[<prev] [next>] [thread-next>] [day] [month] [year] [list]
Message-ID: <1476937498.5650.102.camel@edumazet-glaptop3.roam.corp.google.com>
Date: Wed, 19 Oct 2016 21:24:58 -0700
From: Eric Dumazet <eric.dumazet@...il.com>
To: David Miller <davem@...emloft.net>
Cc: netdev <netdev@...r.kernel.org>
Subject: [PATCH net-next] tcp: relax listening_hash operations
From: Eric Dumazet <edumazet@...gle.com>
softirq handlers use RCU protection to lookup listeners,
and write operations all happen from process context.
We do not need to block BH for dump operations.
Also SYN_RECV since request sockets are stored in the ehash table :
1) inet_diag_dump_icsk() no longer need to clear
cb->args[3] and cb->args[4] that were used as cursors while
iterating the old per listener hash table.
2) Also factorize a test : No need to scan listening_hash[]
if r->id.idiag_dport is not zero.
Signed-off-by: Eric Dumazet <edumazet@...gle.com>
---
net/ipv4/inet_diag.c | 16 ++++------------
net/ipv4/tcp_ipv4.c | 6 +++---
2 files changed, 7 insertions(+), 15 deletions(-)
diff --git a/net/ipv4/inet_diag.c b/net/ipv4/inet_diag.c
index e4d16fc5bbb3fec077f8e3bb3ff8811a66046edc..0a1d4a896a261dcb67e69a2d8c9aff83332fadd7 100644
--- a/net/ipv4/inet_diag.c
+++ b/net/ipv4/inet_diag.c
@@ -863,7 +863,7 @@ void inet_diag_dump_icsk(struct inet_hashinfo *hashinfo, struct sk_buff *skb,
s_num = num = cb->args[2];
if (cb->args[0] == 0) {
- if (!(idiag_states & TCPF_LISTEN))
+ if (!(idiag_states & TCPF_LISTEN) || r->id.idiag_dport)
goto skip_listen_ht;
for (i = s_i; i < INET_LHTABLE_SIZE; i++) {
@@ -872,7 +872,7 @@ void inet_diag_dump_icsk(struct inet_hashinfo *hashinfo, struct sk_buff *skb,
num = 0;
ilb = &hashinfo->listening_hash[i];
- spin_lock_bh(&ilb->lock);
+ spin_lock(&ilb->lock);
sk_for_each(sk, &ilb->head) {
struct inet_sock *inet = inet_sk(sk);
@@ -892,26 +892,18 @@ void inet_diag_dump_icsk(struct inet_hashinfo *hashinfo, struct sk_buff *skb,
r->id.idiag_sport)
goto next_listen;
- if (r->id.idiag_dport ||
- cb->args[3] > 0)
- goto next_listen;
-
if (inet_csk_diag_dump(sk, skb, cb, r,
bc, net_admin) < 0) {
- spin_unlock_bh(&ilb->lock);
+ spin_unlock(&ilb->lock);
goto done;
}
next_listen:
- cb->args[3] = 0;
- cb->args[4] = 0;
++num;
}
- spin_unlock_bh(&ilb->lock);
+ spin_unlock(&ilb->lock);
s_num = 0;
- cb->args[3] = 0;
- cb->args[4] = 0;
}
skip_listen_ht:
cb->args[0] = 1;
diff --git a/net/ipv4/tcp_ipv4.c b/net/ipv4/tcp_ipv4.c
index bd5e8d10893fb6abffa6b0aa65de239b0000fe5b..83b3d0b8c4812d734728360a694a596a17257d02 100644
--- a/net/ipv4/tcp_ipv4.c
+++ b/net/ipv4/tcp_ipv4.c
@@ -1893,7 +1893,7 @@ static void *listening_get_next(struct seq_file *seq, void *cur)
if (!sk) {
get_head:
ilb = &tcp_hashinfo.listening_hash[st->bucket];
- spin_lock_bh(&ilb->lock);
+ spin_lock(&ilb->lock);
sk = sk_head(&ilb->head);
st->offset = 0;
goto get_sk;
@@ -1911,7 +1911,7 @@ static void *listening_get_next(struct seq_file *seq, void *cur)
return sk;
icsk = inet_csk(sk);
}
- spin_unlock_bh(&ilb->lock);
+ spin_unlock(&ilb->lock);
st->offset = 0;
if (++st->bucket < INET_LHTABLE_SIZE)
goto get_head;
@@ -2119,7 +2119,7 @@ static void tcp_seq_stop(struct seq_file *seq, void *v)
switch (st->state) {
case TCP_SEQ_STATE_LISTENING:
if (v != SEQ_START_TOKEN)
- spin_unlock_bh(&tcp_hashinfo.listening_hash[st->bucket].lock);
+ spin_unlock(&tcp_hashinfo.listening_hash[st->bucket].lock);
break;
case TCP_SEQ_STATE_ESTABLISHED:
if (v)
Powered by blists - more mailing lists