From: Oscar Salvador <osalvador@suse.de>
To: akpm@linux-foundation.org
Cc: n-horiguchi@ah.jp.nec.com, linux-mm@kvack.org,
linux-kernel@vger.kernel.org, Oscar Salvador <osalvador@suse.de>
Subject: [PATCH 0/7] HWPoison: Refactor get page interface
Date: Thu, 19 Nov 2020 11:57:09 +0100 [thread overview]
Message-ID: <20201119105716.5962-1-osalvador@suse.de> (raw)
Hi,
following up on previous fix-ups an refactors, this patchset simplifies
the get page interface and removes the MF_COUNT_INCREASED trick we have
for soft offline.
Please, note that this patchset is on top of [1] and [2].
This patchset does three things:
1) Drops MF_COUNT_INCREASED trick
2) Refactors get page interface
3) Places a common entry for grabbin a page from both hard offline
and soft offline guarded by zone_pcp_{disable/enable}, so we do not
have to drain pcplists by ourself and retry again.
Note that the MF_COUNT_INCREASED trick was left because if get_hwpoison_page
races with put_page (e.g:)
CPU0 CPU1
put_page (refcount decremented to 0)
__put_single_page
free_unref_page
free_unref_page_prepare
free_pcp_prepare
free_pages_prepare soft_offline_page
:page->flags &= ~PAGE_FLAGS_CHECK_AT_PREP get_any_page
get_hwpoison_page
free_unref_page_commit
free_one_page
__free_one_page (place it in buddy)
get_hwpoison_page sees that page has a refcount of 0, but since it was not placed
in buddy yet we cannot really handle it.
We now have a sort of maximum passes in get_any_page, so in case we race
with either an allocation or a put_page, we retry again.
After an off-list discussion with Naoya, he agreed to proceed.
[1] https://patchwork.kernel.org/project/linux-mm/list/?series=364009
[2] https://patchwork.kernel.org/project/linux-mm/list/?series=381903
Naoya Horiguchi (3):
mm,madvise: call soft_offline_page() without MF_COUNT_INCREASED
mm,hwpoison: remove MF_COUNT_INCREASED
mm,hwpoison: remove flag argument from soft offline functions
Oscar Salvador (4):
mm,hwpoison: Refactor get_any_page
mm,hwpoison: Drop pfn parameter
mm,hwpoison: Disable pcplists before grabbing a refcount
mm,hwpoison: Remove drain_all_pages from shake_page
drivers/base/memory.c | 2 +-
include/linux/mm.h | 9 +--
mm/madvise.c | 19 +++--
mm/memory-failure.c | 168 +++++++++++++++++-------------------------
4 files changed, 85 insertions(+), 113 deletions(-)
--
2.26.2
next reply other threads:[~2020-11-19 10:57 UTC|newest]
Thread overview: 26+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-11-19 10:57 Oscar Salvador [this message]
2020-11-19 10:57 ` [PATCH 1/7] mm,hwpoison: Refactor get_any_page Oscar Salvador
2020-11-20 1:33 ` HORIGUCHI NAOYA(堀口 直也)
2020-11-25 16:54 ` Vlastimil Babka
2020-11-19 10:57 ` [PATCH 2/7] mm,hwpoison: Drop pfn parameter Oscar Salvador
2020-11-20 1:33 ` HORIGUCHI NAOYA(堀口 直也)
2020-11-25 16:55 ` Vlastimil Babka
2020-11-19 10:57 ` [PATCH 3/7] mm,madvise: call soft_offline_page() without MF_COUNT_INCREASED Oscar Salvador
2020-11-25 18:20 ` Vlastimil Babka
2020-12-01 11:35 ` Oscar Salvador
2020-12-04 17:25 ` Vlastimil Babka
2020-12-05 15:34 ` Oscar Salvador
2020-12-07 2:34 ` HORIGUCHI NAOYA(堀口 直也)
2020-12-07 7:24 ` Oscar Salvador
2020-11-19 10:57 ` [PATCH 4/7] mm,hwpoison: remove MF_COUNT_INCREASED Oscar Salvador
2020-11-19 10:57 ` [PATCH 5/7] mm,hwpoison: remove flag argument from soft offline functions Oscar Salvador
2020-11-19 10:57 ` [PATCH 6/7] mm,hwpoison: Disable pcplists before grabbing a refcount Oscar Salvador
2020-11-20 1:33 ` HORIGUCHI NAOYA(堀口 直也)
2020-11-26 13:45 ` Vlastimil Babka
2020-11-28 0:51 ` Andrew Morton
2020-11-19 10:57 ` [PATCH 7/7] mm,hwpoison: Remove drain_all_pages from shake_page Oscar Salvador
2020-11-20 1:33 ` HORIGUCHI NAOYA(堀口 直也)
2020-11-26 13:52 ` Vlastimil Babka
2020-11-27 7:20 ` Oscar Salvador
2020-12-02 13:34 ` [PATCH 0/7] HWPoison: Refactor get page interface Qian Cai
2020-12-02 13:41 ` Oscar Salvador
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=20201119105716.5962-1-osalvador@suse.de \
--to=osalvador@suse.de \
--cc=akpm@linux-foundation.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=n-horiguchi@ah.jp.nec.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).