With IORING_FEAT_FAST_POLL in place, io_put_req_find_next() for poll
requests doesn't make much sense, and in any case re-adding it
shouldn't be a problem considering batching in tctx_task_work(). We can
remove it.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/15699682bf81610ec901d4e79d6da64baa9f70be.1639605189.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
In previous patches, we have already gathered some tw with
io_req_task_complete() as callback in prior_task_list, let's complete
them in batch while we cannot grab uring lock. In this way, we batch
the req_complete_post path.
Signed-off-by: Hao Xu <haoxu@linux.alibaba.com>
Link: https://lore.kernel.org/r/20211208052125.351587-1-haoxu@linux.alibaba.com
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now we have a lot of task_work users, some are just to complete a req
and generate a cqe. Let's put the work to a new tw list which has a
higher priority, so that it can be handled quickly and thus to reduce
avg req latency and users can issue next round of sqes earlier.
An explanatory case:
origin timeline:
submit_sqe-->irq-->add completion task_work
-->run heavy work0~n-->run completion task_work
now timeline:
submit_sqe-->irq-->add completion task_work
-->run completion task_work-->run heavy work0~n
Limitation: this optimization is only for those that submission and
reaping process are in different threads. Otherwise anyhow we have to
submit new sqes after returning to userspace, then the order of TWs
doesn't matter.
Tested this patch(and the following ones) by manually replace
__io_queue_sqe() in io_queue_sqe() by io_req_task_queue() to construct
'heavy' task works. Then test with fio:
ioengine=io_uring
sqpoll=1
thread=1
bs=4k
direct=1
rw=randread
time_based=1
runtime=600
randrepeat=0
group_reporting=1
filename=/dev/nvme0n1
Tried various iodepth.
The peak IOPS for this patch is 710K, while the old one is 665K.
For avg latency, difference shows when iodepth grow:
depth and avg latency(usec):
depth new old
1 7.05 7.10
2 8.47 8.60
4 10.42 10.42
8 13.78 13.22
16 27.41 24.33
32 49.40 53.08
64 102.53 103.36
128 196.98 205.61
256 372.99 414.88
512 747.23 791.30
1024 1472.59 1538.72
2048 3153.49 3329.01
4096 6387.86 6682.54
8192 12150.25 12774.14
16384 23085.58 26044.71
Signed-off-by: Hao Xu <haoxu@linux.alibaba.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/20211207093951.247840-3-haoxu@linux.alibaba.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When iopolling the userspace specifies the minimum number of "events" it
expects. Previously, we had one CQE per request, so the definition of
an "event" was unequivocal, but that's not more the case anymore with
REQ_F_CQE_SKIP.
Currently it counts the number of completed requests, replace it with
the number of posted CQEs. This allows users of the "one CQE per link"
scheme to wait for all N links in a single syscall, which is not
possible without the patch and requires extra context switches.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/d5a965c4d2249827392037bbd0186f87fea49c55.1638714983.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Current IOSQE_IO_DRAIN implementation doesn't work well with CQE
skipping and it's not allowed, otherwise some requests might be not
executed until the ring is destroyed and the userspace would hang.
Let's fail all drain requests after seeing IOSQE_CQE_SKIP_SUCCESS at
least once. All drained requests prior to that will get run normally,
so there should be no stalls. However, even though such mixing wouldn't
lead to issues at the moment, it's still not allowed as the behaviour
may change.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/bcf7164f8bf3eb54b7bb7b4fd119907fa4d4d43b.1636559119.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Emitting a CQE is expensive from the kernel perspective. Often, it's
also not convenient for the userspace, spends some cycles on processing
and just complicates the logic. A similar problems goes for linked
requests, where we post an CQE for each request in the link.
Introduce a new flags, IOSQE_CQE_SKIP_SUCCESS, trying to help with it.
When set and a request completed successfully, it won't generate a CQE.
When fails, it produces an CQE, but all following linked requests will
be CQE-less, regardless whether they have IOSQE_CQE_SKIP_SUCCESS or not.
The notion of "fail" is the same as for link failing-cancellation, where
it's opcode dependent, and _usually_ result >= 0 is a success, but not
always.
Linked timeouts are a bit special. When the requests it's linked to was
not attempted to be executed, e.g. failing linked requests, it follows
the description above. Otherwise, whether a linked timeout will post a
completion or not solely depends on IOSQE_CQE_SKIP_SUCCESS of that
linked timeout request. Linked timeout never "fail" during execution, so
for them it's unconditional. It's expected for users to not really care
about the result of it but rely solely on the result of the master
request. Another reason for such a treatment is that it's racy, and the
timeout callback may be running awhile the master request posts its
completion.
use case 1:
If one doesn't care about results of some requests, e.g. normal
timeouts, just set IOSQE_CQE_SKIP_SUCCESS. Error result will still be
posted and need to be handled.
use case 2:
Set IOSQE_CQE_SKIP_SUCCESS for all requests of a link but the last,
and it'll post a completion only for the last one if everything goes
right, otherwise there will be one only one CQE for the first failed
request.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/0220fbe06f7cf99e6fc71b4297bb1cb6c0e89c2c.1636559119.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Split io_cqring_fill_event() into a couple of more targeted functions.
The first on is io_fill_cqe_aux() for completions that are not
associated with request completions and doing the ->cq_extra accounting.
Examples are additional CQEs from multishot poll and rsrc notifications.
The second is io_fill_cqe_req(), should be called when it's a normal
request completion. Nothing more to it at the moment, will be used in
later patches.
The last one is inlined __io_fill_cqe() for a finer grained control,
should be used with caution and in hottest places.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/59a9117a4a44fc9efcf04b3afa51e0d080f5943c.1636559119.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Simplify failed resubmission prep in kiocb_done(), it's a bit ugly with
conditional logic and hand handling cflags / select buffers. Instead,
punt to tw and use io_req_task_complete() already handling all the
cases.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/667c33484b05b612e9420e1b1d5f4dc46d0ee9ce.1637524285.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When we pass in zero as an io-wq worker number limit it shouldn't
actually change the limits but return the old value, follow that
behaviour with deferred limits setup as well.
Cc: stable@kernel.org # 5.15
Reported-by: Beld Zhang <beldzhang@gmail.com>
Fixes: e139a1ec92 ("io_uring: apply max_workers limit to all future users")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/1b222a92f7a78a24b042763805e891a4cdd4b544.1636384034.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The kernel test robot correctly identifies that we store sqe twice,
remove the earlier one that is done before validating the index.
Fixes: f75d118349 ("io_uring: harder fdinfo sq/cq ring iterating")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
After the assignment, only exit path with label 'err' uses ret as
return value. However,before exiting through this path with label 'err',
ret is assigned with the return value of io_wq_max_workers(). Hence, the
initial assignment is redundant and can be removed.
Signed-off-by: Nghia Le <nghialm78@gmail.com>
Link: https://lore.kernel.org/r/20211102190521.28291-1-nghialm78@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
-----BEGIN PGP SIGNATURE-----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=sb8J
-----END PGP SIGNATURE-----
Merge tag 'selinux-pr-20211101' of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/selinux
Pull selinux updates from Paul Moore:
- Add LSM/SELinux/Smack controls and auditing for io-uring.
As usual, the individual commit descriptions have more detail, but we
were basically missing two things which we're adding here:
+ establishment of a proper audit context so that auditing of
io-uring ops works similarly to how it does for syscalls (with
some io-uring additions because io-uring ops are *not* syscalls)
+ additional LSM hooks to enable access control points for some of
the more unusual io-uring features, e.g. credential overrides.
The additional audit callouts and LSM hooks were done in conjunction
with the io-uring folks, based on conversations and RFC patches
earlier in the year.
- Fixup the binder credential handling so that the proper credentials
are used in the LSM hooks; the commit description and the code
comment which is removed in these patches are helpful to understand
the background and why this is the proper fix.
- Enable SELinux genfscon policy support for securityfs, allowing
improved SELinux filesystem labeling for other subsystems which make
use of securityfs, e.g. IMA.
* tag 'selinux-pr-20211101' of git://git.kernel.org/pub/scm/linux/kernel/git/pcmoore/selinux:
security: Return xattr name from security_dentry_init_security()
selinux: fix a sock regression in selinux_ip_postroute_compat()
binder: use cred instead of task for getsecid
binder: use cred instead of task for selinux checks
binder: use euid from cred instead of using task
LSM: Avoid warnings about potentially unused hook variables
selinux: fix all of the W=1 build warnings
selinux: make better use of the nf_hook_state passed to the NF hooks
selinux: fix race condition when computing ocontext SIDs
selinux: remove unneeded ipv6 hook wrappers
selinux: remove the SELinux lockdown implementation
selinux: enable genfscon labeling for securityfs
Smack: Brutalist io_uring support
selinux: add support for the io_uring access controls
lsm,io_uring: add LSM hooks to io_uring
io_uring: convert io_uring to the secure anon inode interface
fs: add anon_inode_getfile_secure() similar to anon_inode_getfd_secure()
audit: add filtering for io_uring records
audit,io_uring,io-wq: add some basic audit support to io_uring
audit: prepare audit_context for use in calling contexts beyond syscalls
-----BEGIN PGP SIGNATURE-----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=4vl/
-----END PGP SIGNATURE-----
Merge tag 'for-5.16/ki_complete-2021-10-29' of git://git.kernel.dk/linux-block
Pull kiocb->ki_complete() cleanup from Jens Axboe:
"This removes the res2 argument from kiocb->ki_complete().
Only the USB gadget code used it, everybody else passes 0. The USB
guys checked the user gadget code they could find, and everybody just
uses res as expected for the async interface"
* tag 'for-5.16/ki_complete-2021-10-29' of git://git.kernel.dk/linux-block:
fs: get rid of the res2 iocb->ki_complete argument
usb: remove res2 argument from gadget code completions
-----BEGIN PGP SIGNATURE-----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=5U4u
-----END PGP SIGNATURE-----
Merge tag 'for-5.16/io_uring-2021-10-29' of git://git.kernel.dk/linux-block
Pull io_uring updates from Jens Axboe:
"Light on new features - basically just the hybrid mode support.
Outside of that it's just fixes, cleanups, and performance
improvements.
In detail:
- Add ring related information to the fdinfo output (Hao)
- Hybrid async mode (Hao)
- Support for batched issue on block (me)
- sqe error trace improvement (me)
- IOPOLL efficiency improvements (Pavel)
- submit state cleanups and improvements (Pavel)
- Completion side improvements (Pavel)
- Drain improvements (Pavel)
- Buffer selection cleanups (Pavel)
- Fixed file node improvements (Pavel)
- io-wq setup cancelation fix (Pavel)
- Various other performance improvements and cleanups (Pavel)
- Misc fixes (Arnd, Bixuan, Changcheng, Hao, me, Noah)"
* tag 'for-5.16/io_uring-2021-10-29' of git://git.kernel.dk/linux-block: (97 commits)
io-wq: remove worker to owner tw dependency
io_uring: harder fdinfo sq/cq ring iterating
io_uring: don't assign write hint in the read path
io_uring: clusterise ki_flags access in rw_prep
io_uring: kill unused param from io_file_supports_nowait
io_uring: clean up timeout async_data allocation
io_uring: don't try io-wq polling if not supported
io_uring: check if opcode needs poll first on arming
io_uring: clean iowq submit work cancellation
io_uring: clean io_wq_submit_work()'s main loop
io-wq: use helper for worker refcounting
io_uring: implement async hybrid mode for pollable requests
io_uring: Use ERR_CAST() instead of ERR_PTR(PTR_ERR())
io_uring: split logic of force_nonblock
io_uring: warning about unused-but-set parameter
io_uring: inform block layer of how many requests we are submitting
io_uring: simplify io_file_supports_nowait()
io_uring: combine REQ_F_NOWAIT_{READ,WRITE} flags
io_uring: arm poll for non-nowait files
fs/io_uring: Prioritise checking faster conditions first in io_write
...
-----BEGIN PGP SIGNATURE-----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=QWC+
-----END PGP SIGNATURE-----
Merge tag 'for-5.16/block-2021-10-29' of git://git.kernel.dk/linux-block
Pull block updates from Jens Axboe:
- mq-deadline accounting improvements (Bart)
- blk-wbt timer fix (Andrea)
- Untangle the block layer includes (Christoph)
- Rework the poll support to be bio based, which will enable adding
support for polling for bio based drivers (Christoph)
- Block layer core support for multi-actuator drives (Damien)
- blk-crypto improvements (Eric)
- Batched tag allocation support (me)
- Request completion batching support (me)
- Plugging improvements (me)
- Shared tag set improvements (John)
- Concurrent queue quiesce support (Ming)
- Cache bdev in ->private_data for block devices (Pavel)
- bdev dio improvements (Pavel)
- Block device invalidation and block size improvements (Xie)
- Various cleanups, fixes, and improvements (Christoph, Jackie,
Masahira, Tejun, Yu, Pavel, Zheng, me)
* tag 'for-5.16/block-2021-10-29' of git://git.kernel.dk/linux-block: (174 commits)
blk-mq-debugfs: Show active requests per queue for shared tags
block: improve readability of blk_mq_end_request_batch()
virtio-blk: Use blk_validate_block_size() to validate block size
loop: Use blk_validate_block_size() to validate block size
nbd: Use blk_validate_block_size() to validate block size
block: Add a helper to validate the block size
block: re-flow blk_mq_rq_ctx_init()
block: prefetch request to be initialized
block: pass in blk_mq_tags to blk_mq_rq_ctx_init()
block: add rq_flags to struct blk_mq_alloc_data
block: add async version of bio_set_polled
block: kill DIO_MULTI_BIO
block: kill unused polling bits in __blkdev_direct_IO()
block: avoid extra iter advance with async iocb
block: Add independent access ranges support
blk-mq: don't issue request directly in case that current is to be blocked
sbitmap: silence data race warning
blk-cgroup: synchronize blkg creation against policy deactivation
block: refactor bio_iov_bvec_set()
block: add single bio async direct IO helper
...
Add memory folios, a new type to represent either order-0 pages or
the head page of a compound page. This should be enough infrastructure
to support filesystems converting from pages to folios.
-----BEGIN PGP SIGNATURE-----
iQEzBAABCgAdFiEEejHryeLBw/spnjHrDpNsjXcpgj4FAmF9uI0ACgkQDpNsjXcp
gj7MUAf/R7LCZ+xFiIedw7SAgb/DGK0C9uVjuBEIZgAw21ZUw/GuPI6cuKBMFGGf
rRcdtlvMpwi7yZJcoNXxaqU/xPaaJMjf2XxscIvYJP1mjlZVuwmP9dOx0neNvWOc
T+8lqR6c1TLl82lpqIjGFLwvj2eVowq2d3J5jsaIJFd4odmmYVInrhJXOzC/LQ54
Niloj5ksehf+KUIRLDz7ycppvIHhlVsoAl0eM2dWBAtL0mvT7Nyn/3y+vnMfV2v3
Flb4opwJUgTJleYc16oxTn9svT2yS8q2uuUemRDLW8ABghoAtH3fUUk43RN+5Krd
LYCtbeawtkikPVXZMfWybsx5vn0c3Q==
=7SBe
-----END PGP SIGNATURE-----
Merge tag 'folio-5.16' of git://git.infradead.org/users/willy/pagecache
Pull memory folios from Matthew Wilcox:
"Add memory folios, a new type to represent either order-0 pages or the
head page of a compound page. This should be enough infrastructure to
support filesystems converting from pages to folios.
The point of all this churn is to allow filesystems and the page cache
to manage memory in larger chunks than PAGE_SIZE. The original plan
was to use compound pages like THP does, but I ran into problems with
some functions expecting only a head page while others expect the
precise page containing a particular byte.
The folio type allows a function to declare that it's expecting only a
head page. Almost incidentally, this allows us to remove various calls
to VM_BUG_ON(PageTail(page)) and compound_head().
This converts just parts of the core MM and the page cache. For 5.17,
we intend to convert various filesystems (XFS and AFS are ready; other
filesystems may make it) and also convert more of the MM and page
cache to folios. For 5.18, multi-page folios should be ready.
The multi-page folios offer some improvement to some workloads. The
80% win is real, but appears to be an artificial benchmark (postgres
startup, which isn't a serious workload). Real workloads (eg building
the kernel, running postgres in a steady state, etc) seem to benefit
between 0-10%. I haven't heard of any performance losses as a result
of this series. Nobody has done any serious performance tuning; I
imagine that tweaking the readahead algorithm could provide some more
interesting wins. There are also other places where we could choose to
create large folios and currently do not, such as writes that are
larger than PAGE_SIZE.
I'd like to thank all my reviewers who've offered review/ack tags:
Christoph Hellwig, David Howells, Jan Kara, Jeff Layton, Johannes
Weiner, Kirill A. Shutemov, Michal Hocko, Mike Rapoport, Vlastimil
Babka, William Kucharski, Yu Zhao and Zi Yan.
I'd also like to thank those who gave feedback I incorporated but
haven't offered up review tags for this part of the series: Nick
Piggin, Mel Gorman, Ming Lei, Darrick Wong, Ted Ts'o, John Hubbard,
Hugh Dickins, and probably a few others who I forget"
* tag 'folio-5.16' of git://git.infradead.org/users/willy/pagecache: (90 commits)
mm/writeback: Add folio_write_one
mm/filemap: Add FGP_STABLE
mm/filemap: Add filemap_get_folio
mm/filemap: Convert mapping_get_entry to return a folio
mm/filemap: Add filemap_add_folio()
mm/filemap: Add filemap_alloc_folio
mm/page_alloc: Add folio allocation functions
mm/lru: Add folio_add_lru()
mm/lru: Convert __pagevec_lru_add_fn to take a folio
mm: Add folio_evictable()
mm/workingset: Convert workingset_refault() to take a folio
mm/filemap: Add readahead_folio()
mm/filemap: Add folio_mkwrite_check_truncate()
mm/filemap: Add i_blocks_per_folio()
mm/writeback: Add folio_redirty_for_writepage()
mm/writeback: Add folio_account_redirty()
mm/writeback: Add folio_clear_dirty_for_io()
mm/writeback: Add folio_cancel_dirty()
mm/writeback: Add folio_account_cleaned()
mm/writeback: Add filemap_dirty_folio()
...
The ring iteration is racy, which isn't necessarily a problem except it
can cause us to iterate the whole thing. That isn't desired or ideal,
and it can lead to excessive runtimes of reading fdinfo.
Cap the iteration at tail - head OR the ring size. While in there, clean
up the ring masking and just dump the raw values along with the masks.
That provides more useful debug info.
Fixes: 83f84356bc ("io_uring: add more uring info to fdinfo for debug")
Reported-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The second argument was only used by the USB gadget code, yet everyone
pays the overhead of passing a zero to be passed into aio, where it
ends up being part of the aio res2 value.
Now that everybody is passing in zero, kill off the extra argument.
Reviewed-by: Darrick J. Wong <djwong@kernel.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
ioprio setup doesn't depend on other fields that are modified in
io_prep_rw() and we can move it down in the function without worrying
about performance. It's useful as it makes iocb->ki_flags
accesses/modifications closer together, so it's more likely the compiler
will cache it in a register and avoid extra reloads.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/8ee98779c06f1b59f6039b1e292db4332efd664b.1634987320.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
opcode prep functions are one of the first things that are called, we
can't have ->async_data allocated at this point and it's certainly a
bug. Reflect this assumption in io_timeout_prep() and add a WARN_ONCE
just in case.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/75a28ca7dbcc5af8b6cd9092819e8384c24dedd4.1634987320.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If an opcode doesn't support polling, just let it be executed
synchronously in iowq, otherwise it will do a nonblock attempt just to
fail in io_arm_poll_handler() and return back to blocking execution.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/6401256db01b88f448f15fcd241439cb76f5b940.1634987320.git.asml.silence@gmail.com
Reviewed-by: Hao Xu <haoxu@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
->pollout or ->pollin are set only for opcodes that need a file, so if
io_arm_poll_handler() tests them first we can be sure that the request
has file set and the ->file check can be removed.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/9adfe4f543d984875e516fce6da35348aab48668.1634987320.git.asml.silence@gmail.com
Reviewed-by: Hao Xu <haoxu@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we've got IO_WQ_WORK_CANCEL in io_wq_submit_work(), handle the error
on the same lines as the check instead of having a weird code flow. The
main loop doesn't change but goes one indention left.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/ff4a09cf41f7a22bbb294b6f1faea721e21fe615.1634987320.git.asml.silence@gmail.com
Reviewed-by: Hao Xu <haoxu@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Do a bit of cleaning for the main loop of io_wq_submit_work(). Get rid
of switch, just replace it with a single if as we're retrying in both
other cases. Kill issue_sqe label, Get rid of needs_poll nesting and
disambiguate a bit the comment.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/ed12ce0c64e051f9a6b8a37a24f8ea554d299c29.1634987320.git.asml.silence@gmail.com
Reviewed-by: Hao Xu <haoxu@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The current logic of requests with IOSQE_ASYNC is first queueing it to
io-worker, then execute it in a synchronous way. For unbound works like
pollable requests(e.g. read/write a socketfd), the io-worker may stuck
there waiting for events for a long time. And thus other works wait in
the list for a long time too.
Let's introduce a new way for unbound works (currently pollable
requests), with this a request will first be queued to io-worker, then
executed in a nonblock try rather than a synchronous way. Failure of
that leads it to arm poll stuff and then the worker can begin to handle
other works.
The detail process of this kind of requests is:
step1: original context:
queue it to io-worker
step2: io-worker context:
nonblock try(the old logic is a synchronous try here)
|
|--fail--> arm poll
|
|--(fail/ready)-->synchronous issue
|
|--(succeed)-->worker finish it's job, tw
take over the req
This works much better than the old IOSQE_ASYNC logic in cases where
unbound max_worker is relatively small. In this case, number of
io-worker eazily increments to max_worker, new worker cannot be created
and running workers stuck there handling old works in IOSQE_ASYNC mode.
In my 64-core machine, set unbound max_worker to 20, run echo-server,
turns out:
(arguments: register_file, connetion number is 1000, message size is 12
Byte)
original IOSQE_ASYNC: 76664.151 tps
after this patch: 166934.985 tps
Suggested-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Hao Xu <haoxu@linux.alibaba.com>
Link: https://lore.kernel.org/r/20211018133445.103438-1-haoxu@linux.alibaba.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Another change to the API io-wq worker limitation API added in 5.15,
apply the limit to all prior users that already registered a tctx. It
may be confusing as it's now, in particular the change covers the
following 2 cases:
TASK1 | TASK2
_________________________________________________
ring = create() |
| limit_iowq_workers()
*not limited* |
TASK1 | TASK2
_________________________________________________
ring = create() |
| issue_requests()
limit_iowq_workers() |
| *not limited*
A note on locking, it's safe to traverse ->tctx_list as we hold
->uring_lock, but do that after dropping sqd->lock to avoid possible
problems. It's also safe to access tctx->io_wq there because tasks
kill it only after removing themselves from tctx_list, see
io_uring_cancel_generic() -> io_uring_clean_tctx()
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/d6e09ecc3545e4dc56e43c906ee3d71b7ae21bed.1634818641.git.asml.silence@gmail.com
Reviewed-by: Hao Xu <haoxu@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_unprep_linked_timeout() is broken, first it needs to return back
REQ_F_ARM_LTIMEOUT, so the linked timeout is enqueued and disarmed. But
now we refcounted it, and linked timeouts may get not executed at all,
leaking a request.
Just kill the unprep optimisation.
Fixes: 906c6caaf5 ("io_uring: optimise io_prep_linked_timeout()")
Reported-by: Beld Zhang <beldzhang@gmail.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/51b8e2bfc4bea8ee625cf2ba62b2a350cc9be031.1634719585.git.asml.silence@gmail.com
Link: https://github.com/axboe/liburing/issues/460
Reported-by: Beld Zhang <beldzhang@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Currently, IORING_REGISTER_IOWQ_MAX_WORKERS applies only to the task
that issued it, it's unexpected for users. If one task creates a ring,
limits workers and then passes it to another task the limit won't be
applied to the other task.
Another pitfall is that a task should either create a ring or submit at
least one request for IORING_REGISTER_IOWQ_MAX_WORKERS to work at all,
furher complicating the picture.
Change the API, save the limits and apply to all future users. Note, it
should be done first before giving away the ring or submitting new
requests otherwise the result is not guaranteed.
Fixes: 2e480058dd ("io-wq: provide a way to limit max number of workers")
Link: https://github.com/axboe/liburing/issues/460
Reported-by: Beld Zhang <beldzhang@gmail.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/51d0bae97180e08ab722c0d5c93e7439cfb6f697.1634683237.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Use ERR_CAST() instead of ERR_PTR(PTR_ERR()).
This makes it more readable and also fix this warning detected by
err_cast.cocci:
./fs/io_uring.c: WARNING: 3208: 11-18: ERR_CAST can be used with buf
Reported-by: Zeal Robot <zealci@zte.com.cn>
Signed-off-by: Changcheng Deng <deng.changcheng@zte.com.cn>
Link: https://lore.kernel.org/r/20211020084948.1038420-1-deng.changcheng@zte.com.cn
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Currently force_nonblock stands for three meanings:
- nowait or not
- in an io-worker or not(hold uring_lock or not)
Let's split the logic to two flags, IO_URING_F_NONBLOCK and
IO_URING_F_UNLOCKED for convenience of the next patch.
Suggested-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Hao Xu <haoxu@linux.alibaba.com>
Reviewed-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/20211018133431.103298-1-haoxu@linux.alibaba.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When enabling -Wunused warnings by building with W=1, I get an
instance of the -Wunused-but-set-parameter warning in the io_uring code:
fs/io_uring.c: In function 'io_queue_async_work':
fs/io_uring.c:1445:61: error: parameter 'locked' set but not used [-Werror=unused-but-set-parameter]
1445 | static void io_queue_async_work(struct io_kiocb *req, bool *locked)
| ~~~~~~^~~~~~
There are very few warnings of this type, so it would be nice to enable
this by default and fix all the existing instances. As the assignment
serves no purpose by itself other than to prevent developers from using
the variable, an easy workaround is to remove the assignment and just
rename the argument to "dont_use".
Fixes: f237c30a56 ("io_uring: batch task work locking")
Link: https://lore.kernel.org/lkml/20210920121352.93063-1-arnd@kernel.org/
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20211019153507.348480-1-arnd@kernel.org
Signed-off-by: Jens Axboe <axboe@kernel.dk>
