io_uring/cancel: re-grab ctx mutex after finishing wait

If we have a signal pending during cancelations, it'll cause the
task_work run to return an error. Since we didn't run task_work, the
current task is left in TASK_INTERRUPTIBLE state when we need to
re-grab the ctx mutex, and the kernel will rightfully complain about
that.

Move the lock grabbing for the error cases outside the loop to avoid
that issue.

Reported-by: syzbot+7df055631cd1be4586fd@syzkaller.appspotmail.com
Link: https://lore.kernel.org/io-uring/0000000000003a14a905f05050b0@google.com/Signed-off-by: Jens Axboe <axboe@kernel.dk>

io_uring/cancel.c

@@ -288,24 +288,23 @@ int io_sync_cancel(struct io_ring_ctx *ctx, void __user *arg)
 
 		ret = __io_sync_cancel(current->io_uring, &cd, sc.fd);
 
+		mutex_unlock(&ctx->uring_lock);
 		if (ret != -EALREADY)
 			break;
 
-		mutex_unlock(&ctx->uring_lock);
 		ret = io_run_task_work_sig(ctx);
-		if (ret < 0) {
-			mutex_lock(&ctx->uring_lock);
+		if (ret < 0)
 			break;
-		}
 		ret = schedule_hrtimeout(&timeout, HRTIMER_MODE_ABS);
-		mutex_lock(&ctx->uring_lock);
 		if (!ret) {
 			ret = -ETIME;
 			break;
 		}
+		mutex_lock(&ctx->uring_lock);
 	} while (1);
 
 	finish_wait(&ctx->cq_wait, &wait);
+	mutex_lock(&ctx->uring_lock);
 
 	if (ret == -ENOENT || ret > 0)
 		ret = 0;