linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: David Hildenbrand <david@redhat.com>
To: "Michael S. Tsirkin" <mst@redhat.com>
Cc: Alexander Duyck <alexander.duyck@gmail.com>,
	wei.w.wang@intel.com, Nitesh Narayan Lal <nitesh@redhat.com>,
	kvm list <kvm@vger.kernel.org>,
	Dave Hansen <dave.hansen@intel.com>,
	LKML <linux-kernel@vger.kernel.org>,
	linux-mm <linux-mm@kvack.org>,
	Andrew Morton <akpm@linux-foundation.org>,
	Yang Zhang <yang.zhang.wz@gmail.com>,
	pagupta@redhat.com, Rik van Riel <riel@surriel.com>,
	Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>,
	lcapitulino@redhat.com, Andrea Arcangeli <aarcange@redhat.com>,
	Paolo Bonzini <pbonzini@redhat.com>,
	dan.j.williams@intel.com,
	Alexander Duyck <alexander.h.duyck@linux.intel.com>
Subject: Re: use of shrinker in virtio balloon free page hinting
Date: Wed, 17 Jul 2019 16:38:22 +0200	[thread overview]
Message-ID: <99c05ccc-5ce8-a192-ede8-40417a3f45dc@redhat.com> (raw)
In-Reply-To: <20190717103208-mutt-send-email-mst@kernel.org>

On 17.07.19 16:34, Michael S. Tsirkin wrote:
> On Wed, Jul 17, 2019 at 04:10:47PM +0200, David Hildenbrand wrote:
>> On 17.07.19 13:20, Michael S. Tsirkin wrote:
>>> Wei, others,
>>>
>>> ATM virtio_balloon_shrinker_scan will only get registered
>>> when deflate on oom feature bit is set.
>>>
>>> Not sure whether that's intentional.  Assuming it is:
>>>
>>> virtio_balloon_shrinker_scan will try to locate and free
>>> pages that are processed by host.
>>> The above seems broken in several ways:
>>> - count ignores the free page list completely
>>> - if free pages are being reported, pages freed
>>>   by shrinker will just get re-allocated again
>>
>> Trying to answer your questions (not sure if I fully understood what you
>> mean)
>>
>> virtio_balloon_shrinker_scan() will not be called due to inflation
>> requests (balloon_page_alloc()). It will be called whenever the system
>> is OOM, e.g., when starting a new application.
>>
>> I assume you were expecting the shrinker getting called due to
>> balloon_page_alloc(). however, that is not the case as we pass
>> "__GFP_NORETRY".
> 
> Right but it's possible we exhaust all memory, then
> someone else asks for a single page and that invokes
> the shrinker.

Yes, I think that can happen.

> 
>>
>> To test, something like:
>>
>> 1. Start a VM with
>>
>> -device virtio-balloon-pci,deflate-on-oom=true
>>
>> 2. Inflate the balloon, e.g.,
>>
>> QMP: balloon 1024
>> QMP: info balloon
>> -> 1024
>>
>> See how "MemTotal" in /proc/meminfo in the guest won't change
>>
>> 3. Run a workload that exhausts memory in the guest (OOM).
>>
>> See how the balloon was automatically deflated
>>
>> QMP: info balloon
>> -> Something bigger than 1024
>>
>>
>> Not sure if it is broken, last time I played with it, it worked, but
>> that was ~1-2 years ago.
>>
>> -- 
>>
>> Thanks,
>>
>> David / dhildenb
> 
> Sorry I was unclear.  The question was about
> VIRTIO_BALLOON_F_FREE_PAGE_HINT specifically.

Ah, I see. Never used both things together.

-- 

Thanks,

David / dhildenb


  reply	other threads:[~2019-07-17 14:38 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-07-17 11:20 Michael S. Tsirkin
2019-07-17 11:32 ` David Hildenbrand
2019-07-17 14:10 ` David Hildenbrand
2019-07-17 14:34   ` Michael S. Tsirkin
2019-07-17 14:38     ` David Hildenbrand [this message]
2019-07-17 15:46 ` Wang, Wei W
2019-07-18  4:13   ` Michael S. Tsirkin
2019-07-18  5:57     ` Wei Wang
2019-07-18  5:58       ` Michael S. Tsirkin
2019-07-18  6:30         ` Wei Wang
2019-07-18  6:47           ` Michael S. Tsirkin
2019-07-18  9:08             ` Wei Wang

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=99c05ccc-5ce8-a192-ede8-40417a3f45dc@redhat.com \
    --to=david@redhat.com \
    --cc=aarcange@redhat.com \
    --cc=akpm@linux-foundation.org \
    --cc=alexander.duyck@gmail.com \
    --cc=alexander.h.duyck@linux.intel.com \
    --cc=dan.j.williams@intel.com \
    --cc=dave.hansen@intel.com \
    --cc=konrad.wilk@oracle.com \
    --cc=kvm@vger.kernel.org \
    --cc=lcapitulino@redhat.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mst@redhat.com \
    --cc=nitesh@redhat.com \
    --cc=pagupta@redhat.com \
    --cc=pbonzini@redhat.com \
    --cc=riel@surriel.com \
    --cc=wei.w.wang@intel.com \
    --cc=yang.zhang.wz@gmail.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