From: "Huang, Ying" <firstname.lastname@example.org>
To: Andrew Morton <email@example.com>
Cc: Peter Xu <firstname.lastname@example.org>,
Hugh Dickins <email@example.com>,
"Kirill A . Shutemov" <firstname.lastname@example.org>,
Alistair Popple <email@example.com>,
Andrea Arcangeli <firstname.lastname@example.org>,
Minchan Kim <email@example.com>,
David Hildenbrand <firstname.lastname@example.org>,
Andi Kleen <email@example.com>,
Nadav Amit <firstname.lastname@example.org>,
Vlastimil Babka <email@example.com>
Subject: Re: [PATCH v4 5/7] mm: Remember young/dirty bit for page migrations
Date: Tue, 13 Sep 2022 08:55:35 +0800 [thread overview]
Message-ID: <firstname.lastname@example.org> (raw)
In-Reply-To: <email@example.com> (Andrew Morton's message of "Sun, 11 Sep 2022 16:48:28 -0700")
Andrew Morton <firstname.lastname@example.org> writes:
> On Thu, 11 Aug 2022 12:13:29 -0400 Peter Xu <email@example.com> wrote:
>> When page migration happens, we always ignore the young/dirty bit settings
>> in the old pgtable, and marking the page as old in the new page table using
>> either pte_mkold() or pmd_mkold(), and keeping the pte clean.
>> That's fine from functional-wise, but that's not friendly to page reclaim
>> because the moving page can be actively accessed within the procedure. Not
>> to mention hardware setting the young bit can bring quite some overhead on
>> some systems, e.g. x86_64 needs a few hundreds nanoseconds to set the bit.
>> The same slowdown problem to dirty bits when the memory is first written
>> after page migration happened.
>> Actually we can easily remember the A/D bit configuration and recover the
>> information after the page is migrated. To achieve it, define a new set of
>> bits in the migration swap offset field to cache the A/D bits for old pte.
>> Then when removing/recovering the migration entry, we can recover the A/D
>> bits even if the page changed.
>> One thing to mention is that here we used max_swapfile_size() to detect how
>> many swp offset bits we have, and we'll only enable this feature if we know
>> the swp offset is big enough to store both the PFN value and the A/D bits.
>> Otherwise the A/D bits are dropped like before.
> There was some discussion over v3 of this patch, but none over v4.
> Can people please review this patch series so we can get moving with it?
Most discussions over v3 are for migrate_device.c code. There are some
bugs and they have been fixed by Alistair via .
This patch itself is good. Sorry for bothering.
Reviewed-by: "Huang, Ying" <firstname.lastname@example.org>
next prev parent reply other threads:[~2022-09-13 0:55 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-08-11 16:13 [PATCH v4 0/7] mm: Remember a/d bits for migration entries Peter Xu
2022-08-11 16:13 ` [PATCH v4 1/7] mm/x86: Use SWP_TYPE_BITS in 3-level swap macros Peter Xu
2022-08-11 16:13 ` [PATCH v4 2/7] mm/swap: Comment all the ifdef in swapops.h Peter Xu
2022-08-15 6:03 ` Alistair Popple
2022-08-11 16:13 ` [PATCH v4 3/7] mm/swap: Add swp_offset_pfn() to fetch PFN from swap entry Peter Xu
2022-08-12 2:33 ` Huang, Ying
2022-08-23 21:01 ` Yu Zhao
2022-08-23 22:04 ` Peter Xu
2022-08-11 16:13 ` [PATCH v4 4/7] mm/thp: Carry over dirty bit when thp splits on pmd Peter Xu
2022-10-21 16:06 ` dpkg fails on sparc64 (was: [PATCH v4 4/7] mm/thp: Carry over dirty bit when thp splits on pmd) Anatoly Pugachev
2022-10-23 13:33 ` dpkg fails on sparc64 (was: [PATCH v4 4/7] mm/thp: Carry over dirty bit when thp splits on pmd) #forregzbot Thorsten Leemhuis
2022-11-04 10:39 ` Thorsten Leemhuis
2022-11-13 17:56 ` Thorsten Leemhuis
2022-10-23 19:52 ` dpkg fails on sparc64 (was: [PATCH v4 4/7] mm/thp: Carry over dirty bit when thp splits on pmd) Peter Xu
2022-10-25 10:22 ` Anatoly Pugachev
2022-10-25 14:43 ` Peter Xu
2022-11-01 13:13 ` Anatoly Pugachev
2022-11-02 18:34 ` Peter Xu
2022-11-02 18:47 ` Andrew Morton
2022-08-11 16:13 ` [PATCH v4 5/7] mm: Remember young/dirty bit for page migrations Peter Xu
2022-09-11 23:48 ` Andrew Morton
2022-09-13 0:55 ` Huang, Ying [this message]
2022-08-11 16:13 ` [PATCH v4 6/7] mm/swap: Cache maximum swapfile size when init swap Peter Xu
2022-08-12 2:34 ` Huang, Ying
2022-08-11 16:13 ` [PATCH v4 7/7] mm/swap: Cache swap migration A/D bits support Peter Xu
2022-11-21 5:15 ` [PATCH v4 0/7] mm: Remember a/d bits for migration entries Raghavendra K T
2022-11-21 14:57 ` Peter Xu
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:
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
* 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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.