When running certain routing protocols like VRRP, VF guests need the
ability to set the unicast address of the interface.  Extend the new ndo
trust feature to let the hypervisor trust a guest to set/update its own
unicast address.
Signed-off-by: NChas Williams <3chas3@gmail.com>
Tested-by: NPhil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Move the reset flags to adapter->state in order to make use of bit
operations.

This is an alternative patch to the one previously submitted by
John Greene.
Suggested-by: NAlexander Duyck <aduyck@mirantis.com>
Reported-by: NScott Otto <otts62@yahoo.com>
Reported-by: NJohn Greene <jogreene@redhat.com>
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NPhil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

It seem to be non intentionally changed to Tx in
commit adc81090 ("ixgbe: Refactor busy poll socket code to address
multiple issues")

Lock is taken from ixgbe_low_latency_recv, and there under this
lock we use ixgbe_clean_rx_irq so it looks wrong for me to increment
Tx counter.

Yield stats can be shown through ethtool:
ethtool -S enp129s0 | grep yield
Signed-off-by: NPavel Tikhomirov <ptikhomirov@virtuozzo.com>
Tested-by: NPhil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This patch just updates the driver to the version fully
tested on STi platforms. This version is Jan_2016.
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch adds the whole GMAC4 support inside the
stmmac d.d. now able to use the new HW and some new features
i.e.: TSO.
It is missing the multi-queue and split Header support at this
stage.
This patch also updates the driver version and the stmmac.txt.
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is to support the snps,dwmac-4.00 and snps,dwmac-4.10a
and related features on the platform driver.
See binding doc for further details.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

For gmac3, the MMC addr map is: 0x100 - 0x2fc
For gmac4, the MMC addr map is: 0x700 - 0x8fc

So instead of adding 0x600 to the IO address when setup the mmc,
the RMON base address is saved inside the private structure and
then used to manage the counters.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is the initial support for GMAC4 that includes
the main callbacks to setup the core module: including
Csum, basic filtering, mac address and interrupt (MMC,
MTL, PMT) No LPI added.
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

DMA behavior is linked to descriptor management:

-descriptor mechanism (Tx for example, but it is exactly the same for RX):
-useful registers:
-DMA_CH#_TxDesc_Ring_Len: length of transmit descriptor ring
-DMA_CH#_TxDesc_List_Address: start address of the ring
	-DMA_CH#_TxDesc_Tail_Pointer: address of the last
					      descriptor to send + 1.
	-DMA_CH#_TxDesc_Current_App_TxDesc: address of the current
						    descriptor

-The descriptor Tail Pointer register contains the pointer to the
 descriptor address (N). The base address and the current
 descriptor decide the address of the current descriptor that the
 DMA can process. The descriptors up to one location less than the
 one indicated by the descriptor tail pointer (N-1) are owned by
 the DMA. The DMA continues to process the descriptors until the
 following condition occurs:
 "current descriptor pointer == Descriptor Tail pointer"

Then the DMA goes into suspend mode. The application must perform
a write to descriptor tail pointer register and update the tail
pointer to have the following condition and to start a new transfer:
"current descriptor pointer < Descriptor tail pointer"

The DMA automatically wraps around the base address when the end
of ring is reached.

Up to 8 DMA could be use but currently we only use one (channel0)
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is the main header file to define all the
macro used for GMAC4 DMA and CORE parts.
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

One of main changes of GMAC 4.xx IP is descriptors management.
-descriptors are only used in ring mode.
-A descriptor is composed of 4 32bits registers (no more extended
 descriptors)
-descriptor mechanism (Tx for example, but it is exactly the same for RX):
-useful registers:
	-DMA_CH#_TxDesc_Ring_Len: length of transmit descriptor
				   ring
	-DMA_CH#_TxDesc_List_Address: start address of the ring
	-DMA_CH#_TxDesc_Tail_Pointer: address of the last
				      descriptor to send + 1.
	-DMA_CH#_TxDesc_Current_App_TxDesc: address of the current
					    descriptor

-The descriptor Tail Pointer register contains the pointer to the
 descriptor address (N). The base address and the current
 descriptor decide the address of the current descriptor that the
 DMA can process. The descriptors up to one location less than the
 one indicated by the descriptor tail pointer (N-1) are owned by
 the DMA. The DMA continues to process the descriptors until the
 following condition occurs:
 "current descriptor pointer == Descriptor Tail pointer"

  Then the DMA goes into suspend mode. The application must perform
  a write to descriptor tail pointer register and update the tail
  pointer to have the following condition and to start a new
      transfer:
  "current descriptor pointer < Descriptor tail pointer"

  The DMA automatically wraps around the base address when the end
  of ring is reached.

-New features are available on IP:
-TSO (TCP Segmentation Offload) for TX only
-Split header: to have header and payload in 2 different buffers
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

synopsys_uid is only used once after setup, to get synopsys_id
by using shitf/mask operation. It's no longer used then.
So, remove this temporary variable and directly compute
synopsys_id from setup routine.
Acked-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NFabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

To avoid lot of check in stmmac_main for display ring management
and support the GMAC4 chip, the display_ring function is moved
into dedicated descriptor file.
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

On next GMAC IP generation (4.xx), the way to get hw feature
is not the same than on previous 3.xx. As it is hardware
dependent, the way to get hw capabilities should be defined in dma ops of
each MAC IP. It will avoid also a huge computation of hw capabilities in
stmmac_main.
Signed-off-by: NAlexandre TORGUE <alexandre.torgue@st.com>
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The patch adds support of pause ctrl for HNS V2, and this feature is lost
by HNS V1:
       1) service ports can disable rx pause frame,
       2) debug ports can open tx/rx pause frame.

And this patch updates the REGs about the pause ctrl when updated
status function called by upper layer routine.
Signed-off-by: NLisheng <lisheng011@huawei.com>
Signed-off-by: NYisen Zhuang <Yisen.Zhuang@huawei.com>
Reviewed-by: NAndy Shevchenko <andriy.shevchenko@linux.intel.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After enabling per-cpu processing it appeared that under heavy load
changing MTU can result in blocking all port's interrupts and
transmitting data is not possible after the change.

This commit fixes above issue by disabling percpu interrupts for the
time, when TXQs and RXQs are reconfigured.
Signed-off-by: NMarcin Wojtas <mw@semihalf.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Initially the phy_bus_name was added to manipulate the
driver name but it was recently just used to manage the
fixed-link and then to take some decision at run-time.
So the patch uses the is_pseudo_fixed_link and removes
the phy_bus_name variable not necessary anymore.

The driver can manage the mdio registration by using phy-handle,
dwmac-mdio and own parameter e.g. snps,phy-addr.
This patch takes care about all these possible configurations
and fixes the mdio registration in case of there is a real
transceiver or a switch (that needs to be managed by using
fixed-link).
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Reviewed-by: NAndreas Färber <afaerber@suse.de>
Tested-by: NFrank Schäfer <fschaefer.oss@googlemail.com>
Cc: Gabriel Fernandez <gabriel.fernandez@linaro.org>
Cc: Dinh Nguyen <dinh.linux@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Cc: Phil Reid <preid@electromag.com.au>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This reverts commit 88f8b1bb.
due to problems on GeekBox and Banana Pi M1 board when
connected to a real transceiver instead of a switch via
fixed-link.
Signed-off-by: NGiuseppe Cavallaro <peppe.cavallaro@st.com>
Cc: Gabriel Fernandez <gabriel.fernandez@linaro.org>
Cc: Andreas Färber <afaerber@suse.de>
Cc: Frank Schäfer <fschaefer.oss@googlemail.com>
Cc: Dinh Nguyen <dinh.linux@gmail.com>
Cc: David S. Miller <davem@davemloft.net>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fixs a regression raised when test on chips that use
the normal descriptor layout. In fact, no len bits were set for
the TDES1 and no OWN bit inside the TDES0.
Signed-off-by: NGiuseppe CAVALLARO <peppe.cavallaro@st.com>
Tested-by: NAndreas Färber <afaerber@suse.de>
Cc: Fabrice Gasnier <fabrice.gasnier@st.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

L1_CACHE_BYTES may not be the real cacheline size, use cache_line_size
to determine the cacheline size in runtime.
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Suggested-by: NMarcin Wojtas <mw@semihalf.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

L1_CACHE_BYTES may not be the real cacheline size, use cache_line_size
to determine the cacheline size in runtime.
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Suggested-by: NMarcin Wojtas <mw@semihalf.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is to fix the following maybe-uninitialized warning:

drivers/net/ethernet/marvell/mvpp2.c:6007:18: warning: 'err' may be
used uninitialized in this function [-Wmaybe-uninitialized]
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 55cd48c8 ("net: fec: stop the "rcv is not +last, " error
messages") introduces a write to a register that does not exist in
Coldfire.

Move the FEC_FTRL register access inside the FEC_QUIRK_HAS_RACC 'if' block,
so that we guarantee it will not be used on Coldfire CPUs.
Reported-by: NGreg Ungerer <gerg@uclinux.org>
Signed-off-by: NFabio Estevam <fabio.estevam@nxp.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The mvneta is also used in some Marvell berlin family SoCs which may
have 64bytes cacheline size. Replace the MVNETA_CPU_D_CACHE_LINE_SIZE
usage with L1_CACHE_BYTES.

And since dma_alloc_coherent() is always cacheline size aligned, so
remove the align checks.
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The mvpp2 ip maybe used in SoCs which may have have 64bytes cacheline
size. Replace the MVPP2_CPU_D_CACHE_LINE_SIZE with L1_CACHE_BYTES.

And since dma_alloc_coherent() is always cacheline size aligned, so
remove the align checks.
Signed-off-by: NJisheng Zhang <jszhang@marvell.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

To report flow control tx/rx settings accurately regardless of autoneg
setting, we should use link_info->req_flow_ctrl.  Before this patch,
the reported settings were only correct when autoneg was on.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The typo caused the wrong flow control bit to be set.

Reported by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The size of every padded firmware message is specified in the first
HWRM_VER_GET response message.  Use this value to pad every message
after that.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The existing code does the following:
    allocate completion ring
    initialize completion ring doorbell
    disable interrupts on this completion ring by writing to the doorbell

We can have a race where firmware sends an asynchronous event to the host
after completion ring allocation and before doorbell is initialized.
When this happens driver can crash while ringing the doorbell using
uninitialized value as part of handling the IRQ/napi request.
Signed-off-by: NPrashant Sreedharan <prashant.sreedharan@broadcom.com>
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

in the case where qed_slowpath_irq_req is not called, rc is not
assigned and so qed_int_igu_enable will return a garbage value.
Fix this by initializing rc to 0.
Signed-off-by: NColin Ian King <colin.king@canonical.com>
Acked-by: NYuval Mintz <Yuval.Mintz@qlogic.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Fix support for 16 bit source/dest port matches in ixgbe model.
u32 uses a single 32-bit key value for both source and destination ports
starting at offset 0. So replace the 2 functions with a single function
that takes this key value/mask to program both source and dest ports.

Verified with the following filter:

 #tc qdisc add dev p4p1 ingress
 #tc filter add dev p4p1 parent ffff: protocol ip prio 99 \
	handle 1: u32 divisor 1
 #tc filter add dev p4p1 parent ffff: protocol ip prio 99 \
	handle 800:0:10 u32 ht 800: link 1: \
	offset at 0 mask 0f00 shift 6 plus 0 eat match ip protocol 6 ff
 #tc filter add dev p4p1 parent ffff: protocol ip prio 99 \
	handle 1:0:10 u32 ht 1: \
	match tcp src 1024 ffff match tcp dst 80 ffff action drop
 #tc filter add dev p4p1 parent ffff: protocol ip prio 99 \
	handle 1:0:11 u32 ht 1: \
	match tcp src 1025 ffff action drop
 #tc filter add dev p4p1 parent ffff: protocol ip prio 99 \
	handle 1:0:12 u32 ht 1: \
	match tcp dst 81 ffff action drop
Signed-off-by: NSridhar Samudrala <sridhar.samudrala@intel.com>
Acked-by: NJohn Fastabend <john.r.fastabend@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Remove the incorrect check for mask in ixgbe_configure_clsu32 and
drop the 'mask' field that is not required in struct ixgbe_mat_field

Verified with the following filters:

 #tc qdisc add dev p4p1 ingress
 #tc filter add dev p4p1 parent ffff: protocol ip prio 99 \
	handle 800:0:1 u32 ht 800: \
	match ip dst 10.0.0.1/8 match ip src 10.0.0.2/8 action drop
 #tc filter add dev p4p1 parent ffff: protocol ip prio 99 \
	handle 800:0:2 u32 ht 800: \
	match ip dst 11.0.0.1/16 match ip src 11.0.0.2/16 action drop
 #tc filter add dev p4p1 parent ffff: protocol ip prio 99 \
	handle 800:0:3 u32 ht 800: \
	match ip dst 12.0.0.1/24 match ip src 12.0.0.2/24 action drop
 #tc filter add dev p4p1 parent ffff: protocol ip prio 99 \
	handle 800:0:4 u32 ht 800: \
	match ip dst 13.0.0.1/32 match ip src 13.0.0.2/32 action drop
Signed-off-by: NSridhar Samudrala <sridhar.samudrala@intel.com>
Acked-by: NJohn Fastabend <john.r.fastabend@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Check for handle ids when adding/deleting hash nodes OR adding/deleting
filter entries and limit them to max number of links or header nodes
supported(IXGBE_MAX_LINK_HANDLE).

Start from bit 0 when setting hash table bit-map.(adapter->tables)
Signed-off-by: NSridhar Samudrala <sridhar.samudrala@intel.com>
Acked-by: NJohn Fastabend <john.r.fastabend@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

This function is only used in ixgbe_main.c
Resolves a "missing prototype" warning when building the driver with W=1
Reported-by: NPhil Schmitt <phillip.j.schmitt@intel.com>
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Acked-by: NJohn Fastabend <john.r.fastabend@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Return error when a MAC address change is rejected by the PF.

This will prevent the user from modifying the MAC address when
that operation is not permitted.
Signed-off-by: NEmil Tantilov <emil.s.tantilov@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Calling dev_close() causes IFF_UP to be cleared which will remove the
interfaces routes and some addresses. That's probably not what the user
intended when running the offline selftest. Besides this does not happen
if the interface is brought down before the test, so the current
behaviour is inconsistent.
Instead call the net_device_ops ndo_stop function directly and avoid
touching IFF_UP at all.
Signed-off-by: NStefan Assmann <sassmann@kpanic.de>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Calling dev_close() causes IFF_UP to be cleared which will remove the
interfaces routes and some addresses. That's probably not what the user
intended when running the offline selftest. Besides this does not happen
if the interface is brought down before the test, so the current
behaviour is inconsistent.
Instead call the net_device_ops ndo_stop function directly and avoid
touching IFF_UP at all.
Signed-off-by: NStefan Assmann <sassmann@kpanic.de>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

Use udelay instead of usleep_range because this can be called while
a lock is held.
Signed-off-by: NMark Rustad <mark.d.rustad@intel.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The ATR code was assuming that it would be able to use tcp_hdr for
every TCP frame that came through.  However this isn't the case as it
is possible for a frame to arrive that is TCP but sent through something
like a raw socket.  As a result the driver was setting up bad filters in
which tcp_hdr was really pointing to the network header so the data was
all invalid.

In order to correct this I have added a bit of parsing logic that will
determine the TCP header location based off of the network header and
either the offset in the case of the IPv4 header, or a walk through the
IPv6 extension headers until it encounters the header that indicates
IPPROTO_TCP.  In addition I have added checks to verify that the lowest
protocol provided is recognized as IPv4 or IPv6 to help mitigate raw
sockets using ETH_P_ALL from having ATR applied to them.
Signed-off-by: NAlexander Duyck <aduyck@mirantis.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>

The VXLAN port number should be stored in network order instead of in host
order as it is accessed from the hot-path in ATR.  This way we can avoid
having to do any byte swaps in order to validate the port number.

I moved the vxlan_port value into a hole in the read-mostly region of the
adapter struct.  This way it should be in a warm cache-line instead of in
some isolated region in memory when it needs to be accessed.

In addition I went through and stripped a bunch of unneeded ifdef flags
since having an extra variable present doesn't really hurt anything and
makes the code easier to read.  I also went through and dropped the
NETIF_F_RXCSUM flag which was being set in hw_encap_features but provides
no value as the flag is not evaluated in the Rx path.
Signed-off-by: NAlexander Duyck <aduyck@mirantis.com>
Tested-by: NAndrew Bowers <andrewx.bowers@intel.com>
Signed-off-by: NJeff Kirsher <jeffrey.t.kirsher@intel.com>