From: Hugh Dickins <hughd@google.com>
To: Huang Ying <ying.huang@intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
Jan Kara <jack@suse.cz>,
linux-mm@kvack.org, linux-kernel@vger.kernel.org,
linux-fsdevel@vger.kernel.org, Zi Yan <ziy@nvidia.com>,
Yang Shi <shy828301@gmail.com>,
Baolin Wang <baolin.wang@linux.alibaba.com>,
Oscar Salvador <osalvador@suse.de>,
Matthew Wilcox <willy@infradead.org>,
Bharata B Rao <bharata@amd.com>,
Alistair Popple <apopple@nvidia.com>,
Xin Hao <xhao@linux.alibaba.com>,
Minchan Kim <minchan@kernel.org>,
Mike Kravetz <mike.kravetz@oracle.com>,
Hyeonggon Yoo <42.hyeyoo@gmail.com>
Subject: Re: [PATCH -v5 0/9] migrate_pages(): batch TLB flushing
Date: Fri, 17 Feb 2023 13:47:48 -0800 (PST) [thread overview]
Message-ID: <87a6c8c-c5c1-67dc-1e32-eb30831d6e3d@google.com> (raw)
In-Reply-To: <20230213123444.155149-1-ying.huang@intel.com>
On Mon, 13 Feb 2023, Huang Ying wrote:
> From: "Huang, Ying" <ying.huang@intel.com>
>
> Now, migrate_pages() migrate folios one by one, like the fake code as
> follows,
>
> for each folio
> unmap
> flush TLB
> copy
> restore map
>
> If multiple folios are passed to migrate_pages(), there are
> opportunities to batch the TLB flushing and copying. That is, we can
> change the code to something as follows,
>
> for each folio
> unmap
> for each folio
> flush TLB
> for each folio
> copy
> for each folio
> restore map
>
> The total number of TLB flushing IPI can be reduced considerably. And
> we may use some hardware accelerator such as DSA to accelerate the
> folio copying.
>
> So in this patch, we refactor the migrate_pages() implementation and
> implement the TLB flushing batching. Base on this, hardware
> accelerated folio copying can be implemented.
>
> If too many folios are passed to migrate_pages(), in the naive batched
> implementation, we may unmap too many folios at the same time. The
> possibility for a task to wait for the migrated folios to be mapped
> again increases. So the latency may be hurt. To deal with this
> issue, the max number of folios be unmapped in batch is restricted to
> no more than HPAGE_PMD_NR in the unit of page. That is, the influence
> is at the same level of THP migration.
>
> We use the following test to measure the performance impact of the
> patchset,
>
> On a 2-socket Intel server,
>
> - Run pmbench memory accessing benchmark
>
> - Run `migratepages` to migrate pages of pmbench between node 0 and
> node 1 back and forth.
>
> With the patch, the TLB flushing IPI reduces 99.1% during the test and
> the number of pages migrated successfully per second increases 291.7%.
>
> Xin Hao helped to test the patchset on an ARM64 server with 128 cores,
> 2 NUMA nodes. Test results show that the page migration performance
> increases up to 78%.
>
> This patchset is based on mm-unstable 2023-02-10.
And back in linux-next this week: I tried next-20230217 overnight.
There is a deadlock in this patchset (and in previous versions: sorry
it's taken me so long to report), but I think one that's easily solved.
I've not bisected to precisely which patch (load can take several hours
to hit the deadlock), but it doesn't really matter, and I expect that
you can guess.
My root and home filesystems are ext4 (4kB blocks with 4kB PAGE_SIZE),
and so is the filesystem I'm testing, ext4 on /dev/loop0 on tmpfs.
So, plenty of ext4 page cache and buffer_heads.
Again and again, the deadlock is seen with buffer_migrate_folio_norefs(),
either in kcompactd0 or in khugepaged trying to compact, or in both:
it ends up calling __lock_buffer(), and that schedules away, waiting
forever to get BH_lock. I have not identified who is holding BH_lock,
but I imagine a jbd2 journalling thread, and presume that it wants one
of the folio locks which migrate_pages_batch() is already holding; or
maybe it's all more convoluted than that. Other tasks then back up
waiting on those folio locks held in the batch.
Never a problem with buffer_migrate_folio(), always with the "more
careful" buffer_migrate_folio_norefs(). And the patch below fixes
it for me: I've had enough hours with it now, on enough occasions,
to be confident of that.
Cc'ing Jan Kara, who knows buffer_migrate_folio_norefs() and jbd2
very well, and I hope can assure us that there is an understandable
deadlock here, from holding several random folio locks, then trying
to lock buffers. Cc'ing fsdevel, because there's a risk that mm
folk think something is safe, when it's not sufficient to cope with
the diversity of filesystems. I hope nothing more than the below is
needed (and I've had no other problems with the patchset: good job),
but cannot be sure.
[PATCH next] migrate_pages: fix deadlock on buffer heads
When __buffer_migrate_folio() is called from buffer_migrate_folio_norefs(),
force MIGRATE_ASYNC mode so that buffer_migrate_lock_buffers() will only
trylock_buffer(), failing with -EAGAIN as usual if that does not succeed.
Signed-off-by: Hugh Dickins <hughd@google.com>
--- next-20230217/mm/migrate.c
+++ fixed/mm/migrate.c
@@ -748,7 +748,8 @@ static int __buffer_migrate_folio(struct
if (folio_ref_count(src) != expected_count)
return -EAGAIN;
- if (!buffer_migrate_lock_buffers(head, mode))
+ if (!buffer_migrate_lock_buffers(head,
+ check_refs ? MIGRATE_ASYNC : mode))
return -EAGAIN;
if (check_refs) {
next prev parent reply other threads:[~2023-02-17 21:48 UTC|newest]
Thread overview: 23+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-13 12:34 [PATCH -v5 0/9] migrate_pages(): batch TLB flushing Huang Ying
2023-02-13 12:34 ` [PATCH -v5 1/9] migrate_pages: organize stats with struct migrate_pages_stats Huang Ying
2023-02-13 12:34 ` [PATCH -v5 2/9] migrate_pages: separate hugetlb folios migration Huang Ying
2023-02-13 12:34 ` [PATCH -v5 3/9] migrate_pages: restrict number of pages to migrate in batch Huang Ying
2023-02-13 12:34 ` [PATCH -v5 4/9] migrate_pages: split unmap_and_move() to _unmap() and _move() Huang Ying
2023-02-13 12:34 ` [PATCH -v5 5/9] migrate_pages: batch _unmap and _move Huang Ying
2023-02-13 12:34 ` [PATCH -v5 6/9] migrate_pages: move migrate_folio_unmap() Huang Ying
2023-02-13 12:34 ` [PATCH -v5 7/9] migrate_pages: share more code between _unmap and _move Huang Ying
2023-02-13 12:34 ` [PATCH -v5 8/9] migrate_pages: batch flushing TLB Huang Ying
2023-02-13 12:34 ` [PATCH -v5 9/9] migrate_pages: move THP/hugetlb migration support check to simplify code Huang Ying
2023-02-17 21:47 ` Hugh Dickins [this message]
2023-02-20 9:28 ` [PATCH -v5 0/9] migrate_pages(): batch TLB flushing Huang, Ying
2023-02-21 2:48 ` Hugh Dickins
2023-02-21 3:34 ` Huang, Ying
2023-02-21 22:15 ` Hugh Dickins
2023-02-21 4:30 ` Matthew Wilcox
2023-02-21 4:38 ` Huang, Ying
2023-02-21 14:04 ` Huang, Ying
2023-02-21 22:25 ` Hugh Dickins
2023-02-22 1:02 ` Huang, Ying
2023-02-27 11:06 ` Jan Kara
2023-02-28 1:13 ` Huang, Ying
2023-02-28 5:59 ` Hugh Dickins
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=87a6c8c-c5c1-67dc-1e32-eb30831d6e3d@google.com \
--to=hughd@google.com \
--cc=42.hyeyoo@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=apopple@nvidia.com \
--cc=baolin.wang@linux.alibaba.com \
--cc=bharata@amd.com \
--cc=jack@suse.cz \
--cc=linux-fsdevel@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mike.kravetz@oracle.com \
--cc=minchan@kernel.org \
--cc=osalvador@suse.de \
--cc=shy828301@gmail.com \
--cc=willy@infradead.org \
--cc=xhao@linux.alibaba.com \
--cc=ying.huang@intel.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).