Paolo Valente [Wed, 19 Apr 2017 14:29:02 +0000 (08:29 -0600)]
block, bfq: introduce the BFQ-v0 I/O scheduler as an extra scheduler
We tag as v0 the version of BFQ containing only BFQ's engine plus
hierarchical support. BFQ's engine is introduced by this commit, while
hierarchical support is added by next commit. We use the v0 tag to
distinguish this minimal version of BFQ from the versions containing
also the features and the improvements added by next commits. BFQ-v0
coincides with the version of BFQ submitted a few years ago [1], apart
from the introduction of preemption, described below.
BFQ is a proportional-share I/O scheduler, whose general structure,
plus a lot of code, are borrowed from CFQ.
- Each process doing I/O on a device is associated with a weight and a
(bfq_)queue.
- BFQ grants exclusive access to the device, for a while, to one queue
(process) at a time, and implements this service model by
associating every queue with a budget, measured in number of
sectors.
- After a queue is granted access to the device, the budget of the
queue is decremented, on each request dispatch, by the size of the
request.
- The in-service queue is expired, i.e., its service is suspended,
only if one of the following events occurs: 1) the queue finishes
its budget, 2) the queue empties, 3) a "budget timeout" fires.
- The budget timeout prevents processes doing random I/O from
holding the device for too long and dramatically reducing
throughput.
- Actually, as in CFQ, a queue associated with a process issuing
sync requests may not be expired immediately when it empties. In
contrast, BFQ may idle the device for a short time interval,
giving the process the chance to go on being served if it issues
a new request in time. Device idling typically boosts the
throughput on rotational devices, if processes do synchronous
and sequential I/O. In addition, under BFQ, device idling is
also instrumental in guaranteeing the desired throughput
fraction to processes issuing sync requests (see [2] for
details).
- With respect to idling for service guarantees, if several
processes are competing for the device at the same time, but
all processes (and groups, after the following commit) have
the same weight, then BFQ guarantees the expected throughput
distribution without ever idling the device. Throughput is
thus as high as possible in this common scenario.
- Queues are scheduled according to a variant of WF2Q+, named
B-WF2Q+, and implemented using an augmented rb-tree to preserve an
O(log N) overall complexity. See [2] for more details. B-WF2Q+ is
also ready for hierarchical scheduling. However, for a cleaner
logical breakdown, the code that enables and completes
hierarchical support is provided in the next commit, which focuses
exactly on this feature.
- B-WF2Q+ guarantees a tight deviation with respect to an ideal,
perfectly fair, and smooth service. In particular, B-WF2Q+
guarantees that each queue receives a fraction of the device
throughput proportional to its weight, even if the throughput
fluctuates, and regardless of: the device parameters, the current
workload and the budgets assigned to the queue.
- The last, budget-independence, property (although probably
counterintuitive in the first place) is definitely beneficial, for
the following reasons:
- First, with any proportional-share scheduler, the maximum
deviation with respect to an ideal service is proportional to
the maximum budget (slice) assigned to queues. As a consequence,
BFQ can keep this deviation tight not only because of the
accurate service of B-WF2Q+, but also because BFQ *does not*
need to assign a larger budget to a queue to let the queue
receive a higher fraction of the device throughput.
- Second, BFQ is free to choose, for every process (queue), the
budget that best fits the needs of the process, or best
leverages the I/O pattern of the process. In particular, BFQ
updates queue budgets with a simple feedback-loop algorithm that
allows a high throughput to be achieved, while still providing
tight latency guarantees to time-sensitive applications. When
the in-service queue expires, this algorithm computes the next
budget of the queue so as to:
- Let large budgets be eventually assigned to the queues
associated with I/O-bound applications performing sequential
I/O: in fact, the longer these applications are served once
got access to the device, the higher the throughput is.
- Let small budgets be eventually assigned to the queues
associated with time-sensitive applications (which typically
perform sporadic and short I/O), because, the smaller the
budget assigned to a queue waiting for service is, the sooner
B-WF2Q+ will serve that queue (Subsec 3.3 in [2]).
- Weights can be assigned to processes only indirectly, through I/O
priorities, and according to the relation:
weight = 10 * (IOPRIO_BE_NR - ioprio).
The next patch provides, instead, a cgroups interface through which
weights can be assigned explicitly.
- If several processes are competing for the device at the same time,
but all processes and groups have the same weight, then BFQ
guarantees the expected throughput distribution without ever idling
the device. It uses preemption instead. Throughput is then much
higher in this common scenario.
- ioprio classes are served in strict priority order, i.e.,
lower-priority queues are not served as long as there are
higher-priority queues. Among queues in the same class, the
bandwidth is distributed in proportion to the weight of each
queue. A very thin extra bandwidth is however guaranteed to the Idle
class, to prevent it from starving.
- If the strict_guarantees parameter is set (default: unset), then BFQ
- always performs idling when the in-service queue becomes empty;
- forces the device to serve one I/O request at a time, by
dispatching a new request only if there is no outstanding
request.
In the presence of differentiated weights or I/O-request sizes,
both the above conditions are needed to guarantee that every
queue receives its allotted share of the bandwidth (see
Documentation/block/bfq-iosched.txt for more details). Setting
strict_guarantees may evidently affect throughput.
[2] P. Valente and M. Andreolini, "Improving Application
Responsiveness with the BFQ Disk I/O Scheduler", Proceedings of
the 5th Annual International Systems and Storage Conference
(SYSTOR '12), June 2012.
Slightly extended version:
http://algogroup.unimore.it/people/paolo/disk_sched/bfq-v1-suite-
results.pdf
Josef Bacik [Tue, 18 Apr 2017 20:22:51 +0000 (16:22 -0400)]
nbd: set the max segment size to UINT_MAX
NBD doesn't care about limiting the segment size, let the user push the
largest bio's they want. This allows us to control the request size
solely through max_sectors_kb.
Signed-off-by: Josef Bacik <jbacik@fb.com> Reviewed-by: Ming Lei <ming.lei@redhat.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Marc Olson [Tue, 11 Apr 2017 19:24:09 +0000 (12:24 -0700)]
blkfront: add uevent for size change
When a blkfront device is resized from dom0, emit a KOBJ_CHANGE uevent to
notify the guest about the change. This allows for custom udev rules, such
as automatically resizing a filesystem, when an event occurs.
Josef Bacik [Thu, 6 Apr 2017 21:02:07 +0000 (17:02 -0400)]
nbd: add a flag to destroy an nbd device on disconnect
For ease of management it would be nice for users to specify that the
device node for a nbd device is destroyed once it is disconnected and
there are no more users. Add a client flag and enable this operation to
happen.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:02:06 +0000 (17:02 -0400)]
nbd: add device refcounting
In order to support deleting the device on disconnect we need to
refcount the actual nbd_device struct. So add the refcounting framework
and change how we free the normal devices at rmmod time so we can catch
reference leaks.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:02:05 +0000 (17:02 -0400)]
nbd: add a status netlink command
Allow users to query the status of existing nbd devices. Right now this
only returns whether or not the device is connected, but could be
extended in the future to include more information.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:02:04 +0000 (17:02 -0400)]
nbd: handle dead connections
Sometimes we like to upgrade our server without making all of our
clients freak out and reconnect. This patch provides a way to specify a
dead connection timeout to allow us to pause all requests and wait for
new connections to be opened. With this in place I can take down the
nbd server for less than the dead connection timeout time and bring it
back up and everything resumes gracefully.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:02:03 +0000 (17:02 -0400)]
nbd: only clear the queue on device teardown
When running a disconnect torture test I noticed that sometimes we would
crash with a negative ref count on our queue. This was because we were
ending the same request twice. Turns out we were racing with
NBD_CLEAR_SOCK clearing the requests as well as the teardown of the
device clearing the requests. So instead make the ioctl only shutdown
the sockets and make it so that we only ever run nbd_clear_que from the
device teardown.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:02:02 +0000 (17:02 -0400)]
nbd: multicast dead link notifications
Provide a mechanism to notify userspace that there's been a link problem
on a NBD device. This will allow userspace to re-establish a connection
and provide the new socket to the device without disrupting the device.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:02:01 +0000 (17:02 -0400)]
nbd: add a reconfigure netlink command
We want to be able to reconnect dead connections to existing block
devices, so add a reconfigure netlink command. We will also allow users
to change their timeout on the fly, but everything else will require a
disconnect and reconnect. You won't be able to add more connections
either, simply replace dead connections with new more lively
connections.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:02:00 +0000 (17:02 -0400)]
nbd: add a basic netlink interface
The existing ioctl interface for configuring NBD devices is a bit
cumbersome and hard to extend. The other problem is we leave a
userspace app sitting in it's syscall until the device disconnects,
which is less than ideal.
This patch introduces a netlink interface for adding and disconnecting
nbd devices. This has the benefits of being easily extendable without
breaking older userspace applications, and allows us to configure a nbd
device without leaving a userspace app sitting waiting for the device to
disconnect.
With this interface we also gain the ability to configure more devices
than are preallocated at insmod time. We also have gained the ability
to not specify a particular device and be provided one for us so that
userspace doesn't need to find a free device to configure.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:01:59 +0000 (17:01 -0400)]
nbd: stop using the bdev everywhere
In preparation for the upcoming netlink interface we need to not rely on
already having the bdev for the NBD device we are doing operations on.
Instead of passing the bdev around, just use it in places where we know
we already have the bdev.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:01:58 +0000 (17:01 -0400)]
nbd: separate out the config information
In order to properly refcount the various aspects of a NBD device we
need to separate out the configuration elements of the nbd device. The
configuration of a NBD device has a different lifetime from the actual
device, so it doesn't make sense to bundle these two concepts. Add a
config_refs to keep track of the configuration structure, that way we
can be sure that we never access it when we've torn down the device.
Add a new nbd_config structure to hold all of the transient
configuration information. Finally create this when we open the device
so that it is in place when we start to configure the device. This has
a nice side-effect of fixing a long standing problem where you could end
up with a half-configured nbd device that needed to be "disconnected" in
order to be usable again. Now once we close our device the
configuration will be discarded.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:01:57 +0000 (17:01 -0400)]
nbd: handle single path failures gracefully
Currently if we have multiple connections and one of them goes down we will tear
down the whole device. However there's no reason we need to do this as we
could have other connections that are working fine. Deal with this by keeping
track of the state of the different connections, and if we lose one we mark it
as dead and send all IO destined for that socket to one of the other healthy
sockets. Any outstanding requests that were on the dead socket will timeout and
be re-submitted properly.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Josef Bacik [Thu, 6 Apr 2017 21:01:56 +0000 (17:01 -0400)]
nbd: put socket in error cases
When adding a new socket we look it up and then try to add it to our
configuration. If any of those steps fail we need to make sure we put
the socket so we don't leak them.
Signed-off-by: Josef Bacik <jbacik@fb.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Dan Carpenter [Sat, 15 Apr 2017 18:55:53 +0000 (20:55 +0200)]
lightnvm: fix some error code in pblk-init.c
There were a bunch of places in pblk_lines_init() where we didn't set an
error code. And in pblk_writer_init() we accidentally return 1 instead
of a correct error code, which would result in a Oops later.
Javier González [Sat, 15 Apr 2017 18:55:50 +0000 (20:55 +0200)]
lightnvm: physical block device (pblk) target
This patch introduces pblk, a host-side translation layer for
Open-Channel SSDs to expose them like block devices. The translation
layer allows data placement decisions, and I/O scheduling to be
managed by the host, enabling users to optimize the SSD for their
specific workloads.
An open-channel SSD has a set of LUNs (parallel units) and a
collection of blocks. Each block can be read in any order, but
writes must be sequential. Writes may also fail, and if a block
requires it, must also be reset before new writes can be
applied.
To manage the constraints, pblk maintains a logical to
physical address (L2P) table, write cache, garbage
collection logic, recovery scheme, and logic to rate-limit
user I/Os versus garbage collection I/Os.
The L2P table is fully-associative and manages sectors at a
4KB granularity. Pblk stores the L2P table in two places, in
the out-of-band area of the media and on the last page of a
line. In the cause of a power failure, pblk will perform a
scan to recover the L2P table.
The user data is organized into lines. A line is data
striped across blocks and LUNs. The lines enable the host to
reduce the amount of metadata to maintain besides the user
data and makes it easier to implement RAID or erasure coding
in the future.
pblk implements multi-tenant support and can be instantiated
multiple times on the same drive. Each instance owns a
portion of the SSD - both regarding I/O bandwidth and
capacity - providing I/O isolation for each case.
Finally, pblk also exposes a sysfs interface that allows
user-space to peek into the internals of pblk. The interface
is available at /dev/block/*/pblk/ where * is the block
device name exposed.
This work also contains contributions from:
Matias Bjørling <matias@cnexlabs.com>
Simon A. F. Lund <slund@cnexlabs.com>
Young Tack Jin <youngtack.jin@gmail.com>
Huaicheng Li <huaicheng@cs.uchicago.edu>
Javier González [Sat, 15 Apr 2017 18:55:45 +0000 (20:55 +0200)]
lightnvm: allow to init targets on factory mode
Target initialization has two responsibilities: creating the target
partition and instantiating the target. This patch enables to create a
factory partition (e.g., do not trigger recovery on the given target).
This is useful for target development and for being able to restore the
device state at any moment in time without requiring a full-device
erase.
Javier González [Sat, 15 Apr 2017 18:55:42 +0000 (20:55 +0200)]
lightnvm: double-clear of dev->lun_map on target init error
The dev->lun_map bits are cleared twice if an target init error occurs.
First in the target clean routine, and then next in the nvm_tgt_create
error function. Make sure that it is only cleared once by extending
nvm_remove_tgt_devi() with a clear bit, such that clearing of bits can
ignored when cleaning up a successful initialized target.
The asserts in _nvme_nvm_check_size are not compiled due to the function
not begin called. Make sure that it is called, and also fix the wrong
sizes of asserts for nvme_nvm_addr_format, and nvme_nvm_bb_tbl, which
checked for number of bits instead of bytes.
Javier González [Sat, 15 Apr 2017 18:55:38 +0000 (20:55 +0200)]
lightnvm: rename scrambler controller hint
According to the OCSSD 1.2 specification, the 0x200 hint enables the
media scrambler for the read/write opcode, providing that the controller
has been correctly configured by the firmware. Rename the macro to
represent this meaning.
Javier González [Sat, 15 Apr 2017 18:55:37 +0000 (20:55 +0200)]
lightnvm: submit erases using the I/O path
Until now erases have been submitted as synchronous commands through a
dedicated erase function. In order to enable targets implementing
asynchronous erases, refactor the erase path so that it uses the normal
async I/O submission functions. If a target requires sync I/O, it can
implement it internally. Also, adapt rrpc to use the new erase path.
Scott Bauer [Sat, 15 Apr 2017 18:55:36 +0000 (20:55 +0200)]
nvme/lightnvm: Prevent small buffer overflow in nvme_nvm_identify
There are two closely named structs in lightnvm:
struct nvme_nvm_addr_format and
struct nvme_addr_format.
The first struct has 4 reserved bytes at the end, the second does not.
(gdb) p sizeof(struct nvme_nvm_addr_format)
$1 = 16
(gdb) p sizeof(struct nvm_addr_format)
$2 = 12
In the nvme_nvm_identify function we memcpy from the larger struct to the
smaller struct. We incorrectly pass the length of the larger struct
and overflow by 4 bytes, lets not do that.
The Kyber I/O scheduler is an I/O scheduler for fast devices designed to
scale to multiple queues. Users configure only two knobs, the target
read and synchronous write latencies, and the scheduler tunes itself to
achieve that latency goal.
The implementation is based on "tokens", built on top of the scalable
bitmap library. Tokens serve as a mechanism for limiting requests. There
are two tiers of tokens: queueing tokens and dispatch tokens.
A queueing token is required to allocate a request. In fact, these
tokens are actually the blk-mq internal scheduler tags, but the
scheduler manages the allocation directly in order to implement its
policy.
Dispatch tokens are device-wide and split up into two scheduling
domains: reads vs. writes. Each hardware queue dispatches batches
round-robin between the scheduling domains as long as tokens are
available for that domain.
These tokens can be used as the mechanism to enable various policies.
The policy Kyber uses is inspired by active queue management techniques
for network routing, similar to blk-wbt. The scheduler monitors
latencies and scales the number of dispatch tokens accordingly. Queueing
tokens are used to prevent starvation of synchronous requests by
asynchronous requests.
Various extensions are possible, including better heuristics and ionice
support. The new scheduler isn't set as the default yet.
blk-mq-sched: make completed_request() callback more useful
Currently, this callback is called right after put_request() and has no
distinguishable purpose. Instead, let's call it before put_request() as
soon as I/O has completed on the request, before we account it in
blk-stat. With this, Kyber can enable stats when it sees a latency
outlier and make sure the outlier gets accounted.
blk_mq_finish_request() is required for schedulers that define their own
put_request(). blk_mq_run_hw_queue() is required for schedulers that
hold back requests to be run later.
This operation supports the use case of limiting the number of bits that
can be allocated for a given operation. Rather than setting aside some
bits at the end of the bitmap, we can set aside bits in each word of the
bitmap. This means we can keep the allocation hints spread out and
support sbitmap_resize() nicely at the cost of lower granularity for the
allowed depth.
Jan Kara [Tue, 11 Apr 2017 09:29:01 +0000 (11:29 +0200)]
block: Fix list corruption of blk stats callback list
When CFQ calls wbt_disable_default(), it will call
blk_stat_remove_callback() to stop gathering IO statistics for the
purposes of writeback throttling. Later, when request_queue is
unregistered, wbt_exit() will call blk_stat_remove_callback() again
which will try to delete callback from the list again and possibly cause
list corruption.
Fix the problem by making wbt_disable_default() called wbt_exit() which
is properly guarded against being called multiple times.
Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Jens Axboe <axboe@fb.com>
Separating discards and zeroout operations allows us to remove the LBPRZ
block zeroing constraints from discards and honor the device preferences
for UNMAP commands.
If supported by the device, we'll also choose UNMAP over one of the
WRITE SAME variants for discards.
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
scsi: sd: Separate zeroout and discard command choices
Now that zeroout and discards are distinct operations we need to
separate the policy of choosing the appropriate command. Create a
zeroing_mode which can be one of:
write: Zeroout assist not present, use regular WRITE
writesame: Allow WRITE SAME(10/16) with a zeroed payload
writesame_16_unmap: Allow WRITE SAME(16) with UNMAP
writesame_10_unmap: Allow WRITE SAME(10) with UNMAP
The last two are conditional on the device being thin provisioned with
LBPRZ=1 and LBPWS=1 or LBPWS10=1 respectively.
Whether to set the UNMAP bit or not depends on the REQ_NOUNMAP flag. And
if none of the _unmap variants are supported, regular WRITE SAME will be
used if the device supports it.
The zeroout_mode is exported in sysfs and the detected mode for a given
device can be overridden using the string constants above.
With this change in place we can now issue WRITE SAME(16) with UNMAP set
for block zeroing applications that require hard guarantees and
logical_block_size granularity. And at the same time use the UNMAP
command with the device's preferred granulary and alignment for discard
operations.
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Now that we use the proper REQ_OP_WRITE_ZEROES operation everywhere we can
kill this hack.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
drbd: make intelligent use of blkdev_issue_zeroout
drbd always wants its discard wire operations to zero the blocks, so
use blkdev_issue_zeroout with the BLKDEV_ZERO_UNMAP flag instead of
reinventing it poorly.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Now that we have REQ_OP_WRITE_ZEROES implemented for all devices that
support efficient zeroing, we can remove the call to blkdev_issue_discard.
This means we only have two ways of zeroing left and can simplify the
code.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
But now for the real NVMe Write Zeroes yet, just to get rid of the
discard abuse for zeroing. Also rename the quirk flag to be a bit
more self-explanatory.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Try to use a write same with unmap bit variant if the device supports it
and the caller allows for it.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
block_dev: use blkdev_issue_zerout for hole punches
This gets us support for non-discard efficient write of zeroes (e.g. NVMe)
and prepares for removing the discard_zeroes_data flag.
Also remove a pointless discard support check, which is done in
blkdev_issue_discard already.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
This avoids fallbacks to explicit zeroing in (__)blkdev_issue_zeroout if
the caller doesn't want them.
Also clean up the convoluted check for the return condition that this
new flag is added to.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
block: add a REQ_NOUNMAP flag for REQ_OP_WRITE_ZEROES
If this flag is set logical provisioning capable device should
release space for the zeroed blocks if possible, if it is not set
devices should keep the blocks anchored.
Also remove an out of sync kerneldoc comment for a static function
that would have become even more out of data with this change.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
block: add a flags argument to (__)blkdev_issue_zeroout
Turn the existing discard flag into a new BLKDEV_ZERO_UNMAP flag with
similar semantics, but without referring to diѕcard.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
block: stop using blkdev_issue_write_same for zeroing
We'll always use the WRITE ZEROES code for zeroing now.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
It seems like the code currently passes whatever it was using for writes
to WRITE SAME. Just switch it to WRITE ZEROES, although that doesn't
need any payload.
Untested, and confused by the code, maybe someone who understands it
better than me can help..
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
block: implement splitting of REQ_OP_WRITE_ZEROES bios
Copy and past the REQ_OP_WRITE_SAME code to prepare to implementations
that limit the write zeroes size.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Make life easy for implementations that needs to send a data buffer
to the device (e.g. SCSI) by numbering it as a data out command.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Split sd_setup_discard_cmnd into one function per provisioning type. While
this creates some very slight duplication of boilerplate code it keeps the
code modular for additions of new provisioning types, and for reusing the
write same functions for the upcoming scsi implementation of the Write Zeroes
operation.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Reviewed-by: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
blk-mq: Clarify comments in blk_mq_dispatch_rq_list()
The blk_mq_dispatch_rq_list() implementation got modified several
times but the comments in that function were not updated every
time. Since it is nontrivial what is going on, update the comments
in blk_mq_dispatch_rq_list().
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> Cc: Omar Sandoval <osandov@fb.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.de> Signed-off-by: Jens Axboe <axboe@fb.com>
blk-mq: Make it safe to use RCU to iterate over blk_mq_tag_set.tag_list
Since the next patch in this series will use RCU to iterate over
tag_list, make this safe. Add lockdep_assert_held() statements
in functions that iterate over tag_list to make clear that using
list_for_each_entry() instead of list_for_each_entry_rcu() is
fine in these functions.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
blk-mq-sched: provide hooks for initializing hardware queue data
Schedulers need to be informed when a hardware queue is added or removed
at runtime so they can allocate/free per-hardware queue data. So,
replace the blk_mq_sched_init_hctx_data() helper, which only makes sense
at init time, with .init_hctx() and .exit_hctx() hooks.
We've added a considerable amount of fixes for stalls and issues
with the blk-mq scheduling in the 4.11 series since forking
off the for-4.12/block branch. We need to do improvements on
top of that for 4.12, so pull in the previous fixes to make
our lives easier going forward.
blk-mq: Restart a single queue if tag sets are shared
To improve scalability, if hardware queues are shared, restart
a single hardware queue in round-robin fashion. Rename
blk_mq_sched_restart_queues() to reflect the new semantics.
Remove blk_mq_sched_mark_restart_queue() because this function
has no callers. Remove flag QUEUE_FLAG_RESTART because this
patch removes the code that uses this flag.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.com> Signed-off-by: Jens Axboe <axboe@fb.com>
dm rq: Avoid that request processing stalls sporadically
While running the srp-test software I noticed that request
processing stalls sporadically at the beginning of a test, namely
when mkfs is run against a dm-mpath device. Every time when that
happened the following command was sufficient to resume request
processing:
echo run >/sys/kernel/debug/block/dm-0/state
This patch avoids that such request processing stalls occur. The
test I ran is as follows:
while srp-test/run_tests -d -r 30 -t 02-mq; do :; done
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> Cc: Mike Snitzer <snitzer@redhat.com> Cc: dm-devel@redhat.com Signed-off-by: Jens Axboe <axboe@fb.com>
If a .queue_rq() function returns BLK_MQ_RQ_QUEUE_BUSY then the block
driver that implements that function is responsible for rerunning the
hardware queue once requests can be queued again successfully.
commit 52d7f1b5c2f3 ("blk-mq: Avoid that requeueing starts stopped
queues") removed the blk_mq_stop_hw_queue() call from scsi_queue_rq()
for the BLK_MQ_RQ_QUEUE_BUSY case. Hence change all calls to functions
that are intended to rerun a busy queue such that these examine all
hardware queues instead of only stopped queues.
Since no other functions than scsi_internal_device_block() and
scsi_internal_device_unblock() should ever stop or restart a SCSI
queue, change the blk_mq_delay_queue() call into a
blk_mq_delay_run_hw_queue() call.
Fixes: commit 52d7f1b5c2f3 ("blk-mq: Avoid that requeueing starts stopped queues") Fixes: commit 7e79dadce222 ("blk-mq: stop hardware queue in blk_mq_delay_queue()") Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> Cc: Martin K. Petersen <martin.petersen@oracle.com> Cc: James Bottomley <James.Bottomley@HansenPartnership.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.de> Cc: Sagi Grimberg <sagi@grimberg.me> Cc: Long Li <longli@microsoft.com> Cc: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Introduce a function that runs a hardware queue unconditionally
after a delay. Note: there is already a function that stops and
restarts a hardware queue after a delay, namely blk_mq_delay_queue().
This function will be used in the next patch in this series.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> Cc: Christoph Hellwig <hch@lst.de> Cc: Hannes Reinecke <hare@suse.de> Cc: Long Li <longli@microsoft.com> Cc: K. Y. Srinivasan <kys@microsoft.com> Signed-off-by: Jens Axboe <axboe@fb.com>
Currently only dm and md/raid5 bios trigger
trace_block_bio_complete(). Now that we have bio_chain() and
bio_inc_remaining(), it is not possible, in general, for a driver to
know when the bio is really complete. Only bio_endio() knows that.
So move the trace_block_bio_complete() call to bio_endio().
Now trace_block_bio_complete() pairs with trace_block_bio_queue().
Any bio for which a 'queue' event is traced, will subsequently
generate a 'complete' event.
There are a few cases where completion tracing is not wanted.
1/ If blk_update_request() has already generated a completion
trace event at the 'request' level, there is no point generating
one at the bio level too. In this case the bi_sector and bi_size
will have changed, so the bio level event would be wrong
2/ If the bio hasn't actually been queued yet, but is being aborted
early, then a trace event could be confusing. Some filesystems
call bio_endio() but do not want tracing.
3/ The bio_integrity code interposes itself by replacing bi_end_io,
then restoring it and calling bio_endio() again. This would produce
two identical trace events if left like that.
To handle these, we introduce a flag BIO_TRACE_COMPLETION and only
produce the trace event when this is set.
We address point 1 above by clearing the flag in blk_update_request().
We address point 2 above by only setting the flag when
generic_make_request() is called.
We address point 3 above by clearing the flag after generating a
completion event.
When bio_split() is used on a bio, particularly in blk_queue_split(),
there is an extra complication. A new bio is split off the front, and
may be handle directly without going through generic_make_request().
The old bio, which has been advanced, is passed to
generic_make_request(), so it will trigger a trace event a second
time.
Probably the best result when a split happens is to see a single
'queue' event for the whole bio, then multiple 'complete' events - one
for each component. To achieve this was can:
- copy the BIO_TRACE_COMPLETION flag to the new bio in bio_split()
- avoid generating a 'queue' event if BIO_TRACE_COMPLETION is already set.
This way, the split-off bio won't create a queue event, the original
won't either even if it re-submitted to generic_make_request(),
but both will produce completion events, each for their own range.
So if generic_make_request() is called (which generates a QUEUED
event), then bi_endio() will create a single COMPLETE event for each
range that the bio is split into, unless the driver has explicitly
requested it not to.
blk-mq: remap queues when adding/removing hardware queues
blk_mq_update_nr_hw_queues() used to remap hardware queues, which is the
behavior that drivers expect. However, commit 4e68a011428a changed
blk_mq_queue_reinit() to not remap queues for the case of CPU
hotplugging, inadvertently making blk_mq_update_nr_hw_queues() not remap
queues as well. This breaks, for example, NBD's multi-connection mode,
leaving the added hardware queues unused. Fix it by making
blk_mq_update_nr_hw_queues() explicitly remap the queues.
Fixes: 4e68a011428a ("blk-mq: don't redistribute hardware queues on a CPU hotplug event") Reviewed-by: Keith Busch <keith.busch@intel.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Sagi Grimberg <sagi@grimberg.me> Signed-off-by: Omar Sandoval <osandov@fb.com> Signed-off-by: Jens 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.
blk-mq-sched: set up scheduler tags when bringing up new queues
If a new hardware queue is added at runtime, we don't allocate scheduler
tags for it, leading to a crash. This hooks up the scheduler framework
to blk_mq_{init,exit}_hctx() to make sure everything gets properly
initialized/freed.
blk-mq: use the right hctx when getting a driver tag fails
While dispatching requests, if we fail to get a driver tag, we mark the
hardware queue as waiting for a tag and put the requests on a
hctx->dispatch list to be run later when a driver tag is freed. However,
blk_mq_dispatch_rq_list() may dispatch requests from multiple hardware
queues if using a single-queue scheduler with a multiqueue device. If
blk_mq_get_driver_tag() fails, it doesn't update the hardware queue we
are processing. This means we end up using the hardware queue of the
previous request, which may or may not be the same as that of the
current request. If it isn't, the wrong hardware queue will end up
waiting for a tag, and the requests will be on the wrong dispatch list,
leading to a hang.
The fix is twofold:
1. Make sure we save which hardware queue we were trying to get a
request for in blk_mq_get_driver_tag() regardless of whether it
succeeds or not.
2. Make blk_mq_dispatch_rq_list() take a request_queue instead of a
blk_mq_hw_queue to make it clear that it must handle multiple
hardware queues, since I've already messed this up on a couple of
occasions.
This didn't appear in testing with nvme and mq-deadline because nvme has
more driver tags than the default number of scheduler tags. However,
with the blk_mq_update_nr_hw_queues() fix, it showed up with nbd.
block: move timeout field in struct request to pack better
After commit 64c7f1d1572c, we went from 1 to 2 holes in my
test setup. If we move the timeout field a bit, we remove
both of those holes and shrink struct request by 8 bytes.
Reviewed-by: Christoph Hellwig <hch@lst.de> Signed-off-by: Jens Axboe <axboe@fb.com>
->retries is counting the number of times a command is resubmitted, and
be cleared on the first time we see the command. We currently don't do
that for non-PCIe command, which is easily fixed by moving the setup
to common code.
This driver is for pre-IDE hardisk that are only found in PC from the
stoneage of personal computing, and which we don't support elsewhere
in the kernel these days.
It's also been marked broken forever.
Signed-off-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Jens Axboe <axboe@fb.com>
The block layer core sets blk_mq_queue_data.list but no block
drivers read that member. Hence remove it and also the code that
is used to set this member.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com> Reviewed-by: Christoph Hellwig <hch@lst.de> Reviewed-by: Sagi Grimberg <sagi@grimberg.me> Signed-off-by: Jens Axboe <axboe@fb.com>
Jan Kara [Tue, 4 Apr 2017 12:31:30 +0000 (14:31 +0200)]
cfq: Disable writeback throttling by default
Writeback throttling does not play well with CFQ since that also tries
to throttle async writes. As a result async writeback can get starved in
presence of readers. As an example take a benchmark simulating
postgreSQL database running over a standard rotating SATA drive. There
are 16 processes doing random reads from a huge file (2*machine memory),
1 process doing random writes to the huge file and calling fsync once
per 50000 writes and 1 process doing sequential 8k writes to a
relatively small file wrapping around at the end of the file and calling
fsync every 5 writes. Under this load read latency easily exceeds the
target latency of 75 ms (just because there are so many reads happening
against a relatively slow disk) and thus writeback is throttled to a
point where only 1 write request is allowed at a time. Blktrace data
then looks like:
8,0 1 0 8.347751764 0 m N cfq workload slice:40000000
8,0 1 0 8.347755256 0 m N cfq293A / set_active wl_class: 0 wl_type:0
8,0 1 0 8.347784100 0 m N cfq293A / Not idling. st->count:1
8,0 1 3814 8.347763916 5839 UT N [kworker/u9:2] 1
8,0 0 0 8.347777605 0 m N cfq293A / Not idling. st->count:1
8,0 1 0 8.347784100 0 m N cfq293A / Not idling. st->count:1
8,0 3 1596 8.354364057 0 C R 156109528 + 8 (6906954) [0]
8,0 3 0 8.354383193 0 m N cfq6196SN / complete rqnoidle 0
8,0 3 0 8.354386476 0 m N cfq schedule dispatch
8,0 3 0 8.354399397 0 m N cfq293A / Not idling. st->count:1
8,0 3 0 8.354404705 0 m N cfq293A / dispatch_insert
8,0 3 0 8.354409454 0 m N cfq293A / dispatched a request
8,0 3 0 8.354412527 0 m N cfq293A / activate rq, drv=1
8,0 3 1597 8.354414692 0 D W 145961400 + 24 (6718452) [swapper/0]
8,0 3 0 8.354484184 0 m N cfq293A / Not idling. st->count:1
8,0 3 0 8.354487536 0 m N cfq293A / slice expired t=0
8,0 3 0 8.354498013 0 m N / served: vt=5888102466265088 min_vt=5888074869387264
8,0 3 0 8.354502692 0 m N cfq293A / sl_used=6737519 disp=1 charge=6737519 iops=0 sect=24
8,0 3 0 8.354505695 0 m N cfq293A / del_from_rr
...
8,0 0 1810 8.354728768 0 C W 145961400 + 24 (314076) [0]
8,0 0 0 8.354746927 0 m N cfq293A / complete rqnoidle 0
...
8,0 1 3829 8.389886102 5839 G W 145962968 + 24 [kworker/u9:2]
8,0 1 3830 8.389888127 5839 P N [kworker/u9:2]
8,0 1 3831 8.389908102 5839 A W 145978336 + 24 <- (8,4) 44000
8,0 1 3832 8.389910477 5839 Q W 145978336 + 24 [kworker/u9:2]
8,0 1 3833 8.389914248 5839 I W 145962968 + 24 (28146) [kworker/u9:2]
8,0 1 0 8.389919137 0 m N cfq293A / insert_request
8,0 1 0 8.389924305 0 m N cfq293A / add_to_rr
8,0 1 3834 8.389933175 5839 UT N [kworker/u9:2] 1
...
8,0 0 0 9.455290997 0 m N cfq workload slice:40000000
8,0 0 0 9.455294769 0 m N cfq293A / set_active wl_class:0 wl_type:0
8,0 0 0 9.455303499 0 m N cfq293A / fifo=ffff880003166090
8,0 0 0 9.455306851 0 m N cfq293A / dispatch_insert
8,0 0 0 9.455311251 0 m N cfq293A / dispatched a request
8,0 0 0 9.455314324 0 m N cfq293A / activate rq, drv=1
8,0 0 2043 9.455316210 6204 D W 145962968 + 24 (1065401962) [pgioperf]
8,0 0 0 9.455392407 0 m N cfq293A / Not idling. st->count:1
8,0 0 0 9.455395969 0 m N cfq293A / slice expired t=0
8,0 0 0 9.455404210 0 m N / served: vt=5888958194597888 min_vt=5888941810597888
8,0 0 0 9.455410077 0 m N cfq293A / sl_used=4000000 disp=1 charge=4000000 iops=0 sect=24
8,0 0 0 9.455416851 0 m N cfq293A / del_from_rr
...
8,0 0 2045 9.455648515 0 C W 145962968 + 24 (332305) [0]
8,0 0 0 9.455668350 0 m N cfq293A / complete rqnoidle 0
...
8,0 1 4371 9.455710115 5839 G W 145978336 + 24 [kworker/u9:2]
8,0 1 4372 9.455712350 5839 P N [kworker/u9:2]
8,0 1 4373 9.455730159 5839 A W 145986616 + 24 <- (8,4) 52280
8,0 1 4374 9.455732674 5839 Q W 145986616 + 24 [kworker/u9:2]
8,0 1 4375 9.455737563 5839 I W 145978336 + 24 (27448) [kworker/u9:2]
8,0 1 0 9.455742871 0 m N cfq293A / insert_request
8,0 1 0 9.455747550 0 m N cfq293A / add_to_rr
8,0 1 4376 9.455756629 5839 UT N [kworker/u9:2] 1
So we can see a Q event for a write request, then IO is blocked by
writeback throttling and G and I events for the request happen only once
other writeback IO is completed. Thus CFQ always sees only one write
request. When it sees it, it queues the async queue behind all the read
queues and the async queue gets scheduled after about one second. When
it is scheduled, that one request gets dispatched and async queue is
expired as it has no more requests to submit. Overall we submit about
one write request per second.
Although this scheduling is beneficial for read latency, writes are
heavily starved and this causes large delays all over the system (due to
processes blocking on page lock, transaction starts, etc.). When
writeback throttling is disabled, write throughput is about one fifth of
a read throughput which roughly matches readers/writers ratio and
overall the system stalls are much shorter.
Mixing writeback throttling logic with CFQ throttling logic is always a
recipe for surprises as CFQ assumes it sees the big part of the picture
which is not necessarily true when writeback throttling is blocking
requests. So disable writeback throttling logic by default when CFQ is
used as an IO scheduler.
Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Jens Axboe <axboe@fb.com>
Adam Manzanares [Tue, 4 Apr 2017 15:25:14 +0000 (08:25 -0700)]
block: fix inheriting request priority from bio
In 4.10 I introduced a patch that associates the ioc priority with
each request in the block layer. This work was done in the single queue
block layer code. This patch unifies ioc priority to request mapping across
the single/multi queue block layers.
I have tested this patch with the null block device driver with the following
parameters.
projects / mirror_ubuntu-kernels.git / log