This reverts a series committed earlier due to null pointer exception
bug report in [1]. It seems there are edge case interactions that I did
not consider and will need some time to understand what causes the
adverse interactions.

The original series can be found in [2] with a follow up series in [3].

[1] https://www.spinics.net/lists/cgroups/msg20719.html
[2] https://lore.kernel.org/lkml/20180911184137.35897-1-dennisszhou@gmail.com/
[3] https://lore.kernel.org/lkml/20181020185612.51587-1-dennis@kernel.org/

This reverts the following commits:
d459d853, b2c3fa54, 101246ec, b3b9f24f, e2b09899,
f0fcb3ec, c839e7a0, bdc24917, 74b7c02a, 5bf9a1f3,
a7b39b4e, 07b05bcc, 49f4c2dc, 27e6fa99Signed-off-by: NDennis Zhou <dennis@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

bfq defines as asymmetric a scenario where an active entity, say E
(representing either a single bfq_queue or a group of other entities),
has a higher weight than some other entities.  If the entity E does sync
I/O in such a scenario, then bfq plugs the dispatch of the I/O of the
other entities in the following situation: E is in service but
temporarily has no pending I/O request.  In fact, without this plugging,
all the times that E stops being temporarily idle, it may find the
internal queues of the storage device already filled with an
out-of-control number of extra requests, from other entities. So E may
have to wait for the service of these extra requests, before finally
having its own requests served. This may easily break service
guarantees, with E getting less than its fair share of the device
throughput.  Usually, the end result is that E gets the same fraction of
the throughput as the other entities, instead of getting more, according
to its higher weight.

Yet there are two other more subtle cases where E, even if its weight is
actually equal to or even lower than the weight of any other active
entities, may get less than its fair share of the throughput in case the
above I/O plugging is not performed:
1. other entities issue larger requests than E;
2. other entities contain more active child entities than E (or in
   general tend to have more backlog than E).

In the first case, other entities may get more service than E because
they get larger requests, than those of E, served during the temporary
idle periods of E.  In the second case, other entities get more service
because, by having many child entities, they have many requests ready
for dispatching while E is temporarily idle.

This commit addresses this issue by extending the definition of
asymmetric scenario: a scenario is asymmetric when
- active entities representing bfq_queues have differentiated weights,
  as in the original definition
or (inclusive)
- one or more entities representing groups of entities are active.

This broader definition makes sure that I/O plugging will be performed
in all the above cases, provided that there is at least one active
group.  Of course, this definition is very coarse, so it will trigger
I/O plugging also in cases where it is not needed, such as, e.g.,
multiple active entities with just one child each, and all with the same
I/O-request size.  The reason for this coarse definition is just that a
finer-grained definition would be rather heavy to compute.

On the opposite end, even this new definition does not trigger I/O
plugging in all cases where there is no active group, and all bfq_queues
have the same weight.  So, in these cases some unfairness may occur if
there are asymmetries in I/O-request sizes.  We made this choice because
I/O plugging may lower throughput, and probably a user that has not
created any group cares more about throughput than about perfect
fairness.  At any rate, as for possible applications that may care about
service guarantees, bfq already guarantees a high responsiveness and a
low latency to soft real-time applications automatically.
Signed-off-by: NFederico Motta <federico@willer.it>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

The accessor function bio_blkcg either returns the blkcg associated with
the bio or finds one in the current context. This can cause an issue
when trying to associate a bio with a blkcg. Particularly, it's the
third case that is problematic:

	return css_to_blkcg(task_css(current, io_cgrp_id));

As the above may race against task migration and the cgroup exiting, it
is not always ok to take a reference on the blkcg returned from
bio_blkcg.

This patch adds association ahead of calling bio_blkcg rather than
after. This makes association a required and explicit step along the
code paths for calling bio_blkcg. blk_get_rl is modified as well to get
a reference to the blkcg it may use and blk_put_rl will always put the
reference back. Association is also moved above the bio_blkcg call to
ensure it will not return NULL in blk-iolatency.

BFQ and CFQ utilize this flaw, but due to the complexity, I do not want
to address this in this series. I've created a private version of the
function with notes not to use it describing the flaw. Hopefully soon,
that code can be cleaned up.
Signed-off-by: NDennis Zhou <dennisszhou@gmail.com>
Acked-by: NTejun Heo <tj@kernel.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

To reduce latency for interactive and soft real-time applications, bfq
privileges the bfq_queues containing the I/O of these
applications. These privileged queues, referred-to as weight-raised
queues, get a much higher share of the device throughput
w.r.t. non-privileged queues. To preserve this higher share, the I/O
of any non-weight-raised queue must be plugged whenever a sync
weight-raised queue, while being served, remains temporarily empty. To
attain this goal, bfq simply plugs any I/O (from any queue), if a sync
weight-raised queue remains empty while in service.

Unfortunately, this plugging typically lowers throughput with random
I/O, on devices with internal queueing (because it reduces the filling
level of the internal queues of the device).

This commit addresses this issue by restricting the cases where
plugging is performed: if a sync weight-raised queue remains empty
while in service, then I/O plugging is performed only if some of the
active bfq_queues are *not* weight-raised (which is actually the only
circumstance where plugging is needed to preserve the higher share of
the throughput of weight-raised queues). This restriction proved able
to boost throughput in really many use cases needing only maximum
throughput.
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

The Achilles' heel of BFQ is its failing to reach a high throughput
with sync random I/O on flash storage with internal queueing, in case
the processes doing I/O have differentiated weights.

The cause of this failure is as follows. If at least two processes do
sync I/O, and have a different weight from each other, then BFQ plugs
I/O dispatching every time one of these processes, while it is being
served, remains temporarily without pending I/O requests. This
plugging is necessary to guarantee that every process enjoys a
bandwidth proportional to its weight; but it empties the internal
queue(s) of the drive. And this kills throughput with random I/O. So,
if some processes have differentiated weights and do both sync and
random I/O, the end result is a throughput collapse.

This commit tries to counter this problem by injecting the service of
other processes, in a controlled way, while the process in service
happens to have no I/O. This injection is performed only if the medium
is non rotational and performs internal queueing, and the process in
service does random I/O (service injection might be beneficial for
sequential I/O too, we'll work on that).

As an example of the benefits of this commit, on a PLEXTOR PX-256M5S
SSD, and with five processes having differentiated weights and doing
sync random 4KB I/O, this commit makes the throughput with bfq grow by
400%, from 25 to 100MB/s. This higher throughput is 10MB/s lower than
that reached with none. As some less random I/O is added to the mix,
the throughput becomes equal to or higher than that with none.

This commit is a very first attempt to recover throughput without
losing control, and certainly has many limitations. One is, e.g., that
the processes whose service is injected are not chosen so as to
distribute the extra bandwidth they receive in accordance to their
weights. Thus there might be loss of weighted fairness in some
cases. Anyway, this loss concerns extra service, which would not have
been received at all without this commit. Other limitations and issues
will probably show up with usage.
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

When a sync request is dispatched, the queue that contains that
request, and all the ancestor entities of that queue, are charged with
the number of sectors of the request. In constrast, if the request is
async, then the queue and its ancestor entities are charged with the
number of sectors of the request, multiplied by an overcharge
factor. This throttles the bandwidth for async I/O, w.r.t. to sync
I/O, and it is done to counter the tendency of async writes to steal
I/O throughput to reads.

On the opposite end, the lower this parameter, the stabler I/O
control, in the following respect.  The lower this parameter is, the
less the bandwidth enjoyed by a group decreases
- when the group does writes, w.r.t. to when it does reads;
- when other groups do reads, w.r.t. to when they do writes.

The fixes "block, bfq: always update the budget of an entity when
needed" and "block, bfq: readd missing reset of parent-entity service"
improved I/O control in bfq to such an extent that it has been
possible to revise this overcharge factor downwards.  This commit
introduces the resulting, new value.
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

The received-service counter needs to be equal to 0 when an entity is
set in service. Unfortunately, commit "block, bfq: fix service being
wrongly set to zero in case of preemption" mistakenly removed the
resetting of this counter for the parent entities of the bfq_queue
being set in service. This commit fixes this issue by resetting
service for parent entities, directly on the expiration of the
in-service bfq_queue.

Fixes: 9fae8dd5 ("block, bfq: fix service being wrongly set to zero in case of preemption")
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

The actual goal of the function bfq_bfqq_may_idle is to tell whether
it is better to perform device idling (more precisely: I/O-dispatch
plugging) for the input bfq_queue, either to boost throughput or to
preserve service guarantees. This commit improves the name of the
function accordingly.
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

If
- a bfq_queue Q preempts another queue, because one request of Q
arrives in time,
- but, after this preemption, Q is not the queue that is set in service,
then Q->entity.service is set to 0 when Q is eventually set in
service. But Q should have continued receiving service with its old
budget (which is why preemption has occurred) and its old service.

This commit addresses this issue by resetting service on queue real
expiration.
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

For some bfq_queues, BFQ plugs I/O dispatching when the queue becomes
idle, and keeps the plug until a new request of the queue arrives, or
a timeout fires. BFQ does so either to boost throughput or to preserve
service guarantees for the queue.

More precisely, for such a queue, plugging starts when the queue
happens to have either no request enqueued, or no request in flight,
that is, no request already dispatched but not yet completed.

On the opposite end, BFQ may happen to expire a queue with no request
enqueued, without doing any plugging, if the queue still has some
request in flight. Unfortunately, such a premature expiration causes
the queue to lose its chance to enjoy dispatch plugging a moment
later, i.e., when its in-flight requests finally get completed. This
breaks service guarantees for the queue.

This commit prevents BFQ from expiring an empty queue if the latter
still has in-flight requests.
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

To keep I/O throughput high as often as possible, BFQ performs
I/O-dispatch plugging (aka device idling) only when beneficial exactly
for throughput, or when needed for service guarantees (low latency,
fairness). An important case where the latter condition holds is when
the scenario is 'asymmetric' in terms of weights: i.e., when some
bfq_queue or whole group of queues has a higher weight, and thus has
to receive more service, than other queues or groups. Without dispatch
plugging, lower-weight queues/groups may unjustly steal bandwidth to
higher-weight queues/groups.

To detect asymmetric scenarios, BFQ checks some sufficient
conditions. One of these conditions is that active groups have
different weights. BFQ controls this condition by maintaining a
special set of unique weights of active groups
(group_weights_tree). To this purpose, in the function
bfq_active_insert/bfq_active_extract BFQ adds/removes the weight of a
group to/from this set.

Unfortunately, the function bfq_active_extract may happen to be
invoked also for a group that is still active (to preserve the correct
update of the next queue to serve, see comments in function
bfq_no_longer_next_in_service() for details). In this case, removing
the weight of the group makes the set group_weights_tree
inconsistent. Service-guarantee violations follow.

This commit addresses this issue by moving group_weights_tree
insertions from their previous location (in bfq_active_insert) into
the function __bfq_activate_entity, and by moving group_weights_tree
extractions from bfq_active_extract to when the entity that represents
a group remains throughly idle, i.e., with no request either enqueued
or dispatched.
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

BFQ can deem a bfq_queue as soft real-time only if the queue
- periodically becomes completely idle, i.e., empty and with
  no still-outstanding I/O request;
- after becoming idle, gets new I/O only after a special reference
  time soft_rt_next_start.

In this respect, after commit "block, bfq: consider also past I/O in
soft real-time detection", the value of soft_rt_next_start can never
decrease. This causes a problem with the following special updating
case for soft_rt_next_start: to prevent queues that are not completely
idle to be wrongly detected as soft real-time (when they become
non-empty again), soft_rt_next_start is temporarily set to infinity
for empty queues with still outstanding I/O requests. But, if such an
update is actually performed, then, because of the above commit,
soft_rt_next_start will be stuck at infinity forever, and the queue
will have no more chance to be considered soft real-time.

On slow systems, this problem does cause actual soft real-time
applications to be occasionally not detected as such.

This commit addresses this issue by eliminating the pushing of
soft_rt_next_start to infinity, and by changing the way non-empty
queues are prevented from being wrongly detected as soft
real-time. Simply, a queue that becomes non-empty again can now be
detected as soft real-time only if it has no outstanding I/O request.
Signed-off-by: NDavide Sapienza <sapienza.dav@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

The maximum possible duration of the weight-raising period for
interactive applications is limited to 13 seconds, as this is the time
needed to load the largest application that we considered when tuning
weight raising. Unfortunately, in such an evaluation, we did not
consider the case of very slow virtual machines.

For example, on a QEMU/KVM virtual machine
- running in a slow PC;
- with a virtual disk stacked on a slow low-end 5400rpm HDD;
- serving a heavy I/O workload, such as the sequential reading of
several files;
mplayer takes 23 seconds to start, if constantly weight-raised.

To address this issue, this commit conservatively sets the upper limit
for weight-raising duration to 25 seconds.
Signed-off-by: NDavide Sapienza <sapienza.dav@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

BFQ computes the duration of weight raising for interactive
applications automatically, using some reference parameters. In
particular, BFQ uses the best durations (see comments in the code for
how these durations have been assessed) for two classes of systems:
slow and fast ones. Examples of slow systems are old phones or systems
using micro HDDs. Fast systems are all the remaining ones. Using these
parameters, BFQ computes the actual duration of the weight raising,
for the system at hand, as a function of the relative speed of the
system w.r.t. the speed of a reference system, belonging to the same
class of systems as the system at hand.

This slow vs fast differentiation proved to be useful in the past, but
happens to have little meaning with current hardware. Even worse, it
does cause problems in virtual systems, where the speed of the system
can vary frequently, and so widely to just confuse the class-detection
mechanism, and, as we have verified experimentally, to cause BFQ to
compute non-sensical weight-raising durations.

This commit addresses this issue by removing the slow class and the
class-detection mechanism.
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

A description of how weight raising works is missing in BFQ
sources. In addition, the code for handling weight raising is
scattered across a few functions. This makes it rather hard to
understand the mechanism and its rationale. This commits adds such a
description at the beginning of the main source file.
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Since bfq_finish_request() is always called on the request 'next',
after bfq_requests_merged() is finished, and bfq_finish_request()
removes 'next' from its bfq_queue if needed, it isn't necessary to do
such a removal in advance in bfq_merged_requests().

This commit removes such a useless 'next' removal.
Signed-off-by: NFilippo Muzzini <filippo.muzzini@outlook.it>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

The request rq passed to the function bfq_requests_merged is always in
a bfq_queue, so the check !RB_EMPTY_NODE(&rq->rb_node) at the
beginning of bfq_requests_merged always succeeds, and the control
flow systematically skips to the end of the function.  This implies
that the body of the function is never executed, i.e., the
repositioning of rq is never performed.

On the opposite end, a control is missing in the body of the function:
'next' must be removed only if it is inside a bfq_queue.

This commit removes the wrong check on rq, and adds the missing check
on 'next'. In addition, this commit adds comments on
bfq_requests_merged.
Signed-off-by: NFilippo Muzzini <filippo.muzzini@outlook.it>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

In bfq_requests_merged(), there is a deadlock because the lock on
bfqq->bfqd->lock is held by the calling function, but the code of
this function tries to grab the lock again.

This deadlock is currently hidden by another bug (fixed by next commit
for this source file), which causes the body of bfq_requests_merged()
to be never executed.

This commit removes the deadlock by removing the lock/unlock pair.
Signed-off-by: NFilippo Muzzini <filippo.muzzini@outlook.it>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

If our shallow depth is smaller than the wake batching of sbitmap,
we can introduce hangs. Ensure that sbitmap knows how low we'll go.
Acked-by: NPaolo Valente <paolo.valente@linaro.org>
Reviewed-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

bfqd->sb_shift was attempted used as a cache for the sbitmap queue
shift, but we don't need it, as it never changes. Kill it with fire.
Acked-by: NPaolo Valente <paolo.valente@linaro.org>
Reviewed-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

It doesn't change, so don't put it in the per-IO hot path.
Acked-by: NPaolo Valente <paolo.valente@linaro.org>
Reviewed-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Reserved tags are used for error handling, we don't need to
care about them for regular IO. The core won't call us for these
anyway.
Acked-by: NPaolo Valente <paolo.valente@linaro.org>
Reviewed-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

When invoked for an I/O request rq, the prepare_request hook of bfq
increments reference counters in the destination bfq_queue for rq. In
this respect, after this hook has been invoked, rq may still be
transformed into a request with no icq attached, i.e., for bfq, a
request not associated with any bfq_queue. No further hook is invoked
to signal this tranformation to bfq (in general, to the destination
elevator for rq). This leads bfq into an inconsistent state, because
bfq has no chance to correctly lower these counters back. This
inconsistency may in its turn cause incorrect scheduling and hangs. It
certainly causes memory leaks, by making it impossible for bfq to free
the involved bfq_queue.

On the bright side, no transformation can still happen for rq after rq
has been inserted into bfq, or merged with another, already inserted,
request. Exploiting this fact, this commit addresses the above issue
by delaying the preparation of an I/O request to when the request is
inserted or merged.

This change also gives a performance bonus: a lock-contention point
gets removed. To prepare a request, bfq needs to hold its scheduler
lock. After postponing request preparation to insertion or merging, no
lock needs to be grabbed any longer in the prepare_request hook, while
the lock already taken to perform insertion or merging is used to
preparare the request as well.
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Tested-by: NBart Van Assche <bart.vanassche@wdc.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Currently, struct request has four timestamp fields:

- A start time, set at get_request time, in jiffies, used for iostats
- An I/O start time, set at start_request time, in ktime nanoseconds,
  used for blk-stats (i.e., wbt, kyber, hybrid polling)
- Another start time and another I/O start time, used for cfq and bfq

These can all be consolidated into one start time and one I/O start
time, both in ktime nanoseconds, shaving off up to 16 bytes from struct
request depending on the kernel config.
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Even if we don't have an IO context attached to a request, we still
need to clear the priv[0..1] pointers, as they could be pointing
to previously used bic/bfqq structures. If we don't do so, we'll
either corrupt memory on dispatching a request, or cause an
imbalance in counters.

Inspired by a fix from Kees.
Reported-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Reported-by: NKees Cook <keescook@chromium.org>
Cc: stable@vger.kernel.org
Fixes: aee69d78 ("block, bfq: introduce the BFQ-v0 I/O scheduler as an extra scheduler")
Signed-off-by: NJens Axboe <axboe@kernel.dk>

If a storage device handled by BFQ happens to be slower than 7.5 KB/s
for a certain amount of time (in the order of a second), then the
estimated peak rate of the device, maintained in BFQ, becomes equal to
0. The reason is the limited precision with which the rate is
represented (details on the range of representable values in the
comments introduced by this commit). This leads to a division-by-zero
error where the estimated peak rate is used as divisor. Such a type of
failure has been reported in [1].

This commit addresses this issue by:
1. Lower-bounding the estimated peak rate to 1
2. Adding and improving comments on the range of rates representable

[1] https://www.spinics.net/lists/kernel/msg2739205.htmlSigned-off-by: NKonstantin Khlebnikov <khlebnikov@yandex-team.ru>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Commit 'a6a252e6 ("blk-mq-sched: decide how to handle flush rq via
RQF_FLUSH_SEQ")' makes all non-flush re-prepared requests for a device
be re-inserted into the active I/O scheduler for that device. As a
consequence, I/O schedulers may get the same request inserted again,
even several times, without a finish_request invoked on that request
before each re-insertion.

This fact is the cause of the failure reported in [1]. For an I/O
scheduler, every re-insertion of the same re-prepared request is
equivalent to the insertion of a new request. For schedulers like
mq-deadline or kyber, this fact causes no harm. In contrast, it
confuses a stateful scheduler like BFQ, which keeps state for an I/O
request, until the finish_request hook is invoked on the request. In
particular, BFQ may get stuck, waiting forever for the number of
request dispatches, of the same request, to be balanced by an equal
number of request completions (while there will be one completion for
that request). In this state, BFQ may refuse to serve I/O requests
from other bfq_queues. The hang reported in [1] then follows.

However, the above re-prepared requests undergo a requeue, thus the
requeue_request hook of the active elevator is invoked for these
requests, if set. This commit then addresses the above issue by
properly implementing the hook requeue_request in BFQ.

[1] https://marc.info/?l=linux-block&m=151211117608676Reported-by: NIvan Kozik <ivan@ludios.org>
Reported-by: NAlban Browaeys <alban.browaeys@gmail.com>
Tested-by: NMike Galbraith <efault@gmx.de>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NSerena Ziviani <ziviani.serena@gmail.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

To maximise responsiveness, BFQ raises the weight, and performs device
idling, for bfq_queues associated with processes deemed as
interactive. In particular, weight raising has a maximum duration,
equal to the time needed to start a large application. If a
weight-raised process goes on doing I/O beyond this maximum duration,
it loses weight-raising.

This mechanism is evidently vulnerable to the following false
positives: I/O-bound applications that will go on doing I/O for much
longer than the duration of weight-raising. These applications have
basically no benefit from being weight-raised at the beginning of
their I/O. On the opposite end, while being weight-raised, these
applications
a) unjustly steal throughput to applications that may truly need
low latency;
b) make BFQ uselessly perform device idling; device idling results
in loss of device throughput with most flash-based storage, and may
increase latencies when used purposelessly.

This commit adds a countermeasure to reduce both the above
problems. To introduce this countermeasure, we provide the following
extra piece of information (full details in the comments added by this
commit). During the start-up of the large application used as a
reference to set the duration of weight-raising, involved processes
transfer at most ~110K sectors each. Accordingly, a process initially
deemed as interactive has no right to be weight-raised any longer,
once transferred 110K sectors or more.

Basing on this consideration, this commit early-ends weight-raising
for a bfq_queue if the latter happens to have received an amount of
service at least equal to 110K sectors (actually, a little bit more,
to keep a safety margin). I/O-bound applications that reach a high
throughput, such as file copy, get to this threshold much before the
allowed weight-raising period finishes. Thus this early ending of
weight-raising reduces the amount of time during which these
applications cause the problems described above.
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Asynchronous I/O can easily starve synchronous I/O (both sync reads
and sync writes), by consuming all request tags. Similarly, storms of
synchronous writes, such as those that sync(2) may trigger, can starve
synchronous reads. In their turn, these two problems may also cause
BFQ to loose control on latency for interactive and soft real-time
applications. For example, on a PLEXTOR PX-256M5S SSD, LibreOffice
Writer takes 0.6 seconds to start if the device is idle, but it takes
more than 45 seconds (!) if there are sequential writes in the
background.

This commit addresses this issue by limiting the maximum percentage of
tags that asynchronous I/O requests and synchronous write requests can
consume. In particular, this commit grants a higher threshold to
synchronous writes, to prevent the latter from being starved by
asynchronous I/O.

According to the above test, LibreOffice Writer now starts in about
1.2 seconds on average, regardless of the background workload, and
apart from some rare outlier. To check this improvement, run, e.g.,
sudo ./comm_startup_lat.sh bfq 5 5 seq 10 "lowriter --terminate_after_init"
for the comm_startup_lat benchmark in the S suite [1].

[1] https://github.com/Algodev-github/STested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Commit '7b9e9361' ("blk-mq-sched: unify request finished methods")
changed the old name of current bfq_finish_request method, but left it
unchanged elsewhere in the code (related comments, part of function
name bfq_put_rq_priv_body).

This commit fixes all occurrences of the old name of this method by
changing them into the current name.

Fixes: 7b9e9361 ("blk-mq-sched: unify request finished methods")
Reviewed-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NFederico Motta <federico@willer.it>
Signed-off-by: NChiara Bruschi <bruschi.chiara@outlook.it>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

It's not available if we don't have group io scheduling set, and
there's no need to call it.

Fixes: 0d52af59 ("block, bfq: release oom-queue ref to root group on exit")
Signed-off-by: NJens Axboe <axboe@kernel.dk>

On scheduler init, a reference to the root group, and a reference to
its corresponding blkg are taken for the oom queue. Yet these
references are not released on scheduler exit, which prevents these
objects from be freed. This commit adds the missing reference
releases.
Reported-by: NDavide Ferrari <davideferrari8@gmail.com>
Tested-by: NHolger Hoffstätte <holger@applied-asynchrony.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Commit a33801e8 ("block, bfq: move debug blkio stats behind
CONFIG_DEBUG_BLK_CGROUP") introduced two batches of confusing ifdefs:
one reported in [1], plus a similar one in another function. This
commit removes both batches, in the way suggested in [1].

[1] https://www.spinics.net/lists/linux-block/msg20043.html

Fixes: a33801e8 ("block, bfq: move debug blkio stats behind CONFIG_DEBUG_BLK_CGROUP")
Reported-by: NLinus Torvalds <torvalds@linux-foundation.org>
Tested-by: NLuca Miccio <lucmiccio@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

BFQ privileges the I/O of soft real-time applications, such as video
players, to guarantee to these application a high bandwidth and a low
latency. In this respect, it is not easy to correctly detect when an
application is soft real-time. A particularly nasty false positive is
that of an I/O-bound application that occasionally happens to meet all
requirements to be deemed as soft real-time. After being detected as
soft real-time, such an application monopolizes the device. Fortunately,
BFQ will realize soon that the application is actually not soft
real-time and suspend every privilege. Yet, the application may happen
again to be wrongly detected as soft real-time, and so on.

As highlighted by our tests, this problem causes BFQ to occasionally
fail to guarantee a high responsiveness, in the presence of heavy
background I/O workloads. The reason is that the background workload
happens to be detected as soft real-time, more or less frequently,
during the execution of the interactive task under test. To give an
idea, because of this problem, Libreoffice Writer occasionally takes 8
seconds, instead of 3, to start up, if there are sequential reads and
writes in the background, on a Kingston SSDNow V300.

This commit addresses this issue by leveraging the following facts.

The reason why some applications are detected as soft real-time despite
all BFQ checks to avoid false positives, is simply that, during high
CPU or storage-device load, I/O-bound applications may happen to do
I/O slowly enough to meet all soft real-time requirements, and pass
all BFQ extra checks. Yet, this happens only for limited time periods:
slow-speed time intervals are usually interspersed between other time
intervals during which these applications do I/O at a very high speed.
To exploit these facts, this commit introduces a little change, in the
detection of soft real-time behavior, to systematically consider also
the recent past: the higher the speed was in the recent past, the
later next I/O should arrive for the application to be considered as
soft real-time. At the beginning of a slow-speed interval, the minimum
arrival time allowed for the next I/O usually happens to still be so
high, to fall *after* the end of the slow-speed period itself. As a
consequence, the application does not risk to be deemed as soft
real-time during the slow-speed interval. Then, during the next
high-speed interval, the application cannot, evidently, be deemed as
soft real-time (exactly because of its speed), and so on.

This extra filtering proved to be rather effective: in the above test,
the frequency of false positives became so low that the start-up time
was 3 seconds in all iterations (apart from occasional outliers,
caused by page-cache-management issues, which are out of the scope of
this commit, and cannot be solved by an I/O scheduler).
Tested-by: NLee Tibbert <lee.tibbert@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NAngelo Ruocco <angeloruocco90@gmail.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

When two or more processes do I/O in a way that the their requests are
sequential in respect to one another, BFQ merges the bfq_queues associated
with the processes. This way the overall I/O pattern becomes sequential,
and thus there is a boost in througput.
These cooperating processes usually start or restart to do I/O shortly
after each other. So, in order to avoid merging non-cooperating processes,
BFQ ensures that none of these queues has been in weight raising for too
long.

In this respect, from commit "block, bfq-sq, bfq-mq: let a queue be merged
only shortly after being created", BFQ checks whether any queue (and not
only weight-raised ones) is doing I/O continuously from too long to be
merged.

This new additional check makes the first one useless: a queue doing
I/O from long enough, if being weight-raised, is also a queue in
weight raising for too long to be merged. Accordingly, this commit
removes the first check.
Signed-off-by: NAngelo Ruocco <angeloruocco90@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

In BFQ and CFQ, two processes are said to be cooperating if they do
I/O in such a way that the union of their I/O requests yields a
sequential I/O pattern. To get such a sequential I/O pattern out of
the non-sequential pattern of each cooperating process, BFQ and CFQ
merge the queues associated with these processes. In more detail,
cooperating processes, and thus their associated queues, usually
start, or restart, to do I/O shortly after each other. This is the
case, e.g., for the I/O threads of KVM/QEMU and of the dump
utility. Basing on this assumption, this commit allows a bfq_queue to
be merged only during a short time interval (100ms) after it starts,
or re-starts, to do I/O.  This filtering provides two important
benefits.

First, it greatly reduces the probability that two non-cooperating
processes have their queues merged by mistake, if they just happen to
do I/O close to each other for a short time interval. These spurious
merges cause loss of service guarantees. A low-weight bfq_queue may
unjustly get more than its expected share of the throughput: if such a
low-weight queue is merged with a high-weight queue, then the I/O for
the low-weight queue is served as if the queue had a high weight. This
may damage other high-weight queues unexpectedly.  For instance,
because of this issue, lxterminal occasionally took 7.5 seconds to
start, instead of 6.5 seconds, when some sequential readers and
writers did I/O in the background on a FUJITSU MHX2300BT HDD.  The
reason is that the bfq_queues associated with some of the readers or
the writers were merged with the high-weight queues of some processes
that had to do some urgent but little I/O. The readers then exploited
the inherited high weight for all or most of their I/O, during the
start-up of terminal. The filtering introduced by this commit
eliminated any outlier caused by spurious queue merges in our start-up
time tests.

This filtering also provides a little boost of the throughput
sustainable by BFQ: 3-4%, depending on the CPU. The reason is that,
once a bfq_queue cannot be merged any longer, this commit makes BFQ
stop updating the data needed to handle merging for the queue.
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NAngelo Ruocco <angeloruocco90@gmail.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

A just-created bfq_queue will certainly be deemed as interactive on
the arrival of its first I/O request, if the low_latency flag is
set. Yet, if the queue is merged with another queue on the arrival of
its first I/O request, it will not have the chance to be flagged as
interactive. Nevertheless, if the queue is then split soon enough, it
has to be flagged as interactive after the split.

To handle this early-merge scenario correctly, BFQ saves the state of
the queue, on the merge, as if the latter had already been deemed
interactive. So, if the queue is split soon, it will get
weight-raised, because the previous state of the queue is resumed on
the split.

Unfortunately, in the act of saving the state of the newly-created
queue, BFQ doesn't check whether the low_latency flag is set, and this
causes early-merged queues to be then weight-raised, on queue splits,
even if low_latency is off. This commit addresses this problem by
adding the missing check.
Signed-off-by: NAngelo Ruocco <angeloruocco90@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

If two processes do I/O close to each other, then BFQ merges the
bfq_queues associated with these processes, to get a more sequential
I/O, and thus a higher throughput.  In this respect, to detect whether
two processes are doing I/O close to each other, BFQ keeps a list of
the head-of-line I/O requests of all active bfq_queues.  The list is
ordered by initial sectors, and implemented through a red-black tree
(rq_pos_tree).

Unfortunately, the update of the rq_pos_tree was incomplete, because
the tree was not updated on the removal of the head-of-line I/O
request of a bfq_queue, in case the queue did not remain empty. This
commit adds the missing update.
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NAngelo Ruocco <angeloruocco90@gmail.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

If two processes do I/O close to each other, i.e., are cooperating
processes in BFQ (and CFQ'S) nomenclature, then BFQ merges their
associated bfq_queues, so as to get sequential I/O from the union of
the I/O requests of the processes, and thus reach a higher
throughput. A merged queue is then split if its I/O stops being
sequential. In this respect, BFQ deems the I/O of a bfq_queue as
(mostly) sequential only if less than 4 I/O requests are random, out
of the last 32 requests inserted into the queue.

Unfortunately, extensive testing (with the interleaved_io benchmark of
the S suite [1], and with real applications spawning cooperating
processes) has clearly shown that, with such a low threshold, only a
rather low I/O throughput may be reached when several cooperating
processes do I/O. In particular, the outcome of each test run was
bimodal: if queue merging occurred and was stable during the test,
then the throughput was close to the peak rate of the storage device,
otherwise the throughput was arbitrarily low (usually around 1/10 of
the peak rate with a rotational device). The probability to get the
unlucky outcomes grew with the number of cooperating processes: it was
already significant with 5 processes, and close to one with 7 or more
processes.

The cause of the low throughput in the unlucky runs was that the
merged queues containing the I/O of these cooperating processes were
soon split, because they contained more random I/O requests than those
tolerated by the 4/32 threshold, but
- that I/O would have however allowed the storage device to reach
  peak throughput or almost peak throughput;
- in contrast, the I/O of these processes, if served individually
  (from separate queues) yielded a rather low throughput.

So we repeated our tests with increasing values of the threshold,
until we found the minimum value (19) for which we obtained maximum
throughput, reliably, with at least up to 9 cooperating
processes. Then we checked that the use of that higher threshold value
did not cause any regression for any other benchmark in the suite [1].
This commit raises the threshold to such a higher value.

[1] https://github.com/Algodev-github/SSigned-off-by: NAngelo Ruocco <angeloruocco90@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

BFQ currently creates, and updates, its own instance of the whole
set of blkio statistics that cfq creates. Yet, from the comments
of Tejun Heo in [1], it turned out that most of these statistics
are meant/useful only for debugging. This commit makes BFQ create
the latter, debugging statistics only if the option
CONFIG_DEBUG_BLK_CGROUP is set.

By doing so, this commit also enables BFQ to enjoy a high perfomance
boost. The reason is that, if CONFIG_DEBUG_BLK_CGROUP is not set, then
BFQ has to update far fewer statistics, and, in particular, not the
heaviest to update.  To give an idea of the benefits, if
CONFIG_DEBUG_BLK_CGROUP is not set, then, on an Intel i7-4850HQ, and
with 8 threads doing random I/O in parallel on null_blk (configured
with 0 latency), the throughput of BFQ grows from 310 to 400 KIOPS
(+30%). We have measured similar or even much higher boosts with other
CPUs: e.g., +45% with an ARM CortexTM-A53 Octa-core. Our results have
been obtained and can be reproduced very easily with the script in [1].

[1] https://www.spinics.net/lists/linux-block/msg18943.htmlSuggested-by: NTejun Heo <tj@kernel.org>
Suggested-by: NUlf Hansson <ulf.hansson@linaro.org>
Tested-by: NLee Tibbert <lee.tibbert@gmail.com>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Signed-off-by: NLuca Miccio <lucmiccio@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>