From: Florian Westphal <fw@strlen.de>
To: Eric Dumazet <eric.dumazet@gmail.com>
Cc: David Miller <davem@davemloft.net>,
sebastian.poehn@gmail.com, netdev@vger.kernel.org
Subject: Re: [FYI] xfrm: Don't lookup sk_policy for timewait sockets
Date: Thu, 9 Apr 2015 21:25:40 +0200 [thread overview]
Message-ID: <20150409192540.GF20653@breakpoint.cc> (raw)
In-Reply-To: <1428607284.25985.269.camel@edumazet-glaptop2.roam.corp.google.com>
Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Thu, 2015-04-09 at 14:37 -0400, David Miller wrote:
> > From: Eric Dumazet <eric.dumazet@gmail.com>
> > Date: Thu, 09 Apr 2015 02:07:41 -0700
> >
> > > TCP stack never sends packets attached to a socket in timewait state.
> >
> > TPROXY assigns timewait sockets to skb->sk, that's the bug.
> >
> > In net/netfilter/xt_TPROXY.c:
> >
> > tproxy_tg4()
> > ...
> > sk = nf_tproxy_get_sock_v4(dev_net(skb->dev), iph->protocol,
> > iph->saddr, iph->daddr,
> > hp->source, hp->dest,
> > skb->dev, NFT_LOOKUP_ESTABLISHED);
> > /* NOTE: assign_sock consumes our sk reference */
> > if (sk && tproxy_sk_is_transparent(sk)) {
> > /* This should be in a separate target, but we don't do multiple
> > targets on the same rule yet */
> > skb->mark = (skb->mark & ~mark_mask) ^ mark_value;
> >
> > pr_debug("redirecting: proto %hhu %pI4:%hu -> %pI4:%hu, mark: %x\n",
> > iph->protocol, &iph->daddr, ntohs(hp->dest),
> > &laddr, ntohs(lport), skb->mark);
> >
> > nf_tproxy_assign_sock(skb, sk);
> > ...
> > /* assign a socket to the skb -- consumes sk */
> > static void
> > nf_tproxy_assign_sock(struct sk_buff *skb, struct sock *sk)
> > {
> > skb_orphan(skb);
> > skb->sk = sk;
> > skb->destructor = sock_edemux;
> > }
>
>
> Right, but stack trace shown by Sebastian seems to be an input frame,
> and we transmit a frame. This is the part I do not understand, yet.
Maybe policy routing is wedged so skb is erronously entering forward
path?
After all TPROXY depends on correct marking of skb + ip rule magic
to redirect skb to localhost.
prev parent reply other threads:[~2015-04-09 19:25 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-04-09 8:09 [FYI] xfrm: Don't lookup sk_policy for timewait sockets Sebastian Poehn
2015-04-09 9:07 ` Eric Dumazet
2015-04-09 9:24 ` Sebastian Poehn
2015-04-09 18:37 ` David Miller
2015-04-09 19:14 ` Florian Westphal
2015-04-09 21:07 ` David Miller
2015-04-09 21:21 ` Florian Westphal
2015-04-10 11:14 ` Sebastian Poehn
2015-04-13 8:04 ` Sebastian Poehn
2015-04-13 15:09 ` Sebastian Poehn
2015-04-13 15:39 ` Eric Dumazet
2015-04-13 17:25 ` David Miller
2015-04-13 16:04 ` Florian Westphal
2015-04-09 19:21 ` Eric Dumazet
2015-04-09 19:25 ` Florian Westphal [this message]
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=20150409192540.GF20653@breakpoint.cc \
--to=fw@strlen.de \
--cc=davem@davemloft.net \
--cc=eric.dumazet@gmail.com \
--cc=netdev@vger.kernel.org \
--cc=sebastian.poehn@gmail.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).