Re: [syzbot] [reiserfs?] possible deadlock in __run_timers

[Lizhi Xu]

#syz test https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git 88035e5694a8

diff --git a/kernel/workqueue.c b/kernel/workqueue.c
index 2989b57e154a..9daa5d695dbd 100644
--- a/kernel/workqueue.c
+++ b/kernel/workqueue.c
@@ -941,8 +941,11 @@ static void worker_enter_idle(struct worker *worker)
 	/* idle_list is LIFO */
 	list_add(&worker->entry, &pool->idle_list);
 
-	if (too_many_workers(pool) && !timer_pending(&pool->idle_timer))
+	if (too_many_workers(pool) && !timer_pending(&pool->idle_timer)) {
+		raw_spin_unlock_irq(&pool->lock);
 		mod_timer(&pool->idle_timer, jiffies + IDLE_WORKER_TIMEOUT);
+		raw_spin_lock_irq(&pool->lock);
+	}
 
 	/* Sanity check nr_running. */
 	WARN_ON_ONCE(pool->nr_workers == pool->nr_idle && pool->nr_running);