This is a file private kind of request. io_uring doesn't know what's in this command type, it's for the file_operations->uring_cmd() handler to deal with. Signed-off-by: Jens Axboe <axboe@xxxxxxxxx> --- fs/io_uring.c | 59 +++++++++++++++++++++++++++++++++++ include/linux/io_uring.h | 12 +++++++ include/uapi/linux/io_uring.h | 1 + 3 files changed, 72 insertions(+) diff --git a/fs/io_uring.c b/fs/io_uring.c index 03748faa5295..55c2714a591e 100644 --- a/fs/io_uring.c +++ b/fs/io_uring.c @@ -712,6 +712,7 @@ struct io_kiocb { struct io_shutdown shutdown; struct io_rename rename; struct io_unlink unlink; + struct io_uring_cmd uring_cmd; /* use only after cleaning per-op data, see io_clean_op() */ struct io_completion compl; }; @@ -805,6 +806,8 @@ struct io_op_def { unsigned needs_async_data : 1; /* should block plug */ unsigned plug : 1; + /* doesn't support personality */ + unsigned no_personality : 1; /* size of async data needed, if any */ unsigned short async_size; unsigned work_flags; @@ -998,6 +1001,11 @@ static const struct io_op_def io_op_defs[] = { .work_flags = IO_WQ_WORK_MM | IO_WQ_WORK_FILES | IO_WQ_WORK_FS | IO_WQ_WORK_BLKCG, }, + [IORING_OP_URING_CMD] = { + .needs_file = 1, + .no_personality = 1, + .work_flags = IO_WQ_WORK_MM | IO_WQ_WORK_BLKCG, + }, }; enum io_mem_account { @@ -3797,6 +3805,47 @@ static int io_unlinkat(struct io_kiocb *req, bool force_nonblock) return 0; } +static void io_uring_cmd_done(struct io_uring_cmd *cmd, ssize_t ret) +{ + struct io_kiocb *req = container_of(cmd, struct io_kiocb, uring_cmd); + + if (ret < 0) + req_set_fail_links(req); + io_req_complete(req, ret); +} + +static int io_uring_cmd_prep(struct io_kiocb *req, + const struct io_uring_sqe *sqe) +{ + struct io_uring_cmd *cmd = &req->uring_cmd; + + if (!req->file->f_op->uring_cmd) + return -EOPNOTSUPP; + + memcpy(&cmd->pdu, (void *) &sqe->off, sizeof(cmd->pdu)); + cmd->done = io_uring_cmd_done; + return 0; +} + +static int io_uring_cmd(struct io_kiocb *req, bool force_nonblock) +{ + enum io_uring_cmd_flags flags = 0; + struct file *file = req->file; + int ret; + + if (force_nonblock) + flags |= IO_URING_F_NONBLOCK; + + ret = file->f_op->uring_cmd(&req->uring_cmd, flags); + /* queued async, consumer will call ->done() when complete */ + if (ret == -EIOCBQUEUED) + return 0; + else if (ret < 0) + req_set_fail_links(req); + io_req_complete(req, ret); + return 0; +} + static int io_shutdown_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) { @@ -6093,6 +6142,8 @@ static int io_req_prep(struct io_kiocb *req, const struct io_uring_sqe *sqe) return io_renameat_prep(req, sqe); case IORING_OP_UNLINKAT: return io_unlinkat_prep(req, sqe); + case IORING_OP_URING_CMD: + return io_uring_cmd_prep(req, sqe); } printk_once(KERN_WARNING "io_uring: unhandled opcode %d\n", @@ -6351,6 +6402,9 @@ static int io_issue_sqe(struct io_kiocb *req, bool force_nonblock, case IORING_OP_UNLINKAT: ret = io_unlinkat(req, force_nonblock); break; + case IORING_OP_URING_CMD: + ret = io_uring_cmd(req, force_nonblock); + break; default: ret = -EINVAL; break; @@ -6865,6 +6919,9 @@ static int io_init_req(struct io_ring_ctx *ctx, struct io_kiocb *req, if (id) { struct io_identity *iod; + if (io_op_defs[req->opcode].no_personality) + return -EINVAL; + iod = idr_find(&ctx->personality_idr, id); if (unlikely(!iod)) return -EINVAL; @@ -10260,6 +10317,8 @@ static int __init io_uring_init(void) BUILD_BUG_SQE_ELEM(40, __u16, buf_index); BUILD_BUG_SQE_ELEM(42, __u16, personality); BUILD_BUG_SQE_ELEM(44, __s32, splice_fd_in); + BUILD_BUG_ON(offsetof(struct io_uring_sqe, user_data) != + offsetof(struct io_uring_pdu, reserved)); BUILD_BUG_ON(ARRAY_SIZE(io_op_defs) != IORING_OP_LAST); BUILD_BUG_ON(__REQ_F_LAST_BIT >= 8 * sizeof(int)); diff --git a/include/linux/io_uring.h b/include/linux/io_uring.h index 35b2d845704d..e4e822d86e22 100644 --- a/include/linux/io_uring.h +++ b/include/linux/io_uring.h @@ -34,6 +34,18 @@ struct io_uring_task { bool sqpoll; }; +struct io_uring_pdu { + __u64 data[4]; /* available for free use */ + __u64 reserved; /* can't be used by application! */ + __u64 data2; /* available or free use */ +}; + +struct io_uring_cmd { + struct file *file; + struct io_uring_pdu pdu; + void (*done)(struct io_uring_cmd *, ssize_t); +}; + #if defined(CONFIG_IO_URING) struct sock *io_uring_get_socket(struct file *file); void __io_uring_task_cancel(void); diff --git a/include/uapi/linux/io_uring.h b/include/uapi/linux/io_uring.h index ac4e1738a9af..0a0de40a3a5c 100644 --- a/include/uapi/linux/io_uring.h +++ b/include/uapi/linux/io_uring.h @@ -137,6 +137,7 @@ enum { IORING_OP_SHUTDOWN, IORING_OP_RENAMEAT, IORING_OP_UNLINKAT, + IORING_OP_URING_CMD, /* this goes last, obviously */ IORING_OP_LAST, -- 2.30.0