path: root/include/linux/skbuff.h
diff options
authorQian Cai <cai@lca.pw>2020-02-04 13:40:29 -0500
committerDavid S. Miller <davem@davemloft.net>2020-02-06 13:59:10 +0100
commit86b18aaa2b5b5bb48e609cd591b3d2d0fdbe0442 (patch)
tree39d9ae29d9d944e6a14ed4ddc323b145f1b6b9c0 /include/linux/skbuff.h
parentnet: mvneta: move rx_dropped and rx_errors in per-cpu stats (diff)
skbuff: fix a data race in skb_queue_len()
sk_buff.qlen can be accessed concurrently as noticed by KCSAN, BUG: KCSAN: data-race in __skb_try_recv_from_queue / unix_dgram_sendmsg read to 0xffff8a1b1d8a81c0 of 4 bytes by task 5371 on cpu 96: unix_dgram_sendmsg+0x9a9/0xb70 include/linux/skbuff.h:1821 net/unix/af_unix.c:1761 ____sys_sendmsg+0x33e/0x370 ___sys_sendmsg+0xa6/0xf0 __sys_sendmsg+0x69/0xf0 __x64_sys_sendmsg+0x51/0x70 do_syscall_64+0x91/0xb47 entry_SYSCALL_64_after_hwframe+0x49/0xbe write to 0xffff8a1b1d8a81c0 of 4 bytes by task 1 on cpu 99: __skb_try_recv_from_queue+0x327/0x410 include/linux/skbuff.h:2029 __skb_try_recv_datagram+0xbe/0x220 unix_dgram_recvmsg+0xee/0x850 ____sys_recvmsg+0x1fb/0x210 ___sys_recvmsg+0xa2/0xf0 __sys_recvmsg+0x66/0xf0 __x64_sys_recvmsg+0x51/0x70 do_syscall_64+0x91/0xb47 entry_SYSCALL_64_after_hwframe+0x49/0xbe Since only the read is operating as lockless, it could introduce a logic bug in unix_recvq_full() due to the load tearing. Fix it by adding a lockless variant of skb_queue_len() and unix_recvq_full() where READ_ONCE() is on the read while WRITE_ONCE() is on the write similar to the commit d7d16a89350a ("net: add skb_queue_empty_lockless()"). Signed-off-by: Qian Cai <cai@lca.pw> Signed-off-by: David S. Miller <davem@davemloft.net>
Diffstat (limited to '')
1 files changed, 13 insertions, 1 deletions
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 3d13a4b717e9..ca8806b69388 100644
--- a/include/linux/skbuff.h
+++ b/include/linux/skbuff.h
@@ -1822,6 +1822,18 @@ static inline __u32 skb_queue_len(const struct sk_buff_head *list_)
+ * skb_queue_len_lockless - get queue length
+ * @list_: list to measure
+ *
+ * Return the length of an &sk_buff queue.
+ * This variant can be used in lockless contexts.
+ */
+static inline __u32 skb_queue_len_lockless(const struct sk_buff_head *list_)
+ return READ_ONCE(list_->qlen);
* __skb_queue_head_init - initialize non-spinlock portions of sk_buff_head
* @list: queue to initialize
@@ -2026,7 +2038,7 @@ static inline void __skb_unlink(struct sk_buff *skb, struct sk_buff_head *list)
struct sk_buff *next, *prev;
- list->qlen--;
+ WRITE_ONCE(list->qlen, list->qlen - 1);
next = skb->next;
prev = skb->prev;
skb->next = skb->prev = NULL;