Preparation for adding more declarations.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

A later patch in this series will modify blk_mq_debugfs_register()
such that it uses q->kobj.parent to determine the name of a
request queue. Hence make sure that that pointer is initialized
before blk_mq_debugfs_register() is called. To avoid lock inversion,
protect sysfs / debugfs registration with the queue sysfs_lock
instead of the global mutex all_q_mutex.
Signed-off-by: NBart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Reviewed-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

When CFQ is used as an elevator, it disables writeback throttling
because they don't play well together. Later when a different elevator
is chosen for the device, writeback throttling doesn't get enabled
again as it should. Make sure CFQ enables writeback throttling (if it
should be enabled by default) when we switch from it to another IO
scheduler.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

Now that we use the proper REQ_OP_WRITE_ZEROES operation everywhere we can
kill this hack.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NMartin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

In elevator_switch(), if blk_mq_init_sched() fails, we attempt to fall
back to the original scheduler. However, at this point, we've already
torn down the original scheduler's tags, so this causes a crash. Doing
the fallback like the legacy elevator path is much harder for mq, so fix
it by just falling back to none, instead.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

CONFIG_DEBUG_TEST_DRIVER_REMOVE found a possible leak of q->rq_wb when a
request queue is reregistered. This has been a problem since wbt was
introduced, but the WARN_ON(!list_empty(&stats->callbacks)) in the
blk-stat rework exposed it. Fix it by cleaning up wbt when we unregister
the queue.

Fixes: 87760e5e ("block: hook up writeback throttling")
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Now that the remaining drivers have been converted to one request queue
per gendisk, let's warn if a request queue gets registered more than
once. This will catch future drivers which might do it inadvertently or
any old drivers that I may have missed.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

The throtl_slice is 100ms by default. This is a long time for SSD, a lot
of IO can run. To make cgroups have smoother throughput, we choose a
small value (20ms) for SSD.
Signed-off-by: NShaohua Li <shli@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

throtl_slice is important for blk-throttling. It's called slice
internally but it really is a time window blk-throttling samples data.
blk-throttling will make decision based on the samplings. An example is
bandwidth measurement. A cgroup's bandwidth is measured in the time
interval of throtl_slice.

A small throtl_slice meanse cgroups have smoother throughput but burn
more CPUs. It has 100ms default value, which is not appropriate for all
disks. A fast SSD can dispatch a lot of IOs in 100ms. This patch makes
it tunable.

Since throtl_slice isn't a time slice, the sysfs name
'throttle_sample_time' reflects its character better.
Signed-off-by: NShaohua Li <shli@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently, statistics are gathered in ~0.13s windows, and users grab the
statistics whenever they need them. This is not ideal for both in-tree
users:

1. Writeback throttling wants its own dynamically sized window of
   statistics. Since the blk-stats statistics are reset after every
   window and the wbt windows don't line up with the blk-stats windows,
   wbt doesn't see every I/O.
2. Polling currently grabs the statistics on every I/O. Again, depending
   on how the window lines up, we may miss some I/Os. It's also
   unnecessary overhead to get the statistics on every I/O; the hybrid
   polling heuristic would be just as happy with the statistics from the
   previous full window.

This reworks the blk-stats infrastructure to be callback-based: users
register a callback that they want called at a given time with all of
the statistics from the window during which the callback was active.
Users can dynamically bucketize the statistics. wbt and polling both
currently use read vs. write, but polling can be extended to further
subdivide based on request size.

The callbacks are kept on an RCU list, and each callback has percpu
stats buffers. There will only be a few users, so the overhead on the
I/O completion side is low. The stats flushing is also simplified
considerably: since the timer function is responsible for clearing the
statistics, we don't have to worry about stale statistics.

wbt is a trivial conversion. After the conversion, the windowing problem
mentioned above is fixed.

For polling, we register an extra callback that caches the previous
window's statistics in the struct request_queue for the hybrid polling
heuristic to use.

Since we no longer have a single stats buffer for the request queue,
this also removes the sysfs and debugfs stats entries. To replace those,
we add a debugfs entry for the poll statistics.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

The stats buckets will become generic soon, so make the existing users
use the common READ and WRITE definitions instead of one internal to
blk-stat.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

For legacy scheduling, we always call ioc_exit_icq() with both the
ioc and queue lock held. This poses a problem for blk-mq with
scheduling, since the queue lock isn't what we use in the scheduler.
And since we don't need the queue lock held for ioc exit there,
don't grab it and leave any extra locking up to the blk-mq scheduler.
Reported-by: NPaolo Valente <paolo.valente@linaro.org>
Tested-by: NPaolo Valente <paolo.valente@linaro.org>
Reviewed-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

When a new disk shows up, sysfs queue directory is created before elevator
is registered. This allows a user to attempt a scheduler switch even though
the initial registration hasn't completed yet.

In one scenario, blk_register_queue() calls elv_register_queue() and
right before cfq_registered_queue() is called, another process executes
elevator_switch() and replaces q->elevator with deadline scheduler. When
cfq_registered_queue() executes it interprets e->elevator_data as struct
cfq_data even though it is actually struct deadline_data.

Grab q->sysfs_lock in blk_register_queue() to synchronize with sysfs
callers.
Signed-off-by: NTahsin Erdogan <tahsin@google.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Add a new merge strategy that merges discard bios into a request until the
maximum number of discard ranges (or the maximum discard size) is reached
from the plug merging code.  I/O scheduler merging is not wired up yet
but might also be useful, although not for fast devices like NVMe which
are the only user for now.

Note that for now we don't support limiting the size of each discard range,
but if needed that can be added later.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

I noticed that when booting with a default blk-mq I/O scheduler, the
/sys/block/*/queue/iosched directory was missing. However, switching
after boot did create the directory. This is because we skip the initial
elevator register/unregister when we don't have a ->request_fn(), but we
should still do it for the ->mq_ops case.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

This needs to happen after we tear down blktrace.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Instead of storing backing_dev_info inside struct request_queue,
allocate it dynamically, reference count it, and free it when the last
reference is dropped. Currently only request_queue holds the reference
but in the following patch we add other users referencing
backing_dev_info.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

We will want to have struct backing_dev_info allocated separately from
struct request_queue. As the first step add pointer to backing_dev_info
to request_queue and convert all users touching it. No functional
changes in this patch.
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@fb.com>

This mirrors the blk-mq capabilities to allocate extra drivers-specific
data behind struct request by setting a cmd_size field, as well as having
a constructor / destructor for it.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NMartin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: NHannes Reinecke <hare@suse.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

We ran into a funky issue, where someone doing 256K buffered reads saw
128K requests at the device level.  Turns out it is read-ahead capping
the request size, since we use 128K as the default setting.  This
doesn't make a lot of sense - if someone is issuing 256K reads, they
should see 256K reads, regardless of the read-ahead setting, if the
underlying device can support a 256K read in a single command.

This patch introduces a bdi hint, io_pages.  This is the soft max IO
size for the lower level, I've hooked it up to the bdev settings here.
Read-ahead is modified to issue the maximum of the user request size,
and the read-ahead max size, but capped to the max request size on the
device side.  The latter is done to avoid reading ahead too much, if the
application asks for a huge read.  With this patch, the kernel behaves
like the application expects.

Link: http://lkml.kernel.org/r/1479498073-8657-1-git-send-email-axboe@fb.comSigned-off-by: NJens Axboe <axboe@fb.com>
Acked-by: NJohannes Weiner <hannes@cmpxchg.org>
Signed-off-by: NAndrew Morton <akpm@linux-foundation.org>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

This adds a new block layer operation to zero out a range of
LBAs. This allows to implement zeroing for devices that don't use
either discard with a predictable zero pattern or WRITE SAME of zeroes.
The prominent example of that is NVMe with the Write Zeroes command,
but in the future, this should also help with improving the way
zeroing discards work. For this operation, suitable entry is exported in
sysfs which indicate the number of maximum bytes allowed in one
write zeroes operation by the device.
Signed-off-by: NChaitanya Kulkarni <chaitanya.kulkarni@hgst.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently there's no way to enable wbt if it's not enabled in the
kernel config by default for a device. Allow a write to the
'wbt_lat_usec' queue sysfs file to enable wbt.

This is useful for both the kernel config case, but also if the
device is CFQ managed and it was turned off by default.
Signed-off-by: NJens Axboe <axboe@fb.com>

Allow a write of '-1' to reset the default latency target for
a given device. This removes knowledge of the different default
settings for rotational vs non-rotational from user space.
Signed-off-by: NJens Axboe <axboe@fb.com>

The previous commit introduced the hybrid sleep/poll mode. Take
that one step further, and use the completion latencies to
automatically sleep for half the mean completion time. This is
a good approximation.

This changes the 'io_poll_delay' sysfs file a bit to expose the
various options. Depending on the value, the polling code will
behave differently:

-1	Never enter hybrid sleep mode
 0	Use half of the completion mean for the sleep delay
>0	Use this specific value as the sleep delay
Signed-off-by: NJens Axboe <axboe@fb.com>
Tested-By: NStephen Bates <sbates@raithlin.com>
Reviewed-By: NStephen Bates <sbates@raithlin.com>

This patch enables a hybrid polling mode. Instead of polling after IO
submission, we can induce an artificial delay, and then poll after that.
For example, if the IO is presumed to complete in 8 usecs from now, we
can sleep for 4 usecs, wake up, and then do our polling. This still puts
a sleep/wakeup cycle in the IO path, but instead of the wakeup happening
after the IO has completed, it'll happen before. With this hybrid
scheme, we can achieve big latency reductions while still using the same
(or less) amount of CPU.
Signed-off-by: NJens Axboe <axboe@fb.com>
Tested-By: NStephen Bates <sbates@raithlin.com>
Reviewed-By: NStephen Bates <sbates@raithlin.com>

Again a leftover from when the throttling code was generic. Now that we
just have the block user, get rid of the stat ops and indirections.
Signed-off-by: NJens Axboe <axboe@fb.com>

Enable throttling of buffered writeback to make it a lot
more smooth, and has way less impact on other system activity.
Background writeback should be, by definition, background
activity. The fact that we flush huge bundles of it at the time
means that it potentially has heavy impacts on foreground workloads,
which isn't ideal. We can't easily limit the sizes of writes that
we do, since that would impact file system layout in the presence
of delayed allocation. So just throttle back buffered writeback,
unless someone is waiting for it.

The algorithm for when to throttle takes its inspiration in the
CoDel networking scheduling algorithm. Like CoDel, blk-wb monitors
the minimum latencies of requests over a window of time. In that
window of time, if the minimum latency of any request exceeds a
given target, then a scale count is incremented and the queue depth
is shrunk. The next monitoring window is shrunk accordingly. Unlike
CoDel, if we hit a window that exhibits good behavior, then we
simply increment the scale count and re-calculate the limits for that
scale value. This prevents us from oscillating between a
close-to-ideal value and max all the time, instead remaining in the
windows where we get good behavior.

Unlike CoDel, blk-wb allows the scale count to to negative. This
happens if we primarily have writes going on. Unlike positive
scale counts, this doesn't change the size of the monitoring window.
When the heavy writers finish, blk-bw quickly snaps back to it's
stable state of a zero scale count.

The patch registers a sysfs entry, 'wb_lat_usec'. This sets the latency
target to me met. It defaults to 2 msec for non-rotational storage, and
75 msec for rotational storage. Setting this value to '0' disables
blk-wb. Generally, a user would not have to touch this setting.

We don't enable WBT on devices that are managed with CFQ, and have
a non-root block cgroup attached. If we have a proportional share setup
on this particular disk, then the wbt throttling will interfere with
that. We don't have a strong need for wbt for that case, since we will
rely on CFQ doing that for us.
Signed-off-by: NJens Axboe <axboe@fb.com>

For legacy block, we simply track them in the request queue. For
blk-mq, we track them on a per-sw queue basis, which we can then
sum up through the hardware queues and finally to a per device
state.

The stats are tracked in, roughly, 0.1s interval windows.

Add sysfs files to display the stats.

The feature is off by default, to avoid any extra overhead. In-kernel
users of it can turn it on by setting QUEUE_FLAG_STATS in the queue
flags. We currently don't turn it on if someone just reads any of
the stats files, that is something we could add as well.
Signed-off-by: NJens Axboe <axboe@fb.com>

The queue limits already have a 'chunk_sectors' setting, so
we should be presenting it via sysfs.
Signed-off-by: NHannes Reinecke <hare@suse.de>

[Damien: Updated Documentation/ABI/testing/sysfs-block]
Signed-off-by: NDamien Le Moal <damien.lemoal@hgst.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NMartin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: NShaun Tancheff <shaun.tancheff@seagate.com>
Tested-by: NShaun Tancheff <shaun.tancheff@seagate.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Add the zoned queue limit to indicate the zoning model of a block device.
Defined values are 0 (BLK_ZONED_NONE) for regular block devices,
1 (BLK_ZONED_HA) for host-aware zone block devices and 2 (BLK_ZONED_HM)
for host-managed zone block devices. The standards defined drive managed
model is not defined here since these block devices do not provide any
command for accessing zone information. Drive managed model devices will
be reported as BLK_ZONED_NONE.

The helper functions blk_queue_zoned_model and bdev_zoned_model return
the zoned limit and the functions blk_queue_is_zoned and bdev_is_zoned
return a boolean for callers to test if a block device is zoned.

The zoned attribute is also exported as a string to applications via
sysfs. BLK_ZONED_NONE shows as "none", BLK_ZONED_HA as "host-aware" and
BLK_ZONED_HM as "host-managed".
Signed-off-by: NDamien Le Moal <damien.lemoal@hgst.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NMartin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: NShaun Tancheff <shaun.tancheff@seagate.com>
Tested-by: NShaun Tancheff <shaun.tancheff@seagate.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Enable devices without a gendisk instance to register itself with blk-mq
and expose the associated multi-queue sysfs entries.
Signed-off-by: NMatias Bjørling <m@bjorling.me>
Signed-off-by: NJens Axboe <axboe@fb.com>

Provides the ability to identify DAX enabled devices in userspace.
Signed-off-by: NYigal Korman <yigal@plexistor.com>
Signed-off-by: NToshi Kani <toshi.kani@hpe.com>
Acked-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NMike Snitzer <snitzer@redhat.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Add an internal helper and flag for setting whether a queue has
write back caching, or write through (or none). Add a sysfs file
to show this as well, and make it changeable from user space.

This will replace the (awkward) blk_queue_flush() interface that
drivers currently use to inform the block layer of write cache state
and capabilities.
Signed-off-by: NJens Axboe <axboe@fb.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>

PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time
ago with promise that one day it will be possible to implement page
cache with bigger chunks than PAGE_SIZE.

This promise never materialized.  And unlikely will.

We have many places where PAGE_CACHE_SIZE assumed to be equal to
PAGE_SIZE.  And it's constant source of confusion on whether
PAGE_CACHE_* or PAGE_* constant should be used in a particular case,
especially on the border between fs and mm.

Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much
breakage to be doable.

Let's stop pretending that pages in page cache are special.  They are
not.

The changes are pretty straight-forward:

 - <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;

 - <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>;

 - PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN};

 - page_cache_get() -> get_page();

 - page_cache_release() -> put_page();

This patch contains automated changes generated with coccinelle using
script below.  For some reason, coccinelle doesn't patch header files.
I've called spatch for them manually.

The only adjustment after coccinelle is revert of changes to
PAGE_CAHCE_ALIGN definition: we are going to drop it later.

There are few places in the code where coccinelle didn't reach.  I'll
fix them manually in a separate patch.  Comments and documentation also
will be addressed with the separate patch.

virtual patch

@@
expression E;
@@
- E << (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E

@@
expression E;
@@
- E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT)
+ E

@@
@@
- PAGE_CACHE_SHIFT
+ PAGE_SHIFT

@@
@@
- PAGE_CACHE_SIZE
+ PAGE_SIZE

@@
@@
- PAGE_CACHE_MASK
+ PAGE_MASK

@@
expression E;
@@
- PAGE_CACHE_ALIGN(E)
+ PAGE_ALIGN(E)

@@
expression E;
@@
- page_cache_get(E)
+ get_page(E)

@@
expression E;
@@
- page_cache_release(E)
+ put_page(E)
Signed-off-by: NKirill A. Shutemov <kirill.shutemov@linux.intel.com>
Acked-by: NMichal Hocko <mhocko@suse.com>
Signed-off-by: NLinus Torvalds <torvalds@linux-foundation.org>

We get this right for queue_discard_max_show but not max_hw_show. Follow the
same pattern as queue_discard_max_show instead so that we don't truncate.
Signed-off-by: NAlan Cox <alan@linux.intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Commit 4f258a46 ("sd: Fix maximum I/O size for BLOCK_PC requests")
had the unfortunate side-effect of removing an implicit clamp to
BLK_DEF_MAX_SECTORS for REQ_TYPE_FS requests in the block layer
code. This caused problems for some SMR drives.

Debugging this issue revealed a few problems with the existing
infrastructure since the block layer didn't know how to deal with
device-imposed limits, only limits set by the I/O controller.

 - Introduce a new queue limit, max_dev_sectors, which is used by the
   ULD to signal the maximum sectors for a REQ_TYPE_FS request.

 - Ensure that max_dev_sectors is correctly stacked and taken into
   account when overriding max_sectors through sysfs.

 - Rework sd_read_block_limits() so it saves the max_xfer and opt_xfer
   values for later processing.

 - In sd_revalidate() set the queue's max_dev_sectors based on the
   MAXIMUM TRANSFER LENGTH value in the Block Limits VPD. If this value
   is not reported, fall back to a cap based on the CDB TRANSFER LENGTH
   field size.

 - In sd_revalidate(), use OPTIMAL TRANSFER LENGTH from the Block Limits
   VPD--if reported and sane--to signal the preferred device transfer
   size for FS requests. Otherwise use BLK_DEF_MAX_SECTORS.

 - blk_limits_max_hw_sectors() is no longer used and can be removed.
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>
Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=93581Reviewed-by: NChristoph Hellwig <hch@lst.de>
Tested-by: sweeneygj@gmx.com
Tested-by: NArzeets <anatol.pomozov@gmail.com>
Tested-by: NDavid Eisner <david.eisner@oriel.oxon.org>
Tested-by: NMario Kicherer <dev@kicherer.org>
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>

Add basic support for polling for specific IO to complete. This uses
the cookie that blk-mq passes back, which enables the block layer
to pass this cookie to the driver to spin for a specific request.

This will be combined with request latency tracking, so we can make
qualified decisions about when to poll and when not to. For now, for
benchmark purposes, we add a sysfs file that controls whether polling
is enabled or not.
Signed-off-by: NJens Axboe <axboe@fb.com>
Acked-by: NChristoph Hellwig <hch@lst.de>
Acked-by: NKeith Busch <keith.busch@intel.com>

Allow pmem, and other synchronous/bio-based block drivers, to fallback
on a per-cpu reference count managed by the core for tracking queue
live/dead state.

The existing per-cpu reference count for the blk_mq case is promoted to
be used in all block i/o scenarios.  This involves initializing it by
default, waiting for it to drop to zero at exit, and holding a live
reference over the invocation of q->make_request_fn() in
generic_make_request().  The blk_mq code continues to take its own
reference per blk_mq request and retains the ability to freeze the
queue, but the check that the queue is frozen is moved to
generic_make_request().

This fixes crash signatures like the following:

 BUG: unable to handle kernel paging request at ffff880140000000
 [..]
 Call Trace:
  [<ffffffff8145e8bf>] ? copy_user_handle_tail+0x5f/0x70
  [<ffffffffa004e1e0>] pmem_do_bvec.isra.11+0x70/0xf0 [nd_pmem]
  [<ffffffffa004e331>] pmem_make_request+0xd1/0x200 [nd_pmem]
  [<ffffffff811c3162>] ? mempool_alloc+0x72/0x1a0
  [<ffffffff8141f8b6>] generic_make_request+0xd6/0x110
  [<ffffffff8141f966>] submit_bio+0x76/0x170
  [<ffffffff81286dff>] submit_bh_wbc+0x12f/0x160
  [<ffffffff81286e62>] submit_bh+0x12/0x20
  [<ffffffff813395bd>] jbd2_write_superblock+0x8d/0x170
  [<ffffffff8133974d>] jbd2_mark_journal_empty+0x5d/0x90
  [<ffffffff813399cb>] jbd2_journal_destroy+0x24b/0x270
  [<ffffffff810bc4ca>] ? put_pwq_unlocked+0x2a/0x30
  [<ffffffff810bc6f5>] ? destroy_workqueue+0x225/0x250
  [<ffffffff81303494>] ext4_put_super+0x64/0x360
  [<ffffffff8124ab1a>] generic_shutdown_super+0x6a/0xf0

Cc: Jens Axboe <axboe@kernel.dk>
Cc: Keith Busch <keith.busch@intel.com>
Cc: Ross Zwisler <ross.zwisler@linux.intel.com>
Suggested-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Tested-by: NRoss Zwisler <ross.zwisler@linux.intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

bdi's are initialized in two steps, bdi_init() and bdi_register(), but
destroyed in a single step by bdi_destroy() which, for a bdi embedded
in a request_queue, is called during blk_cleanup_queue() which makes
the queue invisible and starts the draining of remaining usages.

A request_queue's user can access the congestion state of the embedded
bdi as long as it holds a reference to the queue.  As such, it may
access the congested state of a queue which finished
blk_cleanup_queue() but hasn't reached blk_release_queue() yet.
Because the congested state was embedded in backing_dev_info which in
turn is embedded in request_queue, accessing the congested state after
bdi_destroy() was called was fine.  The bdi was destroyed but the
memory region for the congested state remained accessible till the
queue got released.

a13f35e8 ("writeback: don't embed root bdi_writeback_congested in
bdi_writeback") changed the situation.  Now, the root congested state
which is expected to be pinned while request_queue remains accessible
is separately reference counted and the base ref is put during
bdi_destroy().  This means that the root congested state may go away
prematurely while the queue is between bdi_dstroy() and
blk_cleanup_queue(), which was detected by Andrey's KASAN tests.

The root cause of this problem is that bdi doesn't distinguish the two
steps of destruction, unregistration and release, and now the root
congested state actually requires a separate release step.  To fix the
issue, this patch separates out bdi_unregister() and bdi_exit() from
bdi_destroy().  bdi_unregister() is called from blk_cleanup_queue()
and bdi_exit() from blk_release_queue().  bdi_destroy() is now just a
simple wrapper calling the two steps back-to-back.

While at it, the prototype of bdi_destroy() is moved right below
bdi_setup_and_register() so that the counterpart operations are
located together.
Signed-off-by: NTejun Heo <tj@kernel.org>
Fixes: a13f35e8 ("writeback: don't embed root bdi_writeback_congested in bdi_writeback")
Cc: stable@vger.kernel.org # v4.2+
Reported-and-tested-by: NAndrey Konovalov <andreyknvl@google.com>
Link: http://lkml.kernel.org/g/CAAeHK+zUJ74Zn17=rOyxacHU18SgCfC6bsYW=6kCY5GXJBwGfQ@mail.gmail.comReviewed-by: NJan Kara <jack@suse.com>
Reviewed-by: NJeff Moyer <jmoyer@redhat.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

The way the block layer is currently written, it goes to great lengths
to avoid having to split bios; upper layer code (such as bio_add_page())
checks what the underlying device can handle and tries to always create
bios that don't need to be split.

But this approach becomes unwieldy and eventually breaks down with
stacked devices and devices with dynamic limits, and it adds a lot of
complexity. If the block layer could split bios as needed, we could
eliminate a lot of complexity elsewhere - particularly in stacked
drivers. Code that creates bios can then create whatever size bios are
convenient, and more importantly stacked drivers don't have to deal with
both their own bio size limitations and the limitations of the
(potentially multiple) devices underneath them.  In the future this will
let us delete merge_bvec_fn and a bunch of other code.

We do this by adding calls to blk_queue_split() to the various
make_request functions that need it - a few can already handle arbitrary
size bios. Note that we add the call _after_ any call to
blk_queue_bounce(); this means that blk_queue_split() and
blk_recalc_rq_segments() don't need to be concerned with bouncing
affecting segment merging.

Some make_request_fn() callbacks were simple enough to audit and verify
they don't need blk_queue_split() calls. The skipped ones are:

 * nfhd_make_request (arch/m68k/emu/nfblock.c)
 * axon_ram_make_request (arch/powerpc/sysdev/axonram.c)
 * simdisk_make_request (arch/xtensa/platforms/iss/simdisk.c)
 * brd_make_request (ramdisk - drivers/block/brd.c)
 * mtip_submit_request (drivers/block/mtip32xx/mtip32xx.c)
 * loop_make_request
 * null_queue_bio
 * bcache's make_request fns

Some others are almost certainly safe to remove now, but will be left
for future patches.

Cc: Jens Axboe <axboe@kernel.dk>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Ming Lei <ming.lei@canonical.com>
Cc: Neil Brown <neilb@suse.de>
Cc: Alasdair Kergon <agk@redhat.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: dm-devel@redhat.com
Cc: Lars Ellenberg <drbd-dev@lists.linbit.com>
Cc: drbd-user@lists.linbit.com
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Geoff Levand <geoff@infradead.org>
Cc: Jim Paris <jim@jtan.com>
Cc: Philip Kelleher <pjk1939@linux.vnet.ibm.com>
Cc: Minchan Kim <minchan@kernel.org>
Cc: Nitin Gupta <ngupta@vflare.org>
Cc: Oleg Drokin <oleg.drokin@intel.com>
Cc: Andreas Dilger <andreas.dilger@intel.com>
Acked-by: NeilBrown <neilb@suse.de> (for the 'md/md.c' bits)
Acked-by: NMike Snitzer <snitzer@redhat.com>
Reviewed-by: NMartin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: NKent Overstreet <kent.overstreet@gmail.com>
[dpark: skip more mq-based drivers, resolve merge conflicts, etc.]
Signed-off-by: NDongsu Park <dpark@posteo.net>
Signed-off-by: NMing Lin <ming.l@ssi.samsung.com>
Signed-off-by: NJens Axboe <axboe@fb.com>