Currently, blk-mq timeout path synchronizes against the usual
issue/completion path using a complex scheme involving atomic
bitflags, REQ_ATOM_*, memory barriers and subtle memory coherence
rules. Unfortunately, it contains quite a few holes.
There's a complex dancing around REQ_ATOM_STARTED and
REQ_ATOM_COMPLETE between issue/completion and timeout paths; however,
they don't have a synchronization point across request recycle
instances and it isn't clear what the barriers add.
blk_mq_check_expired() can easily read STARTED from N-2'th iteration,
deadline from N-1'th, blk_mark_rq_complete() against Nth instance.
In fact, it's pretty easy to make blk_mq_check_expired() terminate a
later instance of a request. If we induce 5 sec delay before
time_after_eq() test in blk_mq_check_expired(), shorten the timeout to
2s, and issue back-to-back large IOs, blk-mq starts timing out
requests spuriously pretty quickly. Nothing actually timed out. It
just made the call on a recycle instance of a request and then
terminated a later instance long after the original instance finished.
The scenario isn't theoretical either.
This patch replaces the broken synchronization mechanism with a RCU
and generation number based one.
1. Each request has a u64 generation + state value, which can be
updated only by the request owner. Whenever a request becomes
in-flight, the generation number gets bumped up too. This provides
the basis for the timeout path to distinguish different recycle
instances of the request.
Also, marking a request in-flight and setting its deadline are
protected with a seqcount so that the timeout path can fetch both
values coherently.
2. The timeout path fetches the generation, state and deadline. If
the verdict is timeout, it records the generation into a dedicated
request abortion field and does RCU wait.
3. The completion path is also protected by RCU (from the previous
patch) and checks whether the current generation number and state
match the abortion field. If so, it skips completion.
4. The timeout path, after RCU wait, scans requests again and
terminates the ones whose generation and state still match the ones
requested for abortion.
By now, the timeout path knows that either the generation number
and state changed if it lost the race or the completion will yield
to it and can safely timeout the request.
While it's more lines of code, it's conceptually simpler, doesn't
depend on direct use of subtle memory ordering or coherence, and
hopefully doesn't terminate the wrong instance.
While this change makes REQ_ATOM_COMPLETE synchronization unnecessary
between issue/complete and timeout paths, REQ_ATOM_COMPLETE isn't
removed yet as it's still used in other places. Future patches will
move all state tracking to the new mechanism and remove all bitops in
the hot paths.
Note that this patch adds a comment explaining a race condition in
BLK_EH_RESET_TIMER path. The race has always been there and this
patch doesn't change it. It's just documenting the existing race.
v2: - Fixed BLK_EH_RESET_TIMER handling as pointed out by Jianchao.
- s/request->gstate_seqc/request->gstate_seq/ as suggested by Peter.
- READ_ONCE() added in blk_mq_rq_update_state() as suggested by Peter.
v3: - Fixed possible extended seqcount / u64_stats_sync read looping
spotted by Peter.
- MQ_RQ_IDLE was incorrectly being set in complete_request instead
of free_request. Fixed.
v4: - Rebased on top of hctx_lock() refactoring patch.
- Added comment explaining the use of hctx_lock() in completion path.
v5: - Added comments requested by Bart.
- Note the addition of BLK_EH_RESET_TIMER race condition in the
commit message.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: "jianchao.wang" <jianchao.w.wang@oracle.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Christoph Hellwig <hch@lst.de>
Cc: Bart Van Assche <Bart.VanAssche@wdc.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pull core block layer updates from Jens Axboe:
"This is the main pull request for block storage for 4.15-rc1.
Nothing out of the ordinary in here, and no API changes or anything
like that. Just various new features for drivers, core changes, etc.
In particular, this pull request contains:
- A patch series from Bart, closing the whole on blk/scsi-mq queue
quescing.
- A series from Christoph, building towards hidden gendisks (for
multipath) and ability to move bio chains around.
- NVMe
- Support for native multipath for NVMe (Christoph).
- Userspace notifications for AENs (Keith).
- Command side-effects support (Keith).
- SGL support (Chaitanya Kulkarni)
- FC fixes and improvements (James Smart)
- Lots of fixes and tweaks (Various)
- bcache
- New maintainer (Michael Lyle)
- Writeback control improvements (Michael)
- Various fixes (Coly, Elena, Eric, Liang, et al)
- lightnvm updates, mostly centered around the pblk interface
(Javier, Hans, and Rakesh).
- Removal of unused bio/bvec kmap atomic interfaces (me, Christoph)
- Writeback series that fix the much discussed hundreds of millions
of sync-all units. This goes all the way, as discussed previously
(me).
- Fix for missing wakeup on writeback timer adjustments (Yafang
Shao).
- Fix laptop mode on blk-mq (me).
- {mq,name} tupple lookup for IO schedulers, allowing us to have
alias names. This means you can use 'deadline' on both !mq and on
mq (where it's called mq-deadline). (me).
- blktrace race fix, oopsing on sg load (me).
- blk-mq optimizations (me).
- Obscure waitqueue race fix for kyber (Omar).
- NBD fixes (Josef).
- Disable writeback throttling by default on bfq, like we do on cfq
(Luca Miccio).
- Series from Ming that enable us to treat flush requests on blk-mq
like any other request. This is a really nice cleanup.
- Series from Ming that improves merging on blk-mq with schedulers,
getting us closer to flipping the switch on scsi-mq again.
- BFQ updates (Paolo).
- blk-mq atomic flags memory ordering fixes (Peter Z).
- Loop cgroup support (Shaohua).
- Lots of minor fixes from lots of different folks, both for core and
driver code"
* 'for-4.15/block' of git://git.kernel.dk/linux-block: (294 commits)
nvme: fix visibility of "uuid" ns attribute
blk-mq: fixup some comment typos and lengths
ide: ide-atapi: fix compile error with defining macro DEBUG
blk-mq: improve tag waiting setup for non-shared tags
brd: remove unused brd_mutex
blk-mq: only run the hardware queue if IO is pending
block: avoid null pointer dereference on null disk
fs: guard_bio_eod() needs to consider partitions
xtensa/simdisk: fix compile error
nvme: expose subsys attribute to sysfs
nvme: create 'slaves' and 'holders' entries for hidden controllers
block: create 'slaves' and 'holders' entries for hidden gendisks
nvme: also expose the namespace identification sysfs files for mpath nodes
nvme: implement multipath access to nvme subsystems
nvme: track shared namespaces
nvme: introduce a nvme_ns_ids structure
nvme: track subsystems
block, nvme: Introduce blk_mq_req_flags_t
block, scsi: Make SCSI quiesce and resume work reliably
block: Add the QUEUE_FLAG_PREEMPT_ONLY request queue flag
...
Currently we are inconsistent in when we decide to run the queue. Using
blk_mq_run_hw_queues() we check if the hctx has pending IO before
running it, but we don't do that from the individual queue run function,
blk_mq_run_hw_queue(). This results in a lot of extra and pointless
queue runs, potentially, on flush requests and (much worse) on tag
starvation situations. This is observable just looking at top output,
with lots of kworkers active. For the !async runs, it just adds to the
CPU overhead of blk-mq.
Move the has-pending check into the run function instead of having
callers do it.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Several block layer and NVMe core functions accept a combination
of BLK_MQ_REQ_* flags through the 'flags' argument but there is
no verification at compile time whether the right type of block
layer flags is passed. Make it possible for sparse to verify this.
This patch does not change any functionality.
Signed-off-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Tested-by: Oleksandr Natalenko <oleksandr@natalenko.name>
Cc: linux-nvme@lists.infradead.org
Cc: Christoph Hellwig <hch@lst.de>
Cc: Johannes Thumshirn <jthumshirn@suse.de>
Cc: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We need this helper to put the driver tag for flush rq, since we will
not share tag in the flush request sequence in the following patch
in case that I/O scheduler is applied.
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Block flush need this function without running the queue, so add a
parameter controlling whether we run it or not.
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
It is enough to just check if we can get the budget via .get_budget().
And we don't need to deal with device state change in .get_budget().
For SCSI, one issue to be fixed is that we have to call
scsi_mq_uninit_cmd() to free allocated ressources if SCSI device fails
to handle the request. And it isn't enough to simply call
blk_mq_end_request() to do that if this request is marked as
RQF_DONTPREP.
Fixes: 0df21c86bdbf(scsi: implement .get_budget and .put_budget for blk-mq)
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Many source files in the tree are missing licensing information, which
makes it harder for compliance tools to determine the correct license.
By default all files without license information are under the default
license of the kernel, which is GPL version 2.
Update the files which contain no license information with the 'GPL-2.0'
SPDX license identifier. The SPDX identifier is a legally binding
shorthand, which can be used instead of the full boiler plate text.
This patch is based on work done by Thomas Gleixner and Kate Stewart and
Philippe Ombredanne.
How this work was done:
Patches were generated and checked against linux-4.14-rc6 for a subset of
the use cases:
- file had no licensing information it it.
- file was a */uapi/* one with no licensing information in it,
- file was a */uapi/* one with existing licensing information,
Further patches will be generated in subsequent months to fix up cases
where non-standard license headers were used, and references to license
had to be inferred by heuristics based on keywords.
The analysis to determine which SPDX License Identifier to be applied to
a file was done in a spreadsheet of side by side results from of the
output of two independent scanners (ScanCode & Windriver) producing SPDX
tag:value files created by Philippe Ombredanne. Philippe prepared the
base worksheet, and did an initial spot review of a few 1000 files.
The 4.13 kernel was the starting point of the analysis with 60,537 files
assessed. Kate Stewart did a file by file comparison of the scanner
results in the spreadsheet to determine which SPDX license identifier(s)
to be applied to the file. She confirmed any determination that was not
immediately clear with lawyers working with the Linux Foundation.
Criteria used to select files for SPDX license identifier tagging was:
- Files considered eligible had to be source code files.
- Make and config files were included as candidates if they contained >5
lines of source
- File already had some variant of a license header in it (even if <5
lines).
All documentation files were explicitly excluded.
The following heuristics were used to determine which SPDX license
identifiers to apply.
- when both scanners couldn't find any license traces, file was
considered to have no license information in it, and the top level
COPYING file license applied.
For non */uapi/* files that summary was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 11139
and resulted in the first patch in this series.
If that file was a */uapi/* path one, it was "GPL-2.0 WITH
Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 WITH Linux-syscall-note 930
and resulted in the second patch in this series.
- if a file had some form of licensing information in it, and was one
of the */uapi/* ones, it was denoted with the Linux-syscall-note if
any GPL family license was found in the file or had no licensing in
it (per prior point). Results summary:
SPDX license identifier # files
---------------------------------------------------|------
GPL-2.0 WITH Linux-syscall-note 270
GPL-2.0+ WITH Linux-syscall-note 169
((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21
((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17
LGPL-2.1+ WITH Linux-syscall-note 15
GPL-1.0+ WITH Linux-syscall-note 14
((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5
LGPL-2.0+ WITH Linux-syscall-note 4
LGPL-2.1 WITH Linux-syscall-note 3
((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3
((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1
and that resulted in the third patch in this series.
- when the two scanners agreed on the detected license(s), that became
the concluded license(s).
- when there was disagreement between the two scanners (one detected a
license but the other didn't, or they both detected different
licenses) a manual inspection of the file occurred.
- In most cases a manual inspection of the information in the file
resulted in a clear resolution of the license that should apply (and
which scanner probably needed to revisit its heuristics).
- When it was not immediately clear, the license identifier was
confirmed with lawyers working with the Linux Foundation.
- If there was any question as to the appropriate license identifier,
the file was flagged for further research and to be revisited later
in time.
In total, over 70 hours of logged manual review was done on the
spreadsheet to determine the SPDX license identifiers to apply to the
source files by Kate, Philippe, Thomas and, in some cases, confirmation
by lawyers working with the Linux Foundation.
Kate also obtained a third independent scan of the 4.13 code base from
FOSSology, and compared selected files where the other two scanners
disagreed against that SPDX file, to see if there was new insights. The
Windriver scanner is based on an older version of FOSSology in part, so
they are related.
Thomas did random spot checks in about 500 files from the spreadsheets
for the uapi headers and agreed with SPDX license identifier in the
files he inspected. For the non-uapi files Thomas did random spot checks
in about 15000 files.
In initial set of patches against 4.14-rc6, 3 files were found to have
copy/paste license identifier errors, and have been fixed to reflect the
correct identifier.
Additionally Philippe spent 10 hours this week doing a detailed manual
inspection and review of the 12,461 patched files from the initial patch
version early this week with:
- a full scancode scan run, collecting the matched texts, detected
license ids and scores
- reviewing anything where there was a license detected (about 500+
files) to ensure that the applied SPDX license was correct
- reviewing anything where there was no detection but the patch license
was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied
SPDX license was correct
This produced a worksheet with 20 files needing minor correction. This
worksheet was then exported into 3 different .csv files for the
different types of files to be modified.
These .csv files were then reviewed by Greg. Thomas wrote a script to
parse the csv files and add the proper SPDX tag to the file, in the
format that the file expected. This script was further refined by Greg
based on the output to detect more types of files automatically and to
distinguish between header and source .c files (which need different
comment types.) Finally Greg ran the script using the .csv files to
generate the patches.
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
SCSI devices use host-wide tagset, and the shared driver tag space is
often quite big. However, there is also a queue depth for each lun(
.cmd_per_lun), which is often small, for example, on both lpfc and
qla2xxx, .cmd_per_lun is just 3.
So lots of requests may stay in sw queue, and we always flush all
belonging to same hw queue and dispatch them all to driver.
Unfortunately it is easy to cause queue busy because of the small
.cmd_per_lun. Once these requests are flushed out, they have to stay in
hctx->dispatch, and no bio merge can happen on these requests, and
sequential IO performance is harmed.
This patch introduces blk_mq_dequeue_from_ctx for dequeuing a request
from a sw queue, so that we can dispatch them in scheduler's way. We can
then avoid dequeueing too many requests from sw queue, since we don't
flush ->dispatch completely.
This patch improves dispatching from sw queue by using the .get_budget
and .put_budget callbacks.
Reviewed-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
For SCSI devices, there is often a per-request-queue depth, which needs
to be respected before queuing one request.
Currently blk-mq always dequeues the request first, then calls
.queue_rq() to dispatch the request to lld. One obvious issue with this
approach is that I/O merging may not be successful, because when the
per-request-queue depth can't be respected, .queue_rq() has to return
BLK_STS_RESOURCE, and then this request has to stay in hctx->dispatch
list. This means it never gets a chance to be merged with other IO.
This patch introduces .get_budget and .put_budget callback in blk_mq_ops,
then we can try to get reserved budget first before dequeuing request.
If the budget for queueing I/O can't be satisfied, we don't need to
dequeue request at all. Hence the request can be left in the IO
scheduler queue, for more merging opportunities.
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
A NULL pointer crash was reported for the case of having the BFQ IO
scheduler attached to the underlying blk-mq paths of a DM multipath
device. The crash occured in blk_mq_sched_insert_request()'s call to
e->type->ops.mq.insert_requests().
Paolo Valente correctly summarized why the crash occured with:
"the call chain (dm_mq_queue_rq -> map_request -> setup_clone ->
blk_rq_prep_clone) creates a cloned request without invoking
e->type->ops.mq.prepare_request for the target elevator e. The cloned
request is therefore not initialized for the scheduler, but it is
however inserted into the scheduler by blk_mq_sched_insert_request."
All said, a request-based DM multipath device's IO scheduler should be
the only one used -- when the original requests are issued to the
underlying paths as cloned requests they are inserted directly in the
underlying dispatch queue(s) rather than through an additional elevator.
But commit bd166ef18 ("blk-mq-sched: add framework for MQ capable IO
schedulers") switched blk_insert_cloned_request() from using
blk_mq_insert_request() to blk_mq_sched_insert_request(). Which
incorrectly added elevator machinery into a call chain that isn't
supposed to have any.
To fix this introduce a blk-mq private blk_mq_request_bypass_insert()
that blk_insert_cloned_request() calls to insert the request without
involving any elevator that may be attached to the cloned request's
request_queue.
Fixes: bd166ef183 ("blk-mq-sched: add framework for MQ capable IO schedulers")
Cc: stable@vger.kernel.org
Reported-by: Bart Van Assche <Bart.VanAssche@wdc.com>
Tested-by: Mike Snitzer <snitzer@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We don't have to inc/dec some counter, since we can just
iterate the tags. That makes inc/dec a noop, but means we
have to iterate busy tags to get an in-flight count.
Reviewed-by: Bart Van Assche <bart.vanassche@wdc.com>
Reviewed-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pull irq updates from Thomas Gleixner:
"The irq department delivers:
- Expand the generic infrastructure handling the irq migration on CPU
hotplug and convert X86 over to it. (Thomas Gleixner)
Aside of consolidating code this is a preparatory change for:
- Finalizing the affinity management for multi-queue devices. The
main change here is to shut down interrupts which are affine to a
outgoing CPU and reenabling them when the CPU comes online again.
That avoids moving interrupts pointlessly around and breaking and
reestablishing affinities for no value. (Christoph Hellwig)
Note: This contains also the BLOCK-MQ and NVME changes which depend
on the rework of the irq core infrastructure. Jens acked them and
agreed that they should go with the irq changes.
- Consolidation of irq domain code (Marc Zyngier)
- State tracking consolidation in the core code (Jeffy Chen)
- Add debug infrastructure for hierarchical irq domains (Thomas
Gleixner)
- Infrastructure enhancement for managing generic interrupt chips via
devmem (Bartosz Golaszewski)
- Constification work all over the place (Tobias Klauser)
- Two new interrupt controller drivers for MVEBU (Thomas Petazzoni)
- The usual set of fixes, updates and enhancements all over the
place"
* 'irq-core-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip: (112 commits)
irqchip/or1k-pic: Fix interrupt acknowledgement
irqchip/irq-mvebu-gicp: Allocate enough memory for spi_bitmap
irqchip/gic-v3: Fix out-of-bound access in gic_set_affinity
nvme: Allocate queues for all possible CPUs
blk-mq: Create hctx for each present CPU
blk-mq: Include all present CPUs in the default queue mapping
genirq: Avoid unnecessary low level irq function calls
genirq: Set irq masked state when initializing irq_desc
genirq/timings: Add infrastructure for estimating the next interrupt arrival time
genirq/timings: Add infrastructure to track the interrupt timings
genirq/debugfs: Remove pointless NULL pointer check
irqchip/gic-v3-its: Don't assume GICv3 hardware supports 16bit INTID
irqchip/gic-v3-its: Add ACPI NUMA node mapping
irqchip/gic-v3-its-platform-msi: Make of_device_ids const
irqchip/gic-v3-its: Make of_device_ids const
irqchip/irq-mvebu-icu: Add new driver for Marvell ICU
irqchip/irq-mvebu-gicp: Add new driver for Marvell GICP
dt-bindings/interrupt-controller: Add DT binding for the Marvell ICU
genirq/irqdomain: Remove auto-recursive hierarchy support
irqchip/MSI: Use irq_domain_update_bus_token instead of an open coded access
...
Currently we only create hctx for online CPUs, which can lead to a lot
of churn due to frequent soft offline / online operations. Instead
allocate one for each present CPU to avoid this and dramatically simplify
the code.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Cc: Keith Busch <keith.busch@intel.com>
Cc: linux-block@vger.kernel.org
Cc: linux-nvme@lists.infradead.org
Link: http://lkml.kernel.org/r/20170626102058.10200-3-hch@lst.de
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Move most code into blk_mq_rq_ctx_init, and the rest into
blk_mq_get_request.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Merge three functions only tail-called by blk_mq_free_request into
blk_mq_free_request.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Since the blk_mq_debugfs_*register_hctxs() functions register and
unregister all attributes under the "mq" directory, rename these
into blk_mq_debugfs_*register_mq().
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
A later patch will move the call of blk_mq_debugfs_register() to
a function to which the queue name is not passed as an argument.
To avoid having to add a 'name' argument to multiple callers, let
blk_mq_debugfs_register() look up the queue name.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Omar Sandoval <osandov@fb.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
A later patch in this series will modify blk_mq_debugfs_register()
such that it uses q->kobj.parent to determine the name of a
request queue. Hence make sure that that pointer is initialized
before blk_mq_debugfs_register() is called. To avoid lock inversion,
protect sysfs / debugfs registration with the queue sysfs_lock
instead of the global mutex all_q_mutex.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Wire up the sbitmap_get_shallow() operation to the tag code so that a
caller can limit the number of tags available to it.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
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.
Signed-off-by: Jens Axboe <axboe@fb.com>
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.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Currently, statistics are gathered in ~0.13s windows, and users grab the
statistics whenever they need them. This is not ideal for both in-tree
users:
1. Writeback throttling wants its own dynamically sized window of
statistics. Since the blk-stats statistics are reset after every
window and the wbt windows don't line up with the blk-stats windows,
wbt doesn't see every I/O.
2. Polling currently grabs the statistics on every I/O. Again, depending
on how the window lines up, we may miss some I/Os. It's also
unnecessary overhead to get the statistics on every I/O; the hybrid
polling heuristic would be just as happy with the statistics from the
previous full window.
This reworks the blk-stats infrastructure to be callback-based: users
register a callback that they want called at a given time with all of
the statistics from the window during which the callback was active.
Users can dynamically bucketize the statistics. wbt and polling both
currently use read vs. write, but polling can be extended to further
subdivide based on request size.
The callbacks are kept on an RCU list, and each callback has percpu
stats buffers. There will only be a few users, so the overhead on the
I/O completion side is low. The stats flushing is also simplified
considerably: since the timer function is responsible for clearing the
statistics, we don't have to worry about stale statistics.
wbt is a trivial conversion. After the conversion, the windowing problem
mentioned above is fixed.
For polling, we register an extra callback that caches the previous
window's statistics in the struct request_queue for the hybrid polling
heuristic to use.
Since we no longer have a single stats buffer for the request queue,
this also removes the sysfs and debugfs stats entries. To replace those,
we add a debugfs entry for the poll statistics.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Currently from kobject view, both q->mq_kobj and ctx->kobj can
be released during one cycle of blk_mq_register_dev() and
blk_mq_unregister_dev(). Actually, sw queue's lifetime is
same with its request queue's, which is covered by request_queue->kobj.
So we don't need to call kobject_put() for the two kinds of
kobject in __blk_mq_unregister_dev(), instead we do that
in release handler of request queue.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Tested-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Jens Axboe <axboe@fb.com>
When I added the blk-mq debugging information to debugfs, I didn't
notice that blktrace also creates a "block" directory in debugfs. Make
them use the same dentry, now created in the core block code. Based on a
patch from Jens.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
This fixes a couple of problems:
1. In the !CONFIG_DEBUG_FS case, the stub definitions were bogus.
2. In the !CONFIG_BLOCK case, blk-mq-debugfs.c shouldn't be compiled at
all.
Fix the stub definitions and add a CONFIG_BLK_DEBUG_FS Kconfig option.
Fixes: 07e4fead45 ("blk-mq: create debugfs directory tree")
Signed-off-by: Omar Sandoval <osandov@fb.com>
Augment Kconfig description.
Signed-off-by: Jens Axboe <axboe@fb.com>
Instead of letting the caller check this and handle the details
of inserting a flush request, put the logic in the scheduler
insertion function. This fixes direct flush insertion outside
of the usual make_request_fn calls, like from dm via
blk_insert_cloned_request().
Signed-off-by: Jens Axboe <axboe@fb.com>
If we have both multiple hardware queues and shared tag map between
devices, we need to ensure that we propagate the hardware queue
restart bit higher up. This is because we can get into a situation
where we don't have any IO pending on a hardware queue, yet we fail
getting a tag to start new IO. If that happens, it's not enough to
mark the hardware queue as needing a restart, we need to bubble
that up to the higher level queue as well.
Signed-off-by: Jens Axboe <axboe@fb.com>
Reviewed-by: Omar Sandoval <osandov@fb.com>
Tested-by: Hannes Reinecke <hare@suse.com>
In preparation for putting blk-mq debugging information in debugfs,
create a directory tree mirroring the one in sysfs:
# tree -d /sys/kernel/debug/block
/sys/kernel/debug/block
|-- nvme0n1
| `-- mq
| |-- 0
| | `-- cpu0
| |-- 1
| | `-- cpu1
| |-- 2
| | `-- cpu2
| `-- 3
| `-- cpu3
`-- vda
`-- mq
`-- 0
|-- cpu0
|-- cpu1
|-- cpu2
`-- cpu3
Also add the scaffolding for the actual files that will go in here,
either under the hardware queue or software queue directories.
Reviewed-by: Hannes Reinecke <hare@suse.com>
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
This adds a set of hooks that intercepts the blk-mq path of
allocating/inserting/issuing/completing requests, allowing
us to develop a scheduler within that framework.
We reuse the existing elevator scheduler API on the registration
side, but augment that with the scheduler flagging support for
the blk-mq interfce, and with a separate set of ops hooks for MQ
devices.
We split driver and scheduler tags, so we can run the scheduling
independently of device queue depth.
Signed-off-by: Jens Axboe <axboe@fb.com>
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Omar Sandoval <osandov@fb.com>
Prep patch for adding an extra tag map for scheduler requests.
Signed-off-by: Jens Axboe <axboe@fb.com>
Reviewed-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Omar Sandoval <osandov@fb.com>
This is in preparation for having another tag set available. Cleanup
the parameters, and allow passing in of tags for blk_mq_put_tag().
Signed-off-by: Jens Axboe <axboe@fb.com>
[hch: even more cleanups]
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Omar Sandoval <osandov@fb.com>
This update includes the usual round of major driver updates (ncr5380,
lpfc, hisi_sas, megaraid_sas, ufs, ibmvscsis, mpt3sas). There's also
an assortment of minor fixes, mostly in error legs or other not very
user visible stuff. The major change is the pci_alloc_irq_vectors
replacement for the old pci_msix_.. calls; this effectively makes IRQ
mapping generic for the drivers and allows blk_mq to use the
information.
Signed-off-by: James E.J. Bottomley <jejb@linux.vnet.ibm.com>
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2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=sPXh
-----END PGP SIGNATURE-----
Merge tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi
Pull SCSI updates from James Bottomley:
"This update includes the usual round of major driver updates (ncr5380,
lpfc, hisi_sas, megaraid_sas, ufs, ibmvscsis, mpt3sas).
There's also an assortment of minor fixes, mostly in error legs or
other not very user visible stuff. The major change is the
pci_alloc_irq_vectors replacement for the old pci_msix_.. calls; this
effectively makes IRQ mapping generic for the drivers and allows
blk_mq to use the information"
* tag 'scsi-misc' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi: (256 commits)
scsi: qla4xxx: switch to pci_alloc_irq_vectors
scsi: hisi_sas: support deferred probe for v2 hw
scsi: megaraid_sas: switch to pci_alloc_irq_vectors
scsi: scsi_devinfo: remove synchronous ALUA for NETAPP devices
scsi: be2iscsi: set errno on error path
scsi: be2iscsi: set errno on error path
scsi: hpsa: fallback to use legacy REPORT PHYS command
scsi: scsi_dh_alua: Fix RCU annotations
scsi: hpsa: use %phN for short hex dumps
scsi: hisi_sas: fix free'ing in probe and remove
scsi: isci: switch to pci_alloc_irq_vectors
scsi: ipr: Fix runaway IRQs when falling back from MSI to LSI
scsi: dpt_i2o: double free on error path
scsi: cxlflash: Migrate scsi command pointer to AFU command
scsi: cxlflash: Migrate IOARRIN specific routines to function pointers
scsi: cxlflash: Cleanup queuecommand()
scsi: cxlflash: Cleanup send_tmf()
scsi: cxlflash: Remove AFU command lock
scsi: cxlflash: Wait for active AFU commands to timeout upon tear down
scsi: cxlflash: Remove private command pool
...
Takes a list of requests, and dispatches it. Moves any residual
requests to the dispatch list.
Signed-off-by: Jens Axboe <axboe@fb.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
For legacy block, we simply track them in the request queue. For
blk-mq, we track them on a per-sw queue basis, which we can then
sum up through the hardware queues and finally to a per device
state.
The stats are tracked in, roughly, 0.1s interval windows.
Add sysfs files to display the stats.
The feature is off by default, to avoid any extra overhead. In-kernel
users of it can turn it on by setting QUEUE_FLAG_STATS in the queue
flags. We currently don't turn it on if someone just reads any of
the stats files, that is something we could add as well.
Signed-off-by: Jens Axboe <axboe@fb.com>
This will allow SCSI to have a single blk_mq_ops structure that either
lets the LLDD map the queues to PCIe MSIx vectors or use the default.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Multiple functions test the BLK_MQ_S_STOPPED bit so introduce
a helper function that performs this test.
Signed-off-by: Bart Van Assche <bart.vanassche@sandisk.com>
Reviewed-by: Ming Lei <tom.leiming@gmail.com>
Reviewed-by: Hannes Reinecke <hare@suse.com>
Reviewed-by: Johannes Thumshirn <jthumshirn@suse.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
Pull blk-mq CPU hotplug update from Jens Axboe:
"This is the conversion of blk-mq to the new hotplug state machine"
* 'for-4.9/block-smp' of git://git.kernel.dk/linux-block:
blk-mq: fixup "Convert to new hotplug state machine"
blk-mq: Convert to new hotplug state machine
blk-mq/cpu-notif: Convert to new hotplug state machine
Pull blk-mq irq/cpu mapping updates from Jens Axboe:
"This is the block-irq topic branch for 4.9-rc. It's mostly from
Christoph, and it allows drivers to specify their own mappings, and
more importantly, to share the blk-mq mappings with the IRQ affinity
mappings. It's a good step towards making this work better out of the
box"
* 'for-4.9/block-irq' of git://git.kernel.dk/linux-block:
blk_mq: linux/blk-mq.h does not include all the headers it depends on
blk-mq: kill unused blk_mq_create_mq_map()
blk-mq: get rid of the cpumask in struct blk_mq_tags
nvme: remove the post_scan callout
nvme: switch to use pci_alloc_irq_vectors
blk-mq: provide a default queue mapping for PCI device
blk-mq: allow the driver to pass in a queue mapping
blk-mq: remove ->map_queue
blk-mq: only allocate a single mq_map per tag_set
blk-mq: don't redistribute hardware queues on a CPU hotplug event
Replace the block-mq notifier list management with the multi instance
facility in the cpu hotplug state machine.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: linux-block@vger.kernel.org
Cc: rt@linutronix.de
Cc: Christoph Hellwing <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
Allocating your own per-cpu allocation hint separately makes for an
awkward API. Instead, allocate the per-cpu hint as part of the struct
sbitmap_queue. There's no point for a struct sbitmap_queue without the
cache, but you can still use a bare struct sbitmap.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
This is a generally useful data structure, so make it available to
anyone else who might want to use it. It's also a nice cleanup
separating the allocation logic from the rest of the tag handling logic.
The code is behind a new Kconfig option, CONFIG_SBITMAP, which is only
selected by CONFIG_BLOCK for now.
This should be a complete noop functionality-wise.
Signed-off-by: Omar Sandoval <osandov@fb.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
This allows drivers specify their own queue mapping by overriding the
setup-time function that builds the mq_map. This can be used for
example to build the map based on the MSI-X vector mapping provided
by the core interrupt layer for PCI devices.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
All drivers use the default, so provide an inline version of it. If we
ever need other queue mapping we can add an optional method back,
although supporting will also require major changes to the queue setup
code.
This provides better code generation, and better debugability as well.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
The hardware's provided queue count may change at runtime with resource
provisioning. This patch allows a block driver to alter the number of
h/w queues available when its resource count changes.
The main part is a new blk-mq API to request a new number of h/w queues
for a given live tag set. The new API freezes all queues using that set,
then adjusts the allocated count prior to remapping these to CPUs.
The bulk of the rest just shifts where h/w contexts and all their
artifacts are allocated and freed.
The number of max h/w contexts is capped to the number of possible cpus
since there is no use for more than that. As such, all pre-allocated
memory for pointers need to account for the max possible rather than
the initial number of queues.
A side effect of this is that the blk-mq will proceed successfully as
long as it can allocate at least one h/w context. Previously it would
fail request queue initialization if less than the requested number
was allocated.
Signed-off-by: Keith Busch <keith.busch@intel.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Tested-by: Jon Derrick <jonathan.derrick@intel.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
We already have the reserved flag, and a nowait flag awkwardly encoded as
a gfp_t. Add a real flags argument to make the scheme more extensible and
allow for a nicer calling convention.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
Notifier callbacks for CPU_ONLINE action can be run on the other CPU
than the CPU which was just onlined. So it is possible for the
process running on the just onlined CPU to insert request and run
hw queue before establishing new mapping which is done by
blk_mq_queue_reinit_notify().
This can cause a problem when the CPU has just been onlined first time
since the request queue was initialized. At this time ctx->index_hw
for the CPU, which is the index in hctx->ctxs[] for this ctx, is still
zero before blk_mq_queue_reinit_notify() is called by notifier
callbacks for CPU_ONLINE action.
For example, there is a single hw queue (hctx) and two CPU queues
(ctx0 for CPU0, and ctx1 for CPU1). Now CPU1 is just onlined and
a request is inserted into ctx1->rq_list and set bit0 in pending
bitmap as ctx1->index_hw is still zero.
And then while running hw queue, flush_busy_ctxs() finds bit0 is set
in pending bitmap and tries to retrieve requests in
hctx->ctxs[0]->rq_list. But htx->ctxs[0] is a pointer to ctx0, so the
request in ctx1->rq_list is ignored.
Fix it by ensuring that new mapping is established before onlined cpu
starts running.
Signed-off-by: Akinobu Mita <akinobu.mita@gmail.com>
Reviewed-by: Ming Lei <tom.leiming@gmail.com>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Ming Lei <tom.leiming@gmail.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
The kobject memory inside blk-mq hctx/ctx shouldn't have been freed
before the kobject is released because driver core can access it freely
before its release.
We can't do that in all ctx/hctx/mq_kobj's release handler because
it can be run before blk_cleanup_queue().
Given mq_kobj shouldn't have been introduced, this patch simply moves
mq's release into blk_release_queue().
Reported-by: Sasha Levin <sasha.levin@oracle.com>
Signed-off-by: Ming Lei <ming.lei@canonical.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
If it's dying, we can't expect new request to complete and come
in an wake up other tasks waiting for requests. So after we
have marked it as dying, wake up everybody currently waiting
for a request. Once they wake, they will retry their allocation
and fail appropriately due to the state of the queue.
Tested-by: Keith Busch <keith.busch@intel.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
When one hardware queue has no mapped software queues, it
shouldn't have been scheduled. Otherwise WARNING or OOPS
can triggered.
blk_mq_hw_queue_mapped() helper is introduce for fixing
the problem.
Signed-off-by: Ming Lei <ming.lei@canonical.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
These two temporary functions are introduced for holding flush
initialization and de-initialization, so that we can
introduce 'flush queue' easier in the following patch. And
once 'flush queue' and its allocation/free functions are ready,
they will be removed for sake of code readability.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <ming.lei@canonical.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
It is reasonable to allocate flush req in blk_mq_init_flush().
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <ming.lei@canonical.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Moved blk_mq_rq_timed_out() definition to the private blk-mq.h header.
Signed-off-by: Jens Axboe <axboe@fb.com>
blk_mq freezing is entangled with generic bypassing which bypasses
blkcg and io scheduler and lets IO requests fall through the block
layer to the drivers in FIFO order. This allows forward progress on
IOs with the advanced features disabled so that those features can be
configured or altered without worrying about stalling IO which may
lead to deadlock through memory allocation.
However, generic bypassing doesn't quite fit blk-mq. blk-mq currently
doesn't make use of blkcg or ioscheds and it maps bypssing to
freezing, which blocks request processing and drains all the in-flight
ones. This causes problems as bypassing assumes that request
processing is online. blk-mq works around this by conditionally
allowing request processing for the problem case - during queue
initialization.
Another weirdity is that except for during queue cleanup, bypassing
started on the generic side prevents blk-mq from processing new
requests but doesn't drain the in-flight ones. This shouldn't break
anything but again highlights that something isn't quite right here.
The root cause is conflating blk-mq freezing and generic bypassing
which are two different mechanisms. The only intersecting purpose
that they serve is during queue cleanup. Let's properly separate
blk-mq freezing from generic bypassing and simply use it where
necessary.
* request_queue->mq_freeze_depth is added and
blk_mq_[un]freeze_queue() now operate on this counter instead of
->bypass_depth. The replacement for QUEUE_FLAG_BYPASS isn't added
but the counter is tested directly. This will be further updated by
later changes.
* blk_mq_drain_queue() is dropped and "__" prefix is dropped from
blk_mq_freeze_queue(). Queue cleanup path now calls
blk_mq_freeze_queue() directly.
* blk_queue_enter()'s fast path condition is simplified to simply
check @q->mq_freeze_depth. Previously, the condition was
!blk_queue_dying(q) &&
(!blk_queue_bypass(q) || !blk_queue_init_done(q))
mq_freeze_depth is incremented right after dying is set and
blk_queue_init_done() exception isn't necessary as blk-mq doesn't
start frozen, which only leaves the blk_queue_bypass() test which
can be replaced by @q->mq_freeze_depth test.
This change simplifies the code and reduces confusion in the area.
Signed-off-by: Tejun Heo <tj@kernel.org>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Nicholas A. Bellinger <nab@linux-iscsi.org>
Signed-off-by: Jens Axboe <axboe@fb.com>
blk_mq_put_ctx() has to be called before io_schedule() in
bt_get().
This patch fixes the problem by taking similar approach from
percpu_ida allocation for the situation.
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Jens Axboe <axboe@fb.com>
Currently blk-mq registers all the hardware queues in sysfs,
regardless of whether it uses them (e.g. they have CPU mappings)
or not. The unused hardware queues lack the cpux/ directories,
and the other sysfs entries (like active, pending, etc) are all
zeroes.
Change this so that sysfs correctly reflects the current mappings
of the hardware queues.
Signed-off-by: Jens Axboe <axboe@fb.com>
Drivers currently have to figure this out on their own, and they
are missing information to do it properly. The ones that did
attempt to do it, do it wrong.
So just pass in the suggested node directly to the alloc
function.
Signed-off-by: Jens Axboe <axboe@fb.com>
Prepare this for the next patch which adds more smarts in the
plugging logic, so that we can save some memory.
Signed-off-by: Jens Axboe <axboe@fb.com>
For request_fn based devices, the block layer exports a 'nr_requests'
file through sysfs to allow adjusting of queue depth on the fly.
Currently this returns -EINVAL for blk-mq, since it's not wired up.
Wire this up for blk-mq, so that it now also always dynamic
adjustments of the allowed queue depth for any given block device
managed by blk-mq.
Signed-off-by: Jens Axboe <axboe@fb.com>
blk-mq currently uses percpu_ida for tag allocation. But that only
works well if the ratio between tag space and number of CPUs is
sufficiently high. For most devices and systems, that is not the
case. The end result if that we either only utilize the tag space
partially, or we end up attempting to fully exhaust it and run
into lots of lock contention with stealing between CPUs. This is
not optimal.
This new tagging scheme is a hybrid bitmap allocator. It uses
two tricks to both be SMP friendly and allow full exhaustion
of the space:
1) We cache the last allocated (or freed) tag on a per blk-mq
software context basis. This allows us to limit the space
we have to search. The key element here is not caching it
in the shared tag structure, otherwise we end up dirtying
more shared cache lines on each allocate/free operation.
2) The tag space is split into cache line sized groups, and
each context will start off randomly in that space. Even up
to full utilization of the space, this divides the tag users
efficiently into cache line groups, avoiding dirtying the same
one both between allocators and between allocator and freeer.
This scheme shows drastically better behaviour, both on small
tag spaces but on large ones as well. It has been tested extensively
to show better performance for all the cases blk-mq cares about.
Signed-off-by: Jens Axboe <axboe@fb.com>
The blk-mq code is using it's own version of the I/O completion affinity
tunables, which causes a few issues:
- the rq_affinity sysfs file doesn't work for blk-mq devices, even if it
still is present, thus breaking existing tuning setups.
- the rq_affinity = 1 mode, which is the defauly for legacy request based
drivers isn't implemented at all.
- blk-mq drivers don't implement any completion affinity with the default
flag settings.
This patches removes the blk-mq ipi_redirect flag and sysfs file, as well
as the internal BLK_MQ_F_SHOULD_IPI flag and replaces it with code that
respects the queue-wide rq_affinity flags and also implements the
rq_affinity = 1 mode.
This means I/O completion affinity can now only be tuned block-queue wide
instead of per context, which seems more sensible to me anyway.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
If a requeue event races with a timeout, we can get into the
situation where we attempt to complete a request from the
timeout handler when it's not start anymore. This causes a crash.
So have the timeout handler check that REQ_ATOM_STARTED is still
set on the request - if not, we ignore the event. If this happens,
the request has now been marked as complete. As a consequence, we
need to ensure to clear REQ_ATOM_COMPLETE in blk_mq_start_request(),
as to maintain proper request state.
Signed-off-by: Jens Axboe <axboe@fb.com>
Add a new blk_mq_tag_set structure that gets set up before we initialize
the queue. A single blk_mq_tag_set structure can be shared by multiple
queues.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Modular export of blk_mq_{alloc,free}_tagset added by me.
Signed-off-by: Jens Axboe <axboe@fb.com>
Drivers shouldn't have to care about the block layer setting aside a
request to implement the flush state machine. We already override the
mq context and tag to make it more transparent, but so far haven't deal
with the driver private data in the request. Make sure to override this
as well, and while we're at it add a proper helper sitting in blk-mq.c
that implements the full impersonation.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
Drivers can reach their private data easily using the blk_mq_rq_to_pdu
helper and don't need req->special. By not initializing it code can
be simplified nicely, and we also shave off a few more instructions from
the I/O path.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
It's almost identical to blk_mq_insert_request, so fold the two into one
slightly more generic function by making the flush special case a bit
smarted.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
Now that we are out of initial debug/bringup mode, remove
the verbose dump of the mapping table.
Provide the mapping table in sysfs, under the hardware queue
directory, in the cpu_list file.
Signed-off-by: Jens Axboe <axboe@fb.com>
Witch to using a preallocated flush_rq for blk-mq similar to what's done
with the old request path. This allows us to set up the request properly
with a tag from the actually allowed range and ->rq_disk as needed by
some drivers. To make life easier we also switch to dynamic allocation
of ->flush_rq for the old path.
This effectively reverts most of
"blk-mq: fix for flush deadlock"
and
"blk-mq: Don't reserve a tag for flush request"
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
Rework I/O completions to work more like the old code path. blk_mq_end_io
now stays out of the business of deferring completions to others CPUs
and calling blk_mark_rq_complete. The latter is very important to allow
completing requests that have timed out and thus are already marked completed,
the former allows using the IPI callout even for driver specific completions
instead of having to reimplement them.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@fb.com>
__smp_call_function_single already avoids multiple IPIs by internally
queing up the items, and now also is available for non-SMP builds as
a trivially correct stub, so there is no need to wrap it. If the
additional lock roundtrip cause problems my patch to convert the
generic IPI code to llists is waiting to get merged will fix it.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
blk_mq_free_queue() is called from release handler of
queue kobject, so it needn't be called from drivers.
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
blk_mq_drain_queue() is introduced so that we can drain
mq queue inside blk_cleanup_queue().
Also don't accept new requests any more if queue is marked
as dying.
Cc: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ming Lei <tom.leiming@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Linux currently has two models for block devices:
- The classic request_fn based approach, where drivers use struct
request units for IO. The block layer provides various helper
functionalities to let drivers share code, things like tag
management, timeout handling, queueing, etc.
- The "stacked" approach, where a driver squeezes in between the
block layer and IO submitter. Since this bypasses the IO stack,
driver generally have to manage everything themselves.
With drivers being written for new high IOPS devices, the classic
request_fn based driver doesn't work well enough. The design dates
back to when both SMP and high IOPS was rare. It has problems with
scaling to bigger machines, and runs into scaling issues even on
smaller machines when you have IOPS in the hundreds of thousands
per device.
The stacked approach is then most often selected as the model
for the driver. But this means that everybody has to re-invent
everything, and along with that we get all the problems again
that the shared approach solved.
This commit introduces blk-mq, block multi queue support. The
design is centered around per-cpu queues for queueing IO, which
then funnel down into x number of hardware submission queues.
We might have a 1:1 mapping between the two, or it might be
an N:M mapping. That all depends on what the hardware supports.
blk-mq provides various helper functions, which include:
- Scalable support for request tagging. Most devices need to
be able to uniquely identify a request both in the driver and
to the hardware. The tagging uses per-cpu caches for freed
tags, to enable cache hot reuse.
- Timeout handling without tracking request on a per-device
basis. Basically the driver should be able to get a notification,
if a request happens to fail.
- Optional support for non 1:1 mappings between issue and
submission queues. blk-mq can redirect IO completions to the
desired location.
- Support for per-request payloads. Drivers almost always need
to associate a request structure with some driver private
command structure. Drivers can tell blk-mq this at init time,
and then any request handed to the driver will have the
required size of memory associated with it.
- Support for merging of IO, and plugging. The stacked model
gets neither of these. Even for high IOPS devices, merging
sequential IO reduces per-command overhead and thus
increases bandwidth.
For now, this is provided as a potential 3rd queueing model, with
the hope being that, as it matures, it can replace both the classic
and stacked model. That would get us back to having just 1 real
model for block devices, leaving the stacked approach to dm/md
devices (as it was originally intended).
Contributions in this patch from the following people:
Shaohua Li <shli@fusionio.com>
Alexander Gordeev <agordeev@redhat.com>
Christoph Hellwig <hch@infradead.org>
Mike Christie <michaelc@cs.wisc.edu>
Matias Bjorling <m@bjorling.me>
Jeff Moyer <jmoyer@redhat.com>
Acked-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Tejun Heo