From: "Kirill A. Shutemov" <kirill@shutemov.name>
To: Yang Shi <yang.shi@linux.alibaba.com>
Cc: akpm@linux-foundation.org, Andrea Arcangeli <aarcange@redhat.com>,
Zi Yan <ziy@nvidia.com>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
"Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Subject: Re: [PATCHv2 5/8] khugepaged: Allow to callapse a page shared across fork
Date: Wed, 8 Apr 2020 16:10:44 +0300 [thread overview]
Message-ID: <20200408131044.xzlheacvslrbwrja@box> (raw)
In-Reply-To: <b03643ba-8411-8486-737c-1bc29dd10a74@linux.alibaba.com>
On Mon, Apr 06, 2020 at 01:50:56PM -0700, Yang Shi wrote:
>
>
> On 4/3/20 4:29 AM, Kirill A. Shutemov wrote:
> > The page can be included into collapse as long as it doesn't have extra
> > pins (from GUP or otherwise).
> >
> > Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> > ---
> > mm/khugepaged.c | 25 ++++++++++++++-----------
> > 1 file changed, 14 insertions(+), 11 deletions(-)
> >
> > diff --git a/mm/khugepaged.c b/mm/khugepaged.c
> > index 57ff287caf6b..1e7e6543ebca 100644
> > --- a/mm/khugepaged.c
> > +++ b/mm/khugepaged.c
> > @@ -581,11 +581,18 @@ static int __collapse_huge_page_isolate(struct vm_area_struct *vma,
> > }
> > /*
> > - * cannot use mapcount: can't collapse if there's a gup pin.
> > - * The page must only be referenced by the scanned process
> > - * and page swap cache.
> > + * Check if the page has any GUP (or other external) pins.
> > + *
> > + * The page table that maps the page has been already unlinked
> > + * from the page table tree and this process cannot get
> > + * additinal pin on the page.
> > + *
> > + * New pins can come later if the page is shared across fork,
> > + * but not for the this process. It is fine. The other process
> > + * cannot write to the page, only trigger CoW.
> > */
> > - if (page_count(page) != 1 + PageSwapCache(page)) {
> > + if (total_mapcount(page) + PageSwapCache(page) !=
> > + page_count(page)) {
>
> This check looks fine for base page, but what if the page is PTE-mapped THP?
> The following patch made this possible.
>
> If it is PTE-mapped THP and the page is in swap cache, the refcount would be
> 512 + the number of PTE-mapped pages.
>
> Shall we do the below change in the following patch?
>
> extra_pins = PageSwapCache(page) ? nr_ccompound(page) - 1 : 0;
> if (total_mapcount(page) + PageSwapCache(page) != page_count(page) -
> extra_pins) {
> ...
Looks like you're right.
It would be nice to have a test case to demonstrate the issue.
Is there any way to trigger moving the page to swap cache? I don't see it
immediately.
--
Kirill A. Shutemov
next prev parent reply other threads:[~2020-04-08 13:10 UTC|newest]
Thread overview: 44+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-04-03 11:29 [PATCHv2 0/8] thp/khugepaged improvements and CoW semantics Kirill A. Shutemov
2020-04-03 11:29 ` [PATCHv2 1/8] khugepaged: Add self test Kirill A. Shutemov
2020-04-06 14:59 ` Zi Yan
2020-04-06 15:20 ` Kirill A. Shutemov
2020-04-06 18:50 ` Zi Yan
2020-04-08 14:21 ` Kirill A. Shutemov
2020-04-08 15:53 ` Zi Yan
2020-04-10 11:47 ` Kirill A. Shutemov
2020-04-10 14:36 ` Zi Yan
2020-04-10 14:58 ` Kirill A. Shutemov
2020-04-10 15:03 ` Zi Yan
2020-04-06 18:53 ` Ralph Campbell
2020-04-03 11:29 ` [PATCHv2 2/8] khugepaged: Do not stop collapse if less than half PTEs are referenced Kirill A. Shutemov
2020-04-06 18:13 ` Yang Shi
2020-04-06 19:53 ` Ralph Campbell
2020-04-09 13:34 ` Kirill A. Shutemov
2020-04-03 11:29 ` [PATCHv2 3/8] khugepaged: Drain all LRU caches before scanning pages Kirill A. Shutemov
2020-04-06 18:15 ` Yang Shi
2020-04-03 11:29 ` [PATCHv2 4/8] khugepaged: Drain LRU add pagevec after swapin Kirill A. Shutemov
2020-04-06 13:11 ` Zi Yan
2020-04-06 18:29 ` Yang Shi
2020-04-08 13:05 ` Kirill A. Shutemov
2020-04-08 18:42 ` Yang Shi
2020-04-03 11:29 ` [PATCHv2 5/8] khugepaged: Allow to callapse a page shared across fork Kirill A. Shutemov
2020-04-06 20:15 ` Ralph Campbell
2020-04-06 20:50 ` Yang Shi
2020-04-08 13:10 ` Kirill A. Shutemov [this message]
2020-04-08 18:51 ` Yang Shi
2020-04-10 0:03 ` Yang Shi
2020-04-10 15:56 ` Kirill A. Shutemov
2020-04-06 21:30 ` John Hubbard
2020-04-10 15:55 ` Kirill A. Shutemov
2020-04-10 20:59 ` John Hubbard
2020-04-13 9:42 ` Kirill A. Shutemov
2020-04-03 11:29 ` [PATCHv2 6/8] khugepaged: Allow to collapse PTE-mapped compound pages Kirill A. Shutemov
2020-04-06 21:29 ` Yang Shi
2020-04-08 13:29 ` Kirill A. Shutemov
2020-04-08 18:57 ` Yang Shi
2020-04-09 13:47 ` Kirill A. Shutemov
2020-04-03 11:29 ` [PATCHv2 7/8] thp: Change CoW semantics for anon-THP Kirill A. Shutemov
2020-04-07 7:57 ` [thp] db001b7115: vm-scalability.median 8.9% improvement kernel test robot
2020-04-03 11:29 ` [PATCHv2 8/8] khugepaged: Introduce 'max_ptes_shared' tunable Kirill A. Shutemov
2020-04-06 13:17 ` Zi Yan
2020-04-05 23:40 ` [PATCHv2 0/8] thp/khugepaged improvements and CoW semantics William Kucharski
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=20200408131044.xzlheacvslrbwrja@box \
--to=kirill@shutemov.name \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=yang.shi@linux.alibaba.com \
--cc=ziy@nvidia.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).