From: Vlastimil Babka <vbabka@suse.cz>
To: Michal Hocko <mhocko@kernel.org>,
Mel Gorman <mgorman@techsingularity.net>
Cc: Dmitry Vyukov <dvyukov@google.com>, Tejun Heo <tj@kernel.org>,
Christoph Lameter <cl@linux.com>,
"linux-mm@kvack.org" <linux-mm@kvack.org>,
LKML <linux-kernel@vger.kernel.org>,
Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@kernel.org>,
Peter Zijlstra <peterz@infradead.org>,
syzkaller <syzkaller@googlegroups.com>,
Andrew Morton <akpm@linux-foundation.org>
Subject: Re: mm: deadlock between get_online_cpus/pcpu_alloc
Date: Tue, 7 Feb 2017 10:23:31 +0100 [thread overview]
Message-ID: <614e9873-c894-de42-a38a-1798fc0be039@suse.cz> (raw)
In-Reply-To: <20170207084855.GC5065@dhcp22.suse.cz>
On 02/07/2017 09:48 AM, Michal Hocko wrote:
> On Mon 06-02-17 22:05:30, Mel Gorman wrote:
>>> Unfortunately it does not seem to help.
>>
>> I'm a little stuck on how to best handle this. get_online_cpus() can
>> halt forever if the hotplug operation is holding the mutex when calling
>> pcpu_alloc. One option would be to add a try_get_online_cpus() helper which
>> trylocks the mutex. However, given that drain is so unlikely to actually
>> make that make a difference when racing against parallel allocations,
>> I think this should be acceptable.
>>
>> Any objections?
>>
>> diff --git a/mm/page_alloc.c b/mm/page_alloc.c
>> index 3b93879990fd..a3192447e906 100644
>> --- a/mm/page_alloc.c
>> +++ b/mm/page_alloc.c
>> @@ -3432,7 +3432,17 @@ __alloc_pages_direct_reclaim(gfp_t gfp_mask, unsigned int order,
>> */
>> if (!page && !drained) {
>> unreserve_highatomic_pageblock(ac, false);
>> - drain_all_pages(NULL);
>> +
>> + /*
>> + * Only drain from contexts allocating for user allocations.
>> + * Kernel allocations could be holding a CPU hotplug-related
>> + * mutex, particularly hot-add allocating per-cpu structures
>> + * while hotplug-related mutex's are held which would prevent
>> + * get_online_cpus ever returning.
>> + */
>> + if (gfp_mask & __GFP_HARDWALL)
>> + drain_all_pages(NULL);
>> +
>
> This wouldn't work AFAICS. If you look at the lockdep splat, the path
> which reverses the locking order (takes pcpu_alloc_mutex prior to
> cpu_hotplug.lock is bpf_array_alloc_percpu which is GFP_USER and thus
> __GFP_HARDWALL.
>
> I believe we shouldn't pull any dependency on the hotplug locks inside
> the allocator. This is just too fragile! Can we simply drop the
> get_online_cpus()? Why do we need it, anyway? Say we are racing with the
It was added after I noticed in review that queue_work_on() has a
comment that caller must ensure that cpu can't go away, and wondered
about it. Also noted that a similar lru_add_drain_all() does it too.
> cpu offlining. I have to check the code but my impression was that WQ
> code will ignore the cpu requested by the work item when the cpu is
> going offline. If the offline happens while the worker function already
> executes then it has to wait as we run with preemption disabled so we
> should be safe here. Or am I missing something obvious?
Tejun suggested an alternative solution to avoiding get_online_cpus() in
this thread:
https://lkml.kernel.org/r/<20170123170329.GA7820@htj.duckdns.org>
next prev parent reply other threads:[~2017-02-07 9:23 UTC|newest]
Thread overview: 60+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-01-29 12:44 mm: deadlock between get_online_cpus/pcpu_alloc Dmitry Vyukov
2017-01-29 17:22 ` Vlastimil Babka
2017-01-30 15:48 ` Dmitry Vyukov
2017-02-06 19:13 ` Dmitry Vyukov
2017-02-06 22:05 ` Mel Gorman
2017-02-07 8:48 ` Michal Hocko
2017-02-07 9:23 ` Vlastimil Babka [this message]
2017-02-07 9:46 ` Mel Gorman
2017-02-07 9:53 ` Michal Hocko
2017-02-07 10:42 ` Mel Gorman
2017-02-07 11:13 ` Mel Gorman
2017-02-07 9:43 ` Mel Gorman
2017-02-07 9:49 ` Vlastimil Babka
2017-02-07 10:05 ` Michal Hocko
2017-02-07 10:28 ` Mel Gorman
2017-02-07 10:35 ` Michal Hocko
2017-02-07 11:34 ` Mel Gorman
2017-02-07 11:43 ` Michal Hocko
2017-02-07 11:54 ` Vlastimil Babka
2017-02-07 12:08 ` Michal Hocko
2017-02-07 12:37 ` Michal Hocko
2017-02-07 12:43 ` Vlastimil Babka
2017-02-07 12:48 ` Michal Hocko
2017-02-07 13:57 ` Vlastimil Babka
2017-02-07 13:58 ` Mel Gorman
2017-02-07 14:19 ` Michal Hocko
2017-02-07 15:34 ` Michal Hocko
2017-02-07 16:22 ` Mel Gorman
2017-02-07 16:41 ` Michal Hocko
2017-02-07 16:55 ` Christoph Lameter
2017-02-07 22:25 ` Thomas Gleixner
2017-02-08 7:35 ` Michal Hocko
2017-02-08 12:02 ` Thomas Gleixner
2017-02-08 12:21 ` Michal Hocko
2017-02-08 12:26 ` Mel Gorman
2017-02-08 13:23 ` Thomas Gleixner
2017-02-08 14:03 ` Mel Gorman
2017-02-08 14:11 ` Peter Zijlstra
2017-02-08 15:11 ` Christoph Lameter
2017-02-08 15:21 ` Michal Hocko
2017-02-08 16:17 ` Christoph Lameter
2017-02-08 17:46 ` Thomas Gleixner
2017-02-09 3:15 ` Christoph Lameter
2017-02-09 11:42 ` Thomas Gleixner
2017-02-09 14:00 ` Christoph Lameter
2017-02-09 14:53 ` Thomas Gleixner
2017-02-09 15:42 ` Christoph Lameter
2017-02-09 16:12 ` Thomas Gleixner
2017-02-09 17:22 ` Christoph Lameter
2017-02-09 17:40 ` Thomas Gleixner
2017-02-09 19:15 ` Michal Hocko
2017-02-10 17:58 ` Christoph Lameter
2017-02-08 15:06 ` Christoph Lameter
2017-02-07 17:03 ` Tejun Heo
2017-02-07 20:16 ` Michal Hocko
2017-02-07 13:03 ` Mel Gorman
2017-02-07 13:48 ` Michal Hocko
2017-02-07 11:24 ` Tetsuo Handa
2017-02-07 8:43 ` Michal Hocko
2017-02-07 21:53 ` Thomas Gleixner
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=614e9873-c894-de42-a38a-1798fc0be039@suse.cz \
--to=vbabka@suse.cz \
--cc=akpm@linux-foundation.org \
--cc=cl@linux.com \
--cc=dvyukov@google.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mgorman@techsingularity.net \
--cc=mhocko@kernel.org \
--cc=mingo@kernel.org \
--cc=peterz@infradead.org \
--cc=syzkaller@googlegroups.com \
--cc=tglx@linutronix.de \
--cc=tj@kernel.org \
/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).