Received-SPF: pass (google.com: domain of linux-kernel+bounces-85300-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
From: Chengming Zhou <chengming.zhou@linux.dev>
To: akpm@linux-foundation.org
Cc: hannes@cmpxchg.org,
	yosryahmed@google.com,
	nphamcs@gmail.com,
	willy@infradead.org,
	linux-mm@kvack.org,
	linux-kernel@vger.kernel.org,
	Chengming Zhou <chengming.zhou@linux.dev>
Subject: [PATCH mm-unstable v2] mm/zswap: fix zswap_pools_lock usages after changing to percpu_ref
Date: Wed, 28 Feb 2024 15:49:54 +0000
Message-Id: <20240228154954.3028626-1-chengming.zhou@linux.dev>
In-Reply-To: <20240210-zswap-global-lru-v3-2-200495333595@bytedance.com>
References: <20240210-zswap-global-lru-v3-2-200495333595@bytedance.com>
Precedence: bulk
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit

Now the release of zswap pool is controlled by percpu_ref, its release
callback (__zswap_pool_empty()) will be called when percpu_ref hit 0.
But this release callback may potentially be called from RCU callback
context by percpu_ref_kill(), which maybe from the softirq context.

So we need to use spin_lock/unlock_bh() to avoid potential deadlock.

This problem is introduced by the commit f3da427e82c4 ("mm/zswap: change
zswap_pool kref to percpu_ref"), which is in mm-unstable branch now.
It can be reproduced by testing kernel build in tmpfs with zswap and
CONFIG_LOCKDEP enabled, meanwhile changing the zswap compressor setting
dynamically.

Signed-off-by: Chengming Zhou <chengming.zhou@linux.dev>
---
v2:
 - Change to use spin_lock/unlock_bh(), per Matthew.
---
 mm/zswap.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/mm/zswap.c b/mm/zswap.c
index 011e068eb355..da90933c6d20 100644
--- a/mm/zswap.c
+++ b/mm/zswap.c
@@ -459,7 +459,7 @@ static void __zswap_pool_empty(struct percpu_ref *ref)
 
 	pool = container_of(ref, typeof(*pool), ref);
 
-	spin_lock(&zswap_pools_lock);
+	spin_lock_bh(&zswap_pools_lock);
 
 	WARN_ON(pool == zswap_pool_current());
 
@@ -468,7 +468,7 @@ static void __zswap_pool_empty(struct percpu_ref *ref)
 	INIT_WORK(&pool->release_work, __zswap_pool_release);
 	schedule_work(&pool->release_work);
 
-	spin_unlock(&zswap_pools_lock);
+	spin_unlock_bh(&zswap_pools_lock);
 }
 
 static int __must_check zswap_pool_get(struct zswap_pool *pool)
@@ -598,7 +598,7 @@ static int __zswap_param_set(const char *val, const struct kernel_param *kp,
 		return -EINVAL;
 	}
 
-	spin_lock(&zswap_pools_lock);
+	spin_lock_bh(&zswap_pools_lock);
 
 	pool = zswap_pool_find_get(type, compressor);
 	if (pool) {
@@ -607,7 +607,7 @@ static int __zswap_param_set(const char *val, const struct kernel_param *kp,
 		list_del_rcu(&pool->list);
 	}
 
-	spin_unlock(&zswap_pools_lock);
+	spin_unlock_bh(&zswap_pools_lock);
 
 	if (!pool)
 		pool = zswap_pool_create(type, compressor);
@@ -628,7 +628,7 @@ static int __zswap_param_set(const char *val, const struct kernel_param *kp,
 	else
 		ret = -EINVAL;
 
-	spin_lock(&zswap_pools_lock);
+	spin_lock_bh(&zswap_pools_lock);
 
 	if (!ret) {
 		put_pool = zswap_pool_current();
@@ -643,7 +643,7 @@ static int __zswap_param_set(const char *val, const struct kernel_param *kp,
 		put_pool = pool;
 	}
 
-	spin_unlock(&zswap_pools_lock);
+	spin_unlock_bh(&zswap_pools_lock);
 
 	if (!zswap_has_pool && !pool) {
 		/* if initial pool creation failed, and this pool creation also
-- 
2.40.1