From: Jarek Poplawski <jarkao2@gmail.com>
To: minyard@acm.org
Cc: Linux Kernel <linux-kernel@vger.kernel.org>, netdev@vger.kernel.org
Subject: Re: [PATCH 1/1] Use RCU for the UDP hash lock
Date: Thu, 25 Sep 2008 08:45:09 +0000 [thread overview]
Message-ID: <20080925084509.GA5090@ff.dom.local> (raw)
In-Reply-To: <20080924172827.GA1573@minyard.local>
On 24-09-2008 19:28, Corey Minyard wrote:
...
> From: Corey Minyard <cminyard@mvista.com>
>
> Convert access to the udp_hash table to use RCU.
>
> Signed-off-by: Corey Minyard <cminyard@mvista.com>
> ---
...
> diff --git a/include/net/sock.h b/include/net/sock.h
> index 06c5259..ada44ad 100644
> --- a/include/net/sock.h
> +++ b/include/net/sock.h
> @@ -42,6 +42,7 @@
>
> #include <linux/kernel.h>
> #include <linux/list.h>
> +#include <linux/rculist.h>
> #include <linux/timer.h>
> #include <linux/cache.h>
> #include <linux/module.h>
> @@ -361,6 +362,27 @@ static __inline__ int sk_del_node_init(struct sock *sk)
> return rc;
> }
>
> +static inline int __sk_del_node_rcu(struct sock *sk)
> +{
> + if (sk_hashed(sk)) {
> + hlist_del_rcu(&sk->sk_node);
> + return 1;
> + }
> + return 0;
> +}
> +
> +static inline int sk_del_node_rcu(struct sock *sk)
> +{
> + int rc = __sk_del_node_rcu(sk);
Why sk_node_init() part (or hlist_del_init_rcu) isn't used?
> +
> + if (rc) {
> + /* paranoid for a while -acme */
> + WARN_ON(atomic_read(&sk->sk_refcnt) == 1);
> + __sock_put(sk);
> + }
> + return rc;
> +}
> +
...
> diff --git a/include/net/udp.h b/include/net/udp.h
> index addcdc6..04181f8 100644
> --- a/include/net/udp.h
> +++ b/include/net/udp.h
> @@ -51,7 +51,7 @@ struct udp_skb_cb {
> #define UDP_SKB_CB(__skb) ((struct udp_skb_cb *)((__skb)->cb))
>
> extern struct hlist_head udp_hash[UDP_HTABLE_SIZE];
> -extern rwlock_t udp_hash_lock;
> +extern spinlock_t udp_hash_wlock;
>
>
> /* Note: this must match 'valbool' in sock_setsockopt */
> @@ -112,12 +112,13 @@ static inline void udp_lib_hash(struct sock *sk)
>
> static inline void udp_lib_unhash(struct sock *sk)
> {
> - write_lock_bh(&udp_hash_lock);
> - if (sk_del_node_init(sk)) {
> + spin_lock_bh(&udp_hash_wlock);
> + if (sk_del_node_rcu(sk)) {
> inet_sk(sk)->num = 0;
> sock_prot_inuse_add(sock_net(sk), sk->sk_prot, -1);
> }
> - write_unlock_bh(&udp_hash_lock);
> + spin_unlock_bh(&udp_hash_wlock);
> + synchronize_sched();
> }
>
> static inline void udp_lib_close(struct sock *sk, long timeout)
> diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
> index 57e26fa..3aa04da 100644
> --- a/net/ipv4/udp.c
> +++ b/net/ipv4/udp.c
...
> @@ -1094,7 +1103,7 @@ static int __udp4_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
> struct sock *sk;
> int dif;
>
> - read_lock(&udp_hash_lock);
> + rcu_read_lock();
> sk = sk_head(&udptable[udp_hashfn(net, ntohs(uh->dest))]);
Probably sk_head_rcu() is needed too.
> dif = skb->dev->ifindex;
> sk = udp_v4_mcast_next(sk, uh->dest, daddr, uh->source, saddr, dif);
> @@ -1120,7 +1129,7 @@ static int __udp4_lib_mcast_deliver(struct net *net, struct sk_buff *skb,
> } while (sknext);
> } else
> kfree_skb(skb);
> - read_unlock(&udp_hash_lock);
> + rcu_read_unlock();
> return 0;
> }
...
Aren't other functions like sk_next() or sk_unhashed() used on the
read side and need _rcu versions?
Jarek P.
next prev parent reply other threads:[~2008-09-25 8:45 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <48DB2925.2070908@poczta.onet.pl>
2008-09-24 17:28 ` [PATCH 1/1] Use RCU for the UDP hash lock Corey Minyard
2008-09-24 19:40 ` Stephen Hemminger
2008-09-24 20:46 ` Corey Minyard
2008-09-25 15:29 ` Paul E. McKenney
2008-09-25 15:34 ` Stephen Hemminger
2008-09-25 19:21 ` Corey Minyard
2008-09-25 20:34 ` Stephen Hemminger
2008-09-25 8:45 ` Jarek Poplawski [this message]
2008-09-25 19:14 ` Corey Minyard
2008-09-26 3:18 Corey Minyard
2008-09-26 4:09 ` Paul E. McKenney
2008-09-26 13:49 ` Corey Minyard
2008-09-26 19:50 ` Jarek Poplawski
2008-09-26 5:46 ` Jarek Poplawski
2008-09-26 13:37 ` Corey Minyard
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=20080925084509.GA5090@ff.dom.local \
--to=jarkao2@gmail.com \
--cc=linux-kernel@vger.kernel.org \
--cc=minyard@acm.org \
--cc=netdev@vger.kernel.org \
/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).