800 (PST) [thread overview]
Message-ID: <alpine.LSU.2.11.1811211726080.5557@eggly.anvils> (raw)
In-Reply-To: <20181121173123.GS12932@dhcp22.suse.cz>
On Wed, 21 Nov 2018, Michal Hocko wrote:
> On Mon 19-11-18 21:44:41, Hugh Dickins wrote:
> [...]
> > [PATCH] mm: put_and_wait_on_page_locked() while page is migrated
> >
> > We have all assumed that it is essential to hold a page reference while
> > waiting on a page lock: partly to guarantee that there is still a struct
> > page when MEMORY_HOTREMOVE is configured, but also to protect against
> > reuse of the struct page going to someone who then holds the page locked
> > indefinitely, when the waiter can reasonably expect timely unlocking.
>
> I would add the following for the "problem statement". Feel free to
> reuse per your preference:
> "
> An elevated reference count, however, stands in the way of migration and
> forces it to fail with a bad timing. This is especially a problem for
> memory offlining which retries for ever (or until the operation is
> terminated from userspace) because a heavy refault workload can trigger
> essentially an endless loop of migration failures. Therefore
> __migration_entry_wait is essentially harmful for the even it is waiting
> for.
> "
Okay, I do have a lot written from way back when I prepared the
now-abandoned migration_waitqueue patch internally, but I'll factor in
what you say above when I get there - in particular, you highlight the
memory offlining aspect, as in this mailthread: which is very helpful,
because it's outside my experience so I won't have mentioned it - thanks.
I just know that there's some important linkage to do, to the August 2017
WQ_FLAG_BOOKMARK discussion: so it's a research and editing job I have to
work myself up to at the right moment.
>
> > But in fact, so long as wait_on_page_bit_common() does the put_page(),
> > and is careful not to rely on struct page contents thereafter, there is
> > no need to hold a reference to the page while waiting on it. That does
> > mean that this case cannot go back through the loop: but that's fine for
> > the page migration case, and even if used more widely, is limited by the
> > "Stop walking if it's locked" optimization in wake_page_function().
>
> I would appreciate this would be more explicit about the existence of
> the elevated-ref-count problem but it reduces it to a tiny time window
> compared to the whole time the waiter is blocked. So a great
> improvement.
Fair enough, I'll do so. (But that's a bit like when we say we've attached
something and then forget to do so: please check that I've been honest
when I do post.)
>
> > Add interface put_and_wait_on_page_locked() to do this, using negative
> > value of the lock arg to wait_on_page_bit_common() to implement it.
> > No interruptible or killable variant needed yet, but they might follow:
> > I have a vague notion that reporting -EINTR should take precedence over
> > return from wait_on_page_bit_common() without knowing the page state,
> > so arrange it accordingly - but that may be nothing but pedantic.
> >
> > shrink_page_list()'s __ClearPageLocked(): that was a surprise!
>
> and I can imagine a bad one. Do we really have to be so clever here?
> The unlock_page went away in the name of performance (a978d6f521063)
> and I would argue that this is a slow path where this is just not worth
> it.
Do we really have to be so clever here? That's a good question: now we
have PG_waiters, we probably do not need to bother with this cleverness,
and it would save me from having to expand on that comment as I was asked.
I'll try going back to a simple unlock_page() there: and can always restore
the __ClearPageLocked if a reviewer demands, or 0-day notices regression,
>
> > this
> > survived a lot of testing before that showed up. It does raise the
> > question: should is_page_cache_freeable() and __remove_mapping() now
> > treat a PG_waiters page as if an extra reference were held? Perhaps,
> > but I don't think it matters much, since shrink_page_list() already
> > had to win its trylock_page(), so waiters are not very common there: I
> > noticed no difference when trying the bigger change, and it's surely not
> > needed while put_and_wait_on_page_locked() is only for page migration.
> >
> > Signed-off-by: Hugh Dickins <hughd@google.com>
>
> The patch looks good to me - quite ugly but it doesn't make the existing
> code much worse.
>
> With the problem described Vlastimil fixed, feel free to add
> Acked-by: Michal Hocko <mhocko@suse.com>
Thanks!
>
> And thanks for a prompt patch. This is something I've been chasing for
> quite some time. __migration_entry_wait came to my radar only recently
> because this is an extremely volatile area.
You are very gracious to describe a patch promised six months ago as
"prompt". But it does help me a lot to have it fixing a real problem
for someone (thank you Baoquan) - well, it fixed a real problem for us
internally too, but very nice to gather more backing for it like this.
Hugh
next prev parent reply other threads:[~2018-11-22 1:53 UTC|newest]
Thread overview: 55+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-11-14 7:09 Memory hotplug softlock issue Baoquan He
2018-11-14 7:16 ` Baoquan He
2018-11-14 8:18 ` David Hildenbrand
2018-11-14 9:00 ` Baoquan He
2018-11-14 9:25 ` David Hildenbrand
2018-11-14 9:41 ` Michal Hocko
2018-11-14 9:48 ` David Hildenbrand
2018-11-14 10:04 ` Michal Hocko
2018-11-14 9:01 ` Michal Hocko
2018-11-14 9:22 ` David Hildenbrand
2018-11-14 9:37 ` Michal Hocko
2018-11-14 9:39 ` David Hildenbrand
2018-11-14 14:52 ` Baoquan He
2018-11-14 15:00 ` Michal Hocko
2018-11-15 5:10 ` Baoquan He
2018-11-15 7:30 ` Michal Hocko
2018-11-15 7:53 ` Baoquan He
2018-11-15 8:30 ` Michal Hocko
2018-11-15 9:42 ` David Hildenbrand
2018-11-15 9:52 ` Baoquan He
2018-11-15 9:53 ` David Hildenbrand
2018-11-15 13:12 ` Baoquan He
2018-11-15 13:19 ` Michal Hocko
2018-11-15 13:23 ` Baoquan He
2018-11-15 14:25 ` Michal Hocko
2018-11-15 13:38 ` Baoquan He
2018-11-15 14:32 ` Michal Hocko
2018-11-15 14:34 ` Baoquan He
2018-11-16 1:24 ` Baoquan He
2018-11-16 9:14 ` Michal Hocko
2018-11-17 4:22 ` Baoquan He
2018-11-19 10:52 ` Baoquan He
2018-11-19 12:40 ` Michal Hocko
2018-11-19 12:51 ` Michal Hocko
2018-11-19 14:10 ` Michal Hocko
2018-11-19 16:36 ` Vlastimil Babka
2018-11-19 16:46 ` Michal Hocko
2018-11-19 16:46 ` Vlastimil Babka
2018-11-19 16:48 ` Vlastimil Babka
2018-11-19 17:01 ` Michal Hocko
2018-11-19 17:33 ` Michal Hocko
2018-11-19 20:34 ` Hugh Dickins
2018-11-19 20:59 ` Michal Hocko
2018-11-20 1:56 ` Baoquan He
2018-11-20 5:44 ` Hugh Dickins
2018-11-20 13:38 ` Vlastimil Babka
2018-11-20 13:58 ` Baoquan He
2018-11-20 14:05 ` Michal Hocko
2018-11-20 14:12 ` Baoquan He
2018-11-21 1:21 ` Hugh Dickins
2018-11-21 1:08 ` Hugh Dickins
2018-11-21 3:20 ` Hugh Dickins
2018-11-21 17:31 ` Michal Hocko
2018-11-22 1:53 ` Hugh Dickins [this message]
2018-11-14 10:00 ` Michal Hocko
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=alpine.LSU.2.11.1811211726080.5557@eggly.anvils \
--to=hughd@google.com \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=bhe@redhat.com \
--cc=david@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mgorman@suse.de \
--cc=mhocko@kernel.org \
--cc=pifang@redhat.com \
--cc=vbabka@suse.cz \
/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).