This adds support for active queue tracking, meaning that the
blk-mq tagging maintains a count of active users of a tag set.
This allows us to maintain a notion of fairness between users,
so that we can distribute the tag depth evenly without starving
some users while allowing others to try unfair deep queues.

If sharing of a tag set is detected, each hardware queue will
track the depth of its own queue. And if this exceeds the total
depth divided by the number of active queues, the user is actively
throttled down.

The active queue count is done lazily to avoid bouncing that data
between submitter and completer. Each hardware queue gets marked
active when it allocates its first tag, and gets marked inactive
when 1) the last tag is cleared, and 2) the queue timeout grace
period has passed.
Signed-off-by: NJens Axboe <axboe@fb.com>

If a requeue event races with a timeout, we can get into the
situation where we attempt to complete a request from the
timeout handler when it's not start anymore. This causes a crash.
So have the timeout handler check that REQ_ATOM_STARTED is still
set on the request - if not, we ignore the event. If this happens,
the request has now been marked as complete. As a consequence, we
need to ensure to clear REQ_ATOM_COMPLETE in blk_mq_start_request(),
as to maintain proper request state.
Signed-off-by: NJens Axboe <axboe@fb.com>

Martin reported that his test system would not boot with
current git, it oopsed with this:

BUG: unable to handle kernel paging request at ffff88046c6c9e80
IP: [<ffffffff812971e0>] blk_queue_start_tag+0x90/0x150
PGD 1ddf067 PUD 1de2067 PMD 47fc7d067 PTE 800000046c6c9060
Oops: 0002 [#1] SMP DEBUG_PAGEALLOC
Modules linked in: sd_mod lpfc(+) scsi_transport_fc scsi_tgt oracleasm
rpcsec_gss_krb5 ipv6 igb dca i2c_algo_bit i2c_core hwmon
CPU: 3 PID: 87 Comm: kworker/u17:1 Not tainted 3.14.0+ #246
Hardware name: Supermicro X9DRX+-F/X9DRX+-F, BIOS 3.00 07/09/2013
Workqueue: events_unbound async_run_entry_fn
task: ffff8802743c2150 ti: ffff880273d02000 task.ti: ffff880273d02000
RIP: 0010:[<ffffffff812971e0>]  [<ffffffff812971e0>]
blk_queue_start_tag+0x90/0x150
RSP: 0018:ffff880273d03a58  EFLAGS: 00010092
RAX: ffff88046c6c9e78 RBX: ffff880077208e78 RCX: 00000000fffc8da6
RDX: 00000000fffc186d RSI: 0000000000000009 RDI: 00000000fffc8d9d
RBP: ffff880273d03a88 R08: 0000000000000001 R09: ffff8800021c2410
R10: 0000000000000005 R11: 0000000000015b30 R12: ffff88046c5bb8a0
R13: ffff88046c5c0890 R14: 000000000000001e R15: 000000000000001e
FS:  0000000000000000(0000) GS:ffff880277b00000(0000)
knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: ffff88046c6c9e80 CR3: 00000000018f6000 CR4: 00000000000407e0
Stack:
 ffff880273d03a98 ffff880474b18800 0000000000000000 ffff880474157000
 ffff88046c5c0890 ffff880077208e78 ffff880273d03ae8 ffffffff813b9e62
 ffff880200000010 ffff880474b18968 ffff880474b18848 ffff88046c5c0cd8
Call Trace:
 [<ffffffff813b9e62>] scsi_request_fn+0xf2/0x510
 [<ffffffff81293167>] __blk_run_queue+0x37/0x50
 [<ffffffff8129ac43>] blk_execute_rq_nowait+0xb3/0x130
 [<ffffffff8129ad24>] blk_execute_rq+0x64/0xf0
 [<ffffffff8108d2b0>] ? bit_waitqueue+0xd0/0xd0
 [<ffffffff813bba35>] scsi_execute+0xe5/0x180
 [<ffffffff813bbe4a>] scsi_execute_req_flags+0x9a/0x110
 [<ffffffffa01b1304>] sd_spinup_disk+0x94/0x460 [sd_mod]
 [<ffffffff81160000>] ? __unmap_hugepage_range+0x200/0x2f0
 [<ffffffffa01b2b9a>] sd_revalidate_disk+0xaa/0x3f0 [sd_mod]
 [<ffffffffa01b2fb8>] sd_probe_async+0xd8/0x200 [sd_mod]
 [<ffffffff8107703f>] async_run_entry_fn+0x3f/0x140
 [<ffffffff8106a1c5>] process_one_work+0x175/0x410
 [<ffffffff8106b373>] worker_thread+0x123/0x400
 [<ffffffff8106b250>] ? manage_workers+0x160/0x160
 [<ffffffff8107104e>] kthread+0xce/0xf0
 [<ffffffff81070f80>] ? kthread_freezable_should_stop+0x70/0x70
 [<ffffffff815f0bac>] ret_from_fork+0x7c/0xb0
 [<ffffffff81070f80>] ? kthread_freezable_should_stop+0x70/0x70
Code: 48 0f ab 11 72 db 48 81 4b 40 00 00 10 00 89 83 08 01 00 00 48 89
df 49 8b 04 24 48 89 1c d0 e8 f7 a8 ff ff 49 8b 85 28 05 00 00 <48> 89
58 08 48 89 03 49 8d 85 28 05 00 00 48 89 43 08 49 89 9d
RIP  [<ffffffff812971e0>] blk_queue_start_tag+0x90/0x150
 RSP <ffff880273d03a58>
CR2: ffff88046c6c9e80

Martin bisected and found this to be the problem patch;

	commit 6d113398
	Author: Jan Kara <jack@suse.cz>
	Date:   Mon Feb 24 16:39:54 2014 +0100

	    block: Stop abusing rq->csd.list in blk-softirq

and the problem was immediately apparent. The patch states that
it is safe to reuse queuelist at completion time, since it is
no longer used. However, that is not true if a device is using
block enabled tagging. If that is the case, then the queuelist
is reused to keep track of busy tags. If a device also ended
up using softirq completions, we'd reuse ->queuelist for the
IPI handling while block tagging was still using it. Boom.

Fix this by adding a new ipi_list list head, and share the
memory used with the request hash table. The hash table is
never used after the request is moved to the dispatch list,
which happens long before any potential completion of the
request. Add a new request bit for this, so we don't have
cases that check rq->hash while it could potentially have
been reused for the IPI completion.
Reported-by: NMartin K. Petersen <martin.petersen@oracle.com>
Tested-by: NBenjamin Herrenschmidt <benh@kernel.crashing.org>
Signed-off-by: NJens Axboe <axboe@fb.com>

request_queue bypassing is used to suppress higher-level function of a
request_queue so that they can be switched, reconfigured and shut
down.  A request_queue does the followings while bypassing.

* bypasses elevator and io_cq association and queues requests directly
  to the FIFO dispatch queue.

* bypasses block cgroup request_list lookup and always uses the root
  request_list.

Once confirmed to be bypassing, specific elevator and block cgroup
policy implementations can assume that nothing is in flight for them
and perform various operations which would be dangerous otherwise.

Such confirmation is acheived by short-circuiting all new requests
directly to the dispatch queue and waiting for all the requests which
were issued before to finish.  Unfortunately, while the request
allocating and draining sides were properly handled, we forgot to
actually plug the request dispatch path.  Even after bypassing mode is
confirmed, if the attached driver tries to fetch a request and the
dispatch queue is empty, __elv_next_request() would invoke the current
elevator's elevator_dispatch_fn() callback.  As all in-flight requests
were drained, the elevator wouldn't contain any request but once
bypass is confirmed we don't even know whether the elevator is even
there.  It might be in the process of being switched and half torn
down.

Frank Mayhar reports that this actually happened while switching
elevators, leading to an oops.

Let's fix it by making __elv_next_request() avoid invoking the
elevator_dispatch_fn() callback if the queue is bypassing.  It already
avoids invoking the callback if the queue is dying.  As a dying queue
is guaranteed to be bypassing, we can simply replace blk_queue_dying()
check with blk_queue_bypass().
Reported-by: NFrank Mayhar <fmayhar@google.com>
References: http://lkml.kernel.org/g/1390319905.20232.38.camel@bobble.lax.corp.google.com
Cc: stable@vger.kernel.org
Tested-by: NFrank Mayhar <fmayhar@google.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Linux currently has two models for block devices:

- The classic request_fn based approach, where drivers use struct
  request units for IO. The block layer provides various helper
  functionalities to let drivers share code, things like tag
  management, timeout handling, queueing, etc.

- The "stacked" approach, where a driver squeezes in between the
  block layer and IO submitter. Since this bypasses the IO stack,
  driver generally have to manage everything themselves.

With drivers being written for new high IOPS devices, the classic
request_fn based driver doesn't work well enough. The design dates
back to when both SMP and high IOPS was rare. It has problems with
scaling to bigger machines, and runs into scaling issues even on
smaller machines when you have IOPS in the hundreds of thousands
per device.

The stacked approach is then most often selected as the model
for the driver. But this means that everybody has to re-invent
everything, and along with that we get all the problems again
that the shared approach solved.

This commit introduces blk-mq, block multi queue support. The
design is centered around per-cpu queues for queueing IO, which
then funnel down into x number of hardware submission queues.
We might have a 1:1 mapping between the two, or it might be
an N:M mapping. That all depends on what the hardware supports.

blk-mq provides various helper functions, which include:

- Scalable support for request tagging. Most devices need to
  be able to uniquely identify a request both in the driver and
  to the hardware. The tagging uses per-cpu caches for freed
  tags, to enable cache hot reuse.

- Timeout handling without tracking request on a per-device
  basis. Basically the driver should be able to get a notification,
  if a request happens to fail.

- Optional support for non 1:1 mappings between issue and
  submission queues. blk-mq can redirect IO completions to the
  desired location.

- Support for per-request payloads. Drivers almost always need
  to associate a request structure with some driver private
  command structure. Drivers can tell blk-mq this at init time,
  and then any request handed to the driver will have the
  required size of memory associated with it.

- Support for merging of IO, and plugging. The stacked model
  gets neither of these. Even for high IOPS devices, merging
  sequential IO reduces per-command overhead and thus
  increases bandwidth.

For now, this is provided as a potential 3rd queueing model, with
the hope being that, as it matures, it can replace both the classic
and stacked model. That would get us back to having just 1 real
model for block devices, leaving the stacked approach to dm/md
devices (as it was originally intended).

Contributions in this patch from the following people:

Shaohua Li <shli@fusionio.com>
Alexander Gordeev <agordeev@redhat.com>
Christoph Hellwig <hch@infradead.org>
Mike Christie <michaelc@cs.wisc.edu>
Matias Bjorling <m@bjorling.me>
Jeff Moyer <jmoyer@redhat.com>
Acked-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Switch elevator to use the new hashtable implementation. This reduces the
amount of generic unrelated code in the elevator.

This also removes the dymanic allocation of the hash table. The size of the table is
constant so there's no point in paying the price of an extra dereference when accessing
it.

This patch depends on d9b482c8 ("hashtable: introduce a small and naive
hashtable") which was merged in v3.6.
Signed-off-by: NSasha Levin <sasha.levin@oracle.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

A block driver may start cleaning up resources needed by its
request_fn as soon as blk_cleanup_queue() finished, so request_fn
must not be invoked after draining finished. This is important
when blk_run_queue() is invoked without any requests in progress.
As an example, if blk_drain_queue() and scsi_run_queue() run in
parallel, blk_drain_queue() may have finished all requests after
scsi_run_queue() has taken a SCSI device off the starved list but
before that last function has had a chance to run the queue.
Signed-off-by: NBart Van Assche <bvanassche@acm.org>
Cc: James Bottomley <JBottomley@Parallels.com>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Chanho Min <chanho.min@lge.com>
Acked-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

QUEUE_FLAG_DEAD is used to indicate that queuing new requests must
stop. After this flag has been set queue draining starts. However,
during the queue draining phase it is still safe to invoke the
queue's request_fn, so QUEUE_FLAG_DYING is a better name for this
flag.

This patch has been generated by running the following command
over the kernel source tree:

git grep -lEw 'blk_queue_dead|QUEUE_FLAG_DEAD' |
    xargs sed -i.tmp -e 's/blk_queue_dead/blk_queue_dying/g'      \
        -e 's/QUEUE_FLAG_DEAD/QUEUE_FLAG_DYING/g';                \
sed -i.tmp -e "s/QUEUE_FLAG_DYING$(printf \\t)*5/QUEUE_FLAG_DYING$(printf \\t)5/g" \
    include/linux/blkdev.h;                                       \
sed -i.tmp -e 's/ DEAD/ DYING/g' -e 's/dead queue/a dying queue/' \
    -e 's/Dead queue/A dying queue/' block/blk-core.c
Signed-off-by: NBart Van Assche <bvanassche@acm.org>
Acked-by: NTejun Heo <tj@kernel.org>
Cc: James Bottomley <JBottomley@Parallels.com>
Cc: Mike Christie <michaelc@cs.wisc.edu>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Chanho Min <chanho.min@lge.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Remove special-casing of non-rw fs style requests (discard). The nomerge
flags are consolidated in blk_types.h, and rq_mergeable() and
bio_mergeable() have been modified to use them.

bio_is_rw() is used in place of bio_has_data() a few places. This is
done to to distinguish true reads and writes from other fs type requests
that carry a payload (e.g. write same).
Signed-off-by: NMartin K. Petersen <martin.petersen@oracle.com>
Acked-by: NMike Snitzer <snitzer@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

__generic_unplug_device() function is removed with commit
7eaceacc, which forgot to
remove the declaration at meantime. Here remove it.
Signed-off-by: NYuanhan Liu <yuanhan.liu@linux.intel.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Request allocation is about to be made per-blkg meaning that there'll
be multiple request lists.

* Make queue full state per request_list.  blk_*queue_full() functions
  are renamed to blk_*rl_full() and takes @rl instead of @q.

* Rename blk_init_free_list() to blk_init_rl() and make it take @rl
  instead of @q.  Also add @gfp_mask parameter.

* Add blk_exit_rl() instead of destroying rl directly from
  blk_release_queue().

* Add request_list->q and make request alloc/free functions -
  blk_free_request(), [__]freed_request(), __get_request() - take @rl
  instead of @q.

This patch doesn't introduce any functional difference.
Signed-off-by: NTejun Heo <tj@kernel.org>
Acked-by: NVivek Goyal <vgoyal@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Make the following interface updates to prepare for future ioc related
changes.

* create_io_context() returning ioc only works for %current because it
  doesn't increment ref on the ioc.  Drop @task parameter from it and
  always assume %current.

* Make create_io_context_slowpath() return 0 or -errno and rename it
  to create_task_io_context().

* Make ioc_create_icq() take @ioc as parameter instead of assuming
  that of %current.  The caller, get_request(), is updated to create
  ioc explicitly and then pass it into ioc_create_icq().
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Currently block core calls directly into blk-throttle for init, drain
and exit.  This patch adds blkcg_{init|drain|exit}_queue() which wraps
the blk-throttle functions.  This is to give more control and
visiblity to blkcg core layer for proper layering.  Further patches
will add logic common to blkcg policies to the functions.

While at it, collapse blk_throtl_release() into blk_throtl_exit().
There's no reason to keep them separate.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Rename and extend elv_queisce_start/end() to
blk_queue_bypass_start/end() which are exported and supports nesting
via @q->bypass_depth.  Also add blk_queue_bypass() to test bypass
state.

This will be further extended and used for blkio_group management.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

blk_rq_merge_ok() is the elevator-neutral part of merge eligibility
test.  blk_try_merge() determines merge direction and expects the
caller to have tested elv_rq_merge_ok() previously.

elv_rq_merge_ok() now wraps blk_rq_merge_ok() and then calls
elv_iosched_allow_merge().  elv_try_merge() is removed and the two
callers are updated to call elv_rq_merge_ok() explicitly followed by
blk_try_merge().  While at it, make rq_merge_ok() functions return
bool.

This is to prepare for plug merge update and doesn't introduce any
behavior change.

This is based on Jens' patch to skip elevator_allow_merge_fn() from
plug merge.
Signed-off-by: NTejun Heo <tj@kernel.org>
LKML-Reference: <4F16F3CA.90904@kernel.dk>
Original-patch-by: NJens Axboe <axboe@kernel.dk>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

The block layer has some code trying to determine if two CPUs share a
cache, the scheduler has a similar function. Expose the function used
by the scheduler and make the block layer use it, thereby removing the
block layers usage of CONFIG_SCHED* and topology bits.
Signed-off-by: NPeter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: NJens Axboe <axboe@kernel.dk>
Link: http://lkml.kernel.org/r/1327579450.2446.95.camel@twins

Now block layer knows everything necessary to create and associate
icq's with requests.  Move ioc_create_icq() to blk-ioc.c and update
get_request() such that, if elevator_type->icq_size is set, requests
are automatically associated with their matching icq's before
elv_set_request().  io_context reference is also managed by block core
on request alloc/free.

* Only ioprio/cgroup changed handling remains from cfq_get_cic().
  Collapsed into cfq_set_request().

* This removes queue kicking on icq allocation failure (for now).  As
  icq allocation failure is rare and the only effect of queue kicking
  achieved was possibily accelerating queue processing, this change
  shouldn't be noticeable.

  There is a larger underlying problem.  Unlike request allocation,
  icq allocation is not guaranteed to succeed eventually after
  retries.  The number of icq is unbound and thus mempool can't be the
  solution either.  This effectively adds allocation dependency on
  memory free path and thus possibility of deadlock.

  This usually wouldn't happen because icq allocation is not a hot
  path and, even when the condition triggers, it's highly unlikely
  that none of the writeback workers already has icq.

  However, this is still possible especially if elevator is being
  switched under high memory pressure, so we better get it fixed.
  Probably the only solution is just bypassing elevator and appending
  to dispatch queue on any elevator allocation failure.

* Comment added to explain how icq's are managed and synchronized.

This completes cleanup of io_context interface.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

With kmem_cache managed by blk-ioc, io_cq exit/release can be moved to
blk-ioc too.  The odd ->io_cq->exit/release() callbacks are replaced
with elevator_ops->elevator_exit_icq_fn() with unlinking from both ioc
and q, and freeing automatically handled by blk-ioc.  The elevator
operation only need to perform exit operation specific to the elevator
- in cfq's case, exiting the cfqq's.

Also, clearing of io_cq's on q detach is moved to block core and
automatically performed on elevator switch and q release.

Because the q io_cq points to might be freed before RCU callback for
the io_cq runs, blk-ioc code should remember to which cache the io_cq
needs to be freed when the io_cq is released.  New field
io_cq->__rcu_icq_cache is added for this purpose.  As both the new
field and rcu_head are used only after io_cq is released and the
q/ioc_node fields aren't, they are put into unions.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Now that all io_cq related data structures are in block core layer,
io_cq lookup can be moved from cfq-iosched.c to blk-ioc.c.

Lookup logic from cfq_cic_lookup() is moved to ioc_lookup_icq() with
parameter return type changes (cfqd -> request_queue, cfq_io_cq ->
io_cq) and cfq_cic_lookup() becomes thin wrapper around
cfq_cic_lookup().
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

elevator_queue->ops points to the same ops struct ->elevator_type.ops
is pointing to.  The only effect of caching it in elevator_queue is
shorter notation - it doesn't save any indirect derefence.

Relocate elevator_type->list which used only during module init/exit
to the end of the structure, rename elevator_queue->elevator_type to
->type, and replace elevator_queue->ops with elevator_queue->type.ops.

This doesn't introduce any functional difference.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

When called under queue_lock, current_io_context() triggers lockdep
warning if it hits allocation path.  This is because io_context
installation is protected by task_lock which is not IRQ safe, so it
triggers irq-unsafe-lock -> irq -> irq-safe-lock -> irq-unsafe-lock
deadlock warning.

Given the restriction, accessor + creator rolled into one doesn't work
too well.  Drop current_io_context() and let the users access
task->io_context directly inside queue_lock combined with explicit
creation using create_io_context().

Future ioc updates will further consolidate ioc access and the create
interface will be unexported.

While at it, relocate ioc internal interface declarations in blk.h and
add section comments before and after.

This patch does not introduce functional change.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

* blk_get_queue() is peculiar in that it returns 0 on success and 1 on
  failure instead of 0 / -errno or boolean.  Update it such that it
  returns %true on success and %false on failure.

* Make sure the caller checks for the return value.

* Separate out __blk_get_queue() which doesn't check whether @q is
  dead and put it in blk.h.  This will be used later.

This patch doesn't introduce any functional changes.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Ignoring copy_io() during fork, io_context can be allocated from two
places - current_io_context() and set_task_ioprio().  The former is
always called from local task while the latter can be called from
different task.  The synchornization between them are peculiar and
dubious.

* current_io_context() doesn't grab task_lock() and assumes that if it
  saw %NULL ->io_context, it would stay that way until allocation and
  assignment is complete.  It has smp_wmb() between alloc/init and
  assignment.

* set_task_ioprio() grabs task_lock() for assignment and does
  smp_read_barrier_depends() between "ioc = task->io_context" and "if
  (ioc)".  Unfortunately, this doesn't achieve anything - the latter
  is not a dependent load of the former.  ie, if ioc itself were being
  dereferenced "ioc->xxx", it would mean something (not sure what tho)
  but as the code currently stands, the dependent read barrier is
  noop.

As only one of the the two test-assignment sequences is task_lock()
protected, the task_lock() can't do much about race between the two.
Nothing prevents current_io_context() and set_task_ioprio() allocating
its own ioc for the same task and overwriting the other's.

Also, set_task_ioprio() can race with exiting task and create a new
ioc after exit_io_context() is finished.

ioc get/put doesn't have any reason to be complex.  The only hot path
is accessing the existing ioc of %current, which is simple to achieve
given that ->io_context is never destroyed as long as the task is
alive.  All other paths can happily go through task_lock() like all
other task sub structures without impacting anything.

This patch updates ioc get/put so that it becomes more conventional.

* alloc_io_context() is replaced with get_task_io_context().  This is
  the only interface which can acquire access to ioc of another task.
  On return, the caller has an explicit reference to the object which
  should be put using put_io_context() afterwards.

* The functionality of current_io_context() remains the same but when
  creating a new ioc, it shares the code path with
  get_task_io_context() and always goes through task_lock().

* get_io_context() now means incrementing ref on an ioc which the
  caller already has access to (be that an explicit refcnt or implicit
  %current one).

* PF_EXITING inhibits creation of new io_context and once
  exit_io_context() is finished, it's guaranteed that both ioc
  acquisition functions return %NULL.

* All users are updated.  Most are trivial but
  smp_read_barrier_depends() removal from cfq_get_io_context() needs a
  bit of explanation.  I suppose the original intention was to ensure
  ioc->ioprio is visible when set_task_ioprio() allocates new
  io_context and installs it; however, this wouldn't have worked
  because set_task_ioprio() doesn't have wmb between init and install.
  There are other problems with this which will be fixed in another
  patch.

* While at it, use NUMA_NO_NODE instead of -1 for wildcard node
  specification.

-v2: Vivek spotted contamination from debug patch.  Removed.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

cfq allocates per-queue id using ida and uses it to index cic radix
tree from io_context.  Move it to q->id and allocate on queue init and
free on queue release.  This simplifies cfq a bit and will allow for
further improvements of io context life-cycle management.

This patch doesn't introduce any functional difference.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

There are a number of QUEUE_FLAG_DEAD tests.  Add blk_queue_dead()
macro and use it.

This patch doesn't introduce any functional difference.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

request_queue is refcounted but actually depdends on lifetime
management from the queue owner - on blk_cleanup_queue(), block layer
expects that there's no request passing through request_queue and no
new one will.

This is fundamentally broken.  The queue owner (e.g. SCSI layer)
doesn't have a way to know whether there are other active users before
calling blk_cleanup_queue() and other users (e.g. bsg) don't have any
guarantee that the queue is and would stay valid while it's holding a
reference.

With delay added in blk_queue_bio() before queue_lock is grabbed, the
following oops can be easily triggered when a device is removed with
in-flight IOs.

 sd 0:0:1:0: [sdb] Stopping disk
 ata1.01: disabled
 general protection fault: 0000 [#1] PREEMPT SMP
 CPU 2
 Modules linked in:

 Pid: 648, comm: test_rawio Not tainted 3.1.0-rc3-work+ #56 Bochs Bochs
 RIP: 0010:[<ffffffff8137d651>]  [<ffffffff8137d651>] elv_rqhash_find+0x61/0x100
 ...
 Process test_rawio (pid: 648, threadinfo ffff880019efa000, task ffff880019ef8a80)
 ...
 Call Trace:
  [<ffffffff8137d774>] elv_merge+0x84/0xe0
  [<ffffffff81385b54>] blk_queue_bio+0xf4/0x400
  [<ffffffff813838ea>] generic_make_request+0xca/0x100
  [<ffffffff81383994>] submit_bio+0x74/0x100
  [<ffffffff811c53ec>] dio_bio_submit+0xbc/0xc0
  [<ffffffff811c610e>] __blockdev_direct_IO+0x92e/0xb40
  [<ffffffff811c39f7>] blkdev_direct_IO+0x57/0x60
  [<ffffffff8113b1c5>] generic_file_aio_read+0x6d5/0x760
  [<ffffffff8118c1ca>] do_sync_read+0xda/0x120
  [<ffffffff8118ce55>] vfs_read+0xc5/0x180
  [<ffffffff8118cfaa>] sys_pread64+0x9a/0xb0
  [<ffffffff81afaf6b>] system_call_fastpath+0x16/0x1b

This happens because blk_queue_cleanup() destroys the queue and
elevator whether IOs are in progress or not and DEAD tests are
sprinkled in the request processing path without proper
synchronization.

Similar problem exists for blk-throtl.  On queue cleanup, blk-throtl
is shutdown whether it has requests in it or not.  Depending on
timing, it either oopses or throttled bios are lost putting tasks
which are waiting for bio completion into eternal D state.

The way it should work is having the usual clear distinction between
shutdown and release.  Shutdown drains all currently pending requests,
marks the queue dead, and performs partial teardown of the now
unnecessary part of the queue.  Even after shutdown is complete,
reference holders are still allowed to issue requests to the queue
although they will be immmediately failed.  The rest of teardown
happens on release.

This patch makes the following changes to make blk_queue_cleanup()
behave as proper shutdown.

* QUEUE_FLAG_DEAD is now set while holding both q->exit_mutex and
  queue_lock.

* Unsynchronized DEAD check in generic_make_request_checks() removed.
  This couldn't make any meaningful difference as the queue could die
  after the check.

* blk_drain_queue() updated such that it can drain all requests and is
  now called during cleanup.

* blk_throtl updated such that it checks DEAD on grabbing queue_lock,
  drains all throttled bios during cleanup and free td when queue is
  released.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

blk_throtl_bio() and throtl_get_tg() have rather unusual interface.

* throtl_get_tg() returns pointer to a valid tg or ERR_PTR(-ENODEV),
  and drops queue_lock in the latter case.  Different locking context
  depending on return value is error-prone and DEAD state is scheduled
  to be protected by queue_lock anyway.  Move DEAD check inside
  queue_lock and return valid tg or NULL.

* blk_throtl_bio() indicates return status both with its return value
  and in/out param **@bio.  The former is used to indicate whether
  queue is found to be dead during throtl processing.  The latter
  whether the bio is throttled.

  There's no point in returning DEAD check result from
  blk_throtl_bio().  The queue can die after blk_throtl_bio() is
  finished but before make_request_fn() grabs queue lock.

  Make it take *@bio instead and return boolean result indicating
  whether the request is throttled or not.

This patch doesn't cause any visible functional difference.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Reorganize queue draining related code in preparation of queue exit
changes.

* Factor out actual draining from elv_quiesce_start() to
  blk_drain_queue().

* Make elv_quiesce_start/end() responsible for their own locking.

* Replace open-coded ELVSWITCH clearing in elevator_switch() with
  elv_quiesce_end().

This patch doesn't cause any visible functional difference.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

blk_throtl interface is block internal and there's no reason to have
them in linux/blkdev.h.  Move them to block/blk.h.

This patch doesn't introduce any functional change.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Commit ae1b1539, block: reimplement
FLUSH/FUA to support merge, introduced a performance regression when
running any sort of fsyncing workload using dm-multipath and certain
storage (in our case, an HP EVA).  The test I ran was fs_mark, and it
dropped from ~800 files/sec on ext4 to ~100 files/sec.  It turns out
that dm-multipath always advertised flush+fua support, and passed
commands on down the stack, where those flags used to get stripped off.
The above commit changed that behavior:

static inline struct request *__elv_next_request(struct request_queue *q)
{
        struct request *rq;

        while (1) {
-               while (!list_empty(&q->queue_head)) {
+               if (!list_empty(&q->queue_head)) {
                        rq = list_entry_rq(q->queue_head.next);
-                       if (!(rq->cmd_flags & (REQ_FLUSH | REQ_FUA)) ||
-                           (rq->cmd_flags & REQ_FLUSH_SEQ))
-                               return rq;
-                       rq = blk_do_flush(q, rq);
-                       if (rq)
-                               return rq;
+                       return rq;
                }

Note that previously, a command would come in here, have
REQ_FLUSH|REQ_FUA set, and then get handed off to blk_do_flush:

struct request *blk_do_flush(struct request_queue *q, struct request *rq)
{
        unsigned int fflags = q->flush_flags; /* may change, cache it */
        bool has_flush = fflags & REQ_FLUSH, has_fua = fflags & REQ_FUA;
        bool do_preflush = has_flush && (rq->cmd_flags & REQ_FLUSH);
        bool do_postflush = has_flush && !has_fua && (rq->cmd_flags &
        REQ_FUA);
        unsigned skip = 0;
...
        if (blk_rq_sectors(rq) && !do_preflush && !do_postflush) {
                rq->cmd_flags &= ~REQ_FLUSH;
		if (!has_fua)
			rq->cmd_flags &= ~REQ_FUA;
	        return rq;
	}

So, the flush machinery was bypassed in such cases (q->flush_flags == 0
&& rq->cmd_flags & (REQ_FLUSH|REQ_FUA)).

Now, however, we don't get into the flush machinery at all.  Instead,
__elv_next_request just hands a request with flush and fua bits set to
the scsi_request_fn, even if the underlying request_queue does not
support flush or fua.

The agreed upon approach is to fix the flush machinery to allow
stacking.  While this isn't used in practice (since there is only one
request-based dm target, and that target will now reflect the flush
flags of the underlying device), it does future-proof the solution, and
make it function as designed.

In order to make this work, I had to add a field to the struct request,
inside the flush structure (to store the original req->end_io).  Shaohua
had suggested overloading the union with rb_node and completion_data,
but the completion data is used by device mapper and can also be used by
other drivers.  So, I didn't see a way around the additional field.

I tested this patch on an HP EVA with both ext4 and xfs, and it recovers
the lost performance.  Comments and other testers, as always, are
appreciated.

Cheers,
Jeff
Signed-off-by: NJeff Moyer <jmoyer@redhat.com>
Acked-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>

blk_cleanup_queue() calls elevator_exit() and after this, we can't
touch the elevator without oopsing.  __elv_next_request() must check
for this state because in the refcounted queue model, we can still
call it after blk_cleanup_queue() has been called.

This was reported as causing an oops attributable to scsi.
Signed-off-by: NJames Bottomley <James.Bottomley@suse.de>
Cc: stable@kernel.org
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>

In some drives, flush requests are non-queueable. When flush request is
running, normal read/write requests can't run. If block layer dispatches
such request, driver can't handle it and requeue it.  Tejun suggested we
can hold the queue when flush is running. This can avoid unnecessary
requeue.  Also this can improve performance. For example, we have
request flush1, write1, flush 2. flush1 is dispatched, then queue is
hold, write1 isn't inserted to queue. After flush1 is finished, flush2
will be dispatched. Since disk cache is already clean, flush2 will be
finished very soon, so looks like flush2 is folded to flush1.

In my test, the queue holding completely solves a regression introduced by
commit 53d63e6b:

    block: make the flush insertion use the tail of the dispatch list

    It's not a preempt type request, in fact we have to insert it
    behind requests that do specify INSERT_FRONT.

which causes about 20% regression running a sysbench fileio
workload.

Stable: 2.6.39 only

Cc: stable@kernel.org
Signed-off-by: NShaohua Li <shaohua.li@intel.com>
Acked-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>

We are currently using this flag to check whether it's safe
to call into ->request_fn(). If it is set, we punt to kblockd.
But we get a lot of false positives and excessive punts to
kblockd, which hurts performance.

The only real abuser of this infrastructure is SCSI. So export
the async queue run and convert SCSI over to use that. There's
room for improvement in that SCSI need not always use the async
call, but this fixes our performance issue and they can fix that
up in due time.
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>

Instead of overloading __blk_run_queue to force an offload to kblockd
add a new blk_run_queue_async helper to do it explicitly.  I've kept
the blk_queue_stopped check for now, but I suspect it's not needed
as the check we do when the workqueue items runs should be enough.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>

Fixes generated by 'codespell' and manually reviewed.
Signed-off-by: NLucas De Marchi <lucas.demarchi@profusion.mobi>

One of the disadvantages of on-stack plugging is that we potentially
lose out on merging since all pending IO isn't always visible to
everybody. When we flush the on-stack plugs, right now we don't do
any checks to see if potential merge candidates could be utilized.

Correct this by adding a new insert variant, ELEVATOR_INSERT_SORT_MERGE.
It works just ELEVATOR_INSERT_SORT, but first checks whether we can
merge with an existing request before doing the insertion (if we fail
merging).

This fixes a regression with multiple processes issuing IO that
can be merged.

Thanks to Shaohua Li <shaohua.li@intel.com> for testing and fixing
an accounting bug.
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>

Code has been converted over to the new explicit on-stack plugging,
and delay users have been converted to use the new API for that.
So lets kill off the old plugging along with aops->sync_page().
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>

The current FLUSH/FUA support has evolved from the implementation
which had to perform queue draining.  As such, sequencing is done
queue-wide one flush request after another.  However, with the
draining requirement gone, there's no reason to keep the queue-wide
sequential approach.

This patch reimplements FLUSH/FUA support such that each FLUSH/FUA
request is sequenced individually.  The actual FLUSH execution is
double buffered and whenever a request wants to execute one for either
PRE or POSTFLUSH, it queues on the pending queue.  Once certain
conditions are met, a flush request is issued and on its completion
all pending requests proceed to the next sequence.

This allows arbitrary merging of different type of flushes.  How they
are merged can be primarily controlled and tuned by adjusting the
above said 'conditions' used to determine when to issue the next
flush.

This is inspired by Darrick's patches to merge multiple zero-data
flushes which helps workloads with highly concurrent fsync requests.

* As flush requests are never put on the IO scheduler, request fields
  used for flush share space with rq->rb_node.  rq->completion_data is
  moved out of the union.  This increases the request size by one
  pointer.

  As rq->elevator_private* are used only by the iosched too, it is
  possible to reduce the request size further.  However, to do that,
  we need to modify request allocation path such that iosched data is
  not allocated for flush requests.

* FLUSH/FUA processing happens on insertion now instead of dispatch.

- Comments updated as per Vivek and Mike.
Signed-off-by: NTejun Heo <tj@kernel.org>
Cc: "Darrick J. Wong" <djwong@us.ibm.com>
Cc: Shaohua Li <shli@kernel.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Vivek Goyal <vgoyal@redhat.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>

rq == &q->flush_rq was used to determine whether a rq is part of a
flush sequence, which worked because all requests in a flush sequence
were sequenced using the single dedicated request.  This is about to
change, so introduce REQ_FLUSH_SEQ flag to distinguish flush sequence
requests.

This patch doesn't cause any behavior change.
Signed-off-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>

This reverts commit 7681bfee.

Conflicts:

	include/linux/genhd.h

It has numerous issues with the cleanup path and non-elevator
devices. Revert it for now so we can come up with a clean
version without rushing things.
Signed-off-by: NJens Axboe <jaxboe@fusionio.com>