Merge tag 'io_uring-5.8-2020-07-01' of git://git.kernel.dk/linux-block

Pull io_uring fixes from Jens Axboe:
 "One fix in here, for a regression in 5.7 where a task is waiting in
  the kernel for a condition, but that condition won't become true until
  task_work is run. And the task_work can't be run exactly because the
  task is waiting in the kernel, so we'll never make any progress.

  One example of that is registering an eventfd and queueing io_uring
  work, and then the task goes and waits in eventfd read with the
  expectation that it'll get woken (and read an event) when the io_uring
  request completes. The io_uring request is finished through task_work,
  which won't get run while the task is looping in eventfd read"

* tag 'io_uring-5.8-2020-07-01' of git://git.kernel.dk/linux-block:
  io_uring: use signal based task_work running
  task_work: teach task_work_add() to do signal_wake_up()

Author: torvalds

fs/io_uring.c

@@ -4072,6 +4072,21 @@ struct io_poll_table {
 	int error;
 };
 
+static int io_req_task_work_add(struct io_kiocb *req, struct callback_head *cb,
+				int notify)
+{
+	struct task_struct *tsk = req->task;
+	int ret;
+
+	if (req->ctx->flags & IORING_SETUP_SQPOLL)
+		notify = 0;
+
+	ret = task_work_add(tsk, cb, notify);
+	if (!ret)
+		wake_up_process(tsk);
+	return ret;
+}
+
 static int __io_async_wake(struct io_kiocb *req, struct io_poll_iocb *poll,
 			   __poll_t mask, task_work_func_t func)
 {
@@ -4095,13 +4110,13 @@ static int __io_async_wake(struct io_kiocb *req, struct io_poll_iocb *poll,
 	 * of executing it. We can't safely execute it anyway, as we may not
 	 * have the needed state needed for it anyway.
 	 */
-	ret = task_work_add(tsk, &req->task_work, true);
+	ret = io_req_task_work_add(req, &req->task_work, TWA_SIGNAL);
 	if (unlikely(ret)) {
 		WRITE_ONCE(poll->canceled, true);
 		tsk = io_wq_get_task(req->ctx->io_wq);
-		task_work_add(tsk, &req->task_work, true);
+		task_work_add(tsk, &req->task_work, 0);
+		wake_up_process(tsk);
 	}
-	wake_up_process(tsk);
 	return 1;
 }
 
@@ -6182,19 +6197,20 @@ static int io_cqring_wait(struct io_ring_ctx *ctx, int min_events,
 	do {
 		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_should_wake(&iowq, false))
-			break;
-		schedule();
 		if (signal_pending(current)) {
-			ret = -EINTR;
+			ret = -ERESTARTSYS;
 			break;
 		}
+		if (io_should_wake(&iowq, false))
+			break;
+		schedule();
 	} while (1);
 	finish_wait(&ctx->wait, &iowq.wq);
 
-	restore_saved_sigmask_unless(ret == -EINTR);
+	restore_saved_sigmask_unless(ret == -ERESTARTSYS);
 
 	return READ_ONCE(rings->cq.head) == READ_ONCE(rings->cq.tail) ? ret : 0;
 }

include/linux/sched/jobctl.h

@@ -19,6 +19,7 @@ struct task_struct;
 #define JOBCTL_TRAPPING_BIT	21	/* switching to TRACED */
 #define JOBCTL_LISTENING_BIT	22	/* ptracer is listening for events */
 #define JOBCTL_TRAP_FREEZE_BIT	23	/* trap for cgroup freezer */
+#define JOBCTL_TASK_WORK_BIT	24	/* set by TWA_SIGNAL */
 
 #define JOBCTL_STOP_DEQUEUED	(1UL << JOBCTL_STOP_DEQUEUED_BIT)
 #define JOBCTL_STOP_PENDING	(1UL << JOBCTL_STOP_PENDING_BIT)
@@ -28,9 +29,10 @@ struct task_struct;
 #define JOBCTL_TRAPPING		(1UL << JOBCTL_TRAPPING_BIT)
 #define JOBCTL_LISTENING	(1UL << JOBCTL_LISTENING_BIT)
 #define JOBCTL_TRAP_FREEZE	(1UL << JOBCTL_TRAP_FREEZE_BIT)
+#define JOBCTL_TASK_WORK	(1UL << JOBCTL_TASK_WORK_BIT)
 
 #define JOBCTL_TRAP_MASK	(JOBCTL_TRAP_STOP | JOBCTL_TRAP_NOTIFY)
-#define JOBCTL_PENDING_MASK	(JOBCTL_STOP_PENDING | JOBCTL_TRAP_MASK)
+#define JOBCTL_PENDING_MASK	(JOBCTL_STOP_PENDING | JOBCTL_TRAP_MASK | JOBCTL_TASK_WORK)
 
 extern bool task_set_jobctl_pending(struct task_struct *task, unsigned long mask);
 extern void task_clear_jobctl_trapping(struct task_struct *task);

include/linux/task_work.h

@@ -13,7 +13,10 @@ init_task_work(struct callback_head *twork, task_work_func_t func)
 	twork->func = func;
 }
 
-int task_work_add(struct task_struct *task, struct callback_head *twork, bool);
+#define TWA_RESUME	1
+#define TWA_SIGNAL	2
+int task_work_add(struct task_struct *task, struct callback_head *twork, int);
+
 struct callback_head *task_work_cancel(struct task_struct *, task_work_func_t);
 void task_work_run(void);
 

kernel/signal.c

@@ -2529,9 +2529,6 @@ bool get_signal(struct ksignal *ksig)
 	struct signal_struct *signal = current->signal;
 	int signr;
 
-	if (unlikely(current->task_works))
-		task_work_run();
-
 	if (unlikely(uprobe_deny_signal()))
 		return false;
 
@@ -2544,6 +2541,13 @@ bool get_signal(struct ksignal *ksig)
 
 relock:
 	spin_lock_irq(&sighand->siglock);
+	current->jobctl &= ~JOBCTL_TASK_WORK;
+	if (unlikely(current->task_works)) {
+		spin_unlock_irq(&sighand->siglock);
+		task_work_run();
+		goto relock;
+	}
+
 	/*
 	 * Every stopped thread goes here after wakeup. Check to see if
 	 * we should notify the parent, prepare_signal(SIGCONT) encodes

kernel/task_work.c

@@ -25,9 +25,10 @@ static struct callback_head work_exited; /* all we need is ->next == NULL */
  * 0 if succeeds or -ESRCH.
  */
 int
-task_work_add(struct task_struct *task, struct callback_head *work, bool notify)
+task_work_add(struct task_struct *task, struct callback_head *work, int notify)
 {
 	struct callback_head *head;
+	unsigned long flags;
 
 	do {
 		head = READ_ONCE(task->task_works);
@@ -36,8 +37,19 @@ task_work_add(struct task_struct *task, struct callback_head *work, bool notify)
 		work->next = head;
 	} while (cmpxchg(&task->task_works, head, work) != head);
 
-	if (notify)
+	switch (notify) {
+	case TWA_RESUME:
 		set_notify_resume(task);
+		break;
+	case TWA_SIGNAL:
+		if (lock_task_sighand(task, &flags)) {
+			task->jobctl |= JOBCTL_TASK_WORK;
+			signal_wake_up(task, 0);
+			unlock_task_sighand(task, &flags);
+		}
+		break;
+	}
+
 	return 0;
 }