Take advantage of delayed/inline completion flushing and pass right
issue flags for completion of open, open2, fadvise and poll remove
opcodes. All others either already use it or always punted and never
executed inline.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/0badc7512e82f7350b73bb09abbebbecbdd5dab8.1618101759.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Instead of keeping requests in a inflight_list, just track them with a
per tctx atomic counter. Apart from it being much easier and more
consistent with task cancel, it frees ->inflight_entry from being shared
between iopoll and cancel-track, so less headache for us.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/3c2ee0863cd7eeefa605f3eaff4c1c461a6f1157.1618101759.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
hash_del() works well with non-hashed nodes, there's no need to check
if it is hashed first.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Remove error parameter from io_poll_complete(), 0 is always passed,
and do a bit of cleaning on top.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_poll_complete() always fills an event (even an overflowed one), so we
always should do io_cqring_ev_posted() afterwards. And that's what is
currently happening, because second EPOLLONESHOT check is always true,
it can't return !done for oneshots.
Remove those branching, it's much easier to read.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We currently allow racy updates to multishot requests, but we can end up
double adding the poll request if both completion and update does it.
Ensure that we skip re-add on the update side if someone else is
completing it.
Fixes: b69de288e9 ("io_uring: allow events and user_data update of running poll requests")
Reported-by: Joakim Hassila <joj@mac.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add struct io_fixed_file representing a single registered file, first to
hide ugly struct file **, which may be misleading, and secondly to
retype it to unsigned long as conversions to it and back to file * for
handling and masking FFS_* flags are getting nasty.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/78669731a605a7614c577c3de552631cfaf0869a.1617287883.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Introduce a heler io_free_file_tables() doing all the cleaning, there
are several places where it's hand coded. Also move all allocations into
io_sqe_alloc_file_tables() and rename it, so all of it is in one place.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/502a84ebf41ff119b095e59661e678eacb752bf8.1617287883.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Don't put linked timeout req in io_async_find_and_cancel() but do it in
io_link_timeout_fn(), so we have only one point for that and won't have
to do it differently as it's now (put vs put_deferred). Btw, improve a
bit io_async_find_and_cancel()'s locking.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/d75b70957f245275ab7cba83e0ac9c1b86aae78a.1617287883.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Replace a hand-coded overflow check with a specialised function. Even
though compilers are smart enough to generate identical binary (i.e.
check carry bit), but it's more foolproof and conveys the intention
better.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/e437dcdc929bacbb6f11a4824ecbbf17225cb82a.1617287883.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
rsrc_data refs should always be valid for potential submitters,
io_rsrc_ref_quiesce() restores it before unlocking, so
percpu_ref_is_dying() check in io_sqe_files_unregister() does nothing
and misleading. Concurrent quiesce is prevented with
struct io_rsrc_data::quiesce.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/bf97055e1748ee3a382e66daf384a469eb90b931.1617287883.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we're going to ever support multiple types of resources we need
shared rsrc nodes to not bloat requests, that is implemented in this
patch. It also gives a nicer API and saves one pointer dereference
in io_req_set_rsrc_node().
We may say that all requests bound to a resource belong to one and only
one rsrc node, and considering that nodes are removed and recycled
strictly in-order, this separates requests into generations, where
generation are changed on each node switch (i.e. io_rsrc_node_switch()).
The API is simple, io_rsrc_node_switch() switches to a new generation if
needed, and also optionally kills a passed in io_rsrc_data. Each call to
io_rsrc_node_switch() have to be preceded with
io_rsrc_node_switch_start(). The start function is idempotent and should
not necessarily be followed by switch.
One difference is that once a node was set it will always retain a valid
rsrc node, even on unregister. It may be a nuisance at the moment, but
makes much sense for multiple types of resources. Another thing changed
is that nodes are bound to/associated with a io_rsrc_data later just
before killing (i.e. switching).
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/7e9c693b4b9a2f47aa784b616ce29843021bb65a.1617287883.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_rsrc_node_get() and io_rsrc_node_set() are always used together,
merge them into one so most users don't even see io_rsrc_node and don't
need to care about it.
It helped to catch io_sqe_files_register() inferring rsrc data argument
for get and set differently, not a problem but a good sign.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/0827b080b2e61b3dec795380f7e1a1995595d41f.1617287883.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Keep it consistent with update and use io_rsrc_node_prealloc() +
io_rsrc_node_get() in io_sqe_files_register() as well, that will be used
in future patches, not as error prone and allows to deduplicate
rsrc_node init.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/cf87321e6be5e38f4dc7fe5079d2aa6945b1ace0.1617287883.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Joakim reports that in some conditions he sees a multishot poll request
being canceled, and that it coincides with getting -EALREADY on
modification. As part of the poll update procedure, there's a small window
where the request is marked as canceled, and if this coincides with the
event actually triggering, then we can get a spurious -ECANCELED and
termination of the multishot request.
Don't mark the poll request as being canceled for update. We also don't
care if we race on removal unless it's a one-shot request, we can safely
updated for either case.
Fixes: b69de288e9 ("io_uring: allow events and user_data update of running poll requests")
Reported-by: Joakim Hassila <joj@mac.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now that we have any worker being attached to the original task as
threads, accounting of CPU time is directly attributed to the original
task as well. This means that we no longer have to restrict SQPOLL to
needing elevated privileges, as it's really no different from just having
the task spawn a busy looping thread in userspace.
Reported-by: Stefano Garzarella <sgarzare@redhat.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io-wq relies on a manager thread to create/fork new workers, as needed.
But there's really no strong need for it anymore. We have the following
cases that fork a new worker:
1) Work queue. This is done from the task itself always, and it's trivial
to create a worker off that path, if needed.
2) All workers have gone to sleep, and we have more work. This is called
off the sched out path. For this case, use a task_work items to queue
a fork-worker operation.
3) Hashed work completion. Don't think we need to do anything off this
case. If need be, it could just use approach 2 as well.
Part of this change is incrementing the running worker count before the
fork, to avoid cases where we observe we need a worker and then queue
creation of one. Then new work comes in, we fork a new one. That last
queue operation should have waited for the previous worker to come up,
it's quite possible we don't even need it. Hence move the worker running
from before we fork it off to more efficiently handle that case.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This adds two new POLL_ADD flags, IORING_POLL_UPDATE_EVENTS and
IORING_POLL_UPDATE_USER_DATA. As with the other POLL_ADD flag, these are
masked into sqe->len. If set, the POLL_ADD will have the following
behavior:
- sqe->addr must contain the the user_data of the poll request that
needs to be modified. This field is otherwise invalid for a POLL_ADD
command.
- If IORING_POLL_UPDATE_EVENTS is set, sqe->poll_events must contain the
new mask for the existing poll request. There are no checks for whether
these are identical or not, if a matching poll request is found, then it
is re-armed with the new mask.
- If IORING_POLL_UPDATE_USER_DATA is set, sqe->off must contain the new
user_data for the existing poll request.
A POLL_ADD with any of these flags set may complete with any of the
following results:
1) 0, which means that we successfully found the existing poll request
specified, and performed the re-arm procedure. Any error from that
re-arm will be exposed as a completion event for that original poll
request, not for the update request.
2) -ENOENT, if no existing poll request was found with the given
user_data.
3) -EALREADY, if the existing poll request was already in the process of
being removed/canceled/completing.
4) -EACCES, if an attempt was made to modify an internal poll request
(eg not one originally issued ass IORING_OP_POLL_ADD).
The usual -EINVAL cases apply as well, if any invalid fields are set
in the sqe for this command type.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We'll need this helper for another purpose, for now just abstract it
out and have io_poll_cancel() use it for lookups.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we hit overflow and fail to allocate an overflow entry for the
completion, terminate the multishot poll mode.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The default io_uring poll mode is one-shot, where once the event triggers,
the poll command is completed and won't trigger any further events. If
we're doing repeated polling on the same file or socket, then it can be
more efficient to do multishot, where we keep triggering whenever the
event becomes true.
This deviates from the usual norm of having one CQE per SQE submitted. Add
a CQE flag, IORING_CQE_F_MORE, which tells the application to expect
further completion events from the submitted SQE. Right now the only user
of this is POLL_ADD in multishot mode.
Since sqe->poll_events is using the space that we normally use for adding
flags to commands, use sqe->len for the flag space for POLL_ADD. Multishot
mode is selected by setting IORING_POLL_ADD_MULTI in sqe->len. An
application should expect more CQEs for the specificed SQE if the CQE is
flagged with IORING_CQE_F_MORE. In multishot mode, only cancelation or an
error will terminate the poll request, in which case the flag will be
cleared.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We should be including the completion flags for better introspection on
exactly what completion event was logged.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Instead of using a request itself for overflowed CQE stashing, allocate a
separate entry. The disadvantage is that the allocation may fail and it
will be accounted as lost (see rings->cq_overflow), so we lose reliability
in case of memory pressure if the application is driving the CQ ring into
overflow. However, it opens a way for for multiple CQEs per an SQE and
even generating SQE-less CQEs.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
[axboe: use GFP_ATOMIC | __GFP_ACCOUNT]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Instead of masking these in as part of regular POLL_ADD prep, do it in
io_init_poll_iocb(), and include NVAL as that's generally unmaskable,
and RDHUP alongside the HUP that is already set.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Expect read/write to succeed and create a hot path for this case, in
particular hide all error handling with resubmission under a single
check with the desired result.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Move iov_iter_revert() resetting iterator in case of -EIOCBQUEUED into
io_resubmit_prep(), so we don't do heavy revert in hot path, also saves
a couple of checks.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When reissue_prep failed in io_complete_rw_iopoll(), we change return
code to -EIO to prevent io_iopoll_complete() from doing resubmission.
Mark requests with a new flag (i.e. REQ_F_DONT_REISSUE) instead and
retain the original return value.
It also removes io_rw_reissue() from io_iopoll_complete() that will be
used later.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
file_end_write() is only for regular files, so the function do a couple
of dereferences to get inode and check for it. However, we already have
REQ_F_ISREG at hand, just use it and inline file_end_write().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
current->files are always valid now even for io-wq threads, so kill not
used anymore REQ_F_NO_FILE_TABLE.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
req->work is mostly unused unless it's punted, and io_init_req() is too
hot for fully initialising it. Fortunately, we can skip init work.next
as it's controlled by io-wq, and can not touch work.flags by moving
everything related into io_prep_async_work(). The only field left is
req->work.creds, but there is nothing can be done, keep maintaining it.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
struct io_uring_task::sqpoll is not used anymore, kill it
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_match_task() matches all requests with PF_EXITING task, even though
those may be valid requests. It was necessary for SQPOLL cancellation,
but now it kills all requests before exiting via
io_uring_cancel_sqpoll(), so it's not needed.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
REQ_F_LINK_TIMEOUT is a hint that to look for linked timeouts to cancel,
we're leaving it even when it's already fired. Hence don't care to clear
it in io_kill_linked_timeout(), it's safe and is called only once.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Inline io_task_work_add() into io_req_task_work_add(). They both work
with a request, so keeping them separate doesn't make things much more
clear, but merging allows optimise it. Apart from small wins like not
reading req->ctx or not calculating @notify in the hot path, i.e. with
tctx->task_state set, it avoids doing wake_up_process() for every single
add, but only after actually done task_work_add().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_put_file() doesn't do a good job at generating a good code. Inline
it, so we can check REQ_F_FIXED_FILE first, prioritising FIXED_FILE case
over requests without files, and saving a memory load in that case.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Reshuffle io_dismantle_req() checks to put most of slow path stuff under
a single if.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Inline io_clean_op(), leaving __io_clean_op() but renaming it. This will
be used in following patches.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Both io_req_complete_failed() and __io_req_task_cancel() do the same
thing: set failure flag, put both req refs and emit an CQE. The former
one is a bit more advance as it puts req back into a req cache, so make
it to take over __io_req_task_cancel() and remove the last one.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a new helper io_flush_cached_locked_reqs() that splices
locked_free_list to free_list, and does it right doing all sync and
invariant reinit.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We don't care about ret value in io_free_req_deferred(), make the code a
bit more concise.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
One big omission is that io_put_req() haven't been marked inline, and at
least gcc 9 doesn't inline it, not to mention that it's really hot and
extra function call is intolerable, especially when it doesn't put a
final ref.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There are two problems:
1) we always allocate refnodes in advance and free them if those
haven't been used. It's expensive, takes two allocations, where one of
them is percpu. And it may be pretty common not actually using them.
2) Current API with allocating a refnode and setting some of the fields
is error prone, we don't ever want to have a file node runninng fixed
buffer callback...
Solve both with pre-init/get API. Pre-init just leaves the node for
later if not used, and for get (i.e. io_rsrc_refnode_get()), you need to
explicitly pass all arguments setting callbacks/etc., so it's more
resilient.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Emphasize that return value of io_flush_cached_reqs() depends on number
of requests in the cache. It looks nicer and might help tools from
false-negative analyses.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Move the case of successfully issued request by doing that check first.
It's not much of a difference, just generates slightly better code for
me.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Inline __io_queue_linked_timeout(), we don't need it
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Don't do a function call (io_dismantle_req()) in the middle and place it
to near other function calls, otherwise may lead to excessive register
spilling.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
First of all, w need to set tctx->sqpoll only when we add a new entry
into ->xa, so move it from the hot path. Also extract a hot path for
io_uring_add_task_file() as an inline helper.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add unlikely annotations, because my compiler pretty much mispredicts
every first check, and apart jumping around in the fast path, it also
generates extra instructions, like in advance setting ret value.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
__tctx_task_work() guarantees that ctx won't be killed while running
task_works, so we can remove now unnecessary ctx pinning for internally
armed polling.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We can set canceled == true and complete out-of-line, ensure that we catch
that and correctly return -ECANCELED if the poll operation got canceled.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The correct function is io_iopoll_complete(), which deals with completions
of IOPOLL requests, not io_poll_complete().
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We have to dig quite deep to check for particularly whether or not a
file supports a fast-path nonblock attempt. For fixed files, we can do
this lookup once and cache the state instead.
This adds two new bits to track whether we support async read/write
attempt, and lines up the REQ_F_ISREG bit with those two. The file slot
re-uses the last 3 (or 2, for 32-bit) of the file pointer to cache that
state, and then we mask it in when we go and use a fixed file.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We don't allow them at registration time, so limit the check for needing
inflight tracking in io_file_get() to the non-fixed path.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Use a more comprehensible() max instead of hand coding it with ifs in
io_sqd_update_thread_idle().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_uring manipulates references twice for each request, and hence is very
sensitive to performance of the reference count. This commit borrows a
trick from:
commit f958d7b528
Author: Linus Torvalds <torvalds@linux-foundation.org>
Date: Thu Apr 11 10:06:20 2019 -0700
mm: make page ref count overflow check tighter and more explicit
and switches to atomic_t for references, while still retaining overflow
and underflow checks.
This is good for a 2-3% increase in peak IOPS on a single core. Before:
IOPS=2970879, IOS/call=31/31, inflight=128 (128)
IOPS=2952597, IOS/call=31/31, inflight=128 (128)
IOPS=2943904, IOS/call=31/31, inflight=128 (128)
IOPS=2930006, IOS/call=31/31, inflight=96 (96)
and after:
IOPS=3054354, IOS/call=31/31, inflight=128 (128)
IOPS=3059038, IOS/call=31/31, inflight=128 (128)
IOPS=3060320, IOS/call=31/31, inflight=128 (128)
IOPS=3068256, IOS/call=31/31, inflight=96 (96)
Signed-off-by: Jens Axboe <axboe@kernel.dk>
No functional changes in this patch, just in preparation for handling the
references a bit more efficiently.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If not for async_data NULL check, io_resubmit_prep() is already an rw
specific version of io_req_prep_async(), but slower because 1) it always
goes through io_import_iovec() even if following io_setup_async_rw() the
result 2) instead of initialising iovec/iter in-place it does it
on-stack and then copies with io_setup_async_rw().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Merge two function and do renaming in favour of the second one, it
relays the meaning better.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
needs_async_data controls allocation of async_data, and used in two
cases. 1) when async setup requires it (by io_req_prep_async() or
handler themselves), and 2) when op always needs additional space to
operate, like timeouts do.
Opcode preps already don't bother about the second case and do
allocation unconditionally, restrict needs_async_data to the first case
only and rename it into needs_async_setup.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
[axboe: update for IOPOLL fix]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
All opcode handlers pretty well know whether they need async data or
not, and can skip testing for needs_async_data. The exception is rw
the generic path, but those test the flag by hand anyway. So, check the
flag and make io_alloc_async_data() allocating unconditionally.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
IORING_OP_[SEND,RECV] don't need async setup neither will get into
io_req_prep_async(). Remove them from io_req_prep_async() and remove
needs_async_data checks from the related setup functions.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
__io_cqring_fill_event() takes cflags as long to squeeze it into u32 in
an CQE, awhile all users pass int or unsigned. Replace it with unsigned
int and store it as u32 in struct io_completion to match CQE.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Always complete request holding the mutex instead of doing that strange
dancing with conditional ordering.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Add a simple helper doing CQE posting, marking request for link-failure,
and putting both submission and completion references.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_fixed_file_slot() and io_file_from_index() behave pretty similarly,
DRY and call one from another.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Use io_req_task_queue_fail() on the fail path of io_req_task_queue().
It's unlikely to happen, so don't care about additional overhead, but
allows to keep all the req->result invariant in a single function.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Don't bother to take a ctx->refs for io_req_task_cancel() because it
take uring_lock before putting a request, and the context is promised to
stay alive until unlock happens.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
WARNING: at fs/io_uring.c:8578 io_ring_exit_work.cold+0x0/0x18
As reissuing is now passed back by REQ_F_REISSUE and kiocb_done()
internally uses __io_complete_rw(), it may stop after setting the flag
so leaving a dangling request.
There are tricky edge cases, e.g. reading beyound file, boundary, so
the easiest way is to hand code reissue in kiocb_done() as
__io_complete_rw() was doing for us before.
Fixes: 230d50d448 ("io_uring: move reissue into regular IO path")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/f602250d292f8a84cca9a01d747744d1e797be26.1617842918.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There are lots of ways r/w request may continue its path after getting
REQ_F_REISSUE, it's not necessarily io-wq and can be, e.g. apoll,
and submitted via io_async_task_func() -> __io_req_task_submit()
Clear the flag right after getting it, so the next attempt is well
prepared regardless how the request will be executed.
Fixes: 230d50d448 ("io_uring: move reissue into regular IO path")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/11dcead939343f4e27cab0074d34afcab771bfa4.1617842918.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
kernel test robot correctly pinpoints a compilation failure if
CONFIG_BLOCK isn't set:
fs/io_uring.c: In function '__io_complete_rw':
>> fs/io_uring.c:2509:48: error: implicit declaration of function 'io_rw_should_reissue'; did you mean 'io_rw_reissue'? [-Werror=implicit-function-declaration]
2509 | if ((res == -EAGAIN || res == -EOPNOTSUPP) && io_rw_should_reissue(req)) {
| ^~~~~~~~~~~~~~~~~~~~
| io_rw_reissue
cc1: some warnings being treated as errors
Ensure that we have a stub declaration of io_rw_should_reissue() for
!CONFIG_BLOCK.
Fixes: 230d50d448 ("io_uring: move reissue into regular IO path")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
It's non-obvious how retry is done for block backed files, when it happens
off the kiocb done path. It also makes it tricky to deal with the iov_iter
handling.
Just mark the req as needing a reissue, and handling it from the
submission path instead. This makes it directly obvious that we're not
re-importing the iovec from userspace past the submit point, and it means
that we can just reuse our usual -EAGAIN retry path from the read/write
handling.
At some point in the future, we'll gain the ability to always reliably
return -EAGAIN through the stack. A previous attempt on the block side
didn't pan out and got reverted, hence the need to check for this
information out-of-band right now.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
iov_iter_revert() is done in completion handlers that happensf before
read/write returns -EIOCBQUEUED, no need to repeat reverting afterwards.
Moreover, even though it may appear being just a no-op, it's actually
races with 1) user forging a new iovec of a different size 2) reissue,
that is done via io-wq continues completely asynchronously.
Fixes: 3e6a0d3c75 ("io_uring: fix -EAGAIN retry with IOPOLL")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
task_pid may be large enough to not fit into the left space of
TASK_COMM_LEN-sized buffers and overflow in sprintf. We not so care
about uniqueness, so replace it with safer snprintf().
Reported-by: Alexey Dobriyan <adobriyan@gmail.com>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/1702c6145d7e1c46fbc382f28334c02e1a3d3994.1617267273.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
S_ISBLK is marked as unbounded work for async preparation, because it
doesn't match S_ISREG. That is incorrect, as any read/write to a block
device is also a bounded operation. Fix it up and ensure that S_ISBLK
isn't marked unbounded.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Don't call into get_signal() with the sqd mutex held, it'll fail if we're
freezing the task and we'll get complaints on locks still being held:
====================================
WARNING: iou-sqp-8386/8387 still has locks held!
5.12.0-rc4-syzkaller #0 Not tainted
------------------------------------
1 lock held by iou-sqp-8386/8387:
#0: ffff88801e1d2470 (&sqd->lock){+.+.}-{3:3}, at: io_sq_thread+0x24c/0x13a0 fs/io_uring.c:6731
stack backtrace:
CPU: 1 PID: 8387 Comm: iou-sqp-8386 Not tainted 5.12.0-rc4-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
__dump_stack lib/dump_stack.c:79 [inline]
dump_stack+0x141/0x1d7 lib/dump_stack.c:120
try_to_freeze include/linux/freezer.h:66 [inline]
get_signal+0x171a/0x2150 kernel/signal.c:2576
io_sq_thread+0x8d2/0x13a0 fs/io_uring.c:6748
Fold the get_signal() case in with the parking checks, as we need to drop
the lock in both cases, and since we need to be checking for parking when
juggling the lock anyway.
Reported-by: syzbot+796d767eb376810256f5@syzkaller.appspotmail.com
Fixes: dbe1bdbb39 ("io_uring: handle signals for IO threads like a normal thread")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
general protection fault, probably for non-canonical address
0xdffffc0000000018: 0000 [#1] KASAN: null-ptr-deref
in range [0x00000000000000c0-0x00000000000000c7]
RIP: 0010:io_commit_cqring+0x37f/0xc10 fs/io_uring.c:1318
Call Trace:
io_kill_timeouts+0x2b5/0x320 fs/io_uring.c:8606
io_ring_ctx_wait_and_kill+0x1da/0x400 fs/io_uring.c:8629
io_uring_create fs/io_uring.c:9572 [inline]
io_uring_setup+0x10da/0x2ae0 fs/io_uring.c:9599
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xae
It can get into wait_and_kill() before setting up ctx->rings, and hence
io_commit_cqring() fails. Mimic poll cancel and do it only when we
completed events, there can't be any requests if it failed before
initialising rings.
Fixes: 80c4cbdb5e ("io_uring: do post-completion chore on t-out cancel")
Reported-by: syzbot+0e905eb8228070c457a0@syzkaller.appspotmail.com
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/660261a48f0e7abf260c8e43c87edab3c16736fa.1617014345.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Always try to do cancellation in __io_uring_task_cancel() at least once,
so it actually goes and cleans its sqpoll tasks (i.e. via
io_sqpoll_cancel_sync()), otherwise sqpoll task may submit new requests
after cancellation and it's racy for many reasons.
Fixes: 521d6a737a ("io_uring: cancel sqpoll via task_work")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/0a21bd6d794bb1629bc906dd57a57b2c2985a8ac.1616839147.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There is an assignment to io that is never read after the assignment,
the assignment is redundant and can be removed.
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
As tasks always wait and kill their io-wq on exec/exit, files are of no
more concern to us, so we don't need to specifically cancel them by hand
in those cases. Moreover we should not, because io_match_task() looks at
req->task->files now, which is always true and so leads to extra
cancellations, that wasn't a case before per-task io-wq.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/0566c1de9b9dd417f5de345c817ca953580e0e2e.1616696997.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Don't forget about io_commit_cqring() + io_cqring_ev_posted() after
exit/exec cancelling timeouts. Both functions declared only after
io_kill_timeouts(), so to avoid tons of forward declarations move
it down.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/72ace588772c0f14834a6a4185d56c445a366fb4.1616696997.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We go through various hoops to disallow signals for the IO threads, but
there's really no reason why we cannot just allow them. The IO threads
never return to userspace like a normal thread, and hence don't go through
normal signal processing. Instead, just check for a pending signal as part
of the work loop, and call get_signal() to handle it for us if anything
is pending.
With that, we can support receiving signals, including special ones like
SIGSTOP.
Acked-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Arguably we want CQEs of linked requests be in a strict order of
submission as it always was. Now if init of a request fails its CQE may
be posted before all prior linked requests including the head of the
link. Fix it by failing it last.
Fixes: de59bc104c ("io_uring: fail links more in io_submit_sqe()")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/b7a96b05832e7ab23ad55f84092a2548c4a888b0.1616699075.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
WARNING: CPU: 1 PID: 27907 at fs/io_uring.c:7147 io_sq_thread_park+0xb5/0xd0 fs/io_uring.c:7147
CPU: 1 PID: 27907 Comm: iou-sqp-27905 Not tainted 5.12.0-rc4-syzkaller #0
RIP: 0010:io_sq_thread_park+0xb5/0xd0 fs/io_uring.c:7147
Call Trace:
io_ring_ctx_wait_and_kill+0x214/0x700 fs/io_uring.c:8619
io_uring_release+0x3e/0x50 fs/io_uring.c:8646
__fput+0x288/0x920 fs/file_table.c:280
task_work_run+0xdd/0x1a0 kernel/task_work.c:140
io_run_task_work fs/io_uring.c:2238 [inline]
io_run_task_work fs/io_uring.c:2228 [inline]
io_uring_try_cancel_requests+0x8ec/0xc60 fs/io_uring.c:8770
io_uring_cancel_sqpoll+0x1cf/0x290 fs/io_uring.c:8974
io_sqpoll_cancel_cb+0x87/0xb0 fs/io_uring.c:8907
io_run_task_work_head+0x58/0xb0 fs/io_uring.c:1961
io_sq_thread+0x3e2/0x18d0 fs/io_uring.c:6763
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294
May happen that last ctx ref is killed in io_uring_cancel_sqpoll(), so
fput callback (i.e. io_uring_release()) is enqueued through task_work,
and run by same cancellation. As it's deeply nested we can't do parking
or taking sqd->lock there, because its state is unclear. So avoid
ctx ejection from sqd list from io_ring_ctx_wait_and_kill() and do it
in a clear context in io_ring_exit_work().
Fixes: f6d54255f4 ("io_uring: halt SQO submission on ctx exit")
Reported-by: syzbot+e3a3f84f5cecf61f0583@syzkaller.appspotmail.com
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/e90df88b8ff2cabb14a7534601d35d62ab4cb8c7.1616496707.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_provide_buffers_prep()'s "p->len * p->nbufs" to sign extension
problems. Not a huge problem as it's only used for access_ok() and
increases the checked length, but better to keep typing right.
Reported-by: Colin Ian King <colin.king@canonical.com>
Fixes: efe68c1ca8 ("io_uring: validate the full range of provided buffers for access")
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Reviewed-by: Colin Ian King <colin.king@canonical.com>
Link: https://lore.kernel.org/r/562376a39509e260d8532186a06226e56eb1f594.1616149233.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Linus correctly points out that this is both unnecessary and generates
much worse code on some archs as going from current to thread_info is
actually backwards - and obviously just wasteful, since the thread_info
is what we care about.
Since io_uring only operates on current for these operations, just use
test_thread_flag() instead. For io-wq, we can further simplify and use
tracehook_notify_signal() to handle the TIF_NOTIFY_SIGNAL work and clear
the flag. The latter isn't an actual bug right now, but it may very well
be in the future if we place other work items under TIF_NOTIFY_SIGNAL.
Reported-by: Linus Torvalds <torvalds@linux-foundation.org>
Link: https://lore.kernel.org/io-uring/CAHk-=wgYhNck33YHKZ14mFB5MzTTk8gqXHcfj=RWTAXKwgQJgg@mail.gmail.com/
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Without that it's not safe to use them in a linked combination with
others.
Now combinations like IORING_OP_SENDMSG followed by IORING_OP_SPLICE
should be possible.
We already handle short reads and writes for the following opcodes:
- IORING_OP_READV
- IORING_OP_READ_FIXED
- IORING_OP_READ
- IORING_OP_WRITEV
- IORING_OP_WRITE_FIXED
- IORING_OP_WRITE
- IORING_OP_SPLICE
- IORING_OP_TEE
Now we have it for these as well:
- IORING_OP_SENDMSG
- IORING_OP_SEND
- IORING_OP_RECVMSG
- IORING_OP_RECV
For IORING_OP_RECVMSG we also check for the MSG_TRUNC and MSG_CTRUNC
flags in order to call req_set_fail_links().
There might be applications arround depending on the behavior
that even short send[msg]()/recv[msg]() retuns continue an
IOSQE_IO_LINK chain.
It's very unlikely that such applications pass in MSG_WAITALL,
which is only defined in 'man 2 recvmsg', but not in 'man 2 sendmsg'.
It's expected that the low level sock_sendmsg() call just ignores
MSG_WAITALL, as MSG_ZEROCOPY is also ignored without explicitly set
SO_ZEROCOPY.
We also expect the caller to know about the implicit truncation to
MAX_RW_COUNT, which we don't detect.
cc: netdev@vger.kernel.org
Link: https://lore.kernel.org/r/c4e1a4cc0d905314f4d5dc567e65a7b09621aab3.1615908477.git.metze@samba.org
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Attaching to already dead/dying SQPOLL task is disallowed in
io_sq_offload_create(), but cleanup is hand coded by calling
io_put_sq_data()/etc., that miss to put ctx->sq_creds.
Defer everything to error-path io_sq_thread_finish(), adding
ctx->sqd_list in the error case as well as finish will handle it.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Running sqpoll cancellations via task_work_run() is a bad idea because
it depends on other task works to be run, but those may be locked in
currently running task_work_run() because of how it's (splicing the list
in batches).
Enqueue and run them through a separate callback head, namely
struct io_sq_data::park_task_work. As a nice bonus we now precisely
control where it's run, that's much safer than guessing where it can
happen as it was before.
Reported-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We already have helpers to run/add callback_head but taking ctx and
working with ctx->exit_task_work. Extract generic versions of them
implemented in terms of struct callback_head, it will be used later.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If io_sq_thread_park() of one task got rescheduled right after
set_bit(), before it gets back to mutex_lock() there can happen
park()/unpark() by another task with SQPOLL locking again and
continuing running never seeing that first set_bit(SHOULD_PARK),
so won't even try to put the mutex down for parking.
It will get parked eventually when SQPOLL drops the lock for reschedule,
but may be problematic and will get in the way of further fixes.
Account number of tasks waiting for parking with a new atomic variable
park_pending and adjust SHOULD_PARK accordingly. It doesn't entirely
replaces SHOULD_PARK bit with this atomic var because it's convenient
to have it as a bit in the state and will help to do optimisations
later.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_sq_thread_finish() is called in io_ring_ctx_free(), so SQPOLL task is
potentially running submitting new requests. It's not a disaster because
of using a "try" variant of percpu_ref_get, but is far from nice.
Remove ctx from the sqd ctx list earlier, before cancellation loop, so
SQPOLL can't find it and so won't submit new requests.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The only user of read-locking of sqd->rw_lock is sq_thread itself, which
is by definition alone, so we don't really need rw_semaphore, but mutex
will do. Replace it with a mutex, and kill read-to-write upgrading and
extra task_work handling in io_sq_thread().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If io_req_complete_post() put not a final ref, we can't rely on the
request's ctx ref, and so ctx may potentially be freed while
complete_post() is in io_cqring_ev_posted()/etc.
In that case get an additional ctx reference, and put it in the end, so
protecting following io_cqring_ev_posted(). And also prolong ctx
lifetime until spin_unlock happens, as we do with mutexes, so added
percpu_ref_get() doesn't race with ctx free.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
It's racy to modify req->flags from a not owning context, e.g. linked
timeout calling req_set_fail_links() for the master request might race
with that request setting/clearing flags while being executed
concurrently. Just remove req_set_fail_links(prev) from
io_link_timeout_fn(), io_async_find_and_cancel() and functions down the
line take care of setting the fail bit.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Like we did for the personality idr, convert the IO buffer idr to use
XArray. This avoids a use-after-free on removal of entries, since idr
doesn't like doing so from inside an iterator, and it nicely reduces
the amount of code we need to support this feature.
Fixes: 5a2e745d4d ("io_uring: buffer registration infrastructure")
Cc: stable@vger.kernel.org
Cc: Matthew Wilcox <willy@infradead.org>
Cc: yangerkun <yangerkun@huawei.com>
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
With the freezer using the proper signaling to notify us of when it's
time to freeze a thread, we can re-enable normal freezer usage for the
IO threads. Ensure that SQPOLL, io-wq, and the io-wq manager call
try_to_freeze() appropriately, and remove the default setting of
PF_NOFREEZE from create_io_thread().
Signed-off-by: Jens Axboe <axboe@kernel.dk>
IORING_OP_ASYNC_CANCEL tries io-wq cancellation only for current task.
If it fails go over tctx_list and try it out for every single tctx.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
1) The first problem is io_uring_cancel_sqpoll() ->
io_uring_cancel_task_requests() basically doing park(); park(); and so
hanging.
2) Another one is more subtle, when the master task is doing cancellations,
but SQPOLL task submits in-between the end of the cancellation but
before finish() requests taking a ref to the ctx, and so eternally
locking it up.
3) Yet another is a dying SQPOLL task doing io_uring_cancel_sqpoll() and
same io_uring_cancel_sqpoll() from the owner task, they race for
tctx->wait events. And there probably more of them.
Instead do SQPOLL cancellations from within SQPOLL task context via
task_work, see io_sqpoll_cancel_sync(). With that we don't need temporal
park()/unpark() during cancellation, which is ugly, subtle and anyway
doesn't allow to do io_run_task_work() properly.
io_uring_cancel_sqpoll() is called only from SQPOLL task context and
under sqd locking, so all parking is removed from there. And so,
io_sq_thread_[un]park() and io_sq_thread_stop() are not used now by
SQPOLL task, and that spare us from some headache.
Also remove ctx->sqd_list early to avoid 2). And kill tctx->sqpoll,
which is not used anymore.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
SQPOLL thread to which we're trying to attach may be going away, it's
not nice but a more serious problem is if io_sq_offload_create() sees
sqd->thread==NULL, and tries to init it with a new thread. There are
tons of ways it can be exploited or fail.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We always do complete(&sqd->startup) almost right after sqd->thread
creation, either in the success path or in io_sq_thread_finish(). It's
specifically created not started for us to be able to set some stuff
like sqd->thread and io_uring_alloc_task_context() before following
right after wake_up_new_task().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
As io_uring_cancel_files() and others let SQO to run between
io_uring_try_cancel_requests(), SQO may generate new deferred requests,
so it's safer to try to cancel them in it.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We bypass IOPOLL completion polling (and reaping) for the SQPOLL thread,
but if it's the thread itself invoking cancelations, then we still need
to perform it or no one will.
Fixes: 9936c7c2bc ("io_uring: deduplicate core cancellations sequence")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Earlier kernels had SQPOLL threads that could share across anything, as
we grabbed the context we needed on a per-ring basis. This is no longer
the case, so only allow attaching directly if we're in the same thread
group. That is the common use case. For non-group tasks, just setup a
new context and thread as we would've done if sharing wasn't set. This
isn't 100% ideal in terms of CPU utilization for the forked and share
case, but hopefully that isn't much of a concern. If it is, there are
plans in motion for how to improve that. Most importantly, we want to
avoid app side regressions where sharing worked before and now doesn't.
With this patch, functionality is equivalent to previous kernels that
supported IORING_SETUP_ATTACH_WQ with SQPOLL.
Reported-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We use ->ctx_new_list to notify sqo about new ctx pending, then sqo
should stop and splice it to its sqd->ctx_list, paired with
->sq_thread_comp.
The last one is broken because nobody reinitialises it, and trying to
fix it would only add more complexity and bugs. And the first isn't
really needed as is done under park(), that protects from races well.
Add ctx into sqd->ctx_list directly (under park()), it's much simpler
and allows to kill both, ctx_new_list and sq_thread_comp.
note: apparently there is no real problem at the moment, because
sq_thread_comp is used only by io_sq_thread_finish() followed by
parking, where list_del(&ctx->sqd_list) removes it well regardless
whether it's in the new or the active list.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We have to set ctx->sq_thread_idle before adding a ring to an SQ task,
otherwise sqd races for seeing zero and accounting it as such.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The io-wq threads were already marked as no-freeze, but the manager was
not. On resume, we perpetually have signal_pending() being true, and
hence the manager will loop and spin 100% of the time.
Just mark the tasks created by create_io_thread() as PF_NOFREEZE by
default, and remove any knowledge of it in io-wq and io_uring.
Reported-by: Kevin Locke <kevin@kevinlocke.name>
Tested-by: Kevin Locke <kevin@kevinlocke.name>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We have a tiny race where io_put_sq_data() calls io_sq_thead_stop()
and finds the thread gone, but the thread has indeed not fully
exited or called complete() yet. Close it up by always having
io_sq_thread_stop() wait on completion of the exit event.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Fix the following coccicheck warning:
./fs/io_uring.c:8984:5-8: Unneeded variable: "ret". Return "0" on line
8998
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Signed-off-by: Yang Li <yang.lee@linux.alibaba.com>
Link: https://lore.kernel.org/r/1615271441-33649-1-git-send-email-yang.lee@linux.alibaba.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we hit an error path in the function, make sure that the io_kiocb is
fully initialized at that point so that freeing the request always sees
a valid state.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Calling io_queue_next() after spin_unlock in io_req_complete_post()
races with the other side extracting and reusing this request. Hand
coded parts of io_req_find_next() considering that io_disarm_next()
and io_req_task_queue() have (and safe) to be called with
completion_lock held.
It already does io_commit_cqring() and io_cqring_ev_posted(), so just
reuse it for post io_disarm_next().
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/5672a62f3150ee7c55849f40c0037655c4f2840f.1615250156.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
A preparation patch placing all preparations before extracting a next
request into a separate helper io_disarm_next().
Also, don't spuriously do ev_posted in a rare case where REQ_F_FAIL_LINK
is set but there are no requests linked (i.e. after cancelling a linked
timeout or setting IOSQE_IO_LINK on a last request of a submission
batch).
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/44ecff68d6b47e1c4e6b891bdde1ddc08cfc3590.1615250156.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Don't set IO_SQ_THREAD_SHOULD_STOP when io_sq_offload_create() has
failed on io_uring_alloc_task_context() but leave everything to
io_sq_thread_finish(), because currently io_sq_thread_finish()
hangs on trying to park it. That's great it stalls there, because
otherwise the following io_sq_thread_stop() would be skipped on
IO_SQ_THREAD_SHOULD_STOP check and the sqo would race for sqd with
freeing ctx.
A simple error injection gives something like this.
[ 245.463955] INFO: task sqpoll-test-hang:523 blocked for more than 122 seconds.
[ 245.463983] Call Trace:
[ 245.463990] __schedule+0x36b/0x950
[ 245.464005] schedule+0x68/0xe0
[ 245.464013] schedule_timeout+0x209/0x2a0
[ 245.464032] wait_for_completion+0x8b/0xf0
[ 245.464043] io_sq_thread_finish+0x44/0x1a0
[ 245.464049] io_uring_setup+0x9ea/0xc80
[ 245.464058] __x64_sys_io_uring_setup+0x16/0x20
[ 245.464064] do_syscall_64+0x38/0x50
[ 245.464073] entry_SYSCALL_64_after_hwframe+0x44/0xae
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
You can't call idr_remove() from within a idr_for_each() callback,
but you can call xa_erase() from an xa_for_each() loop, so switch the
entire personality_idr from the IDR to the XArray. This manifests as a
use-after-free as idr_for_each() attempts to walk the rest of the node
after removing the last entry from it.
Fixes: 071698e13a ("io_uring: allow registering credentials")
Cc: stable@vger.kernel.org # 5.6+
Reported-by: yangerkun <yangerkun@huawei.com>
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
[Pavel: rebased (creds load was moved into io_init_req())]
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Link: https://lore.kernel.org/r/7ccff36e1375f2b0ebf73d957f037b43becc0dde.1615212806.git.asml.silence@gmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There are enough of problems with IORING_SETUP_R_DISABLED, including the
burden of checking and kicking off the SQO task all over the codebase --
for exit/cancel/etc.
Rework it, always start the thread but don't do submit unless the flag
is gone, that's much easier.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io-wq now is per-task, so cancellations now should match against
request's ctx.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We keep running into weird dependency issues between the sqd lock and
the parking state. Disentangle the SQPOLL thread from the last bits of
the kthread parking inheritance, and just replace the parking state,
and two associated locks, with a single rw mutex. The SQPOLL thread
keeps the mutex for read all the time, except if someone has marked us
needing to park. Then we drop/re-acquire and try again.
This greatly simplifies the parking state machine (by just getting rid
of it), and makes it a lot more obvious how it works - if you need to
modify the ctx list, then you simply park the thread which will grab
the lock for writing.
Fold in fix from Hillf Danton on not setting STOP on a fatal signal.
Fixes: e54945ae94 ("io_uring: SQPOLL stop error handling fixes")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This brings the behavior back in line with what 5.11 and earlier did,
and this is no longer needed with the improved handling of creds
not needing to do unshare().
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
With IORING_SETUP_ATTACH_WQ we should let __io_sq_thread() use the
initial creds from each ctx.
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_ring_exit_work() have to cancel all requests, including those staying
in io-wq, however it tries only cancellation of current tctx, which is
NULL. If we've got task==NULL, use the ctx-to-tctx map to go over all
tctx/io-wq and try cancellations on them.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We use system_unbound_wq to run io_ring_exit_work(), so it's hard to
monitor whether removal hang or not. Add WARN_ONCE to catch hangs.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We don't use task file notes anymore, and no need left in indexing
task->io_uring->xa by file, and replace it with ctx. It's better
design-wise, especially since we keep a dangling file, and so have to
keep an eye on not dereferencing it.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
With ->flush() gone we're now leaving all uring file notes until the
task dies/execs, so the ctx will not be freed until all tasks that have
ever submit a request die. It was nicer with flush but not much, we
could have locked as described ctx in many cases.
Now we guarantee that ctx outlives all tctx in a sense that
io_ring_exit_work() waits for all tctxs to drop their corresponding
enties in ->xa, and ctx won't go away until then. Hence, additional
io_uring file reference (a.k.a. task file notes) are not needed anymore.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Another preparation patch. When full quiesce is done on ctx exit, use
task_work infra to remove corresponding to the ctx io_uring->xa entries.
For that we use the back tctx map. Also use ->in_idle to prevent
removing it while we traversing ->xa on cancellation, just ignore it.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
For each pair tcxt-ctx create an object and chain it into ctx, so we
have a way to traverse all tctx that are using current ctx. Preparation
patch, will be used later.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Rework io_uring_del_task_file(), so it accepts an index to delete, and
it's not necessarily have to be in the ->xa. Infer file from xa_erase()
to maintain a single origin of truth.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we go async with a request, grab the creds that the task currently has
assigned and make sure that the async side switches to them. This is
handled in the same way that we do for registered personalities.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
45d189c606 ("io_uring: replace force_nonblock with flags") did
something strange for io_openat() slicing all issue_flags but
IO_URING_F_NONBLOCK. Not a bug for now, but better to just forward the
flags.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We have this weird true/false return from parking, and then some of the
callers decide to look at that. It can lead to unbalanced parks and
sqd locking. Have the callers check the thread status once it's parked.
We know we have the lock at that point, so it's either valid or it's NULL.
Fix race with parking on thread exit. We need to be careful here with
ordering of the sdq->lock and the IO_SQ_THREAD_SHOULD_PARK bit.
Rename sqd->completion to sqd->parked to reflect that this is the only
thing this completion event doesn.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
The callback can only be armed, if we get -EIOCBQUEUED returned. It's
important that we clear the WAITQ bit for other cases, otherwise we can
queue for async retry and filemap will assume that we're armed and
return -EAGAIN instead of just blocking for the IO.
Cc: stable@vger.kernel.org # 5.9+
Signed-off-by: Jens Axboe <axboe@kernel.dk>
It doesn't make sense to wait for more events to come in, if we can't
even flush the overflow we already have to the ring. Return -EBUSY for
that condition, just like we do for attempts to submit with overflow
pending.
Cc: stable@vger.kernel.org # 5.11
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This allows us to do task creation and setup without needing to use
completions to try and synchronize with the starting thread. Get rid of
the old io_wq_fork_thread() wrapper, and the 'wq' and 'worker' startup
completion events - we can now do setup before the task is running.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Linked timeouts are fired asynchronously (i.e. soft-irq), and use
generic cancellation paths to do its stuff, including poking into io-wq.
The problem is that it's racy to access tctx->io_wq, as
io_uring_task_cancel() and others may be happening at this exact moment.
Mark linked timeouts with REQ_F_INLIFGHT for now, making sure there are
no timeouts before io-wq destraction.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Instead of going into request internals, like checking req->file->f_op,
do match them based on REQ_F_INFLIGHT, it's set only when we want it to
be reliably cancelled.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_dismantle_req() is always followed by io_put_task(), which already do
proper in_idle wake ups, so we can skip waking the owner task in
io_dismantle_req(). The rules are simpler now, do io_put_task() shortly
after ending a request, and it will be fine.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
__io_queue_async_work() is only called from io_queue_async_work(),
inline it.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Inline io_req_clean_work(), less code and easier to analyse
tctx dependencies and refs usage.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
When we cancel SQPOLL, @task in io_uring_try_cancel_requests() will
differ from current. Use the right tctx from passed in @task, and don't
forget that it can be NULL when the io_uring ctx exits.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We no longer revert the iovec on -EIOCBQUEUED, see commit ab2125df92,
and this started causing issues for IOPOLL on devies that run out of
request slots. Turns out what outside of needing a revert for those, we
also had a bug where we didn't properly setup retry inside the submission
path. That could cause re-import of the iovec, if any, and that could lead
to spurious results if the application had those allocated on the stack.
Catch -EAGAIN retry and make the iovec stable for IOPOLL, just like we do
for !IOPOLL retries.
Cc: <stable@vger.kernel.org> # 5.9+
Reported-by: Abaci Robot <abaci@linux.alibaba.com>
Reported-by: Xiaoguang Wang <xiaoguang.wang@linux.alibaba.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Now, sqo_task is used only for a warning that is not interesting anymore
since sqo_dead is gone, remove all of that including ctx->sqo_task.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
As SQPOLL task doesn't poke into ->sqo_task anymore, there is no need to
kill the sqo when the master task exits. Before it was necessary to
avoid races accessing sqo_task->files with removing them.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
[axboe: don't forget to enable SQPOLL before exit, if started disabled]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we create it in a disabled state because IORING_SETUP_R_DISABLED is
set on ring creation, we need to ensure that we've kicked the thread if
we're exiting before it's been explicitly disabled. Otherwise we can run
into a deadlock where exit is waiting go park the SQPOLL thread, but the
SQPOLL thread itself is waiting to get a signal to start.
That results in the below trace of both tasks hung, waiting on each other:
INFO: task syz-executor458:8401 blocked for more than 143 seconds.
Not tainted 5.11.0-next-20210226-syzkaller #0
"echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
task:syz-executor458 state:D stack:27536 pid: 8401 ppid: 8400 flags:0x00004004
Call Trace:
context_switch kernel/sched/core.c:4324 [inline]
__schedule+0x90c/0x21a0 kernel/sched/core.c:5075
schedule+0xcf/0x270 kernel/sched/core.c:5154
schedule_timeout+0x1db/0x250 kernel/time/timer.c:1868
do_wait_for_common kernel/sched/completion.c:85 [inline]
__wait_for_common kernel/sched/completion.c:106 [inline]
wait_for_common kernel/sched/completion.c:117 [inline]
wait_for_completion+0x168/0x270 kernel/sched/completion.c:138
io_sq_thread_park fs/io_uring.c:7115 [inline]
io_sq_thread_park+0xd5/0x130 fs/io_uring.c:7103
io_uring_cancel_task_requests+0x24c/0xd90 fs/io_uring.c:8745
__io_uring_files_cancel+0x110/0x230 fs/io_uring.c:8840
io_uring_files_cancel include/linux/io_uring.h:47 [inline]
do_exit+0x299/0x2a60 kernel/exit.c:780
do_group_exit+0x125/0x310 kernel/exit.c:922
__do_sys_exit_group kernel/exit.c:933 [inline]
__se_sys_exit_group kernel/exit.c:931 [inline]
__x64_sys_exit_group+0x3a/0x50 kernel/exit.c:931
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x43e899
RSP: 002b:00007ffe89376d48 EFLAGS: 00000246 ORIG_RAX: 00000000000000e7
RAX: ffffffffffffffda RBX: 00000000004af2f0 RCX: 000000000043e899
RDX: 000000000000003c RSI: 00000000000000e7 RDI: 0000000000000000
RBP: 0000000000000000 R08: ffffffffffffffc0 R09: 0000000010000000
R10: 0000000000008011 R11: 0000000000000246 R12: 00000000004af2f0
R13: 0000000000000001 R14: 0000000000000000 R15: 0000000000000001
INFO: task iou-sqp-8401:8402 can't die for more than 143 seconds.
task:iou-sqp-8401 state:D stack:30272 pid: 8402 ppid: 8400 flags:0x00004004
Call Trace:
context_switch kernel/sched/core.c:4324 [inline]
__schedule+0x90c/0x21a0 kernel/sched/core.c:5075
schedule+0xcf/0x270 kernel/sched/core.c:5154
schedule_timeout+0x1db/0x250 kernel/time/timer.c:1868
do_wait_for_common kernel/sched/completion.c:85 [inline]
__wait_for_common kernel/sched/completion.c:106 [inline]
wait_for_common kernel/sched/completion.c:117 [inline]
wait_for_completion+0x168/0x270 kernel/sched/completion.c:138
io_sq_thread+0x27d/0x1ae0 fs/io_uring.c:6717
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:294
INFO: task iou-sqp-8401:8402 blocked for more than 143 seconds.
Reported-by: syzbot+fb5458330b4442f2090d@syzkaller.appspotmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_run_ctx_fallback() can use xchg() instead of cmpxchg(). It's simpler
and faster.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
There is an unlikely but possible race using a freed context. That's
because req->task_work.func() can free a request, but we won't
necessarily find a completion in submit_state.comp and so all ctx refs
may be put by the time we do mutex_lock(&ctx->uring_ctx);
There are several reasons why it can miss going through
submit_state.comp: 1) req->task_work.func() didn't complete it itself,
but punted to iowq (e.g. reissue) and it got freed later, or a similar
situation with it overflowing and getting flushed by someone else, or
being submitted to IRQ completion, 2) As we don't hold the uring_lock,
someone else can do io_submit_flush_completions() and put our ref.
3) Bugs and code obscurities, e.g. failing to propagate issue_flags
properly.
One example is as follows
CPU1 | CPU2
=======================================================================
@req->task_work.func() |
-> @req overflwed, |
so submit_state.comp,nr==0 |
| flush overflows, and free @req
| ctx refs == 0, free it
ctx is dead, but we do |
lock + flush + unlock |
So take a ctx reference for each new ctx we see in __tctx_task_work(),
and do release it until we do all our flushing.
Fixes: 65453d1efb ("io_uring: enable req cache for task_work items")
Reported-by: syzbot+a157ac7c03a56397f553@syzkaller.appspotmail.com
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
[axboe: fold in my one-liner and fix ref mismatch]
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We already run the fallback task_work in io_uring_try_cancel_requests(),
no need to duplicate at ring exit explicitly.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we move it in there, then we no longer have to care about it in io-wq.
This means we can drop the cred handling in io-wq, and we can drop the
REQ_F_WORK_INITIALIZED flag and async init functions as that was the last
user of it since we moved to the new workers. Then we can also drop
io_wq_work->creds, and just hold the personality u16 in there instead.
Suggested-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We're no longer checking anything that requires the work item to be
initialized, as we're not carrying any file related state there.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Destroy current's io-wq backend and tctx on __io_uring_task_cancel(),
aka exec(). Looks it's not strictly necessary, because it will be done
at some point when the task dies and changes of creds/files/etc. are
handled, but better to do that earlier to free io-wq and not potentially
lock previous mm and other resources for the time being.
It's safe to do because we wait for all requests of the current task to
complete, so no request will use tctx afterwards. Note, that
io_uring_files_cancel() may leave some requests for later reaping, so it
leaves tctx intact, that's ok as the task is dying anyway.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Make sure that we killed an io-wq by the time a task is dead.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We clear the bit marking the ctx task_work as active after having run
the queued work, but we really should be clearing it before. Otherwise
we can hit a tiny race ala:
CPU0 CPU1
io_task_work_add() tctx_task_work()
run_work
add_to_list
test_and_set_bit
clear_bit
already set
and CPU0 will return thinking the task_work is queued, while in reality
it's already being run. If we hit the condition after __tctx_task_work()
found no more work, but before we've cleared the bit, then we'll end up
thinking it's queued and will be run. In reality it is queued, but we
didn't queue the ctx task_work to ensure that it gets run.
Fixes: 7cbf1722d5 ("io_uring: provide FIFO ordering for task_work")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we put the io-wq from io_uring, we really want it to exit. Provide
a helper that does that for us. Couple that with not having the manager
hold a reference to the 'wq' and the normal SQPOLL exit will tear down
the io-wq context appropriate.
On the io-wq side, our wq context is per task, so only the task itself
is manipulating ->manager and hence it's safe to check and clear without
any extra locking. We just need to ensure that the manager task stays
around, in case it exits.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We want to reuse this completion, and a single complete should do just
fine. Ensure that we park ourselves first if requested, as that is what
lead to the initial deadlock in this area. If we've got someone attempting
to park us, then we can't proceed without having them finish first.
Fixes: 37d1e2e364 ("io_uring: move SQPOLL thread io-wq forked worker")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
io_uring_try_cancel_requests() matches not only current's requests, but
also of other exiting tasks, so we need to actively cancel them and not
just wait, especially since the function can be called on flush during
do_exit() -> exit_files().
Even if it's not a problem for now, it's much nicer to know that the
function tries to cancel everything it can.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If we fail to fork an SQPOLL worker, we can hit cancel, and hence
attempted thread stop, with the thread already being stopped. Ensure
we check for that.
Also guard thread stop fully by the sqd mutex, just like we do for
park.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
-----BEGIN PGP SIGNATURE-----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=oc8u
-----END PGP SIGNATURE-----
Merge tag 'io_uring-worker.v3-2021-02-25' of git://git.kernel.dk/linux-block
Pull io_uring thread rewrite from Jens Axboe:
"This converts the io-wq workers to be forked off the tasks in question
instead of being kernel threads that assume various bits of the
original task identity.
This kills > 400 lines of code from io_uring/io-wq, and it's the worst
part of the code. We've had several bugs in this area, and the worry
is always that we could be missing some pieces for file types doing
unusual things (recent /dev/tty example comes to mind, userfaultfd
reads installing file descriptors is another fun one... - both of
which need special handling, and I bet it's not the last weird oddity
we'll find).
With these identical workers, we can have full confidence that we're
never missing anything. That, in itself, is a huge win. Outside of
that, it's also more efficient since we're not wasting space and code
on tracking state, or switching between different states.
I'm sure we're going to find little things to patch up after this
series, but testing has been pretty thorough, from the usual
regression suite to production. Any issue that may crop up should be
manageable.
There's also a nice series of further reductions we can do on top of
this, but I wanted to get the meat of it out sooner rather than later.
The general worry here isn't that it's fundamentally broken. Most of
the little issues we've found over the last week have been related to
just changes in how thread startup/exit is done, since that's the main
difference between using kthreads and these kinds of threads. In fact,
if all goes according to plan, I want to get this into the 5.10 and
5.11 stable branches as well.
That said, the changes outside of io_uring/io-wq are:
- arch setup, simple one-liner to each arch copy_thread()
implementation.
- Removal of net and proc restrictions for io_uring, they are no
longer needed or useful"
* tag 'io_uring-worker.v3-2021-02-25' of git://git.kernel.dk/linux-block: (30 commits)
io-wq: remove now unused IO_WQ_BIT_ERROR
io_uring: fix SQPOLL thread handling over exec
io-wq: improve manager/worker handling over exec
io_uring: ensure SQPOLL startup is triggered before error shutdown
io-wq: make buffered file write hashed work map per-ctx
io-wq: fix race around io_worker grabbing
io-wq: fix races around manager/worker creation and task exit
io_uring: ensure io-wq context is always destroyed for tasks
arch: ensure parisc/powerpc handle PF_IO_WORKER in copy_thread()
io_uring: cleanup ->user usage
io-wq: remove nr_process accounting
io_uring: flag new native workers with IORING_FEAT_NATIVE_WORKERS
net: remove cmsg restriction from io_uring based send/recvmsg calls
Revert "proc: don't allow async path resolution of /proc/self components"
Revert "proc: don't allow async path resolution of /proc/thread-self components"
io_uring: move SQPOLL thread io-wq forked worker
io-wq: make io_wq_fork_thread() available to other users
io-wq: only remove worker from free_list, if it was there
io_uring: remove io_identity
io_uring: remove any grabbing of context
...
-----BEGIN PGP SIGNATURE-----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=Vgca
-----END PGP SIGNATURE-----
Merge tag 'for-5.12/io_uring-2021-02-25' of git://git.kernel.dk/linux-block
Pull more io_uring updates from Jens Axboe:
"A collection of later fixes that we should get into this release:
- Series of submission cleanups (Pavel)
- A few fixes for issues from earlier this merge window (Pavel, me)
- IOPOLL resubmission fix
- task_work locking fix (Hao)"
* tag 'for-5.12/io_uring-2021-02-25' of git://git.kernel.dk/linux-block: (25 commits)
Revert "io_uring: wait potential ->release() on resurrect"
io_uring: fix locked_free_list caches_free()
io_uring: don't attempt IO reissue from the ring exit path
io_uring: clear request count when freeing caches
io_uring: run task_work on io_uring_register()
io_uring: fix leaving invalid req->flags
io_uring: wait potential ->release() on resurrect
io_uring: keep generic rsrc infra generic
io_uring: zero ref_node after killing it
io_uring: make the !CONFIG_NET helpers a bit more robust
io_uring: don't hold uring_lock when calling io_run_task_work*
io_uring: fail io-wq submission from a task_work
io_uring: don't take uring_lock during iowq cancel
io_uring: fail links more in io_submit_sqe()
io_uring: don't do async setup for links' heads
io_uring: do io_*_prep() early in io_submit_sqe()
io_uring: split sqe-prep and async setup
io_uring: don't submit link on error
io_uring: move req link into submit_state
io_uring: move io_init_req() into io_submit_sqe()
...
Just like the changes for io-wq, ensure that we re-fork the SQPOLL
thread if the owner execs. Mark the ctx sq thread as sqo_exec if
it dies, and the ring as needing a wakeup which will force the task
to enter the kernel. When it does, setup the new thread and proceed
as usual.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
exec will cancel any threads, including the ones that io-wq is using. This
isn't a problem, in fact we'd prefer it to be that way since it means we
know that any async work cancels naturally without having to handle it
proactively.
But it does mean that we need to setup a new manager, as the manager and
workers are gone. Handle this at queue time, and cancel work if we fail.
Since the manager can go away without us noticing, ensure that the manager
itself holds a reference to the 'wq' as well. Rename io_wq_destroy() to
io_wq_put() to reflect that.
In the future we can now simplify exec cancelation handling, for now just
leave it the same.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
syzbot reports the following hang:
INFO: task syz-executor.0:12538 can't die for more than 143 seconds.
task:syz-executor.0 state:D stack:28352 pid:12538 ppid: 8423 flags:0x00004004
Call Trace:
context_switch kernel/sched/core.c:4324 [inline]
__schedule+0x90c/0x21a0 kernel/sched/core.c:5075
schedule+0xcf/0x270 kernel/sched/core.c:5154
schedule_timeout+0x1db/0x250 kernel/time/timer.c:1868
do_wait_for_common kernel/sched/completion.c:85 [inline]
__wait_for_common kernel/sched/completion.c:106 [inline]
wait_for_common kernel/sched/completion.c:117 [inline]
wait_for_completion+0x168/0x270 kernel/sched/completion.c:138
io_sq_thread_finish+0x96/0x580 fs/io_uring.c:7152
io_sq_offload_create fs/io_uring.c:7929 [inline]
io_uring_create fs/io_uring.c:9465 [inline]
io_uring_setup+0x1fb2/0x2c20 fs/io_uring.c:9550
do_syscall_64+0x2d/0x70 arch/x86/entry/common.c:46
entry_SYSCALL_64_after_hwframe+0x44/0xae
which is due to exiting after the SQPOLL thread has been created, but
hasn't been started yet. Ensure that we always complete the startup
side when waiting for it to exit.
Reported-by: syzbot+c927c937cba8ef66dd4a@syzkaller.appspotmail.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Before the io-wq thread change, we maintained a hash work map and lock
per-node per-ring. That wasn't ideal, as we really wanted it to be per
ring. But now that we have per-task workers, the hash map ends up being
just per-task. That'll work just fine for the normal case of having
one task use a ring, but if you share the ring between tasks, then it's
considerably worse than it was before.
Make the hash map per ctx instead, which provides full per-ctx buffered
write serialization on hashed writes.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
This reverts commit 88f171ab77.
I ran into a case where the ref resurrect now spins, so revert
this change for now until we can further investigate why it's
broken. The bug seems to indicate spinning on the lock itself,
likely there's some ABBA deadlock involved:
[<0>] __percpu_ref_switch_mode+0x45/0x180
[<0>] percpu_ref_resurrect+0x46/0x70
[<0>] io_refs_resurrect+0x25/0xa0
[<0>] __io_uring_register+0x135/0x10c0
[<0>] __x64_sys_io_uring_register+0xc2/0x1a0
[<0>] do_syscall_64+0x42/0x110
[<0>] entry_SYSCALL_64_after_hwframe+0x44/0xa9
Signed-off-by: Jens Axboe <axboe@kernel.dk>
If the task ends up doing no IO, the context list is empty and we don't
call into __io_uring_files_cancel() when the task exits. This can cause
a leak of the io-wq structures.
Ensure we always call __io_uring_files_cancel(), even if the task
context list is empty.
Fixes: 5aa75ed5b9 ("io_uring: tie async worker side to the task context")
Signed-off-by: Jens Axboe <axboe@kernel.dk>
At this point we're only using it for memory accounting, so there's no
need to have an extra ->limit_mem - we can just set ->user if we do
the accounting, or leave it at NULL if we don't.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
We're now just using fork like we would from userspace, so there's no
need to try and impose extra restrictions or accounting on the user
side of things. That's already being done for us. That also means we
don't have to pass in the user_struct anymore, that's correctly inherited
through ->creds on fork.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
A few reasons to do this:
- The naming of the manager and worker have changed. That's a user visible
change, so makes sense to flag it.
- Opening certain files that use ->signal (like /proc/self or /dev/tty)
now works, and the flag tells the application upfront that this is the
case.
- Related to the above, using signalfd will now work as well.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Pavel Begunkov