Make dwmac4_release_tx_desc() clear all descriptor fields, not just
TDES2 and TDES3.

I'm suspecting that TDES0 and TDES1 wasn't cleared because the DMA
engine uses them to store the tx hardware timestamp (if PTP is enabled).

However, stmmac_tx_clean() calls stmmac_get_tx_hwtstamp(), which reads
and saves the timestamp, before it calls release_tx_desc(), so this
is not an issue.

stmmac_xmit() and stmmac_tso_xmit() both always overwrite TDES0,
however, stmmac_tso_xmit() sometimes sets TDES1, and since neither
stmmac_xmit() nor stmmac_tso_xmit() explicitly clears TDES1, both
functions might reuse a DMA descriptor with old TDES1 data.

I haven't observed any misbehavior even though TDES1 sometimes
point to an old skb, however, explicitly clearing both TDES0 and TDES1
in dwmac4_release_tx_desc() minimizes the chances of undefined behavior.
Signed-off-by: NNiklas Cassel <niklas.cassel@axis.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

According to Documentation/memory-barriers.txt, we need to use a
dma_rmb() after reading the status/own bit, to ensure that all
descriptor fields are read after reading the own bit.

This way, we ensure that the DMA engine is done with the DMA
descriptor before we read the other descriptor fields, e.g. reading
the tx hardware timestamp (if PTP is enabled).
Signed-off-by: NNiklas Cassel <niklas.cassel@axis.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The last memory barrier in stmmac_xmit()/stmmac_tso_xmit() is placed
between a coherent memory write and a MMIO write:

The own bit is written in First Desc (TSO: MSS desc or First Desc).
<barrier>
The DMA engine is started by a write to the tx desc tail pointer/
enable dma transmission register, i.e. a MMIO write.

This barrier cannot be a simple dma_wmb(), since a dma_wmb() is only
used to guarantee the ordering, with respect to other writes,
to cache coherent DMA memory.

To guarantee that the cache coherent memory writes have completed
before we attempt to write to the cache incoherent MMIO region,
we need to use the more heavyweight barrier wmb().
Signed-off-by: NNiklas Cassel <niklas.cassel@axis.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

A dma_wmb() is used to guarantee the ordering, with respect to
other writes, to cache coherent DMA memory.

There is a dma_wmb() in prepare_tx_desc()/prepare_tso_tx_desc() which
ensures that TDES0/1/2 is written before TDES3 (which contains the own
bit), for First Desc.

However, in the rare case that MSS changes, there will be a MSS
context descriptor in front of the regular DMA descriptors:

<MSS desc> <- DMA Next Descriptor
<First Desc>
<desc n>
<Last Desc>

Thus, for this special case, we need a dma_wmb()
after prepare_tso_tx_desc()/before writing the own bit to the MSS desc,
so that we flush the write to TDES3 for First Desc,
in order to ensure that the MSS descriptor is the last descriptor to
set the own bit.
Signed-off-by: NNiklas Cassel <niklas.cassel@axis.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch improves few aspects of interrupt handling:
- update to current interrupt allocation API
  (use pci_alloc_irq_vectors() instead of deprecated pci_enable_msi())
- this implicitly will allocate a MSI-X interrupt if available
- get rid of flag RTL_FEATURE_MSI
- remove some dead code, intentionally disabling (unreliable) MSI
  being partially available on old PCI chips.

The patch works fine on a RTL8168evl (chip version 34) and on a
RTL8169SB (chip version 04).
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds allmulticast option for memac, dtsec
and 10GEC controllers.
Signed-off-by: NRadu Bulie <radu-andrei.bulie@nxp.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Signed-off-by: NMadalin Bucur <madalin.bucur@nxp.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Simplify the code and avoid some Rx errors not being
accounted.
Signed-off-by: NMadalin Bucur <madalin.bucur@nxp.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

An issue in the code mapping the skb fragments into
scatter-gather frames was evidentiated by netperf
TCP_SENDFILE tests. The size was set wrong for all
fragments but the first, affecting the transmission
of any skb with more than one fragment.
Signed-off-by: NMadalin Bucur <madalin.bucur@nxp.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

These two lines are indented too far.
Signed-off-by: NDan Carpenter <dan.carpenter@oracle.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The checks to see if key->dst.s6_addr and key->src.s6_addr are null
pointers are redundant because these are constant size arrays and
so the checks always return true.  Fix this by removing the redundant
checks.   Also replace filter->f with vf, allowing wide lines to be
condensed and to rejoin some split wide lines.

Detected by CoverityScan, CID#1465279 ("Array compared to 0")
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>

Function i40e_find_vsi_from_id can potentially return null, hence
VSI may be null, so defensively check it is non-null before
dereferencing it to check the seid.

Fixes: e284fc28 ("i40e: Add and delete cloud filter")
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Acked-by: NAvinash Dayanand <avinash.dayanand@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Passing struct virtchnl_filter f by value requires a 272 byte copy
on x86_64, so instead pass it by reference is much more efficient. Also
adjust some lines that are over 80 chars.

Detected by CoverityScan, CID#1465285 ("Big parameter passed by value")
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Acked-by: NHarshitha Ramamurthy <harshitha.ramamurthy@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The i40e_detect_recover_hung function uses the i40e_get_tx_pending
function to determine if there are packets stalled on the ring.
i40e_get_tx_pending calculates the pending packets using the head
writeback value and HW tail.  If the queue is stopped and we lose the
interrupt to update our next_to_clean then we a) won't get another
interrupt to clean because queue is stopped b) we won't catch the
problem with i40e_detect_recover_hung because the HW values look like
there's no packets waiting to be transmitted.  Using the SW values we
can catch the issue because next_to_clean will be out of sync with head
writeback.

This has the added benefit being less CPU intensive because we don't
need to reach into the hardware to get the values.
Signed-off-by: NAlan Brady <alan.brady@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Removes the locking of adapter->mac_vlan_list_lock resource in
i40evf_add_filter(). The locking part is moved above i40evf_add_filter().
i40evf_add_filter(), called by i40evf_addr_sync(), was trying to lock the
resource again and double locking generated a kernel panic after bringing
an interface up.

Fixes: 8946b563 ("i40evf: use __dev_[um]c_sync routines in
       .set_rx_mode")
Signed-off-by: NPaweł Jabłoński <pawel.jablonski@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch introduces new ethtool private flag used for
forcing true link state. Function i40e_force_link_state that implements
this functionality was added, it sets phy_type = 0 in order to
work-around firmware's LESM. False positive error messages were
suppressed.

The ndo_open() should not succeed if there were issues with forcing link
state to be UP.

Added I40E_PHY_TYPES_BITMASK define with all phy types OR-ed together in
one bitmask.  Added after phy type definition, so it will be hard to
forget to include new phy types to the bitmask.
Signed-off-by: NMariusz Stachura <mariusz.stachura@intel.com>
Signed-off-by: NMitch Williams <mitch.a.williams@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Eliminate duplicated debug code by moving it into the core driver.
Don't log the only valid silicon revision number (it's in the source).

Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Chris Zankel <chris@zankel.net>
Tested-by: NStan Johnson <userm57@yahoo.com>
Signed-off-by: NFinn Thain <fthain@telegraphics.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add missing printk severity levels by adopting pr_foo() calls for the
platform_driver and dev_foo() calls for the nubus_driver.
Avoid KERN_CONT usage as per advice from checkpatch.
Avoid #ifdef around printk calls.
Don't log driver probe messages after calling register_netdev().

Cc: Thomas Bogendoerfer <tsbogend@alpha.franken.de>
Cc: Chris Zankel <chris@zankel.net>
Tested-by: NStan Johnson <userm57@yahoo.com>
Signed-off-by: NFinn Thain <fthain@telegraphics.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The MACH_IS_MAC test is redundant here because the platform device
won't get registered unless MACH_IS_MAC.

Cc: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: NFinn Thain <fthain@telegraphics.com.au>
Reviewed-by: NGeert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This resolves an old issue preventing any NuBus SONIC NICs from
working in a Mac with an on-board SONIC device.
Tested-by: NStan Johnson <userm57@yahoo.com>
Signed-off-by: NFinn Thain <fthain@telegraphics.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

It appears that the single port Ether controllers having TSU (like SH7734/
R8A7740) need the same kind of treating in sh_eth_tsu_init() as R7S72100
currently has -- they also don't have the TSU registers related e.g. to
passing the frames between ports. Add the 'sh_eth_cpu_data::dual_port'
flag and use it as a new criterion for taking a "short path" in the TSU
init sequence in order to avoid writing to the non-existent registers...

Fixes: f0e81fec ("net: sh_eth: Add support SH7734")
Fixes: 73a0d907 ("net: sh_eth: add support R8A7740")
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Tested-by: NGeert Uytterhoeven <geert+renesas@glider.be>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The TSU_QTAG0/1 registers found in the Gigabit Ether controllers actually
have the same long name  as the TSU_QTAGM0/1 registers in the early Ether
controllers:  Qtag Addition/Deletion Set Register (Port 0/1 to 1/0); thus
there's no need to make a difference in sh_eth_tsu_init() between those
controllers. Unfortunately, we can't just remove TSU_QTAG0/1 from the
register *enum* because that would break the ethtool register dump...

Fixes: b0ca2a21 ("sh_eth: Add support of SH7763 to sh_eth")
Signed-off-by: NSergei Shtylyov <sergei.shtylyov@cogentembedded.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Variable dma is initialized with a value that is never read, later
on it is re-assigned a new value, hence the initialization is redundant
and can be removed.

Cleans up clang warning:
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c:584:13: warning: Value
stored to 'dma' during its initialization is never read
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Acked-by: NAlexander Duyck <alexander.h.duyck@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Add support for build_skb() similar to:
commit 6f429223 ("ixgbe: Add support for build_skb")
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NKrishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Based on commit e0142726 ("igb: Break out Rx buffer page management")

Consolidate Rx code paths to reduce duplication when we expand them in
the future.
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NKrishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Make it so that all rings allocations are made as part of q_vector.
The advantage to this is that we can keep all of the memory related to
a single interrupt in one page.

The goal is to bring the logic of handling rings closer to ixgbe.
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NKrishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Similar to commit a50c29dd
("ixgbe: Make certain that all frames fit minimum size requirements")

Make sure that any packet we attempt to transmit will meet minimum
size requirements.
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NKrishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Following the logic from commit 2de6aa3a
("ixgbe: Add support for padding packet")

Add support for providing a buffer with headroom and tail room
to allow for shared info, NET_SKB_PAD, and NET_IP_ALIGN.  With this
combined with the DMA changes we can start using build_skb to build frames
around an incoming Rx buffer instead of having to memcpy the headers.
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NKrishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Add calls for netif_set_real_num_t/rx_queues() in ixgbevf_open().
Make sure that calls to ixgbevf_open() are rtnl protected and improve
the error handling when setting up multiple queues.
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NKrishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Based on commit 8649aaef
("igb: Add support for using order 1 pages to receive large frames")

Add support for using 3K buffers in order 1 page. We are reserving 1K for
now to have space available for future tail room and head room when we
enable build_skb support.
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NKrishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Introduce legacy-rx private flag that will allow switching between the
old and new (build_skb based) Rx code paths. The implementation is the
same as in commit e0891298
("igb: Add support for ethtool private flag to allow use of legacy Rx")

This provides a means of validating the legacy Rx path in the event that
we are forced to fall back.  At some point in the future when we are
convinced we don't need it anymore we might be able to drop the legacy-rx
flag.
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NKrishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Based on commit 3456fd53
("igb: Use page_address offset from page instead of masking virtual address")

Update the handling of page addresses so that we always refer to them using
a void pointer, and try to use the consistent name of va indicating we are
working with a virtual address.
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NKrishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

On hardware which supports timestamping all packets, the timestamps are
recorded in the packet buffer, and the driver no longer uses or reads
the registers. This makes the logic for checking and clearing Rx
timestamp hangs meaningless.

If we run the ixgbe_ptp_rx_hang() function in this case, then the driver
will continuously spam the log output with "Clearing Rx timestamp hang".
These messages are spurious, and confusing to end users.

The original code in commit a9763f3c ("ixgbe: Update PTP to support
X550EM_x devices", 2015-12-03) did have a flag PTP_RX_TIMESTAMP_IN_REGISTER
which was intended to be used to avoid the Rx timestamp hang check,
however it did not actually check the flag before calling the function.

Do so now in order to stop the checks and prevent the spurious log
messages.

Fixes: a9763f3c ("ixgbe: Update PTP to support X550EM_x devices", 2015-12-03)
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>

If indir == 0 in the ixgbe_set_rxfh(), it is unnecessary
to write the HW. Because redirection table is not changed.
Signed-off-by: NTonghao Zhang <xiangxia.m.yue@gmail.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Bit pattern LOOPBACK_SGMII is being bit-wise or'd twice; remove the
redundant 2nd LOOPBACK_SGMII
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

gcc warns that 'resource_id' is not initialized if we don't come though
any of the three 'case' statements before:

drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c: In function 'mlxsw_sp_kvdl_part_init':
drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.c:275:8: error: 'resource_id' may be used uninitialized in this function [-Werror=maybe-uninitialized]

In the current code, that won't happen, but it's more robust to explicitly
handle this by returning a failure from mlxsw_sp_kvdl_part_init.

Fixes: 887839e6 ("mlxsw: spectrum_kvdl: Add support for dynamic partition set")
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Acked-by: NArkadi Sharshevsky <arkadis@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Calculating the number of entries now uses 64-bit arithmetic that
causes a link error on 32-bit architectures:

drivers/net/ethernet/mellanox/mlxsw/spectrum_kvdl.o: In function `mlxsw_sp_kvdl_init':
spectrum_kvdl.c:(.text+0x51c): undefined reference to `__aeabi_uldivmod'

We could probably use a 32-bit division here as before, but since this is
not in a performance critical path, div_u64() seems cleaner here.

Fixes: 887839e6 ("mlxsw: spectrum_kvdl: Add support for dynamic partition set")
Signed-off-by: NArnd Bergmann <arnd@arndb.de>
Acked-by: NArkadi Sharshevsky <arkadis@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Variable pool is being assigned zero and then in the following for-loop
is it being set to zero again. Remove the redundant first assignment.

Cleans up clang warning:
drivers/net/ethernet/intel/ixgbe/ixgbe_lib.c:61:2: warning: Value stored
to 'pool' is never read
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>

Up until now we only allowed VLAN devices to be put in a VLAN-unaware
bridge, but some users need the ability to enslave physical ports as
well.

This is achieved by mapping the port and VID 1 to the bridge's vFID,
instead of the port and the VID used by the VLAN device.

The above is valid because as long as the port is not enslaved to a
bridge, VID 1 is guaranteed to be configured as PVID and egress
untagged.
Signed-off-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Tested-by: NDavid Ahern <dsahern@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

r8168_check_dash() returns false anyway for all chip versions not
supporting dash. So we can simplify the check conditions.

In addition change the check functions to return bool instead of int,
because they actually return a bool value.
Signed-off-by: NHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>