Add PCI bus implementation for Mellanox Technologies Switch ASICs. This
includes firmware initialization, async queues manipulation and command
interface implementation.

Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Elad Raz <eladr@mellanox.com>
Reviewed-by: Scott Feldman <sfeldma@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add core components of Mellanox switch driver infrastructure.
Core infrastructure is designed so that it can be used by multiple
bus drivers (PCI now, I2C and SGMII are planned to be implemented
in the future). Multiple switch kind drivers can be registered as well.
This core serves as a glue between buses and drivers.

Signed-off-by: Jiri Pirko <jiri@mellanox.com>
Signed-off-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: Elad Raz <eladr@mellanox.com>
Reviewed-by: Scott Feldman <sfeldma@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch adds vxlan offload specific counters to ethtool stats. We
provide tx/rx queue counters to show the number of vxlan offload pkts
sent/received.

Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@emulex.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Current driver always uses vlan-promisc mode, i.e., it receives both
tagged and untagged traffic and lets the network stack drop packets
tagged with unrequested vlan tags.

This patch implements vlan-filtering offload in the driver -
Unless explicitly configured to promisc mode, only untagged packets or
packets tagged with requested vlans would reach the Rx flow.

Signed-off-by: Yuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Amir Vadai says:

====================
net/mlx5e: Driver update 29-Jul-2015

This patchset contain bug fixes and code cleaning patches to the ConnectX-4
Ethernet driver.

Patchset was applied and tested over commit 8c1a91f1

 ("Merge branch
'mlx4-802.1ad-accel'")
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

It was used to update netdev priv parameters that require stopping
and re-opening the device in a generic way - it got the new
parameters and did: ndo_stop(), copy new parameters into current
parameters, ndo_open().

We chose to remove it for two reasons:
1) It requires additional instance of struct mlx5e_params on the
   stack and looking forward we expect this struct to grow.
2) Sometimes we want to do additional operations (besides
   just updating the priv parameters) while the netdev is stopped.
   For example, updating netdev->mtu @mlx5e_change_mtu() should
   be done while the netdev is stopped (done in this commit).

Signed-off-by: Achiad Shochat <achiad@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Introduce access functions to create/destroy RSS indrection table
and use it in the Ethernet driver.

Signed-off-by: Achiad Shochat <achiad@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Since it is un-named at this time.

Signed-off-by: Achiad Shochat <achiad@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Use the already defined rq pointer directly.

Signed-off-by: Achiad Shochat <achiad@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

It is not needed by the mlx5 Eth driver since it has a CQ per RQ/SQ.

Signed-off-by: Achiad Shochat <achiad@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This field already exists under the mlx5e_params struct

Signed-off-by: Achiad Shochat <achiad@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The page size of the device's RQ/SQ/CQ objects is defined in 4K
units regardless of the system pages size.
Thus using the Linux's PAGE_SHIFT macro yields wrong device
configuration in systems where PAGE_SHIFT!=12.

Signed-off-by: Achiad Shochat <achiad@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

mlx5_cmd_exec() might fail - need to check return value.

Signed-off-by: Haggai Abramovsky <hagaya@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This readds the config option CONFIG_OPENVSWITCH_VXLAN to avoid a
hard dependency of OVS on VXLAN. It moves the VXLAN config compat
code to vport-vxlan.c and allows compliation as a module.

Fixes: 614732ea ("openvswitch: Use regular VXLAN net_device device")
Fixes: 2661371a

 ("openvswitch: fix compilation when vxlan is a module")
Cc: Pravin B Shelar <pshelar@nicira.com>
Cc: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Acked-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Field pktgen_dev.allocated_skbs had been written to, but never read
from. The number of allocated skbs can be deduced anyway, from the total
number of sent packets and the 'clone_skb' param.

Signed-off-by: Bogdan Hamciuc <bogdan.hamciuc@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Allocate enough space so as not to force the outgoing net device to do
skb_realloc_headroom().

Signed-off-by: Bogdan Hamciuc <bogdan.hamciuc@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Any external user should use the registration API instead of
accessing this directly.

Cc: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: Thomas Graf <tgraf@suug.ch>
Acked-by: Roopa Prabhu <roopa@cumulusnetworks.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

As part of defconfig consolidation using fragments, we'd like to be
able to have the same drivers enabled on 32-bit and 64-bit.  Gianfar
happens to only exist on 32-bit systems, and when building the
resulting 64-bit kernel warnings were produced.

A couple of the warnings are trivial, but the rfbptr code has deeper
issues.  It uses the virtual address as the DMA address, which again,
happens to work in the environments where this driver is currently
used, but is not the right thing to do.

Fixes: 45b679c9

 ("gianfar: Implement PAUSE frame generation
support")
Signed-off-by: Scott Wood <scottwood@freescale.com>
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Tom Herbert says:

====================
net: Initialize sk_hash to random value and reset for failing cnxs

This patch set implements a common function to simply set sk_txhash to
a random number instead of going through the trouble to call flow
dissector. From dst_negative_advice we now reset the sk_txhash in hopes
of finding a better ECMP path through the network. Changing sk_txhash
affects:
  - IPv6 flow label and UDP source port which affect ECMP in the network
  - Local ECMP route selection (pending changes to use sk_txhash)
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

When a connection is failing a transport protocol calls
dst_negative_advice to try to get a better route. This patch includes
changing the sk_txhash in that function. This provides a rudimentary
method to try to find a different path in the network since sk_txhash
affects ECMP on the local host and through the network (via flow labels
or UDP source port in encapsulation).

Signed-off-by: Tom Herbert <tom@herbertland.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch creates sk_set_txhash and eliminates protocol specific
inet_set_txhash and ip6_set_txhash. sk_set_txhash simply sets a
random number instead of performing flow dissection. sk_set_txash
is also allowed to be called multiple times for the same socket,
we'll need this when redoing the hash for negative routing advice.

Signed-off-by: Tom Herbert <tom@herbertland.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Radha Mohan Chintakuntla says:

====================
Add MDIO support to ThunderX NIC driver

This patch series adds MDIO support to ThunderX NIC driver by making use
of existing mdio-octeon driver. In the process modified the mdio-octeon
driver to work on both Octeon and ThunderX platforms.

* From v1:
  - Removed default selection in Kconfig for MDIO_OCTEON
  - Replace uint64 with u64 as suggested by David Daney
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

The CONFIG_MDIO_OCTEON is required so that the ThunderX NIC driver can
talk to the PHY drivers.

Signed-off-by: Radha Mohan Chintakuntla <rchintakuntla@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch fixes a possible crash in the octeon_mdiobus_probe function
if the return values are not handled properly.

Signed-off-by: Radha Mohan Chintakuntla <rchintakuntla@cavium.com>
Signed-off-by: Tomasz Nowicki <tomasz.nowicki@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch modifies the mdio-octeon driver to work on both ThunderX and
Octeon SoCs from Cavium Inc.

Signed-off-by: Sunil Goutham <sgoutham@cavium.com>
Signed-off-by: Radha Mohan Chintakuntla <rchintakuntla@cavium.com>
Signed-off-by: David Daney <david.daney@cavium.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

On some of the K2E and K2L platforms, the two DWORDs in
efuse occupied by the pre-programmed mac address for
slave port 1 are swapped.  To workaround this issue,
this patch adds a new define NETCP_EFUSE_ADDR_SWAP (2)
which signifies the occurrence of such swapping so that
the driver can take proper action.  The flag can be
enabled in the corresponding netcp interface dts binding
as efuse-mac = <2>  under the corresponding netcp
interface node.

Signed-off-by: WingMan Kwok <w-kwok2@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In commit d999297c


("tipc: reduce locking scope during packet reception") we introduced
a new function tipc_build_bcast_sync_msg(), which carries initial
synchronization data between two nodes at first contact and at
re-contact. In this function, we missed to add synchronization data,
with the effect that the broadcast link endpoints will fail to
synchronize correctly at re-contact between a running and a restarted
node. All other cases work as intended.

With this commit, we fix this bug.

Signed-off-by: Jon Maloy <jon.maloy@ericsson.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Follow e8e85cc5

 ("packet: remove handling of tx_ring") and remove
the tx_ring parameter from prb_shutdown_retire_blk_timer() as it is only
called with tx_ring = 0.

Signed-off-by: Tobias Klauser <tklauser@distanz.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Michael Holzheu says:

====================
s390/bpf: recache skb->data/hlen for skb_vlan_push/pop

Here the s390 backend for Alexei's patch 4e10df9a

 ("bpf: introduce
bpf_skb_vlan_push/pop() helpers") plus two bugfixes and two minor
improvements.

The first patch "s390/bpf: clear correct BPF accumulator register" will
also go upstream via Martin's "fixes" branch.

* v2: Integrated suggestions from Joe Perches
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

Allow eBPF programs attached to TC qdiscs call skb_vlan_push/pop
via helper functions. These functions may change skb->data/hlen.
This data is cached by s390 JIT to improve performance of ld_abs/ld_ind
instructions. Therefore after a change we have to reload the data.

In case of usage of skb_vlan_push/pop, in the prologue we store
the SKB pointer on the stack and restore it after BPF_JMP_CALL
to skb_vlan_push/pop.

Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Only classic BPF programs that have been converted to eBPF need to clear
the A and X registers. We can check for converted programs with:

  bpf_prog->type == BPF_PROG_TYPE_UNSPEC

So add the check and skip initialization for real eBPF programs.

Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently we have the restriction that jitted BPF programs can
have a maximum size of one page. The reason is that we use short
displacements for the literal pool.

The 20 bit displacements are available since z990 and BPF requires
z196 as minimum. Therefore we can remove this restriction and use
everywhere 20 bit signed long displacements.

Acked-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The EMIT6_DISP_LH macro passes the "disp" parameter to the _EMIT6_DISP_LH
macro. The _EMIT6_DISP_LH macro uses the "disp" parameter twice:

 unsigned int __disp_h = ((u32)disp) & 0xff000;
 unsigned int __disp_l = ((u32)disp) & 0x00fff;

The EMIT6_DISP_LH is used several times with EMIT_CONST_U64() as "disp"
parameter. Therefore always two constants are created per usage of
EMIT6_DISP_LH.

Fix this and add variable "_disp" to avoid multiple expansions.

* v2: Move "_disp" to _EMIT6_DISP_LH as suggested by Joe Perches

Fixes: 05462310

 ("s390/bpf: Add s390x eBPF JIT compiler backend")
Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently we assumed the following BPF to eBPF register mapping:

 - BPF_REG_A -> BPF_REG_7
 - BPF_REG_X -> BPF_REG_8

Unfortunately this mapping is wrong. The correct mapping is:

 - BPF_REG_A -> BPF_REG_0
 - BPF_REG_X -> BPF_REG_7

So clear the correct registers and use the BPF_REG_A and BPF_REG_X
macros instead of BPF_REG_0/7.

Fixes: 05462310

 ("s390/bpf: Add s390x eBPF JIT compiler backend")
Cc: stable@vger.kernel.org # 4.0+
Signed-off-by: Michael Holzheu <holzheu@linux.vnet.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Fix mistakenly used, hard coded, port number in get_phv_bit()

Fixes: 77fc29c4

 ("net/mlx4_core: Preparations for 802.1ad VLAN support")
Signed-off-by: Amir Vadai <amirv@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Joachim Eastwood says:

====================
stmmac clean up for 4.3 part2

This patch set creates proper probe functions in the remaining
dwmac-* drivers, lets the drivers manage their own of match data
structure, and cleans up some of the drivers a bit. Note that
should be no functional change in any of the drivers.

First patch from Fengguang fixes a Coccinelle that was found after
the previous patch set when into next.

Probe functions are best placed just above the platform_driver
struct since it usually needs to call other functions in the
driver but this requires some code movement in the drivers. To
keep the conversion itself easier to review the code movement is
done in separate patches. This creates some extra churn but I
think it's worth it in this case.

As only a couple of drivers actually needs to use of match data
this pushed down from stmmac_platform and into the drivers. This
also makes it possible for drivers to have their own match data
structure. dwmac-rk benefits most from this.

After each driver has it's own probe function and manages it's
own match data the setup/free hooks and common match data struct
are removed.

Sorry for the slightly largish patch set this time around.

Since I can only test the core changes on my platform and not the
specific dwmac-drivers I greatly appreciate if people could test
this on their respective platforms. I did spend quite a bit of
time visually checking all these patches, but nothing beats real
hardware testing. Thanks!

Patch set based on net-next can also be picked up from here:
https://github.com/manabian/linux-lpc.git stmmac_4.3_part2

For those who want to have a look at all the patches for
v4.3 a branch based on Linus master can be found here:
https://github.com/manabian/linux-lpc.git stmmac_4.2

Previous parts can be found here:
http://www.spinics.net/lists/netdev/msg328997.html
http://www.spinics.net/lists/netdev/msg329932.html
http://www.spinics.net/lists/netdev/msg336097.html


====================

Signed-off-by: David S. Miller <davem@davemloft.net>

Remove the two platform specific init callbacks and make
them use a common one by creating a function member in
the internal data structure. This allow us to remove the
layer of indirection and simplify the code a bit.

Signed-off-by: Joachim Eastwood <manabian@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

As all dwmac-* drivers have been converted to have a proper probe
function the setup callback can now be removed. Also remove the
free callback that wasn't used by any driver.

New dwmac-* drivers should implement standard probe and remove
functions to preform any needed setup and teardown.

Signed-off-by: Joachim Eastwood <manabian@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

As dwmac-* drivers that need OF match have been converted
to use their own internal OF match data structure this can
now be removed.

Signed-off-by: Joachim Eastwood <manabian@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Remove the setup glue callback and use rk_gmac_ops as OF match
data so it can used directly when calling rk_gmac_setup.

Signed-off-by: Joachim Eastwood <manabian@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>