One of the methods with which bfq boosts throughput is by merging queues.
One of the merging variants in bfq is the stable merge.
This mechanism is activated between two queues only if they are created
within a certain maximum time T1 from each other.
Merging can happen soon or be delayed. In the second case, before
merging, bfq needs to evaluate a throughput-boost parameter that
indicates whether the queue generates a high throughput is served alone.
Merging occurs when this throughput-boost is not high enough.
In particular, this parameter is evaluated and late merging may occur
only after at least a time T2 from the creation of the queue.

Currently T1 and T2 are set to 180ms and 200ms, respectively.
In this way the merging mechanism rarely occurs because time is not
enough. This results in a noticeable lowering of the overall throughput
with some workloads (see the example below).

This commit introduces two constants bfq_activation_stable_merging and
bfq_late_stable_merging in order to increase the duration of T1 and T2.
Both the stable merging activation time and the late merging
time are set to 600ms. This value has been experimentally evaluated
using sqlite benchmark in the Phoronix Test Suite on a HDD.
The duration of the benchmark before this fix was 111.02s, while now
it has reached 97.02s, a better result than that of all the other
schedulers.
Signed-off-by: NPietro Pedroni <pedroni.pietro.96@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Link: https://lore.kernel.org/r/20210619140948.98712-5-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

Since commit 430a67f9 ("block, bfq: merge bursts of newly-created
queues"), BFQ may schedule a merge between a newly created sync
bfq_queue and the last sync bfq_queue created. Such a merging is not
performed immediately, because BFQ needs first to find out whether the
newly created queue actually reaches a higher throughput if not merged
at all (and in that case BFQ will not perform any stable merging). To
check that, a little time must be waited after the creation of the new
queue, so that some I/O can flow in the queue, and statistics on such
I/O can be computed.

Yet, to evaluate the above waiting time, the last split time is
considered as start time, instead of the creation time of the
queue. This is a mistake, because considering the split time is
correct only in the following scenario.

The queue undergoes a non-stable merges on the arrival of its very
first I/O request, due to close I/O with some other queue. While the
queue is merged for close I/O, stable merging is not considered. Yet
the queue may then happen to be split, if the close I/O finishes (or
happens to be a false positive). From this time on, the queue can
again be considered for stable merging. But, again, a little time must
elapse, to let some new I/O flow in the queue and to get updated
statistics. To wait for this time, the split time is to be taken into
account.

Yet, if the queue does not undergo a non-stable merge on the arrival
of its very first request, then BFQ immediately checks whether the
stable merge is to be performed. It happens because the split time for
a queue is initialized to minus infinity when the queue is created.

This commit fixes this mistake by adding the missing condition. Now
the check for delayed stable-merge is performed after a little time is
elapsed not only from the last queue split time, but also from the
creation time of the queue.

Fixes: 430a67f9 ("block, bfq: merge bursts of newly-created queues")
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Link: https://lore.kernel.org/r/20210619140948.98712-4-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

When attempting to schedule a merge of a given bfq_queue with the currently
in-service bfq_queue or with a cooperating bfq_queue among the scheduled
bfq_queues, delayed stable merge is checked for rotational or non-queueing
devs. For this stable merge to be performed, some conditions must be met.
If the current bfq_queue underwent some split from some merged bfq_queue,
one of these conditions is that two hundred milliseconds must elapse from
split, otherwise this condition is always met.

Unfortunately, by mistake, time_is_after_jiffies() was written instead of
time_is_before_jiffies() for this check, verifying that less than two
hundred milliseconds have elapsed instead of verifying that at least two
hundred milliseconds have elapsed.

Fix this issue by replacing time_is_after_jiffies() with
time_is_before_jiffies().
Signed-off-by: NLuca Mariotti <mariottiluca1@hotmail.it>
Signed-off-by: NPaolo Valente <paolo.valente@unimore.it>
Signed-off-by: NPietro Pedroni <pedroni.pietro.96@gmail.com>
Link: https://lore.kernel.org/r/20210619140948.98712-3-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

Merged bfq_queues are kept out of weight-raising (low-latency)
mechanisms. The reason is that these queues are usually created for
non-interactive and non-soft-real-time tasks. Yet this is not the case
for stably-merged queues. These queues are merged just because they
are created shortly after each other. So they may easily serve the I/O
of an interactive or soft-real time application, if the application
happens to spawn multiple processes.

To address this issue, this commits lets also stably-merged queued
enjoy weight raising.
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Link: https://lore.kernel.org/r/20210619140948.98712-2-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

BFQ may merge a new bfq_queue, stably, with the last bfq_queue
created. In particular, BFQ first waits a little bit for some I/O to
flow inside the new queue, say Q2, if this is needed to understand
whether it is better or worse to merge Q2 with the last queue created,
say Q1. This delayed stable merge is performed by assigning
bic->stable_merge_bfqq = Q1, for the bic associated with Q1.

Yet, while waiting for some I/O to flow in Q2, a non-stable queue
merge of Q2 with Q1 may happen, causing the bic previously associated
with Q2 to be associated with exactly Q1 (bic->bfqq = Q1). After that,
Q2 and Q1 may happen to be split, and, in the split, Q1 may happen to
be recycled as a non-shared bfq_queue. In that case, Q1 may then
happen to undergo a stable merge with the bfq_queue pointed by
bic->stable_merge_bfqq. Yet bic->stable_merge_bfqq still points to
Q1. So Q1 would be merged with itself.

This commit fixes this error by intercepting this situation, and
canceling the schedule of the stable merge.

Fixes: 430a67f9 ("block, bfq: merge bursts of newly-created queues")
Signed-off-by: NPietro Pedroni <pedroni.pietro.96@gmail.com>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Link: https://lore.kernel.org/r/20210512094352.85545-2-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

__blk_mq_sched_bio_merge() gets the ctx and hctx for the current CPU and
passes the hctx to ->bio_merge(). kyber_bio_merge() then gets the ctx
for the current CPU again and uses that to get the corresponding Kyber
context in the passed hctx. However, the thread may be preempted between
the two calls to blk_mq_get_ctx(), and the ctx returned the second time
may no longer correspond to the passed hctx. This "works" accidentally
most of the time, but it can cause us to read garbage if the second ctx
came from an hctx with more ctx's than the first one (i.e., if
ctx->index_hw[hctx->type] > hctx->nr_ctx).

This manifested as this UBSAN array index out of bounds error reported
by Jakub:

UBSAN: array-index-out-of-bounds in ../kernel/locking/qspinlock.c:130:9
index 13106 is out of range for type 'long unsigned int [128]'
Call Trace:
 dump_stack+0xa4/0xe5
 ubsan_epilogue+0x5/0x40
 __ubsan_handle_out_of_bounds.cold.13+0x2a/0x34
 queued_spin_lock_slowpath+0x476/0x480
 do_raw_spin_lock+0x1c2/0x1d0
 kyber_bio_merge+0x112/0x180
 blk_mq_submit_bio+0x1f5/0x1100
 submit_bio_noacct+0x7b0/0x870
 submit_bio+0xc2/0x3a0
 btrfs_map_bio+0x4f0/0x9d0
 btrfs_submit_data_bio+0x24e/0x310
 submit_one_bio+0x7f/0xb0
 submit_extent_page+0xc4/0x440
 __extent_writepage_io+0x2b8/0x5e0
 __extent_writepage+0x28d/0x6e0
 extent_write_cache_pages+0x4d7/0x7a0
 extent_writepages+0xa2/0x110
 do_writepages+0x8f/0x180
 __writeback_single_inode+0x99/0x7f0
 writeback_sb_inodes+0x34e/0x790
 __writeback_inodes_wb+0x9e/0x120
 wb_writeback+0x4d2/0x660
 wb_workfn+0x64d/0xa10
 process_one_work+0x53a/0xa80
 worker_thread+0x69/0x5b0
 kthread+0x20b/0x240
 ret_from_fork+0x1f/0x30

Only Kyber uses the hctx, so fix it by passing the request_queue to
->bio_merge() instead. BFQ and mq-deadline just use that, and Kyber can
map the queues itself to avoid the mismatch.

Fixes: a6088845 ("block: kyber: make kyber more friendly with merging")
Reported-by: NJakub Kicinski <kuba@kernel.org>
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Link: https://lore.kernel.org/r/c7598605401a48d5cfeadebb678abd10af22b83f.1620691329.git.osandov@fb.comSigned-off-by: NJens Axboe <axboe@kernel.dk>

Since commit 01e99aec 'blk-mq: insert passthrough request into
hctx->dispatch directly', passthrough request should not appear in
IO-scheduler any more, so blk_rq_is_passthrough checking in addon IO
schedulers is redundant.

(Notes: this patch passes generic IO load test with hdds under SAS
controller and hdds under AHCI controller but obviously not covers all.
Not sure if passthrough request can still escape into IO scheduler from
blk_mq_sched_insert_requests, which is used by blk_mq_flush_plug_list and
has lots of indirect callers.)
Signed-off-by: NLin Feng <linf@wangsu.com>
Reviewed-by: NMing Lei <ming.lei@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Many throughput-sensitive workloads are made of several parallel I/O
flows, with all flows generated by the same application, or more
generically by the same task (e.g., system boot). The most
counterproductive action with these workloads is plugging I/O dispatch
when one of the bfq_queues associated with these flows remains
temporarily empty.

To avoid this plugging, BFQ has been using a burst-handling mechanism
for years now. This mechanism has proven effective for throughput, and
not detrimental for service guarantees. This commit pushes this
mechanism a little bit further, basing on the following two facts.

First, all the I/O flows of a the same application or task contribute
to the execution/completion of that common application or task. So the
performance figures that matter are total throughput of the flows and
task-wide I/O latency.  In particular, these flows do not need to be
protected from each other, in terms of individual bandwidth or
latency.

Second, the above fact holds regardless of the number of flows.

Putting these two facts together, this commits merges stably the
bfq_queues associated with these I/O flows, i.e., with the processes
that generate these IO/ flows, regardless of how many the involved
processes are.

To decide whether a set of bfq_queues is actually associated with the
I/O flows of a common application or task, and to merge these queues
stably, this commit operates as follows: given a bfq_queue, say Q2,
currently being created, and the last bfq_queue, say Q1, created
before Q2, Q2 is merged stably with Q1 if
- very little time has elapsed since when Q1 was created
- Q2 has the same ioprio as Q1
- Q2 belongs to the same group as Q1

Merging bfq_queues also reduces scheduling overhead. A fio test with
ten random readers on /dev/nullb shows a throughput boost of 40%, with
a quadcore. Since BFQ's execution time amounts to ~50% of the total
per-request processing time, the above throughput boost implies that
BFQ's overhead is reduced by more than 50%.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Link: https://lore.kernel.org/r/20210304174627.161-7-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

Shared queues are likely to receive I/O at a high rate. This may
deceptively let them be considered as wakers of other queues. But a
false waker will unjustly steal bandwidth to its supposedly woken
queue. So considering also shared queues in the waking mechanism may
cause more control troubles than throughput benefits. This commit
keeps shared queues out of the waker-detection mechanism.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Link: https://lore.kernel.org/r/20210304174627.161-6-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

When the io_latency heuristic is off, bfq_queues must not start to be
weight-raised. Unfortunately, by mistake, this may happen when the
state of a previously weight-raised bfq_queue is resumed after a queue
split. This commit fixes this error.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Link: https://lore.kernel.org/r/20210304174627.161-5-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

Consider a bfq_queue bfqq that is about to be merged with another
bfq_queue new_bfqq. The processes associated with bfqq are cooperators
of the processes associated with new_bfqq. So, if bfqq has a waker,
then it is reasonable (and beneficial for throughput) to assume that
all these processes will be happy to let bfqq's waker freely inject
I/O when they have no I/O. So this commit makes new_bfqq inherit
bfqq's waker.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Link: https://lore.kernel.org/r/20210304174627.161-4-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

Consider a new I/O request that arrives for a bfq_queue bfqq. If, when
this happens, the only active bfq_queues are bfqq and either its waker
bfq_queue or one of its woken bfq_queues, then there is no point in
queueing this new I/O request in bfqq for service. In fact, the
in-service queue and bfqq agree on serving this new I/O request as
soon as possible. So this commit puts this new I/O request directly
into the dispatch list.
Tested-by: NJan Kara <jack@suse.cz>
Acked-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Link: https://lore.kernel.org/r/20210304174627.161-3-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

Suppose that I/O dispatch is plugged, to wait for new I/O for the
in-service bfq-queue, say bfqq.  Suppose then that there is a further
bfq_queue woken by bfqq, and that this woken queue has pending I/O. A
woken queue does not steal bandwidth from bfqq, because it remains
soon without I/O if bfqq is not served. So there is virtually no risk
of loss of bandwidth for bfqq if this woken queue has I/O dispatched
while bfqq is waiting for new I/O. In contrast, this extra I/O
injection boosts throughput. This commit performs this extra
injection.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Tested-by: NOleksandr Natalenko <oleksandr@natalenko.name>
Link: https://lore.kernel.org/r/20210304174627.161-2-paolo.valente@linaro.orgSigned-off-by: NJens Axboe <axboe@kernel.dk>

Correct the comments since bfq_fifo_expire[0] is for async request,
while bfq_fifo_expire[1] is for sync request.
Also update docs, according the source code, the default
fifo_expire_async is 250ms, and fifo_expire_sync is 125ms.
Signed-off-by: NJoseph Qi <joseph.qi@linux.alibaba.com>
Acked-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Get rid of the wrapper for trace_block_rq_insert() and call the function
directly.
Signed-off-by: NChaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Reviewed-by: NJohannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: NDamien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: NChristoph Hellwig <hch@lst.de>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

This reverts commit 6d4d2735.

bfq.limit_depth passes word_depths[] as shallow_depth down to sbitmap core
sbitmap_get_shallow, which uses just the number to limit the scan depth of
each bitmap word, formula:
scan_percentage_for_each_word = shallow_depth / (1 << sbimap->shift) * 100%

That means the comments's percentiles 50%, 75%, 18%, 37% of bfq are correct.
But after commit patch 'bfq: Fix computation of shallow depth', we use
sbitmap.depth instead, as a example in following case:

sbitmap.depth = 256, map_nr = 4, shift = 6; sbitmap_word.depth = 64.
The resulsts of computed bfqd->word_depths[] are {128, 192, 48, 96}, and
three of the numbers exceed core dirver's 'sbitmap_word.depth=64' limit
nothing.
Signed-off-by: NLin Feng <linf@wangsu.com>
Reviewed-by: NJan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Currently whenever bfq queue has a request queued we add now -
last_completion_time to the think time statistics. This is however
misleading in case the process is able to submit several requests in
parallel because e.g. if the queue has request completed at time T0 and
then queues new requests at times T1, T2, then we will add T1-T0 and
T2-T0 to think time statistics which just doesn't make any sence (the
queue's think time is penalized by the queue being able to submit more
IO). So add to think time statistics only time intervals when the queue
had no IO pending.
Signed-off-by: NJan Kara <jack@suse.cz>
Acked-by: NPaolo Valente <paolo.valente@linaro.org>
[axboe: fix whitespace on empty line]
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Use local variable 'ttime' instead of dereferencing bfqq.
Signed-off-by: NJan Kara <jack@suse.cz>
Acked-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

bfq_setup_cooperator() uses bfqd->in_serv_last_pos so detect whether it
makes sense to merge current bfq queue with the in-service queue.
However if the in-service queue is freshly scheduled and didn't dispatch
any requests yet, bfqd->in_serv_last_pos is stale and contains value
from the previously scheduled bfq queue which can thus result in a bogus
decision that the two queues should be merged. This bug can be observed
for example with the following fio jobfile:

[global]
direct=0
ioengine=sync
invalidate=1
size=1g
rw=read

[reader]
numjobs=4
directory=/mnt

where the 4 processes will end up in the one shared bfq queue although
they do IO to physically very distant files (for some reason I was able to
observe this only with slice_idle=1ms setting).

Fix the problem by invalidating bfqd->in_serv_last_pos when switching
in-service queue.

Fixes: 058fdecc ("block, bfq: fix in-service-queue check for queue merging")
CC: stable@vger.kernel.org
Signed-off-by: NJan Kara <jack@suse.cz>
Acked-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

It's only used in the same file, mark is appropriately static.

Fixes: 71217df3 ("block, bfq: make waker-queue detection more robust")
Reported-by: Nkernel test robot <lkp@intel.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

In the presence of many parallel I/O flows, the detection of waker
bfq_queues suffers from false positives. This commits addresses this
issue by making the filtering of actual wakers more selective. In more
detail, a candidate waker must be found to meet waker requirements
three times before being promoted to actual waker.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

To prevent injection information from being lost on bfq_queue merging,
also the amount of service that a bfq_queue receives must be saved and
restored when the bfq_queue is merged and split, respectively.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

To prevent weight-raising information from being lost on bfq_queue merging,
also the amount of service that a bfq_queue receives must be saved and
restored when the bfq_queue is merged and split, respectively.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

A bfq_queue may happen to be deemed as soft real-time while it is
still enjoying interactive weight-raising. If this happens because of
a false positive, then the bfq_queue is likely to loose its soft
real-time status soon. Upon losing such a status, the bfq_queue must
get back its interactive weight-raising, if its interactive period is
not over yet. But this case is not handled. This commit corrects this
error.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Upon an I/O-dispatch attempt, BFQ may detect that it was better to
plug I/O dispatch, and to wait for a new request to arrive for the
currently in-service queue. But the arrival of a new request for an
empty bfq_queue, and thus the switch from idle to busy of the
bfq_queue, may cause the scenario to change, and make plugging no
longer needed for service guarantees, or more convenient for
throughput. In this case, keeping I/O-dispatch plugged would certainly
lower throughput.

To address this issue, this commit makes such a check, and stops
plugging I/O if it is better to stop plugging I/O.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Some BFQ mechanisms make their decisions on a bfq_queue basing also on
whether the bfq_queue is I/O bound. In this respect, the current logic
for evaluating whether a bfq_queue is I/O bound is rather rough. This
commits replaces this logic with a more effective one.

The new logic measures the percentage of time during which a bfq_queue
is active, and marks the bfq_queue as I/O bound if the latter if this
percentage is above a fixed threshold.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

This reverts commit b445547e.

Since both mq-deadline and BFQ completely ignore hctx they are passed to
their dispatch function and dispatch whatever request they deem fit
checking whether any request for a particular hctx is queued is just
pointless since we'll very likely get a request from a different hctx
anyway. In the following commit we'll deal with lock contention in these
IO schedulers in presence of multiple HW queues in a different way.
Signed-off-by: NJan Kara <jack@suse.cz>
Reviewed-by: NMing Lei <ming.lei@redhat.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

This commits preserves I/O-dispatch plugging for a special symmetric
case that may suddenly turn into asymmetric: the case where only one
bfq_queue, say bfqq, is busy. In this case, not expiring bfqq does not
cause any harm to any other queues in terms of service guarantees. In
contrast, it avoids the following unlucky sequence of events: (1) bfqq
is expired, (2) a new queue with a lower weight than bfqq becomes busy
(or more queues), (3) the new queue is served until a new request
arrives for bfqq, (4) when bfqq is finally served, there are so many
requests of the new queue in the drive that the pending requests for
bfqq take a lot of time to be served. In particular, event (2) may
case even already dispatched requests of bfqq to be delayed, inside
the drive. So, to avoid this series of events, the scenario is
preventively declared as asymmetric also if bfqq is the only busy
queues. By doing so, I/O-dispatch plugging is performed for bfqq.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

BFQ tags some bfq_queues as interactive or soft_rt if it deems that
these bfq_queues contain the I/O of, respectively, interactive or soft
real-time applications. BFQ privileges both these special types of
bfq_queues over normal bfq_queues. To privilege a bfq_queue, BFQ
mainly raises the weight of the bfq_queue. In particular, soft_rt
bfq_queues get a higher weight than interactive bfq_queues.

A bfq_queue may turn from interactive to soft_rt. And this leads to a
tricky issue. Soft real-time applications usually start with an
I/O-bound, interactive phase, in which they load themselves into main
memory. BFQ correctly detects this phase, and keeps the bfq_queues
associated with the application in interactive mode for a
while. Problems arise when the I/O pattern of the application finally
switches to soft real-time. One of the conditions for a bfq_queue to
be deemed as soft_rt is that the bfq_queue does not consume too much
bandwidth. But the bfq_queues associated with a soft real-time
application consume as much bandwidth as they can in the loading phase
of the application. So, after the application becomes truly soft
real-time, a lot of time should pass before the average bandwidth
consumed by its bfq_queues finally drops to a value acceptable for
soft_rt bfq_queues. As a consequence, there might be a time gap during
which the application is not privileged at all, because its bfq_queues
are not interactive any longer, but cannot be deemed as soft_rt yet.

To avoid this problem, BFQ pretends that an interactive bfq_queue
consumes zero bandwidth, and allows an interactive bfq_queue to switch
to soft_rt. Yet, this fake zero-bandwidth consumption easily causes
the bfq_queue to often switch to soft_rt deceptively, during its
loading phase. As in soft_rt mode, the bfq_queue gets its bandwidth
correctly computed, and therefore soon switches back to
interactive. Then it switches again to soft_rt, and so on. These
spurious fluctuations usually cause losses of throughput, because they
deceive BFQ's mechanisms for boosting throughput (injection,
I/O-plugging avoidance, ...).

This commit addresses this issue as follows:
1) It does compute actual bandwidth consumption also for interactive
   bfq_queues. This avoids the above false positives.
2) When a bfq_queue switches from interactive to normal mode, the
   consumed bandwidth is reset (forgotten). This allows the
   bfq_queue to enjoy soft_rt very quickly. In particular, two
   alternatives are possible in this switch:
    - the bfq_queue still has backlog, and therefore there is a budget
      already scheduled to serve the bfq_queue; in this case, the
      scheduling of the current budget of the bfq_queue is not
      hindered, because only the scheduling of the next budget will
      be affected by the weight drop. After that, if the bfq_queue is
      actually in a soft_rt phase, and becomes empty during the
      service of its current budget, which is the natural behavior of
      a soft_rt bfq_queue, then the bfq_queue will be considered as
      soft_rt when its next I/O arrives. If, in contrast, the
      bfq_queue remains constantly non-empty, then its next budget
      will be scheduled with a low weight, which is the natural
      treatment for an I/O-bound (non soft_rt) bfq_queue.
    - the bfq_queue is empty; in this case, the bfq_queue may be
      considered unjustly soft_rt when its new I/O arrives. Yet
      the problem is now much smaller than before, because it is
      unlikely that more than one spurious fluctuation occurs.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

BFQ heuristics try to detect interactive I/O, and raise the weight of
the queues containing such an I/O. Yet, if also the user changes the
weight of a queue (i.e., the user changes the ioprio of the process
associated with that queue), then it is most likely better to prevent
BFQ heuristics from silently changing the same weight.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Tests on slower machines showed current window to be way too
small. This commit increases it.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Since commit c5089591c3ba ("block, bfq: detect wakers and
unconditionally inject their I/O"), when the in-service bfq_queue, say
Q, is temporarily empty, BFQ checks whether there are I/O requests to
inject (also) from the waker bfq_queue for Q. To this goal, the value
pointed by bfqq->waker_bfqq->next_rq must be controlled. However, the
current implementation mistakenly looks at bfqq->next_rq, which
instead points to the next request of the currently served queue.

This mistake evidently causes losses of throughput in scenarios with
waker bfq_queues.

This commit corrects this mistake.

Fixes: c5089591c3ba ("block, bfq: detect wakers and unconditionally inject their I/O")
Signed-off-by: NJia Cheng Hu <jia.jiachenghu@gmail.com>
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

The value of the I/O plugging (idling) timeout is used also as the
think-time threshold to decide whether a process has a short think
time.  In this respect, a good value of this timeout for rotational
drives is un the order of several ms. Yet, this is often too long a
time interval to be effective as a think-time threshold. This commit
mitigates this problem (by a lot, according to tests), by halving the
threshold.
Tested-by: NJan Kara <jack@suse.cz>
Signed-off-by: NPaolo Valente <paolo.valente@linaro.org>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

BFQ computes number of tags it allows to be allocated for each request type
based on tag bitmap. However it uses 1 << bitmap.shift as number of
available tags which is wrong. 'shift' is just an internal bitmap value
containing logarithm of how many bits bitmap uses in each bitmap word.
Thus number of tags allowed for some request types can be far to low.
Use proper bitmap.depth which has the number of tags instead.
Signed-off-by: NJan Kara <jack@suse.cz>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Yang Yang reported the following crash caused by requeueing a flush
request in Kyber:

  [    2.517297] Unable to handle kernel paging request at virtual address ffffffd8071c0b00
  ...
  [    2.517468] pc : clear_bit+0x18/0x2c
  [    2.517502] lr : sbitmap_queue_clear+0x40/0x228
  [    2.517503] sp : ffffff800832bc60 pstate : 00c00145
  ...
  [    2.517599] Process ksoftirqd/5 (pid: 51, stack limit = 0xffffff8008328000)
  [    2.517602] Call trace:
  [    2.517606]  clear_bit+0x18/0x2c
  [    2.517619]  kyber_finish_request+0x74/0x80
  [    2.517627]  blk_mq_requeue_request+0x3c/0xc0
  [    2.517637]  __scsi_queue_insert+0x11c/0x148
  [    2.517640]  scsi_softirq_done+0x114/0x130
  [    2.517643]  blk_done_softirq+0x7c/0xb0
  [    2.517651]  __do_softirq+0x208/0x3bc
  [    2.517657]  run_ksoftirqd+0x34/0x60
  [    2.517663]  smpboot_thread_fn+0x1c4/0x2c0
  [    2.517667]  kthread+0x110/0x120
  [    2.517669]  ret_from_fork+0x10/0x18

This happens because Kyber doesn't track flush requests, so
kyber_finish_request() reads a garbage domain token. Only call the
scheduler's requeue_request() hook if RQF_ELVPRIV is set (like we do for
the finish_request() hook in blk_mq_free_request()). Now that we're
handling it in blk-mq, also remove the check from BFQ.
Reported-by: NYang Yang <yang.yang@vivo.com>
Signed-off-by: NOmar Sandoval <osandov@fb.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

High CPU utilization on "native_queued_spin_lock_slowpath" due to lock
contention is possible for mq-deadline and bfq IO schedulers
when nr_hw_queues is more than one.

It is because kblockd work queue can submit IO from all online CPUs
(through blk_mq_run_hw_queues()) even though only one hctx has pending
commands.

The elevator callback .has_work for mq-deadline and bfq scheduler considers
pending work if there are any IOs on request queue but it does not account
hctx context.

Add a per-hctx 'elevator_queued' count to the hctx to avoid triggering
the elevator even though there are no requests queued.

[jpg: Relocated atomic_dec() in dd_dispatch_request(), update commit message per Kashyap]
Signed-off-by: NKashyap Desai <kashyap.desai@broadcom.com>
Signed-off-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Tested-by: NDouglas Gilbert <dgilbert@interlog.com>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Introduce pointers for the blk_mq_tags regular and reserved bitmap tags,
with the goal of later being able to use a common shared tag bitmap across
all HW contexts in a set.
Signed-off-by: NJohn Garry <john.garry@huawei.com>
Tested-by: Don Brace<don.brace@microsemi.com> #SCSI resv cmds patches used
Tested-by: NDouglas Gilbert <dgilbert@interlog.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Signed-off-by: NJens Axboe <axboe@kernel.dk>

Replace the existing /* fall through */ comments and its variants with
the new pseudo-keyword macro fallthrough[1]. Also, remove unnecessary
fall-through markings when it is the case.

[1] https://www.kernel.org/doc/html/v5.7/process/deprecated.html?highlight=fallthrough#implicit-switch-case-fall-throughSigned-off-by: NGustavo A. R. Silva <gustavoars@kernel.org>

Change "at at" to "at a".
Signed-off-by: NRandy Dunlap <rdunlap@infradead.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: linux-block@vger.kernel.org
Signed-off-by: NJens Axboe <axboe@kernel.dk>

None of the I/O schedulers actually needs it.
Signed-off-by: NChristoph Hellwig <hch@lst.de>
Reviewed-by: NMing Lei <ming.lei@redhat.com>
Reviewed-by: NHannes Reinecke <hare@suse.de>
Reviewed-by: NJohannes Thumshirn <johannes.thumshirn@wdc.com>
Reviewed-by: NBart Van Assche <bvanassche@acm.org>
Reviewed-by: NDaniel Wagner <dwagner@suse.de>
Signed-off-by: NJens Axboe <axboe@kernel.dk>