1 files changed, 208 insertions, 57 deletions

diff --git a/fs/io_uring.c b/fs/io_uring.c
index 493e5047e67c..26978630378e 100644
--- a/fs/io_uring.c
+++ b/fs/io_uring.c
@@ -669,12 +669,12 @@ struct io_kiocb {
 		 * restore the work, if needed.
 		 */
 		struct {
-			struct callback_head	task_work;
 			struct hlist_node	hash_node;
 			struct async_poll	*apoll;
 		};
 		struct io_wq_work	work;
 	};
+	struct callback_head	task_work;
 };
 
 #define IO_PLUG_THRESHOLD		2
@@ -1549,12 +1549,9 @@ static void io_req_link_next(struct io_kiocb *req, struct io_kiocb **nxtptr)
 /*
  * Called if REQ_F_LINK_HEAD is set, and we fail the head request
  */
-static void io_fail_links(struct io_kiocb *req)
+static void __io_fail_links(struct io_kiocb *req)
 {
 	struct io_ring_ctx *ctx = req->ctx;
-	unsigned long flags;
-
-	spin_lock_irqsave(&ctx->completion_lock, flags);
 
 	while (!list_empty(&req->link_list)) {
 		struct io_kiocb *link = list_first_entry(&req->link_list,
@@ -1568,13 +1565,29 @@ static void io_fail_links(struct io_kiocb *req)
 			io_link_cancel_timeout(link);
 		} else {
 			io_cqring_fill_event(link, -ECANCELED);
+			link->flags |= REQ_F_COMP_LOCKED;
 			__io_double_put_req(link);
 		}
 		req->flags &= ~REQ_F_LINK_TIMEOUT;
 	}
 
 	io_commit_cqring(ctx);
-	spin_unlock_irqrestore(&ctx->completion_lock, flags);
+}
+
+static void io_fail_links(struct io_kiocb *req)
+{
+	struct io_ring_ctx *ctx = req->ctx;
+
+	if (!(req->flags & REQ_F_COMP_LOCKED)) {
+		unsigned long flags;
+
+		spin_lock_irqsave(&ctx->completion_lock, flags);
+		__io_fail_links(req);
+		spin_unlock_irqrestore(&ctx->completion_lock, flags);
+	} else {
+		__io_fail_links(req);
+	}
+
 	io_cqring_ev_posted(ctx);
 }
 
@@ -1747,6 +1760,17 @@ static int io_put_kbuf(struct io_kiocb *req)
 	return cflags;
 }
 
+static inline bool io_run_task_work(void)
+{
+	if (current->task_works) {
+		__set_current_state(TASK_RUNNING);
+		task_work_run();
+		return true;
+	}
+
+	return false;
+}
+
 static void io_iopoll_queue(struct list_head *again)
 {
 	struct io_kiocb *req;
@@ -1936,6 +1960,7 @@ static int io_iopoll_check(struct io_ring_ctx *ctx, unsigned *nr_events,
 		 */
 		if (!(++iters & 7)) {
 			mutex_unlock(&ctx->uring_lock);
+			io_run_task_work();
 			mutex_lock(&ctx->uring_lock);
 		}
 
@@ -2661,8 +2686,10 @@ static int io_read(struct io_kiocb *req, bool force_nonblock)
 
 		if (req->file->f_op->read_iter)
 			ret2 = call_read_iter(req->file, kiocb, &iter);
-		else
+		else if (req->file->f_op->read)
 			ret2 = loop_rw_iter(READ, req->file, kiocb, &iter);
+		else
+			ret2 = -EINVAL;
 
 		/* Catch -EAGAIN return for forced non-blocking submission */
 		if (!force_nonblock || ret2 != -EAGAIN) {
@@ -2776,8 +2803,10 @@ static int io_write(struct io_kiocb *req, bool force_nonblock)
 
 		if (req->file->f_op->write_iter)
 			ret2 = call_write_iter(req->file, kiocb, &iter);
-		else
+		else if (req->file->f_op->write)
 			ret2 = loop_rw_iter(WRITE, req->file, kiocb, &iter);
+		else
+			ret2 = -EINVAL;
 
 		if (!force_nonblock)
 			current->signal->rlim[RLIMIT_FSIZE].rlim_cur = RLIM_INFINITY;
@@ -4088,22 +4117,22 @@ static int io_req_task_work_add(struct io_kiocb *req, struct callback_head *cb)
 {
 	struct task_struct *tsk = req->task;
 	struct io_ring_ctx *ctx = req->ctx;
-	int ret, notify = TWA_RESUME;
+	int ret, notify;
 
 	/*
-	 * SQPOLL kernel thread doesn't need notification, just a wakeup.
-	 * If we're not using an eventfd, then TWA_RESUME is always fine,
-	 * as we won't have dependencies between request completions for
-	 * other kernel wait conditions.
+	 * SQPOLL kernel thread doesn't need notification, just a wakeup. For
+	 * all other cases, use TWA_SIGNAL unconditionally to ensure we're
+	 * processing task_work. There's no reliable way to tell if TWA_RESUME
+	 * will do the job.
 	 */
-	if (ctx->flags & IORING_SETUP_SQPOLL)
-		notify = 0;
-	else if (ctx->cq_ev_fd)
+	notify = 0;
+	if (!(ctx->flags & IORING_SETUP_SQPOLL))
 		notify = TWA_SIGNAL;
 
 	ret = task_work_add(tsk, cb, notify);
 	if (!ret)
 		wake_up_process(tsk);
+
 	return ret;
 }
 
@@ -4124,6 +4153,8 @@ static int __io_async_wake(struct io_kiocb *req, struct io_poll_iocb *poll,
 	tsk = req->task;
 	req->result = mask;
 	init_task_work(&req->task_work, func);
+	percpu_ref_get(&req->ctx->refs);
+
 	/*
 	 * If this fails, then the task is exiting. When a task exits, the
 	 * work gets canceled, so just cancel this request as well instead
@@ -4160,9 +4191,24 @@ static bool io_poll_rewait(struct io_kiocb *req, struct io_poll_iocb *poll)
 	return false;
 }
 
-static void io_poll_remove_double(struct io_kiocb *req, void *data)
+static struct io_poll_iocb *io_poll_get_double(struct io_kiocb *req)
+{
+	/* pure poll stashes this in ->io, poll driven retry elsewhere */
+	if (req->opcode == IORING_OP_POLL_ADD)
+		return (struct io_poll_iocb *) req->io;
+	return req->apoll->double_poll;
+}
+
+static struct io_poll_iocb *io_poll_get_single(struct io_kiocb *req)
+{
+	if (req->opcode == IORING_OP_POLL_ADD)
+		return &req->poll;
+	return &req->apoll->poll;
+}
+
+static void io_poll_remove_double(struct io_kiocb *req)
 {
-	struct io_poll_iocb *poll = data;
+	struct io_poll_iocb *poll = io_poll_get_double(req);
 
 	lockdep_assert_held(&req->ctx->completion_lock);
 
@@ -4182,7 +4228,7 @@ static void io_poll_complete(struct io_kiocb *req, __poll_t mask, int error)
 {
 	struct io_ring_ctx *ctx = req->ctx;
 
-	io_poll_remove_double(req, req->io);
+	io_poll_remove_double(req);
 	req->poll.done = true;
 	io_cqring_fill_event(req, error ? error : mangle_poll(mask));
 	io_commit_cqring(ctx);
@@ -4208,6 +4254,7 @@ static void io_poll_task_handler(struct io_kiocb *req, struct io_kiocb **nxt)
 static void io_poll_task_func(struct callback_head *cb)
 {
 	struct io_kiocb *req = container_of(cb, struct io_kiocb, task_work);
+	struct io_ring_ctx *ctx = req->ctx;
 	struct io_kiocb *nxt = NULL;
 
 	io_poll_task_handler(req, &nxt);
@@ -4218,13 +4265,14 @@ static void io_poll_task_func(struct callback_head *cb)
 		__io_queue_sqe(nxt, NULL);
 		mutex_unlock(&ctx->uring_lock);
 	}
+	percpu_ref_put(&ctx->refs);
 }
 
 static int io_poll_double_wake(struct wait_queue_entry *wait, unsigned mode,
 			       int sync, void *key)
 {
 	struct io_kiocb *req = wait->private;
-	struct io_poll_iocb *poll = req->apoll->double_poll;
+	struct io_poll_iocb *poll = io_poll_get_single(req);
 	__poll_t mask = key_to_poll(key);
 
 	/* for instances that support it check for an event match first: */
@@ -4238,6 +4286,8 @@ static int io_poll_double_wake(struct wait_queue_entry *wait, unsigned mode,
 		done = list_empty(&poll->wait.entry);
 		if (!done)
 			list_del_init(&poll->wait.entry);
+		/* make sure double remove sees this as being gone */
+		wait->private = NULL;
 		spin_unlock(&poll->head->lock);
 		if (!done)
 			__io_async_wake(req, poll, mask, io_poll_task_func);
@@ -4313,7 +4363,8 @@ static int io_sq_thread_acquire_mm(struct io_ring_ctx *ctx,
 				   struct io_kiocb *req)
 {
 	if (io_op_defs[req->opcode].needs_mm && !current->mm) {
-		if (unlikely(!mmget_not_zero(ctx->sqo_mm)))
+		if (unlikely(!(ctx->flags & IORING_SETUP_SQPOLL) ||
+			     !mmget_not_zero(ctx->sqo_mm)))
 			return -EFAULT;
 		kthread_use_mm(ctx->sqo_mm);
 	}
@@ -4332,6 +4383,7 @@ static void io_async_task_func(struct callback_head *cb)
 
 	if (io_poll_rewait(req, &apoll->poll)) {
 		spin_unlock_irq(&ctx->completion_lock);
+		percpu_ref_put(&ctx->refs);
 		return;
 	}
 
@@ -4346,7 +4398,7 @@ static void io_async_task_func(struct callback_head *cb)
 		}
 	}
 
-	io_poll_remove_double(req, apoll->double_poll);
+	io_poll_remove_double(req);
 	spin_unlock_irq(&ctx->completion_lock);
 
 	/* restore ->work in case we need to retry again */
@@ -4356,7 +4408,6 @@ static void io_async_task_func(struct callback_head *cb)
 	kfree(apoll);
 
 	if (!canceled) {
-		__set_current_state(TASK_RUNNING);
 		if (io_sq_thread_acquire_mm(ctx, req)) {
 			io_cqring_add_event(req, -EFAULT);
 			goto end_req;
@@ -4370,6 +4421,7 @@ end_req:
 		req_set_fail_links(req);
 		io_double_put_req(req);
 	}
+	percpu_ref_put(&ctx->refs);
 }
 
 static int io_async_wake(struct wait_queue_entry *wait, unsigned mode, int sync,
@@ -4472,8 +4524,8 @@ static bool io_arm_poll_handler(struct io_kiocb *req)
 
 	ret = __io_arm_poll_handler(req, &apoll->poll, &ipt, mask,
 					io_async_wake);
-	if (ret) {
-		io_poll_remove_double(req, apoll->double_poll);
+	if (ret || ipt.error) {
+		io_poll_remove_double(req);
 		spin_unlock_irq(&ctx->completion_lock);
 		if (req->flags & REQ_F_WORK_INITIALIZED)
 			memcpy(&req->work, &apoll->work, sizeof(req->work));
@@ -4507,14 +4559,13 @@ static bool io_poll_remove_one(struct io_kiocb *req)
 {
 	bool do_complete;
 
+	io_poll_remove_double(req);
+
 	if (req->opcode == IORING_OP_POLL_ADD) {
-		io_poll_remove_double(req, req->io);
 		do_complete = __io_poll_remove_one(req, &req->poll);
 	} else {
 		struct async_poll *apoll = req->apoll;
 
-		io_poll_remove_double(req, apoll->double_poll);
-
 		/* non-poll requests have submit ref still */
 		do_complete = __io_poll_remove_one(req, &apoll->poll);
 		if (do_complete) {
@@ -4536,6 +4587,7 @@ static bool io_poll_remove_one(struct io_kiocb *req)
 		io_cqring_fill_event(req, -ECANCELED);
 		io_commit_cqring(req->ctx);
 		req->flags |= REQ_F_COMP_LOCKED;
+		req_set_fail_links(req);
 		io_put_req(req);
 	}
 
@@ -4709,6 +4761,23 @@ static enum hrtimer_restart io_timeout_fn(struct hrtimer *timer)
 	return HRTIMER_NORESTART;
 }
 
+static int __io_timeout_cancel(struct io_kiocb *req)
+{
+	int ret;
+
+	list_del_init(&req->list);
+
+	ret = hrtimer_try_to_cancel(&req->io->timeout.timer);
+	if (ret == -1)
+		return -EALREADY;
+
+	req_set_fail_links(req);
+	req->flags |= REQ_F_COMP_LOCKED;
+	io_cqring_fill_event(req, -ECANCELED);
+	io_put_req(req);
+	return 0;
+}
+
 static int io_timeout_cancel(struct io_ring_ctx *ctx, __u64 user_data)
 {
 	struct io_kiocb *req;
@@ -4716,7 +4785,6 @@ static int io_timeout_cancel(struct io_ring_ctx *ctx, __u64 user_data)
 
 	list_for_each_entry(req, &ctx->timeout_list, list) {
 		if (user_data == req->user_data) {
-			list_del_init(&req->list);
 			ret = 0;
 			break;
 		}
@@ -4725,14 +4793,7 @@ static int io_timeout_cancel(struct io_ring_ctx *ctx, __u64 user_data)
 	if (ret == -ENOENT)
 		return ret;
 
-	ret = hrtimer_try_to_cancel(&req->io->timeout.timer);
-	if (ret == -1)
-		return -EALREADY;
-
-	req_set_fail_links(req);
-	io_cqring_fill_event(req, -ECANCELED);
-	io_put_req(req);
-	return 0;
+	return __io_timeout_cancel(req);
 }
 
 static int io_timeout_remove_prep(struct io_kiocb *req,
@@ -6082,8 +6143,7 @@ static int io_sq_thread(void *data)
 			if (!list_empty(&ctx->poll_list) || need_resched() ||
 			    (!time_after(jiffies, timeout) && ret != -EBUSY &&
 			    !percpu_ref_is_dying(&ctx->refs))) {
-				if (current->task_works)
-					task_work_run();
+				io_run_task_work();
 				cond_resched();
 				continue;
 			}
@@ -6115,8 +6175,7 @@ static int io_sq_thread(void *data)
 					finish_wait(&ctx->sqo_wait, &wait);
 					break;
 				}
-				if (current->task_works) {
-					task_work_run();
+				if (io_run_task_work()) {
 					finish_wait(&ctx->sqo_wait, &wait);
 					continue;
 				}
@@ -6145,8 +6204,7 @@ static int io_sq_thread(void *data)
 		timeout = jiffies + ctx->sq_thread_idle;
 	}
 
-	if (current->task_works)
-		task_work_run();
+	io_run_task_work();
 
 	io_sq_thread_drop_mm(ctx);
 	revert_creds(old_cred);
@@ -6211,9 +6269,8 @@ static int io_cqring_wait(struct io_ring_ctx *ctx, int min_events,
 	do {
 		if (io_cqring_events(ctx, false) >= min_events)
 			return 0;
-		if (!current->task_works)
+		if (!io_run_task_work())
 			break;
-		task_work_run();
 	} while (1);
 
 	if (sig) {
@@ -6235,8 +6292,8 @@ static int io_cqring_wait(struct io_ring_ctx *ctx, int min_events,
 		prepare_to_wait_exclusive(&ctx->wait, &iowq.wq,
 						TASK_INTERRUPTIBLE);
 		/* make sure we run task_work before checking for signals */
-		if (current->task_works)
-			task_work_run();
+		if (io_run_task_work())
+			continue;
 		if (signal_pending(current)) {
 			if (current->jobctl & JOBCTL_TASK_WORK) {
 				spin_lock_irq(&current->sighand->siglock);
@@ -7086,6 +7143,9 @@ static unsigned long rings_size(unsigned sq_entries, unsigned cq_entries,
 		return SIZE_MAX;
 #endif
 
+	if (sq_offset)
+		*sq_offset = off;
+
 	sq_array_size = array_size(sizeof(u32), sq_entries);
 	if (sq_array_size == SIZE_MAX)
 		return SIZE_MAX;
@@ -7093,9 +7153,6 @@ static unsigned long rings_size(unsigned sq_entries, unsigned cq_entries,
 	if (check_add_overflow(off, sq_array_size, &off))
 		return SIZE_MAX;
 
-	if (sq_offset)
-		*sq_offset = off;
-
 	return off;
 }
 
@@ -7488,6 +7545,98 @@ static bool io_wq_files_match(struct io_wq_work *work, void *data)
 	return work->files == files;
 }
 
+/*
+ * Returns true if 'preq' is the link parent of 'req'
+ */
+static bool io_match_link(struct io_kiocb *preq, struct io_kiocb *req)
+{
+	struct io_kiocb *link;
+
+	if (!(preq->flags & REQ_F_LINK_HEAD))
+		return false;
+
+	list_for_each_entry(link, &preq->link_list, link_list) {
+		if (link == req)
+			return true;
+	}
+
+	return false;
+}
+
+/*
+ * We're looking to cancel 'req' because it's holding on to our files, but
+ * 'req' could be a link to another request. See if it is, and cancel that
+ * parent request if so.
+ */
+static bool io_poll_remove_link(struct io_ring_ctx *ctx, struct io_kiocb *req)
+{
+	struct hlist_node *tmp;
+	struct io_kiocb *preq;
+	bool found = false;
+	int i;
+
+	spin_lock_irq(&ctx->completion_lock);
+	for (i = 0; i < (1U << ctx->cancel_hash_bits); i++) {
+		struct hlist_head *list;
+
+		list = &ctx->cancel_hash[i];
+		hlist_for_each_entry_safe(preq, tmp, list, hash_node) {
+			found = io_match_link(preq, req);
+			if (found) {
+				io_poll_remove_one(preq);
+				break;
+			}
+		}
+	}
+	spin_unlock_irq(&ctx->completion_lock);
+	return found;
+}
+
+static bool io_timeout_remove_link(struct io_ring_ctx *ctx,
+				   struct io_kiocb *req)
+{
+	struct io_kiocb *preq;
+	bool found = false;
+
+	spin_lock_irq(&ctx->completion_lock);
+	list_for_each_entry(preq, &ctx->timeout_list, list) {
+		found = io_match_link(preq, req);
+		if (found) {
+			__io_timeout_cancel(preq);
+			break;
+		}
+	}
+	spin_unlock_irq(&ctx->completion_lock);
+	return found;
+}
+
+static bool io_cancel_link_cb(struct io_wq_work *work, void *data)
+{
+	return io_match_link(container_of(work, struct io_kiocb, work), data);
+}
+
+static void io_attempt_cancel(struct io_ring_ctx *ctx, struct io_kiocb *req)
+{
+	enum io_wq_cancel cret;
+
+	/* cancel this particular work, if it's running */
+	cret = io_wq_cancel_work(ctx->io_wq, &req->work);
+	if (cret != IO_WQ_CANCEL_NOTFOUND)
+		return;
+
+	/* find links that hold this pending, cancel those */
+	cret = io_wq_cancel_cb(ctx->io_wq, io_cancel_link_cb, req, true);
+	if (cret != IO_WQ_CANCEL_NOTFOUND)
+		return;
+
+	/* if we have a poll link holding this pending, cancel that */
+	if (io_poll_remove_link(ctx, req))
+		return;
+
+	/* final option, timeout link is holding this req pending */
+	io_timeout_remove_link(ctx, req);
+}
+
 static void io_uring_cancel_files(struct io_ring_ctx *ctx,
 				  struct files_struct *files)
 {
@@ -7529,10 +7678,10 @@ static void io_uring_cancel_files(struct io_ring_ctx *ctx,
 				clear_bit(0, &ctx->cq_check_overflow);
 				ctx->rings->sq_flags &= ~IORING_SQ_CQ_OVERFLOW;
 			}
-			spin_unlock_irq(&ctx->completion_lock);
-
 			WRITE_ONCE(ctx->rings->cq_overflow,
 				atomic_inc_return(&ctx->cached_cq_overflow));
+			io_commit_cqring(ctx);
+			spin_unlock_irq(&ctx->completion_lock);
 
 			/*
 			 * Put inflight ref and overflow ref. If that's
@@ -7544,7 +7693,8 @@ static void io_uring_cancel_files(struct io_ring_ctx *ctx,
 				continue;
 			}
 		} else {
-			io_wq_cancel_work(ctx->io_wq, &cancel_req->work);
+			/* cancel this request, or head link requests */
+			io_attempt_cancel(ctx, cancel_req);
 			io_put_req(cancel_req);
 		}
 
@@ -7655,8 +7805,7 @@ SYSCALL_DEFINE6(io_uring_enter, unsigned int, fd, u32, to_submit,
 	int submitted = 0;
 	struct fd f;
 
-	if (current->task_works)
-		task_work_run();
+	io_run_task_work();
 
 	if (flags & ~(IORING_ENTER_GETEVENTS | IORING_ENTER_SQ_WAKEUP))
 		return -EINVAL;
@@ -7828,6 +7977,10 @@ static int io_allocate_scq_urings(struct io_ring_ctx *ctx,
 	struct io_rings *rings;
 	size_t size, sq_array_offset;
 
+	/* make sure these are sane, as we already accounted them */
+	ctx->sq_entries = p->sq_entries;
+	ctx->cq_entries = p->cq_entries;
+
 	size = rings_size(p->sq_entries, p->cq_entries, &sq_array_offset);
 	if (size == SIZE_MAX)
 		return -EOVERFLOW;
@@ -7844,8 +7997,6 @@ static int io_allocate_scq_urings(struct io_ring_ctx *ctx,
 	rings->cq_ring_entries = p->cq_entries;
 	ctx->sq_mask = rings->sq_ring_mask;
 	ctx->cq_mask = rings->cq_ring_mask;
-	ctx->sq_entries = rings->sq_ring_entries;
-	ctx->cq_entries = rings->cq_ring_entries;
 
 	size = array_size(sizeof(struct io_uring_sqe), p->sq_entries);
 	if (size == SIZE_MAX) {