From: Nicolas Dichtel <nicolas.dichtel@6wind•com>
To: Vlad Yasevich <vyasevich@gmail•com>
Cc: davem@davemloft•net, eric.dumazet@gmail•com, sri@us•ibm.com,
linux-sctp@vger•kernel.org, netdev@vger•kernel.org
Subject: Re: [PATCH net-next v2 3/4] ipv6: use net->rt_genid to check dst validity
Date: Mon, 10 Sep 2012 16:44:57 +0200 [thread overview]
Message-ID: <504DFCE9.9020100@6wind.com> (raw)
In-Reply-To: <504DFC9D.9010402@gmail.com>
Le 10/09/2012 16:43, Vlad Yasevich a écrit :
> On 09/10/2012 10:34 AM, Nicolas Dichtel wrote:
>> Le 10/09/2012 16:29, Vlad Yasevich a écrit :
>>> On 09/10/2012 09:22 AM, Nicolas Dichtel wrote:
>>>> IPv6 dst should take care of rt_genid too. When a xfrm policy is
>>>> inserted or
>>>> deleted, all dst should be invalidated.
>>>> To force the validation, dst entries should be created with
>>>> ->obsolete set to
>>>> DST_OBSOLETE_FORCE_CHK. This was already the case for all functions
>>>> calling
>>>> ip6_dst_alloc(), except for ip6_rt_copy().
>>>>
>>>> As a consequence, we can remove the specific code in
>>>> inet6_connection_sock.
>>>>
>>>> Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind•com>
>>>> ---
>>>> include/net/ip6_fib.h | 2 +-
>>>> net/ipv6/inet6_connection_sock.c | 23 +----------------------
>>>> net/ipv6/route.c | 17 +++++++++++++----
>>>> 3 files changed, 15 insertions(+), 27 deletions(-)
>>>>
>>>> diff --git a/include/net/ip6_fib.h b/include/net/ip6_fib.h
>>>> index cd64cf3..5eb93f4 100644
>>>> --- a/include/net/ip6_fib.h
>>>> +++ b/include/net/ip6_fib.h
>>>> @@ -113,7 +113,7 @@ struct rt6_info {
>>>> unsigned long _rt6i_peer;
>>>>
>>>> #ifdef CONFIG_XFRM
>>>> - u32 rt6i_flow_cache_genid;
>>>> + u32 rt6i_genid;
>>>> #endif
>>>> /* more non-fragment space at head required */
>>>> unsigned short rt6i_nfheader_len;
>>>> diff --git a/net/ipv6/inet6_connection_sock.c
>>>> b/net/ipv6/inet6_connection_sock.c
>>>> index 0251a60..c4f9341 100644
>>>> --- a/net/ipv6/inet6_connection_sock.c
>>>> +++ b/net/ipv6/inet6_connection_sock.c
>>>> @@ -175,33 +175,12 @@ void __inet6_csk_dst_store(struct sock *sk, struct
>>>> dst_entry *dst,
>>>> const struct in6_addr *saddr)
>>>> {
>>>> __ip6_dst_store(sk, dst, daddr, saddr);
>>>> -
>>>> -#ifdef CONFIG_XFRM
>>>> - {
>>>> - struct rt6_info *rt = (struct rt6_info *)dst;
>>>> - rt->rt6i_flow_cache_genid = atomic_read(&flow_cache_genid);
>>>> - }
>>>> -#endif
>>>> }
>>>>
>>>> static inline
>>>> struct dst_entry *__inet6_csk_dst_check(struct sock *sk, u32 cookie)
>>>> {
>>>> - struct dst_entry *dst;
>>>> -
>>>> - dst = __sk_dst_check(sk, cookie);
>>>> -
>>>> -#ifdef CONFIG_XFRM
>>>> - if (dst) {
>>>> - struct rt6_info *rt = (struct rt6_info *)dst;
>>>> - if (rt->rt6i_flow_cache_genid !=
>>>> atomic_read(&flow_cache_genid)) {
>>>> - __sk_dst_reset(sk);
>>>> - dst = NULL;
>>>> - }
>>>> - }
>>>> -#endif
>>>> -
>>>> - return dst;
>>>> + return __sk_dst_check(sk, cookie);
>>>> }
>>>>
>>>> static struct dst_entry *inet6_csk_route_socket(struct sock *sk,
>>>> diff --git a/net/ipv6/route.c b/net/ipv6/route.c
>>>> index 339d921..db7b78f 100644
>>>> --- a/net/ipv6/route.c
>>>> +++ b/net/ipv6/route.c
>>>> @@ -281,13 +281,16 @@ static inline struct rt6_info
>>>> *ip6_dst_alloc(struct net
>>>> *net,
>>>> struct fib6_table *table)
>>>> {
>>>> struct rt6_info *rt = dst_alloc(&net->ipv6.ip6_dst_ops, dev,
>>>> - 0, DST_OBSOLETE_NONE, flags);
>>>> + 0, DST_OBSOLETE_FORCE_CHK, flags);
>>>>
>>>> if (rt) {
>>>> struct dst_entry *dst = &rt->dst;
>>>>
>>>> memset(dst + 1, 0, sizeof(*rt) - sizeof(*dst));
>>>> rt6_init_peer(rt, table ? &table->tb6_peers :
>>>> net->ipv6.peers);
>>>> +#ifdef CONFIG_XFRM
>>>> + rt->rt6i_genid = rt_genid(net);
>>>> +#endif
>>>
>>> This isn't XFRM dependent any more, is it?
>> Not dependent, but for IPv6, it's only usefull when xfrm is set. Goal of
>> this ifdef was to avoid the test if xfrm is not used.
>
> It's not the usage, it's enable at build time and that's almost always on. Now
> the cache behavior is different when XFRM is excluded from the kernel build.
>
> Before the ifdef was needed since you were actually looking at xfrm variable.
> Not anymore. The ifdef doesn't make sense.
Ok, I will remove it.
next prev parent reply other threads:[~2012-09-10 14:45 UTC|newest]
Thread overview: 52+ messages / expand[flat|nested] mbox.gz Atom feed top
2012-09-06 17:40 [PATCH] sctp: check dst validity after IPsec operations Nicolas Dichtel
2012-09-06 16:04 ` Vlad Yasevich
2012-09-06 16:40 ` Nicolas Dichtel
2012-09-06 17:03 ` Vlad Yasevich
2012-09-07 12:24 ` Nicolas Dichtel
2012-09-07 12:07 ` Nicolas Dichtel
2012-09-06 18:10 ` David Miller
2012-09-07 13:47 ` Nicolas Dichtel
2012-09-07 15:57 ` [PATCH RFC net-next 0/2] Take care of xfrm policy when checking dst entries Nicolas Dichtel
2012-09-07 15:57 ` [PATCH 1/2] dst: take into account policy update on check() Nicolas Dichtel
2012-09-07 14:20 ` Vlad Yasevich
2012-09-07 14:35 ` Eric Dumazet
2012-09-07 14:47 ` Nicolas Dichtel
2012-09-07 15:09 ` Eric Dumazet
2012-09-07 15:13 ` Nicolas Dichtel
2012-09-07 15:21 ` Eric Dumazet
2012-09-07 18:48 ` David Miller
2012-09-07 18:48 ` David Miller
2012-09-10 12:47 ` Nicolas Dichtel
2012-09-10 13:10 ` Eric Dumazet
2012-09-10 13:22 ` [PATCH net-next v2] Take care of xfrm policy when checking dst entries Nicolas Dichtel
2012-09-10 13:22 ` [PATCH net-next v2 1/4] netns: move net->ipv4.rt_genid to net->rt_genid Nicolas Dichtel
2012-09-10 13:22 ` [PATCH net-next v2 2/4] xfrm: invalidate dst on policy insertion/deletion Nicolas Dichtel
2012-09-10 14:21 ` Vlad Yasevich
2012-09-10 14:56 ` Nicolas Dichtel
2012-09-11 8:09 ` [PATCH net-next v3 0/4] Take care of xfrm policy when checking dst entries Nicolas Dichtel
2012-09-11 8:09 ` [PATCH net-next v3 1/4] netns: move net->ipv4.rt_genid to net->rt_genid Nicolas Dichtel
2012-09-11 8:09 ` [PATCH net-next v3 2/4] xfrm: invalidate dst on policy insertion/deletion Nicolas Dichtel
2012-09-11 8:09 ` [PATCH net-next v3 3/4] ipv6: use net->rt_genid to check dst validity Nicolas Dichtel
2012-09-11 8:09 ` [PATCH net-next v3 4/4] ipv6: use DST_* macro to set obselete field Nicolas Dichtel
2012-09-12 7:40 ` Eric Dumazet
2012-09-17 16:49 ` [PATCH net-next v3 0/4] Take care of xfrm policy when checking dst entries David Miller
2012-09-17 18:14 ` Vlad Yasevich
2012-09-17 18:25 ` David Miller
2012-09-17 19:52 ` Nicolas Dichtel
2012-09-17 19:54 ` David Miller
2012-09-18 20:08 ` David Miller
2012-09-10 13:22 ` [PATCH net-next v2 3/4] ipv6: use net->rt_genid to check dst validity Nicolas Dichtel
2012-09-10 14:29 ` Vlad Yasevich
2012-09-10 14:34 ` Nicolas Dichtel
2012-09-10 14:43 ` Vlad Yasevich
2012-09-10 14:44 ` Nicolas Dichtel [this message]
2012-09-10 13:22 ` [PATCH net-next v2 4/4] ipv6: use DST_* macro to set obselete field Nicolas Dichtel
2012-09-10 14:35 ` [PATCH net-next v2] Take care of xfrm policy when checking dst entries Vlad Yasevich
2012-09-10 14:38 ` Nicolas Dichtel
2012-09-10 17:18 ` David Miller
2012-09-10 17:59 ` Vlad Yasevich
2012-09-10 18:01 ` David Miller
2012-09-22 16:49 ` Jan Engelhardt
2012-09-07 14:51 ` [PATCH 1/2] dst: take into account policy update on check() Vlad Yasevich
2012-09-07 15:08 ` Eric Dumazet
2012-09-07 15:57 ` [PATCH 2/2] ipv6: remove rt6i_flow_cache_genid field in rt6_info Nicolas Dichtel
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=504DFCE9.9020100@6wind.com \
--to=nicolas.dichtel@6wind$(echo .)com \
--cc=davem@davemloft$(echo .)net \
--cc=eric.dumazet@gmail$(echo .)com \
--cc=linux-sctp@vger$(echo .)kernel.org \
--cc=netdev@vger$(echo .)kernel.org \
--cc=sri@us$(echo .)ibm.com \
--cc=vyasevich@gmail$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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