The Concurrent Positioning Ranges VPD page (for SCSI) and data log page
(for ATA) contain parameters describing the set of contiguous LBAs that
can be served independently by a single LUN multi-actuator hard-disk.
Similarly, a logically defined block device composed of multiple disks
can in some cases execute requests directed at different sector ranges
in parallel. A dm-linear device aggregating 2 block devices together is
an example.
This patch implements support for exposing a block device independent
access ranges to the user through sysfs to allow optimizing device
accesses to increase performance.
To describe the set of independent sector ranges of a device (actuators
of a multi-actuator HDDs or table entries of a dm-linear device),
The type struct blk_independent_access_ranges is introduced. This
structure describes the sector ranges using an array of
struct blk_independent_access_range structures. This range structure
defines the start sector and number of sectors of the access range.
The ranges in the array cannot overlap and must contain all sectors
within the device capacity.
The function disk_set_independent_access_ranges() allows a device
driver to signal to the block layer that a device has multiple
independent access ranges. In this case, a struct
blk_independent_access_ranges is attached to the device request queue
by the function disk_set_independent_access_ranges(). The function
disk_alloc_independent_access_ranges() is provided for drivers to
allocate this structure.
struct blk_independent_access_ranges contains kobjects (struct kobject)
to expose to the user through sysfs the set of independent access ranges
supported by a device. When the device is initialized, sysfs
registration of the ranges information is done from blk_register_queue()
using the block layer internal function
disk_register_independent_access_ranges(). If a driver calls
disk_set_independent_access_ranges() for a registered queue, e.g. when a
device is revalidated, disk_set_independent_access_ranges() will execute
disk_register_independent_access_ranges() to update the sysfs attribute
files. The sysfs file structure created starts from the
independent_access_ranges sub-directory and contains the start sector
and number of sectors of each range, with the information for each range
grouped in numbered sub-directories.
E.g. for a dual actuator HDD, the user sees:
$ tree /sys/block/sdk/queue/independent_access_ranges/
/sys/block/sdk/queue/independent_access_ranges/
|-- 0
| |-- nr_sectors
| `-- sector
`-- 1
|-- nr_sectors
`-- sector
For a regular device with a single access range, the
independent_access_ranges sysfs directory does not exist.
Device revalidation may lead to changes to this structure and to the
attribute values. When manipulated, the queue sysfs_lock and
sysfs_dir_lock mutexes are held for atomicity, similarly to how the
blk-mq and elevator sysfs queue sub-directories are protected.
The code related to the management of independent access ranges is
added in the new file block/blk-ia-ranges.c.
Signed-off-by: Damien Le Moal <damien.lemoal@wdc.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Reviewed-by: Keith Busch <kbusch@kernel.org>
Link: https://lore.kernel.org/r/20211027022223.183838-2-damien.lemoal@wdc.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If CONFIG_BLOCK isn't set, then it's an empty struct anyway. Just make
it generally available, so we don't break the compile:
kernel/sched/core.c: In function ‘sched_submit_work’:
kernel/sched/core.c:6346:35: error: ‘struct task_struct’ has no member named ‘plug’
6346 | blk_flush_plug(tsk->plug, true);
| ^~
kernel/sched/core.c: In function ‘io_schedule_prepare’:
kernel/sched/core.c:8357:20: error: ‘struct task_struct’ has no member named ‘plug’
8357 | if (current->plug)
| ^~
kernel/sched/core.c:8358:39: error: ‘struct task_struct’ has no member named ‘plug’
8358 | blk_flush_plug(current->plug, true);
| ^~
Reported-by: Nathan Chancellor <nathan@kernel.org>
Fixes: 008f75a20e ("block: cleanup the flush plug helpers")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
blk_keyslot_manager is misnamed because it doesn't necessarily manage
keyslots. It actually does several different things:
- Contains the crypto capabilities of the device.
- Provides functions to control the inline encryption hardware.
Originally these were just for programming/evicting keyslots;
however, new functionality (hardware-wrapped keys) will require new
functions here which are unrelated to keyslots. Moreover,
device-mapper devices already (ab)use "keyslot_evict" to pass key
eviction requests to their underlying devices even though
device-mapper devices don't have any keyslots themselves (so it
really should be "evict_key", not "keyslot_evict").
- Sometimes (but not always!) it manages keyslots. Originally it
always did, but device-mapper devices don't have keyslots
themselves, so they use a "passthrough keyslot manager" which
doesn't actually manage keyslots. This hack works, but the
terminology is unnatural. Also, some hardware doesn't have keyslots
and thus also uses a "passthrough keyslot manager" (support for such
hardware is yet to be upstreamed, but it will happen eventually).
Let's stop having keyslot managers which don't actually manage keyslots.
Instead, rename blk_keyslot_manager to blk_crypto_profile.
This is a fairly big change, since for consistency it also has to update
keyslot manager-related function names, variable names, and comments --
not just the actual struct name. However it's still a fairly
straightforward change, as it doesn't change any actual functionality.
Acked-by: Ulf Hansson <ulf.hansson@linaro.org> # For MMC
Reviewed-by: Mike Snitzer <snitzer@redhat.com>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Eric Biggers <ebiggers@google.com>
Link: https://lore.kernel.org/r/20211018180453.40441-4-ebiggers@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Consolidate the various helpers into a single blk_flush_plug helper that
takes a plk_plug and the from_scheduler bool and switch all callsites to
call it directly. Checks that the plug is non-NULL must be performed by
the caller, something that most already do anyway.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211020144119.142582-5-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Inline BIO_NO_PAGE_REF check of bio_release_pages() to avoid function
call.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Replace bio_set_dev() with an identical inline helper and move it
further to fix a dependency problem with bio_associate_blkg(). Do the
same for bio_copy_dev().
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we have just one queue type in the plug list, then we can extend our
direct issue to cover a full plug list as well. This allows sending a
batch of requests for direct issue, which is more efficient than doing
one-at-a-time kind of issue.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Use a singly linked list for the blk_plug. This saves 8 bytes in the
blk_plug struct, and makes for faster list manipulations than doubly
linked lists. As we don't use the doubly linked lists for anything,
singly linked is just fine.
This yields a bump in default (merging enabled) performance from 7.0
to 7.1M IOPS, and ~7.5M IOPS with merging disabled.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This is in the fast path of driver issue or completion, and it's a single
array index operation. Move it inline to avoid a function call for it.
This does mean making struct blk_mq_tags block layer public, but there's
not really much in there.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We only need to call it to resolve the blk_status_t -> errno mapping for
tracing, so move the conversion into the tracepoints that are not called
at all when tracing isn't enabled.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This is called for every write in the fast path, move it inline next
to get_disk_ro() which is called internally.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Instead of calling blk_mq_end_request() on a single request, add a helper
that takes the new struct io_comp_batch and completes any request stored
in there.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
sbitmap currently only supports clearing tags one-by-one, add a helper
that allows the caller to pass in an array of tags to clear.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
struct io_comp_batch contains a list head and a completion handler, which
will allow completions to more effciently completed batches of IO.
For now, no functional changes in this patch, we just define the
io_comp_batch structure and add the argument to the file_operations iopoll
handler.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Instead of open-coding the list additions, traversal, and removal,
provide a basic set of helpers.
Suggested-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Just like the blk_mq_ctx counterparts, we've got a bunch of counters
in here that are only for debugfs and are of questionnable value. They
are:
- dispatched, index of how many requests were dispatched in one go
- poll_{considered,invoked,success}, which track poll sucess rates. We're
confident in the iopoll implementation at this point, don't bother
tracking these.
As a bonus, this shrinks each hardware queue from 576 bytes to 512 bytes,
dropping a whole cacheline.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add an rq private RQF_ELV flag, which tells the block layer that this
request was initialized on a queue that has an IO scheduler attached.
This allows for faster checking in the fast path, rather than having to
deference rq->q later on.
Elevator switching does full quiesce of the queue before detaching an
IO scheduler, so it's safe to cache this in the request itself.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
It's been a while since this was analyzed, move some members around to
better flow with the use case. Initial state up top, and queued state
after that. This improves my peak case by about 1.5%, from 7750K to
7900K IOPS.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we're completing nbytes and nbytes is the size of the bio, don't bother
with calling into the iterator increment helpers. Just clear the bio
size and we're done.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Replace the blk_poll interface that requires the caller to keep a queue
and cookie from the submissions with polling based on the bio.
Polling for the bio itself leads to a few advantages:
- the cookie construction can made entirely private in blk-mq.c
- the caller does not need to remember the request_queue and cookie
separately and thus sidesteps their lifetime issues
- keeping the device and the cookie inside the bio allows to trivially
support polling BIOs remapping by stacking drivers
- a lot of code to propagate the cookie back up the submission path can
be removed entirely.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-15-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
'struct bvec_iter' is embedded into 'struct bio', define it as packed
so that we can get one extra 4bytes for other uses without expanding
bio.
'struct bvec_iter' is often allocated on stack, so making it packed
doesn't affect performance. Also I have run io_uring on both
nvme/null_blk, and not observe performance effect in this way.
Suggested-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-14-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The block layer tag allocation batching still calls into sbitmap to get
each tag, but we can improve on that. Add __sbitmap_queue_get_batch(),
which returns a mask of tags all at once, along with an offset for
those tags.
An example return would be 0xff, where bits 0..7 are set, with
tag_offset == 128. The valid tags in this case would be 128..135.
A batch is specific to an individual sbitmap_map, hence it cannot be
larger than that. The requested number of tags is automatically reduced
to the max that can be satisfied with a single map.
On failure, 0 is returned. Caller should fall back to single tag
allocation at that point/
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The caller typically has a good (or even exact) idea of how many requests
it needs to submit. We can make the request/tag allocation a lot more
efficient if we just allocate N requests/tags upfront when we queue the
first bio from the batch.
Provide a new plug start helper that allows the caller to specify how many
IOs are expected. This sets plug->nr_ios, and we can use that for smarter
request allocation. The plug provides a holding spot for requests, and
request allocation will check it before calling into the normal request
allocation path.
The blk_finish_plug() is called, check if there are unused requests and
free them. This should not happen in normal operations. The exception is
if we get merging, then we may be left with requests that need freeing
when done.
This raises the per-core performance on my setup from ~5.8M to ~6.1M
IOPS.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Particularly for NVMe with efficient deferred submission for many
requests, there are nice benefits to be seen by bumping the default max
plug count from 16 to 32. This is especially true for virtualized setups,
where the submit part is more expensive. But can be noticed even on
native hardware.
Reduce the multiple queue factor from 4 to 2, since we're changing the
default size.
While changing it, move the defines into the block layer private header.
These aren't values that anyone outside of the block layer uses, or
should use.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Currently we use separate sbitmap pairs and active_queues atomic_t for
shared sbitmap support.
However a full sets of static requests are used per HW queue, which is
quite wasteful, considering that the total number of requests usable at
any given time across all HW queues is limited by the shared sbitmap depth.
As such, it is considerably more memory efficient in the case of shared
sbitmap to allocate a set of static rqs per tag set or request queue, and
not per HW queue.
So replace the sbitmap pairs and active_queues atomic_t with a shared
tags per tagset and request queue, which will hold a set of shared static
rqs.
Since there is now no valid HW queue index to be passed to the blk_mq_ops
.init and .exit_request callbacks, pass an invalid index token. This
changes the semantics of the APIs, such that the callback would need to
validate the HW queue index before using it. Currently no user of shared
sbitmap actually uses the HW queue index (as would be expected).
Signed-off-by: John Garry <john.garry@huawei.com>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Link: https://lore.kernel.org/r/1633429419-228500-13-git-send-email-john.garry@huawei.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
