Always lookup to see if we have an existing inetpeer entry for
a route. Let FLOWI_FLAG_PRECOW_METRICS merely influence the
"create" argument to rt_bind_peer().
Also, call rt_bind_peer() unconditionally since it is not
possible for rt->peer to be non-NULL at this point.
Signed-off-by: David S. Miller <davem@davemloft.net>
Both fib_trie and fib_hash have a local implementation of
fib_table_select_default(). This is completely unnecessary
code duplication.
Since we now remember the fib_table and the head of the fib
alias list of the default route, we can implement one single
generic version of this routine.
Looking at the fib_hash implementation you may get the impression
that it's possible for there to be multiple top-level routes in
the table for the default route. The truth is, it isn't, the
insert code will only allow one entry to exist in the zero
prefix hash table, because all keys evaluate to zero and all
keys in a hash table must be unique.
Signed-off-by: David S. Miller <davem@davemloft.net>
When an IPSEC SA is still being set up, __xfrm_lookup() will return
-EREMOTE and so ip_route_output_flow() will return a blackhole route.
This can happen in a sndmsg call, and after d33e455337 ("net: Abstract
default MTU metric calculation behind an accessor.") this leads to a
crash in ip_append_data() because the blackhole dst_ops have no
default_mtu() method and so dst_mtu() calls a NULL pointer.
Fix this by adding default_mtu() methods (that simply return 0, matching
the old behavior) to the blackhole dst_ops.
The IPv4 part of this patch fixes a crash that I saw when using an IPSEC
VPN; the IPv6 part is untested because I don't have an IPv6 VPN, but it
looks to be needed as well.
Signed-off-by: Roland Dreier <roland@purestorage.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The fib metric memory in this case is static in the kernel image,
so we don't need to reference count it since it's never going
to go away on us.
Signed-off-by: David S. Miller <davem@davemloft.net>
TCP is going to record metrics for the connection,
so pre-COW the route metrics at route cache entry
creation time.
This avoids several atomic operations that have to
occur if we COW the metrics after the entry reaches
global visibility.
Signed-off-by: David S. Miller <davem@davemloft.net>
Please note that the IPSEC dst entry metrics keep using
the generic metrics COW'ing mechanism using kmalloc/kfree.
This gives the IPSEC routes an opportunity to use metrics
which are unique to their encapsulated paths.
Signed-off-by: David S. Miller <davem@davemloft.net>
Routing metrics are now copy-on-write.
Initially a route entry points it's metrics at a read-only location.
If a routing table entry exists, it will point there. Else it will
point at the all zero metric place-holder called 'dst_default_metrics'.
The writeability state of the metrics is stored in the low bits of the
metrics pointer, we have two bits left to spare if we want to store
more states.
For the initial implementation, COW is implemented simply via kmalloc.
However future enhancements will change this to place the writable
metrics somewhere else, in order to increase sharing. Very likely
this "somewhere else" will be the inetpeer cache.
Note also that this means that metrics updates may transiently fail
if we cannot COW the metrics successfully.
But even by itself, this patch should decrease memory usage and
increase cache locality especially for routing workloads. In those
cases the read-only metric copies stay in place and never get written
to.
TCP workloads where metrics get updated, and those rare cases where
PMTU triggers occur, will take a very slight performance hit. But
that hit will be alleviated when the long-term writable metrics
move to a more sharable location.
Since the metrics storage went from a u32 array of RTAX_MAX entries to
what is essentially a pointer, some retooling of the dst_entry layout
was necessary.
Most importantly, we need to preserve the alignment of the reference
count so that it doesn't share cache lines with the read-mostly state,
as per Eric Dumazet's alignment assertion checks.
The only non-trivial bit here is the move of the 'flags' member into
the writeable cacheline. This is OK since we are always accessing the
flags around the same moment when we made a modification to the
reference count.
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix dependencies of netfilter realm match: it depends on NET_CLS_ROUTE,
which itself depends on NET_SCHED; this dependency is missing from netfilter.
Since matching on realms is also useful without having NET_SCHED enabled and
the option really only controls whether the tclassid member is included in
route and dst entries, rename the config option to IP_ROUTE_CLASSID and move
it outside of traffic scheduling context to get rid of the NET_SCHED dependeny.
Reported-by: Vladis Kletnieks <Valdis.Kletnieks@vt.edu>
Signed-off-by: Patrick McHardy <kaber@trash.net>
The preferred source address is currently ignored for local routes,
which results in all local connections having a src address that is the
same as the local dst address. Fix this by respecting the preferred source
address when it is provided for local routes.
This bug can be demonstrated as follows:
# ifconfig dummy0 192.168.0.1
# ip route show table local | grep local.*dummy0
local 192.168.0.1 dev dummy0 proto kernel scope host src 192.168.0.1
# ip route change table local local 192.168.0.1 dev dummy0 \
proto kernel scope host src 127.0.0.1
# ip route show table local | grep local.*dummy0
local 192.168.0.1 dev dummy0 proto kernel scope host src 127.0.0.1
We now establish a local connection and verify the source IP
address selection:
# nc -l 192.168.0.1 3128 &
# nc 192.168.0.1 3128 &
# netstat -ant | grep 192.168.0.1:3128.*EST
tcp 0 0 192.168.0.1:3128 192.168.0.1:33228 ESTABLISHED
tcp 0 0 192.168.0.1:33228 192.168.0.1:3128 ESTABLISHED
Signed-off-by: Joel Sing <jsing@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In ip_route_output_slow(), instead of allowing a route to be created on
a not UPed device, report -ENETUNREACH immediately.
# ip tunnel add mode ipip remote 10.16.0.164 local
10.16.0.72 dev eth0
# (Note : tunl1 is down)
# ping -I tunl1 10.1.2.3
PING 10.1.2.3 (10.1.2.3) from 192.168.18.5 tunl1: 56(84) bytes of data.
(nothing)
# ./a.out tunl1
# ip tunnel del tunl1
Message from syslogd@shelby at Dec 22 10:12:08 ...
kernel: unregister_netdevice: waiting for tunl1 to become free.
Usage count = 3
After patch:
# ping -I tunl1 10.1.2.3
connect: Network is unreachable
Reported-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Reviewed-by: Octavian Purdila <opurdila@ixiacom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Flush the routing cache only of entries that match the
network namespace in which the purge event occurred.
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Like RTAX_ADVMSS, make the default calculation go through a dst_ops
method rather than caching the computation in the routing cache
entries.
Now dst metrics are pretty much left as-is when new entries are
created, thus optimizing metric sharing becomes a real possibility.
Signed-off-by: David S. Miller <davem@davemloft.net>
Make all RTAX_ADVMSS metric accesses go through a new helper function,
dst_metric_advmss().
Leave the actual default metric as "zero" in the real metric slot,
and compute the actual default value dynamically via a new dst_ops
AF specific callback.
For stacked IPSEC routes, we use the advmss of the path which
preserves existing behavior.
Unlike ipv4/ipv6, DecNET ties the advmss to the mtu and thus updates
advmss on pmtu updates. This inconsistency in advmss handling
results in more raw metric accesses than I wish we ended up with.
Signed-off-by: David S. Miller <davem@davemloft.net>
Always go through a new ip4_dst_hoplimit() helper, just like ipv6.
This allowed several simplifications:
1) The interim dst_metric_hoplimit() can go as it's no longer
userd.
2) The sysctl_ip_default_ttl entry no longer needs to use
ipv4_doint_and_flush, since the sysctl is not cached in
routing cache metrics any longer.
3) ipv4_doint_and_flush no longer needs to be exported and
therefore can be marked static.
When ipv4_doint_and_flush_strategy was removed some time ago,
the external declaration in ip.h was mistakenly left around
so kill that off too.
We have to move the sysctl_ip_default_ttl declaration into
ipv4's route cache definition header net/route.h, because
currently net/ip.h (where the declaration lives now) has
a back dependency on net/route.h
Signed-off-by: David S. Miller <davem@davemloft.net>
Use helper functions to hide all direct accesses, especially writes,
to dst_entry metrics values.
This will allow us to:
1) More easily change how the metrics are stored.
2) Implement COW for metrics.
In particular this will help us put metrics into the inetpeer
cache if that is what we end up doing. We can make the _metrics
member a pointer instead of an array, initially have it point
at the read-only metrics in the FIB, and then on the first set
grab an inetpeer entry and point the _metrics member there.
Signed-off-by: David S. Miller <davem@davemloft.net>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Use the macros defined for the members of flowi to clean the code up.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When we test rt->fl.iif against zero, we're seeing if it's
an output or an input route.
Make that explicit with some helper functions.
Signed-off-by: David S. Miller <davem@davemloft.net>
It seems idev field in struct rtable has no special purpose, but adding
extra atomic ops.
We hold refcounts on the device itself (using percpu data, so pretty
cheap in current kernel).
infiniband case is solved using dst.dev instead of idev->dev
Removal of this field means routing without route cache is now using
shared data, percpu data, and only potential contention is a pair of
atomic ops on struct neighbour per forwarded packet.
About 5% speedup on routing test.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Roland Dreier <rolandd@cisco.com>
Cc: Sean Hefty <sean.hefty@intel.com>
Cc: Hal Rosenstock <hal.rosenstock@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add __rcu annotations to :
(struct dst_entry)->rt_next
(struct rt_hash_bucket)->chain
And use appropriate rcu primitives to reduce sparse warnings if
CONFIG_SPARSE_RCU_POINTER=y
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is no point using RCU for dst we allocate for a very short time
(used once).
Change dst_release() to take DST_NOCACHE into account, but also change
skb_dst_set_noref() to force a refcount increment for such dst.
This is a _huge_ gain, because we dont waste memory to store xx thousand
of dsts. Instead of queueing them to RCU, we can free them instantly.
CPU caches can stay hot, re-using same memory blocks to hold temporary
dsts.
Note : remove unneeded smp_mb__before_atomic_dec(); in dst_release(),
since atomic_dec_return() implies a full memory barrier.
Stress test, 160.000.000 udp frames sent, IP route cache disabled
(DDOS).
Before:
real 0m38.091s
user 0m13.189s
sys 7m53.018s
After:
real 0m29.946s
user 0m12.157s
sys 7m40.605s
For reference, if IP route cache was enabled :
real 0m32.030s
user 0m10.521s
sys 8m15.243s
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change a few checks against the hardcoded broadcast address,
0xffffffff, to ipv4_is_lbcast(). Remove some existing checks
using ipv4_is_lbcast() that are now obviously superfluous.
Signed-off-by: Andy Walls <awalls@md.metrocast.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
struct dst_ops tracks number of allocated dst in an atomic_t field,
subject to high cache line contention in stress workload.
Switch to a percpu_counter, to reduce number of time we need to dirty a
central location. Place it on a separate cache line to avoid dirtying
read only fields.
Stress test :
(Sending 160.000.000 UDP frames,
IP route cache disabled, dual E5540 @2.53GHz,
32bit kernel, FIB_TRIE, SLUB/NUMA)
Before:
real 0m51.179s
user 0m15.329s
sys 10m15.942s
After:
real 0m45.570s
user 0m15.525s
sys 9m56.669s
With a small reordering of struct neighbour fields, subject of a
following patch, (to separate refcnt from other read mostly fields)
real 0m41.841s
user 0m15.261s
sys 8m45.949s
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit "fib: RCU conversion of fib_lookup()" removed rcu_read_lock() from
__mkroute_output but left a couple of calls to rcu_read_unlock() in there.
This causes lockdep to complain that the rcu_read_unlock() call in
__ip_route_output_key causes a lock inbalance and quickly crashes the
kernel. The below fixes this for me.
Signed-off-by: Dimitris Michailidis <dm@chelsio.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While doing stress tests with IP route cache disabled, and multi queue
devices, I noticed a very high contention on one rwlock used in
neighbour code.
When many cpus are trying to send frames (possibly using a high
performance multiqueue device) to the same neighbour, they fight for the
neigh->lock rwlock in order to call neigh_hh_init(), and fight on
hh->hh_refcnt (a pair of atomic_inc/atomic_dec_and_test())
But we dont need to call neigh_hh_init() for dst that are used only
once. It costs four atomic operations at least, on two contended cache
lines, plus the high contention on neigh->lock rwlock.
Introduce a new dst flag, DST_NOCACHE, that is set when dst was not
inserted in route cache.
With the stress test bench, sending 160000000 frames on one neighbour,
results are :
Before patch:
real 2m28.406s
user 0m11.781s
sys 36m17.964s
After patch:
real 1m26.532s
user 0m12.185s
sys 20m3.903s
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While doing stress tests with a disabled IP route cache, I found
__mkroute_output() was touching three times in_device atomic refcount.
Use RCU to touch it once to reduce cache line ping pongs.
Before patch
time to perform the test
real 1m42.009s
user 0m12.545s
sys 25m0.726s
Profile :
16109.00 26.4% ip_route_output_slow vmlinux
7434.00 12.2% dst_destroy vmlinux
3280.00 5.4% fib_rules_lookup vmlinux
3252.00 5.3% fib_semantic_match vmlinux
2622.00 4.3% fib_table_lookup vmlinux
2535.00 4.1% dst_alloc vmlinux
1750.00 2.9% _raw_read_lock vmlinux
1532.00 2.5% rt_set_nexthop vmlinux
After patch
real 1m36.503s
user 0m12.977s
sys 23m25.608s
14234.00 22.4% ip_route_output_slow vmlinux
8717.00 13.7% dst_destroy vmlinux
4052.00 6.4% fib_rules_lookup vmlinux
3951.00 6.2% fib_semantic_match vmlinux
3191.00 5.0% dst_alloc vmlinux
1764.00 2.8% fib_table_lookup vmlinux
1692.00 2.7% _raw_read_lock vmlinux
1605.00 2.5% rt_set_nexthop vmlinux
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
IPv4 and IPv6 have separate neighbour tables, so
the warning messages should be distinguishable.
[ Add a suitable message prefix on the ipv4 side as well -DaveM ]
Signed-off-by: Ulrich Weber <uweber@astaro.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
__in_dev_get_rtnl(dev_out) is called while RTNL is not held, thus
triggers a lockdep fault.
At this point, we only perform a raw test of dev_out->ip_ptr being NULL,
we dont need to make sure ip_ptr cant changed right after.
We can use rcu_dereference_raw() for this.
Reported-by: Andrew Morton <akpm@linux-foundation.org>
Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change "return (EXPR);" to "return EXPR;"
return is not a function, parentheses are not required.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Blackhole routes are used when xfrm_lookup() returns -EREMOTE (error
triggered by IKE for example), hence this kind of route is always
temporary and so we should check if a better route exists for next
packets.
Bug has been introduced by commit d11a4dc18b.
Signed-off-by: Jianzhao Wang <jianzhao.wang@6wind.com>
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now cmpxchg() is available on all arches, we can use it in
build_ehash_secret() and rt_bind_peer() instead of using spinlocks.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
CC: Mathieu Desnoyers <mathieu.desnoyers@polymtl.ca>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add a new rt attribute, RTA_MARK, and use it in
rt_fill_info()/inet_rtm_getroute() to support following commands :
ip route get 192.168.20.110 mark NUMBER
ip route get 192.168.20.108 from 192.168.20.110 iif eth1 mark NUMBER
ip route list cache [192.168.20.110] mark NUMBER
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
CodingStyle cleanups
EXPORT_SYMBOL should immediately follow the symbol declaration.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Addition of rcu_head to struct inet_peer added 16bytes on 64bit arches.
Thats a bit unfortunate, since old size was exactly 64 bytes.
This can be solved, using an union between this rcu_head an four fields,
that are normally used only when a refcount is taken on inet_peer.
rcu_head is used only when refcnt=-1, right before structure freeing.
Add a inet_peer_refcheck() function to check this assertion for a while.
We can bring back SLAB_HWCACHE_ALIGN qualifier in kmem cache creation.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
remove useless union keyword in rtable, rt6_info and dn_route.
Since there is only one member in a union, the union keyword isn't useful.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
in_dev_get() -> __in_dev_get_rcu() in a rcu protected function.
[ Fix build with CONFIG_IP_ROUTE_VERBOSE disabled. -DaveM ]
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Avoid two atomic ops on output device refcount
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Avoid two atomic ops on struct in_device refcount per incoming packet,
if slow path taken, (or route cache disabled)
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Christoph Lameter mentioned that packets could be dropped in input path
because of rp_filter settings, without any SNMP counter being
incremented. System administrator can have a hard time to track the
problem.
This patch introduces a new counter, LINUX_MIB_IPRPFILTER, incremented
each time we drop a packet because Reverse Path Filter triggers.
(We receive an IPv4 datagram on a given interface, and find the route to
send an answer would use another interface)
netstat -s | grep IPReversePathFilter
IPReversePathFilter: 21714
Reported-by: Christoph Lameter <cl@linux-foundation.org>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch saves 224 bytes of text on my machine.
__this_cpu_inc() generates a single instruction, using no scratch
registers :
65 ff 04 25 a8 30 01 00 incl %gs:0x130a8
instead of :
48 c7 c2 80 30 01 00 mov $0x13080,%rdx
65 48 8b 04 25 88 ea 00 00 mov %gs:0xea88,%rax
83 44 10 28 01 addl $0x1,0x28(%rax,%rdx,1)
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
ip_route_input() is the version returning a refcounted dst, while
ip_route_input_noref() returns a non refcounted one.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller