Received-SPF: pass (google.com: best guess record for domain of linux-kernel-owner@vger.kernel.org designates 209.132.180.67 as permitted sender) client-ip=209.132.180.67;
From:   Qian Cai <cai@lca.pw>
To:     davem@davemloft.net
Cc:     alexander.h.duyck@linux.intel.com, kuznet@ms2.inr.ac.ru,
        kuba@kernel.org, yoshfuji@linux-ipv6.org, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org, Qian Cai <cai@lca.pw>
Subject: [PATCH] ipv4: fix a RCU-list bug in inet_dump_fib()
Date:   Thu, 19 Mar 2020 18:11:41 -0400
Message-Id: <20200319221141.8814-1-cai@lca.pw>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Sender: linux-kernel-owner@vger.kernel.org
Precedence: bulk

There is a place,

inet_dump_fib()
  fib_table_dump
    fn_trie_dump_leaf()
      hlist_for_each_entry_rcu()

without rcu_read_lock() triggers a warning,

 WARNING: suspicious RCU usage
 -----------------------------
 net/ipv4/fib_trie.c:2216 RCU-list traversed in non-reader section!!

 other info that might help us debug this:

 rcu_scheduler_active = 2, debug_locks = 1
 1 lock held by ip/1923:
  #0: ffffffff8ce76e40 (rtnl_mutex){+.+.}, at: netlink_dump+0xd6/0x840

 Call Trace:
  dump_stack+0xa1/0xea
  lockdep_rcu_suspicious+0x103/0x10d
  fn_trie_dump_leaf+0x581/0x590
  fib_table_dump+0x15f/0x220
  inet_dump_fib+0x4ad/0x5d0
  netlink_dump+0x350/0x840
  __netlink_dump_start+0x315/0x3e0
  rtnetlink_rcv_msg+0x4d1/0x720
  netlink_rcv_skb+0xf0/0x220
  rtnetlink_rcv+0x15/0x20
  netlink_unicast+0x306/0x460
  netlink_sendmsg+0x44b/0x770
  __sys_sendto+0x259/0x270
  __x64_sys_sendto+0x80/0xa0
  do_syscall_64+0x69/0xf4
  entry_SYSCALL_64_after_hwframe+0x49/0xb3

Signed-off-by: Qian Cai <cai@lca.pw>
---
 net/ipv4/fib_frontend.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/net/ipv4/fib_frontend.c b/net/ipv4/fib_frontend.c
index 577db1d50a24..5e441282d647 100644
--- a/net/ipv4/fib_frontend.c
+++ b/net/ipv4/fib_frontend.c
@@ -987,6 +987,8 @@ static int inet_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
 	if (filter.flags & RTM_F_PREFIX)
 		return skb->len;
 
+	rcu_read_lock();
+
 	if (filter.table_id) {
 		tb = fib_get_table(net, filter.table_id);
 		if (!tb) {
@@ -1004,8 +1006,6 @@ static int inet_dump_fib(struct sk_buff *skb, struct netlink_callback *cb)
 	s_h = cb->args[0];
 	s_e = cb->args[1];
 
-	rcu_read_lock();
-
 	for (h = s_h; h < FIB_TABLE_HASHSZ; h++, s_e = 0) {
 		e = 0;
 		head = &net->ipv4.fib_table_hash[h];
-- 
2.21.0 (Apple Git-122.2)