2005-04-16 16:20:36 -06:00
|
|
|
/*
|
2007-02-09 07:24:47 -07:00
|
|
|
* Linux NET3: GRE over IP protocol decoder.
|
2005-04-16 16:20:36 -06:00
|
|
|
*
|
|
|
|
* Authors: Alexey Kuznetsov (kuznet@ms2.inr.ac.ru)
|
|
|
|
*
|
|
|
|
* This program is free software; you can redistribute it and/or
|
|
|
|
* modify it under the terms of the GNU General Public License
|
|
|
|
* as published by the Free Software Foundation; either version
|
|
|
|
* 2 of the License, or (at your option) any later version.
|
|
|
|
*
|
|
|
|
*/
|
|
|
|
|
2012-03-12 01:03:32 -06:00
|
|
|
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
|
|
|
|
|
2006-01-11 13:17:47 -07:00
|
|
|
#include <linux/capability.h>
|
2005-04-16 16:20:36 -06:00
|
|
|
#include <linux/module.h>
|
|
|
|
#include <linux/types.h>
|
|
|
|
#include <linux/kernel.h>
|
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h
percpu.h is included by sched.h and module.h and thus ends up being
included when building most .c files. percpu.h includes slab.h which
in turn includes gfp.h making everything defined by the two files
universally available and complicating inclusion dependencies.
percpu.h -> slab.h dependency is about to be removed. Prepare for
this change by updating users of gfp and slab facilities include those
headers directly instead of assuming availability. As this conversion
needs to touch large number of source files, the following script is
used as the basis of conversion.
http://userweb.kernel.org/~tj/misc/slabh-sweep.py
The script does the followings.
* Scan files for gfp and slab usages and update includes such that
only the necessary includes are there. ie. if only gfp is used,
gfp.h, if slab is used, slab.h.
* When the script inserts a new include, it looks at the include
blocks and try to put the new include such that its order conforms
to its surrounding. It's put in the include block which contains
core kernel includes, in the same order that the rest are ordered -
alphabetical, Christmas tree, rev-Xmas-tree or at the end if there
doesn't seem to be any matching order.
* If the script can't find a place to put a new include (mostly
because the file doesn't have fitting include block), it prints out
an error message indicating which .h file needs to be added to the
file.
The conversion was done in the following steps.
1. The initial automatic conversion of all .c files updated slightly
over 4000 files, deleting around 700 includes and adding ~480 gfp.h
and ~3000 slab.h inclusions. The script emitted errors for ~400
files.
2. Each error was manually checked. Some didn't need the inclusion,
some needed manual addition while adding it to implementation .h or
embedding .c file was more appropriate for others. This step added
inclusions to around 150 files.
3. The script was run again and the output was compared to the edits
from #2 to make sure no file was left behind.
4. Several build tests were done and a couple of problems were fixed.
e.g. lib/decompress_*.c used malloc/free() wrappers around slab
APIs requiring slab.h to be added manually.
5. The script was run on all .h files but without automatically
editing them as sprinkling gfp.h and slab.h inclusions around .h
files could easily lead to inclusion dependency hell. Most gfp.h
inclusion directives were ignored as stuff from gfp.h was usually
wildly available and often used in preprocessor macros. Each
slab.h inclusion directive was examined and added manually as
necessary.
6. percpu.h was updated not to include slab.h.
7. Build test were done on the following configurations and failures
were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my
distributed build env didn't work with gcov compiles) and a few
more options had to be turned off depending on archs to make things
build (like ipr on powerpc/64 which failed due to missing writeq).
* x86 and x86_64 UP and SMP allmodconfig and a custom test config.
* powerpc and powerpc64 SMP allmodconfig
* sparc and sparc64 SMP allmodconfig
* ia64 SMP allmodconfig
* s390 SMP allmodconfig
* alpha SMP allmodconfig
* um on x86_64 SMP allmodconfig
8. percpu.h modifications were reverted so that it could be applied as
a separate patch and serve as bisection point.
Given the fact that I had only a couple of failures from tests on step
6, I'm fairly confident about the coverage of this conversion patch.
If there is a breakage, it's likely to be something in one of the arch
headers which should be easily discoverable easily on most builds of
the specific arch.
Signed-off-by: Tejun Heo <tj@kernel.org>
Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 02:04:11 -06:00
|
|
|
#include <linux/slab.h>
|
2005-04-16 16:20:36 -06:00
|
|
|
#include <asm/uaccess.h>
|
|
|
|
#include <linux/skbuff.h>
|
|
|
|
#include <linux/netdevice.h>
|
|
|
|
#include <linux/in.h>
|
|
|
|
#include <linux/tcp.h>
|
|
|
|
#include <linux/udp.h>
|
|
|
|
#include <linux/if_arp.h>
|
|
|
|
#include <linux/mroute.h>
|
|
|
|
#include <linux/init.h>
|
|
|
|
#include <linux/in6.h>
|
|
|
|
#include <linux/inetdevice.h>
|
|
|
|
#include <linux/igmp.h>
|
|
|
|
#include <linux/netfilter_ipv4.h>
|
2008-10-09 13:00:17 -06:00
|
|
|
#include <linux/etherdevice.h>
|
2006-01-05 17:35:42 -07:00
|
|
|
#include <linux/if_ether.h>
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
#include <net/sock.h>
|
|
|
|
#include <net/ip.h>
|
|
|
|
#include <net/icmp.h>
|
|
|
|
#include <net/protocol.h>
|
|
|
|
#include <net/ipip.h>
|
|
|
|
#include <net/arp.h>
|
|
|
|
#include <net/checksum.h>
|
|
|
|
#include <net/dsfield.h>
|
|
|
|
#include <net/inet_ecn.h>
|
|
|
|
#include <net/xfrm.h>
|
2008-04-16 02:08:53 -06:00
|
|
|
#include <net/net_namespace.h>
|
|
|
|
#include <net/netns/generic.h>
|
2008-10-09 12:59:55 -06:00
|
|
|
#include <net/rtnetlink.h>
|
2010-08-22 00:05:39 -06:00
|
|
|
#include <net/gre.h>
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2011-12-10 02:48:31 -07:00
|
|
|
#if IS_ENABLED(CONFIG_IPV6)
|
2005-04-16 16:20:36 -06:00
|
|
|
#include <net/ipv6.h>
|
|
|
|
#include <net/ip6_fib.h>
|
|
|
|
#include <net/ip6_route.h>
|
|
|
|
#endif
|
|
|
|
|
|
|
|
/*
|
|
|
|
Problems & solutions
|
|
|
|
--------------------
|
|
|
|
|
|
|
|
1. The most important issue is detecting local dead loops.
|
|
|
|
They would cause complete host lockup in transmit, which
|
|
|
|
would be "resolved" by stack overflow or, if queueing is enabled,
|
|
|
|
with infinite looping in net_bh.
|
|
|
|
|
|
|
|
We cannot track such dead loops during route installation,
|
|
|
|
it is infeasible task. The most general solutions would be
|
|
|
|
to keep skb->encapsulation counter (sort of local ttl),
|
2010-09-30 00:35:10 -06:00
|
|
|
and silently drop packet when it expires. It is a good
|
2012-02-24 01:08:20 -07:00
|
|
|
solution, but it supposes maintaining new variable in ALL
|
2005-04-16 16:20:36 -06:00
|
|
|
skb, even if no tunneling is used.
|
|
|
|
|
2010-09-30 00:35:10 -06:00
|
|
|
Current solution: xmit_recursion breaks dead loops. This is a percpu
|
|
|
|
counter, since when we enter the first ndo_xmit(), cpu migration is
|
|
|
|
forbidden. We force an exit if this counter reaches RECURSION_LIMIT
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
2. Networking dead loops would not kill routers, but would really
|
|
|
|
kill network. IP hop limit plays role of "t->recursion" in this case,
|
|
|
|
if we copy it from packet being encapsulated to upper header.
|
|
|
|
It is very good solution, but it introduces two problems:
|
|
|
|
|
|
|
|
- Routing protocols, using packets with ttl=1 (OSPF, RIP2),
|
|
|
|
do not work over tunnels.
|
|
|
|
- traceroute does not work. I planned to relay ICMP from tunnel,
|
|
|
|
so that this problem would be solved and traceroute output
|
|
|
|
would even more informative. This idea appeared to be wrong:
|
|
|
|
only Linux complies to rfc1812 now (yes, guys, Linux is the only
|
|
|
|
true router now :-)), all routers (at least, in neighbourhood of mine)
|
|
|
|
return only 8 bytes of payload. It is the end.
|
|
|
|
|
|
|
|
Hence, if we want that OSPF worked or traceroute said something reasonable,
|
|
|
|
we should search for another solution.
|
|
|
|
|
|
|
|
One of them is to parse packet trying to detect inner encapsulation
|
|
|
|
made by our node. It is difficult or even impossible, especially,
|
2012-02-24 01:08:20 -07:00
|
|
|
taking into account fragmentation. TO be short, ttl is not solution at all.
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
Current solution: The solution was UNEXPECTEDLY SIMPLE.
|
|
|
|
We force DF flag on tunnels with preconfigured hop limit,
|
|
|
|
that is ALL. :-) Well, it does not remove the problem completely,
|
|
|
|
but exponential growth of network traffic is changed to linear
|
|
|
|
(branches, that exceed pmtu are pruned) and tunnel mtu
|
2012-02-24 01:08:20 -07:00
|
|
|
rapidly degrades to value <68, where looping stops.
|
2005-04-16 16:20:36 -06:00
|
|
|
Yes, it is not good if there exists a router in the loop,
|
|
|
|
which does not force DF, even when encapsulating packets have DF set.
|
|
|
|
But it is not our problem! Nobody could accuse us, we made
|
|
|
|
all that we could make. Even if it is your gated who injected
|
|
|
|
fatal route to network, even if it were you who configured
|
|
|
|
fatal static route: you are innocent. :-)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
3. Really, ipv4/ipip.c, ipv4/ip_gre.c and ipv6/sit.c contain
|
|
|
|
practically identical code. It would be good to glue them
|
|
|
|
together, but it is not very evident, how to make them modular.
|
|
|
|
sit is integral part of IPv6, ipip and gre are naturally modular.
|
|
|
|
We could extract common parts (hash table, ioctl etc)
|
|
|
|
to a separate module (ip_tunnel.c).
|
|
|
|
|
|
|
|
Alexey Kuznetsov.
|
|
|
|
*/
|
|
|
|
|
2012-09-25 05:02:48 -06:00
|
|
|
static bool log_ecn_error = true;
|
|
|
|
module_param(log_ecn_error, bool, 0644);
|
|
|
|
MODULE_PARM_DESC(log_ecn_error, "Log packets received with corrupted ECN");
|
|
|
|
|
2008-10-09 12:59:55 -06:00
|
|
|
static struct rtnl_link_ops ipgre_link_ops __read_mostly;
|
2005-04-16 16:20:36 -06:00
|
|
|
static int ipgre_tunnel_init(struct net_device *dev);
|
|
|
|
static void ipgre_tunnel_setup(struct net_device *dev);
|
2008-10-09 12:59:32 -06:00
|
|
|
static int ipgre_tunnel_bind_dev(struct net_device *dev);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
/* Fallback tunnel: no source, no destination, no key, no options */
|
|
|
|
|
2008-04-16 02:10:26 -06:00
|
|
|
#define HASH_SIZE 16
|
|
|
|
|
2009-11-17 03:42:49 -07:00
|
|
|
static int ipgre_net_id __read_mostly;
|
2008-04-16 02:08:53 -06:00
|
|
|
struct ipgre_net {
|
2010-09-15 05:07:53 -06:00
|
|
|
struct ip_tunnel __rcu *tunnels[4][HASH_SIZE];
|
2008-04-16 02:10:26 -06:00
|
|
|
|
2008-04-16 02:10:05 -06:00
|
|
|
struct net_device *fb_tunnel_dev;
|
2008-04-16 02:08:53 -06:00
|
|
|
};
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/* Tunnel hash table */
|
|
|
|
|
|
|
|
/*
|
|
|
|
4 hash tables:
|
|
|
|
|
|
|
|
3: (remote,local)
|
|
|
|
2: (remote,*)
|
|
|
|
1: (*,local)
|
|
|
|
0: (*,*)
|
|
|
|
|
|
|
|
We require exact key match i.e. if a key is present in packet
|
|
|
|
it will match only tunnel with the same key; if it is not present,
|
|
|
|
it will match only keyless tunnel.
|
|
|
|
|
|
|
|
All keysless packets, if not matched configured keyless tunnels
|
|
|
|
will match fallback tunnel.
|
|
|
|
*/
|
|
|
|
|
2006-11-08 01:23:14 -07:00
|
|
|
#define HASH(addr) (((__force u32)addr^((__force u32)addr>>4))&0xF)
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2008-04-16 02:10:26 -06:00
|
|
|
#define tunnels_r_l tunnels[3]
|
|
|
|
#define tunnels_r tunnels[2]
|
|
|
|
#define tunnels_l tunnels[1]
|
|
|
|
#define tunnels_wc tunnels[0]
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2012-04-12 00:31:16 -06:00
|
|
|
static struct rtnl_link_stats64 *ipgre_get_stats64(struct net_device *dev,
|
|
|
|
struct rtnl_link_stats64 *tot)
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
{
|
|
|
|
int i;
|
|
|
|
|
|
|
|
for_each_possible_cpu(i) {
|
|
|
|
const struct pcpu_tstats *tstats = per_cpu_ptr(dev->tstats, i);
|
2012-04-12 00:31:16 -06:00
|
|
|
u64 rx_packets, rx_bytes, tx_packets, tx_bytes;
|
|
|
|
unsigned int start;
|
|
|
|
|
|
|
|
do {
|
|
|
|
start = u64_stats_fetch_begin_bh(&tstats->syncp);
|
|
|
|
rx_packets = tstats->rx_packets;
|
|
|
|
tx_packets = tstats->tx_packets;
|
|
|
|
rx_bytes = tstats->rx_bytes;
|
|
|
|
tx_bytes = tstats->tx_bytes;
|
|
|
|
} while (u64_stats_fetch_retry_bh(&tstats->syncp, start));
|
|
|
|
|
|
|
|
tot->rx_packets += rx_packets;
|
|
|
|
tot->tx_packets += tx_packets;
|
|
|
|
tot->rx_bytes += rx_bytes;
|
|
|
|
tot->tx_bytes += tx_bytes;
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
}
|
2012-04-12 00:31:16 -06:00
|
|
|
|
|
|
|
tot->multicast = dev->stats.multicast;
|
|
|
|
tot->rx_crc_errors = dev->stats.rx_crc_errors;
|
|
|
|
tot->rx_fifo_errors = dev->stats.rx_fifo_errors;
|
|
|
|
tot->rx_length_errors = dev->stats.rx_length_errors;
|
2012-09-25 05:02:48 -06:00
|
|
|
tot->rx_frame_errors = dev->stats.rx_frame_errors;
|
2012-04-12 00:31:16 -06:00
|
|
|
tot->rx_errors = dev->stats.rx_errors;
|
2012-09-25 05:02:48 -06:00
|
|
|
|
2012-04-12 00:31:16 -06:00
|
|
|
tot->tx_fifo_errors = dev->stats.tx_fifo_errors;
|
|
|
|
tot->tx_carrier_errors = dev->stats.tx_carrier_errors;
|
|
|
|
tot->tx_dropped = dev->stats.tx_dropped;
|
|
|
|
tot->tx_aborted_errors = dev->stats.tx_aborted_errors;
|
|
|
|
tot->tx_errors = dev->stats.tx_errors;
|
|
|
|
|
|
|
|
return tot;
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
}
|
|
|
|
|
2012-09-24 12:12:23 -06:00
|
|
|
/* Does key in tunnel parameters match packet */
|
|
|
|
static bool ipgre_key_match(const struct ip_tunnel_parm *p,
|
2012-09-30 23:21:14 -06:00
|
|
|
__be16 flags, __be32 key)
|
2012-09-24 12:12:23 -06:00
|
|
|
{
|
|
|
|
if (p->i_flags & GRE_KEY) {
|
|
|
|
if (flags & GRE_KEY)
|
|
|
|
return key == p->i_key;
|
|
|
|
else
|
|
|
|
return false; /* key expected, none present */
|
|
|
|
} else
|
|
|
|
return !(flags & GRE_KEY);
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/* Given src, dst and key, find appropriate for input tunnel. */
|
|
|
|
|
2012-04-14 19:34:41 -06:00
|
|
|
static struct ip_tunnel *ipgre_tunnel_lookup(struct net_device *dev,
|
|
|
|
__be32 remote, __be32 local,
|
2012-09-30 23:21:14 -06:00
|
|
|
__be16 flags, __be32 key,
|
2012-09-24 12:12:23 -06:00
|
|
|
__be16 gre_proto)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2009-01-19 18:22:12 -07:00
|
|
|
struct net *net = dev_net(dev);
|
|
|
|
int link = dev->ifindex;
|
2010-09-15 05:07:53 -06:00
|
|
|
unsigned int h0 = HASH(remote);
|
|
|
|
unsigned int h1 = HASH(key);
|
2009-01-26 21:56:10 -07:00
|
|
|
struct ip_tunnel *t, *cand = NULL;
|
2008-04-16 02:10:05 -06:00
|
|
|
struct ipgre_net *ign = net_generic(net, ipgre_net_id);
|
2008-10-09 13:00:17 -06:00
|
|
|
int dev_type = (gre_proto == htons(ETH_P_TEB)) ?
|
|
|
|
ARPHRD_ETHER : ARPHRD_IPGRE;
|
2009-01-26 21:56:10 -07:00
|
|
|
int score, cand_score = 4;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2012-11-11 14:52:34 -07:00
|
|
|
for_each_ip_tunnel_rcu(t, ign->tunnels_r_l[h0 ^ h1]) {
|
2009-01-19 18:22:12 -07:00
|
|
|
if (local != t->parms.iph.saddr ||
|
|
|
|
remote != t->parms.iph.daddr ||
|
|
|
|
!(t->dev->flags & IFF_UP))
|
|
|
|
continue;
|
|
|
|
|
2012-09-24 12:12:23 -06:00
|
|
|
if (!ipgre_key_match(&t->parms, flags, key))
|
|
|
|
continue;
|
|
|
|
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->dev->type != ARPHRD_IPGRE &&
|
|
|
|
t->dev->type != dev_type)
|
|
|
|
continue;
|
|
|
|
|
2009-01-26 21:56:10 -07:00
|
|
|
score = 0;
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->parms.link != link)
|
2009-01-26 21:56:10 -07:00
|
|
|
score |= 1;
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->dev->type != dev_type)
|
2009-01-26 21:56:10 -07:00
|
|
|
score |= 2;
|
|
|
|
if (score == 0)
|
2009-01-19 18:22:12 -07:00
|
|
|
return t;
|
2009-01-26 21:56:10 -07:00
|
|
|
|
|
|
|
if (score < cand_score) {
|
|
|
|
cand = t;
|
|
|
|
cand_score = score;
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
2008-10-09 13:00:17 -06:00
|
|
|
|
2012-11-11 14:52:34 -07:00
|
|
|
for_each_ip_tunnel_rcu(t, ign->tunnels_r[h0 ^ h1]) {
|
2009-01-19 18:22:12 -07:00
|
|
|
if (remote != t->parms.iph.daddr ||
|
|
|
|
!(t->dev->flags & IFF_UP))
|
|
|
|
continue;
|
|
|
|
|
2012-09-24 12:12:23 -06:00
|
|
|
if (!ipgre_key_match(&t->parms, flags, key))
|
|
|
|
continue;
|
|
|
|
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->dev->type != ARPHRD_IPGRE &&
|
|
|
|
t->dev->type != dev_type)
|
|
|
|
continue;
|
|
|
|
|
2009-01-26 21:56:10 -07:00
|
|
|
score = 0;
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->parms.link != link)
|
2009-01-26 21:56:10 -07:00
|
|
|
score |= 1;
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->dev->type != dev_type)
|
2009-01-26 21:56:10 -07:00
|
|
|
score |= 2;
|
|
|
|
if (score == 0)
|
2009-01-19 18:22:12 -07:00
|
|
|
return t;
|
2009-01-26 21:56:10 -07:00
|
|
|
|
|
|
|
if (score < cand_score) {
|
|
|
|
cand = t;
|
|
|
|
cand_score = score;
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
2008-10-09 13:00:17 -06:00
|
|
|
|
2012-11-11 14:52:34 -07:00
|
|
|
for_each_ip_tunnel_rcu(t, ign->tunnels_l[h1]) {
|
2009-01-19 18:22:12 -07:00
|
|
|
if ((local != t->parms.iph.saddr &&
|
|
|
|
(local != t->parms.iph.daddr ||
|
|
|
|
!ipv4_is_multicast(local))) ||
|
|
|
|
!(t->dev->flags & IFF_UP))
|
|
|
|
continue;
|
|
|
|
|
2012-09-24 12:12:23 -06:00
|
|
|
if (!ipgre_key_match(&t->parms, flags, key))
|
|
|
|
continue;
|
|
|
|
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->dev->type != ARPHRD_IPGRE &&
|
|
|
|
t->dev->type != dev_type)
|
|
|
|
continue;
|
|
|
|
|
2009-01-26 21:56:10 -07:00
|
|
|
score = 0;
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->parms.link != link)
|
2009-01-26 21:56:10 -07:00
|
|
|
score |= 1;
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->dev->type != dev_type)
|
2009-01-26 21:56:10 -07:00
|
|
|
score |= 2;
|
|
|
|
if (score == 0)
|
2009-01-19 18:22:12 -07:00
|
|
|
return t;
|
2009-01-26 21:56:10 -07:00
|
|
|
|
|
|
|
if (score < cand_score) {
|
|
|
|
cand = t;
|
|
|
|
cand_score = score;
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
2008-10-09 13:00:17 -06:00
|
|
|
|
2012-11-11 14:52:34 -07:00
|
|
|
for_each_ip_tunnel_rcu(t, ign->tunnels_wc[h1]) {
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->parms.i_key != key ||
|
|
|
|
!(t->dev->flags & IFF_UP))
|
|
|
|
continue;
|
|
|
|
|
|
|
|
if (t->dev->type != ARPHRD_IPGRE &&
|
|
|
|
t->dev->type != dev_type)
|
|
|
|
continue;
|
|
|
|
|
2009-01-26 21:56:10 -07:00
|
|
|
score = 0;
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->parms.link != link)
|
2009-01-26 21:56:10 -07:00
|
|
|
score |= 1;
|
2009-01-19 18:22:12 -07:00
|
|
|
if (t->dev->type != dev_type)
|
2009-01-26 21:56:10 -07:00
|
|
|
score |= 2;
|
|
|
|
if (score == 0)
|
2009-01-19 18:22:12 -07:00
|
|
|
return t;
|
2009-01-26 21:56:10 -07:00
|
|
|
|
|
|
|
if (score < cand_score) {
|
|
|
|
cand = t;
|
|
|
|
cand_score = score;
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2009-01-26 21:56:10 -07:00
|
|
|
if (cand != NULL)
|
|
|
|
return cand;
|
2008-10-09 13:00:17 -06:00
|
|
|
|
2009-10-23 00:14:38 -06:00
|
|
|
dev = ign->fb_tunnel_dev;
|
|
|
|
if (dev->flags & IFF_UP)
|
|
|
|
return netdev_priv(dev);
|
2009-01-19 18:22:12 -07:00
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
2010-09-15 05:07:53 -06:00
|
|
|
static struct ip_tunnel __rcu **__ipgre_bucket(struct ipgre_net *ign,
|
2008-04-16 02:09:22 -06:00
|
|
|
struct ip_tunnel_parm *parms)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2007-04-24 05:44:48 -06:00
|
|
|
__be32 remote = parms->iph.daddr;
|
|
|
|
__be32 local = parms->iph.saddr;
|
|
|
|
__be32 key = parms->i_key;
|
2010-09-15 05:07:53 -06:00
|
|
|
unsigned int h = HASH(key);
|
2005-04-16 16:20:36 -06:00
|
|
|
int prio = 0;
|
|
|
|
|
|
|
|
if (local)
|
|
|
|
prio |= 1;
|
2007-12-16 14:45:43 -07:00
|
|
|
if (remote && !ipv4_is_multicast(remote)) {
|
2005-04-16 16:20:36 -06:00
|
|
|
prio |= 2;
|
|
|
|
h ^= HASH(remote);
|
|
|
|
}
|
|
|
|
|
2008-04-16 02:10:26 -06:00
|
|
|
return &ign->tunnels[prio][h];
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2010-09-15 05:07:53 -06:00
|
|
|
static inline struct ip_tunnel __rcu **ipgre_bucket(struct ipgre_net *ign,
|
2008-04-16 02:09:22 -06:00
|
|
|
struct ip_tunnel *t)
|
2007-04-24 05:44:48 -06:00
|
|
|
{
|
2008-04-16 02:09:22 -06:00
|
|
|
return __ipgre_bucket(ign, &t->parms);
|
2007-04-24 05:44:48 -06:00
|
|
|
}
|
|
|
|
|
2008-04-16 02:09:22 -06:00
|
|
|
static void ipgre_tunnel_link(struct ipgre_net *ign, struct ip_tunnel *t)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2010-09-15 05:07:53 -06:00
|
|
|
struct ip_tunnel __rcu **tp = ipgre_bucket(ign, t);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2010-09-15 05:07:53 -06:00
|
|
|
rcu_assign_pointer(t->next, rtnl_dereference(*tp));
|
2009-10-23 00:14:38 -06:00
|
|
|
rcu_assign_pointer(*tp, t);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2008-04-16 02:09:22 -06:00
|
|
|
static void ipgre_tunnel_unlink(struct ipgre_net *ign, struct ip_tunnel *t)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2010-09-15 05:07:53 -06:00
|
|
|
struct ip_tunnel __rcu **tp;
|
|
|
|
struct ip_tunnel *iter;
|
|
|
|
|
|
|
|
for (tp = ipgre_bucket(ign, t);
|
|
|
|
(iter = rtnl_dereference(*tp)) != NULL;
|
|
|
|
tp = &iter->next) {
|
|
|
|
if (t == iter) {
|
|
|
|
rcu_assign_pointer(*tp, t->next);
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
static struct ip_tunnel *ipgre_tunnel_find(struct net *net,
|
|
|
|
struct ip_tunnel_parm *parms,
|
|
|
|
int type)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2006-11-08 01:23:14 -07:00
|
|
|
__be32 remote = parms->iph.daddr;
|
|
|
|
__be32 local = parms->iph.saddr;
|
|
|
|
__be32 key = parms->i_key;
|
2009-01-19 18:22:12 -07:00
|
|
|
int link = parms->link;
|
2010-09-15 05:07:53 -06:00
|
|
|
struct ip_tunnel *t;
|
|
|
|
struct ip_tunnel __rcu **tp;
|
2008-10-09 13:00:17 -06:00
|
|
|
struct ipgre_net *ign = net_generic(net, ipgre_net_id);
|
|
|
|
|
2010-09-15 05:07:53 -06:00
|
|
|
for (tp = __ipgre_bucket(ign, parms);
|
|
|
|
(t = rtnl_dereference(*tp)) != NULL;
|
|
|
|
tp = &t->next)
|
2008-10-09 13:00:17 -06:00
|
|
|
if (local == t->parms.iph.saddr &&
|
|
|
|
remote == t->parms.iph.daddr &&
|
|
|
|
key == t->parms.i_key &&
|
2009-01-19 18:22:12 -07:00
|
|
|
link == t->parms.link &&
|
2008-10-09 13:00:17 -06:00
|
|
|
type == t->dev->type)
|
|
|
|
break;
|
|
|
|
|
|
|
|
return t;
|
|
|
|
}
|
|
|
|
|
2010-09-15 05:07:53 -06:00
|
|
|
static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
|
2008-10-09 13:00:17 -06:00
|
|
|
struct ip_tunnel_parm *parms, int create)
|
|
|
|
{
|
|
|
|
struct ip_tunnel *t, *nt;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct net_device *dev;
|
|
|
|
char name[IFNAMSIZ];
|
2008-04-16 02:09:22 -06:00
|
|
|
struct ipgre_net *ign = net_generic(net, ipgre_net_id);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
t = ipgre_tunnel_find(net, parms, ARPHRD_IPGRE);
|
|
|
|
if (t || !create)
|
|
|
|
return t;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (parms->name[0])
|
|
|
|
strlcpy(name, parms->name, IFNAMSIZ);
|
2008-02-23 21:19:20 -07:00
|
|
|
else
|
2010-11-29 02:47:47 -07:00
|
|
|
strcpy(name, "gre%d");
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
|
|
|
|
if (!dev)
|
2010-11-29 02:47:47 -07:00
|
|
|
return NULL;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2008-04-16 02:11:13 -06:00
|
|
|
dev_net_set(dev, net);
|
|
|
|
|
2006-01-08 23:05:26 -07:00
|
|
|
nt = netdev_priv(dev);
|
2005-04-16 16:20:36 -06:00
|
|
|
nt->parms = *parms;
|
2008-10-09 12:59:55 -06:00
|
|
|
dev->rtnl_link_ops = &ipgre_link_ops;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2008-10-09 12:59:32 -06:00
|
|
|
dev->mtu = ipgre_tunnel_bind_dev(dev);
|
|
|
|
|
[INET]: Don't create tunnels with '%' in name.
Four tunnel drivers (ip_gre, ipip, ip6_tunnel and sit) can receive a
pre-defined name for a device from the userspace. Since these drivers
call the register_netdevice() (rtnl_lock, is held), which does _not_
generate the device's name, this name may contain a '%' character.
Not sure how bad is this to have a device with a '%' in its name, but
all the other places either use the register_netdev(), which call the
dev_alloc_name(), or explicitly call the dev_alloc_name() before
registering, i.e. do not allow for such names.
This had to be prior to the commit 34cc7b, but I forgot to number the
patches and this one got lost, sorry.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-27 00:51:04 -07:00
|
|
|
if (register_netdevice(dev) < 0)
|
|
|
|
goto failed_free;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
ipv6: Fix ip_gre lockless xmits.
Tunnel devices set NETIF_F_LLTX to bypass HARD_TX_LOCK. Sit and
ipip set this unconditionally in ops->setup, but gre enables it
conditionally after parameter passing in ops->newlink. This is
not called during tunnel setup as below, however, so GRE tunnels are
still taking the lock.
modprobe ip_gre
ip tunnel add test0 mode gre remote 10.5.1.1 dev lo
ip link set test0 up
ip addr add 10.6.0.1 dev test0
# cat /sys/class/net/test0/features
# $DIR/test_tunnel_xmit 10 10.5.2.1
ip route add 10.5.2.0/24 dev test0
ip tunnel del test0
The newlink callback is only called in rtnl_netlink, and only if
the device is new, as it calls register_netdevice internally. Gre
tunnels are created at 'ip tunnel add' with ioctl SIOCADDTUNNEL,
which calls ipgre_tunnel_locate, which calls register_netdev.
rtnl_newlink is called at 'ip link set', but skips ops->newlink
and the device is up with locking still enabled. The equivalent
ipip tunnel works fine, btw (just substitute 'method gre' for
'method ipip').
On kernels before /sys/class/net/*/features was removed [1],
the first commented out line returns 0x6000 with method gre,
which indicates that NETIF_F_LLTX (0x1000) is not set. With ipip,
it reports 0x7000. This test cannot be used on recent kernels where
the sysfs file is removed (and ETHTOOL_GFEATURES does not currently
work for tunnel devices, because they lack dev->ethtool_ops).
The second commented out line calls a simple transmission test [2]
that sends on 24 cores at maximum rate. Results of a single run:
ipip: 19,372,306
gre before patch: 4,839,753
gre after patch: 19,133,873
This patch replicates the condition check in ipgre_newlink to
ipgre_tunnel_locate. It works for me, both with oseq on and off.
This is the first time I looked at rtnetlink and iproute2 code,
though, so someone more knowledgeable should probably check the
patch. Thanks.
The tail of both functions is now identical, by the way. To avoid
code duplication, I'll be happy to rework this and merge the two.
[1] http://patchwork.ozlabs.org/patch/104610/
[2] http://kernel.googlecode.com/files/xmit_udp_parallel.c
Signed-off-by: Willem de Bruijn <willemb@google.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-01-26 03:34:35 -07:00
|
|
|
/* Can use a lockless transmit, unless we generate output sequences */
|
|
|
|
if (!(nt->parms.o_flags & GRE_SEQ))
|
|
|
|
dev->features |= NETIF_F_LLTX;
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
dev_hold(dev);
|
2008-04-16 02:09:22 -06:00
|
|
|
ipgre_tunnel_link(ign, nt);
|
2005-04-16 16:20:36 -06:00
|
|
|
return nt;
|
|
|
|
|
[INET]: Don't create tunnels with '%' in name.
Four tunnel drivers (ip_gre, ipip, ip6_tunnel and sit) can receive a
pre-defined name for a device from the userspace. Since these drivers
call the register_netdevice() (rtnl_lock, is held), which does _not_
generate the device's name, this name may contain a '%' character.
Not sure how bad is this to have a device with a '%' in its name, but
all the other places either use the register_netdev(), which call the
dev_alloc_name(), or explicitly call the dev_alloc_name() before
registering, i.e. do not allow for such names.
This had to be prior to the commit 34cc7b, but I forgot to number the
patches and this one got lost, sorry.
Signed-off-by: Pavel Emelyanov <xemul@openvz.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2008-02-27 00:51:04 -07:00
|
|
|
failed_free:
|
|
|
|
free_netdev(dev);
|
2005-04-16 16:20:36 -06:00
|
|
|
return NULL;
|
|
|
|
}
|
|
|
|
|
|
|
|
static void ipgre_tunnel_uninit(struct net_device *dev)
|
|
|
|
{
|
2008-04-16 02:09:22 -06:00
|
|
|
struct net *net = dev_net(dev);
|
|
|
|
struct ipgre_net *ign = net_generic(net, ipgre_net_id);
|
|
|
|
|
|
|
|
ipgre_tunnel_unlink(ign, netdev_priv(dev));
|
2005-04-16 16:20:36 -06:00
|
|
|
dev_put(dev);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
static void ipgre_err(struct sk_buff *skb, u32 info)
|
|
|
|
{
|
|
|
|
|
2008-05-21 18:47:54 -06:00
|
|
|
/* All the routers (except for Linux) return only
|
2005-04-16 16:20:36 -06:00
|
|
|
8 bytes of packet payload. It means, that precise relaying of
|
|
|
|
ICMP in the real Internet is absolutely infeasible.
|
|
|
|
|
|
|
|
Moreover, Cisco "wise men" put GRE key to the third word
|
|
|
|
in GRE header. It makes impossible maintaining even soft state for keyed
|
|
|
|
GRE tunnels with enabled checksum. Tell them "thank you".
|
|
|
|
|
|
|
|
Well, I wonder, rfc1812 was written by Cisco employee,
|
2012-02-24 01:08:20 -07:00
|
|
|
what the hell these idiots break standards established
|
|
|
|
by themselves???
|
2005-04-16 16:20:36 -06:00
|
|
|
*/
|
|
|
|
|
2011-04-21 22:53:02 -06:00
|
|
|
const struct iphdr *iph = (const struct iphdr *)skb->data;
|
2012-04-14 19:34:41 -06:00
|
|
|
__be16 *p = (__be16 *)(skb->data+(iph->ihl<<2));
|
2005-04-16 16:20:36 -06:00
|
|
|
int grehlen = (iph->ihl<<2) + 4;
|
2007-03-13 11:43:18 -06:00
|
|
|
const int type = icmp_hdr(skb)->type;
|
|
|
|
const int code = icmp_hdr(skb)->code;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct ip_tunnel *t;
|
2006-11-08 01:23:14 -07:00
|
|
|
__be16 flags;
|
2012-09-24 12:12:23 -06:00
|
|
|
__be32 key = 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
flags = p[0];
|
|
|
|
if (flags&(GRE_CSUM|GRE_KEY|GRE_SEQ|GRE_ROUTING|GRE_VERSION)) {
|
|
|
|
if (flags&(GRE_VERSION|GRE_ROUTING))
|
|
|
|
return;
|
|
|
|
if (flags&GRE_KEY) {
|
|
|
|
grehlen += 4;
|
|
|
|
if (flags&GRE_CSUM)
|
|
|
|
grehlen += 4;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
/* If only 8 bytes returned, keyed message will be dropped here */
|
|
|
|
if (skb_headlen(skb) < grehlen)
|
|
|
|
return;
|
|
|
|
|
2012-09-24 12:12:23 -06:00
|
|
|
if (flags & GRE_KEY)
|
|
|
|
key = *(((__be32 *)p) + (grehlen / 4) - 1);
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
switch (type) {
|
|
|
|
default:
|
|
|
|
case ICMP_PARAMETERPROB:
|
|
|
|
return;
|
|
|
|
|
|
|
|
case ICMP_DEST_UNREACH:
|
|
|
|
switch (code) {
|
|
|
|
case ICMP_SR_FAILED:
|
|
|
|
case ICMP_PORT_UNREACH:
|
|
|
|
/* Impossible event. */
|
|
|
|
return;
|
|
|
|
default:
|
|
|
|
/* All others are translated to HOST_UNREACH.
|
|
|
|
rfc2003 contains "deep thoughts" about NET_UNREACH,
|
|
|
|
I believe they are just ether pollution. --ANK
|
|
|
|
*/
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
break;
|
|
|
|
case ICMP_TIME_EXCEEDED:
|
|
|
|
if (code != ICMP_EXC_TTL)
|
|
|
|
return;
|
|
|
|
break;
|
2012-07-11 22:27:49 -06:00
|
|
|
|
|
|
|
case ICMP_REDIRECT:
|
|
|
|
break;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2009-01-19 18:22:12 -07:00
|
|
|
t = ipgre_tunnel_lookup(skb->dev, iph->daddr, iph->saddr,
|
2012-09-24 12:12:23 -06:00
|
|
|
flags, key, p[1]);
|
|
|
|
|
2012-06-14 23:21:46 -06:00
|
|
|
if (t == NULL)
|
2012-09-24 12:12:24 -06:00
|
|
|
return;
|
2012-06-14 23:21:46 -06:00
|
|
|
|
|
|
|
if (type == ICMP_DEST_UNREACH && code == ICMP_FRAG_NEEDED) {
|
|
|
|
ipv4_update_pmtu(skb, dev_net(skb->dev), info,
|
|
|
|
t->parms.link, 0, IPPROTO_GRE, 0);
|
2012-09-24 12:12:24 -06:00
|
|
|
return;
|
2012-06-14 23:21:46 -06:00
|
|
|
}
|
2012-07-11 22:27:49 -06:00
|
|
|
if (type == ICMP_REDIRECT) {
|
|
|
|
ipv4_redirect(skb, dev_net(skb->dev), t->parms.link, 0,
|
|
|
|
IPPROTO_GRE, 0);
|
2012-09-24 12:12:24 -06:00
|
|
|
return;
|
2012-07-11 22:27:49 -06:00
|
|
|
}
|
2012-06-14 23:21:46 -06:00
|
|
|
if (t->parms.iph.daddr == 0 ||
|
2007-12-16 14:45:43 -07:00
|
|
|
ipv4_is_multicast(t->parms.iph.daddr))
|
2012-09-24 12:12:24 -06:00
|
|
|
return;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (t->parms.iph.ttl == 0 && type == ICMP_TIME_EXCEEDED)
|
2012-09-24 12:12:24 -06:00
|
|
|
return;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2009-02-25 00:34:48 -07:00
|
|
|
if (time_before(jiffies, t->err_time + IPTUNNEL_ERR_TIMEO))
|
2005-04-16 16:20:36 -06:00
|
|
|
t->err_count++;
|
|
|
|
else
|
|
|
|
t->err_count = 1;
|
|
|
|
t->err_time = jiffies;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline u8
|
2011-04-21 22:53:02 -06:00
|
|
|
ipgre_ecn_encapsulate(u8 tos, const struct iphdr *old_iph, struct sk_buff *skb)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
|
|
|
u8 inner = 0;
|
|
|
|
if (skb->protocol == htons(ETH_P_IP))
|
|
|
|
inner = old_iph->tos;
|
|
|
|
else if (skb->protocol == htons(ETH_P_IPV6))
|
2011-04-21 22:53:02 -06:00
|
|
|
inner = ipv6_get_dsfield((const struct ipv6hdr *)old_iph);
|
2005-04-16 16:20:36 -06:00
|
|
|
return INET_ECN_encapsulate(tos, inner);
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ipgre_rcv(struct sk_buff *skb)
|
|
|
|
{
|
2011-04-21 22:53:02 -06:00
|
|
|
const struct iphdr *iph;
|
2005-04-16 16:20:36 -06:00
|
|
|
u8 *h;
|
2006-11-08 01:23:14 -07:00
|
|
|
__be16 flags;
|
2006-11-14 22:24:49 -07:00
|
|
|
__sum16 csum = 0;
|
2006-11-08 01:23:14 -07:00
|
|
|
__be32 key = 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
u32 seqno = 0;
|
|
|
|
struct ip_tunnel *tunnel;
|
|
|
|
int offset = 4;
|
2008-10-09 13:00:17 -06:00
|
|
|
__be16 gre_proto;
|
2012-09-25 05:02:48 -06:00
|
|
|
int err;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (!pskb_may_pull(skb, 16))
|
2012-09-24 12:12:24 -06:00
|
|
|
goto drop;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-04-20 23:47:35 -06:00
|
|
|
iph = ip_hdr(skb);
|
2005-04-16 16:20:36 -06:00
|
|
|
h = skb->data;
|
2012-04-14 19:34:41 -06:00
|
|
|
flags = *(__be16 *)h;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (flags&(GRE_CSUM|GRE_KEY|GRE_ROUTING|GRE_SEQ|GRE_VERSION)) {
|
|
|
|
/* - Version must be 0.
|
|
|
|
- We do not support routing headers.
|
|
|
|
*/
|
|
|
|
if (flags&(GRE_VERSION|GRE_ROUTING))
|
2012-09-24 12:12:24 -06:00
|
|
|
goto drop;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (flags&GRE_CSUM) {
|
2005-11-10 14:01:24 -07:00
|
|
|
switch (skb->ip_summed) {
|
2006-08-29 17:44:56 -06:00
|
|
|
case CHECKSUM_COMPLETE:
|
2006-11-14 22:24:49 -07:00
|
|
|
csum = csum_fold(skb->csum);
|
2005-11-10 14:01:24 -07:00
|
|
|
if (!csum)
|
|
|
|
break;
|
|
|
|
/* fall through */
|
|
|
|
case CHECKSUM_NONE:
|
|
|
|
skb->csum = 0;
|
|
|
|
csum = __skb_checksum_complete(skb);
|
2006-08-29 17:44:56 -06:00
|
|
|
skb->ip_summed = CHECKSUM_COMPLETE;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
offset += 4;
|
|
|
|
}
|
|
|
|
if (flags&GRE_KEY) {
|
2012-04-14 19:34:41 -06:00
|
|
|
key = *(__be32 *)(h + offset);
|
2005-04-16 16:20:36 -06:00
|
|
|
offset += 4;
|
|
|
|
}
|
|
|
|
if (flags&GRE_SEQ) {
|
2012-04-14 19:34:41 -06:00
|
|
|
seqno = ntohl(*(__be32 *)(h + offset));
|
2005-04-16 16:20:36 -06:00
|
|
|
offset += 4;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
gre_proto = *(__be16 *)(h + 2);
|
|
|
|
|
2012-09-24 12:12:23 -06:00
|
|
|
tunnel = ipgre_tunnel_lookup(skb->dev,
|
|
|
|
iph->saddr, iph->daddr, flags, key,
|
|
|
|
gre_proto);
|
|
|
|
if (tunnel) {
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
struct pcpu_tstats *tstats;
|
2008-05-21 15:14:22 -06:00
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
secpath_reset(skb);
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
skb->protocol = gre_proto;
|
2005-04-16 16:20:36 -06:00
|
|
|
/* WCCP version 1 and 2 protocol decoding.
|
|
|
|
* - Change protocol to IP
|
|
|
|
* - When dealing with WCCPv2, Skip extra 4 bytes in GRE header
|
|
|
|
*/
|
2008-10-09 13:00:17 -06:00
|
|
|
if (flags == 0 && gre_proto == htons(ETH_P_WCCP)) {
|
2006-10-10 20:41:21 -06:00
|
|
|
skb->protocol = htons(ETH_P_IP);
|
2007-02-09 07:24:47 -07:00
|
|
|
if ((*(h + offset) & 0xF0) != 0x40)
|
2005-04-16 16:20:36 -06:00
|
|
|
offset += 4;
|
|
|
|
}
|
|
|
|
|
2007-12-20 01:10:33 -07:00
|
|
|
skb->mac_header = skb->network_header;
|
2007-03-10 14:42:03 -07:00
|
|
|
__pskb_pull(skb, offset);
|
2007-04-25 19:04:18 -06:00
|
|
|
skb_postpull_rcsum(skb, skb_transport_header(skb), offset);
|
2005-04-16 16:20:36 -06:00
|
|
|
skb->pkt_type = PACKET_HOST;
|
|
|
|
#ifdef CONFIG_NET_IPGRE_BROADCAST
|
2007-12-16 14:45:43 -07:00
|
|
|
if (ipv4_is_multicast(iph->daddr)) {
|
2005-04-16 16:20:36 -06:00
|
|
|
/* Looped back packet, drop it! */
|
2010-11-11 18:07:48 -07:00
|
|
|
if (rt_is_output_route(skb_rtable(skb)))
|
2005-04-16 16:20:36 -06:00
|
|
|
goto drop;
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
tunnel->dev->stats.multicast++;
|
2005-04-16 16:20:36 -06:00
|
|
|
skb->pkt_type = PACKET_BROADCAST;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
if (((flags&GRE_CSUM) && csum) ||
|
|
|
|
(!(flags&GRE_CSUM) && tunnel->parms.i_flags&GRE_CSUM)) {
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
tunnel->dev->stats.rx_crc_errors++;
|
|
|
|
tunnel->dev->stats.rx_errors++;
|
2005-04-16 16:20:36 -06:00
|
|
|
goto drop;
|
|
|
|
}
|
|
|
|
if (tunnel->parms.i_flags&GRE_SEQ) {
|
|
|
|
if (!(flags&GRE_SEQ) ||
|
|
|
|
(tunnel->i_seqno && (s32)(seqno - tunnel->i_seqno) < 0)) {
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
tunnel->dev->stats.rx_fifo_errors++;
|
|
|
|
tunnel->dev->stats.rx_errors++;
|
2005-04-16 16:20:36 -06:00
|
|
|
goto drop;
|
|
|
|
}
|
|
|
|
tunnel->i_seqno = seqno + 1;
|
|
|
|
}
|
2008-10-09 13:00:17 -06:00
|
|
|
|
|
|
|
/* Warning: All skb pointers will be invalidated! */
|
|
|
|
if (tunnel->dev->type == ARPHRD_ETHER) {
|
|
|
|
if (!pskb_may_pull(skb, ETH_HLEN)) {
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
tunnel->dev->stats.rx_length_errors++;
|
|
|
|
tunnel->dev->stats.rx_errors++;
|
2008-10-09 13:00:17 -06:00
|
|
|
goto drop;
|
|
|
|
}
|
|
|
|
|
|
|
|
iph = ip_hdr(skb);
|
|
|
|
skb->protocol = eth_type_trans(skb, tunnel->dev);
|
|
|
|
skb_postpull_rcsum(skb, eth_hdr(skb), ETH_HLEN);
|
|
|
|
}
|
|
|
|
|
2012-09-25 05:02:48 -06:00
|
|
|
__skb_tunnel_rx(skb, tunnel->dev);
|
|
|
|
|
|
|
|
skb_reset_network_header(skb);
|
|
|
|
err = IP_ECN_decapsulate(iph, skb);
|
|
|
|
if (unlikely(err)) {
|
|
|
|
if (log_ecn_error)
|
|
|
|
net_info_ratelimited("non-ECT from %pI4 with TOS=%#x\n",
|
|
|
|
&iph->saddr, iph->tos);
|
|
|
|
if (err > 1) {
|
|
|
|
++tunnel->dev->stats.rx_frame_errors;
|
|
|
|
++tunnel->dev->stats.rx_errors;
|
|
|
|
goto drop;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
tstats = this_cpu_ptr(tunnel->dev->tstats);
|
2012-04-12 00:31:16 -06:00
|
|
|
u64_stats_update_begin(&tstats->syncp);
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
tstats->rx_packets++;
|
|
|
|
tstats->rx_bytes += skb->len;
|
2012-04-12 00:31:16 -06:00
|
|
|
u64_stats_update_end(&tstats->syncp);
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
|
2012-09-26 20:48:50 -06:00
|
|
|
gro_cells_receive(&tunnel->gro_cells, skb);
|
2010-09-19 18:12:11 -06:00
|
|
|
return 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
2006-04-05 23:31:19 -06:00
|
|
|
icmp_send(skb, ICMP_DEST_UNREACH, ICMP_PORT_UNREACH, 0);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
drop:
|
|
|
|
kfree_skb(skb);
|
2010-09-22 14:43:57 -06:00
|
|
|
return 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2013-02-24 13:05:05 -07:00
|
|
|
static struct sk_buff *handle_offloads(struct ip_tunnel *tunnel, struct sk_buff *skb)
|
2013-02-14 07:02:41 -07:00
|
|
|
{
|
|
|
|
int err;
|
|
|
|
|
|
|
|
if (skb_is_gso(skb)) {
|
|
|
|
err = skb_unclone(skb, GFP_ATOMIC);
|
|
|
|
if (unlikely(err))
|
|
|
|
goto error;
|
|
|
|
skb_shinfo(skb)->gso_type |= SKB_GSO_GRE;
|
|
|
|
return skb;
|
2013-02-24 13:05:05 -07:00
|
|
|
} else if (skb->ip_summed == CHECKSUM_PARTIAL &&
|
|
|
|
tunnel->parms.o_flags&GRE_CSUM) {
|
|
|
|
err = skb_checksum_help(skb);
|
|
|
|
if (unlikely(err))
|
|
|
|
goto error;
|
|
|
|
} else if (skb->ip_summed != CHECKSUM_PARTIAL)
|
2013-02-18 02:50:52 -07:00
|
|
|
skb->ip_summed = CHECKSUM_NONE;
|
2013-02-14 07:02:41 -07:00
|
|
|
|
|
|
|
return skb;
|
|
|
|
|
|
|
|
error:
|
|
|
|
kfree_skb(skb);
|
|
|
|
return ERR_PTR(err);
|
|
|
|
}
|
|
|
|
|
2009-08-31 13:50:41 -06:00
|
|
|
static netdev_tx_t ipgre_tunnel_xmit(struct sk_buff *skb, struct net_device *dev)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2013-02-14 07:02:41 -07:00
|
|
|
struct pcpu_tstats *tstats = this_cpu_ptr(dev->tstats);
|
2006-01-08 23:05:26 -07:00
|
|
|
struct ip_tunnel *tunnel = netdev_priv(dev);
|
2013-01-25 13:34:37 -07:00
|
|
|
const struct iphdr *old_iph;
|
2011-04-21 22:53:02 -06:00
|
|
|
const struct iphdr *tiph;
|
2011-05-04 13:33:34 -06:00
|
|
|
struct flowi4 fl4;
|
2005-04-16 16:20:36 -06:00
|
|
|
u8 tos;
|
2006-11-08 01:23:14 -07:00
|
|
|
__be16 df;
|
2005-04-16 16:20:36 -06:00
|
|
|
struct rtable *rt; /* Route to the other host */
|
2010-09-15 05:07:53 -06:00
|
|
|
struct net_device *tdev; /* Device to other host */
|
2005-04-16 16:20:36 -06:00
|
|
|
struct iphdr *iph; /* Our new IP header */
|
2007-10-23 22:07:32 -06:00
|
|
|
unsigned int max_headroom; /* The extra header space needed */
|
2005-04-16 16:20:36 -06:00
|
|
|
int gre_hlen;
|
2006-11-08 01:23:14 -07:00
|
|
|
__be32 dst;
|
2005-04-16 16:20:36 -06:00
|
|
|
int mtu;
|
2012-12-20 09:00:27 -07:00
|
|
|
u8 ttl;
|
2013-02-14 07:02:41 -07:00
|
|
|
int err;
|
|
|
|
int pkt_len;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2013-02-24 13:05:05 -07:00
|
|
|
skb = handle_offloads(tunnel, skb);
|
2013-02-14 07:02:41 -07:00
|
|
|
if (IS_ERR(skb)) {
|
|
|
|
dev->stats.tx_dropped++;
|
|
|
|
return NETDEV_TX_OK;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!skb->encapsulation) {
|
|
|
|
skb_reset_inner_headers(skb);
|
|
|
|
skb->encapsulation = 1;
|
|
|
|
}
|
2012-09-13 15:25:33 -06:00
|
|
|
|
2013-01-25 13:34:37 -07:00
|
|
|
old_iph = ip_hdr(skb);
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
if (dev->type == ARPHRD_ETHER)
|
|
|
|
IPCB(skb)->flags = 0;
|
|
|
|
|
|
|
|
if (dev->header_ops && dev->type == ARPHRD_IPGRE) {
|
2005-04-16 16:20:36 -06:00
|
|
|
gre_hlen = 0;
|
2012-12-20 08:12:52 -07:00
|
|
|
if (skb->protocol == htons(ETH_P_IP))
|
|
|
|
tiph = (const struct iphdr *)skb->data;
|
|
|
|
else
|
|
|
|
tiph = &tunnel->parms.iph;
|
2005-04-16 16:20:36 -06:00
|
|
|
} else {
|
|
|
|
gre_hlen = tunnel->hlen;
|
|
|
|
tiph = &tunnel->parms.iph;
|
|
|
|
}
|
|
|
|
|
|
|
|
if ((dst = tiph->daddr) == 0) {
|
|
|
|
/* NBMA tunnel */
|
|
|
|
|
2009-06-01 23:19:30 -06:00
|
|
|
if (skb_dst(skb) == NULL) {
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
dev->stats.tx_fifo_errors++;
|
2005-04-16 16:20:36 -06:00
|
|
|
goto tx_error;
|
|
|
|
}
|
|
|
|
|
2012-01-24 16:23:30 -07:00
|
|
|
if (skb->protocol == htons(ETH_P_IP)) {
|
2009-06-01 23:14:27 -06:00
|
|
|
rt = skb_rtable(skb);
|
2012-07-13 06:03:45 -06:00
|
|
|
dst = rt_nexthop(rt, old_iph->daddr);
|
2012-01-24 16:23:30 -07:00
|
|
|
}
|
2011-12-10 02:48:31 -07:00
|
|
|
#if IS_ENABLED(CONFIG_IPV6)
|
2005-04-16 16:20:36 -06:00
|
|
|
else if (skb->protocol == htons(ETH_P_IPV6)) {
|
2011-04-21 22:53:02 -06:00
|
|
|
const struct in6_addr *addr6;
|
2012-01-27 16:01:08 -07:00
|
|
|
struct neighbour *neigh;
|
|
|
|
bool do_tx_error_icmp;
|
2005-04-16 16:20:36 -06:00
|
|
|
int addr_type;
|
|
|
|
|
2012-01-27 16:01:08 -07:00
|
|
|
neigh = dst_neigh_lookup(skb_dst(skb), &ipv6_hdr(skb)->daddr);
|
2005-04-16 16:20:36 -06:00
|
|
|
if (neigh == NULL)
|
|
|
|
goto tx_error;
|
|
|
|
|
2011-04-21 22:53:02 -06:00
|
|
|
addr6 = (const struct in6_addr *)&neigh->primary_key;
|
2005-04-16 16:20:36 -06:00
|
|
|
addr_type = ipv6_addr_type(addr6);
|
|
|
|
|
|
|
|
if (addr_type == IPV6_ADDR_ANY) {
|
2007-04-25 18:54:47 -06:00
|
|
|
addr6 = &ipv6_hdr(skb)->daddr;
|
2005-04-16 16:20:36 -06:00
|
|
|
addr_type = ipv6_addr_type(addr6);
|
|
|
|
}
|
|
|
|
|
|
|
|
if ((addr_type & IPV6_ADDR_COMPATv4) == 0)
|
2012-01-27 16:01:08 -07:00
|
|
|
do_tx_error_icmp = true;
|
|
|
|
else {
|
|
|
|
do_tx_error_icmp = false;
|
|
|
|
dst = addr6->s6_addr32[3];
|
|
|
|
}
|
|
|
|
neigh_release(neigh);
|
|
|
|
if (do_tx_error_icmp)
|
2005-04-16 16:20:36 -06:00
|
|
|
goto tx_error_icmp;
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
else
|
|
|
|
goto tx_error;
|
|
|
|
}
|
|
|
|
|
2012-12-20 09:00:27 -07:00
|
|
|
ttl = tiph->ttl;
|
2005-04-16 16:20:36 -06:00
|
|
|
tos = tiph->tos;
|
2013-01-27 06:04:58 -07:00
|
|
|
if (tos & 0x1) {
|
|
|
|
tos &= ~0x1;
|
2005-04-16 16:20:36 -06:00
|
|
|
if (skb->protocol == htons(ETH_P_IP))
|
|
|
|
tos = old_iph->tos;
|
2010-07-08 22:35:58 -06:00
|
|
|
else if (skb->protocol == htons(ETH_P_IPV6))
|
2011-04-21 22:53:02 -06:00
|
|
|
tos = ipv6_get_dsfield((const struct ipv6hdr *)old_iph);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2011-05-04 13:33:34 -06:00
|
|
|
rt = ip_route_output_gre(dev_net(dev), &fl4, dst, tiph->saddr,
|
2011-03-11 22:00:52 -07:00
|
|
|
tunnel->parms.o_key, RT_TOS(tos),
|
|
|
|
tunnel->parms.link);
|
|
|
|
if (IS_ERR(rt)) {
|
|
|
|
dev->stats.tx_carrier_errors++;
|
|
|
|
goto tx_error;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
2010-06-11 00:31:35 -06:00
|
|
|
tdev = rt->dst.dev;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (tdev == dev) {
|
|
|
|
ip_rt_put(rt);
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
dev->stats.collisions++;
|
2005-04-16 16:20:36 -06:00
|
|
|
goto tx_error;
|
|
|
|
}
|
|
|
|
|
|
|
|
df = tiph->frag_off;
|
|
|
|
if (df)
|
2010-06-11 00:31:35 -06:00
|
|
|
mtu = dst_mtu(&rt->dst) - dev->hard_header_len - tunnel->hlen;
|
2005-04-16 16:20:36 -06:00
|
|
|
else
|
2009-06-01 23:19:30 -06:00
|
|
|
mtu = skb_dst(skb) ? dst_mtu(skb_dst(skb)) : dev->mtu;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2009-06-01 23:19:30 -06:00
|
|
|
if (skb_dst(skb))
|
2012-07-17 04:29:28 -06:00
|
|
|
skb_dst(skb)->ops->update_pmtu(skb_dst(skb), NULL, skb, mtu);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (skb->protocol == htons(ETH_P_IP)) {
|
|
|
|
df |= (old_iph->frag_off&htons(IP_DF));
|
|
|
|
|
2013-02-14 07:02:41 -07:00
|
|
|
if (!skb_is_gso(skb) &&
|
|
|
|
(old_iph->frag_off&htons(IP_DF)) &&
|
2005-04-16 16:20:36 -06:00
|
|
|
mtu < ntohs(old_iph->tot_len)) {
|
|
|
|
icmp_send(skb, ICMP_DEST_UNREACH, ICMP_FRAG_NEEDED, htonl(mtu));
|
|
|
|
ip_rt_put(rt);
|
|
|
|
goto tx_error;
|
|
|
|
}
|
|
|
|
}
|
2011-12-10 02:48:31 -07:00
|
|
|
#if IS_ENABLED(CONFIG_IPV6)
|
2005-04-16 16:20:36 -06:00
|
|
|
else if (skb->protocol == htons(ETH_P_IPV6)) {
|
2009-06-01 23:19:30 -06:00
|
|
|
struct rt6_info *rt6 = (struct rt6_info *)skb_dst(skb);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2009-06-01 23:19:30 -06:00
|
|
|
if (rt6 && mtu < dst_mtu(skb_dst(skb)) && mtu >= IPV6_MIN_MTU) {
|
2007-12-16 14:45:43 -07:00
|
|
|
if ((tunnel->parms.iph.daddr &&
|
|
|
|
!ipv4_is_multicast(tunnel->parms.iph.daddr)) ||
|
2005-04-16 16:20:36 -06:00
|
|
|
rt6->rt6i_dst.plen == 128) {
|
|
|
|
rt6->rt6i_flags |= RTF_MODIFIED;
|
2010-12-08 22:16:57 -07:00
|
|
|
dst_metric_set(skb_dst(skb), RTAX_MTU, mtu);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-02-14 07:02:41 -07:00
|
|
|
if (!skb_is_gso(skb) &&
|
|
|
|
mtu >= IPV6_MIN_MTU &&
|
|
|
|
mtu < skb->len - tunnel->hlen + gre_hlen) {
|
2010-02-18 01:25:24 -07:00
|
|
|
icmpv6_send(skb, ICMPV6_PKT_TOOBIG, 0, mtu);
|
2005-04-16 16:20:36 -06:00
|
|
|
ip_rt_put(rt);
|
|
|
|
goto tx_error;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
#endif
|
|
|
|
|
|
|
|
if (tunnel->err_count > 0) {
|
2009-02-25 00:34:48 -07:00
|
|
|
if (time_before(jiffies,
|
|
|
|
tunnel->err_time + IPTUNNEL_ERR_TIMEO)) {
|
2005-04-16 16:20:36 -06:00
|
|
|
tunnel->err_count--;
|
|
|
|
|
|
|
|
dst_link_failure(skb);
|
|
|
|
} else
|
|
|
|
tunnel->err_count = 0;
|
|
|
|
}
|
|
|
|
|
2010-06-11 00:31:35 -06:00
|
|
|
max_headroom = LL_RESERVED_SPACE(tdev) + gre_hlen + rt->dst.header_len;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-07-09 16:33:40 -06:00
|
|
|
if (skb_headroom(skb) < max_headroom || skb_shared(skb)||
|
|
|
|
(skb_cloned(skb) && !skb_clone_writable(skb, 0))) {
|
2005-04-16 16:20:36 -06:00
|
|
|
struct sk_buff *new_skb = skb_realloc_headroom(skb, max_headroom);
|
2011-11-17 19:20:06 -07:00
|
|
|
if (max_headroom > dev->needed_headroom)
|
|
|
|
dev->needed_headroom = max_headroom;
|
2005-04-16 16:20:36 -06:00
|
|
|
if (!new_skb) {
|
|
|
|
ip_rt_put(rt);
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
dev->stats.tx_dropped++;
|
2005-04-16 16:20:36 -06:00
|
|
|
dev_kfree_skb(skb);
|
2009-06-23 00:03:08 -06:00
|
|
|
return NETDEV_TX_OK;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
if (skb->sk)
|
|
|
|
skb_set_owner_w(new_skb, skb->sk);
|
|
|
|
dev_kfree_skb(skb);
|
|
|
|
skb = new_skb;
|
2007-04-20 23:47:35 -06:00
|
|
|
old_iph = ip_hdr(skb);
|
2012-12-20 09:00:27 -07:00
|
|
|
/* Warning : tiph value might point to freed memory */
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2007-04-10 21:46:21 -06:00
|
|
|
skb_push(skb, gre_hlen);
|
|
|
|
skb_reset_network_header(skb);
|
2012-12-24 09:51:03 -07:00
|
|
|
skb_set_transport_header(skb, sizeof(*iph));
|
2005-04-16 16:20:36 -06:00
|
|
|
memset(&(IPCB(skb)->opt), 0, sizeof(IPCB(skb)->opt));
|
2006-02-15 16:10:22 -07:00
|
|
|
IPCB(skb)->flags &= ~(IPSKB_XFRM_TUNNEL_SIZE | IPSKB_XFRM_TRANSFORMED |
|
|
|
|
IPSKB_REROUTED);
|
2009-06-01 23:19:30 -06:00
|
|
|
skb_dst_drop(skb);
|
2010-06-11 00:31:35 -06:00
|
|
|
skb_dst_set(skb, &rt->dst);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Push down and install the IPIP header.
|
|
|
|
*/
|
|
|
|
|
2007-04-20 23:47:35 -06:00
|
|
|
iph = ip_hdr(skb);
|
2005-04-16 16:20:36 -06:00
|
|
|
iph->version = 4;
|
|
|
|
iph->ihl = sizeof(struct iphdr) >> 2;
|
|
|
|
iph->frag_off = df;
|
|
|
|
iph->protocol = IPPROTO_GRE;
|
|
|
|
iph->tos = ipgre_ecn_encapsulate(tos, old_iph, skb);
|
2011-05-04 13:33:34 -06:00
|
|
|
iph->daddr = fl4.daddr;
|
|
|
|
iph->saddr = fl4.saddr;
|
2012-12-20 09:00:27 -07:00
|
|
|
iph->ttl = ttl;
|
2013-02-22 00:30:30 -07:00
|
|
|
|
|
|
|
tunnel_ip_select_ident(skb, old_iph, &rt->dst);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2012-12-20 09:00:27 -07:00
|
|
|
if (ttl == 0) {
|
2005-04-16 16:20:36 -06:00
|
|
|
if (skb->protocol == htons(ETH_P_IP))
|
|
|
|
iph->ttl = old_iph->ttl;
|
2011-12-10 02:48:31 -07:00
|
|
|
#if IS_ENABLED(CONFIG_IPV6)
|
2005-04-16 16:20:36 -06:00
|
|
|
else if (skb->protocol == htons(ETH_P_IPV6))
|
2011-04-21 22:53:02 -06:00
|
|
|
iph->ttl = ((const struct ipv6hdr *)old_iph)->hop_limit;
|
2005-04-16 16:20:36 -06:00
|
|
|
#endif
|
|
|
|
else
|
2010-12-12 22:55:08 -07:00
|
|
|
iph->ttl = ip4_dst_hoplimit(&rt->dst);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
((__be16 *)(iph + 1))[0] = tunnel->parms.o_flags;
|
|
|
|
((__be16 *)(iph + 1))[1] = (dev->type == ARPHRD_ETHER) ?
|
|
|
|
htons(ETH_P_TEB) : skb->protocol;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (tunnel->parms.o_flags&(GRE_KEY|GRE_CSUM|GRE_SEQ)) {
|
2012-04-14 19:34:41 -06:00
|
|
|
__be32 *ptr = (__be32 *)(((u8 *)iph) + tunnel->hlen - 4);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
if (tunnel->parms.o_flags&GRE_SEQ) {
|
|
|
|
++tunnel->o_seqno;
|
|
|
|
*ptr = htonl(tunnel->o_seqno);
|
|
|
|
ptr--;
|
|
|
|
}
|
|
|
|
if (tunnel->parms.o_flags&GRE_KEY) {
|
|
|
|
*ptr = tunnel->parms.o_key;
|
|
|
|
ptr--;
|
|
|
|
}
|
2013-02-14 07:02:41 -07:00
|
|
|
/* Skip GRE checksum if skb is getting offloaded. */
|
|
|
|
if (!(skb_shinfo(skb)->gso_type & SKB_GSO_GRE) &&
|
|
|
|
(tunnel->parms.o_flags&GRE_CSUM)) {
|
2013-01-23 04:45:42 -07:00
|
|
|
int offset = skb_transport_offset(skb);
|
|
|
|
|
2013-02-14 07:02:41 -07:00
|
|
|
if (skb_has_shared_frag(skb)) {
|
|
|
|
err = __skb_linearize(skb);
|
2013-02-19 05:14:10 -07:00
|
|
|
if (err)
|
2013-02-14 07:02:41 -07:00
|
|
|
goto tx_error;
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
*ptr = 0;
|
2013-01-23 04:45:42 -07:00
|
|
|
*(__sum16 *)ptr = csum_fold(skb_checksum(skb, offset,
|
|
|
|
skb->len - offset,
|
|
|
|
0));
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2013-02-14 07:02:41 -07:00
|
|
|
nf_reset(skb);
|
|
|
|
|
|
|
|
pkt_len = skb->len - skb_transport_offset(skb);
|
|
|
|
err = ip_local_out(skb);
|
|
|
|
if (likely(net_xmit_eval(err) == 0)) {
|
|
|
|
u64_stats_update_begin(&tstats->syncp);
|
|
|
|
tstats->tx_bytes += pkt_len;
|
|
|
|
tstats->tx_packets++;
|
|
|
|
u64_stats_update_end(&tstats->syncp);
|
|
|
|
} else {
|
|
|
|
dev->stats.tx_errors++;
|
|
|
|
dev->stats.tx_aborted_errors++;
|
|
|
|
}
|
2009-06-23 00:03:08 -06:00
|
|
|
return NETDEV_TX_OK;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2012-01-11 17:46:32 -07:00
|
|
|
#if IS_ENABLED(CONFIG_IPV6)
|
2005-04-16 16:20:36 -06:00
|
|
|
tx_error_icmp:
|
|
|
|
dst_link_failure(skb);
|
2012-01-11 17:46:32 -07:00
|
|
|
#endif
|
2005-04-16 16:20:36 -06:00
|
|
|
tx_error:
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
dev->stats.tx_errors++;
|
2005-04-16 16:20:36 -06:00
|
|
|
dev_kfree_skb(skb);
|
2009-06-23 00:03:08 -06:00
|
|
|
return NETDEV_TX_OK;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2008-10-09 12:59:32 -06:00
|
|
|
static int ipgre_tunnel_bind_dev(struct net_device *dev)
|
2007-12-13 10:46:32 -07:00
|
|
|
{
|
|
|
|
struct net_device *tdev = NULL;
|
|
|
|
struct ip_tunnel *tunnel;
|
2011-04-21 22:53:02 -06:00
|
|
|
const struct iphdr *iph;
|
2007-12-13 10:46:32 -07:00
|
|
|
int hlen = LL_MAX_HEADER;
|
|
|
|
int mtu = ETH_DATA_LEN;
|
|
|
|
int addend = sizeof(struct iphdr) + 4;
|
|
|
|
|
|
|
|
tunnel = netdev_priv(dev);
|
|
|
|
iph = &tunnel->parms.iph;
|
|
|
|
|
2008-10-09 12:58:54 -06:00
|
|
|
/* Guess output device to choose reasonable mtu and needed_headroom */
|
2007-12-13 10:46:32 -07:00
|
|
|
|
|
|
|
if (iph->daddr) {
|
2011-05-04 13:33:34 -06:00
|
|
|
struct flowi4 fl4;
|
|
|
|
struct rtable *rt;
|
|
|
|
|
|
|
|
rt = ip_route_output_gre(dev_net(dev), &fl4,
|
|
|
|
iph->daddr, iph->saddr,
|
|
|
|
tunnel->parms.o_key,
|
|
|
|
RT_TOS(iph->tos),
|
|
|
|
tunnel->parms.link);
|
2011-03-02 15:31:35 -07:00
|
|
|
if (!IS_ERR(rt)) {
|
2010-06-11 00:31:35 -06:00
|
|
|
tdev = rt->dst.dev;
|
2007-12-13 10:46:32 -07:00
|
|
|
ip_rt_put(rt);
|
|
|
|
}
|
2008-10-09 13:00:17 -06:00
|
|
|
|
|
|
|
if (dev->type != ARPHRD_ETHER)
|
|
|
|
dev->flags |= IFF_POINTOPOINT;
|
2007-12-13 10:46:32 -07:00
|
|
|
}
|
|
|
|
|
|
|
|
if (!tdev && tunnel->parms.link)
|
2008-04-16 02:10:44 -06:00
|
|
|
tdev = __dev_get_by_index(dev_net(dev), tunnel->parms.link);
|
2007-12-13 10:46:32 -07:00
|
|
|
|
|
|
|
if (tdev) {
|
2008-10-09 12:58:54 -06:00
|
|
|
hlen = tdev->hard_header_len + tdev->needed_headroom;
|
2007-12-13 10:46:32 -07:00
|
|
|
mtu = tdev->mtu;
|
|
|
|
}
|
|
|
|
dev->iflink = tunnel->parms.link;
|
|
|
|
|
|
|
|
/* Precalculate GRE options length */
|
|
|
|
if (tunnel->parms.o_flags&(GRE_CSUM|GRE_KEY|GRE_SEQ)) {
|
|
|
|
if (tunnel->parms.o_flags&GRE_CSUM)
|
|
|
|
addend += 4;
|
|
|
|
if (tunnel->parms.o_flags&GRE_KEY)
|
|
|
|
addend += 4;
|
|
|
|
if (tunnel->parms.o_flags&GRE_SEQ)
|
|
|
|
addend += 4;
|
|
|
|
}
|
2008-10-09 12:58:54 -06:00
|
|
|
dev->needed_headroom = addend + hlen;
|
2009-08-14 17:33:56 -06:00
|
|
|
mtu -= dev->hard_header_len + addend;
|
2008-10-09 12:59:32 -06:00
|
|
|
|
|
|
|
if (mtu < 68)
|
|
|
|
mtu = 68;
|
|
|
|
|
2007-12-13 10:46:32 -07:00
|
|
|
tunnel->hlen = addend;
|
2013-02-14 07:02:41 -07:00
|
|
|
/* TCP offload with GRE SEQ is not supported. */
|
|
|
|
if (!(tunnel->parms.o_flags & GRE_SEQ)) {
|
2013-02-24 13:05:12 -07:00
|
|
|
dev->features |= NETIF_F_GSO_SOFTWARE;
|
|
|
|
dev->hw_features |= NETIF_F_GSO_SOFTWARE;
|
2013-02-14 07:02:41 -07:00
|
|
|
}
|
2007-12-13 10:46:32 -07:00
|
|
|
|
2008-10-09 12:59:32 -06:00
|
|
|
return mtu;
|
2007-12-13 10:46:32 -07:00
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
static int
|
|
|
|
ipgre_tunnel_ioctl (struct net_device *dev, struct ifreq *ifr, int cmd)
|
|
|
|
{
|
|
|
|
int err = 0;
|
|
|
|
struct ip_tunnel_parm p;
|
|
|
|
struct ip_tunnel *t;
|
2008-04-16 02:09:22 -06:00
|
|
|
struct net *net = dev_net(dev);
|
|
|
|
struct ipgre_net *ign = net_generic(net, ipgre_net_id);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
switch (cmd) {
|
|
|
|
case SIOCGETTUNNEL:
|
|
|
|
t = NULL;
|
2008-04-16 02:10:05 -06:00
|
|
|
if (dev == ign->fb_tunnel_dev) {
|
2005-04-16 16:20:36 -06:00
|
|
|
if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p))) {
|
|
|
|
err = -EFAULT;
|
|
|
|
break;
|
|
|
|
}
|
2008-04-16 02:09:22 -06:00
|
|
|
t = ipgre_tunnel_locate(net, &p, 0);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
if (t == NULL)
|
2006-01-08 23:05:26 -07:00
|
|
|
t = netdev_priv(dev);
|
2005-04-16 16:20:36 -06:00
|
|
|
memcpy(&p, &t->parms, sizeof(p));
|
|
|
|
if (copy_to_user(ifr->ifr_ifru.ifru_data, &p, sizeof(p)))
|
|
|
|
err = -EFAULT;
|
|
|
|
break;
|
|
|
|
|
|
|
|
case SIOCADDTUNNEL:
|
|
|
|
case SIOCCHGTUNNEL:
|
|
|
|
err = -EPERM;
|
net: Allow userns root to control ipv4
Allow an unpriviled user who has created a user namespace, and then
created a network namespace to effectively use the new network
namespace, by reducing capable(CAP_NET_ADMIN) and
capable(CAP_NET_RAW) calls to be ns_capable(net->user_ns,
CAP_NET_ADMIN), or capable(net->user_ns, CAP_NET_RAW) calls.
Settings that merely control a single network device are allowed.
Either the network device is a logical network device where
restrictions make no difference or the network device is hardware NIC
that has been explicity moved from the initial network namespace.
In general policy and network stack state changes are allowed
while resource control is left unchanged.
Allow creating raw sockets.
Allow the SIOCSARP ioctl to control the arp cache.
Allow the SIOCSIFFLAG ioctl to allow setting network device flags.
Allow the SIOCSIFADDR ioctl to allow setting a netdevice ipv4 address.
Allow the SIOCSIFBRDADDR ioctl to allow setting a netdevice ipv4 broadcast address.
Allow the SIOCSIFDSTADDR ioctl to allow setting a netdevice ipv4 destination address.
Allow the SIOCSIFNETMASK ioctl to allow setting a netdevice ipv4 netmask.
Allow the SIOCADDRT and SIOCDELRT ioctls to allow adding and deleting ipv4 routes.
Allow the SIOCADDTUNNEL, SIOCCHGTUNNEL and SIOCDELTUNNEL ioctls for
adding, changing and deleting gre tunnels.
Allow the SIOCADDTUNNEL, SIOCCHGTUNNEL and SIOCDELTUNNEL ioctls for
adding, changing and deleting ipip tunnels.
Allow the SIOCADDTUNNEL, SIOCCHGTUNNEL and SIOCDELTUNNEL ioctls for
adding, changing and deleting ipsec virtual tunnel interfaces.
Allow setting the MRT_INIT, MRT_DONE, MRT_ADD_VIF, MRT_DEL_VIF, MRT_ADD_MFC,
MRT_DEL_MFC, MRT_ASSERT, MRT_PIM, MRT_TABLE socket options on multicast routing
sockets.
Allow setting and receiving IPOPT_CIPSO, IP_OPT_SEC, IP_OPT_SID and
arbitrary ip options.
Allow setting IP_SEC_POLICY/IP_XFRM_POLICY ipv4 socket option.
Allow setting the IP_TRANSPARENT ipv4 socket option.
Allow setting the TCP_REPAIR socket option.
Allow setting the TCP_CONGESTION socket option.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-11-15 20:03:05 -07:00
|
|
|
if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
|
2005-04-16 16:20:36 -06:00
|
|
|
goto done;
|
|
|
|
|
|
|
|
err = -EFAULT;
|
|
|
|
if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
|
|
|
|
goto done;
|
|
|
|
|
|
|
|
err = -EINVAL;
|
|
|
|
if (p.iph.version != 4 || p.iph.protocol != IPPROTO_GRE ||
|
|
|
|
p.iph.ihl != 5 || (p.iph.frag_off&htons(~IP_DF)) ||
|
|
|
|
((p.i_flags|p.o_flags)&(GRE_VERSION|GRE_ROUTING)))
|
|
|
|
goto done;
|
|
|
|
if (p.iph.ttl)
|
|
|
|
p.iph.frag_off |= htons(IP_DF);
|
|
|
|
|
|
|
|
if (!(p.i_flags&GRE_KEY))
|
|
|
|
p.i_key = 0;
|
|
|
|
if (!(p.o_flags&GRE_KEY))
|
|
|
|
p.o_key = 0;
|
|
|
|
|
2008-04-16 02:09:22 -06:00
|
|
|
t = ipgre_tunnel_locate(net, &p, cmd == SIOCADDTUNNEL);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2008-04-16 02:10:05 -06:00
|
|
|
if (dev != ign->fb_tunnel_dev && cmd == SIOCCHGTUNNEL) {
|
2005-04-16 16:20:36 -06:00
|
|
|
if (t != NULL) {
|
|
|
|
if (t->dev != dev) {
|
|
|
|
err = -EEXIST;
|
|
|
|
break;
|
|
|
|
}
|
|
|
|
} else {
|
2010-09-15 05:07:53 -06:00
|
|
|
unsigned int nflags = 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2006-01-08 23:05:26 -07:00
|
|
|
t = netdev_priv(dev);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-12-16 14:45:43 -07:00
|
|
|
if (ipv4_is_multicast(p.iph.daddr))
|
2005-04-16 16:20:36 -06:00
|
|
|
nflags = IFF_BROADCAST;
|
|
|
|
else if (p.iph.daddr)
|
|
|
|
nflags = IFF_POINTOPOINT;
|
|
|
|
|
|
|
|
if ((dev->flags^nflags)&(IFF_POINTOPOINT|IFF_BROADCAST)) {
|
|
|
|
err = -EINVAL;
|
|
|
|
break;
|
|
|
|
}
|
2008-04-16 02:09:22 -06:00
|
|
|
ipgre_tunnel_unlink(ign, t);
|
2010-10-26 23:43:53 -06:00
|
|
|
synchronize_net();
|
2005-04-16 16:20:36 -06:00
|
|
|
t->parms.iph.saddr = p.iph.saddr;
|
|
|
|
t->parms.iph.daddr = p.iph.daddr;
|
|
|
|
t->parms.i_key = p.i_key;
|
|
|
|
t->parms.o_key = p.o_key;
|
|
|
|
memcpy(dev->dev_addr, &p.iph.saddr, 4);
|
|
|
|
memcpy(dev->broadcast, &p.iph.daddr, 4);
|
2008-04-16 02:09:22 -06:00
|
|
|
ipgre_tunnel_link(ign, t);
|
2005-04-16 16:20:36 -06:00
|
|
|
netdev_state_change(dev);
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if (t) {
|
|
|
|
err = 0;
|
|
|
|
if (cmd == SIOCCHGTUNNEL) {
|
|
|
|
t->parms.iph.ttl = p.iph.ttl;
|
|
|
|
t->parms.iph.tos = p.iph.tos;
|
|
|
|
t->parms.iph.frag_off = p.iph.frag_off;
|
2007-12-13 10:46:32 -07:00
|
|
|
if (t->parms.link != p.link) {
|
|
|
|
t->parms.link = p.link;
|
2008-10-09 12:59:32 -06:00
|
|
|
dev->mtu = ipgre_tunnel_bind_dev(dev);
|
2007-12-13 10:46:32 -07:00
|
|
|
netdev_state_change(dev);
|
|
|
|
}
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
if (copy_to_user(ifr->ifr_ifru.ifru_data, &t->parms, sizeof(p)))
|
|
|
|
err = -EFAULT;
|
|
|
|
} else
|
|
|
|
err = (cmd == SIOCADDTUNNEL ? -ENOBUFS : -ENOENT);
|
|
|
|
break;
|
|
|
|
|
|
|
|
case SIOCDELTUNNEL:
|
|
|
|
err = -EPERM;
|
net: Allow userns root to control ipv4
Allow an unpriviled user who has created a user namespace, and then
created a network namespace to effectively use the new network
namespace, by reducing capable(CAP_NET_ADMIN) and
capable(CAP_NET_RAW) calls to be ns_capable(net->user_ns,
CAP_NET_ADMIN), or capable(net->user_ns, CAP_NET_RAW) calls.
Settings that merely control a single network device are allowed.
Either the network device is a logical network device where
restrictions make no difference or the network device is hardware NIC
that has been explicity moved from the initial network namespace.
In general policy and network stack state changes are allowed
while resource control is left unchanged.
Allow creating raw sockets.
Allow the SIOCSARP ioctl to control the arp cache.
Allow the SIOCSIFFLAG ioctl to allow setting network device flags.
Allow the SIOCSIFADDR ioctl to allow setting a netdevice ipv4 address.
Allow the SIOCSIFBRDADDR ioctl to allow setting a netdevice ipv4 broadcast address.
Allow the SIOCSIFDSTADDR ioctl to allow setting a netdevice ipv4 destination address.
Allow the SIOCSIFNETMASK ioctl to allow setting a netdevice ipv4 netmask.
Allow the SIOCADDRT and SIOCDELRT ioctls to allow adding and deleting ipv4 routes.
Allow the SIOCADDTUNNEL, SIOCCHGTUNNEL and SIOCDELTUNNEL ioctls for
adding, changing and deleting gre tunnels.
Allow the SIOCADDTUNNEL, SIOCCHGTUNNEL and SIOCDELTUNNEL ioctls for
adding, changing and deleting ipip tunnels.
Allow the SIOCADDTUNNEL, SIOCCHGTUNNEL and SIOCDELTUNNEL ioctls for
adding, changing and deleting ipsec virtual tunnel interfaces.
Allow setting the MRT_INIT, MRT_DONE, MRT_ADD_VIF, MRT_DEL_VIF, MRT_ADD_MFC,
MRT_DEL_MFC, MRT_ASSERT, MRT_PIM, MRT_TABLE socket options on multicast routing
sockets.
Allow setting and receiving IPOPT_CIPSO, IP_OPT_SEC, IP_OPT_SID and
arbitrary ip options.
Allow setting IP_SEC_POLICY/IP_XFRM_POLICY ipv4 socket option.
Allow setting the IP_TRANSPARENT ipv4 socket option.
Allow setting the TCP_REPAIR socket option.
Allow setting the TCP_CONGESTION socket option.
Signed-off-by: "Eric W. Biederman" <ebiederm@xmission.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-11-15 20:03:05 -07:00
|
|
|
if (!ns_capable(net->user_ns, CAP_NET_ADMIN))
|
2005-04-16 16:20:36 -06:00
|
|
|
goto done;
|
|
|
|
|
2008-04-16 02:10:05 -06:00
|
|
|
if (dev == ign->fb_tunnel_dev) {
|
2005-04-16 16:20:36 -06:00
|
|
|
err = -EFAULT;
|
|
|
|
if (copy_from_user(&p, ifr->ifr_ifru.ifru_data, sizeof(p)))
|
|
|
|
goto done;
|
|
|
|
err = -ENOENT;
|
2008-04-16 02:09:22 -06:00
|
|
|
if ((t = ipgre_tunnel_locate(net, &p, 0)) == NULL)
|
2005-04-16 16:20:36 -06:00
|
|
|
goto done;
|
|
|
|
err = -EPERM;
|
2008-04-16 02:10:05 -06:00
|
|
|
if (t == netdev_priv(ign->fb_tunnel_dev))
|
2005-04-16 16:20:36 -06:00
|
|
|
goto done;
|
|
|
|
dev = t->dev;
|
|
|
|
}
|
2007-02-07 01:09:58 -07:00
|
|
|
unregister_netdevice(dev);
|
|
|
|
err = 0;
|
2005-04-16 16:20:36 -06:00
|
|
|
break;
|
|
|
|
|
|
|
|
default:
|
|
|
|
err = -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
done:
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ipgre_tunnel_change_mtu(struct net_device *dev, int new_mtu)
|
|
|
|
{
|
2006-01-08 23:05:26 -07:00
|
|
|
struct ip_tunnel *tunnel = netdev_priv(dev);
|
2008-10-09 12:58:54 -06:00
|
|
|
if (new_mtu < 68 ||
|
|
|
|
new_mtu > 0xFFF8 - dev->hard_header_len - tunnel->hlen)
|
2005-04-16 16:20:36 -06:00
|
|
|
return -EINVAL;
|
|
|
|
dev->mtu = new_mtu;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* Nice toy. Unfortunately, useless in real life :-)
|
|
|
|
It allows to construct virtual multiprotocol broadcast "LAN"
|
|
|
|
over the Internet, provided multicast routing is tuned.
|
|
|
|
|
|
|
|
|
|
|
|
I have no idea was this bicycle invented before me,
|
|
|
|
so that I had to set ARPHRD_IPGRE to a random value.
|
|
|
|
I have an impression, that Cisco could make something similar,
|
|
|
|
but this feature is apparently missing in IOS<=11.2(8).
|
2007-02-09 07:24:47 -07:00
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
I set up 10.66.66/24 and fec0:6666:6666::0/96 as virtual networks
|
|
|
|
with broadcast 224.66.66.66. If you have access to mbone, play with me :-)
|
|
|
|
|
|
|
|
ping -t 255 224.66.66.66
|
|
|
|
|
|
|
|
If nobody answers, mbone does not work.
|
|
|
|
|
|
|
|
ip tunnel add Universe mode gre remote 224.66.66.66 local <Your_real_addr> ttl 255
|
|
|
|
ip addr add 10.66.66.<somewhat>/24 dev Universe
|
|
|
|
ifconfig Universe up
|
|
|
|
ifconfig Universe add fe80::<Your_real_addr>/10
|
|
|
|
ifconfig Universe add fec0:6666:6666::<Your_real_addr>/96
|
|
|
|
ftp 10.66.66.66
|
|
|
|
...
|
|
|
|
ftp fec0:6666:6666::193.233.7.65
|
|
|
|
...
|
|
|
|
|
|
|
|
*/
|
|
|
|
|
2007-10-09 02:40:57 -06:00
|
|
|
static int ipgre_header(struct sk_buff *skb, struct net_device *dev,
|
|
|
|
unsigned short type,
|
2010-09-15 05:07:53 -06:00
|
|
|
const void *daddr, const void *saddr, unsigned int len)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2006-01-08 23:05:26 -07:00
|
|
|
struct ip_tunnel *t = netdev_priv(dev);
|
2005-04-16 16:20:36 -06:00
|
|
|
struct iphdr *iph = (struct iphdr *)skb_push(skb, t->hlen);
|
2012-04-14 19:34:41 -06:00
|
|
|
__be16 *p = (__be16 *)(iph+1);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
memcpy(iph, &t->parms.iph, sizeof(struct iphdr));
|
|
|
|
p[0] = t->parms.o_flags;
|
|
|
|
p[1] = htons(type);
|
|
|
|
|
|
|
|
/*
|
2007-02-09 07:24:47 -07:00
|
|
|
* Set the source hardware address.
|
2005-04-16 16:20:36 -06:00
|
|
|
*/
|
2007-02-09 07:24:47 -07:00
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
if (saddr)
|
|
|
|
memcpy(&iph->saddr, saddr, 4);
|
2010-03-02 21:01:13 -07:00
|
|
|
if (daddr)
|
2005-04-16 16:20:36 -06:00
|
|
|
memcpy(&iph->daddr, daddr, 4);
|
2010-03-02 21:01:13 -07:00
|
|
|
if (iph->daddr)
|
2005-04-16 16:20:36 -06:00
|
|
|
return t->hlen;
|
2007-02-09 07:24:47 -07:00
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
return -t->hlen;
|
|
|
|
}
|
|
|
|
|
2007-10-23 21:31:53 -06:00
|
|
|
static int ipgre_header_parse(const struct sk_buff *skb, unsigned char *haddr)
|
|
|
|
{
|
2011-04-21 22:53:02 -06:00
|
|
|
const struct iphdr *iph = (const struct iphdr *) skb_mac_header(skb);
|
2007-10-23 21:31:53 -06:00
|
|
|
memcpy(haddr, &iph->saddr, 4);
|
|
|
|
return 4;
|
|
|
|
}
|
|
|
|
|
2007-10-09 02:40:57 -06:00
|
|
|
static const struct header_ops ipgre_header_ops = {
|
|
|
|
.create = ipgre_header,
|
2007-10-23 21:31:53 -06:00
|
|
|
.parse = ipgre_header_parse,
|
2007-10-09 02:40:57 -06:00
|
|
|
};
|
|
|
|
|
2007-10-23 21:31:53 -06:00
|
|
|
#ifdef CONFIG_NET_IPGRE_BROADCAST
|
2005-04-16 16:20:36 -06:00
|
|
|
static int ipgre_open(struct net_device *dev)
|
|
|
|
{
|
2006-01-08 23:05:26 -07:00
|
|
|
struct ip_tunnel *t = netdev_priv(dev);
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2007-12-16 14:45:43 -07:00
|
|
|
if (ipv4_is_multicast(t->parms.iph.daddr)) {
|
2011-05-04 13:33:34 -06:00
|
|
|
struct flowi4 fl4;
|
|
|
|
struct rtable *rt;
|
|
|
|
|
|
|
|
rt = ip_route_output_gre(dev_net(dev), &fl4,
|
|
|
|
t->parms.iph.daddr,
|
|
|
|
t->parms.iph.saddr,
|
|
|
|
t->parms.o_key,
|
|
|
|
RT_TOS(t->parms.iph.tos),
|
|
|
|
t->parms.link);
|
2011-03-02 15:31:35 -07:00
|
|
|
if (IS_ERR(rt))
|
2005-04-16 16:20:36 -06:00
|
|
|
return -EADDRNOTAVAIL;
|
2010-06-11 00:31:35 -06:00
|
|
|
dev = rt->dst.dev;
|
2005-04-16 16:20:36 -06:00
|
|
|
ip_rt_put(rt);
|
2005-10-03 15:35:55 -06:00
|
|
|
if (__in_dev_get_rtnl(dev) == NULL)
|
2005-04-16 16:20:36 -06:00
|
|
|
return -EADDRNOTAVAIL;
|
|
|
|
t->mlink = dev->ifindex;
|
2005-10-03 15:35:55 -06:00
|
|
|
ip_mc_inc_group(__in_dev_get_rtnl(dev), t->parms.iph.daddr);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ipgre_close(struct net_device *dev)
|
|
|
|
{
|
2006-01-08 23:05:26 -07:00
|
|
|
struct ip_tunnel *t = netdev_priv(dev);
|
2008-11-20 21:34:29 -07:00
|
|
|
|
2007-12-16 14:45:43 -07:00
|
|
|
if (ipv4_is_multicast(t->parms.iph.daddr) && t->mlink) {
|
2008-01-21 18:32:38 -07:00
|
|
|
struct in_device *in_dev;
|
2008-03-25 06:47:49 -06:00
|
|
|
in_dev = inetdev_by_index(dev_net(dev), t->mlink);
|
2010-10-18 18:39:26 -06:00
|
|
|
if (in_dev)
|
2005-04-16 16:20:36 -06:00
|
|
|
ip_mc_dec_group(in_dev, t->parms.iph.daddr);
|
|
|
|
}
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|
|
|
|
|
2008-11-20 21:34:29 -07:00
|
|
|
static const struct net_device_ops ipgre_netdev_ops = {
|
|
|
|
.ndo_init = ipgre_tunnel_init,
|
|
|
|
.ndo_uninit = ipgre_tunnel_uninit,
|
|
|
|
#ifdef CONFIG_NET_IPGRE_BROADCAST
|
|
|
|
.ndo_open = ipgre_open,
|
|
|
|
.ndo_stop = ipgre_close,
|
|
|
|
#endif
|
|
|
|
.ndo_start_xmit = ipgre_tunnel_xmit,
|
|
|
|
.ndo_do_ioctl = ipgre_tunnel_ioctl,
|
|
|
|
.ndo_change_mtu = ipgre_tunnel_change_mtu,
|
2012-04-12 00:31:16 -06:00
|
|
|
.ndo_get_stats64 = ipgre_get_stats64,
|
2008-11-20 21:34:29 -07:00
|
|
|
};
|
|
|
|
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
static void ipgre_dev_free(struct net_device *dev)
|
|
|
|
{
|
2012-09-26 20:48:50 -06:00
|
|
|
struct ip_tunnel *tunnel = netdev_priv(dev);
|
|
|
|
|
|
|
|
gro_cells_destroy(&tunnel->gro_cells);
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
free_percpu(dev->tstats);
|
|
|
|
free_netdev(dev);
|
|
|
|
}
|
|
|
|
|
2012-09-13 15:25:33 -06:00
|
|
|
#define GRE_FEATURES (NETIF_F_SG | \
|
|
|
|
NETIF_F_FRAGLIST | \
|
|
|
|
NETIF_F_HIGHDMA | \
|
|
|
|
NETIF_F_HW_CSUM)
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
static void ipgre_tunnel_setup(struct net_device *dev)
|
|
|
|
{
|
2008-11-20 21:34:29 -07:00
|
|
|
dev->netdev_ops = &ipgre_netdev_ops;
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
dev->destructor = ipgre_dev_free;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
|
|
|
dev->type = ARPHRD_IPGRE;
|
2008-10-09 12:58:54 -06:00
|
|
|
dev->needed_headroom = LL_MAX_HEADER + sizeof(struct iphdr) + 4;
|
2006-01-05 17:35:42 -07:00
|
|
|
dev->mtu = ETH_DATA_LEN - sizeof(struct iphdr) - 4;
|
2005-04-16 16:20:36 -06:00
|
|
|
dev->flags = IFF_NOARP;
|
|
|
|
dev->iflink = 0;
|
|
|
|
dev->addr_len = 4;
|
2008-04-16 02:11:13 -06:00
|
|
|
dev->features |= NETIF_F_NETNS_LOCAL;
|
2009-05-28 16:35:10 -06:00
|
|
|
dev->priv_flags &= ~IFF_XMIT_DST_RELEASE;
|
2012-09-13 15:25:33 -06:00
|
|
|
|
|
|
|
dev->features |= GRE_FEATURES;
|
|
|
|
dev->hw_features |= GRE_FEATURES;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
static int ipgre_tunnel_init(struct net_device *dev)
|
|
|
|
{
|
|
|
|
struct ip_tunnel *tunnel;
|
|
|
|
struct iphdr *iph;
|
2012-09-26 20:48:50 -06:00
|
|
|
int err;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2006-01-08 23:05:26 -07:00
|
|
|
tunnel = netdev_priv(dev);
|
2005-04-16 16:20:36 -06:00
|
|
|
iph = &tunnel->parms.iph;
|
|
|
|
|
|
|
|
tunnel->dev = dev;
|
|
|
|
strcpy(tunnel->parms.name, dev->name);
|
|
|
|
|
|
|
|
memcpy(dev->dev_addr, &tunnel->parms.iph.saddr, 4);
|
|
|
|
memcpy(dev->broadcast, &tunnel->parms.iph.daddr, 4);
|
|
|
|
|
|
|
|
if (iph->daddr) {
|
|
|
|
#ifdef CONFIG_NET_IPGRE_BROADCAST
|
2007-12-16 14:45:43 -07:00
|
|
|
if (ipv4_is_multicast(iph->daddr)) {
|
2005-04-16 16:20:36 -06:00
|
|
|
if (!iph->saddr)
|
|
|
|
return -EINVAL;
|
|
|
|
dev->flags = IFF_BROADCAST;
|
2007-10-09 02:40:57 -06:00
|
|
|
dev->header_ops = &ipgre_header_ops;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
#endif
|
2007-12-13 10:46:32 -07:00
|
|
|
} else
|
2007-10-23 21:31:53 -06:00
|
|
|
dev->header_ops = &ipgre_header_ops;
|
2005-04-16 16:20:36 -06:00
|
|
|
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
dev->tstats = alloc_percpu(struct pcpu_tstats);
|
|
|
|
if (!dev->tstats)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2012-09-26 20:48:50 -06:00
|
|
|
err = gro_cells_init(&tunnel->gro_cells, dev);
|
|
|
|
if (err) {
|
|
|
|
free_percpu(dev->tstats);
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-11-20 21:34:29 -07:00
|
|
|
static void ipgre_fb_tunnel_init(struct net_device *dev)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2006-01-08 23:05:26 -07:00
|
|
|
struct ip_tunnel *tunnel = netdev_priv(dev);
|
2005-04-16 16:20:36 -06:00
|
|
|
struct iphdr *iph = &tunnel->parms.iph;
|
|
|
|
|
|
|
|
tunnel->dev = dev;
|
|
|
|
strcpy(tunnel->parms.name, dev->name);
|
|
|
|
|
|
|
|
iph->version = 4;
|
|
|
|
iph->protocol = IPPROTO_GRE;
|
|
|
|
iph->ihl = 5;
|
|
|
|
tunnel->hlen = sizeof(struct iphdr) + 4;
|
|
|
|
|
|
|
|
dev_hold(dev);
|
|
|
|
}
|
|
|
|
|
|
|
|
|
2010-08-22 00:05:39 -06:00
|
|
|
static const struct gre_protocol ipgre_protocol = {
|
|
|
|
.handler = ipgre_rcv,
|
|
|
|
.err_handler = ipgre_err,
|
2005-04-16 16:20:36 -06:00
|
|
|
};
|
|
|
|
|
2009-10-27 01:07:16 -06:00
|
|
|
static void ipgre_destroy_tunnels(struct ipgre_net *ign, struct list_head *head)
|
2008-04-16 02:10:26 -06:00
|
|
|
{
|
|
|
|
int prio;
|
|
|
|
|
|
|
|
for (prio = 0; prio < 4; prio++) {
|
|
|
|
int h;
|
|
|
|
for (h = 0; h < HASH_SIZE; h++) {
|
2010-09-15 05:07:53 -06:00
|
|
|
struct ip_tunnel *t;
|
|
|
|
|
|
|
|
t = rtnl_dereference(ign->tunnels[prio][h]);
|
2009-10-27 01:07:16 -06:00
|
|
|
|
|
|
|
while (t != NULL) {
|
|
|
|
unregister_netdevice_queue(t->dev, head);
|
2010-09-15 05:07:53 -06:00
|
|
|
t = rtnl_dereference(t->next);
|
2009-10-27 01:07:16 -06:00
|
|
|
}
|
2008-04-16 02:10:26 -06:00
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2010-01-16 20:35:32 -07:00
|
|
|
static int __net_init ipgre_init_net(struct net *net)
|
2008-04-16 02:08:53 -06:00
|
|
|
{
|
2009-11-29 08:46:13 -07:00
|
|
|
struct ipgre_net *ign = net_generic(net, ipgre_net_id);
|
2008-04-16 02:08:53 -06:00
|
|
|
int err;
|
|
|
|
|
2008-04-16 02:10:05 -06:00
|
|
|
ign->fb_tunnel_dev = alloc_netdev(sizeof(struct ip_tunnel), "gre0",
|
|
|
|
ipgre_tunnel_setup);
|
|
|
|
if (!ign->fb_tunnel_dev) {
|
|
|
|
err = -ENOMEM;
|
|
|
|
goto err_alloc_dev;
|
|
|
|
}
|
2008-11-23 18:26:26 -07:00
|
|
|
dev_net_set(ign->fb_tunnel_dev, net);
|
2008-04-16 02:10:05 -06:00
|
|
|
|
2008-11-20 21:34:29 -07:00
|
|
|
ipgre_fb_tunnel_init(ign->fb_tunnel_dev);
|
2008-10-09 12:59:55 -06:00
|
|
|
ign->fb_tunnel_dev->rtnl_link_ops = &ipgre_link_ops;
|
2008-04-16 02:10:05 -06:00
|
|
|
|
|
|
|
if ((err = register_netdev(ign->fb_tunnel_dev)))
|
|
|
|
goto err_reg_dev;
|
|
|
|
|
2010-10-30 17:21:28 -06:00
|
|
|
rcu_assign_pointer(ign->tunnels_wc[0],
|
|
|
|
netdev_priv(ign->fb_tunnel_dev));
|
2008-04-16 02:08:53 -06:00
|
|
|
return 0;
|
|
|
|
|
2008-04-16 02:10:05 -06:00
|
|
|
err_reg_dev:
|
2010-10-30 17:21:28 -06:00
|
|
|
ipgre_dev_free(ign->fb_tunnel_dev);
|
2008-04-16 02:10:05 -06:00
|
|
|
err_alloc_dev:
|
2008-04-16 02:08:53 -06:00
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
2010-01-16 20:35:32 -07:00
|
|
|
static void __net_exit ipgre_exit_net(struct net *net)
|
2008-04-16 02:08:53 -06:00
|
|
|
{
|
|
|
|
struct ipgre_net *ign;
|
2009-10-27 01:07:16 -06:00
|
|
|
LIST_HEAD(list);
|
2008-04-16 02:08:53 -06:00
|
|
|
|
|
|
|
ign = net_generic(net, ipgre_net_id);
|
2008-04-16 02:10:05 -06:00
|
|
|
rtnl_lock();
|
2009-10-27 01:07:16 -06:00
|
|
|
ipgre_destroy_tunnels(ign, &list);
|
|
|
|
unregister_netdevice_many(&list);
|
2008-04-16 02:10:05 -06:00
|
|
|
rtnl_unlock();
|
2008-04-16 02:08:53 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
static struct pernet_operations ipgre_net_ops = {
|
|
|
|
.init = ipgre_init_net,
|
|
|
|
.exit = ipgre_exit_net,
|
2009-11-29 08:46:13 -07:00
|
|
|
.id = &ipgre_net_id,
|
|
|
|
.size = sizeof(struct ipgre_net),
|
2008-04-16 02:08:53 -06:00
|
|
|
};
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2008-10-09 12:59:55 -06:00
|
|
|
static int ipgre_tunnel_validate(struct nlattr *tb[], struct nlattr *data[])
|
|
|
|
{
|
|
|
|
__be16 flags;
|
|
|
|
|
|
|
|
if (!data)
|
|
|
|
return 0;
|
|
|
|
|
|
|
|
flags = 0;
|
|
|
|
if (data[IFLA_GRE_IFLAGS])
|
|
|
|
flags |= nla_get_be16(data[IFLA_GRE_IFLAGS]);
|
|
|
|
if (data[IFLA_GRE_OFLAGS])
|
|
|
|
flags |= nla_get_be16(data[IFLA_GRE_OFLAGS]);
|
|
|
|
if (flags & (GRE_VERSION|GRE_ROUTING))
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
static int ipgre_tap_validate(struct nlattr *tb[], struct nlattr *data[])
|
|
|
|
{
|
|
|
|
__be32 daddr;
|
|
|
|
|
|
|
|
if (tb[IFLA_ADDRESS]) {
|
|
|
|
if (nla_len(tb[IFLA_ADDRESS]) != ETH_ALEN)
|
|
|
|
return -EINVAL;
|
|
|
|
if (!is_valid_ether_addr(nla_data(tb[IFLA_ADDRESS])))
|
|
|
|
return -EADDRNOTAVAIL;
|
|
|
|
}
|
|
|
|
|
|
|
|
if (!data)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
if (data[IFLA_GRE_REMOTE]) {
|
|
|
|
memcpy(&daddr, nla_data(data[IFLA_GRE_REMOTE]), 4);
|
|
|
|
if (!daddr)
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
|
|
|
|
|
|
|
out:
|
|
|
|
return ipgre_tunnel_validate(tb, data);
|
|
|
|
}
|
|
|
|
|
2008-10-09 12:59:55 -06:00
|
|
|
static void ipgre_netlink_parms(struct nlattr *data[],
|
|
|
|
struct ip_tunnel_parm *parms)
|
|
|
|
{
|
2008-10-11 13:20:15 -06:00
|
|
|
memset(parms, 0, sizeof(*parms));
|
2008-10-09 12:59:55 -06:00
|
|
|
|
|
|
|
parms->iph.protocol = IPPROTO_GRE;
|
|
|
|
|
|
|
|
if (!data)
|
|
|
|
return;
|
|
|
|
|
|
|
|
if (data[IFLA_GRE_LINK])
|
|
|
|
parms->link = nla_get_u32(data[IFLA_GRE_LINK]);
|
|
|
|
|
|
|
|
if (data[IFLA_GRE_IFLAGS])
|
|
|
|
parms->i_flags = nla_get_be16(data[IFLA_GRE_IFLAGS]);
|
|
|
|
|
|
|
|
if (data[IFLA_GRE_OFLAGS])
|
|
|
|
parms->o_flags = nla_get_be16(data[IFLA_GRE_OFLAGS]);
|
|
|
|
|
|
|
|
if (data[IFLA_GRE_IKEY])
|
|
|
|
parms->i_key = nla_get_be32(data[IFLA_GRE_IKEY]);
|
|
|
|
|
|
|
|
if (data[IFLA_GRE_OKEY])
|
|
|
|
parms->o_key = nla_get_be32(data[IFLA_GRE_OKEY]);
|
|
|
|
|
|
|
|
if (data[IFLA_GRE_LOCAL])
|
2008-10-10 13:11:06 -06:00
|
|
|
parms->iph.saddr = nla_get_be32(data[IFLA_GRE_LOCAL]);
|
2008-10-09 12:59:55 -06:00
|
|
|
|
|
|
|
if (data[IFLA_GRE_REMOTE])
|
2008-10-10 13:11:06 -06:00
|
|
|
parms->iph.daddr = nla_get_be32(data[IFLA_GRE_REMOTE]);
|
2008-10-09 12:59:55 -06:00
|
|
|
|
|
|
|
if (data[IFLA_GRE_TTL])
|
|
|
|
parms->iph.ttl = nla_get_u8(data[IFLA_GRE_TTL]);
|
|
|
|
|
|
|
|
if (data[IFLA_GRE_TOS])
|
|
|
|
parms->iph.tos = nla_get_u8(data[IFLA_GRE_TOS]);
|
|
|
|
|
|
|
|
if (!data[IFLA_GRE_PMTUDISC] || nla_get_u8(data[IFLA_GRE_PMTUDISC]))
|
|
|
|
parms->iph.frag_off = htons(IP_DF);
|
|
|
|
}
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
static int ipgre_tap_init(struct net_device *dev)
|
|
|
|
{
|
|
|
|
struct ip_tunnel *tunnel;
|
|
|
|
|
|
|
|
tunnel = netdev_priv(dev);
|
|
|
|
|
|
|
|
tunnel->dev = dev;
|
|
|
|
strcpy(tunnel->parms.name, dev->name);
|
|
|
|
|
|
|
|
ipgre_tunnel_bind_dev(dev);
|
|
|
|
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
dev->tstats = alloc_percpu(struct pcpu_tstats);
|
|
|
|
if (!dev->tstats)
|
|
|
|
return -ENOMEM;
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
2008-11-20 21:34:29 -07:00
|
|
|
static const struct net_device_ops ipgre_tap_netdev_ops = {
|
|
|
|
.ndo_init = ipgre_tap_init,
|
|
|
|
.ndo_uninit = ipgre_tunnel_uninit,
|
|
|
|
.ndo_start_xmit = ipgre_tunnel_xmit,
|
|
|
|
.ndo_set_mac_address = eth_mac_addr,
|
|
|
|
.ndo_validate_addr = eth_validate_addr,
|
|
|
|
.ndo_change_mtu = ipgre_tunnel_change_mtu,
|
2012-04-12 00:31:16 -06:00
|
|
|
.ndo_get_stats64 = ipgre_get_stats64,
|
2008-11-20 21:34:29 -07:00
|
|
|
};
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
static void ipgre_tap_setup(struct net_device *dev)
|
|
|
|
{
|
|
|
|
|
|
|
|
ether_setup(dev);
|
|
|
|
|
2009-10-29 23:51:48 -06:00
|
|
|
dev->netdev_ops = &ipgre_tap_netdev_ops;
|
ip_gre: percpu stats accounting
Le lundi 27 septembre 2010 à 14:29 +0100, Ben Hutchings a écrit :
> > diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> > index 5d6ddcb..de39b22 100644
> > --- a/net/ipv4/ip_gre.c
> > +++ b/net/ipv4/ip_gre.c
> [...]
> > @@ -377,7 +405,7 @@ static struct ip_tunnel *ipgre_tunnel_locate(struct net *net,
> > if (parms->name[0])
> > strlcpy(name, parms->name, IFNAMSIZ);
> > else
> > - sprintf(name, "gre%%d");
> > + strcpy(name, "gre%d");
> >
> > dev = alloc_netdev(sizeof(*t), name, ipgre_tunnel_setup);
> > if (!dev)
> [...]
>
> This is a valid fix, but doesn't belong in this patch!
>
Sorry ? It was not a fix, but at most a cleanup ;)
Anyway I forgot the gretap case...
[PATCH 2/4 v2] ip_gre: percpu stats accounting
Maintain per_cpu tx_bytes, tx_packets, rx_bytes, rx_packets.
Other seldom used fields are kept in netdev->stats structure, possibly
unsafe.
This is a preliminary work to support lockless transmit path, and
correct RX stats, that are already unsafe.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2010-09-26 21:57:11 -06:00
|
|
|
dev->destructor = ipgre_dev_free;
|
2008-10-09 13:00:17 -06:00
|
|
|
|
|
|
|
dev->iflink = 0;
|
|
|
|
dev->features |= NETIF_F_NETNS_LOCAL;
|
2013-02-14 07:02:41 -07:00
|
|
|
|
|
|
|
dev->features |= GRE_FEATURES;
|
|
|
|
dev->hw_features |= GRE_FEATURES;
|
2008-10-09 13:00:17 -06:00
|
|
|
}
|
|
|
|
|
2009-11-08 01:53:51 -07:00
|
|
|
static int ipgre_newlink(struct net *src_net, struct net_device *dev, struct nlattr *tb[],
|
2008-10-09 12:59:55 -06:00
|
|
|
struct nlattr *data[])
|
|
|
|
{
|
|
|
|
struct ip_tunnel *nt;
|
|
|
|
struct net *net = dev_net(dev);
|
|
|
|
struct ipgre_net *ign = net_generic(net, ipgre_net_id);
|
|
|
|
int mtu;
|
|
|
|
int err;
|
|
|
|
|
|
|
|
nt = netdev_priv(dev);
|
|
|
|
ipgre_netlink_parms(data, &nt->parms);
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
if (ipgre_tunnel_find(net, &nt->parms, dev->type))
|
2008-10-09 12:59:55 -06:00
|
|
|
return -EEXIST;
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
if (dev->type == ARPHRD_ETHER && !tb[IFLA_ADDRESS])
|
2012-02-14 23:45:39 -07:00
|
|
|
eth_hw_addr_random(dev);
|
2008-10-09 13:00:17 -06:00
|
|
|
|
2008-10-09 12:59:55 -06:00
|
|
|
mtu = ipgre_tunnel_bind_dev(dev);
|
|
|
|
if (!tb[IFLA_MTU])
|
|
|
|
dev->mtu = mtu;
|
|
|
|
|
2010-09-27 17:05:47 -06:00
|
|
|
/* Can use a lockless transmit, unless we generate output sequences */
|
|
|
|
if (!(nt->parms.o_flags & GRE_SEQ))
|
|
|
|
dev->features |= NETIF_F_LLTX;
|
|
|
|
|
2008-10-09 12:59:55 -06:00
|
|
|
err = register_netdevice(dev);
|
|
|
|
if (err)
|
|
|
|
goto out;
|
|
|
|
|
|
|
|
dev_hold(dev);
|
|
|
|
ipgre_tunnel_link(ign, nt);
|
|
|
|
|
|
|
|
out:
|
|
|
|
return err;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ipgre_changelink(struct net_device *dev, struct nlattr *tb[],
|
|
|
|
struct nlattr *data[])
|
|
|
|
{
|
|
|
|
struct ip_tunnel *t, *nt;
|
|
|
|
struct net *net = dev_net(dev);
|
|
|
|
struct ipgre_net *ign = net_generic(net, ipgre_net_id);
|
|
|
|
struct ip_tunnel_parm p;
|
|
|
|
int mtu;
|
|
|
|
|
|
|
|
if (dev == ign->fb_tunnel_dev)
|
|
|
|
return -EINVAL;
|
|
|
|
|
|
|
|
nt = netdev_priv(dev);
|
|
|
|
ipgre_netlink_parms(data, &p);
|
|
|
|
|
|
|
|
t = ipgre_tunnel_locate(net, &p, 0);
|
|
|
|
|
|
|
|
if (t) {
|
|
|
|
if (t->dev != dev)
|
|
|
|
return -EEXIST;
|
|
|
|
} else {
|
|
|
|
t = nt;
|
|
|
|
|
2009-10-29 23:51:48 -06:00
|
|
|
if (dev->type != ARPHRD_ETHER) {
|
2010-09-15 05:07:53 -06:00
|
|
|
unsigned int nflags = 0;
|
2008-10-09 12:59:55 -06:00
|
|
|
|
2009-10-29 23:51:48 -06:00
|
|
|
if (ipv4_is_multicast(p.iph.daddr))
|
|
|
|
nflags = IFF_BROADCAST;
|
|
|
|
else if (p.iph.daddr)
|
|
|
|
nflags = IFF_POINTOPOINT;
|
|
|
|
|
|
|
|
if ((dev->flags ^ nflags) &
|
|
|
|
(IFF_POINTOPOINT | IFF_BROADCAST))
|
|
|
|
return -EINVAL;
|
|
|
|
}
|
2008-10-09 12:59:55 -06:00
|
|
|
|
|
|
|
ipgre_tunnel_unlink(ign, t);
|
|
|
|
t->parms.iph.saddr = p.iph.saddr;
|
|
|
|
t->parms.iph.daddr = p.iph.daddr;
|
|
|
|
t->parms.i_key = p.i_key;
|
2009-10-29 23:51:48 -06:00
|
|
|
if (dev->type != ARPHRD_ETHER) {
|
|
|
|
memcpy(dev->dev_addr, &p.iph.saddr, 4);
|
|
|
|
memcpy(dev->broadcast, &p.iph.daddr, 4);
|
|
|
|
}
|
2008-10-09 12:59:55 -06:00
|
|
|
ipgre_tunnel_link(ign, t);
|
|
|
|
netdev_state_change(dev);
|
|
|
|
}
|
|
|
|
|
|
|
|
t->parms.o_key = p.o_key;
|
|
|
|
t->parms.iph.ttl = p.iph.ttl;
|
|
|
|
t->parms.iph.tos = p.iph.tos;
|
|
|
|
t->parms.iph.frag_off = p.iph.frag_off;
|
|
|
|
|
|
|
|
if (t->parms.link != p.link) {
|
|
|
|
t->parms.link = p.link;
|
|
|
|
mtu = ipgre_tunnel_bind_dev(dev);
|
|
|
|
if (!tb[IFLA_MTU])
|
|
|
|
dev->mtu = mtu;
|
|
|
|
netdev_state_change(dev);
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static size_t ipgre_get_size(const struct net_device *dev)
|
|
|
|
{
|
|
|
|
return
|
|
|
|
/* IFLA_GRE_LINK */
|
|
|
|
nla_total_size(4) +
|
|
|
|
/* IFLA_GRE_IFLAGS */
|
|
|
|
nla_total_size(2) +
|
|
|
|
/* IFLA_GRE_OFLAGS */
|
|
|
|
nla_total_size(2) +
|
|
|
|
/* IFLA_GRE_IKEY */
|
|
|
|
nla_total_size(4) +
|
|
|
|
/* IFLA_GRE_OKEY */
|
|
|
|
nla_total_size(4) +
|
|
|
|
/* IFLA_GRE_LOCAL */
|
|
|
|
nla_total_size(4) +
|
|
|
|
/* IFLA_GRE_REMOTE */
|
|
|
|
nla_total_size(4) +
|
|
|
|
/* IFLA_GRE_TTL */
|
|
|
|
nla_total_size(1) +
|
|
|
|
/* IFLA_GRE_TOS */
|
|
|
|
nla_total_size(1) +
|
|
|
|
/* IFLA_GRE_PMTUDISC */
|
|
|
|
nla_total_size(1) +
|
|
|
|
0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static int ipgre_fill_info(struct sk_buff *skb, const struct net_device *dev)
|
|
|
|
{
|
|
|
|
struct ip_tunnel *t = netdev_priv(dev);
|
|
|
|
struct ip_tunnel_parm *p = &t->parms;
|
|
|
|
|
2012-04-01 18:39:02 -06:00
|
|
|
if (nla_put_u32(skb, IFLA_GRE_LINK, p->link) ||
|
|
|
|
nla_put_be16(skb, IFLA_GRE_IFLAGS, p->i_flags) ||
|
|
|
|
nla_put_be16(skb, IFLA_GRE_OFLAGS, p->o_flags) ||
|
|
|
|
nla_put_be32(skb, IFLA_GRE_IKEY, p->i_key) ||
|
|
|
|
nla_put_be32(skb, IFLA_GRE_OKEY, p->o_key) ||
|
|
|
|
nla_put_be32(skb, IFLA_GRE_LOCAL, p->iph.saddr) ||
|
|
|
|
nla_put_be32(skb, IFLA_GRE_REMOTE, p->iph.daddr) ||
|
|
|
|
nla_put_u8(skb, IFLA_GRE_TTL, p->iph.ttl) ||
|
|
|
|
nla_put_u8(skb, IFLA_GRE_TOS, p->iph.tos) ||
|
|
|
|
nla_put_u8(skb, IFLA_GRE_PMTUDISC,
|
|
|
|
!!(p->iph.frag_off & htons(IP_DF))))
|
|
|
|
goto nla_put_failure;
|
2008-10-09 12:59:55 -06:00
|
|
|
return 0;
|
|
|
|
|
|
|
|
nla_put_failure:
|
|
|
|
return -EMSGSIZE;
|
|
|
|
}
|
|
|
|
|
|
|
|
static const struct nla_policy ipgre_policy[IFLA_GRE_MAX + 1] = {
|
|
|
|
[IFLA_GRE_LINK] = { .type = NLA_U32 },
|
|
|
|
[IFLA_GRE_IFLAGS] = { .type = NLA_U16 },
|
|
|
|
[IFLA_GRE_OFLAGS] = { .type = NLA_U16 },
|
|
|
|
[IFLA_GRE_IKEY] = { .type = NLA_U32 },
|
|
|
|
[IFLA_GRE_OKEY] = { .type = NLA_U32 },
|
2008-10-10 13:11:06 -06:00
|
|
|
[IFLA_GRE_LOCAL] = { .len = FIELD_SIZEOF(struct iphdr, saddr) },
|
|
|
|
[IFLA_GRE_REMOTE] = { .len = FIELD_SIZEOF(struct iphdr, daddr) },
|
2008-10-09 12:59:55 -06:00
|
|
|
[IFLA_GRE_TTL] = { .type = NLA_U8 },
|
|
|
|
[IFLA_GRE_TOS] = { .type = NLA_U8 },
|
|
|
|
[IFLA_GRE_PMTUDISC] = { .type = NLA_U8 },
|
|
|
|
};
|
|
|
|
|
|
|
|
static struct rtnl_link_ops ipgre_link_ops __read_mostly = {
|
|
|
|
.kind = "gre",
|
|
|
|
.maxtype = IFLA_GRE_MAX,
|
|
|
|
.policy = ipgre_policy,
|
|
|
|
.priv_size = sizeof(struct ip_tunnel),
|
|
|
|
.setup = ipgre_tunnel_setup,
|
|
|
|
.validate = ipgre_tunnel_validate,
|
|
|
|
.newlink = ipgre_newlink,
|
|
|
|
.changelink = ipgre_changelink,
|
|
|
|
.get_size = ipgre_get_size,
|
|
|
|
.fill_info = ipgre_fill_info,
|
|
|
|
};
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
static struct rtnl_link_ops ipgre_tap_ops __read_mostly = {
|
|
|
|
.kind = "gretap",
|
|
|
|
.maxtype = IFLA_GRE_MAX,
|
|
|
|
.policy = ipgre_policy,
|
|
|
|
.priv_size = sizeof(struct ip_tunnel),
|
|
|
|
.setup = ipgre_tap_setup,
|
|
|
|
.validate = ipgre_tap_validate,
|
|
|
|
.newlink = ipgre_newlink,
|
|
|
|
.changelink = ipgre_changelink,
|
|
|
|
.get_size = ipgre_get_size,
|
|
|
|
.fill_info = ipgre_fill_info,
|
|
|
|
};
|
|
|
|
|
2005-04-16 16:20:36 -06:00
|
|
|
/*
|
|
|
|
* And now the modules code and kernel interface.
|
|
|
|
*/
|
|
|
|
|
|
|
|
static int __init ipgre_init(void)
|
|
|
|
{
|
|
|
|
int err;
|
|
|
|
|
2012-03-11 12:36:11 -06:00
|
|
|
pr_info("GRE over IPv4 tunneling driver\n");
|
2005-04-16 16:20:36 -06:00
|
|
|
|
2009-11-29 08:46:13 -07:00
|
|
|
err = register_pernet_device(&ipgre_net_ops);
|
2008-04-16 02:08:53 -06:00
|
|
|
if (err < 0)
|
2010-02-16 00:57:44 -07:00
|
|
|
return err;
|
|
|
|
|
2010-08-22 00:05:39 -06:00
|
|
|
err = gre_add_protocol(&ipgre_protocol, GREPROTO_CISCO);
|
2010-02-16 00:57:44 -07:00
|
|
|
if (err < 0) {
|
2012-03-11 12:36:11 -06:00
|
|
|
pr_info("%s: can't add protocol\n", __func__);
|
2010-02-16 00:57:44 -07:00
|
|
|
goto add_proto_failed;
|
|
|
|
}
|
2008-04-16 02:10:05 -06:00
|
|
|
|
2008-10-09 12:59:55 -06:00
|
|
|
err = rtnl_link_register(&ipgre_link_ops);
|
|
|
|
if (err < 0)
|
|
|
|
goto rtnl_link_failed;
|
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
err = rtnl_link_register(&ipgre_tap_ops);
|
|
|
|
if (err < 0)
|
|
|
|
goto tap_ops_failed;
|
|
|
|
|
2008-10-09 12:59:55 -06:00
|
|
|
out:
|
2005-04-16 16:20:36 -06:00
|
|
|
return err;
|
2008-10-09 12:59:55 -06:00
|
|
|
|
2008-10-09 13:00:17 -06:00
|
|
|
tap_ops_failed:
|
|
|
|
rtnl_link_unregister(&ipgre_link_ops);
|
2008-10-09 12:59:55 -06:00
|
|
|
rtnl_link_failed:
|
2010-08-22 00:05:39 -06:00
|
|
|
gre_del_protocol(&ipgre_protocol, GREPROTO_CISCO);
|
2010-02-16 00:57:44 -07:00
|
|
|
add_proto_failed:
|
|
|
|
unregister_pernet_device(&ipgre_net_ops);
|
2008-10-09 12:59:55 -06:00
|
|
|
goto out;
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
2005-07-30 18:46:44 -06:00
|
|
|
static void __exit ipgre_fini(void)
|
2005-04-16 16:20:36 -06:00
|
|
|
{
|
2008-10-09 13:00:17 -06:00
|
|
|
rtnl_link_unregister(&ipgre_tap_ops);
|
2008-10-09 12:59:55 -06:00
|
|
|
rtnl_link_unregister(&ipgre_link_ops);
|
2010-08-22 00:05:39 -06:00
|
|
|
if (gre_del_protocol(&ipgre_protocol, GREPROTO_CISCO) < 0)
|
2012-03-11 12:36:11 -06:00
|
|
|
pr_info("%s: can't remove protocol\n", __func__);
|
2010-02-16 00:57:44 -07:00
|
|
|
unregister_pernet_device(&ipgre_net_ops);
|
2005-04-16 16:20:36 -06:00
|
|
|
}
|
|
|
|
|
|
|
|
module_init(ipgre_init);
|
|
|
|
module_exit(ipgre_fini);
|
|
|
|
MODULE_LICENSE("GPL");
|
2008-10-10 13:11:06 -06:00
|
|
|
MODULE_ALIAS_RTNL_LINK("gre");
|
|
|
|
MODULE_ALIAS_RTNL_LINK("gretap");
|
2011-03-01 14:33:13 -07:00
|
|
|
MODULE_ALIAS_NETDEV("gre0");
|