From: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
To: Mike Kravetz <mike.kravetz@oracle.com>
Cc: "Kirill A. Shutemov" <kirill@shutemov.name>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
"Jérôme Glisse" <jglisse@redhat.com>,
"Vlastimil Babka" <vbabka@suse.cz>,
"Naoya Horiguchi" <n-horiguchi@ah.jp.nec.com>,
"Davidlohr Bueso" <dave@stgolabs.net>,
"Michal Hocko" <mhocko@kernel.org>,
"Andrew Morton" <akpm@linux-foundation.org>
Subject: Re: [PATCH] mm: migration: fix migration of huge PMD shared pages
Date: Tue, 14 Aug 2018 11:48:37 +0300 [thread overview]
Message-ID: <20180814084837.nl7dkea7aov2pzao@black.fi.intel.com> (raw)
In-Reply-To: <d4cf0f85-e010-36f2-3fae-f7983e4f6505@oracle.com>
On Mon, Aug 13, 2018 at 11:21:41PM +0000, Mike Kravetz wrote:
> On 08/13/2018 03:58 AM, Kirill A. Shutemov wrote:
> > On Sun, Aug 12, 2018 at 08:41:08PM -0700, Mike Kravetz wrote:
> >> The page migration code employs try_to_unmap() to try and unmap the
> >> source page. This is accomplished by using rmap_walk to find all
> >> vmas where the page is mapped. This search stops when page mapcount
> >> is zero. For shared PMD huge pages, the page map count is always 1
> >> not matter the number of mappings. Shared mappings are tracked via
> >> the reference count of the PMD page. Therefore, try_to_unmap stops
> >> prematurely and does not completely unmap all mappings of the source
> >> page.
> >>
> >> This problem can result is data corruption as writes to the original
> >> source page can happen after contents of the page are copied to the
> >> target page. Hence, data is lost.
> >>
> >> This problem was originally seen as DB corruption of shared global
> >> areas after a huge page was soft offlined. DB developers noticed
> >> they could reproduce the issue by (hotplug) offlining memory used
> >> to back huge pages. A simple testcase can reproduce the problem by
> >> creating a shared PMD mapping (note that this must be at least
> >> PUD_SIZE in size and PUD_SIZE aligned (1GB on x86)), and using
> >> migrate_pages() to migrate process pages between nodes.
> >>
> >> To fix, have the try_to_unmap_one routine check for huge PMD sharing
> >> by calling huge_pmd_unshare for hugetlbfs huge pages. If it is a
> >> shared mapping it will be 'unshared' which removes the page table
> >> entry and drops reference on PMD page. After this, flush caches and
> >> TLB.
> >>
> >> Signed-off-by: Mike Kravetz <mike.kravetz@oracle.com>
> >> ---
> >> I am not %100 sure on the required flushing, so suggestions would be
> >> appreciated. This also should go to stable. It has been around for
> >> a long time so still looking for an appropriate 'fixes:'.
> >
> > I believe we need flushing. And huge_pmd_unshare() usage in
> > __unmap_hugepage_range() looks suspicious: I don't see how we flush TLB in
> > that case.
>
> Thanks Kirill,
>
> __unmap_hugepage_range() has two callers:
> 1) unmap_hugepage_range, which wraps the call with tlb_gather_mmu and
> tlb_finish_mmu on the range. IIUC, this should cause an appropriate
> TLB flush.
> 2) __unmap_hugepage_range_final via unmap_single_vma. unmap_single_vma
> has three callers:
> - unmap_vmas which assumes the caller will flush the whole range after
> return.
> - zap_page_range wraps the call with tlb_gather_mmu/tlb_finish_mmu
> - zap_page_range_single wraps the call with tlb_gather_mmu/tlb_finish_mmu
>
> So, it appears we are covered. But, I could be missing something.
My problem here is that the mapping that moved by huge_pmd_unshare() in
not accounted into mmu_gather and can be missed on tlb_finish_mmu().
--
Kirill A. Shutemov
next prev parent reply other threads:[~2018-08-14 8:48 UTC|newest]
Thread overview: 10+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-08-13 3:41 [PATCH] mm: migration: fix migration of huge PMD shared pages Mike Kravetz
2018-08-13 4:10 ` kbuild test robot
2018-08-14 0:30 ` [PATCH v2] " Mike Kravetz
2018-08-14 6:28 ` Greg KH
2018-08-13 4:17 ` [PATCH] " kbuild test robot
2018-08-13 10:58 ` Kirill A. Shutemov
2018-08-13 23:21 ` Mike Kravetz
2018-08-14 8:48 ` Kirill A. Shutemov [this message]
2018-08-15 0:15 ` Mike Kravetz
2018-08-15 8:47 ` Kirill A. Shutemov
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=20180814084837.nl7dkea7aov2pzao@black.fi.intel.com \
--to=kirill.shutemov@linux.intel.com \
--cc=akpm@linux-foundation.org \
--cc=dave@stgolabs.net \
--cc=jglisse@redhat.com \
--cc=kirill@shutemov.name \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@kernel.org \
--cc=mike.kravetz@oracle.com \
--cc=n-horiguchi@ah.jp.nec.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).