mbox series

[-v5,0/9] migrate_pages(): batch TLB flushing

Message ID 20230213123444.155149-1-ying.huang@intel.com (mailing list archive)
Headers show
Series migrate_pages(): batch TLB flushing | expand

Message

Huang, Ying Feb. 13, 2023, 12:34 p.m. UTC
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.

Changes:

v5:

- Added Xin's test results on ARM 64.  Thanks Xin!

- Fixed many comments.  Thanks Zi and Xin!

- Collected reviewed-by and tested-by.

v4:

- Fixed another bug about non-LRU folio migration.  Thanks Hyeonggon!

v3:

- Rebased on v6.2-rc4

- Fixed a bug about non-LRU folio migration.  Thanks Mike!

- Fixed some comments.  Thanks Baolin!

- Collected reviewed-by.

v2:

- Rebased on v6.2-rc3

- Fixed type force cast warning.  Thanks Kees!

- Added more comments and cleaned up the code.  Thanks Andrew, Zi, Alistair, Dan!

- Collected reviewed-by.

from rfc to v1:

- Rebased on v6.2-rc1

- Fix the deadlock issue caused by locking multiple pages synchronously
  per Alistair's comments.  Thanks!

- Fix the autonumabench panic per Rao's comments and fix.  Thanks!

- Other minor fixes per comments. Thanks!

Best Regards,
Huang, Ying

Comments

Hugh Dickins Feb. 17, 2023, 9:47 p.m. UTC | #1
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) {
Huang, Ying Feb. 20, 2023, 9:28 a.m. UTC | #2
Hi, Hugh,

Hugh Dickins <hughd@google.com> writes:

> 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) {

Thank you very much for pointing this out and the fix patch.  Today, my
colleague Pengfei reported a deadlock bug to me.  It seems that we
cannot wait the writeback to complete when we have locked some folios.
Below patch can fix that deadlock.  I don't know whether this is related
to the deadlock you run into.  It appears that we should avoid to
lock/wait synchronously if we have locked more than one folios.

Best Regards,
Huang, Ying

------------------------------------8<------------------------------------
From 0699fa2f80a67e863107d49a25909c92b900a9be Mon Sep 17 00:00:00 2001
From: Huang Ying <ying.huang@intel.com>
Date: Mon, 20 Feb 2023 14:56:34 +0800
Subject: [PATCH] migrate_pages: fix deadlock on waiting writeback

Pengfei reported a system soft lockup issue with Syzkaller.  The stack
traces are as follows,

...
[  300.124933] INFO: task kworker/u4:3:73 blocked for more than 147 seconds.
[  300.125214]       Not tainted 6.2.0-rc4-kvm+ #1314
[  300.125408] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[  300.125736] task:kworker/u4:3    state:D stack:0     pid:73    ppid:2      flags:0x00004000
[  300.126059] Workqueue: writeback wb_workfn (flush-7:3)
[  300.126282] Call Trace:
[  300.126378]  <TASK>
[  300.126464]  __schedule+0x43b/0xd00
[  300.126601]  ? __blk_flush_plug+0x142/0x180
[  300.126765]  schedule+0x6a/0xf0
[  300.126912]  io_schedule+0x4a/0x80
[  300.127051]  folio_wait_bit_common+0x1b5/0x4e0
[  300.127227]  ? __pfx_wake_page_function+0x10/0x10
[  300.127403]  __folio_lock+0x27/0x40
[  300.127541]  write_cache_pages+0x350/0x870
[  300.127699]  ? __pfx_iomap_do_writepage+0x10/0x10
[  300.127889]  iomap_writepages+0x3f/0x80
[  300.128037]  xfs_vm_writepages+0x94/0xd0
[  300.128192]  ? __pfx_xfs_vm_writepages+0x10/0x10
[  300.128370]  do_writepages+0x10a/0x240
[  300.128514]  ? lock_is_held_type+0xe6/0x140
[  300.128675]  __writeback_single_inode+0x9f/0xa90
[  300.128854]  writeback_sb_inodes+0x2fb/0x8d0
[  300.129030]  __writeback_inodes_wb+0x68/0x150
[  300.129212]  wb_writeback+0x49c/0x770
[  300.129357]  wb_workfn+0x6fb/0x9d0
[  300.129500]  process_one_work+0x3cc/0x8d0
[  300.129669]  worker_thread+0x66/0x630
[  300.129824]  ? __pfx_worker_thread+0x10/0x10
[  300.129989]  kthread+0x153/0x190
[  300.130116]  ? __pfx_kthread+0x10/0x10
[  300.130264]  ret_from_fork+0x29/0x50
[  300.130409]  </TASK>
[  300.179347] INFO: task repro:1023 blocked for more than 147 seconds.
[  300.179905]       Not tainted 6.2.0-rc4-kvm+ #1314
[  300.180317] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[  300.180955] task:repro           state:D stack:0     pid:1023  ppid:360    flags:0x00004004
[  300.181660] Call Trace:
[  300.181879]  <TASK>
[  300.182085]  __schedule+0x43b/0xd00
[  300.182407]  schedule+0x6a/0xf0
[  300.182694]  io_schedule+0x4a/0x80
[  300.183020]  folio_wait_bit_common+0x1b5/0x4e0
[  300.183506]  ? compaction_alloc+0x77/0x1150
[  300.183892]  ? __pfx_wake_page_function+0x10/0x10
[  300.184304]  folio_wait_bit+0x30/0x40
[  300.184640]  folio_wait_writeback+0x2e/0x1e0
[  300.185034]  migrate_pages_batch+0x555/0x1ac0
[  300.185462]  ? __pfx_compaction_alloc+0x10/0x10
[  300.185808]  ? __pfx_compaction_free+0x10/0x10
[  300.186022]  ? __this_cpu_preempt_check+0x17/0x20
[  300.186234]  ? lock_is_held_type+0xe6/0x140
[  300.186423]  migrate_pages+0x100e/0x1180
[  300.186603]  ? __pfx_compaction_free+0x10/0x10
[  300.186800]  ? __pfx_compaction_alloc+0x10/0x10
[  300.187011]  compact_zone+0xe10/0x1b50
[  300.187182]  ? lock_is_held_type+0xe6/0x140
[  300.187374]  ? check_preemption_disabled+0x80/0xf0
[  300.187588]  compact_node+0xa3/0x100
[  300.187755]  ? __sanitizer_cov_trace_const_cmp8+0x1c/0x30
[  300.187993]  ? _find_first_bit+0x7b/0x90
[  300.188171]  sysctl_compaction_handler+0x5d/0xb0
[  300.188376]  proc_sys_call_handler+0x29d/0x420
[  300.188583]  proc_sys_write+0x2b/0x40
[  300.188749]  vfs_write+0x3a3/0x780
[  300.188912]  ksys_write+0xb7/0x180
[  300.189070]  __x64_sys_write+0x26/0x30
[  300.189260]  do_syscall_64+0x3b/0x90
[  300.189424]  entry_SYSCALL_64_after_hwframe+0x72/0xdc
[  300.189654] RIP: 0033:0x7f3a2471f59d
[  300.189815] RSP: 002b:00007ffe567f7288 EFLAGS: 00000217 ORIG_RAX: 0000000000000001
[  300.190137] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f3a2471f59d
[  300.190397] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000005
[  300.190653] RBP: 00007ffe567f72a0 R08: 0000000000000010 R09: 0000000000000010
[  300.190910] R10: 0000000000000010 R11: 0000000000000217 R12: 00000000004012e0
[  300.191172] R13: 00007ffe567f73e0 R14: 0000000000000000 R15: 0000000000000000
[  300.191440]  </TASK>
...

To migrate a folio, we may wait the writeback of a folio to complete
when we already have held the lock of some folios.  But the writeback
code may wait to lock some folio we held lock.  This causes the
deadlock.  To fix the issue, we will avoid to wait the writeback to
complete if we have locked some folios.  After moving the locked
folios and unlocked, we will retry.

Signed-off-by: "Huang, Ying" <ying.huang@intel.com>
Reported-by: "Xu, Pengfei" <pengfei.xu@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Stefan Roesch <shr@devkernel.io>
Cc: Tejun Heo <tj@kernel.org>
Cc: Xin Hao <xhao@linux.alibaba.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
---
 mm/migrate.c | 12 ++++++++++++
 1 file changed, 12 insertions(+)

diff --git a/mm/migrate.c b/mm/migrate.c
index 28b435cdeac8..bc9a8050f1b0 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -1205,6 +1205,18 @@ static int migrate_folio_unmap(new_page_t get_new_page, free_page_t put_new_page
 		}
 		if (!force)
 			goto out;
+		/*
+		 * We have locked some folios and are going to wait the
+		 * writeback of this folio to complete.  But it's possible for
+		 * the writeback to wait to lock the folios we have locked.  To
+		 * avoid a potential deadlock, let's bail out and not do that.
+		 * The locked folios will be moved and unlocked, then we
+		 * can wait the writeback of this folio.
+		 */
+		if (avoid_force_lock) {
+			rc = -EDEADLOCK;
+			goto out;
+		}
 		folio_wait_writeback(src);
 	}
Hugh Dickins Feb. 21, 2023, 2:48 a.m. UTC | #3
On Mon, 20 Feb 2023, Huang, Ying wrote:

> Hi, Hugh,
> 
> Hugh Dickins <hughd@google.com> writes:
> 
> > 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) {
> 
> Thank you very much for pointing this out and the fix patch.  Today, my
> colleague Pengfei reported a deadlock bug to me.  It seems that we
> cannot wait the writeback to complete when we have locked some folios.
> Below patch can fix that deadlock.  I don't know whether this is related
> to the deadlock you run into.  It appears that we should avoid to
> lock/wait synchronously if we have locked more than one folios.

Thanks, I've checked now, on next-20230217 without my patch but
with your patch below: it took a few hours, but still deadlocks
as I described above, so it's not the same issue.

Yes, that's a good principle, that we should avoid to lock/wait
synchronously once we have locked one folio (hmm, above you say
"more than one": I think we mean the same thing, we're just
stating it differently, given how the code runs at present).

I'm not a great fan of migrate_folio_unmap()'s arguments,
"force" followed by "oh, but don't force" (but applaud the recent
"avoid_force_lock" as much better than the original "force_lock").
I haven't tried, but I wonder if you can avoid both those arguments,
and both of these patches, by passing down an adjusted mode (perhaps
MIGRATE_ASYNC, or perhaps a new mode) to all callees, once the first
folio of a batch has been acquired (then restore to the original mode
when starting a new batch).

(My patch is weak in that it trylocks for buffer_head even on the
first folio of a MIGRATE_SYNC norefs batch, although that has never
given a problem historically: adjusting the mode after acquiring
the first folio would correct that weakness.)

Hugh

> 
> Best Regards,
> Huang, Ying
> 
> ------------------------------------8<------------------------------------
> From 0699fa2f80a67e863107d49a25909c92b900a9be Mon Sep 17 00:00:00 2001
> From: Huang Ying <ying.huang@intel.com>
> Date: Mon, 20 Feb 2023 14:56:34 +0800
> Subject: [PATCH] migrate_pages: fix deadlock on waiting writeback
> 
> Pengfei reported a system soft lockup issue with Syzkaller.  The stack
> traces are as follows,
> 
> ...
> [  300.124933] INFO: task kworker/u4:3:73 blocked for more than 147 seconds.
> [  300.125214]       Not tainted 6.2.0-rc4-kvm+ #1314
> [  300.125408] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> [  300.125736] task:kworker/u4:3    state:D stack:0     pid:73    ppid:2      flags:0x00004000
> [  300.126059] Workqueue: writeback wb_workfn (flush-7:3)
> [  300.126282] Call Trace:
> [  300.126378]  <TASK>
> [  300.126464]  __schedule+0x43b/0xd00
> [  300.126601]  ? __blk_flush_plug+0x142/0x180
> [  300.126765]  schedule+0x6a/0xf0
> [  300.126912]  io_schedule+0x4a/0x80
> [  300.127051]  folio_wait_bit_common+0x1b5/0x4e0
> [  300.127227]  ? __pfx_wake_page_function+0x10/0x10
> [  300.127403]  __folio_lock+0x27/0x40
> [  300.127541]  write_cache_pages+0x350/0x870
> [  300.127699]  ? __pfx_iomap_do_writepage+0x10/0x10
> [  300.127889]  iomap_writepages+0x3f/0x80
> [  300.128037]  xfs_vm_writepages+0x94/0xd0
> [  300.128192]  ? __pfx_xfs_vm_writepages+0x10/0x10
> [  300.128370]  do_writepages+0x10a/0x240
> [  300.128514]  ? lock_is_held_type+0xe6/0x140
> [  300.128675]  __writeback_single_inode+0x9f/0xa90
> [  300.128854]  writeback_sb_inodes+0x2fb/0x8d0
> [  300.129030]  __writeback_inodes_wb+0x68/0x150
> [  300.129212]  wb_writeback+0x49c/0x770
> [  300.129357]  wb_workfn+0x6fb/0x9d0
> [  300.129500]  process_one_work+0x3cc/0x8d0
> [  300.129669]  worker_thread+0x66/0x630
> [  300.129824]  ? __pfx_worker_thread+0x10/0x10
> [  300.129989]  kthread+0x153/0x190
> [  300.130116]  ? __pfx_kthread+0x10/0x10
> [  300.130264]  ret_from_fork+0x29/0x50
> [  300.130409]  </TASK>
> [  300.179347] INFO: task repro:1023 blocked for more than 147 seconds.
> [  300.179905]       Not tainted 6.2.0-rc4-kvm+ #1314
> [  300.180317] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> [  300.180955] task:repro           state:D stack:0     pid:1023  ppid:360    flags:0x00004004
> [  300.181660] Call Trace:
> [  300.181879]  <TASK>
> [  300.182085]  __schedule+0x43b/0xd00
> [  300.182407]  schedule+0x6a/0xf0
> [  300.182694]  io_schedule+0x4a/0x80
> [  300.183020]  folio_wait_bit_common+0x1b5/0x4e0
> [  300.183506]  ? compaction_alloc+0x77/0x1150
> [  300.183892]  ? __pfx_wake_page_function+0x10/0x10
> [  300.184304]  folio_wait_bit+0x30/0x40
> [  300.184640]  folio_wait_writeback+0x2e/0x1e0
> [  300.185034]  migrate_pages_batch+0x555/0x1ac0
> [  300.185462]  ? __pfx_compaction_alloc+0x10/0x10
> [  300.185808]  ? __pfx_compaction_free+0x10/0x10
> [  300.186022]  ? __this_cpu_preempt_check+0x17/0x20
> [  300.186234]  ? lock_is_held_type+0xe6/0x140
> [  300.186423]  migrate_pages+0x100e/0x1180
> [  300.186603]  ? __pfx_compaction_free+0x10/0x10
> [  300.186800]  ? __pfx_compaction_alloc+0x10/0x10
> [  300.187011]  compact_zone+0xe10/0x1b50
> [  300.187182]  ? lock_is_held_type+0xe6/0x140
> [  300.187374]  ? check_preemption_disabled+0x80/0xf0
> [  300.187588]  compact_node+0xa3/0x100
> [  300.187755]  ? __sanitizer_cov_trace_const_cmp8+0x1c/0x30
> [  300.187993]  ? _find_first_bit+0x7b/0x90
> [  300.188171]  sysctl_compaction_handler+0x5d/0xb0
> [  300.188376]  proc_sys_call_handler+0x29d/0x420
> [  300.188583]  proc_sys_write+0x2b/0x40
> [  300.188749]  vfs_write+0x3a3/0x780
> [  300.188912]  ksys_write+0xb7/0x180
> [  300.189070]  __x64_sys_write+0x26/0x30
> [  300.189260]  do_syscall_64+0x3b/0x90
> [  300.189424]  entry_SYSCALL_64_after_hwframe+0x72/0xdc
> [  300.189654] RIP: 0033:0x7f3a2471f59d
> [  300.189815] RSP: 002b:00007ffe567f7288 EFLAGS: 00000217 ORIG_RAX: 0000000000000001
> [  300.190137] RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f3a2471f59d
> [  300.190397] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000005
> [  300.190653] RBP: 00007ffe567f72a0 R08: 0000000000000010 R09: 0000000000000010
> [  300.190910] R10: 0000000000000010 R11: 0000000000000217 R12: 00000000004012e0
> [  300.191172] R13: 00007ffe567f73e0 R14: 0000000000000000 R15: 0000000000000000
> [  300.191440]  </TASK>
> ...
> 
> To migrate a folio, we may wait the writeback of a folio to complete
> when we already have held the lock of some folios.  But the writeback
> code may wait to lock some folio we held lock.  This causes the
> deadlock.  To fix the issue, we will avoid to wait the writeback to
> complete if we have locked some folios.  After moving the locked
> folios and unlocked, we will retry.
> 
> Signed-off-by: "Huang, Ying" <ying.huang@intel.com>
> Reported-by: "Xu, Pengfei" <pengfei.xu@intel.com>
> Cc: Hugh Dickins <hughd@google.com>
> Cc: Christoph Hellwig <hch@lst.de>
> Cc: Stefan Roesch <shr@devkernel.io>
> Cc: Tejun Heo <tj@kernel.org>
> Cc: Xin Hao <xhao@linux.alibaba.com>
> Cc: Zi Yan <ziy@nvidia.com>
> Cc: Yang Shi <shy828301@gmail.com>
> Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
> Cc: Matthew Wilcox <willy@infradead.org>
> Cc: Mike Kravetz <mike.kravetz@oracle.com>
> ---
>  mm/migrate.c | 12 ++++++++++++
>  1 file changed, 12 insertions(+)
> 
> diff --git a/mm/migrate.c b/mm/migrate.c
> index 28b435cdeac8..bc9a8050f1b0 100644
> --- a/mm/migrate.c
> +++ b/mm/migrate.c
> @@ -1205,6 +1205,18 @@ static int migrate_folio_unmap(new_page_t get_new_page, free_page_t put_new_page
>  		}
>  		if (!force)
>  			goto out;
> +		/*
> +		 * We have locked some folios and are going to wait the
> +		 * writeback of this folio to complete.  But it's possible for
> +		 * the writeback to wait to lock the folios we have locked.  To
> +		 * avoid a potential deadlock, let's bail out and not do that.
> +		 * The locked folios will be moved and unlocked, then we
> +		 * can wait the writeback of this folio.
> +		 */
> +		if (avoid_force_lock) {
> +			rc = -EDEADLOCK;
> +			goto out;
> +		}
>  		folio_wait_writeback(src);
>  	}
>  
> -- 
> 2.39.1
> 
>
Huang, Ying Feb. 21, 2023, 3:34 a.m. UTC | #4
Hugh Dickins <hughd@google.com> writes:

> On Mon, 20 Feb 2023, Huang, Ying wrote:
>
>> Hi, Hugh,
>> 
>> Hugh Dickins <hughd@google.com> writes:
>> 
>> > 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) {
>> 
>> Thank you very much for pointing this out and the fix patch.  Today, my
>> colleague Pengfei reported a deadlock bug to me.  It seems that we
>> cannot wait the writeback to complete when we have locked some folios.
>> Below patch can fix that deadlock.  I don't know whether this is related
>> to the deadlock you run into.  It appears that we should avoid to
>> lock/wait synchronously if we have locked more than one folios.
>
> Thanks, I've checked now, on next-20230217 without my patch but
> with your patch below: it took a few hours, but still deadlocks
> as I described above, so it's not the same issue.
>
> Yes, that's a good principle, that we should avoid to lock/wait
> synchronously once we have locked one folio (hmm, above you say
> "more than one": I think we mean the same thing, we're just
> stating it differently, given how the code runs at present).
>
> I'm not a great fan of migrate_folio_unmap()'s arguments,
> "force" followed by "oh, but don't force" (but applaud the recent
> "avoid_force_lock" as much better than the original "force_lock").
> I haven't tried, but I wonder if you can avoid both those arguments,
> and both of these patches, by passing down an adjusted mode (perhaps
> MIGRATE_ASYNC, or perhaps a new mode) to all callees, once the first
> folio of a batch has been acquired (then restore to the original mode
> when starting a new batch).

Thanks for suggestion!  I think it's a good idea, but it will take me
some time to think how to implement.

> (My patch is weak in that it trylocks for buffer_head even on the
> first folio of a MIGRATE_SYNC norefs batch, although that has never
> given a problem historically: adjusting the mode after acquiring
> the first folio would correct that weakness.)

Further digging shows that this is related to loop device.  That can be
shown in the following trace,

[  300.109786] INFO: task kworker/u4:0:9 blocked for more than 147 seconds.
[  300.110616]       Not tainted 6.2.0-rc4-kvm+ #1314
[  300.111143] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[  300.111985] task:kworker/u4:0    state:D stack:0     pid:9     ppid:2      flags:0x00004000
[  300.112964] Workqueue: loop4 loop_rootcg_workfn
[  300.113658] Call Trace:
[  300.113996]  <TASK>
[  300.114315]  __schedule+0x43b/0xd00
[  300.114848]  schedule+0x6a/0xf0
[  300.115319]  io_schedule+0x4a/0x80
[  300.115860]  folio_wait_bit_common+0x1b5/0x4e0
[  300.116430]  ? __pfx_wake_page_function+0x10/0x10
[  300.116615]  __filemap_get_folio+0x73d/0x770
[  300.116790]  shmem_get_folio_gfp+0x1fd/0xc80
[  300.116963]  shmem_write_begin+0x91/0x220
[  300.117121]  generic_perform_write+0x10e/0x2e0
[  300.117312]  __generic_file_write_iter+0x17e/0x290
[  300.117498]  ? generic_write_checks+0x12b/0x1a0
[  300.117693]  generic_file_write_iter+0x97/0x180
[  300.117881]  ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
[  300.118087]  do_iter_readv_writev+0x13c/0x210
[  300.118256]  ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
[  300.118463]  do_iter_write+0xf6/0x330
[  300.118608]  vfs_iter_write+0x46/0x70
[  300.118754]  loop_process_work+0x723/0xfe0
[  300.118922]  loop_rootcg_workfn+0x28/0x40
[  300.119078]  process_one_work+0x3cc/0x8d0
[  300.119240]  worker_thread+0x66/0x630
[  300.119384]  ? __pfx_worker_thread+0x10/0x10
[  300.119551]  kthread+0x153/0x190
[  300.119681]  ? __pfx_kthread+0x10/0x10
[  300.119833]  ret_from_fork+0x29/0x50
[  300.119984]  </TASK>

When a folio of the loop device is written back, the underlying shmem
needs to be written back.  Which will acquire the lock of the folio of
shmem.  While that folio may be locked by migrate_pages_batch() already.

Your testing involves the loop device too.  So is it related to loop?
For example, after the buffer head was locked, is it possible to acquire
lock for folios of the underlying file system?

Best Regards,
Huang, Ying
Matthew Wilcox Feb. 21, 2023, 4:30 a.m. UTC | #5
On Mon, Feb 20, 2023 at 06:48:38PM -0800, Hugh Dickins wrote:
> Yes, that's a good principle, that we should avoid to lock/wait
> synchronously once we have locked one folio (hmm, above you say
> "more than one": I think we mean the same thing, we're just
> stating it differently, given how the code runs at present).

I suspect the migrate page code is disobeying the locking ordering
rules for multiple folios.  if two folios belong to the same file,
they must be locked by folio->index order, low to high.  If two folios
belong to different files, they must be ordered by folio->mapping, the
mapping lowest in memory first.  You can see this locking rule embedded
in vfs_lock_two_folios() in fs/remap_range.c.

I don't know what the locking rules are for two folios which are not file
folios, or for two folios when one is anonymous and the other belongs
to a file.  Maybe it's the same; you can lock them ordered by ->mapping
first, then by ->index.

Or you can just trylock multiple folios and skip the ones that don't work.
Huang, Ying Feb. 21, 2023, 4:38 a.m. UTC | #6
Matthew Wilcox <willy@infradead.org> writes:

> On Mon, Feb 20, 2023 at 06:48:38PM -0800, Hugh Dickins wrote:
>> Yes, that's a good principle, that we should avoid to lock/wait
>> synchronously once we have locked one folio (hmm, above you say
>> "more than one": I think we mean the same thing, we're just
>> stating it differently, given how the code runs at present).
>
> I suspect the migrate page code is disobeying the locking ordering
> rules for multiple folios.  if two folios belong to the same file,
> they must be locked by folio->index order, low to high.  If two folios
> belong to different files, they must be ordered by folio->mapping, the
> mapping lowest in memory first.  You can see this locking rule embedded
> in vfs_lock_two_folios() in fs/remap_range.c.
>
> I don't know what the locking rules are for two folios which are not file
> folios, or for two folios when one is anonymous and the other belongs
> to a file.  Maybe it's the same; you can lock them ordered by ->mapping
> first, then by ->index.
>
> Or you can just trylock multiple folios and skip the ones that don't work.

Yes.  We will only trylock when we have locked some folios (including
one).  And retry locking only after having processed and unlocked the
already locked folios.

Best Regards,
Huang, Ying
Huang, Ying Feb. 21, 2023, 2:04 p.m. UTC | #7
Hugh Dickins <hughd@google.com> writes:

> On Mon, 20 Feb 2023, Huang, Ying wrote:
>
>> Hi, Hugh,
>> 
>> Hugh Dickins <hughd@google.com> writes:
>> 
>> > 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) {
>> 
>> Thank you very much for pointing this out and the fix patch.  Today, my
>> colleague Pengfei reported a deadlock bug to me.  It seems that we
>> cannot wait the writeback to complete when we have locked some folios.
>> Below patch can fix that deadlock.  I don't know whether this is related
>> to the deadlock you run into.  It appears that we should avoid to
>> lock/wait synchronously if we have locked more than one folios.
>
> Thanks, I've checked now, on next-20230217 without my patch but
> with your patch below: it took a few hours, but still deadlocks
> as I described above, so it's not the same issue.
>
> Yes, that's a good principle, that we should avoid to lock/wait
> synchronously once we have locked one folio (hmm, above you say
> "more than one": I think we mean the same thing, we're just
> stating it differently, given how the code runs at present).
>
> I'm not a great fan of migrate_folio_unmap()'s arguments,
> "force" followed by "oh, but don't force" (but applaud the recent
> "avoid_force_lock" as much better than the original "force_lock").
> I haven't tried, but I wonder if you can avoid both those arguments,
> and both of these patches, by passing down an adjusted mode (perhaps
> MIGRATE_ASYNC, or perhaps a new mode) to all callees, once the first
> folio of a batch has been acquired (then restore to the original mode
> when starting a new batch).
>
> (My patch is weak in that it trylocks for buffer_head even on the
> first folio of a MIGRATE_SYNC norefs batch, although that has never
> given a problem historically: adjusting the mode after acquiring
> the first folio would correct that weakness.)

On second thought, I think that it may be better to provide a fix as
simple as possible firstly.  Then we can work on a more complex fix as
we discussed above.  The simple fix is easy to review now.  And, we will
have more time to test and review the complex fix.

In the following fix, I disabled the migration batching except for the
MIGRATE_ASYNC mode, or the split folios of a THP folio.  After that, I
will work on the complex fix to enable migration batching for all modes.

What do you think about that?

Best Regards,
Huang, Ying

-------------------------------8<---------------------------------
From 8e475812eacd9f2eeac76776c2b1a17af3e59b89 Mon Sep 17 00:00:00 2001
From: Huang Ying <ying.huang@intel.com>
Date: Tue, 21 Feb 2023 16:37:50 +0800
Subject: [PATCH] migrate_pages: fix deadlock in batched migration

Two deadlock bugs were reported for the migrate_pages() batching
series.  Thanks Hugh and Pengfei!  For example, in the following
deadlock trace snippet,

 INFO: task kworker/u4:0:9 blocked for more than 147 seconds.
       Not tainted 6.2.0-rc4-kvm+ #1314
 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
 task:kworker/u4:0    state:D stack:0     pid:9     ppid:2      flags:0x00004000
 Workqueue: loop4 loop_rootcg_workfn
 Call Trace:
  <TASK>
  __schedule+0x43b/0xd00
  schedule+0x6a/0xf0
  io_schedule+0x4a/0x80
  folio_wait_bit_common+0x1b5/0x4e0
  ? __pfx_wake_page_function+0x10/0x10
  __filemap_get_folio+0x73d/0x770
  shmem_get_folio_gfp+0x1fd/0xc80
  shmem_write_begin+0x91/0x220
  generic_perform_write+0x10e/0x2e0
  __generic_file_write_iter+0x17e/0x290
  ? generic_write_checks+0x12b/0x1a0
  generic_file_write_iter+0x97/0x180
  ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
  do_iter_readv_writev+0x13c/0x210
  ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
  do_iter_write+0xf6/0x330
  vfs_iter_write+0x46/0x70
  loop_process_work+0x723/0xfe0
  loop_rootcg_workfn+0x28/0x40
  process_one_work+0x3cc/0x8d0
  worker_thread+0x66/0x630
  ? __pfx_worker_thread+0x10/0x10
  kthread+0x153/0x190
  ? __pfx_kthread+0x10/0x10
  ret_from_fork+0x29/0x50
  </TASK>

 INFO: task repro:1023 blocked for more than 147 seconds.
       Not tainted 6.2.0-rc4-kvm+ #1314
 "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
 task:repro           state:D stack:0     pid:1023  ppid:360    flags:0x00004004
 Call Trace:
  <TASK>
  __schedule+0x43b/0xd00
  schedule+0x6a/0xf0
  io_schedule+0x4a/0x80
  folio_wait_bit_common+0x1b5/0x4e0
  ? compaction_alloc+0x77/0x1150
  ? __pfx_wake_page_function+0x10/0x10
  folio_wait_bit+0x30/0x40
  folio_wait_writeback+0x2e/0x1e0
  migrate_pages_batch+0x555/0x1ac0
  ? __pfx_compaction_alloc+0x10/0x10
  ? __pfx_compaction_free+0x10/0x10
  ? __this_cpu_preempt_check+0x17/0x20
  ? lock_is_held_type+0xe6/0x140
  migrate_pages+0x100e/0x1180
  ? __pfx_compaction_free+0x10/0x10
  ? __pfx_compaction_alloc+0x10/0x10
  compact_zone+0xe10/0x1b50
  ? lock_is_held_type+0xe6/0x140
  ? check_preemption_disabled+0x80/0xf0
  compact_node+0xa3/0x100
  ? __sanitizer_cov_trace_const_cmp8+0x1c/0x30
  ? _find_first_bit+0x7b/0x90
  sysctl_compaction_handler+0x5d/0xb0
  proc_sys_call_handler+0x29d/0x420
  proc_sys_write+0x2b/0x40
  vfs_write+0x3a3/0x780
  ksys_write+0xb7/0x180
  __x64_sys_write+0x26/0x30
  do_syscall_64+0x3b/0x90
  entry_SYSCALL_64_after_hwframe+0x72/0xdc
 RIP: 0033:0x7f3a2471f59d
 RSP: 002b:00007ffe567f7288 EFLAGS: 00000217 ORIG_RAX: 0000000000000001
 RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f3a2471f59d
 RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000005
 RBP: 00007ffe567f72a0 R08: 0000000000000010 R09: 0000000000000010
 R10: 0000000000000010 R11: 0000000000000217 R12: 00000000004012e0
 R13: 00007ffe567f73e0 R14: 0000000000000000 R15: 0000000000000000
  </TASK>

The page migration task has held the lock of the shmem folio A, and is
waiting the writeback of the folio B of the file system on the loop
block device to complete.  While the loop worker task which writes
back the folio B is waiting to lock the shmem folio A, because the
folio A backs the folio B in the loop device.  Thus deadlock is
triggered.

In general, if we have locked some other folios except the one we are
migrating, it's not safe to wait synchronously, for example, to wait
the writeback to complete or wait to lock the buffer head.

To fix the deadlock, in this patch, we avoid to batch the page
migration except for MIGRATE_ASYNC mode or the split folios of a THP
folio.  In MIGRATE_ASYNC mode, synchronous waiting is avoided.  And
there isn't any dependency relationship among the split folios of a
THP folio.

The fix can be improved via converting migration mode from synchronous
to asynchronous if we have locked some other folios except the one we
are migrating.  We will do that in the near future.

Link: https://lore.kernel.org/linux-mm/87a6c8c-c5c1-67dc-1e32-eb30831d6e3d@google.com/
Link: https://lore.kernel.org/linux-mm/874jrg7kke.fsf@yhuang6-desk2.ccr.corp.intel.com/
Signed-off-by: "Huang, Ying" <ying.huang@intel.com>
Reported-by: Hugh Dickins <hughd@google.com>
Reported-by: "Xu, Pengfei" <pengfei.xu@intel.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Stefan Roesch <shr@devkernel.io>
Cc: Tejun Heo <tj@kernel.org>
Cc: Xin Hao <xhao@linux.alibaba.com>
Cc: Zi Yan <ziy@nvidia.com>
Cc: Yang Shi <shy828301@gmail.com>
Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
Cc: Matthew Wilcox <willy@infradead.org>
Cc: Mike Kravetz <mike.kravetz@oracle.com>
---
 mm/migrate.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/mm/migrate.c b/mm/migrate.c
index ef68a1aff35c..bc04c34543f3 100644
--- a/mm/migrate.c
+++ b/mm/migrate.c
@@ -1937,7 +1937,7 @@ int migrate_pages(struct list_head *from, new_page_t get_new_page,
 		enum migrate_mode mode, int reason, unsigned int *ret_succeeded)
 {
 	int rc, rc_gather;
-	int nr_pages;
+	int nr_pages, batch;
 	struct folio *folio, *folio2;
 	LIST_HEAD(folios);
 	LIST_HEAD(ret_folios);
@@ -1951,6 +1951,11 @@ int migrate_pages(struct list_head *from, new_page_t get_new_page,
 				     mode, reason, &stats, &ret_folios);
 	if (rc_gather < 0)
 		goto out;
+
+	if (mode == MIGRATE_ASYNC)
+		batch = NR_MAX_BATCHED_MIGRATION;
+	else
+		batch = 1;
 again:
 	nr_pages = 0;
 	list_for_each_entry_safe(folio, folio2, from, lru) {
@@ -1961,11 +1966,11 @@ int migrate_pages(struct list_head *from, new_page_t get_new_page,
 		}
 
 		nr_pages += folio_nr_pages(folio);
-		if (nr_pages > NR_MAX_BATCHED_MIGRATION)
+		if (nr_pages >= batch)
 			break;
 	}
-	if (nr_pages > NR_MAX_BATCHED_MIGRATION)
-		list_cut_before(&folios, from, &folio->lru);
+	if (nr_pages >= batch)
+		list_cut_before(&folios, from, &folio2->lru);
 	else
 		list_splice_init(from, &folios);
 	rc = migrate_pages_batch(&folios, get_new_page, put_new_page, private,
Hugh Dickins Feb. 21, 2023, 10:15 p.m. UTC | #8
On Tue, 21 Feb 2023, Huang, Ying wrote:
> Hugh Dickins <hughd@google.com> writes:
> > On Mon, 20 Feb 2023, Huang, Ying wrote:
> >
> >> Hi, Hugh,
> >> 
> >> Hugh Dickins <hughd@google.com> writes:
> >> 
> >> > 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) {
> >> 
> >> Thank you very much for pointing this out and the fix patch.  Today, my
> >> colleague Pengfei reported a deadlock bug to me.  It seems that we
> >> cannot wait the writeback to complete when we have locked some folios.
> >> Below patch can fix that deadlock.  I don't know whether this is related
> >> to the deadlock you run into.  It appears that we should avoid to
> >> lock/wait synchronously if we have locked more than one folios.
> >
> > Thanks, I've checked now, on next-20230217 without my patch but
> > with your patch below: it took a few hours, but still deadlocks
> > as I described above, so it's not the same issue.
> >
> > Yes, that's a good principle, that we should avoid to lock/wait
> > synchronously once we have locked one folio (hmm, above you say
> > "more than one": I think we mean the same thing, we're just
> > stating it differently, given how the code runs at present).
> >
> > I'm not a great fan of migrate_folio_unmap()'s arguments,
> > "force" followed by "oh, but don't force" (but applaud the recent
> > "avoid_force_lock" as much better than the original "force_lock").
> > I haven't tried, but I wonder if you can avoid both those arguments,
> > and both of these patches, by passing down an adjusted mode (perhaps
> > MIGRATE_ASYNC, or perhaps a new mode) to all callees, once the first
> > folio of a batch has been acquired (then restore to the original mode
> > when starting a new batch).
> 
> Thanks for suggestion!  I think it's a good idea, but it will take me
> some time to think how to implement.
> 
> > (My patch is weak in that it trylocks for buffer_head even on the
> > first folio of a MIGRATE_SYNC norefs batch, although that has never
> > given a problem historically: adjusting the mode after acquiring
> > the first folio would correct that weakness.)
> 
> Further digging shows that this is related to loop device.  That can be
> shown in the following trace,
> 
> [  300.109786] INFO: task kworker/u4:0:9 blocked for more than 147 seconds.
> [  300.110616]       Not tainted 6.2.0-rc4-kvm+ #1314
> [  300.111143] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
> [  300.111985] task:kworker/u4:0    state:D stack:0     pid:9     ppid:2      flags:0x00004000
> [  300.112964] Workqueue: loop4 loop_rootcg_workfn
> [  300.113658] Call Trace:
> [  300.113996]  <TASK>
> [  300.114315]  __schedule+0x43b/0xd00
> [  300.114848]  schedule+0x6a/0xf0
> [  300.115319]  io_schedule+0x4a/0x80
> [  300.115860]  folio_wait_bit_common+0x1b5/0x4e0
> [  300.116430]  ? __pfx_wake_page_function+0x10/0x10
> [  300.116615]  __filemap_get_folio+0x73d/0x770
> [  300.116790]  shmem_get_folio_gfp+0x1fd/0xc80
> [  300.116963]  shmem_write_begin+0x91/0x220
> [  300.117121]  generic_perform_write+0x10e/0x2e0
> [  300.117312]  __generic_file_write_iter+0x17e/0x290
> [  300.117498]  ? generic_write_checks+0x12b/0x1a0
> [  300.117693]  generic_file_write_iter+0x97/0x180
> [  300.117881]  ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
> [  300.118087]  do_iter_readv_writev+0x13c/0x210
> [  300.118256]  ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
> [  300.118463]  do_iter_write+0xf6/0x330
> [  300.118608]  vfs_iter_write+0x46/0x70
> [  300.118754]  loop_process_work+0x723/0xfe0
> [  300.118922]  loop_rootcg_workfn+0x28/0x40
> [  300.119078]  process_one_work+0x3cc/0x8d0
> [  300.119240]  worker_thread+0x66/0x630
> [  300.119384]  ? __pfx_worker_thread+0x10/0x10
> [  300.119551]  kthread+0x153/0x190
> [  300.119681]  ? __pfx_kthread+0x10/0x10
> [  300.119833]  ret_from_fork+0x29/0x50
> [  300.119984]  </TASK>
> 
> When a folio of the loop device is written back, the underlying shmem
> needs to be written back.  Which will acquire the lock of the folio of
> shmem.  While that folio may be locked by migrate_pages_batch() already.
> 
> Your testing involves the loop device too.  So is it related to loop?
> For example, after the buffer head was locked, is it possible to acquire
> lock for folios of the underlying file system?

To lock (other than trylock) a folio while holding a buffer head lock
would violate lock ordering: they are both bit spin locks, so lockdep
would not notice, but I'm sure any such violation would have been
caught long ago (unless on some very unlikely path).

There was nothing in the stack traces I looked at to implicate loop,
though admittedly I paid much more attention to the kcompactd0 and
khugepaged pattern which was emerging - the few times I looked at the
loop0 thread, IIRC it was either idle or waiting on the lower level.

If it had a stack trace like you show above, I would just have ignored
that as of little interest, as probably waiting on one of the migrate
batch's locked folios, which were themselves waiting on a buffer_head
lock somewhere.  (And shmem itself doesn't use buffer_heads.)

But it's still an interesting idea that these deadlocks might be
related to loop: loop by its nature certainly has a propensity for
deadlocks, though I think those have all been ironed out years ago.

I've run my load overnight using a small ext4 partition instead of
ext4 on loop on tmpfs, and that has not deadlocked.  I hesitate to
say that confirms your idea that loop is somehow causing the deadlock
- the timings will have changed, so I've no idea how long such a
modified load needs to run to give confidence; but it does suggest
that you're right.

Though I still think your principle, that we should avoid to lock/wait
synchronously once the batch holds one folio, is the safe principle to
follow anyway, whether or not the deadlocks can be pinned on loop.

Hugh
Hugh Dickins Feb. 21, 2023, 10:25 p.m. UTC | #9
On Tue, 21 Feb 2023, Huang, Ying wrote:
> 
> On second thought, I think that it may be better to provide a fix as
> simple as possible firstly.  Then we can work on a more complex fix as
> we discussed above.  The simple fix is easy to review now.  And, we will
> have more time to test and review the complex fix.
> 
> In the following fix, I disabled the migration batching except for the
> MIGRATE_ASYNC mode, or the split folios of a THP folio.  After that, I
> will work on the complex fix to enable migration batching for all modes.
> 
> What do you think about that?

I don't think there's a need to rush in the wrong fix so quickly.
Your series was in (though sometimes out of) linux-next for some
while, without causing any widespread problems.  Andrew did send
it to Linus yesterday, I expect he'll be pushing it out later today
or tomorrow, but I don't think it's going to cause big problems.
Aiming for a fix in -rc2 would be good.  Why would it be complex?

Hugh

> 
> Best Regards,
> Huang, Ying
> 
> -------------------------------8<---------------------------------
> From 8e475812eacd9f2eeac76776c2b1a17af3e59b89 Mon Sep 17 00:00:00 2001
> From: Huang Ying <ying.huang@intel.com>
> Date: Tue, 21 Feb 2023 16:37:50 +0800
> Subject: [PATCH] migrate_pages: fix deadlock in batched migration
> 
> Two deadlock bugs were reported for the migrate_pages() batching
> series.  Thanks Hugh and Pengfei!  For example, in the following
> deadlock trace snippet,
> 
>  INFO: task kworker/u4:0:9 blocked for more than 147 seconds.
>        Not tainted 6.2.0-rc4-kvm+ #1314
>  "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>  task:kworker/u4:0    state:D stack:0     pid:9     ppid:2      flags:0x00004000
>  Workqueue: loop4 loop_rootcg_workfn
>  Call Trace:
>   <TASK>
>   __schedule+0x43b/0xd00
>   schedule+0x6a/0xf0
>   io_schedule+0x4a/0x80
>   folio_wait_bit_common+0x1b5/0x4e0
>   ? __pfx_wake_page_function+0x10/0x10
>   __filemap_get_folio+0x73d/0x770
>   shmem_get_folio_gfp+0x1fd/0xc80
>   shmem_write_begin+0x91/0x220
>   generic_perform_write+0x10e/0x2e0
>   __generic_file_write_iter+0x17e/0x290
>   ? generic_write_checks+0x12b/0x1a0
>   generic_file_write_iter+0x97/0x180
>   ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
>   do_iter_readv_writev+0x13c/0x210
>   ? __sanitizer_cov_trace_const_cmp4+0x1a/0x20
>   do_iter_write+0xf6/0x330
>   vfs_iter_write+0x46/0x70
>   loop_process_work+0x723/0xfe0
>   loop_rootcg_workfn+0x28/0x40
>   process_one_work+0x3cc/0x8d0
>   worker_thread+0x66/0x630
>   ? __pfx_worker_thread+0x10/0x10
>   kthread+0x153/0x190
>   ? __pfx_kthread+0x10/0x10
>   ret_from_fork+0x29/0x50
>   </TASK>
> 
>  INFO: task repro:1023 blocked for more than 147 seconds.
>        Not tainted 6.2.0-rc4-kvm+ #1314
>  "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
>  task:repro           state:D stack:0     pid:1023  ppid:360    flags:0x00004004
>  Call Trace:
>   <TASK>
>   __schedule+0x43b/0xd00
>   schedule+0x6a/0xf0
>   io_schedule+0x4a/0x80
>   folio_wait_bit_common+0x1b5/0x4e0
>   ? compaction_alloc+0x77/0x1150
>   ? __pfx_wake_page_function+0x10/0x10
>   folio_wait_bit+0x30/0x40
>   folio_wait_writeback+0x2e/0x1e0
>   migrate_pages_batch+0x555/0x1ac0
>   ? __pfx_compaction_alloc+0x10/0x10
>   ? __pfx_compaction_free+0x10/0x10
>   ? __this_cpu_preempt_check+0x17/0x20
>   ? lock_is_held_type+0xe6/0x140
>   migrate_pages+0x100e/0x1180
>   ? __pfx_compaction_free+0x10/0x10
>   ? __pfx_compaction_alloc+0x10/0x10
>   compact_zone+0xe10/0x1b50
>   ? lock_is_held_type+0xe6/0x140
>   ? check_preemption_disabled+0x80/0xf0
>   compact_node+0xa3/0x100
>   ? __sanitizer_cov_trace_const_cmp8+0x1c/0x30
>   ? _find_first_bit+0x7b/0x90
>   sysctl_compaction_handler+0x5d/0xb0
>   proc_sys_call_handler+0x29d/0x420
>   proc_sys_write+0x2b/0x40
>   vfs_write+0x3a3/0x780
>   ksys_write+0xb7/0x180
>   __x64_sys_write+0x26/0x30
>   do_syscall_64+0x3b/0x90
>   entry_SYSCALL_64_after_hwframe+0x72/0xdc
>  RIP: 0033:0x7f3a2471f59d
>  RSP: 002b:00007ffe567f7288 EFLAGS: 00000217 ORIG_RAX: 0000000000000001
>  RAX: ffffffffffffffda RBX: 0000000000000000 RCX: 00007f3a2471f59d
>  RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000005
>  RBP: 00007ffe567f72a0 R08: 0000000000000010 R09: 0000000000000010
>  R10: 0000000000000010 R11: 0000000000000217 R12: 00000000004012e0
>  R13: 00007ffe567f73e0 R14: 0000000000000000 R15: 0000000000000000
>   </TASK>
> 
> The page migration task has held the lock of the shmem folio A, and is
> waiting the writeback of the folio B of the file system on the loop
> block device to complete.  While the loop worker task which writes
> back the folio B is waiting to lock the shmem folio A, because the
> folio A backs the folio B in the loop device.  Thus deadlock is
> triggered.
> 
> In general, if we have locked some other folios except the one we are
> migrating, it's not safe to wait synchronously, for example, to wait
> the writeback to complete or wait to lock the buffer head.
> 
> To fix the deadlock, in this patch, we avoid to batch the page
> migration except for MIGRATE_ASYNC mode or the split folios of a THP
> folio.  In MIGRATE_ASYNC mode, synchronous waiting is avoided.  And
> there isn't any dependency relationship among the split folios of a
> THP folio.
> 
> The fix can be improved via converting migration mode from synchronous
> to asynchronous if we have locked some other folios except the one we
> are migrating.  We will do that in the near future.
> 
> Link: https://lore.kernel.org/linux-mm/87a6c8c-c5c1-67dc-1e32-eb30831d6e3d@google.com/
> Link: https://lore.kernel.org/linux-mm/874jrg7kke.fsf@yhuang6-desk2.ccr.corp.intel.com/
> Signed-off-by: "Huang, Ying" <ying.huang@intel.com>
> Reported-by: Hugh Dickins <hughd@google.com>
> Reported-by: "Xu, Pengfei" <pengfei.xu@intel.com>
> Cc: Christoph Hellwig <hch@lst.de>
> Cc: Stefan Roesch <shr@devkernel.io>
> Cc: Tejun Heo <tj@kernel.org>
> Cc: Xin Hao <xhao@linux.alibaba.com>
> Cc: Zi Yan <ziy@nvidia.com>
> Cc: Yang Shi <shy828301@gmail.com>
> Cc: Baolin Wang <baolin.wang@linux.alibaba.com>
> Cc: Matthew Wilcox <willy@infradead.org>
> Cc: Mike Kravetz <mike.kravetz@oracle.com>
> ---
>  mm/migrate.c | 13 +++++++++----
>  1 file changed, 9 insertions(+), 4 deletions(-)
> 
> diff --git a/mm/migrate.c b/mm/migrate.c
> index ef68a1aff35c..bc04c34543f3 100644
> --- a/mm/migrate.c
> +++ b/mm/migrate.c
> @@ -1937,7 +1937,7 @@ int migrate_pages(struct list_head *from, new_page_t get_new_page,
>  		enum migrate_mode mode, int reason, unsigned int *ret_succeeded)
>  {
>  	int rc, rc_gather;
> -	int nr_pages;
> +	int nr_pages, batch;
>  	struct folio *folio, *folio2;
>  	LIST_HEAD(folios);
>  	LIST_HEAD(ret_folios);
> @@ -1951,6 +1951,11 @@ int migrate_pages(struct list_head *from, new_page_t get_new_page,
>  				     mode, reason, &stats, &ret_folios);
>  	if (rc_gather < 0)
>  		goto out;
> +
> +	if (mode == MIGRATE_ASYNC)
> +		batch = NR_MAX_BATCHED_MIGRATION;
> +	else
> +		batch = 1;
>  again:
>  	nr_pages = 0;
>  	list_for_each_entry_safe(folio, folio2, from, lru) {
> @@ -1961,11 +1966,11 @@ int migrate_pages(struct list_head *from, new_page_t get_new_page,
>  		}
>  
>  		nr_pages += folio_nr_pages(folio);
> -		if (nr_pages > NR_MAX_BATCHED_MIGRATION)
> +		if (nr_pages >= batch)
>  			break;
>  	}
> -	if (nr_pages > NR_MAX_BATCHED_MIGRATION)
> -		list_cut_before(&folios, from, &folio->lru);
> +	if (nr_pages >= batch)
> +		list_cut_before(&folios, from, &folio2->lru);
>  	else
>  		list_splice_init(from, &folios);
>  	rc = migrate_pages_batch(&folios, get_new_page, put_new_page, private,
> -- 
> 2.39.1
Huang, Ying Feb. 22, 2023, 1:02 a.m. UTC | #10
Hugh Dickins <hughd@google.com> writes:

> On Tue, 21 Feb 2023, Huang, Ying wrote:
>> 
>> On second thought, I think that it may be better to provide a fix as
>> simple as possible firstly.  Then we can work on a more complex fix as
>> we discussed above.  The simple fix is easy to review now.  And, we will
>> have more time to test and review the complex fix.
>> 
>> In the following fix, I disabled the migration batching except for the
>> MIGRATE_ASYNC mode, or the split folios of a THP folio.  After that, I
>> will work on the complex fix to enable migration batching for all modes.
>> 
>> What do you think about that?
>
> I don't think there's a need to rush in the wrong fix so quickly.
> Your series was in (though sometimes out of) linux-next for some
> while, without causing any widespread problems.  Andrew did send
> it to Linus yesterday, I expect he'll be pushing it out later today
> or tomorrow, but I don't think it's going to cause big problems.
> Aiming for a fix in -rc2 would be good.

Sure, I will target to fix in -rc2.  Thanks for suggestion!

> Why would it be complex?

Now, I think the big picture could be,

if (MIGRATE_ASYNC) {
        migrate_pages_batch(from,);
} else {
        migrate_pages_batch(from,, MIGRATE_ASYNC,);
        list_for_each_entry_safe (folio,, from) {
                migrate_pages_batch(one_folio, , MIGRATE_SYNC,);
        }
}

That is, for synchronous migration, try asynchronous batched migration
firstly, then fall back to synchronous migration one by one.  This will
make the retry logic easier to be understood.

This needs some code change.  Anyway, I will try to do that and show the
code.

Best Regards,
Huang, Ying
Jan Kara Feb. 27, 2023, 11:06 a.m. UTC | #11
On Fri 17-02-23 13:47:48, Hugh Dickins wrote:
> 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.

I suspect it can indeed be caused by the presence of the loop device as
Huang Ying has suggested. What filesystems using buffer_heads do is a
pattern like:

bh = page_buffers(loop device page cache page);
lock_buffer(bh);
submit_bh(bh);
- now on loop dev this ends up doing:
  lo_write_bvec()
    vfs_iter_write()
      ...
      folio_lock(backing file folio);

So if migration code holds "backing file folio" lock and at the same time
waits for 'bh' lock (while trying to migrate loop device page cache page), it
is a deadlock.

Proposed solution of never waiting for locks in batched mode looks like a
sensible one to me...

								Honza
Huang, Ying Feb. 28, 2023, 1:13 a.m. UTC | #12
Hi, Honza,

Jan Kara <jack@suse.cz> writes:

> On Fri 17-02-23 13:47:48, Hugh Dickins wrote:
>> 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.
>
> I suspect it can indeed be caused by the presence of the loop device as
> Huang Ying has suggested. What filesystems using buffer_heads do is a
> pattern like:
>
> bh = page_buffers(loop device page cache page);
> lock_buffer(bh);
> submit_bh(bh);
> - now on loop dev this ends up doing:
>   lo_write_bvec()
>     vfs_iter_write()
>       ...
>       folio_lock(backing file folio);
>
> So if migration code holds "backing file folio" lock and at the same time
> waits for 'bh' lock (while trying to migrate loop device page cache page), it
> is a deadlock.
>
> Proposed solution of never waiting for locks in batched mode looks like a
> sensible one to me...

Thank you very much for detail explanation!

Best Regards,
Huang, Ying
Hugh Dickins Feb. 28, 2023, 5:59 a.m. UTC | #13
On Tue, 28 Feb 2023, Huang, Ying wrote:
> Jan Kara <jack@suse.cz> writes:
> > On Fri 17-02-23 13:47:48, Hugh Dickins wrote:
> >> 
> >> 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.
> >
> > I suspect it can indeed be caused by the presence of the loop device as
> > Huang Ying has suggested. What filesystems using buffer_heads do is a
> > pattern like:
> >
> > bh = page_buffers(loop device page cache page);
> > lock_buffer(bh);
> > submit_bh(bh);
> > - now on loop dev this ends up doing:
> >   lo_write_bvec()
> >     vfs_iter_write()
> >       ...
> >       folio_lock(backing file folio);
> >
> > So if migration code holds "backing file folio" lock and at the same time
> > waits for 'bh' lock (while trying to migrate loop device page cache page), it
> > is a deadlock.
> >
> > Proposed solution of never waiting for locks in batched mode looks like a
> > sensible one to me...
> 
> Thank you very much for detail explanation!

Yes, thanks a lot, Jan, for elucidating the deadlocks.

I was running with the 1/3,2/3,3/3 series for 48 hours on two machines
at the weekend, and hit no problems with all of them on.

I did try to review them this evening, but there's too much for me to
take in there to give an Acked-by: but I'll ask a couple of questions.

Hugh