Commit Graph

760 Commits

Author SHA1 Message Date
Tomasz Majchrzak 1532d9e87e raid5-ppl: PPL support for disks with write-back cache enabled
In order to provide data consistency with PPL for disks with write-back
cache enabled all data has to be flushed to disks before next PPL
entry. The disks to be flushed are marked in the bitmap. It's modified
under a mutex and it's only read after PPL io unit is submitted.

A limitation of 64 disks in the array has been introduced to keep data
structures and implementation simple. RAID5 arrays with so many disks are
not likely due to high risk of multiple disks failure. Such restriction
should not be a real life limitation.

With write-back cache disabled next PPL entry is submitted when data write
for current one completes. Data flush defers next log submission so trigger
it when there are no stripes for handling found.

As PPL assures all data is flushed to disk at request completion, just
acknowledge flush request when PPL is enabled.

Signed-off-by: Tomasz Majchrzak <tomasz.majchrzak@intel.com>
Signed-off-by: Shaohua Li <sh.li@alibaba-inc.com>
2018-01-15 14:29:42 -08:00
Song Liu d5d885fd51 md: introduce new personality funciton start()
In do_md_run(), md threads should not wake up until the array is fully
initialized in md_run(). However, in raid5_run(), raid5-cache may wake
up mddev->thread to flush stripes that need to be written back. This
design doesn't break badly right now. But it could lead to bad bug in
the future.

This patch tries to resolve this problem by splitting start up work
into two personality functions, run() and start(). Tasks that do not
require the md threads should go into run(), while task that require
the md threads go into start().

r5l_load_log() is moved to raid5_start(), so it is not called until
the md threads are started in do_md_run().

Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-12-11 08:52:34 -08:00
bingjingc aff69d89bd md/raid5: correct degraded calculation in raid5_error
When disk failure occurs on new disks for reshape, mddev->degraded
is not calculated correctly. Faulty bit of the failure device is not
set before raid5_calc_degraded(conf).

mdadm --create /dev/md0 --level=5 --raid-devices=3 /dev/loop[012]
mdadm /dev/md0 -a /dev/loop3
mdadm /dev/md0 --grow -n4
mdadm /dev/md0 -f /dev/loop3 # simulating disk failure

cat /sys/block/md0/md/degraded # it outputs 0, but it should be 1.

However, mdadm -D /dev/md0 will show that it is degraded. It's a bug.
It can be fixed by moving the resources raid5_calc_degraded() depends
on before it.

Reported-by: Roy Chung <roychung@synology.com>
Reviewed-by: Alex Wu <alexwu@synology.com>
Signed-off-by: BingJing Chang <bingjingc@synology.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-12-01 11:27:32 -08:00
Linus Torvalds 47f521ba18 Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md
Pull MD update from Shaohua Li:
 "This update mostly includes bug fixes:

   - md-cluster now supports raid10 from Guoqing

   - raid5 PPL fixes from Artur

   - badblock regression fix from Bo

   - suspend hang related fixes from Neil

   - raid5 reshape fixes from Neil

   - raid1 freeze deadlock fix from Nate

   - memleak fixes from Zdenek

   - bitmap related fixes from Me and Tao

   - other fixes and cleanups"

* 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md: (33 commits)
  md: free unused memory after bitmap resize
  md: release allocated bitset sync_set
  md/bitmap: clear BITMAP_WRITE_ERROR bit before writing it to sb
  md: be cautious about using ->curr_resync_completed for ->recovery_offset
  badblocks: fix wrong return value in badblocks_set if badblocks are disabled
  md: don't check MD_SB_CHANGE_CLEAN in md_allow_write
  md-cluster: update document for raid10
  md: remove redundant variable q
  raid1: remove obsolete code in raid1_write_request
  md-cluster: Use a small window for raid10 resync
  md-cluster: Suspend writes in RAID10 if within range
  md-cluster/raid10: set "do_balance = 0" if area is resyncing
  md: use lockdep_assert_held
  raid1: prevent freeze_array/wait_all_barriers deadlock
  md: use TASK_IDLE instead of blocking signals
  md: remove special meaning of ->quiesce(.., 2)
  md: allow metadata update while suspending.
  md: use mddev_suspend/resume instead of ->quiesce()
  md: move suspend_hi/lo handling into core md code
  md: don't call bitmap_create() while array is quiesced.
  ...
2017-11-14 16:07:26 -08:00
NeilBrown db0505d320 md: be cautious about using ->curr_resync_completed for ->recovery_offset
The ->recovery_offset shows how much of a non-InSync device is actually
in sync - how much has been recoveryed.

When performing a recovery, ->curr_resync and ->curr_resync_completed
follow the device address being recovered and so can be used to update
->recovery_offset.

When performing a reshape, ->curr_resync* might follow the device
addresses (raid5) or might follow array addresses (raid10), so cannot
in general be used to set ->recovery_offset.  When reshaping backwards,
->curre_resync* measures from the *end* of the array-or-device, so is
particularly unhelpful.

So change the common code in md.c to only use ->curr_resync_complete
for the simple recovery case, and add code to raid5.c to update
->recovery_offset during a forwards reshape.

Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-11-09 07:29:40 -08:00
Mikulas Patocka ae89fd3de4 md: use TASK_IDLE instead of blocking signals
Hi - I submit this patch for the next merge window:

Some times ago, I made a patch f9c79bc05a that blocks signals around the
schedule() calls in MD. The MD subsystem needs to do an uninterruptible
sleep that is not accounted in load average - so we block signals and use
interruptible sleep.

The kernel has a special TASK_IDLE state for this purpose, so we can use
it instead of blocking signals. This patch doesn't fix any bug, it just
makes the code simpler.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Acked-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-11-01 21:32:21 -07:00
NeilBrown b03e0ccb5a md: remove special meaning of ->quiesce(.., 2)
The '2' argument means "wake up anything that is waiting".
This is an inelegant part of the design and was added
to help support management of suspend_lo/suspend_hi setting.
Now that suspend_lo/hi is managed in mddev_suspend/resume,
that need is gone.
These is still a couple of places where we call 'quiesce'
with an argument of '2', but they can safely be changed to
call ->quiesce(.., 1); ->quiesce(.., 0) which
achieve the same result at the small cost of pausing IO
briefly.

This removes a small "optimization" from suspend_{hi,lo}_store,
but it isn't clear that optimization served a useful purpose.
The code now is a lot clearer.

Suggested-by: Shaohua Li <shli@kernel.org>
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-11-01 21:32:20 -07:00
NeilBrown b3143b9a38 md: move suspend_hi/lo handling into core md code
responding to ->suspend_lo and ->suspend_hi is similar
to responding to ->suspended.  It is best to wait in
the common core code without incrementing ->active_io.
This allows mddev_suspend()/mddev_resume() to work while
requests are waiting for suspend_lo/hi to change.
This is will be important after a subsequent patch
which uses mddev_suspend() to synchronize updating for
suspend_lo/hi.

So move the code for testing suspend_lo/hi out of raid1.c
and raid5.c, and place it in md.c

Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-11-01 21:32:19 -07:00
NeilBrown 230b55fa8d md: forbid a RAID5 from having both a bitmap and a journal.
Having both a bitmap and a journal is pointless.
Attempting to do so can corrupt the bitmap if the journal
replay happens before the bitmap is initialized.
Rather than try to avoid this corruption, simply
refuse to allow arrays with both a bitmap and a journal.
So:
 - if raid5_run sees both are present, fail.
 - if adding a bitmap finds a journal is present, fail
 - if adding a journal finds a bitmap is present, fail.

Cc: stable@vger.kernel.org (4.10+)
Signed-off-by: NeilBrown <neilb@suse.com>
Tested-by: Joshua Kinard <kumba@gentoo.org>
Acked-by: Joshua Kinard <kumba@gentoo.org>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-11-01 21:32:18 -07:00
Mark Rutland 6aa7de0591 locking/atomics: COCCINELLE/treewide: Convert trivial ACCESS_ONCE() patterns to READ_ONCE()/WRITE_ONCE()
Please do not apply this to mainline directly, instead please re-run the
coccinelle script shown below and apply its output.

For several reasons, it is desirable to use {READ,WRITE}_ONCE() in
preference to ACCESS_ONCE(), and new code is expected to use one of the
former. So far, there's been no reason to change most existing uses of
ACCESS_ONCE(), as these aren't harmful, and changing them results in
churn.

However, for some features, the read/write distinction is critical to
correct operation. To distinguish these cases, separate read/write
accessors must be used. This patch migrates (most) remaining
ACCESS_ONCE() instances to {READ,WRITE}_ONCE(), using the following
coccinelle script:

----
// Convert trivial ACCESS_ONCE() uses to equivalent READ_ONCE() and
// WRITE_ONCE()

// $ make coccicheck COCCI=/home/mark/once.cocci SPFLAGS="--include-headers" MODE=patch

virtual patch

@ depends on patch @
expression E1, E2;
@@

- ACCESS_ONCE(E1) = E2
+ WRITE_ONCE(E1, E2)

@ depends on patch @
expression E;
@@

- ACCESS_ONCE(E)
+ READ_ONCE(E)
----

Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Signed-off-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: davem@davemloft.net
Cc: linux-arch@vger.kernel.org
Cc: mpe@ellerman.id.au
Cc: shuah@kernel.org
Cc: snitzer@redhat.com
Cc: thor.thayer@linux.intel.com
Cc: tj@kernel.org
Cc: viro@zeniv.linux.org.uk
Cc: will.deacon@arm.com
Link: http://lkml.kernel.org/r/1508792849-3115-19-git-send-email-paulmck@linux.vnet.ibm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-10-25 11:01:08 +02:00
NeilBrown 235b6003fb raid5: Set R5_Expanded on parity devices as well as data.
When reshaping a fully degraded raid5/raid6 to a larger
nubmer of devices, the new device(s) are not in-sync
and so that can make the newly grown stripe appear to be
"failed".
To avoid this, we set the R5_Expanded flag to say "Even though
this device is not fully in-sync, this block is safe so
don't treat the device as failed for this stripe".
This flag is set for data devices, not not for parity devices.

Consequently, if you have a RAID6 with two devices that are partly
recovered and a spare, and start a reshape to include the spare,
then when the reshape gets past the point where the recovery was
up to, it will think the stripes are failed and will get into
an infinite loop, failing to make progress.

So when contructing parity on an EXPAND_READY stripe,
set R5_Expanded.

Reported-by: Curt <lightspd@gmail.com>
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-10-18 20:04:06 -07:00
Mike Snitzer 935fe0983e md: rename some drivers/md/ files to have an "md-" prefix
Motivated by the desire to illiminate the imprecise nature of
DM-specific patches being unnecessarily sent to both the MD maintainer
and mailing-list.  Which is born out of the fact that DM files also
reside in drivers/md/

Now all MD-specific files in drivers/md/ start with either "raid" or
"md-" and the MAINTAINERS file has been updated accordingly.

Shaohua: don't change module name

Signed-off-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-10-16 19:06:36 -07:00
Shaohua Li 7d5d7b5058 md/raid5: cap worker count
static checker reports a potential integer overflow. Cap the worker count to
avoid the overflow.

Reported:-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-09-27 20:08:44 -07:00
Linus Torvalds 12fcf66e74 Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md
Pull MD fixes from Shaohua Li:
 "Two small patches to fix long-lived raid5 stripe batch bugs, one from
  Dennis and the other from me"

* 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md:
  md/raid5: preserve STRIPE_ON_UNPLUG_LIST in break_stripe_batch_list
  md/raid5: fix a race condition in stripe batch
2017-09-19 08:39:32 -07:00
Linus Torvalds 3645e6d0dc Merge tag 'md/4.14-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md
Pull MD updates from Shaohua Li:
 "This update mainly fixes bugs:

   - Make raid5 ppl support several ppl from Pawel

   - Several raid5-cache bug fixes from Song

   - Bitmap fixes from Neil and Me

   - One raid1/10 regression fix since 4.12 from Me

   - Other small fixes and cleanup"

* tag 'md/4.14-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md:
  md/bitmap: disable bitmap_resize for file-backed bitmaps.
  raid5-ppl: Recovery support for multiple partial parity logs
  md: Runtime support for multiple ppls
  md/raid0: attach correct cgroup info in bio
  lib/raid6: align AVX512 constants to 512 bits, not bytes
  raid5: remove raid5_build_block
  md/r5cache: call mddev_lock/unlock() in r5c_journal_mode_show
  md: replace seq_release_private with seq_release
  md: notify about new spare disk in the container
  md/raid1/10: reset bio allocated from mempool
  md/raid5: release/flush io in raid5_do_work()
  md/bitmap: copy correct data for bitmap super
2017-09-07 12:41:48 -07:00
Dennis Yang 184a09eb9a md/raid5: preserve STRIPE_ON_UNPLUG_LIST in break_stripe_batch_list
In release_stripe_plug(), if a stripe_head has its STRIPE_ON_UNPLUG_LIST
set, it indicates that this stripe_head is already in the raid5_plug_cb
list and release_stripe() would be called instead to drop a reference
count. Otherwise, the STRIPE_ON_UNPLUG_LIST bit would be set for this
stripe_head and it will get queued into the raid5_plug_cb list.

Since break_stripe_batch_list() did not preserve STRIPE_ON_UNPLUG_LIST,
A stripe could be re-added to plug list while it is still on that list
in the following situation. If stripe_head A is added to another
stripe_head B's batch list, in this case A will have its
batch_head != NULL and be added into the plug list. After that,
stripe_head B gets handled and called break_stripe_batch_list() to
reset all the batched stripe_head(including A which is still on
the plug list)'s state and reset their batch_head to NULL.
Before the plug list gets processed, if there is another write request
comes in and get stripe_head A, A will have its batch_head == NULL
(cleared by calling break_stripe_batch_list() on B) and be added to
plug list once again.

Signed-off-by: Dennis Yang <dennisyang@qnap.com>
Cc: stable@vger.kernel.org (v4.1+)
Signed-off-by: Shaohua Li <shli@fb.com>
2017-09-05 22:51:13 -07:00
Shaohua Li 3664847d95 md/raid5: fix a race condition in stripe batch
We have a race condition in below scenario, say have 3 continuous stripes, sh1,
sh2 and sh3, sh1 is the stripe_head of sh2 and sh3:

CPU1				CPU2				CPU3
handle_stripe(sh3)
				stripe_add_to_batch_list(sh3)
				-> lock(sh2, sh3)
				-> lock batch_lock(sh1)
				-> add sh3 to batch_list of sh1
				-> unlock batch_lock(sh1)
								clear_batch_ready(sh1)
								-> lock(sh1) and batch_lock(sh1)
								-> clear STRIPE_BATCH_READY for all stripes in batch_list
								-> unlock(sh1) and batch_lock(sh1)
->clear_batch_ready(sh3)
-->test_and_clear_bit(STRIPE_BATCH_READY, sh3)
--->return 0 as sh->batch == NULL
				-> sh3->batch_head = sh1
				-> unlock (sh2, sh3)

In CPU1, handle_stripe will continue handle sh3 even it's in batch stripe list
of sh1. By moving sh3->batch_head assignment in to batch_lock, we make it
impossible to clear STRIPE_BATCH_READY before batch_head is set.

Thanks Stephane for helping debug this tricky issue.

Reported-and-tested-by: Stephane Thiell <sthiell@stanford.edu>
Cc: stable@vger.kernel.org (v4.1+)
Signed-off-by: Shaohua Li <shli@fb.com>
2017-09-05 10:57:49 -07:00
Pawel Baldysiak ddc088238c md: Runtime support for multiple ppls
Increase PPL area to 1MB and use it as circular buffer to store PPL. The
entry with highest generation number is the latest one. If PPL to be
written is larger then space left in a buffer, rewind the buffer to the
start (don't wrap it).

Signed-off-by: Pawel Baldysiak <pawel.baldysiak@intel.com>
Signed-off-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-08-28 07:45:48 -07:00
Guoqing Jiang 27a4ff8f49 raid5: remove raid5_build_block
Now raid5_build_block is just called to set the
sector of r5dev, raid5_compute_blocknr can be
used directly for the purpose.

Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-08-25 10:21:47 -07:00
Song Liu 9c72a18e46 md/raid5: release/flush io in raid5_do_work()
In raid5, there are scenarios where some ios are deferred to a later
time, and some IO need a flush to complete. To make sure we make
progress with these IOs, we need to call the following functions:

    flush_deferred_bios(conf);
    r5l_flush_stripe_to_raid(conf->log);

Both of these functions are called in raid5d(), but missing in
raid5_do_work(). As a result, these functions are not called
when multi-threading (group_thread_cnt > 0) is enabled. This patch
adds calls to these function to raid5_do_work().

Note for stable branches:

  r5l_flush_stripe_to_raid(conf->log) is need for 4.4+
  flush_deferred_bios(conf) is only needed for 4.11+

Cc: stable@vger.kernel.org (4.4+)
Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-08-24 10:06:19 -07:00
Christoph Hellwig 74d46992e0 block: replace bi_bdev with a gendisk pointer and partitions index
This way we don't need a block_device structure to submit I/O.  The
block_device has different life time rules from the gendisk and
request_queue and is usually only available when the block device node
is open.  Other callers need to explicitly create one (e.g. the lightnvm
passthrough code, or the new nvme multipathing code).

For the actual I/O path all that we need is the gendisk, which exists
once per block device.  But given that the block layer also does
partition remapping we additionally need a partition index, which is
used for said remapping in generic_make_request.

Note that all the block drivers generally want request_queue or
sometimes the gendisk, so this removes a layer of indirection all
over the stack.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2017-08-23 12:49:55 -06:00
Christoph Hellwig 10433d04b8 raid5: remove a call to get_start_sect
The block layer always remaps partitions before calling into the
->make_request methods of drivers.  Thus the call to get_start_sect in
in_chunk_boundary will always return 0 and can be removed.

Reviewed-by: Shaohua Li <shli@fb.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2017-08-23 12:49:49 -06:00
Ofer Heifetz 7e96d55963 md/raid5: add thread_group worker async_tx_issue_pending_all
Since thread_group worker and raid5d kthread are not in sync, if
worker writes stripe before raid5d then requests will be waiting
for issue_pendig.

Issue observed when building raid5 with ext4, in some build runs
jbd2 would get hung and requests were waiting in the HW engine
waiting to be issued.

Fix this by adding a call to async_tx_issue_pending_all in the
raid5_do_work.

Signed-off-by: Ofer Heifetz <oferh@marvell.com>
Cc: stable@vger.kernel.org
Signed-off-by: Shaohua Li <shli@fb.com>
2017-07-24 07:49:15 -07:00
Guoqing Jiang 6308d8e3d4 md: simplify code with bio_io_error
Since bio_io_error sets bi_status to BLK_STS_IOERR,
and calls bio_endio, so we can use it directly.

And as mentioned by Shaohua, there are also two
places in raid5.c can use bio_io_error either.

Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-07-21 13:16:52 -07:00
Xiao Ni b5d27718f3 Raid5 should update rdev->sectors after reshape
The raid5 md device is created by the disks which we don't use the total size. For example,
the size of the device is 5G and it just uses 3G of the devices to create one raid5 device.
Then change the chunksize and wait reshape to finish. After reshape finishing stop the raid
and assemble it again. It fails.
mdadm -CR /dev/md0 -l5 -n3 /dev/loop[0-2] --size=3G --chunk=32 --assume-clean
mdadm /dev/md0 --grow --chunk=64
wait reshape to finish
mdadm -S /dev/md0
mdadm -As
The error messages:
[197519.814302] md: loop1 does not have a valid v1.2 superblock, not importing!
[197519.821686] md: md_import_device returned -22

After reshape the data offset is changed. It selects backwards direction in this condition.
In function super_1_load it compares the available space of the underlying device with
sb->data_size. The new data offset gets bigger after reshape. So super_1_load returns -EINVAL.
rdev->sectors is updated in md_finish_reshape. Then sb->data_size is set in super_1_sync based
on rdev->sectors. So add md_finish_reshape in end_reshape.

Signed-off-by: Xiao Ni <xni@redhat.com>
Acked-by: Guoqing Jiang <gqjiang@suse.com>
Cc: stable@vger.kernel.org
Signed-off-by: Shaohua Li <shli@fb.com>
2017-07-10 10:43:58 -07:00
Linus Torvalds 026d15f6b9 Merge branch 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md
Pull MD update from Shaohua Li:

 - fixed deadlock in MD suspend and a potential bug in bio allocation
   (Neil Brown)

 - fixed signal issue (Mikulas Patocka)

 - fixed typo in FailFast test (Guoqing Jiang)

 - other trival fixes

* 'for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/shli/md:
  MD: fix sleep in atomic
  MD: fix a null dereference
  md: use a separate bio_set for synchronous IO.
  md: change the initialization value for a spare device spot to MD_DISK_ROLE_SPARE
  md/raid1: remove unused bio in sync_request_write
  md/raid10: fix FailFast test for wrong device
  md: don't use flush_signals in userspace processes
  md: fix deadlock between mddev_suspend() and md_write_start()
2017-07-08 12:50:18 -07:00
NeilBrown 011067b056 blk: replace bioset_create_nobvec() with a flags arg to bioset_create()
"flags" arguments are often seen as good API design as they allow
easy extensibility.
bioset_create_nobvec() is implemented internally as a variation in
flags passed to __bioset_create().

To support future extension, make the internal structure part of the
API.
i.e. add a 'flags' argument to bioset_create() and discard
bioset_create_nobvec().

Note that the bio_split allocations in drivers/md/raid* do not need
the bvec mempool - they should have used bioset_create_nobvec().

Suggested-by: Christoph Hellwig <hch@infradead.org>
Reviewed-by: Christoph Hellwig <hch@infradead.org>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
2017-06-18 12:40:59 -06:00
Mikulas Patocka f9c79bc05a md: don't use flush_signals in userspace processes
The function flush_signals clears all pending signals for the process. It
may be used by kernel threads when we need to prepare a kernel thread for
responding to signals. However using this function for an userspaces
processes is incorrect - clearing signals without the program expecting it
can cause misbehavior.

The raid1 and raid5 code uses flush_signals in its request routine because
it wants to prepare for an interruptible wait. This patch drops
flush_signals and uses sigprocmask instead to block all signals (including
SIGKILL) around the schedule() call. The signals are not lost, but the
schedule() call won't respond to them.

Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Cc: stable@vger.kernel.org
Acked-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-06-13 10:18:02 -07:00
NeilBrown cc27b0c78c md: fix deadlock between mddev_suspend() and md_write_start()
If mddev_suspend() races with md_write_start() we can deadlock
with mddev_suspend() waiting for the request that is currently
in md_write_start() to complete the ->make_request() call,
and md_write_start() waiting for the metadata to be updated
to mark the array as 'dirty'.
As metadata updates done by md_check_recovery() only happen then
the mddev_lock() can be claimed, and as mddev_suspend() is often
called with the lock held, these threads wait indefinitely for each
other.

We fix this by having md_write_start() abort if mddev_suspend()
is happening, and ->make_request() aborts if md_write_start()
aborted.
md_make_request() can detect this abort, decrease the ->active_io
count, and wait for mddev_suspend().

Reported-by: Nix <nix@esperi.org.uk>
Fix: 68866e425be2(MD: no sync IO while suspended)
Cc: stable@vger.kernel.org
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-06-13 10:18:01 -07:00
Jens Axboe 8f66439eec Linux 4.12-rc5
-----BEGIN PGP SIGNATURE-----
 
 iQEcBAABAgAGBQJZPdbLAAoJEHm+PkMAQRiGx4wH/1nCjfnl6fE8oJ24/1gEAOUh
 biFdqJkYZmlLYHVtYfLm4Ueg4adJdg0wx6qM/4RaAzmQVvLfDV34bc1qBf1+P95G
 kVF+osWyXrZo5cTwkwapHW/KNu4VJwAx2D1wrlxKDVG5AOrULH1pYOYGOpApEkZU
 4N+q5+M0ce0GJpqtUZX+UnI33ygjdDbBxXoFKsr24B7eA0ouGbAJ7dC88WcaETL+
 2/7tT01SvDMo0jBSV0WIqlgXwZ5gp3yPGnklC3F4159Yze6VFrzHMKS/UpPF8o8E
 W9EbuzwxsKyXUifX2GY348L1f+47glen/1sedbuKnFhP6E9aqUQQJXvEO7ueQl4=
 =m2Gx
 -----END PGP SIGNATURE-----

Merge tag 'v4.12-rc5' into for-4.13/block

We've already got a few conflicts and upcoming work depends on some of the
changes that have gone into mainline as regression fixes for this series.

Pull in 4.12-rc5 to resolve these conflicts and make it easier on down stream
trees to continue working on 4.13 changes.

Signed-off-by: Jens Axboe <axboe@kernel.dk>
2017-06-12 08:30:13 -06:00
Christoph Hellwig 4e4cbee93d block: switch bios to blk_status_t
Replace bi_error with a new bi_status to allow for a clear conversion.
Note that device mapper overloaded bi_error with a private value, which
we'll have to keep arround at least for now and thus propagate to a
proper blk_status_t value.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-06-09 09:27:32 -06:00
NeilBrown a415c0f106 md: initialise ->writes_pending in personality modules.
The new per-cpu counter for writes_pending is initialised in
md_alloc(), which is not called by dm-raid.
So dm-raid fails when md_write_start() is called.

Move the initialization to the personality modules
that need it.  This way it is always initialised when needed,
but isn't unnecessarily initialized (requiring memory allocation)
when the personality doesn't use writes_pending.

Reported-by: Heinz Mauelshagen <heinzm@redhat.com>
Fixes: 4ad23a9764 ("MD: use per-cpu counter for writes_pending")
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-06-05 16:04:35 -07:00
Nix e153903686 md: report sector of stripes with check mismatches
This makes it possible, with appropriate filesystem support, for a
sysadmin to tell what is affected by the mismatch, and whether
it should be ignored (if it's inside a swap partition, for
instance).

We ratelimit to prevent log flooding: if there are so many
mismatches that ratelimiting is necessary, the individual messages
are relatively unlikely to be important (either the machine is
swapping like crazy or something is very wrong with the disk).

Signed-off-by: Nick Alcock <nick.alcock@oracle.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-05-24 16:31:13 -07:00
Song Liu 5ddf0440a1 md/r5cache: handle sync with data in write back cache
Currently, sync of raid456 array cannot make progress when hitting
data in writeback r5cache.

This patch fixes this issue by flushing cached data of the stripe
before processing the sync request. This is achived by:

1. In handle_stripe(), do not set STRIPE_SYNCING if the stripe is
   in write back cache;
2. In r5c_try_caching_write(), handle the stripe in sync with write
   through;
3. In do_release_stripe(), make stripe in sync write out and send
   it to the state machine.

Shaohua: explictly set STRIPE_HANDLE after write out completed

Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-05-11 22:14:40 -07:00
Song Liu 70d466f760 md/r5cache: gracefully handle journal device errors for writeback mode
For the raid456 with writeback cache, when journal device failed during
normal operation, it is still possible to persist all data, as all
pending data is still in stripe cache. However, it is necessary to handle
journal failure gracefully.

During journal failures, the following logic handles the graceful shutdown
of journal:
1. raid5_error() marks the device as Faulty and schedules async work
   log->disable_writeback_work;
2. In disable_writeback_work (r5c_disable_writeback_async), the mddev is
   suspended, set to write through, and then resumed. mddev_suspend()
   flushes all cached stripes;
3. All cached stripes need to be flushed carefully to the RAID array.

This patch fixes issues within the process above:
1. In r5c_update_on_rdev_error() schedule disable_writeback_work for
   journal failures;
2. In r5c_disable_writeback_async(), wait for MD_SB_CHANGE_PENDING,
   since raid5_error() updates superblock.
3. In handle_stripe(), allow stripes with data in journal (s.injournal > 0)
   to make progress during log_failed;
4. In delay_towrite(), if log failed only process data in the cache (skip
   new writes in dev->towrite);
5. In __get_priority_stripe(), process loprio_list during journal device
   failures.
6. In raid5_remove_disk(), wait for all cached stripes are flushed before
   calling log_exit().

Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-05-11 22:11:11 -07:00
Artur Paszkiewicz 2214c260c7 md: don't return -EAGAIN in md_allow_write for external metadata arrays
This essentially reverts commit b5470dc5fc ("md: resolve external
metadata handling deadlock in md_allow_write") with some adjustments.

Since commit 6791875e2e ("md: make reconfig_mutex optional for writes
to md sysfs files.") changing array_state to 'active' does not use
mddev_lock() and will not cause a deadlock with md_allow_write(). This
revert simplifies userspace tools that write to sysfs attributes like
"stripe_cache_size" or "consistency_policy" because it removes the need
for special handling for external metadata arrays, checking for EAGAIN
and retrying the write.

Signed-off-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-05-08 10:32:59 -07:00
Julia Cartwright 3d05f3aed5 md/raid5: make use of spin_lock_irq over local_irq_disable + spin_lock
On mainline, there is no functional difference, just less code, and
symmetric lock/unlock paths.

On PREEMPT_RT builds, this fixes the following warning, seen by
Alexander GQ Gerasiov, due to the sleeping nature of spinlocks.

   BUG: sleeping function called from invalid context at kernel/locking/rtmutex.c:993
   in_atomic(): 0, irqs_disabled(): 1, pid: 58, name: kworker/u12:1
   CPU: 5 PID: 58 Comm: kworker/u12:1 Tainted: G        W       4.9.20-rt16-stand6-686 #1
   Hardware name: Supermicro SYS-5027R-WRF/X9SRW-F, BIOS 3.2a 10/28/2015
   Workqueue: writeback wb_workfn (flush-253:0)
   Call Trace:
    dump_stack+0x47/0x68
    ? migrate_enable+0x4a/0xf0
    ___might_sleep+0x101/0x180
    rt_spin_lock+0x17/0x40
    add_stripe_bio+0x4e3/0x6c0 [raid456]
    ? preempt_count_add+0x42/0xb0
    raid5_make_request+0x737/0xdd0 [raid456]

Reported-by: Alexander GQ Gerasiov <gq@redlab-i.ru>
Tested-by: Alexander GQ Gerasiov <gq@redlab-i.ru>
Signed-off-by: Julia Cartwright <julia@ni.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-05-04 13:44:23 -07:00
Shaohua Li e265eb3a30 Merge branch 'md-next' into md-linus 2017-05-01 14:09:21 -07:00
Guoqing Jiang e5bc9c3c54 md: clear WantReplacement once disk is removed
We can clear 'WantReplacement' flag directly no
matter it's replacement existed or not since the
semantic is same as before.

Also since the disk is removed from array, then
it is straightforward to remove 'WantReplacement'
flag and the comments in raid10/5 can be removed
as well.

Signed-off-by: Guoqing Jiang <gqjiang@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-04-25 09:36:29 -07:00
NeilBrown dd7a8f5dee md/raid5: make chunk_aligned_read() split bios more cleanly.
chunk_aligned_read() currently uses fs_bio_set - which is meant for
filesystems to use - and loops if multiple splits are needed, which is
not best practice.
As this is only used for READ requests, not writes, it is unlikely
to cause a problem.  However it is best to be consistent in how
we split bios, and to follow the pattern used in raid1/raid10.

So create a private bioset, bio_split, and use it to perform a single
split, submitting the remainder to generic_make_request() for later
processing.

Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-04-11 10:16:50 -07:00
Artur Paszkiewicz ae1713e296 raid5-ppl: partial parity calculation optimization
In case of read-modify-write, partial partity is the same as the result
of ops_run_prexor5(), so we can just copy sh->dev[pd_idx].page into
sh->ppl_page instead of calculating it again.

Signed-off-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-04-10 12:01:37 -07:00
Artur Paszkiewicz 845b9e229f raid5-ppl: use resize_stripes() when enabling or disabling ppl
Use resize_stripes() instead of raid5_reset_stripe_cache() to allocate
or free sh->ppl_page at runtime for all stripes in the stripe cache.
raid5_reset_stripe_cache() required suspending the mddev and could
deadlock because of GFP_KERNEL allocations.

Move the 'newsize' check to check_reshape() to allow reallocating the
stripes with the same number of disks. Allocate sh->ppl_page in
alloc_stripe() instead of grow_buffers(). Pass 'struct r5conf *conf' as
a parameter to alloc_stripe() because it is needed to check whether to
allocate ppl_page. Add free_stripe() and use it to free stripes rather
than directly call kmem_cache_free(). Also free sh->ppl_page in
free_stripe().

Set MD_HAS_PPL at the end of ppl_init_log() instead of explicitly
setting it in advance and add another parameter to log_init() to allow
calling ppl_init_log() without the bit set. Don't try to calculate
partial parity or add a stripe to log if it does not have ppl_page set.

Enabling ppl can now be performed without suspending the mddev, because
the log won't be used until new stripes are allocated with ppl_page.
Calling mddev_suspend/resume is still necessary when disabling ppl,
because we want all stripes to finish before stopping the log, but
resize_stripes() can be called after mddev_resume() when ppl is no
longer active.

Suggested-by: NeilBrown <neilb@suse.com>
Signed-off-by: Artur Paszkiewicz <artur.paszkiewicz@intel.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-04-10 12:00:49 -07:00
NeilBrown 7471fb77ce md/raid6: Fix anomily when recovering a single device in RAID6.
When recoverying a single missing/failed device in a RAID6,
those stripes where the Q block is on the missing device are
handled a bit differently.  In these cases it is easy to
check that the P block is correct, so we do.  This results
in the P block be destroy.  Consequently the P block needs
to be read a second time in order to compute Q.  This causes
lots of seeks and hurts performance.

It shouldn't be necessary to re-read P as it can be computed
from the DATA.  But we only compute blocks on missing
devices, since c337869d95 ("md: do not compute parity
unless it is on a failed drive").

So relax the change made in that commit to allow computing
of the P block in a RAID6 which it is the only missing that
block.

This makes RAID6 recovery run much faster as the disk just
"before" the recovering device is no longer seeking
back-and-forth.

Reported-by-tested-by: Brad Campbell <lists2009@fnarfbargle.com>
Reviewed-by: Dan Williams <dan.j.williams@intel.com>
Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-04-10 10:35:27 -07:00
Dennis Yang 583da48e38 md: update slab_cache before releasing new stripes when stripes resizing
When growing raid5 device on machine with small memory, there is chance that
mdadm will be killed and the following bug report can be observed. The same
bug could also be reproduced in linux-4.10.6.

[57600.075774] BUG: unable to handle kernel NULL pointer dereference at           (null)
[57600.083796] IP: [<ffffffff81a6aa87>] _raw_spin_lock+0x7/0x20
[57600.110378] PGD 421cf067 PUD 4442d067 PMD 0
[57600.114678] Oops: 0002 [#1] SMP
[57600.180799] CPU: 1 PID: 25990 Comm: mdadm Tainted: P           O    4.2.8 #1
[57600.187849] Hardware name: To be filled by O.E.M. To be filled by O.E.M./MAHOBAY, BIOS QV05AR66 03/06/2013
[57600.197490] task: ffff880044e47240 ti: ffff880043070000 task.ti: ffff880043070000
[57600.204963] RIP: 0010:[<ffffffff81a6aa87>]  [<ffffffff81a6aa87>] _raw_spin_lock+0x7/0x20
[57600.213057] RSP: 0018:ffff880043073810  EFLAGS: 00010046
[57600.218359] RAX: 0000000000000000 RBX: 000000000000000c RCX: ffff88011e296dd0
[57600.225486] RDX: 0000000000000001 RSI: ffffe8ffffcb46c0 RDI: 0000000000000000
[57600.232613] RBP: ffff880043073878 R08: ffff88011e5f8170 R09: 0000000000000282
[57600.239739] R10: 0000000000000005 R11: 28f5c28f5c28f5c3 R12: ffff880043073838
[57600.246872] R13: ffffe8ffffcb46c0 R14: 0000000000000000 R15: ffff8800b9706a00
[57600.253999] FS:  00007f576106c700(0000) GS:ffff88011e280000(0000) knlGS:0000000000000000
[57600.262078] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[57600.267817] CR2: 0000000000000000 CR3: 00000000428fe000 CR4: 00000000001406e0
[57600.274942] Stack:
[57600.276949]  ffffffff8114ee35 ffff880043073868 0000000000000282 000000000000eb3f
[57600.284383]  ffffffff81119043 ffff880043073838 ffff880043073838 ffff88003e197b98
[57600.291820]  ffffe8ffffcb46c0 ffff88003e197360 0000000000000286 ffff880043073968
[57600.299254] Call Trace:
[57600.301698]  [<ffffffff8114ee35>] ? cache_flusharray+0x35/0xe0
[57600.307523]  [<ffffffff81119043>] ? __page_cache_release+0x23/0x110
[57600.313779]  [<ffffffff8114eb53>] kmem_cache_free+0x63/0xc0
[57600.319344]  [<ffffffff81579942>] drop_one_stripe+0x62/0x90
[57600.324915]  [<ffffffff81579b5b>] raid5_cache_scan+0x8b/0xb0
[57600.330563]  [<ffffffff8111b98a>] shrink_slab.part.36+0x19a/0x250
[57600.336650]  [<ffffffff8111e38c>] shrink_zone+0x23c/0x250
[57600.342039]  [<ffffffff8111e4f3>] do_try_to_free_pages+0x153/0x420
[57600.348210]  [<ffffffff8111e851>] try_to_free_pages+0x91/0xa0
[57600.353959]  [<ffffffff811145b1>] __alloc_pages_nodemask+0x4d1/0x8b0
[57600.360303]  [<ffffffff8157a30b>] check_reshape+0x62b/0x770
[57600.365866]  [<ffffffff8157a4a5>] raid5_check_reshape+0x55/0xa0
[57600.371778]  [<ffffffff81583df7>] update_raid_disks+0xc7/0x110
[57600.377604]  [<ffffffff81592b73>] md_ioctl+0xd83/0x1b10
[57600.382827]  [<ffffffff81385380>] blkdev_ioctl+0x170/0x690
[57600.388307]  [<ffffffff81195238>] block_ioctl+0x38/0x40
[57600.393525]  [<ffffffff811731c5>] do_vfs_ioctl+0x2b5/0x480
[57600.399010]  [<ffffffff8115e07b>] ? vfs_write+0x14b/0x1f0
[57600.404400]  [<ffffffff811733cc>] SyS_ioctl+0x3c/0x70
[57600.409447]  [<ffffffff81a6ad97>] entry_SYSCALL_64_fastpath+0x12/0x6a
[57600.415875] Code: 00 00 00 00 55 48 89 e5 8b 07 85 c0 74 04 31 c0 5d c3 ba 01 00 00 00 f0 0f b1 17 85 c0 75 ef b0 01 5d c3 90 31 c0 ba 01 00 00 00 <f0> 0f b1 17 85 c0 75 01 c3 55 89 c6 48 89 e5 e8 85 d1 63 ff 5d
[57600.435460] RIP  [<ffffffff81a6aa87>] _raw_spin_lock+0x7/0x20
[57600.441208]  RSP <ffff880043073810>
[57600.444690] CR2: 0000000000000000
[57600.448000] ---[ end trace cbc6b5cc4bf9831d ]---

The problem is that resize_stripes() releases new stripe_heads before assigning new
slab cache to conf->slab_cache. If the shrinker function raid5_cache_scan() gets called
after resize_stripes() starting releasing new stripes but right before new slab cache
being assigned, it is possible that these new stripe_heads will be freed with the old
slab_cache which was already been destoryed and that triggers this bug.

Signed-off-by: Dennis Yang <dennisyang@qnap.com>
Fixes: edbe83ab4c ("md/raid5: allow the stripe_cache to grow and shrink.")
Cc: stable@vger.kernel.org (4.1+)
Reviewed-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-04-10 09:27:12 -07:00
Christoph Hellwig 48920ff2a5 block: remove the discard_zeroes_data flag
Now that we use the proper REQ_OP_WRITE_ZEROES operation everywhere we can
kill this hack.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-08 11:25:38 -06:00
Christoph Hellwig 3deff1a70d md: support REQ_OP_WRITE_ZEROES
Copy & paste from the REQ_OP_WRITE_SAME code.

Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-08 11:25:38 -06:00
NeilBrown fbbaf700e7 block: trace completion of all bios.
Currently only dm and md/raid5 bios trigger
trace_block_bio_complete().  Now that we have bio_chain() and
bio_inc_remaining(), it is not possible, in general, for a driver to
know when the bio is really complete.  Only bio_endio() knows that.

So move the trace_block_bio_complete() call to bio_endio().

Now trace_block_bio_complete() pairs with trace_block_bio_queue().
Any bio for which a 'queue' event is traced, will subsequently
generate a 'complete' event.

There are a few cases where completion tracing is not wanted.
1/ If blk_update_request() has already generated a completion
   trace event at the 'request' level, there is no point generating
   one at the bio level too.  In this case the bi_sector and bi_size
   will have changed, so the bio level event would be wrong

2/ If the bio hasn't actually been queued yet, but is being aborted
   early, then a trace event could be confusing.  Some filesystems
   call bio_endio() but do not want tracing.

3/ The bio_integrity code interposes itself by replacing bi_end_io,
   then restoring it and calling bio_endio() again.  This would produce
   two identical trace events if left like that.

To handle these, we introduce a flag BIO_TRACE_COMPLETION and only
produce the trace event when this is set.
We address point 1 above by clearing the flag in blk_update_request().
We address point 2 above by only setting the flag when
generic_make_request() is called.
We address point 3 above by clearing the flag after generating a
completion event.

When bio_split() is used on a bio, particularly in blk_queue_split(),
there is an extra complication.  A new bio is split off the front, and
may be handle directly without going through generic_make_request().
The old bio, which has been advanced, is passed to
generic_make_request(), so it will trigger a trace event a second
time.
Probably the best result when a split happens is to see a single
'queue' event for the whole bio, then multiple 'complete' events - one
for each component.  To achieve this was can:
- copy the BIO_TRACE_COMPLETION flag to the new bio in bio_split()
- avoid generating a 'queue' event if BIO_TRACE_COMPLETION is already set.
This way, the split-off bio won't create a queue event, the original
won't either even if it re-submitted to generic_make_request(),
but both will produce completion events, each for their own range.

So if generic_make_request() is called (which generates a QUEUED
event), then bi_endio() will create a single COMPLETE event for each
range that the bio is split into, unless the driver has explicitly
requested it not to.

Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
2017-04-07 09:40:52 -06:00
Song Liu 0bb0c10500 md/raid5: use consistency_policy to remove journal feature
When journal device of an array fails, the array is forced into read-only
mode. To make the array normal without adding another journal device, we
need to remove journal _feature_ from the array.

This patch allows remove journal _feature_ from an array, For journal
existing journal should be either missing or faulty.

To remove journal feature, it is necessary to remove the journal device
first:

  mdadm --fail /dev/md0 /dev/sdb
  mdadm: set /dev/sdb faulty in /dev/md0
  mdadm --remove /dev/md0 /dev/sdb
  mdadm: hot removed /dev/sdb from /dev/md0

Then the journal feature can be removed by echoing into the sysfs file:

 cat /sys/block/md0/md/consistency_policy
 journal

 echo resync > /sys/block/md0/md/consistency_policy
 cat /sys/block/md0/md/consistency_policy
 resync

Signed-off-by: Song Liu <songliubraving@fb.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-03-27 12:02:33 -07:00
Zhilong Liu 3560741e31 md: fix several trivial typos in comments
Signed-off-by: Zhilong Liu <zlliu@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-03-23 22:54:57 -07:00
NeilBrown 84dd97a690 md/raid5: don't test ->writes_pending in raid5_remove_disk
This test on ->writes_pending cannot be safe as the counter
can be incremented at any moment and cannot be locked against.

Change it to test conf->active_stripes, which at least
can be locked against.  More changes are still needed.

A future patch will change ->writes_pending, and testing it here will
be very inconvenient.

Signed-off-by: NeilBrown <neilb@suse.com>
Signed-off-by: Shaohua Li <shli@fb.com>
2017-03-22 19:18:05 -07:00