From: Cong Wang <xiyou.wangcong@gmail.com>
To: netdev@vger.kernel.org
Cc: bpf@vger.kernel.org, duanxiongchun@bytedance.com,
wangdongdong.6@bytedance.com, jiang.wang@bytedance.com,
Cong Wang <cong.wang@bytedance.com>,
Daniel Borkmann <daniel@iogearbox.net>,
Lorenz Bauer <lmb@cloudflare.com>,
Jakub Sitnicki <jakub@cloudflare.com>,
John Fastabend <john.fastabend@gmail.com>
Subject: [Patch bpf-next v8 02/16] skmsg: introduce a spinlock to protect ingress_msg
Date: Tue, 30 Mar 2021 19:32:23 -0700 [thread overview]
Message-ID: <20210331023237.41094-3-xiyou.wangcong@gmail.com> (raw)
In-Reply-To: <20210331023237.41094-1-xiyou.wangcong@gmail.com>
From: Cong Wang <cong.wang@bytedance.com>
Currently we rely on lock_sock to protect ingress_msg,
it is too big for this, we can actually just use a spinlock
to protect this list like protecting other skb queues.
__tcp_bpf_recvmsg() is still special because of peeking,
it still has to use lock_sock.
Cc: Daniel Borkmann <daniel@iogearbox.net>
Cc: Lorenz Bauer <lmb@cloudflare.com>
Acked-by: Jakub Sitnicki <jakub@cloudflare.com>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Cong Wang <cong.wang@bytedance.com>
---
include/linux/skmsg.h | 46 +++++++++++++++++++++++++++++++++++++++++++
net/core/skmsg.c | 3 +++
net/ipv4/tcp_bpf.c | 18 ++++++-----------
3 files changed, 55 insertions(+), 12 deletions(-)
diff --git a/include/linux/skmsg.h b/include/linux/skmsg.h
index 6c09d94be2e9..f2d45a73b2b2 100644
--- a/include/linux/skmsg.h
+++ b/include/linux/skmsg.h
@@ -89,6 +89,7 @@ struct sk_psock {
#endif
struct sk_buff_head ingress_skb;
struct list_head ingress_msg;
+ spinlock_t ingress_lock;
unsigned long state;
struct list_head link;
spinlock_t link_lock;
@@ -284,7 +285,45 @@ static inline struct sk_psock *sk_psock(const struct sock *sk)
static inline void sk_psock_queue_msg(struct sk_psock *psock,
struct sk_msg *msg)
{
+ spin_lock_bh(&psock->ingress_lock);
list_add_tail(&msg->list, &psock->ingress_msg);
+ spin_unlock_bh(&psock->ingress_lock);
+}
+
+static inline struct sk_msg *sk_psock_dequeue_msg(struct sk_psock *psock)
+{
+ struct sk_msg *msg;
+
+ spin_lock_bh(&psock->ingress_lock);
+ msg = list_first_entry_or_null(&psock->ingress_msg, struct sk_msg, list);
+ if (msg)
+ list_del(&msg->list);
+ spin_unlock_bh(&psock->ingress_lock);
+ return msg;
+}
+
+static inline struct sk_msg *sk_psock_peek_msg(struct sk_psock *psock)
+{
+ struct sk_msg *msg;
+
+ spin_lock_bh(&psock->ingress_lock);
+ msg = list_first_entry_or_null(&psock->ingress_msg, struct sk_msg, list);
+ spin_unlock_bh(&psock->ingress_lock);
+ return msg;
+}
+
+static inline struct sk_msg *sk_psock_next_msg(struct sk_psock *psock,
+ struct sk_msg *msg)
+{
+ struct sk_msg *ret;
+
+ spin_lock_bh(&psock->ingress_lock);
+ if (list_is_last(&msg->list, &psock->ingress_msg))
+ ret = NULL;
+ else
+ ret = list_next_entry(msg, list);
+ spin_unlock_bh(&psock->ingress_lock);
+ return ret;
}
static inline bool sk_psock_queue_empty(const struct sk_psock *psock)
@@ -292,6 +331,13 @@ static inline bool sk_psock_queue_empty(const struct sk_psock *psock)
return psock ? list_empty(&psock->ingress_msg) : true;
}
+static inline void kfree_sk_msg(struct sk_msg *msg)
+{
+ if (msg->skb)
+ consume_skb(msg->skb);
+ kfree(msg);
+}
+
static inline void sk_psock_report_error(struct sk_psock *psock, int err)
{
struct sock *sk = psock->sk;
diff --git a/net/core/skmsg.c b/net/core/skmsg.c
index bebf84ed4e30..305dddc51857 100644
--- a/net/core/skmsg.c
+++ b/net/core/skmsg.c
@@ -592,6 +592,7 @@ struct sk_psock *sk_psock_init(struct sock *sk, int node)
INIT_WORK(&psock->work, sk_psock_backlog);
INIT_LIST_HEAD(&psock->ingress_msg);
+ spin_lock_init(&psock->ingress_lock);
skb_queue_head_init(&psock->ingress_skb);
sk_psock_set_state(psock, SK_PSOCK_TX_ENABLED);
@@ -638,7 +639,9 @@ static void sk_psock_zap_ingress(struct sk_psock *psock)
skb_bpf_redirect_clear(skb);
kfree_skb(skb);
}
+ spin_lock_bh(&psock->ingress_lock);
__sk_psock_purge_ingress_msg(psock);
+ spin_unlock_bh(&psock->ingress_lock);
}
static void sk_psock_link_destroy(struct sk_psock *psock)
diff --git a/net/ipv4/tcp_bpf.c b/net/ipv4/tcp_bpf.c
index 17c322b875fd..ae980716d896 100644
--- a/net/ipv4/tcp_bpf.c
+++ b/net/ipv4/tcp_bpf.c
@@ -18,9 +18,7 @@ int __tcp_bpf_recvmsg(struct sock *sk, struct sk_psock *psock,
struct sk_msg *msg_rx;
int i, copied = 0;
- msg_rx = list_first_entry_or_null(&psock->ingress_msg,
- struct sk_msg, list);
-
+ msg_rx = sk_psock_peek_msg(psock);
while (copied != len) {
struct scatterlist *sge;
@@ -68,22 +66,18 @@ int __tcp_bpf_recvmsg(struct sock *sk, struct sk_psock *psock,
} while (i != msg_rx->sg.end);
if (unlikely(peek)) {
- if (msg_rx == list_last_entry(&psock->ingress_msg,
- struct sk_msg, list))
+ msg_rx = sk_psock_next_msg(psock, msg_rx);
+ if (!msg_rx)
break;
- msg_rx = list_next_entry(msg_rx, list);
continue;
}
msg_rx->sg.start = i;
if (!sge->length && msg_rx->sg.start == msg_rx->sg.end) {
- list_del(&msg_rx->list);
- if (msg_rx->skb)
- consume_skb(msg_rx->skb);
- kfree(msg_rx);
+ msg_rx = sk_psock_dequeue_msg(psock);
+ kfree_sk_msg(msg_rx);
}
- msg_rx = list_first_entry_or_null(&psock->ingress_msg,
- struct sk_msg, list);
+ msg_rx = sk_psock_peek_msg(psock);
}
return copied;
--
2.25.1
next prev parent reply other threads:[~2021-03-31 2:33 UTC|newest]
Thread overview: 35+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-03-31 2:32 [Patch bpf-next v8 00/16] sockmap: introduce BPF_SK_SKB_VERDICT and support UDP Cong Wang
2021-03-31 2:32 ` [Patch bpf-next v8 01/16] skmsg: lock ingress_skb when purging Cong Wang
2021-03-31 22:00 ` John Fastabend
2021-03-31 2:32 ` Cong Wang [this message]
2021-03-31 2:32 ` [Patch bpf-next v8 03/16] net: introduce skb_send_sock() for sock_map Cong Wang
2021-04-01 8:10 ` Jakub Sitnicki
2021-03-31 2:32 ` [Patch bpf-next v8 04/16] skmsg: avoid lock_sock() in sk_psock_backlog() Cong Wang
2021-03-31 2:32 ` [Patch bpf-next v8 05/16] skmsg: use rcu work for destroying psock Cong Wang
2021-03-31 2:32 ` [Patch bpf-next v8 06/16] skmsg: use GFP_KERNEL in sk_psock_create_ingress_msg() Cong Wang
2021-03-31 2:32 ` [Patch bpf-next v8 07/16] sock_map: simplify sock_map_link() a bit Cong Wang
2021-04-01 5:48 ` John Fastabend
2021-03-31 2:32 ` [Patch bpf-next v8 08/16] sock_map: kill sock_map_link_no_progs() Cong Wang
2021-03-31 2:32 ` [Patch bpf-next v8 09/16] sock_map: introduce BPF_SK_SKB_VERDICT Cong Wang
2021-04-01 5:51 ` John Fastabend
2021-03-31 2:32 ` [Patch bpf-next v8 10/16] sock: introduce sk->sk_prot->psock_update_sk_prot() Cong Wang
2021-04-02 10:16 ` Jakub Sitnicki
2021-04-03 5:13 ` Cong Wang
2021-04-05 8:25 ` Eric Dumazet
2021-04-06 18:12 ` John Fastabend
2021-04-06 18:30 ` Cong Wang
2021-04-06 21:07 ` John Fastabend
2021-03-31 2:32 ` [Patch bpf-next v8 11/16] udp: implement ->read_sock() for sockmap Cong Wang
2021-04-01 6:00 ` John Fastabend
2021-04-03 5:08 ` Cong Wang
2021-04-03 6:45 ` Alexei Starovoitov
2021-03-31 2:32 ` [Patch bpf-next v8 12/16] skmsg: extract __tcp_bpf_recvmsg() and tcp_bpf_wait_data() Cong Wang
2021-04-01 16:36 ` John Fastabend
2021-03-31 2:32 ` [Patch bpf-next v8 13/16] udp: implement udp_bpf_recvmsg() for sockmap Cong Wang
2021-04-01 16:24 ` John Fastabend
2021-03-31 2:32 ` [Patch bpf-next v8 14/16] sock_map: update sock type checks for UDP Cong Wang
2021-04-01 6:02 ` John Fastabend
2021-03-31 2:32 ` [Patch bpf-next v8 15/16] selftests/bpf: add a test case for udp sockmap Cong Wang
2021-03-31 2:32 ` [Patch bpf-next v8 16/16] selftests/bpf: add a test case for loading BPF_SK_SKB_VERDICT Cong Wang
2021-04-01 16:51 ` [Patch bpf-next v8 00/16] sockmap: introduce BPF_SK_SKB_VERDICT and support UDP John Fastabend
2021-04-01 18:03 ` Alexei Starovoitov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20210331023237.41094-3-xiyou.wangcong@gmail.com \
--to=xiyou.wangcong@gmail.com \
--cc=bpf@vger.kernel.org \
--cc=cong.wang@bytedance.com \
--cc=daniel@iogearbox.net \
--cc=duanxiongchun@bytedance.com \
--cc=jakub@cloudflare.com \
--cc=jiang.wang@bytedance.com \
--cc=john.fastabend@gmail.com \
--cc=lmb@cloudflare.com \
--cc=netdev@vger.kernel.org \
--cc=wangdongdong.6@bytedance.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).