[linux-block.git] / include / uapi / linux / io_uring.h

/* SPDX-License-Identifier: (GPL-2.0 WITH Linux-syscall-note) OR MIT */
/*
 * Header file for the io_uring interface.
 *
 * Copyright (C) 2019 Jens Axboe
 * Copyright (C) 2019 Christoph Hellwig
 */
#ifndef LINUX_IO_URING_H
#define LINUX_IO_URING_H

#include <linux/fs.h>
#include <linux/types.h>
#include <linux/time_types.h>

/*
 * IO submission data structure (Submission Queue Entry)
 */
struct io_uring_sqe {
	__u8	opcode;		/* type of operation for this sqe */
	__u8	flags;		/* IOSQE_ flags */
	__u16	ioprio;		/* ioprio for the request */
	__s32	fd;		/* file descriptor to do IO on */
	union {
		__u64	off;	/* offset into file */
		__u64	addr2;
		struct {
			__u32	cmd_op;
			__u32	__pad1;
		};
	};
	union {
		__u64	addr;	/* pointer to buffer or iovecs */
		__u64	splice_off_in;
	};
	__u32	len;		/* buffer size or number of iovecs */
	union {
		__kernel_rwf_t	rw_flags;
		__u32		fsync_flags;
		__u16		poll_events;	/* compatibility */
		__u32		poll32_events;	/* word-reversed for BE */
		__u32		sync_range_flags;
		__u32		msg_flags;
		__u32		timeout_flags;
		__u32		accept_flags;
		__u32		cancel_flags;
		__u32		open_flags;
		__u32		statx_flags;
		__u32		fadvise_advice;
		__u32		splice_flags;
		__u32		rename_flags;
		__u32		unlink_flags;
		__u32		hardlink_flags;
		__u32		xattr_flags;
		__u32		msg_ring_flags;
	};
	__u64	user_data;	/* data to be passed back at completion time */
	/* pack this to avoid bogus arm OABI complaints */
	union {
		/* index into fixed buffers, if used */
		__u16	buf_index;
		/* for grouped buffer selection */
		__u16	buf_group;
	} __attribute__((packed));
	/* personality to use, if used */
	__u16	personality;
	union {
		__s32	splice_fd_in;
		__u32	file_index;
	};
	union {
		struct {
			__u64	addr3;
			__u64	__pad2[1];
		};
		/*
		 * If the ring is initialized with IORING_SETUP_SQE128, then
		 * this field is used for 80 bytes of arbitrary command data
		 */
		__u8	cmd[0];
	};
};

/*
 * If sqe->file_index is set to this for opcodes that instantiate a new
 * direct descriptor (like openat/openat2/accept), then io_uring will allocate
 * an available direct descriptor instead of having the application pass one
 * in. The picked direct descriptor will be returned in cqe->res, or -ENFILE
 * if the space is full.
 */
#define IORING_FILE_INDEX_ALLOC		(~0U)

enum {
	IOSQE_FIXED_FILE_BIT,
	IOSQE_IO_DRAIN_BIT,
	IOSQE_IO_LINK_BIT,
	IOSQE_IO_HARDLINK_BIT,
	IOSQE_ASYNC_BIT,
	IOSQE_BUFFER_SELECT_BIT,
	IOSQE_CQE_SKIP_SUCCESS_BIT,
};

/*
 * sqe->flags
 */
/* use fixed fileset */
#define IOSQE_FIXED_FILE	(1U << IOSQE_FIXED_FILE_BIT)
/* issue after inflight IO */
#define IOSQE_IO_DRAIN		(1U << IOSQE_IO_DRAIN_BIT)
/* links next sqe */
#define IOSQE_IO_LINK		(1U << IOSQE_IO_LINK_BIT)
/* like LINK, but stronger */
#define IOSQE_IO_HARDLINK	(1U << IOSQE_IO_HARDLINK_BIT)
/* always go async */
#define IOSQE_ASYNC		(1U << IOSQE_ASYNC_BIT)
/* select buffer from sqe->buf_group */
#define IOSQE_BUFFER_SELECT	(1U << IOSQE_BUFFER_SELECT_BIT)
/* don't post CQE if request succeeded */
#define IOSQE_CQE_SKIP_SUCCESS	(1U << IOSQE_CQE_SKIP_SUCCESS_BIT)

/*
 * io_uring_setup() flags
 */
#define IORING_SETUP_IOPOLL	(1U << 0)	/* io_context is polled */
#define IORING_SETUP_SQPOLL	(1U << 1)	/* SQ poll thread */
#define IORING_SETUP_SQ_AFF	(1U << 2)	/* sq_thread_cpu is valid */
#define IORING_SETUP_CQSIZE	(1U << 3)	/* app defines CQ size */
#define IORING_SETUP_CLAMP	(1U << 4)	/* clamp SQ/CQ ring sizes */
#define IORING_SETUP_ATTACH_WQ	(1U << 5)	/* attach to existing wq */
#define IORING_SETUP_R_DISABLED	(1U << 6)	/* start with ring disabled */
#define IORING_SETUP_SUBMIT_ALL	(1U << 7)	/* continue submit on error */
/*
 * Cooperative task running. When requests complete, they often require
 * forcing the submitter to transition to the kernel to complete. If this
 * flag is set, work will be done when the task transitions anyway, rather
 * than force an inter-processor interrupt reschedule. This avoids interrupting
 * a task running in userspace, and saves an IPI.
 */
#define IORING_SETUP_COOP_TASKRUN	(1U << 8)
/*
 * If COOP_TASKRUN is set, get notified if task work is available for
 * running and a kernel transition would be needed to run it. This sets
 * IORING_SQ_TASKRUN in the sq ring flags. Not valid with COOP_TASKRUN.
 */
#define IORING_SETUP_TASKRUN_FLAG	(1U << 9)
#define IORING_SETUP_SQE128		(1U << 10) /* SQEs are 128 byte */
#define IORING_SETUP_CQE32		(1U << 11) /* CQEs are 32 byte */
/*
 * Only one task is allowed to submit requests
 */
#define IORING_SETUP_SINGLE_ISSUER	(1U << 12)

enum io_uring_op {
	IORING_OP_NOP,
	IORING_OP_READV,
	IORING_OP_WRITEV,
	IORING_OP_FSYNC,
	IORING_OP_READ_FIXED,
	IORING_OP_WRITE_FIXED,
	IORING_OP_POLL_ADD,
	IORING_OP_POLL_REMOVE,
	IORING_OP_SYNC_FILE_RANGE,
	IORING_OP_SENDMSG,
	IORING_OP_RECVMSG,
	IORING_OP_TIMEOUT,
	IORING_OP_TIMEOUT_REMOVE,
	IORING_OP_ACCEPT,
	IORING_OP_ASYNC_CANCEL,
	IORING_OP_LINK_TIMEOUT,
	IORING_OP_CONNECT,
	IORING_OP_FALLOCATE,
	IORING_OP_OPENAT,
	IORING_OP_CLOSE,
	IORING_OP_FILES_UPDATE,
	IORING_OP_STATX,
	IORING_OP_READ,
	IORING_OP_WRITE,
	IORING_OP_FADVISE,
	IORING_OP_MADVISE,
	IORING_OP_SEND,
	IORING_OP_RECV,
	IORING_OP_OPENAT2,
	IORING_OP_EPOLL_CTL,
	IORING_OP_SPLICE,
	IORING_OP_PROVIDE_BUFFERS,
	IORING_OP_REMOVE_BUFFERS,
	IORING_OP_TEE,
	IORING_OP_SHUTDOWN,
	IORING_OP_RENAMEAT,
	IORING_OP_UNLINKAT,
	IORING_OP_MKDIRAT,
	IORING_OP_SYMLINKAT,
	IORING_OP_LINKAT,
	IORING_OP_MSG_RING,
	IORING_OP_FSETXATTR,
	IORING_OP_SETXATTR,
	IORING_OP_FGETXATTR,
	IORING_OP_GETXATTR,
	IORING_OP_SOCKET,
	IORING_OP_URING_CMD,

	/* this goes last, obviously */
	IORING_OP_LAST,
};

/*
 * sqe->fsync_flags
 */
#define IORING_FSYNC_DATASYNC	(1U << 0)

/*
 * sqe->timeout_flags
 */
#define IORING_TIMEOUT_ABS		(1U << 0)
#define IORING_TIMEOUT_UPDATE		(1U << 1)
#define IORING_TIMEOUT_BOOTTIME		(1U << 2)
#define IORING_TIMEOUT_REALTIME		(1U << 3)
#define IORING_LINK_TIMEOUT_UPDATE	(1U << 4)
#define IORING_TIMEOUT_ETIME_SUCCESS	(1U << 5)
#define IORING_TIMEOUT_CLOCK_MASK	(IORING_TIMEOUT_BOOTTIME | IORING_TIMEOUT_REALTIME)
#define IORING_TIMEOUT_UPDATE_MASK	(IORING_TIMEOUT_UPDATE | IORING_LINK_TIMEOUT_UPDATE)
/*
 * sqe->splice_flags
 * extends splice(2) flags
 */
#define SPLICE_F_FD_IN_FIXED	(1U << 31) /* the last bit of __u32 */

/*
 * POLL_ADD flags. Note that since sqe->poll_events is the flag space, the
 * command flags for POLL_ADD are stored in sqe->len.
 *
 * IORING_POLL_ADD_MULTI	Multishot poll. Sets IORING_CQE_F_MORE if
 *				the poll handler will continue to report
 *				CQEs on behalf of the same SQE.
 *
 * IORING_POLL_UPDATE		Update existing poll request, matching
 *				sqe->addr as the old user_data field.
 *
 * IORING_POLL_LEVEL		Level triggered poll.
 */
#define IORING_POLL_ADD_MULTI	(1U << 0)
#define IORING_POLL_UPDATE_EVENTS	(1U << 1)
#define IORING_POLL_UPDATE_USER_DATA	(1U << 2)
#define IORING_POLL_ADD_LEVEL		(1U << 3)

/*
 * ASYNC_CANCEL flags.
 *
 * IORING_ASYNC_CANCEL_ALL	Cancel all requests that match the given key
 * IORING_ASYNC_CANCEL_FD	Key off 'fd' for cancelation rather than the
 *				request 'user_data'
 * IORING_ASYNC_CANCEL_ANY	Match any request
 * IORING_ASYNC_CANCEL_FD_FIXED	'fd' passed in is a fixed descriptor
 */
#define IORING_ASYNC_CANCEL_ALL	(1U << 0)
#define IORING_ASYNC_CANCEL_FD	(1U << 1)
#define IORING_ASYNC_CANCEL_ANY	(1U << 2)
#define IORING_ASYNC_CANCEL_FD_FIXED	(1U << 3)

/*
 * send/sendmsg and recv/recvmsg flags (sqe->ioprio)
 *
 * IORING_RECVSEND_POLL_FIRST	If set, instead of first attempting to send
 *				or receive and arm poll if that yields an
 *				-EAGAIN result, arm poll upfront and skip
 *				the initial transfer attempt.
 */
#define IORING_RECVSEND_POLL_FIRST	(1U << 0)

/*
 * accept flags stored in sqe->ioprio
 */
#define IORING_ACCEPT_MULTISHOT	(1U << 0)

/*
 * IORING_OP_MSG_RING command types, stored in sqe->addr
 */
enum {
	IORING_MSG_DATA,	/* pass sqe->len as 'res' and off as user_data */
	IORING_MSG_SEND_FD,	/* send a registered fd to another ring */
};

/*
 * IORING_OP_MSG_RING flags (sqe->msg_ring_flags)
 *
 * IORING_MSG_RING_CQE_SKIP	Don't post a CQE to the target ring. Not
 *				applicable for IORING_MSG_DATA, obviously.
 */
#define IORING_MSG_RING_CQE_SKIP	(1U << 0)

/*
 * IO completion data structure (Completion Queue Entry)
 */
struct io_uring_cqe {
	__u64	user_data;	/* sqe->data submission passed back */
	__s32	res;		/* result code for this event */
	__u32	flags;

	/*
	 * If the ring is initialized with IORING_SETUP_CQE32, then this field
	 * contains 16-bytes of padding, doubling the size of the CQE.
	 */
	__u64 big_cqe[];
};

/*
 * cqe->flags
 *
 * IORING_CQE_F_BUFFER	If set, the upper 16 bits are the buffer ID
 * IORING_CQE_F_MORE	If set, parent SQE will generate more CQE entries
 * IORING_CQE_F_SOCK_NONEMPTY	If set, more data to read after socket recv
 */
#define IORING_CQE_F_BUFFER		(1U << 0)
#define IORING_CQE_F_MORE		(1U << 1)
#define IORING_CQE_F_SOCK_NONEMPTY	(1U << 2)

enum {
	IORING_CQE_BUFFER_SHIFT		= 16,
};

/*
 * Magic offsets for the application to mmap the data it needs
 */
#define IORING_OFF_SQ_RING		0ULL
#define IORING_OFF_CQ_RING		0x8000000ULL
#define IORING_OFF_SQES			0x10000000ULL

/*
 * Filled with the offset for mmap(2)
 */
struct io_sqring_offsets {
	__u32 head;
	__u32 tail;
	__u32 ring_mask;
	__u32 ring_entries;
	__u32 flags;
	__u32 dropped;
	__u32 array;
	__u32 resv1;
	__u64 resv2;
};

/*
 * sq_ring->flags
 */
#define IORING_SQ_NEED_WAKEUP	(1U << 0) /* needs io_uring_enter wakeup */
#define IORING_SQ_CQ_OVERFLOW	(1U << 1) /* CQ ring is overflown */
#define IORING_SQ_TASKRUN	(1U << 2) /* task should enter the kernel */

struct io_cqring_offsets {
	__u32 head;
	__u32 tail;
	__u32 ring_mask;
	__u32 ring_entries;
	__u32 overflow;
	__u32 cqes;
	__u32 flags;
	__u32 resv1;
	__u64 resv2;
};

/*
 * cq_ring->flags
 */

/* disable eventfd notifications */
#define IORING_CQ_EVENTFD_DISABLED	(1U << 0)

/*
 * io_uring_enter(2) flags
 */
#define IORING_ENTER_GETEVENTS		(1U << 0)
#define IORING_ENTER_SQ_WAKEUP		(1U << 1)
#define IORING_ENTER_SQ_WAIT		(1U << 2)
#define IORING_ENTER_EXT_ARG		(1U << 3)
#define IORING_ENTER_REGISTERED_RING	(1U << 4)

/*
 * Passed in for io_uring_setup(2). Copied back with updated info on success
 */
struct io_uring_params {
	__u32 sq_entries;
	__u32 cq_entries;
	__u32 flags;
	__u32 sq_thread_cpu;
	__u32 sq_thread_idle;
	__u32 features;
	__u32 wq_fd;
	__u32 resv[3];
	struct io_sqring_offsets sq_off;
	struct io_cqring_offsets cq_off;
};

/*
 * io_uring_params->features flags
 */
#define IORING_FEAT_SINGLE_MMAP		(1U << 0)
#define IORING_FEAT_NODROP		(1U << 1)
#define IORING_FEAT_SUBMIT_STABLE	(1U << 2)
#define IORING_FEAT_RW_CUR_POS		(1U << 3)
#define IORING_FEAT_CUR_PERSONALITY	(1U << 4)
#define IORING_FEAT_FAST_POLL		(1U << 5)
#define IORING_FEAT_POLL_32BITS 	(1U << 6)
#define IORING_FEAT_SQPOLL_NONFIXED	(1U << 7)
#define IORING_FEAT_EXT_ARG		(1U << 8)
#define IORING_FEAT_NATIVE_WORKERS	(1U << 9)
#define IORING_FEAT_RSRC_TAGS		(1U << 10)
#define IORING_FEAT_CQE_SKIP		(1U << 11)
#define IORING_FEAT_LINKED_FILE		(1U << 12)

/*
 * io_uring_register(2) opcodes and arguments
 */
enum {
	IORING_REGISTER_BUFFERS			= 0,
	IORING_UNREGISTER_BUFFERS		= 1,
	IORING_REGISTER_FILES			= 2,
	IORING_UNREGISTER_FILES			= 3,
	IORING_REGISTER_EVENTFD			= 4,
	IORING_UNREGISTER_EVENTFD		= 5,
	IORING_REGISTER_FILES_UPDATE		= 6,
	IORING_REGISTER_EVENTFD_ASYNC		= 7,
	IORING_REGISTER_PROBE			= 8,
	IORING_REGISTER_PERSONALITY		= 9,
	IORING_UNREGISTER_PERSONALITY		= 10,
	IORING_REGISTER_RESTRICTIONS		= 11,
	IORING_REGISTER_ENABLE_RINGS		= 12,

	/* extended with tagging */
	IORING_REGISTER_FILES2			= 13,
	IORING_REGISTER_FILES_UPDATE2		= 14,
	IORING_REGISTER_BUFFERS2		= 15,
	IORING_REGISTER_BUFFERS_UPDATE		= 16,

	/* set/clear io-wq thread affinities */
	IORING_REGISTER_IOWQ_AFF		= 17,
	IORING_UNREGISTER_IOWQ_AFF		= 18,

	/* set/get max number of io-wq workers */
	IORING_REGISTER_IOWQ_MAX_WORKERS	= 19,

	/* register/unregister io_uring fd with the ring */
	IORING_REGISTER_RING_FDS		= 20,
	IORING_UNREGISTER_RING_FDS		= 21,

	/* register ring based provide buffer group */
	IORING_REGISTER_PBUF_RING		= 22,
	IORING_UNREGISTER_PBUF_RING		= 23,

	/* sync cancelation API */
	IORING_REGISTER_SYNC_CANCEL		= 24,

	/* this goes last */
	IORING_REGISTER_LAST
};

/* io-wq worker categories */
enum {
	IO_WQ_BOUND,
	IO_WQ_UNBOUND,
};

/* deprecated, see struct io_uring_rsrc_update */
struct io_uring_files_update {
	__u32 offset;
	__u32 resv;
	__aligned_u64 /* __s32 * */ fds;
};

/*
 * Register a fully sparse file space, rather than pass in an array of all
 * -1 file descriptors.
 */
#define IORING_RSRC_REGISTER_SPARSE	(1U << 0)

struct io_uring_rsrc_register {
	__u32 nr;
	__u32 flags;
	__u64 resv2;
	__aligned_u64 data;
	__aligned_u64 tags;
};

struct io_uring_rsrc_update {
	__u32 offset;
	__u32 resv;
	__aligned_u64 data;
};

struct io_uring_rsrc_update2 {
	__u32 offset;
	__u32 resv;
	__aligned_u64 data;
	__aligned_u64 tags;
	__u32 nr;
	__u32 resv2;
};

/* Skip updating fd indexes set to this value in the fd table */
#define IORING_REGISTER_FILES_SKIP	(-2)

#define IO_URING_OP_SUPPORTED	(1U << 0)

struct io_uring_probe_op {
	__u8 op;
	__u8 resv;
	__u16 flags;	/* IO_URING_OP_* flags */
	__u32 resv2;
};

struct io_uring_probe {
	__u8 last_op;	/* last opcode supported */
	__u8 ops_len;	/* length of ops[] array below */
	__u16 resv;
	__u32 resv2[3];
	struct io_uring_probe_op ops[];
};

struct io_uring_restriction {
	__u16 opcode;
	union {
		__u8 register_op; /* IORING_RESTRICTION_REGISTER_OP */
		__u8 sqe_op;      /* IORING_RESTRICTION_SQE_OP */
		__u8 sqe_flags;   /* IORING_RESTRICTION_SQE_FLAGS_* */
	};
	__u8 resv;
	__u32 resv2[3];
};

struct io_uring_buf {
	__u64	addr;
	__u32	len;
	__u16	bid;
	__u16	resv;
};

struct io_uring_buf_ring {
	union {
		/*
		 * To avoid spilling into more pages than we need to, the
		 * ring tail is overlaid with the io_uring_buf->resv field.
		 */
		struct {
			__u64	resv1;
			__u32	resv2;
			__u16	resv3;
			__u16	tail;
		};
		struct io_uring_buf	bufs[0];
	};
};

/* argument for IORING_(UN)REGISTER_PBUF_RING */
struct io_uring_buf_reg {
	__u64	ring_addr;
	__u32	ring_entries;
	__u16	bgid;
	__u16	pad;
	__u64	resv[3];
};

/*
 * io_uring_restriction->opcode values
 */
enum {
	/* Allow an io_uring_register(2) opcode */
	IORING_RESTRICTION_REGISTER_OP		= 0,

	/* Allow an sqe opcode */
	IORING_RESTRICTION_SQE_OP		= 1,

	/* Allow sqe flags */
	IORING_RESTRICTION_SQE_FLAGS_ALLOWED	= 2,

	/* Require sqe flags (these flags must be set on each submission) */
	IORING_RESTRICTION_SQE_FLAGS_REQUIRED	= 3,

	IORING_RESTRICTION_LAST
};

struct io_uring_getevents_arg {
	__u64	sigmask;
	__u32	sigmask_sz;
	__u32	pad;
	__u64	ts;
};

/*
 * Argument for IORING_REGISTER_SYNC_CANCEL
 */
struct io_uring_sync_cancel_reg {
	__u64				addr;
	__s32				fd;
	__u32				flags;
	struct __kernel_timespec	timeout;
	__u64				pad[4];
};

#endif
Commit	Line	Data
9f5834c8	1	/* SPDX-License-Identifier: (GPL-2.0 WITH Linux-syscall-note) OR MIT */
2b188cc1 JA	2	/*
	3	* Header file for the io_uring interface.
	4	*
	5	* Copyright (C) 2019 Jens Axboe
	6	* Copyright (C) 2019 Christoph Hellwig
	7	*/
	8	#ifndef LINUX_IO_URING_H
	9	#define LINUX_IO_URING_H
	10
	11	#include <linux/fs.h>
	12	#include <linux/types.h>
78a861b9	13	#include <linux/time_types.h>
2b188cc1 JA	14
	15	/*
	16	* IO submission data structure (Submission Queue Entry)
	17	*/
	18	struct io_uring_sqe {
	19	__u8 opcode; /* type of operation for this sqe */
6b06314c	20	__u8 flags; /* IOSQE_ flags */
2b188cc1 JA	21	__u16 ioprio; /* ioprio for the request */
2b188cc1 JA	22	__s32 fd; /* file descriptor to do IO on */
17f2fe35 JA	23	union {
	24	__u64 off; /* offset into file */
	25	__u64 addr2;
bdb2c48e PB	26	struct {
	27	__u32 cmd_op;
	28	__u32 __pad1;
	29	};
17f2fe35	30	};
7d67af2c PB	31	union {
	32	__u64 addr; /* pointer to buffer or iovecs */
	33	__u64 splice_off_in;
	34	};
2b188cc1 JA	35	__u32 len; /* buffer size or number of iovecs */
	36	union {
	37	__kernel_rwf_t rw_flags;
c992fe29	38	__u32 fsync_flags;
5769a351 JX	39	__u16 poll_events; /* compatibility */
5769a351 JX	40	__u32 poll32_events; /* word-reversed for BE */
5d17b4a4	41	__u32 sync_range_flags;
0fa03c62	42	__u32 msg_flags;
5262f567	43	__u32 timeout_flags;
17f2fe35	44	__u32 accept_flags;
62755e35	45	__u32 cancel_flags;
15b71abe	46	__u32 open_flags;
eddc7ef5	47	__u32 statx_flags;
4840e418	48	__u32 fadvise_advice;
7d67af2c	49	__u32 splice_flags;
80a261fd	50	__u32 rename_flags;
14a1143b	51	__u32 unlink_flags;
cf30da90	52	__u32 hardlink_flags;
e9621e2b	53	__u32 xattr_flags;
e6130eba	54	__u32 msg_ring_flags;
2b188cc1 JA	55	};
2b188cc1 JA	56	__u64 user_data; /* data to be passed back at completion time */
9ba6a1c0	57	/* pack this to avoid bogus arm OABI complaints */
edafccee	58	union {
9ba6a1c0 PB	59	/* index into fixed buffers, if used */
	60	__u16 buf_index;
	61	/* for grouped buffer selection */
	62	__u16 buf_group;
	63	} __attribute__((packed));
	64	/* personality to use, if used */
	65	__u16 personality;
b9445598 PB	66	union {
	67	__s32 splice_fd_in;
	68	__u32 file_index;
	69	};
ee692a21 JA	70	union {
	71	struct {
	72	__u64 addr3;
	73	__u64 __pad2[1];
	74	};
	75	/*
	76	* If the ring is initialized with IORING_SETUP_SQE128, then
	77	* this field is used for 80 bytes of arbitrary command data
	78	*/
	79	__u8 cmd[0];
	80	};
2b188cc1 JA	81	};
2b188cc1 JA	82
1339f24b JA	83	/*
	84	* If sqe->file_index is set to this for opcodes that instantiate a new
	85	* direct descriptor (like openat/openat2/accept), then io_uring will allocate
	86	* an available direct descriptor instead of having the application pass one
	87	* in. The picked direct descriptor will be returned in cqe->res, or -ENFILE
	88	* if the space is full.
	89	*/
	90	#define IORING_FILE_INDEX_ALLOC (~0U)
	91
6b47ee6e PB	92	enum {
	93	IOSQE_FIXED_FILE_BIT,
	94	IOSQE_IO_DRAIN_BIT,
	95	IOSQE_IO_LINK_BIT,
	96	IOSQE_IO_HARDLINK_BIT,
	97	IOSQE_ASYNC_BIT,
bcda7baa	98	IOSQE_BUFFER_SELECT_BIT,
04c76b41	99	IOSQE_CQE_SKIP_SUCCESS_BIT,
6b47ee6e PB	100	};
6b47ee6e PB	101
6b06314c JA	102	/*
	103	* sqe->flags
	104	*/
6b47ee6e PB	105	/* use fixed fileset */
	106	#define IOSQE_FIXED_FILE (1U << IOSQE_FIXED_FILE_BIT)
	107	/* issue after inflight IO */
	108	#define IOSQE_IO_DRAIN (1U << IOSQE_IO_DRAIN_BIT)
	109	/* links next sqe */
	110	#define IOSQE_IO_LINK (1U << IOSQE_IO_LINK_BIT)
	111	/* like LINK, but stronger */
	112	#define IOSQE_IO_HARDLINK (1U << IOSQE_IO_HARDLINK_BIT)
	113	/* always go async */
	114	#define IOSQE_ASYNC (1U << IOSQE_ASYNC_BIT)
bcda7baa JA	115	/* select buffer from sqe->buf_group */
bcda7baa JA	116	#define IOSQE_BUFFER_SELECT (1U << IOSQE_BUFFER_SELECT_BIT)
04c76b41 PB	117	/* don't post CQE if request succeeded */
04c76b41 PB	118	#define IOSQE_CQE_SKIP_SUCCESS (1U << IOSQE_CQE_SKIP_SUCCESS_BIT)
6b06314c	119
def596e9 JA	120	/*
	121	* io_uring_setup() flags
	122	*/
	123	#define IORING_SETUP_IOPOLL (1U << 0) /* io_context is polled */
6c271ce2 JA	124	#define IORING_SETUP_SQPOLL (1U << 1) /* SQ poll thread */
6c271ce2 JA	125	#define IORING_SETUP_SQ_AFF (1U << 2) /* sq_thread_cpu is valid */
33a107f0	126	#define IORING_SETUP_CQSIZE (1U << 3) /* app defines CQ size */
8110c1a6	127	#define IORING_SETUP_CLAMP (1U << 4) /* clamp SQ/CQ ring sizes */
24369c2e	128	#define IORING_SETUP_ATTACH_WQ (1U << 5) /* attach to existing wq */
7e84e1c7	129	#define IORING_SETUP_R_DISABLED (1U << 6) /* start with ring disabled */
bcbb7bf6	130	#define IORING_SETUP_SUBMIT_ALL (1U << 7) /* continue submit on error */
e1169f06 JA	131	/*
	132	* Cooperative task running. When requests complete, they often require
	133	* forcing the submitter to transition to the kernel to complete. If this
	134	* flag is set, work will be done when the task transitions anyway, rather
	135	* than force an inter-processor interrupt reschedule. This avoids interrupting
	136	* a task running in userspace, and saves an IPI.
	137	*/
	138	#define IORING_SETUP_COOP_TASKRUN (1U << 8)
ef060ea9 JA	139	/*
	140	* If COOP_TASKRUN is set, get notified if task work is available for
	141	* running and a kernel transition would be needed to run it. This sets
	142	* IORING_SQ_TASKRUN in the sq ring flags. Not valid with COOP_TASKRUN.
	143	*/
	144	#define IORING_SETUP_TASKRUN_FLAG (1U << 9)
ebdeb7c0	145	#define IORING_SETUP_SQE128 (1U << 10) /* SQEs are 128 byte */
7a51e5b4	146	#define IORING_SETUP_CQE32 (1U << 11) /* CQEs are 32 byte */
97bbdc06 PB	147	/*
	148	* Only one task is allowed to submit requests
	149	*/
	150	#define IORING_SETUP_SINGLE_ISSUER (1U << 12)
ebdeb7c0	151
cc51eaa8	152	enum io_uring_op {
9e3aa61a JA	153	IORING_OP_NOP,
	154	IORING_OP_READV,
	155	IORING_OP_WRITEV,
	156	IORING_OP_FSYNC,
	157	IORING_OP_READ_FIXED,
	158	IORING_OP_WRITE_FIXED,
	159	IORING_OP_POLL_ADD,
	160	IORING_OP_POLL_REMOVE,
	161	IORING_OP_SYNC_FILE_RANGE,
	162	IORING_OP_SENDMSG,
	163	IORING_OP_RECVMSG,
	164	IORING_OP_TIMEOUT,
	165	IORING_OP_TIMEOUT_REMOVE,
	166	IORING_OP_ACCEPT,
	167	IORING_OP_ASYNC_CANCEL,
	168	IORING_OP_LINK_TIMEOUT,
	169	IORING_OP_CONNECT,
d63d1b5e	170	IORING_OP_FALLOCATE,
15b71abe	171	IORING_OP_OPENAT,
b5dba59e	172	IORING_OP_CLOSE,
05f3fb3c	173	IORING_OP_FILES_UPDATE,
eddc7ef5	174	IORING_OP_STATX,
3a6820f2 JA	175	IORING_OP_READ,
3a6820f2 JA	176	IORING_OP_WRITE,
4840e418	177	IORING_OP_FADVISE,
c1ca757b	178	IORING_OP_MADVISE,
fddaface JA	179	IORING_OP_SEND,
fddaface JA	180	IORING_OP_RECV,
cebdb986	181	IORING_OP_OPENAT2,
3e4827b0	182	IORING_OP_EPOLL_CTL,
7d67af2c	183	IORING_OP_SPLICE,
ddf0322d	184	IORING_OP_PROVIDE_BUFFERS,
067524e9	185	IORING_OP_REMOVE_BUFFERS,
f2a8d5c7	186	IORING_OP_TEE,
36f4fa68	187	IORING_OP_SHUTDOWN,
80a261fd	188	IORING_OP_RENAMEAT,
14a1143b	189	IORING_OP_UNLINKAT,
e34a02dc	190	IORING_OP_MKDIRAT,
7a8721f8	191	IORING_OP_SYMLINKAT,
cf30da90	192	IORING_OP_LINKAT,
4f57f06c	193	IORING_OP_MSG_RING,
e9621e2b SR	194	IORING_OP_FSETXATTR,
e9621e2b SR	195	IORING_OP_SETXATTR,
a56834e0 SR	196	IORING_OP_FGETXATTR,
a56834e0 SR	197	IORING_OP_GETXATTR,
1374e08e	198	IORING_OP_SOCKET,
ee692a21	199	IORING_OP_URING_CMD,
9e3aa61a JA	200
	201	/* this goes last, obviously */
	202	IORING_OP_LAST,
	203	};
c992fe29 CH	204
	205	/*
	206	* sqe->fsync_flags
	207	*/
	208	#define IORING_FSYNC_DATASYNC (1U << 0)
2b188cc1	209
a41525ab JA	210	/*
	211	* sqe->timeout_flags
	212	*/
f1042b6c PB	213	#define IORING_TIMEOUT_ABS (1U << 0)
	214	#define IORING_TIMEOUT_UPDATE (1U << 1)
	215	#define IORING_TIMEOUT_BOOTTIME (1U << 2)
	216	#define IORING_TIMEOUT_REALTIME (1U << 3)
	217	#define IORING_LINK_TIMEOUT_UPDATE (1U << 4)
6224590d	218	#define IORING_TIMEOUT_ETIME_SUCCESS (1U << 5)
50c1df2b	219	#define IORING_TIMEOUT_CLOCK_MASK (IORING_TIMEOUT_BOOTTIME \| IORING_TIMEOUT_REALTIME)
f1042b6c	220	#define IORING_TIMEOUT_UPDATE_MASK (IORING_TIMEOUT_UPDATE \| IORING_LINK_TIMEOUT_UPDATE)
7d67af2c PB	221	/*
	222	* sqe->splice_flags
	223	* extends splice(2) flags
	224	*/
	225	#define SPLICE_F_FD_IN_FIXED (1U << 31) /* the last bit of __u32 */
	226
88e41cf9 JA	227	/*
	228	* POLL_ADD flags. Note that since sqe->poll_events is the flag space, the
	229	* command flags for POLL_ADD are stored in sqe->len.
	230	*
	231	* IORING_POLL_ADD_MULTI Multishot poll. Sets IORING_CQE_F_MORE if
	232	* the poll handler will continue to report
	233	* CQEs on behalf of the same SQE.
b69de288 JA	234	*
	235	* IORING_POLL_UPDATE Update existing poll request, matching
	236	* sqe->addr as the old user_data field.
b9ba8a44 JA	237	*
b9ba8a44 JA	238	* IORING_POLL_LEVEL Level triggered poll.
88e41cf9 JA	239	*/
88e41cf9 JA	240	#define IORING_POLL_ADD_MULTI (1U << 0)
b69de288 JA	241	#define IORING_POLL_UPDATE_EVENTS (1U << 1)
b69de288 JA	242	#define IORING_POLL_UPDATE_USER_DATA (1U << 2)
b9ba8a44	243	#define IORING_POLL_ADD_LEVEL (1U << 3)
88e41cf9	244
8e29da69 JA	245	/*
	246	* ASYNC_CANCEL flags.
	247	*
	248	* IORING_ASYNC_CANCEL_ALL Cancel all requests that match the given key
4bf94615 JA	249	* IORING_ASYNC_CANCEL_FD Key off 'fd' for cancelation rather than the
4bf94615 JA	250	* request 'user_data'
970f256e	251	* IORING_ASYNC_CANCEL_ANY Match any request
7d8ca725	252	* IORING_ASYNC_CANCEL_FD_FIXED 'fd' passed in is a fixed descriptor
8e29da69 JA	253	*/
8e29da69 JA	254	#define IORING_ASYNC_CANCEL_ALL (1U << 0)
4bf94615	255	#define IORING_ASYNC_CANCEL_FD (1U << 1)
970f256e	256	#define IORING_ASYNC_CANCEL_ANY (1U << 2)
7d8ca725	257	#define IORING_ASYNC_CANCEL_FD_FIXED (1U << 3)
8e29da69	258
0455d4cc	259	/*
29c1ac23	260	* send/sendmsg and recv/recvmsg flags (sqe->ioprio)
0455d4cc JA	261	*
	262	* IORING_RECVSEND_POLL_FIRST If set, instead of first attempting to send
	263	* or receive and arm poll if that yields an
	264	* -EAGAIN result, arm poll upfront and skip
	265	* the initial transfer attempt.
	266	*/
	267	#define IORING_RECVSEND_POLL_FIRST (1U << 0)
	268
390ed29b HX	269	/*
	270	* accept flags stored in sqe->ioprio
	271	*/
	272	#define IORING_ACCEPT_MULTISHOT (1U << 0)
	273
e6130eba JA	274	/*
	275	* IORING_OP_MSG_RING command types, stored in sqe->addr
	276	*/
	277	enum {
	278	IORING_MSG_DATA, /* pass sqe->len as 'res' and off as user_data */
	279	IORING_MSG_SEND_FD, /* send a registered fd to another ring */
	280	};
	281
	282	/*
	283	* IORING_OP_MSG_RING flags (sqe->msg_ring_flags)
	284	*
	285	* IORING_MSG_RING_CQE_SKIP Don't post a CQE to the target ring. Not
	286	* applicable for IORING_MSG_DATA, obviously.
	287	*/
	288	#define IORING_MSG_RING_CQE_SKIP (1U << 0)
	289
2b188cc1 JA	290	/*
	291	* IO completion data structure (Completion Queue Entry)
	292	*/
	293	struct io_uring_cqe {
	294	__u64 user_data; /* sqe->data submission passed back */
	295	__s32 res; /* result code for this event */
	296	__u32 flags;
7a51e5b4 SR	297
	298	/*
	299	* If the ring is initialized with IORING_SETUP_CQE32, then this field
	300	* contains 16-bytes of padding, doubling the size of the CQE.
	301	*/
	302	__u64 big_cqe[];
2b188cc1 JA	303	};
2b188cc1 JA	304
bcda7baa JA	305	/*
	306	* cqe->flags
	307	*
	308	* IORING_CQE_F_BUFFER If set, the upper 16 bits are the buffer ID
88e41cf9	309	* IORING_CQE_F_MORE If set, parent SQE will generate more CQE entries
f548a12e	310	* IORING_CQE_F_SOCK_NONEMPTY If set, more data to read after socket recv
bcda7baa JA	311	*/
bcda7baa JA	312	#define IORING_CQE_F_BUFFER (1U << 0)
88e41cf9	313	#define IORING_CQE_F_MORE (1U << 1)
f548a12e	314	#define IORING_CQE_F_SOCK_NONEMPTY (1U << 2)
bcda7baa JA	315
	316	enum {
	317	IORING_CQE_BUFFER_SHIFT = 16,
	318	};
	319
2b188cc1 JA	320	/*
	321	* Magic offsets for the application to mmap the data it needs
	322	*/
	323	#define IORING_OFF_SQ_RING 0ULL
	324	#define IORING_OFF_CQ_RING 0x8000000ULL
	325	#define IORING_OFF_SQES 0x10000000ULL
	326
	327	/*
	328	* Filled with the offset for mmap(2)
	329	*/
	330	struct io_sqring_offsets {
	331	__u32 head;
	332	__u32 tail;
	333	__u32 ring_mask;
	334	__u32 ring_entries;
	335	__u32 flags;
	336	__u32 dropped;
	337	__u32 array;
	338	__u32 resv1;
	339	__u64 resv2;
	340	};
	341
6c271ce2 JA	342	/*
	343	* sq_ring->flags
	344	*/
	345	#define IORING_SQ_NEED_WAKEUP (1U << 0) /* needs io_uring_enter wakeup */
6d5f9049	346	#define IORING_SQ_CQ_OVERFLOW (1U << 1) /* CQ ring is overflown */
ef060ea9	347	#define IORING_SQ_TASKRUN (1U << 2) /* task should enter the kernel */
6c271ce2	348
2b188cc1 JA	349	struct io_cqring_offsets {
	350	__u32 head;
	351	__u32 tail;
	352	__u32 ring_mask;
	353	__u32 ring_entries;
	354	__u32 overflow;
	355	__u32 cqes;
0d9b5b3a SG	356	__u32 flags;
	357	__u32 resv1;
	358	__u64 resv2;
2b188cc1 JA	359	};
2b188cc1 JA	360
7e55a19c SG	361	/*
	362	* cq_ring->flags
	363	*/
	364
	365	/* disable eventfd notifications */
	366	#define IORING_CQ_EVENTFD_DISABLED (1U << 0)
	367
2b188cc1 JA	368	/*
	369	* io_uring_enter(2) flags
	370	*/
e7a6c00d JA	371	#define IORING_ENTER_GETEVENTS (1U << 0)
	372	#define IORING_ENTER_SQ_WAKEUP (1U << 1)
	373	#define IORING_ENTER_SQ_WAIT (1U << 2)
	374	#define IORING_ENTER_EXT_ARG (1U << 3)
	375	#define IORING_ENTER_REGISTERED_RING (1U << 4)
2b188cc1 JA	376
	377	/*
	378	* Passed in for io_uring_setup(2). Copied back with updated info on success
	379	*/
	380	struct io_uring_params {
	381	__u32 sq_entries;
	382	__u32 cq_entries;
	383	__u32 flags;
6c271ce2 JA	384	__u32 sq_thread_cpu;
6c271ce2 JA	385	__u32 sq_thread_idle;
ac90f249	386	__u32 features;
24369c2e PB	387	__u32 wq_fd;
24369c2e PB	388	__u32 resv[3];
2b188cc1 JA	389	struct io_sqring_offsets sq_off;
	390	struct io_cqring_offsets cq_off;
	391	};
	392
ac90f249 JA	393	/*
	394	* io_uring_params->features flags
	395	*/
	396	#define IORING_FEAT_SINGLE_MMAP (1U << 0)
1d7bb1d5	397	#define IORING_FEAT_NODROP (1U << 1)
da8c9690	398	#define IORING_FEAT_SUBMIT_STABLE (1U << 2)
ba04291e	399	#define IORING_FEAT_RW_CUR_POS (1U << 3)
cccf0ee8	400	#define IORING_FEAT_CUR_PERSONALITY (1U << 4)
d7718a9d	401	#define IORING_FEAT_FAST_POLL (1U << 5)
5769a351	402	#define IORING_FEAT_POLL_32BITS (1U << 6)
28cea78a	403	#define IORING_FEAT_SQPOLL_NONFIXED (1U << 7)
c73ebb68	404	#define IORING_FEAT_EXT_ARG (1U << 8)
1c0aa1fa	405	#define IORING_FEAT_NATIVE_WORKERS (1U << 9)
9690557e	406	#define IORING_FEAT_RSRC_TAGS (1U << 10)
04c76b41	407	#define IORING_FEAT_CQE_SKIP (1U << 11)
c4212f3e	408	#define IORING_FEAT_LINKED_FILE (1U << 12)
ac90f249	409
edafccee JA	410	/*
	411	* io_uring_register(2) opcodes and arguments
	412	*/
9d4a75ef SG	413	enum {
	414	IORING_REGISTER_BUFFERS = 0,
	415	IORING_UNREGISTER_BUFFERS = 1,
	416	IORING_REGISTER_FILES = 2,
	417	IORING_UNREGISTER_FILES = 3,
	418	IORING_REGISTER_EVENTFD = 4,
	419	IORING_UNREGISTER_EVENTFD = 5,
	420	IORING_REGISTER_FILES_UPDATE = 6,
	421	IORING_REGISTER_EVENTFD_ASYNC = 7,
	422	IORING_REGISTER_PROBE = 8,
	423	IORING_REGISTER_PERSONALITY = 9,
	424	IORING_UNREGISTER_PERSONALITY = 10,
21b55dbc	425	IORING_REGISTER_RESTRICTIONS = 11,
7e84e1c7	426	IORING_REGISTER_ENABLE_RINGS = 12,
992da01a PB	427
	428	/* extended with tagging */
	429	IORING_REGISTER_FILES2 = 13,
	430	IORING_REGISTER_FILES_UPDATE2 = 14,
	431	IORING_REGISTER_BUFFERS2 = 15,
	432	IORING_REGISTER_BUFFERS_UPDATE = 16,
9d4a75ef	433
fe76421d JA	434	/* set/clear io-wq thread affinities */
	435	IORING_REGISTER_IOWQ_AFF = 17,
	436	IORING_UNREGISTER_IOWQ_AFF = 18,
	437
dd47c104	438	/* set/get max number of io-wq workers */
2e480058 JA	439	IORING_REGISTER_IOWQ_MAX_WORKERS = 19,
2e480058 JA	440
e7a6c00d JA	441	/* register/unregister io_uring fd with the ring */
	442	IORING_REGISTER_RING_FDS = 20,
	443	IORING_UNREGISTER_RING_FDS = 21,
	444
c7fb1942 JA	445	/* register ring based provide buffer group */
	446	IORING_REGISTER_PBUF_RING = 22,
	447	IORING_UNREGISTER_PBUF_RING = 23,
	448
78a861b9 JA	449	/* sync cancelation API */
	450	IORING_REGISTER_SYNC_CANCEL = 24,
	451
9d4a75ef SG	452	/* this goes last */
	453	IORING_REGISTER_LAST
	454	};
c3a31e60	455
dd47c104 ES	456	/* io-wq worker categories */
	457	enum {
	458	IO_WQ_BOUND,
	459	IO_WQ_UNBOUND,
	460	};
	461
269bbe5f	462	/* deprecated, see struct io_uring_rsrc_update */
c3a31e60 JA	463	struct io_uring_files_update {
c3a31e60 JA	464	__u32 offset;
1292e972 ES	465	__u32 resv;
1292e972 ES	466	__aligned_u64 /* __s32 * */ fds;
c3a31e60	467	};
edafccee	468
a8da73a3 JA	469	/*
	470	* Register a fully sparse file space, rather than pass in an array of all
	471	* -1 file descriptors.
	472	*/
	473	#define IORING_RSRC_REGISTER_SPARSE (1U << 0)
	474
792e3582	475	struct io_uring_rsrc_register {
792e3582	476	__u32 nr;
a8da73a3	477	__u32 flags;
992da01a	478	__u64 resv2;
792e3582 PB	479	__aligned_u64 data;
	480	__aligned_u64 tags;
	481	};
	482
c3bdad02 PB	483	struct io_uring_rsrc_update {
	484	__u32 offset;
	485	__u32 resv;
	486	__aligned_u64 data;
	487	};
	488
	489	struct io_uring_rsrc_update2 {
	490	__u32 offset;
	491	__u32 resv;
	492	__aligned_u64 data;
	493	__aligned_u64 tags;
c3bdad02	494	__u32 nr;
992da01a	495	__u32 resv2;
c3bdad02 PB	496	};
c3bdad02 PB	497
4e0377a1	498	/* Skip updating fd indexes set to this value in the fd table */
	499	#define IORING_REGISTER_FILES_SKIP (-2)
	500
66f4af93 JA	501	#define IO_URING_OP_SUPPORTED (1U << 0)
	502
	503	struct io_uring_probe_op {
	504	__u8 op;
	505	__u8 resv;
	506	__u16 flags; /* IO_URING_OP_* flags */
	507	__u32 resv2;
	508	};
	509
	510	struct io_uring_probe {
	511	__u8 last_op; /* last opcode supported */
	512	__u8 ops_len; /* length of ops[] array below */
	513	__u16 resv;
	514	__u32 resv2[3];
8fcf4c48	515	struct io_uring_probe_op ops[];
66f4af93 JA	516	};
66f4af93 JA	517
21b55dbc SG	518	struct io_uring_restriction {
	519	__u16 opcode;
	520	union {
	521	__u8 register_op; /* IORING_RESTRICTION_REGISTER_OP */
	522	__u8 sqe_op; /* IORING_RESTRICTION_SQE_OP */
	523	__u8 sqe_flags; /* IORING_RESTRICTION_SQE_FLAGS_* */
	524	};
	525	__u8 resv;
	526	__u32 resv2[3];
	527	};
	528
c7fb1942 JA	529	struct io_uring_buf {
	530	__u64 addr;
	531	__u32 len;
	532	__u16 bid;
	533	__u16 resv;
	534	};
	535
	536	struct io_uring_buf_ring {
	537	union {
	538	/*
	539	* To avoid spilling into more pages than we need to, the
	540	* ring tail is overlaid with the io_uring_buf->resv field.
	541	*/
	542	struct {
	543	__u64 resv1;
	544	__u32 resv2;
	545	__u16 resv3;
	546	__u16 tail;
	547	};
	548	struct io_uring_buf bufs[0];
	549	};
	550	};
	551
	552	/* argument for IORING_(UN)REGISTER_PBUF_RING */
	553	struct io_uring_buf_reg {
	554	__u64 ring_addr;
	555	__u32 ring_entries;
	556	__u16 bgid;
	557	__u16 pad;
	558	__u64 resv[3];
	559	};
	560
21b55dbc SG	561	/*
	562	* io_uring_restriction->opcode values
	563	*/
	564	enum {
	565	/* Allow an io_uring_register(2) opcode */
	566	IORING_RESTRICTION_REGISTER_OP = 0,
	567
	568	/* Allow an sqe opcode */
	569	IORING_RESTRICTION_SQE_OP = 1,
	570
	571	/* Allow sqe flags */
	572	IORING_RESTRICTION_SQE_FLAGS_ALLOWED = 2,
	573
	574	/* Require sqe flags (these flags must be set on each submission) */
	575	IORING_RESTRICTION_SQE_FLAGS_REQUIRED = 3,
	576
	577	IORING_RESTRICTION_LAST
	578	};
	579
c73ebb68 HX	580	struct io_uring_getevents_arg {
	581	__u64 sigmask;
	582	__u32 sigmask_sz;
	583	__u32 pad;
	584	__u64 ts;
	585	};
	586
78a861b9 JA	587	/*
	588	* Argument for IORING_REGISTER_SYNC_CANCEL
	589	*/
	590	struct io_uring_sync_cancel_reg {
	591	__u64 addr;
	592	__s32 fd;
	593	__u32 flags;
	594	struct __kernel_timespec timeout;
	595	__u64 pad[4];
	596	};
	597
2b188cc1	598	#endif