We know for sure pktgen is going to write skb->data right after
*_alloc_skb, causing unnecessary cache misses.
Idea is to add a prefetchw() call to prefetch the first cache line
indicated by skb->data. On systems with Adjacent Cache Line Prefetch,
it's probably two cache lines are prefetched.
With this patch, pktgen on Intel SR1625 server with two E5530
quad-core processors and a single ixgbe-based NIC went from 8.63Mpps
to 9.03Mpps, with 4.6% improvement.
Signed-off-by: Junchang Wang <junchangwang@gmail.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
__load_pointer() checks data we fetch from skb is included in head
portion, but assumes we fetch one byte, instead of up to four.
This wont crash because we have extra bytes (struct skb_shared_info)
after head, but this can read uninitialized bytes.
Fix this using size of the data (1, 2, 4 bytes) in the test.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Followup of commit b178bb3dfc (net: reorder struct sock fields)
Optimize INET input path a bit further, by :
1) moving sk_refcnt close to sk_lock.
This reduces number of dirtied cache lines by one on 64bit arches (and
64 bytes cache line size).
2) moving inet_daddr & inet_rcv_saddr at the beginning of sk
(same cache line than hash / family / bound_dev_if / nulls_node)
This reduces number of accessed cache lines in lookups by one, and dont
increase size of inet and timewait socks.
inet and tw sockets now share same place-holder for these fields.
Before patch :
offsetof(struct sock, sk_refcnt) = 0x10
offsetof(struct sock, sk_lock) = 0x40
offsetof(struct sock, sk_receive_queue) = 0x60
offsetof(struct inet_sock, inet_daddr) = 0x270
offsetof(struct inet_sock, inet_rcv_saddr) = 0x274
After patch :
offsetof(struct sock, sk_refcnt) = 0x44
offsetof(struct sock, sk_lock) = 0x48
offsetof(struct sock, sk_receive_queue) = 0x68
offsetof(struct inet_sock, inet_daddr) = 0x0
offsetof(struct inet_sock, inet_rcv_saddr) = 0x4
compute_score() (udp or tcp) now use a single cache line per ignored
item, instead of two.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Avoid some atomic ops on dst refcount, calling dev_queue_xmit_nit()
after skb_dst_drop() in dev_hard_start_xmit().
When queueing a packet into af_packet socket, we drop dst anyway.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
sk_run_filter() doesnt write on skb, change its prototype to reflect
this.
Fix two af_packet comments.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Le dimanche 05 décembre 2010 à 09:19 +0100, Eric Dumazet a écrit :
> Hmm..
>
> If somebody can explain why RTNL is held in arp_ioctl() (and therefore
> in arp_req_delete()), we might first remove RTNL use in arp_ioctl() so
> that your patch can be applied.
>
> Right now it is not good, because RTNL wont be necessarly held when you
> are going to call arp_invalidate() ?
While doing this analysis, I found a refcount bug in llc, I'll send a
patch for net-2.6
Meanwhile, here is the patch for net-next-2.6
Your patch then can be applied after mine.
Thanks
[PATCH] net: RCU conversion of dev_getbyhwaddr() and arp_ioctl()
dev_getbyhwaddr() was called under RTNL.
Rename it to dev_getbyhwaddr_rcu() and change all its caller to now use
RCU locking instead of RTNL.
Change arp_ioctl() to use RCU instead of RTNL locking.
Note: this fix a dev refcount bug in llc
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The dev field of ingress queue is forgot to initialized, then NULL
pointer dereference happens in qdisc_alloc().
Move inits of tx queues to netif_alloc_netdev_queues().
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Hi,
This patch fixes a typo in net/core/datagram.c and in net/sctp/socket.c
Regards,
David Shwartz
Signed-off-by: David Shwartz <dshwatrz@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We added some security checks in commit 57fe93b374
(filter: make sure filters dont read uninitialized memory) to close a
potential leak of kernel information to user.
This added a potential extra cost at run time, while we can perform a
check of the filter itself, to make sure a malicious user doesnt try to
abuse us.
This patch adds a check_loads() function, whole unique purpose is to
make this check, allocating a temporary array of mask. We scan the
filter and propagate a bitmask information, telling us if a load M(K) is
allowed because a previous store M(K) is guaranteed. (So that
sk_run_filter() can possibly not read unitialized memory)
Note: this can uncover application bug, denying a filter attach,
previously allowed.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Dan Rosenberg <drosenberg@vsecurity.com>
Cc: Changli Gao <xiaosuo@gmail.com>
Acked-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add SKF_AD_RXHASH and SKF_AD_CPU to filter ancillary mechanism,
to be able to build advanced filters.
This can help spreading packets on several sockets with a fast
selection, after RPS dispatch to N cpus for example, or to catch a
percentage of flows in one queue.
tcpdump -s 500 "cpu = 1" :
[0] ld CPU
[1] jeq #1 jt 2 jf 3
[2] ret #500
[3] ret #0
# take 12.5 % of flows (average)
tcpdump -s 1000 "rxhash & 7 = 2" :
[0] ld RXHASH
[1] and #7
[2] jeq #2 jt 3 jf 4
[3] ret #1000
[4] ret #0
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Rui <wirelesser@gmail.com>
Acked-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
NETIF_F_HW_CSUM is a superset of NETIF_F_IP_CSUM+NETIF_F_IPV6_CSUM, but
some drivers miss the difference. Fix this and also fix UFO dependency
on checksumming offload as it makes the same mistake in assumptions.
Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Acked-by: Jon Mason <jon.mason@exar.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Pavel Emelyanov tried to fix a race between sk_filter_(de|at)tach and
sk_clone() in commit 47e958eac2
Problem is we can have several clones sharing a common sk_filter, and
these clones might want to sk_filter_attach() their own filters at the
same time, and can overwrite old_filter->rcu, corrupting RCU queues.
We can not use filter->rcu without being sure no other thread could do
the same thing.
Switch code to a more conventional ref-counting technique : Do the
atomic decrement immediately and queue one rcu call back when last
reference is released.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
skb head being allocated by kmalloc(), it might be larger than what
actually requested because of discrete kmem caches sizes. Before
reallocating a new skb head, check if the current one has the needed
extra size.
Do this check only if skb head is not shared.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Cc: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This will also improve handling of ipv6 tcp socket request
backlog when syncookies are not enabled. When backlog
becomes very deep, last quarter of backlog is limited to
validated destinations. Previously only ipv4 implemented
this logic, but now ipv6 does too.
Now we are only one step away from enabling timewait
recycling for ipv6, and that step is simply filling in
the implementation of tcp_v6_get_peer() and
tcp_v6_tw_get_peer().
Signed-off-by: David S. Miller <davem@davemloft.net>
Allocate qdisc memory according to NUMA properties of cpus included in
xps map.
To be effective, qdisc should be (re)setup after changes
of /sys/class/net/eth<n>/queues/tx-<n>/xps_cpus
I added a numa_node field in struct netdev_queue, containing NUMA node
if all cpus included in xps_cpus share same node, else -1.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Ben Hutchings <bhutchings@solarflare.com>
Cc: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Avoid sparse warnings : add __rcu annotations and use
rcu_dereference_protected() where necessary.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
store_xps_map() allocates maps that are used by single cpu, it makes
sense to use NUMA allocations.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Cc: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch adds XPS_CONFIG option to enable and disable XPS. This is
done in the same manner as RPS_CONFIG. This is also fixes build
failure in XPS code when SMP is not enabled.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When testing struct netdev_queue state against FROZEN bit, we also test
XOFF bit. We can test both bits at once and save some cycles.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
As David pointed out correctly, updates to af-specific attributes
are currently not atomic. If multiple changes are requested and
one of them fails, previous updates may have been applied already
leaving the link behind in a undefined state.
This patch splits the function parse_link_af() into two functions
validate_link_af() and set_link_at(). validate_link_af() is placed
to validate_linkmsg() check for errors as early as possible before
any changes to the link have been made. set_link_af() is called to
commit the changes later.
This method is not fail proof, while it is currently sufficient
to make set_link_af() inerrable and thus 100% atomic, the
validation function method will not be able to detect all error
scenarios in the future, there will likely always be errors
depending on states which are f.e. not protected by rtnl_mutex
and thus may change between validation and setting.
Also, instead of silently ignoring unknown address families and
config blocks for address families which did not register a set
function the errors EAFNOSUPPORT respectively EOPNOSUPPORT are
returned to avoid comitting 4 out of 5 update requests without
notifying the user.
Signed-off-by: Thomas Graf <tgraf@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch implements transmit packet steering (XPS) for multiqueue
devices. XPS selects a transmit queue during packet transmission based
on configuration. This is done by mapping the CPU transmitting the
packet to a queue. This is the transmit side analogue to RPS-- where
RPS is selecting a CPU based on receive queue, XPS selects a queue
based on the CPU (previously there was an XPS patch from Eric
Dumazet, but that might more appropriately be called transmit completion
steering).
Each transmit queue can be associated with a number of CPUs which will
use the queue to send packets. This is configured as a CPU mask on a
per queue basis in:
/sys/class/net/eth<n>/queues/tx-<n>/xps_cpus
The mappings are stored per device in an inverted data structure that
maps CPUs to queues. In the netdevice structure this is an array of
num_possible_cpu structures where each structure holds and array of
queue_indexes for queues which that CPU can use.
The benefits of XPS are improved locality in the per queue data
structures. Also, transmit completions are more likely to be done
nearer to the sending thread, so this should promote locality back
to the socket on free (e.g. UDP). The benefits of XPS are dependent on
cache hierarchy, application load, and other factors. XPS would
nominally be configured so that a queue would only be shared by CPUs
which are sharing a cache, the degenerative configuration woud be that
each CPU has it's own queue.
Below are some benchmark results which show the potential benfit of
this patch. The netperf test has 500 instances of netperf TCP_RR test
with 1 byte req. and resp.
bnx2x on 16 core AMD
XPS (16 queues, 1 TX queue per CPU) 1234K at 100% CPU
No XPS (16 queues) 996K at 100% CPU
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In dev_pick_tx, don't do work in calculating queue
index or setting
the index in the sock unless the device has more than one queue. This
allows the sock to be set only with a queue index of a multi-queue
device which is desirable if device are stacked like in a tunnel.
We also allow the mapping of a socket to queue to be changed. To
maintain in order packet transmission a flag (ooo_okay) has been
added to the sk_buff structure. If a transport layer sets this flag
on a packet, the transmit queue can be changed for the socket.
Presumably, the transport would set this if there was no possbility
of creating OOO packets (for instance, there are no packets in flight
for the socket). This patch includes the modification in TCP output
for setting this flag.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Lower SCM_MAX_FD from 255 to 253 so that allocations for scm_fp_list are
halved. (commit f8d570a4 added two pointers in this structure)
scm_fp_dup() should not copy whole structure (and trigger kmemcheck
warnings), but only the used part. While we are at it, only allocate
needed size.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Unloading pktgen module needs ~6 seconds on a 64 cpus machine, to stop
64 kthreads.
Add a pktgen_exiting variable to let kernel threads die faster, so that
kthread_stop() doesnt have to wait too long for them. This variable is
not tested in fast path.
Note : Before exiting from pktgen_thread_worker(), we must make sure
kthread_stop() is waiting for this thread to be stopped, like its done
in kernel/softirq.c
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We forgot to use __GFP_HIGHMEM in several __vmalloc() calls.
In ceph, add the missing flag.
In fib_trie.c, xfrm_hash.c and request_sock.c, using vzalloc() is
cleaner and allows using HIGHMEM pages as well.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
At compile time, we can replace the DIV_K instruction (divide by a
constant value) by a reciprocal divide.
At exec time, the expensive divide is replaced by a multiply, a less
expensive operation on most processors.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Starting the translated instruction to 1 instead of 0 allows us to
remove one descrement at check time and makes codes[] array init
cleaner.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove pc variable to avoid arithmetic to compute fentry at each filter
instruction. Jumps directly manipulate fentry pointer.
As the last instruction of filter[] is guaranteed to be a RETURN, and
all jumps are before the last instruction, we dont need to check filter
bounds (number of instructions in filter array) at each iteration, so we
remove it from sk_run_filter() params.
On x86_32 remove f_k var introduced in commit 57fe93b374
(filter: make sure filters dont read uninitialized memory)
Note : We could use a CONFIG_ARCH_HAS_{FEW|MANY}_REGISTERS in order to
avoid too many ifdefs in this code.
This helps compiler to use cpu registers to hold fentry and A
accumulator.
On x86_32, this saves 401 bytes, and more important, sk_run_filter()
runs much faster because less register pressure (One less conditional
branch per BPF instruction)
# size net/core/filter.o net/core/filter_pre.o
text data bss dec hex filename
2948 0 0 2948 b84 net/core/filter.o
3349 0 0 3349 d15 net/core/filter_pre.o
on x86_64 :
# size net/core/filter.o net/core/filter_pre.o
text data bss dec hex filename
5173 0 0 5173 1435 net/core/filter.o
5224 0 0 5224 1468 net/core/filter_pre.o
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Changli Gao <xiaosuo@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix kernel-doc warning for sk_filter_rcu_release():
Warning(net/core/filter.c:586): missing initial short description on line:
* sk_filter_rcu_release: Release a socket filter by rcu_head
Signed-off-by: Randy Dunlap <randy.dunlap@oracle.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: netdev@vger.kernel.org
Signed-off-by: David S. Miller <davem@davemloft.net>
BPF_S_* are used internally, should not be exposed to the others.
Signed-off-by: Changli Gao <xiaosuo@gmail.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Hagen Paul Pfeifer <hagen@jauu.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Since repeating u16 value to u8 value conversion using switch() clause's
case statement is wasteful, this patch introduces u16 to u8 mapping table
and removes most of case statements. As a result, the size of net/core/filter.o
is reduced by about 29% on x86.
Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add option to set skb priority to pktgen. Useful for testing
QOS features. Also by running pktgen on the vlan device the
qdisc on the real device can be tested.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
netif_set_real_num_rx_queues() can decrement and increment
the number of rx queues. For example ixgbe does this as
features and offloads are toggled. Presumably this could
also happen across down/up on most devices if the available
resources changed (cpu offlined).
The kobject needs to be zero'd in this case so that the
state is not preserved across kobject_put()/kobject_init_and_add().
This resolves the following error report.
ixgbe 0000:03:00.0: eth2: NIC Link is Up 10 Gbps, Flow Control: RX/TX
kobject (ffff880324b83210): tried to init an initialized object, something is seriously wrong.
Pid: 1972, comm: lldpad Not tainted 2.6.37-rc18021qaz+ #169
Call Trace:
[<ffffffff8121c940>] kobject_init+0x3a/0x83
[<ffffffff8121cf77>] kobject_init_and_add+0x23/0x57
[<ffffffff8107b800>] ? mark_lock+0x21/0x267
[<ffffffff813c6d11>] net_rx_queue_update_kobjects+0x63/0xc6
[<ffffffff813b5e0e>] netif_set_real_num_rx_queues+0x5f/0x78
[<ffffffffa0261d49>] ixgbe_set_num_queues+0x1c6/0x1ca [ixgbe]
[<ffffffffa0262509>] ixgbe_init_interrupt_scheme+0x1e/0x79c [ixgbe]
[<ffffffffa0274596>] ixgbe_dcbnl_set_state+0x167/0x189 [ixgbe]
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
netif_set_real_num_rx_queues() can decrement and increment
the number of rx queues. For example ixgbe does this as
features and offloads are toggled. Presumably this could
also happen across down/up on most devices if the available
resources changed (cpu offlined).
The kobject needs to be zero'd in this case so that the
state is not preserved across kobject_put()/kobject_init_and_add().
This resolves the following error report.
ixgbe 0000:03:00.0: eth2: NIC Link is Up 10 Gbps, Flow Control: RX/TX
kobject (ffff880324b83210): tried to init an initialized object, something is seriously wrong.
Pid: 1972, comm: lldpad Not tainted 2.6.37-rc18021qaz+ #169
Call Trace:
[<ffffffff8121c940>] kobject_init+0x3a/0x83
[<ffffffff8121cf77>] kobject_init_and_add+0x23/0x57
[<ffffffff8107b800>] ? mark_lock+0x21/0x267
[<ffffffff813c6d11>] net_rx_queue_update_kobjects+0x63/0xc6
[<ffffffff813b5e0e>] netif_set_real_num_rx_queues+0x5f/0x78
[<ffffffffa0261d49>] ixgbe_set_num_queues+0x1c6/0x1ca [ixgbe]
[<ffffffffa0262509>] ixgbe_init_interrupt_scheme+0x1e/0x79c [ixgbe]
[<ffffffffa0274596>] ixgbe_dcbnl_set_state+0x167/0x189 [ixgbe]
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Each net_device contains address family specific data such as
per device settings and statistics. We already expose this data
via procfs/sysfs and partially netlink.
The netlink method requires the requester to send one RTM_GETLINK
request for each address family it wishes to receive data of
and then merge this data itself.
This patch implements a new API which combines all address family
specific link data in a new netlink attribute IFLA_AF_SPEC.
IFLA_AF_SPEC contains a sequence of nested attributes, one for each
address family which in turn defines the structure of its own
attribute. Example:
[IFLA_AF_SPEC] = {
[AF_INET] = {
[IFLA_INET_CONF] = ...,
},
[AF_INET6] = {
[IFLA_INET6_FLAGS] = ...,
[IFLA_INET6_CONF] = ...,
}
}
The API also allows for address families to implement a function
which parses the IFLA_AF_SPEC attribute sent by userspace to
implement address family specific link options.
Signed-off-by: Thomas Graf <tgraf@infradead.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
ERROR: "netif_get_vlan_features" [drivers/net/xen-netfront.ko] undefined!
Reported-by: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch move RX queue allocation to alloc_netdev_mq and freeing of
the queues to free_netdev (symmetric to TX queue allocation). Each
kobject RX queue takes a reference to the queue's device so that the
device can't be freed before all the kobjects have been released-- this
obviates the need for reference counts specific to RX queues.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
TX queues are now allocated in alloc_netdev_mq and freed in
free_netdev.
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We currently use vlan_features to check for TSO support if there is
a vlan tag. However, it's quite likely that the NIC is not able to
do TSO when there is an arbitrary number of tags. Therefore if there
is more than one tag (in-band or out-of-band), fall back to software
emulation.
Signed-off-by: Jesse Gross <jesse@nicira.com>
CC: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We assume that hardware TSO can't support multiple levels of vlan tags
but we allow it to be done. Therefore, enable GSO to parse these tags
so we can fallback to software.
Signed-off-by: Jesse Gross <jesse@nicira.com>
CC: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When checking if it is necessary to linearize a packet, we currently
use vlan_features if the packet contains either an in-band or out-
of-band vlan tag. However, in-band tags aren't special in any way
for scatter/gather since they are part of the packet buffer and are
simply more data to DMA. Therefore, only use vlan_features for out-
of-band tags, which could potentially have some interaction with
scatter/gather.
Signed-off-by: Jesse Gross <jesse@nicira.com>
CC: Ben Hutchings <bhutchings@solarflare.com>
Reviewed-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
nlmsg_total_size() calculates the length of a netlink message
including header and alignment. nla_total_size() calculates the
space an individual attribute consumes which was meant to be used
in this context.
Also, ensure to account for the attribute header for the
IFLA_INFO_XSTATS attribute as implementations of get_xstats_size()
seem to assume that we do so.
The addition of two message headers minus the missing attribute
header resulted in a calculated message size that was larger than
required. Therefore we never risked running out of skb tailroom.
Signed-off-by: Thomas Graf <tgraf@infradead.org>
Acked-by: Patrick McHardy <kaber@trash.net>
Signed-off-by: David S. Miller <davem@davemloft.net>
Robin Holt tried to boot a 16TB machine and found some limits were
reached : sysctl_tcp_mem[2], sysctl_udp_mem[2]
We can switch infrastructure to use long "instead" of "int", now
atomic_long_t primitives are available for free.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Reported-by: Robin Holt <holt@sgi.com>
Reviewed-by: Robin Holt <holt@sgi.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is a possibility malicious users can get limited information about
uninitialized stack mem array. Even if sk_run_filter() result is bound
to packet length (0 .. 65535), we could imagine this can be used by
hostile user.
Initializing mem[] array, like Dan Rosenberg suggested in his patch is
expensive since most filters dont even use this array.
Its hard to make the filter validation in sk_chk_filter(), because of
the jumps. This might be done later.
In this patch, I use a bitmap (a single long var) so that only filters
using mem[] loads/stores pay the price of added security checks.
For other filters, additional cost is a single instruction.
[ Since we access fentry->k a lot now, cache it in a local variable
and mark filter entry pointer as const. -DaveM ]
Reported-by: Dan Rosenberg <drosenberg@vsecurity.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Followup of commit ef885afbf8 (net: use rcu_barrier() in
rollback_registered_many)
dst_dev_event() scans a garbage dst list that might be feeded by various
network notifiers at device dismantle time.
Its important to call dst_dev_event() after other notifiers, or we might
enter the infamous msleep(250) in netdev_wait_allrefs(), and wait one
second before calling again call_netdevice_notifiers(NETDEV_UNREGISTER,
dev) to properly remove last device references.
Use priority -10 to let dst_dev_notifier be called after other network
notifiers (they have the default 0 priority)
Reported-by: Ben Greear <greearb@candelatech.com>
Reported-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Reported-by: Octavian Purdila <opurdila@ixiacom.com>
Reported-by: Benjamin LaHaise <bcrl@kvack.org>
Tested-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This fix a bug reported by backyes.
Right the first time pktgen's using queue_map that's not been initialized
by set_cur_queue_map(pkt_dev);
Signed-off-by: Junchang Wang <junchangwang@gmail.com>
Signed-off-by: Backyes <backyes@mail.ustc.edu.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
This should fix the following warning:
net/core/pktgen.c: In function ‘pktgen_if_write’:
net/core/pktgen.c:890: warning: comparison of distinct pointer types lacks a cast
Signed-off-by: Dmitry Torokhov <dtor@mail.ru>
Reviewed-by: Nelson Elhage <nelhage@ksplice.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In dev_pick_tx recompute the queue index if the value stored in the
socket is greater than or equal to the number of real queues for the
device. The saved index in the sock structure is not guaranteed to
be appropriate for the egress device (this could happen on a route
change or in presence of tunnelling). The result of the queue index
being bad would be to return a bogus queue (crash could prersumably
follow).
Signed-off-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A program that accidentally writes too much data to the pktgen file can overflow
the kernel stack and oops the machine. This is only triggerable by root, so
there's no security issue, but it's still an unfortunate bug.
printk() won't print more than 1024 bytes in a single call, anyways, so let's
just never copy more than that much data. We're on a fairly shallow stack, so
that should be safe even with CONFIG_4KSTACKS.
Signed-off-by: Nelson Elhage <nelhage@ksplice.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This helps protect us from overflow issues down in the
individual protocol sendmsg/recvmsg handlers. Once
we hit INT_MAX we truncate out the rest of the iovec
by setting the iov_len members to zero.
This works because:
1) For SOCK_STREAM and SOCK_SEQPACKET sockets, partial
writes are allowed and the application will just continue
with another write to send the rest of the data.
2) For datagram oriented sockets, where there must be a
one-to-one correspondance between write() calls and
packets on the wire, INT_MAX is going to be far larger
than the packet size limit the protocol is going to
check for and signal with -EMSGSIZE.
Based upon a patch by Linus Torvalds.
Signed-off-by: David S. Miller <davem@davemloft.net>
Adds __rcu annotation to (struct fib_rule)->ctarget
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
NETIF_F_HW_CSUM indicates the ability to update an TCP/IP-style 16-bit
checksum with the checksum of an arbitrary part of the packet data,
whereas the FCoE CRC is something entirely different.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Cc: stable@kernel.org [2.6.32+]
Signed-off-by: David S. Miller <davem@davemloft.net>
dev_can_checksum() incorrectly returns true in these cases:
1. The skb has both out-of-band and in-band VLAN tags and the device
supports checksum offload for the encapsulated protocol but only with
one layer of encapsulation.
2. The skb has a VLAN tag and the device supports generic checksumming
but not in conjunction with VLAN encapsulation.
Rearrange the VLAN tag checks to avoid these.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Some panic reports in fib_rules_lookup() show a rule could have a NULL
pointer as a next pointer in the rules_list.
This can actually happen because of a bug in fib_nl_newrule() : It
checks if current rule is the destination of unresolved gotos. (Other
rules have gotos to this about to be inserted rule)
Problem is it does the resolution of the gotos before the rule is
inserted in the rules_list (and has a valid next pointer)
Fix this by moving the rules_list insertion before the changes on gotos.
A lockless reader can not any more follow a ctarget pointer, unless
destination is ready (has a valid next pointer)
Reported-by: Oleg A. Arkhangelsky <sysoleg@yandex.ru>
Reported-by: Joe Buehler <aspam@cox.net>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add __rcu annotation to :
(struct sock)->sk_filter
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>
add __rcu annotation to (struct net)->gen, and use
rcu_dereference_protected() in net_assign_generic()
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add __rcu annotations to :
(struct netdev_rx_queue)->rps_map
(struct netdev_rx_queue)->rps_flow_table
struct rps_sock_flow_table *rps_sock_flow_table;
And use appropriate rcu primitives.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
(struct net_device)->ip6_ptr is rcu protected :
add __rcu annotation and proper rcu primitives.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Three is definitely too low, and we know from reports that GRE tunnels
stacked as deeply as 37 levels cause stack overflows, so pick some
reasonable value between those two.
Signed-off-by: David S. Miller <davem@davemloft.net>
The temporary variable "i" is needlessly initialized to zero
in two distinct cases in this file:
1) where it is set to zero and then used as an argument in an addition
before being assigned a non-zero value.
2) where it is only used in a standard/typical loop counter
For (1), simply delete assignment to zero and usages while still
zero; for (2) simply make the loop start at zero as per standard
practice as seen everywhere else in the same file.
Signed-off-by: Paul Gortmaker <paul.gortmaker@windriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-next-2.6: (1699 commits)
bnx2/bnx2x: Unsupported Ethtool operations should return -EINVAL.
vlan: Calling vlan_hwaccel_do_receive() is always valid.
tproxy: use the interface primary IP address as a default value for --on-ip
tproxy: added IPv6 support to the socket match
cxgb3: function namespace cleanup
tproxy: added IPv6 support to the TPROXY target
tproxy: added IPv6 socket lookup function to nf_tproxy_core
be2net: Changes to use only priority codes allowed by f/w
tproxy: allow non-local binds of IPv6 sockets if IP_TRANSPARENT is enabled
tproxy: added tproxy sockopt interface in the IPV6 layer
tproxy: added udp6_lib_lookup function
tproxy: added const specifiers to udp lookup functions
tproxy: split off ipv6 defragmentation to a separate module
l2tp: small cleanup
nf_nat: restrict ICMP translation for embedded header
can: mcp251x: fix generation of error frames
can: mcp251x: fix endless loop in interrupt handler if CANINTF_MERRF is set
can-raw: add msg_flags to distinguish local traffic
9p: client code cleanup
rds: make local functions/variables static
...
Fix up conflicts in net/core/dev.c, drivers/net/pcmcia/smc91c92_cs.c and
drivers/net/wireless/ath/ath9k/debug.c as per David
The function napi_reuse_skb is only used inside core.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When there are VLANs on a VETH device, the packets being transmitted
through the VETH device may be 4 bytes bigger than MTU. A check
in dev_forward_skb did not take this into account and so dropped
these packets.
This patch is needed at least as far back as 2.6.34.7 and should
be considered for -stable.
Signed-off-by: Ben Greear <greearb@candelatech.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The function rtnl_kill_links is defined but never used.
Signed-off-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Now that vlan acceleration is handled consistently regardless of usage,
it is possible to enable and disable it at will. This adds support for
Ethtool operations that change the offloading status for debugging
purposes, similar to other forms of hardware acceleration.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently each driver that is capable of vlan hardware acceleration
must be aware of the vlan groups that are configured and then pass
the stripped tag to a specialized receive function. This is
different from other types of hardware offload in that it places a
significant amount of knowledge in the driver itself rather keeping
it in the networking core.
This makes vlan offloading function more similarly to other forms
of offloading (such as checksum offloading or TSO) by doing the
following:
* On receive, stripped vlans are passed directly to the network
core, without attempting to check for vlan groups or reconstructing
the header if no group
* vlans are made less special by folding the logic into the main
receive routines
* On transmit, the device layer will add the vlan header in software
if the hardware doesn't support it, instead of spreading that logic
out in upper layers, such as bonding.
There are a number of advantages to this:
* Fixes all bugs with drivers incorrectly dropping vlan headers at once.
* Avoids having to disable VLAN acceleration when in promiscuous mode
(good for bridging since it always puts devices in promiscuous mode).
* Keeps VLAN tag separate until given to ultimate consumer, which
avoids needing to do header reconstruction as in tg3 unless absolutely
necessary.
* Consolidates common code in core networking.
Signed-off-by: Jesse Gross <jesse@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently users of hardware vlan accleration need to know whether
the device supports it before generating packets. However, vlan
acceleration will soon be available in a more flexible manner so
knowing ahead of time becomes much more difficult. This adds
a software fallback path for vlan packets on devices without the
necessary offloading support, similar to other types of hardware
accleration.
Signed-off-by: Jesse Gross <jesse@nicira.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>
This patch introduces netif_alloc_netdev_queues which is called from
register_device instead of alloc_netdev_mq. This makes TX queue
allocation symmetric with RX allocation. Also, queue locks allocation
is done in netdev_init_one_queue. Change set_real_num_tx_queues to
fail if requested number < 1 or greater than number of allocated
queues.
Signed-off-by: Tom Herbert <therbert@google.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Clean up in RX queue allocation. In netif_set_real_num_rx_queues
return error on attempt to set zero queues, or requested number is
greater than number of allocated queues. In netif_alloc_rx_queues,
do BUG_ON if queue_count is zero.
Signed-off-by: Tom Herbert <therbert@google.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In alloc_netdev_mq fail if requested queue_count < 1.
Signed-off-by: Tom Herbert <therbert@google.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In an erlier patch I modified napi_poll so that devices with IFF_MASTER polled
the per_cpu list instead of the device list for napi. I did this because the
bonding driver has no napi instances to poll, it instead expects to check the
slave devices napi instances, which napi_poll was unaware of. Looking at this
more closely however, I now see this isn't strictly needed. As the bond driver
poll_controller calls the slaves poll_controller via netpoll_poll_dev, which
recursively calls poll_napi on each slave, allowing those napi instances to get
serviced. The earlier patch isn't at all harmfull, its just not needed, so lets
revert it to make the code cleaner. Sorry for the noise,
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Reviewed-by: WANG Cong <amwang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Usually the netpoll path, when preforming a napi poll can get away with just
polling all the napi instances of the configured device. Thats not the case for
the bonding driver however, as the napi instances which may wind up getting
flagged as needing polling after the poll_controller call don't belong to the
bonded device, but rather to the slave devices. Fix this by checking the device
in question for the IFF_MASTER flag, if set, we know we need to check the full
poll list for this cpu, rather than just the devices napi instance list.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The bonding driver currently modifies the netpoll structure in its xmit path
while sending frames from netpoll. This is racy, as other cpus can access the
netpoll structure in parallel. Since the bonding driver points np->dev to a
slave device, other cpus can inadvertently attempt to send data directly to
slave devices, leading to improper locking with the bonding master, lost frames,
and deadlocks. This patch fixes that up.
This patch also removes the real_dev pointer from the netpoll structure as that
data is really only used by bonding in the poll_controller, and we can emulate
its behavior by check each slave for IS_UP.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
fib_nl_delrule() calls synchronize_rcu() for no apparent reason,
while rtnl is held.
I suspect it was done to avoid an atomic_inc_not_zero() in
fib_rules_lookup(), which commit 7fa7cb7109 added anyway.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
commit b30973f877 (node-aware skb allocation) spread a wrong habit of
allocating net drivers skbs on a given memory node : The one closest to
the NIC hardware. This is wrong because as soon as we try to scale
network stack, we need to use many cpus to handle traffic and hit
slub/slab management on cross-node allocations/frees when these cpus
have to alloc/free skbs bound to a central node.
skb allocated in RX path are ephemeral, they have a very short
lifetime : Extra cost to maintain NUMA affinity is too expensive. What
appeared as a nice idea four years ago is in fact a bad one.
In 2010, NIC hardwares are multiqueue, or we use RPS to spread the load,
and two 10Gb NIC might deliver more than 28 million packets per second,
needing all the available cpus.
Cost of cross-node handling in network and vm stacks outperforms the
small benefit hardware had when doing its DMA transfert in its 'local'
memory node at RX time. Even trying to differentiate the two allocations
done for one skb (the sk_buff on local node, the data part on NIC
hardware node) is not enough to bring good performance.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Acked-by: Tom Herbert <therbert@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We tried very hard to remove all possible dev_hold()/dev_put() pairs in
network stack, using RCU conversions.
There is still an unavoidable device refcount change for every dst we
create/destroy, and this can slow down some workloads (routers or some
app servers, mmap af_packet)
We can switch to a percpu refcount implementation, now dynamic per_cpu
infrastructure is mature. On a 64 cpus machine, this consumes 256 bytes
per device.
On x86, dev_hold(dev) code :
before
lock incl 0x280(%ebx)
after:
movl 0x260(%ebx),%eax
incl fs:(%eax)
Stress bench :
(Sending 160.000.000 UDP frames,
IP route cache disabled, dual E5540 @2.53GHz,
32bit kernel, FIB_TRIE)
Before:
real 1m1.662s
user 0m14.373s
sys 12m55.960s
After:
real 0m51.179s
user 0m15.329s
sys 10m15.942s
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
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>
Add a seqlock in struct neighbour to protect neigh->ha[], and avoid
dirtying neighbour in stress situation (many different flows / dsts)
Dirtying takes place because of read_lock(&n->lock) and n->used writes.
Switching to a seqlock, and writing n->used only on jiffies changes
permits less dirtying.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Several other ethtool functions leave heap uncleared (potentially) by
drivers. Some interfaces appear safe (eeprom, etc), in that the sizes
are well controlled. In some situations (e.g. unchecked error conditions),
the heap will remain unchanged in areas before copying back to userspace.
Note that these are less of an issue since these all require CAP_NET_ADMIN.
Cc: stable@kernel.org
Signed-off-by: Kees Cook <kees.cook@canonical.com>
Acked-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When a new dst is used to send a frame, neigh_resolve_output() tries to
associate an struct hh_cache to this dst, calling neigh_hh_init() with
the neigh rwlock write locked.
Most of the time, hh_cache is already known and linked into neighbour,
so we find it and increment its refcount.
This patch changes the logic so that we call neigh_hh_init() with
neighbour lock read locked only, so that fast path can be run in
parallel by concurrent cpus.
This brings part of the speedup we got with commit c7d4426a98
(introduce DST_NOCACHE flag) for non cached dsts, even for cached ones,
removing one of the contention point that routers hit on multiqueue
enabled machines.
Further improvements would need to use a seqlock instead of an rwlock to
protect neigh->ha[], to not dirty neigh too often and remove two atomic
ops.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The rx->count reference is used to track reference counts to the
number of rx-queue kobjects created for the device. This patch
eliminates initialization of the counter in netif_alloc_rx_queues
and instead increments the counter each time a kobject is created.
This is now symmetric with the decrement that is done when an object is
released.
Signed-off-by: Tom Herbert <therbert@google.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Calling ETHTOOL_GRXCLSRLALL with a large rule_cnt will allocate kernel
heap without clearing it. For the one driver (niu) that implements it,
it will leave the unused portion of heap unchanged and copy the full
contents back to userspace.
Signed-off-by: Kees Cook <kees.cook@canonical.com>
Acked-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Synchronise the comment with the preceding implementation change.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Do not set num_rx_queues in netif_set_real_num_rx_queues() some
drivers will increase the real_num_rx_queues later due to a feature
changes or available interrupts increasing. By setting num_rx_queues
here this ends up creating a cap on the number of rx queues
available.
For example the ixgbe driver sets the max number of queues it intends
to use ever then sets the current number in use with the
netif_set_num_{rx|tx}_queues calls. With the current implementation
the number of rx queues gets limited so when a feature such as DCB
or FCoE is enabled the queues are no longer available.
kobjects will only be allocated for real_num_rx_queues so the waste
in memory is minimal.
Signed-off-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Acked-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This is the second step for neighbour RCU conversion.
(first was commit d6bf7817 : RCU conversion of neigh hash table)
neigh_lookup() becomes lockless, but still take a reference on found
neighbour. (no more read_lock()/read_unlock() on tbl->lock)
struct neighbour gets an additional rcu_head field and is freed after an
RCU grace period.
Future work would need to eventually not take a reference on neighbour
for temporary dst (DST_NOCACHE), but this would need dst->_neighbour to
use a noref bit like we did for skb->_dst.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
