So far the mcast tvlv handler did not anticipate the processing of
multiple incoming OGMs from the same originator at the same time. This
can lead to various issues:

* Broken refcounting: For instance two mcast handlers might both assume
  that an originator just got multicast capabilities and will together
  wrongly decrease mcast.num_disabled by two, potentially leading to
  an integer underflow.

* Potential kernel panic on hlist_del_rcu(): Two mcast handlers might
  one after another try to do an
  hlist_del_rcu(&orig->mcast_want_all_*_node). The second one will
  cause memory corruption / crashes.
  (Reported by: Sven Eckelmann <sven@narfation.org>)

Right in the beginning the code path makes assumptions about the current
multicast related state of an originator and bases all updates on that. The
easiest and least error prune way to fix the issues in this case is to
serialize multiple mcast handler invocations with a spinlock.

Fixes: 60432d75 ("batman-adv: Announce new capability via multicast TVLV")
Signed-off-by: NLinus Lüssing <linus.luessing@c0d3.blue>
Signed-off-by: NMarek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: NAntonio Quartulli <antonio@meshcoding.com>

Bitwise OR/AND assignments in C aren't guaranteed to be atomic. One
OGM handler might undo the set/clear of a specific bit from another
handler run in between.

Fix this by using the atomic set_bit()/clear_bit()/test_bit() functions.

Fixes: 60432d75 ("batman-adv: Announce new capability via multicast TVLV")
Signed-off-by: NLinus Lüssing <linus.luessing@c0d3.blue>
Signed-off-by: NMarek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: NAntonio Quartulli <antonio@meshcoding.com>

Bitwise OR/AND assignments in C aren't guaranteed to be atomic. One
OGM handler might undo the set/clear of a specific bit from another
handler run in between.

Fix this by using the atomic set_bit()/clear_bit()/test_bit() functions.

Fixes: e17931d1 ("batman-adv: introduce capability initialization bitfield")
Signed-off-by: NLinus Lüssing <linus.luessing@c0d3.blue>
Signed-off-by: NMarek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: NAntonio Quartulli <antonio@meshcoding.com>

Bitwise OR/AND assignments in C aren't guaranteed to be atomic. One
OGM handler might undo the set/clear of a specific bit from another
handler run in between.

Fix this by using the atomic set_bit()/clear_bit()/test_bit() functions.

Fixes: 3f4841ff ("batman-adv: tvlv - add network coding container")
Signed-off-by: NLinus Lüssing <linus.luessing@c0d3.blue>
Signed-off-by: NMarek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: NAntonio Quartulli <antonio@meshcoding.com>

Bitwise OR/AND assignments in C aren't guaranteed to be atomic. One
OGM handler might undo the set/clear of a specific bit from another
handler run in between.

Fix this by using the atomic set_bit()/clear_bit()/test_bit() functions.

Fixes: 17cf0ea4 ("batman-adv: tvlv - add distributed arp table container")
Signed-off-by: NLinus Lüssing <linus.luessing@c0d3.blue>
Signed-off-by: NMarek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: NAntonio Quartulli <antonio@meshcoding.com>

The gateway selection based on fast connections is using a single value
calculated from the average tq (0-255) and the download bandwidth (in
100Kibit). The formula for the first step (tq ** 2 * 10000 * bandwidth)
tends to overflow a u32 with low bandwidth settings like 50 [100KiBit]
and a tq value of over 92.

Changing this to a 64 bit unsigned integer allows to support a
bandwidth_down with up to ~2.8e10 [100KiBit] and a perfect tq of 255. This
is ~6.6 times higher than the maximum possible value of the gateway
announcement TVLV.

This problem only affects the non-default gw_sel_class 1.
Signed-off-by: NRuben Wisniewsi <ruben@vfn-nrw.de>
[sven@narfation.org: rewritten commit message]
Signed-off-by: NSven Eckelmann <sven@narfation.org>
Signed-off-by: NMarek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: NAntonio Quartulli <antonio@meshcoding.com>

The gw_factor is divided by BATADV_TQ_LOCAL_WINDOW_SIZE ** 2 * 64. But the
rest of the calculation has nothing to do with the tq window size and
therefore the calculation is just (tmp_gw_factor / (64 ** 3)).

Replace it with a simple shift to avoid a costly 64-bit divide when the
max_gw_factor is changed from u32 to u64. This type change is necessary
to avoid an overflow bug.
Signed-off-by: NSven Eckelmann <sven@narfation.org>
Signed-off-by: NMarek Lindner <mareklindner@neomailbox.ch>
Signed-off-by: NAntonio Quartulli <antonio@meshcoding.com>

When too many remotes are bound to an FDB entry, index may not be increased.
This problem will be caused on the large scale environment that is based on
the unicast default destination, for instance.
Signed-off-by: NAtzm Watanabe <atzm@iij.ad.jp>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use '%zx' to print size_t format in order to fix the following build warning:

drivers/net/ethernet/mellanox/mlxsw/item.h:65:3: warning: format '%lx' expects argument of type 'long unsigned int', but argument 6 has type 'size_t' [-Wformat=]
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Acked-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This change isdn driver, remove reverse_bits() function,
use the generic revbit8() function instead.
Signed-off-by: Nyalin wang <yalin.wang2010@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add or remove some tabs so that statements line up correctly.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When using a cluster of switches, some topologies will have an MDIO
bus per switch, not one for the whole cluster. Allow this to be
represented in the device tree, by adding an optional mii-bus property
at the switch level. The old platform_device method of instantiation
supports this already, so only the device tree binding needs extending
with an additional optional phandle.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Reviewed-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add register definitions #defines for accessing the EEPROM.
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Acked-by: NGuenter Roeck <linux@roeck-us.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Pravin B Shelar says:

====================
GRE: Use flow based tunneling for OVS GRE vport.

Following patches make use of new Using GRE tunnel meta data
collection feature. This allows us to directly use netdev
based GRE tunnel implementation. While doing so I have
removed GRE demux API which were targeted for OVS. Most
of GRE protocol code is now consolidated in ip_gre module.

v5-v4:
Fixed Kconfig dependency for vport-gre module.

v3-v4:
Added interface to ip-gre device to enable meta data collection.
While doing this I split second patch into two patches.

v2-v3:
Add API to create GRE flow based device.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Support for sharing GREPROTO_CISCO port was added so that
OVS gre port and kernel GRE devices can co-exist. After
flow-based tunneling patches OVS GRE protocol processing
is completely moved to ip_gre module. so there is no need
for GRE protocol hook. Following patch consolidates
GRE protocol related functions into ip_gre module.
Signed-off-by: NPravin B Shelar <pshelar@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Using GRE tunnel meta data collection feature, we can implement
OVS GRE vport. This patch removes all of the OVS
specific GRE code and make OVS use a ip_gre net_device.
Minimal GRE vport is kept to handle compatibility with
current userspace application.
Signed-off-by: NPravin B Shelar <pshelar@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Following patch create new tunnel flag which enable
tunnel metadata collection on given device.
Signed-off-by: NPravin B Shelar <pshelar@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This function will be used in gre and geneve vport implementations.
Signed-off-by: NPravin B Shelar <pshelar@nicira.com>
Acked-by: NThomas Graf <tgraf@suug.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add an explicit neighbour table overflow message (ratelimited) and
statistic to make diagnosing neighbour table overflows tractable in
the wild.

Diagnosing a neighbour table overflow can be quite difficult in the wild
because there is no explicit dmesg logged.  Callers to neighbour code
seem to use net_dbg_ratelimit when the neighbour call fails which means
the "base message" is not emitted and the callback suppressed messages
from the ratelimiting can end-up juxtaposed with unrelated messages.
Further, a forced garbage collection will increment a stat on each call
whether it was successful in freeing-up a table entry or not, so that
statistic is only a hint.  So, add a net_info_ratelimited message and
explicit statistic to the neighbour code.
Signed-off-by: NRick Jones <rick.jones2@hp.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds the ability to toggle the vlan filtering support via
netlink. Since we're already running with rtnl in .changelink() we don't
need to take any additional locks.
Signed-off-by: NNikolay Aleksandrov <nikolay@cumulusnetworks.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Shahed Shaikh says:

====================
qlcnic: enhancements

This series adds few enhancements.

  o Patch from Harish reorders the sequence of header files inclusion,
    keeping kernel's header files on top.

  o Firmware introduced a new feature which allows driver to increases
    the size of firmware dump of iSCSI function which is being collected
    by NIC driver.

  o Print buffer address which is holding a firmware dump.

  o Use vzalloc() instead kzalloc() for allocating large chunk of memory
    which will avoid potential memory allocation failure.

  o Add new device ID for 0x8C30 which is a 83xx series based VF function.

Please apply this series to net-next.
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NShahed Shaikh <shahed.shaikh@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Driver allocates a large chunk of temporary buffer using kzalloc
to copy FW image. As there is no real need of this memory to be
physically contiguous, use vzalloc instead.
Signed-off-by: NShahed Shaikh <shahed.shaikh@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is a 83xx series based VF device
Signed-off-by: NShahed Shaikh <shahed.shaikh@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NShahed Shaikh <shahed.shaikh@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In some cases it is required to capture minidump for iSCSI functions
as part of default minidump collection process. To enable this, firmware
exports it's capability and driver need to enable that capability
by issuing a mailbox command.

With this feature, firmware can provide additional iSCSI function's
minidump with smaller minidump capture mask (0x1f).
Signed-off-by: NShahed Shaikh <shahed.shaikh@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Include local headers files after kernel's header files.
Signed-off-by: NHarish Patil <harish.patil@qlogic.com>
Signed-off-by: NShahed Shaikh <shahed.shaikh@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For the Marvel 88e1111 PHY only two SGMII modes are available, both
allowing only SGMII to copper mode (with or without clock). SGMII
to fiber mode is not supported. Make sure the fiber/copper registers
selector bits are cleared for selecting copper mode.
Signed-off-by: NMadalin Bucur <madalin.bucur@freescale.com>
Signed-off-by: NShaohui Xie <Shaohui.Xie@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When we transmit a fragmented skb, we may run into a race like the
following scenario (assume txq->cur_tx is next to txq->dirty_tx):
           cpu 0                                          cpu 1
  fec_enet_txq_submit_skb
    reserve a bdp for the first fragment
    fec_enet_txq_submit_frag_skb
       update the bdp for the other fragment
       update txq->cur_tx
                                                   fec_enet_tx_queue
                                                     bdp = fec_enet_get_nextdesc(txq->dirty_tx, fep, queue_id);
                                                     This bdp is the bdp reserved for the first segment. Given
                                                     that this bdp BD_ENET_TX_READY bit is not set and txq->cur_tx
                                                     is already pointed to a bdp beyond this one. We think this is a
                                                     completed bdp and try to reclaim it.
    update the bdp for the first segment
    update txq->cur_tx

So we shouldn't update the txq->cur_tx until all the update to the
bdps used for fragments are performed. Also add the corresponding
memory barrier to guarantee that the update to the bdps, dirty_tx and
cur_tx performed in the proper order.
Signed-off-by: NKevin Hao <haokexin@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Jiri Pirko says:

====================
mlxsw: Couple of fixes/adjustments

Ido Schimmel (5):
  mlxsw: Call free_netdev when removing port
  mlxsw: Make system port to local port mapping explicit
  mlxsw: Simplify mlxsw_sx_port_xmit function
  mlxsw: Use correct skb length when dumping payload
  mlxsw: Fix use-after-free bug in mlxsw_sx_port_xmit

Jiri Pirko (2):
  mlxsw: Make pci module dependent on HAS_DMA and HAS_IOMEM
  mlxsw: Strip FCS from incoming packets
====================
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Store the length of the skb before transmitting it and use it for stats
instead of skb->len, since skb might have been freed already.

This issue was discovered using the Kernel Address sanitizer (KASan).
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Do not use the length of the transmitted skb (which was freed), but
that of the response skb.

This issue was discovered using the Kernel Address sanitizer (KASan).
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Previously we only checked if the transmission queue is not full in the
middle of the xmit function. This lead to complex logic due to the fact
that sometimes we need to reallocate the headroom for our Tx header.

Allow the switch driver to know if the transmission queue is not full
before sending the packet and remove this complex logic.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

FCS of incoming packets is already checked by HW. Just strip it out.
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This resolves compile errors on um-allyesconfig.

Note that there are many other drivers which have the same issue.
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

System ports are unique identifiers in a multi-ASIC environment that
represent all the available ports in the system. Local ports on the
other hand, are unique only within the local ASIC.

Since system port to local port mapping is not part of the HW-SW
contract and since only single-ASIC configurations are currently
supported, set an explicit 1:1 mapping by configuring the Switch System
Port Record (SSPR) register.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When removing a port's netdevice we should also free the memory
allocated by alloc_etherdev(). Do this by calling free_netdev() at the
end of the teardown sequence.
Reported-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fix double word "the the" in
Documentation/DocBook/networking/API-eth-get-headlen.html
Documentation/DocBook/networking/netdev.html
Documentation/DocBook/networking.xml

These files are generated from comment in source,
so I have to fix comment in net/ethernet/eth.c.
Signed-off-by: NMasanari Iida <standby24x7@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

RTL8211DN is compatible with RTL8211E.
Signed-off-by: NShaohui Xie <Shaohui.Xie@freescale.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

RFC 4182 s2 states that if an IPv4 Explicit NULL label is the only
label on the stack, then after popping the resulting packet must be
treated as a IPv4 packet and forwarded based on the IPv4 header. The
same is true for IPv6 Explicit NULL with an IPv6 packet following.

Therefore, when installing the IPv4/IPv6 Explicit NULL label routes,
add an attribute that specifies the expected payload type for use at
forwarding time for determining the type of the encapsulated packet
instead of inspecting the first nibble of the packet.
Signed-off-by: NRobert Shearman <rshearma@brocade.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>