path: root/man/io_uring_register.2

.\" Copyright (C) 2019 Jens Axboe <axboe@kernel.dk>
.\" Copyright (C) 2019 Red Hat, Inc.
.\"
.\" SPDX-License-Identifier: LGPL-2.0-or-later
.\"
.TH IO_URING_REGISTER 2 2019-01-17 "Linux" "Linux Programmer's Manual"
.SH NAME
io_uring_register \- register files or user buffers for asynchronous I/O 
.SH SYNOPSIS
.nf
.BR "#include <linux/io_uring.h>"
.PP
.BI "int io_uring_register(unsigned int " fd ", unsigned int " opcode ,
.BI "                      void *" arg ", unsigned int " nr_args );
.fi
.PP
.SH DESCRIPTION
.PP

The
.BR io_uring_register ()
system call registers resources (e.g. user buffers, files, eventfd,
personality, restrictions) for use in an
.BR io_uring (7)
instance referenced by
.IR fd .
Registering files or user buffers allows the kernel to take long term
references to internal data structures or create long term mappings of
application memory, greatly reducing per-I/O overhead.

.I fd
is the file descriptor returned by a call to
.BR io_uring_setup (2).
.I opcode
can be one of:

.TP
.B IORING_REGISTER_BUFFERS
.I arg
points to a
.I struct iovec
array of
.I nr_args
entries.  The buffers associated with the iovecs will be locked in
memory and charged against the user's
.B RLIMIT_MEMLOCK
resource limit.  See
.BR getrlimit (2)
for more information.  Additionally, there is a size limit of 1GiB per
buffer.  Currently, the buffers must be anonymous, non-file-backed
memory, such as that returned by
.BR malloc (3)
or
.BR mmap (2)
with the
.B MAP_ANONYMOUS
flag set.  It is expected that this limitation will be lifted in the
future. Huge pages are supported as well. Note that the entire huge
page will be pinned in the kernel, even if only a portion of it is
used.

After a successful call, the supplied buffers are mapped into the
kernel and eligible for I/O.  To make use of them, the application
must specify the
.B IORING_OP_READ_FIXED
or
.B IORING_OP_WRITE_FIXED
opcodes in the submission queue entry (see the
.I struct io_uring_sqe
definition in
.BR io_uring_enter (2)),
and set the
.I buf_index
field to the desired buffer index.  The memory range described by the
submission queue entry's
.I addr
and
.I len
fields must fall within the indexed buffer.

It is perfectly valid to setup a large buffer and then only use part
of it for an I/O, as long as the range is within the originally mapped
region.

An application can increase or decrease the size or number of
registered buffers by first unregistering the existing buffers, and
then issuing a new call to
.BR io_uring_register ()
with the new buffers.

Note that before 5.13 registering buffers would wait for the ring to idle.
If the application currently has requests in-flight, the registration will
wait for those to finish before proceeding.

An application need not unregister buffers explicitly before shutting
down the io_uring instance. Available since 5.1.

.TP
.B IORING_REGISTER_BUFFERS2
Register buffers for I/O. Similar to
.B IORING_REGISTER_BUFFERS
but aims to have a more extensible ABI.

.I arg
points to a
.I struct io_uring_rsrc_register,
and
.I nr_args
should be set to the number of bytes in the structure.

.PP
.in +8n
.EX
struct io_uring_rsrc_register {
    __u32 nr;
    __u32 resv;
    __u64 resv2;
    __aligned_u64 data;
    __aligned_u64 tags;
};

.EE
.in
.PP

.in +8n

The
.I data
field contains a pointer to a
.I struct iovec
array of
.I nr
entries.
The
.I tags
field should either be 0, then tagging is disabled, or point to an array
of
.I nr
"tags" (unsigned 64 bit integers). If a tag is zero, then tagging for this
particular resource (a buffer in this case) is disabled. Otherwise, after the
resource had been unregistered and it's not used anymore, a CQE will be
posted with
.I user_data
set to the specified tag and all other fields zeroed.

Note that resource updates, e.g.
.B IORING_REGISTER_BUFFERS_UPDATE,
don't necessarily deallocate resources by the time it returns, but they might
be held alive until all requests using it complete.

Available since 5.13.

.TP
.B IORING_REGISTER_BUFFERS_UPDATE
Updates registered buffers with new ones, either turning a sparse entry into
a real one, or replacing an existing entry.

.I arg
must contain a pointer to a struct io_uring_rsrc_update2, which contains
an offset on which to start the update, and an array of
.I struct iovec.
.I tags
points to an array of tags.
.I nr
must contain the number of descriptors in the passed in arrays.
See
.B IORING_REGISTER_BUFFERS2
for the resource tagging description.

.PP
.in +8n
.EX

struct io_uring_rsrc_update2 {
    __u32 offset;
    __u32 resv;
    __aligned_u64 data;
    __aligned_u64 tags;
    __u32 nr;
    __u32 resv2;
};
.EE
.in
.PP

.in +8n

Available since 5.13.

.TP
.B IORING_UNREGISTER_BUFFERS
This operation takes no argument, and
.I arg
must be passed as NULL.  All previously registered buffers associated
with the io_uring instance will be released. Available since 5.1.

.TP
.B IORING_REGISTER_FILES
Register files for I/O.
.I arg
contains a pointer to an array of
.I nr_args
file descriptors (signed 32 bit integers).

To make use of the registered files, the
.B IOSQE_FIXED_FILE
flag must be set in the
.I flags
member of the
.IR "struct io_uring_sqe" ,
and the
.I fd
member is set to the index of the file in the file descriptor array.

The file set may be sparse, meaning that the
.B fd
field in the array may be set to
.B -1.
See
.B IORING_REGISTER_FILES_UPDATE
for how to update files in place.

Note that before 5.13 registering files would wait for the ring to idle.
If the application currently has requests in-flight, the registration will
wait for those to finish before proceeding. See
.B IORING_REGISTER_FILES_UPDATE
for how to update an existing set without that limitation.

Files are automatically unregistered when the io_uring instance is
torn down. An application needs only unregister if it wishes to
register a new set of fds. Available since 5.1.

.TP
.B IORING_REGISTER_FILES2
Register files for I/O. Similar to
.B IORING_REGISTER_FILES.

.I arg
points to a
.I struct io_uring_rsrc_register,
and
.I nr_args
should be set to the number of bytes in the structure.

The
.I data
field contains a pointer to an array of
.I nr
file descriptors (signed 32 bit integers).
.I tags
field should either be 0 or or point to an array of
.I nr
"tags" (unsigned 64 bit integers). See
.B IORING_REGISTER_BUFFERS2
for more info on resource tagging.

Note that resource updates, e.g.
.B IORING_REGISTER_FILES_UPDATE,
don't necessarily deallocate resources, they might be held until all requests
using that resource complete.

Available since 5.13.

.TP
.B IORING_REGISTER_FILES_UPDATE
This operation replaces existing files in the registered file set with new
ones, either turning a sparse entry (one where fd is equal to
.B -1
) into a real one, removing an existing entry (new one is set to
.B -1
), or replacing an existing entry with a new existing entry.

.I arg
must contain a pointer to a
.I struct io_uring_files_update,
which contains
an offset on which to start the update, and an array of file descriptors to
use for the update.
.I nr_args
must contain the number of descriptors in the passed in array. Available
since 5.5.

File descriptors can be skipped if they are set to
.B IORING_REGISTER_FILES_SKIP.
Skipping an fd will not touch the file associated with the previous
fd at that index. Available since 5.12.

.TP
.B IORING_REGISTER_FILES_UPDATE2
Similar to IORING_REGISTER_FILES_UPDATE, replaces existing files in the
registered file set with new ones, either turning a sparse entry (one where
fd is equal to
.B -1
) into a real one, removing an existing entry (new one is set to
.B -1
), or replacing an existing entry with a new existing entry.

.I arg
must contain a pointer to a
.I struct io_uring_rsrc_update2,
which contains
an offset on which to start the update, and an array of file descriptors to
use for the update stored in
.I data.
.I tags
points to an array of tags.
.I nr
must contain the number of descriptors in the passed in arrays.
See
.B IORING_REGISTER_BUFFERS2
for the resource tagging description.

Available since 5.13.

.TP
.B IORING_UNREGISTER_FILES
This operation requires no argument, and
.I arg
must be passed as NULL.  All previously registered files associated
with the io_uring instance will be unregistered. Available since 5.1.

.TP
.B IORING_REGISTER_EVENTFD
It's possible to use eventfd(2) to get notified of completion events on an
io_uring instance. If this is desired, an eventfd file descriptor can be
registered through this operation.
.I arg
must contain a pointer to the eventfd file descriptor, and
.I nr_args
must be 1. Note that while io_uring generally takes care to avoid spurious
events, they can occur. Similarly, batched completions of CQEs may only trigger
a single eventfd notification even if multiple CQEs are posted. The application
should make no assumptions on number of events being available having a direct
correlation to eventfd notifications posted. An eventfd notification must thus
only be treated as a hint to check the CQ ring for completions. Available since
5.2.

An application can temporarily disable notifications, coming through the
registered eventfd, by setting the
.B IORING_CQ_EVENTFD_DISABLED
bit in the
.I flags
field of the CQ ring.
Available since 5.8.

.TP
.B IORING_REGISTER_EVENTFD_ASYNC
This works just like
.B IORING_REGISTER_EVENTFD
, except notifications are only posted for events that complete in an async
manner. This means that events that complete inline while being submitted
do not trigger a notification event. The arguments supplied are the same as
for
.B IORING_REGISTER_EVENTFD.
Available since 5.6.

.TP
.B IORING_UNREGISTER_EVENTFD
Unregister an eventfd file descriptor to stop notifications. Since only one
eventfd descriptor is currently supported, this operation takes no argument,
and
.I arg
must be passed as NULL and
.I nr_args
must be zero. Available since 5.2.

.TP
.B IORING_REGISTER_PROBE
This operation returns a structure, io_uring_probe, which contains information
about the opcodes supported by io_uring on the running kernel.
.I arg
must contain a pointer to a struct io_uring_probe, and
.I nr_args
must contain the size of the ops array in that probe struct. The ops array
is of the type io_uring_probe_op, which holds the value of the opcode and
a flags field. If the flags field has
.B IO_URING_OP_SUPPORTED
set, then this opcode is supported on the running kernel. Available since 5.6.

.TP
.B IORING_REGISTER_PERSONALITY
This operation registers credentials of the running application with io_uring,
and returns an id associated with these credentials. Applications wishing to
share a ring between separate users/processes can pass in this credential id
in the sqe
.B personality
field. If set, that particular sqe will be issued with these credentials. Must
be invoked with
.I arg
set to NULL and
.I nr_args
set to zero. Available since 5.6.

.TP
.B IORING_UNREGISTER_PERSONALITY
This operation unregisters a previously registered personality with io_uring.
.I nr_args
must be set to the id in question, and
.I arg
must be set to NULL. Available since 5.6.

.TP
.B IORING_REGISTER_ENABLE_RINGS
This operation enables an io_uring ring started in a disabled state
.RB (IORING_SETUP_R_DISABLED
was specified in the call to
.BR io_uring_setup (2)).
While the io_uring ring is disabled, submissions are not allowed and
registrations are not restricted.

After the execution of this operation, the io_uring ring is enabled:
submissions and registration are allowed, but they will
be validated following the registered restrictions (if any).
This operation takes no argument, must be invoked with
.I arg
set to NULL and
.I nr_args
set to zero. Available since 5.10.

.TP
.B IORING_REGISTER_RESTRICTIONS
.I arg
points to a
.I struct io_uring_restriction
array of
.I nr_args
entries.

With an entry it is possible to allow an
.BR io_uring_register ()
.I opcode,
or specify which
.I opcode
and
.I flags
of the submission queue entry are allowed,
or require certain
.I flags
to be specified (these flags must be set on each submission queue entry).

All the restrictions must be submitted with a single
.BR io_uring_register ()
call and they are handled as an allowlist (opcodes and flags not registered,
are not allowed).

Restrictions can be registered only if the io_uring ring started in a disabled
state
.RB (IORING_SETUP_R_DISABLED
must be specified in the call to
.BR io_uring_setup (2)).

Available since 5.10.

.TP
.B IORING_REGISTER_IOWQ_AFF
By default, async workers created by io_uring will inherit the CPU mask of its
parent. This is usually all the CPUs in the system, unless the parent is being
run with a limited set. If this isn't the desired outcome, the application
may explicitly tell io_uring what CPUs the async workers may run on.
.I arg
must point to a
.B cpu_set_t
mask, and
.I nr_args
the byte size of that mask.

Available since 5.14.

.TP
.B IORING_UNREGISTER_IOWQ_AFF
Undoes a CPU mask previously set with
.B IORING_REGISTER_IOWQ_AFF.
Must not have
.I arg
or
.I nr_args
set.

Available since 5.14.

.TP
.B IORING_REGISTER_IOWQ_MAX_WORKERS
By default, io_uring limits the unbounded workers created to the maximum
processor count set by
.I RLIMIT_NPROC
and the bounded workers is a function of the SQ ring size and the number
of CPUs in the system. Sometimes this can be excessive (or too little, for
bounded), and this command provides a way to change the count per ring (per NUMA
node) instead.

.I arg
must be set to an
.I unsigned int
pointer to an array of two values, with the values in the array being set to
the maximum count of workers per NUMA node. Index 0 holds the bounded worker
count, and index 1 holds the unbounded worker count. On successful return, the
passed in array will contain the previous maximum valyes for each type. If the
count being passed in is 0, then this command returns the current maximum values
and doesn't modify the current setting.
.I nr_args
must be set to 2, as the command takes two values.

Available since 5.15.

.TP
.B IORING_REGISTER_RING_FDS
Whenever
.BR io_uring_enter (2)
is called to submit request or wait for completions, the kernel must grab a
reference to the file descriptor. If the application using io_uring is threaded,
the file table is marked as shared, and the reference grab and put of the file
descriptor count is more expensive than it is for a non-threaded application.

Similarly to how io_uring allows registration of files, this allow registration
of the ring file descriptor itself. This reduces the overhead of the
.BR io_uring_enter (2)
system call.

.I arg
must be set to an unsigned int pointer to an array of type
.I struct io_uring_rsrc_register
of
.I nr_args
number of entries. The
.B data
field of this struct must point to an io_uring file descriptor, and the
.B offset
field can be either
.B -1
or an explicit offset desired for the registered file descriptor value. If
.B -1
is used, then upon successful return of this system call, the field will
contain the value of the registered file descriptor to be used for future
.BR io_uring_enter (2)
system calls.

On successful completion of this request, the returned descriptors may be used
instead of the real file descriptor for
.BR io_uring_enter (2),
provided that
.B IORING_ENTER_REGISTERED_RING
is set in the
.I flags
for the system call. This flag tells the kernel that a registered descriptor
is used rather than a real file descriptor.

Each thread or process using a ring must register the file descriptor directly
by issuing this request.o

The maximum number of supported registered ring descriptors is currently
limited to
.B 16.

Available since 5.18.

.TP
.B IORING_UNREGISTER_RING_FDS
Unregister descriptors previously registered with
.B IORING_REGISTER_RING_FDS.

.I arg
must be set to an unsigned int pointer to an array of type
.I struct io_uring_rsrc_register
of
.I nr_args
number of entries. Only the
.B offset
field should be set in the structure, containing the registered file descriptor
offset previously returned from
.B IORING_REGISTER_RING_FDS
that the application wishes to unregister.

Note that this isn't done automatically on ring exit, if the thread or task
that previously registered a ring file descriptor isn't exiting. It is
recommended to manually unregister any previously registered ring descriptors
if the ring is closed and the task persists. This will free up a registration
slot, making it available for future use.

Available since 5.18.

.SH RETURN VALUE

On success,
.BR io_uring_register ()
returns 0.  On error,
.B -1
is returned, and
.I errno
is set accordingly.

.SH ERRORS
.TP
.B EACCES
The
.I opcode
field is not allowed due to registered restrictions.
.TP
.B EBADF
One or more fds in the
.I fd
array are invalid.
.TP
.B EBADFD
.B IORING_REGISTER_ENABLE_RINGS
or
.B IORING_REGISTER_RESTRICTIONS
was specified, but the io_uring ring is not disabled.
.TP
.B EBUSY
.B IORING_REGISTER_BUFFERS
or
.B IORING_REGISTER_FILES
or
.B IORING_REGISTER_RESTRICTIONS
was specified, but there were already buffers, files, or restrictions
registered.
.TP
.B EFAULT
buffer is outside of the process' accessible address space, or
.I iov_len
is greater than 1GiB.
.TP
.B EINVAL
.B IORING_REGISTER_BUFFERS
or
.B IORING_REGISTER_FILES
was specified, but
.I nr_args
is 0.
.TP
.B EINVAL
.B IORING_REGISTER_BUFFERS
was specified, but
.I nr_args
exceeds
.B UIO_MAXIOV
.TP
.B EINVAL
.B IORING_UNREGISTER_BUFFERS
or
.B IORING_UNREGISTER_FILES
was specified, and
.I nr_args
is non-zero or
.I arg
is non-NULL.
.TP
.B EINVAL
.B IORING_REGISTER_RESTRICTIONS
was specified, but
.I nr_args
exceeds the maximum allowed number of restrictions or restriction
.I opcode
is invalid.
.TP
.B EMFILE
.B IORING_REGISTER_FILES
was specified and
.I nr_args
exceeds the maximum allowed number of files in a fixed file set.
.TP
.B EMFILE
.B IORING_REGISTER_FILES
was specified and adding
.I nr_args
file references would exceed the maximum allowed number of files the user
is allowed to have according to the
.B
RLIMIT_NOFILE
resource limit and the caller does not have
.B CAP_SYS_RESOURCE
capability. Note that this is a per user limit, not per process.
.TP
.B ENOMEM
Insufficient kernel resources are available, or the caller had a
non-zero
.B RLIMIT_MEMLOCK
soft resource limit, but tried to lock more memory than the limit
permitted.  This limit is not enforced if the process is privileged
.RB ( CAP_IPC_LOCK ).
.TP
.B ENXIO
.B IORING_UNREGISTER_BUFFERS
or
.B IORING_UNREGISTER_FILES
was specified, but there were no buffers or files registered.
.TP
.B ENXIO
Attempt to register files or buffers on an io_uring instance that is already
undergoing file or buffer registration, or is being torn down.
.TP
.B EOPNOTSUPP
User buffers point to file-backed memory.