The Hyper-V netvsc driver was looking at the incorrect status bits
in the checksum info. It was setting the receive checksum unnecessary
flag based on the IP header checksum being correct. The checksum
flag is skb is about TCP and UDP checksum status. Because of this
bug, any packet received with bad TCP checksum would be passed
up the stack and to the application causing data corruption.
The problem is reproducible via netcat and netem.

This had a side effect of not doing receive checksum offload
on IPv6. The driver was also also always doing checksum offload
independent of the checksum setting done via ethtool.
Signed-off-by: NStephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix to return error code -EINVAL from the error handling
case instead of 0, as done elsewhere in this function.

Fixes: e420114e ("rocker: introduce worlds infrastructure")
Signed-off-by: NWei Yongjun <weiyongjun1@huawei.com>
Acked-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 3ac72b7b ("net: fec: align IP header in hardware") breaks
networking on mx28.

There is an erratum on mx28 (ENGR121613 - ENET big endian mode
not compatible with ARM little endian) that requires an additional
byte-swap operation to workaround this problem.

So call swap_buffer() prior to performing the IP header alignment
to restore network functionality on mx28.

Fixes: 3ac72b7b ("net: fec: align IP header in hardware")
Reported-and-tested-by: NHenri Roosen <henri.roosen@ginzinger.com>
Signed-off-by: NFabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Time Sync (PTP) implementation uses the divisor/shift value for converting
the clock ticks to nanoseconds. Driver currently defines shift value as 1,
this results in the nanoseconds value to be calculated as half the actual
value. Hence the user application fails to synchronize the device clock
value with the PTP master device clock. Need to use the 'shift' value of 0.
Signed-off-by: NSony.Chacko <Sony.Chacko@cavium.com>
Signed-off-by: NSudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Driver uses incorrect APIs to unmap DMA memory which were
mapped using dma_map_single(). This patch fixes it to use
appropriate APIs for un-mapping DMA memory.
Signed-off-by: NManish Chopra <manish.chopra@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NSudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

qed_dcbx_query_params() implementation populate the values to input
buffer based on the dcbx mode and, the current negotiated state/params,
the caller of this API need to memset the buffer to zero.
Signed-off-by: NSudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Rx indirection table entries are in the range [0, (rss_count - 1)]. If
user reduces the rss count, the table entries may not be in the ccorrect
range. Need to reconfigure the table with new rss_count as a basis.
Signed-off-by: NSudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With the current default values for Rx path i.e., 8 queues of 8Kb entries
each with 4Kb size, interface will consume 256Mb for Rx. The default values
causing the driver probe to fail when the system memory is low. Based on
the perforamnce results, rx-ring count value of 1Kb gives the comparable
performance with Rx coalesce timeout of 12 seconds. Updating the default
values.
Signed-off-by: NSudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

During the execution of loopback test, driver may receive the packets which
are not originated by this test, loopback implementation need to skip those
packets.
Signed-off-by: NSudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

RSS configuration is not supported for 100G adapters.
Signed-off-by: NSudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Recent changes in kernel ethtool implementation requires the driver
callback for get_channels() has to populate the values for max tx/rx
coalesce fields.
Signed-off-by: NSudarsana Reddy Kalluru <Sudarsana.Kalluru@cavium.com>
Signed-off-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After discussing with Eric, it turns out that, while using
kexec_in_progress is a nice optimization, which prevents us from always
powering on the integrated PHY, let's just turn it on in the shutdown
path.

This removes a dependency on kexec_in_progress which, according to Eric
should not be used by modules

Fixes: 2399d614 ("net: dsa: bcm_sf2: Prevent GPHY shutdown for kexec'd kernels")
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For a kernel that is being kexec'd we re-enable the integrated GPHY in
order for the subsequent MDIO bus scan to succeed and properly bind to
the bcm7xxx PHY driver. If we did not do that, the GPHY would be shut
down by the time the MDIO driver is probing the bus, and it would fail
to read the correct PHY OUI and therefore bind to an appropriate PHY
driver. Later on, this would cause DSA not to be able to successfully
attach to the PHY, and the interface would not be created at all.
Signed-off-by: NFlorian Fainelli <f.fainelli@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, GRO can do unlimited recursion through the gro_receive
handlers.  This was fixed for tunneling protocols by limiting tunnel GRO
to one level with encap_mark, but both VLAN and TEB still have this
problem.  Thus, the kernel is vulnerable to a stack overflow, if we
receive a packet composed entirely of VLAN headers.

This patch adds a recursion counter to the GRO layer to prevent stack
overflow.  When a gro_receive function hits the recursion limit, GRO is
aborted for this skb and it is processed normally.  This recursion
counter is put in the GRO CB, but could be turned into a percpu counter
if we run out of space in the CB.

Thanks to Vladimír Beneš <vbenes@redhat.com> for the initial bug report.

Fixes: CVE-2016-7039
Fixes: 9b174d88 ("net: Add Transparent Ethernet Bridging GRO support.")
Fixes: 66e5133f ("vlan: Add GRO support for non hardware accelerated vlan")
Signed-off-by: NSabrina Dubroca <sd@queasysnail.net>
Reviewed-by: NJiri Benc <jbenc@redhat.com>
Acked-by: NHannes Frederic Sowa <hannes@stressinduktion.org>
Acked-by: NTom Herbert <tom@herbertland.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It makes sense to display the descriptors even if
DES0 is zero. This helps for example in case of it
is needed to dump rx write-back descriptors to get
timestamp status.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Cc: Alexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

SwitchX2 firmware does not implement reset done yet. Moreover, when
busy-polled for ready magic, that slows down firmware and reset takes
longer than the defined timeout, causing initialization to fail.
So restore the previous behaviour and just sleep in this case.

Fixes: 233fa44b ("mlxsw: pci: Implement reset done check")
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Reviewed-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When creating an ethernet port fails, we must move the port to disable,
otherwise putting the port in switch partition 0 (ETH) or 1 (IB) will
always fails.

Fixes: 31557f0f ("mlxsw: Introduce Mellanox SwitchX-2 ASIC support")
Signed-off-by: NElad Raz <eladr@mellanox.com>
Reviewed-by: NJiri Pirko <jiri@mellanox.com>
Reviewed-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The function return value is not checked anywhere. Also, the warning
causes huge slowdown when removing large number of FIB entries which
were not offloaded, because of ordering issue. Ido's preparing
a patchset to fix the ordering issue, but that is definitelly not
net tree material.

Fixes: b45f64d1 ("mlxsw: spectrum_router: Use FIB notifications instead of switchdev calls")
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

By a mistake, there is tree index 0 passed to RALTB. Should be
MLXSW_SP_LPM_TREE_MIN.

Fixes: b45f64d1 ("mlxsw: spectrum_router: Use FIB notifications instead of switchdev calls")
Reported-by: NYotam Gigi <yotamg@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Probe functions should return 0 on success. This driver's probe
returns the value returned by hci_register_dev(), which is the hci
index. This works for systems with only one hci device (id = 0) but
for systems where the btwilink device ends up with an id larger than
0, things will start to fall apart.

Make the probe function return 0 on success.
Signed-off-by: NJacob Siverskog <jacob@teenage.engineering>
Signed-off-by: NMarcel Holtmann <marcel@holtmann.org>

The commit commit 7086605a ("stmmac: fix error check when init ptp")
breaks the procedure added by the
commit efee95f4 ("ptp_clock: future-proofing drivers against PTP
subsystem becoming optional")

So this patch tries to re-import the logic added by the latest
commit above: it makes sense to have the stmmac_ptp_register
as void function and, inside the main, the stmmac_init_ptp can fails
in case of the capability cannot be supported by the HW.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Cc: Alexandre TORGUE <alexandre.torgue@st.com>
Cc: Rayagond Kokatanur <rayagond@vayavyalabs.com>
Cc: Dan Carpenter <dan.carpenter@oracle.com>
Cc: Nicolas Pitre <nico@linaro.org>
Acked-by: NNicolas Pitre <nico@linaro.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With unified images, we need to make sure the net-detect scan is
stopped after resuming, since we don't restart the FW.  Also, we need
to make sure we check if there are enough scan slots available to run
it, as we do with other scans.

Fixes: commit 23ae6128 ("iwlwifi: mvm: Do not switch to D3 image on suspend")
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

The SPLC data parsing is too restrictive and was not trying find the
correct element for WiFi.  This causes problems with some BIOSes where
the SPLC method exists, but doesn't have a WiFi entry on the first
element of the list.  The domain type values are also incorrect
according to the specification.

Fix this by complying with the actual specification.

Additionally, replace all occurrences of SPLX to SPLC, since SPLX is
only a structure internal to the ACPI tables, and may not even exist.

Fixes: bcb079a1 ("iwlwifi: pcie: retrieve and parse ACPI power limitations")
Reported-by: NChris Rorvick <chris@rorvick.com>
Tested-by: NPaul Bolle <pebolle@tiscali.nl>
Tested-by: NChris Rorvick <chris@rorvick.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

When we sync the RX queues the driver waits to receive echo
notification on all the RX queues.
The wait queue is set with timeout until all queues have received
the notification.
However, iwl_mvm_rx_queue_notif() never woke up the wait queue,
with the result of the counter value being checked only when the
timeout expired.
This may cause a latency of up to 1 second.

Fixes: 0636b938 ("iwlwifi: mvm: implement driver RX queues sync command")
Signed-off-by: NSara Sharon <sara.sharon@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

If the suspend flow fails, we restart the hardware to go back to
the D0 image (with non-unified images), but we don't comply with
the fw_restart module parameter.  If something goes wrong when
starting the D3 image, we may want to debug it, so we should
comply with the fw_restart flag to avoid clearing everything up
and losing the firmware state when the error occurred.
Signed-off-by: NHaim Dreyfuss <haim.dreyfuss@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

When a unified D0/D3 image is used, we don't restart the FW in the
D0->D3->D0 transitions.  Therefore, the d3_test functionality should
not call ieee8021_restart_hw() when the resuming either.

Fixes: commit 23ae6128 ("iwlwifi: mvm: Do not switch to D3 image on suspend")
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

In iwl_dbgfs_mem_read(), the len variable may become negative and is
compared to < 0 (an error case).  Comparing size_t (which is unsigned)
to < 0 causes a warning on certain platforms (like i386):

drivers/net/wireless/intel/iwlwifi/mvm/debugfs.c:1561:5-8: WARNING: Unsigned expression compared with zero: len < 0

To prevent that, use ssize_t for len instead.

Fixes: commit 2b55f43f ("iwlwifi: mvm: Add mem debugfs entry")
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

Emmanuel reports that when CMD_WANT_ASYNC_CALLBACK is used by mvm,
the callback will be called with the command queue lock held, and
mvm will try to stop all (other) TX queues, which acquires their
locks - this caused a false lockdep recursive locking report.

Suppress this report by marking the command queue lock with a new,
separate, lock class so lockdep can tell the difference between
the two types of queues.

Fixes: 156f92f2 ("iwlwifi: block the queues when we send ADD_STA for uAPSD")
Reported-by: NEmmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: NJohannes Berg <johannes.berg@intel.com>
Signed-off-by: NLuca Coelho <luciano.coelho@intel.com>

Using list_move_tail() instead of list_del() + list_add_tail().
Signed-off-by: NWei Yongjun <weiyongjun1@huawei.com>
Acked-by: NYuval Mintz <Yuval.Mintz@cavium.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

In some rare configurations, we get a warning about the 'index' variable
being used without an initialization:

drivers/net/ethernet/rocker/rocker_ofdpa.c: In function ‘ofdpa_port_fib_ipv4.isra.16.constprop’:
drivers/net/ethernet/rocker/rocker_ofdpa.c:2425:92: warning: ‘index’ may be used uninitialized in this function [-Wmaybe-uninitialized]

This is a false positive, the logic is just a bit too complex for gcc
to follow here. Moving the intialization of 'index' a little further
down makes it clear to gcc that the function always returns an error
if it is not initialized.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The hdr_offset variable is only if we deal with a TCP or UDP packet,
but as the check surrounding its usage tests for skb_is_gso()
instead, the compiler has no idea if the variable is initialized
or not at that point:

drivers/net/hyperv/netvsc_drv.c: In function ‘netvsc_start_xmit’:
drivers/net/hyperv/netvsc_drv.c:494:42: error: ‘hdr_offset’ may be used uninitialized in this function [-Werror=maybe-uninitialized]

This adds an additional check for the transport type, which
tells the compiler that this path cannot happen. Since the
get_net_transport_info() function should always be inlined
here, I don't expect this to result in additional runtime
checks.
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

gcc found a reference to an uninitialized variable in the error handling
of bcm_enet_open, introduced by a recent cleanup:

drivers/net/ethernet/broadcom/bcm63xx_enet.c: In function 'bcm_enet_open'
drivers/net/ethernet/broadcom/bcm63xx_enet.c:1129:2: warning: 'phydev' may be used uninitialized in this function [-Wmaybe-uninitialized]

This makes the use of that variable conditional, so we only reference it
here after it has been used before. Unlike my normal patches, I have not
build-tested this one, as I don't currently have mips test in my
randconfig setup.

Fixes: 625eb866 ("net: ethernet: broadcom: bcm63xx: use phydev from struct net_device")
Cc: Philippe Reynes <tremyfr@gmail.com>
Reported-by: Nkbuild test robot <fengguang.wu@intel.com>
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It is possible for the MTU to be changed during the initialization
process with the VNIC Server.  Ensure that the net device is updated
to reflect the new MTU.
Signed-off-by: NThomas Falcon <tlfalcon@linux.vnet.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NThomas Falcon <tlfalcon@linux.vnet.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Increment driver version to reflect features that have
been added since release.
Signed-off-by: NThomas Falcon <tlfalcon@linux.vnet.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Do not let number of offload queue sets to go more than
MAX_OFLD_QSETS, which would otherwise crash the driver
on machines with cores more than MAX_OFLD_QSETS.
Signed-off-by: NGanesh Goudar <ganeshgr@chelsio.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Allow autoneg to enable flow control by default.
The behavior when autoneg is off has not changed.
Signed-off-by: NNiklas Cassel <niklas.cassel@axis.com>
Signed-off-by: NJesper Nilsson <jespern@axis.com>
Acked-by: NLars Persson <larper@axis.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

phy_device->supported is originally set by the PHY driver.
The ethernet driver should filter phy_device->supported to only contain
flags supported by the IP.
The IP supports setting rx and tx flow control independently,
therefore SUPPORTED_Pause and SUPPORTED_Asym_Pause should not be cleared.
If the flags are cleared, pause frames cannot be enabled (even if they
are supported by the PHY).
Signed-off-by: NNiklas Cassel <niklas.cassel@axis.com>
Signed-off-by: NJesper Nilsson <jespern@axis.com>
Acked-by: NLars Persson <larper@axis.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Functions bnx2_reg_rd_ind(), bnx2_reg_wr_ind() and bnx2_ctx_wr()
can be called with IRQs disabled when netconsole is enabled. So they
should use spin_{,un}lock_irq{save,restore} instead of _bh variants.

Example call flow:
bnx2_poll()
  ->bnx2_poll_link()
    ->bnx2_phy_int()
      ->bnx2_set_remote_link()
        ->bnx2_shmem_rd()
          ->bnx2_reg_rd_ind()
            -> spin_lock_bh(&bp->indirect_lock);
               spin_unlock_bh(&bp->indirect_lock);
               ...
               -> __local_bh_enable_ip

static inline void __local_bh_enable_ip(unsigned long ip)
      WARN_ON_ONCE(in_irq() || irqs_disabled());   <<<<<< WARN

Cc: Sony Chacko <sony.chacko@qlogic.com>
Cc: Dept-HSGLinuxNICDev@qlogic.com
Signed-off-by: NIvan Vecera <ivecera@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If the driver is built as a module, autoload won't work because the module
alias information is not filled. So user-space can't match the registered
device with the corresponding module.

Export the module alias information using the MODULE_DEVICE_TABLE() macro.

Before this patch:

$ modinfo drivers/net/dsa/bcm_sf2.ko | grep alias
alias:          platform:brcm-sf2

After this patch:

$ modinfo drivers/net/dsa/bcm_sf2.ko | grep alias
alias:          platform:brcm-sf2
alias:          of:N*T*Cbrcm,bcm7445-switch-v4.0C*
alias:          of:N*T*Cbrcm,bcm7445-switch-v4.0
Signed-off-by: NJavier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>