From 423d5081d0451faa59a707e57373801da5b40141 Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 19 Jan 2023 09:01:27 -0700 Subject: io_uring/msg_ring: move double lock/unlock helpers higher up In preparation for needing them somewhere else, move them and get rid of the unused 'issue_flags' for the unlock side. No functional changes in this patch. Signed-off-by: Jens Axboe --- io_uring/msg_ring.c | 47 +++++++++++++++++++++++------------------------ 1 file changed, 23 insertions(+), 24 deletions(-) diff --git a/io_uring/msg_ring.c b/io_uring/msg_ring.c index 2d3cd945a531..321f5eafef99 100644 --- a/io_uring/msg_ring.c +++ b/io_uring/msg_ring.c @@ -25,6 +25,28 @@ struct io_msg { u32 flags; }; +static void io_double_unlock_ctx(struct io_ring_ctx *octx) +{ + mutex_unlock(&octx->uring_lock); +} + +static int io_double_lock_ctx(struct io_ring_ctx *octx, + unsigned int issue_flags) +{ + /* + * To ensure proper ordering between the two ctxs, we can only + * attempt a trylock on the target. If that fails and we already have + * the source ctx lock, punt to io-wq. + */ + if (!(issue_flags & IO_URING_F_UNLOCKED)) { + if (!mutex_trylock(&octx->uring_lock)) + return -EAGAIN; + return 0; + } + mutex_lock(&octx->uring_lock); + return 0; +} + void io_msg_ring_cleanup(struct io_kiocb *req) { struct io_msg *msg = io_kiocb_to_cmd(req, struct io_msg); @@ -77,29 +99,6 @@ static int io_msg_ring_data(struct io_kiocb *req) return -EOVERFLOW; } -static void io_double_unlock_ctx(struct io_ring_ctx *octx, - unsigned int issue_flags) -{ - mutex_unlock(&octx->uring_lock); -} - -static int io_double_lock_ctx(struct io_ring_ctx *octx, - unsigned int issue_flags) -{ - /* - * To ensure proper ordering between the two ctxs, we can only - * attempt a trylock on the target. If that fails and we already have - * the source ctx lock, punt to io-wq. - */ - if (!(issue_flags & IO_URING_F_UNLOCKED)) { - if (!mutex_trylock(&octx->uring_lock)) - return -EAGAIN; - return 0; - } - mutex_lock(&octx->uring_lock); - return 0; -} - static struct file *io_msg_grab_file(struct io_kiocb *req, unsigned int issue_flags) { struct io_msg *msg = io_kiocb_to_cmd(req, struct io_msg); @@ -148,7 +147,7 @@ static int io_msg_install_complete(struct io_kiocb *req, unsigned int issue_flag if (!io_post_aux_cqe(target_ctx, msg->user_data, msg->len, 0)) ret = -EOVERFLOW; out_unlock: - io_double_unlock_ctx(target_ctx, issue_flags); + io_double_unlock_ctx(target_ctx); return ret; } -- cgit v1.2.3 From e12d7a46f65ae4b7d58a5e0c1cbfa825cf8d830d Mon Sep 17 00:00:00 2001 From: Jens Axboe Date: Thu, 19 Jan 2023 09:04:40 -0700 Subject: io_uring/msg_ring: fix missing lock on overflow for IOPOLL If the target ring is configured with IOPOLL, then we always need to hold the target ring uring_lock before posting CQEs. We could just grab it unconditionally, but since we don't expect many target rings to be of this type, make grabbing the uring_lock conditional on the ring type. Link: https://lore.kernel.org/io-uring/Y8krlYa52%2F0YGqkg@ip-172-31-85-199.ec2.internal/ Reported-by: Xingyuan Mo Signed-off-by: Jens Axboe --- io_uring/msg_ring.c | 39 ++++++++++++++++++++++++++++++--------- 1 file changed, 30 insertions(+), 9 deletions(-) diff --git a/io_uring/msg_ring.c b/io_uring/msg_ring.c index 321f5eafef99..a333781565d3 100644 --- a/io_uring/msg_ring.c +++ b/io_uring/msg_ring.c @@ -65,20 +65,33 @@ static void io_msg_tw_complete(struct callback_head *head) struct io_ring_ctx *target_ctx = req->file->private_data; int ret = 0; - if (current->flags & PF_EXITING) + if (current->flags & PF_EXITING) { ret = -EOWNERDEAD; - else if (!io_post_aux_cqe(target_ctx, msg->user_data, msg->len, 0)) - ret = -EOVERFLOW; + } else { + /* + * If the target ring is using IOPOLL mode, then we need to be + * holding the uring_lock for posting completions. Other ring + * types rely on the regular completion locking, which is + * handled while posting. + */ + if (target_ctx->flags & IORING_SETUP_IOPOLL) + mutex_lock(&target_ctx->uring_lock); + if (!io_post_aux_cqe(target_ctx, msg->user_data, msg->len, 0)) + ret = -EOVERFLOW; + if (target_ctx->flags & IORING_SETUP_IOPOLL) + mutex_unlock(&target_ctx->uring_lock); + } if (ret < 0) req_set_fail(req); io_req_queue_tw_complete(req, ret); } -static int io_msg_ring_data(struct io_kiocb *req) +static int io_msg_ring_data(struct io_kiocb *req, unsigned int issue_flags) { struct io_ring_ctx *target_ctx = req->file->private_data; struct io_msg *msg = io_kiocb_to_cmd(req, struct io_msg); + int ret; if (msg->src_fd || msg->dst_fd || msg->flags) return -EINVAL; @@ -93,10 +106,18 @@ static int io_msg_ring_data(struct io_kiocb *req) return IOU_ISSUE_SKIP_COMPLETE; } - if (io_post_aux_cqe(target_ctx, msg->user_data, msg->len, 0)) - return 0; - - return -EOVERFLOW; + ret = -EOVERFLOW; + if (target_ctx->flags & IORING_SETUP_IOPOLL) { + if (unlikely(io_double_lock_ctx(target_ctx, issue_flags))) + return -EAGAIN; + if (io_post_aux_cqe(target_ctx, msg->user_data, msg->len, 0)) + ret = 0; + io_double_unlock_ctx(target_ctx); + } else { + if (io_post_aux_cqe(target_ctx, msg->user_data, msg->len, 0)) + ret = 0; + } + return ret; } static struct file *io_msg_grab_file(struct io_kiocb *req, unsigned int issue_flags) @@ -223,7 +244,7 @@ int io_msg_ring(struct io_kiocb *req, unsigned int issue_flags) switch (msg->cmd) { case IORING_MSG_DATA: - ret = io_msg_ring_data(req); + ret = io_msg_ring_data(req, issue_flags); break; case IORING_MSG_SEND_FD: ret = io_msg_send_fd(req, issue_flags); -- cgit v1.2.3 From 56d8e3180c065c9b78ed77afcd0cf99677a4e22f Mon Sep 17 00:00:00 2001 From: Pavel Begunkov Date: Fri, 20 Jan 2023 16:38:05 +0000 Subject: io_uring/msg_ring: fix flagging remote execution There is a couple of problems with queueing a tw in io_msg_ring_data() for remote execution. First, once we queue it the target ring can go away and so setting IORING_SQ_TASKRUN there is not safe. Secondly, the userspace might not expect IORING_SQ_TASKRUN. Extract a helper and uniformly use TWA_SIGNAL without TWA_SIGNAL_NO_IPI tricks for now, just as it was done in the original patch. Cc: stable@vger.kernel.org Fixes: 6d043ee1164ca ("io_uring: do msg_ring in target task via tw") Signed-off-by: Pavel Begunkov Signed-off-by: Jens Axboe --- io_uring/msg_ring.c | 40 +++++++++++++++++++++++----------------- 1 file changed, 23 insertions(+), 17 deletions(-) diff --git a/io_uring/msg_ring.c b/io_uring/msg_ring.c index a333781565d3..bb868447dcdf 100644 --- a/io_uring/msg_ring.c +++ b/io_uring/msg_ring.c @@ -58,6 +58,25 @@ void io_msg_ring_cleanup(struct io_kiocb *req) msg->src_file = NULL; } +static inline bool io_msg_need_remote(struct io_ring_ctx *target_ctx) +{ + if (!target_ctx->task_complete) + return false; + return current != target_ctx->submitter_task; +} + +static int io_msg_exec_remote(struct io_kiocb *req, task_work_func_t func) +{ + struct io_ring_ctx *ctx = req->file->private_data; + struct io_msg *msg = io_kiocb_to_cmd(req, struct io_msg); + + init_task_work(&msg->tw, func); + if (task_work_add(ctx->submitter_task, &msg->tw, TWA_SIGNAL)) + return -EOWNERDEAD; + + return IOU_ISSUE_SKIP_COMPLETE; +} + static void io_msg_tw_complete(struct callback_head *head) { struct io_msg *msg = container_of(head, struct io_msg, tw); @@ -96,15 +115,8 @@ static int io_msg_ring_data(struct io_kiocb *req, unsigned int issue_flags) if (msg->src_fd || msg->dst_fd || msg->flags) return -EINVAL; - if (target_ctx->task_complete && current != target_ctx->submitter_task) { - init_task_work(&msg->tw, io_msg_tw_complete); - if (task_work_add(target_ctx->submitter_task, &msg->tw, - TWA_SIGNAL_NO_IPI)) - return -EOWNERDEAD; - - atomic_or(IORING_SQ_TASKRUN, &target_ctx->rings->sq_flags); - return IOU_ISSUE_SKIP_COMPLETE; - } + if (io_msg_need_remote(target_ctx)) + return io_msg_exec_remote(req, io_msg_tw_complete); ret = -EOVERFLOW; if (target_ctx->flags & IORING_SETUP_IOPOLL) { @@ -202,14 +214,8 @@ static int io_msg_send_fd(struct io_kiocb *req, unsigned int issue_flags) req->flags |= REQ_F_NEED_CLEANUP; } - if (target_ctx->task_complete && current != target_ctx->submitter_task) { - init_task_work(&msg->tw, io_msg_tw_fd_complete); - if (task_work_add(target_ctx->submitter_task, &msg->tw, - TWA_SIGNAL)) - return -EOWNERDEAD; - - return IOU_ISSUE_SKIP_COMPLETE; - } + if (io_msg_need_remote(target_ctx)) + return io_msg_exec_remote(req, io_msg_tw_fd_complete); return io_msg_install_complete(req, issue_flags); } -- cgit v1.2.3 From 8579538c89e33ce78be2feb41e07489c8cbf8f31 Mon Sep 17 00:00:00 2001 From: Pavel Begunkov Date: Fri, 20 Jan 2023 16:38:06 +0000 Subject: io_uring/msg_ring: fix remote queue to disabled ring IORING_SETUP_R_DISABLED rings don't have the submitter task set, so it's not always safe to use ->submitter_task. Disallow posting msg_ring messaged to disabled rings. Also add task NULL check for loosy sync around testing for IORING_SETUP_R_DISABLED. Cc: stable@vger.kernel.org Fixes: 6d043ee1164ca ("io_uring: do msg_ring in target task via tw") Signed-off-by: Pavel Begunkov Signed-off-by: Jens Axboe --- io_uring/io_uring.c | 4 ++-- io_uring/msg_ring.c | 8 ++++++++ 2 files changed, 10 insertions(+), 2 deletions(-) diff --git a/io_uring/io_uring.c b/io_uring/io_uring.c index 2ac1cd8d23ea..0a4efada9b3c 100644 --- a/io_uring/io_uring.c +++ b/io_uring/io_uring.c @@ -3674,7 +3674,7 @@ static __cold int io_uring_create(unsigned entries, struct io_uring_params *p, if (ctx->flags & IORING_SETUP_SINGLE_ISSUER && !(ctx->flags & IORING_SETUP_R_DISABLED)) - ctx->submitter_task = get_task_struct(current); + WRITE_ONCE(ctx->submitter_task, get_task_struct(current)); file = io_uring_get_file(ctx); if (IS_ERR(file)) { @@ -3868,7 +3868,7 @@ static int io_register_enable_rings(struct io_ring_ctx *ctx) return -EBADFD; if (ctx->flags & IORING_SETUP_SINGLE_ISSUER && !ctx->submitter_task) - ctx->submitter_task = get_task_struct(current); + WRITE_ONCE(ctx->submitter_task, get_task_struct(current)); if (ctx->restrictions.registered) ctx->restricted = 1; diff --git a/io_uring/msg_ring.c b/io_uring/msg_ring.c index bb868447dcdf..15602a136821 100644 --- a/io_uring/msg_ring.c +++ b/io_uring/msg_ring.c @@ -69,6 +69,10 @@ static int io_msg_exec_remote(struct io_kiocb *req, task_work_func_t func) { struct io_ring_ctx *ctx = req->file->private_data; struct io_msg *msg = io_kiocb_to_cmd(req, struct io_msg); + struct task_struct *task = READ_ONCE(ctx->submitter_task); + + if (unlikely(!task)) + return -EOWNERDEAD; init_task_work(&msg->tw, func); if (task_work_add(ctx->submitter_task, &msg->tw, TWA_SIGNAL)) @@ -114,6 +118,8 @@ static int io_msg_ring_data(struct io_kiocb *req, unsigned int issue_flags) if (msg->src_fd || msg->dst_fd || msg->flags) return -EINVAL; + if (target_ctx->flags & IORING_SETUP_R_DISABLED) + return -EBADFD; if (io_msg_need_remote(target_ctx)) return io_msg_exec_remote(req, io_msg_tw_complete); @@ -206,6 +212,8 @@ static int io_msg_send_fd(struct io_kiocb *req, unsigned int issue_flags) if (target_ctx == ctx) return -EINVAL; + if (target_ctx->flags & IORING_SETUP_R_DISABLED) + return -EBADFD; if (!src_file) { src_file = io_msg_grab_file(req, issue_flags); if (!src_file) -- cgit v1.2.3