This generates a lot better code for me, and bumps performance from
7650K IOPS to 7750K IOPS. Looking at profiles for the run and running
perf diff, it confirms that we're now sending a lot less time there:
6.38% -2.80% [kernel.vmlinux] [k] blkdev_direct_IO
Taking it from the 2nd most cycle consumer to only the 9th most at
3.35% of the CPU time.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
bdev = &BDEV_I(file->f_mapping->host)->bdev
Getting struct block_device from a file requires 2 memory dereferences
as illustrated above, that takes a toll on performance, so cache it in
yet unused file->private_data. That gives a noticeable peak performance
improvement.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/8415f9fe12e544b9da89593dfbca8de2b52efe03.1634115360.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Set the poll queue flag to enable polling, given that the multipath
node just dispatches the bios to a lower queue.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-17-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The poll attribute is a historic artefact from before when we had
explicit poll queues that require driver specific configuration.
Just print a warning when writing to the attribute.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-16-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>
This flags ensures that the pages will not be reused for non-bio
allocations before the end of an RCU grace period. With that we can
safely use a RCU lookup for bio polling as long as we are fine with
occasionally polling the wrong device.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-13-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Unlike the RWF_HIPRI userspace ABI which is intentionally kept vague,
the bio flag is specific to the polling implementation, so rename and
document it properly.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-12-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There is no point in sleeping for the expected I/O completion timeout
in the io_uring async polling model as we never poll for a specific
I/O.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-11-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Switch the boolean spin argument to blk_poll to passing a set of flags
instead. This will allow to control polling behavior in a more fine
grained way.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-10-hch@lst.de
[axboe: adapt to changed io_uring iopoll]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Move the trivial check into the only caller.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Reviewed-by: Chaitanya Kulkarni <chaitanya.kulkarni@wdc.com>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-9-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Merge both functions into their only caller to keep the blk-mq tag to
blk_qc_t mapping as private as possible in blk-mq.c.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-8-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Factor the code to do the classic full metal polling out of blk_poll into
a separate blk_mq_poll_classic helper.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-7-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a helper to get the hctx from a request_queue and cookie, and fold
the blk_qc_t_to_queue_num helper into it as no other callers are left.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Sagi Grimberg <sagi@grimberg.me>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-6-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
syscall-level code can't just poke into the details of the poll cookie,
which is private information of the block layer.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211012111226.760968-5-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If an iocb is split into multiple bios we can't poll for both. So don't
bother to even try to poll in that case.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-4-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If an iocb is split into multiple bios we can't poll for both. So don't
even bother to try to poll in that case.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211012111226.760968-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The polling support in the legacy direct-io support is a little crufty.
It already doesn't support the asynchronous polling needed for io_uring
polling, and is hard to adopt to upcoming changes in the polling
interfaces. Given that all the major file systems already use the iomap
direct I/O code, just drop the polling support.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Tested-by: Mark Wunderlich <mark.wunderlich@intel.com>
Link: https://lore.kernel.org/r/20211012111226.760968-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Currently we scan the entire plug list, which is potentially very
expensive. In an IOPS bound workload, we can drive about 5.6M IOPS with
merging enabled, and profiling shows that the plug merge check is the
(by far) most expensive thing we're doing:
Overhead Command Shared Object Symbol
+ 20.89% io_uring [kernel.vmlinux] [k] blk_attempt_plug_merge
+ 4.98% io_uring [kernel.vmlinux] [k] io_submit_sqes
+ 4.78% io_uring [kernel.vmlinux] [k] blkdev_direct_IO
+ 4.61% io_uring [kernel.vmlinux] [k] blk_mq_submit_bio
Instead of browsing the whole list, just check the previously inserted
entry. That is enough for a naive merge check and will catch most cases,
and for devices that need full merging, the IO scheduler attached to
such devices will do that anyway. The plug merge is meant to be an
inexpensive check to avoid getting a request, but if we repeatedly
scan the list for every single insert, it is very much not a cheap
check.
With this patch, the workload instead runs at ~7.0M IOPS, providing
a 25% improvement. Disabling merging entirely yields another 5%
improvement.
Reviewed-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Every object under block/ depends on CONFIG_BLOCK.
Move the guard to the top Makefile since there is no point to
descend into block/ if CONFIG_BLOCK=n.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20210927140000.866249-5-masahiroy@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Everything under block/ depends on BLOCK. BLOCK_HOLDER_DEPRECATED is
selected from drivers/md/Kconfig, which is entirely dependent on BLOCK.
Extend the 'if BLOCK' ... 'endif' so it covers the whole block/Kconfig.
Also, clean up the definition of BLOCK_COMPAT and BLK_MQ_PCI because
COMPAT and PCI are boolean.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20210927140000.866249-3-masahiroy@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
CONFIG_BLK_CGROUP is a boolean option, that is, its value is 'y' or 'n'.
The comparison to 'y' is redundant.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20210927140000.866249-2-masahiroy@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a blk_mq_get_tags() helper, which uses the new sbitmap API for
allocating a batch of tags all at once. This both simplifies the block
code for batched allocation, and it is also more efficient than just
doing repeated calls into __sbitmap_queue_get().
This reduces the sbitmap overhead in peak runs from ~3% to ~1% and
yields a performanc increase from 6.6M IOPS to 6.8M IOPS for a single
CPU core.
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>
We already have a blk_mq_need_time_stamp() check in
__blk_mq_end_request() to get a timestamp, hide all the statistics
accounting under it. It cuts some cycles for requests that don't need
stats, and is free otherwise.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/e0f2ea812e93a8adcd07101212e7d7e70ca304e7.1634115360.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
bio_get_first_bvec and bio_get_last_bvec are only used in blk-merge.c,
so move them there.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211012161804.991559-8-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Mark __bio_try_merge_page static and move it up a bit to avoid the need
for a forward declaration.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211012161804.991559-7-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
BIO_DEBUG is always defined, so just switch the two instances to use
BUG_ON directly.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211012161804.991559-2-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We could have a race here, where the request gets freed before we call
into blk_mq_run_hw_queue(). If this happens, we cannot rely on the state
of the request.
Grab the hardware context before inserting the flush.
Fixes: 0f38d76646 ("blk-mq: cleanup blk_mq_submit_bio")
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Move the blk_mq_alloc_data stack allocation only into the branch
that actually needs it, and use rq->mq_hctx instead of data.hctx
to refer to the hctx.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211012104045.658051-3-hch@lst.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The newly added loop for the cached requests in __blk_mq_alloc_request
is a little too convoluted for my taste, so unwind it a bit. Also
rename the function to __blk_mq_alloc_requests now that it can allocate
more than a single request.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Link: https://lore.kernel.org/r/20211012104045.658051-2-hch@lst.de
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>
Doing high IOPS testing with blk-cgroups enabled spends ~15-20% of the
time just doing ktime_get_ns() -> readtsc. We essentially read and
set the start time twice, one for the bio and then again when that bio
is mapped to a request.
Given that the time between the two is very short, inherit the bio
start time instead of reading it again. This cuts 1/3rd of the overhead
of the time keeping.
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Even if no policies are defined, we spend ~2% of the total IO time
checking. Move the fast path inline.
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now that we use shared tags for shared sbitmap support, we don't require
the tags sbitmap pointers, so drop them.
This essentially reverts commit 222a5ae03c ("blk-mq: Use pointers for
blk_mq_tags bitmap tags").
Function blk_mq_init_bitmap_tags() is removed also, since it would be only
a wrappper for blk_mq_init_bitmaps().
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Signed-off-by: John Garry <john.garry@huawei.com>
Link: https://lore.kernel.org/r/1633429419-228500-14-git-send-email-john.garry@huawei.com
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>
Refactor blk_mq_free_map_and_requests() such that it can be used at many
sites at which the tag map and rqs are freed.
Also rename to blk_mq_free_map_and_rqs(), which is shorter and matches the
alloc equivalent.
Suggested-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: John Garry <john.garry@huawei.com>
Reviewed-by: Hannes Reinecke <hare@suse.de>
Link: https://lore.kernel.org/r/1633429419-228500-12-git-send-email-john.garry@huawei.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Jens Axboe