* Re: [PATCH] tun: orphan an skb on tx
From: Eric Dumazet @ 2010-04-13 18:31 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: Jan Kiszka, David S. Miller, Herbert Xu, Paul Moore,
David Woodhouse, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org, qemu-devel
In-Reply-To: <20100413173919.GC26011@redhat.com>
Le mardi 13 avril 2010 à 20:39 +0300, Michael S. Tsirkin a écrit :
> > When a socket with inflight tx packets is closed, we dont block the
> > close, we only delay the socket freeing once all packets were delivered
> > and freed.
> >
>
> Which is wrong, since this is under userspace control, so you get
> unkillable processes.
>
We do not get unkillable processes, at least with sockets I was thinking
about (TCP/UDP ones).
Maybe tun sockets can behave the same ?
Herbert Acked your patch, so I guess its OK, but I think it can be
dangerous.
Anyway my feeling is that we try to add various mechanisms to keep a
hostile user flooding another one.
For example, UDP got memory accounting quite recently, and we added
socket backlog limits very recently. It was considered not needed few
years ago.
^ permalink raw reply
* Re: [PATCH] tun: orphan an skb on tx
From: Michael S. Tsirkin @ 2010-04-13 17:39 UTC (permalink / raw)
To: Eric Dumazet
Cc: Jan Kiszka, David S. Miller, Herbert Xu, Paul Moore,
David Woodhouse, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org, qemu-devel
In-Reply-To: <1271176838.16881.537.camel@edumazet-laptop>
On Tue, Apr 13, 2010 at 06:40:38PM +0200, Eric Dumazet wrote:
> Le mardi 13 avril 2010 à 17:36 +0200, Jan Kiszka a écrit :
> > Michael S. Tsirkin wrote:
> > > The following situation was observed in the field:
> > > tap1 sends packets, tap2 does not consume them, as a result
> > > tap1 can not be closed.
> >
> > And before that, tap1 may not be able to send further packets to anyone
> > else on the bridge as its TX resources were blocked by tap2 - that's
> > what we saw in the field.
> >
>
> After the patch, tap1 is able to flood tap2, and tap3/tap4 not able to
> send one single frame. Is it OK ?
Yes :) This was always possible. Number of senders needed to flood
a receiver might vary depending on send/recv queue size
that you set. External sources can also fill your RX queue
if you let them. In the end, we need to rely on the scheduler for fairness,
or apply packet shaping.
> Back to the problem : tap1 cannot be closed.
>
> Why ? because of refcounts ?
Yes.
> When a socket with inflight tx packets is closed, we dont block the
> close, we only delay the socket freeing once all packets were delivered
> and freed.
>
Which is wrong, since this is under userspace control, so you get
unkillable processes.
--
MST
^ permalink raw reply
* Re: [Bonding-devel] [v3 Patch 2/3] bridge: make bridge support netpoll
From: Stephen Hemminger @ 2010-04-13 17:33 UTC (permalink / raw)
To: Jay Vosburgh
Cc: Cong Wang, Eric Dumazet, Neil Horman, netdev, Andy Gospodarek,
bridge, linux-kernel, bonding-devel, Jeff Moyer, Matt Mackall,
David Miller
In-Reply-To: <8304.1271177567@death.nxdomain.ibm.com>
On Tue, 13 Apr 2010 09:52:47 -0700
Jay Vosburgh <fubar@us.ibm.com> wrote:
> Cong Wang <amwang@redhat.com> wrote:
>
> >Stephen Hemminger wrote:
> >> On Mon, 12 Apr 2010 12:38:57 +0200
> >> Eric Dumazet <eric.dumazet@gmail.com> wrote:
> >>
> >>> Le lundi 12 avril 2010 à 18:37 +0800, Cong Wang a écrit :
> >>>> Stephen Hemminger wrote:
> >>>>> There is no protection on dev->priv_flags for SMP access.
> >>>>> It would better bit value in dev->state if you are using it as control flag.
> >>>>>
> >>>>> Then you could use
> >>>>> if (unlikely(test_and_clear_bit(__IN_NETPOLL, &skb->dev->state)))
> >>>>> netpoll_send_skb(...)
> >>>>>
> >>>>>
> >>>> Hmm, I think we can't use ->state here, it is not for this kind of purpose,
> >>>> according to its comments.
> >>>>
> >>>> Also, I find other usages of IFF_XXX flags of ->priv_flags are also using
> >>>> &, | to set or clear the flags. So there must be some other things preventing
> >>>> the race...
> >>> Yes, its RTNL that protects priv_flags changes, hopefully...
> >>>
> >>
> >> The patch was not protecting priv_flags with RTNL.
> >> For example..
> >>
> >>
> >> @@ -308,7 +312,9 @@ static void netpoll_send_skb(struct netp
> >> tries > 0; --tries) {
> >> if (__netif_tx_trylock(txq)) {
> >> if (!netif_tx_queue_stopped(txq)) {
> >> + dev->priv_flags |= IFF_IN_NETPOLL;
> >> status = ops->ndo_start_xmit(skb, dev);
> >> + dev->priv_flags &= ~IFF_IN_NETPOLL;
> >> if (status == NETDEV_TX_OK)
> >> txq_trans_update(txq);
> >
> >Hmm, but I checked the bonding case (IFF_BONDING), it doesn't
> >hold rtnl_lock. Strange.
>
> I looked, and there are a couple of cases in bonding that don't
> have RTNL for adjusting priv_flags (in bond_ab_arp_probe when no slaves
> are up, and a couple of cases in 802.3ad). I think the solution there
> is to move bonding away from priv_flags for some of this (e.g., convert
> bonding to use a frame hook like bridge and macvlan, and greatly
> simplify skb_bond_should_drop), but that's a separate topic.
>
> The majority of the cases, however, do hold RTNL. Bonding
> generally doesn't have to acquire RTNL itself, since whatever called
> into bonding is holding it already. For example, the slave add and
> remove paths (bond_enslave, bond_release) are called either via sysfs or
> ioctl, both of which acquire RTNL. All of the set and clear operations
> for IFF_BONDING fall into this category; look at bonding_store_slaves
> for an example.
>
> Bonding does acquire RTNL itself when performing failovers,
> e.g., bond_mii_monitor holds RTNL prior to calling bond_miimon_commit,
> which will change priv_flags.
>
All this was related to netpoll. And netpoll processing often needs to occur
in hard IRQ context. Therefor netpoll stuff and RTNL (which is a mutex),
really don't mix well. Keep RTNL for what it was meant for network
reconfiguration. Don't turn it into a network special BKL.
--
^ permalink raw reply
* Re: Network protocol (IP,IPv6,...) and TC actions
From: Jan Ceuleers @ 2010-04-13 17:31 UTC (permalink / raw)
To: Grégoire Baron, netdev
In-Reply-To: <20100411190845.GA18383@n7mm.org>
Grégoire Baron wrote:
> As this .protocol member seems to be used at different moments when a
> packet is received, forwared or sent, and could contain something like
> ETH_P_8021Q which isn't a network protocol Id, can we say the struct
> sk_buff .protocol member is guaranteed to contain a network protocol Id
> in the struct sb_buff used in the TC action executions ?
Grégoire,
I suggest that you ask your question on the netdev mailing list (netdev@vger.kernel.org).
Cheers, Jan
^ permalink raw reply
* Re: forcedeth driver hangs under heavy load
From: Xose Vazquez Perez @ 2010-04-13 17:22 UTC (permalink / raw)
To: netdev
stephen mulcahy wrote:
> running Hadoop[1] TeraSort[2] but I haven't identified a simpler
> reproducer. I tried to recreate this with iperf and ping -f but neither
> helped - it may be that the problem only occurs when systems are passing
> large amounts of traffic and have very high cpu utilisation (when
Did you try ISIC(IP Stack Integrity Checker)[1] tools ?
Net-drivers usually break running these tools.
[1] http://isic.sf.net needs libnet[2]
[2] http://github.com/sam-github/libnet
--
«Allá muevan feroz guerra, ciegos reyes por un palmo más de tierra;
que yo aquí tengo por mío cuanto abarca el mar bravío, a quien nadie
impuso leyes. Y no hay playa, sea cualquiera, ni bandera de esplendor,
que no sienta mi derecho y dé pecho a mi valor.»
^ permalink raw reply
* Re: [PATCH] tun: orphan an skb on tx
From: Jan Kiszka @ 2010-04-13 16:52 UTC (permalink / raw)
To: Eric Dumazet
Cc: Michael S. Tsirkin, David S. Miller, Herbert Xu, Paul Moore,
David Woodhouse, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org, qemu-devel
In-Reply-To: <1271176838.16881.537.camel@edumazet-laptop>
Eric Dumazet wrote:
> Le mardi 13 avril 2010 à 17:36 +0200, Jan Kiszka a écrit :
>> Michael S. Tsirkin wrote:
>>> The following situation was observed in the field:
>>> tap1 sends packets, tap2 does not consume them, as a result
>>> tap1 can not be closed.
>> And before that, tap1 may not be able to send further packets to anyone
>> else on the bridge as its TX resources were blocked by tap2 - that's
>> what we saw in the field.
>>
>
> After the patch, tap1 is able to flood tap2, and tap3/tap4 not able to
> send one single frame. Is it OK ?
I think if that's a real issue, you have to apply traffic shaping to the
untrusted nodes. The existing flow-control scheme was fragile anyway as
you had to translate packet lengths on TX side to packet counts on RX.
Jan
--
Siemens AG, Corporate Technology, CT T DE IT 1
Corporate Competence Center Embedded Linux
^ permalink raw reply
* Re: [Bonding-devel] [v3 Patch 2/3] bridge: make bridge support netpoll
From: Jay Vosburgh @ 2010-04-13 16:52 UTC (permalink / raw)
To: Cong Wang
Cc: Stephen Hemminger, Eric Dumazet, Neil Horman, netdev,
Andy Gospodarek, bridge, linux-kernel, bonding-devel, Jeff Moyer,
Matt Mackall, David Miller
In-Reply-To: <4BC43214.6030009@redhat.com>
Cong Wang <amwang@redhat.com> wrote:
>Stephen Hemminger wrote:
>> On Mon, 12 Apr 2010 12:38:57 +0200
>> Eric Dumazet <eric.dumazet@gmail.com> wrote:
>>
>>> Le lundi 12 avril 2010 à 18:37 +0800, Cong Wang a écrit :
>>>> Stephen Hemminger wrote:
>>>>> There is no protection on dev->priv_flags for SMP access.
>>>>> It would better bit value in dev->state if you are using it as control flag.
>>>>>
>>>>> Then you could use
>>>>> if (unlikely(test_and_clear_bit(__IN_NETPOLL, &skb->dev->state)))
>>>>> netpoll_send_skb(...)
>>>>>
>>>>>
>>>> Hmm, I think we can't use ->state here, it is not for this kind of purpose,
>>>> according to its comments.
>>>>
>>>> Also, I find other usages of IFF_XXX flags of ->priv_flags are also using
>>>> &, | to set or clear the flags. So there must be some other things preventing
>>>> the race...
>>> Yes, its RTNL that protects priv_flags changes, hopefully...
>>>
>>
>> The patch was not protecting priv_flags with RTNL.
>> For example..
>>
>>
>> @@ -308,7 +312,9 @@ static void netpoll_send_skb(struct netp
>> tries > 0; --tries) {
>> if (__netif_tx_trylock(txq)) {
>> if (!netif_tx_queue_stopped(txq)) {
>> + dev->priv_flags |= IFF_IN_NETPOLL;
>> status = ops->ndo_start_xmit(skb, dev);
>> + dev->priv_flags &= ~IFF_IN_NETPOLL;
>> if (status == NETDEV_TX_OK)
>> txq_trans_update(txq);
>
>Hmm, but I checked the bonding case (IFF_BONDING), it doesn't
>hold rtnl_lock. Strange.
I looked, and there are a couple of cases in bonding that don't
have RTNL for adjusting priv_flags (in bond_ab_arp_probe when no slaves
are up, and a couple of cases in 802.3ad). I think the solution there
is to move bonding away from priv_flags for some of this (e.g., convert
bonding to use a frame hook like bridge and macvlan, and greatly
simplify skb_bond_should_drop), but that's a separate topic.
The majority of the cases, however, do hold RTNL. Bonding
generally doesn't have to acquire RTNL itself, since whatever called
into bonding is holding it already. For example, the slave add and
remove paths (bond_enslave, bond_release) are called either via sysfs or
ioctl, both of which acquire RTNL. All of the set and clear operations
for IFF_BONDING fall into this category; look at bonding_store_slaves
for an example.
Bonding does acquire RTNL itself when performing failovers,
e.g., bond_mii_monitor holds RTNL prior to calling bond_miimon_commit,
which will change priv_flags.
-J
---
-Jay Vosburgh, IBM Linux Technology Center, fubar@us.ibm.com
^ permalink raw reply
* Re: [PATCH] tun: orphan an skb on tx
From: Eric Dumazet @ 2010-04-13 16:40 UTC (permalink / raw)
To: Jan Kiszka
Cc: Michael S. Tsirkin, David S. Miller, Herbert Xu, Paul Moore,
David Woodhouse, netdev@vger.kernel.org,
linux-kernel@vger.kernel.org, qemu-devel
In-Reply-To: <4BC48F79.5090409@siemens.com>
Le mardi 13 avril 2010 à 17:36 +0200, Jan Kiszka a écrit :
> Michael S. Tsirkin wrote:
> > The following situation was observed in the field:
> > tap1 sends packets, tap2 does not consume them, as a result
> > tap1 can not be closed.
>
> And before that, tap1 may not be able to send further packets to anyone
> else on the bridge as its TX resources were blocked by tap2 - that's
> what we saw in the field.
>
After the patch, tap1 is able to flood tap2, and tap3/tap4 not able to
send one single frame. Is it OK ?
Back to the problem : tap1 cannot be closed.
Why ? because of refcounts ?
When a socket with inflight tx packets is closed, we dont block the
close, we only delay the socket freeing once all packets were delivered
and freed.
^ permalink raw reply
* Re: SO_REUSEADDR with UDP (again)
From: Eric Dumazet @ 2010-04-13 16:36 UTC (permalink / raw)
To: Michal Svoboda; +Cc: netdev
In-Reply-To: <20100413162326.GD16595@myhost.felk.cvut.cz>
Le mardi 13 avril 2010 à 18:23 +0200, Michal Svoboda a écrit :
> Eric Dumazet wrote:
> > sock1 = socket(AF_INET, SOCK_DGRAM, 0);
> > setsockopt(sock1, SOL_SOCKET, SO_REUSEADDR, &on, sizeof(on));
> > addr.sin_addr.s_addr = htonl(0x7f000001);
> > if (bind(sock1, (struct sockaddr *)&addr, sizeof(addr)))
> > perror("bind1");
> >
> > sock2 = socket(AF_INET, SOCK_DGRAM, 0);
> > setsockopt(sock2, SOL_SOCKET, SO_REUSEADDR, &on, sizeof(on));
> > addr.sin_addr.s_addr = INADDR_ANY; /* or htonl(0x7f000001); */
> > if (bind(sock2, (struct sockaddr *)&addr, sizeof(addr)))
> > perror("bind2");
> > }
>
> Well, now if I send to 127.0.0.1, who gets the datagram? I guess sock2,
> so it steals from sock1. What practical use does this have?
>
No, sock1 will get the frame.
In udp receive (kernel), we chose the socket with highest score.
A socket bound to an IP address (not 0.0.0.0) has a bonus.
A connected socket has an extra bonus.
A socket bound to a device has an extra bonus.
> > Therefore, applications should not use REUSEADDR on unicast UDP, unless
> > it is a non security issue (for example, if it is able to react to any
> > new IP addresses added by the administrator on the machine, and complain
> > loudly if another application could bind() before itself)
>
> I don't think that in that case REUSEADDR would be useful because you
> can already claim new addresses without it, either by binding a separate
> socket to each IP or by binding to 0.0.0.0. Moreover the detection of
> the "complain" case would be very tricky, at least on first sight.
>
> > REUSADDR has a meaning for multicast, but for unicast... this is hardly
> > useful ?
>
> So would it be somehow possible to deliver the datagram to both sockets
> (for example if they would be SO_BROADCAST as well)?
Not without a change in kernel. AFAIK no other OS do that anyway.
^ permalink raw reply
* RE: [Patch 3/3] net: reserve ports for applications using fixed port numbers
From: Sean Hefty @ 2010-04-13 16:32 UTC (permalink / raw)
To: 'Tetsuo Handa', amwang, rolandd
Cc: opurdila, eric.dumazet, netdev, nhorman, davem, ebiederm,
linux-kernel
In-Reply-To: <201004132207.GAJ52684.OJFtMQVFHOSFLO@I-love.SAKURA.ne.jp>
>Sean and Roland, is below patch correct?
>inet_is_reserved_local_port() is the new function proposed in this patchset.
It looks correct to me. I didn't test the patch series, but if I comment out
the call to inet_is_reserved_local_port() in the provided below, the changes
worked fine for me.
Acked-by: Sean Hefty <sean.hefty@intel.com>
^ permalink raw reply
* Re: SO_REUSEADDR with UDP (again)
From: Michal Svoboda @ 2010-04-13 16:23 UTC (permalink / raw)
To: netdev
In-Reply-To: <1271161297.16881.293.camel@edumazet-laptop>
[-- Attachment #1: Type: text/plain, Size: 1469 bytes --]
Eric Dumazet wrote:
> sock1 = socket(AF_INET, SOCK_DGRAM, 0);
> setsockopt(sock1, SOL_SOCKET, SO_REUSEADDR, &on, sizeof(on));
> addr.sin_addr.s_addr = htonl(0x7f000001);
> if (bind(sock1, (struct sockaddr *)&addr, sizeof(addr)))
> perror("bind1");
>
> sock2 = socket(AF_INET, SOCK_DGRAM, 0);
> setsockopt(sock2, SOL_SOCKET, SO_REUSEADDR, &on, sizeof(on));
> addr.sin_addr.s_addr = INADDR_ANY; /* or htonl(0x7f000001); */
> if (bind(sock2, (struct sockaddr *)&addr, sizeof(addr)))
> perror("bind2");
> }
Well, now if I send to 127.0.0.1, who gets the datagram? I guess sock2,
so it steals from sock1. What practical use does this have?
> Therefore, applications should not use REUSEADDR on unicast UDP, unless
> it is a non security issue (for example, if it is able to react to any
> new IP addresses added by the administrator on the machine, and complain
> loudly if another application could bind() before itself)
I don't think that in that case REUSEADDR would be useful because you
can already claim new addresses without it, either by binding a separate
socket to each IP or by binding to 0.0.0.0. Moreover the detection of
the "complain" case would be very tricky, at least on first sight.
> REUSADDR has a meaning for multicast, but for unicast... this is hardly
> useful ?
So would it be somehow possible to deliver the datagram to both sockets
(for example if they would be SO_BROADCAST as well)?
Michal Svoboda
[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]
^ permalink raw reply
* Re: [PATCH v2] net: batch skb dequeueing from softnet input_pkt_queue
From: Paul E. McKenney @ 2010-04-13 15:52 UTC (permalink / raw)
To: Changli Gao; +Cc: Eric Dumazet, David S. Miller, netdev
In-Reply-To: <o2v412e6f7f1004130250rdbe17c5dr5c11c3cbbb2f6dbb@mail.gmail.com>
On Tue, Apr 13, 2010 at 05:50:29PM +0800, Changli Gao wrote:
> On Tue, Apr 13, 2010 at 4:08 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> >
> > Probably not necessary.
> >
> >> + volatile bool flush_processing_queue;
> >
> > Use of 'volatile' is strongly discouraged, I would say, forbidden.
>
> volatile is used to avoid compiler optimization.
Would it be reasonable to use ACCESS_ONCE() where this variable is used?
Thanx, Paul
> > Its usually a sign of 'I dont exactly what memory ordering I need, so I
> > throw a volatile just in case'. We live in a world full of RCU, read ,
> > write, full barriers. And these apis are well documented.
> >
>
> There isn't memory accessing order problem.
>
> >> @@ -2803,6 +2808,7 @@ static void flush_backlog(void *arg)
> >> __skb_unlink(skb, &queue->input_pkt_queue);
> >> kfree_skb(skb);
> >> }
> >> + queue->flush_processing_queue = true;
> >
> > Probably not necessary
> >
>
> If flush_backlog() is called when there are still packets in
> processing_queue, there maybe some packets refer to the netdev gone,
> if we remove this line.
>
> >> rps_unlock(queue);
> >> }
> >>
> >> @@ -3112,14 +3118,23 @@ static int process_backlog(struct napi_struct *napi, int quota)
> >> struct softnet_data *queue = &__get_cpu_var(softnet_data);
> >> unsigned long start_time = jiffies;
> >>
> >> + if (queue->flush_processing_queue) {
> >
> > Really... this is bloat IMHO
>
>
> Any better idea?
>
> >
> >>
> >
> > I advise to keep it simple.
> >
> > My suggestion would be to limit this patch only to process_backlog().
> >
> > Really if you touch other areas, there is too much risk.
> >
> > Perform sort of skb_queue_splice_tail_init() into a local (stack) queue,
> > but the trick is to not touch input_pkt_queue.qlen, so that we dont slow
> > down enqueue_to_backlog().
> >
> > Process at most 'quota' skbs (or jiffies limit).
> >
> > relock queue.
> > input_pkt_queue.qlen -= number_of_handled_skbs;
> >
>
> Oh no, in order to let latter packets in as soon as possible, we have
> to update qlen immediately.
>
> --
> Regards，
> Changli Gao(xiaosuo@gmail.com)
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [PATCH] tun: orphan an skb on tx
From: Jan Kiszka @ 2010-04-13 15:36 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: David S. Miller, Herbert Xu, Paul Moore, David Woodhouse,
netdev@vger.kernel.org, linux-kernel@vger.kernel.org, qemu-devel
In-Reply-To: <20100413145944.GA7716@redhat.com>
Michael S. Tsirkin wrote:
> The following situation was observed in the field:
> tap1 sends packets, tap2 does not consume them, as a result
> tap1 can not be closed.
And before that, tap1 may not be able to send further packets to anyone
else on the bridge as its TX resources were blocked by tap2 - that's
what we saw in the field.
> This happens because
> tun/tap devices can hang on to skbs undefinitely.
>
> As noted by Herbert, possible solutions include a timeout followed by a
> copy/change of ownership of the skb, or always copying/changing
> ownership if we're going into a hostile device.
>
> This patch implements the second approach.
>
> Note: one issue still remaining is that since skbs
> keep reference to tun socket and tun socket has a
> reference to tun device, we won't flush backlog,
> instead simply waiting for all skbs to get transmitted.
> At least this is not user-triggerable, and
> this was not reported in practice, my assumption is
> other devices besides tap complete an skb
> within finite time after it has been queued.
>
> A possible solution for the second issue
> would not to have socket reference the device,
> instead, implement dev->destructor for tun, and
> wait for all skbs to complete there, but this
> needs some thought, probably too risky for 2.6.34.
>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> Tested-by: Yan Vugenfirer <yvugenfi@redhat.com>
>
> ---
>
> Please review the below, and consider for 2.6.34,
> and stable trees.
>
> drivers/net/tun.c | 4 ++++
> 1 files changed, 4 insertions(+), 0 deletions(-)
>
> diff --git a/drivers/net/tun.c b/drivers/net/tun.c
> index 96c39bd..4326520 100644
> --- a/drivers/net/tun.c
> +++ b/drivers/net/tun.c
> @@ -387,6 +387,10 @@ static netdev_tx_t tun_net_xmit(struct sk_buff *skb, struct net_device *dev)
> }
> }
>
> + /* Orphan the skb - required as we might hang on to it
> + * for indefinite time. */
> + skb_orphan(skb);
> +
> /* Enqueue packet */
> skb_queue_tail(&tun->socket.sk->sk_receive_queue, skb);
> dev->trans_start = jiffies;
Nice solution, thanks!
Jan
--
Siemens AG, Corporate Technology, CT T DE IT 1
Corporate Competence Center Embedded Linux
^ permalink raw reply
* Re: forcedeth driver hangs under heavy load
From: stephen mulcahy @ 2010-04-13 15:25 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Ben Hutchings, netdev, Ben Hutchings, Ayaz Abdulla, 572201
In-Reply-To: <1271172172.16881.484.camel@edumazet-laptop>
Eric Dumazet wrote:
> OK, thanks for clarification.
>
> Last question, did you tried a vanilla kernel, aka 2.6.33.2 for
> example ?
I built a Debian package from the vanilla 2.6.33.2 and installed that on
all nodes and tried my reproducer with the same results - nodes becoming
unresponsive.
I didn't try changing the tso and tx settings with the 2.6.33.2 kernel
though. Let me know if that would be useful (and/or if there is another
kernel that you would like me to test with) and I'll try to fit it in.
Thanks again for your help.
-stephen
^ permalink raw reply
* Re: forcedeth driver hangs under heavy load
From: Eric Dumazet @ 2010-04-13 15:22 UTC (permalink / raw)
To: stephen mulcahy
Cc: Ben Hutchings, netdev, Ben Hutchings, Ayaz Abdulla, 572201
In-Reply-To: <4BC488DA.3020508@gmail.com>
Le mardi 13 avril 2010 à 16:08 +0100, stephen mulcahy a écrit :
> Eric Dumazet wrote:
>
> > I am scratching my head, but I thought you told me that
> >
> > ethtool -K eth0 tso off
> > ethtool -K eth0 tx on
> >
> > was working ?
>
> No, sorry for the confusion.
>
> ethtool -K eth0 tx off
>
> fixes the problem.
>
>
> Setting only
>
> ethtool -K eth0 tso off
> ethtool -K eth0 tx on
>
> still results in failures.
OK, thanks for clarification.
Last question, did you tried a vanilla kernel, aka 2.6.33.2 for
example ?
^ permalink raw reply
* Re: [PATCH] tun: orphan an skb on tx
From: Herbert Xu @ 2010-04-13 15:12 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: David S. Miller, Paul Moore, David Woodhouse, netdev,
linux-kernel, Jan Kiszka, qemu-devel
In-Reply-To: <20100413145944.GA7716@redhat.com>
On Tue, Apr 13, 2010 at 05:59:44PM +0300, Michael S. Tsirkin wrote:
> The following situation was observed in the field:
> tap1 sends packets, tap2 does not consume them, as a result
> tap1 can not be closed. This happens because
> tun/tap devices can hang on to skbs undefinitely.
>
> As noted by Herbert, possible solutions include a timeout followed by a
> copy/change of ownership of the skb, or always copying/changing
> ownership if we're going into a hostile device.
>
> This patch implements the second approach.
>
> Note: one issue still remaining is that since skbs
> keep reference to tun socket and tun socket has a
> reference to tun device, we won't flush backlog,
> instead simply waiting for all skbs to get transmitted.
> At least this is not user-triggerable, and
> this was not reported in practice, my assumption is
> other devices besides tap complete an skb
> within finite time after it has been queued.
>
> A possible solution for the second issue
> would not to have socket reference the device,
> instead, implement dev->destructor for tun, and
> wait for all skbs to complete there, but this
> needs some thought, probably too risky for 2.6.34.
>
> Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
> Tested-by: Yan Vugenfirer <yvugenfi@redhat.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Thanks,
--
Visit Openswan at http://www.openswan.org/
Email: Herbert Xu ~{PmV>HI~} <herbert@gondor.apana.org.au>
Home Page: http://gondor.apana.org.au/~herbert/
PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt
^ permalink raw reply
* Re: forcedeth driver hangs under heavy load
From: stephen mulcahy @ 2010-04-13 15:08 UTC (permalink / raw)
To: Eric Dumazet; +Cc: Ben Hutchings, netdev, Ben Hutchings, Ayaz Abdulla, 572201
In-Reply-To: <1271171107.16881.465.camel@edumazet-laptop>
Eric Dumazet wrote:
> I am scratching my head, but I thought you told me that
>
> ethtool -K eth0 tso off
> ethtool -K eth0 tx on
>
> was working ?
No, sorry for the confusion.
ethtool -K eth0 tx off
fixes the problem.
Setting only
ethtool -K eth0 tso off
ethtool -K eth0 tx on
still results in failures.
-stephen
^ permalink raw reply
* Re: [Bugme-new] [Bug 15720] New: IPv6's ipv4-compatibility addresses don't bind
From: Brian Haley @ 2010-04-13 15:06 UTC (permalink / raw)
To: charles; +Cc: Andrew Morton, netdev, bugzilla-daemon, bugme-daemon
In-Reply-To: <20100412212241.d8bf4185.akpm@linux-foundation.org>
Andrew Morton wrote:
>> When attempting to bind to an address using ipv4-compatibility, for example,
>> "::ffff:127.0.0.1", Linux refuses to bind to that address when
>> /proc/sys/net/ipv6/bindv6only is set.
This is actually not a (deprecated) IPv4-compatible IPv6 address (::127.0.0.1),
but an IPv4-mapped IPv6 address.
>> It seems to me that the intent of "bindv6only" was to not bind to the ipv4
>> address when you bind to all addresses (specifically ipv6 address "::"). So
>> when you bind to ::, an ipv4 client connects to you, and it appears to be
>> connecting from ::ffff:192.168.5.5. I don't think its intent was to effectively
>> disable binding to ::ffff:x.x.x.x addresses - just breaking that feature makes
>> no sense.
It is documented in ip-sysctl.txt:
bindv6only - BOOLEAN
Default value for IPV6_V6ONLY socket option,
which restricts use of the IPv6 socket to IPv6 communication
only.
TRUE: disable IPv4-mapped address feature
FALSE: enable IPv4-mapped address feature
Default: FALSE (as specified in RFC2553bis)
-Brian
^ permalink raw reply
* [PATCH iproute] ip: add support for multicast rules
From: Patrick McHardy @ 2010-04-13 15:06 UTC (permalink / raw)
To: Stephen Hemminger; +Cc: Linux Netdev List
[-- Attachment #1: Type: text/plain, Size: 207 bytes --]
This patch adds support for a "ip mrule" command, which is used
to configure multicast routing rules.
The corresponding kernel patches have been sent to Dave and
should (hopefully) appear in net-next soon.
[-- Attachment #2: mrule.diff --]
[-- Type: text/x-patch, Size: 2934 bytes --]
>From 1c21ad3ad0a5cd10d1f8cf9c7c5f5b28ae558103 Mon Sep 17 00:00:00 2001
From: Patrick McHardy <kaber@trash.net>
Date: Tue, 13 Apr 2010 17:03:47 +0200
Subject: [PATCH] ip: add support for multicast rules
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
include/linux/fib_rules.h | 8 ++++++++
ip/ip.c | 3 ++-
ip/ip_common.h | 1 +
ip/iprule.c | 12 ++++++++++++
4 files changed, 23 insertions(+), 1 deletions(-)
diff --git a/include/linux/fib_rules.h b/include/linux/fib_rules.h
index 51da65b..04a3976 100644
--- a/include/linux/fib_rules.h
+++ b/include/linux/fib_rules.h
@@ -15,6 +15,14 @@
/* try to find source address in routing lookups */
#define FIB_RULE_FIND_SADDR 0x00010000
+/* fib_rules families. values up to 127 are reserved for real address
+ * families, values above 128 may be used arbitrarily.
+ */
+#define FIB_RULES_IPV4 AF_INET
+#define FIB_RULES_IPV6 AF_INET6
+#define FIB_RULES_DECNET AF_DECnet
+#define FIB_RULES_IPMR 128
+
struct fib_rule_hdr {
__u8 family;
__u8 dst_len;
diff --git a/ip/ip.c b/ip/ip.c
index e0cf175..9f29533 100644
--- a/ip/ip.c
+++ b/ip/ip.c
@@ -42,7 +42,7 @@ static void usage(void)
"Usage: ip [ OPTIONS ] OBJECT { COMMAND | help }\n"
" ip [ -force ] -batch filename\n"
"where OBJECT := { link | addr | addrlabel | route | rule | neigh | ntable |\n"
-" tunnel | tuntap | maddr | mroute | monitor | xfrm }\n"
+" tunnel | tuntap | maddr | mroute | mrule | monitor | xfrm }\n"
" OPTIONS := { -V[ersion] | -s[tatistics] | -d[etails] | -r[esolve] |\n"
" -f[amily] { inet | inet6 | ipx | dnet | link } |\n"
" -o[neline] | -t[imestamp] | -b[atch] [filename] |\n"
@@ -76,6 +76,7 @@ static const struct cmd {
{ "monitor", do_ipmonitor },
{ "xfrm", do_xfrm },
{ "mroute", do_multiroute },
+ { "mrule", do_multirule },
{ "help", do_help },
{ 0 }
};
diff --git a/ip/ip_common.h b/ip/ip_common.h
index c857667..a114186 100644
--- a/ip/ip_common.h
+++ b/ip/ip_common.h
@@ -37,6 +37,7 @@ extern int do_iplink(int argc, char **argv);
extern int do_ipmonitor(int argc, char **argv);
extern int do_multiaddr(int argc, char **argv);
extern int do_multiroute(int argc, char **argv);
+extern int do_multirule(int argc, char **argv);
extern int do_xfrm(int argc, char **argv);
static inline int rtm_get_table(struct rtmsg *r, struct rtattr **tb)
diff --git a/ip/iprule.c b/ip/iprule.c
index 7140375..abb8752 100644
--- a/ip/iprule.c
+++ b/ip/iprule.c
@@ -436,3 +436,15 @@ int do_iprule(int argc, char **argv)
exit(-1);
}
+int do_multirule(int argc, char **argv)
+{
+ if (preferred_family == AF_UNSPEC ||
+ preferred_family == AF_INET)
+ preferred_family = FIB_RULES_IPMR;
+ else {
+ fprintf(stderr, "Multicast rules are only supported for IPv4\n");
+ exit(-1);
+ }
+
+ return do_iprule(argc, argv);
+}
--
1.7.0.4
^ permalink raw reply related
* Re: forcedeth driver hangs under heavy load
From: Eric Dumazet @ 2010-04-13 15:05 UTC (permalink / raw)
To: stephen mulcahy
Cc: Ben Hutchings, netdev, Ben Hutchings, Ayaz Abdulla, 572201
In-Reply-To: <4BC48460.4040001@gmail.com>
Le mardi 13 avril 2010 à 15:49 +0100, stephen mulcahy a écrit :
> Eric Dumazet wrote:
> > Le mardi 13 avril 2010 à 15:27 +0100, stephen mulcahy a écrit :
> >> Ok, I've tried both of the following with my reproducer
> >>
> >> 1. ethtool -K eth0 tso off
> >>
> >> RESULT: reproducer causes multiple hosts to be come unresponsive on
> >> first run.
> >>
> >> 2. ethtool -K eth0 tx off
> >>
> >> RESULT: reproducer runs three times without any hosts becoming unresponsive.
> >>
> >> -stephen
> >
> > Thanks Stephen !
> >
> > Now some brave fouls to check the 6410 lines of this driver ? ;)
> >
> > Question of the day : Why TSO is broken in forcedeth ?
> > Is it generically broken or is it broken for specific NICS ?
> >
>
> Actually, it is only when tx-checksumming is turned off that the problem
> doesn't occur (so I'm not sure TSO is the problem).
>
> Additionally, a google also turns up this existing Debian bug
> http://bugs.debian.org/506419 which seems to be related.
>
> -stephen
>
I am scratching my head, but I thought you told me that
ethtool -K eth0 tso off
ethtool -K eth0 tx on
was working ?
And, given that when you perform "ethtool -K eth0 tx off", TSO is
automatically off...
tx checksums work, TSO doesnt work, therefore TSO is broken ?
^ permalink raw reply
* [PATCH 5/9] ipv4: ipmr: move unres_queue and timer to per-namespace data
From: Patrick McHardy @ 2010-04-13 15:03 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1271171003-11901-1-git-send-email-kaber@trash.net>
The unres_queue is currently shared between all namespaces. Following patches
will additionally allow to create multiple multicast routing tables in each
namespace. Having a single shared queue for all these users seems to excessive,
move the queue and the cleanup timer to the per-namespace data to unshare it.
As a side-effect, this fixes a bug in the seq file iteration functions: the
first entry returned is always from the current namespace, entries returned
after that may belong to any namespace.
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
include/net/netns/ipv4.h | 2 +
net/ipv4/ipmr.c | 70 +++++++++++++++++++---------------------------
2 files changed, 31 insertions(+), 41 deletions(-)
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
index 2764994..b15e518 100644
--- a/include/net/netns/ipv4.h
+++ b/include/net/netns/ipv4.h
@@ -60,6 +60,8 @@ struct netns_ipv4 {
#ifdef CONFIG_IP_MROUTE
struct sock *mroute_sk;
+ struct timer_list ipmr_expire_timer;
+ struct mfc_cache *mfc_unres_queue;
struct mfc_cache **mfc_cache_array;
struct vif_device *vif_table;
int maxvif;
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 9d4f6d1..d6aa65e 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -80,8 +80,6 @@ static DEFINE_RWLOCK(mrt_lock);
#define VIF_EXISTS(_net, _idx) ((_net)->ipv4.vif_table[_idx].dev != NULL)
-static struct mfc_cache *mfc_unres_queue; /* Queue of unresolved entries */
-
/* Special spinlock for queue of unresolved entries */
static DEFINE_SPINLOCK(mfc_unres_lock);
@@ -100,8 +98,6 @@ static int ipmr_cache_report(struct net *net,
struct sk_buff *pkt, vifi_t vifi, int assert);
static int ipmr_fill_mroute(struct sk_buff *skb, struct mfc_cache *c, struct rtmsg *rtm);
-static struct timer_list ipmr_expire_timer;
-
/* Service routines creating virtual interfaces: DVMRP tunnels and PIMREG */
static void ipmr_del_tunnel(struct net_device *dev, struct vifctl *v)
@@ -364,25 +360,26 @@ static void ipmr_destroy_unres(struct mfc_cache *c)
}
-/* Single timer process for all the unresolved queue. */
+/* Timer process for the unresolved queue. */
-static void ipmr_expire_process(unsigned long dummy)
+static void ipmr_expire_process(unsigned long arg)
{
+ struct net *net = (struct net *)arg;
unsigned long now;
unsigned long expires;
struct mfc_cache *c, **cp;
if (!spin_trylock(&mfc_unres_lock)) {
- mod_timer(&ipmr_expire_timer, jiffies+HZ/10);
+ mod_timer(&net->ipv4.ipmr_expire_timer, jiffies+HZ/10);
return;
}
- if (mfc_unres_queue == NULL)
+ if (net->ipv4.mfc_unres_queue == NULL)
goto out;
now = jiffies;
expires = 10*HZ;
- cp = &mfc_unres_queue;
+ cp = &net->ipv4.mfc_unres_queue;
while ((c=*cp) != NULL) {
if (time_after(c->mfc_un.unres.expires, now)) {
@@ -398,8 +395,8 @@ static void ipmr_expire_process(unsigned long dummy)
ipmr_destroy_unres(c);
}
- if (mfc_unres_queue != NULL)
- mod_timer(&ipmr_expire_timer, jiffies + expires);
+ if (net->ipv4.mfc_unres_queue != NULL)
+ mod_timer(&net->ipv4.ipmr_expire_timer, jiffies + expires);
out:
spin_unlock(&mfc_unres_lock);
@@ -708,9 +705,8 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
const struct iphdr *iph = ip_hdr(skb);
spin_lock_bh(&mfc_unres_lock);
- for (c=mfc_unres_queue; c; c=c->next) {
- if (net_eq(mfc_net(c), net) &&
- c->mfc_mcastgrp == iph->daddr &&
+ for (c=net->ipv4.mfc_unres_queue; c; c=c->next) {
+ if (c->mfc_mcastgrp == iph->daddr &&
c->mfc_origin == iph->saddr)
break;
}
@@ -751,10 +747,10 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
}
atomic_inc(&net->ipv4.cache_resolve_queue_len);
- c->next = mfc_unres_queue;
- mfc_unres_queue = c;
+ c->next = net->ipv4.mfc_unres_queue;
+ net->ipv4.mfc_unres_queue = c;
- mod_timer(&ipmr_expire_timer, c->mfc_un.unres.expires);
+ mod_timer(&net->ipv4.ipmr_expire_timer, c->mfc_un.unres.expires);
}
/*
@@ -849,18 +845,17 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
* need to send on the frames and tidy up.
*/
spin_lock_bh(&mfc_unres_lock);
- for (cp = &mfc_unres_queue; (uc=*cp) != NULL;
+ for (cp = &net->ipv4.mfc_unres_queue; (uc=*cp) != NULL;
cp = &uc->next) {
- if (net_eq(mfc_net(uc), net) &&
- uc->mfc_origin == c->mfc_origin &&
+ if (uc->mfc_origin == c->mfc_origin &&
uc->mfc_mcastgrp == c->mfc_mcastgrp) {
*cp = uc->next;
atomic_dec(&net->ipv4.cache_resolve_queue_len);
break;
}
}
- if (mfc_unres_queue == NULL)
- del_timer(&ipmr_expire_timer);
+ if (net->ipv4.mfc_unres_queue == NULL)
+ del_timer(&net->ipv4.ipmr_expire_timer);
spin_unlock_bh(&mfc_unres_lock);
if (uc) {
@@ -912,14 +907,9 @@ static void mroute_clean_tables(struct net *net)
struct mfc_cache *c, **cp;
spin_lock_bh(&mfc_unres_lock);
- cp = &mfc_unres_queue;
+ cp = &net->ipv4.mfc_unres_queue;
while ((c = *cp) != NULL) {
- if (!net_eq(mfc_net(c), net)) {
- cp = &c->next;
- continue;
- }
*cp = c->next;
-
ipmr_destroy_unres(c);
}
spin_unlock_bh(&mfc_unres_lock);
@@ -1819,11 +1809,10 @@ static struct mfc_cache *ipmr_mfc_seq_idx(struct net *net,
return mfc;
read_unlock(&mrt_lock);
- it->cache = &mfc_unres_queue;
+ it->cache = &net->ipv4.mfc_unres_queue;
spin_lock_bh(&mfc_unres_lock);
- for (mfc = mfc_unres_queue; mfc; mfc = mfc->next)
- if (net_eq(mfc_net(mfc), net) &&
- pos-- == 0)
+ for (mfc = net->ipv4.mfc_unres_queue; mfc; mfc = mfc->next)
+ if (pos-- == 0)
return mfc;
spin_unlock_bh(&mfc_unres_lock);
@@ -1857,7 +1846,7 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
if (mfc->next)
return mfc->next;
- if (it->cache == &mfc_unres_queue)
+ if (it->cache == &net->ipv4.mfc_unres_queue)
goto end_of_list;
BUG_ON(it->cache != net->ipv4.mfc_cache_array);
@@ -1870,13 +1859,11 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
/* exhausted cache_array, show unresolved */
read_unlock(&mrt_lock);
- it->cache = &mfc_unres_queue;
+ it->cache = &net->ipv4.mfc_unres_queue;
it->ct = 0;
spin_lock_bh(&mfc_unres_lock);
- mfc = mfc_unres_queue;
- while (mfc && !net_eq(mfc_net(mfc), net))
- mfc = mfc->next;
+ mfc = net->ipv4.mfc_unres_queue;
if (mfc)
return mfc;
@@ -1892,7 +1879,7 @@ static void ipmr_mfc_seq_stop(struct seq_file *seq, void *v)
struct ipmr_mfc_iter *it = seq->private;
struct net *net = seq_file_net(seq);
- if (it->cache == &mfc_unres_queue)
+ if (it->cache == &net->ipv4.mfc_unres_queue)
spin_unlock_bh(&mfc_unres_lock);
else if (it->cache == net->ipv4.mfc_cache_array)
read_unlock(&mrt_lock);
@@ -1915,7 +1902,7 @@ static int ipmr_mfc_seq_show(struct seq_file *seq, void *v)
(unsigned long) mfc->mfc_origin,
mfc->mfc_parent);
- if (it->cache != &mfc_unres_queue) {
+ if (it->cache != &net->ipv4.mfc_unres_queue) {
seq_printf(seq, " %8lu %8lu %8lu",
mfc->mfc_un.res.pkt,
mfc->mfc_un.res.bytes,
@@ -1992,6 +1979,9 @@ static int __net_init ipmr_net_init(struct net *net)
goto fail_mfc_cache;
}
+ setup_timer(&net->ipv4.ipmr_expire_timer, ipmr_expire_process,
+ (unsigned long)net);
+
#ifdef CONFIG_IP_PIMSM
net->ipv4.mroute_reg_vif_num = -1;
#endif
@@ -2047,7 +2037,6 @@ int __init ip_mr_init(void)
if (err)
goto reg_pernet_fail;
- setup_timer(&ipmr_expire_timer, ipmr_expire_process, 0);
err = register_netdevice_notifier(&ip_mr_notifier);
if (err)
goto reg_notif_fail;
@@ -2065,7 +2054,6 @@ add_proto_fail:
unregister_netdevice_notifier(&ip_mr_notifier);
#endif
reg_notif_fail:
- del_timer(&ipmr_expire_timer);
unregister_pernet_subsys(&ipmr_net_ops);
reg_pernet_fail:
kmem_cache_destroy(mrt_cachep);
--
1.7.0.4
^ permalink raw reply related
* [PATCH 6/9] ipv4: ipmr: remove net pointer from struct mfc_cache
From: Patrick McHardy @ 2010-04-13 15:03 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1271171003-11901-1-git-send-email-kaber@trash.net>
Now that cache entries in unres_queue don't need to be distinguished by their
network namespace pointer anymore, we can remove it from struct mfc_cache
add pass the namespace as function argument to the functions that need it.
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
include/linux/mroute.h | 15 -----------
net/ipv4/ipmr.c | 65 +++++++++++++++++++++++------------------------
2 files changed, 32 insertions(+), 48 deletions(-)
diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index c5f3d53..de7780a 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -192,9 +192,6 @@ struct vif_device {
struct mfc_cache {
struct mfc_cache *next; /* Next entry on cache line */
-#ifdef CONFIG_NET_NS
- struct net *mfc_net;
-#endif
__be32 mfc_mcastgrp; /* Group the entry belongs to */
__be32 mfc_origin; /* Source of packet */
vifi_t mfc_parent; /* Source interface */
@@ -217,18 +214,6 @@ struct mfc_cache {
} mfc_un;
};
-static inline
-struct net *mfc_net(const struct mfc_cache *mfc)
-{
- return read_pnet(&mfc->mfc_net);
-}
-
-static inline
-void mfc_net_set(struct mfc_cache *mfc, struct net *net)
-{
- write_pnet(&mfc->mfc_net, hold_net(net));
-}
-
#define MFC_STATIC 1
#define MFC_NOTIFY 2
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index d6aa65e..f8e25c8 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -93,10 +93,12 @@ static DEFINE_SPINLOCK(mfc_unres_lock);
static struct kmem_cache *mrt_cachep __read_mostly;
-static int ip_mr_forward(struct sk_buff *skb, struct mfc_cache *cache, int local);
+static int ip_mr_forward(struct net *net, struct sk_buff *skb,
+ struct mfc_cache *cache, int local);
static int ipmr_cache_report(struct net *net,
struct sk_buff *pkt, vifi_t vifi, int assert);
-static int ipmr_fill_mroute(struct sk_buff *skb, struct mfc_cache *c, struct rtmsg *rtm);
+static int ipmr_fill_mroute(struct net *net, struct sk_buff *skb,
+ struct mfc_cache *c, struct rtmsg *rtm);
/* Service routines creating virtual interfaces: DVMRP tunnels and PIMREG */
@@ -325,7 +327,6 @@ static int vif_delete(struct net *net, int vifi, int notify,
static inline void ipmr_cache_free(struct mfc_cache *c)
{
- release_net(mfc_net(c));
kmem_cache_free(mrt_cachep, c);
}
@@ -333,11 +334,10 @@ static inline void ipmr_cache_free(struct mfc_cache *c)
and reporting error to netlink readers.
*/
-static void ipmr_destroy_unres(struct mfc_cache *c)
+static void ipmr_destroy_unres(struct net *net, struct mfc_cache *c)
{
struct sk_buff *skb;
struct nlmsgerr *e;
- struct net *net = mfc_net(c);
atomic_dec(&net->ipv4.cache_resolve_queue_len);
@@ -392,7 +392,7 @@ static void ipmr_expire_process(unsigned long arg)
*cp = c->next;
- ipmr_destroy_unres(c);
+ ipmr_destroy_unres(net, c);
}
if (net->ipv4.mfc_unres_queue != NULL)
@@ -404,10 +404,10 @@ out:
/* Fill oifs list. It is called under write locked mrt_lock. */
-static void ipmr_update_thresholds(struct mfc_cache *cache, unsigned char *ttls)
+static void ipmr_update_thresholds(struct net *net, struct mfc_cache *cache,
+ unsigned char *ttls)
{
int vifi;
- struct net *net = mfc_net(cache);
cache->mfc_un.res.minvif = MAXVIFS;
cache->mfc_un.res.maxvif = 0;
@@ -547,24 +547,22 @@ static struct mfc_cache *ipmr_cache_find(struct net *net,
/*
* Allocate a multicast cache entry
*/
-static struct mfc_cache *ipmr_cache_alloc(struct net *net)
+static struct mfc_cache *ipmr_cache_alloc(void)
{
struct mfc_cache *c = kmem_cache_zalloc(mrt_cachep, GFP_KERNEL);
if (c == NULL)
return NULL;
c->mfc_un.res.minvif = MAXVIFS;
- mfc_net_set(c, net);
return c;
}
-static struct mfc_cache *ipmr_cache_alloc_unres(struct net *net)
+static struct mfc_cache *ipmr_cache_alloc_unres(void)
{
struct mfc_cache *c = kmem_cache_zalloc(mrt_cachep, GFP_ATOMIC);
if (c == NULL)
return NULL;
skb_queue_head_init(&c->mfc_un.unres.unresolved);
c->mfc_un.unres.expires = jiffies + 10*HZ;
- mfc_net_set(c, net);
return c;
}
@@ -572,7 +570,8 @@ static struct mfc_cache *ipmr_cache_alloc_unres(struct net *net)
* A cache entry has gone into a resolved state from queued
*/
-static void ipmr_cache_resolve(struct mfc_cache *uc, struct mfc_cache *c)
+static void ipmr_cache_resolve(struct net *net, struct mfc_cache *uc,
+ struct mfc_cache *c)
{
struct sk_buff *skb;
struct nlmsgerr *e;
@@ -585,7 +584,7 @@ static void ipmr_cache_resolve(struct mfc_cache *uc, struct mfc_cache *c)
if (ip_hdr(skb)->version == 0) {
struct nlmsghdr *nlh = (struct nlmsghdr *)skb_pull(skb, sizeof(struct iphdr));
- if (ipmr_fill_mroute(skb, c, NLMSG_DATA(nlh)) > 0) {
+ if (ipmr_fill_mroute(net, skb, c, NLMSG_DATA(nlh)) > 0) {
nlh->nlmsg_len = (skb_tail_pointer(skb) -
(u8 *)nlh);
} else {
@@ -597,9 +596,9 @@ static void ipmr_cache_resolve(struct mfc_cache *uc, struct mfc_cache *c)
memset(&e->msg, 0, sizeof(e->msg));
}
- rtnl_unicast(skb, mfc_net(c), NETLINK_CB(skb).pid);
+ rtnl_unicast(skb, net, NETLINK_CB(skb).pid);
} else
- ip_mr_forward(skb, c, 0);
+ ip_mr_forward(net, skb, c, 0);
}
}
@@ -717,7 +716,7 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
*/
if (atomic_read(&net->ipv4.cache_resolve_queue_len) >= 10 ||
- (c = ipmr_cache_alloc_unres(net)) == NULL) {
+ (c = ipmr_cache_alloc_unres()) == NULL) {
spin_unlock_bh(&mfc_unres_lock);
kfree_skb(skb);
@@ -814,7 +813,7 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
if (c != NULL) {
write_lock_bh(&mrt_lock);
c->mfc_parent = mfc->mfcc_parent;
- ipmr_update_thresholds(c, mfc->mfcc_ttls);
+ ipmr_update_thresholds(net, c, mfc->mfcc_ttls);
if (!mrtsock)
c->mfc_flags |= MFC_STATIC;
write_unlock_bh(&mrt_lock);
@@ -824,14 +823,14 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
if (!ipv4_is_multicast(mfc->mfcc_mcastgrp.s_addr))
return -EINVAL;
- c = ipmr_cache_alloc(net);
+ c = ipmr_cache_alloc();
if (c == NULL)
return -ENOMEM;
c->mfc_origin = mfc->mfcc_origin.s_addr;
c->mfc_mcastgrp = mfc->mfcc_mcastgrp.s_addr;
c->mfc_parent = mfc->mfcc_parent;
- ipmr_update_thresholds(c, mfc->mfcc_ttls);
+ ipmr_update_thresholds(net, c, mfc->mfcc_ttls);
if (!mrtsock)
c->mfc_flags |= MFC_STATIC;
@@ -859,7 +858,7 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
spin_unlock_bh(&mfc_unres_lock);
if (uc) {
- ipmr_cache_resolve(uc, c);
+ ipmr_cache_resolve(net, uc, c);
ipmr_cache_free(uc);
}
return 0;
@@ -910,7 +909,7 @@ static void mroute_clean_tables(struct net *net)
cp = &net->ipv4.mfc_unres_queue;
while ((c = *cp) != NULL) {
*cp = c->next;
- ipmr_destroy_unres(c);
+ ipmr_destroy_unres(net, c);
}
spin_unlock_bh(&mfc_unres_lock);
}
@@ -1221,9 +1220,9 @@ static inline int ipmr_forward_finish(struct sk_buff *skb)
* Processing handlers for ipmr_forward
*/
-static void ipmr_queue_xmit(struct sk_buff *skb, struct mfc_cache *c, int vifi)
+static void ipmr_queue_xmit(struct net *net, struct sk_buff *skb,
+ struct mfc_cache *c, int vifi)
{
- struct net *net = mfc_net(c);
const struct iphdr *iph = ip_hdr(skb);
struct vif_device *vif = &net->ipv4.vif_table[vifi];
struct net_device *dev;
@@ -1335,11 +1334,11 @@ static int ipmr_find_vif(struct net_device *dev)
/* "local" means that we should preserve one skb (for local delivery) */
-static int ip_mr_forward(struct sk_buff *skb, struct mfc_cache *cache, int local)
+static int ip_mr_forward(struct net *net, struct sk_buff *skb,
+ struct mfc_cache *cache, int local)
{
int psend = -1;
int vif, ct;
- struct net *net = mfc_net(cache);
vif = cache->mfc_parent;
cache->mfc_un.res.pkt++;
@@ -1396,7 +1395,7 @@ static int ip_mr_forward(struct sk_buff *skb, struct mfc_cache *cache, int local
if (psend != -1) {
struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
if (skb2)
- ipmr_queue_xmit(skb2, cache, psend);
+ ipmr_queue_xmit(net, skb2, cache, psend);
}
psend = ct;
}
@@ -1405,9 +1404,9 @@ static int ip_mr_forward(struct sk_buff *skb, struct mfc_cache *cache, int local
if (local) {
struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
if (skb2)
- ipmr_queue_xmit(skb2, cache, psend);
+ ipmr_queue_xmit(net, skb2, cache, psend);
} else {
- ipmr_queue_xmit(skb, cache, psend);
+ ipmr_queue_xmit(net, skb, cache, psend);
return 0;
}
}
@@ -1488,7 +1487,7 @@ int ip_mr_input(struct sk_buff *skb)
return -ENODEV;
}
- ip_mr_forward(skb, cache, local);
+ ip_mr_forward(net, skb, cache, local);
read_unlock(&mrt_lock);
@@ -1602,11 +1601,11 @@ drop:
#endif
static int
-ipmr_fill_mroute(struct sk_buff *skb, struct mfc_cache *c, struct rtmsg *rtm)
+ipmr_fill_mroute(struct net *net, struct sk_buff *skb, struct mfc_cache *c,
+ struct rtmsg *rtm)
{
int ct;
struct rtnexthop *nhp;
- struct net *net = mfc_net(c);
u8 *b = skb_tail_pointer(skb);
struct rtattr *mp_head;
@@ -1686,7 +1685,7 @@ int ipmr_get_route(struct net *net,
if (!nowait && (rtm->rtm_flags&RTM_F_NOTIFY))
cache->mfc_flags |= MFC_NOTIFY;
- err = ipmr_fill_mroute(skb, cache, rtm);
+ err = ipmr_fill_mroute(net, skb, cache, rtm);
read_unlock(&mrt_lock);
return err;
}
--
1.7.0.4
^ permalink raw reply related
* [PATCH 7/9] ipv4: ipmr: convert struct mfc_cache to struct list_head
From: Patrick McHardy @ 2010-04-13 15:03 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1271171003-11901-1-git-send-email-kaber@trash.net>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
include/linux/mroute.h | 2 +-
include/net/netns/ipv4.h | 4 +-
net/ipv4/ipmr.c | 125 ++++++++++++++++++++++-----------------------
3 files changed, 64 insertions(+), 67 deletions(-)
diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index de7780a..7ff6c77 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -191,7 +191,7 @@ struct vif_device {
#define VIFF_STATIC 0x8000
struct mfc_cache {
- struct mfc_cache *next; /* Next entry on cache line */
+ struct list_head list;
__be32 mfc_mcastgrp; /* Group the entry belongs to */
__be32 mfc_origin; /* Source of packet */
vifi_t mfc_parent; /* Source interface */
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
index b15e518..5d06429 100644
--- a/include/net/netns/ipv4.h
+++ b/include/net/netns/ipv4.h
@@ -61,8 +61,8 @@ struct netns_ipv4 {
#ifdef CONFIG_IP_MROUTE
struct sock *mroute_sk;
struct timer_list ipmr_expire_timer;
- struct mfc_cache *mfc_unres_queue;
- struct mfc_cache **mfc_cache_array;
+ struct list_head mfc_unres_queue;
+ struct list_head *mfc_cache_array;
struct vif_device *vif_table;
int maxvif;
atomic_t cache_resolve_queue_len;
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index f8e25c8..21b5edc 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -367,35 +367,32 @@ static void ipmr_expire_process(unsigned long arg)
struct net *net = (struct net *)arg;
unsigned long now;
unsigned long expires;
- struct mfc_cache *c, **cp;
+ struct mfc_cache *c, *next;
if (!spin_trylock(&mfc_unres_lock)) {
mod_timer(&net->ipv4.ipmr_expire_timer, jiffies+HZ/10);
return;
}
- if (net->ipv4.mfc_unres_queue == NULL)
+ if (list_empty(&net->ipv4.mfc_unres_queue))
goto out;
now = jiffies;
expires = 10*HZ;
- cp = &net->ipv4.mfc_unres_queue;
- while ((c=*cp) != NULL) {
+ list_for_each_entry_safe(c, next, &net->ipv4.mfc_unres_queue, list) {
if (time_after(c->mfc_un.unres.expires, now)) {
unsigned long interval = c->mfc_un.unres.expires - now;
if (interval < expires)
expires = interval;
- cp = &c->next;
continue;
}
- *cp = c->next;
-
+ list_del(&c->list);
ipmr_destroy_unres(net, c);
}
- if (net->ipv4.mfc_unres_queue != NULL)
+ if (!list_empty(&net->ipv4.mfc_unres_queue))
mod_timer(&net->ipv4.ipmr_expire_timer, jiffies + expires);
out:
@@ -537,11 +534,11 @@ static struct mfc_cache *ipmr_cache_find(struct net *net,
int line = MFC_HASH(mcastgrp, origin);
struct mfc_cache *c;
- for (c = net->ipv4.mfc_cache_array[line]; c; c = c->next) {
- if (c->mfc_origin==origin && c->mfc_mcastgrp==mcastgrp)
- break;
+ list_for_each_entry(c, &net->ipv4.mfc_cache_array[line], list) {
+ if (c->mfc_origin == origin && c->mfc_mcastgrp == mcastgrp)
+ return c;
}
- return c;
+ return NULL;
}
/*
@@ -699,18 +696,21 @@ static int ipmr_cache_report(struct net *net,
static int
ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
{
+ bool found = false;
int err;
struct mfc_cache *c;
const struct iphdr *iph = ip_hdr(skb);
spin_lock_bh(&mfc_unres_lock);
- for (c=net->ipv4.mfc_unres_queue; c; c=c->next) {
+ list_for_each_entry(c, &net->ipv4.mfc_unres_queue, list) {
if (c->mfc_mcastgrp == iph->daddr &&
- c->mfc_origin == iph->saddr)
+ c->mfc_origin == iph->saddr) {
+ found = true;
break;
+ }
}
- if (c == NULL) {
+ if (!found) {
/*
* Create a new entry if allowable
*/
@@ -746,8 +746,7 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
}
atomic_inc(&net->ipv4.cache_resolve_queue_len);
- c->next = net->ipv4.mfc_unres_queue;
- net->ipv4.mfc_unres_queue = c;
+ list_add(&c->list, &net->ipv4.mfc_unres_queue);
mod_timer(&net->ipv4.ipmr_expire_timer, c->mfc_un.unres.expires);
}
@@ -774,16 +773,15 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
static int ipmr_mfc_delete(struct net *net, struct mfcctl *mfc)
{
int line;
- struct mfc_cache *c, **cp;
+ struct mfc_cache *c, *next;
line = MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);
- for (cp = &net->ipv4.mfc_cache_array[line];
- (c = *cp) != NULL; cp = &c->next) {
+ list_for_each_entry_safe(c, next, &net->ipv4.mfc_cache_array[line], list) {
if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr) {
write_lock_bh(&mrt_lock);
- *cp = c->next;
+ list_del(&c->list);
write_unlock_bh(&mrt_lock);
ipmr_cache_free(c);
@@ -795,22 +793,24 @@ static int ipmr_mfc_delete(struct net *net, struct mfcctl *mfc)
static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
{
+ bool found = false;
int line;
- struct mfc_cache *uc, *c, **cp;
+ struct mfc_cache *uc, *c;
if (mfc->mfcc_parent >= MAXVIFS)
return -ENFILE;
line = MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);
- for (cp = &net->ipv4.mfc_cache_array[line];
- (c = *cp) != NULL; cp = &c->next) {
+ list_for_each_entry(c, &net->ipv4.mfc_cache_array[line], list) {
if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
- c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr)
+ c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr) {
+ found = true;
break;
+ }
}
- if (c != NULL) {
+ if (found) {
write_lock_bh(&mrt_lock);
c->mfc_parent = mfc->mfcc_parent;
ipmr_update_thresholds(net, c, mfc->mfcc_ttls);
@@ -835,8 +835,7 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
c->mfc_flags |= MFC_STATIC;
write_lock_bh(&mrt_lock);
- c->next = net->ipv4.mfc_cache_array[line];
- net->ipv4.mfc_cache_array[line] = c;
+ list_add(&c->list, &net->ipv4.mfc_cache_array[line]);
write_unlock_bh(&mrt_lock);
/*
@@ -844,16 +843,15 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
* need to send on the frames and tidy up.
*/
spin_lock_bh(&mfc_unres_lock);
- for (cp = &net->ipv4.mfc_unres_queue; (uc=*cp) != NULL;
- cp = &uc->next) {
+ list_for_each_entry(uc, &net->ipv4.mfc_unres_queue, list) {
if (uc->mfc_origin == c->mfc_origin &&
uc->mfc_mcastgrp == c->mfc_mcastgrp) {
- *cp = uc->next;
+ list_del(&uc->list);
atomic_dec(&net->ipv4.cache_resolve_queue_len);
break;
}
}
- if (net->ipv4.mfc_unres_queue == NULL)
+ if (list_empty(&net->ipv4.mfc_unres_queue))
del_timer(&net->ipv4.ipmr_expire_timer);
spin_unlock_bh(&mfc_unres_lock);
@@ -872,6 +870,7 @@ static void mroute_clean_tables(struct net *net)
{
int i;
LIST_HEAD(list);
+ struct mfc_cache *c, *next;
/*
* Shut down all active vif entries
@@ -885,17 +884,12 @@ static void mroute_clean_tables(struct net *net)
/*
* Wipe the cache
*/
- for (i=0; i<MFC_LINES; i++) {
- struct mfc_cache *c, **cp;
-
- cp = &net->ipv4.mfc_cache_array[i];
- while ((c = *cp) != NULL) {
- if (c->mfc_flags&MFC_STATIC) {
- cp = &c->next;
+ for (i = 0; i < MFC_LINES; i++) {
+ list_for_each_entry_safe(c, next, &net->ipv4.mfc_cache_array[i], list) {
+ if (c->mfc_flags&MFC_STATIC)
continue;
- }
write_lock_bh(&mrt_lock);
- *cp = c->next;
+ list_del(&c->list);
write_unlock_bh(&mrt_lock);
ipmr_cache_free(c);
@@ -903,12 +897,9 @@ static void mroute_clean_tables(struct net *net)
}
if (atomic_read(&net->ipv4.cache_resolve_queue_len) != 0) {
- struct mfc_cache *c, **cp;
-
spin_lock_bh(&mfc_unres_lock);
- cp = &net->ipv4.mfc_unres_queue;
- while ((c = *cp) != NULL) {
- *cp = c->next;
+ list_for_each_entry_safe(c, next, &net->ipv4.mfc_unres_queue, list) {
+ list_del(&c->list);
ipmr_destroy_unres(net, c);
}
spin_unlock_bh(&mfc_unres_lock);
@@ -1789,7 +1780,7 @@ static const struct file_operations ipmr_vif_fops = {
struct ipmr_mfc_iter {
struct seq_net_private p;
- struct mfc_cache **cache;
+ struct list_head *cache;
int ct;
};
@@ -1799,18 +1790,18 @@ static struct mfc_cache *ipmr_mfc_seq_idx(struct net *net,
{
struct mfc_cache *mfc;
- it->cache = net->ipv4.mfc_cache_array;
read_lock(&mrt_lock);
- for (it->ct = 0; it->ct < MFC_LINES; it->ct++)
- for (mfc = net->ipv4.mfc_cache_array[it->ct];
- mfc; mfc = mfc->next)
+ for (it->ct = 0; it->ct < MFC_LINES; it->ct++) {
+ it->cache = &net->ipv4.mfc_cache_array[it->ct];
+ list_for_each_entry(mfc, it->cache, list)
if (pos-- == 0)
return mfc;
+ }
read_unlock(&mrt_lock);
- it->cache = &net->ipv4.mfc_unres_queue;
spin_lock_bh(&mfc_unres_lock);
- for (mfc = net->ipv4.mfc_unres_queue; mfc; mfc = mfc->next)
+ it->cache = &net->ipv4.mfc_unres_queue;
+ list_for_each_entry(mfc, it->cache, list)
if (pos-- == 0)
return mfc;
spin_unlock_bh(&mfc_unres_lock);
@@ -1842,18 +1833,19 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
if (v == SEQ_START_TOKEN)
return ipmr_mfc_seq_idx(net, seq->private, 0);
- if (mfc->next)
- return mfc->next;
+ if (mfc->list.next != it->cache)
+ return list_entry(mfc->list.next, struct mfc_cache, list);
if (it->cache == &net->ipv4.mfc_unres_queue)
goto end_of_list;
- BUG_ON(it->cache != net->ipv4.mfc_cache_array);
+ BUG_ON(it->cache != &net->ipv4.mfc_cache_array[it->ct]);
while (++it->ct < MFC_LINES) {
- mfc = net->ipv4.mfc_cache_array[it->ct];
- if (mfc)
- return mfc;
+ it->cache = &net->ipv4.mfc_cache_array[it->ct];
+ if (list_empty(it->cache))
+ continue;
+ return list_first_entry(it->cache, struct mfc_cache, list);
}
/* exhausted cache_array, show unresolved */
@@ -1862,9 +1854,8 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
it->ct = 0;
spin_lock_bh(&mfc_unres_lock);
- mfc = net->ipv4.mfc_unres_queue;
- if (mfc)
- return mfc;
+ if (!list_empty(it->cache))
+ return list_first_entry(it->cache, struct mfc_cache, list);
end_of_list:
spin_unlock_bh(&mfc_unres_lock);
@@ -1880,7 +1871,7 @@ static void ipmr_mfc_seq_stop(struct seq_file *seq, void *v)
if (it->cache == &net->ipv4.mfc_unres_queue)
spin_unlock_bh(&mfc_unres_lock);
- else if (it->cache == net->ipv4.mfc_cache_array)
+ else if (it->cache == &net->ipv4.mfc_cache_array[it->ct])
read_unlock(&mrt_lock);
}
@@ -1960,6 +1951,7 @@ static const struct net_protocol pim_protocol = {
*/
static int __net_init ipmr_net_init(struct net *net)
{
+ unsigned int i;
int err = 0;
net->ipv4.vif_table = kcalloc(MAXVIFS, sizeof(struct vif_device),
@@ -1971,13 +1963,18 @@ static int __net_init ipmr_net_init(struct net *net)
/* Forwarding cache */
net->ipv4.mfc_cache_array = kcalloc(MFC_LINES,
- sizeof(struct mfc_cache *),
+ sizeof(struct list_head),
GFP_KERNEL);
if (!net->ipv4.mfc_cache_array) {
err = -ENOMEM;
goto fail_mfc_cache;
}
+ for (i = 0; i < MFC_LINES; i++)
+ INIT_LIST_HEAD(&net->ipv4.mfc_cache_array[i]);
+
+ INIT_LIST_HEAD(&net->ipv4.mfc_unres_queue);
+
setup_timer(&net->ipv4.ipmr_expire_timer, ipmr_expire_process,
(unsigned long)net);
--
1.7.0.4
^ permalink raw reply related
* [PATCH 8/9] ipv4: ipmr: move mroute data into seperate structure
From: Patrick McHardy @ 2010-04-13 15:03 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1271171003-11901-1-git-send-email-kaber@trash.net>
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
include/net/netns/ipv4.h | 13 +--
net/ipv4/ipmr.c | 369 +++++++++++++++++++++++++---------------------
2 files changed, 200 insertions(+), 182 deletions(-)
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
index 5d06429..72e762a 100644
--- a/include/net/netns/ipv4.h
+++ b/include/net/netns/ipv4.h
@@ -59,18 +59,7 @@ struct netns_ipv4 {
atomic_t rt_genid;
#ifdef CONFIG_IP_MROUTE
- struct sock *mroute_sk;
- struct timer_list ipmr_expire_timer;
- struct list_head mfc_unres_queue;
- struct list_head *mfc_cache_array;
- struct vif_device *vif_table;
- int maxvif;
- atomic_t cache_resolve_queue_len;
- int mroute_do_assert;
- int mroute_do_pim;
-#if defined(CONFIG_IP_PIMSM_V1) || defined(CONFIG_IP_PIMSM_V2)
- int mroute_reg_vif_num;
-#endif
+ struct mr_table *mrt;
#endif
};
#endif
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 21b5edc..498f4e9 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -68,6 +68,21 @@
#define CONFIG_IP_PIMSM 1
#endif
+struct mr_table {
+ struct sock *mroute_sk;
+ struct timer_list ipmr_expire_timer;
+ struct list_head mfc_unres_queue;
+ struct list_head mfc_cache_array[MFC_LINES];
+ struct vif_device vif_table[MAXVIFS];
+ int maxvif;
+ atomic_t cache_resolve_queue_len;
+ int mroute_do_assert;
+ int mroute_do_pim;
+#if defined(CONFIG_IP_PIMSM_V1) || defined(CONFIG_IP_PIMSM_V2)
+ int mroute_reg_vif_num;
+#endif
+};
+
/* Big lock, protecting vif table, mrt cache and mroute socket state.
Note that the changes are semaphored via rtnl_lock.
*/
@@ -78,7 +93,7 @@ static DEFINE_RWLOCK(mrt_lock);
* Multicast router control variables
*/
-#define VIF_EXISTS(_net, _idx) ((_net)->ipv4.vif_table[_idx].dev != NULL)
+#define VIF_EXISTS(_mrt, _idx) ((_mrt)->vif_table[_idx].dev != NULL)
/* Special spinlock for queue of unresolved entries */
static DEFINE_SPINLOCK(mfc_unres_lock);
@@ -93,11 +108,12 @@ static DEFINE_SPINLOCK(mfc_unres_lock);
static struct kmem_cache *mrt_cachep __read_mostly;
-static int ip_mr_forward(struct net *net, struct sk_buff *skb,
- struct mfc_cache *cache, int local);
-static int ipmr_cache_report(struct net *net,
+static int ip_mr_forward(struct net *net, struct mr_table *mrt,
+ struct sk_buff *skb, struct mfc_cache *cache,
+ int local);
+static int ipmr_cache_report(struct mr_table *mrt,
struct sk_buff *pkt, vifi_t vifi, int assert);
-static int ipmr_fill_mroute(struct net *net, struct sk_buff *skb,
+static int ipmr_fill_mroute(struct mr_table *mrt, struct sk_buff *skb,
struct mfc_cache *c, struct rtmsg *rtm);
/* Service routines creating virtual interfaces: DVMRP tunnels and PIMREG */
@@ -199,12 +215,12 @@ failure:
static netdev_tx_t reg_vif_xmit(struct sk_buff *skb, struct net_device *dev)
{
struct net *net = dev_net(dev);
+ struct mr_table *mrt = net->ipv4.mrt;
read_lock(&mrt_lock);
dev->stats.tx_bytes += skb->len;
dev->stats.tx_packets++;
- ipmr_cache_report(net, skb, net->ipv4.mroute_reg_vif_num,
- IGMPMSG_WHOLEPKT);
+ ipmr_cache_report(mrt, skb, mrt->mroute_reg_vif_num, IGMPMSG_WHOLEPKT);
read_unlock(&mrt_lock);
kfree_skb(skb);
return NETDEV_TX_OK;
@@ -274,17 +290,17 @@ failure:
* @notify: Set to 1, if the caller is a notifier_call
*/
-static int vif_delete(struct net *net, int vifi, int notify,
+static int vif_delete(struct mr_table *mrt, int vifi, int notify,
struct list_head *head)
{
struct vif_device *v;
struct net_device *dev;
struct in_device *in_dev;
- if (vifi < 0 || vifi >= net->ipv4.maxvif)
+ if (vifi < 0 || vifi >= mrt->maxvif)
return -EADDRNOTAVAIL;
- v = &net->ipv4.vif_table[vifi];
+ v = &mrt->vif_table[vifi];
write_lock_bh(&mrt_lock);
dev = v->dev;
@@ -296,17 +312,17 @@ static int vif_delete(struct net *net, int vifi, int notify,
}
#ifdef CONFIG_IP_PIMSM
- if (vifi == net->ipv4.mroute_reg_vif_num)
- net->ipv4.mroute_reg_vif_num = -1;
+ if (vifi == mrt->mroute_reg_vif_num)
+ mrt->mroute_reg_vif_num = -1;
#endif
- if (vifi+1 == net->ipv4.maxvif) {
+ if (vifi+1 == mrt->maxvif) {
int tmp;
for (tmp=vifi-1; tmp>=0; tmp--) {
- if (VIF_EXISTS(net, tmp))
+ if (VIF_EXISTS(mrt, tmp))
break;
}
- net->ipv4.maxvif = tmp+1;
+ mrt->maxvif = tmp+1;
}
write_unlock_bh(&mrt_lock);
@@ -334,12 +350,13 @@ static inline void ipmr_cache_free(struct mfc_cache *c)
and reporting error to netlink readers.
*/
-static void ipmr_destroy_unres(struct net *net, struct mfc_cache *c)
+static void ipmr_destroy_unres(struct mr_table *mrt, struct mfc_cache *c)
{
+ struct net *net = NULL; //mrt->net;
struct sk_buff *skb;
struct nlmsgerr *e;
- atomic_dec(&net->ipv4.cache_resolve_queue_len);
+ atomic_dec(&mrt->cache_resolve_queue_len);
while ((skb = skb_dequeue(&c->mfc_un.unres.unresolved))) {
if (ip_hdr(skb)->version == 0) {
@@ -364,23 +381,23 @@ static void ipmr_destroy_unres(struct net *net, struct mfc_cache *c)
static void ipmr_expire_process(unsigned long arg)
{
- struct net *net = (struct net *)arg;
+ struct mr_table *mrt = (struct mr_table *)arg;
unsigned long now;
unsigned long expires;
struct mfc_cache *c, *next;
if (!spin_trylock(&mfc_unres_lock)) {
- mod_timer(&net->ipv4.ipmr_expire_timer, jiffies+HZ/10);
+ mod_timer(&mrt->ipmr_expire_timer, jiffies+HZ/10);
return;
}
- if (list_empty(&net->ipv4.mfc_unres_queue))
+ if (list_empty(&mrt->mfc_unres_queue))
goto out;
now = jiffies;
expires = 10*HZ;
- list_for_each_entry_safe(c, next, &net->ipv4.mfc_unres_queue, list) {
+ list_for_each_entry_safe(c, next, &mrt->mfc_unres_queue, list) {
if (time_after(c->mfc_un.unres.expires, now)) {
unsigned long interval = c->mfc_un.unres.expires - now;
if (interval < expires)
@@ -389,11 +406,11 @@ static void ipmr_expire_process(unsigned long arg)
}
list_del(&c->list);
- ipmr_destroy_unres(net, c);
+ ipmr_destroy_unres(mrt, c);
}
- if (!list_empty(&net->ipv4.mfc_unres_queue))
- mod_timer(&net->ipv4.ipmr_expire_timer, jiffies + expires);
+ if (!list_empty(&mrt->mfc_unres_queue))
+ mod_timer(&mrt->ipmr_expire_timer, jiffies + expires);
out:
spin_unlock(&mfc_unres_lock);
@@ -401,7 +418,7 @@ out:
/* Fill oifs list. It is called under write locked mrt_lock. */
-static void ipmr_update_thresholds(struct net *net, struct mfc_cache *cache,
+static void ipmr_update_thresholds(struct mr_table *mrt, struct mfc_cache *cache,
unsigned char *ttls)
{
int vifi;
@@ -410,8 +427,8 @@ static void ipmr_update_thresholds(struct net *net, struct mfc_cache *cache,
cache->mfc_un.res.maxvif = 0;
memset(cache->mfc_un.res.ttls, 255, MAXVIFS);
- for (vifi = 0; vifi < net->ipv4.maxvif; vifi++) {
- if (VIF_EXISTS(net, vifi) &&
+ for (vifi = 0; vifi < mrt->maxvif; vifi++) {
+ if (VIF_EXISTS(mrt, vifi) &&
ttls[vifi] && ttls[vifi] < 255) {
cache->mfc_un.res.ttls[vifi] = ttls[vifi];
if (cache->mfc_un.res.minvif > vifi)
@@ -422,16 +439,17 @@ static void ipmr_update_thresholds(struct net *net, struct mfc_cache *cache,
}
}
-static int vif_add(struct net *net, struct vifctl *vifc, int mrtsock)
+static int vif_add(struct net *net, struct mr_table *mrt,
+ struct vifctl *vifc, int mrtsock)
{
int vifi = vifc->vifc_vifi;
- struct vif_device *v = &net->ipv4.vif_table[vifi];
+ struct vif_device *v = &mrt->vif_table[vifi];
struct net_device *dev;
struct in_device *in_dev;
int err;
/* Is vif busy ? */
- if (VIF_EXISTS(net, vifi))
+ if (VIF_EXISTS(mrt, vifi))
return -EADDRINUSE;
switch (vifc->vifc_flags) {
@@ -441,7 +459,7 @@ static int vif_add(struct net *net, struct vifctl *vifc, int mrtsock)
* Special Purpose VIF in PIM
* All the packets will be sent to the daemon
*/
- if (net->ipv4.mroute_reg_vif_num >= 0)
+ if (mrt->mroute_reg_vif_num >= 0)
return -EADDRINUSE;
dev = ipmr_reg_vif(net);
if (!dev)
@@ -519,22 +537,22 @@ static int vif_add(struct net *net, struct vifctl *vifc, int mrtsock)
v->dev = dev;
#ifdef CONFIG_IP_PIMSM
if (v->flags&VIFF_REGISTER)
- net->ipv4.mroute_reg_vif_num = vifi;
+ mrt->mroute_reg_vif_num = vifi;
#endif
- if (vifi+1 > net->ipv4.maxvif)
- net->ipv4.maxvif = vifi+1;
+ if (vifi+1 > mrt->maxvif)
+ mrt->maxvif = vifi+1;
write_unlock_bh(&mrt_lock);
return 0;
}
-static struct mfc_cache *ipmr_cache_find(struct net *net,
+static struct mfc_cache *ipmr_cache_find(struct mr_table *mrt,
__be32 origin,
__be32 mcastgrp)
{
int line = MFC_HASH(mcastgrp, origin);
struct mfc_cache *c;
- list_for_each_entry(c, &net->ipv4.mfc_cache_array[line], list) {
+ list_for_each_entry(c, &mrt->mfc_cache_array[line], list) {
if (c->mfc_origin == origin && c->mfc_mcastgrp == mcastgrp)
return c;
}
@@ -567,8 +585,8 @@ static struct mfc_cache *ipmr_cache_alloc_unres(void)
* A cache entry has gone into a resolved state from queued
*/
-static void ipmr_cache_resolve(struct net *net, struct mfc_cache *uc,
- struct mfc_cache *c)
+static void ipmr_cache_resolve(struct net *net, struct mr_table *mrt,
+ struct mfc_cache *uc, struct mfc_cache *c)
{
struct sk_buff *skb;
struct nlmsgerr *e;
@@ -581,7 +599,7 @@ static void ipmr_cache_resolve(struct net *net, struct mfc_cache *uc,
if (ip_hdr(skb)->version == 0) {
struct nlmsghdr *nlh = (struct nlmsghdr *)skb_pull(skb, sizeof(struct iphdr));
- if (ipmr_fill_mroute(net, skb, c, NLMSG_DATA(nlh)) > 0) {
+ if (ipmr_fill_mroute(mrt, skb, c, NLMSG_DATA(nlh)) > 0) {
nlh->nlmsg_len = (skb_tail_pointer(skb) -
(u8 *)nlh);
} else {
@@ -595,7 +613,7 @@ static void ipmr_cache_resolve(struct net *net, struct mfc_cache *uc,
rtnl_unicast(skb, net, NETLINK_CB(skb).pid);
} else
- ip_mr_forward(net, skb, c, 0);
+ ip_mr_forward(net, mrt, skb, c, 0);
}
}
@@ -606,7 +624,7 @@ static void ipmr_cache_resolve(struct net *net, struct mfc_cache *uc,
* Called under mrt_lock.
*/
-static int ipmr_cache_report(struct net *net,
+static int ipmr_cache_report(struct mr_table *mrt,
struct sk_buff *pkt, vifi_t vifi, int assert)
{
struct sk_buff *skb;
@@ -639,7 +657,7 @@ static int ipmr_cache_report(struct net *net,
memcpy(msg, skb_network_header(pkt), sizeof(struct iphdr));
msg->im_msgtype = IGMPMSG_WHOLEPKT;
msg->im_mbz = 0;
- msg->im_vif = net->ipv4.mroute_reg_vif_num;
+ msg->im_vif = mrt->mroute_reg_vif_num;
ip_hdr(skb)->ihl = sizeof(struct iphdr) >> 2;
ip_hdr(skb)->tot_len = htons(ntohs(ip_hdr(pkt)->tot_len) +
sizeof(struct iphdr));
@@ -671,7 +689,7 @@ static int ipmr_cache_report(struct net *net,
skb->transport_header = skb->network_header;
}
- if (net->ipv4.mroute_sk == NULL) {
+ if (mrt->mroute_sk == NULL) {
kfree_skb(skb);
return -EINVAL;
}
@@ -679,7 +697,7 @@ static int ipmr_cache_report(struct net *net,
/*
* Deliver to mrouted
*/
- ret = sock_queue_rcv_skb(net->ipv4.mroute_sk, skb);
+ ret = sock_queue_rcv_skb(mrt->mroute_sk, skb);
if (ret < 0) {
if (net_ratelimit())
printk(KERN_WARNING "mroute: pending queue full, dropping entries.\n");
@@ -694,7 +712,7 @@ static int ipmr_cache_report(struct net *net,
*/
static int
-ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
+ipmr_cache_unresolved(struct mr_table *mrt, vifi_t vifi, struct sk_buff *skb)
{
bool found = false;
int err;
@@ -702,7 +720,7 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
const struct iphdr *iph = ip_hdr(skb);
spin_lock_bh(&mfc_unres_lock);
- list_for_each_entry(c, &net->ipv4.mfc_unres_queue, list) {
+ list_for_each_entry(c, &mrt->mfc_unres_queue, list) {
if (c->mfc_mcastgrp == iph->daddr &&
c->mfc_origin == iph->saddr) {
found = true;
@@ -715,7 +733,7 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
* Create a new entry if allowable
*/
- if (atomic_read(&net->ipv4.cache_resolve_queue_len) >= 10 ||
+ if (atomic_read(&mrt->cache_resolve_queue_len) >= 10 ||
(c = ipmr_cache_alloc_unres()) == NULL) {
spin_unlock_bh(&mfc_unres_lock);
@@ -733,7 +751,7 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
/*
* Reflect first query at mrouted.
*/
- err = ipmr_cache_report(net, skb, vifi, IGMPMSG_NOCACHE);
+ err = ipmr_cache_report(mrt, skb, vifi, IGMPMSG_NOCACHE);
if (err < 0) {
/* If the report failed throw the cache entry
out - Brad Parker
@@ -745,10 +763,10 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
return err;
}
- atomic_inc(&net->ipv4.cache_resolve_queue_len);
- list_add(&c->list, &net->ipv4.mfc_unres_queue);
+ atomic_inc(&mrt->cache_resolve_queue_len);
+ list_add(&c->list, &mrt->mfc_unres_queue);
- mod_timer(&net->ipv4.ipmr_expire_timer, c->mfc_un.unres.expires);
+ mod_timer(&mrt->ipmr_expire_timer, c->mfc_un.unres.expires);
}
/*
@@ -770,14 +788,14 @@ ipmr_cache_unresolved(struct net *net, vifi_t vifi, struct sk_buff *skb)
* MFC cache manipulation by user space mroute daemon
*/
-static int ipmr_mfc_delete(struct net *net, struct mfcctl *mfc)
+static int ipmr_mfc_delete(struct mr_table *mrt, struct mfcctl *mfc)
{
int line;
struct mfc_cache *c, *next;
line = MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);
- list_for_each_entry_safe(c, next, &net->ipv4.mfc_cache_array[line], list) {
+ list_for_each_entry_safe(c, next, &mrt->mfc_cache_array[line], list) {
if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr) {
write_lock_bh(&mrt_lock);
@@ -791,7 +809,8 @@ static int ipmr_mfc_delete(struct net *net, struct mfcctl *mfc)
return -ENOENT;
}
-static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
+static int ipmr_mfc_add(struct net *net, struct mr_table *mrt,
+ struct mfcctl *mfc, int mrtsock)
{
bool found = false;
int line;
@@ -802,7 +821,7 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
line = MFC_HASH(mfc->mfcc_mcastgrp.s_addr, mfc->mfcc_origin.s_addr);
- list_for_each_entry(c, &net->ipv4.mfc_cache_array[line], list) {
+ list_for_each_entry(c, &mrt->mfc_cache_array[line], list) {
if (c->mfc_origin == mfc->mfcc_origin.s_addr &&
c->mfc_mcastgrp == mfc->mfcc_mcastgrp.s_addr) {
found = true;
@@ -813,7 +832,7 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
if (found) {
write_lock_bh(&mrt_lock);
c->mfc_parent = mfc->mfcc_parent;
- ipmr_update_thresholds(net, c, mfc->mfcc_ttls);
+ ipmr_update_thresholds(mrt, c, mfc->mfcc_ttls);
if (!mrtsock)
c->mfc_flags |= MFC_STATIC;
write_unlock_bh(&mrt_lock);
@@ -830,12 +849,12 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
c->mfc_origin = mfc->mfcc_origin.s_addr;
c->mfc_mcastgrp = mfc->mfcc_mcastgrp.s_addr;
c->mfc_parent = mfc->mfcc_parent;
- ipmr_update_thresholds(net, c, mfc->mfcc_ttls);
+ ipmr_update_thresholds(mrt, c, mfc->mfcc_ttls);
if (!mrtsock)
c->mfc_flags |= MFC_STATIC;
write_lock_bh(&mrt_lock);
- list_add(&c->list, &net->ipv4.mfc_cache_array[line]);
+ list_add(&c->list, &mrt->mfc_cache_array[line]);
write_unlock_bh(&mrt_lock);
/*
@@ -843,20 +862,20 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
* need to send on the frames and tidy up.
*/
spin_lock_bh(&mfc_unres_lock);
- list_for_each_entry(uc, &net->ipv4.mfc_unres_queue, list) {
+ list_for_each_entry(uc, &mrt->mfc_unres_queue, list) {
if (uc->mfc_origin == c->mfc_origin &&
uc->mfc_mcastgrp == c->mfc_mcastgrp) {
list_del(&uc->list);
- atomic_dec(&net->ipv4.cache_resolve_queue_len);
+ atomic_dec(&mrt->cache_resolve_queue_len);
break;
}
}
- if (list_empty(&net->ipv4.mfc_unres_queue))
- del_timer(&net->ipv4.ipmr_expire_timer);
+ if (list_empty(&mrt->mfc_unres_queue))
+ del_timer(&mrt->ipmr_expire_timer);
spin_unlock_bh(&mfc_unres_lock);
if (uc) {
- ipmr_cache_resolve(net, uc, c);
+ ipmr_cache_resolve(net, mrt, uc, c);
ipmr_cache_free(uc);
}
return 0;
@@ -866,7 +885,7 @@ static int ipmr_mfc_add(struct net *net, struct mfcctl *mfc, int mrtsock)
* Close the multicast socket, and clear the vif tables etc
*/
-static void mroute_clean_tables(struct net *net)
+static void mroute_clean_tables(struct mr_table *mrt)
{
int i;
LIST_HEAD(list);
@@ -875,9 +894,9 @@ static void mroute_clean_tables(struct net *net)
/*
* Shut down all active vif entries
*/
- for (i = 0; i < net->ipv4.maxvif; i++) {
- if (!(net->ipv4.vif_table[i].flags&VIFF_STATIC))
- vif_delete(net, i, 0, &list);
+ for (i = 0; i < mrt->maxvif; i++) {
+ if (!(mrt->vif_table[i].flags&VIFF_STATIC))
+ vif_delete(mrt, i, 0, &list);
}
unregister_netdevice_many(&list);
@@ -885,7 +904,7 @@ static void mroute_clean_tables(struct net *net)
* Wipe the cache
*/
for (i = 0; i < MFC_LINES; i++) {
- list_for_each_entry_safe(c, next, &net->ipv4.mfc_cache_array[i], list) {
+ list_for_each_entry_safe(c, next, &mrt->mfc_cache_array[i], list) {
if (c->mfc_flags&MFC_STATIC)
continue;
write_lock_bh(&mrt_lock);
@@ -896,11 +915,11 @@ static void mroute_clean_tables(struct net *net)
}
}
- if (atomic_read(&net->ipv4.cache_resolve_queue_len) != 0) {
+ if (atomic_read(&mrt->cache_resolve_queue_len) != 0) {
spin_lock_bh(&mfc_unres_lock);
- list_for_each_entry_safe(c, next, &net->ipv4.mfc_unres_queue, list) {
+ list_for_each_entry_safe(c, next, &mrt->mfc_unres_queue, list) {
list_del(&c->list);
- ipmr_destroy_unres(net, c);
+ ipmr_destroy_unres(mrt, c);
}
spin_unlock_bh(&mfc_unres_lock);
}
@@ -909,16 +928,17 @@ static void mroute_clean_tables(struct net *net)
static void mrtsock_destruct(struct sock *sk)
{
struct net *net = sock_net(sk);
+ struct mr_table *mrt = net->ipv4.mrt;
rtnl_lock();
- if (sk == net->ipv4.mroute_sk) {
+ if (sk == mrt->mroute_sk) {
IPV4_DEVCONF_ALL(net, MC_FORWARDING)--;
write_lock_bh(&mrt_lock);
- net->ipv4.mroute_sk = NULL;
+ mrt->mroute_sk = NULL;
write_unlock_bh(&mrt_lock);
- mroute_clean_tables(net);
+ mroute_clean_tables(mrt);
}
rtnl_unlock();
}
@@ -936,9 +956,10 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
struct vifctl vif;
struct mfcctl mfc;
struct net *net = sock_net(sk);
+ struct mr_table *mrt = net->ipv4.mrt;
if (optname != MRT_INIT) {
- if (sk != net->ipv4.mroute_sk && !capable(CAP_NET_ADMIN))
+ if (sk != mrt->mroute_sk && !capable(CAP_NET_ADMIN))
return -EACCES;
}
@@ -951,7 +972,7 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
return -ENOPROTOOPT;
rtnl_lock();
- if (net->ipv4.mroute_sk) {
+ if (mrt->mroute_sk) {
rtnl_unlock();
return -EADDRINUSE;
}
@@ -959,7 +980,7 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
ret = ip_ra_control(sk, 1, mrtsock_destruct);
if (ret == 0) {
write_lock_bh(&mrt_lock);
- net->ipv4.mroute_sk = sk;
+ mrt->mroute_sk = sk;
write_unlock_bh(&mrt_lock);
IPV4_DEVCONF_ALL(net, MC_FORWARDING)++;
@@ -967,7 +988,7 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
rtnl_unlock();
return ret;
case MRT_DONE:
- if (sk != net->ipv4.mroute_sk)
+ if (sk != mrt->mroute_sk)
return -EACCES;
return ip_ra_control(sk, 0, NULL);
case MRT_ADD_VIF:
@@ -980,9 +1001,9 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
return -ENFILE;
rtnl_lock();
if (optname == MRT_ADD_VIF) {
- ret = vif_add(net, &vif, sk == net->ipv4.mroute_sk);
+ ret = vif_add(net, mrt, &vif, sk == mrt->mroute_sk);
} else {
- ret = vif_delete(net, vif.vifc_vifi, 0, NULL);
+ ret = vif_delete(mrt, vif.vifc_vifi, 0, NULL);
}
rtnl_unlock();
return ret;
@@ -999,9 +1020,9 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
return -EFAULT;
rtnl_lock();
if (optname == MRT_DEL_MFC)
- ret = ipmr_mfc_delete(net, &mfc);
+ ret = ipmr_mfc_delete(mrt, &mfc);
else
- ret = ipmr_mfc_add(net, &mfc, sk == net->ipv4.mroute_sk);
+ ret = ipmr_mfc_add(net, mrt, &mfc, sk == mrt->mroute_sk);
rtnl_unlock();
return ret;
/*
@@ -1012,7 +1033,7 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
int v;
if (get_user(v,(int __user *)optval))
return -EFAULT;
- net->ipv4.mroute_do_assert = (v) ? 1 : 0;
+ mrt->mroute_do_assert = (v) ? 1 : 0;
return 0;
}
#ifdef CONFIG_IP_PIMSM
@@ -1026,9 +1047,9 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
rtnl_lock();
ret = 0;
- if (v != net->ipv4.mroute_do_pim) {
- net->ipv4.mroute_do_pim = v;
- net->ipv4.mroute_do_assert = v;
+ if (v != mrt->mroute_do_pim) {
+ mrt->mroute_do_pim = v;
+ mrt->mroute_do_assert = v;
}
rtnl_unlock();
return ret;
@@ -1052,6 +1073,7 @@ int ip_mroute_getsockopt(struct sock *sk, int optname, char __user *optval, int
int olr;
int val;
struct net *net = sock_net(sk);
+ struct mr_table *mrt = net->ipv4.mrt;
if (optname != MRT_VERSION &&
#ifdef CONFIG_IP_PIMSM
@@ -1073,10 +1095,10 @@ int ip_mroute_getsockopt(struct sock *sk, int optname, char __user *optval, int
val = 0x0305;
#ifdef CONFIG_IP_PIMSM
else if (optname == MRT_PIM)
- val = net->ipv4.mroute_do_pim;
+ val = mrt->mroute_do_pim;
#endif
else
- val = net->ipv4.mroute_do_assert;
+ val = mrt->mroute_do_assert;
if (copy_to_user(optval, &val, olr))
return -EFAULT;
return 0;
@@ -1093,16 +1115,17 @@ int ipmr_ioctl(struct sock *sk, int cmd, void __user *arg)
struct vif_device *vif;
struct mfc_cache *c;
struct net *net = sock_net(sk);
+ struct mr_table *mrt = net->ipv4.mrt;
switch (cmd) {
case SIOCGETVIFCNT:
if (copy_from_user(&vr, arg, sizeof(vr)))
return -EFAULT;
- if (vr.vifi >= net->ipv4.maxvif)
+ if (vr.vifi >= mrt->maxvif)
return -EINVAL;
read_lock(&mrt_lock);
- vif = &net->ipv4.vif_table[vr.vifi];
- if (VIF_EXISTS(net, vr.vifi)) {
+ vif = &mrt->vif_table[vr.vifi];
+ if (VIF_EXISTS(mrt, vr.vifi)) {
vr.icount = vif->pkt_in;
vr.ocount = vif->pkt_out;
vr.ibytes = vif->bytes_in;
@@ -1120,7 +1143,7 @@ int ipmr_ioctl(struct sock *sk, int cmd, void __user *arg)
return -EFAULT;
read_lock(&mrt_lock);
- c = ipmr_cache_find(net, sr.src.s_addr, sr.grp.s_addr);
+ c = ipmr_cache_find(mrt, sr.src.s_addr, sr.grp.s_addr);
if (c) {
sr.pktcnt = c->mfc_un.res.pkt;
sr.bytecnt = c->mfc_un.res.bytes;
@@ -1143,16 +1166,17 @@ static int ipmr_device_event(struct notifier_block *this, unsigned long event, v
{
struct net_device *dev = ptr;
struct net *net = dev_net(dev);
+ struct mr_table *mrt = net->ipv4.mrt;
struct vif_device *v;
int ct;
LIST_HEAD(list);
if (event != NETDEV_UNREGISTER)
return NOTIFY_DONE;
- v = &net->ipv4.vif_table[0];
- for (ct = 0; ct < net->ipv4.maxvif; ct++, v++) {
+ v = &mrt->vif_table[0];
+ for (ct = 0; ct < mrt->maxvif; ct++, v++) {
if (v->dev == dev)
- vif_delete(net, ct, 1, &list);
+ vif_delete(mrt, ct, 1, &list);
}
unregister_netdevice_many(&list);
return NOTIFY_DONE;
@@ -1211,11 +1235,11 @@ static inline int ipmr_forward_finish(struct sk_buff *skb)
* Processing handlers for ipmr_forward
*/
-static void ipmr_queue_xmit(struct net *net, struct sk_buff *skb,
- struct mfc_cache *c, int vifi)
+static void ipmr_queue_xmit(struct net *net, struct mr_table *mrt,
+ struct sk_buff *skb, struct mfc_cache *c, int vifi)
{
const struct iphdr *iph = ip_hdr(skb);
- struct vif_device *vif = &net->ipv4.vif_table[vifi];
+ struct vif_device *vif = &mrt->vif_table[vifi];
struct net_device *dev;
struct rtable *rt;
int encap = 0;
@@ -1229,7 +1253,7 @@ static void ipmr_queue_xmit(struct net *net, struct sk_buff *skb,
vif->bytes_out += skb->len;
vif->dev->stats.tx_bytes += skb->len;
vif->dev->stats.tx_packets++;
- ipmr_cache_report(net, skb, vifi, IGMPMSG_WHOLEPKT);
+ ipmr_cache_report(mrt, skb, vifi, IGMPMSG_WHOLEPKT);
goto out_free;
}
#endif
@@ -1312,12 +1336,12 @@ out_free:
return;
}
-static int ipmr_find_vif(struct net_device *dev)
+static int ipmr_find_vif(struct mr_table *mrt, struct net_device *dev)
{
- struct net *net = dev_net(dev);
int ct;
- for (ct = net->ipv4.maxvif-1; ct >= 0; ct--) {
- if (net->ipv4.vif_table[ct].dev == dev)
+
+ for (ct = mrt->maxvif-1; ct >= 0; ct--) {
+ if (mrt->vif_table[ct].dev == dev)
break;
}
return ct;
@@ -1325,8 +1349,9 @@ static int ipmr_find_vif(struct net_device *dev)
/* "local" means that we should preserve one skb (for local delivery) */
-static int ip_mr_forward(struct net *net, struct sk_buff *skb,
- struct mfc_cache *cache, int local)
+static int ip_mr_forward(struct net *net, struct mr_table *mrt,
+ struct sk_buff *skb, struct mfc_cache *cache,
+ int local)
{
int psend = -1;
int vif, ct;
@@ -1338,7 +1363,7 @@ static int ip_mr_forward(struct net *net, struct sk_buff *skb,
/*
* Wrong interface: drop packet and (maybe) send PIM assert.
*/
- if (net->ipv4.vif_table[vif].dev != skb->dev) {
+ if (mrt->vif_table[vif].dev != skb->dev) {
int true_vifi;
if (skb_rtable(skb)->fl.iif == 0) {
@@ -1357,26 +1382,26 @@ static int ip_mr_forward(struct net *net, struct sk_buff *skb,
}
cache->mfc_un.res.wrong_if++;
- true_vifi = ipmr_find_vif(skb->dev);
+ true_vifi = ipmr_find_vif(mrt, skb->dev);
- if (true_vifi >= 0 && net->ipv4.mroute_do_assert &&
+ if (true_vifi >= 0 && mrt->mroute_do_assert &&
/* pimsm uses asserts, when switching from RPT to SPT,
so that we cannot check that packet arrived on an oif.
It is bad, but otherwise we would need to move pretty
large chunk of pimd to kernel. Ough... --ANK
*/
- (net->ipv4.mroute_do_pim ||
+ (mrt->mroute_do_pim ||
cache->mfc_un.res.ttls[true_vifi] < 255) &&
time_after(jiffies,
cache->mfc_un.res.last_assert + MFC_ASSERT_THRESH)) {
cache->mfc_un.res.last_assert = jiffies;
- ipmr_cache_report(net, skb, true_vifi, IGMPMSG_WRONGVIF);
+ ipmr_cache_report(mrt, skb, true_vifi, IGMPMSG_WRONGVIF);
}
goto dont_forward;
}
- net->ipv4.vif_table[vif].pkt_in++;
- net->ipv4.vif_table[vif].bytes_in += skb->len;
+ mrt->vif_table[vif].pkt_in++;
+ mrt->vif_table[vif].bytes_in += skb->len;
/*
* Forward the frame
@@ -1386,7 +1411,8 @@ static int ip_mr_forward(struct net *net, struct sk_buff *skb,
if (psend != -1) {
struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
if (skb2)
- ipmr_queue_xmit(net, skb2, cache, psend);
+ ipmr_queue_xmit(net, mrt, skb2, cache,
+ psend);
}
psend = ct;
}
@@ -1395,9 +1421,9 @@ static int ip_mr_forward(struct net *net, struct sk_buff *skb,
if (local) {
struct sk_buff *skb2 = skb_clone(skb, GFP_ATOMIC);
if (skb2)
- ipmr_queue_xmit(net, skb2, cache, psend);
+ ipmr_queue_xmit(net, mrt, skb2, cache, psend);
} else {
- ipmr_queue_xmit(net, skb, cache, psend);
+ ipmr_queue_xmit(net, mrt, skb, cache, psend);
return 0;
}
}
@@ -1417,6 +1443,7 @@ int ip_mr_input(struct sk_buff *skb)
{
struct mfc_cache *cache;
struct net *net = dev_net(skb->dev);
+ struct mr_table *mrt = net->ipv4.mrt;
int local = skb_rtable(skb)->rt_flags & RTCF_LOCAL;
/* Packet is looped back after forward, it should not be
@@ -1437,9 +1464,9 @@ int ip_mr_input(struct sk_buff *skb)
that we can forward NO IGMP messages.
*/
read_lock(&mrt_lock);
- if (net->ipv4.mroute_sk) {
+ if (mrt->mroute_sk) {
nf_reset(skb);
- raw_rcv(net->ipv4.mroute_sk, skb);
+ raw_rcv(mrt->mroute_sk, skb);
read_unlock(&mrt_lock);
return 0;
}
@@ -1448,7 +1475,7 @@ int ip_mr_input(struct sk_buff *skb)
}
read_lock(&mrt_lock);
- cache = ipmr_cache_find(net, ip_hdr(skb)->saddr, ip_hdr(skb)->daddr);
+ cache = ipmr_cache_find(mrt, ip_hdr(skb)->saddr, ip_hdr(skb)->daddr);
/*
* No usable cache entry
@@ -1466,9 +1493,9 @@ int ip_mr_input(struct sk_buff *skb)
skb = skb2;
}
- vif = ipmr_find_vif(skb->dev);
+ vif = ipmr_find_vif(mrt, skb->dev);
if (vif >= 0) {
- int err = ipmr_cache_unresolved(net, vif, skb);
+ int err = ipmr_cache_unresolved(mrt, vif, skb);
read_unlock(&mrt_lock);
return err;
@@ -1478,7 +1505,7 @@ int ip_mr_input(struct sk_buff *skb)
return -ENODEV;
}
- ip_mr_forward(net, skb, cache, local);
+ ip_mr_forward(net, mrt, skb, cache, local);
read_unlock(&mrt_lock);
@@ -1500,6 +1527,7 @@ static int __pim_rcv(struct sk_buff *skb, unsigned int pimlen)
struct net_device *reg_dev = NULL;
struct iphdr *encap;
struct net *net = dev_net(skb->dev);
+ struct mr_table *mrt = net->ipv4.mrt;
encap = (struct iphdr *)(skb_transport_header(skb) + pimlen);
/*
@@ -1514,8 +1542,8 @@ static int __pim_rcv(struct sk_buff *skb, unsigned int pimlen)
return 1;
read_lock(&mrt_lock);
- if (net->ipv4.mroute_reg_vif_num >= 0)
- reg_dev = net->ipv4.vif_table[net->ipv4.mroute_reg_vif_num].dev;
+ if (mrt->mroute_reg_vif_num >= 0)
+ reg_dev = mrt->vif_table[mrt->mroute_reg_vif_num].dev;
if (reg_dev)
dev_hold(reg_dev);
read_unlock(&mrt_lock);
@@ -1550,13 +1578,14 @@ int pim_rcv_v1(struct sk_buff * skb)
{
struct igmphdr *pim;
struct net *net = dev_net(skb->dev);
+ struct mr_table *mrt = net->ipv4.mrt;
if (!pskb_may_pull(skb, sizeof(*pim) + sizeof(struct iphdr)))
goto drop;
pim = igmp_hdr(skb);
- if (!net->ipv4.mroute_do_pim ||
+ if (!mrt->mroute_do_pim ||
pim->group != PIM_V1_VERSION || pim->code != PIM_V1_REGISTER)
goto drop;
@@ -1592,7 +1621,7 @@ drop:
#endif
static int
-ipmr_fill_mroute(struct net *net, struct sk_buff *skb, struct mfc_cache *c,
+ipmr_fill_mroute(struct mr_table *mrt, struct sk_buff *skb, struct mfc_cache *c,
struct rtmsg *rtm)
{
int ct;
@@ -1604,19 +1633,19 @@ ipmr_fill_mroute(struct net *net, struct sk_buff *skb, struct mfc_cache *c,
if (c->mfc_parent > MAXVIFS)
return -ENOENT;
- if (VIF_EXISTS(net, c->mfc_parent))
- RTA_PUT(skb, RTA_IIF, 4, &net->ipv4.vif_table[c->mfc_parent].dev->ifindex);
+ if (VIF_EXISTS(mrt, c->mfc_parent))
+ RTA_PUT(skb, RTA_IIF, 4, &mrt->vif_table[c->mfc_parent].dev->ifindex);
mp_head = (struct rtattr *)skb_put(skb, RTA_LENGTH(0));
for (ct = c->mfc_un.res.minvif; ct < c->mfc_un.res.maxvif; ct++) {
- if (VIF_EXISTS(net, ct) && c->mfc_un.res.ttls[ct] < 255) {
+ if (VIF_EXISTS(mrt, ct) && c->mfc_un.res.ttls[ct] < 255) {
if (skb_tailroom(skb) < RTA_ALIGN(RTA_ALIGN(sizeof(*nhp)) + 4))
goto rtattr_failure;
nhp = (struct rtnexthop *)skb_put(skb, RTA_ALIGN(sizeof(*nhp)));
nhp->rtnh_flags = 0;
nhp->rtnh_hops = c->mfc_un.res.ttls[ct];
- nhp->rtnh_ifindex = net->ipv4.vif_table[ct].dev->ifindex;
+ nhp->rtnh_ifindex = mrt->vif_table[ct].dev->ifindex;
nhp->rtnh_len = sizeof(*nhp);
}
}
@@ -1634,11 +1663,12 @@ int ipmr_get_route(struct net *net,
struct sk_buff *skb, struct rtmsg *rtm, int nowait)
{
int err;
+ struct mr_table *mrt = net->ipv4.mrt;
struct mfc_cache *cache;
struct rtable *rt = skb_rtable(skb);
read_lock(&mrt_lock);
- cache = ipmr_cache_find(net, rt->rt_src, rt->rt_dst);
+ cache = ipmr_cache_find(mrt, rt->rt_src, rt->rt_dst);
if (cache == NULL) {
struct sk_buff *skb2;
@@ -1652,7 +1682,7 @@ int ipmr_get_route(struct net *net,
}
dev = skb->dev;
- if (dev == NULL || (vif = ipmr_find_vif(dev)) < 0) {
+ if (dev == NULL || (vif = ipmr_find_vif(mrt, dev)) < 0) {
read_unlock(&mrt_lock);
return -ENODEV;
}
@@ -1669,14 +1699,14 @@ int ipmr_get_route(struct net *net,
iph->saddr = rt->rt_src;
iph->daddr = rt->rt_dst;
iph->version = 0;
- err = ipmr_cache_unresolved(net, vif, skb2);
+ err = ipmr_cache_unresolved(mrt, vif, skb2);
read_unlock(&mrt_lock);
return err;
}
if (!nowait && (rtm->rtm_flags&RTM_F_NOTIFY))
cache->mfc_flags |= MFC_NOTIFY;
- err = ipmr_fill_mroute(net, skb, cache, rtm);
+ err = ipmr_fill_mroute(mrt, skb, cache, rtm);
read_unlock(&mrt_lock);
return err;
}
@@ -1694,11 +1724,13 @@ static struct vif_device *ipmr_vif_seq_idx(struct net *net,
struct ipmr_vif_iter *iter,
loff_t pos)
{
- for (iter->ct = 0; iter->ct < net->ipv4.maxvif; ++iter->ct) {
- if (!VIF_EXISTS(net, iter->ct))
+ struct mr_table *mrt = net->ipv4.mrt;
+
+ for (iter->ct = 0; iter->ct < mrt->maxvif; ++iter->ct) {
+ if (!VIF_EXISTS(mrt, iter->ct))
continue;
if (pos-- == 0)
- return &net->ipv4.vif_table[iter->ct];
+ return &mrt->vif_table[iter->ct];
}
return NULL;
}
@@ -1717,15 +1749,16 @@ static void *ipmr_vif_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
struct ipmr_vif_iter *iter = seq->private;
struct net *net = seq_file_net(seq);
+ struct mr_table *mrt = net->ipv4.mrt;
++*pos;
if (v == SEQ_START_TOKEN)
return ipmr_vif_seq_idx(net, iter, 0);
- while (++iter->ct < net->ipv4.maxvif) {
- if (!VIF_EXISTS(net, iter->ct))
+ while (++iter->ct < mrt->maxvif) {
+ if (!VIF_EXISTS(mrt, iter->ct))
continue;
- return &net->ipv4.vif_table[iter->ct];
+ return &mrt->vif_table[iter->ct];
}
return NULL;
}
@@ -1739,6 +1772,7 @@ static void ipmr_vif_seq_stop(struct seq_file *seq, void *v)
static int ipmr_vif_seq_show(struct seq_file *seq, void *v)
{
struct net *net = seq_file_net(seq);
+ struct mr_table *mrt = net->ipv4.mrt;
if (v == SEQ_START_TOKEN) {
seq_puts(seq,
@@ -1749,7 +1783,7 @@ static int ipmr_vif_seq_show(struct seq_file *seq, void *v)
seq_printf(seq,
"%2Zd %-10s %8ld %7ld %8ld %7ld %05X %08X %08X\n",
- vif - net->ipv4.vif_table,
+ vif - mrt->vif_table,
name, vif->bytes_in, vif->pkt_in,
vif->bytes_out, vif->pkt_out,
vif->flags, vif->local, vif->remote);
@@ -1788,11 +1822,12 @@ struct ipmr_mfc_iter {
static struct mfc_cache *ipmr_mfc_seq_idx(struct net *net,
struct ipmr_mfc_iter *it, loff_t pos)
{
+ struct mr_table *mrt = net->ipv4.mrt;
struct mfc_cache *mfc;
read_lock(&mrt_lock);
for (it->ct = 0; it->ct < MFC_LINES; it->ct++) {
- it->cache = &net->ipv4.mfc_cache_array[it->ct];
+ it->cache = &mrt->mfc_cache_array[it->ct];
list_for_each_entry(mfc, it->cache, list)
if (pos-- == 0)
return mfc;
@@ -1800,7 +1835,7 @@ static struct mfc_cache *ipmr_mfc_seq_idx(struct net *net,
read_unlock(&mrt_lock);
spin_lock_bh(&mfc_unres_lock);
- it->cache = &net->ipv4.mfc_unres_queue;
+ it->cache = &mrt->mfc_unres_queue;
list_for_each_entry(mfc, it->cache, list)
if (pos-- == 0)
return mfc;
@@ -1827,6 +1862,7 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
struct mfc_cache *mfc = v;
struct ipmr_mfc_iter *it = seq->private;
struct net *net = seq_file_net(seq);
+ struct mr_table *mrt = net->ipv4.mrt;
++*pos;
@@ -1836,13 +1872,13 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
if (mfc->list.next != it->cache)
return list_entry(mfc->list.next, struct mfc_cache, list);
- if (it->cache == &net->ipv4.mfc_unres_queue)
+ if (it->cache == &mrt->mfc_unres_queue)
goto end_of_list;
- BUG_ON(it->cache != &net->ipv4.mfc_cache_array[it->ct]);
+ BUG_ON(it->cache != &mrt->mfc_cache_array[it->ct]);
while (++it->ct < MFC_LINES) {
- it->cache = &net->ipv4.mfc_cache_array[it->ct];
+ it->cache = &mrt->mfc_cache_array[it->ct];
if (list_empty(it->cache))
continue;
return list_first_entry(it->cache, struct mfc_cache, list);
@@ -1850,7 +1886,7 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
/* exhausted cache_array, show unresolved */
read_unlock(&mrt_lock);
- it->cache = &net->ipv4.mfc_unres_queue;
+ it->cache = &mrt->mfc_unres_queue;
it->ct = 0;
spin_lock_bh(&mfc_unres_lock);
@@ -1868,10 +1904,11 @@ static void ipmr_mfc_seq_stop(struct seq_file *seq, void *v)
{
struct ipmr_mfc_iter *it = seq->private;
struct net *net = seq_file_net(seq);
+ struct mr_table *mrt = net->ipv4.mrt;
- if (it->cache == &net->ipv4.mfc_unres_queue)
+ if (it->cache == &mrt->mfc_unres_queue)
spin_unlock_bh(&mfc_unres_lock);
- else if (it->cache == &net->ipv4.mfc_cache_array[it->ct])
+ else if (it->cache == &mrt->mfc_cache_array[it->ct])
read_unlock(&mrt_lock);
}
@@ -1879,6 +1916,7 @@ static int ipmr_mfc_seq_show(struct seq_file *seq, void *v)
{
int n;
struct net *net = seq_file_net(seq);
+ struct mr_table *mrt = net->ipv4.mrt;
if (v == SEQ_START_TOKEN) {
seq_puts(seq,
@@ -1892,14 +1930,14 @@ static int ipmr_mfc_seq_show(struct seq_file *seq, void *v)
(unsigned long) mfc->mfc_origin,
mfc->mfc_parent);
- if (it->cache != &net->ipv4.mfc_unres_queue) {
+ if (it->cache != &mrt->mfc_unres_queue) {
seq_printf(seq, " %8lu %8lu %8lu",
mfc->mfc_un.res.pkt,
mfc->mfc_un.res.bytes,
mfc->mfc_un.res.wrong_if);
for (n = mfc->mfc_un.res.minvif;
n < mfc->mfc_un.res.maxvif; n++ ) {
- if (VIF_EXISTS(net, n) &&
+ if (VIF_EXISTS(mrt, n) &&
mfc->mfc_un.res.ttls[n] < 255)
seq_printf(seq,
" %2d:%-3d",
@@ -1951,35 +1989,27 @@ static const struct net_protocol pim_protocol = {
*/
static int __net_init ipmr_net_init(struct net *net)
{
+ struct mr_table *mrt;
unsigned int i;
int err = 0;
- net->ipv4.vif_table = kcalloc(MAXVIFS, sizeof(struct vif_device),
- GFP_KERNEL);
- if (!net->ipv4.vif_table) {
+ mrt = kzalloc(sizeof(*mrt), GFP_KERNEL);
+ if (mrt == NULL) {
err = -ENOMEM;
goto fail;
}
/* Forwarding cache */
- net->ipv4.mfc_cache_array = kcalloc(MFC_LINES,
- sizeof(struct list_head),
- GFP_KERNEL);
- if (!net->ipv4.mfc_cache_array) {
- err = -ENOMEM;
- goto fail_mfc_cache;
- }
-
for (i = 0; i < MFC_LINES; i++)
- INIT_LIST_HEAD(&net->ipv4.mfc_cache_array[i]);
+ INIT_LIST_HEAD(&mrt->mfc_cache_array[i]);
- INIT_LIST_HEAD(&net->ipv4.mfc_unres_queue);
+ INIT_LIST_HEAD(&mrt->mfc_unres_queue);
- setup_timer(&net->ipv4.ipmr_expire_timer, ipmr_expire_process,
+ setup_timer(&mrt->ipmr_expire_timer, ipmr_expire_process,
(unsigned long)net);
#ifdef CONFIG_IP_PIMSM
- net->ipv4.mroute_reg_vif_num = -1;
+ mrt->mroute_reg_vif_num = -1;
#endif
#ifdef CONFIG_PROC_FS
@@ -1989,16 +2019,16 @@ static int __net_init ipmr_net_init(struct net *net)
if (!proc_net_fops_create(net, "ip_mr_cache", 0, &ipmr_mfc_fops))
goto proc_cache_fail;
#endif
+
+ net->ipv4.mrt = mrt;
return 0;
#ifdef CONFIG_PROC_FS
proc_cache_fail:
proc_net_remove(net, "ip_mr_vif");
proc_vif_fail:
- kfree(net->ipv4.mfc_cache_array);
+ kfree(mrt);
#endif
-fail_mfc_cache:
- kfree(net->ipv4.vif_table);
fail:
return err;
}
@@ -2009,8 +2039,7 @@ static void __net_exit ipmr_net_exit(struct net *net)
proc_net_remove(net, "ip_mr_cache");
proc_net_remove(net, "ip_mr_vif");
#endif
- kfree(net->ipv4.mfc_cache_array);
- kfree(net->ipv4.vif_table);
+ kfree(net->ipv4.mrt);
}
static struct pernet_operations ipmr_net_ops = {
--
1.7.0.4
^ permalink raw reply related
* [PATCH 9/9] ipv4: ipmr: support multiple tables
From: Patrick McHardy @ 2010-04-13 15:03 UTC (permalink / raw)
To: davem; +Cc: netdev
In-Reply-To: <1271171003-11901-1-git-send-email-kaber@trash.net>
This patch adds support for multiple independant multicast routing instances,
named "tables".
Userspace multicast routing daemons can bind to a specific table instance by
issuing a setsockopt call using a new option MRT_TABLE. The table number is
stored in the raw socket data and affects all following ipmr setsockopt(),
getsockopt() and ioctl() calls. By default, a single table (RT_TABLE_DEFAULT)
is created with a default routing rule pointing to it. Newly created pimreg
devices have the table number appended ("pimregX"), with the exception of
devices created in the default table, which are named just "pimreg" for
compatibility reasons.
Packets are directed to a specific table instance using routing rules,
similar to how regular routing rules work. Currently iif, oif and mark
are supported as keys, source and destination addresses could be supported
additionally.
Example usage:
- bind pimd/xorp/... to a specific table:
uint32_t table = 123;
setsockopt(fd, IPPROTO_IP, MRT_TABLE, &table, sizeof(table));
- create routing rules directing packets to the new table:
# ip mrule add iif eth0 lookup 123
# ip mrule add oif eth0 lookup 123
Signed-off-by: Patrick McHardy <kaber@trash.net>
---
include/linux/fib_rules.h | 1 +
include/linux/mroute.h | 3 +-
include/net/netns/ipv4.h | 5 +
include/net/raw.h | 1 +
net/ipv4/Kconfig | 14 ++
net/ipv4/ipmr.c | 399 ++++++++++++++++++++++++++++++++++++++-------
6 files changed, 361 insertions(+), 62 deletions(-)
diff --git a/include/linux/fib_rules.h b/include/linux/fib_rules.h
index 405e411..04a3976 100644
--- a/include/linux/fib_rules.h
+++ b/include/linux/fib_rules.h
@@ -21,6 +21,7 @@
#define FIB_RULES_IPV4 AF_INET
#define FIB_RULES_IPV6 AF_INET6
#define FIB_RULES_DECNET AF_DECnet
+#define FIB_RULES_IPMR 128
struct fib_rule_hdr {
__u8 family;
diff --git a/include/linux/mroute.h b/include/linux/mroute.h
index 7ff6c77..fa04b24 100644
--- a/include/linux/mroute.h
+++ b/include/linux/mroute.h
@@ -27,7 +27,8 @@
#define MRT_DEL_MFC (MRT_BASE+5) /* Delete a multicast forwarding entry */
#define MRT_VERSION (MRT_BASE+6) /* Get the kernel multicast version */
#define MRT_ASSERT (MRT_BASE+7) /* Activate PIM assert mode */
-#define MRT_PIM (MRT_BASE+8) /* enable PIM code */
+#define MRT_PIM (MRT_BASE+8) /* enable PIM code */
+#define MRT_TABLE (MRT_BASE+9) /* Specify mroute table ID */
#define SIOCGETVIFCNT SIOCPROTOPRIVATE /* IP protocol privates */
#define SIOCGETSGCNT (SIOCPROTOPRIVATE+1)
diff --git a/include/net/netns/ipv4.h b/include/net/netns/ipv4.h
index 72e762a..ae07fee 100644
--- a/include/net/netns/ipv4.h
+++ b/include/net/netns/ipv4.h
@@ -59,7 +59,12 @@ struct netns_ipv4 {
atomic_t rt_genid;
#ifdef CONFIG_IP_MROUTE
+#ifndef CONFIG_IP_MROUTE_MULTIPLE_TABLES
struct mr_table *mrt;
+#else
+ struct list_head mr_tables;
+ struct fib_rules_ops *mr_rules_ops;
+#endif
#endif
};
#endif
diff --git a/include/net/raw.h b/include/net/raw.h
index 67cc643..43c5750 100644
--- a/include/net/raw.h
+++ b/include/net/raw.h
@@ -61,6 +61,7 @@ struct raw_sock {
/* inet_sock has to be the first member */
struct inet_sock inet;
struct icmp_filter filter;
+ u32 ipmr_table;
};
static inline struct raw_sock *raw_sk(const struct sock *sk)
diff --git a/net/ipv4/Kconfig b/net/ipv4/Kconfig
index c9a1c68..be59774 100644
--- a/net/ipv4/Kconfig
+++ b/net/ipv4/Kconfig
@@ -250,6 +250,20 @@ config IP_MROUTE
<file:Documentation/networking/multicast.txt>. If you haven't heard
about it, you don't need it.
+config IP_MROUTE_MULTIPLE_TABLES
+ bool "IP: multicast policy routing"
+ depends on IP_ADVANCED_ROUTER
+ select FIB_RULES
+ help
+ Normally, a multicast router runs a userspace daemon and decides
+ what to do with a multicast packet based on the source and
+ destination addresses. If you say Y here, the multicast router
+ will also be able to take interfaces and packet marks into
+ account and run multiple instances of userspace daemons
+ simultaneously, each one handling a single table.
+
+ If unsure, say N.
+
config IP_PIMSM_V1
bool "IP: PIM-SM version 1 support"
depends on IP_MROUTE
diff --git a/net/ipv4/ipmr.c b/net/ipv4/ipmr.c
index 498f4e9..5df5fd7 100644
--- a/net/ipv4/ipmr.c
+++ b/net/ipv4/ipmr.c
@@ -63,12 +63,15 @@
#include <net/ipip.h>
#include <net/checksum.h>
#include <net/netlink.h>
+#include <net/fib_rules.h>
#if defined(CONFIG_IP_PIMSM_V1) || defined(CONFIG_IP_PIMSM_V2)
#define CONFIG_IP_PIMSM 1
#endif
struct mr_table {
+ struct list_head list;
+ u32 id;
struct sock *mroute_sk;
struct timer_list ipmr_expire_timer;
struct list_head mfc_unres_queue;
@@ -83,6 +86,14 @@ struct mr_table {
#endif
};
+struct ipmr_rule {
+ struct fib_rule common;
+};
+
+struct ipmr_result {
+ struct mr_table *mrt;
+};
+
/* Big lock, protecting vif table, mrt cache and mroute socket state.
Note that the changes are semaphored via rtnl_lock.
*/
@@ -108,6 +119,7 @@ static DEFINE_SPINLOCK(mfc_unres_lock);
static struct kmem_cache *mrt_cachep __read_mostly;
+static struct mr_table *ipmr_new_table(struct net *net, u32 id);
static int ip_mr_forward(struct net *net, struct mr_table *mrt,
struct sk_buff *skb, struct mfc_cache *cache,
int local);
@@ -115,6 +127,206 @@ static int ipmr_cache_report(struct mr_table *mrt,
struct sk_buff *pkt, vifi_t vifi, int assert);
static int ipmr_fill_mroute(struct mr_table *mrt, struct sk_buff *skb,
struct mfc_cache *c, struct rtmsg *rtm);
+static void ipmr_expire_process(unsigned long arg);
+
+#ifdef CONFIG_IP_MROUTE_MULTIPLE_TABLES
+#define ipmr_for_each_table(mrt, net) \
+ list_for_each_entry_rcu(mrt, &net->ipv4.mr_tables, list)
+
+static struct mr_table *ipmr_get_table(struct net *net, u32 id)
+{
+ struct mr_table *mrt;
+
+ ipmr_for_each_table(mrt, net) {
+ if (mrt->id == id)
+ return mrt;
+ }
+ return NULL;
+}
+
+static int ipmr_fib_lookup(struct net *net, struct flowi *flp,
+ struct mr_table **mrt)
+{
+ struct ipmr_result res;
+ struct fib_lookup_arg arg = { .result = &res, };
+ int err;
+
+ err = fib_rules_lookup(net->ipv4.mr_rules_ops, flp, 0, &arg);
+ if (err < 0)
+ return err;
+ *mrt = res.mrt;
+ return 0;
+}
+
+static int ipmr_rule_action(struct fib_rule *rule, struct flowi *flp,
+ int flags, struct fib_lookup_arg *arg)
+{
+ struct ipmr_result *res = arg->result;
+ struct mr_table *mrt;
+
+ switch (rule->action) {
+ case FR_ACT_TO_TBL:
+ break;
+ case FR_ACT_UNREACHABLE:
+ return -ENETUNREACH;
+ case FR_ACT_PROHIBIT:
+ return -EACCES;
+ case FR_ACT_BLACKHOLE:
+ default:
+ return -EINVAL;
+ }
+
+ mrt = ipmr_get_table(rule->fr_net, rule->table);
+ if (mrt == NULL)
+ return -EAGAIN;
+ res->mrt = mrt;
+ return 0;
+}
+
+static int ipmr_rule_match(struct fib_rule *rule, struct flowi *fl, int flags)
+{
+ return 1;
+}
+
+static const struct nla_policy ipmr_rule_policy[FRA_MAX + 1] = {
+ FRA_GENERIC_POLICY,
+};
+
+static int ipmr_rule_configure(struct fib_rule *rule, struct sk_buff *skb,
+ struct fib_rule_hdr *frh, struct nlattr **tb)
+{
+ return 0;
+}
+
+static int ipmr_rule_compare(struct fib_rule *rule, struct fib_rule_hdr *frh,
+ struct nlattr **tb)
+{
+ return 1;
+}
+
+static int ipmr_rule_fill(struct fib_rule *rule, struct sk_buff *skb,
+ struct fib_rule_hdr *frh)
+{
+ frh->dst_len = 0;
+ frh->src_len = 0;
+ frh->tos = 0;
+ return 0;
+}
+
+static struct fib_rules_ops ipmr_rules_ops_template = {
+ .family = FIB_RULES_IPMR,
+ .rule_size = sizeof(struct ipmr_rule),
+ .addr_size = sizeof(u32),
+ .action = ipmr_rule_action,
+ .match = ipmr_rule_match,
+ .configure = ipmr_rule_configure,
+ .compare = ipmr_rule_compare,
+ .default_pref = fib_default_rule_pref,
+ .fill = ipmr_rule_fill,
+ .nlgroup = RTNLGRP_IPV4_RULE,
+ .policy = ipmr_rule_policy,
+ .owner = THIS_MODULE,
+};
+
+static int __net_init ipmr_rules_init(struct net *net)
+{
+ struct fib_rules_ops *ops;
+ struct mr_table *mrt;
+ int err;
+
+ ops = fib_rules_register(&ipmr_rules_ops_template, net);
+ if (IS_ERR(ops))
+ return PTR_ERR(ops);
+
+ INIT_LIST_HEAD(&net->ipv4.mr_tables);
+
+ mrt = ipmr_new_table(net, RT_TABLE_DEFAULT);
+ if (mrt == NULL) {
+ err = -ENOMEM;
+ goto err1;
+ }
+
+ err = fib_default_rule_add(ops, 0x7fff, RT_TABLE_DEFAULT, 0);
+ if (err < 0)
+ goto err2;
+
+ net->ipv4.mr_rules_ops = ops;
+ return 0;
+
+err2:
+ kfree(mrt);
+err1:
+ fib_rules_unregister(ops);
+ return err;
+}
+
+static void __net_exit ipmr_rules_exit(struct net *net)
+{
+ struct mr_table *mrt, *next;
+
+ list_for_each_entry_safe(mrt, next, &net->ipv4.mr_tables, list)
+ kfree(mrt);
+ fib_rules_unregister(net->ipv4.mr_rules_ops);
+}
+#else
+#define ipmr_for_each_table(mrt, net) \
+ for (mrt = net->ipv4.mrt; mrt; mrt = NULL)
+
+static struct mr_table *ipmr_get_table(struct net *net, u32 id)
+{
+ return net->ipv4.mrt;
+}
+
+static int ipmr_fib_lookup(struct net *net, struct flowi *flp,
+ struct mr_table **mrt)
+{
+ *mrt = net->ipv4.mrt;
+ return 0;
+}
+
+static int __net_init ipmr_rules_init(struct net *net)
+{
+ net->ipv4.mrt = ipmr_new_table(net, RT_TABLE_DEFAULT);
+ return net->ipv4.mrt ? 0 : -ENOMEM;
+}
+
+static void __net_exit ipmr_rules_exit(struct net *net)
+{
+ kfree(net->ipv4.mrt);
+}
+#endif
+
+static struct mr_table *ipmr_new_table(struct net *net, u32 id)
+{
+ struct mr_table *mrt;
+ unsigned int i;
+
+ mrt = ipmr_get_table(net, id);
+ if (mrt != NULL)
+ return mrt;
+
+ mrt = kzalloc(sizeof(*mrt), GFP_KERNEL);
+ if (mrt == NULL)
+ return NULL;
+ mrt->id = id;
+
+ /* Forwarding cache */
+ for (i = 0; i < MFC_LINES; i++)
+ INIT_LIST_HEAD(&mrt->mfc_cache_array[i]);
+
+ INIT_LIST_HEAD(&mrt->mfc_unres_queue);
+
+ setup_timer(&mrt->ipmr_expire_timer, ipmr_expire_process,
+ (unsigned long)mrt);
+
+#ifdef CONFIG_IP_PIMSM
+ mrt->mroute_reg_vif_num = -1;
+#endif
+#ifdef CONFIG_IP_MROUTE_MULTIPLE_TABLES
+ list_add_tail_rcu(&mrt->list, &net->ipv4.mr_tables);
+#endif
+ return mrt;
+}
/* Service routines creating virtual interfaces: DVMRP tunnels and PIMREG */
@@ -215,7 +427,17 @@ failure:
static netdev_tx_t reg_vif_xmit(struct sk_buff *skb, struct net_device *dev)
{
struct net *net = dev_net(dev);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt;
+ struct flowi fl = {
+ .oif = dev->ifindex,
+ .iif = skb->skb_iif,
+ .mark = skb->mark,
+ };
+ int err;
+
+ err = ipmr_fib_lookup(net, &fl, &mrt);
+ if (err < 0)
+ return err;
read_lock(&mrt_lock);
dev->stats.tx_bytes += skb->len;
@@ -240,12 +462,18 @@ static void reg_vif_setup(struct net_device *dev)
dev->features |= NETIF_F_NETNS_LOCAL;
}
-static struct net_device *ipmr_reg_vif(struct net *net)
+static struct net_device *ipmr_reg_vif(struct net *net, struct mr_table *mrt)
{
struct net_device *dev;
struct in_device *in_dev;
+ char name[IFNAMSIZ];
+
+ if (mrt->id == RT_TABLE_DEFAULT)
+ sprintf(name, "pimreg");
+ else
+ sprintf(name, "pimreg%u", mrt->id);
- dev = alloc_netdev(0, "pimreg", reg_vif_setup);
+ dev = alloc_netdev(0, name, reg_vif_setup);
if (dev == NULL)
return NULL;
@@ -461,7 +689,7 @@ static int vif_add(struct net *net, struct mr_table *mrt,
*/
if (mrt->mroute_reg_vif_num >= 0)
return -EADDRINUSE;
- dev = ipmr_reg_vif(net);
+ dev = ipmr_reg_vif(net, mrt);
if (!dev)
return -ENOBUFS;
err = dev_set_allmulti(dev, 1);
@@ -928,17 +1156,19 @@ static void mroute_clean_tables(struct mr_table *mrt)
static void mrtsock_destruct(struct sock *sk)
{
struct net *net = sock_net(sk);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt;
rtnl_lock();
- if (sk == mrt->mroute_sk) {
- IPV4_DEVCONF_ALL(net, MC_FORWARDING)--;
+ ipmr_for_each_table(mrt, net) {
+ if (sk == mrt->mroute_sk) {
+ IPV4_DEVCONF_ALL(net, MC_FORWARDING)--;
- write_lock_bh(&mrt_lock);
- mrt->mroute_sk = NULL;
- write_unlock_bh(&mrt_lock);
+ write_lock_bh(&mrt_lock);
+ mrt->mroute_sk = NULL;
+ write_unlock_bh(&mrt_lock);
- mroute_clean_tables(mrt);
+ mroute_clean_tables(mrt);
+ }
}
rtnl_unlock();
}
@@ -956,7 +1186,11 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
struct vifctl vif;
struct mfcctl mfc;
struct net *net = sock_net(sk);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt;
+
+ mrt = ipmr_get_table(net, raw_sk(sk)->ipmr_table ? : RT_TABLE_DEFAULT);
+ if (mrt == NULL)
+ return -ENOENT;
if (optname != MRT_INIT) {
if (sk != mrt->mroute_sk && !capable(CAP_NET_ADMIN))
@@ -1055,6 +1289,27 @@ int ip_mroute_setsockopt(struct sock *sk, int optname, char __user *optval, unsi
return ret;
}
#endif
+#ifdef CONFIG_IP_MROUTE_MULTIPLE_TABLES
+ case MRT_TABLE:
+ {
+ u32 v;
+
+ if (optlen != sizeof(u32))
+ return -EINVAL;
+ if (get_user(v, (u32 __user *)optval))
+ return -EFAULT;
+ if (sk == mrt->mroute_sk)
+ return -EBUSY;
+
+ rtnl_lock();
+ ret = 0;
+ if (!ipmr_new_table(net, v))
+ ret = -ENOMEM;
+ raw_sk(sk)->ipmr_table = v;
+ rtnl_unlock();
+ return ret;
+ }
+#endif
/*
* Spurious command, or MRT_VERSION which you cannot
* set.
@@ -1073,7 +1328,11 @@ int ip_mroute_getsockopt(struct sock *sk, int optname, char __user *optval, int
int olr;
int val;
struct net *net = sock_net(sk);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt;
+
+ mrt = ipmr_get_table(net, raw_sk(sk)->ipmr_table ? : RT_TABLE_DEFAULT);
+ if (mrt == NULL)
+ return -ENOENT;
if (optname != MRT_VERSION &&
#ifdef CONFIG_IP_PIMSM
@@ -1115,7 +1374,11 @@ int ipmr_ioctl(struct sock *sk, int cmd, void __user *arg)
struct vif_device *vif;
struct mfc_cache *c;
struct net *net = sock_net(sk);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt;
+
+ mrt = ipmr_get_table(net, raw_sk(sk)->ipmr_table ? : RT_TABLE_DEFAULT);
+ if (mrt == NULL)
+ return -ENOENT;
switch (cmd) {
case SIOCGETVIFCNT:
@@ -1166,17 +1429,20 @@ static int ipmr_device_event(struct notifier_block *this, unsigned long event, v
{
struct net_device *dev = ptr;
struct net *net = dev_net(dev);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt;
struct vif_device *v;
int ct;
LIST_HEAD(list);
if (event != NETDEV_UNREGISTER)
return NOTIFY_DONE;
- v = &mrt->vif_table[0];
- for (ct = 0; ct < mrt->maxvif; ct++, v++) {
- if (v->dev == dev)
- vif_delete(mrt, ct, 1, &list);
+
+ ipmr_for_each_table(mrt, net) {
+ v = &mrt->vif_table[0];
+ for (ct = 0; ct < mrt->maxvif; ct++, v++) {
+ if (v->dev == dev)
+ vif_delete(mrt, ct, 1, &list);
+ }
}
unregister_netdevice_many(&list);
return NOTIFY_DONE;
@@ -1443,8 +1709,9 @@ int ip_mr_input(struct sk_buff *skb)
{
struct mfc_cache *cache;
struct net *net = dev_net(skb->dev);
- struct mr_table *mrt = net->ipv4.mrt;
int local = skb_rtable(skb)->rt_flags & RTCF_LOCAL;
+ struct mr_table *mrt;
+ int err;
/* Packet is looped back after forward, it should not be
forwarded second time, but still can be delivered locally.
@@ -1452,6 +1719,10 @@ int ip_mr_input(struct sk_buff *skb)
if (IPCB(skb)->flags&IPSKB_FORWARDED)
goto dont_forward;
+ err = ipmr_fib_lookup(net, &skb_rtable(skb)->fl, &mrt);
+ if (err < 0)
+ return err;
+
if (!local) {
if (IPCB(skb)->opt.router_alert) {
if (ip_call_ra_chain(skb))
@@ -1522,12 +1793,11 @@ dont_forward:
}
#ifdef CONFIG_IP_PIMSM
-static int __pim_rcv(struct sk_buff *skb, unsigned int pimlen)
+static int __pim_rcv(struct mr_table *mrt, struct sk_buff *skb,
+ unsigned int pimlen)
{
struct net_device *reg_dev = NULL;
struct iphdr *encap;
- struct net *net = dev_net(skb->dev);
- struct mr_table *mrt = net->ipv4.mrt;
encap = (struct iphdr *)(skb_transport_header(skb) + pimlen);
/*
@@ -1578,18 +1848,21 @@ int pim_rcv_v1(struct sk_buff * skb)
{
struct igmphdr *pim;
struct net *net = dev_net(skb->dev);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt;
if (!pskb_may_pull(skb, sizeof(*pim) + sizeof(struct iphdr)))
goto drop;
pim = igmp_hdr(skb);
+ if (ipmr_fib_lookup(net, &skb_rtable(skb)->fl, &mrt) < 0)
+ goto drop;
+
if (!mrt->mroute_do_pim ||
pim->group != PIM_V1_VERSION || pim->code != PIM_V1_REGISTER)
goto drop;
- if (__pim_rcv(skb, sizeof(*pim))) {
+ if (__pim_rcv(mrt, skb, sizeof(*pim))) {
drop:
kfree_skb(skb);
}
@@ -1601,6 +1874,8 @@ drop:
static int pim_rcv(struct sk_buff * skb)
{
struct pimreghdr *pim;
+ struct net *net = dev_net(skb->dev);
+ struct mr_table *mrt;
if (!pskb_may_pull(skb, sizeof(*pim) + sizeof(struct iphdr)))
goto drop;
@@ -1612,7 +1887,10 @@ static int pim_rcv(struct sk_buff * skb)
csum_fold(skb_checksum(skb, 0, skb->len, 0))))
goto drop;
- if (__pim_rcv(skb, sizeof(*pim))) {
+ if (ipmr_fib_lookup(net, &skb_rtable(skb)->fl, &mrt) < 0)
+ goto drop;
+
+ if (__pim_rcv(mrt, skb, sizeof(*pim))) {
drop:
kfree_skb(skb);
}
@@ -1663,10 +1941,14 @@ int ipmr_get_route(struct net *net,
struct sk_buff *skb, struct rtmsg *rtm, int nowait)
{
int err;
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt;
struct mfc_cache *cache;
struct rtable *rt = skb_rtable(skb);
+ mrt = ipmr_get_table(net, RT_TABLE_DEFAULT);
+ if (mrt == NULL)
+ return -ENOENT;
+
read_lock(&mrt_lock);
cache = ipmr_cache_find(mrt, rt->rt_src, rt->rt_dst);
@@ -1717,6 +1999,7 @@ int ipmr_get_route(struct net *net,
*/
struct ipmr_vif_iter {
struct seq_net_private p;
+ struct mr_table *mrt;
int ct;
};
@@ -1724,7 +2007,7 @@ static struct vif_device *ipmr_vif_seq_idx(struct net *net,
struct ipmr_vif_iter *iter,
loff_t pos)
{
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt = iter->mrt;
for (iter->ct = 0; iter->ct < mrt->maxvif; ++iter->ct) {
if (!VIF_EXISTS(mrt, iter->ct))
@@ -1738,7 +2021,15 @@ static struct vif_device *ipmr_vif_seq_idx(struct net *net,
static void *ipmr_vif_seq_start(struct seq_file *seq, loff_t *pos)
__acquires(mrt_lock)
{
+ struct ipmr_vif_iter *iter = seq->private;
struct net *net = seq_file_net(seq);
+ struct mr_table *mrt;
+
+ mrt = ipmr_get_table(net, RT_TABLE_DEFAULT);
+ if (mrt == NULL)
+ return ERR_PTR(-ENOENT);
+
+ iter->mrt = mrt;
read_lock(&mrt_lock);
return *pos ? ipmr_vif_seq_idx(net, seq->private, *pos - 1)
@@ -1749,7 +2040,7 @@ static void *ipmr_vif_seq_next(struct seq_file *seq, void *v, loff_t *pos)
{
struct ipmr_vif_iter *iter = seq->private;
struct net *net = seq_file_net(seq);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt = iter->mrt;
++*pos;
if (v == SEQ_START_TOKEN)
@@ -1771,8 +2062,8 @@ static void ipmr_vif_seq_stop(struct seq_file *seq, void *v)
static int ipmr_vif_seq_show(struct seq_file *seq, void *v)
{
- struct net *net = seq_file_net(seq);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct ipmr_vif_iter *iter = seq->private;
+ struct mr_table *mrt = iter->mrt;
if (v == SEQ_START_TOKEN) {
seq_puts(seq,
@@ -1814,6 +2105,7 @@ static const struct file_operations ipmr_vif_fops = {
struct ipmr_mfc_iter {
struct seq_net_private p;
+ struct mr_table *mrt;
struct list_head *cache;
int ct;
};
@@ -1822,7 +2114,7 @@ struct ipmr_mfc_iter {
static struct mfc_cache *ipmr_mfc_seq_idx(struct net *net,
struct ipmr_mfc_iter *it, loff_t pos)
{
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt = it->mrt;
struct mfc_cache *mfc;
read_lock(&mrt_lock);
@@ -1850,7 +2142,13 @@ static void *ipmr_mfc_seq_start(struct seq_file *seq, loff_t *pos)
{
struct ipmr_mfc_iter *it = seq->private;
struct net *net = seq_file_net(seq);
+ struct mr_table *mrt;
+
+ mrt = ipmr_get_table(net, RT_TABLE_DEFAULT);
+ if (mrt == NULL)
+ return ERR_PTR(-ENOENT);
+ it->mrt = mrt;
it->cache = NULL;
it->ct = 0;
return *pos ? ipmr_mfc_seq_idx(net, seq->private, *pos - 1)
@@ -1862,7 +2160,7 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
struct mfc_cache *mfc = v;
struct ipmr_mfc_iter *it = seq->private;
struct net *net = seq_file_net(seq);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt = it->mrt;
++*pos;
@@ -1903,8 +2201,7 @@ static void *ipmr_mfc_seq_next(struct seq_file *seq, void *v, loff_t *pos)
static void ipmr_mfc_seq_stop(struct seq_file *seq, void *v)
{
struct ipmr_mfc_iter *it = seq->private;
- struct net *net = seq_file_net(seq);
- struct mr_table *mrt = net->ipv4.mrt;
+ struct mr_table *mrt = it->mrt;
if (it->cache == &mrt->mfc_unres_queue)
spin_unlock_bh(&mfc_unres_lock);
@@ -1915,8 +2212,6 @@ static void ipmr_mfc_seq_stop(struct seq_file *seq, void *v)
static int ipmr_mfc_seq_show(struct seq_file *seq, void *v)
{
int n;
- struct net *net = seq_file_net(seq);
- struct mr_table *mrt = net->ipv4.mrt;
if (v == SEQ_START_TOKEN) {
seq_puts(seq,
@@ -1924,6 +2219,7 @@ static int ipmr_mfc_seq_show(struct seq_file *seq, void *v)
} else {
const struct mfc_cache *mfc = v;
const struct ipmr_mfc_iter *it = seq->private;
+ const struct mr_table *mrt = it->mrt;
seq_printf(seq, "%08lX %08lX %-3hd",
(unsigned long) mfc->mfc_mcastgrp,
@@ -1989,28 +2285,11 @@ static const struct net_protocol pim_protocol = {
*/
static int __net_init ipmr_net_init(struct net *net)
{
- struct mr_table *mrt;
- unsigned int i;
- int err = 0;
+ int err;
- mrt = kzalloc(sizeof(*mrt), GFP_KERNEL);
- if (mrt == NULL) {
- err = -ENOMEM;
+ err = ipmr_rules_init(net);
+ if (err < 0)
goto fail;
- }
-
- /* Forwarding cache */
- for (i = 0; i < MFC_LINES; i++)
- INIT_LIST_HEAD(&mrt->mfc_cache_array[i]);
-
- INIT_LIST_HEAD(&mrt->mfc_unres_queue);
-
- setup_timer(&mrt->ipmr_expire_timer, ipmr_expire_process,
- (unsigned long)net);
-
-#ifdef CONFIG_IP_PIMSM
- mrt->mroute_reg_vif_num = -1;
-#endif
#ifdef CONFIG_PROC_FS
err = -ENOMEM;
@@ -2019,15 +2298,13 @@ static int __net_init ipmr_net_init(struct net *net)
if (!proc_net_fops_create(net, "ip_mr_cache", 0, &ipmr_mfc_fops))
goto proc_cache_fail;
#endif
-
- net->ipv4.mrt = mrt;
return 0;
#ifdef CONFIG_PROC_FS
proc_cache_fail:
proc_net_remove(net, "ip_mr_vif");
proc_vif_fail:
- kfree(mrt);
+ ipmr_rules_exit(net);
#endif
fail:
return err;
@@ -2039,7 +2316,7 @@ static void __net_exit ipmr_net_exit(struct net *net)
proc_net_remove(net, "ip_mr_cache");
proc_net_remove(net, "ip_mr_vif");
#endif
- kfree(net->ipv4.mrt);
+ ipmr_rules_exit(net);
}
static struct pernet_operations ipmr_net_ops = {
--
1.7.0.4
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox