From: Anshuman Khandual <khandual@linux.vnet.ibm.com>
To: Nitin Gupta <nitin.m.gupta@oracle.com>, linux-mm@kvack.org
Cc: steven.sistare@oracle.com,
"Andrew Morton
(commit_signer:14/26=54%,commit_signer:10/16=62%,commit_signer:24/26=92%,commit_signer:48/63=76%)"
<akpm@linux-foundation.org>,
"Ingo Molnar
(commit_signer:6/26=23%,authored:4/26=15%,added_lines:17/189=9%,removed_lines:52/150=35%,authored:2/16=12%,added_lines:2/25=8%,authored:4/63=6%)"
<mingo@kernel.org>,
"Mel Gorman (commit_signer:5/26=19%,authored:2/26=8%)"
<mgorman@suse.de>,
"Nadav Amit
(commit_signer:5/26=19%,authored:2/26=8%,added_lines:32/189=17%,removed_lines:13/150=9%)"
<namit@vmware.com>,
"Minchan Kim
(commit_signer:4/26=15%,authored:3/26=12%,added_lines:14/189=7%,removed_lines:21/150=14%,removed_lines:2/40=5%,commit_signer:5/26=19%,authored:4/63=6%,added_lines:83/883=9%,removed_lines:34/354=10%)"
<minchan@kernel.org>,
"Kirill A. Shutemov
(authored:3/26=12%,commit_signer:4/16=25%,authored:2/16=12%,commit_signer:12/63=19%,authored:8/63=13%,added_lines:214/883=24%,removed_lines:56/354=16%)"
<kirill.shutemov@linux.intel.com>,
"Peter Zijlstra
(authored:2/26=8%,added_lines:72/189=38%,removed_lines:39/150=26%)"
<peterz@infradead.org>,
"Vegard Nossum (added_lines:21/189=11%)"
<vegard.nossum@oracle.com>,
"Levin,
Alexander (Sasha Levin) (removed_lines:8/150=5%)"
<alexander.levin@verizon.com>,
"Michal Hocko
(commit_signer:7/16=44%,authored:2/16=12%,added_lines:4/25=16%,removed_lines:4/40=10%,commit_signer:7/26=27%,commit_signer:15/63=24%,removed_lines:32/354=9%)"
<mhocko@suse.com>,
"David Rientjes
(commit_signer:3/16=19%,authored:2/16=12%,added_lines:3/25=12%,removed_lines:5/40=12%,added_lines:42/189=22%,removed_lines:9/73=12%)"
<rientjes@google.com>,
"Vlastimil Babka (commit_signer:3/16=19%)" <vbabka@suse.cz>,
"SeongJae Park (authored:1/16=6%,added_lines:3/25=12%)"
<sj38.park@gmail.com>,
"Shaohua Li
(added_lines:3/25=12%,removed_lines:5/40=12%,authored:4/26=15%,removed_lines:11/73=15%)"
<shli@fb.com>,
"Aneesh Kumar K.V (removed_lines:19/40=48%)"
<aneesh.kumar@linux.vnet.ibm.com>,
"Andrea Arcangeli
(commit_signer:5/26=19%,authored:2/26=8%,added_lines:42/189=22%,removed_lines:4/73=5%)"
<aarcange@redhat.com>,
"Mike Rapoport
(commit_signer:5/26=19%,authored:3/26=12%,added_lines:24/189=13%,removed_lines:21/73=29%)"
<rppt@linux.vnet.ibm.com>,
"Anshuman Khandual
(authored:2/26=8%,added_lines:29/189=15%,removed_lines:18/73=25%)"
<khandual@linux.vnet.ibm.com>,
"Rik van Riel (added_lines:13/189=7%)" <riel@redhat.com>,
"Ross Zwisler
(commit_signer:8/63=13%,authored:4/63=6%,added_lines:105/883=12%)"
<ross.zwisler@linux.intel.com>,
"Jan Kara (commit_signer:7/63=11%)" <jack@suse.cz>,
"Dave Jiang (authored:5/63=8%)" <dave.jiang@intel.com>,
"Jérôme Glisse (added_lines:128/883=14%)" <jglisse@redhat.com>,
"Matthew Wilcox (added_lines:81/883=9%)" <willy@linux.intel.com>,
"Hugh Dickins (removed_lines:65/354=18%)" <hughd@google.com>,
"Tobin C Harding (removed_lines:34/354=10%)" <me@tobin.cc>,
"open list" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH] mm: Reduce memory bloat with THP
Date: Fri, 15 Dec 2017 11:25:11 +0530 [thread overview]
Message-ID: <cb4f46ed-6ca9-4dd3-a21d-7a87ec348da1@linux.vnet.ibm.com> (raw)
In-Reply-To: <1513301359-117568-1-git-send-email-nitin.m.gupta@oracle.com>
On 12/15/2017 06:58 AM, Nitin Gupta wrote:
> Currently, if the THP enabled policy is "always", or the mode
> is "madvise" and a region is marked as MADV_HUGEPAGE, a hugepage
> is allocated on a page fault if the pud or pmd is empty. This
> yields the best VA translation performance, but increases memory
> consumption if some small page ranges within the huge page are
> never accessed.
Right, thats as per design.
>
> An alternate behavior for such page faults is to install a
> hugepage only when a region is actually found to be (almost)
> fully mapped and active. This is a compromise between
That is the async method by analyzing page table segment for
the process by khugepaged and evaluate if a huge page can be
installed replacing the existing pages.
> translation performance and memory consumption. Currently there
> is no way for an application to choose this compromise for the
> page fault conditions above.
Cant we mark the THP enablement mode as "madvise", then switch
between MADV_HUGEPAGE/MADV_NOHUGEPAGE to implement this ?
>
> With this change, when an application issues MADV_DONTNEED on a
> memory region, the region is marked as "space-efficient". For
Isn't it that MADV_DONTNEED should be used for a region where
there are already pages faulted in and page table populated ?
Are you suggesting that MADV_DONTNEED should be called upon
a region just after creation to control it's fault behavior ?
Thats not what MADV_DONTNEED was meant to be.
> such regions, a hugepage is not immediately allocated on first
> write. Instead, it is left to the khugepaged thread to do
> delayed hugepage promotion depending on whether the region is
> actually mapped and active. When application issues
> MADV_HUGEPAGE, the region is marked again as non-space-efficient
> wherein hugepage is allocated on first touch
But MADV_HUGEPAGE/MADV_NOHUGEPAGE combination should do the trick
as well.
--
To unsubscribe, send a message with 'unsubscribe linux-mm' in
the body to majordomo@kvack.org. For more info on Linux MM,
see: http://www.linux-mm.org/ .
Don't email: <a href=mailto:"dont@kvack.org"> email@kvack.org </a>
next prev parent reply other threads:[~2017-12-15 5:55 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-12-15 1:28 Nitin Gupta
2017-12-15 5:55 ` Anshuman Khandual [this message]
2017-12-16 7:18 ` Nitin Gupta
2017-12-15 10:00 ` Kirill A. Shutemov
2017-12-16 7:04 ` Nitin Gupta
2017-12-18 13:53 ` Peter Zijlstra
2017-12-15 10:01 ` Kirill A. Shutemov
2017-12-16 7:21 ` Nitin Gupta
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=cb4f46ed-6ca9-4dd3-a21d-7a87ec348da1@linux.vnet.ibm.com \
--to=khandual@linux.vnet.ibm.com \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=alexander.levin@verizon.com \
--cc=aneesh.kumar@linux.vnet.ibm.com \
--cc=dave.jiang@intel.com \
--cc=hughd@google.com \
--cc=jack@suse.cz \
--cc=jglisse@redhat.com \
--cc=kirill.shutemov@linux.intel.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=me@tobin.cc \
--cc=mgorman@suse.de \
--cc=mhocko@suse.com \
--cc=minchan@kernel.org \
--cc=mingo@kernel.org \
--cc=namit@vmware.com \
--cc=nitin.m.gupta@oracle.com \
--cc=peterz@infradead.org \
--cc=riel@redhat.com \
--cc=rientjes@google.com \
--cc=ross.zwisler@linux.intel.com \
--cc=rppt@linux.vnet.ibm.com \
--cc=shli@fb.com \
--cc=sj38.park@gmail.com \
--cc=steven.sistare@oracle.com \
--cc=vbabka@suse.cz \
--cc=vegard.nossum@oracle.com \
--cc=willy@linux.intel.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