From: Arnaldo Carvalho de Melo <acme@redhat.com>
To: KOVACS Krisztian <hidden@sch.bme.hu>
Cc: David Miller <davem@davemloft.net>,
kaber@trash.net, netdev@vger.kernel.org,
netfilter-devel@vger.kernel.org
Subject: Re: [net-next PATCH] Don't lookup the socket if there's a socket attached to the skb (was: Re: [net-next PATCH 10/16] Don't lookup the socket if there's a socket attached to the skb)
Date: Tue, 7 Oct 2008 09:36:26 -0300 [thread overview]
Message-ID: <20081007123626.GB29610@ghostprotocols.net> (raw)
In-Reply-To: <1223366369.35518.13.camel@nessa.odu>
Em Tue, Oct 07, 2008 at 09:59:29AM +0200, KOVACS Krisztian escreveu:
> Hi,
>
> On Fri, 2008-10-03 at 10:47 -0300, Arnaldo Carvalho de Melo wrote:
> > > Those functions don't have access to the skb so unless we change the
> > > signature they won't be able to steal the reference.
> >
> > Indeed, but we should try to have the main TCP code flow clean, ditto for
> > DCCP, free of such details, so after this activitity settles down I'll
> > submit something like the patch below.
> >
> > If Dave agrees and you feel like merging it on your current patchset,
> > feel free to do it.
>
> And here's the modified use-cached-sk-if-present patch built on top of
> the previous two x_lookup_skb() helper patches.
>
> --
> KOVACS Krisztian
>
>
> Don't lookup the socket if there's a socket attached to the skb
>
> Use the socket cached in the skb if it's present.
>
> Signed-off-by: KOVACS Krisztian <hidden@sch.bme.hu>
> ---
>
> include/net/inet6_hashtables.h | 12 ++++++++----
> include/net/inet_hashtables.h | 9 ++++++---
> include/net/sock.h | 12 ++++++++++++
> net/ipv4/udp.c | 9 ++++++---
> net/ipv6/udp.c | 9 ++++++---
> 5 files changed, 38 insertions(+), 13 deletions(-)
>
>
> diff --git a/include/net/inet6_hashtables.h b/include/net/inet6_hashtables.h
> index 995efbb..f74665d 100644
> --- a/include/net/inet6_hashtables.h
> +++ b/include/net/inet6_hashtables.h
> @@ -96,10 +96,14 @@ static inline struct sock *__inet6_lookup_skb(struct inet_hashinfo *hashinfo,
> const __be16 sport,
> const __be16 dport)
> {
> - return __inet6_lookup(dev_net(skb->dst->dev), hashinfo,
> - &ipv6_hdr(skb)->saddr, sport,
> - &ipv6_hdr(skb)->daddr, ntohs(dport),
> - inet6_iif(skb));
> + struct sock *sk;
> +
> + if (unlikely(sk = skb_steal_sock(skb)))
> + return sk;
> + else return __inet6_lookup(dev_net(skb->dst->dev), hashinfo,
> + &ipv6_hdr(skb)->saddr, sport,
> + &ipv6_hdr(skb)->daddr, ntohs(dport),
> + inet6_iif(skb));
> }
>
> extern struct sock *inet6_lookup(struct net *net, struct inet_hashinfo *hashinfo,
> diff --git a/include/net/inet_hashtables.h b/include/net/inet_hashtables.h
> index 3522bbc..72b9ba5 100644
> --- a/include/net/inet_hashtables.h
> +++ b/include/net/inet_hashtables.h
> @@ -378,11 +378,14 @@ static inline struct sock *__inet_lookup_skb(struct inet_hashinfo *hashinfo,
> const __be16 sport,
> const __be16 dport)
> {
> + struct sock *sk;
> const struct iphdr *iph = ip_hdr(skb);
>
> - return __inet_lookup(dev_net(skb->dst->dev), hashinfo,
> - iph->saddr, sport,
> - iph->daddr, dport, inet_iif(skb));
> + if (unlikely(sk = skb_steal_sock(skb)))
> + return sk;
> + else return __inet_lookup(dev_net(skb->dst->dev), hashinfo,
> + iph->saddr, sport,
> + iph->daddr, dport, inet_iif(skb));
return on a different line, please:
else
return
> }
>
> extern int __inet_hash_connect(struct inet_timewait_death_row *death_row,
> diff --git a/include/net/sock.h b/include/net/sock.h
> index 75a312d..18f9670 100644
> --- a/include/net/sock.h
> +++ b/include/net/sock.h
> @@ -1324,6 +1324,18 @@ static inline void sk_change_net(struct sock *sk, struct net *net)
> sock_net_set(sk, hold_net(net));
> }
>
> +static inline struct sock *skb_steal_sock(struct sk_buff *skb)
> +{
> + if (unlikely(skb->sk)) {
> + struct sock *sk = skb->sk;
> +
> + skb->destructor = NULL;
> + skb->sk = NULL;
> + return sk;
> + }
> + return NULL;
> +}
> +
> extern void sock_enable_timestamp(struct sock *sk);
> extern int sock_get_timestamp(struct sock *, struct timeval __user *);
> extern int sock_get_timestampns(struct sock *, struct timespec __user *);
> diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
> index 8369f4d..219a4aa 100644
> --- a/net/ipv4/udp.c
> +++ b/net/ipv4/udp.c
> @@ -306,11 +306,14 @@ static inline struct sock *__udp4_lib_lookup_skb(struct sk_buff *skb,
> __be16 sport, __be16 dport,
> struct hlist_head udptable[])
> {
> + struct sock *sk;
> const struct iphdr *iph = ip_hdr(skb);
>
> - return __udp4_lib_lookup(dev_net(skb->dst->dev), iph->saddr, sport,
> - iph->daddr, dport, inet_iif(skb),
> - udptable);
> + if (unlikely(sk = skb_steal_sock(skb)))
> + return sk;
> + else return __udp4_lib_lookup(dev_net(skb->dst->dev), iph->saddr, sport,
> + iph->daddr, dport, inet_iif(skb),
> + udptable);
ditto
> }
>
> struct sock *udp4_lib_lookup(struct net *net, __be32 saddr, __be16 sport,
> diff --git a/net/ipv6/udp.c b/net/ipv6/udp.c
> index ce26c41..95a2b56 100644
> --- a/net/ipv6/udp.c
> +++ b/net/ipv6/udp.c
> @@ -111,11 +111,14 @@ static struct sock *__udp6_lib_lookup_skb(struct sk_buff *skb,
> __be16 sport, __be16 dport,
> struct hlist_head udptable[])
> {
> + struct sock *sk;
> struct ipv6hdr *iph = ipv6_hdr(skb);
>
> - return __udp6_lib_lookup(dev_net(skb->dst->dev), &iph->saddr, sport,
> - &iph->daddr, dport, inet6_iif(skb),
> - udptable);
> + if (unlikely(sk = skb_steal_sock(skb)))
> + return sk;
> + else return __udp6_lib_lookup(dev_net(skb->dst->dev), &iph->saddr, sport,
> + &iph->daddr, dport, inet6_iif(skb),
> + udptable);
ditto
After you fix this up, please add:
Acked-by: Arnaldo Carvalho de Melo <acme@redhat.com>
next prev parent reply other threads:[~2008-10-07 12:36 UTC|newest]
Thread overview: 61+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-10-01 14:24 [net-next PATCH 00/16] Transparent proxying patches, take six KOVACS Krisztian
2008-10-01 14:24 ` [net-next PATCH 15/16] iptables TPROXY target KOVACS Krisztian
2008-10-02 9:28 ` Patrick McHardy
2008-10-01 14:24 ` [net-next PATCH 11/16] Don't lookup the socket if there's a socket attached to the skb KOVACS Krisztian
2008-10-01 14:24 ` [net-next PATCH 08/16] Port redirection support for TCP KOVACS Krisztian
2008-10-01 14:47 ` David Miller
2008-10-01 14:24 ` [net-next PATCH 16/16] Add documentation KOVACS Krisztian
2008-10-01 16:22 ` Randy Dunlap
2008-10-02 9:37 ` [RESEND net-next " KOVACS Krisztian
2008-10-02 9:38 ` Patrick McHardy
2008-10-03 14:01 ` [net-next " Jan Engelhardt
2008-10-07 7:01 ` KOVACS Krisztian
2008-10-07 19:50 ` David Miller
2008-10-07 20:02 ` KOVACS Krisztian
2008-10-07 20:47 ` Patrick McHardy
2008-10-07 20:53 ` David Miller
2008-10-08 0:32 ` Philip Craig
2008-10-01 14:24 ` [net-next PATCH 02/16] Implement IP_TRANSPARENT socket option KOVACS Krisztian
2008-10-01 14:30 ` David Miller
2008-10-01 14:24 ` [net-next PATCH 04/16] Make inet_sock.h independent of route.h KOVACS Krisztian
2008-10-01 14:34 ` David Miller
2008-10-01 14:24 ` [net-next PATCH 10/16] Don't lookup the socket if there's a socket attached to the skb KOVACS Krisztian
2008-10-01 14:50 ` David Miller
2008-10-01 15:38 ` KOVACS Krisztian
2008-10-01 15:51 ` David Miller
2008-10-02 15:43 ` KOVACS Krisztian
2008-10-02 17:09 ` Arnaldo Carvalho de Melo
2008-10-02 19:58 ` David Miller
2008-10-03 8:57 ` KOVACS Krisztian
2008-10-03 13:47 ` Arnaldo Carvalho de Melo
2008-10-07 7:36 ` KOVACS Krisztian
2008-10-07 12:36 ` Arnaldo Carvalho de Melo
2008-10-07 18:42 ` David Miller
2008-10-07 7:42 ` [net-next PATCH] Add udplib_lookup_skb() helpers (was: [net-next PATCH 10/16] Don't lookup the socket if there's a socket attached to the skb) KOVACS Krisztian
2008-10-07 12:34 ` Arnaldo Carvalho de Melo
2008-10-07 19:39 ` [net-next PATCH] Add udplib_lookup_skb() helpers David Miller
2008-10-07 7:59 ` [net-next PATCH] Don't lookup the socket if there's a socket attached to the skb (was: Re: [net-next PATCH 10/16] Don't lookup the socket if there's a socket attached to the skb) KOVACS Krisztian
2008-10-07 12:36 ` Arnaldo Carvalho de Melo [this message]
2008-10-07 19:41 ` [net-next PATCH] Don't lookup the socket if there's a socket attached to the skb David Miller
2008-10-01 14:24 ` [net-next PATCH 01/16] Loosen source address check on IPv4 output KOVACS Krisztian
2008-10-01 14:28 ` David Miller
2008-10-01 14:24 ` [net-next PATCH 05/16] Conditionally enable transparent flow flag when connecting KOVACS Krisztian
2008-10-01 14:36 ` David Miller
2008-10-01 14:24 ` [net-next PATCH 03/16] Allow binding to non-local addresses if IP_TRANSPARENT is set KOVACS Krisztian
2008-10-01 14:31 ` David Miller
2008-10-01 14:24 ` [net-next PATCH 13/16] iptables tproxy core KOVACS Krisztian
2008-10-02 9:19 ` Patrick McHardy
2008-10-01 14:24 ` [net-next PATCH 07/16] Make Netfilter's ip_route_me_harder() non-local address compatible KOVACS Krisztian
2008-10-01 14:45 ` David Miller
2008-10-01 14:24 ` [net-next PATCH 06/16] Handle TCP SYN+ACK/ACK/RST transparency KOVACS Krisztian
2008-10-01 14:42 ` David Miller
2008-10-01 14:46 ` KOVACS Krisztian
2008-10-01 14:24 ` [net-next PATCH 12/16] Split Netfilter IPv4 defragmentation into a separate module KOVACS Krisztian
2008-10-02 9:18 ` Patrick McHardy
2008-10-01 14:24 ` [net-next PATCH 14/16] iptables socket match KOVACS Krisztian
2008-10-02 9:26 ` Patrick McHardy
2008-10-02 10:26 ` KOVACS Krisztian
2008-10-02 10:35 ` Patrick McHardy
2008-10-03 14:04 ` Jan Engelhardt
2008-10-01 14:24 ` [net-next PATCH 09/16] Export UDP socket lookup function KOVACS Krisztian
2008-10-01 14:48 ` David Miller
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=20081007123626.GB29610@ghostprotocols.net \
--to=acme@redhat.com \
--cc=davem@davemloft.net \
--cc=hidden@sch.bme.hu \
--cc=kaber@trash.net \
--cc=netdev@vger.kernel.org \
--cc=netfilter-devel@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).