The hardware's provided queue count may change at runtime with resource
provisioning. This patch allows a block driver to alter the number of
h/w queues available when its resource count changes.

The main part is a new blk-mq API to request a new number of h/w queues
for a given live tag set. The new API freezes all queues using that set,
then adjusts the allocated count prior to remapping these to CPUs.

The bulk of the rest just shifts where h/w contexts and all their
artifacts are allocated and freed.

The number of max h/w contexts is capped to the number of possible cpus
since there is no use for more than that. As such, all pre-allocated
memory for pointers need to account for the max possible rather than
the initial number of queues.

A side effect of this is that the blk-mq will proceed successfully as
long as it can allocate at least one h/w context. Previously it would
fail request queue initialization if less than the requested number
was allocated.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Tested-by: NJon Derrick <jonathan.derrick@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Currently we don't allow sync workload of one cgroup to preempt sync
workload of any other cgroup. This is because we want to achieve service
separation between cgroups. However in cases where cgroup preempting is
ancestor of the current cgroup, there is no need of separation and
idling introduces unnecessary overhead. This hurts for example the case
when workload is isolated within a cgroup but journalling threads are in
root cgroup. Simple way to demostrate the issue is using:

dbench4 -c /usr/share/dbench4/client.txt -t 10 -D /mnt 1

on ext4 filesystem on plain SATA drive (mounted with barrier=0 to make
difference more visible). When all processes are in the root cgroup,
reported throughput is 153.132 MB/sec. When dbench process gets its own
blkio cgroup, reported throughput drops to 26.1006 MB/sec.

Fix the problem by making check in cfq_should_preempt() more benevolent
and allow preemption by ancestor cgroup. This improves the throughput
reported by dbench4 to 48.9106 MB/sec.
Acked-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJan Kara <jack@suse.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

The original idea with preemption of sync noidle queues (introduced in
commit 718eee05 "cfq-iosched: fairness for sync no-idle queues") was
that we service all sync noidle queues together, we don't idle on any of
the queues individually and we idle only if there is no sync noidle
queue to be served. This intention also matches the original test:

	if (cfqd->serving_type == SYNC_NOIDLE_WORKLOAD
	   && new_cfqq->service_tree == cfqq->service_tree)
		return true;

However since at that time cfqq->service_tree was not set for idling
queues, this test was unreliable and was replaced in commit e4a22919
"cfq-iosched: fix no-idle preemption logic" by:

	if (cfqd->serving_type == SYNC_NOIDLE_WORKLOAD &&
	    cfqq_type(new_cfqq) == SYNC_NOIDLE_WORKLOAD &&
	    new_cfqq->service_tree->count == 1)
		return true;

That was a reliable test but was actually doing something different -
now we preempt sync noidle queue only if the new queue is the only one
busy in the service tree.

These days cfq queue is kept in service tree even if it is idling and
thus the original check would be safe again. But since we actually check
that cfq queues are in the same cgroup, of the same priority class and
workload type (sync noidle), we know that new_cfqq is fine to preempt
cfqq. So just remove the service tree check.
Acked-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJan Kara <jack@suse.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Move check for preemption by rt class up. There is no functional change
but it makes arguing about conditions simpler since we can be sure both
cfq queues are from the same ioprio class.
Acked-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJan Kara <jack@suse.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

There is no point in idling on a cfq group if the only cfq queue that is
there has too big thinktime.
Signed-off-by: NJan Kara <jack@suse.com>
Acked-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@fb.com>

After commit e36f6204(block: split bios to maxpossible length),
bio can be splitted in the middle of a vector entry, then it
is easy to split out one bio which size isn't aligned with block
size, especially when the block size is bigger than 512.

This patch fixes the issue by making the max io size aligned
to logical block size.

Fixes: e36f6204(block: split bios to maxpossible length)
Reported-by: NStefan Haberland <sth@linux.vnet.ibm.com>
Cc: Keith Busch <keith.busch@intel.com>
Suggested-by: NLinus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: NMing Lei <tom.leiming@gmail.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

parallel to mutex_{lock,unlock,trylock,is_locked,lock_nested},
inode_foo(inode) being mutex_foo(&inode->i_mutex).

Please, use those for access to ->i_mutex; over the coming cycle
->i_mutex will become rwsem, with ->lookup() done with it held
only shared.
Signed-off-by: NAl Viro <viro@zeniv.linux.org.uk>

This splits bio in the middle of a vector to form the largest possible
bio at the h/w's desired alignment, and guarantees the bio being split
will have some data.

The criteria for splitting is changed from the max sectors to the h/w's
optimal sector alignment if it is provided. For h/w that advertise their
block storage's underlying chunk size, it's a big performance win to not
submit commands that cross them. If sector alignment is not provided,
this patch uses the max sectors as before.

This addresses the performance issue commit d3805611 attempted to
fix, but was reverted due to splitting logic error.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Cc: Jens Axboe <axboe@fb.com>
Cc: Ming Lei <tom.leiming@gmail.com>
Cc: Kent Overstreet <kent.overstreet@gmail.com>
Cc: <stable@vger.kernel.org> # 4.4.x-
Signed-off-by: NJens Axboe <axboe@fb.com>

These actions are completely managed by a block driver or can use the
badblocks api directly.
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

Longer term teach dax to punch "error" holes in mapping requests and
deliver SIGBUS to applications that consume a bad pmem page.  For now,
simply disable the dax performance optimization in the presence of known
errors.
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

Provide a devres interface for initializing a badblocks instance.  The
pmem driver has several scenarios where it will be beneficial to have
this structure automatically freed when the device is disabled / fails
probe.
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

The badblocks list attached to a gendisk is allocated by the driver
which equates to the driver owning the lifetime of the object.  Do not
automatically free it in del_gendisk(). This is in preparation for
expanding the use of badblocks in libnvdimm drivers and introducing
devm_init_badblocks().
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

For symmetry with badblocks_init() make it clear that this path only
destroys incremental allocations of a badblocks instance, and does not
free the badblocks instance itself.
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

NVDIMM devices, which can behave more like DRAM rather than block
devices, may develop bad cache lines, or 'poison'. A block device
exposed by the pmem driver can then consume poison via a read (or
write), and cause a machine check. On platforms without machine
check recovery features, this would mean a crash.

The block device maintaining a runtime list of all known sectors that
have poison can directly avoid this, and also provide a path forward
to enable proper handling/recovery for DAX faults on such a device.

Use the new badblock management interfaces to add a badblocks list to
gendisks.
Signed-off-by: NVishal Verma <vishal.l.verma@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

Take the core badblocks implementation from md, and make it generally
available. This follows the same style as kernel implementations of
linked lists, rb-trees etc, where you can have a structure that can be
embedded anywhere, and accessor functions to manipulate the data.

The only changes in this copy of the code are ones to generalize
function/variable names from md-specific ones. Also add init and free
functions.
Signed-off-by: NVishal Verma <vishal.l.verma@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

When tearing down a block device early in its lifetime, userspace may
still be performing discovery actions like blkdev_ioctl() to re-read
partitions.

The nvdimm_revalidate_disk() implementation depends on
disk->driverfs_dev to be valid at entry.  However, it is set to NULL in
del_gendisk() and fatally this is happening *before* the disk device is
deleted from userspace view.

There's no reason for del_gendisk() to clear ->driverfs_dev.  That
device is the parent of the disk.  It is guaranteed to not be freed
until the disk, as a child, drops its ->parent reference.

We could also fix this issue locally in nvdimm_revalidate_disk() by
using disk_to_dev(disk)->parent, but lets fix it globally since
->driverfs_dev follows the lifetime of the parent.  Longer term we
should probably just add a @parent parameter to add_disk(), and stop
carrying this pointer in the gendisk.

 BUG: unable to handle kernel NULL pointer dereference at           (null)
 IP: [<ffffffffa00340a8>] nvdimm_revalidate_disk+0x18/0x90 [libnvdimm]
 CPU: 2 PID: 538 Comm: systemd-udevd Tainted: G           O    4.4.0-rc5 #2257
 [..]
 Call Trace:
  [<ffffffff8143e5c7>] rescan_partitions+0x87/0x2c0
  [<ffffffff810f37f9>] ? __lock_is_held+0x49/0x70
  [<ffffffff81438c62>] __blkdev_reread_part+0x72/0xb0
  [<ffffffff81438cc5>] blkdev_reread_part+0x25/0x40
  [<ffffffff8143982d>] blkdev_ioctl+0x4fd/0x9c0
  [<ffffffff811246c9>] ? current_kernel_time64+0x69/0xd0
  [<ffffffff812916dd>] block_ioctl+0x3d/0x50
  [<ffffffff81264c38>] do_vfs_ioctl+0x308/0x560
  [<ffffffff8115dbd1>] ? __audit_syscall_entry+0xb1/0x100
  [<ffffffff810031d6>] ? do_audit_syscall_entry+0x66/0x70
  [<ffffffff81264f09>] SyS_ioctl+0x79/0x90
  [<ffffffff81902672>] entry_SYSCALL_64_fastpath+0x12/0x76
Reported-by: NRobert Hu <robert.hu@intel.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

If an application wants exclusive access to all of the persistent memory
provided by an NVDIMM namespace it can use this raw-block-dax facility
to forgo establishing a filesystem.  This capability is targeted
primarily to hypervisors wanting to provision persistent memory for
guests.  It can be disabled / enabled dynamically via the new BLKDAXSET
ioctl.

Cc: Jeff Moyer <jmoyer@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Ross Zwisler <ross.zwisler@linux.intel.com>
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Reviewed-by: NJan Kara <jack@suse.com>
Signed-off-by: NDan Williams <dan.j.williams@intel.com>

This reverts commit d3805611.

If we end up splitting on the first segment, we don't adjust
the sector count. That results in hitting a BUG() with attempting
to split 0 sectors.

As this is just a performance issue and not a regression since
4.3 release, let's just rever this change. That gives us more
time to test a real fix for 4.5, which would be marked for
stable anyway.

We currently only have an inline/sync helper to restart a stopped
queue. If drivers need an async version, they have to roll their
own. Add a generic helper instead.
Signed-off-by: NJens Axboe <axboe@fb.com>

For h/w that advertise their block storage's underlying chunk size, it's
a big performance win to not submit commands that cross them. This patch
uses that criteria if it is provided. If it is not provided, this patch
uses the max sectors as before.
Signed-off-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

blk_queue_bio() does split then bounce, which makes the segment
counting based on pages before bouncing and could go wrong. Move
the split to after bouncing, like we do for blk-mq, and the we
fix the issue of having the bio count for segments be wrong.

Fixes: 54efd50b ("block: make generic_make_request handle arbitrarily sized bios")
Cc: stable@vger.kernel.org
Tested-by: NArtem S. Tashkinov <t.artem@lycos.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

This was added for the 'magic' AEN requests in the NVMe driver that never
return.  We now handle them purely inside the driver and don't need this
core hack any more.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Acked-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Timer context is not very useful for drivers to perform any meaningful abort
action from.  So instead of calling the driver from this useless context
defer it to a workqueue as soon as possible.

Note that while a delayed_work item would seem the right thing here I didn't
dare to use it due to the magic in blk_add_timer that pokes deep into timer
internals.  But maybe this encourages Tejun to add a sensible API for that to
the workqueue API and we'll all be fine in the end :)

Contains a major update from Keith Bush:

"This patch removes synchronizing the timeout work so that the timer can
 start a freeze on its own queue. The timer enters the queue, so timer
 context can only start a freeze, but not wait for frozen."
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Acked-by: NKeith Busch <keith.busch@intel.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

The new name is irq_poll as iopoll is already taken.  Better suggestions
welcome.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NBart Van Assche <bart.vanassche@sandisk.com>

We recently changed bio_integrity_alloc() to return ERR_PTRs instead of
NULL but these calls were missed.

Fixes: 06c1e390 ('blk-integrity: empty implementation when disabled')
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: NJohannes Thumshirn <jthumshirn@suse.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

The routines in scsi_pm.c assume that if a runtime-PM callback is
invoked for a SCSI device, it can only mean that the device's driver
has asked the block layer to handle the runtime power management (by
calling blk_pm_runtime_init(), which among other things sets q->dev).

However, this assumption turns out to be wrong for things like the ses
driver.  Normally ses devices are not allowed to do runtime PM, but
userspace can override this setting.  If this happens, the kernel gets
a NULL pointer dereference when blk_post_runtime_resume() tries to use
the uninitialized q->dev pointer.

This patch fixes the problem by checking q->dev in block layer before
handle runtime PM. Since ses doesn't define any PM callbacks and call
blk_pm_runtime_init(), the crash won't occur.

This fixes Bugzilla #101371.
https://bugzilla.kernel.org/show_bug.cgi?id=101371

More discussion can be found from below link.
http://marc.info/?l=linux-scsi&m=144163730531875&w=2Signed-off-by: NKen Xue <Ken.Xue@amd.com>
Acked-by: NAlan Stern <stern@rowland.harvard.edu>
Cc: Xiangliang Yu <Xiangliang.Yu@amd.com>
Cc: James E.J. Bottomley <JBottomley@odin.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Michael Terry <Michael.terry@canonical.com>
Cc: stable@vger.kernel.org
Signed-off-by: NJens Axboe <axboe@fb.com>

There is a split tracepoint that is supposed to be called when
bio is splitted, and it was called in bio_split function until
commit 4b1faf93 ("block: Kill bio_pair_split()").
But now, no one reports splits, so this patch adds call to
trace_block_split back in blk_queue_split right after split.
Signed-off-by: NMike Krinkin <krinkin.m.u@gmail.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

In architecture like powerpc, we can have cpus without any local memory
attached to it (a.k.a memoryless nodes). In such cases cpu to node mapping
can result in memory allocation hints for block hctx->numa_node populated
with node values which does not have real memory.

Instead use local_memory_node(), which is guaranteed to have memory.
local_memory_node is a noop in other architectures that does not support
memoryless nodes.
Signed-off-by: NRaghavendra K T <raghavendra.kt@linux.vnet.ibm.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

hctx->cpumask is already populated and let the tag cpumask follow that
instead of going through a new for loop.
Signed-off-by: NRaghavendra K T <raghavendra.kt@linux.vnet.ibm.com>
Reviewed-by: NSagi Grimberg <sagig@mellanox.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

This patch moves the blk_integrity_payload definition outside the
CONFIG_BLK_DEV_INTERITY dependency and provides empty function
implementations when the kernel configuration disables integrity
extensions. This simplifies drivers that make use of these to map user
data so they don't need to repeat the same configuration checks.
Signed-off-by: NKeith Busch <keith.busch@intel.com>

Updated by Jens to pass an error pointer return from
bio_integrity_alloc(), otherwise if CONFIG_BLK_DEV_INTEGRITY isn't
set, we return a weird ENOMEM from __nvme_submit_user_cmd()
if a meta buffer is set.
Signed-off-by: NJens Axboe <axboe@fb.com>

Consider the following v2 hierarchy.

  P0 (+memory) --- P1 (-memory) --- A
                                 \- B
       
P0 has memory enabled in its subtree_control while P1 doesn't.  If
both A and B contain processes, they would belong to the memory css of
P1.  Now if memory is enabled on P1's subtree_control, memory csses
should be created on both A and B and A's processes should be moved to
the former and B's processes the latter.  IOW, enabling controllers
can cause atomic migrations into different csses.

The core cgroup migration logic has been updated accordingly but the
controller migration methods haven't and still assume that all tasks
migrate to a single target css; furthermore, the methods were fed the
css in which subtree_control was updated which is the parent of the
target csses.  pids controller depends on the migration methods to
move charges and this made the controller attribute charges to the
wrong csses often triggering the following warning by driving a
counter negative.

 WARNING: CPU: 1 PID: 1 at kernel/cgroup_pids.c:97 pids_cancel.constprop.6+0x31/0x40()
 Modules linked in:
 CPU: 1 PID: 1 Comm: systemd Not tainted 4.4.0-rc1+ #29
 ...
  ffffffff81f65382 ffff88007c043b90 ffffffff81551ffc 0000000000000000
  ffff88007c043bc8 ffffffff810de202 ffff88007a752000 ffff88007a29ab00
  ffff88007c043c80 ffff88007a1d8400 0000000000000001 ffff88007c043bd8
 Call Trace:
  [<ffffffff81551ffc>] dump_stack+0x4e/0x82
  [<ffffffff810de202>] warn_slowpath_common+0x82/0xc0
  [<ffffffff810de2fa>] warn_slowpath_null+0x1a/0x20
  [<ffffffff8118e031>] pids_cancel.constprop.6+0x31/0x40
  [<ffffffff8118e0fd>] pids_can_attach+0x6d/0xf0
  [<ffffffff81188a4c>] cgroup_taskset_migrate+0x6c/0x330
  [<ffffffff81188e05>] cgroup_migrate+0xf5/0x190
  [<ffffffff81189016>] cgroup_attach_task+0x176/0x200
  [<ffffffff8118949d>] __cgroup_procs_write+0x2ad/0x460
  [<ffffffff81189684>] cgroup_procs_write+0x14/0x20
  [<ffffffff811854e5>] cgroup_file_write+0x35/0x1c0
  [<ffffffff812e26f1>] kernfs_fop_write+0x141/0x190
  [<ffffffff81265f88>] __vfs_write+0x28/0xe0
  [<ffffffff812666fc>] vfs_write+0xac/0x1a0
  [<ffffffff81267019>] SyS_write+0x49/0xb0
  [<ffffffff81bcef32>] entry_SYSCALL_64_fastpath+0x12/0x76

This patch fixes the bug by removing @css parameter from the three
migration methods, ->can_attach, ->cancel_attach() and ->attach() and
updating cgroup_taskset iteration helpers also return the destination
css in addition to the task being migrated.  All controllers are
updated accordingly.

* Controllers which don't care whether there are one or multiple
  target csses can be converted trivially.  cpu, io, freezer, perf,
  netclassid and netprio fall in this category.

* cpuset's current implementation assumes that there's single source
  and destination and thus doesn't support v2 hierarchy already.  The
  only change made by this patchset is how that single destination css
  is obtained.

* memory migration path already doesn't do anything on v2.  How the
  single destination css is obtained is updated and the prep stage of
  mem_cgroup_can_attach() is reordered to accomodate the change.

* pids is the only controller which was affected by this bug.  It now
  correctly handles multi-destination migrations and no longer causes
  counter underflow from incorrect accounting.
Signed-off-by: NTejun Heo <tj@kernel.org>
Reported-and-tested-by: NDaniel Wagner <daniel.wagner@bmw-carit.de>
Cc: Aleksa Sarai <cyphar@cyphar.com>

We already have the reserved flag, and a nowait flag awkwardly encoded as
a gfp_t.  Add a real flags argument to make the scheme more extensible and
allow for a nicer calling convention.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@fb.com>

When bio has only one physical segment, we should set bio's
bi_seg_front_size as the real(final) size of the single segment.

Fixes: 02e70742(blk-merge: fix blk_bio_segment_split)
Reported-by: NMarkus Trippelsdorf <markus@trippelsdorf.de>
Tested-by: NMarkus Trippelsdorf <markus@trippelsdorf.de>
Signed-off-by: NMing Lei <ming.lei@canonical.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

When a cloned request is retried on other queues it always needs
to be checked against the queue limits of that queue.
Otherwise the calculations for nr_phys_segments might be wrong,
leading to a crash in scsi_init_sgtable().

To clarify this the patch renames blk_rq_check_limits()
to blk_cloned_rq_check_limits() and removes the symbol
export, as the new function should only be used for
cloned requests and never exported.

Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Ewan Milne <emilne@redhat.com>
Cc: Jeff Moyer <jmoyer@redhat.com>
Signed-off-by: NHannes Reinecke <hare@suse.de>
Fixes: e2a60da7 ("block: Clean up special command handling logic")
Cc: stable@vger.kernel.org # 3.7+
Acked-by: NMike Snitzer <snitzer@redhat.com>
Signed-off-by: NJens Axboe <axboe@fb.com>

Today, blockdev --rereadpt /dev/sda will fail with EBUSY if any
partition of sda is mounted (and will fail with EINVAL if pointed
at a partition).  But it will pass if the entire block device is
formatted with a filesystem and mounted.  I don't think this makes
sense; partitioning should surely not ever change out from under
a mounted device.

So check for bdev->bd_super, and fail that with -EBUSY as well.
Signed-off-by: NEric Sandeen <sandeen@redhat.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
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>

This reverts commit 1b2ff19e.

Jan writes:

--

Thanks for report! After some investigation I found out we allocate
elevator specific data in __get_request() only for non-flush requests. And
this is actually required since the flush machinery uses the space in
struct request for something else. Doh. So my patch is just wrong and not
easy to fix since at the time __get_request() is called we are not sure
whether the flush machinery will be used in the end. Jens, please revert
1b2ff19e. Thanks!

I'm somewhat surprised that you can reliably hit the race where flushing
gets disabled for the device just while the request is in flight. But I
guess during boot it makes some sense.

--

So let's just revert it, we can fix the queue run manually after the
fact. This race is rare enough that it didn't trigger in testing, it
requires the specific disable-while-in-flight scenario to trigger.

This reverts commit 1b2ff19e.

Jan writes:

--

Thanks for report! After some investigation I found out we allocate
elevator specific data in __get_request() only for non-flush requests. And
this is actually required since the flush machinery uses the space in
struct request for something else. Doh. So my patch is just wrong and not
easy to fix since at the time __get_request() is called we are not sure
whether the flush machinery will be used in the end. Jens, please revert
1b2ff19e. Thanks!

I'm somewhat surprised that you can reliably hit the race where flushing
gets disabled for the device just while the request is in flight. But I
guess during boot it makes some sense.

--

So let's just revert it, we can fix the queue run manually after the
fact. This race is rare enough that it didn't trigger in testing, it
requires the specific disable-while-in-flight scenario to trigger.

Just a comment update on not needing queue_lock, and that we aren't
really adding the request to a timeout list for !mq.
Signed-off-by: NJens Axboe <axboe@fb.com>

Use offset_in_page macro instead of (addr & ~PAGE_MASK).
Signed-off-by: NGeliang Tang <geliangtang@163.com>
Signed-off-by: NJens Axboe <axboe@fb.com>