[嵌入式]Unable to handle kernel NULL pointer dereference at virtual address 分析

Unable to handle kernel NULL pointer dereference at virtual address 分析

现象

[  136.847780] br-lan: received packet on eth0.1 with own address as source address (addr:00:02:e7:f5:02:02, vlan:0)
[  138.026055] Unable to handle kernel NULL pointer dereference at virtual address 00000010
[  138.034175] pgd = c0003000
[  138.036888] [00000010] *pgd=80000040004003, *pmd=00000000
[  138.042311] Internal error: Oops: 207 [#1] PREEMPT SMP ARM
[  138.047790] Modules linked in: pppoe ppp_async l2tp_ppp pptp pppox ppp_mppe ppp_generic iptable_nat ipt_REJECT ipt_MASQUERADE xt_time xt_tcpudp xt_state xt_recent xt_policy xt_nat xt_multiport xt_mark xt_mac xt_limit xt_helper xt_esp xt_conntrack xt_connmark xt_connlimit xt_connbytes xt_comment xt_TCPMSS xt_REDIRECT xt_LOG xt_FLOWOFFLOAD slhc openvswitch nf_reject_ipv4 nf_nat_redirect nf_nat_masquerade_ipv6 nf_nat_masquerade_ipv4 nf_conntrack_ipv6 nf_nat_ipv6 nf_conntrack_ipv4 nf_nat_ipv4 nf_nat nf_log_ipv4 nf_flow_table_hw nf_flow_table nf_defrag_ipv6 nf_defrag_ipv4 nf_conntrack_rtcache nf_conntrack libcrc32c iptable_mangle iptable_filter ipt_ah ip_tables crc_ccitt rtc_sunxi ledtrig_heartbeat ip6t_REJECT nf_reject_ipv6 nf_log_ipv6 nf_log_common ip6table_mangle ip6table_filter ip6_tables x_tables
[  138.118356]  ip_gre gre l2tp_netlink l2tp_core udp_tunnel ip6_udp_tunnel ipcomp6 xfrm6_tunnel xfrm6_mode_tunnel xfrm6_mode_transport xfrm6_mode_beet esp6 ah6 ipcomp xfrm4_tunnel tunnel6 tun mpls_gso mpls_iptunnel mpls_router af_key xfrm_user xfrm_ipcomp algif_skcipher algif_hash af_alg ecb deflate zlib_deflate crypto_acompress
[  138.147315] CPU: 3 PID: 0 Comm: swapper/3 Not tainted 4.14.63 #0
[  138.153312] Hardware name: Allwinner sun8i Family
[  138.158012] task: cf844440 task.stack: cf86c000
[  138.162548] PC is at br_forward+0x4/0x74
[  138.166467] LR is at br_dev_xmit+0x264/0x290
[  138.170733] pc : [<c07725a8>]    lr : [<c07702f0>]    psr: a0000113
[  138.176990] sp : cf86db6c  ip : 00000202  fp : 00000000
[  138.182207] r10: 00000027  r9 : c0c04a54  r8 : cf2f2540
[  138.187425] r7 : cf2f2000  r6 : cc95dd80  r5 : 00000000  r4 : 00000000
[  138.193942] r3 : 00000001  r2 : 00000000  r1 : cc95dd80  r0 : 00000000

分析

根据PC is at br_forward+0x4/0x74 可知道出错在 br_forward 函数中。

/* 位于 linux-4.14.63\net\bridge\br_forward.c */
void br_forward(const struct net_bridge_port *to,
		struct sk_buff *skb, bool local_rcv, bool local_orig)
{
	if (to->flags & BR_ISOLATE_MODE && !local_orig)
		to = NULL;

	if (to && should_deliver(to, skb)) {
		if (local_rcv)
			deliver_clone(to, skb, local_orig);
		else
			__br_forward(to, skb, local_orig);
		return;
	}

	if (!local_rcv)
		kfree_skb(skb);
}

已知在arm 中 r0~rx 寄存器一般用来给函数传参，再结合log:

[ 138.193942] r3 : 00000001 r2 : 00000000 r1 : cc95dd80 r0 : 00000000

可知：br_forward ( r0 ,r1, r2 , r3) 即为 br_forward ( 0 ,cc95dd80, 0 , 1) 。

显然 第一个参数 *to 是 0 ，是个空指针。

而代码中 to-> flags 对空指针进行了引用，flags 相对于 to 的偏移正好是 16 （0x10）, to->flags 就等于 0+0x10 = 0x10

访问 0x10 触发了 Unable to handle kernel NULL pointer dereference at virtual address 00000010

解决办法

在代码中加入空指针判断

void br_forward(const struct net_bridge_port *to,
		struct sk_buff *skb, bool local_rcv, bool local_orig)
{
+	if(unlikely(!to))
+		goto out;
+		
	if (to->flags & BR_ISOLATE_MODE && !local_orig)
		to = NULL;

	if (to && should_deliver(to, skb)) {
		if (local_rcv)
			deliver_clone(to, skb, local_orig);
		else
			__br_forward(to, skb, local_orig);
		return;
	}
+out:
	if (!local_rcv)
		kfree_skb(skb);
}