Replace the mac_filter_list with a static size hash table of 8bits. The
primary advantage of this is a decrease in latency of operations related
to searching for specific MAC filters, including .set_rx_mode. Using
a linked list resulted in several locations which were O(n^2). Using
a hash table should give us latency growth closer to O(n*log(n)).

Change-ID: I5330bd04053b880e670210933e35830b95948ebb
Signed-off-by: NJacob Keller <jacob.e.keller@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>

When inside a loop where we call i40e_del_filter we use an O(n^2)
pattern where i40e_del_filter calls i40e_find_filter for us. We can
avoid this O(n^2) logic by factoring a function, __i40e_del_filter() out
from the i40e_del_filter code. This allows us to re-use the delete logic
where appropriate without having to search for the filter twice.

This new function benefits several functions including i40e_vsi_add_vlan,
i40e_vsi_kill_vlan, i40e_del_mac_vlan_all, and i40e_vsi_release.

Change-ID: I75fabe0f53bf73f56b80d342e5fdcfcc28f4d3eb
Signed-off-by: NJacob Keller <jacob.e.keller@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

When adding new MAC address filters, the driver determines if it should
behave in VLAN mode (where all MAC addresses get assigned to every
existing VLAN) or in non-VLAN mode where MAC addresses get assigned the
VLAN_ANY identifier. Under some circumstances it is possible that a VLAN
has been marked for removal (such that all filters of that VLAN are set
to I40E_FILTER_REMOVE), and a subsequent call to i40e_put_mac_in_vlan
may occur prior to the driver subtask that syncs filters to the
hardware.

In this case, we may add filters to the new removed VLAN, even though it
should have been removed. This is most obvious when first adding a new
VLAN. We will delete all filters which are in I40E_VLAN_ANY (-1) and
then re-add them as in VLAN 0 (untagged). Then before we sync filters,
we will add new MAC address filter, which will be added to every VLAN
that exists. Unfortunately, this will include I40E_VLAN_ANY, so we will
end up incorrectly adding filters to the -1 VLAN. This can be fixed by
simply skipping all filters which are marked for removal.

A similar check is not necessary in i40e_del_mac_all_vlan, since we are
deleting, and any filter which we find already marked for removal would
simply be deleted again, which doesn't cause any issues.

Change-Id: I7962154013ce02fe950584690aeeb3ed853d0086
Signed-off-by: NJacob Keller <jacob.e.keller@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

When a PVID has been assigned to a VSI, the function
i40e_put_mac_in_vlan arbitrarily modifies all filters
to have the same VLAN. This is obviously incorrect
because it could be modifying active filters without
putting them into the NEW state. The correct method
is to remove then re-add filters which is already done
in the code where we assign the PVID.

Fix this issue and a few other minor nits at the same
time. First, when we have a PVID don't even bother
looping and simply add the filter with the PVID immediately.

In the case of the loop, we now can remove several checks.
We also don't need to use i40e_find_filter first before
calling i40e_add_filter, since i40e_add_filter implicitly
does a lookup already.

Finally, update the return semantics of this function so
that on failure to add a filter it returns NULL, but on
success, it returns the last filter added. Otherwise,
we're just returning the last filter in the list. An
alternative fix might be to return 0 or an error code,
but this is pretty invasive to every call site.

Change-ID: I2325dfd843aec76d89fb0d7cb0e7c4f290a34840
Signed-off-by: NJacob Keller <jacob.e.keller@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

A future patch will be modifying these functions and making a call to
a static function which currently is defined after these functions. Move
them in a separate patch to ease review and ensure the moved code is
correct.

Change-ID: I2ca7fd4e10c0c07ed2291db1ea41bf5987fc6474
Signed-off-by: NJacob Keller <jacob.e.keller@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The kernel provides __dev_uc_sync and __dev_mc_sync in order for drivers
which need individual notification of add and delete for each filter.
These functions allow us to vastly simplify our .set_rx_mode handler. We
need to implement two functions for sync and unsync which add and remove
filters respectively.

This change avoids a very complex and inefficient algorithm which
resulted in an abnormal latency for the .set_rx_mode NDO operation. The
resulting code after this change is more readable, more efficient, and
less code.

Due to the callback signature used by these functions we also must
update several other functions to take a const u8 * pointer.

Change-Id: I2ca7fd4e10c0c07ed2291db1ea41bf5987fc6474
Signed-off-by: NJacob Keller <jacob.e.keller@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Originally the is_vf and is_netdev fields were added in order to
distinguish between VF and netdev filters in a single VSI. However, it
can be noted that we use separate VSI for SRIOV VFs and for netdev VSI.
Thus, since a single VSI should only ever have one type of filter, we
can simply remove the checks and remove the typing.

In a similar fashion, we can note that the only remaining way to get
multiple filters of a single type is through a debug command that was
added to debugfs. This command is useless in practice, and results in
causing bugs if we keep counter tracking but lose the is_vf and
is_netdev protections as desired above.

Since the only time we'd actually have a counter value besides 0 and
1 is through use of this debugfs hook, we can remove this unnecessary
command, and the entire counter logic it required.

We vastly simplify mac filters by removing

(a) the distinction between VF and netdev filters
(b) counting logic
(c) the ability to add and remove filters bypassing the stack via debugfs

Change-ID: Idf916dd2a1159b1188ddbab5bef6b85ea6bf27d9
Signed-off-by: NJacob Keller <jacob.e.keller@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Trival fix, dev_err message is missing a \n, so add it.
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Nikolay Aleksandrov says:

====================
bridge: add support for PIM hello router ports

The first 3 patches of this set do minor cleanups and add some helpers to
the PIM header file. Patch 4 adds a way to detect mcast router ports via
PIM hello messages, they're marked as temporary and are not considered for
querier. There's more detailed information in patch 4's commit message.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When we receive a PIM Hello message on a port we can consider that it
has a multicast router attached, thus it is correct to add it to the
router list. The only catch is it shouldn't be considered for a querier.

Using Daniel's description:
leaf-11  leaf-12  leaf-13
       \   |    /
        bridge-1
         /    \
    host-11  host-12

 - all ports in bridge-1 are in a single vlan aware bridge
 - leaf-11 is the IGMP querier
 - leaf-13 is the PIM DR
 - host-11 TXes packets to 226.10.10.10
 - bridge-1 only forwards the 226.10.10.10 traffic out the port to
   leaf-11, it should also forward this traffic out the port to leaf-13
Suggested-by: NDaniel Walton <dwalton@cumulusnetworks.com>
Signed-off-by: NNikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NNikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NNikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add the common pimhdr structure and helpers to access it, also cleanup the
format of the header file.
Signed-off-by: NNikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Yuval Mintz says:

====================
qed*: Patch series

This series does several things. The bigger changes:

 - Add new notification APIs [& Defaults] for various fields.
The series then utilizes some of those qed <-> qede APIs to bass WoL
support upon.

 - Change the resource allocation scheme to receive the values from
management firmware, instead of equally sharing resources between
functions [that might not need those]. That would, e.g., allow us to
configure additional filters to network interfaces in presence of
storage [PCI] functions from same adapter.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, each interfaces assumes it receives an equal portion
of HW/FW resources, but this is wasteful - different partitions
[and specifically, parititions exposing different protocol support]
might require different resources.

Implement a new resource learning scheme where the information is
received directly from the management firmware [which has knowledge
of all of the functions and can serve as arbiter].
Signed-off-by: NTomer Tayar <Tomer.Tayar@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Driver sets several restrictions about the number of supported VFs
according to available HW/FW resources.
This creates a problem as there are constellations which can't be
supported [as limitation don't accurately describe the resources],
as well as holes where enabling IOV would fail due to supposed
lack of resources.

This introduces a new interal feature - vf-queues, which would
be used to lift some of the restriction and accurately enumerate
the queues that can be used by a given PF's VFs.
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Today, RDMA capabilities are learned from management firmware
which provides a per-device indication for all interfaces.
Newer management firmware is capable of providing a per-device
indication [would later be extended to either RoCE/iWARP].

Try using this newer learning mechanism, but fallback in case
management firmware is too old to retain current functionality.
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While the qed_lm_maps is closely tied with the QED_LM_* defines,
when iterating over the array use actual size instead of the qed
define to prevent future possible issues.
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Management firmware is interested in various tidbits about
the driver - including the driver state & several configuration
related fields [MTU, primtary MAC, etc.].
This adds the necessray logic to update MFW with such configurations,
some of which are passed directly via qed while for others APIs
are provide so that qede would be able to later configure if needed.

This also introduces a new default configuration for MTU which would
replace the default inherited by being an ethernet device.
Signed-off-by: NSudarsana Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use DEVICE_ATTR_RW for read-write attributes.  This simplifies the
source code, improves readbility, and reduces the chance of
inconsistencies.

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@rw@
declarer name DEVICE_ATTR;
identifier x,x_show,x_store;
@@

DEVICE_ATTR(x, \(0644\|S_IRUGO|S_IWUSR\), x_show, x_store);

@script:ocaml@
x << rw.x;
x_show << rw.x_show;
x_store << rw.x_store;
@@

if not (x^"_show" = x_show && x^"_store" = x_store)
then Coccilib.include_match false

@@
declarer name DEVICE_ATTR_RW;
identifier rw.x,rw.x_show,rw.x_store;
@@

- DEVICE_ATTR(x, \(0644\|S_IRUGO|S_IWUSR\), x_show, x_store);
+ DEVICE_ATTR_RW(x);
// </smpl>
Signed-off-by: NJulia Lawall <Julia.Lawall@lip6.fr>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use DEVICE_ATTR_RO for read only attributes.  This simplifies the
source code, improves readbility, and reduces the chance of
inconsistencies.

The semantic patch that makes this change is as follows:
(http://coccinelle.lip6.fr/)

// <smpl>
@ro@
declarer name DEVICE_ATTR;
identifier x,x_show;
@@

DEVICE_ATTR(x, \(0444\|S_IRUGO\), x_show, NULL);

@script:ocaml@
x << ro.x;
x_show << ro.x_show;
@@

if not (x^"_show" = x_show) then Coccilib.include_match false

@@
declarer name DEVICE_ATTR_RO;
identifier ro.x,ro.x_show;
@@

- DEVICE_ATTR(x, \(0444\|S_IRUGO\), x_show, NULL);
+ DEVICE_ATTR_RO(x);
// </smpl>
Signed-off-by: NJulia Lawall <Julia.Lawall@lip6.fr>
Acked-by: NRichard Cochran <richardcochran@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

While commit bb35a6ef ("bpf, inode: allow for rename and link ops")
added support for hard links that can be used for prog and map nodes,
this work adds simple symlink support, which can be used f.e. for
directories also when unpriviledged and works with cmdline tooling that
understands S_IFLNK anyway. Since the switch in e27f4a94 ("bpf: Use
mount_nodev not mount_ns to mount the bpf filesystem"), there can be
various mount instances with mount_nodev() and thus hierarchy can be
flattened to facilitate object sharing. Thus, we can keep bpf tooling
also working by repointing paths.

Most of the functionality can be used from vfs library operations. The
symlink is stored in the inode itself, that is in i_link, which is
sufficient in our case as opposed to storing it in the page cache.
While at it, I noticed that bpf_mkdir() and bpf_mkobj() don't update
the directories mtime and ctime, so add a common helper for it called
bpf_dentry_finalize() that takes care of it for all cases now.
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>
Acked-by: NAlexei Starovoitov <ast@kernel.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The following patch fixes an issue with the ldmvsw driver where
the network connection of a guest domain becomes non-functional after
the guest domain has panic'd and rebooted.

The root cause was determined to be from the following series of
events:

1. Guest domain panics - resulting in the guest no longer processing
   network packets (from ldmvsw driver)
2. The ldmvsw driver (in the control domain) eventually exerts flow
   control due to no more available tx drings and stops the tx queue
   for the guest domain
3. The LDC of the network connection for the guest is reset when
   the guest domain reboots after the panic.
4. The LDC reset event is received by the ldmvsw driver and the ldmvsw
   responds by clearing the tx queue for the guest.
5. ldmvsw waits indefinitely for a DATA ACK from the guest - which is
   the normal method to re-enable the tx queue. But the ACK never comes
   because the tx queue was cleared due to the LDC reset.

To fix this issue, in addition to clearing the tx queue, re-enable the
tx queue on a LDC reset. This prevents the ldmvsw from getting caught in
this deadlocked state of waiting for a DATA ACK which will never come.
Signed-off-by: NAaron Young <Aaron.Young@oracle.com>
Acked-by: NSowmini Varadhan <sowmini.varadhan@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Alexander Duyck says:

====================
Add support for XPS when using DCB

This patch series enables proper isolation between traffic classes when
using XPS while DCB is enabled.  Previously enabling XPS would cause the
traffic to be potentially pulled from one traffic class into another on
egress.  This change essentially multiplies the XPS map by the number of
traffic classes and allows us to do a lookup per traffic class for a given
CPU.

To guarantee the isolation I invalidate the XPS map for any queues that are
moved from one traffic class to another, or if we change the number of
traffic classes.

v2: Added sysfs to display traffic class
    Replaced do/while with for loop
    Cleaned up several other for for loops throughout the patch
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds support for setting and using XPS when QoS via traffic
classes is enabled.  With this change we will factor in the priority and
traffic class mapping of the packet and use that information to correctly
select the queue.

This allows us to define a set of queues for a given traffic class via
mqprio and then configure the XPS mapping for those queues so that the
traffic flows can avoid head-of-line blocking between the individual CPUs
if so desired.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch updates the code for removing queues from the XPS map and makes
it so that we can apply the code any time we change either the number of
traffic classes or the mapping of a given block of queues.  This way we
avoid having queues pulling traffic from a foreign traffic class.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add a sysfs attribute for a Tx queue that allows us to determine the
traffic class for a given queue.  This will allow us to more easily
determine this in the future.  It is needed as XPS will take the traffic
class for a group of queues into account in order to avoid pulling traffic
from one traffic class into another.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The functions for configuring the traffic class to queue mappings have
other effects that need to be addressed.  Instead of trying to export a
bunch of new functions just relocate the functions so that we can
instrument them directly with the functionality they will need.
Signed-off-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The current codes use _IOC_TYPE(cmd) == 0x89 to check if the cmd is one
socket ioctl command like SIOCGIFHWADDR. But the literal number 0x89 may
confuse readers. So create one macro SOCK_IOC_TYPE to enhance the readability.
Signed-off-by: NGao Feng <fgao@ikuai8.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Each socket operates in a network namespace where it has been created,
so if we want to dump and restore a socket, we have to know its network
namespace.

We have a socket_diag to get information about sockets, it doesn't
report sockets which are not bound or connected.

This patch introduces a new socket ioctl, which is called SIOCGSKNS
and used to get a file descriptor for a socket network namespace.

A task must have CAP_NET_ADMIN in a target network namespace to
use this ioctl.

Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric W. Biederman <ebiederm@xmission.com>
Signed-off-by: NAndrei Vagin <avagin@openvz.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The second SET_NETDEV_DEV() in the hunk should be
removed.
Reported-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

One merge conflict block wasn't resolved.
Reported-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Saeed Mahameed says:

====================
Mellanox mlx5 core driver updates 2016-10-25

This series contains some updates and fixes of mlx5 core and
IB drivers with the addition of two features that demand
new low level commands and infrastructure updates.
 - SRIOV VF max rate limit support
 - mlx5e tc support for FWD rules with counter.

Needed for both net and rdma subsystems.

Updates and Fixes:
From Saeed Mahameed (2):
  - mlx5 IB: Skip handling unknown mlx5 events
  - Add ConnectX-5 PCIe 4.0 VF device ID

From Artemy Kovalyov (2):
  - Update struct mlx5_ifc_xrqc_bits
  - Ensure SRQ physical address structure endianness

From Eugenia Emantayev (1):
  - Fix length of async_event_mask

New Features:
From Mohamad Haj Yahia (3): mlx5 SRIOV VF max rate limit support
  - Introduce TSAR manipulation firmware commands
  - Introduce E-switch QoS management
  - Add SRIOV VF max rate configuration support

From Mark Bloch (7): mlx5e Tc support for FWD rule with counter
  - Don't unlock fte while still using it
  - Use fte status to decide on firmware command
  - Refactor find_flow_rule
  - Group similar rules under the same fte
  - Add multi dest support
  - Add option to add fwd rule with counter
  - mlx5e tc support for FWD rule with counter
  Mark here fixed two trivial issues with the flow steering core, and did
  some refactoring in the flow steering API to support adding mulit destination
  rules to the same hardware flow table entry at once.  In the last two patches
  added the ability to populate a flow rule with a flow counter to the same flow entry.

V2: Dropped some patches that added new structures without adding any usage of them.
    Added SRIOV VF max rate configuration support patch that introduces
    the usage of the TSAR infrastructure.
    Added flow steering fixes and refactoring in addition to mlx5 tc
    support for forward rule with counter.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The ethtool api {get|set}_settings is deprecated.
We move this driver to new api {get|set}_link_ksettings.
Signed-off-by: NPhilippe Reynes <tremyfr@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Jiri Pirko says:

====================
mlxsw: Add Infiniband support for Mellanox switches

This patchset adds basic Infiniband support for SwitchX-2, Switch-IB
and Switch-IB-2 ASIC drivers.

SwitchX-2 ASIC is VPI capable, which means each port can be either
Ethernet or Infiniband. When the port is configured as Infiniband,
the Subnet Management Agent (SMA) is managed by the SwitchX-2 firmware
and not by the host. Port configuration, MTU and more are configured
remotely by the Subnet Manager (SM).

Usage:
        $ devlink port show
        pci/0000:03:00.0/1: type eth netdev eth0
        pci/0000:03:00.0/3: type eth netdev eth1
        pci/0000:03:00.0/5: type eth netdev eth2
        pci/0000:03:00.0/6: type eth netdev eth3
        pci/0000:03:00.0/8: type eth netdev eth4

        $ devlink port set pci/0000:03:00.0/1 type ib

        $ devlink port show
        pci/0000:03:00.0/1: type ib

Switch-IB (FDR) and Switch-IB-2 (EDR 100Gbs) ASICs are Infiniband-only
switches. The support provided in the mlxsw_switchib.ko driver is port
initialization only. The firmware running in the Silicon implements
the SMA.

Please note that this patchset does only very basic port initialization.
ib_device or RDMA implementations are not part of this patchset.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

SwitchIB and SwitchIB-2 are Infiniband switches with up to 36 ports. This
driver initialize the hardware and Firmware which implements the IB
management and connection with the SM.
Signed-off-by: NElad Raz <eladr@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

SwitchX-2 is IB capable device. This patch add a support to change the
port type between Ethernet and Infiniband.

When the port is set to IB, the FW implements the Subnet Management Agent
(SMA) manage the port. All port attributes can be control remotely by
the SM.

Usage:
	$ devlink port show
	pci/0000:03:00.0/1: type eth netdev eth0
	pci/0000:03:00.0/3: type eth netdev eth1
	pci/0000:03:00.0/5: type eth netdev eth2
	pci/0000:03:00.0/6: type eth netdev eth3
	pci/0000:03:00.0/8: type eth netdev eth4

	$ devlink port set pci/0000:03:00.0/1 type ib

	$ devlink port show
	pci/0000:03:00.0/1: type ib
Signed-off-by: NElad Raz <eladr@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Since we are about to add Infiniband port remove and create we will add
"eth" prefix to port create and remove APIs.
Signed-off-by: NElad Raz <eladr@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>